USF Libraries
USF Digital Collections

Gender stereotypes of citizenship performance

MISSING IMAGE

Material Information

Title:
Gender stereotypes of citizenship performance
Physical Description:
Book
Language:
English
Creator:
Wilkinson, Lisa
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Sex role   ( lcsh )
Stereotypes (Social psychology)   ( lcsh )
Organizational behavior   ( lcsh )
organizational citizenship performance
shifting standards model
objective scales
reward recommendations
overall performance
Dissertations, Academic -- Psychology -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: The relationship between citizenship performance and overall performance and the relationship between citizenship performance and reward recommendations were investigated, with gender and scale type as possible moderators. Two hundred and fifty-four University of South Florida students were used in this study. The majority of these participants were undergraduate, psychology majors, female, and between the ages of 17 and 23. Participants were given statements describing a teacher's performance and were asked to evaluate the professor on citizenship and overall performance and recommend them for rewards. No support was found for the hypothesis that men would have a stronger relationship than women between citizenship and overall performance. No support was found for the hypothesis that men would have a stronger relationship than women between citizenship performance and reward recommendations. Scale type was not found to influence these relationships. These results are not consistent with the shifting standards model. Numerous suggestions are made for changes to the experiment, including performing a field study instead of a lab study.
Thesis:
Thesis (M.A.)--University of South Florida, 2003.
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 Lisa Wilkinson.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 106 pages.

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 - 001430580
oclc - 52376115
notis - AJL4041
usfldc doi - E14-SFE0000098
usfldc handle - e14.98
System ID:
SFS0024794:00001


This item is only available as the following downloads:


Full Text

PAGE 1

GENDER STEREOTYPES OF CITIZENSHIP PERFORMANCE by LISA WILKINSON A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Tammy Allen, Ph.D. Louis Penner, Ph.D. Judith Bryant, Ph.D. Date of Approval: April 8, 2003 Keywords: Organizational Citizenship Performa nce, Shifting Standards Model, Objective Scales, Reward Recommendations, Overall Performance Copyright 2003, Lisa Wilkinson

PAGE 2

i Table of Contents List of Tables iii List of Figures v Abstract vi Chapter 1 – Introduction 1 Citizenship Performance 3 Construct Development 3 Dimensions of Citizenship Performance 6 Consequences of Citizenship Performance 8 Stereotypes of Citizenship Pe rformance 9 Shifting Standards Model 12 Overall Performance 17 Reward Recommendations 25 Citizenship Performance Dimensions 31 Chapter 2 – Method 32 Participants 32 Design 32 Procedure 33 Materials 34 Measures 38 Citizenship Performance 38 Overall Performance 39 Reward Recommendations 40 Chapter 3 – Results 41 Pilot Study 41 Manipulation Check 41 Order Effects 42 Rater Gender Effects 43 Descriptive Statistics 44 Hypothesis Testing 47 Exploratory Questions 53 Post Hoc Analysis 57

PAGE 3

ii Chapter 4 – Discussion 60 Gender and Citizenship Performance 61 Overall Performance 64 Reward Recommendations 66 Citizenship Performance Dimensions 67 Post Hoc 67 General Limitations 68 Implications and Future Research 71 References 74 Appendices 85 Appendix A: Consent Form 86 Appendix B: Cover Story 88 Appendix C: Performance Log 89 Appendix D: Citizenship Performance Scale 91 Appendix E: Overall Performance Scale 94 Appendix F: Reward Recommendations 96 Appendix G: Demographic Information 98 Appendix H: Participants Debrie fing 99

PAGE 4

iii List of Tables Table 1 Distribution of Participants in to 4 Conditions 32 Table 2 Perceived Number of Males and Fe males in 14 Occupations 34 Table 3 Descriptive Statistics (z-scores) by Gender and Scale Type 43 Table 4 Descriptive Statistics (raw) by Ge nder and Scale Type 43 Table 5 Zero-Order Correlation Coefficients Among Variables with Objective Citizenship Performance 45 Table 6 Zero-Order Correlation Coefficients Among Variables with Subjective Citizenship Performa nce 45 Table 7 Zero-Order Correlation Coefficients Among Variables with Objective Citizenship Performance, Split by Professor Gender 47 Table 8 Zero-Order Correlation Coefficients Among Variables with Subjective Citizenship Performance, Split by Professor Gender 47 Table 9 Hierarchical Regression Analysis wi th Objective Overall Performance as the Dependent Variable 48 Table 10 Hierarchical Regre ssion Analysis with Subjective Overall Performance as the Dependent Variable 50 Table 11 Hierarchical Regre ssion Analysis with Objective Reward Recommendations as the Dependent Variable 51 Table 12 Hierarchical Regre ssion Analysis with Subjective Reward Recommendations as the Dependent Variable 52 Table 13. Hierarchical Regression Analysis with Subjective Measures of Both Overall and Citizenship Performance. 53 Table 14. Hierarchical Regression Anal ysis with Subjective Measures of Both Reward Recommendations and Citizenship Performance. 54 Table 15 Promax Oblique Factor Rotati on Pattern Matrix 56

PAGE 5

vi Table 16 Mean Factor Scores by Gender 56 Table 17 Outliers and their Influence on the Variables 57

PAGE 6

v List of Figures Figure 1. Predicted Results for Hypothesis 1 17 Figure 2. Predicted Results for Hypothesis 2a 23 Figure 3. Predicted Results for Hypothesis 2b 24 Figure 4. Predicted Results for Hypothesis 3 24 Figure 5. Predicted Results for Hypothesis 4a 28 Figure 6. Predicted Results for Hypothesis 4b 29 Figure 7. Predicted Results for Hypothesis 5 30 Figure 8. Scree Plot of Eigenvalues for Factor Analysis of Citizenship Performance 55 Figure 9. Interaction Plotted According to Ai ken and West’s procedure 58

PAGE 7

vi Gender Stereotypes of Citizenship Performance Lisa Wilkinson ABSTRACT The relationship between citi zenship performance and ove rall performance and the relationship between citizen ship performance and reward recommendations were investigated, with gender and scale type as possible moderators. Two hundred and fiftyfour University of South Florid a students were used in this study. The majority of these participants were undergraduate, psychology ma jors, female, and between the ages of 17 and 23. Participants were given statements describing a teacher’s performance and were asked to evaluate the professor on citizensh ip and overall performance and recommend them for rewards. No support was found for the hypothesis that men would have a stronger relationship than wo men between citizenship and overall performance. No support was found for the hypothesis that men would have a stronger relationship than women between citizenship performance and reward recommendations. Scale type was not found to influence these relationships. Th ese results are not consistent with the shifting standards model. Numerous suggestions are made for changes to the experiment, including performing a field study instead of a lab study.

PAGE 8

1 Chapter 1 Introduction Relative to men, women have been historically underval ued in the workplace with regard to promotions (Ros en & Jerdee, 1973), compensation (Judge, Cable, Boudreau, & Bretz, 1995) and performance evaluations (Hamner, Kim, Baird, & Bigoness, 1974). Stereotypes of performance held by raters have been studied as a possible explanation for why women’s contributions are devalued. Ge nder stereotypes (beliefs about male or female behaviors that cause generalizations to all members of that se x) can dictate raters’ expectations of male and female behavior within the workplace. Consequently, these expectations may influence how raters evaluate performance and make promotion decisions. The focus of the current study is to examine the consequences of gender stereotypes of citizenship performance for men and wome n. Citizenship performance represents behaviors that are important in shaping the work environment and in supporting task performance (Borman, Penne r, Allen, & Motowidlo, 2001). There is some evidence to suggest that female stereot ypes that characterize women to be unselfish and supportive (Eagly & Crowley, 1986) may pl ace higher expectations on women to be organizational citizens (Allen & Rush, 2001). The shifting standards model provides a theory of how raters’ stereo types effect ratings (Biernat Manis, & Nelson, 1991). The theory predicts that we judge people’s beha viors based on expectations we have for the group (race, gender, etc.) they belong to a nd we shift our standards accordingly. For

PAGE 9

2 example, when rating a woman’s athleticism, raters will compare a female to other females, not to males. Therefore, even if the best female athlete is not as athletic as the average male athlete, she will still be rated as an above average athl ete because raters will compare her to the lower athletic standard they have for women. The shifting standards model provides the framework used in the present study for investigating whether men and women are equa lly evaluated and rewa rded for citizenship performance. Considering that citizenshi p performance has been found to contribute variance to both performance evaluations (Allen & Rush, 1998; Borman, White, & Dorsey, 1995; Conway, 1999; MacKenzie, Podsakoff, & Fetter, 1991; Motowidlo & Van Scotter, 1994) and different types of or ganizational rewards (A llen, 2000; Allen & Rush, 1998; Chen & Heilman, 2001; Hui, Lam, & Law, 2000; Van Scotter et al., 2000), it seems important to study ge nder as a possible moderator of the relationships between citizenship performance and evaluations and rewards. In what follows, a review on the development of the citizenship performance construct will be provided. Following th e construct review, the development of dimensions and consequences of citizenship performance will be detailed. Next, there will be a discussion on gender stereotypes of citizenship performance. A detailed description of the shif ting standards theory will follow and then a section on the relative contribution of the personal support dimension. Finally, the consequences of differential rewards and evaluations of citizenship perfor mance will be discussed in the context of both performance evaluations and promotions.

PAGE 10

3 Citizenship Performance Construct development References of helping behavi ors in the work place can be found as far back as 1939 in Chester Barnard’ s book, “The Functions of the Executive.” In his book, Barnard described i ndividuals who were responsible for giving the office a collective atmosphere (Barnard, 1939) The topic of altruistic behaviors in the workplace, however, was only sparsely menti oned throughout the literature for the next forty years (e.g., Katz, 1964; Roethlisberger & Dickson, 1939). In 1977, Organ resumed interest in helping behaviors in a discussi on about the causal relationship between job performance and job satisfaction. In his ar ticle, Organ attempted to provide support for the dying theory that a satisfied employee is a productive employee. Organ claimed that if theorists considered a broader definition of performance, one that included behaviors that assist in the ease of or ganizational functioning, then th e model of satis faction causing performance would find more empirical and theo retical support. Some examples of the behaviors described by Organ are, arriving to work on time, not breaking the rules, and going along with company decisions a nd action without raising objections. Beginning in the 1980’s, discussion of citi zenship behaviors began to permeate the organizational literature. Throughout its evolution, the citizenship performance construct has been given many titles and diffe rent variations on its definition. In 1983, Smith, Organ, and Near introduced the construc t of Organizational Citizenship Behavior (OCB). The authors described OCB as a “m yriad of acts of cooperation, helpfulness, suggestions, gestures of goodwill, altruism, a nd other instances of what we might call citizenship behavior” (p. 653). In 1988, Orga n modified his original definition of OCB by defining it as a behavior th at does not receive formal rewa rds, but that helps in the

PAGE 11

4 functioning of the organizati on. As Organ worked on the construct of OCB, other authors developed related constructs. In 1986, Brief and Motowidlo introduced a similar construct called Prosocial Organizational Behaviors (POB). These author s defined the construct as “positive social acts carried out to produce and maintain the we ll-being and integrity of others” (p. 710). In their article, Brief and Motowidlo examin ed different ways to break the concept of prosocial behaviors into parts and came up with the in-role an d extra-role di stinction. Inrole behaviors are defined as behaviors that are written in an employee’s job requirements, whereas extra-role behaviors are defined as acts that are not included in an employee’s job tasks or duties. However, Morrison (1994) found th e in-role and extrarole division to be overlapping and hard to distinguish. In a study of 317 clerical workers, Morrison found that employees were more likely to perform OCB if they defined the behaviors as in-role rather than extra-role. Th ey also found that employees and supervisors didn’t agree, above chance, on what behaviors would be defined as extrarole and which would be defined as in-ro le. Subsequently, Borman and Motowidlo (1997) derived another distincti on that avoided the problems of in-role versus extra-role by focusing on whether the behavi or represented a “core” task. Borman and Motowidlo (1997) defined cont extual performance as behaviors that improve or maintain the environment of the organization. Beha viors that typify contextual performance are important because they surround and support the core tasks being performed. The term task performance is used in contrast to contextual performance and represents behaviors that ha ve a direct contributi on to the productivity of the company. Since job requirements vary between jobs and companies, this

PAGE 12

5 distinction can be easily genera lized across situations. Re sults from Motowidlo and Van Scotter (1994) supported the contextual perf ormance and task performance distinction, showing that each type of performance contri buted uniquely to performance evaluations. Task performance contributed 13% of the variance in overall performance and contextual performance contributed 11% above the va riance accounted for by task performance Further, the researchers found th at the two constructs had th eir own unique antecedents. Van Scotter, Motowidlo, and Cross (2000) similarly found support for the division of performance into task and contextual perf ormance. In this study, the authors were interested in accounti ng for the variance found in systematic rewards. The authors found that the two different types of performan ce each differentially predicted variance in medals, promotability rati ngs, rewards, and rank. In light of the recent research on citize nship performance, Organ (1997) wrote a review in order to, in part, consolidate the different terms for helping behaviors into one concept and one definition. Organ agreed with the usefulness of the contextual and task performance distinction, but argued that th e name, contextual performance, does not provide the reader with rea dy knowledge of what the concep t means. He described the concept of contextual performance as “cold, gray, and bloodless” (p. 91). In a recent article by Borman et al. (2001) the authors use the term citizenship performance in place of the term contextual performance. Citizenship performance is defined as behaviors that contribute pos itively to the work place environment and enhance workers' task performance. C itizenship performance takes on the same definition as contextual performance, but its meaning can more readily be understood from the name alone than contextual perfor mance. Examples of these behaviors are

PAGE 13

6 assisting co-workers with their jobs, follo wing the rules, being friendly and having a positive attitude, staying late to finish one’s work, showing support of the organization, and giving extra effort on the job. In li ght of the empirical findings supporting the contextual and task distinction, the definiti on of citizenship performance given will be used in the present study. Dimensions of citizenship performance. Beyond the problems of naming the construct citizenship performance, there have been a number of different dimensions proposed. Smith et al. (1983) used inte rviews with managers from different organizations to create 16-items that ope rationalized citizenship behaviors. The managers or supervisors were asked to id entify behaviors that contributed to the organization, but that were not explicit re quirements of the job. These 16 items were factor analyzed and loaded on two factors: altruism and generalized compliance. Altruism was defined as help given to other individuals in the organization. Generalized compliance was defined as “a more impersona l sort of conscientiousness, more of a ‘good soldier’ or ‘good citizen’ s yndrome of doing things that are ‘right and proper’ but for the sake of the system rather than for specific persons” (p. 662). This two-factor model has also been referred to as OCBO and OCBI (e.g., Skarlicki & Latham, 1996). The distinction between OCBO and OCBI lies in the object to which the behavior is being directed. OCBO is or ganizational citizenship behavior directed at the organization and OCBI is OCB directed at the employee’s co-workers. Organ (1988) proposed 3 new dimensions to add to the previous 2-dimension model of altruism and generali zed compliance: sportsmanship, ci vic virtue, and courtesy. Organ felt that these five dimensions more ad equately covered the entire breadth of the

PAGE 14

7 OCB concept. The following are definitions of these constructs according to Organ (1988). Sportsmanship is the willingness of an employee to ignore the small problems that arise and not complain. Civic virt ue is when an employee takes personal responsibility for how the orga nization performs. Courtesy is the amount one tries to ameliorate the situation when conflict arises between co-workers or personal problems develop for co-workers. Podsakoff and Mackenzie (1997) used a 3-dimension model and introduced the new dimension of helping behavior. The aut hors felt that raters in their studies were unable to distinguish between the altruism a nd courtesy dimensions provided by Organ. The authors chose to use the term helping behaviors to encompass both altruism and courtesies. They define help ing behaviors as behaviors that help co-workers solve workrelated dilemmas (Podsakoff & MacKenzie, 1994). Recently, Coleman and Borman (2000) conducted a study for the purpose of bringing together the different dimensions of OCB used throughout the literature and conducting several types of analyses to find th e best factor structure. These authors found the most consistent and successful m odel to be a 3-factor model. The three dimensions of this model were person al support, organizational support, and conscientious initiative. Personal support enco mpasses the previous concepts of altruism and helping behaviors and involves assist ing co-workers with problems, being considerate of co-workers needs, and c ooperating with co-worke rs. Organizational support includes the earlier co ncept of compliance and represents behaviors that demonstrate support for the organization, bot h by following the rules and making one’s organizational commitment evident with fellow co-workers and people outside the

PAGE 15

8 company. Conscientious initiative involves a desire and perseverance to fulfill one’s job duties and create the best oppor tunities for self and company. The current study will utilize the three dimensions (p ersonal support, conscientious initiative, and organizational support) derived by Borman and Coleman (2000). Consequences of citizenship behavior One reason for the growing research on citizenship performance in the past decade is the consequences it holds for employees. For example, it has consistently been f ound that employees are evaluated for their citizenship performance as well as for th eir task performance (Allen & Rush, 1998; Borman, White, & Dorsey, 1995; Conway, 1999; MacKenzie, Podsakoff, & Fetter, 1991; Motowidlo & Van Scotter, 1994). More specifically, organizational citizens receive higher performance appraisals than do employees who don’t perform citizenship behaviors. For instance, MacKenzie et al. (1991) found that OCB accounted for about 30% of the variance in manager’s performance evaluations of salespeople. Podsakoff and MacKenzie (1994) found similar results in a field study of insurance salesmen. These researchers found that OCB accounted for 48% of the variance in employee evaluations. Further demonstration of the impact of citizenship performance on employees is illustrated with findings that supervisors consider citizenship performance when making various reward recommendations (Allen, 2000; Allen & Rush, 1998; Chen & Heilman, 2001; Hui, Lam, & Law, 2000; Van Scotter, 2000; Van Scotter et al., 2000). In a study of military personnel, Van Scotter et al. (2000) found that contextual performance predicted promotability ratings above the variance explained by task performance. Further, the authors found that contextual performance explained variance in informal rewards, whereas task performance did not. Informal rewards are rewards given to

PAGE 16

9 employees that are not recorded in an em ployee’s personnel file. Examples of these rewards include special assignments, new positions, training, and aiding in career advancement. Similarly, in a field study, A llen (2000) tested the relationships between OCB and promotions and OCB and salary. Th e results showed that both salaries and promotions were significantly correlated with the amount of OCB exhibited by the employee. Allen and Rush (1998) used a five-item measure of reward recommendations that was created to reflect common or ganizational rewards. The fi ve items were increases in salary, promotions, public recognition, hi gh profile project, and opportunities for professional development. The researchers found that OCB correlated with both reward recommendations and performance evaluati ons. The growing evidence demonstrating the importance of citizenship performance to individual evaluations and rewards, underscores the need to accurate ly evaluate citizenship perf ormance. For example, it has been found, in the past, that stereotypes held by the raters can impact the accuracy of overall performance evaluations (Dobbins, Cardy, & Truxillo, 1988). Stereotypes of citiz enship performance. Stereotypes can have a crucial impact on evaluations of employee performance in the wo rkplace. They can be especially harmful because as Heilman (1995) states, once we have associated particular behaviors with a group of people, we generalize that behavi or to all individual group members. For example, Dobbins et al. (1988) found that raters with more traditional gender stereotypes rated women less favorably than men on overall performance evaluations. In her review on the effects of sex stereotypes in the wor kplace, Heilman describes four factors that work to maintain and reinforce stereotypes. These four influences are perceptions,

PAGE 17

10 interpretation, memory, and infe rences. First, perceptions are affected when we focus our attention on information that is consistent with stereotypes and ignore inconsistent information. For example, observers are more likely to recognize a man that is acting aggressively than a woman that is acti ng aggressively. Seco nd, people interpret information in a way that is consistent with stereotypes. For instance, if a woman is observed acting aggressively, the witness ma y interpret a woman's motives differently from a man’s motives. The viewer may assume that the woman was provoked but assume that the man started the conflict. Third, people tend to remember what is c onsistent with previ ous stereotypes or even remember events that did not occur becau se they are consistent with stereotypes. For example, when someone witnesses a fema le acting aggressively toward a male, the observer might falsely remember that the ma n provoked the attack. Finally, inferences are made when there is little or no information given about an individual. When lacking information people tend to rely on shortcut s based on superficial characteristics of someone. For example, when meeting a man and a woman for the first time, it might be assumed that the man is more aggressive than the female, based on their sex alone. Resilience of stereotypes was demonstr ated in a study by Ne lson, Biernat, and Manis (1990), in which the researchers attempte d to reduce various gender stereotypes. In an effort to increase accuracy of judgments, participants were placed in one of three groups. They were either told the truth that there were no gender differences between the ratees, were given monetary incentives for accuracy, or were educated on gender stereotypes. Only in the condition where the participants were told the truth, that men and women ratees were matche d for height, was there a signi ficant decrease of stereotype

PAGE 18

11 effects. In this condition men were stil l rated significantly taller than were women, however, there was a significan t decrease in the rating differences found between men and women. The cash incentives and the training on stereotypes had no significant impact on the ratings made by the participants. In a business setting, the pers istence of the ster eotypes held by supervisors could have implications for how they view their subordinates’ performance. Allen and Rush (2001) found support for the theory that people possess gender stereo types of citizenship performance. In this study, participants were given a list of 10 ci tizenship behaviors and 10 task behaviors, and were asked to determ ine how likely a male or a female employee would be to perform the behavior, to predic t the percentage of males or females who would perform these behaviors, and to provide an expected salary for the job. Results showed that expectations for ratees to engage in citizensh ip behavior were greater for women than for men. Allen and Rush’s (2001) findings were not surprising cons idering the past findings on gender and helping be havior stereotypes. In their meta-analysis on helping behaviors Eagly and Crowley (1986) discuss co mmon stereotypes that are attributed to women regarding their altruistic or social role. The aut hors assert that “women are expected to care for the personal and emotiona l needs of others, to deliver routine forms of personal service, and, more generally, to fa cilitate the progress of others toward their goals” (p. 284). The results supported th e contention that thoughtfulness and nurturing traits are considered female characteristic s. Similarly, Eagly, Makhijani and Klonsky (1992) described women as having “comm unal qualities, such as being friendly, unselfish, concerned with others, and emotiona lly expressive” (p. 6). These behaviors are

PAGE 19

12 similar to the behaviors of an organizational citizen. Examples of citizenship behaviors that resemble female descri ptions are helping co-workers when they need emotional support, demonstrating consider ation for co-workers, encouraging co-workers successes, and putting the needs of the team first. The current research is an attempt to extend the findings of Allen and Rush by investigating if women and me n are rated the same for thei r citizenship performance. The shifting standards model, introduced in the following se ction, provides a theory of how gender expectancies can create diffe rential evaluation of men’s and women’s citizenship performance. Shifting Standards Model The shifting standards model, borne out of the social psychology literature, describes how stereotypes can be hidden when a particular type of measurement is used. In the previously described study by Nelson et al. (1990), the authors developed the shifting standards model while st udying the resilience of the st ereotype that men are taller than are women. The authors found that men we re consistently rated taller than were women, even though height was controlled and there were no actual differences, on average, between men and women. Nelson et al. (1990) attempted to explain the large and consistent difference found between men and women on height. The authors suggested that scale type could have an impact on how raters make differential ratings for men and women. Feet and inches were used to measure height in this particular st udy. The authors claimed that objective scales “have a special virtue, in that there is uni versal agreement that a man of 5’7” and a women of 5’7” are in fact e qual in height” (p. 673). Objective scales are familiar and

PAGE 20

13 quantifiable and therefore provide a more accurate description of rater stereotypes. When height is provided in feet and inches, th ere is no need for interpretation because the meaning of 5’7” is the same for all raters. The objective scale, just described, was consid ered in contrast to the Likert scale. When a Likert scale is used, Nelson et al (1990) hypothesized that the endpoints of the scale could have a different meaning depe nding on the rater’s comparison group. Likert or continuum scales were considered subjec tive, because the meaning of these ratings could vary depending on the rater’s standard of comparison. For example, there is a common stereotype that men are better at ma th than women. Due to this stereotype, when raters judge a woman to be an above average math performer, this judgment may not be equivalent to a man receiving an above average rating. Because of the stereotype that men are better mathematicians than ar e women, male performance may be judged against a higher standard. Therefore, the rater shifts the standard down when rating a female as compared to when rating a male Consequently, since women are compared against a lower standard, they would not need the same math expertise in order to receive an equivalent evaluation, as would men. These different rating patterns, based on s cale type, led to the development of the shifting standards theory The premise of the theory is that when subjective ratings are used, a rater’s standard of comparison is unknown. However, when objective scales are used, raters are forced to compare the men to the women. In a test of the shifting standards theo ry, Biernat, Manis, and Nelson (1991) examined ratings of height, weight, and income across gender. Participants looked at 44 pictures and were asked to judge the height, weight, and income of the person in the

PAGE 21

14 photograph. The participants ra ted (subjective scale) and rank ed (objective scale) the people in the pictures. Specifically, participan ts rated the pictures on a Likert scale with seven choices (e.g., very tall to very short) and ranked them in order from most to least (e.g., tallest to shortest). The authors expected to find that when using the Likert scale, there would be less evidence of a stereotype influence on ratings. However, the results were not the same for the three independent va riables. For height and weight, men were rated significantly taller and heavier than were women on both subjective and objective measures. When the subjects used objective m easures to rate the people in the pictures, the difference between men and women was la rger than when rating on the subjective scale. For income, the results demonstrated, what the authors refer to as a reversed stereotyping effect. When rating on a subjec tive scale of financial success, women were rated as more financially successful than were men, even though the only information provided was a picture. When providing salary amounts using an objective measure, raters gave men higher salaries than th ey gave to women. A woman who makes $40,000 a year is considered successful. Howe ver, a man who earns $40,000 may receive an average rating because the raters are compari ng the man to the standard they have for men, which is higher than the standard they have for women That is to say, even though the man and the woman have the same salary, the value is being rated against a different standard depending upon the sex of the ra tee. These findings support the shifting standards model, because with the subjective ra tings, "for a man to be labeled financially very successful, he had to earn much mo re money than a woman who was similarly labeled" (p. 5).

PAGE 22

15 Biernat and Manis (1994) test ed the shifting standards th eory with more variables including aggressiveness, asse rtiveness, verbal ability, and mathematical ability. The researchers found similar results across all stereotyped variables (Beirnat & Manis, 1994). Recently, the shifting standards mode l was introduced into the industrial/ organizational psychology literature by Mart el and DeSmet (2000). The authors studied gender stereotypes in leadershi p. In order to avoid subjec tive scales that might mask stereotypes, the authors asked participants to rank the ratees. The study was designed for participants to identify the abilities that are expected of a good leader and to discover if people have the same expectations of male and female leaders. Further, the authors wanted to determine if gender stereotypes are the reason why wome n are not promoted into the managerial positions at equal rates as men. Martel and DeSmet found that some behaviors were not subjected to gender stereotyping, but that there were several behaviors that were considered to be more characteristically male or female. For example, behaviors such as mentoring, suppor ting, and rewarding were considered to be more likely performed by women. Behaviors considered characteristic of males were delegating and upward influence. Accordi ng to the shifting standards model, these stereotyped behaviors were evident because an objective scale was used that forced the raters to compare the ratee to everyone, not just the members of the ratee’s gender. The proposed study will test the applicability of the sh ifting standards model to gender stereotypes of citizenship performance. The stereotype that women are more likely than men to perform as organizational ci tizens is expected to increase expectations for women to perform citizenship behaviors. In their meta-analysis, Organ and Ryan (1995) found no evidence that gender related to ratings of OCB (r (1300)=.03, ns ).

PAGE 23

16 Podsakoff, Mackenzie, Paine, and Bachrach (2 000) discussed research ers’ surprise with the null findings from the meta-analysis. Th ey provided examples of characteristics of citizenship performance that share features wi th stereotypes of men and women. Further, most OCB measures are Likert scales (subjectiv e). It is reasonable that raters could be using a different reference group when rati ng men than when rating women, which would mask perceived differences between men and women. Therefore, even though women may be perceived to be greater organizati onal citizens than men, they will not receive higher ratings than will men on subjective scales. For the present study, it is pr edicted that when rating citizenship performance on an objective scale, raters will compare me n and women and will consequently rate women higher because they are expected to be better organizational citizens. However, when using subjective scales, raters are expected to compare women against other women, thereby, hiding the gender stereotype As a consequence, when rating on a subjective scale, there will be no significant difference between men and women on citizenship performance. Figure 1 displa ys the expected results for Hypothesis 1. Hypothesis 1 : When participants rate employees on a subjective citizenship performance scale there will be no significant difference between men and women. However, when participants rate employees using an objectiv e citizenship performance scale, women will be given significantly highe r ratings than will men.

PAGE 24

17 Figure 1. Predicted Results for Hypothesis 1. OCB Standardized Scores Objective Subjective Rating Scale Overall Performance Evaluation In an organizational setting, it is impo rtant to understand the way citizenship behavior may affect overall performance appr aisals and how the relationship may differ across gender. Research studying gender bias in performance evaluation has yielded inconsistent results (Landy & Farr, 1980). St udies have produced results supporting male favored bias, (Hamner et al., 1974), female fa vored bias (Atwater & Van Fleet, 1997), and no bias (Schwab & Grams, 1985). Cons idering that performance is generally measured using subjective scales, the shifti ng standards theory would postulate that men and women are not being rated against each other, but against members of their own group. In other words, women are rated re lative to other women and men relative to other men. According to the shifting sta ndards model, in orde r to explore raters’ stereotypes and biases, an obj ective scale should be used (Biernat et al., 1991). Biernat, Crandall, Young, Kobrynowicz, and Halpin (1998) tested the shifting standards model using peer and self-ratings of officer performance in a leadershiptraining course. For the objective scale, the pa rticipants ranked their groupmates from most capable to least capable. According to the shifting standard s model, rankings are objective because raters must compare the target groups, in this case men and women. Women Men Women Men

PAGE 25

18 For the subjective scale, subjects rated thei r groupmates on a Likert 5-point scale of capability from “excellent” to “needs much improvement.” Men were rated higher than women were on both rankings and the Likert scales. However, when subjects ranked their peers, the difference between men and wo men was greater than when a Likert scale was used. In other words, when raters were forced to compare men and women by ranking them, sex differences were even more pronounced. Further, these findings provide support for the applic ation of the shifting standa rds model in the realm of performance evaluations. These findings also support the hypothesis th at when objective s cales are used to evaluate overall performance, the difference raters perceive betw een men and women are more apparent. An important considerati on, for the proposed study, is how citizenship performance influences overall performance ev aluations. As discussed earlier, both task performance and citizenship performance have been found to contribute to the variance associated with overall ratings of perfor mance (e.g. Allen & Rush, 1998). Therefore, gender stereotypes of citizenship perform ance could subsequently impact overall performance evaluations. As will be described in the following paragraphs, only a few studies (Allen & Rush, 2001; Chen & He ilman, 2001; Lovell, et al., 1999) have attempted to investigate how gender influe nces the relationship between citizenship performance and overall performance evaluations. Chen and Heilman (2001) tested differe nces in performance evaluations and reward recommendations for men and women wh ile manipulating the level of OCB in the ratee performance descriptions provided. Vignettes were used to describe the performance of either a male or a female employee, who either performed OCB, who

PAGE 26

19 chose not to perform OCB, or who only exhibited task behavi ors. When subjects were rating vignettes with only task performan ce information, no significant differences were found between men and women. However, in the two groups wher e the ratee either performed OCB or did not perf orm OCB, men were rated si gnificantly higher than were women on both overall performance evalua tions and reward recommendations. The results also showed that the female ratees were rated significantly lower when they chose not to perform OCB than when only task behaviors were provided, but no differences were found for men. Furthermore, wome n who did perform OCB were not rated significantly higher than were women with no OCB information provided. This was in contrast to men who were rated higher when performing OCB than when no OCB information was provided. In brief, men were positively evaluated for performing OCB and women were given lower evalua tions for not performing OCB. Research has consistently demonstrated that citizenship performance contributes to overall performance apprai sals (e.g., Allen & Rush, 1998). In the Chen and Heilman (2001) study, citizenship perfor mance did not contribute to overall performance ratings the same way for men as for women. The no tion that women engage in citizenship behavior more frequently than do men is predicted to create hi gher expectations for women to perform citizenship behaviors relative to men. As a consequence, women may need to perform more citizensh ip performance in order to be equally recognized for their citizenship performance. In order to understand how the ci tizenship performance ratings contribute to overall performance differen tly for men and for women, both overall and citizenship performances need to be evaluated. Further, Chen and Heilman used subjective ratings to evaluate overall performance. As unc overed by Biernat, Manis, and

PAGE 27

20 Nelson (1991), when objective ratings are use d, a rater’s comparison group is more easily distinguished than when subjective ratings are used. Even though Chen and Heilman found a significant difference between men a nd women on overall performance, if an objective scale had been used the difference may have been greater. Lovell et al. (1999) tested the effect of gender on the relationship between OCB and performance evaluations usi ng resident advisors (RA) in college dorms. A survey of the five dimensions of OCB provided by Or gan (1988) was created by the researchers and then factor analyzed. The factor anal ysis produced a thre e-factor solution of altruism, sportsmanship, and mediation. The experimenters asked the RAs to rate each other on OCB and asked the dorm directors to rate the RAs on overall performance. There was a significant correlation of .38 found between OCB and performance evaluations. In other words, higher levels of OCB were associated with higher scores on performance appraisals. Further, women were given marginally higher OCB ratings than were men (p = .063). However, even t hough women received higher OCB ratings than did men, there were no differences f ound between women and men on overall performance evaluations. Although the patt ern of results concerning overall OCB was not statistically significant, they suggest that the effect of gender on ratings of citizenship performance is worth further investigation. A weakness of Lovell et al.’ s (1999) study is that no task ratings were provided (i.e., the men could have been significantly bett er task performers than were the women). True levels of citizenship pe rformance were also not availa ble. The proposed study is designed to address this limitation by controllin g for task and citizenship performance in a laboratory setting. Further, subjective ra tings were used to measure both OCB and

PAGE 28

21 overall performance. If the experimenters ha d used an objective scale, it is possible that they would have found a significant differen ce between men and wome n. Lastly, peers provided the ratings of citi zenship performance and supervisors provided the overall performance ratings. Consequently, there is no way of knowing if the supervisors and peers perceive the same behavi ors and if the supervisor w ould have given the resident advisors the same ratings on OCB as did the peers. Allen and Rush (2001) provided a third study that looked at gender, OCB, and overall performance evaluations. The author s manipulated levels of task performance and OCB and manipulated the gender of the ratee. Results indicated no significant differences on ratings of overall performa nce or reward recommendations between men and women. However, for both dependent meas ures subjective scales were used. This could be the reason why there were no significant differences found. Perhaps raters were rating women against women and men against men as predicted by the shifting standards theory. In the present study, objective scales will be used in orde r to see how raters compare men and women without masking perceived differences with a subjective scale. Further, participants will evaluate the ratee’ s citizenship performance as well as overall performance, in order to look for a moderating effect of gender. In two of the three studies described in this section (Lovell et al., 1999; Chen & Heilman, 2001), men and women were not equa lly evaluated for performing citizenship behaviors. The current study pl ans to add to this line of re search by using an objective scale as well as a subjective scale to m easure overall performance. In the previous hypothesis, citizenship performance was measur ed using subjective and objective scales. Therefore, the following hypotheses will be divided by the type of scale used to measure

PAGE 29

22 citizenship performance to better understand under what conditions gender differences may emerge. First, when an objective scale is used to measure overall performance, gender is expected to moderate the relati onship between citizensh ip performance and overall performance. The relationship betw een citizenship performance and overall performance is expected to be greater for me n than for women. When we break down the results by the scale used to measure citizenship performance, the natu re of the interaction is expected to vary depending on the scale us ed. When an objective scale is used to measure citizenship performance, the relatio nship between citizenship performance and overall performance is predicted to be weak est for women and the strongest for the men. Figure 2 displays the expected results of Hypothesis 2a. Hypothesis 2a : Gender is predicted to moderate the relationship be tween citizenship performance and overall performance when both are measured using an objective scale. It is predicted that citizenship performe d will be more highly related with overall performance for men than it will for women. The difference for men and women, in the strength of the relation between citizenship and overall performance, is proposed to be greatest when both types of performance are measured using objective scales.

PAGE 30

23 Figure 2. Predicted Results for Hypothesis 2a. Objective (OP) and Objective (CP) Objective Overall Performance Standardized Scores Objective Citizenship Performance Men Women When overall performance is measured us ing an objective scale and citizenship performance is measured using a subjective sc ale, gender is expect ed to moderate the relationship between the two t ypes of performance. However, the difference in the strength of the relationship between overall and citizenship performance for men and for women is not expected to be as great as wh en objective scales are used to measure both types of performance. Figure 3 displays the expected results of Hypothesis 2b. Hypothesis 2b : Gender is predicted to moderate the relationship between citizenship performance and overall performance when citi zenship performance is measured using a subjective scale and overall performance is measured using an objective scale. The relationship between citizenship and overall pe rformance is expected to be stronger for men than it is for women.

PAGE 31

24 Figure 3. Predicted Results for Hypothesis 2b. Objective (OP) and Subjective (CP) Objective Overall Performance Standardized Scores Subjective Citizenship Performance Men Women Next, how gender influences the relati onship between citizenship and overall performance when a subjective measure of overall performance is used will be discussed. When citizenship performance is measured using an objective scale and overall performance is measured using a subjective sc ale, gender is predicted to moderate the relationship between the two t ypes of performance. As predicted with Hypothesis 2b, citizenship performance for men is expected to have a stronger relationship with overall performance than it does for women. Figur e 4 displays the expected results for Hypothesis 3. Hypothesis 3 : When citizenship performance is measured using an objective scale and overall performance is measured using a subjective one, gender is predicted to moderate the relationship. Citizenship performance is predicted to have a stronger relationship with overall performance for men than it does for women. The difference between men and women in this case is not predicted to be as great as when both types of performance are measured using objective scales.

PAGE 32

25 Figure 4. Predicted Results for Hypothesis 3. Objective (OP) and Subjective (CP) Subjective Overall Performance Standardized Scores Objective Citizenship Performance Men Women When both citizenship performance and overall performance are measured using a subjective scale, gender is not expected to impact the relationship between citizenship performance and overall performance. Ther efore, the following research question will be an exploratory one. Exploratory question 1 : Does gender moderate the re lationship between citizenship performance and overall performance when both are measured using subjective scales? Reward Recommendations Also of significance is the question of how citizenship performance ratings relate to reward recommendations. As described earlier, Allen and Rush (1998) used the concept reward recommendations to encomp ass five organizational rewards: salary, promotions, public recognition, high profile project, and opportunities for professional development. Citizenship performance ha s been found to contribute to reward recommendations (e.g., Allen & Rush, 1998), as we ll as to promotions (Hui, et al., 2000), informal rewards (Van Scotter et al., 2000), salary (Allen, 2000), and recommendations

PAGE 33

26 for a fast-track development program (Kik er & Motowidlo, 1999). However, only a few studies (Allen, 2000; Allen & Rush, 2001; Chen & Heilman, 2001) have looked at the influence of gender on the relationship between citizenship performance and organizational rewards. The previously discussed articles by Chen and Heilman (2001) and Allen and Rush (2001) tested the effect of gender on the relationship between OCB and reward recommendations. Even though these two studies used a similar dependent measure, their results were quite diffe rent. Allen and Rush (2001) found no effect of gender on reward recommendations, whereas, Chen and Heilman (2001) found that men were rated significantly higher on reward recommendations than were women. Both of these studies used subjective scales. Chen and Heilman ( 2001) used a different design than did Allen and Rush (2001). Chen and Heilman provided pa rticipants with quantit ative values of the ratees task performance and wrote a paragra ph describing a time when the ratee either performed OCB or decided not to perform OCB. Therefore, th e written, story style of the OCB information may have been more salient th an the task information. The salience of the OCB information may have given it a larger impact on the ratings of overall performance than was the case with the study by Allen and Rush (2001) who provided both task behaviors and OCB together in a vi deotape of job performance. Providing task and citizenship behaviors together seems to be more applicable to an organizational setting. These study design differences may e xplain the discrepant results between the two studies and will be addressed in the follo wing way, task and citizenship performance will be provided together in a performance log in an attempt to capture a more realistic

PAGE 34

27 setting. Further, both subject ive and objective measures will be used, in order to more accurately gauge raters’ per ceptions of men and women (B iernat et al., 1998). Allen (2000) tested the hypot heses that gender moderate s the relationship between OCB and promotions and OCB and salary. In a field study in which self-reports of OCB and promotions were used, Allen found that OCB was correlated with employee salary and number of promotions. More importa ntly, gender moderated the relationship between OCB and promotions, but not OCB a nd salary. The correlation for men between OCBO and promotions was .21 and between OCBI and promotions was .20. For women, neither type of OCB significan tly correlated with promoti ons. Objective measures of promotions were used in this study by aski ng the participants to include the number of promotions they have received to date. Th is study supports the cl aim that women are not rewarded for their citizenship be havior to the same extent as are men and further supports the use of objectives scales to identify these differences One feature of the Allen study that should be noted is that self-ratings were used for OCB, therefore, the rater evaluating OCB did not make the salary or promotion decisions. To better understand how OCB correlates with reward recommendations differently for men than for women, the same rater needs to be used for both OCB and re ward recommendations. This is important because the rater recommending rewards coul d have perceived the employee’s OCB level differently than the rater providing informati on about the employee’s OCB. The present study will ask participants to rate the employee on citizen ship performance and also decide on reward recommendations for the employee. It is predicted that when raters are us ing a subjective scale to measure reward recommendations, there will be no differen ce between men and women, but when using

PAGE 35

28 an objective scale, there will be a significant difference. As with Hypothesis 2, the following hypotheses will be divided based on the scale type used to measure citizenship performance and reward recommendations. First, when reward recommendations and citizenship performance are measured with objective scales, gender is expected to moderate the relationship with the relationship being str onger for men than for women. Further, when both reward recommendations and citizenship performance are measured with objective scales, the diffe rence between men and women is predicted to be greatest. Figure 5 displays the expected results for Hypothesis 4a. Hypothesis 4a : Gender is predicted to moderate the relationship between citizenship performance and reward recommendations when both are measured with objective scales. The relationship between reward recomme ndations and citizenship performance is predicted to be stronger for men than for women. Figure 5. Predicted results for Hypothesis 4a. Objective (RR) and Objective (CP) Objective Reward Recommend. Standardized Scores Objective Citizenship Performance Men Women Next, the effects of gender when reward recommendations are measured using an objective scale and citizenship performance is measured using a subjective scale, will be discussed. Gender is expected to moderate this relationship, however, the difference

PAGE 36

29 between men and women is not expected to be as great as when objective measures are used for both reward recommendations and ci tizenship performance. Figure 6 displays the expected results for Hypothesis 4b. Hypothesis 4b : When an objective scale is used to measure reward recommendations and a subjective scale is used to measure citizen ship performance, gender is predicted to moderate the relationship between reward r ecommendations and citizenship performance. The relationship between citi zenship performance and re ward recommendations is predicted to be greater for men than for women. Figure 6. Predicted Results for Hypothesis 4b. Objective (RR) and Subjective (CP) Objective Reward Recommend. Standardized Scores Subjective Citizenship Performance Men Women Next will be a discussion of the expectations when using a subjective scale of reward recommendations. First, when rewa rd recommendations is measured with a subjective scale and citizenship is measured using an objective scale, gender is predicted to moderate the relationship between reward recommendations and citizenship performance. Men are expected to have a stronger relationship between reward recommendations and citizenship performance than are women, but the difference is not expected to be as great as when the two ar e measured with objective scales. Figure 7 displays the expected results for Hypothesis 5.

PAGE 37

30 Hypothesis 5 : When citizenship performance is measured using an objective scale and reward recommendations are measured using a subjective scale, the relationship between the two is predicted to be moderated by gende r. There is predicted to be a stronger relationship for men than for women. Figure 7. Predicted Results for Hypothesis 5. Subjective (RR) and Objective (CP) Subjective Reward Recommend. Standardized Scores Objective Citizenship Performance Men Women Gender is not expected to moderate the relationship between reward recommendations and citizenship performance when they are measured with subjective scales. However, since these results are unknown, this question will be an exploratory one. Exploratory question 2: Does gender moderate the relationship between citizenship performance and reward recommendations?

PAGE 38

31 Citizenship Performance Dimensions Of the three dimensions proposed by Co leman and Borman (2000), it seems likely that the personal support dimension may be the strongest link to gender stereotypes. The citizenship behaviors included in the persona l support dimension a ppear to have the strongest similarities to gender stereotype s that describe women as unselfish and supportive. If personal support is more repr esentative of gender stereotypes than are initiating structure or organizational support, it will likely provide the largest rating differences between male and fe male citizenship behaviors. Lovell et al. (1999) provided support fo r the hypothesis that gender has more impact on the personal support dimension than the other two dimensions. The researchers found that of the three dimensions (altruism, sportsmanship, and mediation) identified in the study, the altruism dimens ion was the only one that demonstrated a significant difference between ratings of men and women. Because of the lack of research on the di fferential influence of dimensions, this issue will be investigated in an exploratory manner. Exploratory question 3: Is there a larger difference in ratings of citizenship performance between men and women when th e personal support dimension is measured rather than initiating struct ure and organizational support?

PAGE 39

32 Chapter 2 Method Participants Two hundred and seventy-two University of South Florida students received one extra credit point for participating in this study. Of those 272, 254 (93%) answered the manipulation check correctly. The following de mographics are based on the sample of 254 who passed the manipulation check. Most of the respondents were females (N = 210, 83%) between the ages of 17 and 23 years (N = 194, 76%). The majority were white (N = 150, 59%), psychology majors (N = 138, 55%), and were juniors or higher (N = 165, 65%). Almost all (N = 244, 96%) respondents had work experience and 43% (N = 110) had worked as supervisors. Design A 2x2 between subjects factorial design was applied. The independent variables were ratee gender and citizenship performan ce scale type (objecti ve or subjective). Participants were randomly assigned to one of the following conditions, male target rated on a subjective Citizenship Performance (CP) scale (N= 64), male target rated on an objective CP scale (N = 63), female target ra ted on a subjective CP scale (N = 63), or female target rated on an objective CP scale (N = 64).

PAGE 40

33 Table 1. Distribution of Part icipants into 4 Conditions. Scale Type Gender Objective Subjective Female N = 64 13 Men 51 Women N = 63 9 Men 54 Women Male N = 63 9 Men 54 Women N = 64 13 Men 51 Women Procedure Participants completed the materials al one or in groups no larger than five. Participants began by signing an inform ed consent form (Appendix A). Next, participants were asked to read a cover stor y (Appendix B). The st ory explained that we have created a professor evaluation and reward system and that students are needed to test the new items. They were given a perfor mance log (Appendix C) of either a male or a female professor. It was explained in the instructions and verbal ly by the experimenter that these statements were collected last se mester from students and other members of the psychology department. Then, participants were asked to evalua te the professor’s performance and recommend rewards. The gender manipulation was evident throughout the descriptions by the frequent use of gende r specific pronouns. The development and piloting of these materials will be described in the next section. After reading the performance log, part icipants filled out an evaluation of citizenship performance. Citizenship perfor mance scale type was a between-subjects variable, with half the partic ipants filling out an objectiv e scale and half filling out a subjective scale.

PAGE 41

34 The citizenship performance evaluation was followed by either an overall performance evaluation or reward recommendati ons. The order of presentation of overall performance and reward recommendations was counterbalanced with half the subjects receiving overall performance fi rst and half receiving reward recommendations first. The remaining scale (reward recommendations or performance evaluation) was presented as the final measure. The order of the s ubjective and objective scales, for reward recommendations and overall performance, was also counterbalanced. Statistical analyses were done to test for order effects and are discussed in th e results section. Finally, participants were asked to provide demographic, employment, and education information about themselves. Fo llowing the demographic information was a one-item manipulation check aski ng participants to indicate the gender of the professor they evaluated (Appendix G). After the partic ipants had completed the materials, they were given written informati on that debriefed them about the true intent of the study (Appendix H). Materials In order to identify a job that students pe rceive to be occupied by an equivalent number of men and women, a pilot study was conducted. It was important to use a job not preconceived to be domina ted by one sex and therefore gender stereotyped. Fortytwo University of South Flor ida undergraduate psychology stude nts took part in this pilot study. Participants were provided with a li st of 14 occupations and were asked to indicate “What percentage of men and women do you think are employed in these occupations?” Subjects filled in the percenta ge of men they believed occupied each of the 14 jobs and the corresponding percentage of women they felt occupied the 14 jobs. A

PAGE 42

35 t-test was used to test wh ether the perceived percenta ges of men and women were significantly different. Nine out of the 14 jobs were found to be significantly different, with either men or women thought to occupy a larger percentage of the positions in that field. Table 2 displays the findi ngs including the means, standa rd deviations, and t-values for all 14 comparisons. Of the five jobs that were not significantly different, three of them were professors incl uding a sociology professor (t (42) = -.513, n.s. ), a psychology professor (t (42) = .16, n.s. ), and a communicatio ns professor (t (42) = -1.07, n.s. ). A psychology professor was chosen because ps ychology students were the planned study participants and would therefor e be familiar with this job.

PAGE 43

36 Table 2. Perceived Number of Males and Females in 14 Occupations Men Women T-Test Occupation M SD M SD T p English Prof 78.60 16.38 21.40 16.38 11.45** p < .000 Video Store Clerk 55.30 10.63 44.70 10.63 3.27** p = .002 Psychology Prof 50.42 17.56 49.58 17.56 .156 p = .877 Fast Food Worker 51.81 7.93 48.19 7.93 1.50 p = .141 Communications Prof 47.72 14.00 52.28 14.00 -1.07 p = .292 Nurse 28.60 23.77 71.40 23.77 -5.90** p < .000 Real Estate Agent 42.49 14.02 57.52 14.02 -3.51** p = .001 Sociology Prof 48.74 16.04 51.26 16.04 -.513 p = .610 Manager 62.05 12.18 37.95 12.18 6.49** p < .000 English Professor 38.61 14.45 61.40 14.45 -5.17** p < .000 Veterinarian 55.14 13.14 44.86 13.14 2.57* p = .014 Bartender 59.70 16.00 40.30 16.00 3.99** p < .000 Social Worker 27.47 12.84 72.53 12.84 -11.51** p < .000 Bus Driver 51.80 19.38 48.21 19.38 .606 p = .548 N = 43, *significant at .05, **significant at .01 Participants were provided with the name of the professor, either Michelle Smith or Michael Smith, and informed that he/she is a psychology professor. Then they received the performance log with 19 statements about the professor (Appendix B). The list included 11 task performance and 9 citi zenship performance comments. The task performance statements were obtained from Sauser, Evans, and Champion’s (1979) critical incidents of professors. Sauser et al. used undergraduate stud ents to develop a list of 251 critical incidents (samples of good or bad performance) of college professors teaching performance. All incidents were c oded on their perceived level of effective

PAGE 44

37 teacher performance. The items chosen for th is study were appropriate for the job of a professor and were average in effectiv eness (4 to 7 on an 11 point scale). Citizenship performance statements were based on Borman, Buck, Hanson, Motowidlo, Stark, and Drasgow (2001). Bo rman et al. developed 124 citizenship performance statements that corresponded to one of Coleman and Borman’s (2000) three dimensions (personal support, organizational support, and conscien tious initiative). These statements were evaluated by two sets of raters, 37 officers in the Air Force and 26 employees of Personnel Decisions Research Inst itute’s (PDRI). Both groups sorted the statements into the three dimensions a nd provided effectiveness ratings for each statement. The effectiveness ratings were a way to identify the effective and ineffective citizenship behaviors. The dimensions were correctly sorted into the three dimensions with 90 and 96 percent agreemen t for the two groups, respectiv ely. Effectiveness scores were given on a 4-point scale. Raters agreed within .5 points with 80 and 96 percent agreement for the two groups. Seventeen items were taken from the 124 behavioral statements created by Borman et al. All 17 it ems had an effectiveness rating of three. Six items were taken from both the pe rsonal support and or ganizational support dimensions, and five items from th e conscientious initiative dimension. A pilot study was conducted of the 35 ( 17 citizenship and 18 task performance) performance statements in order to verify th at they could consistently be identified as task and citizenship performance. Ten gradua te students (5 females and 5 males) in the Industrial/Organizational Ph.D. program at the University of South Florida participated. Students were provided with de finitions of task and citizenship performance and were asked to sort the behavioral statements into ei ther task or citizenship performance. Half

PAGE 45

38 of the students (3 females and 2 males) receive d female performance logs and half of the students (3 males and 2 females) received ma le performance logs, in order to balance ratee gender effects. The graduate student s were able to sort most of the items successfully. The final list for this study in cluded 10 task and 9 citizenship performance statements. Of the ten task performance stat ements chosen, six of these statements were sorted with 100% accuracy a nd four with 90% accuracy. Of the nine citizenship performance statements chosen (three from each dimension), seven items were sorted with 100% accuracy and two items (1 organization support dimension and 1 conscientious initiative dimension) we re sorted with 80% accuracy. The final profile developed is that of an average task performer, but a slightly above average citizenship performer. Prior to data collection, another pilot study was conducted in order to test the cr edibility of the cove r story and the flow of the materials. The results of this last pilot study will be discussed in the results section. Measures Citizenship Performance. Nine subjective citizenship performance items were developed for the current study (Appendix D). These items were created from Coleman and Borman’s (2000) taxonomy of three dimensions, described earlier. Each dimension emphasized three factors in their definiti on. The personal support dimension included helping, cooperating, and courtesy. The organizational support dimension included representing, loyalty, and compliance. Fina lly, the conscientious initiative dimension included persistence, initiative, and self-devel opment. These factors were used to create items for the citizenship performance scale. The intent was to deve lop a composite score of subjective citizenship pe rformance and an average dimension score for the three

PAGE 46

39 dimensions. The resulting coefficient alpha for the overall scale was .77. The scale was factor analyzed and the three dimension sc ales all produced coefficient alphas of .65. The factor analysis is discussed further in the results section. For the objective measure of citizenship performance, participants were asked to give the professor a letter gr ade, a percentage score from 0 to 100, and to indicate what they would rank the professor if they were comparing him/her to 9 other professors. Biernat (1995) claims that, “the re is general agreement in our culture that a letter grade should be based on meeting some unwavering criteria for performance and that an A means an A, regardless of who has receive d it.” (p. 93). Along the same lines, a percentage is expected to have the same meaning for all ratees. These measures are common and familiar, especially to a college student populati on, and is expected to have universal meaning. Finally, because participants were asked to rank the professor, they directly compared men and women. The inte rnal consistency of the objective scale was = .83. Overall Performance. The following five items were used as a subjective measure of overall performance (Appendix E) : “This professor makes an important contribution to the university,” “This professor is extremely valuable to the University of South Florida,” “This professor would be extr emely costly for the University of South Florida to replace,” “This professo r is a vital part of the Univ ersity of South Florida,” and “This professor is indispensabl e to the University of South Florida.” These five items were developed by Allen and Rush (2001) to m easure overall performance of professors. Participants responded on a 5point scale ranging from st rongly disagree to strongly

PAGE 47

40 agree. Responses from Allen and Rush’s st udy resulted in a high internal reliability ( = .92). The present study resulted in an internal reliability of .86. The same items used as objective measures of citizenship performance, letter grade, a percentage score, and ranking, were used as objective measures of overall performance. The objective overall performan ce scale resulted in a coefficient alpha of .86. Reward Recommendations. Allen and Rush (1998) developed a five-item measure of reward recommendations, which dem onstrated high internal reliability (alpha = .90). Three out of the five items were chosen for the subjective measure of reward recommendations (Appendix F) because of their a pplicability to the job of a professor. Participants were asked to indicate the exte nt that they would “recommend the professor for each of the following rewards”: a promoti on to a more prestigious teaching position, a teaching award, or a salary bonus. Participants provided their answers on a 5-point scale from “would definitely not recommend” to “would recommend wit hout reservation.” The coefficient alpha obtained for the present study was .82. For the objective measure of reward recommendations, three items were used. Participants were asked to indicate whic h salary bonus they would recommend with 6 choices ranging from $0 to $500. Next, partic ipants were asked if 10 professors were competing for a more prestigious teaching position and a teaching award where they would rank this professor. Participants we re given 10 choices ranging from first (most deserving) to tenth (least dese rving). The resulting internal consistency for this scale was = .80.

PAGE 48

41 Chapter 3 Results Pilot Study The final pilot study was conducted to dete rmine the credibility of the materials and the time needed to complete the 8 to 9 paged questionnaire. Twelve undergraduate psychology students at the Univ ersity of South Florida we re asked to complete the questionnaires. After the que stionnaire was completed, participants were asked three questions about the materials. First, they were asked if there was anything in the questionnaire that they did not understand. All participants responded that the materials were clear. Next, the respondents were asked about the believability of the questionnaire and all the participants agreed th at the materials were realistic. Finally, the students were asked what th eir reference point was in making the ratings on the subjective and then on the objective scales. The common response was that they were comparing the professor in th e materials to the best professor they have had and that they did not change their refere nce point from the subjective to the objective scale. The pilot study provided support for th e materials credibility and helped determine the length of time (10 to 15 minutes) n eeded to complete the questionnaire. Manipulation Check The manipulation check consisted of one question, “What was the gender of the professor that you just evaluated? ” Participants were instructed not to flip back to answer this question and were given the choice to circle either a male or a female professor. Of the 272 participants, 93% (254) correctly identi fied the gender of the professor. The 18

PAGE 49

42 participants who failed to correctly identify th e gender of the professor were not used in further analyses. Order Effects The first step in analyzing the data was converting them into z-scores for a standard unit of comparison across different scales (subjective and objective). There were two types of scales that could cause orde r effects in this study. The first was the presentation order of reward recommendations and overall performance. T-tests were run to determine if there was a significant diffe rence on any of the four dependent measures between those who had reward recommenda tions first and those who had overall performance first. There were no significant differences found for order of performance and rewards on subjective performance ratings (t (252) = .95, ns ), subjective reward recommendations (t (252) = 1.47, ns ), or objective reward recommendations (t (252) = 1.68, ns ). There was a significant differen ce found on objective performance ratings (t (252) = 2.45, p < .025). When participants rated the professor on ove rall performance before reward recommendations, they rated objective performance significantly higher (M = .13, SD = .84) than when they rated the professor on reward recommendations first (M = -.13, SD = .91). The significant order effect could potent ially affect Hypothesis 2a and 2b, which both have objective overall performance as the dependent variab le. Hypothesis 2a involves participants who evaluated the citizen ship performance of the professor with an objective measure and Hypothesis 2b involves pa rticipants who evalua ted the citizenship performance of the professor w ith a subjective measure. In other words, the independent variable for Hypothesis 2a is objective citizenship performance ratings and the

PAGE 50

43 independent variable for Hypot hesis 2b is subjective citizen ship performance ratings. A t-test was computed to check for order e ffects on the dependent measure (objective overall performance). There were no signi ficant differences found that would affect Hypothesis 2a (objective CP; t (125) = .51, ns ). However, when a t-test was computed to check for order effects that would affect H ypothesis 2b (subjective CP), it was significant (t (125) = 2.90, p < .01). Therefore, the order of outco me variables for participants who filled out a subjective evaluation of citizenship performance impacted their ratings on the objective measure of overall perf ormance. Participants who rated the profe ssor’s overall performance first gave the professor highe r ratings on objective overall performance (M = .22, SD = .71) than did participants who r ecommended rewards before evaluating performance (M = -.22, SD = .97). As a consequence, when Hypothesis 2b was evaluated, order of overall performance and reward recommendations was entered first (step 1) to remove the variance at tributable to the order effect. T-tests were also run to test for order effects of subjective and objective scales. There were no significant diffe rences found for order of subj ective and objective scales on objective overall performance ratings (t (252) = .62, ns ), subjective overall performance (t (252) = .32, ns ), objective reward recommendations (t (252) = -1.28, ns ) or subjective reward recommendations (t (252) = -1.72, ns ). Rater Gender Effects One possible influence on raters’ stereotype s is the gender of the rater. Although rater gender effects on performance evaluations are generally considered to be minimal (Pulakos, White, Oppler, & Borman, 1989), there is still the possibility that rater gender could influence the results. Analysis of vari ance was used to test for gender differences

PAGE 51

44 on ratings of citizenship performance, overa ll performance, and reward recommendations by testing for an interaction between rater gender and the ra tee gender. There were no significant interactions found for ra tings on citizenship performance (F (1,250) = .16, ns ), objective overall performance (F (1,250) = .08, ns ), subjective overall performance (F (1,250) = .11, ns ), objective reward recommendations (F (1,250) = .04, ns ) or subjective reward recommendations (F (1,250) = .13, ns ). T-tests were also run to test for main effects of rater gender on ratings of citizenship performance, overall performance, and reward recommendations. There were no significant differences found between th e ratings provided by male and female participants on citizen ship performance (t (252) = -.26, ns ), overall performance (t(252) = 1.47, ns ), or reward recommendations (t (252) = -1.27, ns ). Descriptive statistics Tables 3 and 4 display the means and standard deviations for the outcome variables separated by scale type and professor gender. In Table 3, the values are given in z-scores and in Table 4 the values are give n in raw scores. The means are provided for each item on the objective scales because the items were measured with different scales and the objective scale means cannot be comput ed without converting the values first to z-scores.

PAGE 52

45 Table 3. Descriptive Statistics By Ge nder and Scale Type Using Z-Scores. Gender of Professor Female Male Measures M SD M SD Subjective Scales Citizenship Performance .0006 .59 -.0006 .60 Overall Performance .010 .89 -.010 .71 Reward Recommendations .051 .88 -.051 .84 Objective Scales Citizenship Performance -.049 .93 .050 .80 Overall Performance .0029 .84 -.0029 .92 Reward Recommendations -.0024 .84 .0024 .85 Total Citizenship Performance -.025 .78 .025 .71 Overall Performance .0063 .79 -.0063 .72 Reward Recommendations .024 .81 -.024 .79 Note: Values are in z-scores. Table 4. Descriptive Statistics By Gender and Scale Type Using Raw Data. Gender of Professor Female Male Measures M SD M SD Subjective Scales Citizenship Performance 4.14 .42 4.14 .42 Overall Performance 3.61 .75 3.59 .59 Reward Recommendations 3.54 .85 3.44 .81 Objective Scales Citizenship Performance Letter Grade (A+ to F, 1 to 13) 3.31 1.65 3.41 1.66 Citizenship Performance Percentage score 86.25 8.82 88.29 6.22 Citizenship Performance Ranking (1 to 10, 1 = the best) 3.56 1.63 3.41 1.59 Overall Performance Letter Grade(A+ to F, 1 to 13) 4.11 1.85 4.20 1.97 Overall Performance Percentage score 84.17 8.65 84.60 10.53 Overall Performance Ranking(1 to 10, 1=the best) 4.13 1.88 4.17 1.96 Reward Recomm Ranking teacher award (1 to 10, 1 = the best) 4.37 1.89 4.21 2.03 Reward Recomm Salary bonus 261.42 118.22 255.18 120.49 Reward Recomm. Ranking promotion(1 to 10, 1=the best) 4.30 1.94 4.33 2.04

PAGE 53

46 Table 5 contains the zero order correlations between objective citizenship performance and the dependent measures, and Table 6 cont ains the zero order correlations between subjective citizenship performance and the dependent measures. Table 5. Zero-Order Correlation Coefficients Among Variables with Objective Citizenship Performance. Variable 1 2 3 4 5 6 7 1. Prof Gender 2. Rater Gender -.08 3. Objective Citizenship Performance .06 -.03 4. Objective Overall Performance -.02 .08 .49** 5. Subjective Overall Performance -.04 .00 .44** .59** 6. Objective Reward Recomm. -.05 .05 .43** .70** .66** 7. Subjective Reward Recomm. -.11 .06 .41** .65** .66** .73** Mean 0 0 3.58 0 3.47 SD .87 .8 9 .64 .84 .82 Note: Gender was dummy coded (females = 0, males = 1). Means for the objective scales are in z-scores and th e means for the subjective scales are based on the raw values. = p < .05, ** = p < .01. Table 6. Zero-Order Correlation Coefficients Among Variables with Subjective Citizenship Performance. Variable 1 2 3 4 5 6 7 1. Prof Gender 2. Rater Gender .08 3. Subjective Citizenship Performance .00 -.18* 4. Objective Overall Performance .01 .09 .44** 5. Subjective Overall Performance .01 -.13 .55** .61** 6. Objective Reward Recomm. .05 .11 .40** .81** .53** 7. Subjective Reward Recomm. -.01 .07 .43** .76** .64** .76** Mean 4.14 0 3.62 0 3.51 SD .42 .8 7 .68 .85 .84 Note: Gender was dummy coded (females = 0, males = 1). Means for the objective scales are in z-scores and th e means for the subjective scales are based on the raw values. = p < .05, ** = p < .01.

PAGE 54

47 Hypothesis testing For each hypothesis, the first step in an alyzing the data was converting them into z-scores. Hypothesis 1 was tested with a 2 x 2 analysis of variance. The dependent variable was citizenship performance scor es. The two independent between-subject variables were gender and scale type. This first hypothesis pr edicted that there would be an interaction between scale type and gender on ratings of citizenship performance. More specifically, it was predicted that fema les would receive higher objective ratings of citizenship performance than males and th at no differences would be found between males and females on the subjective scales. Ta bles 3 and 4, displayed earlier, have the means for both types of citizenship perfor mance by gender. The interaction between gender and scale type was not significant (F (1, 250) = .29, ns ). Additionally there were no main effects for scale type (F (1, 250) = .00, ns ) or gender (F (1, 250) = .275, ns ). Therefore, no support was found for Hypothesis 1. Hierarchical multiple regression was used to test the remaining hypotheses. At step one, citizenship performance scores were entered. At step two, gender was entered into the equation. At the fi nal step, the intera ction term was entered. For significant interactions, Aiken and West’s (1991) proced ure for plotting and tes ting slopes was used. Table 7 displays the correlations between objective citizenship performance, overall performance, and reward recommendations, sepa rated by gender. Table 7 can be used to help interpret the findings for Hypotheses 2a, 2b, 4a, and 4b. Table 8 displays the correlations between subjective citizenshi p performance, overall performance, and reward recommendations, separated by gender. Th is table will assist in interpretation of Hypotheses 3 and 5 and expl oratory questions 1 and 2.

PAGE 55

48Table 7. Zero-Order Correlation Coefficients Among Variables with Objective Citizenship Performance, Split by Professor Gender. Variable 1 2 3 4 5 Means (SD) 1. Objective Citizenship Performance -.05 (.93) 2. Objective Overall Performance .51** .02 (.84) 3. Subjective Overall Performance .46** .63** 3.60 (.71) 4. Objective Reward Recommendations .49** .75** .70** .04 (.80) Female Prof 5. Subjective Reward Recommendations .49** .62** .64** .71** 3.56 (.87) 1. Objective Citizenship Performance .05 (.80) 2. Objective Overall Performance .49** -.02 (.95) 3. Subjective Overall Performance .43** .55** 3.55 (.56) 4. Objective Reward Recommendations .38** .67** .63** -.04 (.88) Male Prof 5. Subjective Reward Recommendations .34* .69** .70** .75** 3.38 (.77) Note: Means for the objectiv e scales are in z-scores and the means for the subjective scales are based on the raw values. = p < .05, ** = p < .01. Table 8. Zero-Order Correlation Coefficients Among Variables with Subjective Citizenship Performance, Split by Professor Gender. Variable 1 2 3 4 5 Mean (SD) 1. Subjective Citizenship Performance 4.14 (.42) 2. Objective Overall Performance .50** -.01 (.85) 3. Subjective Overall Performance .60** .70** 3.61 (.75) 4. Objective Reward Recommendations .39** .84** .56** -.04 (.88) Female Prof 5. Subjective Reward Recommendations .45** .80** .70** .80** 3.51 (.83) 1. Subjective Citizenship Performance 4.14 (.42) 2. Objective Overall Performance .39** .01 (.90) 3. Subjective Overall Performance .49** .53** 3.63 .62 4. Objective Reward Recommendations .40** .78** .49** .04 (.83) Male Prof 5. Subjective Reward Recommendations .41* .72** .59** .71** 3.50 (.85) Note: Means for the objectiv e scales are in z-scores and the means for the subjective scales are based on the raw values. = p < .05, ** = p < .01.

PAGE 56

49 Hypothesis 2a stated that when overall performance and citizenship performance were measured with an objective scale, the strength of the relationship between citizenship and overall performance would be stronger for men than for women. The regression results for Hypothesis 2a are disp layed in Table 9. Th e interaction between citizenship performance and gender was not significant ( = .07, t (125) = .71, ns ) providing no support for Hypothesis 2a. The onl y variable that cont ributed a significant amount of variance in expl aining objective overall perfor mance scores was objective citizenship performance (step 3; = .45, t (125) = 4.34, p < .05). Citizenship performance accounted for 24% of the varian ce in objective overall performance at step 1. The main effect of gender was not significant (step 3; = -.05, t (125) = -.66, ns ). Table 9. Hierarchical Regression Analysis with Obj ective Overall Performance as the Dependent Variable. Standardized regression weights Objective Overall Performance Step 1 Step 2 Step 3 Hypothesis 2a Objective Citizenship Performance .49** .50** .45** Gender -.05 -.05 Interaction .07 R2 at each step .24 .25 .25 R2 change .003 .003 F 39.95** .44 .50 Hypothesis 2b Step 1 St ep 2 Step 3 Step 4 Presentation order of reward recomm. and overall performance -.25** -.29** -.29** -.29** Subjective Citizenship Performance .46** .46** .52** Gender .01 .01 Interaction -.07 R2 at each step .06 .27 .26 .26 R2 change .21 .00 .002 F 8.40** 36.59** .02 .42 Note: Gender and Presentation order were dummy coded (female/overall performance first = 0 and male/reward recommendations first = 1). = p < .05, ** = p < .01. Hypothesis 2b proposed that when citi zenship performance was measured with an objective scale and overall performance was measured with a subjective scale, men

PAGE 57

50 would have a stronger relationship between their citizenship performance and overall performance than would women. Table 9 show s the regression results for Hypothesis 2b. Because presentation order of overall perf ormance and reward recommendations had a significant impact on objective overall performan ce ratings when a subjective citizenship performance scale was used, the variable, order of the dependent measure, was entered first. Presentation order wa s dummy coded with particip ants receiving performance evaluation first coded “0” a nd participants receiving rewa rd recommendation first coded “1.” Presentation order was significant (step 4; = -.29, t (125) = -3.78, p < .05) and accounted for 6% (F (1, 123) = 8.40, p < .05) of the variance in objective overall performance at step 1. The interaction between ge nder and citizenship perfor mance was not significant ( = -.07, t (125) = -.64, ns ). Subjective citizenship perf ormance was significant (step 4; = .52, t (125) = 4.66, p < .05) and it accounted for 21% (F (1, 123) = 36.59, p < .05) of the variance in objective ov erall performance ratings, beyond the order effect. The main effect of gender (step 4; = .01, t (125) = .15, ns ) was not significant. Therefore, no support was found for Hypothesis 2b. Hypothesis 3 predicted that when citizen ship performance was measured using an objective scale and overall performance was measured using a subjective scale, the relationship between citizensh ip performance and overall performance would be greater for males than for females. Table 10 displays the results for Hypothesis 3. The interaction term was not significant ( = -.05, t (125) = -.48, ns ). The only variable that was a significant predictor of subjective overal l performance was objective citizenship performance (step 3; = .44, t (125) = 4.52, p < .05) and it accounted 20% of the variance

PAGE 58

51 in subjective overall performance (step 1; F (1, 123) = 30.62, p < .05). The gender term was not significant (step 3; = -.06, t (125) = -.79, ns ). Consequently, the results did not support Hypothesis 3. Table 10. Hierarchical Regression Analysis with Subjective Overall Performance as the Dependent Variable. Standardized regression weights Subjective Overall Performance Step 1 Step 2 Step 3 Hypothesis 3 Objective Citizenship Performance .44** .44** .48** Gender -.06 -.06 Interaction -.05 R2 at each step .20 .20 .20 R2 change .004 .002 F 30.62** .63 .23 Note: Gender was dummy coded (females = 0, males = 1). = p < .05, ** = p < .01, = .15. The remaining hypotheses concern the interaction of gender and citizenship performance on reward recommendations. Hypothesis 4a predicted that when both citizenship performance and reward recomme ndations were measured using an objective scale, the correlation between citizenship performance and reward recommendations would be higher for men than for women. Ta ble 11 displays the regression results for Hypothesis 4a. The beta weights were not significant for the interaction term ( = -.007, t (125) = -.07, ns ) or for gender (step 3; = -.07, t (125) = -.88, ns ). Citizenship performance was significant at all stages (step 3; = .44, p < .05) and accounted for 18% (step 1; F (1, 125) = 27.77, p <.05) of the variance in objective reward recommendations. Therefore, there was no support found for Hypothesis 4a.

PAGE 59

52Table 11. Hierarchical Regression Analysis with Objective Reward Recommendations as the Dependent Variable. Standardized regression weights Objective Reward Recommendations Step 1 Step 2 Step 3 Hypothesis 4a Objective Citizenship Performance .43** .43** .44** Gender -.07 -.07 Interaction -.007 R2 at each step .18 .19 .19 R2 change .005 .00 F 27.77** .78 .004 Hypothesis 4b Subjective Citizenship Performance .40** .40** .41** Gender .05 .05 Interaction -.02 R2 at each step .16 .16 .16 R2 change .003 .00 F 23.28** .37 .02 Note: Gender was dummy coded (females = 0, males = 1). = p < .05, ** = p < .01. Hypothesis 4b stated that when citize nship performance was measured with a subjective scale and rewards were measur ed with an objective scale, a stronger relationship between reward recommendations and citizenship pe rformance would be expected for men than for women. As seen in Table 11, the beta coefficient did not support a significant interaction ( = -.02, t (125) = -.15, ns ). The gender main effect was similarly not significant (step 3; = .05, t (125) = .61, ns ). Subjective citizenship performance did predict objective reward recommendations (step 3; = .41, t (125) = 3.45, p < .05) and accounted for 16% of the variance in objective reward recommendations. The final hypothesis concerns subjectiv e reward recommendations. Hypothesis 5 predicted that when objective citizensh ip performance and subjective reward recommendations were used, there would be a stronger relationship for men between citizenship performance and reward recomm endations than there would be for women.

PAGE 60

53 Table 12 displays the regre ssion results for Hypothesis 5. The beta weight did not support a significant interaction between citizenship performance and gender ( = -.09, t (125) = -.85, ns ) nor was the effect of ge nder significant (step 3; = -.13, t (125) = -1.66, p = .10). With the addition of gender at step two, the mode l explained an additional 2% (F (1, 125) = 2.78, p = .10) of the variance in subjec tive reward recommendations, which was marginally significant at an alpha level of .10 (Females: M = .09, SD = .91, Males = M = -.10, SD = .80). Subjective citizenship perfo rmance did predict objective measures of reward recommendations (step 3; = .48, t (125) = 4.49, p < .05) and accounted for 17% (step 1; F (1, 125) = 25.66, p < .05) of the variance in objective reward recommendations. However, no support was found for Hypothesis 5. Table 12. Hierarchical Regression Analysis with Subjective Reward Recommendations as the Dependent Variable. Standardized regression weights Subjective Reward Recommendations Step 1 Step 2 Step 3 Hypothesis 5 Objective Citizenship Performance .41** .42** .48** Gender .-.14 -.13 Interaction -.09 R2 at each step .17 .19 .19 R2 change .02 .005 F 25.66** 2.78 .72 Note: Gender was dummy coded (females = 0, males = 1). = p < .05, ** = p < .01, = p = .10. Exploratory Questions The first two exploratory questions were analyzed using hierarchical multiple regression. The first explorat ory question concerned the re lationship between subjective measures of both citizenship and overall performance and how the relationship differed for men and women. The result s are displayed in Table 13. The beta weight for the interaction was not significant ( = -.17, t (125) = -1.57, p = .15) and neither was the beta weight for gender (step 3; = .01, t (125) = .15, ns ). The incremental validity from the

PAGE 61

54 addition of the interactio n term was 1%, which was not a significant addition (F (1, 123) = 2.47, p = .15. Citizenship performance was a significant predictor (step 3; = .67, t (125) = 6.22, p < .05) and it accounted for 30% (step 1; F (1, 125) = 53.19, p <.05) of the variance in subjective overall performance. Table 13. Hierarchical Regression Analysis with Subjective Measures of Both Overall Performance and Citizenship Performance. Standardized regression weights Subjective Overall Performance Step 1 Step 2 Step 3 Exploratory question 1 Subjective Citizenship Performance .55** .55** .67** Gender .01 .01 Interaction -.17 R2 at each step .30 .30 .31 R2 change .00 .01 F 53.19** .02 2.47 Note: Gender was dummy coded (females = 0, males = 1). = p < .05, ** = p < .01, = p = .10. The second exploratory question investig ated differences for men and women in the relationship between subjec tive measures of reward recommendations and citizenship performance. Table 11 displays the results. Both the interaction term and the gender variable were not significant ( = -.03, t (125) = -.25, ns & = -.01, t (125) = -.10, ns respectively). Subjective citizenship performa nce was a significant predictor of subjective reward recommendations (step 3; = .45, t (125) = 3.88, p < .05) and accounted for 19% of the variance (F (1,125) = 28.53, p < .05).

PAGE 62

55Table 14. Hierarchical Regression Analysis with Subjective Measures of Both Reward Recommendations and Citizenship Performance. Standardized regression weights Subjective Overall Performance Step 1 Step 2 Step 3 Exploratory Question 2 Subjective Citizenship Performance .43** .43** .45** Gender -.01 -.01 Interaction -.03 R2 at each step .19 .19 .19 R2 change .00 .00 F 28.53** .01 .06 Note: Gender was dummy coded (females = 0, males = 1). = p < .05, ** = p < .01. The third exploratory qu estion investigated if men and women are rated differently on the three types of citizenshi p performance. If differences were found on the three types of citizenship performance, th en the magnitude of those differences would be contrasted with three t-tests. Alth ough there were no significant differences found between men and women on the composite subjective measure of citizenship performance (t (125) = .01, ns ; Males: M = .00, SD = .59; Females: M = .00, SD = .60), is it possible that differences could be found on one of the dimensions. A principle factor analysis was done to verify that the citizenship performance dimensions (personal support, organizational su pport, and conscientious initiative) do, in fact, load on three separate factors. The f actor analysis was followed by a promax oblique rotation with a forced three-factor soluti on. The eigenvalues before rotation were 3.23, 1.22, 1.04 and accounted for 61% (36%, 14% and 12%, respectively) percent of the variance. The scree plot (see Fi gure 8 for details) appears to support a one to four factor solution, with the line leveling out after th e fifth eigenvalue. Af ter the rotation, the eigenvalues became 2.10, 2.02, and 1.88.

PAGE 63

56 Figure 8. Scree Plot of Eigenvalues for F actor Analysis of Citizenship Performance Dimensions. 4 3 Eigenvalues 2 1 0 1 2 3 4 5 6 7 Factors Table 15 displays the pattern matrix of the factor loadings after the promax rotation. With a sample size of approximat ely 140, the criterion for significant factor loadings was set at .43 (Stevens, 2002). A ll three organizational support items loaded on factor 1. All three conscienti ous initiative items loaded on factor 2. Only two of the three personal support items loaded on factor 3. The third personal support item, “Dr. Smith is cooperative when working with co lleagues and students” did not sufficiently load on the personal support f actor (.27). Therefore, this item was dropped leaving only two items in the personal support factor. Th e first personal support item, “Dr. Smith is helpful to co-workers and st udents,” had a high loading on bo th conscientious initiative (.42) and personal support (.50). Because the personal support loading was larger, met the loading criterion, and was in accordance with past research (Coleman & Borman, 2000), this item was included in the personal support factor. The final scales all yielded coefficient alphas of .65.

PAGE 64

57Table 15. Promax Oblique Factor Rotation Pattern Matrix. Items Factor 1 Factor 2 Factor 3 5. Represents school well to outsiders (OS2) .65 .02 .00 6. Complies with school rules and procedures (OS3) .62 -.08 -.06 4. Loyalty to school (OS1) .59 .08 .10 3. Cooperation with co-workers and students (PS3) .37 .15 .27 7. Persistence and extra effort given when needed (CI1) .07 .76 -.18 9. Dedicated to self development (CI3) -.07 .56 .04 8. Takes initiative to get things done (CI2) .07 .46 .03 2. Courteous with co-workers and students (PS2) .05 -.16 .92 1. Helpful with co-workers and students (PS1) -.15 .42 .50 Eigenvalues after rotation 2.10 2.02 1.88 Table 16 displays the mean factor sc ores by gender. Differences between citizenship ratings for men and women were evaluated with t-test s. There were no significant differences found between men and women on any of the citizenship performance dimensions (ps: t (125) = -.66, ns ; os: t (125) = .62, ns ; ci: t (125) = -.24, ns ). The effect sizes were calculated and a ll three effect sizes were small (ps: d = .12, os: d = .11, and ci: d = .04) according to Cohen’s effect size criteria (d<.2 = small effect; Stevens, 2002). Table 16. Mean Factor Scores by Gender. Gender Factor 1 Organizational Support Factor 2 Conscientious Initiative Factor 3 Personal Support M SD M SD M SD Female 4.21 .56 4.03 .60 4.20 .47 Male 4.15 .52 4.06 .61 4.26 .52 Coefficient Alpha .65 .65 .65 Eigenvalues 2.10 2.02 1.88 Post Hoc Analysis Following the main analyses, outliers were identified and evaluated on the hypotheses and exploratory questions. Outliers were defined as studentized residuals greater than two. Following the identification of the outliers, an influence analysis was conducted, including Cook’s distan ce, leverage, and DF betas. DF Betas with scores over .18 (2/ N = 2/ 127) were considered to have si gnificant impact on the intercept or

PAGE 65

58 slopes. Leverage scores near .047 (2(k+1)/ N = 2(2+1)/127) and cook’s distance values above .2 were considered to have impact on th e variables. The outli er identification and influence criteria were based on recommendations by Pedhazur (1997). There were no changes in significance as a result of the removal of outliers, but exploratory question one had ma rginal significance as a resu lt of removing outliers. Exploratory question one was interested in how the relationshi p between subjective citizenship performance and subjective overa ll performance was different for men than it was for women. Before the outliers were deleted, the incremental variance attributed to the interaction term was 1% (F (1, 123) = 2.47, p = .15). Several outli ers were identified and three outliers were deleted based on the in fluence analysis detailed in Table 17. After the removal of these outliers, the incremental R2 for the interaction term increased to 2% (F (1, 120) = 3.56, p = .06) and was marginally signi ficant at the .10 alpha level. Table 17. Outliers and their Influence on the Variables. DF Betas ID Number Studentized Residual Cook’s D Leverage Intercept Subj. CP Gender Interaction Exploratory Question 1 64 -2.13 .05 .04 -.28 -.37 .20 .27 105 -3.35 .05 .01 -.45 .14 .32 -.06 231 -2.78 .03 .01 -.36 .03 .26 -.07 The regression equation now predicts 37% of the variance in subjective reward recommendations. The beta weight for the in teraction term is -.20 and is marginally significant (t (122) = -1.89, p = .06). The correlation between subjective citizenship and overall performance was stronger for women (r = .69, p <.05) than it was for men (r = .49, p < .05). To explore this finding further, the pr ocedures described by Aiken and West (1991) for plotting and interpreting interactions was employed. Figure 9 shows the

PAGE 66

59 results from plotting the intera ction. A t-test was performed to determine if the slopes for women and men were signifi cantly different from zero. The female slope was significantly different from zero (b = .96; t(118) = 6.98, p < .001), but the male slope was only marginally significant (b = .60; t(118) = 1.98, p < .10). Therefore, the relationship between citizenship performance and overall performance, when they were measured with subjective scales, is signifi cant for females but not for males. Figure 9. Interaction plo tted according to Aiken and West’s procedure. Citizenship Performance Female: Male: Overall Performance

PAGE 67

60 Chapter 4 Discussion Several lines of recent research suggest that employees who perform citizenship behaviors receive higher performance evaluatio ns and more rewards when compared to employees who do not perform citizenship behaviors (e.g. Allen & Rush, 1998; Van Scotter et al., 2000 ) Further research was needed to determine whether all types of people receive equal consequences for their citizenship performance or whether this phenomenon holds true for only a specific por tion of the general population. The present study was designed to determine whether perf orming citizenship behaviors would have different consequences for wo men when compared to men. More specifically, the present study investigated whether men and women are equally evaluated and rewarded for their citizenship performance. As a first step, the differences in ratings of citizenship performance for men and women were studied. Next, differences in the relationship between citizenship performance and overall performance for men and for women were tested. Along similar lines, differences in the relationship betw een citizenship performance and reward recommendations were investigated. Finally, two different scales were used to measure citizenship performance, overall performance, and reward recommendations to determine whether scale type was a moderator of thes e relationships as proposed by the shifting standards model (Nelson et al., 1990). Unfortunately, th e differential relationships hypothesized for men and women we re not supported by this study.

PAGE 68

61 Gender and Citizenship Performance Eagly and Crowley (1986) found that women ar e expected to assist others in their goals and nurture them through their personal problems. These stereotype expectations of women resemble the expectations of an organizational citizen. Further, Allen and Rush (2001) found that there are greater exp ectations for women to display citizenship performance than for men. Building on this stereotype, the shif ting standards model (Nelson et al., 1990) would pred ict that when raters use a subjective scale to rate citizenship performance, they will rate me n relative to men and women relative to women. The subjective scale w ould consequently hide the st ereotypes held by the raters. However, when an objective scale is used, ra ters’ stereotypes are expected to be more evident. In the present study, it was hypothesi zed that when citizen ship performance was measured with an objective scale, women would be rated higher than would men on citizenship performance. Conversely, when citizenship performan ce was measured with a subjective scale, no differences were e xpected between men and women. This hypothesis was not supported. Several explanations are offered to e xplain these null results. The first consideration is the shifting standards model, which has be en widely supported through numerous studies (Biernat & Manis, 1994; Biernat et al ., 1991; Nelson et al., 1990). Because of the growing support of this model, it seems unlikely that the null results are the outcome of a faulty model, but rather a problem with the presen t studies application of that model. Three other possible explanati ons include that there is no gender stereotype of citizenship performance, the gender stereo type was not triggered by the materials, and

PAGE 69

62 the objective scale did not force raters to co mpare men and women. In what follows, each of these possibilities will be considered. The explanation that there is no gender stereotype of citizenship performance is contrary to the findings by Allen and Rush (2001). Allen and Rush (2001) found that there were more expectations for women to perform citizenship performance than for men. These results were found for male ster eotyped jobs and for gender neutral jobs. Therefore, perhaps the gender stereotypes of citizenship performance are dependent on the type of job. The present study tested th e gender stereotypes of psychology professors and found that students did not expect one gende r to be more prevalent than the other for that profession. However, there may be other as pects of the job that can affect the gender stereotype of citizensh ip performance. One example might be jobs in which citi zenship performance overlaps with task performance. In this case, the job might not possess the gender ster eotypes of citizenship performance because these behaviors are equa lly expected of both genders. Especially from students’ perspective, citizenship performance might be an expectation of all teachers and therefore would not be considered extra-role. This idea will be discussed further in the limitations section. The second explanation is that the study did not trigger the gender citizenship performance stereotype. Firs t, the definition of citizensh ip performance might not have been clear enough for participants. In order for the stereotype to be triggered, participants needed to un derstand the meaning of citi zenship performance. The participants might not have attended to the definition given or it might not have been clear. One example of participants not attending to the definition was observed during

PAGE 70

63 the pilot study. After filling out the materials, participan ts were asked if there was anything about the questionnaire they did not understand. One person commented that a couple of the measures were repetitious. Sh e identified the citizen ship performance and overall performance measures as repetitious and it was evident that she had disregarded the directions. One suggestion for future rese arch is to orally e xplain the meaning of citizenship performance and provide more examples of it. Another reason why the stereotype might not have been triggered is that the professor’s performance was given in a vignett e. In these circumstances, the gender of the professor might not have been a salient ch aracteristic. In a field setting, the gender of the employee is typically an unmistakable char acteristic that is noticed about someone. However, a lab setting might not trigger the same stereotype as a field setting. One suggestion for future research is to provide a picture with the vignette. Chen and Heilman (2001) used vignettes to descri be citizenship performance and overall performance. These researchers provided a picture with performance descriptions and they found support for their hypothesis that wo men and men are not equally evaluated or rewarded for their citizenship performance. The final explanation for the null results suggests there could be problems with the objective measure of citizenship performan ce. A number of participants expressed confusion with the objective item that asks them how th is professor would compare against nine other USF professors. Part of Nelson et al.’s (1990) definition of an objective scale is that it is familiar. It is supposed to be a type of measurement that is common to raters. Therefore, it is possible that this item was not acting as an objective measure that unmasks the stereotype. To test this possibility further, the results were

PAGE 71

64 rerun with the objective ranking item exclude d. There was no change in significance after this item was deleted. Therefore, probl ems with the ranking objective measure seem like an unlikely explanation for the unsupported hypothesis. Along the same lines, there is no evidence that raters using the objective scales were rating men and women against the same standard, as implied by the objective measure. When raters were using letter gr ades and percentages they could have been rating men against men and women against wo men. Biernat and Mani s (1994) used letter grades in their study of stereotypes of verb al ability. These researchers found support for the shifting standards model, with women r eceiving a significantly higher letter grade for verbal ability (objective meas ure) than men. However, wh en a subjective measure was used, there was no significant difference f ound between men and women. This finding provides support for the use of letter grades as an object ive measure and reduces the likelihood that the explanation fo r these results is a problem with the objective measures. However, future research should consider havi ng participants rate several employees so that the objective measure can be ranking them and therefore forcing raters to compare men and women directly. Overall Performance The present study found that both meas ures (objective and subjective) of citizenship performance expl ained a significant amount of variance in both measures (objective and subjec tive) of overall performance. This result is consistent with results found in previous research (e.g. Allen & Rush, 1998), indicating that citizenship performance is related to rati ngs of overall performance.

PAGE 72

65 Several studies have investigated if the correlation between citizenship performance and overall performance is di fferent for men and women (Allen & Rush, 2001; Chen & Heilman, 2001; Lovell et al., 1999), with mixed results. The present study was designed to extend these findings by testi ng whether the scale t ype used to measure performance effects the relationship between citizenship and overall performance for men and women. The hypotheses predicted that the relationship between citizenship and overall performance would be larger for me n than for women and the relationship would vary between objective and s ubjective scale types. These hypotheses were not supported. Further, no significant differences were found between men and women on their relationship between citizenship and overall performance. The explanations provided for the null results from the previous section are applicable here. Another possible explanat ion is that the two scales, subjective and objective, were presented as a within subjec ts variable with one following the other, rather than between. Participants may not ha ve differentiated between the scales. This explanation was tested with factor analysis First, the correlat ion between objective overall performance and subjective overall performance was .60 and the reliability coefficient when the two scales were comb ined was .89. A principal factor analysis extracted two factors and with a promax ob lique rotation, the subj ective items loaded on factor one and the objective items loaded on factor two. Theref ore, it appears as if the two scales were distinct and the explanation that these two scales were not discriminated between is not likely.

PAGE 73

66 Reward Recommendations Several studies have found a signifi cant relationship between citizenship performance and organizational rewards (e .g., Van Scotter et al ., 2000). The present study is consistent with these findings with both measures of citizenship performance relating to both measures of reward recomm endations. There have been mixed results in studies that tested whether citizenship pe rformance relates to reward recommendations differently for men than for women (Allen, 2000; Allen & Rush, 2001; Chen & Heilman, 2001). For the current study, these results were further investigated by testing whether the type of measurement used affects the relationship between ci tizenship performance and reward recommendations differently for men than for women. Based on the shifting standards model (Nelson et al., 1990), it was predicted th at the relationship between citizenship performance and reward recomme ndations, for the different scales, would be stronger for men than for women. These hypot heses were not supported. No significant differences were found in the relationship between citizenship performance and reward recommendations for men and for women. The explanations provided for the earlier hypotheses, there is no gender stereotype, the objective scales were not performing as de fined, scale type should be treated as a between subjects variable, a nd the stereotype was not triggered by the materials, are applicable for the reward recommendations hypotheses as well. The proposed explanation that the nu ll results were a consequence of scale type being a within subjects variable is explored further w ith reward recommendations. The correlation between objective and subject ive reward recommendations was .74, and the reliability coefficient for the scale as a whole was .88. A factor analysis was performed and only

PAGE 74

67 one component was extracted. Therefore, it appears as if the two types of reward recommendation scales were not clearly distinguished by participants. Future studies should use scale type as a be tween subjects factor or conduc t a field study where current salary and position are previously established objective measures and subjective measures can be filled out by participants, th erefore reducing possible carryover effects. Citizenship Performance Dimensions There are three dimensions of citizenship performance used in this study. As opposed to the organizational support and conscientious initiative dimensions, the personal support dimension most closely rese mbles the female stereotype of women as nurturers. Therefore, the final research question investigated whether men and women received different ratings on the citizenship performance dimensions and whether the effect sizes were different acros s dimensions. The factor anal ysis verified that the scale truly did capture three distinct dimensions and that the factors correspond, for the most part, with the expected items. However, no significant differences were found between men and women on any of the dimensions. Two previously mentioned explanations ar e relevant for these null findings. The first is that there are no ge nder stereotypes of citizenship performance. The second explanation is that the gende r stereotype was not trigge red by the materials. Post Hoc Following the hypothesis testing, outliers we re identified that may be influencing the results. Exploratory ques tion one investigated how the correlation between subjective measures of citizenship and overall performa nce was different for men and women. With the removal of three outliers, there was a marginally significant result (p = .06). The

PAGE 75

68 correlation between citizenship performance and overall performance was stronger for women (r = .69) than for men (r = .49). After plotting the interactions, the slope was significant for women and not for men. Furthe r, women received a higher correlation than did men on all but one of the hypotheses and exploratory questi ons. This finding was contrary to expectations. More research is needed to determine the cause of these findings. One explanation could be the job used in this study. This idea will be discussed further in th e following section. General Limitations One limitation of this study was the partic ipant pool. Participants were taken from a student population and therefore, might not have been invested in this project since they were doing it for extra credit. If th e participants were not invested, they would have been less likely to remember the perfor mance statements, read the directions and definitions carefully, or take their time filling out the materials. If participants are not focused on the task, the stereotype might not be triggered and, consequently, no differences would be expected. Another possible limitation with the study is that the task statements were given from the perspective of students, and the ci tizenship performance statements were given from the perspective of co-workers or supe rvisors. This is a problem because the participants were students and might have fo cused more on the task performance than on the citizenship performance, si nce students wrote the task st atements. Another possibility is that the performance statements reminded the participants of a particular teacher, making it hard to distinguish between the prof essor triggered and the performance of the professor in the vignette. As evidence, se veral students asked whom the professor was

PAGE 76

69 they had evaluated because they thought they recognized the professor. In one case, the participant said she was thinking of a female professor at USF when she was making the ratings even though she knew the professor in the vignette was a man. Therefore, this participant got the manipulation check correct because she knew th e vignette was about a male professor, but she was thinking about a particular female professor while filling out the materials. Students are likely to have role schemas (expectations about how a person should act in a particular role; Baron, Byrne, & Johnson, 1998) for the performance of a professor. Statements that fit in that sche ma are more likely to be paid attention to, encoded, and recalled (Baron, Byrne, & Johns on, 1998). Therefore, the citizenship performance statements may not have been atte nded to as much as the task statements because the task statements met the stude nts’ schema of teacher performance. A third limitation concerns the job chos en for this study. The stereotype that women are expected to be “concerned with others” and possess “communal qualities, such as being friendly” might be expectations of a teacher and not extra-role behaviors. Sauser et al. (1979) identified five dimensions of teacher performance from their list of college professor behavioral incidents. Some examples of effective performance from this dimension are “this professor offers help at night” and “t his professor made appointments at her students ’ convenience to discuss problems with classwork.” Citizenship performance might be consid ered a requirement of a teacher and consequently, expectations for citizenshi p performance might be equal for men and women.

PAGE 77

70 The hypotheses proposed for the present study might have had different results in an occupation where citizenship performance is more clearly extra role. Consistent with this idea, Allen and Rush (2001) used th e job of a teacher and found no significant differences in ratings of overall performa nce or reward recommendations between men and women. Further, Lovell et al. (1998) studied college re sident advisors in their investigation of citizenship and overall performance ratings for men and women. Task expectations of a resident a dvisor are similar to the expect ations of an organizational citizen. Resident advisors ar e expected to help the reside nts when needed and act as a counselor when the residents need someone to talk to. Further, Lovell et al. (1998) claimed that one possible reason for their findi ngs was that resident advisors may have been chosen for the job based on their abili ties to perform citizen ship performance. Therefore, citizenship performance for the re sident advisor job mi ght also have equal expectations for both men and women. Chen and Heilman (2001) provided an ex ample of a job that did exhibit gender differences in evaluating and rewarding citi zenship performance. The name of the job was not given, but the description of the pro cedures made it clear that it was an office job. The OCB provided was either staying late to help a co-worke r with an important copy job or not staying late to help the co-w orker. This sample of OCB is more clearly extra-role. In a more traditional office setting than a university, citizenship performance might be considered extra-role. Another limitation of the present study is using a multiple regression interaction as the main level of analysis. A great deal of power is needed to detect an interaction in multiple regression (Aiken & West, 1991). Th e important considerations for power are

PAGE 78

71 sample size, effect size, and alpha level. E ffect sizes were calculated with a formula provided by Aiken and West {1991; 2 = (r2 Y.MI r2 Y.M)/(1 r2 Y.MI) MI = interaction and main effects, M = main effects} and they ranged from .01 to .0001. The largest effect was .01, which is considered small by Aike n and West, (1991). Therefore, no further power analysis was done on this sample. It is un likely that lack of pow er is the reason for the null results. Another limitation is that there was not an equal number of men (N = 44) and women (N = 210) in the sample. Although this une qual number is stated as a limitation, there was no evidence of a difference between male and female raters. However, one consideration is that differences might be f ound if the stereotypes he ld by the raters are studied rather than their gender. Dobbins, Cardy, and Truxillo ( 1988) found that raters who held traditional stereotypes about women’s roles in society rated women less accurately than raters who held nontraditi onal stereotypes of women. There was no effect found for rater gender. Future research might want to test the stereotypes held by raters, rather than rater gender, as a po ssible moderator of ci tizenship performance ratings. Implications and Future Research The null results of the current study left two general explanations. The first possibility is that the design or method of the study was flawed. This possible explanation was discussed previously. The s econd is that raters do evaluate and reward men and women the same for their citizenship performance. Therefore, the relationship between citizenship performance and overall performance and citizenship performance and reward recommendations is not different for men than it is for women. These

PAGE 79

72 conclusions would be supporte d by findings from Allen and Rush (2001) and partially by Lovell et al. (1998). If this conclusion is accurate and there are no significant differences in performance ratings for men and women (La ndy & Far, 1988), then it is likely that the men and women perform the same amount of ci tizenship behaviors in the work place. Future research might expand this research further by measuring baserates of male and female citizenship behaviors. Perhaps women perform one t ype of citizenship behaviors more and men perform another type of citizenship behaviors more. It is possible that the pres ent results may have been di fferent if not for the above mentioned limitations. Future research should consider adding to the knowledge in this area with a field study. Based on the limitations mentioned, several suggestions are made for future research. First, scale type shoul d be a between-subjects variable. Second, the type of job chosen should be one in which ci tizenship performance is not an expectation of the job. Third, the difficulty in choosing an appropriate objective measure should be realized when putting the design together. Finally, make sure that the citizenship performance definition is clea r and that the rater understa nds its meaning. With these limitations met in a field study, a more clear understanding can be reached about whether the null results are a consequence of th e design of the study or the hypotheses. Two variables were identified as possible moderators of the gender stereotype of citizenship performance. The first possibl e moderator is the stereotype of job. As mentioned earlier, Allen and Rush (2001) found that the citizenship performance stereotype was moderated by the stereotype of the job. Future research could perform the present study with a gende r-neutral occupation, a male stereotyped occupation and a female stereotyped occupation. The sec ond possible moderator suggests that gender

PAGE 80

73 stereotypes held by the rater may have an im pact the gender stereotypes of citizenship performance (Dobbins, Cardy, & Truxillo, 1988). Future research could consider rater stereotypes in their analyses. In conclusion, the concept of citi zenship performance has important considerations for individuals because it influences their performance evaluations and rewards. The current study was unable to find differences between males and females, which might lead to the conclusion that ther e are no differences between ratings for males and females on citizenship performance. Fu ture research is needed to test these hypotheses in a field setting creating a realistic setting for gender citizenship performance stereotypes to materialize.

PAGE 81

74 References Aiken, L. S. & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Allen, T. (2000, April). Rewarding good citizens: The relationship between citizenship behavior, gender and organizational rewards Paper presented at the meeting of the Society for Industrial and Orga nizational Psychology, New Orleans, LA. Allen, T. D., & Rush, M. C. (2001). The influence of ratee gender on ratings of organizational citizenship behavior. Journal of Applied Social Psychology, 31 25612587. Allen, T. D. & Rush, M. C. (1998). Th e effects of organizational citizenship behavior on performance judgments: A fiel d study and a laboratory experiment. Journal of Applied Psychology, 83 (2), 247-260. Atwater, L. E. & Van Fleet, D. D. (1997) Another ceiling? Can males compete for traditionally female jobs ? Journal of Management, 23 (5), 603-626. Auster, E. R. (1999). Task characteri stics as a bridge between macroand microlevel research on salary inequality between men and women. Journal of Management Review, 14, (2), 173-193. Barnard, C. I. (1938). The functions of the executive Cambridge, MA: Harvard University Press. Baron, Byrne, & Johnson (1998). Exploring Social Psychology (4th ed.). Needham Heights, MA: Allyn and Bacon.

PAGE 82

75 Basket, G. D. (1973). Interview decisi ons as determined by competency and attitude similarity Journal of Applied Psychology, 57 (3), 343-345. Biernat, M. (1995). The sh ifting standards model: Imp lications of stereotype accuracy for social judgment. In Y. T. Lee, L. J. Jussim, & C. R. McCauley (Eds.), Stereotype accuracy: Toward appreciating group differences (pp. 87-114). Washington, DC: American Psychological Association. Biernat, M., Crandall, C. S., Young, L. V., Kobrynowicz, D., & Halpin, S. M. (1998). All that you can be: Stereotyping of self and others in a military context. Journal of Personality and Social Psychology, 75 (2), 301-317. Beirnat, M. & Manis, M. (1994). Shifting standard s and stereotype-based judgements. Journal of Personality and Social Psychology, 66, (1), 5-20. Biernat, M., Manis, M., & Nelson, T. E. (1991). Stereotypes and standards of judgement Journal of Personality and Social Psychology, 60, (4), 485-499. Borman, W. C., Buck, D. E., Hanson, M. A., Motowidlo, S. J., Stark, S., and Drasgow, F. (2001). An examination of the comparative reli ability, validity, and accuracy of performance ratings made usi ng computerized adaptive rating scales. Journal of Applied Psychology, 86, (5), 965-973. Borman, W. C. & Motowildo, S. J. (1997) Task performance and contextual performance: The meaning for personnel selection research. Human Performance 10 (2), p. 99-109. Borman, W. C., Penner, L. A., Allen, T. D., & Motowidlo, S. J. (2001). Personality predictors of citizenship performance International Journal of Selection and Assessment, 9, (1-2), 52-69.

PAGE 83

76 Borman, W. C., White, L. A., & Dorsey, D. W. (1995). Effects of ratee talk performance and interpersonal factors on supervisor and peer ratings. Journal of Applied Psychology, 80 (1), 168-177. Brief, A. P. & Motowidlo, S. J. (1986) Prosocial organizational behaviors. Academy of Management Review, 11 (4), 710-725. Byrne, D. & Griffitt, W. (1966). Similar ity versus liking: A clarification. Psychometric Science, 6, (6), 295-296. Cardy, R. L. & Dobbins, G. H. (1994). Performance appraisal: Alternative perspectives Cincinnati, OH: South-Western Publishing. Chen, J. J. & Heilman, M. E. (2001, April). Need some help? Gender-specific rewards for organizational citizenship behavior s. Paper presented at the meeting of the Society for the Industrial and Orga nizational Psychology, San Diego, CA. Coleman, V. I. & Borman, W. C. (2000) Investigating the underlying structure of the citizenship performance domain. Human Resource Management Review, 10 2544. Dobbins, G. H., Cardy, R. L., & Truxillo, D. M. (1988). The effects of purpose of appraisal and individu al differences in stereotypes of women on sex differences in performance ratings: A la boratory and field study. Journal of Applied Psychology, 73, (3), 551-558. Eagley, A. H. & Crowley, M. (1986). Gender and helping behavior: A metaanalytic review of the soci al psychological literature. Psychological Bulletin 100(3), 283-308.

PAGE 84

77 Eagly, A. H., Makhijani, M. G., & Kl onsky, B. G. (1992). Gender and the evaluation of leaders: A meta-analysis. Psychological Bulletin 111 (1), p. 3-22. Field, A. (2000). Discover statistics, using SPSS for windows. Thousand Oaks, CA: Sage. Golightly, C., Huffman, D. M., & Byrne, D. (1972). Liking and loaning. Journal of Applied Psychology, 56, 521-523. Graen, G. B., Novak, M. A., Sommerkamp, P. (1982). The effects of leadermember exchange and job design on producti vity and satisfacti on: Testing a dual attachment model. Organizational Behavior and Human Decision Processes, 30, (1), 109-131. Hamner, W. C., Kim, J. S., Baird, L., & Bigoness, W. J. (1974). Race and sex determinants of ratings by potential employe rs in a simulated work-sampling task. Journal of Applied Psychology, 59 (6), 705-711. Heilman, Madeline, E. (1995). Sex stereotype s and their effects in the workplace: What we know and what we don’t know. Journal of Social Behavior and Personality, 10 (6), 3-26. Heilman, M. E., Block, C. J., Martell, R. F., & Simon, M. C. (1989). Has anything changed? Current characteri zations of men, women, and managers. Journal of Applied Psychology, 74 (6), 935-942. Heilman, M. E., Martel, R. F., & Simon, M. C. (1988). The vagaries of sex bias: Conditions regulating the underv aluation, equivalation, and ove rvaluation of female job applicants. Organizational Behavior and Huma n Decision Processes, 41, 98-110.

PAGE 85

78 Hui, C., Lam, S., & Law, K. (2000). Instrumental values of organizational citizenship behavior for promoti on: A field quasi-experiment. Journal of Applied Psychology, 85, (5), 822-828. Judge, T. A., Cable, D. M., Boudreau, J. W., & Bretz, Jr., R. D. (1995). An empirical investigation of the pr edictors of executive success. Personnel Psychology, 48 485-519. Katz, D. (1964). The motivational ba sis of organizational behavior. Behavioral Science, 9, 131-46. Kiker, D. S. & Motowidlo, S. J. (1999) Main and interaction effects of task and contextual performance on supervisory reward decisions. Journal of Applied Psychology, 84 (4), 602-609. Landy, F. J. & Farr, J. L. (1980). Performance rating. Psychological Bulletin, 87, (1), 72-107. Lewis, K. & Taylor, M. S. (2001, April) Reciprocity from the organization’s side: Manager reactions to employee psycholog ical contract breach. Paper presented at the meeting of Society for Industrial Organizational Psychologist, San Diego. Liden, R. C., Wayne, S. J., & Stilwell, D. (1993). A longitudinal study on the early development of leader-member exchanges. Journal of Applied Psychology, 78, (4), 662-674. Lovell, S. E., Kahn, A. S., Anton, J., Davidson, A., Dowling, E., Post, D., & Mason, C. (1999). Does gender affect th e link between organi zational citizenship behavior and performance evaluation? Sex Roles 41 (5-6), p. 469-478.

PAGE 86

79 Lyness, K. S. & Thompson, D. E. (1997). Above the glass ceiling? A comparison of matched samples of female and male executives. Journal of Applied Psychology, 82 (3), 359-375. MacKenzie, S. B., Podsadoff, P. M ., & Ahearne, M. (1998) Some possible antecedents and consequences of in-role a nd extra-role salesperson performance. Journal of Marketing, 62, 87-98. MacKenzie, S. B., Podsakoff, P. M ., & Fetter, R. (1991). Organizational citizenship behavior and object ive productivity as determinants of managerial evaluations of salesperson’ performance. Organizational Behavior and Human Decision Processes, 50 (1), 123-150. Manis, M., Biernat, M., & Nelson, T. (1990). Comparison and expectancy processes in human judgment Journal of Personality and Social Psychology, 61, (2), 203-211. Martel, R. F. & DeSmet, A. L. (2000, April). Gender stereotyping in the managerial ranks: A Bayesian approach to measuring beliefs about the leadership abilities of male and female managers. Paper presented at the meeting of the Society for Industrial and Organizational Psychology, New Orleans, LA. Massengill, D. & di Marco, N. (1979). Sex-role stereotypes and requisite management characteristics: A current replication. Sex Roles, 5 (5), 561-570. Miles, R. E. (1964). Attitudes toward management theory as a factor in managers’ relationships with their superiors. Academy of Management Journal, 7, 303313.

PAGE 87

80 Morrison, E. W. (1994). Role definitions and organizational citi zenship behavior: The importance of the employee’s perspective. Academy of Manage ment Journal, 37 (6), 1543-1567. Motowildo, S. J. & Van Scotter, J. R. (1994). Evidence that task performance should be distinguished from contextual performance. Journal of Applied Psychology, 79, 475-480. Mount, M. K., Judge, T. A., Scullen, S. E., Sytsma, M. R., & Hezlett, S. Al. (1998). Trait, rater, and level effect s in 360-degree performance ratings. Personnel Psychology, 51, 557-576. Nelson, T. E., Biernat, M. R., & Manis, M. (1990). Everyday base rates (sex stereotypes): Pote nt and resilient. Journal of Personality and Social Psychology, 59, (4), 664-675. Organ, D. W. (1997). Organizational citiz enship behavior: It’s construct cleanup time. Human Performance, 10, 85-97. `Organ, D. W. (1988). Organizational citizenship behav ior: The good soldier syndrome. Lexington, MA: Lexington Books. Organ, D. W. (1977). A reappraisal and re interpretation of th e satisfaction causes performance hypothesis. Academy of Management Review, 2 46-53. Organ, D. & Moorman, R. (1993). Fa irness and organiza tional citizenship behavior: What are the connections? Social Justice Research, 6, p. 5-18.

PAGE 88

81 Organ, D. W. & Paine, B. J. (1999). A new kind of performance for industrial and organizational psychology: Recent contri butions to the study of organizational citizenship behavior. International Review of Indust rial and Organizational Psychology, 14, p. 337-368. Organ, D. W. & Ryan, K. (1995). A meta -analytic review of attitudinal and dispositional predictors of orga nizational citizensh ip behavior. Personnel Psychology, 48, (4), 775-802. Orr, J. M., Sackett, P. R., & Mercer, M. (1989). The role of prescribed and nonprescribed behaviors in estimating the dollar value of performance. Journal of Applied Pschology, 74 (1), 34-40. Pedhazor, E. J. (1997). Multiple regression in behavioral research (3rd ed.). Orlando, FL: Harcourt Brace College. Podsakoff, P. M., Ahearne, M. & MacK enzie, S. M. (1997). Organizational Citizenship Behavior and th e Quantity and Quality of Work Group Performance. Journal of Applied Psychology 82(2), 262-270. Podsakoff, P. M. & Mackenzie, S. B. ( 1997). Impact of orga nizational citizenship behavior on organizational performance: A re view and suggestions for future research. Human Performance 10(2), 133-151. Podsakoff, P. M. & MacKenzie, S. B. (1994). Organizational citizenship behavior and sales un it effectiveness. Journal of Marketing Research 31, 351-363. Podsakoff, P. M, MacKenzie, S. B., Paine, J. B., & Bachrach, D. G. (2000). Organizational citizenship behaviors: A critical review of the theoretical and empirical literature and suggestions for future research. Journal of Management, 26, (3), 513-563.

PAGE 89

82 Powell, G. N. & Butterfield, D. A. (1989). The “good manager:” Did androgyny fare better in the 1980s? Group and Organization Stuides, 14 (2), 216-233. Pulakos, E. D. & Wexley, K. N. ( 1983). The relationship among preceptual similarity, sex, and performance rati ngs in manager-subordinate dyads. Academy of Management Journal, 26, (1), 129-139. Pulakos, E. D., White, L. A., Oppler S. H. & Borman, W. C. (1989). Examination of race and sex eff ects on performance ratings. Journal of Applied Psychology 74 (5), p. 770-780. Robbins, T. L. and Denisi, A. S. ( 1993). Moderators of sex bias in the performance appraisal process: A cognitive analysis Journal of Management, 19, (1), 113-126. Rosen, B. & Jerdee, T. H. (1973). The influence of sex-ro le stereotypes on evaluations of male and fema le supervisory behavior. Journal of Applied Psychology, 57, 44-48. Roethlisberger, F. J. & Dickson, W. J. (1939). Management and Worker. Cambridge, MA: Harvard University Press. Sauser, W. I., Evans, K. L., & Champion, C. H. (1979, March). Two hundred and fifty scaled incidents of college classroom behavior. Paper presented at the Southeastern Psychological Association annual conference, New Orleans. Schwab, D. P. & Grams, R. (1985). Sex-re lated errors in job evaluation: A “realworld” test. Journal of Applied Psychology, 70 (3), 533-539. Shein, V. E. (1973). The relationship betw een sex role stereotypes and requisite management characteristics. Journal of Applied Psychology, 57 (2), 95-100.

PAGE 90

83 Schein, V. E., Mueller, R., & Jacobson, C. (1989). The relationship between sex role stereotypes and requisite management characteristics among college students. Sex Roles, 20, 103-110. Senger, J. (1971). Mangers’ percepti ons of subordinates’ competence as a function of personal value orientations. Academy of Management Journal, 14, 415-423. Skarlicki, D. P. & Latham, G. P (1996). Increasing citizenship behavior within a labor union: A test of orga nizational justice theory. Journal of Applied Psychology, 81, 161-169. Smith, A., Organ, D., & Near, J. (1983). Organizational citizenship behavior: Its nature and antecedents Journal of Applied Psychology, 68, p. 653-663. Stevens, J. P. (2002). Applied multivariate statistics for the social sciences (4th ed.). New Jersey: Lawrence Erlbaum Associates. Stroh, L. K., Brett, J. M., & Reilly, A. H. (1992). All the right stuff: A comparison of female and male managers’ career progression. Journal of Applied Psychology, 77, (3), 251-260. Sutton, M. (1998). Does the type of behavior and gender make a difference? Unpublished master’s thesis, Universi ty of South Florida, Florida. Swim, J., Borgida, E., Maruyama, G., & Myers, D. G. (1989). Joan mckay versus john mckay: Do gender ster eotypes bias evaluations? Psychological Bulletin 105(3), 409-429. Turban, D. B. & Jones, A. P. (1988). S upervisor-subordinate similarity: Types, effects, and mechanisms. Journal of Applied Psychology, 73, (2), 228-234.

PAGE 91

84 Van Scotter, J. R. (2000). Relationships of task performance and contextual performance with turnover, job sati sfaction, and affective commitment. Human Resource Management Review, 10, (1), 79-85. Van Scotter, J. R., Motowidlo, S. J., & Cross, T. C. (2000). Effects of task peformance on contextual performance on systematic rewards. Journal of Applied Psychology, 85, (4), 526-535. Wakabayashi, M., Graen, G., Graen, M., & Graen, M. (1988). Japanese management progress: Mobility into middle management. Journal of Applied Psychology, 73, (2), 217-227. Wayne, S. J. & Ferris, G. R. (1990). Influence tactics, affect, and exchange quality in supervisor-subordina te interactions: A laboratory experiment and field study. Journal of Applied Psychology 75(5), 487-499. Wayne, S. J. & Green, S. A. (1993). The effects of leader-member exchange on employee citizenship and impression management behavior. Human Relations, 46, (12), 1431-1440. Wexley, K. N., Alexander, R. A., Green awalt, J. P., & Couch, M. A. (1980). Attitudinal congruence and similarity as relate d to interpersonal evaluations in managersubordinate dyads. Academy of Management Journal, 23, (2), 320-330. Zalesny, M. D. & Highhouse, S. (1992). Accuracy in performance evaluations. Organizational Behavior and Huma n Decision Processes, 51 22-50.

PAGE 92

85 Appendices

PAGE 93

86 Appendix A Consent Form The purpose of this research study is to test a new system for evaluating professor performance in the psychology department at USF. You will be helping to pilot the new evaluation materials. You are being asked to participate because as a st udent, your input is important to the pilot process. You will be placed After reading a list of performance statements, you will be asked to evaluate the professor’s performance and recommend rewards for the prof essor. Following the evaluations, you will be asked a few demographic and educational questions. The entire process will take approximately 20 minutes to complete. There are no risks for your participation. You will receive no compensation for your participa tion. Your privacy and research records will be kept confidential to the extent of the law. Authorized research personnel, employees of the Department of Health and Human Services and the USF Institutional Review Board may inspect the records from this research project. The results of this study may be published. However, the data obtained from you will be combined with data from other people in the publ ication. The published results will not include your name or any other information that would in any way personally identify you. Numbers will be used to identify your survey and only the i nvestigators will have access to the questionnaires, which will be kept in a locked office on campus. Your decision to participate in this research st udy is completely voluntary. You are free to participate in this research study or to withdraw at any time. If you choose not to participate, or if you withdraw, there will be no penalty or loss of benefits that you are entitled to receive. If you have any questions about this research study, contact Lisa Wilkinson at: lvwilkin@helios.acomp.usf.edu or 974-5034. If you have questions about your rights as a pers on who is taking part in a research study, you may contact a member of the Division of Research Compliance of the University of South Florida at 813-974-5638.

PAGE 94

87 Appendix A: (Continued) Your Consent—By signing this form I agree that: I have fully read or have had read and explaine d to me this informed consent form describing a research project. r tunity to question one of the persons in charge of this research and have o ry answe r s. I understand that I am being asked to particip ate in research. I understand the risks and benefits, and I freely give my consent to partic ipate in the research project outlined in this form, under the conditions indicated in it. I have been given a signed copy of this info rmed consent form, which is mine to keep. Signature of Participant Prin ted Name of Participant Date Investigator Statement I have carefully explained to the subject the nature of the above protocol. I hereby certify that to the best of my knowledge the subject signing this consent form understands the nature, demands, risks and benefits involved in participating in this study. Signature of Investigator Printed Name of Investigator Date Institutional Approval of Study and Informed Consent This research project/study and informed c onsent form were reviewed and approved by the University of South Florida Institutional Review Board for the protection of human subjects. This approval is valid until the date provided below. The board may be contacted at (813) 974-5638. Approval Consent Form Expiration Date: Revision Date:_______________

PAGE 95

88 Appendix B Cover Story Professor Evaluations We are working on the development of a new system for evaluating and rewarding professors in the psychology depart ment at USF. We are currently collecting pilot data that will be used to help make decisions about the implementation of the new system. As students, your input is important to the pilot process. On the next several page s you will see comments rega rding the performance of "Dr. Smith" (for confidentiality purposes, the real name of the professor is protected). This information was collected from st udents and other members of the psychology department during the last semester. Followi ng the performance statements, you will be asked to review these comments and then ev aluate various aspects of the professor's performance and provide your opinion on rewards that could be offered to this professor. A fair and objective evaluation and reward process for professors is key to the university system. Please read and respond to the materials carefully. Thanks in advance for your assistance with this project.

PAGE 96

89 Appendix C Performance Log Psychology Professor Michael Smith or Michelle Smith These comments are taken from evaluations provided by students and colleagues of Dr. Smith during Fall 2001 semester. The following comments are in no particular order. Please read them carefully because you will be asked to evaluate Dr. Smith’s performance. He gave details about the material in class but never elaborated beyond them. His test questions are usually reas onable, but are sometimes tricky. He expresses his own personal satisfaction in being a faculty member at USF when asked by outsiders. He can generally be persuaded to sacrifi ce own personal interests for the good of the psychology department. He looks for additional productive work to do when his own normally scheduled duties are completed. He accepts invitations to attend teaching a nd research enhancement courses offered by the university. When students or other faculty ask for help, he can usually be c ounted on to suggest solutions to their problems. He assigned ten pages of read ing before each class period. He leaves promptly after giving his lecture. He required a term paper, oral presentation, and weekly tests. He continuously referred back to hi s notes while attempting to lecture. When others ask for his help because they are overloaded, he can usually be counted on to take on some additional tasks. When students or faculty ask for help, he can usually be counted on to listen to their personal problems and provide emotional support.

PAGE 97

90 Appendix C: (Continued) He requires a lot of memorization for his class. He generally completes work on time, unless deadlines are very short His tests usually cover 3 or 4 chapters of the book. He might offer suggestions for changes to university procedures to make them more efficient. He marks off for poor class attendance. He takes advantage of available opportuni ties to develop own research and teaching skills when such opportunities present themselves.

PAGE 98

91 Appendix D Citizenship Performance Scale We are interested in evaluating professors at the University of South Florida on their citizenship performance. Please read the following definition of citizenship performance carefully and circle th e number that corresponds to the answer you think is appropriate for Dr. Smith. Citizenship Performance: behaviors that go beyond an employee’s normal job duties. These behaviors involve doing extra tasks or making an extra effort that supports the organization and assists co-workers in performing their duties. Subjective Scale 1. Dr. Smith is helpful to co-workers and students: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 2. Dr. Smith is courteous with co-workers and students: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 3. Dr. Smith is cooperative when wo rking with colleagues and students: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 4. Dr. Smith is loyal to USF: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 5. Dr. Smith represents the USF well to outsiders: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5

PAGE 99

92 Appendix D: (Continued) 6. Dr. Smith complies with Univ ersity rules and procedures: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 7. Dr. Smith is persistent and gives extra effort when needed: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 8. Dr. Smith takes initiative to ge t things done when it is needed: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 9. Dr. Smith is dedicated to self-dev elopment in order to improve his own performance: Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5

PAGE 100

93 Appendix D: (Continued) Objective Scale of Citizenship Performance 1. Compare Dr. Smith’s performance agains t the average citizenship performance of all professors and circle the letter grad e you feel that Dr. Smith deserves for his citizenship performance: A AB+ B BC+ C CD+ D DF 2. Compare Dr. Smith’s performance agai nst the average citizenship performance of all professors and place an x in the box that corresponds to the percentage you would give to Dr. Smith for his citizens hip performance, with one hundred percent equaling the best citizenship perfor mance that a professor can have. 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 3. If you were to review the citizenship p erformance of nine other USF psychology professors in addition to this one, how would you guess this professor would compare? 1 2 3 4 5 6 7 8 9 10 Professor Professor Would Would Rank First Rank Tenth (The Best) (The Worst)

PAGE 101

94 Appendix E Overall Performance Scale Use the information from the performan ce log of Dr. Smith to evaluate his performance. Please circle the num ber that corresponds to your answer. 1. This professor makes an important co ntribution to the University of South Florida. Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 2. This professor is extremely valuabl e to the University of South Florida. Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 3. This professor would be extremely costly for the University of South Florida to replace. Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 4. This professor is a vital part of the University of South Florida. Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5 5. This professor is indispensable to the University of South Florida. Strongly Disagree Disagree Neither Agree Nor Disagree Agree Strongly Agree 1 2 3 4 5

PAGE 102

95 Appendix E: (Continued) 6. Circle the letter grade you feel th at Dr. Smith deserves for his overall performance: A AB+ B BC+ C CD+ D DF 7. Compare Dr. Smith’s performance agai nst the average performance of all professors and place an x in the box that corresponds to the percentage you would give to Dr. Smith for his performance, with one hundred percent equaling the best performance that a professor can have. 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 8. If you were to rank the overall performanc e of nine other professors in addition to this one, how would you gue ss this professor would compare? 1 2 3 4 5 6 7 8 9 10 Professor Professor Would Would Rank First Rank Tenth (The Best) (The Worst)

PAGE 103

96 Appendix F Reward Recommendations Scale Directions: Please indicate the extent that you would recommend the professor for each of the following rewards. Circle th e number that corresponds to your choice. 1. Would you recommend this prof essor for a teaching award? Would definitely not recommend Would probably not recommend Neutral Would recommend with some minor reservations Would recommend without reservation 1 2 3 4 5 2. Would you recommend this professor fo r a promotion into a more prestigious teaching position? Would definitely not recommend Would probably not recommend Neutral Would recommend with some minor reservations Would recommend without reservation 1 2 3 4 5 3. Would you recommend this professor for a salary bonus? Would definitely not recommend Would probably not recommend Neutral Would recommend with some minor reservations Would recommend without reservation 1 2 3 4 5 4. If 10 professors were competing for a teaching award, where would you rank this professor? 1 2 3 4 5 6 7 8 9 10 Most Least Deserving Deserving Of the Of the Award Award (Best) (Worst)

PAGE 104

97 Appendix F: (Continued) 5. Compare Dr. Smith’s performance agai nst the average performance of all professors, and circle the amount you would recommend for this professor to receive as a salary bonus: 0 100 200 300 400 500 6. If 10 professors were competing for a more prestigious teaching position, where would you rank this professor? 1 2 3 4 5 6 7 8 9 10 Most Least Deserving Deserving Of the Of the Position Position (Best) (Worst)

PAGE 105

98 Appendix G Demographic Information Please provide some information about yourself: 1. Age: _____ 2. Gender: M ____ F ____ 3. Race or ethnicity: African-American ____ White, non-Hispanic ____ Hispanic/Latina ____ Asian/Pacific Islander ____ Native American/Alaskan ____ 4. What is your college grade level? Freshman ____ Sophomore ____ Junior ____ Senior ____ Graduate Student _____ 5. Please list your major: _____________________________ 6. Do you have any work experience? Yes No 7. Do you have any experience as a supervisor? Yes No When answering the following question, do not flip back. 8. What was the gender of the professor that you just evaluated? Male Female Thank you for your time, we ap preciate your participation!

PAGE 106

99 Appendix H Participant Debriefing The purpose of this study was to investigat e how raters differentiate ratees when evaluating performance and when recomme nding rewards. Specifically, we are interested in how citizenship performance scores effect performance ratings and reward recommendations. However, in past studies there have been mixed results with studies finding that people are differe ntially evaluated for their citizenship performance (eg. Lovell, et. al., 1999) and studies that did not find a significant difference in how people are evaluated for their citizenship performance (Organ & Ryan, 1995). On account of these mixed findings the shifting standards model was applied. This social psychology theory predicts that st ereotyped differences wi ll be more apparent when objective rather than s ubjective scales are used (Biernat, Manis, & Nelson, 1991). Therefore, someone who is stereotyped to receive higher ra tings on citizenship performance will receive higher ratings on objec tive scales, but not subjective scales. Further, this study will investigate how citiz enship performance expectations will affect their ratings on overall performance eval uations and reward recommendations. On account of the nature of this study, plea se do not reveal the purpose of this study to other USF students because th ey may be future participants. Please view the following references for more information: Biernat, M., Manis, M., & Nelson, T. E. (1991). Stereotypes and standards of judgement. Journal of Persona lity and Social Psychology, 60, (4), 485-499. Lovell, S. E., Kahn, A. S., Anton, J., Davidson, A., Dowling, E., Post, D., & Mason, C. (1999). Does gender affect th e link between organi zational citizenship behavior and performan ce evaluation? Sex Roles 41 (5-6), p. 469-478. Organ, D. W. & Ryan, K. (1995). A meta -analytic review of attitudinal and dispositional predictors of organizational citiz enship behavior. Personnel Psychology, 48, (4), 775-802.


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 001430580
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 031007s2003 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000098
035
(OCoLC)52376115
9
AJL4041
b SE
SFE0000098
040
FHM
c FHM
049
FHME
090
BF121
1 100
Wilkinson, Lisa.
0 245
Gender stereotypes of citizenship performance
h [electronic resource] /
by Lisa Wilkinson.
260
[Tampa, Fla.] :
University of South Florida,
2003.
502
Thesis (M.A.)--University of South Florida, 2003.
504
Includes bibliographical references.
516
Text (Electronic thesis) 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 106 pages.
520
ABSTRACT: The relationship between citizenship performance and overall performance and the relationship between citizenship performance and reward recommendations were investigated, with gender and scale type as possible moderators. Two hundred and fifty-four University of South Florida students were used in this study. The majority of these participants were undergraduate, psychology majors, female, and between the ages of 17 and 23. Participants were given statements describing a teacher's performance and were asked to evaluate the professor on citizenship and overall performance and recommend them for rewards. No support was found for the hypothesis that men would have a stronger relationship than women between citizenship and overall performance. No support was found for the hypothesis that men would have a stronger relationship than women between citizenship performance and reward recommendations. Scale type was not found to influence these relationships. These results are not consistent with the shifting standards model. Numerous suggestions are made for changes to the experiment, including performing a field study instead of a lab study.
590
Adviser: Allen, Tammy
650
Sex role.
Stereotypes (Social psychology)
Organizational behavior.
653
organizational citizenship performance.
shifting standards model.
objective scales.
reward recommendations.
overall performance.
690
Dissertations, Academic
z USF
x Psychology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.98