USF Libraries
USF Digital Collections

The relationship between emotional intelligence and satisfaction with life after controlling for self-esteem, depression...

MISSING IMAGE

Material Information

Title:
The relationship between emotional intelligence and satisfaction with life after controlling for self-esteem, depression, and locus of control among community college students
Physical Description:
Book
Language:
English
Creator:
Murphy, Kevin T
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Higher education
Cognitive ability
Validity
Personality
Subjective well being
Dissertations, Academic -- Higher Education -- Doctoral -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: This study investigated the relationship between Emotional Intelligence (EI) and Satisfaction with Life (SWL) among community college students. Some researchers suggest a relationship exists between EI and important outcome variables (e.g., occupational success & satisfaction with life). However, other researchers suggest measures of EI may simply assess personality variables known to predict these variables. I used the Mayer, Salovey, and Caruso Emotional Intelligence Test (MSCEIT) to investigate how much additional variance in SWL, EI predicts after three personality variables (self-esteem,depression, and locus of control). A convenience sample of 200 Central Florida Community College Students completed the following instruments: 1) MSCEIT(Mayer, Salovey, and Caruso Emotional Intelligence Test, 2002) to assess EI. 2) RSES (Rosenberg Self-Esteem Scale, 1965) to assess self-esteem. 3) BDI-II (Beck Depression Inventory ll) Beck, Steer, and Brown (1997) to assess depression . 4) I-E Scale (Internal-External Locus of Control Scale) Rotter (1966) to assess locus of control. 5) SWLS (Satisfaction with Life Scale) Diener, Emmons, Larsen, and Griffin (1985) to assess overall (global) satisfaction with life. Bivariate correlations between the known predictor variables (self-esteem, depression, and locus of control) and the dependant measure (SWL) are in agreement (size and direction) with prior research. However, correlational analysis suggested no correlation between EI as well as all four components of EI with SWL or the known predictor variables. These findings agree with prior research reporting correlations between EI or components of EI with SWL. A series of five hierarchical regression analyses was conducted to investigate whether EI or any of the four components of EI contributes in the prediction of SWL after accounting for known predictors (self-esteem, depression, and locus of control). The results of all five hierarchical regression analysis suggest s EI as well as the components of EI do not account for additional variance in SWL among community college students.Therefore, results of the study suggest EI is not an important predictor of SWLamong community college students. Limitations of the study as well as suggestions for future research are discussed. In the final sections conclusions as well as some implications for practice in higher education are presented.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Kevin T. Murphy.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 220 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001914382
oclc - 176143665
usfldc doi - E14-SFE0001753
usfldc handle - e14.1753
System ID:
SFS0026071:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001914382
003 fts
005 20071023141015.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071023s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001753
040
FHM
c FHM
035
(OCoLC)176143665
049
FHMM
090
LB2322 (ONLINE)
1 100
Murphy, Kevin T.
4 245
The relationship between emotional intelligence and satisfaction with life after controlling for self-esteem, depression, and locus of control among community college students
h [electronic resource] /
by Kevin T. Murphy.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: This study investigated the relationship between Emotional Intelligence (EI) and Satisfaction with Life (SWL) among community college students. Some researchers suggest a relationship exists between EI and important outcome variables (e.g., occupational success & satisfaction with life). However, other researchers suggest measures of EI may simply assess personality variables known to predict these variables. I used the Mayer, Salovey, and Caruso Emotional Intelligence Test (MSCEIT) to investigate how much additional variance in SWL, EI predicts after three personality variables (self-esteem,depression, and locus of control). A convenience sample of 200 Central Florida Community College Students completed the following instruments: 1) MSCEIT(Mayer, Salovey, and Caruso Emotional Intelligence Test, 2002) to assess EI. 2) RSES (Rosenberg Self-Esteem Scale, 1965) to assess self-esteem. 3) BDI-II (Beck Depression Inventory ll) Beck, Steer, and Brown (1997) to assess depression 4) I-E Scale (Internal-External Locus of Control Scale) Rotter (1966) to assess locus of control. 5) SWLS (Satisfaction with Life Scale) Diener, Emmons, Larsen, and Griffin (1985) to assess overall (global) satisfaction with life. Bivariate correlations between the known predictor variables (self-esteem, depression, and locus of control) and the dependant measure (SWL) are in agreement (size and direction) with prior research. However, correlational analysis suggested no correlation between EI as well as all four components of EI with SWL or the known predictor variables. These findings agree with prior research reporting correlations between EI or components of EI with SWL. A series of five hierarchical regression analyses was conducted to investigate whether EI or any of the four components of EI contributes in the prediction of SWL after accounting for known predictors (self-esteem, depression, and locus of control). The results of all five hierarchical regression analysis suggest s EI as well as the components of EI do not account for additional variance in SWL among community college students.Therefore, results of the study suggest EI is not an important predictor of SWLamong community college students. Limitations of the study as well as suggestions for future research are discussed. In the final sections conclusions as well as some implications for practice in higher education are presented.
502
Dissertation (Ph.D.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 220 pages.
Includes vita.
590
Adviser: James Eison, Ph.D.
653
Higher education.
Cognitive ability.
Validity.
Personality.
Subjective well being.
690
Dissertations, Academic
z USF
x Higher Education
Doctoral.
773
t USF Electronic Theses and Dissertations.
0 856
u http://digital.lib.usf.edu/?e14.1753



PAGE 1

The Relationship Between Emotional Intelligence and Satisfaction With Life After Controlling for Self-Esteem Depression, and Locus of Control Among Community College Students by Kevin T. Murphy A dissertation submitted in partial fulfillment of the requirement s for the degree of Doctor of Philosophy Department of Higher Education College of Education University of South Florida Co-Major Professor: James Eison, Ph.D. Co-Major Professor: Donald Dellow, Ed.D. Michael Mills, Ph.D. John Ferron, Ph.D. Date of Approval: October 25, 2006 Keywords: higher education, cognitive abilit y, validity, personality, subjective well-being Copyright 2006, Kevin T. Murphy

PAGE 2

Acknowledgments I wo uld like to express my gratitude to those that hav e supported and assisted me throughout my doctoral program. Specifically, I extend my thanks to the members of my committee. Co-majo r Professors Dr. James Eison and Dr. Donald Dellow provided expert advice and encouragement throughout a difficult process. Dr Michael Mills and Dr John Ferron members of my committee, helped in immeasurable ways in the completion of my study.

PAGE 3

Dedication In recognition of her steadfast suppor t and encouragement, I dedicate this dissertation to my loving wife Aireen Laragan Murphy for the countless hours of editing and proofreading of my study. Aireen has been my partner in this project as well as in life and I am forever grateful.

PAGE 4

i Table of Contents List of Tables iv ABSTRACT v Chapter One Introduction 1 Statement of the Problem 1 Theoretical Basis of Study 3 Purpose of t he Study 8 Research Questions 9 Hypotheses 10 Definition of Terms 12 Delimitations of t he Study 17 Limitations of the Study 18 Significance of the Study 22 Organization of Re maining Chapter 27 Chapter Two Literature Review 28 Theoretical Deve lopment of Satisfaction with Life 30 Relationship of Satisfaction with Life to the Present Study 33 Measurement of Satisfaction with Life and In struments 35 Theoretical De velopment of Emotional Intelligence 47 Relationship of Emot ional Intelligence to the Present Study 51 Measurement of Emotional Intelligence and Instrument 52

PAGE 5

ii Theoretical Development of Self-Esteem 64 Relationship of Self-Esteem to the Present Study 70 Measurement of Self-Esteem and Instru ments 71 Theoretical Developm ent of Depression 76 Relationship of Depression to the Present Study 78 Measurement of Depression and Instruments 79 Theoretical Devel opment of Locus of Control 82 Relationship of Locus of Control to the Present Study 85 Measurement of Locus of Control and Instru ments 86 Chapter ThreeMethodology 90 Introducti on to Methodology 90 Restatement of Research Questions 90 Population Size /Characteristics 91 Selection Elig ibility Characteristics 93 Sampling Scheme/Size/ Characteristics 93 Ethical Nature of Da ta Collection 102 Instruments 103 Research Design 109 Procedures 109 Data Anal ysis 111 Chapter Four Results 115 Restatement of Research Questions 115

PAGE 6

iii Univariate Statistics 116 Bivariate Co rrelations 120 Assump tions of Regression Analysis 126 Hierarchical Regression Analysis 134 Summary of Results 147 Chapter FiveDiscussion 151 Overview of the st udy 151 Major Findings and Comparisons with Previ ous Research 155 Summary of Finding 158 Suggestions for Future Research 159 Limitations of the Study 162 Threats to Inter nal Validity 163 Threats to External Validity 165 Conclusions 165 Implications for Practice in Higher Education 167 References: 172 Appendices: 217 Appendix A: Instrum ents 218 Appendix B: Institut ional Review Board 219 Appendix C: Letter of Voluntary Research Participation 220 About the Author End Page

PAGE 7

iv List of Tables Tables # 1 Univariate Statistics for all Variabl es 117 Table # 2 Bivariate Statistics for all Scaled Variables 121 Table # 3 Cronbachs Coefficient Alpha 130 Table # 4 Stem Equation 136 Table # 5 Stem Regression Analysis Plus EI (total score) 138 Table # 6 Stem Regression Analysis Plus EI1 ( perceiving emotions) 140 Table # 7 Stem Regression Analysis Plus EI2 (f acilitating thought) 142 Table # 8 Stem Regression Analysis Plus EI3 (understanding emotions) 144 Table # 9 Stem Regression Analysis Plus EI4 (managing emotions) 146

PAGE 8

v The Relationship Between Emotional Inte lligence and Satisfaction With Life After Accounting for Self-Esteem, Depression, and Locus of Control Among Community College Students Kevin T. Murphy ABSTRACT This study investigated the relationshi p between Emotional In telligence (EI) and Satisfaction with Life (SWL) among community college students. Some researchers suggest a relationship exists between EI and important outcome variables (e.g., occupational success & sa tisfaction with life). However, other researchers suggest measures of EI may simply assess personality variables known to predict these variables. I us ed the Mayer, Salovey, and Caruso Emotional Intelligence Test (MSCEIT) to investigate how much additional variance in SWL, EI predicts after th ree personality variables (self-esteem, depression, and locus of control). A conveni ence sample of 200 Central Florida Community College Students completed the following instruments: 1) MSCEIT (Mayer, Salovey, and Caruso Emotional In telligence Test, 2002) to assess EI. 2) RSES (Rosenberg Self-Esteem Scale, 1965) to assess self-esteem. 3) BDIII (Beck Depression Inventory ll) Beck, Steer, and Brown (1997) to assess depression. 4) I-E Scale (I nternal-External Lo cus of Control Scale) Rotter (1966) to assess locus of control. 5) SWLS (Satisfaction with Life Scale) Diener, Emmons, Larsen, and Griffin ( 1985) to assess overall (global)

PAGE 9

vi satisfaction with life. Bivariate correl ations between the known predictor variables (self-esteem, depression, and locus of control) and the dependant measure (SWL) are in agreement (size and direction) with prior research. However, correlational analysis suggested no correlation between EI as well as all four components of EI with SWL or the known predictor variables. These findings agree with prior research r eporting correlations between EI or components of EI with SWL. A series of five hierarchical regression analyses was conducted to investigate whether EI or any of the four components of EI contributes in the prediction of SWL afte r accounting for known predictors (selfesteem, depression, and locus of control). The results of all five hierarchical regression analysis suggests EI as we ll as the components of EI do not account for additional variance in SW L among community college students. Therefore, results of the study suggest EI is not an im portant predictor of SWL among community college students. Limita tions of the study as well as suggestions for future resear ch are discussed. In the final sections conclusions as well as some implications for practice in higher education are presented.

PAGE 10

1 Chapter One Introduction Statement of the Problem Interest in emotional intelligence (EI) has re mained high in both the professional literature and the popular press sinc e Daniel Goleman (1995) popularized the concept with publication of the book Emotional Intelligence. During the past decade, much emotional intelligence res earch has focused on both theoretical development (e.g., Mayer & Salovey, 1997; Cobb & Mayer, 2000), as well as the creation of several assessment measures (e.g., Bar-on, 1997; Mayer, Salovey, & Caruso, 2000a; Mayer, Salovey, & Caruso, 2002) A review of this literature (e.g., Bar-on, 1997; Goleman, 1995; Palmer, Walls Burgess, & Stough, 2001) revealed that many authors have a ssumed a relationship exists between emotional intelligence and several important human val ues such as life satisfaction, the quality of interpersonal relationshi ps, academic success and success in occupations that involve considerable reas oning with emotional information (e.g., psychotherapy). Gibbs (1995) noted t hat on its October 2, 1995 cover, Time magazine declared that Emotional Intellig ence may be the best predictor of success in life, redefining what it means to be smart (p. 60). The problem is that some educators attempti ng to increase EI have implemented emotional intelligence programs or incorporated elements of emotional

PAGE 11

2 intelligence within existing programs with littl e or no empirical research to inform such decisions. Elksnin and Elksnin (2003) stated that Within two years after publication of Golemans book, more than 700 school districts across the nation implemented social emotional learni ng (SEL) programs designed to teach students social-emotional skills (p. 65). Co bb and Mayer (2000) noted that For the most part emotional intelligence is fi nding its way into schools in small doses, through social-emotional learning and c haracter education programs (p. 75). However, some schools have revised or attempted to revise their entire curriculum around emotional intelligence. Fo r example, the state of Rhode Island attempted to integrate emotional learning into all its social, health, and education programs (Elias, Zins, Weissberg, Gr eenberg, Haynes, Keggler Schwab-Stone and Schriver, 1997). Cobb and Mayer (2000) stated, To date there has been relatively little research suggesting the va lidity of emotional intelligence within educational, occupational, and other important life domains (p. 397). Before the utilit y (usefulness) of emotional intelligence can be established in any educational context, it must demons trate predictive validity (account for variance) in important human values (e .g., academic success, interpersonal relations, life satisfaction, etc.) greater than existing known predictors. For a construct to possess utility it must demonstr ate it is more than old wine in a new bottle, it must suggest some increment of additional usefulness. From this perspective the degree to which variance ac counted for by a construct that has already been accounted for by related cons tructs is a measure of its redundancy

PAGE 12

3 and a serious threat to its utilit y. Thus, the real test of a constructs utility is in its ability to increase prediction of import ant human values (account for additional variance). At present the problem is that there is little empirical research to suggest how important or how useful emotional intelligence is in the prediction of important human values (e.g., life satisfaction, interpersonal relations, and academic performance). Theoretical Basis of the Study Since the publication of Golemans (1995) Emotional Intelligence, the construct has evolved along two distinct paths. One path, the more popularly oriented (mixed model) is based largely on Golemans (1995) book. This model broadly conceptualizes emotional intelligence incorporating both cognitive abilities as well as non-cognitive elements. In cont rast to the mixed model, the second path (cognitive ability model) the more academically oriented and narrowly defined model of emotional intelligence builds upon Mayer and Saloveys (1990, 1993, 1997) publications. This model concept ualizes emotional intelligence as a specific type of intelligence Cobb and Maye r (2000) noted, The mixed mode l mixes EI as a cognitive ability, with social competencies, pers onality traits, and behaviors (p. 75). Goleman (1995) described EI as compos ed of five dimensions: (a) selfawareness, (b) self-regulation, (c) motivation, (d) empathy, and (e) social skills. Goleman (1995) summarized what he called the collection of emotional intelligence qualities as character. This model makes broad claims regarding the

PAGE 13

4 importance of emotional intelligence to a va riety of important human qualities (e.g., life satisfaction, interpersonal relationships, academic success, and occupational success). For example, Cherniss and Gole man (2001) noted, EI provides the basis for competencies important in almo st any job (p. 10). Goleman (2001) asserted that EI more than any other asset is the most important overall success factor in careers and EI accounts for 85% to 90% of the success of organizational leaders (p. xv). The second pat h, the more academically oriented cognitive ability model, is led primarily by John Mayer, Peter Salove y and associates (e.g., Mayer & Salovey 1990, 993). This model conceptualizes emoti onal intelligence as distinct yet somewhat similar to traditional in telligences. Cobb and Mayer (2000) stated that EI is distinct because it involves information co ming from our feelings and similar because it involves perceiving and reasoni ng abstractly with this emotional information (p. 74). Using this framework Mayer and Geher (1996) studied 321 undergraduates concluding that Emotional inte lligence is distinct from general intelligence, and yet the two intelligences are correlated to a degree (p. 89). Mayer and Salovey (1997) described emotional intelligence as composed of four abilities: the ability to (a) perceive emoti on, (b) integrate emotion to facilitate thought, (c) understand emotions, and (d) regulat e emotions to promote personal growth. Unlike the mixed model which makes impressive claims of importance, Cobb and Mayer (2000) noted th at The cognitive ability model is somewhat more conservative in its claims about the succe ss this intelligence may lead to (p. 75).

PAGE 14

5 I conceptua lized EI as described by Maye r and Salovey (1997) for the following reasons. First, conceptualizing EI as a relatively distinct intelligence is consistent with much of the intelligence literature. Emotional intelligence has its roots in E. L. Thorndikes (1920) discussion of social intelligence (the ability to understand people). Howard Gardner ( 1983) elaborated on the theme of understanding people in his discussion of personal intelligences. Pfeiffer (2001) noted that Gardners writi ng on interpersonal and intrapersonal intelligences specifically set the stage for subsequent more elaborate theorizing on EI as a type of intelligence. Thus, Mayer and Salovey (1993) defined EI as A type of social intelligence that involves the ability to monitor ones own and others emotions, to discriminate among them, and to use the information to guide ones thinking and actions (p. 432). This model was revised in 1997 in order to clearly set it apart from Daniel Golemans (1995) mixed model of EI. Second, t he Mayer and Salovey (1997) ability model demonstrates greater definitional clarity than Golemans (1995) mixed model of emotional intelligence. Pfeiffer (2001) stated EI suffers from a lack of conc eptual precision (p. 140). For example, Goleman (1995) argued that empathy, optimism, assertiveness, and delay of gratification ar e all abilities that consti tute EI. Golemans (1995) popular version of EI expanded Mayer and Sa loveys (1990) conceptualization to include motivational elements as well as personality traits (e.g., zeal, persistence). Goleman (1995) himself equated EI with c haracter (p. 285). The problem with this conceptualization is that if EI (according to the mi xed model) is almost any

PAGE 15

6 thing then it may well be nothing. Unlike t he mixed model of EI the Mayer and Salovey (1997) model quite narrowly defines EI as composed of four cognitive abilities: the ability to (1) accurately perceive emotions; (2) use emotions to facilitate thinking, problem solving, and creativity; (3) understand emotions; and (4) manage emotions for personal growth. Several res earchers (e.g., Bar-On, 1997; Goleman, 1995; Palmer, Walls, Burgess, & Stough, 2001; Mayer & Salovey, 1997) noted that t he popularity of emotional intelligence in both the popular and professional literature has resulted in a plethora of assumed relationships between emotional inte lligence and other important human qualities (e.g., life sati sfaction, the quality of interpersonal relationships, and success in occupations t hat involve consider able reasoning with emotional information such as those invo lving creativity, leadership, sales and conducting psychotherapy). However, a revi ew of this literature also rev ealed that speculation regarding proposed relationships has far exceeded t he empirical research. Some researchers (e.g., Mayer, Salovey & Caruso 2000a) a ssert that the utility of emotional intelligence remains unknown largely because it s validity has not yet been established. However, some researchers (e.g., Palmer Donaldson, and Stough 2002) note that EI has reached a stage of theoretical and in strument development now supportive of research intended to investigate such rela tionships. A review of the EI literature (e.g., Goleman, 1995; Bar-On, 1997; Mayer, Caruso, & Salovey, 2000) suggested that EI has often been theoretically linked with satisfac tion with life. Therefore, the

PAGE 16

7 literature suggested an empiri cal study of the theoretically proposed relationship between EI and satisfaction with life among community college students. Some researchers (e.g., Cia rrochi, Chan, & Caputi, 2000; Mayer, Caruso, & Salovey, 1999; Palmer, Donaldson, & Stough, 2002) have investigated the relationship between individual differences in satisfaction with life and EI and reported correlations ranging from r = .11 to .45. Other researchers (e.g., Mayer, Caruso, & Salovey, 2000; Newsome, Day, & Catano, 2000; Petrides & Furnham, 2000) reported results that suggest emotional intelligence may predict important human values such as satisfaction with life because it essentially measures other personality traits already known to predict th ese criteria. Therefor e, the predictive validity of emotional intelligence can be clearly established only when it is disentangled from related and overlapping constructs such as self-esteem, depression, and locus of control. Many res earchers have conducted empirica l investigations of life satisfaction (e.g., Diener, 1984; Huebner, 1991; Ramanaiah, Detwiler & Byravan, 1997; Hong & Giannakopoulos,1994; Kopp & Ruzicka, 1993) and report findings that suggest significant correlations between life satisfaction and such personality traits as locus of control, self-esteem, depression, extraversion, optimism, neuroticism and anxiety. Some of the liter ature (e.g., Hong & Giannakopoulos, 1994) suggests that three of the most frequently cited predict ors of life satisfaction are self-esteem, depression, and locus of control respectively. Several researchers (e.g., Diener, 1984; Emmons & Diener, 1985; Lewinsohn, Redner, &

PAGE 17

8 Seeley, 1991; Parkerson, Broadhead, & Tse, 1990: Schmitt & Bedeian, 1982; Vermunt, Spaans, & Zorge, 1989; Weiner, Muczyk, & Gable, 1987) have reported results that suggest a positive relationshi p between self-esteem and satisfaction with life. Other researchers (e.g., Hyer, Harrison, & Warsaw, 1987; Kammann & Flett, 1983; Evans, Kleinman, Halar, & Herz er, 1984; Martinez-Pons, 1997) have reported results that suggest a negative relationship between depression and satisfaction with life. Related empirical studi es (e.g., Hickson, Housley, & Boyle, 1988; Klein, Tatone, & Lindsay, 1989; Lewinsohn, et al., 1991; Morganti, Nehrke, Hulicka, & Cataldo, 1988; Raphael, 1988; Schulz, Tompkins, Wood, & Decker, 1987) have reported results that suggest internal locus of control is positively related to satisfaction with life. The curr ent study investigated the relationship between emotional intelligence and satisfac tion with life among community college students after accounting for the followin g known predictors: self-esteem, depression, and locus of control. Purpose of Study The purpose of the present study was to provi de additional evidence to help distinguish between what is theoretically assumed and what may be empirically demonstrated about the relationship between emotional intelligence and life satisfaction. Thus, this empirical study may hel p further establish (or not) the utility of emotional intelligence. Block (1995) assert ed that To the ex tent a variable correlates with other variables it is said to be explainable by t hese other variables and conveys no unique information (p.188). The utility of emotional intelligence

PAGE 18

9 resides in whether it accounts for vari ance in important human values (e.g. satisfaction with life) above the level of variance explained by other personality constructs such as self-esteem, locus of control, and depression. The popularizat ion of EI in both the popular as well as academic literature has resulted in a level of speculation regar ding EI and its relationship with other variables not supported by the empirical research. However, Palmer, Donaldson, and Stough (2002) argue that The advent of assessment measures has provided a platform for research to examine the relationship betw een emotional intelligence and theoretically related life criteria (p. 1092). Thus, 10 y ears of theoretical and instrument development sinc e Goleman (1995) published Emotional Intelligence now makes it possible to empirically inve stigate the relationship between EI and theoretically related life criteria. The current st udy is important for two reasons. First, because it empirically investigated the relationship between EI and an important life criter ia (satisfaction with life) among community college students. Second, because decisions about educational practices regarding emotional intelligence should be based on solid research, empirical investigations that suggest relationships, rather than on sensationalistic claims such as Emotional in telligence is at times as powerful, and even twice as powerful as IQ (Goleman, 1995, p. 34). Research Questions 1) Does emoti onal intelligence conceptualized as a cognitive ability and measured by the MSCEIT account for greater variance in satisfaction with life

PAGE 19

10 among community college students than self-esteem, depression, and locus of control? 2) Does the ability to perceive and accurately express emotion (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than self esteem, depression, and locus of control? 3) Does the ability to use emotion to facilitate thought (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 4) Does the ability to understand emotions (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 5) Does the ability to mana ge emotions for emotional growth (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? Hypotheses Null hypothesis 1. Emotional Intelligence as measured by the MSCEIT (total score) does not account for variance in satisfaction with life among community college students greater than self-esteem, depression, and locus of control.

PAGE 20

11 Research hy pothesis 1. Emotional Intelligence as measured by the MSCEIT (total score) accounts for variance in satisfaction with life among community college students greater than self-esteem, depression, and locus of control. Null hypothesis 2. The ability to perceive and accurately express emotion, a component of emotional intelligence as measured by the MSCEIT does not account for variance in satisfaction wit h life among community college students greater than self-esteem, depres sion, and locus of control. Research hypothesis 2. The ability to perceive and accurately express emotion, a component of emotional intelligence as measured by the MSCEIT accounts for variance in satisfaction with life among community college students greater than self-esteem, depres sion, and locus of control. Null hypothesis 3. The ability to use emot ion to facilitate thought, a component of emotional intelligence as measured by the MSCEIT does not account for variance in satisfaction wit h life among community college students greater than self-esteem, depres sion, and locus of control. Research hypothesis 3. The abi lity to use emotion to facilitate thought, a component of emotional intelligence as measured by the MSCEIT accounts for variance in satisfaction with life among co mmunity college students greater than self-esteem, depression, and locus of control. Null hypothesis 4. The ability to understand emot ions, a component of emotional intelligence as measured by the MSCEIT does not account for

PAGE 21

12 variance in satisfaction with life among co mmunity college students greater than self-esteem, depression, and locus of control. Research hypothesis 4. The ability to understand emot ions, a component of emotional intelligence as measured by t he MSCEIT accounts for variance in satisfaction with life among community coll ege students greater than self-esteem, depression, and locus of control. Null hypothesis 5. The ability to manage emotions for emotional growth, a component of emotional intelligence as measured by the MSCEIT does not account for variance in satisfaction wit h life among community college students greater than self-esteem, depres sion, and locus of control. Research hypothesis 5. T he ability to manage emotions for emotional growth, a component of emotional intel ligence as measured by the MSCEIT accounts for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? Definition of Terms Cognitive ability model of emotional intelligence. The ability to use information in regards to emotions in order to enhance decision making. Mayer and Salovey (1997) defined emoti onal intelligence in terms of four factors: (a) ability to perceive accurately, appraise and express emotions (e.g., the degree to which a person can identify emotion in self and others), (b) ability to access and generate feelings in order to facilitate t hought (e.g., the degree to which a person can use his or her emotions to improve thinking), (c) ability to understand emotion

PAGE 22

13 and emotional knowledge (e.g ., the degree to which a person can understand the complexities of emotional meanings, emotional transitions, and emotional situations), and (d) ability to regulate emot ions in both self and others to promote emotional and intellectual growth (e.g., a persons level of control over their emotions). The Mayer and Salovey (1997) conceptualization of emotional intelligence is referred to as the cognitive ability model of em otional intelligence because it focuses exclusively on cognitive abilities related to processing emotional information and manag ing emotions. The cognitive ability model uses performance or ability measures to i ndex an individuals level of EI. Performance Measure. Sometimes referred to as an ability measure because it asks people to solve problems with so me objective criteria that divides responses into right and wrong responses (e.g., what is the sum of 7 + 7). Mixed model of emotional intelligence. All cognitive abilities and personality traits that enhance decision making. Go leman (1995) broadly describes EI as composed of five dimensions: a) self-awareness, b) self-regulati on, c) motivation, d) empathy, and d) social skills (p. 15). Th is model mixes cognitive abilities with social competencies, personality traits, behaviors and even motivational concepts (e.g., persistence), equating EI with character (Goleman, 1995; p. 285). Self-Report Measures. These measures ask people to self evaluate and self report their level of important human qualities (e.g., intelligence). The problem with such measures are that they may re flect subjective rather than objective qualities (e. g., How intelligent are you?) Mayer, Salovey, and Caruso (2000a)

PAGE 23

14 stated that Early ev idence suggests that self-report ed EI is fairly unrelated to actual ability. (p. 397). Sati sfaction with Life (SWL). The degree to which an individual (in general) is satisfied with his life Diener, Emmons, Larsen, and Griffin (1985) discussed satisfaction with life as an overall (Global ) satisfaction with life. Pavot and Diener (1993) stated that Life satisfaction refers to a judgmental process, in which individuals assess the quality of their lives on the basis of their own unique set of criteria (p. 164). Self-Esteem The attitude a person has towa rd the self. Blascovich and Tomaka (1991) noted that self-esteem is generally considered the evaluative component of ones self-concept, a broader repres entation of the self that includes cognitive and behavioral aspects as well as evaluative or affective ones. The most broad and frequently cited definit ion of self-esteem is Rosenbergs (1965) who described self-esteem as a favorable or unfav orable attitude toward the self (p. 9). Depression An abnormally low and persistent mood that significantly disrupts previously established levels of functional behavior. The American Psychiatric Association (1994 ) Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV), defines depression as a mood disorder with five or more of the follo wing symptoms present during the same two week period: a) depressed mood, b) feelings of sadness or emptiness, c) significant decrease in interest or satisfaction from previously enjoyed activities, d) significant changes in appetite, e) sleep disturbances, f) p sychomotor agitation or retardation,

PAGE 24

15 g) fatigue or loss of energy, (h) feelings of worthlessness or inappropriate guilt, i) cognitive disturbances, j) recurrent thoughts of death or su icide. The effect of all symptoms must represent a significant decrease from previous functioning. Locus of control. Rotter (1966) defined locus of control as a Generalized expectancy of the extent to which a person perceives that events in ones life are consequences of ones behavior (p. 1). S hapiro, Schwartz, and Astin (1996) stated An individuals beliefs about the contro llability of what happens to them is a core element of their underst anding of how they live in the world (p. 1214). Construct validity. Judd, Smith, and Kidder ( 1991) discussed construct validity as the extent to which the conc rete measures in a study successfully duplicate the theoretical constructs in t he hypotheses. Thus, construct validity may be thought of as an inde x of the extent to which the test may be said to measure the theoretical construct or trai t it purports to meas ure. Campbell and Fiske (1959) noted that Construct validit y is validated using both convergent and discriminant validity (p. 80). Convergent validity. Campbell and Fiske (1959) stat ed that Measures of the same variable made by different methods should agree (converge) and certainly should agree better than measures of differ ent variables made by those several methods (p. 81). Discriminant validity. Campbell and Fiske (1959) noted that Discriminant validity refers to the degree to which measures of different constructs are unique (p. 81).

PAGE 25

16 Criterion validity. Cronbach and Meehl (1955) state that Criterion validity has two sub-components: predictive valid ity and concurrent validity (p. 287). Predictive validity Cronbach and Meehl (1955) note t hat predictive validity refers to how well a construct or measur ement instrument forecasts or predicts a future behavior (criterion) or outcome (e .g., college GPA from high school GPA). Concurrent validity. Cook and Campbell (1979) argued that concurrent validity is an index of the correlation betwe en instrument meas urement items and known and accepted standard measur es or criteria. Essentiall y, it is an index of how well the instrument com pares with other tests know n to measure the same domain in question (e.g., ACT and SAT scores). Incremental validity. Dawes (2001) as well as Ha ynes and OBrien (2000) noted that incremental validity refers to the degree to which a measure accounts for variance in a criterion beyond that whic h is already accounted for by other predictors. Haynes and Lench ( 2003) stated that Increment al validity supplements traditional dimensions of c ontent, convergent, predictive, and discriminant validity (e.g., Foster & Cone, 1995; Nunnally & Bernstein, 1994; Haynes, Nelson, and Blaine, 1999; Silva, 1993), because it addr esses the performance of a measure relative to others (p. 456). Internal validity. Cook and Campbell (1979) defined internal validity as the Approximate validity with which we infer t hat a relationship between two variables is causal (p. 37). Gay and Airasian (2003) di scussed internal validity as The con

PAGE 26

17 dition that observed differences on the dependent variable are a direct result of the independent variable, not some other variable (p. 345). External validity. Johnson and Christensen (2000) defined external validity as The extent to which the results of a study can be generalized to and across populations, settings, and times (p. 200). Population validity. Onwuegbuzie (2003) noted that population validity refers to the Extent to which findings are generalizable from the sample of individuals on which a study was conducted to the larger target population of individuals, as well as across different subpopulations within the larger target population (p. 80). Ecological validity. Onwuegbuzie (2003) proposed t hat Ecological validity refers to the extent to which findings from a study can be generalized across settings, conditions, variables and contexts (p. 80). Temporal validity. Onwuegbuzie (2003) noted, Tem poral validity refers to the extent to which research findings can be generalized across time (p. 80). Delimitations of the Study This study deliberately limited itse lf to community colle ge students enrolled in at least one, three credit hour college leve l course at Central Florida Community College (CFCC). This delimitation (one comm unity college) somewhat diminishes the degree to which results from the pres ent study may be generalized beyond the present study. However, re sults from the present study may be generalized to the population of interest, students attending credit courses at CFCC. The focus of the present study also limited itse lf theoretically to the cognitive ability model of EI

PAGE 27

18 developed by Mayer and Salovey (1990; 1993) and revised by Mayer and Salovey (1997). The two relati vely distinct models of EI (cognitive ability and mixed model) in general employ two equally distinct measurement methods. First, the mixed model generally employs self-report method s to assess EI. Self-report measures ask people to evaluate and report their leve l of a quality (e.g., How well do you solve problems?). Second, the cognitive ability model employs ability or performance measures. Ability or performance me asures ask people to solve problems and then their responses are evaluated agains t some criterion (e.g., expert or general consensus scoring) in order index t heir level of a quality (e.g., How many degrees are there in a right angle?). Maye r, Salovey, and Ca ruso (2000) noted that the relationship between self-report measures of EI and actual ability like the relationship between self-report intelligence and actual intellectual ability is low. Thus, in the present study I assessed EI with the Mayer, Salovey, Caruso, Emotional Intelligence Test (MSCEIT), a per formance measure. However, future studies that include assessment of EI with bo th self-report measur es for example, the Self-Report Emotional Intelligence Te st (SREIT) as well as performance measures are recommended. Limitations of the Study Onwuegbuzie (2003) noted that Threats to internal and external validity may take place at the data collection, data analysis or data interpretation stage of all investigations (p. 74). At the data collection stage of the present study, one

PAGE 28

19 potential threat to internal validity is hi story. Unique experiences or significantly different experiences among par ticipants or groups can th reaten internal validity by providing rival explanations of findi ngs (e.g., surveys completed one day before and one day after 9-11-01, surveys completed toward the beginning and toward the end of a semester). One po ssible threat to the external validit y of the present study and virtually all educational studies at the data collection stage is population validity (Onwuegbuzie, 2003). This threat regardi ng population validity according to Johnson and Christensen (2000) have two caus es. First, all members of the target population rarely are availabl e for selection in a study. Second, random samples are difficult to obtain due to practical consi derations such as time, resources, and logistics. In the present study both of t hese considerations we re important to external validity. All member s of the target population (CFCC students) were not available for selection in the study, and limited resources and logistics precluded the use of a random sample. Thus, population validity in t he present study as well as in most non-experimental research invo lving college students presents a threat to external validity. At the dat a analysis stage of the present st udy, population validity once again presented a possible threat to external validity. Any type of sub-sampling from the original sample decreases populat ion validity. Therefor e, in order to minimize this threat to external validity from di screpancies between the sample and population I did not conduct any sub-samp le analysis. Furthermore, the total

PAGE 29

20 sample of 200 participants was inspected fo r representativeness or how well the sample actually reflected the population (a ll students enrolled in credit courses at CFCC). No significant discrepancies (e .g., gender, age, race) between the obtained sample and target population was detected. Life satisf action is generally considered both an important outcome goal of higher education as well as an important hum an value (e.g., Argyle, 1987; Myers, 1992). At the data analysis stage of the present study, the choice of life satisfaction as the criterion variable was an im portant limitation. Un like more clearly defined and more stable variables (e.g ., age, gender, grade point average) life satisfaction is expected to change over ti me. The study of im portant yet less stable constructs (e.g., happiness, spirituality, life satisfaction) often involves the use of assessment instruments that dem onstrate relatively low to moderate reliabilities. The satisfaction with life scale (SWLS) used to measure satisfaction with life in the present study demonstrated less reliability than typically reported by many other instruments assessing other more stable constructs (e.g., age, race, gender). However, prior research on t he relationship between satisfaction with life and other import ant variables (e.g., happiness, academic success, occupational success) suggests it is an important area of investigation. At the data interpretation stage of the pres ent study there are several possible threats to external validity (e.g., population, ec ological, and temporal). Onwuegbuzie (2003) argued that Only if findings are consistent across different populations, locations, settings, times, and c ontexts can researchers be justified

PAGE 30

21 in making generalizations from samples to ta rget populations (p. 74). In order to minimize the threat to exter nal validity at the data interp retation stage from threats to population, ecological, and temporal va lidity, I acknowledged the limits of the present study, avoided the inclination to over generalize, and at best proposed qualified conclusions. A furt her limitation of the present study and a serious threat to the extent one may reasonably generalize findings from the sample to a population (external validity) is small sample size. The present study is a correlation study, in summary I am interested in the re lationship between the dependent variable (satisfaction with life) and the independent variable (emotional intelligence) after controlling for the independent variables self-esteem, depre ssion and locus of control. At the heart of correlation research is predicti on, how well does one variable, or in regression analysis a combination of vari ables, predict another variable. The present study is particularly interested in how much (if any) emotional intelligence adds to the prediction of satisfaction wit h life among community college students over other known predictors (self-est eem, depression, and locus of control). Previous research (e.g ., Hong & Giannakopoulos, 1994) suggested the effect size between self-esteem and sa tisfaction with life is high medium ( R = .21; effect size = .26) However, this same study reported the addition of a second variable, depression resulted in a small effect size ( R = .03; effect size = .03). Likewise the addition of a third variabl e locus of control resulted in an even smaller effect size ( R = .01; effect size = .01).

PAGE 31

22 With a samp le size of 200 participants and a pre-set alpha of .05, the present study should have adequate power (.80) to detect a moderate to large effect size (as large or larger than self-esteem). However, it must be remembered that at the current stage of construct development we can only estimate the associated effect size between EI and other important variables. Much research is constrained by the availabili ty of resources and logistics, thus many independent studies utilize less than desirable sample size s. However, the value of these small sample size studies are realized when subsequent meta-studies pool data from many smaller studies. Significance of the Study Some public and private K schools (e.g., La Salle Academy, R.I.; Nueva School in Hillsborough, C.A.) as well as colleges (e.g., Northern Kentucky University Business School; Department of Educational Leadership, East Carolina University (ECU); Texas A & M Univ ersity-Kingsville) across the nation have already revised their curriculum and/or revised their instructional practices to include elements of emotional intelligenc e. Elias, Zins, Weissberg, Frey, Greenberg, Haynes, Kessler, Schwab-Stone and Schriver (1997) noted that the state of Rhode Island attempted to integrate emotional learning into all its social, health, and education programs. O Shea (2002) as well as Nelson and Low (2002) concluded that many colleges and un iversities offer freshman seminar classes designed to orient students to the campus and integrate components of emotional and social learning. Matthews, Zeidner, and Rober ts (2002) wrote that

PAGE 32

23 The collaborative for social and emotional learning at the University of Illinois reports that today thousands of U.S. sc hools are using more than 150 emotional literacy programs (p. 222). However, Cobb and Mayer (2000) argued that Early claims of the benefits of emotional intelligence to students, schools, and beyond were made without much empirical justification (p. 75). Ma yer, Salovey, and Ca ruso (2004b) noted that Such claims suggest that EI predict s major life outcomes at levels virtually unheard of in psychological science (p. 206) Contrary to such claims several researchers (e.g., Ashkanasy & Dasbor ough, 2003; Barchard, 2003; Brackett & Mayer, 2003; Lam, & Kirby, 2002) investi gated the relationship between EI and problem solving ability or school grades and reported correlations that ranged between r =.20 and .25. Other preliminary research (e.g., Schutte, Malouff, Hall, Haggerty, Cooper, Golden & Dornheim, 1998) suggest a positive yet more moderate relationship between EI and academic performance. Interest in EI will remain high in higher education for the following three reasons. First, Springer, Terenzini, and Pasca rella (1995) stated that Historically the mission of American higher educati on encompassed more than intellectual development and The Socratic imperative to know thyself continues to represent an educational outcome of intrin sic value to many American college students (p. 5). The need to integrate t he intellectual, social, and emotional aspects of undergraduate st udent learning in higher education has been voiced periodically during the last half-century (e.g., Williamson, 1957; Brown, 1972;

PAGE 33

24 Boyer, 1987; Pascarella &Terenzini, 1991; As tin, 1993; Tinto, 19 93). The central mission of higher education remains the educ ation of the whole student which includes cognitive, social, and emotional elements. The traditional yet often elusive goal of holistic education continues to be an important educational outcome. Second, other researchers (e. g., Chickering & Reisser, 1993; Brower, 1990; Upcraft & Gardner, 1989) argue that emotional skills are major factors in college student development (e.g., learning, grades, and retention). Tinto (1987) asserts that 57% of college students leave their fi rst college choice without receiving a degree and 43% of college students leave al together without obtaining a degree. Levitz and Noel (1989) noted that although students leave for a variety of reasons, most attrition is preventable. Other re searchers (e.g., Szulecka, Springett, and De Pauw, 1987) have reported results that suggest the major causes of attrition among college freshman are emotional rath er than academic. Sylvester (1994) stated that Emotion is import ant in education because it drives attention, which in turn drives learning and memory (p. 60). Love and Love (1995) noted that A students development can be enhanced by acti vely bringing the dimensions of affect and cognition together (p. 15). Emoti onal skill is valued both as an outcome goal of higher education as well as an important element of the total undergraduate learning experience. Third, Golem an (1998) asserted that EI accounts for over 85 percent of outstanding performance in top leaders and C ompared to IQ and expertise, EI is twice as important to job per formance (p. 31). In additi on, the recent publication

PAGE 34

25 of popular books such as The Emotionally Intelligent Workplace, by Cherniss and Goleman (2001), as well as recent res earch (e.g., Abraham, 2000; Ashforth & Humphrey, 1995; Ashkanasy & Daus, 2002; Janovics, & Christiansen, 2002) suggest a positive relationship between EI and worker performance. Also, the publication of Primal Leadership by Goleman, Boyatzis, and Mckee (2002) as well as other recent research (e.g ., Atwater & Yammarino, 1993; Gibbons, 1986; Howell & Avolio, 1993; Southwick, 1998; Mandell & Pherwani, 2003) suggest a positive relationship between EI and effectiv e leadership. However, other research (e.g., Mayer & Cobb, 2000) suggests there is little or no direct evidence to support such claims. Thus, the assumption relati ng EI with both worker performance and effective leadership continues despite the lack of and relatively mixed results reported in the research. Palmer, Donaldson, and Stough (2002) proposed that emotional intelligence has reached a stage of theor etical and instrument development now supportive of research intended to estab lish its utility (usefulness). On ly by investigating the level of variance emotional intelligence acco unts for in important outcomes (e.g., satisfaction with life) over known predict ors may we establish the utility of emotional intelligence in the pr ediction of those outcomes. I hope that the present investigation helps further est ablish the relationship or lack of relationship between emotional in telligence and satisfaction with life among community college students. Second, I hope the present study helps reveal which if any of the four relatively independent com ponents of the Mayer and

PAGE 35

26 Salovey (1997) cognitive ability model of emotional intelligence most strongly accounts for variance in satisfaction with lif e. Third, I hope the results from the present study adds to the empi rical research base used to inform decisions in both curriculum development and instructional design within educational settings. For example, Salovey, Stroud, and Woolery (200 2) reported results from their study (community sample) that suggested a m oderate negative relationship between EI and later adult undesirable behaviors (e.g., smoking, alcohol abuse, and fighting). The conclusion suggested by the above studies and similar investigations (e. g., Rubin, 1999; Trinidad & Johnson, 2002) is that higher EI predicts lower incidents of undesirable behavior. This res earch supports other research (e.g., Chickering & Reisser 1993) that sugges ts a positive relationship between emotional skills development and college student development. Other researchers (e.g., Barefoot & Fidler, 1996) note that in general the goals of freshmen seminar programs nationally emphasize the development of emotional skills. Nelson and Nelson (2003) reported from their study with135 first semester university students that Emotional skills are very important factors in the achievement and retention of university freshmen (p. 4). Thus, fr eshmen seminar program s across the nation may influence college student achievement as well as retention by improving student emotional skills and t hus reducing undesirable behavior. Given the positive relationship between emotional ski lls and college student achievement and retention as well as the negative relati onship between emotional skills and undesirable behavior EI may be an important consideration in curriculum develop

PAGE 36

27 ment and instructional design. Fourth, I hope the results of this study suggests additional studies to further enrich t he emotional intelligence literature. Organization of Remaining Chapters Chapter 2 in cludes an examination of the exis ting literature on emotional intelligence, satisfaction with life, self -esteem, depression, and locus of control. Chapter 3 includes a description of the research design and procedures I utilized in the present study to investi gate the relationship between emotional intelligence and satisfaction with life, after accounting the following personality constructs self-esteem, depression, and locu s of control. Chapter 4 contains a description of the procedures used and re sults of the data analysis. Chapter 5 contains an overview of the study; majo r findings are discussed within the context of previous research. Some suggestions for future research as well as limitations of the present study are identif ied. Conclusions as well as implications for practice in higher education are discussed.

PAGE 37

28 Chapter Two Review of the Literature In Western culture the relationship between intellect (rational thought) and affect (emotion) has historically been viewed as somewhat ambiguous. The ambiguity is not in the relative worth of either rational thought or emotion, but rather in deciding whether emotions should be excluded or integrated with rational thought. Traditionally educators have recognized or at least paid lip service to the importance of the emotional domain in the teaching and learning process. Beck and Kosnik ( 1995) noted that Education in Western culture, in general, acknowledges the importance of emotions, and yet may best be described as preoccupied with intellectual skills (p. 161). Zeidner, Roberts and Matthews (2002) similarly proposed that in educational practice, and to a somewhat lesser extent in educational resear ch, emotions have been neglected or at best overshadowed by the cognitive domain. On the other hand, Freshwater and Stickley ( 2004) argued that the concept of emotional intelligence reminds us that we conceptualize the Mind as composed of two minds, a rational mind t hat thinks, and an emotional mi nd that feels (p. 91). Salovey, Woolery, and Mayer (2001) assert the construct emotional intelligence has gained prominence partly because it represents emerging contemporary cultural values. Continuing this line of r easoning, Zeidner et al., (2002) proposed that increasing recent interest in emotional Intelligence is in part a reflection of the

PAGE 38

29 times, the zeitgeist of c ontemporary western society, which is increasingly recognizing the importance of emotions across a variety of important life domains (e.g., academic, occupational, and social) a ll of which contribute to ones global satisfaction with life. Research on satisfaction with life over the past thirty years suggests satisfaction with life is an im portant human value for two reasons. First, Argyle (1987) noted that higher levels of satisfac tion with life are associated with higher levels of positive affect. Second, Myers (1992) stated that high levels of satisfaction with life are associated wit h other important and much desired characteristics (e.g. greater sense of cont rol, higher self-esteem, and less stress (p. 5). Several well st udied personality constructs in psychology (e.g., self-esteem, depression, and locus of control) have been c onsistently found to be predictive of satisfaction with life. Many researchers (e.g., Bar-On, 1997; Ciarrochi, Chan, & Caputi, 2000; Martinez-Pons, 1997, 1999; Maye r, Caruso, & Salovey, 2000) have investigated the relationship between sa tisfaction with life and emotional intelligence (EI) and reported findings that sugges t a low to moderate positive relationship. Bar-On (1997) reported results from his study employing a self-report measure of EI the EQi (Emotional Quotient Inventory) suggesting r = .41, p < .001. Martinez-Pons (1997) reported results from his study that employed another well known self-report measure of EI, the Trai t Meta-Mood Scale (TMMS) that suggested r = .51. Other researchers such as Ciarrochi et al., ( 2000) as well as Mayer,

PAGE 39

30 Salovey and Caruso (2000) employed perform ance based measures of EI such as the Multi-Factor Emotional Intelligence Scale (MEIS). They reported findings suggesting a positive correlation between emot ional intelligence and satisfaction with life r = .28, p < .001 and r = .11, p = .001 respectively. The pres ent study investigated the re lationship between emotional intelligence (total score) as well as each of the four components of the Mayer and Salovey (1997) cognitive ability model of emotional intelligence and satisfaction with life. This chapter reviews the relev ant research and theor y related to the present study. The chapter is organized into five parts: life satisfaction, emotional intelligence, self-esteem, lo cus of control, and depression. Each part addressed an important variable related to the present investigation. A similar outline has been followed within each section so that t he relationships among the individual variables may be better understood. Theoretical Development of Satisfaction With Life Gilman and Huebner (2003) suggested that re search on the nature and correlates of satisfaction with life had become a focus of attention among researchers in a variety of areas of inqui ry (e.g., occupational functioning, physical and mental health, education, retirement, an d interpersonal relationships) during the past thirty years. Other researchers, such as St rack, Argyle, and Schwarz (1991) suggested achieving greater satisfaction in life is important not only because it is a goal for which all indivi duals strive but because increased life satisfaction appears to contribute to health a ttributes (e.g., less stress and reduced

PAGE 40

31 high risk behaviors such as substance abuse). Myers (1992) as well as Veenhoven (1988) reported findings that suggested people with greater satisfaction with life generally are more social, loving, forgiving, trusting, helpful, energetic, decisive and creative as well as less self-focused, hostile and vulnerable to disease. Therefore, increas ing an individuals satisfaction with life may buffer the impact of negative life events, broaden perception, increase creativity, encourage active living, foster social contact, and improve mental health. Early satisfaction with life research (e.g., Fordyce, 1983) suggested everyone strives for personal happiness or satisfaction with life. More recent satisfaction with life research such as Scollon, Diener, Oishi, and Biswas-Diener (2004) reported similar findings from an in ternational study of both Eastern and Western college student samples suggesting the vast majority of college students around the world consider satisfaction with life to be extremely important (more important than money). Diener (1984) proposed that both satisfaction with life and the affective components of well-being are influenced by t he appraisals individuals make of their life circumstances. Lawton (1983) as well as Liang (1985) suggested that while the cognitive and affective components of subjective well-being are distinct, they are also moderately correlated. Emm ons and Diener (1985) as well as Bryant and Veroff (1982) suggested that satisf action with life and the affective components of well-being are qualitatively di fferent. Several researchers (e.g.,

PAGE 41

32 Costa & McCrae, 1980; Michalos, 1991) sugges ted that while satisfaction with life and affective well-being are m oderately correlated, both may act differently across time and have different correlates. Gilman and Huebner (2003) as well as McCullough, Huebner, and Laughlin (2000) proposed that al though the cognitive component (satisfaction with life) and affe ctive components (emotion) are not exclusive of each other, they are relative ly distinct in both adults and children. Gilman and Huebner (2003) argued that Given the degree of independence between the cognitive and affective com ponents of subjective well-being, discussions of subjective well-being s hould focus on each component separately (p. 198). Andrews and Withey (1976) asserted that in the fi eld of subjective well-being research, three relatively independent components have been identified: (a) positive affect, (b) negative affect, and (c) sa tisfaction with life. However, Diener (1984) argued that life satisfaction is one of two components of subjective wellbeing. Based upon Dieners concept ualization, satisfaction with life is the cognitive evaluation an individual makes regarding hi s or her global satisfaction with life across multiple domains. Moods and emotions, which together constitute the affective component represent peoples mo mentary evaluations of the events that occur in their lives. Diener, Emmons, Larsen, and Griffin (1 985) as well as Shin and Johnson (1978) defined satisfacti on with life as an individuals personal judgment of well-being and quality of life bas ed on his or her own chosen criteria. Diener (1984) stated that The hallmark of sati sfaction with life is that it centers on

PAGE 42

33 personal judgments, not upon some criteria that is judged to be important by the researchers (p. 546). Diener (1994) noted that the more global construct of subjective well-being is a multidimensi onal construct, composed of cognitive appraisals (life satisfaction) and affective components. Diener, Suh, Oishi, Lucas, and Smith (1999) suggested that the most commonly accepted model of subjective well-being conceptualizes it as having an emotional component (e.g., sadness, anxiety, and joy) and a cognitive component (satisfaction with life). Altho ugh much of the quality of life literat ure fails to distinguish between subjective well-being and satisfaction with life, it should be noted that the constructs are not equivalent. Subjective we ll-being is a more broadly defined construct having both cognitive and affective components. Life sati sfaction, on the other hand, is limited to the cognitive com ponent of subjective well-being and thus tends to be more stable. Satisfaction with life is the criteri on variable (dependent measure) in the present study. I chos e satisfaction with life because some research (e.g., Diener, 1984; Diener & Lars en, 1984) suggested satisfaction with life demonstrates greater stability over subjective well-being. Relationship of Satisfaction With Life to the Present Study Satisfacti on with life was chosen for the criterion variable (dependent measure) in the present study for the following reasons. Firs t, as previously stated, several researchers (e.g., Bar-On, 1997; Ciarrochi, Chan, & Caputi, 2000; Martinez-Pons 1997, 1999; Mayer, Caruso, & Salovey, 2000) reported finding a positive relationship between emotional inte lligence and satisfaction with life.

PAGE 43

34 Second, previous research suggests seve ral well known personality constructs such as self-esteem, depression, and locus of control, are related (correlated) to satisfaction. Third, Di ener (1984) as well as Die ner and Larsen (1993) have reported similar findings suggesting satisfac tion with life may be the most stable component of subjective well-being. Four th, Pavot and Diener (1993) as well as Schuessler and Fisher (1985) suggested satisf action with life is relatively stable and consistent over time. In support of these findings, Pavot, Diener, Colvin, and Sandvik (1991) noted that although day to day fluctuations in mood and daily events can slightly influence subjective reports of satisfaction with life, the consensus is that considerable stability exists in satisfaction with life. In a more recent study, Diener et al., (1999) asserted that Defined as an individuals overall appraisal of the quality of her or his life, satisfaction with life incorporates but also transcends the immediate effects of life events and mood states (p. 276). In summary, the affective components of subjective well-being are important. However, satisfac tion with life (the cognitive component) was chosen as the dependent measure for the present investigation rather than affective wellbeing. Satisfaction with life was chosen fo r the following reasons, previous research suggested that: (a) emotional intelligence is related to satisfaction with life, (b) self-esteem, depression, and locus of control are related to satisfaction with life, (c) satisfaction with life has greater stab ility than affective well-being, (d) satisfaction with life is related to many other important variables (e.g., health

PAGE 44

35 attributes) and (e) satisfaction with life may be the key indicator of the more global construct subjective well-being. Measurement of Satisfaction With Life and Instruments A review of the literature suggested that at least some of the published studies failed to adequately differentiate between satisfaction with life and related constructs (e.g., quality of life and subjecti ve well-being). Diener (1994) stated that The definitions of satisfaction with life are often not made explicit in the literature and are only implied by the ty pes of measures t hat are used (p. 104). Both subjective well-being and satisfac tion with life are quality of life measures. However, subjective well-being is composed of two elements, cognitive and affective. Diener (1994) stated that Life satisfaction, the cognitive component of subjective wellbeing, refers to a global judgment of a lif e as a whole. And The affective component of subjective well-being consists of ongoing reactions to events (p. 104). Gurin, Vero ff, and Feld (1960) conducted t he first American quality of life research. This study and similar studies (e.g ., Braburn & Caplovitz, 1965) typically used objective measures of quality of life (e.g., income, place of residence, food supply, crime rates, and education level) However, Andrews and Robinson (1991) as well as Argyle (1987) Diener (1994) and Diener and Suh (1997) noted that studies with adults as well as children demonstrated only a weak relationship exists between objective factors and an indivi duals life satisfaction. For example, Diener, Sandvik, Seidlitz, and Diener (1993) reported finding a correlation between income and life satisfaction of r = .12 in a nationally r epresentative sample of

PAGE 45

36 adults in the United States. However, this small yet statistically significant and interesting relationship was not reported by Clark and Oswald (1994) in their study of the effect of income on satisfaction with life in a nationally representative sample from Britain. Thus, even the small relationship between income and satisfaction with life reported by Diener, et al., (1993) wa s not found in this cross cultural study. Shmotkin (1990) as well as Okma and Veenhoven (1996) noted that any small decline in satisfac tion with life with increasing age is eliminated when other variables such as income are controll ed for. Other studies of satisfaction with life have also suggested limitat ions of using only objective measures. Campbell, Converse, and Rogers (1976) argued that obj ective measures (e.g., age, sex, income, race, education, and marital status) accounted for less than 20% of the variance in satisfaction wit h life in their study. Altho ugh much of this research is more than twenty years old, more recent investigations such as Diener and Suh ( 1997) as well as Diener et al. (1999) reported similar small correlations between objective factors and satisfaction with life. Further research such as Pinquart and Sorensen (2000) reported a relationship between satisfaction with life and num erous demographic variables (e.g., education, income, and social class). Th is research suggested that social economic status explains 2.2% to 3.2% of the variance in satisfaction with life (p. 197). In this study, the combined influence of socioeconomic status, social support, and activity levels were found to be significantly and positively related to satisfaction with life. However, when a hi erarchical analysis was performed, the

PAGE 46

37 demographic variables when considered toge ther accounted for less than 15% of the variance in satisfaction with life. Becaus e of the small effect sizes reported in many of the published studies, researcher s have turned away from the exclusive use of demographic or objective variables in investigations of satisfaction with life. Bearsley and Cummins (1999) as well as Argyle (1999) reported findings that suggest satisfaction with life is largely regulated by internal mechanisms rather than objective factors. The Bearsley and Cummins (1999) study (N = 524) compared two groups of youths, one group c onsisted of homeless youths while the other group consisted of youths with hom es. Their results suggested that the level of satisfaction with life reported by bot h groups of youths is largely regulated by internal mechanisms (p. 208). All of this research taken together suggests objective predictors of satisfaction with life may not account for much of the variance in satisfaction with life among children and adults. One explanation for the relatively small amount of variance in satisfaction with life accounted fo r by objective variables may be found in the individual rather than the situation. Individuals may give very different personal meaning to the same objective situation. Huebner (1994) as well as Huebner, Gilman, and Laughlin (1999) suggested the limitat ions of objective factors have led to increasing appreciation of the importance of subjective factors in the prediction of sa tisfaction with life. Diener et al. (1999) stated that People react di fferently to the same circumstances, and they evaluate conditions based on their unique expectations, values,

PAGE 47

38 and previous experiences (p. 277). Diner (2000) as well as Diner et al. (1999) defined satisfaction with life (SWL) as an evaluation of ones own happiness and satisfaction. This self assessment typica lly involves four main components: (a) pleasant emotions, (b) unpleasant emotions (c) global life satisfaction, and (d) satisfaction in specific life domains. Other researchers (e.g., Heady & Wearing, 1989) have investigated the stability of sa tisfaction with life across time. Their research suggested that while positive and negative events may influence slight shifts in satisfaction with life from est ablished baselines, most individuals tend to return to their usual level of satisfacti on with life within a few days. This line of research suggested that ther e may well be both state-like (situational) and trait-like (dispositional) factors involved in the determi nation of ones satisfaction with life. In agreement with this line of research, Stones, Hadjistavropoulos, Tuuko, and Kozma (1995) reported results suggesting that dispositional factors may explain more of the variability in life sati sfaction than situational factors. A review of the SWL literature suggested t here have been two approaches to the investigation of satisfaction with life. Huebner (1996) argued that one approach is the one-dimensional method wh ile the second approach is multidimensional. The one-dimens ional approach measures satisfaction with life as a global construct and measures an individuals subjective evaluation of the quality of his or her life in general. The mult idimensional approach measures life satisfaction within various life domains such as work, school and family. Lewinsohn, Redner, and Seeley (1991) stated that The existence of global life

PAGE 48

39 satisfaction is supported by research findings reporting moderate positive correlations among satisfaction in various domains of a persons life (p. 142). Huebner (1996) stated that Life satisfaction may be assessed as a multidimensional construct, whic h can be separated into satisfaction with various life domains (e.g., school, wo rk, social) (p.131). The advantage of measuring satisfaction with life as a global construct is that a single summed score allows compar isons of group differences in satisfaction with life (e.g., students with learning disorders and students without learning disorders). However, one dis advantage or limitation of measuring satisfaction with life as a global construc t involves ignoring differences in satisfaction with life withi n specific life domains. The number of measures developed to measur e satisfaction with life, the cognitive component of subjective well-bein g, are far fewer t han those developed to measure the affective com ponent of subjective well-bei ng. The majority of the early satisfaction with life measurement scales consisted of single items. For example, Gurin, et al. (1960) simply asked participants to report how happy they were. Bradburn and Caplovitz (1965) asked participants to respond to how would you say things were these days, would you say you are very happy, pretty happy, or not too happy? Another well known sing le item scale is t he Delighted-Terrible Scale (D-T) developed by Andrews and Wi they (1976). The D-T Scale is a oneitem scale that requires subjects to rank how they feel about their current level of happiness on a seven-point Likert-type scale rangi ng from delighted to terrible. It is

PAGE 49

40 not possible to calculate internal reliabiliti es of one item scales (you just cant split one item). However, Diener ( 1984) reported average test-retest reliability for the D-T scale of r = .66, at 15 minutes apart, and a si x-month averaged reliability of r = .40 which suggest as suspected (from si ngle item measures) relatively low testretest reliability. Andrews and Withey (1976) noted that score s on one item measures tend to be skewed with most responses falling in the delighted or satisfied range. The social desirability response set ma y explain many of the low test-retest reliability estimates of single item scales. Participants may not wish to describe themselves as only unsatisfied or satisf ied. The transparency of the measure coupled with very short intervals between te sts may result in very low reliability estimates. Regarding single item scales Diener (1984) stated that They do not offer a differentiated view of a pers ons satisfaction with life (p. 544). An advantage of multi-item scales is that parti cipants may be willing to give more varied and genuine responses. Marsh, Barnes, and Hocevar (1985) as well as Diener (1994) suggested t hat while the interpretation of single item measures are easy, they posses import ant psychometric limitations (e.g., low reliability). Thus, there are two important reasons why multi-item scales were developed. First, because the psychometric properties of multi-item scales are an improvement over those of single item scales (re sponse bias presents less of a threat and internal consistencies can be estimated). Second, because multi-item scales offer a more differentiated or holistic view of satisfaction with life.

PAGE 50

41 Satisfaction With Life Scale The satisf action with life scale (SWLS) was developed by Diener, Emmons, Larsen and Griffin (1985). It is a 5-item scale designed to address the limitations of single item scales by measuring global satisfaction with life as a cognitive judgmental process. Pavot and Diener (1993) st ated that Satisfact ion with life is a global judgment, theoretically predicted to depend on a comparison between ones life circumstances and subjec tive standards (p. 165). This may be accomplished by asking participants to rate their satisfaction with life as a whole (in general) in order to obtain an overall index of life sati sfaction. The SWLS allows participants to subjectively integrate and weigh all of the important life domains. Subjects rate their satisfaction with each item using a seve n point Likert scale that ranges from a score of 1 = strongly dis agree to 7 = strongly agree wit h 4 as a neutral score. Examples of items from the SWLS are; I am satisfied with my life and The conditions of my life are excellent. The origin al SWLS was composed of 48 se lf-report items which included questions that measure both the cognitive and affective domains. Factor analysis allowed elimination of all items with loading of less than .60 and items measuring the affective domain. The result was a revised scale containing ten items. The developers then removed five additional item s because of high semantic similarity. Gilman and Huebner (2003) noted that what emerged was t he five item narrowly focused SWLS that is now widely known and used in social science research today (p. 195).

PAGE 51

42 Stability of m easurement versus sensitivity to change is a critical issue for any assessment instrument. Pavot and Diener (1993) argued that measures of life satisfaction must demonstrate that they are measuring more than momentary changes in emotion. At the same time, they must demonstrate that they are sensitive enough to detect changes in satisfacti on with life, such changes as those occurring during psychotherapy or those due to major life events (e.g., death of a loved one). Diener et al. (1985) stated t hat Regarding the p sychometric properties (construct validity) of the SWLS it seems to measure what it purports to measure (p .74). In the initial validity and reliability test ing of the SWLS, Diener et al. (1985) reported an internal consistency and two-mont h test-retest reliability with a sample of 300 undergraduates from the Un iversity of Illinois, as having a correlation of r = .82, and a coefficient alpha of r = .87. More recent re searchers have examined the satisfaction with life scale for internal consistency and test-ret est reliability. For example, Alfonso and Allison (1992) reported a coefficient alpha of r = .89 and a test-retest correlation of r = .83 with a two week interval Other researchers (e.g., Pavot, Diener, Colvin, & Sandvik, 1991) report ed results from two samples, one (N = 39) composed of elderly persons (age 53-92), the other sample (N = 136) composed of undergraduates (age 18-29). Rese archers in this study reported a coefficient alpha of r = .85 and a test-retest reliability of r = .84 with a four week interval.

PAGE 52

43 In agreement wit h this line of research other researchers (e.g., Yardley & Rice, 1991) investigated the reliability of the SWLS among undergraduate students (N = 65) and reported a coefficient alpha within a range of r = .80 to .86, and a test-retest reliability of r = .50 with a ten week inte rval. Other researchers (e.g., Magnus, Diener, Fujita, & Pavot, 1993) have investigated intervals as long as two hundred and eight weeks (4 years) with a sample of young adults and reported a coefficient alpha of r = .87 and a test-retest reliability of r = .54. Pavot and Diener (1993) reported on the basis of data from several different samples that the SWLS reflects a one dimensional, internally consistent measure of life satisfaction. Vitaliano, Russo, Young, Becker, and Maiuro (1991) as well as Magnus, et al., (1993) have contributed to the research base regarding the re liability of the SWLS as well as the validity of the sa tisfaction with life construct. These researchers noted that result s from their studies suggest ed the SWLS can detect change over time, such as the increase of satisfaction with life after a period of psychotherapy or the decrease in satisfaction with life as ones spouse becomes more debilitated. In general, results report ed from the above cite d studies suggest the satisfaction with life scale demons trates both high moderate internal consistency (range of r = .80 to .89) and moderate test -retest reliability (range of r = .50 to .87) within the cont ext of the above cited studies. In summary, the above cited research sugges ts that there is relative longterm consistency of life satisfaction over time. The research suggests that the

PAGE 53

44 SWLS measures more than momentary mood changes. The SWLS appears to be reliable sensitive enough to detect changes in life satisfaction across time (e.g., after the death of a loved one, after divo rce, after completion of psychotherapy). Anas tasi (1988) noted that c onstruct validation is a process of gradually accumulating information from a variety of sources about a construct and what may influence it. Cronbach and Meehl (1955) ci ted three methods of establishing construct validity. First, demons trate that the internal fact or structure of a measure is consistent and stable. Second, demons trate the measure has adequate convergent validity with measures of theoretically related c onstructs and discriminate validity with measures of cons tructs from which it should be distinct. Third, demonstrate the measure is relat ed to theoretically important external criteria (e.g., College G.P.A., occupational success, and satisfaction with life). Theore tically related evidence of construc t validity for the SWLS begins with the groups scoring lowest on the measur e (e.g., prisoners and psychiatric patients). These groups as well as others (e.g., homeless) are expected to score low on measures of satisfaction with lif e. Pavot and Diener (1993) argued that Satisfaction as we conceptualize it curr ently involves a comparison of our situation with self-imposed subjective standards (p. 164). In essence, we evaluate our level of life satisfaction by comparing ou r perception of our life situation against what we believe our life situation should be. Thus, events or conditions that makes the individuals circumstances better or worse will influence life satisfaction.

PAGE 54

45 In ag reement, Diener et al. (1985) reported findings from their study that included more than 300 undergraduate students and 53 elderly persons (75 years or more of age) suggesting a negative correlation of r = .31 between the SWLS and negative affect. Arrindell, Meeuwesen, and Hu yse (1991) investigated the psychometric properties of the SWLS with a sample (N = 107) of adult medical outpatients (ages 18 65) and r eported similar findings. This study reported results suggesting the SWLS is negatively correlat ed with all eight sympt om dimensions assessed by the Symptom Checklist-90 (SCL-90) including depression ( r = .55), anxiety (r = .54), and general psychological distress (r = .52). In terms of individual difference dimensions, Diener et al. (1985) as well as Pavot and Diener (1993) reported results suggesting a posit ive correlation between the SWLS and extraversion, as well as a negative correlation between the SWLS and neuroticism, which suggested construct validity. Pavot and Diener (1993) suggested the SW LS has demonstrated adequate convergence with related measures of life satisfaction (e.g., Andrews & Withey Scale, r = .68; Fordyce Global Scale r = .58), including studies employing different methodological approaches (e.g., intervie ws & informant ratings). Campbell, Converse and Rogers (1976) reported resu lts from their study of undergraduates that suggest the SWLS correlates with other life satisfaction scales include-ing the Semantic Differential-Like Scale ( r = .75); Well-Be ing Sub-scale of the Differential Personality Questionnaire ( r = .68); Self-Anchoring Ladder ( r = .66) and Affect Balance Scale ( r = .50) for positive affect and ( r = .37) for negative affect.

PAGE 55

46 Pavot and Diene r (1993) reported results sugges ting construct validity for the SWLS after investigating the relations hip between the SWLS and interviewers and informants ratings. The results of this study suggested moderate convergence of self-reports of satisfaction with life with interviewers and informants who were asked to judge their life satisfaction. Other studies which investigated the relationship between the SWLS and interv iewer / informant measures included Pavot et al. (1991) who reported a correlation of r = .54 between t he SWLS and informant reports. Deiner and Allman (1991) reported results of a study with undergraduates (N = 189) at t he University of Illinois, and reported a correlation of r = .58 between the SWLS and informant reports. Frisch (1991) reported a correlation of r = .66 between the SWLS and interviewer ratings as well as a correlation of r = .28 between the SWLS and informant reports. Pavot and Deiner (1991) reported a correlation of r = .46 between the SWLS and informant reports. Judge (1990) reported a correlation of r = .43 between the SW LS and informant reports among medical students The relati onship between satisfaction with lif e and theoretically important external criteria has been investigated by several researchers. For example, Lewinsohn, Redner, and Seeley (1991) repor ted findings from a non-clinical sample suggesting a high moderate relationship ( r = .69) between decreasing satisfaction with life and the onset of depres sion two to three years later. Marks and Flemming (1999) in their study analyzed data from the Australian Youth in Transition study, a longitudinal study of f our nationally represent ative cohorts of

PAGE 56

47 young people (initial total N = 20,000).This study reported findings suggesting a low to moderate positive correlation between satisfaction with life and occupational success. Furr and Funder (1998) reported fi ndings from a study involving undergraduate students (N = 146) suggesting even non-clinical levels of self dissatisfaction may have important consequenc es on quality of interpersonal relationships. Other resear chers investigating the relati onship between satisfaction with life and educational outcomes (e.g., Fri sch, Clark, Rouse, Rudd, Paweleck, Greenstone, & Kopplin, 2005) reported results suggesting a moderate positive relationship between satisfaction with life and school retention. Theoretical Development of Emotional Intelligence While the l abel emotional intelligence may be relatively new to some researchers, the idea has been ar ound for some time. Some researchers as early as the 1920s (e.g., Thorndike, 1920) were sugges ting that social intelligence, The ability to understand others and to act or behav e wisely in relation to others was an important component of intelligence (p. 228). Gardner (1983) published Frames of mind: The theory of multiple intelligences in which he proposed his theory of multiple intelligences. With this model, Gardner proposed that Interpersonal and intrapersonal intelligences comprises an individuals social intelligence (p. 239). Law, Wong, and Song (2004) noted that Salovey and Mayer (1990) were two of the first researchers to build upon this model and conceptualize emotional intelligence as the ability of a per son to deal with his or her emotions.

PAGE 57

48 They defined emotional intelligence as The sub-set of social intelligence that involves the ability to monitor ones own and others feelings and emotions, to discriminate among them and to use this in formation to guide ones thinking and actions (p. 189). Thus, it may be argued that the notion of emotional intelligence began with E. L. Thorndikes concept of so cial intelligence or Howard Gardners concept of multiple intelligences (especially social intelligence). Plucker (2003) argued that the nature of the human intellect has fascinated scholars for centuries. However, the ear liest modern concepts of intelligence and intelligence testing evolved during the first half of the 20th cent ury. Alfred Binet and Theodore Simon (1905/1916) developed t he Binet-Simon Intelligence Scale. During the first three-quarte rs of the 20th century, in telligence and emotion research were largely separate fields. In the case of intelligence, abstract reasoning was stressed to differentiate it from pers onality theories. While in the case of emotion, most investigations pursued one of two distinct paths. One path of investigation focused on biological associations, as earlier researchers beginning with Darwin had argued emotions evolved over time and were primitive impulses to act. The second path of investigation focus ed upon the social adaptive or cultural aspects of emotions. Whether emotions we re a product of biology or culture or some interaction of both, emotions were held separate from the intellect (Mayer, 2001). Several ev ents beginning in the early 1970 s radically influenced how intelligence is both conceptualized and meas ured. First, the cognitive movement

PAGE 58

49 inadvertently stimulated interest in emotions as well as interest in the relationship between thoughts and emotions. Mayer (2001) noted that the investigation of cognition and affect merged to examine how emotions interacted with thoughts. Second, both artificial intelligence and non-verbal communication were examined in connection with cognition and affect (Maye r, 2001). Third, Robert Sternberg (1985) advanced his theory of practical intelligence in his publication, Beyond IQ: A triarchic theory of human intelligence. In this book, Sternberg conceptualized intelligence as comprising three different aspe cts of intelligence the mental abilities necessary for: (a) adaptation to, (b) the shaping of, and (c) the selection of an environment. The key theme of this model is adaptation that he refers to as practical intelligence (Sternberg, 1997, p. 1030). Fourth, Howard Gardner (1983) proposed a theory of multiple intelligences arguing that there were many different ways to be intelligent (Pfeiffer, 2001). Gardners or iginal model of multiple inte lligences contained seven intelligence: however, the model was revised to include an eighth primary intelligence: (a) verbal, (b) mathematical-logical, (c) spatial, (d) kinest hetic, (e) musical, (f)interpersonal, (g) intrapersonal, and (h) naturalistic (Gardner, 1983).The construct of intrapersonal intelligence was used by Gard ner to mean social intelligence that included such components as social skills empathic proficiency, pro-social attitudes, social anxiety, emotionality and sensitivity. Mayer (2001) noted that prior to his and associate Peter Saloveys (1990 ; 1993) publications the term emotional intelligence was sporadically used in refe rence to an intertwining of social know

PAGE 59

50 ledge and access to those social and emoti onal feelings. Tenhouten, Hoppe, and Bogen (1986) noted that brain research began to separate out connections between emotion and cognition. Similar resear ch (e.g., Marlowe, 1986) reported that Empirical research in social intelli gence was discovered to divide into social skills, empathy skills, pro-social attitudes, so cial anxiety, and emotionality (p. 57). The construct emotional intelligence, as we know it today, began with a series of papers published in the professional literature by John Mayer and Peter Salovey (1990, 1993). Mayer and Salovey (1993) argued emotiona l intelligence was a distinct cognitive ability and thus a long overlooked intelligence which promises to meet the standard of a basic inte lligence (pp. 433-434). A pivotal event in the ev olution of emotional in telligence was Daniel Golemans (1995), publication of Emotional Intelligence which soon became a best seller. Golemans theory of emotional in telligence focuses on motivational and social relationship factors. In contras t, the framework of Ma yer and Saloveys (1997) ability model focuses on ability to understand and process emotions. Ciarrochi, Forgas, and Mayer (2001) noted that the second half of the 1990s witnessed an accelerated period of refinem ent of both theore tical models and measures of emotional intelligence. In summary, the term emotional intellig ence first appeared in two academic articles authored by John Mayer and Pete r Salovey (1990; 1993) which at the time generated relatively little interes t. The popularization of the construct

PAGE 60

51 emotional intelligence followed Daniel Gole mans (1995) publication of his bestseller titled Emotional Intelligence. Thus, two si milar yet distinct models of em otional intelligence developed. First, the more academic cognitive ability model developed by Mayer, and Salovey (1990) defined emotional intelligence as The subset of social intelligence that involves the ability to monitor ones own and others feelings and emotions, to discriminate among them and to use this information to guide ones thinking and actions (p. 189). Mayer and Salovey (1997) revised this conceptualization to make more explicit how emotional intelligence represents cognitive abilities. The cognitive ability model narrowly defines emot ional intelligence as a set of cognitive abilities which together constitutes a relatively distinct intelligence. The cognitive ability model makes relatively conservative claims about the importance of EI to important outcomes. Second, is the mixed model which is more popularly oriented and based largely on the work of Daniel Goleman and associates. This model mixes emotional intelligence as an ability with social competencies, personality traits, and behaviors (Cobb & Mayer, 2000). Relationship of Emotional Inte lligence to the Present Study Emotional inte lligence is the primary focus of investigation in the present study. The primary research question in the present study is, does emotional intelligence account for additional variance in life satisfaction not accounted for by other known predictors (e.g., self-e steem, depression, and locus of control)?

PAGE 61

52 Measurement of Emotional Intelligence and Instruments A review of the EI literature revealed that there are several similar yet different operational definitions of emoti onal intelligence. Several authors have contributed to the notion of emotional intelligence, su ch that the multitude of qualities covered by the construct requires specification of the particular model. Among the leading theorists, Mayer and Sa lovey (1997) argued emotional intelligence is a cognitive ability and proposed a four component model which include these abilities: (a) perceive and accurately express emotion, (b) use emotion to facilitate thought, (c) understand emoti ons, and (d) manage emotions for emotional growth. Another l eading theorist, Daniel Goleman (1995) described emotional intelligence as composed of five dimens ions: (a) self-awareness, (b) selfregulation, (c) motivation, (d) empathy, and (e) social skills. In contrast to the more narrowly defined cognitive ability model, Gole man (1995) almost defines emotional intelligence by exclusion. He argued a la rge number of human abilities fall within the emotional intelligence construct frustra tion tolerance, delay of gratification, motivation, zeal, persistence, impulse cont rol, regulation of mood, hopefulness, and optimism (p. 6). In contra st to the ability model of EI Golemans mixed model makes relatively broad claims as to the importance of EI to important outcomes (e. g., leadership). In Gole mans 1995, publication entitled Emotional Intelligence he stated that EI is equal to if not more valuable than IQ as an indicator of ones professional and life success (p. 34). In his second book entitled Working with

PAGE 62

53 Emotional Intelligence Goleman (1998) elaborated on the importance of EI in the work place. In a subsequent publication The Emotionally Intelligent Workplace Goleman (2001) argued that EI more than any other asset is the most important overall success factor in ca reers and EI accounts for 85 % to 90% of the success of organizational leaders (p. xv ). In his most recent book Primal Leadership: Realizing the Power of Emotional Intelligence, Goleman, Boyatzis, and McKee (2002) asserted that The effective use of emotion is basic to successful leadership and The emotional task of the leader is primal: It is both t he original and the most important act of leadership (p. 5). A third leadi ng theorist, Bar-On (2000) defined emot ional intelligence as An array of non-cognitive capabilities, com petencies, and skills that influence ones ability to succeed in coping with environm ental demands and pressures (p. 1108). This model is very similar to the mixed model of emotional intelligence proposed by Goleman (1995). Beyond the relatively distinct mixed model proposed by Daniel Goleman and the cognitive ability model proposed by Mayer and Salovey, all other models of emotional intelligenc e (e.g., Bar-On, 1997; Wong & Law, 2002) share a great deal of similarity. In their revi ew of the EI literatur e, Ciarrochi, Chan, and Caputi (2000) stated that W hile the definitions of EI are often varied for different researchers, they nevertheless tend to be co mplementary rather than contradictory (p. 540). Law, Wong, and Song (2004) argued that Although definitions of emotional in telligence are not identical, the differences between definitions tend to be minor (p. 484).

PAGE 63

54 The resear ch on ability emotional intelli gence began in the early 1990s (e.g., Salovey & Mayer, 1990, 1993). However, within half a decade, Goleman (1995) expanded the initial versio n of the concept such that it included traits such as frustration tolerance, del ay of gratification, optim ism, motivation, and wellbeing along with aspects of ability emotional intelligence. T he broad nature of Golemans theorizing in the final analysis defined emotional intelligence as very similar to character (Goleman publishes primarily in the popular press). Thus, emotional intelligence is typically conceptualized in the professional literature from three different models. First, a four com ponent cognitive ability model (e.g., Mayer & Salovey,1997). Second, a five component model (e.g., Bar-On, 1997). Third, a model consisting of five dimens ions (e.g.,Goleman, 1995). The literatu re suggested two primary ways re searchers measure emotional intelligence and each reflects a different mo del of emotional in telligence. Carroll (1993) noted that researchers investigating emotional intelligence as a cognitive ability, a distinct intelligence, utilize standard performance scales because they are based on the capacity to solve mental tasks (e.g., MSCEIT). However, researchers investigating emot ional intelligence from the mixed model perspective utilize self-report scales for example, the Bar-On Emotional Quot ient Inventory (EQ-i) and the Self-Report Emotional Intelli gence Test (SREIT). These self-report scales are based on subjective endorsements of descriptive statements regarding themselves (e.g., I feel su re of myself in most situat ions). Paulhus, Lysy, and Yik

PAGE 64

55 (1998) stated that Empirical studies sugge st that the correlations between ability and self-report measures of intelligence are relatively low ( r = .00 to .35) (p. 550). There are tw o arguments against self-report me asures of EI discussed in the EI literature. First, Geher, Warner, and Brown (2001) argued that the social desirability bias may skew reporting. Subjec ts may simply respond in such a way as to appear in a more positive light or avoid appearing socially undesirable. Second, Mayer, and Geher (1996) as well as Ma yer et al., (2001) asse rt that if the subjects reporting do not have an accurate understanding of themselves and their abilities, then the data gathered will not render an accurate measure of the subjects ability. Bracket and Mayer (2003) stated Therefore with respect to emotional intelligence, it is likely that ability a nd self-report models wil l yield different representations of the same person (p 1147). Also, they noted that At the present date there are only th ree full-scale tests of emotional intelligence (EQ-i; SREIT; MSCEIT) in the scholarly literature for which preliminary empirical data are now available (p. 1148). An extensive s earch of the literature revealed four additional instruments frequently used to measure EI. Howeve r, it is no surprise that of the many instrum ents purporting to measure emot ional intelligence, these same three emotional intelligence tests, (a) the Emotional Q uotient Inventory (EQ-i) (Bar-On, 1997); (b) Self-Report Emotional Intelligence Test (SREIT) Schutte, Malouff, Hall, Haggerty, Cooper, and Golden (1998); and (c) the MayerSalovey-Caruso Emotional Intelligence Test (MSCEIT) (Mayer, Salovey, & Caruso, 2002) are the most fr equently cited as well as the best known. Each of

PAGE 65

56 these EI instruments will be discussed in turn beginning with the four instruments lacking empirical validation of their psychometric properties in the literature followed by those with psyc hometric data available. The Emotional Compet ence Inventory (ECI). The ECI is a self-report scale designed to measure emoti onal competencies. This scale was developed by Boyatzis and Goleman (1998) and based on Golemans (1995) mixed model of EI. This model links personality with per formance and presents a relatively noncognitive conceptualization of EI. The psychom etric properties of this instrument remain largely unknown. The Emotional Intelligence Inventory (EII). Tapia (2001) developed this scale based upon Mayer and Saloveys (1997) ability model of emotional intelligence. The EII is a 45 item self-report inventory designed to measure the emotional intelligence of hi gh school students. Some preliminary research (e.g., Tapia, 2001) suggest high internal consiste ncy (r = .81) however the instrument has not been validated with colle ge students. The psychometri c properties of this instrument remains largely unknown. Trait Meta-Mood Scale (TMMS). The TMMS was developed by Salovey, Mayer, Goldman, Turvey, and Palfai (1995). Although desi gned to measure reflective mood the TMMS has been used by several researchers (e.g., Salovey, et. al., 2001; Palmer, Walls, Burgees, & Stough, 2001; Palmer, Donaldson, & Stough, 2002) as a measure of perceived emot ional intelligence. Salovey et al. 2001 report the TMMS demonstrates conver gent validity and evidence of divergent validity among subscales; however, t he psychometric properties (reliability,

PAGE 66

57 construct validity) of this instrument remain largely unknown. The TMMS is a measure of perceived emotional intelligenc e and thus not an appropriate measure of EI (cognitive ability ) for the present study. The Wong and Law 16-Item Emotional Intelligence Measure. This 16-item measure was developed by Wong and Law (2002) based upon Mayer and Saloveys (1997) ability model of EI. The scale is a self-report measure primarily used in organizational research however the psychometric properties (reliability, construct validity) of this m easure remain largely unknown. The Bar-On Emotional Intel ligence Inventory (EQ-i). is a 133 item self-report measure of emotional intelligence. Res pondents answer questions using a five point Likert-type scale (1 = very seldom or not true of me, 5 = very often true of me). The test publisher provides scoring whic h consists of a total EQ-i score and five composite scores. The composite sco res consist of (a) intrapersonal EQ, (b) interpersonal EQ, (c) adaptability, (d) stress management, and (e) general mood. Three examples of items from the EQi include I feel sure of myself in most situations, I have strong im pulses that are hard to contro l, and It is easy for me to make friends. The EQ-i generally takes about 35 40 minutes to complete and is appropriate for individuals 16 years of age and above. Several res earchers, including Dawda and Hart (2000); Newsome, Day, and Catano (2000); Parker, Taylor, and Bagby (2001) have published recent studies that suggest the EQi is strongly correlated with se veral personality constructs, such as depression, anxiety, and alexithymi a (a disorder which involves the inability to understand and or express emotions). Bar-On (2000) reported in a

PAGE 67

58 study of normal college students that The EQi substantially overlapped with several measures of anxiety and the Symp tom Checklist-90 (SCL-90), which is a general indicator of social and emotional functioning (p. 364). Other researchers, (e.g., Davies, Stankov, & Roberts 1998; Roberts, Zeidner, & Matthews, 2001) have suggested self-report measures like the EQ-i and the SREIT may simply reassess basic personality. Self-Report Emotional In telligence (SREIT). The SREIT is a brief self-report measure of emotional intelli gence developed by Schutte et al., (1998). The authors of the SREIT developed this instrument based primarily upon Mayer and Saloveys (1990; 1993) model which conceptualized em otional intelligence as the ability to monitor and discriminate emotions and to use emotions to guide ones thinking and actions (Bracket & Mayer, 2003). Partic ipants respond to 33 self-report items such as I know why my emotions change using a 5-point Likerttype scale, in which 1 represents strongly disagree and a 5 represents strongly agree. For example, some of the inst ruments items measure a persons self-perceived ability to monitor private feelings or the feelings of others. Schutte et al., (1998) noted that the SREIT correlates moderately to strongly with a number of personality constr ucts, including alexithymia, r = .65, p < .001; optimism, r = .52, p < .006; impulse control, r = .39, p < .003; and openness to experience, r = .63, p = < .001 (p. 171). Brackett and Ma yer (2003) argued that Most of the attributes measured by the EQi and SREIT substantially overlap with existing measures, which suggests that t hese scales have a breath of coverage

PAGE 68

59 that is not all that diffe rent from well-studied pers onality and well-being scales (p. 1150). The research cited above suggests the EQi and the SREIT are selfreport instruments which do not appear to be valid measures of emotional intelligence when conceptualized from the four part cognitive ability model proposed by Mayer and Salovey (1997). Multifactor Emotional Intelligence Scale (MEIS ). Mayer et al., (1999) developed the (MEIS), the first in strument designed to measur e emotional intelligence when conceptualized as a cognitive ability. This instrument was intended to measure EI according to Mayer and Salo veys (1997) four component cognitive ability model that includes the following: (a) the ability to perceive emotions in oneself and others, as well as in objects, art, and stories (perception of emotion), (b) the ability to generate emotions in order to make use of them in other mental processes (e.g., emotional facilitation of thought), (c) the abilit y to understand and reason about emotional information and ho w emotions combine and progress through relationship transiti ons (understanding emotions), and (d) the ability to be open to emotions and moderate them in one self and others (managing emotions). However, fa ctor analysis performed on the ME IS, by the instrument developers Mayer et al., (1999) as well as other researchers (e.g., Roberts, Zeidner, & Matthews, 2001) reported recove ring only three of four fact ors: (a) Perception, (b) understanding, and (c) regulati on of emotion. This lack of psychometric validity coupled with the length (402 it ems) of the MEIS contribut ed to the development of the MSCEIT.

PAGE 69

60 Mayer, Salovey, and Caruso, Emotional Intellige nce Test (MSCEIT) is the direct descendent of the MEIS. Like the MEIS (1999) the MSCEIT (2000) was developed by the original authors, John Mayer and Peter Salovey along with colleague David Caruso. Upon initial psychom etric evaluation Mayer, Caruso and Salovey, (2000) as well as Mayer et al., (2003) noted that the MSCEIT appears to be content valid and possesses a factor structure congruent with the Mayer and Salovey (1997) four-component cognitive ability model of emotional intelligence. Participants respond to 141 items, endorsing one of five choice alternatives on a Likert-type scale for different problems with 1 = indicating no happiness, and 5 = indicating extreme happiness. The MS CEIT yields 5 different scores of interest in the present study. First, a tota l score, which is an overall index of the respondents level of emotional intelligence according to the model. Second, the MSCEIT yields four branch scores (com ponent scores): (a) perceiving emotions score, which provides an index of how well the respondent can identify emotions in himself or herself and others, (b) fac ilitating thinking score which indicates the degree to which the respondent can use his or her emotions to improve thinking, (c) an understanding emotions score indica tes how well the respondent understands the complexities of emotional meanings, emotional transitions, and emotional situations, and (d) an emotional management score measures how well the respondent is able to manage emotions in his or her own life and in the life of others.

PAGE 70

61 Brac kett, Mayer, and Warner (2003) noted that the MSCEIT measures the ability to perceive emotions by showing people faces and designs and asking them to identify emotions in them. The us e of emotions to facilitate thought is measured by assessing peoples ability to describe emotional sensations and their parallels to other sensory modalities. Understanding emotions is measured by asking participants how emotions comb ine to form other emotions, and how emotions change over time. Emotion m anagement is measured by having testtakers choose among more or less effect ive means of emotional management in private and interpersonal emotional situations. There are two types of scori ng available for the MSCE IT general scoring and expert scoring. The developers of the MSCEIT (Mayer, Salovey, & Caruso, 2002) Recommend that most users employ the general scoring method rather than the expert scoring method (p 33). T he general scoring method utilizes the entire normative sample of 5,000 to score it em responses. For example, if 65% of the norming sample selected option B, as th eir choice for an individual item then the choice of B for that item would yield a score of .65. Similarly, if 15% choose option A, and 10% option C, as well as 10% option D then each of these responses would be scored .15, .10, and .10 respectively. Expert scoring was developed in a similar fashion however, instead of utilizing the normative sample, a sample of 21 emotion experts drawn from membership in the International Society for Research in Emotions (ISRE) was utilized. The sample of experts consisted of 10 men and 11 women aged 30 to 52 with a mean of just under 40

PAGE 71

62 and a standard deviation of 6.4. The general normative sample consisted of 5,000 individuals of which approximately 40% were males and 60% females, 73% were between 17 and 30 years of age and 58% reported having at least some college education. I have used t he general consensus method of scoring for the present study for the following two reasons. First, the te st developers Mayer, Salovey, and Caruso (2002) recommend the general consensus method of scoring in most settings. Second, the descriptors of the no rmative sample rather than the expert sample more closely resembled the obtained sample as well as the target population (CFCC students) of the present study. Mayer, Salovey, and Caruso (2002) reported the following correlations between MSCEIT general and expert scoring: total score r = .98; perceiving emotions r = .98; facilitating thought r = .97; understanding emotions r = .98; and m anaging emotions r = .96 (p 33). The above reported correlations between MSCEIT gener al consensus and expert scoring suggest a high degree of correspondence between expert and the general population sample. However, given the si milarity between methods of scoring I believe the general consensus method of scor ing is most suited to the present study. Brackett and Mayer (2003) report findings from their study with 207 predominantly Caucasian (97%) college student s. In this study the split-half reliability coefficients for the four branches ranged from r = .80 to .91, and for the total score r = .91. In the same study, test-ret est reliability was estimated by

PAGE 72

63 having 60 college students (18 men, 42 women) return 3 weeks after initial testing to retake the MSCEIT. The test-retest reliability was relatively high, r = .86, ( p < .001). Mayer, Salovey, Caruso, and Si tarenios (2003) conducted a study with 2,112 college age participants (58.6% women; 41.4% m en; (52.9%) of the participants were drawn from 36 academic settings from several different countries, in which confirmatory factor analysis supported the theory-driven four factor model of emotional intelligence. These analyses also found support for a general factor of emotional intelligence encompassing all four branches. Other researchers (e.g., Day & Carroll, 2004 p. 1451) reported similar fi ndings from their study with 246 undergraduate st udents (70 men and 176 women) from a Canadian University, suggesting overall the MSCEIT showed low correlations with the big five personality factors: (a) extraversion, (b) neuroticism, (c) conscientiousness, (d) agreeableness, and (e) openness to experience (r values ranged from .13 to 23, all significant at p < .05). These rela tively low correlations between the MSCEIT and measures of personality contribute to the establishment of the MSCEITs construct validity. Davies, Stankov, and Roberts (1998) as well as Newsome, Day, and Catano (2000) noted that their findings highlight the differences between the trait-based self-report measures of emot ional intelligence (e.g., EQ-i and SREIT) and ability based measures (e.g., MEIS & MSCEIT) which typically show greater discriminant validity with personality tr aits. In the present study, the MSCEIT was chosen as the instrument to measur e emotional intelligence for two reasons. First, the

PAGE 73

64 MSCEIT was chosen because it closely fi ts Mayer and Saloveys (1997) cognitive ability model of emotional intelligenc e. Second, unlike self-report measures, ability measures such as the MSCEIT have low correlations with personality constructs. Thus, ability measures are more suitable for assessing additional variance in a criteri on over personality. Theoretical Developm ent of Self-Esteem Since the c oncept of self-esteem first enter ed the discourse of social sciences more than 100 years ago, it has be come both an important and prolific research topic. Brown and Dutton (1995) stated that Self-esteem has become the panacea of modern life. It has been touted as the antidote to poverty, drug use, and under-achievement, and lauded as the royal road to financial success, health, and personal fulfillment (p. 712). Acco rding to Wells and Marwell (1976) ther e are four ways of defining selfesteem. First in the attitudinal definition, t he self is treated as an object. Just as people have cognitive, emotional, and behaviora l responses to objects, they can have them toward the self. Second, is a definition developed by social scientists to understand self-esteem that relies on attitudes, howeve r it is more formal focusing on the relation between different sets of attitudes (e.g., the differences between ones attitude toward goals and ac complishments, such as the importance one attaches to being loved and how much a person feels loved). The third method of defining self-est eem focuses on the psychological responses a person holds toward himself. The fourth method of defining self-esteem discussed by

PAGE 74

65 Wells and Marwell conceptualizes self-esteem as a component of personality, thus self-esteem becomes concerned with or a component of motivation and/or regulation. Mruk (1995) i dentified six major contributors to the development of the concept of self-esteem spanning more than 100 years. William James (1890) made the first reference to self-esteem; he defi ned self-esteem as Determined by the ratio of our actualities to our supposed pot entialities (p. 292). This conceptualization defines self-esteem as a fraction of which pretensions (self-imposed subjective demands) are the denominator and t he numerator our successes. Thus, James framed self-esteem as affective (it is lived as a feeling or emotion), as well as a dynamic process, affected by su ccesses and failures and thus open to enhancement or decay. During the next 60 years very little was said about selfesteem, its popularity declined mostly because of the behavioral insistence on observation and measurement which domi nated American psychology until after mid-century. The second majo r contributor to the theoretical development of self-esteem according to Mruk (1995) was White (1963) who conceptualized self-esteem as emerging from a complex dev elopmental framework char acterized by primitive impulses that are modified into the higher f unctions of the self over time. Like James, White conceptualized self-esteem as a developmental phenomenon, but more so in that self-esteem develops gradually, affected by and effecting both experience and behavior. White argued that self-esteem has two sources: an

PAGE 75

66 internal source (e.g. ones subjective accomplishments) and an external source (e.g., the affirmati ons of others). The third majo r contributor to the theoretical development of self-esteem is Rosenberg (1965) who takes a socio-cultur al approach by stating that By selfesteem we refer to the evaluation whic h the individual makes and customarily maintains with regard to himself, which expr esses an attitude of approval or disapproval (p. 5). This definition frames self -esteem as the product of culture, society, family, and interpersonal relationships The amount of self-esteem an individual has is proportional to the degree to whic h they positively measure up to a core set of self values. Rosenberg developed hi s theory after analyzing data from a large sample (N = 5,000) of adolescenc e 13 -17 years of age. Rosenberg introduced the notion of the impor tance of values in self-esteem, and thus opened the door to another important dimension of se lf-esteem. In additi on to self-esteem being a personal and psychological pheno menon, Rosenberg recognized selfesteem as a social phenomenon. Another signifi cant contributor to the theoretical developm ent of self-esteem is Coopersmith (1967) who defined self-e steem from a behavioral perspective noting that Self-esteem is a personal judgment of worthine ss that is expressed in the attitude the individual holds toward himself (p. 7). Coppersmiths (1967) publication of the Antecedents of Self-Esteem, was especially important because it represents the return of self-esteem to mainstream academic psychology. From this perspective, self-esteem is a construc t or an acquired trait. Thus, individuals

PAGE 76

67 learn how worthy they are init ially from their parent s or other caretakers. This initial self worth is reinforced by others; thus, children model the respect and worthiness of self they observe in t heir parents and others. This defin ition includes success as well as self-worth as an indicator of self-esteem. The next import ant contributor to the theoretical development of the selfesteem construct was Branden (1969) who defined self-esteem from the humanistic perspective. He was the first to descr ibe self-esteem in terms of two basic components: worthiness and competence. Mruk (1995) noted that This definition adds a new dimension of self-esteem to consider, the relationship between the components or how competence and worthiness interact with one another (p.139). Branden was one of the first to di scuss self-esteem as a basic human need and propose the lack of self esteem often has serious consequences (e.g., substance abuse, suicide, anxiety and depression). He considered competence, sense of personal worth, and self respec t all important values effecting selfesteem. In summary, Branden defined self-est eem as a measure of ones ability to live in such a way as to honor our view of ourselves. He seems to bridge the distinction between the cognitive and affe ctive evaluative components of selfesteem, which are imbedded in ot her definitions. The limitati ons of this theory are that the findings were derived exclusivel y from case studies and driven by a philosophy rather than empirical data. Although the decade of the sixties witnessed an increased interest in definitional work regarding self-esteem, subsequent decades have not been as pro

PAGE 77

68 ductive. Mruk (1995) noted that more recent work either repeat ed the themes of earlier works or used one of the existing def initions, choosing to focus on what factors influence self-esteem rath er than self-esteem itself. Another significant contributor to the theoret ical development of self esteem was Epstein (1985) who defined self-est eem from a cognitive-experiential perspective. Epstein (1985) argued that Sel f-esteem is a subjective and enduring sense of realistic self-approval. It reflects how the individual views and values the self at the most fundament al levels of psychological experiencing (p. 284). Like Branden, Epstein considered self-esteem a basic human need equating self-esteem with worthiness which motivates us both consciously and unconsciously. Epstein argued self-esteem is a consequence of an individuals understanding of the world and others we are in relation with. Thus, Epstein noted that we strive to maintain equilibrium of self. An import ant new dimension added to self-esteem by Epstein is the notion of levels of self-esteem. Epstein proposed there are three different levels of self-esteem: (a) global or general overall self-esteem, (b) intermediate self-e steem which is specific to certain domains (e.g., personal power) and (c) si tuational self-esteem which are the everyday manifestations of self-esteem. In addition to Epstein, self-esteem has been defined from a range of perspectives by numerous recent theorists such as Kernberg (1975) emphasizing primitive libidinal impul ses. Solomon, Greenberg, and Pyszczynski (1991) emphasized feelings of existential security in a meaningful universe. Other researchers

PAGE 78

69 (e.g., Baumeister, Tice, & Hutton, 1989) noted that within normal populations, high self-esteem is characterized by a general fondness for oneself while low selfesteem is characteristic of ambival ent or negative feelings toward oneself. In agreement with this line of reasoning, a review of the self-esteem literature suggests some relative consensus among researchers. Several researchers (e.g., Coopersmith, 1967; Harter, 1990; Baumeister, 19 93; & Rosenberg, 1979) proposed a cognitive model of self-esteem They assume self-esteem develops from a judgmental process in which people evaluate their various qualities, weight them by personal importance, and then sum up these values to derive an overall index of self-esteem. One constant which runs through much of the recent self-esteem research is the work of Morris Rosenberg (1965), a sociologist that conducted a study of self esteem with over 5,000 subjects. Rosenber gs brief definition of self-esteem is Simply a positive or negative attitude toward a particular object, namely the self (p. 3). Following Rosenbergs lead other re searchers (e.g., Joubert, 1990) proposed similar simplistic definitions such as Self-esteem is a personal judgment of general self-worth that is a product of an implicit evaluation of self-approval or self-disapproval made by the individual (p. 1147). However, Rosenbergs simplicity in definition, coupled with the simplicity of his 10 item Rose nberg Self-Esteem Scale makes him a prominent figure in both the theory and meas urement of selfesteem. Meisenhelder (1986) no ted that Self-esteem ma y be broadly defined as

PAGE 79

70 the degree to which one values oneself, and almost universally, self-esteem is measured using Rosenbergs Se lf-Esteem Scale (p. 8). Relationship of Self-Esteem to the Present Study It has consis tently been reported that self-esteem is positively related to satisfaction with life, the dependent variabl e in the present study (e.g., Diener, 1984; Emmons & Diener, 1985; Lewinsohn, Redner, & Seeley, 1991; Parkerson, Broadhead, & Tse, 1990: Schmitt & Bedei an, 1982; Vermunt, Spaans, & Zorge, 1989; Weiner, Muczyk, & Gable, 1987). Other researchers such as Lewinsohn et al. (1991) as well as Sekaran (1986) repor ted findings suggesting self-esteem to be the best predictor of satisfaction with life. Other researchers (e.g., Huebner & Alderman, 1993; Dew & Huebn er, 1994; Gilman, Huebner & Laughlin, 2000; Terry & Huebner, 1995) investigat ing the relationship between self-esteem and satisfaction with life among U. S. students reported correlat ions within a range of r = .40 to .60. Cultural differences in satisfaction with life have been well documented (e.g., Michalos, 1991; Myers & Diener, 1995). Several explanations have been offered for these cultural differences incl uding relative import ance of predictors that contribute to satisfaction with life, such as interpersonal relations and selfesteem. Kwan, Bond, and Singelis (1997) re ported findings from their study of college students in both the United States and Hong Kong that suggest selfesteem is a better predictor of satisfac tion with life among college students in the U.S. However, their findings also sugges t self-esteem is at least equal in

PAGE 80

71 importance with interpersonal relations among college students in Hong Kong. Their results were replicated by Uchida, Kitayama, Mesquita, and Reyes (2001) utilizing college students in the U.S., Japan, and the Philippines. They reported results that suggest self-esteem is the bes t predictor of satisfaction with life among college students in the U.S. Ho wever, they also reported findings that suggest selfesteem is at least equal to happiness and per ceived social support in predicting satisfaction with life among both Japanese and Filipino st udents. This research suggests that self-esteem is an important pr edictor of satisfaction with life cross culturally. However, just how important self-esteem is in the prediction of satisfaction with life may be influenced culturally. Measurement of Self-E steem and Instruments The types of methods used to study self-est eem is fairly standard throughout the social sciences. Mecca, Smelser, and Vasconcellos (1989) list the following: Epstein, (1979); as well as Jame s, (1890) used introspection, Bednar, Wells, and Peterson, (1989) used case studies, Branden, (1969); as well as Pope, McHale, and Craighead (1988) used surve ys, Rosenberg (1965); employed an experimental design, Coopersmith (1967); Jackson (1984) as well as Mruk (1983) used phenomenological methods. Scales meas uring self-esteem and related cons tructs (e.g., self-concept) suffer from a lack of consensus regarding definitions. However, some recent researchers (e.g., Harter, 1990; Rosenb erg, Schooler, Schoenbach, & Rosenberg, 1995; Willoughby, King & Polatajko, 1995) hav e noted that self-esteem in general

PAGE 81

72 reflects an overall evaluation of ones self, whereas self-concept represents ones selfdescription. A review of the literatur e (e.g. Mruk, 1995; Winters, Myers, & Proud, 2002) suggest in general most scale s address self-esteem as a global construct. Mruk, (1995) as well as Winters et al., (2002) r eported that three of the best known and most widely used instruments for assessing self-esteem include the Coopersmith (1967) adult version of the Self-Esteem Inventor y, (SEI); PiersHarris Childrens Self-Concept Scale (PHC SCS), (Piers, 1984; Piers & Harris, 1969); and the Rosenberg (1965) Se lf-Esteem Scale (RSES). The Self-Esteem Inventory (SEI). Coopersmith (1967) developed this 25 item paper and pencil forced choice self -report questionnaire to measure selfesteem. The respondents are presented with straightforward questions and asked to choose between either like me or unlike me Scores are interpreted in terms of ranges such as low, medium, and high self -esteem. The SEI provides six scores: total self-esteem, lie scale, school-academic life, social-peers, home parents, and general self. Overall, the instrument (SEI) seems to possess adequate psychometric properties. Franklin, Duley, Rousseau, & Sabers (1981) report in their study with undergraduates an internal reliabilit y (split-half) within a range of r = .75 to .92, and a seven day test-retest coefficient between a range of r = .72 and .84. More recent research such as Winters et al., (2002) reported similar findings, with a community sample (internal reliability ranged between r = .75 and r =.95, with a seven day test-retest coefficient of r = .88) suggesting a dequate reliability within the context of the above cited studies.

PAGE 82

73 Other res earchers have studied the relati on between the SEI and other measures of self-esteem (e.g., Gri ffiths, Beumont, & Giannakopoulos 1999; Johnson, Redfield, Miller, & Simpson 1983; Wood, Hillman, & Sawilowsky, 1996) reported findings in support of concurr ent validity for t he SEI. Additional researchers (e.g., Fendrich, Weissman, & Warner 1990; Marciano & Kazdin 1994; Miller, Warner, Wickramaratne, & Wei ssman 1999; Mullis & Mullis, 1997; Vila, Robert, & Nollet-Clemencon 19 95) investigated the relati onship between the SEI and depression, suicidality, hopel essness, locus of control, and social competence and reported results in support of convergent validity for the SEI. Overall, wit hin the context of the above ment ioned studies the SEI appears to be a relatively reliable and valid measur e of self-esteem (c onsistent with the Coopersmith model). Mruk ( 1995) noted that there is an independent body of research using the SEI that supports its cred ibility. However, t he instrument also suffers from the following serious weak nesses: (a) the instrument does not provide a way to estimate how much re spondents distort their responses in a desired direction, (b) the ceiling effect is strong, and the instrument is relatively transparent. (c) the instrument does not indicate whether global and/or situational self-esteem is being assessed. The Piers-Harris Childrens Se lf-Concept Scale (PHCSCS). measures global self-concept and six component domains: (a ) behavior, (b) intellectual and school status, (c) physical appearance and attributes, (d) anxiety, (e) happiness / satisfaction, and (f) popularity. However, some research (e.g., Pl atten & Williams,

PAGE 83

74 1979) report findings from fa ctor analysis that do not support the sub-scales of this model. Winters et al. (2002) report fi ndings from their review of the recent literature suggesting internal consistency within a range of r = .73 to .81, with a seven day test-retest coefficient between r = .42 and .96.The authors further report a coefficient between the PHCSCS and other measures of self-esteem such as the SEI within a range of r = .42 to .85, and a coefficient between the PHCSCS and other related c onstructs within a range of r = .67 to .75. Thus the instrument appears to have adeq uate concurrent and conv ergent validity. Other researchers (e.g., Franklin, et al, 1981; Pier s, 1984) report findings that internal consistency of the PHCSCS tota l score and sub-scores range between r =.60 and r =.75. In general, within the context of the above mentioned studies the PHCSCS appears to be a relatively reliable and valid measure of self-est eem predominantly validated with community samples under 18 years of age. However, some researchers (e.g., Austin & Huberty 1993; Mannarino, Cohen, & Berman 1994) question the PHCSCS construct validity by noting the strong correlations between the instrument and anxiety, depression, in telligence, and other health measures. Thus, the mixed results reported in the research suggest that just what the instrument measures, does not appear to be clear at this time. The Rosenb erg Self-Esteem Scale (RSES). This instrument consists of ten questions rated on a Likert-type scale wit h 1 representing str ongly agree and 4 representing strongly disagr ee. The tone of the questions are varied to avoid

PAGE 84

75 the confounding influence of response set among participants (e.g., You feel you do not have much to be proud of, and You take a positive attitude toward yourself). Certain questions are then scored in re verse in order to maintain consistent answer values. Kaplan & Pokorny (1969) noted that although the RSES was originally developed for use with high sc hool students, it has become a popular measure of self-esteem with adult samples. The RSES has also been popular among res earchers investigating the stability of self-esteem over time (e.g ., Bachman & OMalley, 1977; Chubb, Fertman, & Ross, 1997; Wigfield, Eccl es, Iver, Reuman, & Midgley, 1991; Zimmerman, Copeland, Shope, & Dielman, 1997). Other researchers, such as Wylie (1989) as well as Lynch (1999) repor ted similar psychometric findings (e.g., internal consistency of r = .77 to .87, test-retest r = .85 to .88, and convergent validity r = .58 to .83). Lewinsohn, Seeley, an d Gotlib, (1997) reported from their study of both clinical and non-clinical adolescents (N = 1,219) that the RSES identified depressed adolescents when other instruments did not. Overall, within the context of the above cited studies the RSES appears to be a relatively reliable and valid measure of self-esteem among adolescents and young adults. The RSES wa s chosen to assess self-esteem in the present study for the following reasons. The RSES, PHCSCS, and SEI may all appear to be relatively reliable and valid measures of self-esteem. However, only the RSES was developed specifically as a global measure of self-esteem. Thus, with the RSES one avoids the questionable task of summing across sub-scales to derive a total score.

PAGE 85

76 When competing instruments fall within an acceptable range of construct and psychometric properties, Winters, et al., ( 2002) noted that Selection of the most appropriate scale then depends upon aspects of the sample and the application (p. 1177). In a review of more than a doz en scales, including the RSES, PHCSCS, and SEI, Cross, McDonald, and Lyons (1997 ) argued that the RSES offers the more powerful multiple response format. It al so has a rich data base, as it is the most frequently cited self-esteem scale. Some relative disadvantages of the PHCSCS is its time for administration, and it is less s ensitive because it utilizes two point scoring. The PHCSCS estimated time for administration is 30 minutes, the SEI requires about 20 minutes, while the RSES requires only about 10 minutes. The SEI is similar to the RSES in length and simplicity of scoring; however, the SEI does not have the extensive data base, especially in regards to college students possessed by the RSES. Theoretical Development of Depression Depression has been recorded since antiquity, and descriptions of what we now refer to as depression can be found in several ancient documents. Kaplan and Sadock (1985) proposed that depression is a broad term with multiple meanings. Depression can denote a variet y of phenomena: a sign, a symptom, a syndrome, an emotional state, a reaction, a disease, or a clinical entity. Webster (2001) defined depression as (a) hollow or low place, (b) low spirits; dejection, (c) a decrease in force, activity, etc., and (d ) a period of reduced business, etc. A review of Rogets International Thesaurus (1992) revealed the following synonyms

PAGE 86

77 for depression: (a) downcast, (b) dejection, (c) melancholia, (d) the blues, (e) in the doldrums, (f) down hearted, (g) moping, and moonstruck madness. A review of the depression literature reveal ed the following definitions: Taylor (1996) defined depression as a morb id sadness, dejection, or melancholy. Keltner, Schwecke, and Bostrom (1995) defined depression as a lowered or saddened mood state or major affective disor der listed as a mood disorder in the American Psychiatric Association Diagnosti c and Statistical Manual of Mental Disorders, Fourth Edition. (DSM-IV). To meet t he DSM-IV criteria for a major depr essive episode requires the presence of at least five of the followin g symptoms within a period of two weeks, and a significant change from a persons prev ious level of functioning. One of the five symptoms must be symptom number 1 or number 2 from the following list: (1) depressed mood most of the day, (2) mark edly diminished interest in all, or almost all, activities, (3) significant weight gain or lo ss when not dieting, greater than 5% per month, (4) insomnia or hypersomnia nearly everyday, (5) psychomotor agitation or retardation nearly everyday (6) fatigue or loss of energy nearly everyday, (7) feelings of worthlessness or excessive/ inappropriate guilt nearly everyday, (8) diminished ability to think, indecisiveness nearly everyday and (9) recurrent thoughts of death or suicide. Additionally the depression cannot be due to a substance condition or general medical condition. It cannot occur wit hin two months of the loss of a loved one. Major depressive disorders are further classified as mild, moderate, or sever.

PAGE 87

78 Keltner, et al. (1995) as well as Valente ( 1994) noted that the common element in all of the definitions present ed, regardless of resource, is a significant change in mood. Haber, Krainovich-Miller, Leach, and Pr ice-Hoskins (1997) defined mood as a sustained, internal, emotional state asso ciated with characteristic emotions and feelings that are refl ected in personality. In an e ffort to be consistent with current th inking, to reduce confusion, and to foster interdisciplinary ex change, the current trend in conceptualizing depression is to use the DSM-IV diagnostic criteria. In the present study, I will follow this trend by using the DSM-IV criteria to define depression. Thus, depression in the present study is defined as the persistence of alte red mood, whether m ild, moderate, or severe, for a time period of two weeks or more. Relationship of Depression to the Present Study Researchers have consistently reported an inverse (negative) relationship between life satisfaction and depression in a variety of samples: with clinical subjects (e.g., Hyer, Gouveia, Harrison, & Warsaw, 1987), non-clinical subjects (e.g., Parkerson et al., 1990), men (e.g ., Kammann & Flett, 1983), women (e.g., Raphael, 1988) and the physically disabled (e.g., Evans, Kleinman, Halar, & Herzer, 1984). Levisohn et al (1991) reported findings that s uggest low life satisfaction tends to precede the onset of depression. A more recent investigation, MartinezPons (1997) utilizing a non-clinical convenience sample (N = 108) and path analysis suggested a negative relationship between depression and life satis

PAGE 88

79 faction. Based on previous research that suggested a negative relationship between depression and life sati sfaction, I expected to find a similar relationship between depression and life sati sfaction in the present study. Also, I expected depression to demonstrate a small effect si ze (account for about 3% or less of the variance in life satisfaction) afte r accounting for self-esteem. Measurement of Depression and Instruments There are many measurement tools available to assess depression. However, a review of the professional li terature revealed the following three instruments are the most frequently used and thus best known instruments in both clinical practice as well as research: (a) Zung Self-Rating Depression Scale (SDS), (b) Center for Epidemiologic Stud ies Depression Scale (CES-D), (c), Beck Depression Inventory (BDI). Zung Self-Rating Depression Scale (SDS). This scale was developed by Zung (1965) it is a 20-item self-rating sca le. The items consist of statements, (e.g., A good part of the time I have crying spells or feel like it and I always feel down-hearted or blue). Subjects are asked to express their degree of agreement with each item on a Likert-type scale with 1 representing completely disagree and 7 representing com pletely agree. Dugan, McD onald, Passik, Rosenfeld, Theobald, and Edgerton (1998) as well as Lane, Shellenber ger, Gresen, and Moore (2000) reported estimates of internal consistency ranging from r = .78 to .92. Tanaka and Huba (1987) noted that a limitation with this instrument is a lack of va lidation among college students

PAGE 89

80 that appears to differ from adult samples. A nother limitation of the SDS according to Sue (1999) is that the instrument has not been validated with samples representative of people of color. Center for Epidemiologic Studi es Depression Scale (CES-D). The CES-D was developed by researchers at the Cent er for Epidemiologic Studies at the National Institute of Mental Health. The CES-D scale consists of 20-items, and may be either self or interviewer adminis tered. The time frame for reporting symptoms is the past week. An example of an item is I have thoughts about hurting myself. Subjects must choose fr om a Likert-type scale beginning with rarely or none of the time (scored 0), some or a little of the time (scored 1), occasionally or a moderate amount of time (scored 2), and most or all of the time (scored 3). The internal psychometrics (internal and test-ret est reliability) of the CES-D scale appear adequate. Several researchers utilizing adult clinical samples (e.g., Craig & Van Natta, 1983; Weissman, S holomska, Pottenger, Prusoff, & Locke, 1977) as well as researchers utilizing non-cl inical samples (e.g., Radloff, 1977; Roberts 1983; Lewinsohn & Teri, 1982) report ed internal consistency reliability within a range of r = .8 to .9, with test-retest reliabilities ranging from r = .5 to .6 over a period ranging from several days to several weeks. Beck Depression Inventory (BDI). The BDI was developed by Beck, Ward, Mendelson, Mock, and Erbaugh (1961). This inst rument is a 21 item self-report depression scale. The items are scored on a 0 to 3 scale. Zero represents not at

PAGE 90

81 all and 3 represents intense. The total BDI score represents the sum of the individual items; scores can range from 0 to 63. The BDI is a widely used measure with a substantial research base. Some researchers (e.g., Beck & Steer, 1984; Beck, Steer, & Garbin, 1988) report adequate internal consistency and test retest reliabilities with the BDI (the r values for internal consistency ranging from r =.72 to .85 and test re-test estimates from r =.65 to .82). Research with adolescents for example, Kaplan, Hong, and Weinhold (1984) as well as research with college students (e.g. Bumberry, Oliver, & McClure 1978) reported internal consistency and test re-test reliability estimates range between r =.80 and .90. However, the original BDI has been revised. T he new instrument the BDI-II was developed by Beck, Steer, and Brow n (1993). The BDI-II is designed to assess depression in persons over 13 years of age. Like the BDI, the BDI-II has a 21-item format, with a choice of four possible answers for each item ranging in value from zero to three. For example, it em 5 asks about guilty feelings: 0 = I dont feel particularly guilty, 1 = I feel guilty over many things I have done or should have done, 2 = I feel guilty most of the time, or 3 = I feel guilty all of the time. Although, a number of changes have been made to successive versions of the original BDI, the general structure of the instrument has not changed. The most significant changes found in the BDI-II are intended to make item content more consistent with the major depressive episode concept as defined in the Diagnostic and Statistical Manual of Mental Disorders, F ourth Edition (DSM-IV). The BDI-II was chosen to assess depression in the present study for the following

PAGE 91

82 reasons; first, because in comparison wit h SDS and CES-D the BDI and BDI-II has the richest research base, including re search with college student samples. Second, the BDI-II appears to be the measur e most consistent with the DSM-IV definition of depression. Theoretical Development of Locus of Control Shapiro, Schwartz, and Astin (1996) stated that Individual s beliefs about the controllability of what happens to them is a core element of t heir understanding of how they live in the world (p. 1217). Rotte r, Seeman, and Liverant (1962) reported from their early investigati ons of these beliefs that some individuals change their beliefs more than others a fter new experiences. T he proposed psychological construct to account for this difference is locus of control which evolved out of social learning theory. Rotte r (1966) defined locus of control as A persons perception of the degree of control he/she has over events that occur in the world (p. 1). Lefcourt (1982) as well as Rotter ( 1990) noted that becaus e of the significance of locus of control in determining beh avior, research of this construct has proliferated in a variet y of areas (e.g., educati on, psychotherapy, management). Rotter (1966) asserted that the importance of reinforcement is universally recognized in the acquisition of skills and knowledge. However, how individuals perceive reward and punishment determines t heir future behavior. Thus, the effect of reinforcement is not simply a mechanical process, but depends upon whether or not the person perceives a causal re lationship between his own behavior and the reward (p. 1).

PAGE 92

83 Rotter (1966) made two additional observatio ns regarding the importance of locus of control to personality. First, depending upon the individ uals history of reinforcement, individuals would differ in the degree to which they attributed reinforcements to their own actions (p. 2) Thus, ones locus of control is both shaped by ones experiences in the world, and ones locus of control shapes ones experiences in the world. Second, Expectancies generalize from specific situations to a series of situations whic h are perceived as similar. Consequently, a generalized expectancy for a class of related events has functional properties and makes up one of the important classes of variables in personality description (p. 2). Rotter (1954) wa s the first to use the term internal locus of control in his social learning theory to describe persons who believe that their own behaviors determine the positive reinforcements they receive. In general, persons who perceive themselves as the cause of their positive reinforcements tend to feel they are in control of their lives and thus take greater re sponsibility for their lives. Some researchers (e.g., Demellow & Imms 1999; Peterson, Maier, & Seligman, 1993; Rothbaum, Weisz, & Snyder, 1982) r eported findings that suggest people with internal locus of control typically eng age in proactive and adaptive behaviors. On the other hand, people who perceive themselves as controlled by external forces (have an external locus of control) tend to feel detached from the positive as well as the negative reinforcements in th eir lives. In agreement with this line of research Gomez (1997; 1998) reported fi ndings suggesting individuals with an

PAGE 93

84 external locus of control tends to be reactive and avoid stressful situations. Thus, the research suggested that people with an external locus of control tend to take less responsibility for their lives. Rotter (1966) a sserted that locus of control wa s originally formulated as a generalized expectancy of reinforcements; where individuals believe that what happens to them is a result of their control or the result of forces beyond their control such as chance, fate, or powerful others. Therefore, locus of control is best conceptualized along a dynamic conti nuum with a range that spans external to internal perceived control. Weiten (1989) stated that Althou gh people are often classified as internals or externals, the concept shoul d not be perceived dichotomously. Rather, it should be viewed as a continuum ranging from highly internal to highly external (p. 39). Rotter (1966) argued that even though locus of control was conceptualized along a continuum, it was a fairly stable psychological construct. Several researchers (e.g., Figurelli, Hartman, & Kawalski 1994; Gaa, 1979; Kim, Omizo, & DAndrea 1998; St. Lawrence, Jefferson, Alleyne & Brasfield,1995; Trice, 1990) reported findings that support Rotters arguments that locus of control is best conceptualized along a continuum and it is also a relatively stable psychological construct. Rotter (1975) warned against falsel y assuming that characteristics of persons with an internal locus of control are all pos itive and the characteristics of persons with an external locus of control are a ll negative. However, some researchers (e.g., Evans, Shapiro, & Lewis, 1993; Furby, 1979) reported that in both locus of

PAGE 94

85 control research as well as in practice, t here is bias from the popular assumption that an internal locus of control is more des irable than an external locus of control. Driven by t he assumption that internality is more desirable than externality, some researchers (e.g., Duke & Nowicki 1974; Young & Shorr, 1986) reported finding a positive relationship between inte rnal locus of control and achievement among male college students. Other res earchers (e.g., R enn & Vandenberg, 1991) reported findings that s uggest employees with an internal locus of control were rated higher than thos e with an external locus of control on important job variables. Koeske and Kirk (1995) r eported that even among mental health professionals, those with a gr eater sense of internal cont rol beliefs report higher satisfaction with their jobs, life and expecte d more favorable outcomes for their clients. Bandura (1989) demons trated a positive relationship between internal locus of control and success in mental healt h therapy. Blumenthal, Matthews, and Weiss, (1994) demonstrated a positive relationship between internal locus of control and physical health. Alfonso, Allis on, and Rader (1996) reported a positive relationship between locus of control and life satisfaction. Relationship of Locus of C ontrol to the Present Study Hong and Giannakopoulos (1994) noted that it has been consistently reported that internal locus of control is positively related to life satisfaction (e.g., Hickson, Housley, & Boyle, 1988; Klein, Tatone, & Lindsay, 1989; Lewinsohn, et al., 1991; Morganti, Nehrke, Hulicka, & Cataldo, 1988; Raphael, 1988; Schulz, Tompkins, Wood, & Decker, 1987). These re searchers reported a range of results

PAGE 95

86 suggesting locus of control accounts for bet ween 4.6% to 23% of the variance in satisfaction with life. Shapiro, et al. (1996) in reviewing the lite rature on locus of control and satisfaction with life concluded that research findings strongly support the importance of an internal locus of contro l in enhancing ones satisfaction with life. Hong and Giannakopoulos (1994) reported that internal locus of control remains an important predictor of sati sfaction with life after accounting for both self-esteem and depression. The present study investigat ed whether emotional intelligence or one or more components of emotional intelligence predicts or accounts for additional variance in life satisfaction great er than self-esteem, locus of control, and depression. Locus of control has been included as one of the independent variables in the present study because prev ious research strongly suggests it is related to satisfaction with life. Measurement of Locus of Control and Instruments Ma rks (1998) stated that Western cultur e has always placed a high value on personal autonomy, and this value has in fluenced the theoretical development and measurement of the locus of control c oncept (p. 251). Fink and Hjelle (1973) as well as Mirels and Garrett (1971) and Lef court (1982) argued that internal locus of control is related to the Protestant ethic and traditional Amer ican values. Therefore, the theoretical developm ent, as well as the measurem ent of locus of control has been influenced from its beginning by Western cultures emphasis on taking personal control in all situations.

PAGE 96

87 Phares (1957) was one of the first to measure individual differences in a generalized expectancy or belief in external control as a psychological construct. The instrument developed by Phares was a 13 item, two point, Likert-type scale. This scale was a crude attempt to measure locus of control utilizing a two point, forced response format. However, the effect of the social desirability response set severely limits the useful ness of this instrument. The next atte mpt to develop an assessment scale for locus of control was in an unpublished dissertation by James (1957). James revised Phares instrument retaining the Likert format, which is now known as the James-Phares Scale. However, Liverant, Rotter, and Seeman revised the James-Phares Scale, developing subscales and using factor analysis reducing the number of items from 100 to 60. The final revisions were made by Rotter, Liverant and Crowne (1961) by changing the wording of some items (making them appropriate for non-college subjects) and eliminating those it ems with high correlations with the Marlowe-Crowne Social Desirability Scale. The final version of the scale is known as the Rotter (1966) Internal-External Locus of Control Scale, or simply the I-E Scale. Internal-External Locus of Control Scale (I-E Scale). The I-E Scale is a 29 item, forced-choice test including 6 filler items intended to make the true purpose of the test somewhat more ambiguous. Ea ch of the 29 items has an a and b part; respondents are asked to choose which one of the pair most accurately reflects their view. Examples of it ems from the I-E Scale includ e: (a) Children get into trouble because their parents punish them too much and (b) The trouble with

PAGE 97

88 most children these days is that their par ents are too easy on them. The I-E Scale appears to have good face validity; a carefu l examination of eac h of the items reveals the items deal exclusively with t he subjects beliefs about how reinforcements are controlled. Cher lin and Bourque (1974) reporte d alpha coefficients of r = .80 for college students and r = .71 for a general popul ation sample. Franklin (1963) reported an alpha of r = .69 (Kuder-Richardson) with a nationally stratified sample (N = 1,000). Other researchers (e.g., Rotter, 1982; Gilman & Huebner, 2000) reported a relatively stable in ternal consistency ranging from r = .65 to .76, and test-retest reliabilities ranging from r = .83 over a 30 day period to r = .49 over a 60 day period. Overall, within the contex t of the above cited studies the final version of the I-E Scale (Rotter, 1966) appea rs to be a relatively reliable and valid measure of locus of control according to Rotters 1962, one dimensional model. Page and Scalor a (2004, p. 527) repor ted that Generally, locus of control scales include several forced choice questions focusing on an individuals beliefs about internal versus external influences in a variety of settings (e.g., Rotters I-E Scale, 1966; and Nowicki & Strickland, 1973). A review of the literature revealed that Rotters I-E Scale and the NS-LOC ar e two instruments often utilized to assess locus of control in the social sciences. Since I have already discussed the I-E Scale, a review of the NS-LOC is in order. Nowicki-Strickland Locus of Control Scale (NS-LOC). Nowicki and Strickland (2000) developed the NS-LOC to a ssess locus of control. The NS-LOC is grounded in Rotters social learning t heory, which conceptualizes locus of

PAGE 98

89 control of reinforcement as an important personality construct. The full scale contains 40 statements concerning perc eptions to which respondents answer yes or no. An example of an item from the NS-LOC is Do y ou feel you have a lot of choice in deciding who your friend s are? The NS-LOC appears to have good construct validity as the items clearly ta rget ones perception of control over consequences. However, the scale is some what transparent and thus may suffer from the social desirability bias. Madsen and Goins (2002) reported findings from their study utilizing a sample of college st udents (N = 120) suggesting relatively good psycho-metric properties for the NS-LOC (split-half reliability ranged from r = .75 to .86; test-retest reliabi lity over a 30 day period was r = .82). Nowicki and Strickland (1973) the developers of the scale reported an internal consistency (the split-half method) of r = .63. Overall, withi n the context of the above cited studies the NS-LOC like the I-E Scale appears to possess relatively good psychometric properties (reliability and validity). Howeve r, while both the NS-LOC and I-E Scale appear to be satisfactory measures of locus of control, the I-E Scale has a much richer data base than the NS-LOC. The instrument chosen to assess locus of control in the present study was the I-E Scale.

PAGE 99

90 Chapter Three Methodology Introduction to Methodology This study evolved from recent research (e.g., Cobb & Mayer, 2000) suggesting that Some educat ors have implemented emotional intelligence programs and policies without much empirica l justification (p. 16). The current study investigated the utility (usefulness) of emotional intelligenc e in the prediction of life satisfaction among community co llege students. Emotional intelligence was conceptualized from the Mayer and Sa lovey (1997) cognitive ability model. The instrum ent chosen to measure emotional intelligence was the Mayer, Salovey, and Caruso Emotional Intelligenc e Test (MSCEIT). The other variables included in the study are self-esteem, depression, and locus of control, have consistently been reported in previous re search (e.g., Hong & Giannakopoulos, 1994) to predict satisfaction with life. I attempted to help establish (or not) the utility of emotional inte lligence by investigating its relationship or lack of a relationship with satisfaction with life among community college students after accounting for variance explained by self-esteem, depression, and locus of control. Restatement of the Research Questions 1. Does emotional intelligence concept ualized as a cognitive ability and measured by the Mayer, Salovey and Caruso Emotional Intelligence Test

PAGE 100

91 (MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 2. Does the ability to perceive and accurately express emotion (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? 3. Does the ability to use emot ion to facilitate thought (a component of emotional intelligence as m easured by the MSCEIT) acc ount for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 4. Does the abilit y to understand emotion (a com ponent of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 5. Does the ability to manage em otion for emotional growth (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? Population Size/Characteristics. Central Fl orida Community College (CFCC) enrolled 28,518 students for credit courses during the 2003-2004 academi c year (Spring -10,378; Summer 7,587; and Fall -10,553). Approx imately sixty-five (65) percent of these students

PAGE 101

92 during the 2003 2004 academic year we re females, and approximately 35 percent were males. Approximately 78 per cent of these student s were White, Non-Hispanic, followed by approximately 13 percent Black, Non-Hispanic. Hispanic students comprised the next largest group with approximately 6 percent, followed by Asians, Native Americans, and others; each comprising approximately 1 percent of the student population. All but 59 students were from the state of Florida and all but 266 students were from Marion, Citrus, and Levy Counties. Average age was 26; however, 43 percent were under 22 years of age and 55 percent were 24 years of age and under. A review of the demographic records for CFCC covering the two prev ious academic years (2001-2002; 20022003) suggested little change from one year to the next in the to tal number or the characteristics (gender, race age) of CFCC students. The study was conduct ed during the 2005 Fall semester. The CFCC student population was approx imately 9,345 students enrolled in one or more credit courses. The present study is anonym ous research. Thus, I did not collect any participant information t hat could personally identify participants. However, I did ask participants to indicate their gender age, and race on two of the instruments (MSCEIT and BDI-II) not included in the appendixes because they are propriety instruments. In or der to evaluate how well t he sample characteristics reflect the population characteristics (sam ple representativeness) I compared the obtained sample characteristics to the Fa ll 2005 population characteristics. First, the percentage of females in the present study (67.5%) is similar to the per

PAGE 102

93 centage in the population (65.5%). Second, the percentage of Whites (89%) in the sample is similar to the percentage of Whites in the population (86.5%). Thus, the sample may somewhat over repr esent Whites in the population. Blacks comprised a smaller percentage of the sample (5%) than in the population (7.5%). The percentage of Hispanics in the sample (4.0%) is similar to the percentage in the population (4%). The perc entage of Asians, Native Americans, and others (1%; 1%; 0% respectively) in t he sample is similar to the percentage found in the population (1%; 1%; 1% respectively). Third, the mean age of students in the sample was 23.5 similar yet somewhat younger than the CFCC population mean 25.3. Overall, within the limits of the above discussion (gender, ethnicity, and age) the obtained sample of 200 participants appears to be representative of the CFCC student population. Selection Eligibility Characteristics All participant s in the study were enrolled in at least one three credit hour course of study at CFCC (Citrus cam pus) during the Fall 2005 semester. In addition, all participants were enrolled in a c ourse section selected to take part in the study. Additionally all par ticipants in the study volu nteered to participate. Sampling Scheme/Size/Characteristics The sampli ng scheme utilized in the pres ent study was convenience sampling. Although, a random sample of all CFCC students would potentially increase external validity by allowing fo r greater generalizability, limited resources and logistical constraints precluded the use of a random sample. Limited

PAGE 103

94 resources and logistics also restricted the si ze of the sample to 200 participants. I began by soliciting students to participate in the study from cla sses taught at the CFCC Citrus campus, where I am employed as an instructor. The Citrus campus is located in Lecanto about 18 miles from the main campus in Ocala. CFCC serves students from Marion, Citrus, and Levy counties. However, over 89 percent of all CFCC students attend classes at the Mari on (Ocala) and/or Citrus (Lecanto) campuses. Although, CFCC does not publish student demographic data by campus I expected there would be little diffe rence between students gender, age, and ethnicity attending the Ocala or Citrus campuses. Many students attend classes on both campuses and many faculties teach at both facilities. I received permission from CFCC office of Institutional Effectiveness to conduct my study on both the Citrus County campus and Ocala campus. My initial plan was to solicit as many participants as possible from the Citrus campus and then solicit the remaining participants from the Ocala camp us. However, I was able to solicit a sufficient number of participants (N = 200) for my study from the Citrus County campus. I began by so liciting the aid of fellow instru ctional faculty for permission to seek volunteer participants from among their students. None of the instructors I made personal contact with declined my request. The test publishers report an estimate of time needed for completing eac h instrument. However, I suspected the actual total time needed to complete all five instruments would be greater for

PAGE 104

95 most students. Thus, in order to determi ne the actual time needed for completion of all five instruments I adm inistered the five instruments to myself, I completed all five in 62 minutes. I then administered all five instruments to two community college students and one high school student. The students completed all five instruments in 65, 55, and 51 minutes respectively. Thus, knowing that most students will need about 60-75 minutes to complete all five instruments I located physical space (classrooms) where students could complete all five instruments without interruption. Students were advis ed they needed about 60-75 minutes to complete all five instruments. All st udents were monitored by me during the completion of the instruments and all inst ruments were inspected for completeness and compliance with instructions. This method of participant selection and data collection continued until the target number (N = 200) of participants as well as completed assessment packets were obtained. During the fi rst week of data collection I so licited participants from three sections of humanities and one section of general psychology; while two students declined to participate in the study; 84 students completed all five assessment instruments. The second week of data collecti on I solicited participants from two sections of introduction to soci al science and two sections of college skills. All students solicited agreed to take part in the study except for four students who had already taken part in the study in other classes. However, 76 students completed all five assessment instruments. During the last week of data collection I solicited participants from one section of general psychology,

PAGE 105

96 two sections of freshmen English skills an d one section of college success skills; although, none of the students solicited declined to take part in the present study more than a dozen had already participated as part of another class. However, forty students did complete all five asse ssment instruments bringing the total number of students taking par t in the study to the target number of 200. The sample si ze in the study is largely the result of limited resources and logistical constraints. The cost of the research instrumentation limited the study to a sample size of 200 participants. A re view of the life sati sfaction literature revealed a number of studies (e.g., Lewinso hn et al., 1991; Schmitt & Bedeian, 1982; Sekaran, 1986) report re sults that suggest self-esteem is one of the most frequently cited predictors of life satisfacti on. For example, a study conducted by Hong and Giannakopoulos (1994) using a la rge sample of 1,749 adults (17-40 years of age) investigated the relationship between life satisfaction and seven other variables: a) psychological reactance, b) self-esteem, c) religiosity, d) trait anger, e) locus of control, f) depression, and g) age. The results of this study suggest that self-esteem, depression, and locu s of control are three of the best predictors of life satisfaction, respecti vely. The above researchers report selfesteem accounts for 21. 4% of th e variance in life satisfaction, ( r = .46, p < .001). This study also revealed an inverse relationship ( r = -.31) between depression and life satisfaction. Depression accounted fo r an additional 2.8% of the variance in life satisfaction ( R = .03).The third strongest predict or of life satisfaction was locus of control (r = .23) whic h accounted for an additional 1% ( R = .01) of

PAGE 106

97 variance in satisfaction with life. The ot her variables included in the Hong and Giannakopoulos (1994) study, trait anger, re ligiosity, psychological reactance, and age contributed less than 1% toward the pr ediction of life satisfaction. Thus, trait anger, religiosity, psyc hological reactance, and age do not significantly contribute in the prediction of satisfac tion with life above the variance accounted for by self-esteem, depression, and locus of control. Accord ing to the Hong and Giannakop oulos (1994) study, the R for selfesteem (first variable entered) in the predi ction of life satisfaction was .214. This R when converted to an f (effect size) equals .27. According to Cohens (1988) scale .27 falls about mid-way betw een a medium (.15) and large (.35) effect size for multiple regression analysis in the social sciences. According to Cohens (1988) sample size chart, studies involving multiple regression analysis with four independent variables, a predetermi ned statistical significance of alpha = .05, and an estimated effect size between medium and large, would need a minimum of between 45 to 97 subjects for a power of .80 (80% chance of rejecting a false null hypothesis). Thus, if the relationship between emotional intelligence and life satisfaction is between medium and large, a sample size of 200 should give me a good chance (equal or great er than .80) of rejecting a false null hypothesis. Re turning to the literature, the Hong and Giannakopoulos (1994) study reported the R for depression and locus of cont rol combined after accounting for self-esteem was .038. When this R value is converted to an f (effect size)

PAGE 107

98 value the result is .04 which according to Cohens (1988) scale of effect sizes is a small effect size. According to Cohens ( 1988) sample size chart a study using multiple regression analysis with four independent variables, a predetermined Alpha of.05 and an estimated effect size between small and medium, the minimum sample size needed for a power of .80 would be between 97 and 599 subjects. If the effect size between emotional intelligence a nd satisfaction with life is small then my sample of 200 subjects would not provide a reasonable expectation of rejecting a false null hypothesis. If the effect size of emotional intelligence on satisfaction with life is greater th an the combined effect size of selfesteem, depression, and locus of control on satisfaction with life then my sample of 200 subjects may well be adequate to provide a reasonable expectation of rejecting a false null hypothesis. Some recent research suggests the effect size between emotional intelligence and satisfaction with life is at least in the medium ( r =.15) range (e.g., Bar-On, 1997). In this study the relationship between emotional intelligence and satisfaction with life was reported to be r = .41 with an estimated effect size index of f = .15 which according to Cohen (1988) is a medium effect size. A more recent study by Ciarrochi et al. (2000) investigated the relationship between emotional intelligence and satisfaction with life among undergraduate students (N = 118) after controlling for general intelligence (IQ) and the following personality variables: a) extraversion, b) neuroticism, c) empathy, d) openness to feelings, and e) self-esteem. The importance of this study is that it reported a

PAGE 108

99 correlation of r = .22, p < 0.05 between emotional intelligence and satisfaction with life after controlling for general IQ, as well as the above mentioned five well known personality variables. The Ciarrochi et al. (2000) study reported that emotional intelligence accounts for additional variance in life sati sfaction over the variance accounted for by IQ, self-esteem, or the other four personality variables (extraversion, neuroticism, empathy, and openness to feelings) included in the study. Thus, the Ciarrochi et al. (2000) study suggests if EI alone accounts for greater variance in satisfaction with life then self-esteem, IQ, and four additional personality variables the effect size between emotional in telligence and satisfaction with life may be large. Other resear chers (e.g., Saklofske, Austin, & Minski, 2003) have elected to investigate the relationship between emotional intelligence and life satisfaction among Canadian undergraduate st udents (N = 354) while accounting for the big five personality dimensions (neurotic ism, extraversion, openness, agreeableness, and conscientiousness). These resear chers report the results of regression modeling shows that emotional intelli gence accounts for additional variance in satisfaction with life not accounted for by personality (p. 707). This study suggests when emotional intelligence is the fi rst variable added to the hierarchical regression analyses (when other variables ar e not controlled for) the result in R = .265. When I transform this value into an estimate of effect size the result is f = 36, which according to Cohens ( 1988) scale is a large effect size.

PAGE 109

100 A recent study by Lopes, Bra ckett, Nezlek, Schutz, Sellin, and Salovey (2004) reported that the effect size between emot ional intelligence and satisfaction with life among college students (N = 118) falls within a range of medium to large. These researchers reported that emotional intelligence as measured by the MSCEIT demonstrated incremental validity by accounting for between 7 and 11 percent of additional variance in satisf action with life over the big five personality dimensions (neuroticism, extr aversion, openness to experience, agreeableness, and conscientiousness). Contemporar y researchers (e.g., Law, W ong, & Song, 2004) reported results from their invest igation with undergraduate univers ity students (N = 202) of the relationship between self-report meas ures of emotional intelligence and satisfaction with life, controlling for personality variables among undergraduate students (N = 202), as well as high school students (N = 560). These researchers reported the results of the hierarchical regression analysis for both samples was similar. When emotional intelligence was added to the regression model, the increases in the model multiple correlation squared was significant ( p < .01), although the absolute magni tude was not large ( R = .05 and .06 for samples 1 and 2, respectively). The authors interpre ted the additional 5% of variance in satisfaction with life accounted for by EI to be of reasonable practical significance (p. 488).

PAGE 110

101 In regards to sample size in multiple regr ession, there is no clear concensus as to what constitutes an adequate or ideal number of participants. In general, there are three school s of thought on this subject. First, many researchers conduct a power analysis by a) esti mating the probable effect size between independent and dependent variables, b) cons ider the number of independent variables, and c) consider a predetermined marg in of error or power (usually.80). Informed by this information the researcher determines the minimal sample size necessary for a desired power by consulting the power analysis tables published by Cohen (1988). Second, some researchers suggest a minimum total sample size, for example, Comfrey & Lee (1992) reported that = very poor; 100 = poor; 200 = fair; 300 = good; 500 = very good; 1,000 or more = excellent (p. 217). The third school of thought suggests a particular ratio between subjects and independent variables. For example, Pedhazur (1997, p. 207) as well as Stevens (2002, p. 72) recommend a nominal number of 15 participants per independent variable. Other researchers recommend different ratios such as 20, 30, or 40 participants per independent variable. The study utiliz ed a sample size of 200 participants that I believe to be an adequate sample size for the following reasons. First, the five par ticular research questions all involve the addition of one additional indep endent variable to the stem multiple regression equation (LS = bo + b1 self-esteem + b2 depression + b3 locus of control). Thus, with 4 independent variables, the ratio of participants to variables is 50 to 1, which exceed s most fixed ratio recommendations. Second, a review of the emotional intell igence literature regar ding the relation

PAGE 111

102 ship between emotional intelligence and sati sfaction with life is both limited (relatively little) and mixed (inconsistent). Therefore, estimati ng effect size between emotional intelligence and satisf action with life is difficult. Ethical Nature of Data Collection Prior to dat a collection or administering any assessment instrument I completed the University of South Florida (USF) required training for researchers utilizing human subjects (see Appendix B). The present study involved minimal risk to participants and I did not collect any personal identifiers. Thus, I elected to make application to the University of South Florida Division of Research Compliance to conduct the study as an ex empted study. My application to conduct the present study as an exempt ed study was approved on October 21, 2005 (see Appendix B). In addition, prior to any data collection I obtained permission from the office of Institutional Effectiveness CFCC to conduct the present study at the Citrus as well as the Ocala campus (see Appendix B). All potential participants received a written reques t from me to take part in the study (see Appendix B). The written request explicitly informed students that I was conducting social science research and that their parti cipation is both voluntary and anonymous. Students were informed of what is ex pected of them as participants (completion of five assessm ent instruments) as well as how much time most students take to complete a ll five assessment instruments (60-75 minutes). In summary students were invited to take part in the study if they had no concerns and wished to do so. Student s were given names, phone numbers,

PAGE 112

103 and e-mail addresses of persons they ma y contact in the event they have questions or concerns at a later time r egarding their participation in the present study. Instruments Five instrum ents were used in the study to m easure emotional intelligence, satisfaction with life, self-esteem, depr ession, and locus of control among community college students. Each of the five instrument s used in the study are now discussed in turn. The Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT). I used the MSCEIT to measure EI for three reasons. First, it was developed by the original authors (Mayer & Salovey, 1997) to gether with a later associate David Caruso to measure emotional intelligenc e according to the Mayer and Salovey (1997) revised model of EI. The MSCEIT measures emotional intelligence according to the authors four components (b ranch), cognitive ability model which includes: a) perceiving emotions, b) facilitating thought, c) understanding emotions, and d) managing emotions. The MS CEIT yields a total score, and the above mentioned four component (branch) scores. Thus, use of the MSCEIT can reveal which if any of the four component s of the cognitive ability model of EI accounts for additional variance in life satisfaction. Second, bot h the paper/pencil and on-line versio ns of the MSCEIT contain 141 multiple choice items, the MEIS c ontains 402 items. Thus, the MSCEIT requires about half the time for administration as the Multi-Factor Emotional

PAGE 113

104 Intelligence Scale (MEIS). I used the paper/ pencil version of the MSCEIT in the present study to maintain format consistenc y among the five instruments. All five of the instruments I employ ed in the present study are of the paper/pencil type. Third, t he developers Mayer, Salovey, and Caruso (2002) report the MSCEIT has a full scale reliability of r = .91 (split-half relia bility). Bracket and Mayer (2003) report a test -retest reliability for t he full scale MSCEIT of r = .86 and branch (component) score reliability between r = .74 and .89. This suggests within the context of the above studies a high ly reliable instrument at the branch and total scale levels. Construct valid ity appears to be high as it gives comprehensive coverage of the four component cognitive ability model developed by Mayer and Salovey (1997). Satisfaction With Life Scale. (SWLS) I chose the SWLS developed by Diener, Emmons, Larsen, and Griffin (1985) to assess global satisfaction with life. The instrument measures satisfaction with life as a cognitive-judgmental process using a five-item scale. The SWLS utilizes a seven-point rating scale ranging from strongly disagree to strongly agree. Short term reliabilities with an interval of up to two weeks have been consis tently reported (e.g., Diener, et al., 1985) to be r = .8 or greater. Other researchers (e.g ., Pavot, Diener, Colvin, & Sandvik, 1991) investigated the reliability and validity of the SWLS with select samples (e.g., elderly persons; college students). This stud y reported test-retes t reliabilities for the SWLS to be r = .7 or greater among the elderly sample and r = .6 or greater

PAGE 114

105 among university students with two week intervals. Alfonso and Allison (1992) reported from their study of 106 university undergraduate students a coefficient alpha of r = .89 and a test-retest correlation of r = .83 with two week intervals. The Pavot et al. (1991) study also inve stigated the predictive and convergent validity of the SWLS. Peer repor ts, a memory measure, and clinical ratings were used as external criteria for validation. In this study the SWLS was compared to other related scales (e.g., Ph iladelphia Geriatric Center Morale Scale). The researchers in this stud y report results that suggest the high convergence of self and peer reported measures of sati sfaction with life, and the SWLS suggest that satisfaction with life is a relatively global and somewhat stable phenomenon. The Rosenberg Self-E steem Scale (RSES). The RSES was chosen by me to measure self-esteem. The RSES (Rosenberg, 1965) provides a global measure of self-esteem. As measured by this scale, high self-esteem indicates cognitive evaluations of self-worth and self-respect. Low self-esteem implies dissatisfaction with oneself and self-reject ion. A review of the self-esteem literature (e.g., Rosenber g, 1965; Crandall, 1973; Goldsm ith, 1986; Blascovich & Tomaka, 1991) revealed the Rosenberg Self -Esteem Scale is one of the most widely utilized measures in social science research and considerable empirical data support its validity. The RSES is a 10 item Likert inventory employing a scale of strongly agree to strongly disagree as response options. Half the items are positively worded and half are negativel y worded, to control for responder

PAGE 115

106 bias. Two examples of items from the RSES ar e on the whole, I am satisfied with my self, and at time s I think I am no good at a ll. Several researchers (e.g., Silbert & Tippett, 1965; Crandall, 1973; McCarthy & H oge, 1982) report findings supporting the RSES one dim ensionality among college students. Multiple st udies have reported results that suggest validity and reliability estimates within the context of particula r studies for the RSES. For example, Silbert and Tippett (1965) report a 2-week te st-retest coefficient of reliability r = .85 (N = 28). Other resear chers such as McCarthy a nd Hoge (1982) report a one year test-retest coefficient r = .77 (N = 1,852). Crandall (1973) investigated the reliability of the RSES and convergent validity between related scales (e.g., Global Self-Worth Scale) and the RSES. Th is research reported a test-retest reliability of r =.76 which suggests overa ll reliability of the scores obtained. The Internal-External Locus of Control Scale (I-E Scale). The I-E Scale was developed by J.B. Rotter (1966). The I-E Scale was used in the study to measure internal vs. external locus of control. This instrument was chosen because it was developed by Rotter ( 1966) who first conceptualized the distinction between internal vs. external locus of control derived from his comprehensive social learning theory. Marsh and Richards (1986) noted that Rotters locus of control instrument has an extensive history and still remains in wide use within the social sciences.

PAGE 116

107 The I-E Scal e measures locus of control as a generalized expectancy of the extent to which a person perceives that events in ones life are consequences of ones behavior. The instrument is a paper and pencil 29 item, forced choice scale. The developer of the I-E Scale (Ro tter, 1966) reported reliability estimates which ranged from r = .69 to .73 using the Split -half Spearman-Brown and KuderRichardson formulas. Other more recent research (e.g., Baumeister, 1991) has investigated the reliability of the I-E Scale with undergraduate students (N = 125) and reported a test-retest reliability of r = 69 with a two week interval. In a review of the locus of control literature, Cher lin and Bourque (1974) not ed that much of the locus of control scale research has em ployed a very specific population (e.g., under 30 years of age). Blau (1984) in vestigated the construct validity of the I-E Scale (N = 267) with undergraduate business students. This st udy compared the I-E Scale with the Levenson Measure of Locus of Control, anot her well known measure of locus of control. The authors report ed a strong positive relationship r = .71 between the I-E Scale and the Levenson measure of locus of control. Thus, this study suggested some evidence supporting both conv ergent and construct validity for the internal-external locus of control construct. The Beck Depression Inventory-II (BDI-II). The BDI-II was chosen to measure depression in the present study. The BDI-II (Beck, Steer, & Brown, 1997) is a revised version of the original Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961).The BDI-II contains 21 items, each of which

PAGE 117

108 assesses a different symptom or attitude by asking the examinee to consider a group of graded statements weighted from 0 to 3 based on levels of severity. The BDI-II is designed for persons 13 years of age and older, and can usually be completed within 5 to 10 minutes. Overall, the psychometric properties of the BD I-II are relatively good. The authors Beck, Steer, and Br own (1997) report estimates of internal reliability (Cronbachs Alpha) with outpatients (N = 500) as well as with a non-clinical population of college st udents (N = 120) of r =.92 and r =.93 respectively. Testretest reliability was assessed over a one week interval (N = 26) among a subsample of outpatients ( r = .93). The authors also report a correlation of r = .71 between the BDI-II and the Hamilton Psych iatric Rating Scale for Depression (HPRSD-R) among psychiatric outpatient s (N = 210), suggesting good convergent validity. The BDI-II was chosen to measure depression in the present study for three reasons. First, because it has a strong theoretical foundation closely fitting the criteria established in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, (DSM-IV) published by the American Psychiatric Association (1994). Second, the BDI-II wa s chosen to measure depression in the present study because of the strong empirical foundat ion upon which it was developed (more than 35 years of research ). Third, in addition to its solid psychometric properties the instrument is relatively easy to administer, score, and interpret.

PAGE 118

109 One or more in dividual items of the BDI-II ask participants to report if they have any thoughts of harming themselves. Da ta will be inspected at time of collection for compliance with instruct ions (e.g., one option chosen for each item). However, the study is anonymous research thus, th e identity of the participants will not be known to me, nor will I have any means of identifying participants. Therefore, r endering any intervention on my part impossible. Research Design A correlati onal research design was used in the present study to assess the relationship between emotional intel ligence and satisfaction with life among community college students after account ing for self-esteem, depression, and locus of control. I employed a hierarchic al regression analyses to investigate each of the five specific research questions discussed in Chapter One. Procedures The type of sample I used for t he study is a convenience sample consisting of 200 participants. All partici pants in the present study were asked to complete the following instruments: The Maye r, Salovey, and Caruso Emotional Intelligence Test (MSCEIT); Dieners (1985) Satisfaction With Life Scale (SWLS); Rotters (1966) Internal-Exter nal Locus of Control Scale (I-E Scale); Rosenbergs (1965) 10 item Self-Esteem Scale (RSES); and Becks (1997) revised 21 item Depression Inventory, the Beck Depression Inventory-II (BDI-II) All five instruments are of the paper/pencil fo rmat. The estimated time for completion of all five instruments ranged between si xty (60) to seventy-five (75) minutes.

PAGE 119

110 Before conducting the present study I made application to the USF Division of Research Compliance for author ization to conduct the present study as an exempt study (applic ation was approved October 21, 2005). I also made application to the office of Institutional Effectiveness CFCC requesting authorization to conduct the present research on both the Ocala and Citrus campuses (I received this authorization on 10-15-05). Having received approval from both institutions (USF and CFCC) I began the study by soliciting the aid of several fellow instructors at the CFCC Citrus C ounty campus. The aid I requested was permission to recruit participants from among their students. I had authorization from Mr. Edwin Goolsby (instructional m anager of the Citrus campus) to meet with students in pre-approved locations (e .g. classrooms, student lounge) for the purpose of having students take part in my study. I had a written script (see Appendix C) which I distri buted and read to students that makes explicit what was expected from participants as well as the voluntary and anonymous nature of the study. The first week of data collection I solicited partici pants from four classes, 84 students agreed to take part in the study, while two students declined. The second week of data collection I solicited particip ants from four classes, 76 students agreed to take part in the study, while 4 students declined to take part noting participation as part of another class. The third and final week of data collection I solicited students from five classes, 40 students agreed to take part in the study,

PAGE 120

111 but more than 12 students declined because t hey participated as part of another class. I recognized most students claiming prior participation as participants. Students that elected to take part in the study received all five instruments along with written instructions for comple ting the instruments. I monitored all students while they completed the instruments. I also co llected and inspected all instruments for compliance with instructions before students exited the room. Data Analysis The SAS (2003 ) system for statistical analysis of data was used to calculate the mean, standard deviation, and internal consistency reliabilities (coefficient alpha) for all measured vari ables and present them in table form. Scatter plots for each pair of variables were examined for linear relationships between each pair of variables. Pears on correlations between each of the measured variables were calculated and pres ented in matrix form in order to evaluate relationships among all variables. It was my in tention to build upon previous research in the present study. Therefore, similar to the Pa lmer et al. (2002) study I in vestigated the relationship between emotional intelligence and satisfaction with life. Similar to the Palmer et al. (2002) study I was primarily interest ed in whether emotional intelligence accounts for additional variance in satisfac tion with life, not accounted for by other predictor variables such as self -esteem, depression, and locus of control.

PAGE 121

112 However, unlike t he Palmer et al. (2002) study that employed a self-report measure of emotional intelligence, I empl oyed an ability measure of emotional intelligence (MSCEIT). I employed the same dependent variable (satisfaction with life) as the Palmer et al. (2002) study. However, my target population in the present study is CFCC students ra ther than the general population. My review of the satisfaction with life literature (e.g., Hong & Giannakopoulos, 1994) revealed that three of the most frequently cited predictors of satisfaction with life is self-esteem, fo llowed by depression and locus of control respectively. In the present study the co mbination of self-esteem, depression, and locus of control in an equation is refe rred to as the stem equation (LS = bo + b1 self-esteem + b2 depression + b3 locus of control). As the prior research suggested, these three variables together accounted for some portion of the variance (44%) in satisfaction with life in the present study. To te st each of the research questions ident ified in the present study it was necessary to add each of the other independent variables individually to the stem equation. The following five research equati ons were investigated; first, SWL = bo + b1 self-esteem + b2 depression + b3 locus of control + b4 EI total score. Second, SWL = bo + b1 self-esteem + b2 depression + b3 locus of control + b4 perceive emotion. Third, SWL = bo + b1 self-esteem + b2 depression + b3 locus of control + b4 facilitate thought. Fourth, SWL = bo + b1 self-esteem + b2 depression + b3 locus of control + b4 understand emotion. Fifth, SWL = bo + b1 self-esteem + b2 depression + b3 locus of control + b4 manage emotion. As each i ndependent variable is added to the stem equation any additional variance accounted for in the depen

PAGE 122

113 dent variable (life satisfaction) will result in changes to the overall R value of the equation. I first added emotional intelligence to tal score to the stem equation as discussed above followed by each of the four EI component variables. Multiple r egression analysis is an extension of simple linear regression. Thus, I began with an evaluation of all uni variate data for violations of assumptions regarding linear regression. The first assumption I consider was whether all variables have been measur ed without error. Since measurement error in multiple regression analyses may lead to overestimates or underestimates of relationships it is critical that measurement error be kept to a minimum. I evaluated measurement error by inspecting the reliability estimates reported for all instruments used in the pr esent study. I also calculated internal reliability estimates for each of the meas ures used in the present study using Cronbachs alpha. All scores from each of the five instruments were available in order to calculate Cronbachs alpha. The second assumption I ev aluated was the assumption of linearity. I inspected the scatter plot of the dependent variable and each independent variable for a linear relationship. The third assumption I evaluated was the assumption of homoscedasticity of errors which is the condition of equality of variance of errors. I evaluated this assumpti on visually by plotting residuals with predicted values looking for equal amounts of scatter all along the regression line. Extreme scores or outliers were evalua ted by calculating Cooks D. Cooks D indicates the influence of an extreme sco re by taking into account both the

PAGE 123

114 size of the residual and leve rage (position). Scores that have a Cooks D greater than 1 or much larger relative to others would be designated outliers. However, in the present study no outlie rs were identified. It is important to recognize that r values in simple linear regression represent the degree of relationship betwe en two variables. However, in the present study I employed a multivariate analysis in order to investigate the relationship among the criterion (dependent ) variable and multiple predictors (independent variables). Unless predictor va riables have zero correlations among them their combined r (s) are always less than additive. The primary fo cus of the present study was the investigation of incremental predictive validity (does the addition of a variable account for additional variance in the criterion variable ) between emotional intelligenc e (including sub-components of emotional intelligence) and satisf action with life after controlling for specific known predictors. In the present study, previous research exist to suggest the order of entering the variables into a prediction equation. I entered the variables logically in the order suggested by prior research. The scope of the present study is limited to the five specific research questions identified in Chapter One.

PAGE 124

115 Chapter Four Results This chapter presents results of statistical analysis related to the five specific research questions discussed in Ch apter One. First, I restate the five specific research questions. Second, univari ate statistics for each of the scaled variables are presented in Table 1. Third, all possible bivariant relationships among the variables as well as their p values are presented in a correlation matrix in Table 2. Fourth, I evaluate the dat a for critical violations of the most important assumptions for multiple regression. Fifth, I present the results of each hierarchical regression analysis employed to test each of the five specific research questions (Does EI or any of the four components of EI account for variance in satisfaction wit h life greater than self-est eem, depression, and locus of control?). I conclude this chapt er with a summary of the results. Restatement of the Research Questions 2. Does emotional intelligence concept ualized as a cognitive ability and measured by the Mayer, Salovey and Caruso Emotional Intelligence Test (MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 2. Does the ability to perceive and accurately express emotion (a component of emotional intelligence as measured by the MSCEIT) account for greater

PAGE 125

116 variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? 3. Does the ability to use emot ion to facilitate thought (a component of emotional intelligence as m easured by the MSCEIT) acc ount for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 4. Does the abilit y to understand emotion (a compon ent of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? 5. Does the ability to manage em otion for emotional growth (a component of emotional intelligence as measured by the MSCEIT) account for greater variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? Univariate Statistics Before conducti ng regression analysis of scores, simple univariate statistics were calculated in order to gain some ov erall understanding of how each variable is distributed. Univariate stat istics are presented in Table 1.

PAGE 126

117 Table # 1 Univariate Statistics for all Scaled Variables Variable N Mean Std Dev Skewness Kurtosis Min Max ________________________________________________________________________ EIT 200 84.79 16.07 -0.03 -0.76 40.00 123.00 EI1 200 98.44 16.01 -0.12 -0.23 51.00 132.00 EI2 200 86.97 16.52 0.06 -1.04 56.00 124.00 EI3 200 81.50 14.68 -0.16 -0.40 41.00 118.00 EI4 200 86.63 13.72 -0.06 -0.59 42.00 114.00 swl 200 22.56 6.29 -0.36 -0.51 8.00 35.00 self 200 20.94 4.67 -0.01 -0.20 9.00 30.00 dep 200 9.60 8.14 1.15 0.90 0.00 34.00 loc 200 10.73 3.60 0.22 0.19 2.00 22.00 ________________________________________________________________________ Note EIT = emotional intelligence total score, EI1 = perceiving emotions, EI2 = facilitating thought, EI3 = understanding emotions, EI4 = ma naging emotions, swl = satisfaction with life, self = self-esteem, dep = de pression, loc = locus of control

PAGE 127

118 I calculated N values (number of observations), mean, standard deviation, skewness, kurtosis, and minimum and maximu m scores of all variables. In addition, the following statistical displays were generated for each variable: box plots, stem and leaf displays, and normal probability plots. The N for each variable was 200 suggesting no observations were missing. The completeness of the data is probably the result of t he method I employed to collect data. Participants were given at least one week prior notice to the administration of the study. T hus, all participants had an opportunity to make necessary arrangements in order to participate in the study. All participants completed the assessment instruments individually (independently) during or immediately after class. All participants were monitored by me while they completed instruments and all instruments were checked by me for compliance with instructions at the time of collection (e .g., one response for each item). The original intended sample size was 160 or more. However, the obtained sample size turned out to be 200. The data collection stage of the pr esent study was completed when the revised target number of 200 complet ed assessment packets were obtained. Some problems with scoring as well as data entry was detected at the data analysis stage. However, these errors once detected were corrected such that no observations were lost from the sample An outlier score on the MSCEIT was found to be an error in data entry, and an unu sual distribution of self-esteem scores (RSES) revealed an error in scoring (some items are reversed scored). Therefore, I replaced the incorrect MSCE IT score with the correct score, re

PAGE 128

119 scored the RSES and entered the corrected scores. An examination of the minimum and maximum values for each of the variables suggested confidence in scoring as well as accuracy in data entry (all scores were within the range of possibility). All variables except the demographic variables (gender, age, race) and depression demonstrated skewness within an acceptable range of normality (SK > -1.0 and < 1.0). Depression demonstr ated a positive skew of 1.15. Thus, depression demonstrated a skew slightly greater than what is normally considered acceptable. However, it is not far enough outside what is normally considered acceptable to constitute an important concern. All non-demographic variables demonstrated kurtosis wit hin an acceptable range of normality (KU > -2.0 and < 2.0). Overall, uni variate statistics discussed above as well as box plots, stem and leaf displays, and normal probability curves suggest all non-demographic variables (except depression) have relative ly normal distributions. My discussion of each variable descriptive statistic is intended to help in the understanding of how individual variables are distributed. However, it should be remembered that normal distribution of individual independent variables is not an assumption of multiple regression analysis. Normal distri bution of errors along the regression line is an assumption of regression analys is and will be discussed later along with other assumptions fo r multiple regression.

PAGE 129

120 Bivariate Correlations The next phas e of data analysis consisted of computing all possible bivariate correlations among the variabl es and presenting them along with associated p values in a correlation matrix (s ee Table 2). This table should be reviewed in order to understand the pattern (level and direction) of correlation between all scaled variables. It is necessary to consider the simple bivariate correlations among all variables in multiple regression analysis for the following reasons. First, multiple regression is an ex tension of simple regression. However, unless all variables in a multiple regr ession are uncorrelated variables, the resulting R (the percent of variance in the dependent variable that is accounted for by the linear combination of predictor variables) are less than additive. This is because intercorrelated variables always demonstrate some redundancy in the prediction of a dependent variable. All bivari ate correlations between each pair of scaled variables are presented in the form of a correlation matrix in Table 2.

PAGE 130

121 Table # 2 Correlation Matrix ________________________ _____________________ ___________________ EIT EI1 EI 2 EI3 EI4 SWL self dep loc __________________________________________ ______________________ EIT 1.00 EI1 .67* 1.00 EI2 .84* .49* 1.00 EI3 .85* .40* .65* 1.00 EI4 .81* .31* .61* .69* 1.00 SWL -.04 .02 -.02 -.03 -.06 1.00 self .00 -.00 .06 -.02 .00 .56* 1.00 dep .07 01 .04 .07 .0 4 -.60* -.58* 1.00 loc .04 -.06 .02 .14 .03 -.12 -.32* .21* 1.00 __________________________________________ ______________________ Note EIT = emotional intelligence total score, EI1 = perceiving emotions, EI2 = facilitating thought, EI3 = understanding emotions, EI4 = managing emotions, SWL = satisfaction with life, self = self-esteem, dep = depression, loc = locus of control, = p < .05

PAGE 131

122 In order to be consistent, I first discuss the relationship among EI total score and the four EI component scores. Second, I discuss the relationship among the four EI components with each other Third, I discuss the relationship between EI total score as well as EI co mponent scores with each of the studys four remaining scaled variables (satisfa ction with life, self -esteem, depression and locus of control). Fourth, I discuss t he relationship between satisfaction with life (dependent variable) and eac h of the three known pr edictor variables selfesteem, depression and locus of control. Fifth, I discuss the relationship among the known predictor variables self-esteem, depression and locus of control. The MSCEIT yields a total score and four component scores reflecting the Mayer and Salovey (1997) model of emotional intelligence. EI total score in the present study demonstrated moderate positiv e bivariate correlations with all four component scores (perceiving emotions r = .67, facilitating thought r =.84, understanding emotions r = .85 and managing emotions r =.81) all p values < .0001 suggesting there is less than 1 chance in 10,000 of obtaining a sample correlation of this size if the population correlati on were zero. Since the pre-set level of statistical significance in the present study is p < .05 all of the above p values are significant. The results presented above are expected since the EI total score is comprised of four component scores. Second, t he models four component score s demonstrated the following relationships among each other a) perceiving emotions with facilitating thought r = .49, understanding emotions r = .40, and m anaging emotions r = .31 all with p

PAGE 132

123 values < .05. b) facilitati ng thought with understanding emotions r = .65 and managing emotions r = .61 both with p values < .05. c) understanding emotions with managing emotions r = .69, p < .05. The obtained intercorrelations between the EI components are consistent with the intercorrelations reported by the authors Mayer, Salovey, and Caruso (2002) in the MSCEIT manual. The authors report a) perceiving emoti ons with facilitating thought r = .54, understanding emotions r = .30, and managing emotions r = .35 b) facilitating thought with understanding emotions r = .43 and managing emotions r = .50 c) understanding emotions with managing emotions r = .51 all with p < .05. Overall, the four component scores are intercorrelated with EI tota l score as well as with each other. This pattern of low to moderate corre lation suggests the four components are related without complete redundancy. Third, EI total score as well as all four component scores demonstrated low or no correlation with each of the studi es four remaining scaled variables. The correlation between EI total and the remaining variables are satisfaction with life, r = -.04, self-esteem, r =.00, depression, r =.07, and locus of control r =.04 all with p > .05). Failure to find even simple correlations between EI total and the dependant variable (SWL) as well as t he other three independent variables suggest the primary research question; Does emotional intelligence concepttualized as a cognitive ability and measur ed by the Mayer, Salovey and Caruso Emotional Intelligence Test (MSCEIT) account for greater variance in satisfaction with life among community college students than self-esteem, depression, and

PAGE 133

124 locus of control can not be answered in the affirmative. Correlations between components of EI with the remaining scaled variables ranged between r = .00 and r = .14, all with p > .05. Therefore, re sults of the present st udy do not support the findings of prior research such as: a) Bar-On (1997) Martinez-Pons (1999) Ciarrochi et al. (2000) Mayer et al. (2000) Pa lmer, et. al. (2002) Law et al. (2004) and Extreme et al. (2005) reporting low to moderate positive correlation between EI and satisfaction with life, b) Ciarroch i et al. (2000) reporting a positive correlation between EI and self-esteem, c) Martinez-Pons (1997) and Schutte et al. (1998) reporting a moderate negative re lationship between EI and depression, and d) Brown and Schutte (2006) reporting a moderate positive relationship between EI and internal locus of control. Fourth, satisfaction with life in the present st udy demonstrated an r = .60 with depression followed by an r = .56 with self-esteem, eac h of the associated p values < .05. This suggests that self -esteem has a low moderate positive relationship with satisfaction with life and depr ession has a low moderate although inverse (negative) relationship with sati sfaction with life. T hus, the bivariate correlation between satisfaction with life and self-esteem as well as satisfaction with life and depression is significant at p < .05 These obtained correlations are in agreement with much of the literature t hat often report both se lf-esteem (e.g., Parkerson et al., 1990; Vermunt et al., 1989) and depression (e.g., Hyer et al., 1987; Martinez-Pons, 1997) as important predict ors of satisfaction with life. It is important to note that the relationship between self-e steem and life satisfaction is

PAGE 134

125 positive however, the relationship bet ween depression and life satisfaction is negative (inverse). The correlation betw een satisfaction with life and the remaining scaled variable locus of control is r = .12 however, its p value is > .05 and thus is not statistically significant. T hus, the obtained correlation between satisfaction with life and locus of control does not support previous research (e.g., Hickson et al., 1988) reporting a small to moderate negative relationship between satisfaction with life and locus of control. Fifth, the method of hierarchical regre ssion analysis employed in the present study enters variables according to research (researcher logically enters variables). My review of the related re search suggests self-esteem, followed by depression, and locus of control respectively are all important predictors of satisfaction with life. Thus, in order to ma intain consistency I discuss correlations among each of these variables (self-esteem depression, and locus of control) in that order. First, self-esteem demonstrated a correlation of r = .58 with depression and r = .32 with locus of control, both p values < .05. It is important to note the direction of each of these correlations. Both of these relationships are negative (inverse) thus, the data suggests that self-esteem increases as depression decreases (low scores reflect less depr ession) and self-esteem increases, when internal locus of control increases (low scores). Rotter (1966) noted that low locus of control scores suggest an internal locus of control and high scores suggest an external locus of control. T hese correlations are in agreement with much of the research reporting an inve rse correlation between self-esteem and

PAGE 135

126 depression as well as locus of control. Depression in the study demonstrated a low positive correlation with locus of control r = .21, p < .05. Assumptions of Regression Analysis Pedhazur (1997) notes that knowledge and unders tanding of the situations when violation of assumptions lead to seri ous biases, and when they are of little consequence, are essential to meaningfu l data analysis (p. 33). The first assumption of regression analysis I discuss in regards to the present study is measurement without error. This assumption is critical to regression analysis; it is not robust to violations of this assump tion regardless of sample size. Measurement error in multiple regression analys is may lead to over-estimate or underestimate of relationships. Thus, it is crit ical that measurement error be kept to a minimum. Pedhazur (1997) discusses two methods of evaluating measurement error. First, a comprehensive review of the related research can suggest how reliable an instrument has been within specif ic contexts. Second, I calculated a well known estimate of internal reliability such as t he Kuder-Richardson formula 20 coefficient or Chronbac hs coefficient alpha. I reviewed the research literature on all of the scaled variables and identified each of the assessm ent instruments employed in the present study. The first consideration, was the history of each instrument, how frequently as well as over what span of time the instrum ent has been used in related research. Second, for each instrument, what level of internal reliability was reported in previous research. The five assessment instruments used in the present study

PAGE 136

127 are often used in related research, have been used for some length of time (are widely known), are often are used in contemporary research, and have been reported to demonstrate adequate internal reliability and validity within the context of specific studies. In acco rd with Pedhazurs second reco mmendation Cronbachs alpha was calculated for each assessment instrum ent. Cronbachs alpha is a measure of the extent to which the individual items t hat constitute a test correlate with one another. The theory behind this is that a re liable test should minimize the measurement error so that the error (inherent in all measures) is not highly correlated with the true score. Cronbachs coefficient alphas can be found in Table 3. The SWLS, RSES, BDI-II, and I-E scale all employ a straight forward method of scoring and interpretation. For example on the I-E scale answer choices are either correct or not and scores are derived by simply adding correct responses. In the case of the BDI-II answer choices are assigned numerical values corresponding to level and score s are derived by simply adding across items. Each individuals cumulative score indicate the level of the variable Therefore, in regards to the SWLS, RSES, BDI-II, and I-E scale individual item responses were used to generate Cronbachs coefficient alphas. The obtained Cronbachs coefficient alphas are SWLS r =.82, RSES r =.86, BDI-II r =.82, and I-E scale r =.64.

PAGE 137

128 The MSCEIT is a proprietary instrument publis hed by Multi Health Systems Inc. All MSCEIT scoring must be done by the publisher either by sending completed response forms or entering the dat a at a secure web page. I entered electronically all 141 MSCEIT item answer choices for all 200 participants. The MSCEIT employs both expert consensus scoring (N =21) and general consensus scoring (N = 5,000). Each MS CEIT response is assigned a score based on the proportion of the consensus samp le (either general or expert) that selected that response. For example if a person selects an alternative chosen by 75% of the norm group, the i ndividuals score is increm ented by .75 and so on (Mayer et al., 2004, p. 200). In the present study the MSCEIT proportional scores were entered in the calculation of Cronbachs coefficient alphas EI total r =.94, perceiving emotions r =.90, facilitating emotions r =.79, understanding emotions r =.85 and managing emotions r =.85. For the present study I ch ose general consensus scori ng as the method of scoring the MSCEIT.. However, their appears to be very little difference between types of scoring the MSCEIT. The authors report a very high correlation between general and expert consensus scoring at the full scale r =.98, and component level, perceiving emotions r =.98, facilitating thought r =.97, understanding emotions r =..98, and managing emotions r =.96. First, I present the estimate of internal reliability published in the MS CEIT manual by the authors Mayer, Salovey, and Caruso (2002). Second, I discuss estimates of MSCEIT internal reliability reported by other researchers.

PAGE 138

129 The internal reliability estimates (split half) reported in the MSCEIT manual by the authors Mayer et. al ., (2002) are EI total r =.93, perceiving emotions r =.91, facilitating thought r =.79, understand ing emotions r =.80, and managing emotions r =.83. Other researchers reporting internal reliabil ity estimates for the MSCEIT include Bracket and Mayer (2003) reporting a test-retest with a two week interval ( r =.86), and Ciarrochi et al. (2000) reporting a full scale split half reliability of r =.90.

PAGE 139

130 Table # 3 Cronbachs Coefficient Alpha ________________________ ________________________________________ Variables Raw ________________________ _____________________ ___________________ MSCEIT (total) .94 EI1 (perceiving emotions) .90 EI2 (facilitating thought) .79 EI3 (understanding emotions) .85 EI4 (managing emotions) .85 SWL .82 RSES .86 BDI-II .82 I-E Scale .64 Note MSCEIT (total) = Mayer, Salovey and Caruso Emotional Intelligence Test, EI1 = perceiving emotions, EI2 = fa cilitating thought, EI3 = understanding emotions, EI4 = managing emotions, SWL = Satisfaction With Life Scale, RSES = Rosenberg Self-Esteem Scale, BDI-II = Beck Depression Inventory-2, and I-E Scale = Internal-External Locus of Control

PAGE 140

131 The Mayer, Salovey and Caruso Emotional Intelligence Test (MSCEIT) demonstrated a full scale raw score Cronbachs alpha r = .94 suggesting relatively high internal consistency. The four components of the Mayer and Salovey (1997) model of EI demonstrat ed the following Cronbachs alpha (EI1) perceiving emotions r = .90; (EI2) fa cilitating thought r = 79; (EI3) understanding emotions r = .85; (EI4) managing emotions r = .85 The satisfaction with life scale (SWLS) demonstrated a raw score Cronbachs alpha r = .83 suggesting relatively moderate internal consistency. Rosenbergs Self-Esteem Scale (RSES) demonstrated a raw score Cronbachs alpha r = .86 suggesting relatively moderate internal consistency. Becks Depression Inventory-II (BDI-II) demonstrated a raw score Cronbachs alpha r = .82 suggesting relatively moderate internal consistency. Rotters Internal-External Locus of Control Scale (I E Scale) demonstrated a raw score Cronbachs alpha r = .64 suggesting a low level of internal consistency. Osborne, Christensen, and Gunter (2001) reported t hat the average alpha reported in top Educational Psychology journals was .83. The question is how large must a reliability coefficient be to be considered acceptable? A widely used rule of thumb of r =.70 has been suggested by Nunnally (1978). However, it should be reme mbered that this is only a rule of thumb and many studies in the social scien ce literature report coefficient alpha reliabilities under .70 and even under .60.Over all, the instruments employed to measure the scaled variables (except the I-E scale) demonstrated adequate internal consistency within the context of the present study.

PAGE 141

132 The second assu mption of multiple regression a nalysis I wish to discuss in regards to the present study is independence of errors. That is, the errors from different observations are independent of eac h other. This assumption is most often violated with studies that employ cluster sampling and/or repeated measures designs. The present study does not employ either cluster sampling or repeated measures design. The assumption of independence of errors is usually met with the proper design of the study. A plot of the errors (residuals) suggested a pattern-less distribution around ze ro. Thus, the design of the present study as well as an evaluation of plott ed residuals suggest the independence of errors assumption has not been violated. The third assu mption of multiple regression analysis I discuss in regards to the present study is linearity of rela tionship between independent and dependent variables. Multiple regression represents the dependent variable as a linear function of a combination of independent variables. Thus, it is critical that the relationship between the independent variables and dependent variable as well as among the independent vari ables be linear. In regards to the present study, two methods of checking for violation of the linearity assumption were employed. First, prior related research was examin ed that suggested the relationship between the independent variables and the dependent variable as well as among the independent variables are linear. Second, scatterplots of the residuals of each regression analysis (EI, Self-Esteem, Depression, Locus of control) and the predicted values of the dependent variable (SWL) were examined for evi

PAGE 142

133 dence of nonlinearity. I evaluated each scatt er plot of the residuals against the predicted values and observer relatively random scatter along a horizontal regression line. Overall prior research (e.g ., Palmer et al., 2002) as well as plots of residuals against predicted values in the present study suggests the relationship between the independent variables and the dependent variab le is linear. The fourth assumption of multiple regressi on analysis I discuss in regards to the present study is equality of or cons tant variance of errors (homoescedasticity). This assumption asserts that for eac h combination of values of the independent variables (predictor) the variance of the errors are the sa me. The method of evaluating data in the present study for violat ions of this assumption was to plot regression residuals against predicted valu es. This assumption was evaluated by looking for evidence of nonconstant vari ance (heteroscedasticity) of residuals across the range of predicted values for each regression analysis. Overall the plots of residuals in the present study suggested relatively constant variance (equal dispersion) of erro rs for each of the independ ent variables. Multiple regression is relatively robus t to minor violations of th is assumption especially with large sample size. Based on an evaluat ion of the residual plots as well as evaluation of sample size (N = 200) the present study does not appear to critically violate the equality of or const ant variance of errors assumptions. The fifth assu mption of multiple regression ana lysis I discuss in regards to the present study is normality of resi duals. Pedhazur (1997) noted that for regression the normality test should be applie d to the residuals rather than the

PAGE 143

134 raw scores. I employed a test available wi th SAS (version 9.0) the Shapiro-Wilk, as one index of the normality of residuals as well as an evaluation of box plots, normal probability plots, and stem and leaf displays. The null hypothesis of a normality test is that there is no departure from normality. Thus, when the p value is greater than .05, it fails to reject t he null hypothesis and t hus the assumption holds. The Shapiro-Wilk suggested p > .05 for each of the regression analysis in the present study. Additionally, an evaluati on of the box plots, stem and leaf displays as well as normal probability plots of the residuals for each regression analysis suggest no critical violations of the normality of residuals assumption. Thus, there does not appear to be a critical violation of the assumption of normality of residuals in the present study. Hierarchical Regression Analysis Previous research suggests self-esteem, depression and locus of control are predictive of satisfaction with life. Ho wever, the present study attempts to determine how much additional variance in satisfaction with life emotional intelligence accounts for over and above thes e known predictors. Thus, the first regression analysis performed consisted of the three known predictors selfesteem, depression, and locus of control with satisfaction with life entered as the dependent variable and will be referred to as the stem equation. The results of this regres-sion analysis (SWL= Self-E steem + Depression + Loc) can be found in Table 4.

PAGE 144

135 Stem Regression Analysis The results of the multiple regression analys is suggested that the linear combination of self-esteem + depression + locus of control accounted for approximately 44% of the varianc e in satisfaction with life ( R = .4375). The significance test associated with this R is F (3, 196) = 50.81, p < .05. Thus, the model is significant at the .05 leve l. Therefore, I may conclude that R = .4375 is probably greater than zero in the population with a 95% confidence level. SAS reports both non-standardized coefficien ts as well as standardized coefficients for each predictor. However, si nce different predictors normally have different standard deviations, and these di fferences affect t he size of nonstandardized coefficients it is more appropriate to review the standardized coefficients often called beta weights. T he standardized coefficient represents the amount of change in the dependent va riable associated with a one-unit standard deviation (SD) change in that predi ctor, while holding constant the remaining predictors. The standardi zed coefficients for the stem model equation (SWL = selfesteem + depression + locus of control) can be found in Table # 4 under the column labeled standardized estimate ( B ). The calculated linear model for the stem equation is (SWL) Y = 0.340 (self-esteem) 0.42280 (depression) + 0.07297 (locus of control). The significant predictors of this model are selfesteem and depression. The most important predictor of sa tisfaction with life is

PAGE 145

136 depression ( B = -0.42), followed by self-esteem ( B = 0.34). locus of control did not significantly predict satisfaction with life ( B = 0.07, p >.05 ). The squared se mi-partial correlation coeffici ent for depression as well as self-esteem is 0.11893 and 0.07258 respectively This statistic suggests that depression uniquely accounts for approximately 12% of the explained variance in life satisfaction, and self-e steem uniquely accounts for approximately 7% of the explained variance in satisfaction with lif e. However, locus of control demonstrated a squared semi-partial correlation coe fficient of 0.00478, not significant at the p < .05 level. Thus, results from t he present study suggest depression and self-esteem are both im portant predictors of life satisf action. However, when both self-esteem and depression were held constant locus of control did not account for additional variance in satisfaction with life. Table # 4 Stem Equation SWL = Self-esteem + Depression + Locus of control ________________________ ________________________________________ b B Squared Parameter St andard Standardized Semi-partial Variable Estimate Error t Value Pr > |t| Esti mate Corr Type II Intercept 14.72224 2.66486 5.52 <.0 001 0 self 0.45879 0.09124 5.03 <.0001 0.34040 0.07258 depress -0 .32668 0.05075 -6.44 <.0001 -0.42280 0.11893 loc 0.12759 0.09883 1.29 0.1982 0.07297 0.00478 Note R = .4375, Raj = .4289, Rms =4.7536

PAGE 146

137 Stem regression analysis plus EI total The remainder of this chapter consists of the results of each regression analysis designed to test the five spec ific research questions discussed in chapter one. The focus of the study is t he investigation of how much if any additional variance in satisfaction with life emot ional intelligence accounts for over other known predictors (self-esteem, depressi on, locus of control) Therefore, the following five regression analysis consist of adding individually emotional intelligence total score as well as each of f our EI component scores to the stem equation (discussed above) and noti ng any significant change in the R (total amount of variance explained by the linear combination of predictors). The first re search question asked whether emotional intelligence, conceptualized as a cognitive ability and measur ed by the MSCEIT, accounts for greater variance in satisfaction with life among community college students than selfesteem, depression, and locus of control? When emotional intelligence total score was added to the stem regression equ ation the results suggest that the linear combination of depressi on, self-esteem, locus of control, and emotional intelligence has an R = 0.4376 suggesting approximatel y 44% of the variance in satisfaction with life is accounted for. The small change in R = 0.0001, suggest emotional intelligence total score a ccounts for little or no variance in life satisfaction over depression, self-esteem, and locus of control.

PAGE 147

138 The linear combination of depression, self-esteem, lo cus of control, and emotional intelligence total score accounts for approximately 44% of the variance in satisfaction with life, significant at the .05 level. However, the relatively small change in R (0.0001) when emotional intelligenc e was added to the stem model suggests that we can not reject the null hypothesis. Therefore, I can conclude that emotional intelligence (total score) does not account for additional variance in satisfaction with life over and above depression, self-esteem and locus of control. Table # 5 Stem Regression Analysis Plus EI (total score) SWL = Self-esteem + Depression + Loc + EIt ________________________ _____________________ ___________________ b B Squared Parameter Standard Standardized Semi-partial Variable Estimate Error t Value Pr > |t| Esti mate Intercept 15.11707 3.11056 4.86 <.0001 0 self 0.46010 0.09161 5.02 <.0001 0.34137 0.07275 depress -0.32566 0.05104 -6.38 <.0001 -0.42148 0.11742 loc 0.12867 0.09917 1.30 0.1960 0.07359 0.00486 EIT -0.00523 0.02112 -0.25 0.8046 -0.01336 0.00017 ________________________ _____________________ ___________________ Note R = .4376, Raj = .4261, Rms = 4.7650, R = .0001

PAGE 148

139 Stem regression analysis plus perception of emotion (EI1) In the present study the second rese arch question asked whether the ability to perceive and accurately expre ss emotion (a component of emotional intelligence as measured by the MSCEIT ), accounts for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? When per ceiving emotions (EI1) was added to the stem regression equation the results suggest that the linear combination of depression, self-esteem, locus of cont rol, and perceiving emotions has an R = 0.4387 suggesting approximately 44% of the total variance in satisfaction with life has been accounted for. Once again the sign ificant predictors were depression ( B = 0.42364) and self-esteem ( B = 0.34068). The non-significant predictors were locus of control ( B = 0.07538) and perceiving emotions ( B = 0.03535). The standardized regression estimates are (S WL) Y = 0.34068 (self-esteem) 0.42364 (depression) + 0.07538 (locus of control) 0.03535 (EI1). The squared se mi-partial correlation coefficients are as follows; depression 0.11936, and self-esteem 0.07269, respecti vely. Once again suggesting depression uniquely accounts for approximately 12% and self-esteem 7% of the total explained variance in satisfaction with life. The squared semi-partial correlation coefficients of the non-significant remain ing predictors are locus of control = 0.00508, and perceiving emot ions (EI1) = 0.00124.

PAGE 149

140 Therefore, the linear combination of depression, self-esteem, locus of control and perceiving emotions (EI1) accounts for approximately 44% of the variance in satisfaction with life. Howe ver, the relatively small change in R (0.0012, p > .05) obtained when per ceiving emotions was added to the stem model suggests that we can not reject the null hypothesis. Therefore, I can conclude that the ability to perceive em otion does not account for additional variance in satisfaction with life over and above depression, self-esteem, and locus of control. Table # 6 Stem Regression Analysis Plus EI1 (perceiving emotions) SWL = Self-esteem + Depression + Loc + EI1 ________________________ ________________________________________ b B Parameter Standar d Stand ardized Squared Variable Estima te Error t Value Pr > |t| Estimate Semi-partial Intercept 13.31056 3.42720 3.88 0.0001 0 self 0.45917 0.09137 5.03 <.0001 0.34068 0.07269 depress -0.32733 0.05083 -6.44 <.0001 -0.42364 0.11936 loc 0.13180 0.09918 1.33 0.1855 0.07538 0.00508 EI1 0.01387 0.02112 0.66 0.5123 0.03530 0.00124 ________________________ _____________________ ___________________ Note R = .4387, R aj = .4272, Rms = 4.7605, R = .0012

PAGE 150

141 Stem regression analysis plus facilitating thought (EI2) The third research question asked whether the ability to use emotion to facilitate thought (a component of emoti onal intelligence as measured by the MSCEIT), accounts for greater variance in life satisfaction among community college students than self-esteem, depression, and locus of control? When facilitating thought (EI2) was added to the stem regression equation the results suggest the following. The linear combinatio n of depression, self-esteem, locus of control, and facilitating thought is R = 0.4379 suggesting approximately 44% of the total variance in satisfaction with life is accounted for. Once again the significant predictors were depression ( B = 0.42035) and self-esteem ( B = 0.34343). The non-significant predi ctors were locus of control (B = 0.07389) and facilitating thought ( B = 0.02072). The standar dized regression estimates are (SWL) Y = 0.34343 (selfesteem) -0.42035 (depression) + 0.07389 (locus of control) 0.02072 (EI2). The squared semi-partial correlation coefficient s were as follows depression and selfesteem 0.11645, and 0.07288, respectively Once again suggesting depression uniquely accounts for approximately 12% and self-esteem 7% of the total explained variance in satisfaction with life. The squared semi-partial correlation coefficients of the non-significant remainin g predictors were locus of control = 0.00489, and facilitating thought (EI2) = 0.00042.

PAGE 151

142 Therefore, the linear combination of depression, self-esteem, locus of control, and facilitating thought (EI2) accounts for approximately 44% of the variance in satisfaction with life. Howe ver, the relatively small change in R (0.0004, p > .05) obtained when facilitating thought was added to the stem regression analysis suggests that we can not reject the null hypothesis. Therefore, I can conclude that facilitating thought does not account for additional variance in satisfaction with life over depression, self-esteem, and locus of control. Table # 7 Stem Regression Analysis Pl us EI2 (facilitating thought) SWL = Self-esteem + Depression + Loc + EI2 ________________________ _____________________ ___________________ b B Squared Parameter Standar d Standa rdized Semi-partial Variable Estimate Error t Value Pr > |t| Estimate Intercept 15.28754 3.05166 5.01 <.0001 0 self 0.46288 0.09206 5.03 <.0001 0.34343 0.07288 depress -0.32479 0.05110 -6.36 <.0001 -0. 42035 0.11645 loc 0.12919 0.09914 1.30 0.1941 0.07389 0.00489 EI2 -0.00789 0.02061 -0.38 0.7022 -0.02072 0.00042 __________________________________________ ______________________ Note R = .4379, R aj = .4264, Rms = 4.76399, R = .0004

PAGE 152

143 Stem regression analysis plus understanding emotions (EI3) The fourth research question asked whether the ability to understand emotions (a component of emotional inte lligence as measured by the MSCEIT), account for greater variance in life satisfaction among community college students than self-esteem, depression, and locu s of control? When understanding emotions (EI3) was added to the stem regression equation the results suggest the following. The linear combination of depr ession, self-esteem, locus of control, and understanding emotions has an R = 0.4375 or accounts for approximately 44% of the total variance in satisfaction wit h life. The associated significance test is F (4,195) = 37.91, p < 05. Once again the significant predictors are depression ( B = 0.42274) and self-esteem ( B = 0.34045). The non-significant predictors are locus of control ( B = 0.07307) and understanding emotions ( B = 0.00063) p >.05. The standar dized regression estimates are (SWL) Y = 0.34045 (selfesteem) 0.42274 (depression) + 0.07307 (locus of control) 0.00064 (EI3). The squared semi-partial correlation coeffici ents are as follows depression and selfesteem 0.11821, and 0.07288, respectively Once again suggesting depression uniquely accounts for approximately 12% and self-esteem 7% of the total explained variance in satisfaction with life. The squared semi-partial correlation coefficients of the non-significant remainin g predictors were locus of control = 0.00471, and understanding emotions (EI3) = 0.00004.

PAGE 153

144 The linear combination of depr ession, self-esteem, lo cus of control, and understanding emotions (EI3) accounts for approximately 44% of the variance in satisfaction with life. However, no change in R (0.0000) was detected when understanding emotions was add ed to the stem regression analysis; thus, suggesting we cannot reject the null hypot hesis. Therefore I can conclude that understanding emotions does not account for additional va riance in satisfaction with life over self-esteem, depression, and locus of control. Table # 8 Stem Regression Analysis Plus EI3 (understanding emotions) SWL = Self-esteem + Depression + Loc + EI3 ________________________ _____________________ ___________________ Squared Parameter Standard Standardiz ed Semi-partial Variable Estimate Error t Value Pr > |t| Estimate Intercept 14.74094 3.10890 4.74 <.0001 0 self 0.45886 0.09167 5.01 <.0001 0.34045 0.07228 depress -0.32663 0.05103 -6.40 <.0001 -0.42274 0.11821 loc 0.12775 0.10002 1.28 0.2030 0.07307 0.00471 EI3 -0.00027 0.02331 -0.01 0.9906 -0.00063 3.98875 ________________________ ________________________________________ Note R = .4375, R aj = .4259, Rms = 4.76578, R = .0000

PAGE 154

145 Stem regression analysis plus the ability to manage emotions (EI4) The fifth re search question asked whether the ability to manage emotions for emotional growth (a component of em otional intelligence as measured by the MSCEIT) accounts for greater variance in satisfaction with life among community college students than self-esteem, depression, and locus of control? When the ability to manage emotions (E I4) was added to the stem regression equation the results suggest t he following. The linear combination of depression, self-esteem, locus of control, and managing emotions (EI4) demonstrated an R = 0.4400 suggesting approximately 44% of the total varianc e in satisfaction with life is accounted for. The associated significance test is F (4,195) = 38.30, p < 0.0001. Again the significant predictors are depression ( B = 0.41953) and selfesteem (B = 0.34300). The non-significant predi ctors are locus of control ( B = 0.0 7482) and managing emotions ( B = -0.05045). The standar dized regression estimates are (SWL) Y = 0.34300 (selfesteem) 0.41953 (depression) + 0.07482 (locus of control) 0.05045 (EI4). The squared semi-partial correlation coeffici ents are as follows depression and selfesteem 0.11677, and 0.07357, respective ly. Again suggesting depression uniquely accounts for approximately 12% and self-esteem 7% of the total explained variance in satisfaction with life. The squared semi-partial correlation coefficients of the non-significant remainin g predictors are locus of control = 0.00502, and the ability to manage emoti ons (EI4) = 0.00253.

PAGE 155

146 The linear combination of depression, self-esteem, lo cus of control, and ability to manage emotions (EI4) accounts for approximately 44% of the variance in satisfaction with life. However, the relatively small change in R (0.0025) obtained when managing emotio ns was added to the stem model regression analysis suggests that we cannot reject t he null hypothesis. Therefore, I cannot conclude that the ability to manage emoti ons accounts for additional variance in satisfaction with life over and above self-est eem, depression and locus of control. Table # 9 Stem Regression Analysis Plus EI4 (managing emotions) SWL = Self-esteem + Depression + Loc + EI4 __________________________ ________________________ ______________ b B Parameter Standard Standardized Squar ed Variable Estima te Error t Value Pr > |t| Estimate Semi-partial Intercept 16.59348 3.32768 4.99 <.0001 0 self 0.46230 0.09134 5.06 <.0001 0.34300 0.07357 depress -0.32416 0.05083 -6.38 <.0001 -0 .41953 0.11677 loc 0.13082 0.09892 1.32 0.1876 0.07482 0.00502 EI4 -0.02313 0.02462 -0.94 0.3487 -0.05045 0.00253 ________________________ _____________________ ___________________ Note R = .4400, Raj = .4285, Rms = 4.75503, R = .0025

PAGE 156

147 Summary of Results Univariate distributions of the scaled vari ables were examined and all found to be within acceptable parameters (skewness 1.00 and kurtosis < 2.0). These distributions reflected the populati on that the sample was drawn from (CFCC students). The bivariat e correlations between emotional intelligence total score and each of the four components of EI ( r = .67, .84, .85, .81) respectively, all significant at p < .05.suggested a pattern of low to moderate positive correlations. Correlations among the components range between r = .31 and r = .68, they are all positive, and significant at p < .05 suggesting the co mponents are related without total redundancy. The correlati on between EI total as well as all four EI components with satisfaction with life (dependent variable) range between r = .01 and r = -.06 and are not significant at p < .05 level. This finding is interesting because it does not support prior research (e.g., Palmer et al., 2002; Ciarrochi, et al., 2000; Law et al., 2004) that report finding correlations between EI or components of EI and satisfaction with life. The correlati on between emotional intelligence to tal as well as each of the four EI components with each of the other predictor variables (self-esteem, depression, locus of control) range between r = .00 and r = .13 and are not significant ( p > .05). This finding is also interesting because it does not support prior research (e.g., Hong & Giannakopoulos, 1994; Kopp & Ruzicka, 1993)

PAGE 157

148 that reported finding significant correla tions between satisfaction with life and self-esteem, depression, and locus of control. The correlation among each of the other predictor variables self-esteem, depression, and locus of control are as follows, self-esteem and depression ( r = 0.58, p < .05), self-esteem and locus of control ( r = 31, p < .05) and between depression and locus of control ( r = .21, p < .05). This finding supports prior research (e.g., Palmer et al, 2002) reporting similar (magnitude & direction) correlations among these variables. The corre lation between the dependent variable (SWL) with self-esteem, depression, and locus of control are as follows, self-esteem with SWL (r = .56, p < .05), depression with SWL ( r = .60, p < .05) and locus of control with SWL ( r = .12, p > .05). This finding supports prior re search (e.g., Palmer, et al., 2002) reporting similar correlations among theses variables. An evaluation of the above data suggests first, that in general the predictor variables self-esteem and depression, dem onstrate a low moderate correlation ( r = .56 and r = .60 respectively) with the dependent variable (SWL) both significant at p < .05. Locus of control sugge sted a small non-significant correlation ( r = .12, p > .05) with SWL. Second, EI as measured with the MSCEIT demonstrated a small nonsignificant correlation with the dependent va riable (SWL). Correlations between EI and EI components with SWL ranged between r = .01 and r = .06 p > .05. Correlations between the known pr edictor variables range between r = .21 and

PAGE 158

149 r = .57, thus, the predicto r variables demonstrate relati vely low correlations with each other. Therefore, the magnitude of intercorrelation among predictor variables suggests in the present study multiple regression analysis is an appropriate method to investigate relationships among these variables. The data were checked for violations of the following important assumptions of multiple regression a) measurement without error (checked with Chronbachs coefficient alpha), b) independence of erro rs, c) linearity of relationship between predictor and dependent variables, d) equality of or constant variance of errors, and e) normality of residuals, with no critical violations of important assumptions discovered. A review of the relevant literature suggest ed the following pr edictors selfesteem, depression, and locus of contro l be included in the first regression analysis with satisfaction with life enter ed as the dependent variable (stem equation). This regression analysis suggested t he linear combination of self-esteem, depression, and locus of control accounts for approximately 44% of the variance in satisfaction with life. The significant pr edictors in the first regression analysis are self-esteem and depression. The five specific resear ch questions ask how much if any additional variance in satisfaction with life does emoti onal intelligence or any one or more components of emotional intelligence a ccount for among college students over self-esteem, depression, and locus of contro l. In order to investigate the above five research questions a series of five regression analysis were conducted. I

PAGE 159

150 added (individually) emotional intelligence total score as well as each EI component score to the stem equation (SWL = Self-est eem + depression + locus of control) and observed any significant change in R (total amount of variance accounted for in the dependent variable). The addition of emotional intelligence total sco re as well as each of the four EI component scores failed to dem onstrate a significant change in R and any small change was not significant at p < .05. Thus, I can not reject the null hypothesis for any of the five research questions. Therefore, I can conclude that emotional intelligence as measured with the MSCEIT does not account for additional variance in satisfaction wit h life among community college students over self-esteem, depression, and locus of control.

PAGE 160

151 Chapter Five Discussion This chapt er begins with an overview of the study and then discusses major findings within the context of prev ious research. Some suggestions for future research as well as limitations of the present study are identified. Conclusions as well as implications fo r practice in higher education are discussed in the final sections. Overview of the Study Since Dani el Goleman (1995) published Emotional Intelligence the construct has become linked with academic and occupational success as well as satisfaction with life. Mayer and Cobb (2000) noted that Education policy experts quickly accepted the idea that EI predicted academic as well as other types of success (p. 170). For exam ple, Pool (1997) reviewed Golemans 1995 publication, and stated that Emotional well-being (skills) is the strongest predictor of achievement in school and on the job and t hat Recent studies have shown that EI predicts about 80 percent of a persons success in life (p.12). Schools have been especially receptive to the EI construct. O Connor and Little (2003) argue The widespread societal acceptance of the EI concept has led some authors (e.g., Gottman & Decl aire, 1998; Shapiro, 1997) to suggest strategies for developing and enhancing EI in our schools (p.189). Elksnin and Elksnin (2003) noted that Within two years after publication of Golemans (1995)

PAGE 161

152 book more than 700 school districts ac ross the nation implemented social emotional learning (SEL) programs designe d to teach students social-emotional skills (p. 65). Barefoot and Fidler (1996) asserted that the goals of freshman seminar programs nationally emphasize the development of emotional skills. Other researchers (e.g., Gardner & Jewler, 2003; Nelson & Low, 2002) noted that the goals of freshman seminar pr ograms often include the development of emotional intelligence. The problem is that much of this specul ation regarding relationships between EI and important life domains has fa r exceeded the empirical research. Cobb and Mayer (2000) stated that To date there has been relatively little research to suggest the relationship betwe en EI and educational, occupational as well as other life domains (p. 397). The present study empirically investigated the relationship between EI and satisfaction with life among community college students. Satisfacti on with life was chosen as the depende nt variable in the present study for the following four reasons: First, some research (e.g., Argyle, 1987) suggest that increasing levels of satisfac tion with life are associated with increasing levels of positive affect and positive affect is a quality rewarding in it self. Second, some research (e.g., Meyers, 1992) reports that high levels of satisfaction with life are associated with ot her important and much desired characteristics (e.g., higher self-esteem; greater sense of control; less stress). Third, some researchers (e.g., Witter, Okun, St ock, & Haring, 1984; Veenhoven, 1994)

PAGE 162

153 report small but significant positive correla tions between satisfaction with life and levels of education. Fourth, some res earchers (e.g., Astin, 1977, 1993; Sanders & Chan, 1996) regard satisf action with life as a key goal and outcome of higher education. Bean and Bradley (1986) reported findings that suggest a small positive relationship ( r = .21, p < .001) between satisfaction with life and academic achievement among undergraduates. K oeske and Koeske (1991) reported a moderate positive relationship between satisfaction with life and retention among undergraduates. Thus, satisfaction wit h life is both an important variable for its affective association as well as it s association with other important life outcomes including those of higher educat ion (e.g., levels of education). Fortunately, there is a rich research base on satisfaction with life. Pavot and Diener (1993) define satisfaction with life as A cognitive judgmental process in which individuals assess the overall quality of their liv es on the basis of their own unique set of criteria (p. 64). Some of the research (e.g., Hong & Giannakopoulos, 1994) suggests that among the best predictors of satisfaction with life are self-esteem, depression, and locus of control respectively. The present study empirically investigates the relationship between EI and satisfaction with life after controlling for self-esteem, depression, and locus of control. Emotional intelligence was conceptualized according to the Mayer and Salovey (1997) four component cognitive ability model. This model conceptualizes EI as composed of four distinct yet related cognitive abilitie s: a) the ability to perceive, appraise, and express emot ions, b) the ability to access and

PAGE 163

154 generate emotions in order to facilitate thought, c) the ability to understand emotion and emotional knowledg e, d) the ability to regulate emotions in both self and others in order to pr omote emotional and inte llectual growth. The present study was conducted on the Lecanto campus (Citrus County) of Central Florida Community College (C FCC) during the Fall, 2005 semester. The method of sampling was convenience accomplished by the primary investigator, an adjunct psychology instructor on the campus asking fellow instructors for permission to solicit participants from among their students. During a three week span of time a total of 200 student participants completed the following five assessment instruments: a) the Mayer Sa lovey and Caruso Emotional Intelligence Test (MSCEIT), b) Rosenbergs Se lf-Esteem Scale (RSES), c) Becks Depression Inventory-II (BDI-II), d) Rotte rs Internal-External Locus of Control Scale (I-E Scale), and e) Dieners Sa tisfaction With Life Scale (SWLS). All participant s completed the assessment package individually during or after class in small groups. I administer ed all the assessments, monitored all sessions and at time of completion I ev aluated all instruments for compliance with instructions. The MSCEIT was sco red by the publisher (Multi-Health Systems Inc.). The remaining assessment s were scored and tabulated by the primary investigator. To investigat e the relative importance of EI as a predictor of satisfaction with life among community college students, a series of hierarchical regression analyses was conducted. Three known predi ctors self-esteem, depression, and

PAGE 164

155 locus of control were entered into the primary regression analyses with satisfaction with life entered as the dependent variab le (stem regression analyses). EI total score as well as each of the m odels four component scores were then added individually and s equentially to the st em regression analyses. As each variable was added to the stem e quation any resulting change in R (total variance in the dependent variable accounted for) was observed. Major Findings and Comparisons with Previous Research The first major finding in the present study is t hat the bivariant relationship between the known predictors self-esteem, depression, and locus of control with satisfaction with life supports much of the prior research. Several researchers (e.g., Diener, 1984; Huebner, 1991; Ramanaiah, Detwiler & Byravan, 1997; Hong & Giannakopoulos, 1994; Kopp & Ruzicka, 1993) reported findings that suggest significant correlations between sati sfaction with life and self-esteem, depression, and locus of cont rol. In the present study the reported bivariant correlations (presented in Table 2) between satisfaction with life and the following predictor variables are self-esteem r = .56, depression r = -.60 and locus of control r = .12. Thus, in the present st udy the predictor variables selfesteem and depression demonstrated statistically significant correlations in the strength and direction suggested by prio r research. The correlation between locus of control and satisfaction with life was small, negative and not significant at p < .05.

PAGE 165

156 Some of t he research (e.g., Hong & Giannak opoulos, 1994) suggested the relationship between self-esteem and SWL is both moderate and positive. The observed correlation was both moderate and positive ( r = .56, p < .05). The same study suggested a moderat e but negative correlation between depression and SWL. The observed correlation between depression and SWL was moderate and negative ( r = -.60, p < .05). Hong and Giannakopoulos (1994) reported a small negative correlation between satisfacti on with life and locus of control. The observed non-significant correlation betwe en locus of control and satisfaction with life is both small and negative ( r = -.12, p > .05). The second majo r finding in the present study is that EI total as well as all four EI components demonstr ated a small, but non-significant correlation with SWL. Several researchers (e.g.,Bar-on, 1997; Ciarrochi, Chan & Caputi, 2000; Martinez-Pons 1997, 1999; Mayer, Caruso, & Salovey, 2000; Law et al., 2004; Cannon & Ranzijn, 2005) reported finding a positive correlation between EI and satisfaction with life. However, in the present study the results of all simple bivariant correlations between EI total as well as all four EI components with satisfaction with life does not support t he above cited findings. None of the correlations between EI total or any of the four EI compo nents and SWL were statistically significance ( p < .05). The instruments us ed are the best available and most widely used. This is an important finding because it suggests little or no correlation between EI conceptualized as a cognitive ability, measured with the MSCEIT, and satisfaction with life among community college students.

PAGE 166

157 The third major finding was that when EI total score was added to the multiple regression SWL = .34 (SE) .42 (Dep) + .07 (L of C) -.01 (EIt) there was little or no change in R ( R = -.0001). It is important to note that even this very small change in R is not significant at the p < .05 level. Thus, in regards to the first research question Does EI concept ualized as a cognitive ability and measured by the MSCEIT account for greater variance in life satisfaction among community college student than self-esteem, depression, and locus of control? I can not reject the null hypotheses and must conclude that EI does not account for additional variance in satisfaction wit h life above self-esteem, depression, and locus of control. Thus, the findings in the present study do not support prior research (e.g., Ciarrochi, Chan & Caputi, 2000; Mayer, Caruso, & Salovey, 1999; Palmer, Donaldson, & Stough, 2002; Saklof ske Austin, & Minski, 2003) that reported EI demonstrates in cremental prediction in satisfaction with life above self-esteem, depression and locus of control. The fourth ma jor finding is all four components of EI (perceiving, facilitating thought, understanding, and regulating emot ions) when added individually and sequentially to the stem regression equation demonstrat ed little or no change in R It is important to note that none of the R associated with the components of EI was significant at the p < .05 level. Thus, in regards to research questions 2 through 5, I can not reject the null hypothes is. Therefore, I conclude that none of the components of the Mayer and Salovey (1997) model of EI accounts for vari

PAGE 167

158 ance in satisfaction with life above self-e steem, depression, and locus of control among community college students. Summary of Findings The predict or variables depression, self-esteem, and locus of control correlated with satisfaction with life. This finding agrees with prior research (e.g., Hong & Giannakopoulos, 1994) that reported a moderate negative (inverse) relationship between depression and SWL, a moderate positive correlation between self-esteem and SWL and a smalle r negative correlation between locus of control and SWL. Emotional in telligence total score as well as all four com ponents of the Mayer and Salovey (1997) EI model demonst rated a small correlation with SWL. However, none of the correlations between EI or the components of EI with SWL are significant at p < .05. In order to investigate EI incremental validit y five sequential hierarchical regression analyses were conducted. EI total score and each EI component score was added individually and sequentially to the stem equation composed of three known predictors of SWL. The resu lt of each regression analyses was a change in R < .01. Therefore, EI total score as well as all four components of the Mayer and Salovey (1997) EI model acc ounted for little or no additional variance in SWL over self-esteem, depression, and locus of control and none of the R are significant at the p < .05 level. In regards to all five research questions the null hypothesis can not be rejected. Thus, results suggest that neither EI nor

PAGE 168

159 the four components of EI accounts for additi onal variance in satisfaction with life among community college students above se lf-esteem, depression, and locus of control. Suggestions for Future Research The concept of EI has evolved along two related yet distinct paths. The first path, the more popularly oriented (mixed m odel) is based largely on Golemans (1995) book. Goleman conceptualizes EI as incorporating both cognitive abilities as well as non-cognitive elements. The second path, the more academically oriented cognitive ability m odel is led primarily by J ohn Mayer, Peter Salovey, and associates (e.g., Mayer, Salovey, & Caruso, 2002). This model conceptualizes EI as distinct yet somewhat similar to traditional intelligence. At the current stage of EI construct an d measurement development future research should address three important issues. First, in crease definitional clarity and consensus, there is little agreement on what is emotional intelligence. Second, improve measurement tools such that research informs conceptual development. Third, generate a research ba se sufficient to evaluate whether EI has incremental validity. Unless EI demonstr ates it can account for variance in some important variable beyond variance ac counted for by known predictors it is simply old wine in a new bottle. Proponents of EI such as Bar-On (2000) argue that EI is a conceptually coherent construct (p. 364). Ciarrochi, C han, and Caputi (2000) note that While the definitions of EI are often varied fo r different researchers they nevertheless

PAGE 169

160 tend to be complementary rather than c ontradictory (p. 540). Law, Wong, and Song (2004) state that A lthough definitions of emoti onal intelligence are not identical the differences between definiti ons tend to be minor (p. 484). A review of the EI literature suggests otherwise (e .g., Goleman, 1995; Mayer & Salovey, 1997; Bar-On, 1997). Matthews, Roberts, and Zeidner (2004) argue that The label emotional intellig ence has been rather haphazardly used to refer to a multitude of distinct constr ucts that may or may not be interrelated (p. 8). Studies that employ competing measures of EI may help determine whether differences between competing models of EI are really complimentary or contradictory. Clearly, the results of the present study s uggest EI as measured by the MSCEIT does not predict life satisf action among community college students. The scope of this study did not include a mixed model measure of EI. However, the results from a competing measure c ould be important to EI conceptual refinement and understanding whic h EI predicts which variables at what level. Palmer et al. (2002) noted that 10 years of theoretical and instrument development since Goleman (1995) published Emotional Intelligence now makes it possible to empirically investigate the relationship between EI and theoretically related life criteria. However, EI constr uct and measurement development is still in its early stages. The most appropriate me thod of measuring EI continues to be an area of controversy. Sakl oske, Austin, and Minski (2003) said It is not clear how, if at all, the two approaches to the measurement of EI should be reconciled (p. 708). At the current stage of EI c onstruct and measurement development

PAGE 170

161 studies that employ different concept ualizations and measurements appear to generate confusion. However, Spector and Johnson (2006) argue that There will eventually be a shakeout in terms of wh ich components and definitions become established in the research community and which are set aside (p. 340). A review of the literature revealed only one study by OConnor and Little (2003) employed both a self-r eport measure of EI, the Ba r-on Emotional Quotient Inventory (EQ-i) and an ability-based ( performance) measure, the Mayer, Salovey, Caruso, Emotional intelligence test (MSCEIT), to investigate the relationship between EI and academic ac hievement or grade point average (GPA) among college students. The results of the O Connor and Little (2003) study suggest EI measured wit h the EQ-i or MSCEIT is not a good predictor of college GPA. Clearly, the O Connor and Little study support an earlier study by Newsome et al. (2000) inve stigating the relationship between EI and GPA among (N=180) undergraduate students. Newsome et al. (2000) reported that EI as measured by the EQ-i was not an im portant predictor of college GPA ( r =.01, p >.05). Futu re research should include measures of EI from both the mixed model and the cognitive ability model. Such a study would employ both self-report and ability (performance) measures of EI. Some research (e.g., Petrides & Furnham, 2000) suggest that mixed models (self r eport measures) and cognitive ability models (performance measures ) are distinct from eac h other. Studies that employ measures from both conceptual models may suggest relationships

PAGE 171

162 between components or dimensions of EI and important life domains as well as suggest conceptual revisions of the EI construct and help refine measurement. Spector and Johnson (2006 ) argue Equally impor tant will be a demonstration of incremental validity over existi ng constructs in order to demonstrate that EI is something unique (p. 338) Gibbs (1995) and Goleman (1995) have made grandiose claims (e.g., EI is twice as important as IQ, and EI is the best predictor of success in life) regarding the relationship between EI and important life outcomes. Mayer, Salovey and Caru so, 2004 stated that Such claims suggest that EI predicts major life outcomes at levels virtually unheard of in psychological science (p. 206). Fu ture research should investigate the relationship between EI and a variety of important life do mains. For example, future studies should include dependent measures of a) intellectual (e.g. GRE scores), b) behavioral (e.g., risk taking), and c) emotional (e.g., depression) life outcomes. Large and comprehensive studies employin g competing models and assessment measures may make it possible to empirica lly investigate what EI actually does predict and at what level. Limitations of the Study The findings presented should be interpreted with caution due to threats to both internal and external validity. Gay and Ai rasian (2003) stated internal validity is The condition that observed differences on the dependent variable are a direct result of the independent variable, not some other variable (p. 345). This study is correlational research and correlation does not imply causation. Johnson and

PAGE 172

163 Christensen (2000) define external validity as The extent to which the results of a study can be generalized to and across populations, settings, and times (p. 200). Onwuegbuzie (2003) noted Findings from every study in the field of education have threats to internal and exte rnal validity (p. 72), and pointed out the importance of discussing threats to both internal and exte rnal validity. First, it allows the reader to place the findings in co ntext. Second, it provides direction for future research (e.g., replication studies that are designed to minimize identified threats to internal and external validity). Threats to Internal Validity An important th reat to internal validity at t he data collection stage of many studies is instrumentation. Onwuegbuzie (2003) proposed that Instrumentation threat to internal validity occurs w hen scores yielded from a measure lack the appropriate level of consistency (e.g., lo w reliability) and/or validity (p. 76). Instrumentation threat to internal validity was not a critical threat in the study. I generated Cronbachs Coefficient Alphas for each instrument as follows MSCEIT r = .94, SWL r = .82, RSES r = .86, BDI-II r = .82, I-E Scale r = .64. In order to evaluate the reliability of each instrument I compared the obtained Cronbachs Coefficient Alphas with estimates of reliability reported in the literature. a) Mayer, Salovey, and Caru so (2002) report the MSCEIT has a full scale reliability of r = .91 (split-half reliability). Br acket and Mayer (2003) report a test-retest reliability for the full scale MSCEIT of r = .86 with a two week interval. b) Short term reliabilities for the SWLS have been consis tently reported by the

PAGE 173

164 authors Diener, et al ., (1985) to be r = .8 or greater. c) For the RSES McCarthy and Hoge (1982) report a one year test-retest coefficient r = .77 (N = 1,852). d) BDI-II the authors Beck, Steer, and Brow n (1997) report estimates of internal reliability (Cronbachs Alpha) with outpatie nts (N = 500) as well as with a nonclinical population of co llege students (N = 120) of r =.92 and r =.93 respectively. And d) The developer of the I-E Scale Rotter (1966) reported reliability estimates which ranged from r = .69 to .73 using the Split -half Spearman-Brown and KuderRichardson formulas. Alfonso and Allison (1992) reported from their study of 106 university students a coefficient alpha of r =.89 The estimates of internal reliability obtained in the study for the MSCEIT, SWLS, and RSES are equal to or higher than estimates reported in the literature. The estimates of internal reliability obtai ned in the study for t he BDI-II and the I-E scale are less than estimates reported in the literature. Howe ver, the obtained r =.86 for the RSES appears adequate and the r =.64 obtained for the I-E scale is not much less than the r = .69 to .73 range r eported by the author (Rotter, 1966). An evaluation of the obtained estimates of reliability and the re liability estimates reported in the literature sugges ts four of the five instru ments used in the present study demonstrated adequate reliability. The IE scale demonstrated a level of internal reliability ( r =.64) less than what is generally considered adequate r =.70. Therefore, Instrumentati on threat is a concern and should be considered however it does not appear to be a critical th reat to the studies internal validity.

PAGE 174

165 Threats to External Validity An important threat to the external validity of many studies at the data interpretation stage is population validity, ecological validity, and temporal validity. Onwuegbuzie (2003) stated When interpre ting findings stemming from small and/or non-random samples, researchers should be very careful not to overgeneralize their conclusions (p. 86). The study is as lar ge or larger (N = 200) then many similar studies. The method of participant selection was convenience. However, a review of the EI literature s uggests many similar studies make use of smaller convenient samples. I collect ed some limited demographic information from participants such as gender, age, and race. A comparison between the sample and population demographics suggest t hat within the limit s of the above discussion the obtained sample of 200 participants appears to be representative of Central Florida Community College students. The obtained sample is probably representat ive of most communi ty colleges in the state of Florida, and yet conceivably non-representative of some. Therefore, population validity, ecological validity and temporal validity while always a threat does not appear to pose any unusual threat to the studies external validity. Conclusions The aim of t he present study is to investigat e whether EI predicts variance in satisfaction with life among community college students beyond that explained by known predictors self-esteem, depression, and locus of control. The results of simple correlation and hierarchical multip le regression analysis suggests clearly

PAGE 175

166 and convincingly that EI as measured with the MSCEIT does not demonstrate a correlation with or an increment in the prediction of SWL above known predictors The MSCEIT is a relatively new and popular proprietar y instrument intended to measure EI as conceptualized from the Mayer and Salovey (1997) cognitive ability model. Clearly, the result s of the study suggest EI as measured by the MSCEIT may not be a useful predi ctor of satisfaction with life among community college students. It is of particular interest to note that the EI construct can be roughly divided into two competing pe rspectives. First, the more broadly defined (inclusive) mixed mode l led primarily by D. Gole man and associates, this model makes somewhat grandiose claims as to the importance of EI. Second, the cognitive ability model led primarily by J. Mayer and P. Salovey and associates that defines EI as a special type of intelligence (set of cognitive abilities) and makes relatively conservative claims as to the importance of EI. By comparis on the grandiose claims as to t he importance of EI made from the mixed model perspective makes the c ognitive ability model of EI palatable. However, results of the present study (f rom the cognitive abili ty model) suggest EI as measured by the MSCEIT may not be a useful predictor of satisfaction with life among community college students. Results of the present study coupled with other studies such as OConnor and Li ttle (2003) that report EI measured from both mixed and ability models (EQ-i, & MSCEIT) is not a good predictor of college GPA. Newsome et al. (2000) repor ted results that suggest EI concept

PAGE 176

167 tualized from the mixed m odel of EI and measured by the EQ-i is not an important predictor of college GPA. Therefore, the cognitive ab ility model of EI may be more palatable to academic researchers and empirically inclined practitioners. However, results of the present study and the above mentioned studies suggests EI is not an important predi ctor of important higher edu cation outcome variables such as satisfaction with life or college gr ade point average regardless of what EI model or type of m easurement employed. Implications for Practice in Higher Education The curricu lum is best conceptualized as a work in progress. Patrick Terenzini and Ernest Pascarella (1999) noted that American colleges and universities have a long history of calls to reform the curriculum (p. 33). However, the history of higher education is no different from histor y in general, what we call change is often little more than re kindling of the past. One such recurrent theme in higher education curriculum reform is holistic education, or at least greater attention to the affective component of education (Beck & Kosnik, 1995). In the 1920s educators were interested in character education. In the 1950s humanistic psychology helped shift educ ators interest toward affective education. Socioemotional learning (SEL) evolved out of the Character and affective education movements. In the 1990s EI helped fuel interest in socioemotional education. The impor tance of socioemotional learning (SEL) in higher education has not gone unnoticed. Pascarella and Terenzini (1991) noted Important changes that occu r during college are probably the cumulative result

PAGE 177

168 of a set of varying, but interrelate d and mutually supporting experiences sustained over an extended period of time and The individual changes as a whole, integrated person dur ing college (p. 21). Americ an colleges to varying degrees have been and continue to be co mmitted to holistic education. Golemans publishing of Emotional Intelligence in 1995 had two important effects. First, he helped popularize the EI c oncept in part because a) traditional variables such as high school GPA, high school class rank, IQ scores, and ACT/SAT scores do not account for all of the variance in college success or other important outcomes. Second, publications such as The Bell Curve published by Herrnstein and Murray (1994) suggested general intelligence was relatively fixed and differentially distributed with respect to racial and socioeconomic lines. The appeal of EI is that it promises to level the playing field, EI is said to be as important or more important than IQ, and teachable, or at least it could be learned (Goleman, 1995). Second, Gole man broadened the definit ion of EI to include a multitude of personality entities, thus pr oviding the link between EI and education. Mayer and Cobb (2000) noted that according to Golem ans conceptualization Virtually any link between personality and good school outc omes could be attributed to EI (p. 170). In the final analysis ten years after Golemans (1995) publication of Emotional Intelligence much has been gained, such as EI conceptual development and instrumentation. Ho wever, the EI construct continues to suffer from a

PAGE 178

169 lack of definitional clarit y and measurement tools are not widely accepted. A review of the literature incl uding the present study fails to suggest what the ability version of EI predicts. A lim itation of the present study is the failure to measure EI with a mixed model instru ment. However, even if Go leman is correct that groups of different variables predict im portant life outcomes, what usefulness does EI have over groups of other well k nown constructs? In order for EI to establish its validity it must demonstrate definitional clarity, accuracy, and reliability of measurement. In order for EI to establish its ut ility it must demonstrate it accounts for variance in impor tant criteria beyo nd other important predictors. Results fr om the present study do not suppor t claims of EI definitional clarity or accuracy of measurement.Results also suggest EI is not an important predictor (does not account for additional variance) of satisfaction with life among community college students. The law of par simony dictates that the simplest of two or more competing theories or explanations is preferable and that an explanation for unknown phenomena should first be attempted in terms of what is already known. Higher education has a rich literature as well as access to related literature such as personality research to inform both curricu lum development and best practices in education. Mayer and Cobb (2000) argue A t present socioemotional programs are implemented with reasonabl e hopes that they will have beneficial effects, independent of empirical research concerni ng EI (p. 179). The state of California experienced a similar situation in the early 1990s when well meaning educational

PAGE 179

170 policy makers incorporated self-esteem programs into their elementary and secondary school curricula with little empirica l justification. Several years later the California self-esteem movem ent in general was judged a failure. The future of the EI construct will take one of two paths. First, with continuing research the EI construct may gain credibility with incr easing definitional clarity and improved measur ement tools. Spector and Johnson (2006) may be correct that There will eventually be a shakeout in terms of which components and definitions become established in rese arch and which are set aside(p. 340). With greater definitional clarity and better measurement tools we may discover EI has incremental validity over existing c onstructs demonstrating EI is something unique (p. 338). In time we may understand what EI predicts and at what levels. Another possib ility is that the EI construct along with its often inflated claims such as EI is equal to if not more valuable than IQ as an indicator of ones professional and life success (Goleman, 1995 p. 34) will be debunked. The exaggerated claims some EI proponents have made to the importance of EI (e.g., job performance & leadership) has helped generate considerable research. However, despite the popularit y of the construct and volume of research EI remains in an early stage of construct dev elopment. The jury is still out on EI, researchers may someday find EI has some measure of usefulness, or researchers may find it is not an educ ationally meaningful significant construct. However, until such time educational policy makers sh ould recall the California self-esteem

PAGE 180

171 movement, and choose to be informed by higher education and personality research rather than good intentions or mass media science journalism.

PAGE 181

172 References Abraham, R. (2000). The role of contro ls as a moderator of emotional dissonance and emot ional intelligence outcomes relations. Journal of Psychology, 134, 169-185. Alfonso, V. C., & Allison, D. B. (1992). Sexual fantasy and satisfaction: A multidimensiona l analysis of gender differences. Journal of Psychology and Human Sexuality, 5 (3), 19-37. Alfonso, V. C., Allison, D. B., & Rader, D. E. (1996). The extended satisfaction with life scale: De velopment and psychometric properties. Social Indicator Research, 38 (3) 275-301. Allman, A., & Diener, E. (1990). Measurem ent issues and the subjective wellbeing of people with disabilities. Personality and Individual Differences, 18, 443-451. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders ( 4th ed.). American Psychiatric Association. Washi ngton, DC: American Psychological Association. Anastasi, A. (1988). Psychological testing (6th ed.). New York: Macmillan Publishing.

PAGE 182

173 Andrews, F. M., & Robinson, J. P. (1991). Measures of subjective well-being. In J. P. Robi nson, & P. R. Shaver (Eds.), Measurement of personality and social psychological attitudes (pp. 61-114). San Diego: Academic Press, Inc. Andrews, F. M., & Wi they, S. B. (1976). Social indicators of well-being. New York: Plenum Press. Argyle, M. (1987). The psychology of happiness. New York: Methuen. Argyle, M. (1999). Causes and correlates of happiness. In D. Kahneman, E. Diener & N. Schwarz (Eds.), Well being: The foundations of hedonic psychology (pp. 353-373). New York: Russell Sage Foundation. Arrindell, W. A., Meeuwesen, L., & Huyse, F. J. (1991). The Satisfaction With Life Scale (SWL S): Psychometric properties in a non-psychiatric medical outpatients sample. Personality and Individual Differences, 12, 117-123. Asforth, B. E., & Humphrey, R. H. (1995). Emotion in the workplace: A reappraisal. Human Relations, 48, 97-125. Ashkanasy, N. M, & Dasborough, M. T. (2003). Emotional awareness and emotional inte lligence in leadership teaching. Journal of Education for Business, 79 (1), 18-22. Ashkanasy, N. M., & Daus, S. D. (2002). Emotion in the workplace: The new challenge for managers. Academy of Management Executive, 16 (1), 76-86.

PAGE 183

174 Astin, A. W. (1993). What matters in college? F our critical years revised San Francisco: Jossey-Bass. Astin, A. W., & Henson, J. W. (1977) New measures of college selectivity. Research in Higher Education, 6 (1), 1-9. Atwater, L. E., & Yammarino, F. J. (1993). Personal attri butes as predictors of superior s and subordinates perceptions of military leadership. Human Relations, 46 645-668. Austin, J., & Huberty, T. J. (1993). Deve lopment of the Child Attitude Toward Il lness Scale (CATIS). Journal of Pediatric Psychology, 18 (4), 467-480. Bachman, J. G., & OMalley, P. M. (1977). Self-esteem in young men: A longitudinal analysis of the impact of educational and occupational attainment. Journal of Personality and Social Psychology, 35 365-380. Bandura, A. (1989). Self-regulation of motivation and action through internal standards and goal syst ems. In L. A. Pervin (Ed.), Goal concepts in personal ity and social psychology (pp.19-85). Hillsdale, NJ: Lawrence Erlbaum Associates. Barchard, K. A. (2003). Does emotional inte lligence assist in the prediction of academic success? Educational and Psycholog ical Measurement, 63, 840-858. Barefoot, B., & Fidler, P. (1996). The national survey of freshmen seminar programs: Continuing innovations in the collegiate curriculum (Monograph No. 20). SC: Un iversity of South Carolina, National Resource Center for

PAGE 184

175 Freshmen Y ear Experience and Students in Tr ansition. (ERIC Document Reproducti on Service No. ED 393 386). Bar-On, R. (1997). Emotional quotient inventory: Technical manual Toronto, Canda: Mu lti-Health Systems, Inc. Bar-on, R. (2000). Emotional and social in telligence: Insight fr om the Emotional Quotient Inventory. In R. Bar-On, & James, D. A. (Eds.), Handbook of emotional intelligence: Theory, development, assessment, and application at home school, and in the workplace. (pp. 363). San Francisco: Jossey-Bass. Baumeister, R. F. (1991). On the stability of variability: Retest reliability of metatraits. Personality and Social Psychology Bulletin, 17 (6), 633-639. Baumeister, R. F. (1993). Self-esteem: The puzzle of low self-regard. New York: Plenum Press. Baumeister, R. F., Tice, D. M., & Hutt on, D. G. (1989). Self-presentational motivations and per sonality differences in self-esteem. Journal of Personality, 57, 547-579. Bean, J. P., & Bradley, R. K. (1986). Untangling the satisfaction-performance relati onship for college students. Journal of Higher Education, 57 (4), 393-412. Bearsley, C., & Cummins, R. A. (1999). No place called home: Life quality and pur pose of homeless youths. Journal of Social Distress and the Homeless, 8, 207-226.

PAGE 185

176 Beck, A.T., Guth, D., Steer R. A., & Ball, R. (1997). Screening for major depression di sorders in medical inpatient s with the Beck Depression Inv entory for primary care. Behavioral Research and Therapy, 35 (8), 785-791. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventor y for measuring depression: Archive of General Psychiatry, 4, 546-571. Beck, A.T., & Steer, R. A. (1984). Internal consistencie s of the original and revised Beck Depression Inventory. Journal of Clinical Psychology, 40, 1365-1367. Beck, A. T., Steer, R. A ., & Brown, G. K. (1993). The Beck Depression Inventory, (2nd ed). San Antonio: Psychol ogical Corporation. Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties of the Beck D epression Inventory: Twenty five years of evaluation. Clinical Psychology Review, 8, 77-100. Beck, C., & Kosnik, C. (1995). Caring for t he emotions: Toward a more balanced schooling. In A. Neiman (Ed.), Philosophy of Education and Society (pp.161-169) Urbana, IL: Free Press. Bednar, R., Wells, G., & Peterson, S. (1989). Self-esteem: Paradoxes and innovations in clinical theory and practice. Washington, DC: American Psyc hological Association.

PAGE 186

177 Binet, A., & Simon, T. (1916). The development of intelligence in children. Baltim ore: William & Wilson. (Original work published 1905). Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P. Robinson, P. R. Shav er, & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes. San Diego: Academic Press. Blau, G. J. (1984). Brief note comparing the Rotter and Levenson measures of locus of control. Perceptual and Motor Skills, 58 (1), 173-174. Block, J. (1995). A contrarian view of the five-factor approach to personality description. Psychological Bulletin, 117, 187-215. Blumenthal, S. J., Matthews, K., & Weiss, S. (1994). New research frontiers in behavioral medicine: Proceedings of the national conference Washington, DC: U. S. Government Printing Office Boyatzis, R. E., Goleman, D., & Rhee, K. S. (1998). Clustering competence in emotional inte lligence: Insights from the Emot ional Competence Inventory. In Bar-On, R., Parker, D, A.; The handbook of emotional intelligence: Theory, Deve lopment, assessment, and application at home, school, and in the workplace. (pp. 343-362). San Francisco: Jossey-Bass. Boyer, E. L. (1987). College: The undergraduate ex perience in America. New Yo rk: Harper & Row. Braburn, N. M., & C oplovitz, D. (1965). Reports of happiness. Chicago: Aldine.

PAGE 187

178 Brackett, M. A., & Mayer, J. D. (2003). Convergent, discriminant, and incremental vali dity of competing measures of emotional intelligence. Personality and Social Psychology Bulletin, 29( 99), 1147-1158. Bracket, M. A., Mayer, J. D. & Warner, R. M. (2003). Em otional intelligence and the predi ction of everyday behavior. Personality and Individual Differences 36 1387-1402. Branden, N. (1969). The psychology of self-esteem. New York: Bantam. Brower, A. M. (1990). Student perceptions of life-task demands as a mediator in the freshman year experience. Journal of the Freshman Year Experience, 2 (2), 7-30. Brown, R. D. (1972). Student development in tomorrows higher education: A return to the academy. Washington, DC: Amer ican College Personnel Association. Brown, J., & Dutton, K. A. (1995). The thrill of victory, the complexity of defeat: Self-esteem and peoples em otional reactions to success and failure. Journal of Personality and Social Psychology, 68( 4), 712-722. Brown, R. F., & Schutte, N. S. (2005). Direct and indirect relationships between emotional in telligence and subjective fatigue in university students. Personality and Individual Differences 38(6), 1353-1364. Bryant, F. B., & Veroff, J. (1982). The st ructure of psychological well-being: A soci o-historical analysis. Journal of Personality and Social Psychology, 43 (4), 653-673.

PAGE 188

179 Bumberry, W., Oliver, J. M. & McClure, J. N. (1978). Validation of the Beck Depression In ventory in a university population using psychiatric estimate as the criterion. Journal of Consultation Clinical Psychology, 46, 150-155. Campbell, D. T., Converse, P. E., & Rogers, W. L. (1976). The quality of American life: Pe rception, evaluations and satisfaction. New York: Russell Sage Foundation. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-m ultimethod matrix. Psychological Bulletin, 56 81-105. Carroll, E. L. (1993). Human cognitive abilities: A su rvey of factor-analytic studies. New York: Cambridge University Press. Cherlin, A., & Bourque, L. B. (1974) Dimensionality and reliability of the Rotter I-E Scale. Sociometry, 37(4), 565-582. Cherniss, C., & Goleman, D. (2001). The emotionally intelligent workplace. San Francisco: Jossey-Bass. Chickering, A., & Reisser, L. (1993). Education and identity (2nd ed .). San Franc isco: Jossey-Bass. Chubb, N. H., Fertman, C. T., & Ross, J. L. (1997). Adolescent self-esteem and locus of control: A longitudinal study of gender and age differences. Adolescent, 32, 113-129. Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation of the emot ional intelligence construct. Personality and Individual Differences, 28 539-561.

PAGE 189

180 Ciarrochi, J., Forgas, J., & Mayer, J. (2001). Emotional intelligence in everyday life: A scientific inquiry. NY: Psychology Press. Clark, A. E., & Oswald, A. J. (1994). Unhappiness and unemployment. Economic Journal, 104, 648-659. Cobb, C. D., & Mayer, J. D. (2000). Emotional intelligence: What the research says. Educational Leadership, 58 (3), 14-18. Cohen, J. (1988). Statistical power analysis for behavioral sciences. (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Comfrey, A. L., & Lee, H. B. (1992). A first course in factor analysis Hillsdale, NJ: La wrence Erlbaum Associates. Cook, J. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis i ssues for field settings. Chicago: Rand McNally. Coopersmith, S. (1967). The antecedents of self esteem. San Francisco: Freeman and Company. Costa, P., & McCrae, R. (1980). Influence of extrav ersion and neuroticism on subjective well-being: Happy and unhappy people Journal of Personality & So cial Psychology, 38 (4), 668-678. Craig, T. J., & Van Natta, P. A. (1 983). Disability and depressive symptoms in two communities. American Journal of Psychiatry, 140(5 ), 598-601. Crandall, R. (1973). The m easurement of self-esteem and related concepts. In J. P. R obinson & P. R. Shaver (Eds.), Measures of social psychological attitudes (pp. 45-167). Ann Arbor: Univ ersity of Michigan Press.

PAGE 190

181 Cronbach, L., & Meehl, P. ( 1955). Construct validity in psychological tests. Psyc hological Bulletin, 52 281-302. Cross, T. P., McDonald, E., & Lyons, H. (1997). Evaluating the outcome of childrens mental services: A guide for the use of available child and family outcomes measures (3rd ed.) Boston: Judge Baker Childrens Center. Davies, M., Stankov, L., & Roberts, R. (1998). Emotional intelligence: In search of an elusive construct. Journal of Personality and Social Psychology, 75, 989-1015. Dawda, D., & Hart, S. D. (2000). Assessing emotional intelligence: Reliability and valid ity of the Bar-On Emotional Quotient Inventory (EQ-i) in university students. Personality and Individual Differences, 28 797-812. Dawes, R. M. (2001). Incr emental validity of the ego impairment index: Its fi ne when its there. Psychological Assessment, 13, 408-409. Day, A., & Carroll, S. A. (2004). Usi ng ability based measur e of emotional intelligence to predict i ndividual performance, group performance, and group citizenship behaviors. Personality and Individual Differences, 36 (96), 1443-1458. Demellow, L. R., & Imms, T. (1999). Self -esteem, locus of control, and coping styles and their relationships to school a ttitudes of adolescents. Psychological Studies, 44 24-34. Dew, T., & Huebner, E.S. (1994) Adolescents perceived quality of life: An expl oratory investigation Journal of School Psychology, 32 (2), 185-199.

PAGE 191

182 Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95 (3) 542-575. Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators Research, 31 (4), 103-157. Diener, E. (2000). Subjective wellbeing: The science of happiness and a proposal for a national index. American Psychologist, 55 (1), 34-43. Diener, E., & Allman, A. ( 1991). The psychic cost of in tense positive affect. Journal of Personality and Social Psychology, 61 (3), 492-503. Diener, E., Emmons, R. A., Lar sen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49 (1), 71-75. Diener, E., & Larsen, R. J. (1984). Temporal stabilit y and cross-situational consistency of affect ive, behavioral, and cognitive responses. Journal of Personalit y and Social Psychology, 47, 580-592. Diener, E., & Larsen, R. J. (1993). The subjective experience of emotional well-being. In M. Lewis., & J. M. Haviland (Eds.), Handbook of emotions (pp. 405-415) New York: Guilford Press. Diener, E., Sandvik, E., Seidlitz, L., & Diener, M. (1993). The relationship between income and subjec tive well-being: Relative or absolute? Social Indicators Research, 28 205-215. Diener, E., & Suh, E. M. (1997). Measuring quality of life: Economic, social, and s ubjective indicators. Social Indicators Research, 40, 189-216. Diener, E., Suh, E., Oishi, S., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychology Bulletin, 125, 276-302.

PAGE 192

183 Dugan, W., Mc Donald., Passik, S., Rosenf eld, B., Theobald, D., & Edgerton, S. (1998). Use of the Zung Self-Rating Depression Scale in cancer patients: Feasib ility as a screening tool. Psycho-Oncology, 7 483-493. Duke, M. P., & Nowicki, S. (1974). Locus of control and achievements the confirmation of a theoretical expectation. The Journal of Psychology, 87, 263-267. Elias, M. J., Zins, J. E., Weissberg, K. S., Frey, A., Greenberg, M. T., Haynes, N. M., Kessler, R. Schwab-Stone., & Schriver, R. (1997). Promoting social and emotional l earning: Guidelines for educators Alexandria, VA: Asso ciation for Supervision and Curriculum Development. Elksnin, L. K., & Elksnin, N. (2003). Fostering social-e motional learning in the classroom. Education, 124( 63), 64-75. Emmons, R. A., & Diener, E. (1985). Personal ity correlates of subjective wellbeing. Personality and Social Psychology Bulletin, 11 89-97. Epstein, S. (1979). The ecological study of emotions in humans. In K. Blankstein (Eds.), Advances in the study of communications and affect (pp. 47-83). New York: Plenum. Epstein, S. (1985). The implications of cognitive-experiential self-theory for research in social psychology and personality. Journal for the Theory of Social Behavior, 15, 283-309

PAGE 193

184 Evans, R. L., Kleinman, L., Halar, E. M. & Herzer, K. (1984). Predicting change in life satisfac tion as a function of group counseling. Psychological Reports, 55, 199-204. Evans, G. W., Shapiro, D. H., & Lewis, M. A. (1993). Specifying dysfunctional mismatches between different control dimensions. British Journal of Psychology, 84, 255-273. Extrema, N., & Fernadez-Berrocal, P. (2005). Perceived emot ional intelligence and life satisfaction: Predict ive and incremental validity using the Trait Meta-Mood Scale. Personality and Individual Differences, 39, 937-948. Fendrich, M., Weissman, N. M., & Warner, V. (1990). Screening for depression disorders in children and adolescent: Validating the center for epidemiologic st udies depression scale for children. American Journal of Epid emiological Studies, 131, 538-51. Figurelli, G. A., Hartman, B. W., & Kowalski Jr., F. X. (1994). Assessment of change in scores on personal control orient ation and use of drugs and alcohol of adolescents who participate in a cognitively oriented pret reatment intervention. Psychological Reports, 75 (2), 939-944. Fink, H. C., & Hjelle, L. A. (1973). Internal-external control and ideology. P sychological Reports, 33, 967-974. Franklin, R. D. (1963). Youths expectancies about internal versus external control of re inforcement related to none variables Unpublished doctoral dissertat ion, Purdue University, IN.

PAGE 194

185 Franklin, R. D., Duley S. M., Rousseau, E.W., & Sabers, D. L. (1981). Construct Va lidation of the PiersHarris Childrens Self Concept Scale. E ducational and Psychological Measurement, 41 (2), 439-443. Freshwater, D., & Stickley, T. (2004). T he heart of the art: Emotional intelligence in nurse education. Nursing Inquiry, 11 (2), 91-98. Frisch, M. B. (1991). A co mparison of formerly and chronically battered women of cogniti ve and situational dimensions. Psychotherapy: Theory, Research, Practice Training, 28 (2), 339-344. Frisch, M., Clark, M. P., Rouse, S. V., R udd, M. D., Paweleck, J., Greenstone, A., & Kopplin, D. (2005). Predictive and treatment va lidity of life satisfaction and the quality of life inventory. Assessment, 12 (1), 66-78. Fordyce, M. W. (1983). A program to increase happiness: Further studies. Journal of Counseling Psychology, 30 (4), 483-498. Foster, S. L., & Cone, J. D. (1995). Valid ity issues in clinical assessment. Psychological Assessment, 7, 248-260. Furby, L. (1979). Individualistic bias in studies of locus of control. In A. R. Buss (Eds.), Psychology in social context (pp.169-190). New York: Irvington. Furr, R. M., & Funder, D. (1998). A multimodal analysis of personal negativity. Journal of Personality and Social Psychology, 74, 1588-1591. Gaa, J. P. (1979). The effect of individ ual goal setting conferences on academic achievement and modifica tion of locus of control orientation. Psychology in the Schools, 16 591-598.

PAGE 195

186 Gannon, N., & Ranzijn, R. (2005). Does emotional intelligence predict unique variance in lif e satisfaction beyond IQ and personality? Personality and Individual Differences, 38 (6), 1353-1364. Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. Ne w York: Basic Books. Gay, L. R., & Airasian, P. (2003). Educational research : Competencies for analysis and application (7th ed.). Upper Saddle River, NJ: Pearson Education. Geher, G., Warner, R. M., & Brown, A. S. (2001). Predictive validity of the emotiona l accuracy research scale. Intelligence, 29, 273-388. Gibbons, T.C. (1986). Revisiting the question of born vs. made: Toward a theory of developm ent of transformational leaders. (Doctoral dissertation, Fielding Institute, Santa Barbara, CA, 1986) Dissertat ion Abstracts International, 48(4), 1167. Gibbs, N. (1995, October 2). The EQ Factor. Time, 146 (14), 60-68. Gilman, R., & Huebner, E. S. (2000). Re view of life satisfaction measures for adolescents. Behavior Change, 17, 178-195. Gilman. R., & Huebner, E. S., & Laughlin (2000). A first stud y of the Multidimensi onal Students Life Satisfac tion Scale with adolescents. Social Indicators Research, 52, 135-160. Gilman, R., & Huebner, E. S. (2003). A revi ew of life satisfaction research with children and adolescents. School Psychology Quarterly, 18, 192-205.

PAGE 196

187 Goldsmith, R. E. (1986). Dimensionality of the Rosenber g Self-Esteem Scale. Journal of Social Behavior and Personality, 1, 253-264. Goleman, D. (1995). Emotional intelligence: Why it can matter more than IQ. New York: Bantam Books. Goleman, D. (1998). Working with emotional intelligence. New York: Bantam Books. Goleman, D., Boyatzis, R., & Mckee, A. (2002). Primal leadership: Realizing the power of emotional intelligence Boston: Harvard Business School Press. Gomez, R. (1997). Locus of control and type A behavior pattern as predictors of copi ng styles among adolescents. Personality and Individual Differences, 23 591-598. Gomez, R. (1998). Locus of control and avoidant coping: Direct, interactional and m ediational effects on mal adjustment in adolescent. Personality and Individual Differences, 24 325-334. Gottman, J. M., & Declaire, J. (1998). Raising an emotionally intelligent child. New York: Simon & Schuster. Griffiths, R. A., Beumon t P. J., & Giannakopoulos, E. (1999). Measuring selfesteem in di eting disordered pati ents: The validation of the Rosenberg and C oppersmith contrasted. International Journal of Eating Disorders, 25, 227-231. Gurin, G., Veroff, J., & Feld, S. (1960). Americans view their mental health. Ne w York: Basic Books.

PAGE 197

188 Haber, J., Krainovich-Miller, B., McMahon, A. L., & Price-Hoski ns, P. (1997). Comprehensive psychiatric nursing. St Louis: Mosby. Haring, M. J., Stock, W. A., & Okun, M. A. (1984). A research synthesis of gender and social class as correlates of subjective well-being. Human Relations, 37, 645-657. Harter, S. (1990). Issues in the assessm ent of the self-concept in children and adolescents. In Annette La Greca, (Eds.), Through the eyes of the child: Obtaining se lf-report from ch ildren and adolescents (pp. 292-325). Needham He ights, MA: Allyn & Bacon. Haynes, S. N., & Lench, H. C. (2003). In cremental validity of new clinical assessment measures. Psychological Assessment, 15( 4), 456-466. Haynes, S. N., Nelson, K., & Blaine, D. D. (1999). Psychometric issues in assessment re search. In P. C. Kendall, J. N. Butcher, & G. N. Holmbeck (Eds.), Handbook of research methods in clinical psychology (2nd ed. pp. 125-154) New York: Wiley. Haynes, S. N., & OBrien, W. B. (2000). Behavioral assessment: A functional approach to psychological assessment. New York: Klower. Headey, B., & Wearing, A. (1989). Persona lity, life events, and subjective wellbeing: Towa rd a dynamic equilibrium model. Journal of Personality and Social Psychology 57, 731-739.

PAGE 198

189 Headey, B., & Wearing, A. (1991). Subjecti ve well-being: A stocks and flows frameworks. In F. Strack, M. Argyle, & N. Schwartz (Eds.), Subjective wellbeing: An interdisciplinary perspective (pp. 41-73). Oxford, England: Pergamon Press. Herrnstein, R., & Mu rray, C, A. (1994). The bell curve: Intelligence and class structure in American life New York: Free Press. Hickson, J., Housley, W. F., & Boyle, C. (1988). The relationship of locus of control, age, and sex to life satisfaction and death anxiety in older persons. Inter national Journal of Agi ng and Human Development, 26 (3), 191-199. Hong, S., & Giannakopoulos, E. (1994). The relationship of satisfaction with life to pe rsonality characteristics. The Journal of Psychology, 128, 547-558. Howell, J. M., & Avolio, B. J. (1993). Transformational leadership, transactional eadership, locus of control, and support for innovation: Key predictors of consolid ated business-unit performance. Journal of Applied Psychology, 78 891-902. Huebner, E. S. (1991). Correlates of life satisfaction in children. School P sychology Quarterly, 6 (2), 103-111. Huebner, E. S. (1994). Preliminary devel opment and validat ion of a multidimensional lif e satisfaction scale for children. Psychological Assessment, 6, 149-158.

PAGE 199

190 Huebner, E. S. (1996). The interrelationships of positiv e affect, negative affect and life satisf action in an adolescent sample. Social Indicators Research, 38 (2), 129-137. Huebner, E. S., & Alderman, G. C. (1993). Convergent and discriminant validation of a childrens life sa tisfaction scale: Its relationship to self-and teache r-reported psychological proble ms and school functioning. Social Indicators Research, 30 (1), 71-82. Huebner, E. S., Gilman R., & Laughlin, J. E. (1999). A multimethod investigati on of the multidimensionality of childrens well-being reports: Discriminant validi ty of life satisfaction and self-esteem. Social Indicators Research, 46 1-22. Hyer, L., Gouveia, I., Harrison, W. R., Wa rsaw, J. (1987). Depression, anxiety, paranoid reactions, hypochondr iasis, and cognitive decline of later life inpatients. Journal of Gerontology, 42, 92-94. Inglehart, R. (1990). Culture shift in advanced industrial society. Princeton, NJ: Prin ceton University Press. Jackson, M. (1984). Self-esteem and meaning: A lif e historical investigation. Albany: State Un iversity of New York Press. James, W. (1890). The principles of psychology. Cambridge, MA: Harvard University Press.

PAGE 200

191 James, W. (1957). Internal versus external control of reinforcement as a basic vari able in learning theory. Unpublished doctoral dissertation, Ohio State University. Janovics, J., & Christiansen, N. D. (2002). Emotional intelligence in the workplace Paper presented at the 16th Annual Conference of the Society of Industrial an d Organizational Psychology, San Diego. Joachim, K. (1996). The po litics of self-esteem. American Educational Research Journal, 33, 3-22. Johnson, B., & Christensen, L. (2000). Educational research: Quantitative and qualitative approaches. Boston, MA: Allyn and Bacon. Johnson B. W., Redfield, D. L., Miller, R. L., & Simpson, R. E. (1983). The Coopersmit h Self-Esteem Inventory: A construct validation study. Educational and Psychological Measurement, 43, 907-913. Joubert, C. E. (1990). Relati onship among self-esteem, psyc hological reactance, and other personality variables. Psychological Reports, 66, 1147-1151. Judd, C., Smith, E., & Kidder, L. (1991). Research methods in social relations (6th ed.) San Diego: Harcourt Brace Jovanovich. Judge, T. (1990). Job satisfaction as a reflection of disposition: Investigating the rela tionship and its effects on employee adaptive behaviors (Doctoral dissertation, University of Illinois at Ur bana-Champaign, 1990).Dissertation Abstract International, 52(3-A), 996.

PAGE 201

192 Kammann, R., & Flett, R. ( 1983). Affectometer 2: A sca le to measure current le vel of general happiness. Australian Journal of Psychology, 35, 259-265. Kaplan, S. L., Hong, G. K. & Weinhold, C. (1984). Epid iomology of depressive sympto matology in adolescents. Journal of American Academy of Child Psychiatry, 23, 91-98. Kaplan, H. I., & Pokorny, M. D. (1969). Self-derogation and psychosocial adjustment. Journal of Nervous and Mental Disease, 149 421434. Kaplan, H. I., & Sadock, B. J. (1985). Synopsis of psychiatry: Behavioral scienc e clinical psychiatry. Baltimore: Willi ams & Wilkins. Keltner, N., Schwecke, L., & Bostron, C. (1995). Psychiatric nursing. St. Louis: Mosby. Kernberg, O. (1975). Borderline conditions and pathological narcissism. New Yo rk: Jason Alexander. Kim, B. S., Omizo, M. M., & D Andrea, M. J. (1998). The effects of culturally consonant gr oup counseling on the self-esteem and internal locus of control among native American adolescents. Journal for Specialists in Group Work, 23, 145-163. Klein, H. A., Tatone, C. L. & Lindsay, N. B. (1989). Correlates of life satisfaction among military wives. Journal of Psychology: Interdisciplinary & Applied, 123 (5), 465475.

PAGE 202

193 Koeske, G. F., & Kirk, S. A. (1995). Direct and buffering effects of internal locus of control among mental health professionals. Journal of Social Science Research (20), 1-28. Koeske, G. F., & Koeske, R. D. (1991) Underestimation of social support buffering. Journal of Applied Behavioral Science, 27 (4), 475-489. Kopp, R. G., & Ruzicka, M. F. (1993). Wo mens multiple roles and psychological well-being. Psychological Reports, 72, 1351-1354. Kwan, V. S. Y., Bond, M. H. & Singelis, T. M. (1997). Pan-cultural explanations for life satisfaction: Adding relations hip harmony to self-esteem. Journal of Personalit y and Social Psychology, 73, 1038-1051. Lam, T. H., & Kirby, S. L., (2002). Is emotional intelligence an advantage? An explorati on of the impact of emotional and general intelligence on i ndividual performance. Journal of Social Psychology, 142 (1), 133-143. Lane, J. E., Shellenberger, S., Gresen, K. W., & Moore, N. C. (2000). Screening for depre ssion within a rural primary care practice by a self-rating depression scale. Primary Care Psychiatry, 6 (2), 73-76. Law, K. S., Wong, Chi-Sum., & Song, L. (2004). The construct and criterion validity of emotional intelligence and its potent ial utility for management studies. Journal of Applied Psychology, 89 (3), 483-496. Lawton, M. P. (1983). Envir onment and other determinants of well-being in older people. Gerontologist, 23 (4), 349-357.

PAGE 203

194 Lefcourt, H. M. (1982). The asse ssment of social intimacy. Journal of Personality Assessment, 46 (95), 514-518. Levitz, R., & Noel, L. (1989). Connecti ng students to institutions: Keys to retention and success. In M. L. Uperaft & J. N. Gardner (Eds.), The fr eshman year experience (pp. 65-81). San Francisco: Jossey-Ba ss. Lewinsohn, P. M., Redner, J. E., & Seel ey, J. R. (1991). The relationship between life sa tisfaction and psychosocial variab les: New perspectives. In F. Strack., M. Argyle., & N. Schwartz (Eds.), Subjective well-being: An interd isciplinary perspective (pp. 141-169). Oxford, England: Pergamon Press. Lewinsohn, P. M., Seeley, J. R., & Gotlib, I. H. (1997). Depression-related psychosocial variables: Are they specific to depression in adolescent? Journal of Abnormal Psychology, 106, 365-375. Lewinsohn, P. M., & Teri, L. (19 82). The selection of depressed and nondepressed subjects on the basis of self-report data. Journal of Consulting and Clinical Psychology, 50, 590-591. Liang, J. (1985). A structural integration of the Affect Balance Scale and the Life Satisfaction Index A. Journal of Gerontology, 40 (5), 552-561. Lopes, P. N., Brackett, M. A., Nezlek, J. B., Schutz, A., Sellin, I. & Salovey, P. (2004) Emotional intelligence and social interaction. Society for Personal ity and Social Psychology, 30 (8), 1018-1034.

PAGE 204

195 Love, P. G., & Love, A. G. (1995). Enhancing students l earning: Intellectual, social, and emotional integration. ASHE-ERIC Higher Education Report No. 4. Washington, DC. Association for the Study of Higher Education. Lucas, R. E., & Gohm, C. L. (2000). A ge and sex differences in subjective well-being across cultures. In E. Die ner & E. M. Suh (Eds.), Subjective we ll-being across culture. Cambridge, MA: University Press. Lynch, D. A. (1999). Sour ces of stress, perceived stress levels, and levels of self-esteem among career oriented mothers who are employed full-time, part-time and those who opted to stay home full-time (Doctoral dissertation, Seton Hall University, NJ, 1999) Dissertation Abstracts International, 60 (3-B) 1307. Madsen, C. K., & Goins, W. E. (2002). Internal versus external locus of control: An a nalysis of music populations. Journal of Music Therapy, 39 (4), 265-273. Magnus, K., Diener, E., Fujita, F., & Pavo t, W. (1993). Extraversion and neuroticism as predictors of objective life ev ents: A longitudinal analysis. Journal of Pers onality and Social Psychology, 65 (5), 1046-1053. Mandell, B., & Pherwani, S. (2003). Relati onship between emotional Intelligence and transformational leadership style: A gender comparison. Journal of Business and Psychology, 17 (3), 387-404.

PAGE 205

196 Mannarino, A. P., Cohen, J. A., & Berm an, S. R. (1994). The relationship between preabuse factors and psychological symptomatology in sexually abused girls. Child Negligence, 18, 63 -71. Marciano, P. L., & Kazdin, A. E. (1994). Self-esteem, depression, hopelessness, and suicidal intent among psychiatrically di sturbed inpatient children. Journal of Clinical Child Psychology, 23, 151-160. Marks, L. (1998). Deconstructing locus of c ontrol: Implications for practitioners. Journal of Counseling and Development, 76 (3), 251-261 Marks, G. N., & Flemming, N. (1999). Influences and consequences of well-being among Aust ralian young people: 1980-1995. Social Indicators Research, 46, 301-323. Marlowe, H. A. (1986). Soci al intelligence: Evidence fo r multidimensi onality and construct independence. Journal of Educational Psychology, 78, 52-58. Marsh, H., Barnes, J., & Hocevar, D. (1985). Self-other agreement on multidimensi onal self-concept ratings: Factor analysis and multitraitmultimethod analysis. Journal of Personality and Social Psychology, 49 (5), 1360-1377. Marsh, H. W., & Richards, G. E. (1987). The Rotter Locus of Control Scale: The comparis on of alternative response formats and implications. Journal of Res earch in Personality, 20 (4), 509-528.

PAGE 206

197 Martinez-Pons, M. (1997). The relation of emotional intelligence with selected areas of personal functioning. Imagination, Cogniti on, and Personality, 17 (1), 3-13. Martinez-Pons, M. (1999). Parental inducement of emotional intelligence. Imagination, Cognition, and Personality, 18, 3-23. Matthews, G., Zeidner M., & Roberts, R. D. (2002). Emotional Intelligence: Science and myth. Cambridge, MA: MIT Press. Matthews, G., Zeidner M., & Roberts, R. D. (2004). Wh at is thing called emotional intelli gence. In K. R. Murphy (Eds.), A critique of emotional intelligence (pp. 3-36 ) New Jersey, Lawrence Erlbaum Associates. Mayer, J. D. (2001). A field guide to emot ional intelligence. In J. Ciarrochi, J. Fo rgas, & J. Mayer (Eds.), Emotional intelligence in everyday life: A scientific Inquiry (pp. 3-24). Phil adelphia: Psychology Press. Mayer, J. D., Caruso, D. R., & Salovey, P. (1999). Em otional intelligence meets traditional standards for an intelligence. Intelligence, 27, 267-298. Mayer, J. D., Caruso, D. R., & Salovey, P. (2000). Se lecting a measure of emotional intelligenc e: The case study of ability sca les. In R. Bar-On., & J. D. A. Parker (Eds.), The handbook of emotional intelligence (pp. 320-342). San Francisco: Jossey-Bass. Mayer, J. D., & Cobb, C. D. (2000). Educat ional policy on emotional intelligence: does it make sense? Educational Psychology Review, 12 (2), 163-183.

PAGE 207

198 Mayer, J. D., & Geher, G. (1996). Emotional intelligence and the identification of emotion. Intelligence, 22, 89-113. Mayer, J. D., & Salovey, P. (1990). Perc eiving affective content in ambiguous visual stimuli: A component of emotional intelligence. Journal of Pe rsonality Assessment, 54 (3-4), 772-781. Mayer, J. D, & Salovey, P. (1993). The intelligence of emotional intelligence. Intelligence, 17, 433-442. Mayer, J. D., & Salovey, P. (1997). What is emotional inte lligence? In P. Salovey, & D. Sluyter (Eds.), Emotional development and emotional implications for educators (pp. 3-31). New York: Basic Book, Inc. Mayer, J. D., & Salovey, P. (2002). Pers onal intelligence, social intelligence, emotional inte lligence: Measures of not inte lligence. In C. Peterson & M. E. P. Seligman (Eds.), The classification of strengt hs and virtues: Values in action manual. Philadelphia: Mayerson Foundation. Mayer, J. D., Salovey, P., & Caruso, D. R. (2000a). Models of emotional intelligence. In R. J. Sternberg (Ed.), The handbook of intelligence (pp. 396-420). New York: Cambridge University Press. Mayer, J. D., Salovey, P., & Caruso, D. R. (2000b). MSCEIT Item Booklet (Version 2.0). Toronto, Canada: MultiHealth Systems Inc. Mayer, J. D., Salovey, P ., & Caruso, D. R. (2002). Mayer-Salovey-Caruso Emoti onal Intelligence Test: Manual. Toronto, Canada: Multi-Health Systems Inc.

PAGE 208

199 Mayer, J. D., Salovey, P., & Caruso, D. R. (2004a). A further consideration of the issues of emotional intelligence. Psychological Inquiry, 15 (3), 249-255. Mayer, J. D., Salovey, P., & Caruso, D. R. (2004b). Emotional intelligence: Theory findings, and implications. Psychological Inquiry, 15 (3), 197-215. Mayer, J., Salovey, P., Caruso, D., & Sitarenios, G. (2001). Emotional intelli gence as a standard intelligence. Emotion, 1, 232-242. Mayer, J. D., Salovey, P., Caruso, D. R, & Sitarenios, G. (2003). Measuring emotional in telligence with the MSCEIT V2.0. Emotion, 31 97-105. McCarthy, J. D., & Hoge, D. R. (1982). Analysis of age effects in longitudinal studies of adolescent self-esteem. Developmental Psychology, 18 (3), 371-379. McCullough, G., Huebner, E. S., & Laughlin, J. E. ( 2000). Life events, selfconcept, and adolescent s positive subjective well-being. Psychology in the School, 37, 281-290. Mecca, A. M., Smelser, N. J. & Vasconcellos, J. (1989). The social importance of self-esteem. (Eds.), Berkeley: University of California Press. Meisenhelder, J. B. (1986). Se lf-esteem in women: The influence of employment and percepti on of husbands appraisals. IMAGE : Journal of Nursing Scholarship, 18 (1), 8-14. Michalos, A. C. (1991). Global report on student well bei ng: Life satisfaction and happiness. New York: Springer.

PAGE 209

200 Miller, L., Warner, V., Wickr amaratne, P., & Weissman, M. (1999). Self-esteem and depre ssion: Ten year follow-up of mothers and offspring. Journal of Affective Disorders, 52, 34-49. Mirels, H. L., & Garre tt, J. (1971). The protestant ethic as a personality variable. Journal of Consulting and Clinical Psychology, 36, 40-44. Morganti, J. B., Nehrke, M. F., Hulicka, I. M., & Cataldo, J. F. (1988). Life-span differences in life sati sfaction, self-concept, and locus of control. International Journal of Aging and Human Development, 26 (1), 45-58. Mruk, C. (1995). Self-esteem: Research, theory, and practice. New York: Springer Publishing. Mruk, C. (1983). Being with oneself in a biographically critical way: An existential phenome nological investigation. Dissertation Abstracts International, 42 (12-B, Pt 1), p. 4937. Mullis, A. K., & Mullis, R. L. (1997). Voca tional interests of adolescents: Relationship between se lf-esteem and locus of control. Psychological Report, 81, 1363-1371. Myers, D. G. (1992). The pursuit of happiness. New York: William Morrow and Company. Myers, D. G., & Diener, E. (1995). Who is happy? Psychological Science, 6, 10-19. Nelson, D. B. & Low, G. R. (2002). Emotional intelligence: Achieving academic and career excellence Upper Saddle River, N J: Prentice Hall.

PAGE 210

201 Nelson, D. B., & Ne lson, K. W. (2003). Emotional intelligence skills: Significant factors in freshmen achievement and retention. Paper presented at the American Counseling Association Conf erence (Anaheim, CA: March 21-25, 2003). Newsome, S., Day, A. L., & Catano, V. M. (2000). Assessi ng the predictive validity of emotional intelligence. Personality and Individual Difference, 29, 1005-1016. Nowicki, S., & Strickland, B. R. (1973). A locus of cont rol scale for children. Journal of Consulting and Clinical Psychology, 40 (1), 148-154. Nunnally, J. C. (1978). Psychometric Theory (2 nd ed.). New York: McGraw-Hill. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed .). New York: McGraw-Hill. OConnor, R. M. Jr., & Little, I. S. (2003) Revisiting the predictive validity of emotional intelligence: Self-report vers us ability-based measures. Personality and Individual Differences, 35 (8), 1893-1902. Okma, P., & Veenhoven, R. (1996). Is a l onger life better?: Happiness of the very old in eight European Union countries. Journal of Personality and Social Psychology, 67, 92-104. Onwuegbuzie, A. J. (2003). Expanding the framework of internal and external validit y in quantitative research. Research in the Schools, 10, 71-90.

PAGE 211

202 Osborne, J. W., Christensen, W. R., & Gunter, J. (2001). Educational Psychology fr om a Statisticians Perspective: A Review of the Power and Goodness of E ducational Psychology Research. Paper presented at the national m eeting of the American Educat ion Research Association (AERA), Seattle, WA. OShea, C. S. (2002). Forward. In J. Higbee, D. Lundell, & I. Duranczyk (Eds.), Developmental education: Policy and practice. Auburn, CA: National Association for Developmental Education. Page, G. L., & Scalora, M. J. (2004). The utility of locus of control for assessing juvenile amenability to treatment. Aggression and Violent Behavior, 95 (5), 523-534. Palmer, B., Donaldson, C. & Stough, C. (2002). Emot ional intelligence and life satisfaction. Personality and Individual Differences, 33 (7), 1091-1100. Palmer, B., Walls, M., Burgess, Z., & St ough, C. (2001). Emoti onal intelligence and effective leadership. Leadership and Organization Development Journal, 22, 5-10. Parker, J. D. A., Taylor, G. I., & Bagby, R. M. (2001). The relationship between alexithy mia and emotional intelligence. Personality and Individual Differences, 30, 107-115. Parkerson, G. R., Broadhead, W. E., & Tse, C. J. (1990). The health status and life satisfacti on of first-year medical students. Academic Medicine, 65 (9), 586-588.

PAGE 212

203 Pascarella, E., & Terenzini, P. T. (1991). How College Affects Students. San Francisco: Jossey-Bass. Paulhus, D., Lysy, D. C., & Yik, M. S. M. (1998). Self-report measures of intelligence: Ar e they useful as proxy IQ Test? Journal of Personality, 66, 525-553. Pavot, W., & Diener, E. (1993). Review of the satisfaction with life scale. Psychological Assessment, 5, 164-172. Pavot, W., Diener, E., Colvin, B., & Sandvik E. (1991). Further validation of the sati sfaction with life scale: Evidence for the cross-method convergence of self -report well-being measures. Journal of Personality Assessment, 57, 149-161. Pedhazur, E. J. (1997). Multiple regression in behavioral research: Explanation and prediction. Fort Worth, TX: Harcourt Brace College Publishers. Peterson, C., Maier, S. F. & Seligman, M. E. P. (1993). Learned helplessness: A theory fo r the age of personal control. New York: Oxford University Press. Petrides, K. V., & Furnham, A. (2000). On the di mensional structure of emotional intelligence. Personality and Individual Differences, 29, 313-320. Pfeiffer, S. (2001). Emotional intellig ence: Popular but elusive construct. Roeper Review, 23 (3), 138-142. Phares, E. J. (1957). Expectancy changes in skill and chance situations. Journal of Abnor mal and Social Psychology, 54, 339-342.

PAGE 213

204 Piers, E. V. (1984). The Piers-Harris Children Self-Concept Scale (The way I feel about myself): (Revised Manual). Los Angeles: Western Psychological Services. Piers, E. V., & Harris, D. B. (1969). Age and other correlates of self-concept in children. Journal of Educational Psychology, 55 (2), 93-95. Pinquart, M., & Sorensen, S. (2000). Influences of soci oeconomic status, social network, and competence on s ubjective well-being in later life: A metaanalysis. Psychology and Aging, 15 (2), 187-224. Platten, M. R., & Williams, L. R. (1979). A comparative analysis of the factorial structures of two administrations of the Pier s-Harris Self-Concept Scale to one group of elementary school children. Education Psychology Measurement, 41, 453-462. Plucker, J. A. (2003). Teaching intervent ion as critical creative processes: A course on technoscientif ic creativity. In Runco, M. (Ed.), Critical creative process (pp. 275-302). Cresskill, NJ: Hampton Press. Pool, C. R. (1997). Up with emotional health. Educational Leader, 54, 43-46. Pope, A., McHale, S., & Craighead, E. (1988). Self-esteem enhancement with children and adolescents. New York: Pergamon Press. Radloff, J. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401.

PAGE 214

205 Ramanaiah, N. V., Detwiler, F. R. J., & Byravan, A. ( 1997). Life satisfaction and the five-f actor model of personality. Psychological Reports, 80 (3), 1208-1210. Raphael, D. (1988). High school conceptual le vel as an indicator of young adult adjustment. Journal of Personality Assessment, 52, 679-690. Renn, R. W., & Vandenberg, R. J. (1991). Differences in employee attitude and behaviors based on Rotters (1966) internal-external locus of control: Are they all valid? Human Relations, 44(11), 1161-1178. Rhee, K. S. (2001). T eaching emotional intelligen ce: Application in an undergraduat e management course. Unpublished masters thesis, Northern Kentu cky University, Highland Heights, KY. (UMI No. 41099). Rhode Island Emotional Compet ency Partnership. (1998). Update on emotional competency. Providence, RI: Rhode Island Partners. Robert, G., & Nollet-Clememcom, J. (1995). Eating and em otional disorders in adolescent obese girl s with insulin-dependent diabetes mellitus. European Child and Adolescent Psychiatry, 4 (4), 270-279. Roberts, R. D., Zeidner, M., & Matthews, G. (2001). Does emotional intelligence meet traditional standards for intelligence? Some new data and conclusions. Emotion, 1, 196-231. Roberts, R. E. (1980). Reliabi lity of the CES-D scale in different ethnic contexts. Psychiatry Research, 2, 125-134.

PAGE 215

206 Roget International Th esaurus Indexed Edition (1992). Barbara, A. K. New York: Harper Collins. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Prince ton University Press. Rosenberg, M. (1979). Conceiving the self. New York: Basic Books. Rosenberg, M. (1986). Conceiving the self (reprint ed.). Melbourne, FL: Academic Press. Rothbaum, F., Weisz, J. R. & Snyder, S. S. (1982). Changing the world and changing the self: A two-process m odel of perceived control. Journal of Personal ity and Social Psychology, 42 (1) 5-37. Rotter, J. B. (1954). Social learning and clinical psychology. Oxford, England: Prentice-Hall. Rotter, J. B. (1955). The role of the p sychological situation in determining the direction of hum an behavior. In M. R. Jones (Ed.), Nebraska Symposium on Motivation. (pp. 245-269.) Lincoln, NB: University of Nebraska Press. Rotter, J. B. (1960). Some implications of a social learning theory for the prediction of good dire cted behavior from testing procedures. Psychological Review, 67, 301-316. Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, 80 (1), 1-28.

PAGE 216

207 Rotter, J. B. (1975). Some problems and misconceptions related to the construct of internal versus external control of reinforcement. Journal of Consulti ng and Clinical Psychology, 43, 56-67. Rotter, J. B. (1982). The development and applications of social learning theory: Selected papers New York: Praeger Publishers. Rotter, J. B. (1990). Internal versus exte rnal control of reinforcement: A case hi story of a variable. American Psychologist, 45, 489-493. Rotter, J. B., Liverant, S., & Crowne, D. P. (1961). The growth and extinction of expectancies in chance controlled and skill tasks. Journal of P sychology: Interdisciplinary and Applied, 52, 161-177. Rotter, J. B., Seeman, M., & Liverant, S. (1962). Internal versus external control of reinforc ement: A major variable in behavior theory. In N. E. Washburne (Ed.), Decision, values and groups: Proceedings of a conference held at the Un iversity of New Mexico, (pp. 473-517). New York: Pergamon. Saklofske, D. H., Austin, E. J., & Minski, P. S. (2003) Factor structure and validity of a trait emotional intelligence measure. Personality and Individual Differences, 34 (4), 707-721. Salovey, P., Stroud, L. R., & Wooler y, A. (2002). Perceived emotional intelligence, st ress reactivity, and symptoms report s: Further explorations using the trait meta-mood scale. Psychology and Health, 17 (5) 611-627. Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cogni tion, and Personality, 9 (3), 185-211.

PAGE 217

208 Salovey, P., Mayer, J. D., Goldman, S. L., Turvey, C., & Palfai, T. P. (1995). Emotional att ention, clarity, and r epair: Exploring emotional intelligence using t he trait Meta-Mood Scale. In J. W. Pennebaker (Ed.), Emotion, disclosure, and health (pp. 125-154). Washington, DC: American Psyc hological Association. Salovey, P., Woolery, A., & Mayer, J. D. (2001). Emoti onal intelligence: Conceptualizat ion and measurement. In G. J. O. Fletcher., & M. S. Clark, (Eds.), Blackwell handbook of social psychology: Interpersonal processes (pp. 279). Malden, MA: Blackwell Publishers. SAS System for Windows 8 (2), (2003). Cary, North Caro lina: SAS Institute Inc. Schmitt, N., & Bedeian, A. G. (1982). A co mparison of LISREL and two-stage least squares analysis of a hypothesized life-job satisfaction reciprocal relationship. Journal of Applied Psychology, 67, 808-817. Schuessler, K. F., & Fisher, G. A. (1985). Qu ality of life research and sociology. Annual Review of Sociology, 11 129-149. Schulz, R., Tompkins, S. L., Wood, D., & Decker, S. (1987). The social psychology of care giving: Physical and psychological costs of providing support to the disabled. Journal of Applied Social Psychology, 17 (4), 401-428. Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty D. J., Cooper, J. T., Golden, C. J., & Dornheim, L. (1998). De velopment and validation of a measure of emotional intelligence. Personality and Individual Differences, 25, 167-177.

PAGE 218

209 Scollon, C., Diener, E., Oish i, S., & Biswas-Diener, R. (2004). Cross situational consistency of af fective experiences across culture. Journal of Personality and Social Psychology, 86 (3), 460-472. Sekaran, V. (1986). Self-est eem and sense of competence as moderators of the job sati sfaction of professionals in dual career families. Journal of Occupational Behavior, 7(4), 341-344. Shapiro, D. H.., Schwartz, C. E., & Asti n, J. A. (1996). Controlling ourselves, controlli ng our world: Psychologys role in understanding positive and negative consequences of seeking and gaining control. American Psychologist, 51, 1213-1230. Shapiro, L. (1997). How to raise a child with a high EQ: A parents guide to em otional intelligence, New York: Harper Collins Publisher. Shin, D. C., & Johnson, D. M. (19 78). Avowed happiness as an overall assessment of the quality of life. Social Indicators Research, 5, 475-492. Shmotkin, D. (1990). Subjective wellbeing as a function of age and gender: A multivariate look for differentiated trends. Social Indicators Research, 23, 201-230. Siegler, R. S. (1992). The other Alfred Binet. Developmental Psychology, 28 (2), 179-190. Silbert, E., & Tippett, J. S. (1965). Self-esteem: Clinical assessment and measurement validation. Psychological Report, 16(3, Pt. 2), 1017.

PAGE 219

210 Silva, F. (1993). Psychometric foundations and behavioral assessment Thousand Oaks, CA: Sage Publications Inc. Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). A terror management theory of so cial behavior: The psychological fu nction of self-esteem and cultural worl dviews. In M. P. Zanna (Ed.), Advances in experimental social psychology, 24, (pp. 93-159). San Diego: Academic Press. Southwick, R. B. (1998). Antecedents of transformational, transactional, and la isser-faire leadership (Doctoral dissertation, University of Georgia, Athens, GA, 1998). Dissertation Abstract International, 59( 6-B), 3107. Spector, P. E., & Johnson, H. M. (2006) Improving the definition, measurement, and applicati on of emotional intelligence. In K. R. Murphy (Ed.), A critique of emotional intelligence (pp. 325-344). Mahwah, NJ : Lawrence Erlbaum Associates Inc. Springer, L., Terenzini, P. T., Pascarella, E. T., & Nora, A. (1995). Influence on college student s orientations toward l earning for self-understanding. Journal of College Student Development, 36 (1), 5-18. Steer, R. A., Ball, R., Ranier i, W. F., & Beck, A.T. ( 1997). Further evidence for the construc t validity of the Beck Depression Inventory-II with psychiatric outpatients. Psychological Reports, 80 (2), 443-446. Sternberg, R. J. (1985). Beyond IQ: A triarchic theory of human intelligence. New York: Cambridge University Press.

PAGE 220

211 Sternberg, R. J. ( 1997). The concept of intelligenc e and its role in lifelong learning and success. American Psychologist, 52 (19), 1030-1037. St. Lawrence, J. S., Jefferson, K. W., Alle yne, E., & Brasfield, T. L. (1995). Comparison of education versus behavioral skills training interventions in lowering se xual HIV-risk behavior of subst ance-dependent adolescents. Journal of Consulting and Clinical Psychology 63 (1), 154-157. Stevens, J. (2002). Applied multivariate statis tics for the social science (3rd ed.). Mahw ah, New Jersey: Lawrenc e Erlbaum Publishers. Stones, M. J., Hadjistavropoulos, T., Tuuko, H., & Kozma, A. (1995). Happiness has trai t-like and state-like properti es: A reply to Veenhoven. Social Indicators Research, 36 (2), 125-144. Strack, F., Argyle, M., & Schwarz, N. (1991). Subjective well-being: An inte rdisciplinary perspective Elmsford, NY: Pergamon Press, Inc. Sue, S. (1999). Science, ethnicity and bias: Where have we gone wrong? American Psychologist, 54, 1070-1077. Sylvester, R. (1994). How emotions affect learning. Educational Leadership, 52, 60-65. Szulecka, T. K., Springett, N. R., & de Pauw, K. W. (1987). General health, psyc hiatric vulnerability and withdrawal from university in first-year undergraduates. British Journal of Guidance and Counseling Special Issue: Couns eling and Health, 15, 82-91.

PAGE 221

212 Tanaka, J. S., & Huba, G. J. (1987). Assessing the stability of depression in college students. Multivariate Behavioral Research, 22, 5-19. Tapia, M. (2001). Measuri ng emotional intelligence. Psychological Reports, 88 (2), 353-364. Taylor, C. (1994). Essentials of psychiatric nursing. St. Louis: Mosby. Tenhouten, W. D., Hoppe, K. D., & Bogen, J. (1986). Alexithymia and the split brain: IV Gottsc halk-Gleser content analysis an overview. Psychotherapy and Psychosomatics, 44 (3),113-121. Terenzini, P. T., & Pascarella, E. T. (1999). Students out-of -class experiences and their influence on learning and cognitive development: A literature review. Journal of College Students Development, 4 (95), 610-623. Terry, T., & Huebner, E. S. (1995). The rela tionship between self-concept and life sati sfaction in children. Social Indicators Research, 35 (91), 39-52. Thorndike, E. L. (1920, Marc h). Intelligence and its uses. Harpers Magazine, 140, 227-235. Tinto, V. (1993). Leaving college: Rethinking the ca uses and cures of student attrition. (2 nd ed.). Chicago: University of Chicago Press. Trice, A. D. (1990). Adolescents locu s of control and compliance with contingency contra cting and counseling interventions. Psychological Reports, 67, 233-234.

PAGE 222

213 Trinidad, D. R., & Johnson, C. A. (2002). The association between emotional intelligence and early adolescent tobacco and alcohol use. Personality and I ndividual Differences, 32 (1), 95-105. Uchida, Y., Kitayama, S., Mesquite B., & Reyes, J. (2001, July). Interpersonal sources of happiness: The relative significanc e in the U.S., the Philippines, and Japan Paper presented at a meeting of the American Psychological Soci ety. Ontario, Canada. Uchida, Y., Kitayama, S., Mesquita, B., & Reyes, J. (2001). Development and validati on of a sympathy scale. Journal of Psychology, 72 (4), 275-282. Upcraft, M. L., & Gardner, J. N. (1989). The freshman year experience: Helping students su rvive and succeed in college. San Francisco: Jossey Bass. Valente, S. (1994). Recognizing d epression in elderly patients. American Journal of Nursing, 94 (12), 19-24. Veenhoven, R. (1988). The utility of happiness. Social Indicators Research, 20, 333-354. Veenhoven, R., & Rotter dam, N. (1994). World database of happiness: Correlates of happiness: 7837 findings from 603 studi es in 69 nations 1911-1994, Vols. 1-3. Erasmus University Rotterdam Vermunt, R., Spaans, E., & Zorge, F. ( 1989). Satisfaction, happiness and well bei ng of Dutch students. Social Indicators Research, 21, 1-33. Veroff, J., Feld, S., & Guri n, G. (1962). Dimensions of subjective adjustment. Journal of Abnormal and Social Psychology, 64 (3), 192-205.

PAGE 223

214 Vila, G., Robert, J. J., & Nollet-Clem encon, C. (1995). Ea ting and emotional disorders in adolescent obese girls with insulin-dependent diabetes mellitus. European Child and Adolescent Psychiatry, 4, 270-279. Vitaliano, P., Russo, J., Young, H., Be cker, J. & Maiuro, J. (1991). The scre en for caregiver burden. Gerontologist, 31(1), 76-93. Websters New Dictionar y of the English Language. (2001). New York: Popular Publishing Company, LLC. Weiner, Y., Muczyk, J. P., & Gable, M. (1987). Relationships between work comm itments and experience of personal well-being. Psychological Reports, 60, 459-466. Weissman, M. M., Sholomska, D., Pottenger, M., Prusoff, B. A. & Locke, B. Z. (1977). Assessing depressive symptoms in five psychiatric populations: A validation study. American Journal of Ep idemiology, 106, 203-214. Weiten, W. (1989). Psychology : Themes and variations. Belmont, CA: Broo ks/Cole Publishing Co. Wells, L.E. & Marwell, G. (1976). Self-esteem: Its conceptualization and measurement. Beverly Hills: Sage Publishing. White, R. W. (1963). Ego and reality in psychoanalytic theory. Psychological Issues, 3 (Whole No.11), 1-210 White, J. M. (1992). Marital status and well-being in Canada. Journal of Family Issues, 13, 390-409.

PAGE 224

215 White, S. H. (2000). Conceptual foundation of IQ testing: Perspectives on intelligence testing. Affirmative Action and Educational Policy, 6 (1), 33-43. Wigfield, A., Eccles, J. S ., Iver, D. M., Reuman, D. A., & Midgley, C. (1991). Transitions during early adolescence: C hanges in childrens domain specific se lf-perception and general self-est eem across the transition to junior high school. Developmental Psychology, 27, 552-565. Williamson, E. G. (1957). T he dean of students as educator. The Educational Record, 230-240. Willoughby, C., King, G., & Polatajko, H. J. (1995). A therapists guide to childrens self-esteem. American Journal of O ccupational Therapy, 50 (2), 124-132. Winters, N. C., Myers, K., & Proud, L. (2002). Ten-year re view of rating scales III: Scales assessing suic idality cognitive style and self-esteem. Journal of the American Academy of Child and Adolescent Psychiatry, 41 (10), 1150-1181. Witter, R. A., Okun, M. A ., & Stock, W. (1984). Education and subjective wellbei ng: A meta-analysis. Education Evaluation and Policy Analysis, 6(2), 165-173. Wong, C. S., & Law, K. S. (2002). The e ffects of leader and follower emotional intelligence on perform ance and attitude: An exploratory study. Leadership Quarterly, 13 (3), 243-274.

PAGE 225

216 Wood, P. C., Hillman, S. B., & Sawilowsky, S. S. (1996). Locus of control, selfconcept, and self-esteem am ong at risk African American adolescents. Adolescence, 31, 597-604. Wood, S. E., Wood, E. G., & Boyd, D. (2004). The world of psychology. (5th ed .). New York: Pearson. Wylie, R. C. (1989). Measures of self-concept. Lincoln, NE: University of Nebraska Press. Yardley, J. K., & Rice, R. W. (1991) The relationship between mood and s ubjective well-being. Social Indicators Research, 24 (1), 101-111. Young, T. W., & Shorr, D. N. (1986). Factors affecting locus of control in school children: Genetic, social, and general. Psychology Monographs, 112, 405-417. Zimmerman, M. A., Copeland, L. A., S hope, J.T., & Dielman, T. E. (1997). A longitudinal study of self-esteem: Implications for adolescent development. Journal of Youth and Adolescence, 26, 117-141. Zung, W. A. (1965). A self rating depression scale. Archive of General Psychiatry, 12, 63-70.

PAGE 226

217 APPENDICES

PAGE 227

Appendix A Instruments 218

PAGE 233

Appendix B Institutional Review Board 219

PAGE 238

Appendix C Letter of Voluntary Research Participation 220

PAGE 240

About the Author Kevin T. Murphy is a native of Queens County New York, where he completed twelve years of public school education before entering the Florida Community College system and beginning a career in law enforcement. In 1976, he earned the Associate of Arts (A.A ) from Central Florida Comm unity College (CFCC), and in 1979, the Bachelor of Science (B.S.) in both Psychology and Criminology from Florida State University (FSU). In 1992 Kevin completed the Master of Science (M.S.) in mental health counseling at Stetso n University. After 25 years of service Kevin retired from law enforcement in 2001. He is now a licensed mental health therapist and adjunct psycholog y instructor at CFCC.