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 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004575
Prioritizing those who follow :
b servant leadership, needs satisfaction, and positive employee outcomes
h [electronic resource] /
by Kristin Saboe.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (MA)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: Servant leaders seek to fulfill the needs of followers and promote their success and well-being through a follower-centric, generative approach to leadership. This study proposes a model to describe the mediating mechanism of follower needs satisfaction, as proposed by Self-Determination Theory (SDT), for the relationship between servant leadership (SL) behaviors and employee outcomes (e.g., job performance, job attitudes, well-being, community prosocial behavior). Supervisor-subordinate dyads (N = 147 pairs) from four diverse organizations completed surveys about the supervisors' leadership behaviors and the subordinates' job experiences. Structural equation modeling and regression analyses were conducted to determine the nature of relationships between SL, SDT needs, and the organizational outcomes. Direct and indirect effects were observed among these variables, suggesting SDT primarily mediates the relationship between supervisors' SL behaviors and subordinates' job attitudes.
Advisor: Russell E. Johnson, PhD
t USF Electronic Theses and Dissertations.
Prioritizing Those Who Follow: Servant Leadership, Needs Satisfaction, and Positive Employee Outcomes By Kristin N. Saboe 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 Co-Major Professor: Russell E. Johnson, Ph.D. Co-Major Professor: Tammy D. Allen, Ph.D. Jennifer K. Bosson, Ph.D. Date of Approval: June 8, 2010 Keywords: Leadership, Servant Leadersh ip, Self-Determination Theory, Needs Satisfaction, Motivation Copyright 2010, Kristin N. Saboe
Dedication It is with much appreciation and admirati on that I dedicate this thesis to my family, mentors and friends who provided much needed support, encouragement and assistance. I would especially like to expr ess my gratitude to my parents who have, through their example, taught me that the path to success is through dedication, hardwork and appreciating those who helped along the way. Special thanks are also due to Kelly Smith, my research assistant and fr iend, who provided much needed motivation and encouragement. In addition, I would like to thank my best friends who spent many hours listening with an open ear, counseli ng me and offering needed mental breaks during the best and worst of times. Lastly, this project would not have been possible without the mentorship of Karen Nelson and Pete DeLisle who helped to ignite my passion in leadership and psychology early in my studies and continue to offer their guidance as my career progresses. Thank you to all of you for being a part of my life!
Acknowledgements The help, encouragement and support from my primary co-chair, Russell E. Johnson, deserves acknowledgement. Additionally, the gui dance offered by key individuals at each of the three organizations sampled was much appreciated. I am thankful to have the support and opportunities for growth provided by these individuals.
i Table of Contents List of Tables ................................................................................................................ .... iii List of Figures ............................................................................................................... ..... iv Abstract ...................................................................................................................... v Introduction .................................................................................................................. ....1 Servant Leadership...................................................................................................5 Distinguishing SL from ot her types of leadership .......................................5 Servant Leadership, Needs, and a Framework for Understanding Needs .............13 Self-Determination Theory ....................................................................................15 Needs..........................................................................................................16 Regulatory motivation ...............................................................................17 Organizational Outcomes.......................................................................................19 Method ....................................................................................................................25 Participants .............................................................................................................25 Measures ................................................................................................................29 Servant Leadership.....................................................................................29 SDT needs ..................................................................................................30 Job performance .........................................................................................31 Well-being..................................................................................................32 Job satisfaction ...........................................................................................33 Social cohesion ..........................................................................................33 Covariates ..................................................................................................34 Exploratory relationships ...........................................................................35 Results ....................................................................................................................36 Results from the Measurement Models .................................................................38 Results from the Structural Models .......................................................................40 Results from the Regression Models .....................................................................46 Exploratory Analyses .............................................................................................51 Exploratory Leadership: Dominance Analysis ..........................................55 Discussion .................................................................................................................... 57 Limitations .............................................................................................................66 Conclusion .............................................................................................................68
ii List of References ............................................................................................................ ..70 Appendices .................................................................................................................... 80 Appendix A: Predictor Survey Scales ....................................................................81 Appendix B: Job Performance Scales ....................................................................83 Appendix C: Well-Being Scales ............................................................................84 Appendix D: Social Cohesion Scales ....................................................................86 Appendix E: Explorator y Leadership Scales .........................................................87 About the Author ............................................................................................... END PAGE
iii List of Tables Table 1. Empirical relationships amongst SD T needs, SL, and organizational outcomes 20 Table 2. Response rates of recruited participants ..............................................................26 Table 3. Zero-order correlati ons amongst focal variables .................................................37 Table 4. Hierarchical mediated regressions for Job Attitudes and Job Performance Criteria ..................................................................................................................47 Table 5. Hierarchical mediat ed regressions for Well-Bei ng and Community Prosocial Behavior outcomes ...............................................................................................48 Table 6. Leadership hierarchical regressions for Job Attitudes and Job Performance outcomes ...............................................................................................................53 Table 7. Leadership hierarchical regressi ons for Well-Being and Community Prosocial Behavior outcomes ...............................................................................................54 Table 8. Dominance Weights Analysis of Th ree Leadership Predictors for Model SDT Needs Satisfaction ................................................................................................56 Table 9. Dominance Weights Analysis of Three Leadership Predictors for Model Outcome Variables ...............................................................................................56
iv List of Figures Figure 1. Hypothesized model of SL, SDT needs, and organizational outcomes .............24 Figure 2. Full mediation model with st andardized path coefficients .................................42 Figure 3. Partial mediation model with standardized path coefficients .............................43
v Prioritizing Those Who Follow: Servant Leadership, Needs Satisfaction, and Positive Employee Outcomes Kristin N. Saboe Abstract Servant leaders seek to fulfill the need s of followers and promote their success and well-being through a follower-centric, genera tive approach to leadership. This study proposes a model to describe the mediating me chanism of follower ne eds satisfaction, as proposed by Self-Determination Theory (SDT ), for the relationship between servant leadership (SL) behaviors and employee outco mes (e.g., job performance, job attitudes, well-being, community prosocial behavi or). Supervisor-subordinate dyads ( N = 147 pairs) from four diverse organizations completed surveys about the supervisorsÂ’ leadership behaviors and the subordinatesÂ’ job experiences. Structural equation modeling and regression analyses were conducted to dete rmine the nature of relationships between SL, SDT needs, and the organizational outco mes. Direct and indi rect effects were observed among these variables, suggesting SDT primarily mediates the relationship between supervisorsÂ’ SL behaviors and subordinatesÂ’ job attitudes.
1 Introduction Fundamentally, leaders are defined by thei r ability to influence others (Kaiser, Hogan & Craig, 2008). Such a basic definition of leadership underscores two critical assumptions about leader-follo wer relationships and motiva tional regulation. First, a leader can only invoke influence in the presen ce of affected followers This addresses the motivational outputs of leadership. Identifyi ng Â“whoÂ” and Â“whatÂ” leaders prioritize within their values hierarc hy is critical for defining the target (e.g. organization, followers, board members) of leadersÂ’ actions Most leadership theories prioritize the needs and growth of the organization and its leadership, whereas others, such as servant leadership (SL), place priority with followe rs (Spears, 1995; Stone, Russell, & Patterson, 2004). SL is defined by the ethical responsibil ity of leaders to pr ioritize the needs, growth, and well-being of their followers above personal and organizational interests (Graham, 1991; Spears). Second, followers may not be motivated to internalize the le adersÂ’ goals and values, thus serving the leader only in deed but not in creed (Graham, 1991). Addressing the motivational inputs of followers, the s econd assumption highlights the importance of intrinsic versus extrinsic regulatory sources for followersÂ’ motivation. Self-determination theory (SDT; Gagn & Deci, 2005; Ryan & Deci, 2000) posits that people experience greater well-being and achieve higher pe rformance levels when their needs for competence, autonomy, and relatedness are fulf illed (Gagn & Deci). By satisfying these
2 needs for followers, Mayer, Bardes, and Picco lo (2008) demonstrated that SL behaviors were related to followersÂ’ job satisfaction. Based on this finding, as well as SDT and the philosophy of SL, followers of leaders who di splay SL behaviors are likely to pursue their leadersÂ’ aims both in creed and in deed, rather than merely the latter. When such a state occurs, SDT suggests that followers will experience enhanced welfare and performance. The purpose of this study is to exam ine the mediating mechanism of needs satisfaction for supervisor-performed SL behaviors and subordinate outcomes. Specifically, a model is proposed in which the fulfillment of subordinatesÂ’ needs, defined by SDT, mediates the relationship between SL behaviors of supervisor s and three clusters of subordinatesÂ’ workplace outcomes: job pe rformance, well-being, and social cohesion. A test of this model provides as least three c ontributions to the litera ture. First, this study provides a holistic approach to the study of leadership, motivation, and follower outcomes by testing all three components w ithin a single model. A large volume of existing leadership research ha s examined relationships of l eadersÂ’ traits and behaviors with followersÂ’ behaviors, while ignoring the likely mediating ro le of followersÂ’ motivation (Lord & Brown, 2004). Several impor tant basic relationships have been identified within the few empirical studies published on SL and organizational outcomes. Thus, support is mounting for the positive impacts of SL, as proposed in GreenleafÂ’s original atheoretical (A volio & Gardner, 2005) philosophy of SL (Greenleaf, 1970, 1991). Specifically, empirical findings support SL Â’s relationships with subordinate job performance (Liden, Wayne, Zhao, & Hende rson, 2008), job satisfaction (Barbuto & Wheeler, 2005; Mayer, Bardes, & Piccolo, 2008), prosocial beha viors (Ehrhart, 2004;
3 Walumbwa, Hartnell, & Oke, 2010), affectiv e commitment (Liden et al.), regulatory focus (Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008) and ju stice perceptions (Ehrhart; Mayer et al.). A dditionally, SL behaviors are associated with reduced counterproductive work behaviors committed by subordinates (Krebs, 2005), enhanced leader-member relations (B arbuto & Wheeler, 2005; Li den et al.), improved subordinatesÂ’ trust in leader s and organizations (Joseph & Winston, 2005; Reinke, 2004) and improved organizational climates (Walumbwa, Hartnell, & Oke). These basic relationships are enlighten ing but fail to address how SL influences followersÂ’ motivation which, then, lend to outcomes. An empirical examination, with a model of leadership behaviors, follower motivation, and outcomes, will provide a more complete picture of how SL motivates subordinate s to think and act in an organizational context. Second, this study contributes by establis hing basic and complex relationships amongst the focal constructs. It also demonstrates mechanisms through which SL enables advantageous workplace outcomes for employ ees and, indirectly, the organization. SL has maintained prominence amongst practitione rs as a conceptual model of behavior since it was coined by Greenleaf in 1970, but remains in its infancy as an empirically defined construct. Providing empirical subs tantiation and theoretical grounding for the effectiveness of SL in organi zations gives credence to the at heoretical and practice-based SL philosophy. In addition, there is a mounti ng interest amongst organizations to promote ethical behaviors and socially responsibl e practices in busine ss (Brown, Trevio, & Harrison, 2005). SL promotes ethi cal leadership behaviors by prioritizing the welfare and regenerative growth of employees. Organiza tions seeking to build a greater ethical framework would be wise to consider the contribution of SL in light of research
4 suggesting the power of leader shipÂ’s cascading effects for an organizationÂ’s ethical tone (Mayer, Kuenzi, Greenbaum, Bardes, & Salvador, 2009). Third, this study provides a framework for SDT by demonstrating its utility within industrial and organiza tional research, an area which has yet to fully integrate SDT as a valuable theory of motivation. SDT ar gues that motivation, regulated by an innerintention to act (intrinsic regulation) or prompted by s ources external (extrinsic regulation) to the individual but largely congr uent with his or her values and beliefs, will maximize well-being and performance (Gagn & Deci, 2005). For the latter conditions to be achieved, the three primary needs of au tonomy, competence, and relatedness must be satisfied by the social environment (Deci & Ryan, 2000; Ryan & Deci, 2000). Leaders play a critical role in sati sfying needs since they can set the tone for a supportive social environment (Kaiser, Hogan & Craig, 2008; Kozlowski & Doherty, 1989). Despite SDT being a potentially fruitful theoretical explanation for how leaders influence their followers, it has been applied to very few leadership studies. Only one study (Mayer, Bardes, & Piccolo, 2008) has utilized SDT w ithin the literature on SL, though several scholars have defined SL as behaviors whic h fulfill followersÂ’ needs (e.g., Graham, 1991; Liden et al, 2008; Spears, 1995). With regard s to SDT, this study will serve as an informative quantitative study, act as an ambassador of SDT for organizational and leadership scholarship, and answer a call by Gagn and Deci (2005) for further research using SDT within organizational contexts. By placing priority with the follower, SL can play a pivotal role in the fulfillment of basic human needs, specifically those defined by SDT which lend to enhanced workplace outcomes.
5 Servant Leadership The concept and term Â“servant leaderÂ” was coined by Greenleaf (1970) as an experienced-based, atheoretical conception of best-practices for leadership behaviors. According to Greenleaf, SL promotes followersÂ’ well-being by fulfilling basic human needs and emphasizing the necessity of moral safeguards to guide responsible leadership behavior. SL stands in stark contrast to ma ny other typologies of leadership behaviors which place leaders as servants to organizations or followers as servants to their leaders. Much of the scholarship in leadership has proposed a lead er-centric perspective in which followers serve the motives, growth, a nd success of leaders a nd their organizations (Stone, Russell, & Patters on, 2004). The leadersÂ’ motives, growth, and success place priority with serving the organization over em ployees, due in part to their ostensible performance being contingent on the perfor mance of the organization (Hogan, Curphy & Hogan, 1994; Smith, Carson, & Alexander, 198 4). Such an approach to improved organizational outcomes may fail to manifest the full capabilities and enthusiasm of its employees, if the employeesÂ’ basic needs and growth are underserved. In contrast to leader-centric perspectives, SL is character ized by behaviors serving followers which promote the followersÂ’ growth and need sa tisfaction through ethical motives and means (Greenleaf, 1970; Graham, 1991; Spears, 1995). Thus, servant leaders espouse a follower-centric leadership approach (B arbuto & Wheeler, 2006; Stone, Russell, & Patterson, 2004) in which followers are valued, in a Kantian se nse, as ends rather than merely as means to an end. Distinguishing SL from other types of leadership Though other types of leadership behaviors and relati onships are associated with each of the criteria examined
6 in the proposed model (e.g., job performan ce, subordinate well-being, and social cohesion), they fail to satisfy all the crit eria simultaneously. Notably, SL has accounted for incremental variance over similar leadersh ip styles in subordinate outcomes with regard to job performance, prosocial work behaviors, job satis faction, leader-member exchange, affective organizational commitme nt and the impact of transformational leadership (Barbuto & Wheeler, 2006; Ehrhar t, 2004; Liden et al., 2008). This suggests that, above all else, servant l eaders are uniquely effective at promoting the well-being and success of their subordinates and encouraging stewardshi p within the organization. Seeking to further define SL as a distin ct set of leadership behaviors begs the question, Â“Do we need yet another construct to describe leader behaviors?Â” The answer to this is largely contingent on the incremental utility that a leadersh ip style, or group of behaviors, offers when understood in the cont ext of its related counterparts. Past theory and research supports the claim that SL is a unique construct, specifically in the context of three of its closest relatives, transforma tional leadership, charismatic leadership, and ethical leadership. Social exchange theory (Blau, 1964) and social learning theory (Bandura, 1977, 1986) will be used to buttress the empirical evidence for the discriminant validity of the construct of SL. Based upon social exchange theory (Blau, 1964), SL followers will be more likely to respond positively to the leaderÂ’s directiv es because they are receiving growth and need fulfillment in return for their performa nce. More specifically, Blau proposed that individuals act according to norms of behavior and self-interes t. These are attributed to a norm of reciprocity in which we expect others to respond to us in a similar fashion as we respond to them. Expecting reciprocation fr om others requires tr ust. Thus, social
7 exchange theory argues that in the presence of trust, we will give so long as we are given to (Blau). A primary tenet of SL is the leaderÂ’s investment in othersÂ’ success. SL has also been linked with improvements in subordinate sÂ’ trust in their l eader and organization (Joseph & Winston, 2005). Thus, it follows that SL provides an environment of trust in which dyadic exchanges are encouraged as serv ant leaders invest in their followers and their followers reciprocate the investment. The organismic social exchange between servant leaders and followers also highlights th e generative nature of SL, a characteristic not common to other leadership styles. Based upon social learning theory (Bandur a, 1977, 1986), servant leaders serve as models of desired values and behaviors, increasing the odds that followers will internalize and adopt similar va lues and behaviors. With the passage of time, followers are promoted within organizations and beco me the organizationsÂ’ newest leaders. SL boasts a generative approach to the prom otion of followersÂ’ growth and well-being. According to social learning theory, as pa st followers seek to achieve success and promotion within an organization, they ar e likely to model the behaviors of their successful leaders. Thus, se rvant leaders will foster SL behaviors amongst their followers. With regard to the model proposed in this study, social learning theory and SDT suggests that as followers model and adop t the behaviors of their servant leaders, greater internalization of SL values and behaviors will occur. According to SDT, enhanced follower outcomes will ensue as followers mirror the values and behaviors of a servant leader. Graham (1991) proposed that leadership guided by a concern for behavioral ethics is a distinctive quality of SL which is not cap tured by other leadership constructs, such as
8 transformational and charismatic leadership. Th e lack of ethical consideration has taken center-stage more recently as scholars (P rice, 2003) have attempted to identify transformational and charismatic leadership behaviors which are a nd are not ethicallydriven (e.g. authentic, inauthentic). Though tr ansformational leadership (Avolio & Bass, 1995; Bass, 2000; Bass & Steidlme ier, 1999) seeks to evoke change in followers, it does not require leaders to consider the ethicality of the aims and the actions prescribed by them. According to social learning theor y, leaders who are transformational and/or charismatic but lack regard for ethics will be less likely to promote ethical behavior amongst followers. Since the leader is not m odeling ethicality, in their values and/or behavior, the follower will not have the opportuni ty to imitate leader-driven ethicality. The absence of an ethical focus in transformational and charismatic leadership is also of concern when considering social exch ange theory. If the follower does not feel the leader acts ethically, then they may be less lik ely to engage in exchange relations which are notably ethical or just. Apathy for ethics within social exchanges may extend beyond the leader-follower dyadic relationship a nd lend to general unethical behavior by followers. This may be manifested in a variet y of organizational outcomes, such as higher occurrences of counterproductive work beha viors, reduced commitment and workplace safety and lower levels of corporate social responsibility. As a specific example of this, Walumbwa, Hartnell, and Ok e (2010) found that SL beha viors increased prosocial behaviors amongst employees and this relationship was amplified when the organizationÂ’s procedural justice climate and positive service climate were strong. SL, unlike transformational and charismatic lead ership, attends to ethics as a central characteristic of the leaderÂ’s motivation a nd behavior. Thus, SL, according to both social
9 learning theory and social exchange theo ry, will engender greater ethical behavior amongst followers and within social interactions, respectively. It should be noted that the initial co nceptualization of transformational and charismatic leadership did address ethicality. BurnsÂ’ (1978) original conceptualization of transformational leadership, which incorpor ated key characteristics of charismatic leadership, included ethics and just behavior as a co mponent of an effective transformational leader. However, the main tenance of ethics as a transformational leadership characteristic has not been mainta ined in its contemporary usage, notably one guided by Avolio and Bass (1995). Focusi ng on ethical behavior, SL provides incremental utility and is notably unique from transformational and charismatic leadership. Beyond ethics, transformational, charismatic and servant leaders seek to engender change by motivating followers to act based upon an internal desire to do so. The latter assumes followers want to or have a desire to change. Graham (1991), in his review of various leadership styles in cluding SL, notes that followers of transformational and charismatic leaders may not want to change. Since transforma tional leadership focuses its primary aims at the organization, with th e exception of the idealized consideration dimension which is partially follower-focused, followers have little incentive to intrinsically desire growth and transformati on in the direction extrinsically promoted by their transformational leader. Social exchange theory suppor ts the notion that when the leaderÂ’s organizational goals are not aligned/focused on the followersÂ’ personal goals, misalignment occurs. Such a misalignment in aims will lead to fewer opportunities for reciprocation since leaders serving their orga nization may not serve the follower in such a
10 way that the follower feels they Â“oweÂ” some thing back to the leader. SL however is follower-focused. Thus, servant leaders will be more likely to have their aims aligned with those of the follower, prompti ng a greater exchange relationship. In the 1940Â’s and 50Â’, the Ohio Stat e Leadership Study identified two complimentary leadership factors: initiati ng structure and indivi dualized consideration (Fleishman, 1953). Building on this and Burn sÂ’ work (1978), Avolio and Bass (1995) developed a model of transformational leadersh ip (as discussed in pr evious paragraphs). They defined transformational leadership acco rding to four subdimensions, one of which is individual consideration. Contemporary usages of consideration le adership, both as a dimension of transformational l eadership and as its own cons truct, define it as Â“the degree to which a leader shows concern and respect for followers, looks out for their welfare, and expresses appreciation and supportÂ” (Judge, Piccolo, & Ilies, 2004, p. 36). Given its groundings in theoretical precursors to and current usages of transformational leadership, it is believed that considerati on leadership is unique from SL in ways mirroring transformational and charismatic leadership. Similar to SL, a supervisor high in cons ideration leadership will show explicit concern for and empathize with their subor dinates. This behavior should foster a relationship high in trust and liking. Meta -analytic evidence demonstrates that consideration leadership is linked with in creased job satisfaction and motivation for followers (Judge, Piccolo, & Ilies, 2004). Soci al exchange theory would suggest that these outcomes are encouraged by the forging of trust between supervisors and subordinates and the fulfillment of the subordina tesÂ’ growth needs. However, considerate leaders are defined by their empathy and forg ing of friendship with their followers, not
11 necessarily by their active role in enabling the followersÂ’ growth and needs fulfillment. The latter are defined as important precurs ors to a successful social exchange (Blau, 1964). In this way, SL is differentiated from the relationship-based model of consideration leadership. Namely, servant leaders not only empathize and build quality relationships with followers, but servant le aders also actively seek out followers to promote their welfare and grow th. Empirical evidence supports this; SL has been linked to behavioral job performance indices (L iden, Wayne, Zhao, & Henderson, 2008) and job attitudes (Barbuto & Wheeler, 2005; Maye r, Bardes, & Piccolo, 2008) whereas consideration leadership is often associated with the latter (job attitudes) and only inconsistently with the former (job perfor mance; Judge, Piccolo, & Ilies). Given these findings, it may be that SL prompts a true exchange relationship by creating a bidirectional and reciprocal relationship be tween supervisors and subordinates whereas consideration leadership is more unidirec tional, with the subor dinate serving the supervisor and not vice versa. More specifi cally, it is unidirecti onal in that though the considerate leader empathizes with the follo wer, the follower does not necessary perceive the leader acting on his or her empathy a nd returning the investment. Consideration leadership may be a necessary component of SL but not sufficient. SL, in this way, describes a higher level of functioning for a l eader in which the leader plays an active role in the supervisor-su bordinate relationship rather than a passive role. Social learning theory provides another explanation for why considerate leaders may promote satisfaction and motivation but not performance, whereas SL enables both. Consideration leadership prioritizes the relatio nal aspects of leader-follower interactions above more explicit performance indices. T hus, followers do not necessarily have a
12 leader exemplifying top performance, but rather high quality relationship building. Without such an example, the follower cannot model high performance behaviors. Conversely, servant leaders ex emplify both top performance within and outside of the organization, high standards and relationshi p building. This provides followers both an environment/organizational culture (e.g., Mayer, Kuenzi, Greenbaum, Bardes, & Salvador, 2009) in which to work and a specific example (e.g. the leader and/or supervisor) of top performance, behavior based on high ethical standards and positive attitudes. Again, this suggests that consideration leadership may be a necessary but not sufficient component of SL, and thus defi nes a different type of follower-focused leadership behaviors. Unlike transformational and consideration leadership which circumvent the topic of whether leaders act ethically, ethical l eadership is fundamentally defined by the standards that guide a leader. By focusing on ethics, ethical leadership is another construct which closely resembles SL. Ethical leaders value and act with a concern for ethical behavior which encourages followers to pursue ethical ends (Brown, Trevino, & Harrison, 2005). However, ethical leadersh ip fails to capture the empowering and transformational qualities of SL and transf ormational leadership. From a motivational perspective, though ethical leaders will model appropriate behavior and encourage ethical exchanges between individuals, predicted by soci al learning and social exchange theories, respectively, they will be less effective at convincing followers to embrace their values. The latter is critical for effective leader ship and explains why certain leadership behaviors are considered more effective (e .g. transformational leadership) at prompting
13 change than others. Thus, while related to et hical leadership via a concern for ethics, SL is unique in its dual-focus on ethical behavior and effec tive motivational guidance. In sum, SL can be differentiated from other leadership styles, most notably transformational, charismatic, consideration and ethical leadership (Graham, 1991) in three primary ways. First, servant leaders s eek to identify with their subordinates and followers whereas other leadership styles (e.g. transformational le adership, charismatic leadership) prioritize identific ation with the organization. S econd, SL seeks to serve the organizationÂ’s employees and the greater co mmunity. Thus, servant leaders have an outward focus which aims to fulfill needs for relatedness amongst their subordinates by forming collective and relational support networks in an orga nizational context and in the context of the larger social community/soc iety. Third, SL behaviors are guided by a moral compass which seeks responsibility a nd accountability. W ith the exception of ethical leadership, SL presents the only leadership approach claiming a moral motivational component. SL goes beyond ethical leadership by modeling ethical behavior and by prompting motivated change to enhan ce the welfare of followers by satisfying their basic needs. Servant Leadership, Needs, and a Framework for Understanding Needs Central to the philosophy of SL (Graham, 1991; Spears, 1995) and consistent with Mayer, Bardes and Piccolo (2008), serv ant leaders impact followersÂ’ needs by prioritizing the welfare and growth of follo wers. Thus, it is somewhat surprising that amongst publications on SL, Mayer and coll eagues are the only scholars to examine needs as an explanatory mechanism thr ough which servant leaders enhance their followersÂ’ welfare.
14 According to SDT, there are three basic needs: autonomy, competence and relatedness. When these needs are met, indi viduals experience maximized performance and well-being (Gagn & Deci, 2005). Serv ant leaders seek to promote follower autonomy by empowering them to achieve both success and challenging goals via conceptualization (Spears, 1995). Follower sÂ’ need for competence is supported by servant leaders as they actively listen to th eir followers, express empathy and awareness of their intrinsic worth and e xperiences and pursue change th rough persuasive rather than coercive means. Notably, using persuasive t echniques, rather than directives, suggests to followers that their leader values their competence and autonomy as an individual with independent thoughts and desires; coercive techniques imply that the follower is not capable of independent thought. Lastly, serv ant leaders encourage relatedness by building a community based upon stewardship, empathic c oncern for others and an interest in the generative growth and success of members of the community (Spears, 1995; Barbuto & Wheeler, 2006). SDT (Deci & Ryan, 2000; Ryan & Deci, 2000) provides a theoretical explanation (e.g. motivational regulation) and needs-ba sed mechanism (e.g. autonomy, relatedness, competence) for explaining how SL lends to heightened job performance, well-being and cohesive social relationships According to SDT, when followers experience greater needs fulfillment, they are more likely to act according to intrinsic motivation or wholly internalized extrinsic motiva tions (e.g., identified and in tegrated regulation; Deci & Ryan, 1985). Behaviors founded in such self-d etermined motivations are more likely to lead to positive and intended outcomes for the individual.
15 Self-Determination Theory When performing voluntary actions, an indi vidualÂ’s motivational state serves as an explanatory mechanism for his or her beha viors. SDT is a needs-based, organismic theory which seeks to explain the relati onship between the process of motivational regulation and subsequent acts performed (Deci & Ryan, 2000; Gagn & Deci, 2005). Motivation occurs when an individual has the volition to and in tends to perform an action or pursue a goal. Conversely, amotivation occurs when an individual lacks intentional regulation (intent) and motiv ation. Intent distinguishes motivated and amotivated regulation. SDT rests on the assumption that humans pursue existential aims such that they are active and growth-seeking beings regula ting behaviors according to intrinsically accepted values and beliefs and extrinsic environmental forces (Deci & Ryan, 1985; 2000). Behaviors are more likely pursued when they are intrinsically enjoyable and allow for interpersonal and intrapersonal feelings of alignment between the self and other values, goals and attitudes. The extent to which individuals have the inner and/or environmental resources to pursue such exis tential aims depends on the satisfaction of their needs for autonomy, competence and re latedness (also termed connectiveness). When these needs are supported, the indivi dual is free to pursue self-determined intentions. However, when situations ar e overly-controlling, re quire awareness or knowledge yet known or are alienating, indivi duals will experience reduced selfmotivation, relying on extrinsic factors to guide and regulate their behavior (Deci & Ryan). The satisfaction of the needs fo r autonomy, competence and relatedness are a primary means by which servant leaders foster growth and well-being in their followers.
16 Needs. SDT posits three basic needs: au tonomy, competence, and relatedness (Deci & Ryan, 1985; 2005). Ryan, Sheldon, Kasse r and Deci (1996) define needs as universal necessities which are not judged by the degr ee of their conseque nces but rather by the extent to which a social environmen t satiates the needs (Gagn & Deci, 2005). Autonomy Â’s central position within SDT ex tends beyond the intrinsic/extrinsic motivation regulatory level to also include its primacy amongst the three needs. The satisfaction of an individualÂ’s need for aut onomy dictates the extent to which motivation is internalized from extrinsic sources or in trinsically regulated. T hus, greater autonomy support lends to greater intrin sic and integrated motivation. Competence is characterized by a need to be challenged and experience ma stery and efficacy over social and physical interactions (Deci & Ryan, 2000) Fulfilling a need for competence instills a sense of purpose and role fulfillment for individuals. A need for relatedness is satisfied when an individual feels a sense of s ecurity, attachment, belonging a nd a level of intimacy with significant others in dyadic pairs and social groups (Deci & Ryan). Relatedness is particularly important when discussing extr insic motivation since a primary external force motivating individuals originates with significant others in their lives. In a work setting, organizational climates which encourage fulfillment of and support autonomy, competence, and relatedness needs Â“will e nhance employeesÂ’ intrinsic motivation and promote full internalization of extrinsic motivation,Â” (Gagne & Deci, 2005, p. 337). Specifically, work climates prom oting the three psychological needs are associated with numerous positive work outco mes, such as increased levels of job satisfaction (Deci, Connell, & Ryan, 1989), job performance (Baard, Deci, & Ryan, 2004), organizational trust (Deci Connell, & Ryan), goal attainment (Koestner, Otis,
17 Powers, Pelletier, & Gagnon, 2008), organiza tional commitment (Gray & Wilson, 2008), physical and psychological well-being (B aard, Deci, & Ryan; Reinboth & Duda, 2006), and prosocial behaviors (Gagn, 2003). Regulatory Motivation Motivation to pursue a goal or act can arise from either or both extrinsic and intrinsi c sources. Behaviors are intrinsically motivated when (1) an individual has an inherent interest in and enj oys a task and (2) when the intent to perform a task originates intern ally (Gagn & Deci, 2005) Conversely, behaviors motivated by environmental and social forces external to the self are extrinsically motivated. Not all extrinsic motivations are created equal, how ever. In some instances, the individual may accept or identify with the motivation prescrib ed by the externally occurring presses, whereas in others, the individual may experien ce dissonance between his or her intrinsic values and extrinsic motivational sources. With time, certain extrinsic motivators may come to be accepted, endorsed, and internalized by individuals. Th e process by which individuals (1 ) identify with the social regulations, (2) endorse and assimilate extrin sically regulated motivations, and (3) accept the regulatory process as one congruent w ith their beliefs and values is termed internalization (Deci & Ryan, 2000). Thus, an employ ee may start a job because of its salary, but over time may come to internaliz e the norms and values of the company such that they are no longer motivated solely by salary but also by a heightened sense of autonomy and intrinsic interest. Importantly, internalized extrinsi c motivation is often just as powerful a motivator as in trinsic motivation for individuals. The level of autonomy gauges the extent to which behavior is self-determined. An individual acting solely according to intr insic motivation is considered wholly
18 autonomous. Conversely, when individuals act according to operant contingencies (e.g. seeking reward or avoiding punishment) defi ned by extrinsic sources, their need for autonomy goes unmet (Gagn & Deci, 2005; Ry an & Deci, 2000) and their motivation is said to be controlled. For example, a superv isor threatening job loss for a subordinate if s/he does not complete a project by a speci fied deadline will likely result in the subordinate experiencing larg ely controlled extrinsic motiv ation. In this example, the extrinsic source of motivation was a Â“signi ficant otherÂ” in a dyadic work-based relationship, namely a supervisor, for the subordinate. As an important significant other, leaders play a key role in regulating followersÂ’ external motivation. Leaders who satiate a followerÂ’s needs for autonomy, competence and relatedness will likely see positive outcome s related to the followerÂ’s well-being and performance compared to leaders who fail to consider the needs of their followers. For example, Bono and Judge (2003) reported me ta-analytic evidence that transformational leaders were effective at promoting autonom y-oriented goals and prompting enhanced work outcomes. This suggests that leadership styles, such as SL and transformational leadership, which promote the internalizati on of goals by followers, may also foster greater support for SDT needs and resulting organizational outcomes (Gagn & Deci, 2005). Mayer, Bardes, and Piccolo (2008) provi de direct support for the relationship between SDT needs fulfillment and SL behaviors by supervisors. Paralleling this finding, Walumbwa, Hartnell, and Oke ( 2010) found that displays of SL behaviors by supervisors led to increased employee self-efficacy; with efficacy being a component of a need for competence. As an additional empirical founda tion for this burgeoni ng area of research, Washington, Sutton, and Feild (2006) reported th at followers who perceived their leader
19 as valuing empathy, competence and integrity (a form of ethical consideration) also attributed more SL behaviors to the leader. Given these three initial studiesÂ’ findings, the time is ripe for research establishing re lationships amongst SL, SDT needs fulfillment and organizational performance, wellbeing and social cohesion outcomes. Organizational Outcomes In sum, few studies have examined SDT and leadership and how SL relates to organizational outcomes. The relationships be tween SDT theory and many organizational outcomes are also in need of empirical verification (Gagn & Deci, 2005). Empirical support is mounting for the claim that SL promotes followersÂ’ SDT need satisfaction, lending to enhanced organizational outcomes. Specifically, SL is related to several positive employee outcomes, including increased job performance and satisfaction, prosocial work behaviors, leader-membe r exchange, and affective organizational commitment (Barbuto & Wheeler, 2006; Ehrhart, 2004; Liden, Wayne, Zhao, & Henderson, 2008). Leader-member exchange and organizational commitment are important for building social cohesiveness w ithin an organization. Additionally, SL has been shown to promote greater SDT needs satisfaction amongst employees (Mayer, Bardes, & Piccolo, 2008). As noted previ ously, employee SDT n eeds satisfaction has been linked with a number of desirable wo rkplace outcomes (e.g. job satisfaction and performance, commitment, well-being and prosocial behaviors). The relationships supported by previous empirical wo rk are highlighted in Table 1.
20 Table 1. Empirical relationships amongst SD T needs, SL, and organizational outcomes Organizational Outcomes SL SDT Needs Job (task) Performance X X CWB X Prosocial Behaviors X X Physical Well-Being X X Psychological Well-Being X Job Satisfaction X X Leader-Member Exchange X Commitment X X Justice Perceptions X Trust in Supervisor/Organization X Goal Attainment X Based upon theoretical considerations and previous empirical evidence of relationships amongst construc ts, the organizational outcomes examined in this study include: in-role task perfor mance, deviant and prosocia l behaviors, physical and psychological well-being, supervisor-subordi nate (leader-member) relationship quality, job satisfaction, and affective organizatio nal commitment. These variables were organized into three clusters for conceptual and empirical reasons: job performance, wellbeing, and social cohesion. Job Performance includes three indicesÂ—in-role ta sk performance, organizational citizenship behaviors, and c ounterproductive work behavior sÂ—of employeesÂ’ fulfillment of in-role and extra-role job tasks. In-role task performance refers to tasks explicitly stated and formally required by the job role an employee fulfills. Prosocial and deviant work behaviors performed beyond oneÂ’s assigned job duties are captured by organizational citizenship behaviors a nd counterproductive work behaviors. Organizational citiz enship behaviors (OCB) are prosocial behaviors performed by employees which are incremental to an indivi dualÂ’s defined job ta sks and roles, thus
21 constituting extra-role voluntary beha viors (Borman, & Motowidlo, 1993; Organ & Ryan, 1995). Counterproductive work behaviors (CWB) include any actions which a person intentionally pe rforms (though it need not be perf ormed with the intent to do harm) that harm the organization by distract ing from its goals, well-being and/or norms (Spector & Fox, 2005). Thus, taken together, ta sk performance assesses the explicitly specified expectations a supervisor has of his or her employee and CWB and OCB offer measures of extra-role behaviors, both deviant and be neficial, an employee may display on the job. Together, the three cr iteria lend to an overall asse ssment of an employeeÂ’s job performance. The second cluster of outcome vari ables includes those addressing the well-being of the subordinate within a work context. Physical well-being, psychological well-being and job satisfaction are hypothe sized as indicators of a subor dinateÂ’s overall well-being. Physical well-being describes the prevalen ce of physical/somatic symptoms perceived by an individual such that reduced well-being is contributed to by the pr evalence of physical symptoms (e.g., headache; Spector & Jex, 1995). Similarly, psychological well-being refers to the psychological components of health as they are perceive d by an individual, such as the presence of depression or a nxiety. Psychological wellbeing can be broken down into three subdimensions: depressive sy mptoms, anxiety symptoms and irritability symptoms (Caplan, Cobb, French, Van Harri son, & Pinneau, 1980). Often, psychological and physical symptoms are intertwined such that the perception of physical symptoms involves a largely psychologi cal component and feeling ps ychological symptoms may be related to the presen ce of a physical symptom (e.g. a he adache and feeling irritable; Spector & Jex). Job satisfaction is frequently referenced as an inde x of oneÂ’s overall life
22 satisfaction and well-being, with work bei ng a major component of oneÂ’s life. Thus, oneÂ’s overall affective experience and appraisal of his or her job, termed job satisfaction is associated with the per ception of physical and psychological symptoms. For example, if headaches at work are frequent for an em ployee, s/he may be less satisfied at work because it takes him or her longer to complete hi s or her job tasks and s/he must deal with additional psychosomatic stressors throughout the day. Together physical, psychological and job satisfaction are hypot hesized to contribute to an overall assessment of a subordinateÂ’s well-being. The final grouping of variables is termed social cohesion This includes the quality of the supervisor-subordinate re lationship (leader-member exchange), an employeeÂ’s affective organizational commitment and prosocial behaviors targeted toward the community. Leader-member exchange (LMX) addresses the de gree to which a leader and his or her follower share a high quality relationship characterized by mutual trust, support, loyalty, and approval among part ners (Graen & Uhl-Bien, 1995). Key antecedents of high-quality LMX include interpersonal liking, fulfilling partnersÂ’ role expectations, and investing high levels of effo rt into relationships (Gerstner & Day, 1997; Graen & Uhl-Bien). Affective organiza tional commitment (AOC) involves an emotional attachment to, involvement in, and identif ication with oneÂ’s organization (Meyer & Allen, 1991). AOC arises from positive soci al exchanges between the employee and organization, which are based, in part, on perceptions of support amongst colleagues, supervisor-subordinate dyads and the organi zation (Eisenberger, Huntington, Hutchison, & Sowa, 1986; Shore, Tetrick, Lynch, & Barksdale, 2006). Community prosocial behaviors (CPB) are acts which an i ndividual performs that se rve the community within
23 which his or her organization is located, such as by volunteering at a food bank or helping to coach little league (Liden Wayne, & Zhao, & Henderson, 2008). Taken together, LMX, AOC and CPB indicate an employeeÂ’s commitment to and perceived membership (e.g., cohesion) in the organiza tion, larger community and relationships within the organization. The three factors of job performance, well-being and social cohesion are expected, based upon theoretical and empirical rationale, to be predicted by supervisorsÂ’ SL behaviors and the fulfillment of the subor dinatesÂ’ SDT needs within their employing organizations. Thus, the following relationships are proposed: Hypothesis 1: Servant leadership behavio rs will be positively related to the satisfaction of subordinatesÂ’ self-r eported needs for (a) autonomy, (b) competence, and (c) relatedness. Hypothesis 2: The fulfillment of subordi natesÂ’ needs will be positively related to subordinatesÂ’ (a) job performance, (b ) well-being, and (c) social cohesion. Hypothesis 3: The relationship betw een servant leadership and (a) job performance, (b) well-be ing, and (c) social cohesi on will be mediated by subordinatesÂ’ needs satisfaction. Taken together, the relationships propos ed by Hypotheses 1-3 are illustrated in Figure 1. To test this model, survey data from supervisor-subor dinate dyads within several organizations were coll ected and analyzed using stru ctural equation modeling and hierarchical regression. Both the use of an applied sample and holistic model analysis will buttress the utility and validity of the re lationships between SL, SDT, and workplace outcomes.
24 Figure 1. Hypothesized model of SL, SD T needs, and organizational outcomes *Subordinate self-report **Supervisor-rating of subordinate Servant Leadershi p SDT Needs -Autonomy -Relatedness -Competence Job Performance -Task ** -CWB -OCB ** Well-being -Physical well-being -Psychological well-being -Job Satisfaction Social Cohesion Interpersonal -Leader-Member Exchange Group -Affective commitment -Community prosocial behavior
25 Method Participants Supervisor-subordinate matched dyads were recruited from three businesses in the Southern US agreeing to take part in this study. Diversity was s ought when recruiting organizations in order to increase the studyÂ’ s external validity. Th e three organizations included a private school serving grades kindergarten through 12th-grade (Organization A), a small private college (Organization B), and a medium-sized law firm (Organization C). Additional surveys were distributed to employees who were enrolled in undergraduate courses at a large public univers ity in the Southeastern US (Organization D). All participants worked a minimum of 20 hours each week, with the majority of participants working full-time. Table 2 repor ts the number of surveys distributed within each organization, the number returned, and the computed response rate. In total, 442 subordinates were contacted via their respec tive organizational affiliation. Of these, 216 subordinates returned completed surveys, for a response rate of 48.87%. In total, 147 surveys were completed by both a subordinate an d his or her supervisor, resulting in an overall response rate (out of the total initially recruite d) of 33.26%. Of the 147 useable supervisor-subordinate pairs, 13.61% were from Organization A, 38.10% were from Organization B, 15.65% came from Organiza tion C, and 32.65% were from Organization D.
26 Table 2. Response rates of recruited participants Organization Total Recruited # of Subordinates Responding (Response Rate) # with Supervisor Responses (Response Rate) Overall Response Rate: Pairs/Total Recruited # of Supervisors Contacted # of Supervisors Responding (Response Rate) A 49 23 (46.93%) 20 (86.96%) 40.82% 4 4 (100%) B 274 85 (31.02%) 57 (67.06%) 20.80% 41 29 (70.73%) C 24 24 (100%) 23 (95.83%) 95.83% 14 14 (93.33%) D 95 84 (88.42%) 47 (55.95%) 49.47% 84 47 (55.95%) Total 442 216 (48.87%) 147 (68.06%) 33.26% 143 94 (65.73%)
27 Subordinates were recruited via e-mail th rough their organizational affiliation to complete a paper and pencil survey packet for this study. Top level administrators (e.g. School Headmaster, College President, Firm Partner) informed their employees at Organizations A, B, and C of their approval and endorsement of this study in order to encourage employeesÂ’ participation. Previous st udies have demonstrated that the support of executive leadership facilitates employees Â’ willingness to partic ipate in extra-role activities at work (Mayer, Kuenzi, Gree nbaum, Bardes, & Salvador, 2009), such as participating in a voluntary survey study. T hus, the executive managementÂ’s support and endorsement of this study were considered crit ical at the organizati on level. In exchange for participating in this study, each organiza tion received a technica l report with this studyÂ’s findings and conclusions. Employees were assured that their orga nization would not have access to their individual responses and that all results would be presen ted to their organization as aggregated, unidentifiable findi ngs only. Thus, subordinates a nd supervisors were assured confidentiality and anonymit y. Each survey package was prestamped with a unique identification number. The first page of the survey included information on the study and informed consent. Subordinates agreeing to pa rticipate provided their name and signature on this first page along with th eir supervisorÂ’s (i.e., the in dividual who would complete a performance evaluation of them) name and e-ma il. Upon receipt of th e surveys, the first page was removed from the survey packet, ensuring the survey re sponses would only be linked with the subordinatesÂ’ unique identification numbers and not their names. This method allowed for subordinatesÂ’ responses to be kept confidential and anonymous. Identified supervisors were then contacted vi a e-mail to participate. In the e-mail, the
28 supervisors were instructed to complete a brief online survey using a provided URL, were provided with the subordinateÂ’s name and hi s/her unique identificat ion number, and were assigned an identification number to en sure anonymity on the online survey. The supervisors entered their identi fication number at the start of the online survey along with the identification number of thei r subordinates, as prompted by the surveyÂ’s instructions. The option was given to supervisors in Orga nization D to complete a paper-and-pencil survey identical to the online version. In the case that a supervisor preferred a paper-andpencil survey, a survey packet was provide d to them along with a preaddressed and Â– stamped envelope. Supervisors were instructed to mail the survey directly back to the researcher to maintain confidentiality. Subordinates, with useable data ( N = 147), were mostly female (72.1%), on average 40.75 years old ( SD = 16.37 years), were predominantly Caucasian (79.6%; African American: 5.4%; Hispanic: 12.2%, As ian: .7%; Native American = .7%), were employed full-time (67.30%), and worked an average of 35.81 hours per week ( SD = 11.27). Subordinates worked mostly in an Educational/Academic Industry (49.3%), a professional industry (19.9%), or in a reta il/service industry (25.3%). Other industries included Manufacturing (.7%), Te chnical (1.4%), Government ( .7%), and othe r industries which failed to fit into the provided categories (2.7%). Overall, participants reported a high education level, with 34.7% holding a gr aduate degree, 20.4% having completed a bachelors degree as their highe st level of education, 22.4% an associateÂ’s degree, and 21.8% reported high school as their last degr ee earned. Because Organization D consisted of employed undergraduates, the la tter percentage is a reflection of this sample such that most undergraduates have yet to complete a degree beyond high school. Subordinates
29 reported being employed by their current orga nization for an average of 7 years and 10 months ( M = 93.94 months, SD = 102.85), working in their current position for 5 years and 10 months ( M = 69.80 months, SD = 73.60), and under their current supervisor for just over 4 years ( M = 48.37 months, SD = 52.08). Supervisors were mostly female (72.1 %), Caucasian (77.3%; African American: 5.0%; Hispanic: 17.0%), highly educated (highest degree earned: High School: 12.2%, AssociateÂ’s: 6.1%, BachelorÂ’s: 17.0%, Gr aduate: 59.2%), and were on average 47.97 years old ( SD = 12.64 years). All but two supervisors worked full-time (98.6%). Supervisors reported working an average of 49.81 hours each week ( SD = 9.24), and indicated the industry they work in as: Educational/Academic (51.7%), Professional (17.7%), Retail/Service (20.4%), Manufacturing (.7%), Techni cal (1.4%), or Government (.7%). Supervisors had worked within their current organization an average of 13 and a half years ( M = 161.26 months, SD = 117.96), in their current position 6 years and 8 months ( M = 80.26 months, SD = 69.16), and as the focal subordinateÂ’s supervisor for just over 4 years ( M = 50.20 months, SD = 48.26). Measures All survey scale responses were made on a 5-point Likert scale (1 = Â“strongly disagreeÂ”, 5 = Â“strongly agre e), unless otherwise noted. A ll scale items are listed in Appendix A with items removed during item-level analyses crossed-out. Servant leadership EhrhartÂ’s (2004) 14-item SL scale was completed by subordinates to assess their pe rceptions of their supervisorsÂ’ SL behaviors. Of the SL scales published in academic journals, Ehrhar tÂ’s was determined to be methodologically strongest based upon a validity study (Ehrhart ) and its use in recent empirical studies
30 (Mayer, Bardes, & Piccolo, 2008; Walumbwa, Hartnell, & Oke, 2010). EhrhartÂ’s Servant Leadership scale demonstrated good discrimi nate validity by accounting for incremental variance in work-relevant outcome variables ov er the conceptually si milar constructs of LMX and transformational leadership. Specifical ly, Ehrhart reported that SL shared a .62 correlation with LMX and correlations between .53 and .61 for the four dimensions of transformational leadership. EhrhartÂ’s reported confirmatory factor analysis (CFA) of the survey items demonstrated adequate fit. Fr om a practical standpoint, the length of other SL scales distracts from their use within applied organizational settings; specifically Page and WongÂ’s (2000), Dennis and WinstonÂ’s ( 2003) revised version of Page and WongÂ’s scale, Liden and colleagues (2008), and Bar buto and WheelerÂ’s (2006) each contain 100items, 20-items, 28-items, and 23-items, respect ively. Thus, EhrhartÂ’s scale is a more succinct measure of SL. Contrary to EhrhartÂ’s reported statistics, th e SL scale did not fair are well in this sample. An initial CFA conducted on all 14 ite ms of the scale showed poor fit. Based upon theoretical and conceptual considerations, a review of the scaleÂ’s item statistics from an exploratory factor analysis (EFA), a nd the factor loadings from the initial CFA, items 7, 8, 13 and 14 were removed. The revised 10-item scaleÂ’s internal reliability was improved by removing these four items ( = .93), as was the fit of the CFA based on the remaining 10 items. An example item is Â“My su pervisor works hard at findings ways to help others be the best they can be.Â” SDT needs. Subordinates self-report ed their needs satisfaction using a 21-item Needs at Work scale (Deci, Ryan, Gagn, Leone, Usunov, & Kornazheva, 2001) comprised of three subdimensions: Autonomy (7-items; Â“My feelings are taken into
31 consideration at work.Â”), Co mpetence (6-items; Â“People at work tell me I am good at what I do.Â”) and Relatedness (8-items; Â“I get along with people at work.Â”). Poor fit statistics and low factor loadings from a CFA and item statistics from a reliability analysis prompted the formation of a revise d version of the SDT scale at the subscale level. The overall SDT scale fared considerably worse than the subscales with regards to the CFA and reliability analysis. Thus, the decision was made to establish the scaleÂ’s factor structure at the subscal e level and, then, to create a composite scale, using subscale means, as an overall indicator of SDT. It ems 5 and 11 were removed from the Autonomy subscale and items 7, 16, and 18 were removed from the Relatedness subscale. These items demonstrated poor loadings and item statis tics. The latter may be a result of their wording, since they were all reverse-scored items. A common artifact of reverse-scored items is the creation of a second factor and/ or reduced loadings when other items are worded counter to reverse-scor ed items. A CFA for the revise d subscales had adequate fit and the internal reliabilities for Autonomy (5 items; = .71), Competence (6 items; = .72), and Relatedness (5 items; = .82) were acceptable. The subscales were used to test the revised version of the hypothesized m odel with SEM. A composite SDT scale ( = .84), using the subscalesÂ’ means, was created and used to test hypotheses. Job performance SubordinatesÂ’ job performanc e was measured using task performance ratings and checklists of CWB and OCB. Task performance was reported by supervisors using Williams and AndersonÂ’s (1991) 7-item scale (Â“Performs tasks expected of him/herÂ”). The initial factor structure, based upon a CFA, of the task performance scale demonstrated poor fit. Further item analysis, based upon factor
32 loadings and a CFA, resulted in the remova l of items 1 and 5. A follow-up reliability analysis ( = .82) and CFA on the revised 5-item scale demonstrated adequate fit. CWB was self-reported by subordinates usi ng a modified version of Spector, Fox, Bruursema, Goh, and KesslerÂ’s (2006) 33item short version CWB checklist. The checklist was shortened to consist of 8 ite ms reported on a frequency scale, such as Â“purposely did your work incorrectlyÂ” and Â“b lamed someone at work for error you madeÂ” ( = .65). Supervisors rated their subor dinatesÂ’ OCBs using the Organizational Citizenship Behavior checklist (OCB-C) developed by Fox, Spector, Goh, Bruursema, and Kessler (2009). Based upon the original 42-item scale, 8-items were selected for study relevance, such as Â“offered suggestions for improving the work environment.Â” and Â“took time to advise, coach, or mentor a co-workerÂ”. Furt her analysis of an initial CFA and item reliability statistics resulted in the rem oval of item 7, Â“volunteered for extra work assignmentÂ”. This item may have been problematic given the sample used since the majority of subordinates were salaried workers. Salaried workers may not view volunteering for an extra work a ssignment as an extra-role be havior, and rather think of it as part of their in-role jo b tasks. Scale reliability ( = .84) and CFA fit for the revised 7item scale were acceptable. Supervisors and s ubordinates were inst ructed to report how often the subordinate performed each OCB a nd CWB, respectively, in their current job using a frequency scale (1 = Â“NeverÂ”, 2 = Â“Once or TwiceÂ”, 3 = Â“Once or twice per monthÂ”, 4 = Â“Once or twice per weekÂ”, and 5 = Â“Every dayÂ”). Well-being Subordinates completed self-repo rt measures of physical and psychological well-being and job satisfacti on. Physical well-being was measured using
33 Spector and Jex (1998) 18-item Physical Symptom Inventory (PSI) checklist ( = .81). Symptoms include Â“trouble sleepingÂ” and Â“l oss of appetite.Â” Consistent with Kessler, Spector, Chang, and Parr (2008), responses were modified from the originally validated dichotomous outcome scale (Spector & Jex) to a Likert-type scale with response options ranging from 1 (Â“ less than once per month or neverÂ” ) to 5 (Â“ several times per dayÂ” ). Psychological well-being was assessed us ing Caplan, Cobb, French, Van Harrison, and PinneauÂ’s (1980) 13-item scale, comprised of subscales for work-related Depression (6 items, = .90; Â“I feel sadÂ”), Anxiety (4 items, = .82; Â“I feel ner vousÂ”) and Irritation (3 items, = .83; Â“I get aggravatedÂ”). The scale was modified from its or iginal 4-point scale to a 5-point Likert-type frequency scale. Th e same response scale was used for both the physical and psychological well-being measures. Job satisfaction was measured using a 3-item self-report scale from the Michigan Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins, & Klesh, 1983; = .89). An example item is: Â“In ge neral, I like working here.Â” Social cohesion LMX, AOC, and CPB were used as indicators of subordinatesÂ’ feelings of social cohesion. S ubordinates self-reporte d their relationship quality with their supervisors using the LMX-7 scale (Gr aen, Novak, & Summerkamp, 1982, 7-items). Items 5 and 6 were removed from the scale following an assessment of a CFA and internal reliability analysis. Both items address ways in which a supervisor and subordinate feel responsible for the other. These items may not be appropriate in the organizations sampled since each was comprised of employees operating relatively autonomously and within horizontal organizatio ns in which subordinates and supervisors work as colleagues more often than as a hi erarchical dyad. The revised 5-item LMX scale
34 had good internal reliability ( = .91). An example item is Â“I feel that my immediate supervisor fully recognizes my potentialÂ”. AOC was measured with s ubordinate self-reports usi ng Meyer and AllenÂ’s (1997) 6-item affective commitment scale ( = .87; e.g., Â“This organiza tion has a great deal of personal meaning to meÂ”). CPB was assessed using the 7-item Co mmunity Prosocial Behavior scale developed by Liden, Wayne, Zhoa and Henderson (2008), = .82. An example item is Â“I believe it is important to give back to the community.Â” Covariates Age, education, subordinatesÂ’ tenu re with their current organization and their organizational affiliation (e.g. employ ee at Organization A, B, C, or D) were controlled for as potential c ovariates in the regression an alyses based upon statistical evidence and theoretical rationale. Previous research has related age and organizational tenure to job satisfaction, amongst other focal variables, and has demonstrated that though age and tenure covary as a result of time, the two are distinct in their effects on job satisfaction (Bedeian, Ferris, & K acmar, 1992). Additionally, focal outcome variables, such as organizational commitment ha ve previously been a ssociated with oneÂ’s organization tenure such that commitment often increases the longer an individual remains with an organization (English, Morrison, & Chalon, 2010). Education and organizational affiliation were controlled for since the two covary within the sample used. For example, Organization D employees reported a lower level of education than did the employees at Organization B. Exam ination of the correlations between the demographic data and the focal variables al ong with follow-up analyses of between group
35 differences supported the use of organizati onal affiliation and education as control variables in this study. Exploratory relationships. To determine the discriminant validity of SL over and above similar leadership st yles, subordinates completed measures of consideration leadership and transformational/charismatic leadership. Consideration leadership was measured using Schriesheim and StogdillÂ’s (1975) Leadership Be havior Description Questionnaire revised version VII (LBDQ-VII). The 10-item scale demonstrated poor fit statistics during an initial CFA. Follow-up analyses examining item-level statistics resulted in the removal of items 3 and 4. Th e remaining 8-items had adequate internal reliability ( = .87) and improved fit statistics. An example item is: Â“My supervisor treats all group members as his/her equals.Â” Tran sformational/charismatic leadership was measured using the Global Transformati onal Leadership Scale (GTL) developed by Carless, Wearing, and Mann (2000). The 7-it em GTL scale had good internal reliability, = .92. An example item is: Â“My supervisor co mmunicates a clear and positive vision of the future.Â”
36 Results Zero-order correlations among the focal variables are reported along with descriptive statistics (e.g., means, standard deviations ) in Table 3. Relationships were generally in the expected directions. It is worth noting the high, positive correlations amongst the three leadership variables: SL with Consideration Leadership, r = .80, p < .01, SL with Transformational Leadership, r = .88, p < .01, and Consideration Leadership with Transformational Leadership, r = .87 p < .01. These high correlations suggest that it may be difficult to tease apart the differen ces in how these three constructs operate uniquely with regards to organizational outcomes.
37 Table 3. Zero-order correlations amongst focal variables. Leadership SDT Needs Job Performance Well-Being Job Attitudes CPB Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Ldr 1. SL (.93) 2. CL .80 (.87) 3. TL .88 .87 (.92) SDT 4. SDT .47 .46 .49 (.84) 5. Autonomy .48 .45 .51 .90 (.71) 6. Competence .38 .32 .35 .87 .69 (.72) 7. Relatedness .36 .42 .43 .85 .63 .60 (.82) J Perf 8. Task Perf. .20 .12 .22 .04 .07 05 -.02 (.82) 9. CWB -.13 -.19 -.15 -.15 -.13 -.15 -. 12 .20 (.65) 10. OCB .16 .10 .13 .11 .08 .11 .10 .31 .14 (.84) W-B 11. Physical .05 -.13 -.06 -.15 -.04 -.20 -. 16 .04 -.02 .20 (.81) 12. Depression -.22 -.26 -.24 -.40 -.29 -.37 -. 40 -.12 -.26 .04 .45 (.90) 13. Anxiety .02 -.11 -.06 -.09 -.02 -.08 -.14 .02 -.05 .14 .53 .53 (.81) 14. Irritation -.14 -.20 -.17 -.16 -.11 -.14 -.17 12 -.09 .14 .34 .38 .45 (.83) J Att 15. LMX .82 .80 .82 .57 .59 .44 .45 .20 -.10 .14 -.02 -.26 -.05 -.09 (.91) 16. AOC .38 .39 .36 .67 .56 66 .53 .12 -.20 .19 -.13 -. 39 -.08 -.12 .37 (.87) 17. Job Sat. .35 .36 .36 .75 65 .65 .67 .00 -.23 .04 -.18 -. 41 -.11 -.18 .44 .71 (.89) 18. CPB .22 .17 .15 .30 .27 27 .23 -.00 -.17 .10 .02 -.08 -.00 -.03 .13 .38 .25 (.82) Mean 4.06 4.09 4.11 4.27 4.14 4.28 4.40 4.67 1.16 3.29 1.50 1. 92 2.07 2.34 4.01 3.90 4.44 3.78 St. Dev. .79 .67 .78 .56 .70 .61 .62 57 .23 .81 .37 .87 .92 1.00 .92 .98 .87 .76 Correlations greater than .16 are significant at p < .05; Correlations greater than .21 are significant at p < .01. N = 147 ; SL = Servant Leadership, CL = Consideration Leadersh ip, TL = Transformational Lead ership, Task Perf = Task Performance, CWB = Counterproductive Workplace Behaviors, O CB = Organizational Citizenship Behaviors, LMX = LeaderMember Exchange, AOC = Affective Organizational Commitme nt, Job Sat = Job Satisfacti on, CPB = Community Prosocial Behavior.
38 Structural equation modeli ng (SEM) was used to test the mediation model. A mediated regression was also conducted as a follow-up test of the relationships amongst individual variables. The complexity and use of maximum likelihood estimation in SEM contributes to the need for large sample si zes. Kline (2005) suggest s that samples with over 200 participants are large, those with 100 to 200 are medium and adequate, and 100 participants is a small but sufficient sample for many models. Based upon this recommendation and related guidelines concerni ng the number of needed participants per indicator (5 to 20 depending on the sour ce; Kline, 2005; Nevitt & Hancock, 2004), a sample size of no less than 140 (minimum of 14 hypothesized i ndicators within the model x 10 = 140) was needed. As discussed below, the proposed model was adjusted based upon item-level and scale-level statistics. The revised model contains 15 indicator variables, suggesting a minimum sample size of 150. The current sample is just shy of 150 supervisor-subordinate pairs. Despite this, a measurement model and SEM were attempted with the understanding that the power to detect effects would be reduced. Results from the Measurement Models Prior to conducting tests of the hypotheses, a series of measurement models were examined in order to determine the most appr opriate, theoretical a nd statistical, way to combine mediation and outcome variables in to higher-order latent factors. Before conducting analyses, all variables were recode d for the SEM model such that high scores indicate positive outcomes and low scores indicate negative outcomes. For example, CWB was recoded such that high scores indica te low levels of CWB (versus the original scale in which high scores indicate more fr equent CWB). CWB and the four well-being
39 indicators were recoded in this way so that the direction of indicatorÂ–factor relationships was consistent acros s all indicators. Initial CFAs, using MPlus version 3.13 with maximum likelihood, were conducted on the hypothesized model structure (F igure 1) with latent factors created for job performance (indicators were Task Perf ormance, OCB, CWB), well-being (indicators were Physical, Psychological, Job Satisfact ion), and social cohesion (indicators were LMX, AOC, and CPB). The hypothesized m easurement model failed to converge, prompting item level analyses and further inve stigation of the latent factorsÂ’ structures. Items were removed from measures that: (1) failed to show adequate factor loadings based on scale and latent factor level CFAs, (2) were flagged as problematic during item level reliability analyses and (3) could be c onceptually distinct from other items on the measure. Items were removed from the scal es for SL, SDT needs satisfaction at work (only for subscales Â“autonomyÂ” and Â“relate dnessÂ”; Â“competenceÂ” was unchanged), task performance, OCB, and consideration leadership. The latent factor structure of the model was adjusted to reflect the observed zeroorder correlations amongst the re vised scales, factor analyzed fit statistics and additional theoretical considerations. This resulted in f our outcome variables in the model, three of which were latent factors. The latent fact ors created were: Job Pe rformance (indicators were Task Performance, OCB, CWB), Job A ttitudes (indicators were Job Satisfaction, LMX, AOC) and Well-Being (indicators we re Physical Well-Being and the three subdimensions of Psychological Well-Being: de pression, anxiety, irritation). In addition, CPB was included as a scale level outcome. Th e final factor struct ure was only slightly changed from the hypothesized model, such that CPB was previously an indicator
40 variable but is now a standal one manifest outcome, the Â“Social CohesionÂ” latent factor was renamed Â“Job AttitudesÂ” to better captu re the nature of its indicators, and job satisfaction was specified as an indicator of Â“Job AttitudesÂ” rather than Â“Well-Being.Â” Additionally, analyses conducte d on the SDT needs at work scale and its subscales resulted in the decision to use subscales scores as indicators of a higher-order latent SDT factor. The original 21-item SDT scale had poo r fit statistics and low internal reliability whereas its subscales demonstrated better mode l fit and internal reliability. Two of the three subscales were revised, by removing revers e-scored items, to enhance model fit. A measurement model with the revised items demonstrated improved fit statistics ( 2 (59) = 160.31, p < .001; CFI = .91, RMSEA = .11, SRMR = .11) and overall factor loadings. Results from the St ructural Models Using the revised scales to create latent factors, an SEM model was tested in MPlus with structural paths defined for both full and partial mediation (see Figures 2 and 3). With regards to fit statistics, neither the full (full mediation: 2 (81) = 318.95, p < .001, CFI = .82, RMSEA = .14, SRMR = .12) or partially ( 2 (77) = 295.42, p < .001, CFI = .84, RMSEA = .14, SRMR = .12) mediated mode ls were at or near commonly agreed upon metrics for establishing ad equate model fit. Ideally, the SRMR value should be below .08 (Hu & Bentler, 1999, 1998). More spec ifically, SRMR is based on discrepancy between the covariance matrix implied by our model and the observed sample covariance matrix. Given that this sample size was not ve ry large, the observed discrepancy may be a result of sampling error. The RMSEA value shows good fit according to Hu and Bentler (1999) when it is below .06, though this is mo re conservative than other metrics which suggest .08 is acceptable. The RMSEA is an absolute fit index which measures the
41 misspecification in the model by each degree of freedom. A low value for a model indicates that for each degree of freedom there is a relatively low degree of misspecification in the model. BentlerÂ’s CFI is an incremental fit index that tests how much better the proposed model reproduces the sample covariance matrix compared to a baseline model where there are no relationships between the variables. Hu and BentlerÂ’s criterion for this index is that it should be no smaller than .95, though .90 is used as a more liberal metric for CFI. Models satisfy ing these fit indices are said to have good Â“model fitÂ” such that they depict the rela tionships amongst variables better than model alternatives. When used in combination, the cut-offs fo r the fit indices may be adjusted since they do not react equally to the sample size, distribution and estimators (Hu & Bentler, 1999). For example, when CFI > .96, the type I er ror rate is not signifi cantly inflated so long as SRMR is less than .10. Similarly, the Type I and Type II error rate s are reduced when RMSEA is below .06 and SRMR is below .10. Unfortunately for the models tested in this study and the scale level CFAÂ’s, few fit statistics fell within the acceptable cut-off metrics suggesting that, in addition to a sma ll sample size, poor model fit and estimation errors may be present.
42 Figure 2. Full mediation model w ith standardized path coefficients N = 147 dyads; 2 (81, 147) = 318.95, p < .001, CFI = .82, RMSEA = .14, SRMR = .12 ^ p < .10; p < .05 ; ** p < .001 +Subordinate self-report ++Supervisor-rating of subordinate Servant Leadershi p + Autonomy+ Self-Determination Theory Needs R2= .29 Relatedness+ Competence+ Job Attitudes R2= .96 Job Performance R2= .02 Subordinate Well-Being R2= .02 Job Satisfaction+ Affective Commitmen t + LMX+ Physical+ Depression+ Anxiety+ Irritation+ OCB++ Task Perf++ CWB+ .53** Community Prosocial Behavior+ R2= .11 .35** .46** .99** .82** .15* .78** .78** .60** .83** .98** .15^.33** .14^.83** .76** .81** 54**
43 Figure 3. Partial mediation model w ith standardized path coefficients N = 147 dyads; 2 (77, 147) = 295.94, CFI = .84, RMSEA = .14, SRMR = .12 ^ p < .10; p < .05 ; ** p < .001 +Subordinate self-report ++Supervisor-rating of subordinate Servant Leadershi p + Autonomy+ Self-Determination Theory Needs R2= .26 Relatedness+ Competence+ Job Attitudes R2= .77 Job Performance R2= .04 Subordinate Well-Being R2= .05 Job Satisfaction+ Affective Commitmen t + LMX+ Physical+ Depression+ Anxiety+ Irritation+ OCB++ Task Perf++ CWB+ .53** Community Prosocial Behavior+ R2= .11 .34** .45** .99** .83** .15* .78** .60** .79** .64** .75** .24* .29** .23* .85** .76** .80** .51** .46** -.15^. 07 -.17^
44 Based upon the standardized beta-weights reported in the SEM analysis and zeroorder correlations, support was found for hypothesi s 1 which states that a significant and positive relationship exists between SL behavi ors and the fulfillment of SDT needs for autonomy, competence and relatedness. Hypothe sis 2 was partially support. Both Job Attitudes and CPB factors shared significant and positive relationships with the fulfillment of SDT needs. Job Performance and Well-Being were significant at an alpha level of .10, but did not meet th e criteria of significance at .05 All relationships were in the expected direction, with an increase in n eeds satisfaction lending to an increase in job performance, CPB, well-being and job attitudes. Hypothesis 3 predicted that the fulfil lment of SDT needs would mediate the relationship between a supervis orÂ’s SL behaviors and subordi nate outcomes such that SLÂ’s relationship with the outcomes would be strengthened in the presence of needs fulfillment. To test this, the fit of a fully mediated SEM (Figure 2; 2 (81) = 318.95, p < .001, CFI = .82, RMSEA = .14, SRMR = .12) was co mpared to the fi t of a partially mediated model (Figure 3; 2 (77) = 295.94, p < .001, CFI = .84, RMSEA = .14, SRMR = .12). Both models had nearly equivalent fit in dices, neither of which met the typical cutoffs (CFI > .95/.90, SRMR < .08/.10, CFI < .06) for acceptable fit (H u & Bentler, 1999). To test whether the partially mediated model pr ovides a better description of the data than the fully mediated model, a test of the chisquare difference was performed. The partially mediated model did not include a direct path from SL to CPB because this relationship was not significant. When direct paths from SL to the outcomes were added, the fit of this partially mediated model was significantly better than the fit of the fully mediated model. Thus, it appears that the relationships that SL has with these criteria are not owing
45 entirely to need satisfaction. The overall poor fit of the part ial and full mediation models may be due to artifacts within the sample its elf, since many of the scalesÂ’ items did not behave as expected during scale level CFAs. The poor fit may also be contributed to by scales which fail to capture the construct of SL itself. Thus, hypothe sis 3 received weak support since the full mediation model contains relationships in the expected directions and at an acceptable level of significance ( p < .05) but failed to satisfy SEM fit indices cut-offs. Interestingly, of the outcomes, SL had the strongest relationship with Job Attitudes ( = .98, p < .001 ) suggesting that SDT needs fulfillment may have the greatest influence on employeeÂ’s job satisfaction, supervisor-subordinate relationships and affective commitment. As a follow-up to the proposed SEM analys es and in light of the high zero-order correlation between LMX and SL ( r = .82), another partially me diated model was tested in which the non-significant path from SL to CPB was removed along with LMX as an indicator of the Job Attitudes latent fact or. This revised part ial mediation model ( 2 (65) = 158.08, p < .001, CFI = .92, RMSEA = .10, SRMR = .11) still did not have adequate fit but did have significantly bette r fit to the data than did the fully mediated model and partially mediated model tested previously (alpha-level was .01). This finding suggests that needs satisfaction and SL behaviors have an impact on the outcomes and that LMX as an outcome statistically blurs the relationships amongs t SL, needs satisfaction and other outcome variables. At a theoretically le vel, this is likely due to both SL and LMX addressing the relational strength of and affi nity between the supervisor and subordinate.
46 Results from the Regression Models A series of hierarchical regressions were performed as a follow-up to the SEM model, in light of the poor m odel fit, in order to examine mediated relationships at the scale level. Results from the mediated regres sions are reported in Tables 4 and 5. In step 1, age, education, the subordinatesÂ’ tenur e with their current organization and organizational membership (i.e. employee at Orga nization A, B, C, or D) were entered as covariates. As a categorical variable, orga nizational membership was dummy coded. In step 2, SL was entered followed by the SDT needs at work scale in step 3. The SDT needs scale was a composite variable computed from the mean values of the original 21item SDT scaleÂ’s subscales for Autonomy, Competence and Relatedness. This allowed for all three subscales to be equally weighted in the regression equa tion, paralleling their use as indicators of a latent SDT needs factor in the SEM model proposed. Given the number of regressions run with the predicto r and mediator, a more conservative alpha value of .01 was adopted to determine significa nce, thus countering an increase in the Type I error rate due to th e number of analyses conducted.
47 Table 4. Hierarchical mediated regressions for Job Attitude s and Job Performance outcomes Criteria Job Attitudes Job Performance Predictors Job Sat LMX AOC Task Perf OCB CWB Step 1: Covariates Age .24 .04 .40** -.02 .32 -.29 Education -.11 -.04 -.04 -.02 .03 .06 Tenure with Current Organiza tion .03 -.11 .11 .15 .08 .08 Org Affiliation: A .10 -.05 .08 .14 -.46** .07 Org Affiliation: B .31 .08 .16 .12 -.39 -.10 Org Affiliation: C .20 -.04 .10 -.09 -.46** -.06 F 4.88*** .35 10.12*** 1.37 3.74** 2.14* R2 .19 .02 .29 .04 .16 .09 Step 2: Servant Leadership .32*** .85*** .35*** .20* .19* -.09 F 18.23*** 287.15*** 28.72*** 5.39** 5.36* 1.17 R2 .10 .68 .12 .04 .04 .01 Step 3:Servant Leadership .00 .71*** .13* .27** .17 -.06 SDT Needs at Work .71*** .29*** .48*** -.15 .04 -.07 F 120.86*** 29.69*** 50.90*** 2.14 .18 .42 R2 .35 .06 .16 .02 .00 .00 Model F 27.66*** 48.82*** 24.02*** 2.02* 3.57*** 1.80 R2 .62 .74 .61 .12 .19 .10 p < .05, ** p < .01, *** p < .001 Note: N = 147 supervisor-subordinate pairs; Standardized regression coefficients are reported in the table. Regression coeffici ents reflect value at each step. Job Sat = Job Satisfaction, LMX = Leader-Member Exchange, AOC = Affective Organizational Commitment, Task Perf= Task Performance, OCB = Organizati onal Citizenship Behaviors, CW B = Counterproductive Workplace Behaviors. Tenure was reported in mont hs, Organizational Affilia tion was dummy coded as 1 = member, 0 = nonmember
48 Table 5. Hierarchical mediated regressions for Well-Being and Commu nity Prosocial Behavior outcomes Criteria Well-Being Predictors Physical Depression Anxiety Irritation CPB Step 1: Covariates Age -.31 -.48** -.55*** -.49** .24 Education -.14 .10 .09 .06 .02 Tenure with Current Organization .05 .05 .02 .12 -.12 Org Affiliation: A 06 .12 .27 .23 .07 Org Affiliation: B 09 .14 .29 .25 .26 Org Affiliation: C -.08 .11 .15 .33** .12 F 2.66* 2.30* 2.61* 2.25* 3.41** R2 .11 .10 .11 .10 .14 Step 2: Servant Leadership .08 -.20* .05 -.13 .19* F .90 5.44* .29 2.31 5.16* R2 .01 .04 .00 .02 .03 Step 3:Servant Leadership .15 -.02 .07 -.07 .13 SDT Needs at Work -.14 -.38*** -.06 -.10 .12 F 2.03 16.21*** .29 .88 1.39 R2 .01 .10 .00 .01 .01 Model F 2.37* 4.79*** 2.01* 2.10* 3.47*** R2 .13 .24 .11 .12 .18 *p < .05, ** p < .01, *** p < .001 Note: N = 147 supervisor-subordinate pairs; Standardized regression coefficients are reported in the table. Regression coeffici ents reflect value at each step. Depression, Anxiety and Irritation are each a subscale fr om Psychological Well-Being; High Well-Bei ng scores indicate more negative symptoms. CPB = Community Prosocial Behavior. Tenur e was reported in months, Organizational Affiliation was dummy coded as 1 = member, 0 = nonmember.
49 In the absence of SDT needs, SL shared few significant relationships with the criteria (at p < .01). Thus, hypothesis 1 was only partia lly supported such that the only significant relationships present between SL and the outcome variables were observed for job satisfaction ( = .32, p < .001), LMX ( = .85, p < .001), and AOC ( = .35, p < .001). Thus, supervisors displaying more SL be haviors tended to ha ve subordinates who reported more favorable job attitudes. SDT needs, when entered in step 3, were significantly related to job satisfaction ( = .71, p < .001), LMX ( = .29, p < .001), AOC ( = .48, p < .001), and the depression subs cale of psychological well-being ( = -.38, p < .001). Thus, fulfillment of SDT needs is associated with increased job attitudes and decreased perceptions of de pressive psychological sympto ms. Similar to Hypothesis 1, Hypothesis 2 received partial support since only a subset of the outcome variables were significantly related to SDT needs fulfillment. To test for mediation, Baron and Kenny ( 1986) recommend four criteria be met: (1) the predictor and outcome variables are correlated, (2) the predictor and mediator variables are correlated, (3) the mediator a nd outcome variables are correlated, and (4) the effect of the predictor on the outcome is attenuated in the presence of the mediator variable. Full mediation is present if the relationship between the predictor and outcome is non-significant when the mediator variable is controlled for. If the predicto r variable is still significant, then partial mediation is pr esent. To determine what outcome variables were eligible for a mediation analysis, a more liberal p -value of .05 was adopted to maximize the inclusion of variables. Additionally not all statisticians agree that the first criteria must be met in order for mediation to be present. The latter provides further
50 justification for a more liberal pvalue despite the large number of analyses conducted and potential for Type I error inflation. Zero-order correlations and beta-weight s were examined to determine which variables fulfilled the first three criteria of mediation at p < .05. Job satisfaction, LMX, AOC, Psychological Well-Being: Depression and CPB shared significant correlations with the predictor (SL) and the mediating va riable (SDT needs). In addition, SDT needs (composite scale) and SL share a significant correlation. A Sobel test was used to assess the presence of mediation fo r the five outcome variables identified as relevant. Full mediation is concluded when (1) the Sobel test is significant, which means that the unstandardized b-weight from the regression for SL was significantly reduced when SDT needs were entered into the regression equa tion in step 3, and (2) when the beta-weight for SL is no longer significant in step 3. Part ial mediation is concl uded if (1) the Sobel test is significant (i.e., the b-weight for SL is significantly reduced in the presence of SDT needs) but (2) the beta-weight for SL remain s significant in step 3. The relationships of SL with job satisfaction (Sobel test = 5.35, p < .001), LMX (Sobel test = 4.65, p < .001), AOC (Sobel test = 4.65, p < .001), and depression (Sobel test = -3.64, p < .001) were significantly attenuated when SDT needs were added to the regression equation, as indicated by a significant Sobe l test. The beta-weights from th e regression analysis were examined to determine whether full or partia l mediation was presen t for each of these four outcomes. Full mediation was observ ed in the case of job satisfaction and depression, whereas partial mediation was obs erved in the case of LMX and AOC (see Tables 4 and 5). The Sobel test was not si gnificant for the relationship between SL and
51 CPB in the presence of SDT needs, meani ng mediation was not present for the outcome of CPB. Similar to the results of the SEM, outcome variables associated with job attitudes (job satisfaction, LMX and AOC) were mediated by SDT needs. In addition, the subscale of depression was affected by the presence of SDT needs fulfillment. The results of the hierarchical mediated regre ssions provided partial support for hypothesis 3, which stated that SL and the outcome variables would be mediated by SDT needs. Exploratory Analyses A series of hierarch ical regressions were performed to test whether SL accounted for incremental variance compared to consideration leadership and transformational leadership. If SL acc ounts for a significant proportion of unique variance, then it suggests that there is valu e added in considering SL in addition to consideration and transformati onal leadership. To test this covariates (same as used previously) were entered in step 1 of a regression, cons ideration and transformational leadership were entered in step 2 and SL was entered in step 3. If the change in R2 is significant following step 3, it can be concl uded that SL accounts for variance in the criterion incremental to the other two type s of leadership (entered in step 2). Consideration and transformational leader ship accounted for a significant amount of variance in job satisfaction ( R2 = 9.47, p < 001), LMX ( R2 = 174.34, p < 001), AOC ( R2 = 9.61, p < 001), task performance ( R2 = 4.64, p < 05), and the depression subscale of psychological well-being ( R2 = 4.73, p < 01). Interestingly, consideration leadership was negatively rela ted to indices of job performance (task performance, = -.33; OCB, = -.18; CWB, = -.25) and well-being (physical, = -.28; depression, = -.31; anxiety, = -.29; irritation, = -09) whereas both SL and
52 transformational leadership generally shared positive relationships with these outcomes. SL accounted for incremental variance be yond that explained by consideration and transformational leadership for LMX ( R2 = 11.34, p < 001), AOC ( R2 = 9.41, p < .01), physical well-being ( R2 = 11.91, p < 001), the anxiety subscale of psychological well-being ( R2 = 5.83, p < 05), and CPB ( R2 = 6.24, p < 01). Results are mixed as to the incremental validity of SL over a nd above consideration and transformational leadership (Tables 6 and 7). Conclusions dr awn concerning the incr emental validity of any of the three leadership constructs are tempered give n the extremely high, positive correlations (between .80 and .90) amongs t SL, consideration leadership and transformational leadership. Across all analyses, SL appears to be closely associated with job attitudes and has a more complex relationship with the other outcome variables examined.
53 Table 6. Leadership hierarchical regressions for Job Attitudes and Job Performance outcomes Criteria Job Attitudes Job Performance Predictors Job Sat LMX AOC Task Perf OCB CWB Step 1: Covariates Age .23 .04 .40** -.02 .32 -.29 Education -.11 -.04 -.04 -.02 .03 .06 Tenure with Current Organiza tion .03 -.11 .11 .15 .08 .08 Org Affiliation: A .10 -.05 .08 .14 -.46** -.06 Org Affiliation: B .31 .08 .16 .12 -.39* -.10 Org Affiliation: C .20 -.04 .10 -.09 -.46*** .07 F 4.88*** .35 10.12*** 1.37 3.74** 2.14* R2 .19 .02 .32 .06 .16 .09 Step 2: Leadership Consideration Leadership .22 .35*** -.07 -.33 -.18 -.25 Transformational Leadership .13 .54*** .36* .50* .33 .10 F 9.47*** 174.34*** 9.61*** 4.64* 2.60 2.06 R2 .11 .72 .09 .07 .04 .03 Step 3: Consideration Leadership .16 .24* -.22 -.40* -.28 -.31 Transformational Leadership .03 .34** .09 .38 .16 -.01 Servant Leadership .16 .33*** .46** .20 .29 .18 F .84 11.34*** 9.41** 1.08 2.28 .89 R2 .02 .02 .04 .01 .02 .01 Model F 3.48*** 44.15*** 6.74*** 2.12* 3.42*** 2.00* R2 .30 .76 .46 .14 .21 .13 p < .05, ** p < .01, *** p < .001 Note: N = 147 supervisor-subordinate pairs; Standardized regression coefficients are reported in the table. Regression coeffici ents reflect value at each step. Tenure was reported in months, Or ganizational Affiliation was dumm y coded as 1 = member, 0 = nonmember.
54 Table 7. Leadership hierarchical regressions for Well-Being and Community Pr osocial Behavior outcomes Criteria Well-Being Predictors Physical Depression Anxiety Irritation CPB Step 1: Covariates Age -.31 -.48* -.55*** -.49** .24 Education -.14 .10 .09 .06 .02 Tenure with Current Organization .05 .05 .02 .12 -.12 Org Affiliation: A 06 .12 .27 .23 .07 Org Affiliation: B 09 .14 .29 .25 .26 Org Affiliation: C -.08 .11 .15 .33** .12 F 2.66* 2.30* 2.61* 2.25* 3.41* R2 .11 .10 .11 .10 .14 Step 2: Servant Leadership Consideration Leadership -.28 -.31 -.29 -.09 .11 Transformational Leadership .22 .06 .23 -.09 -.00 F 1.28 4.73** 1.34 2.30 .90 R2 .02 .06 .02 .03 .01 Step 3: Consideration Leadership -.49** -.33 -.44* -.15 -.04 Transformational Leadership -.15 .03 -.03 -.19 -.27 Servant Leadership .62*** .06 .45* .16 .46** F 11.91*** .10 5.83* .74 6.24** R2 .08 .00 .04 .01 .04 Model F 3.57*** 2.67** 2.78** 2.12* 3.27*** R2 .21 .16 .17 .13 .52 ^ p < .05, p < .01, ** p < .001 Note: N = 147 supervisor-subordinate pairs; Standardized regression coefficients are reported in the table. Regression coeffici ents reflect value at each step. Depression, Anxiety and Irritation are each a subscale fr om Psychological Well-Being; High Well-Bei ng scores indicate more negative symptoms. Tenure was reported in months, Organizational Affiliation was dummy coded as 1 = member, 0 = nonmember
55 Exploratory Leadership Dominance Analysis. The relative contri bution of each of the three leadership behaviors for releva nt predictors was assessed by calculating dominance analysis weights (Azen & Budescu, 2003). Dominance analysis is a statistical technique in which R2 values are assessed using all possible subsets of multiple regression models to rank-order predicto r variables according to their relative contribution to outcomes. The relative contri butions of transforma tional, consideration and servant leadership were assessed usi ng a dominance analysis using LeBretonÂ’s version 4.4 Excel spreadsheet. The relative import ance of the leadership predictors were assessed for the composite measure of needs sa tisfaction, each of th e three subscales of needs satisfaction (e.g., autonomy, competence, relatedness), and for 4 outcome variables (job satisfaction, AOC, LMX, and psychological well-being: depression) Selection of the outcome variables was based upon the presence of significant correlations and mediation. The dominance weights and re-scaled dominan ce weights are reported in Tables 8 (SDT needs satisfaction) and 9 (outcome variable s). The rescaled dominance weight is a percentage value calculated by dividing th e predictorÂ’s dominance weight by the total modelÂ’s R2. Across all outcome variables, the three leadership behaviors displayed little to no variability in terms of their different ial importance in accounting for the variance explained in the criterion ( R2). Paralleling previous conclusions drawn from the analyses, it appears a more general leader ship factor or Â“goodÂ” leader ship behaviors are driving the relationship between the predictor an d criteria rather than SL alone.
56 Table 8. Dominance Weights Analysis of Three Leade rship Predictors for Model SDT Needs Satisfaction Composite Autonomy Competence Relatedness Predictors DW % DW % DW % DW % Servant Leadership .19 33.25 .28 32.29 .20 37.39 .16 30.35 Consideration Leadersh ip .18 31.44 .17 30.22 .16 30.05 .19 35.11 Transformational Leader ship .21 35.31 .21 37.49 .18 32.56 .19 34.55 Model R2 .58 .56 .54 .54 Note: RW = Dominance weights; % = Rescal ed dominance weights (DW divided by model R2). Composite = SDT needs satisfaction composite scale. Covariates were included in the dominance analysis but are not re ported in the table. Table 9. Dominance Weights Analysis of Three Le adership Predictors for Model Outcome Variables W-B: Depression LMX AOC Job Satisfaction Predictors DW % DW % DW % DW % Servant Leadership .12 30.60 .29 33.62 .25 36.47 .18 33.67 Consideration Leadersh ip .15 37.92 .28 32.42 .21 31.21 .18 33.67 Transformational Leader ship .13 31.47 .30 33.96 .22 32.32 .18 32.66 Model R2 .40 .87 .67 .55 Note: DW = Dominance weights; % = Rescal ed dominance weights (DW divided by model R2). W-B: Depression = psychological well-being: depression subs cale, LMX = Leader-Member Exchange, AOC = Affective Organizational Commitment. Covariates were included in the dominance analysis but are not re ported in the table.
57 Discussion Theory and limited empirical support (M ayer, Bardes, & Piccolo, 2008) suggest SDT needs may play a critical mediating ro le between supervisorsÂ’ SL behaviors and their subordinatesÂ’ work-relevant outco mes. This study sought to further our understanding of if and why SL translates into critical subo rdinate-level work criteria. Specifically, this study examined criteria or ganized into four outcome categories: job performance (i.e., task performance, O CB, CWB), well-being (i.e., physical and psychological related to depression, anxiety and irritability), job attitudes (i.e., job satisfaction, LMX and AOC) and prosocial behaviors direct towards the community within which an organization is established (CPB). Correlations suggest that supervisors exhibiting more SL behaviors tended to have employees with more favorable levels, generally, on all of these vari ables. Thus, the answer to if SL behaviors are related to enhanced outcomes for subordinates at work is Â“yesÂ”. To examine why SL behaviors are related to these outcomes, it was proposed that supervisors who exhibit more SL behaviors fulfill their subordinatesÂ’ ba sic needs for autonomy, competence, and relatedness. Self-determination theory pos its (Deci & Ryan, 2000) and empirical evidence suggests (e.g., Baard, Deci, & Ryan, 2004) that when these three needs are met, subordinates exhibit improved experiences at work. This study tested a model using SEM and mediated regression examining whether the fulfillment of subordinatesÂ’ SDT needs
58 mediates the relationship between superviso r-performed SL behaviors and subordinate organizational outcomes. Three primary hypotheses were proposed to establish the relationships amongst SL, SDT needs fulfillment and outcomes. H ypothesis 1 predicted a positive relationship between supervisor SL behaviors and subordi nates reporting the fulf illment of their SDT needs for autonomy, competence and relatedne ss. This hypothesis was fully supported across all analyses. Increased SL behaviors amongst supervisor s, as they were perceived and reported by subordinates, were associat ed with reports by subordinates of higher levels of their SDT needs being met, when examined both overall and at the level of each of the three needs. This finding is consistent with Mayer, Bardes and Piccolo (2008) who proposed a model in which supervisory SL beha viors were a direct a nd indirect (mediated by justice perceptions) antecedent to subordina tes reporting that each of their needs were met for competence, autonomy and relatedne ss. The relationship between SL behaviors and needs fulfillment supports theoretical a nd conceptual arguments for SL being a follower-centric, generative approach to leadership and supervisor-subordinate relationships in which supervisors prioritize the welfare and success of their subordinates (Barbuto & Wheeler, 2006; Graham, 1991; Stone, Russell, & Patterson, 2004). Fulfillment of SDT needs for subordinates should in turn foster more favorable organizational outcomes (Gagn & Deci, 2005). Hypothesis 2 predicted a positive relati onship between subordinates reporting their SDT needs for autonomy, competence and relatedness were met and enhanced workplace outcomes. The workplace outcomes were grouped, in the final revised model, into four factors: job performance, well-b eing, job attitudes and community prosocial
59 behaviors. Hypothesis 2 was partially supporte d such that, across all statistical analyses, when SDT needs fulfillment was rated as hi gh, subordinates reported lower perceptions of depressive psychological well-being symp toms, and higher levels of LMX, AOC, job satisfaction and CPB. These relationships were consistent when SDT was examined globally and when assessed at the level of individual needs. Amongst the outcome factors, SDT needs fulfillment was associated with higher levels of overall job attitudes such that subordinates reporting they felt au tonomous, competent a nd connected (related) at work also reported higher levels of satis faction with their job, emotional connectedness and commitment to their organization and higher quality relationships with their supervisors. Along similar lines, reduced depr essive symptoms were reported when needs were met. Since depressive symptoms were significantly and negativ ely associated with all of the job attitudes indicators, it may be that an overall positive experience at work lends to reductions in depr essive feelings and symptoms at and outside of work. The only behavioral manifestation whic h was upheld, of those hypothesized, was a greater frequency of CPB being performed when needs were fulfilled. These results suggest that needs fulfillment may operate more at an a ttitudinal and affective level which does not always translate into measur eable behavioral indices. The relationship present between SDT needs fulfillment and job attitudes, depressive symptoms and CPB support previous empirical findings and theori zation with regards to social exchange theory (Blau, 1964) and SDT (Ryan & D eci, 2000; Deci & Ryan, 2000). A social exchange theory perspective would argue th at when employees feel their organization and supervisors satiate their basic needs, they reciprocate this investment by affectively committing to the organization and the relationships existing within the organization.
60 Empirical studies examining the mechanis ms by which SL differentially impacts affective workplace experiences, as opposed to workplace behaviors, are warranted given the trends identified in this study. Additiona lly, research integrat ing social exchange theory at a more foundational level, when studying SL behaviors, would allow scholars the ability to better concep tualize the nature of the dyadic supervisor-subordinate relationship. Hypothesis 2 receiving only partial support poses comp lications for SDT since previous empirical work has linked needs sati sfaction with several of the outcomes which were not found to be significantly related here. Specifically, SDT needs satisfaction has been previously related to job performance (Baard, Deci, & Ryan, 2004), physical and overall psychological well-being (Baard, D eci, & Ryan; Reinboth & Duda, 2006) and prosocial organizational beha viors (Gagne, 2003). Job perfor mance, physical well-being, two of the subscales of psychological well-bei ng (e.g., anxiety and irritability), and OCB did not share noteworthy relationships with the fulfillment of SDT needs. Despite this, some relationships previous ly reported in th e literature were upheld, such as organizational commitment (Gray & Wilson, 2008) and job satisfaction (Deci, Connell, & Ryan, 1989). These discrepancies suggest that the holistic claims of the theory of SDT and its relationships to enhanced job outco mes for employees may be more nuanced in reality, at a measurement and theoretical le vel, than conceptually conceived. When examining the fulfillment of SDT needs, it would be wise to consider how needs are differentially related to outcomes and whether an overall assessment of needs is better or worse than relating other constructs to th e subscales of SDT needs. Understanding the
61 interplay of the three needs would allow fo r a more nuanced unde rstanding of SDT and also could explain the inconsis tent outcomes amongst studies. Hypothesis 3 proposed that SDT needs mediate the relationship between SL behaviors and the outcome f actors of job performance, well-being, job attitudes and community prosocial behaviors. The propos ed model did not provide an adequate depiction of the actual relati onships amongst the variables; th is prevents th e interpretation of a holistic model of the motivational m echanisms by which SL behaviors lend to enhanced job experiences. De spite the proposed mediation m odel failing to describe the relationships as a whole, tent ative conclusions can be drawn from individual relationships present within the model. Amongst all four out come factors, job attitudes benefitted the most from SL behaviors and needs fulfillment, followed by CPB. These results compliment those describing the direct re lationships between SL behaviors and the outcomes and SDT needs fulfillment and the outcomes. Follow-up analyses using mediated re gression provide further and more substantive evidence for the relationships s uggested in the revise d SEM model (with SDT needs mediating the relationships between SL behaviors and subordinate outcomes). Specifically, the relationships between increas ed SL behaviors and improved job attitudes and reduced depressive psychological symptoms were mediated by needs fulfillment such that job attitudes were enhanced and depr essive symptoms attenuated when SDT needs were fulfilled. Unlike the trends in hypothesi s 1 and 2 for CPB, the relationship between SL behaviors and CPB was not significantly changed when SDT needs were fulfilled for employees, suggesting SL is a primary driver of employees partic ipating in prosocial behaviors targeting the commun ity outside the walls of th e organization. This supports
62 SL conceptually as a form of leadership which promotes ethical and community-oriented action (Graham, 1991; Spears, 1995). Social learning theory (Bandura, 1977, 1986) would argue that servant leaders exemplify th e importance of CPB a nd, as a result, their subordinates follow their example and engage in similar behaviors, regardless of other intervening factors such as needs being fulfilled. The outcomes which were mediated, namely job satisfaction, LMX, AOC and depr essive psychological symptoms, all have affective components which, again, suggest th at SDT needs fulfillmen t is acting at an affective level. The mediation of SL behaviors and job attitudes by the fulfillment of SDT needs has interesting applied implications. Genera lly, given that some effects of SL are transmitted through need fulfillment, it may be that some of SLÂ’s positive outcomes can be nurtured by simply focusing on followers' needs. More specifically, when effective leadership and SL behaviors are lacking with in an organization, employeesÂ’ job attitudes may still be enhanced through an interventi on that targets the satisfaction of the employeesÂ’ needs for autonomy, competence an d relatedness. Since not all supervisors are Â“leadersÂ”, such an intervention may be cr itical. Thus, even supervisors who are not necessarily servant leaders might be able to foster favorable job attitudes, amongst other outcomes, among their employees if they take steps to ensure that autonomy, competence, and relatedness are being met through non-leadership factors. This idea is consistent with the substitut es-for-leadership theo ry (Kerr & Jermier, 1978), which posits that changes in certain individual-level, ta sk or environmental factors can enhance, neutralize or serve as a substitu te for leadership; thus enabling many of the positive effects that are generally attributed to leadership. Specifically, Kerr and Jermier
63 argue that for types of leadersh ip like SL, subordinate sÂ’ intrinsic satisfact ion (a task-level substitute), need for independence (an i ndividual-level substitute), Â“professionalÂ” orientation, and presence within a c ohesive and connected work group (an environmental-level substitute) can serve as subs titutes for leadership when leadership is absent or lacking. Notably, the last three substitutes parallel SDT needs for autonomy, competence and relatedness, respectively. When these substitutes are available, leadership may not hold the same influence as when mediating factors, such as the fulfillment of needs, are not pr esent. Support for the substitutes-for-leadership theory has been mixed (Avolio, Walumbwa, & Weber, 2009), though job satisfaction has maintained some consistency as an outcome va riable for which leadership substitutes are successful (Podsakoff, MacKenzie, & Bommer 1996; Podsakoff, MacKenzie, & Fetter, 1993; Pool, 1997). The results of this study s upport the latter findings that job attitudes are influenced by leadership substitutes and also the claims of Avolio and colleagues regarding the mixed effects of leadership substitutes since the satisfaction of needs did not mediate the relationship between leadership and either job performance or well-being indices. The discrepancies between what outcome s were and were not influenced by mediating Â“substitutionÂ” factors may also lend credence to critics of the substitution-forleadership theory. For example, Dionne, Ya mmarino, Atwater, and James (2002) argue, via theory and empirical results, that th e moderators and mediators proposed in the substitution-for-leadership theory fail to subst itute and/or trump the effects of leadership. They conclude that leadership does matter, a nd that prior mixed findings in support of the theory can be attributed to statistical arti facts, specifically th e use of oneor two-
64 independent sources for ratings of leadership and leadership-substitutes. If same-source bias does drive some of the findings regarding the effectiveness of leadership substitutes, it may explain why job attitudes and SL were mediated by needs fulfillment (as a substitute), whereas job performance indice s and SL were not mediated by needs. Notably, job attitudes were all self-reported by subordinates whereas job performance included indices rated by multiple sourcesÂ—the subordinate and the supervisor. Future studies employing multiple sources (e.g., supervisor, subordinate, work group, coworkers) may elucidate whether needs sati sfaction is truly acting as a substitute for leadership or whether it is a statistical artifact, as argue d by Dionne and colleagues, of same-source bias. In light of previous research and theory, it is also interesting to note what SL and SDT needs satisfaction were not related to in this study. SL is uniquely defined from other leadership styles by its follower-foc us, generative organi zational approach and concern for ethics. The latter concern for ethi cs is noteworthy since SL behaviors did not share relationships with OCB or CWB. Presum ably these are two areas in which Â“ethical concernÂ” and a generative environment (e.g., mentoring or helping a colleague with a project) would manifest themselves, yet this is not the case presently. Previous studies have reported that SL is rela ted to ethical consideration a nd prosocial behaviors (Ehrhart, 2004). The latter suggests that addi tional factors may be at pla y, such as an individualÂ’s trait levels of regulatory focus (Neubert Kacmar, Carlson, Chonko, & Roberts, 2008) or organizational climate (Walumbwa, Hartne ll, & Oke, 2010). Similarly, SDT needs are boasted to relate to higher levels of several job perform ance indices, such as task performance, and well-being (Baard, D eci, & Ryan, 2004; Reinboth & Duda, 2006).
65 However, such relationships between SDT, well-being and job performance indices were not present for the employees surveyed in this study. The relationship between SL, SDT need s, and the outcomes were closely paralleled by both the reported transformati onal leadership behavi ors and consideration leadership behaviors of the supervisors. Exploratory analyses to investigate the discriminant validity of SL behaviors were le ss conclusive than pr evious studies which have cleanly demonstrated SLÂ’s incrementa l predictive validity for organizational outcomes over and above other leadership st yles (Barbuto & Wheeler, 2006; Ehrhart, 2004; Liden Wayne, Zhao, & Henderson, 2008). This study found that SL predicted physical well-being, LMX and AOC incr emental to both consideration and transformational leadership. However, SL di d not provide any added predictive power for the other outcomes beyond what was accounted for by transformationa l and consideration leadership. One explanation for the lack of discrimination between th e constructs is the large correlations, ranging between .80 a nd .90, shared amongst these three leadership constructs. The large correlati ons suggest that great concep tual overlap exists when subordinates rate their supervis orsÂ’ leadership behaviors. Aside from the indicators of subordinate well-being, SL, consideration leadership and transformational leadership shared nearly identical relationships with the outcome and mediator variables. This may arise for one or more of the fo llowing reasons: (1) the constructs for each leadership type are ill-de fined, (2) the items within the measures for each leadership type fail to capture the cons truct and distinguishing behaviors, and/or (3) the leadership constructs are all moderated by whether a subordinate has a positive or negative appraisal of the supervisor such that a Â“big LÂ” type fact or exists underlying
66 positive leadership behaviors. The latter case is further substantiated by the high correlation LMX shares (.80 to .82) with the three leadership constructs suggesting that the quality of subordinatesÂ’ re lationship with their supervisors is very closely related to their appraisal of the type of behaviors the supervisors exhibit. Unexpectedly, the exploratory analysis which aimed to demonstr ate the discriminant validity of SL may have prompted an argument for revising and reducing the number of leadership typology constructs to reflect a more general appraisal of Â“goodÂ” versus Â“badÂ” leadership. Ultimately it may be that Â“goodÂ” leaders focu s on their followersÂ’ needs which not only demonstrates to followers their worth, but also fulfills certain basic needs, such as those prescribed by SDT. This then could lend to enhanced experiences (e.g., performance, attitudes, well-being) at wo rk. Rather than focusing on sp ecific leadership constructs, future research should seek to understand how the different construc ts relate to each other, in what ways they are different and whether a broader perspective on leadership with it defined as Â“goodÂ” versus Â“badÂ” or some other moderating, appraisal dichotomy would better serve l eadership research. Limitations Several limitations of this study have alre ady been noted, such as the lack of discriminant validity of the SL measure when compared to similar leadership constructs. Relatedly, several of the scales, including fo r SL, failed to demonstrate adequate item analyses and thus were revised for this study. This may have altered some of the relationships amongst variables such that prev iously validated scal es were changed and assumed to behave in the same manner as th e original scale did in previous studies. By
67 revising the scales and removing items, this study may have been prevented from seeing results paralleling past empirical research. In addition to the revision of scales, drawing meaningful conclusions from the results may be made more difficult if the scal es themselves fail to measure the construct it purports to. For example, give n the high correlations amongs t the leadership constructs, it may be that the scale for SL does not ade quately capture the behaviors of a servant leader. Thus, there is debate as to whet her the items on some of the scales were problematic due to the specific sample used pr esently, the lack of sc ale validation rigor, or for conceptual/theoretical reasons. In addition to this the possibility for response biases to impact how supervisors and subor dinates responded to the survey is worth further investigation. Specifically, employees were recruited thr ough their organization and, generally, completed the survey while at work. Some subordinate and/or supervisors may have been uncomfortable responding trut hfully about CWBs, for example, while at work since they are seen as negative, soci ally undesirable behaviors. Related to the limitations of the scales and c onstruct definitions, the size of the sample was adequate but could have been increased. More superviso r-subordinate pairs may result in improved or better defined results as the power to detect effects increases. Since supervisor-subordinate dyads were recruited to participate, common method bias is not a critical concern or limitation si nce both supervisor a nd subordinate rated the subordinateÂ’s experiences. Though this st udy can boast multi-source data, supervisors provided information for only two of the out come variables. Both of the variables supervisors responded to, in-role task perfor mance and OCB, were grouped in the job performance factor. This means that all the ot her constructs were self-report and from a
68 single source. Future research could increa se the rigor of this study by including more other-rated (i.e. rated by supervisors or co workers) measures of the subordinatesÂ’ experiences. Different perspectives may al so shed light on the differences between internal attitudinal outcomes and externally manifested, explicit behaviors at work. Conclusion This study proposed that the satisfaction of SDT needs serves as a mediating mechanism between supervisory SL beha viors and subordinate sÂ’ organizational outcomes. The outcomes which were investig ated and hypothesized to relate to SL behaviors and the fulfillment of SDT needs for autonomy, competence and relatedness were organized into four outcome categories: job performance, well-being, job attitudes and community prosocial behaviors. Results i ndicated that SL behaviors relate to the fulfillment of SDT needs but not necessarily to all of the outcomes. With regards to SDT needs and the outcomes, predominantly at titudinal outcomes (LMX, job satisfaction, AOC and depressive psychologi cal symptoms) related to SDT needs. SDT needs related to only one behavioral outcomeÂ—CPB. Add itionally, SDT needs were found to moderate the relationship between SL behaviors and attitudinal outcomes. Interestingly, neither SDT needs nor SL behaviors shared relati onships with many of the outcomes despite prior empirical work suggesting relationships exist. The discrepanc y in results between this study and previous research suggests th e relationships may be more nuanced than previously acknowledged. The discrepant findi ngs should also prompt discussion as to how the constructs of SL and other leader ship behaviors are de fined and measured. This study sought to build on previous research by proposing and testing a comprehensive model of SL, SDT needs a nd important organizational indices. The
69 results of this study should serve as a call for researchers to bette r define leadership constructs and the measurement of them, understand the nuances of both SL and SDT needs within a work context, and to cons ider whether having a multitude of different leadership typologies is necessary.
70 List of References Avolio, B. J. & Bass, B. M. (1995). Individual consideration viewed at multiple levels of analysis: A multi-level framework for examining the diffusion of transformational leadership. Leadership Quarterly, 6 (2), 199-218. Avolio, B. J. & Gardner, W. L. (2005). Au thentic leadership development: Getting to the root of positive forms of leadership. Leadership Quarterly, 16 315-338. Avolio, B. J., Walumbwa, F. O., & Weber, T. J. (2009). Leadersh ip: Current theories, research, and future directions. Annual Review of Psychology, 60 421-449. Azen, R., & Budescu, D. V. (2003). The dom inance analysis approach for comparing predictors in multiple regression. Psychological Methods, 8 (2), 129-148. Baard, P. P., Deci, E. L., & Ryan, R. M. (2004). Intrinsic need satisfaction: A motivational basis of performance and well-being in two work settings. Journal of Applied Social Psychology, 34 (10), 2045-2068. Bandura, A. (1977). Social Learning Theory New York: General Learning Press. Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory Englewood Cliffs, NJ: Prentice-Hall. Barbuto, J. E. & Wheeler, D. W. (2006). Scale development and constr uct clarification of servant leadership. Group Organization Management, 31 300-326. Baron, R. M., & Kenney, D. A. (1986). The mo derator-mediator variab le distinction in social psychological research: Conceptual, strategic and statistic al considerations. Journal of Personality and Social Psychology, 51 1173-1182. Bass, B. M. (2000). The future of le adership in learning organizations. Journal of Leadership Studies, 7 18-40.
71 Bass, B. M. & Steidlmeier, P. (1999). Ethi cs, character, and authentic transformational leadership. Leadership Quarterly, 10 181-217. Bedeian, A. G., Ferris, G. R., & Kacmar, K. M. (1992). Age, tenure, and job satisfaction: A tale of two perspectives. Journal of Vocational Behavior, 40 33-48. Blau, P. (1964). Exchange and Power in Social Life New York: Wiley. Bono, & Judge, T. (2003). Self-concordan ce at work: Toward understanding the motivational effects of transformational leaders. Academy of Management Journal, 46 (5), 554-571. Borman, W. C., & Motowidlo, S. J. (1993). Expanding the criterion domain to include elements of contextual performance. In N. Schmitt, & W. C. Borman (Eds.), Personnel selection in organizations (pp. 71-98). San Franci sco: Jossey-Bass. Brown, M. E., Trevio, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for constr uct development and testing. Organizational Behavior and Human Decision Processes, 97 117-134. Burns, J. M. (1978). Leadership New York: Harper and Row. Cammann, C., Fichman, M., Jenkins, D., & Kl esh, J. (1983). Assessing the attitudes and perceptions of organizational members. In S. Seashore, E. Lawler, P. Mirvis, & C. Cammann (Eds.), Assessing Organizational Change: A Guide to Methods, Measures and Practices. New York: John Wiley, p. 84. Caplan, R. D., Cobb, S., French, J. R. P ., Van Harrison, R., & Pinneau, S. R. (1980). Job Demands and Worker Health. Ann Arbor: University of Michigan, Institute for Social Research, p. 274.
72 Carless, S. A., Wearing, A. J., & Mann, L. (2000). A short measure of transformational leadership. Journal of Business and Psychology, 14 (3), 389-405. Deci, E. L., Connell, J. P., & Ryan, R. M. (1989). Self-determination in a work organization. Journal of Applied Psychology, 74 (4), 580-590. Deci, E. L. & Ryan, R. M. (2000). The Â“wha tÂ” and Â“whyÂ” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11 (4), 227-268. Deci, E. L. & Ryan, R. M. (1985). Intrinsic Motivation and Self -Determination in Human Behavior New York: Plenum. Deci, E. L., Ryan, R. M., Gagn, M., Leone, D. R., Usunov, J., & Kornazheva, B. P. (2001). Need satisfaction, motivation, and we ll-being in the work organizations of a former eastern bloc country: A crosscultural study of self-determination. Personality and Social Psychology Bulletin, 27 (8), 930-942. Dennis, R. & Winston, B. E. (2003). A factor analysis of Page and WongÂ’s servant leadership instrument. Leadership & Organization Development Journal, 24 (8), 455-459. Dionne, S. D., Yammarino, F. J., Atwater, L. E., & James, L. R. (2002). Neutralizing Substitutes for Leadership Theory: Leadership effects and common-source bias. Journal of Applied Psychology, 87 (3), 454-464. Ehrhart, M. G. (2004). Leadership and proce dural justice climate as antecedents of unitlevel organizational citizenship behavior. Personnel Psychology, 57 61-94. Eisenberger, R., Huntington, R., Hutchi son, S., & Sowa, D. (1986). Perceived organizational support. Journal of Applied Psychology, 71, 500-507.
73 English, B. Morrison, D., & Chalon, C. ( 2010). Moderator effect s of organizational tenure on the relationship between ps ychological climate and affective commitment. Journal of Management Development, 29 (4), 394-408. Fleishman, E. A. (1953). The descri ption of supervisory behavior. Journal of Applied Psychology, 37 1-6. Fox, S., Spector, P. E., Goh, A., Bruursema, K., & Kessler, S. R. (2009). The deviant citizen: Clarifying the measurement of orga nizational citizenship behavior and its relation to counterproductive work beha vior. Loyola University Chicago. Gagn, M. (2003). The role of autonomy suppor t and autonomy orientation in prosocial behavior engagement. Motivation and Emotion, 27 (3), 199-223. Gagn, M. & Deci, E. L. (2005). Self-det ermination theory and work motivation. Journal of Organizational Behavior, 26 331-362. Gerstner, C. R., & Day, D. V. (1997). Meta-a nalytic review of l eader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827844. Graen, G. B., & Uhl-Bien, M. (1995). Rela tionship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. Leadership Quarterly, 6 (2), 219-247. Graen, G. B., Novak, M. A., & Summerkamp, P. (1982). The effects of leader-member exchange and job design on productivity and satisfaction: Testing a dual attachment model. Organizational Behavior and Human Performance, 30, 109131.
74 Graham, J. W. (1991). Servant-leadership in organizations: Inspir ational and moral. Leadership Quarterly, 2 (2), 105-119. Gray, C. E. & Wilson, P. M. (2008). Th e relationship between organizational commitment, perceived relatedness and inten tions to continue in Canadian track and field officials. Journal of Sport Behavior, 31 (1), 44-68. Greenleaf, R. K. (1970). The Servant as Leader Indianapolis, IN: The Robert Greenleaf Center. Greenleaf, R. K. (1991). Servant Leadership: A Journey into the Nature of Legitimate Power and Greatness New York: Paulist Press. Hogan, R., Curphy, G. J., & Hogan, J. ( 1994). What we know about leadership: Effectiveness and personality. American Psychologist, 49 (6), 493-504. Hu, L., & Bentler, P. M. (1999). Cutoff criteri a for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6 (1), 1-55. Hu, L., & Bentler, P. M. (1998) Fit indices in co variance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424453. Joseph, E. E. & Winston, B. E. (2005). A correla tion of servant leadership, leader trust, and organizational trust. Leadership & Organization Development Journal, 26 (1), 6-22. Judge, T. A., Piccolo, R. F., & Ilies, R. (2004). The forgotten ones? The validity of consideration and initiating stru cture in leadership research. Journal of Applied Psychology, 89 (1), 36-51.
75 Kaiser, R. B., Hogan, R. & Craig, S. B. (2008) Leadership and the fate of organizations. American Psychologist, 63 (2), 96-110. Kerr, S. & Jermier, J. M. (1978). Substitutes for leadership: Their meaning and measurement. Organizational Behavior and Human Performance, 22 375-403. Kessler, S. R., Spector, P. E., Chang, C. Â–H., & Parr, A. D. (2008). Organizational violence and aggression: Development of the three-factor Violence Climate Survey. Work & Stress, 22 (2), 108-124. Kline, R. B. (2005). Principles and practice of st ructural equation modeling (2nd ed.). New York: Guilford. Koestner, R., Otis, N., Powers, T. A., Pe lletier, L., & Gagnon, H. (2008). Autonomous motivation, controlled motiv ation, and goal progress. Journal of Personality, 76 (5), 1201-1230. Kozlowski, S. W. J. & Doherty, M. L. ( 1989). Integration of climate and leadership: Examination of a neglected issue. Journal of Applied Psychology, 74 (4), 546-553. Krebs, K. D. (2005). Can servant-leaders be sa fety indicators? Devel opment and test of a model linking servant-leadership to o ccupational safety. Unpublished doctoral dissertation, DePaul University. Liden, R. C., Wayne, S. J., Zhao, H., & Henderson, D. (2008). Servant leadership: Development of a multidimensional measure and multi-level assessment. The Leadership Quarterly, 19 161-177. Lord, R. G., & Brown, D. J. (2004). Leadership processes and follower self-identity. Mahwah, NJ: Lawrence Erlbaum Associates.
76 Mayer, D. M., Bardes, M., & Piccolo, R. F. (2008). Do servant-leaders help satisfy follower needs? An organizational justice perspective. European Journal of Work and Organizational Psychology, 17 (2), 180-197. Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., Salvador, R. (2009). How low does ethical leadership flow? Te st of a trickle-down model. Organizational Behavior and Human Decision Processes, 108 1-13. Meyer, J. P. & Allen, N. J. (1997). Commitment in the Workplace: Theory, Research and Application. Thousand Oaks, CA: Sage Publications. Neubert, M. J., Kacmar, K. M., Carlson, D. S., Chonko, L. B., & Roberts, J. A. (2008). Regulatory focus as a mediator of the infl uence of initiating structure and servant leadership on employee behavior. Journal of Applied Psychology, 93 (6), 12201233. Nevitt, J. & Hancock, G. R. (2004). Evaluati ng small sample approaches for model test statistics in struct ural equation modeling. Multivariate Behav ioral Research, 39 439-478. Organ, R. W., & Ryan, K. (1995). A meta-analytic review of attitudinal and dispositional predictors of organizatio nal citizenship behavior. Personnel Psychology, 48 (4), 775-802. Page, D. & Wong, P. T. P. (2000). A con ceptual framework for measuring servantleadership. In Adjubolosoo, S. (Ed.), The Human Factor in Shaping the Course of History and Development Lanham, MD: University Press of America. Podsakoff, P. M., MacKenzie, S. B., & Fett er, R. (1993). Substitutes for leadership and management of professional. Leadership Quarterly, 4 (1), 1-44.
77 Pool, S. W. (1997). The relati onship of job satisfaction with substitutes for leadership, leadership behaviors, and work motivation. The Journal of Psychology, 131 (3), 271-283. Price, T. L. (2003). The ethics of authentic transforma tional leadership. The Leadership Quarterly, 14 67-81. Reinboth, M. & Duda, J. L. (2006). Perceived motivational climate, need satisfaction and indices of well-being in team s ports: A longitudinal perspective. Psychology of Sport and Exercise, 7 269-286. Reinke, S. J. (2004). Service before self: Towards a theory of servant-leadership. Global Virtue Ethics Review, 5 (3), 30-57. Ryan, R. M. & Deci, E. L. (2000). Self-Det ermination Theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist 55 (1), 68-78. Ryan, R. M., Sheldon, Kasser, & Deci, E. L. (1 996). All goals were not created equal: An organismic perspective on the nature of goals and their regulation. In P. M. Gollwitzer & J. A. Bargh (eds.), The Psychology of Actio n: Linking Cognition and Motivation to Behavior (pp. 7-26). New York: Guilford. Schriesheim, C. A. & Stogdill, R. M. (1975). Differences in factor structure across three versions of the Ohio State Leadership Scales. Personnel Psychology, 28 189-206. Shore, L. M., Tetrick, L. E., Lynch, P., & Barksdale, K. (2006) Social and economic exchange: Construct development and validation. Journal of Applied Social Psychology, 36(4), 837-867.
78 Smith, J. E., Carson, K. P. & Alexander, R. A. (1984). Leadership: It can make a difference. Academy of Manage ment Journal, 27 (4), 765-766. Spears, L. (1995). Reflections on Leadership: How Robe rt K. GreenleafÂ’s Theory of Servant Leadership Influenced TodayÂ’s Top Management Thinkers. New York: John Wiley & Sons. Spector, P. E. & Fox, S. (2005). Chap ter 7: The stressor-emotion model of counterproductive work behavior. In Fox, S. & Spector, P. E. (Eds.) Counterproductive work behavior : Investigations of actors and targets Washington, DC: American Ps ychological Association, pp. 151-174. Spector, P. E., Fox, S., Penney, L. M., Bruurse ma, K., Goh, A., & Kessler, S. (2006). The dimensionality of counterproductivity: Are all counterproductive behaviors created equal? Journal of Vocational Behavior 68 446-460 Spector, P. E. & Jex, S. M. (1998). Developm ent of four self-report measures of job stressors and strain: Interpersonal co nflict at work scale, organizational constraints scale, quantitative workload inventory, and physical symptoms inventory. Journal of Occupationa l Health Psychology, 3 356-367. Stone, A. G., Russell, R. F., & Patterson, K. (2004). Transformational versus servant leadership: A difference in leader focus. Leadership & Organization Development Journal, 25 (3/4), 349-361. Walumbwa, F. O., Hartnell, C. A., & Oke, A. (2010). Servant leadership, procedural justice climate, service climate, employ ee attitudes, and organizational citizenship behavior: A cross-level investigation. Journal of Applied Psychology, 95 (3), 517529.
79 Washington, R. R., Sutton, C. D., & Field, H. S. (2006). Individual difference in servant leadership: The roles of values and personality. Leadership & Organization Development Journal, 27 (8), 700-716 Williams, L. J., & Anderson, S. E. (1991) Job satisfaction and organizational commitment as predictors of organizati onal citizenship and in-role behaviors. Journal of Management, 17, 601-617
81 Appendix A: Predictor Survey Scales Servant Leadership Scale (Ehrhart, 2004) 1. My supervisor spends the time to form quality relationships with his/her employees. 2. My supervisor creates a sense of community among his/her employees. 3. My supervisorÂ’s decisions are infl uenced by his/her employeesÂ’ input. 4. My supervisor tries to reach consensu s amongst his/her employees on important decisions. 5. My supervisor is sensitive to his/her em ployeesÂ’ responsibilities outside the work place. 6. My supervisor makes the personal devel opment of his/her employees a priority. 7. My supervisor holds his/her employees to high ethical standards. 8. My supervisor does what she or he promises to do. 9. My supervisor balances concern for day-to-d ay details with projections for the future. 10. My supervisor displays wide-ranging knowle dge and interests in finding solutions to work problems. 11. My supervisor makes me feel like I wo rk with him/her, not for him/her. 12. My supervisor works hard at finding way to help others be the best they can be. 13. My supervisor encourages his/her employees to be involved in community service and volunteer activities outside of work. 14. My supervisor emphasizes the importance of giving back to the community.
82 SDT Needs: Basic Need Satisfaction at Work (Deci, Ryan, Gagn, Leone, Usunov, & Kornazheva, 2001) When I am at work... 1. I feel like I can make a lot of i nputs to deciding how my job gets done. 2. I really like the people I work with. 3. I do not feel very competent when I am at work. 4. People at work tell me I am good at what I do. 5. I feel pressured at work. 6. I get along with people at work. 7. I pretty much keep to myself when I am at work. 8. I am free to express my ideas and opinions on the job. 9. I consider the people I work with to be my friends. 10. I have been able to learn in teresting new skills on my job. 11. When I am at work, I have to do what I am told. 12. Most days I feel a sense of accomplishment from working. 13. My feelings are taken in to consideration at work. 14. On my job I do not get much of a chance to show how capable I am. 15. People at work care about me. 16. There are not many people at work that I am close to. 17. I feel like I can pretty much be myself at work. 18. T he people I work with do not seem to like me much. 19. When I am working I often do not feel very capable. 20. There is not much opportunity for me to decide for myself how to go about my work. 21. People at work are pretty friendly towards me. Subscale items: Autonomy: 1, 5 8, 11 13, 17, 20 Reverse-coded: items 5, 11, 20 Competence: 3, 4, 10, 12, 14, 19 Reverse-coded: items 3, 14, 19 Relatedness: 2, 6, 7 9, 15, 16 18 21 Reverse-coded: items 7, 16, 18
83 Appendix B: Job Performance Scales Task Performance (Williams & Anderson, 1991) 1. Adequately completes assigned duties 2. Fulfills responsibilities specified in job description 3. Performs tasks that are expected of him/her 4. Meets formal performance requirements of the job 5. Engages in activities that will direc tly affect his/her performance evaluation 6. Neglects aspects of the job he/she is obligated to perform 7. Fails to perform essential duties Reverse-coded: items 6, 7 Counterproductive Work Behaviors checklist (CWB-C) (Spector, Fox, Penney, Bruursema, Goh, & Kessler, 2006) 1. Purposely wasted your employerÂ’s materials/supplies. 2. Purposely did your work incorrectly. 3. Been nasty or rude to a client or customer. 4. Insulted someone about their job performance. 5. Put in to be paid for more hours than you worked. 6. Blamed someone at work for error you made. 7. Started an argument with someone at work. 8. Insulted or made fun of someone at work. Organizational Citizenship Behavior checklist (OCB-C) (Fox, Spector, Goh, Bruursema, & Kessler, 2009) 1. Drove, escorted, or entertained company guests, clients, or out-of-town employees. 2. Took time to advise, coach, or mentor a co-worker. 3. Helped co-worker learn new skills or shared job knowledge. 4. Lent a compassionate ear when someone had a work problem. 5. Offered suggestions for im proving the work environment. 6. Came in early or stayed late without pay to complete a project or task. 7 Volunteered for extra work assignments. 8. Informed manager of co-worker's excellent performance.
84 Appendix C: Well-Being Scales Physical Well-being (Physi cal Symptoms Inventory) (Spector & Jex 1998) During the past 30 days, did you have...? 1. An upset stomach or nausea 2. A backache 3. Trouble sleeping 4. A skin rash 5. Shortness of breath 6. Chest pain 7. Headache 8. Fever 9. Acid indigestion or heartburn 10. Eye strain 11. Diarrhea 12. Stomach cramps (Not menstrual) 13. Constipation 14. Heart pounding when not exercising 15. An infection 16. Loss of appetite 17. Dizziness 18. Tiredness or fatigue
85 Psychological Well-Being: Work-Related Depression, Anxiety, and Irritation (Caplan, Cobb, French, Van Harrison, and Pinneau, 1980) 1. I feel sad 2. I feel unhappy 3. I feel good 4. I feel depressed 5. I feel blue 6. I feel cheerful 7. I feel nervous 8. I feel jittery 9. I feel calm 10. I feel fidgety 11. I get angry 12. I get aggravated 13. I get irritated or annoyed Subscale items: Depression: items 1-6 Anxiety: items 7-10 Irritation: items 11-13 Reverse-coded: items 3, 6, 9 Job Satisfaction (Cammann, Fichman, Jenkins, & Klesh, 1983) 1. All in all, I am satisfied with my job. 2. In general, I donÂ’t like my job. 3. In general, I like working here. Reverse-coded: item 2
86 Appendix D: Social Cohesion Scales Leader-Member Exchange (LMX-7) (Graen, Novak & Summerkamp, 1982) 1. I always know where I stand with my immediate supervisor. 2. I feel that my immediate supervisor completely understand my problems and needs. 3. I feel that my immediate supervisor fully recognizes my potential. 4. Regardless of how much formal authority my immediate supervisor has built into his or her position, s/he would not hesitate to use her/his power to help me solve problems in my work. 5. Again, regardless of the amount of formal authority my immediate supervisor has, I can count on him/her to Â‘bail me outÂ’ at his/her expense when I really need it. 6. I have enough confidence in m y immediate supervisor that I would defend and justify her/his decisions if s/he were not present to do so. 7. I would characterize my working relationship with my immediate supervisor as very effective. Affective Commitment (Meyer & Allen, 1997) 1. I would be happy to spend the rest of my career in th is organization. 2. I really feel as if this or ganizationÂ’s problems are my own. 3. I do not feel like Â“part of th e familyÂ” at my organization. 4. I do not feel Â“emotionally atta chedÂ” to this organization. 5. This organization has a great d eal of personal meaning to me. 6. I do not feel a strong sense of belonging to my organization. Reverse-coded: items 3, 4, 6 Community Prosocial Behavior (Liden, Wayne, Zhoa & Henderson, 2008) 1. I am involved in community service a nd volunteer activitie s outside of work. 2. I believe it is important to give back to the community. 3. I take into consideration the effects of decisions I ma ke in my job on the overall community. 4. I believe that our company has the responsibility to improve the community in which it operates 5. I encourage others in the comp any to volunteer in the community. 6. When possible, I try and get my organizati on involved in community projects that I am involved in. 7. I believe that an organization is obligated to serve the community in which it operates.
87 Appendix E: Explorator y Leadership Scales Consideration Leadership (Schriesheim & Stogdill, 1975) 1. My supervisor lets group members know what is expected of them. 2. My supervisor tries out his/her ideas with the group. 3. My supervisor does little things to make it pleasant to be a member of the group. 4. My supervisor refuses to explain his/her actions. 5. My supervisor treats all group members as his/her equals. 6. My supervisor is willing to make changes. 7. My supervisor is friendly and approachable. 8. My supervisor puts suggestions made by the group into operation. 9. My supervisor gives advance notice of changes. 10. My supervisor looks out for the personal welfare of group members. Reverse coded: item 4 Transformational Leadership (Carless, Wearing, & Mann, 2000) 1. My supervisor communicates a clear and positive vision of the future. 2. My supervisor treats staff as individuals supports and encourages their development. 3. My supervisor gives encourag ement and recognition to staff. 4. My supervisor fosters tr ust, involvement and coope ration among team members. 5. My supervisor encourages thinking a bout problems in new ways and questions assumptions. 6. My supervisor is clear about his/her va lues and practices wh at s/he preaches. 7. My supervisor instills pride and respect in others and inspires me by being highly competent.
About the Author Kristin N. Saboe is a doctoral student in the Industrial/Organizational Psychology program in the Department of Psychology of th e University of South Florida in Tampa, Florida and is a NIOSH Occupational Health Psychology trainee. Her primary research areas, for which she is published and given numerous presentations on, include: leadership, motivation and occupational he alth psychology. Kristin graduated summa cum laude, with honors in Psychology, from Au stin College in Sherman, Texas in 2007 with a B.A. in Psychology and Philosophy. In addition to researching leadership, Kristin has received extensive leadership training a nd experience as a community leader and as a distinguished graduate of the Posey Leadership InstituteÂ—a selective 4-year leadership fellowship at Austin College. Kristin is a member of Phi Beta Kappa Society and is a Rotary International Paul Harris Fellow.