_1 3.. .- ‘ mu‘. :4 0 xi.‘ 2’ I v v .. ‘nifih 1. a“ «g... I" v‘z‘ 1;, \II. ”9;, I; H n .r u '0 o! . an r vOl .I A‘I'h {DW- ...7. fi‘ v’. . Unr Ksfli fl‘ 0‘ -.....)‘ofiiVDX. {:9 l\ in! :a »o LAX I. -..!VIV . ‘.-11.A.v'....v . It’ll .. y; :s I. 3. y!tu..\\ .- n... . 11¢. . l . . . h ,..i . . . . .. ;. l. ....1 ,. . . .. . , y .. .. .. ; . y .. . . ‘ \ ‘ . , ... .. ...; ,. .A...h..tJ.L Hart... ..Aourvh....u.. ‘ . karat.mWarxwrfluMWaHvahmrmwuuncruvlnfllwthuvfl...vuwuunfllh.1 ....YV ,1: {rd ..IIL.|.IK.|§ . A . In L .t . '11" 15.11.; ‘ $.EI '. I. . ‘ - IVERSITY LIBRARIE “2“ lllllllllllllllllll| ll loll 31293 016 ill This is to certify that the dissertation entitled Explaining the extra-role behavior of part-time and full-time workers: A matter of support? presented by Christina L. Daly has been accepted towards fulfillment of the requirements for Ph. D. degree in Organizational Behavior Major professor MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771 LIBRARY UnIversIty Michigan State PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MTE DUE MTE DUE DATE DUE 1M cm.m14 EXPLAINING THE EXTRA-ROLE BEHAVIOR or PART-TIME AND FULL-TIME WORKERS: A MATTER OF SUPPORT? By Christina Lynn Daly A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Management 1998 ' ABSTRACT EXPLAINING THE EXTRA-ROLE BEHAVIOR OF PART-TIME AND FULL-TIME WORKERS: A MATTER OF SUPPORT? By Christina Lynn Daly The current study uses service employees to test a model designed to explain difi‘erences in work behavior among full-time and part-time employees through Social Exchange Theory. Results indicate that, although there are no difl‘erences in perceived organizational support, part-time and full-time workers distinguish between and participate in difl‘erent types of citizenship behavior. Specifically, full-time employees exhibit more overall and interpersonal-focused citizenship behavior than part-time workers, but the two work status groups do not difl‘er on their level of efl‘ort toward organization-focused behavior. Finally, relationships are demonstrated between perceived organizational support and six distinct types of citizenship behavior. To David: my partner, my champion, my hero, my life; I will love you always. ACKNOWLEDGMENTS I would like to acknowledge the efi‘ort and support of a number of special individuals who have helped me in a variety of ways: Thank you to my parents, who have shown me that I have the strength to persevere through anything. Thank you to my aunt, Kathleen Freeman, for being a role-model and so much more. You are truly an inspiration. Thank you to my grandparents, Burt and Margaret Smith, for laying the right foundation for me to succeed, and for setting such high standards to follow. Thank you to Linn, who gave me incredible guidance and allowed me to regain myself; somehow, thank you isn’t enough. Thank you to Pamela, who believed in me when others didn’t. “Without you, I would have never completed this process. Thank you to Chris, Anna, Patty, Mimi, and Ed, my counselors, my support, my comics, my reality checks, my companions, my friends. Thank you to Andrew, who reminds me how important learning is and always will be. I love you. Most importantly, thank you to God for giving me the gifts he has bestowed upon me. May I never take them for granted and always try to put them to His use. TABLE OF CONTENTS List of Tables .............................. List of Figures .............................. Chapter 1: Introduction .............................. Overview .............................. Contributions .............................. Chapter 2: Literature Review .............................. Part-time vs. F ull-time .............................. Definition and Background ........................ Attitudes vs. Behaviors ........................ Lack of Theory .............................. Social Exchange .............................. Perceived Organizational Support ........................ Definition and Background ........................ Work Status and Perceived Organizational Support ....... Hypothesis 1 .............................. Preferred Work Status ................ , .............. Hypothesis 2 .............................. Chapter 3: Work Behavior and Perceived Organizational Support ....... Extra-Role Behavior .............................. Definition and Background ........................ Organizational Citizenship Behaviors .................. Work status and Organizational Citizenship ............ Hypothesis 3 .............................. Perceived Org. Support and Organizational Citizenship . . . . Hypothesis 4 .............................. Specific Citizenship Behaviors ...................... Altruism .............................. Hypothesis 5a ...................... Social Participation ...................... Hypothesis 5b ...................... Advocacy Participation ...................... Hypothesis 5c ...................... Functional Participation ...................... Hypothesis 5d ...................... Loyalty .............................. Hypothesis 5e ...................... Obedience .............................. Hypothesis Sf ...................... U) TABLE OF CONTENTS (con’t) Chapter 3: Work Behavior and Perceived Organizational Support (con’t) Interpersonal vs. Organization Org. Citizenship .......... Hypothesis 6 .............................. Summary Chapter 4: Methods Overview Power Analysis Participants Procedure Variables Data Analysis Chapter 5: Results Overview Descriptive Statistics OLS Assumptions Factor Analyses Reliability Analyses Regressions aaaaaaaaaaaaaaaaaaaaaaaaaaaaaa oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo Chapter 6: Discussion and Future Research ........................ Overview Discussion of Results oooooooooooooooooooooooooooooo Research and Practical Implications ........................ Strengths and Weaknesses Future Research Conclusion Appendices OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Appendix A: Supervisor Cover Letter and Survey ............. Appendix B: Employee Cover Letter and Survey ............. Appendix C: Power Analysis .............................. References oooooooooooooooooooooooooooooo 43 46 46 46 47 47 48 52 61 61 61 62 63 79 79 79 89 97 101 105 107 111 114 115 Table 1: Table 2: Table 3: Table 4: Table 5: Table 6: Table 7: Table 8: Table 9: LIST OF TABLES Part-time Literature Summary ........................ Organizational Citizenship Behaviors ...................... Measurement Summary ........................ Correlations among Primary Variables .................. Result Summary for WS-POS ........................ Result Summary for PWS moderation ...................... Result Summary for WS-OCBS ........................ Result Summary for WS-Individual OCBS ................... Result Summary for POS-OCBS ........................ 31 53 67 69 70 72 74 76 LIST OF FIGURES Figure l: Hypothesized Model .............................. 5 Figure 2: Preferred Work Status efi‘ects on Work Status-Perceived Org. Support relationship ............ 21 Figure 3: Operational Model ........................ 54 Figure 4: Organizational Size effects on the Work Status-OCB relationship .................. 87 viii CHAPTER 1: INTRODUCTION Overview Recent events in today’s world indicate that there are significant differences in the type of employment relationships organizations choose to foster with various employees. Pfefi‘er and Baron (1988) argue that corporations are choosing to externalize some parts of their workforces, creating a lower level of attachment between the organization and certain types of employees (i.e., part-time). In fact, there is a current trend to increase the number of these externalized workers, and subsequently reduce the number of full-time employees, who are considered members of the organization (Pfefi‘er and Baron, 1988); recent estimates of the amount of part-time workers in our society range Horn 16 to 18 percent of the total US labor force (Bureau of Labor Statistics, 1998). Managers in these corporations view changing the nature of their work force'as a way to increase staffing flexibility and decrease labor costs. Employees may see this ofl‘er of non-traditional employment as a way to meet family and/or financial needs which tend to conflict with traditional, fiill-time schedules. However, employees who seek long-term employment with the company may view these forced changes as a barrier to attaining higher, more prestigious positions in the organization. Regardless of the impetus, the increased marginality of some workers should result in redefining the way both employers and employees perceive various facets of the employment relationship. The current study focuses on how one specific type of difference in work status, firll- vs. part-time employment, may impact the perceptions employees have about the extent to which they are valued by their work organizations, and the emfloyee’s subsequent participation in extra-role behavior while on-the-job. In light of the changing perceptions of the employment relatiOnship, both employers and researchers need to examine and understand the way work status influences this relationship, and how work status may impact job behaviors and attitudes. In examining these relationships, I rely on the concepts of perceived organizational support (e. g., Eisenberger, Huntington, Hutchison, and Sowa, 1986) and organizational citizenship behaviors (e. g., Bateman and Organ, 1983). Based on Social Exchange Theory (Blau, 1964), my general proposition is that workers who are not permanent, full—time employees do not feel as supported by their work organizations, and therefore are less likely to perform positive work behaviors that go above-and-beyond the required tasks. A model illustrating the hypothesized relationships may be seen in Figure 1. I develop my arguments in the following manner: first, I introduce the reader to the part-time research literature, including its focus on attitudinal research instead of behavioral studies. I then explain how the concept of perceived organizational support may be direme related to work status, and how it can be used to inform prior and future research findings. I use Social Exchange Theory (Blau, 1964) to show how an employee’s level of perceived organizational support may impact on the participation in extra-role behaviors, specifically positive, organizational citizenship behaviors. Next, I discuss how preferred work status may moderate the relationships mentioned above. I describe the methods I use to study these relationships, and the results of the statistical analyses. I continue with a discussion of the results, including the strengths and weaknesses of the I present study. Finally, I recommend future research related to this study that could help understand possible difl‘erences in full- and part-time workers. Contribution of the Present Research This study will help both researchers and practitioners in several ways. Practically, it examines assumptions widely held by managers about possible differences in work attitudes and behaviors held by full-time and part-time employees. Also, it provides insight into the employment relationships developed by fiill- and part-time workers, which may allow managers to better motivate, direct, and organize their labor forces. For example, if managers are informed that part-time employees generally perceive themselves to be less supported by the organization than firll-time workers, attempts can be made to make part-time employees feel more valued and cared for by the organization; this in turn may increase positive work behaviors of the part-time workers, thus create a benefit for the organization. Conceptually, this Study furthers work status research in the following ways: it examines differences in work behavior, not just work attitudes; it uses theoretical reasoning to explain possible variations in work behaviors; it theoretically compares and difi‘erentiates the employment relationship that develops between the organization and both full- and part-time employees; also, it tests a theory previously suggested by work status researchers (i.e., Social Exchange Theory), which was only used post hoc to explain variations in firll- and part-time workers’ job attitudes. Finally, two ways to measure work status are compared for possible difi’erences in their ability to predict dependent variables. The current study also benefits researchers interested in perceived organizational support and organizational citizenship behavior. It examines an increasingly important but under-researched segment of the workforce: restaurant workers. This type of employee is representative of the growing number of service-oriented jobs in the US. economy. Also, there are a significant number of individuals who work part-time in the restaurant industry. This study allows examination of possible differences in both perceived organizational support and organizational citizenship behaviors among difl‘erent types of workers, which has not been investigated in past research. Finally, a variable which may change the relationships among work status, perceived organizational support, and organizational citizenship behaviors is included in order to better understand the conditions under which this relationship might exist. .832 .3858»: "— 2:»...— 8§§6 5.93..— .eaea .8265... - tonnam oEuéum 1332-535...“qu 382.3590 .m> 43m 338.?“ # 38% etc? Tungsten 38m . 95:52 2.5% x83 33822.23?— BEBE «EDD b: to ‘01 CHAPTER 2: LITERATURE REVIEW AND HYPOTHESES PA T-TIME V . FULL-TIME EMPLOYEE DIFFEREN Definition and Background The largest group of employees working non-traditional schedules consists of part- time workers (Nollen and Axel, 1995: 10). Since World War II, the number of part-time employees has grown to almost 1/5 of the total workforce, or almost 20 million individuals (Ronen, 1984; Nollen and Axel, 1995). According to the US Government, employees are classified as part-time if they generally work under 35 hours per week (Deuterrnann and Brown, 1978; Nardone, 1986; Tilly, 1991). Most part-time positions are found in the service sector, especially among clerical, sales, and food service industries (Nollen and Axel, 1995; Tilly, 1991). The majority of part-time workers are either very young (e. g., students, teens) or older workers (e.g., retirees), and women, (Bureau of Labor Statistics, April, 1996). Popular research samples have included retail workers (e.g., Levanoni and Sales, 1990) and hospital workers (e.g., McGinnis and Morrow, 1990). These retail and hospital workers are both part of the service sector, which has the largest segment of part-time workers and is expected to continue its enormous economic growth (Tilly, 1991); however, researchers have ignored a significant segment of service workers, namely food service and hospitality employees. Studying possible differences in full-time and part-time worker attitudes and behavior is important to both practitioners and researchers. For example, managers tend to hold assumptions about the level of commitment part-time employees bring to their jobs and organizations. Specifically, it is a common belief among managers that part-time ’ workers have lower levels of commitment, greater levels of turnover, and are less willing to perform well (Gannon, 1975; Ronen, 1984; Rotchford and Roberts, 1982). Few researchers have attempted to study the above assumptions. Until recently, organizational behavior researchers have disregarded the possibility that there may be differences in part- time and full-time employees, failing to acknowledge the presence of various types of employment and the impact they may have on well-studied theories and models. Focus on Attitudes, not Behaviors The majority of research conducted on part-time vs. full-time employees has focused on the difl‘erences in job-related attitudes and virtually ignored possible variations in job behaviors. The most widely studied job attitude is job satisfaction (e. g., Allen, Keaveny, and Jackson, 1979; Feldman and Doerpinghaus, 1992; Gannon and Nothem, 1971; Horn, 1979; Jackofsky and Peters, 1987; Logan, O’Reilly, and Roberts, 1973; McGinnis and Morrow, 1990; Miller and Terborg, 1979). However, the results of these studies have been fairly inconsistent and confusing, as illustrated in Table 1 (Pierce, Newstrom, Dunharn, and Barber, 1989; Ronen, 1984). For example, some researchers have found that part-time workers are more satisfied than firll-time employees (Eberhardt and Shani, 1984; Feldman and Doerpinghaus, 1992); other studies have found part-timers less satisfied than full-timers (Miller and Terborg, 1979). Still other investigators have found no difi'erences in general job satisfaction (Levanoni and Sales, 1990; Logan, O’Reilly, and Roberts, 1973; Morrow. McElroy, and Elliott, 1994; Veechio, 1984). These inconsistencies have led researchers to focus on possible moderating variables such as preference for part-time work (Lee and Johnson, 1991; Morrow et al., 1994). Employees may choose to voluntarily work part-time for many reasons: the need to care for small children; the desire to earn more money in addition to their firll-time job, commonly referred to as moonlighting; the desire to earn spending money while attending school; and a need to Stay active when partially retired (Nollen and Martin, 1978). There are also individuals who work part-time because there are no available full-time positions; these individuals are commonly referred to as involuntary part-time workers (Deuterrnann and Brown, 1978; Nollen and Axel, 1995; Tilly, 1991). Whatever the motivation, part- tirne employees have specific reasons for working a non-traditional schedule; these reasons may affect both their workplace attitudes and performance levels. There have been several calls for research on workplace behavior difi‘erences among firll- and part-time employees; however, there has been little investigation of these outcome variables (Feldman, 1990; Hom, 1979; Miller and Terborg, 1979; Rotchford and Roberts, 1982). One study which examined behavioral differences focused on performance and turnover (Jackofsky and Peters, 1987); findings indicated that there were no difi‘erences in either outcome variable among workers in various work status groups. However, a prior study did find differences in turnover among full- and part-time workers (Peters, Jackofsky, and Salter, 1981). To firrther muddy the waters, Werbd (1985) found that family as primary life involvement predicted turnover only for part-time workers. Based on these findings, Werbel concluded that there may be different work attitude- behavior linkages for part-timers than those for full-timers (Gannon, 1975), but called for . further investigation before findings can be considered conclusive. 2.82 >9 33...... 8...... 5.. ...... >o .88 .8... use... .8328 ...—85.80.... .58. 83...»... use. as; so :53. on. .528 ....a. ...a 8. 2.5.1.3 ......eeieiefia. ._ ._ e _ 8... ESE... .2. .33 ... .25. ... as...» .... 83.96 on. .33 332%»... .55.... song... .588... ".255. we 3...... 8...... size»: 5......» .83. ..s. 8.38.. 3...; eases. .. . .... @2332 86.8 23:»? 3.5.8 38.5.5... .955. as... message. ..I s... .880 9.5.3 as... .33 3.... as... a. .556 .25: [Egg . ..5 :6. one ._o>u_.ooe ...—.20 2858.95 8...... .38.».— .8296? 35.8.8.— ..222 2 3.22 .2. m. an. guesses... .... ... 2.5.... 5.6%.... ..o. :8: E332 e 8:56 ,3 agfillluufiu 22.9.3... 8 9.8... .953. .09.... :80 :89. ...—.8“ Eon—E at)... ..c a use 2 2.8... Egon 3.38.5 .3 ...—Ex... {can 22: ... as. 8.. ...... maps: 3.... BE: 2.32 a... ....a ...... ...oo 2.. ...... 2:... as: Bamafisn. e 5.52.... e an... 2.3.. 25.8.. 80.8.9.8 435 85.5 .558“ 8259.52 .e g 8. .85.. so... ... a. ass: ......a ...... .3 .35. ...; 3.... ....am a 6.2.3.. . . Mg... 0.. 22.32.... as .83 .22.». ......58 .90 .58... .8320... 8.8... .9. use. a... R. 8.. .....3. .85» a}. ....a 8. .50 3.8.. .33 .33.. 35.3 a. 33.3 1% .20 3 mega ...oEeB Ami—my 0.95% .323:— bfifix: .8... 32.6... .33 £5.82 a. a as... so... .... a .289. .es§_o> so 25...... .9». .580 33: ES... a. 55523.. .282... d 39...... .38 use... 3.5.. .5338 $0.36 9.38 62:8 fig: 356...: £ £82.30 .3 3.8... .23.»: e... 8.52 38...... ...»... 15282 ....a ...... .8... «ea. .33 2.532.? £839.32 8:232. 2.3.500 80:20.3..— .eoEomaeaE ..o REE... ..ua .8 ass. .83 as. 5.. 9.62 .328 23...... 8. sea. as? as: :83... e ...—3.3. .85. _ z anal lanai... lid .5 .3 >- Hu— had .. u— cash. .00 .889. no“ 0. 000 rum. 920m 833 0.000.: 0003 30302 .003 02.002 .0003 .030. 0.00 0.00 D .000002 80000 02.0-5.0 0.0.00... 0.0000. 800: 0030 800.00 01.. 0. 00002020 0000...... 00.. 020.0 00.03 30.000000 0. 88.2m— 00 02.00.. 8:0... 000008020 00.. 08 .0000 80.00.000.00 02......00 $00: 00.0w.— 000 029—2 002.100.00.008... 80. M...... 500.00.. ... .3... .02.... .0000... 5000000300 30.5. .83 .0082 02:20 0 00000000000 D .000002 VMO .8050 00.00.00 0.0.0.. ..3 .0003 300: 0000.02 :EEE a . .q QB...— e >5m 2 ES:— __228 00.0.3.0 .0000. .828. .w8 0000.00. ......00 00_. .000... .0000... 00.800... 82003 0000000> ..0 00:00.. 0. 00.0.0.0: 0.0.0.0 083 «5 000.02 0000003. 0.00003 .2028 ...03 0.0.0.. :83 000: ..0_=m 0.. $0.502 .3282 .0000 :0 0.000. a. 00m £0.05 00.. .0. 5.00000 ..8 00.. .0000 0.000. 32000.00 00:00< 00 00.00.. ...83 .3 .0000... 000— 0.0 $0. ...—.20 :50: 0.000. ..00000 000 0.0000 000 .0206 x83 85: 900.0... a. .052 333350.200 .0 a 0.02:0 0.0 a 0000 00000 0000.0. 00 $0.0 02.8000 .000. .0008 0.03 000. .0M0 ..00000 82008. 000080> ..0 .0080. :0... 0.. 0000...... 0.0 0.. so .008: 00.. .0050 .... ...... .8... 0...... U....3 .08.. 38.02 0 000.022 .000. 000. . 303.00 2.0000,. .00 0000000080 000 00020000 0008000 2.0080 00. 0000.00080 005 0:008 90 0. .030_ .00 ... n. .000 0.002 .0w0 .8280 20000.0(. 0.00.80 .90 00.0.. 0.03 .000: 0002 Q 00000.02 .0000 0.0.000 0.0... .00 000 .000. 83003. . . ..0 000.00.. .... :00 .... ...... 0.. ..z ... .8 00.08: c 0... .0050 .8 ...... .8... 8.0.. .....3 5.2. 22.3. 0 0000.0 .....03 00000. «00 083.000. 20.300 \3 .00. 0.00. 0.0 0009. 00.. 30.00000 300m ..0 30.00.. ..00 00m 00 mg .3. 02 .0. ..00 0.0.0.0 0.000. ..00000 ..00 00.. .000... 0.0.0.0 083 E: 8.0m 0... 2000.5..— w.0_3 2.09020. 0. 0.0.0 00 0.000. 0.000.000. 0.00000... 0.0 .0000 ..0... 000 003.0m 0008000 00020020 0.20.00 .90 83000.0 .000080> .8 00.00.. , 0... 0.2.0.. ... 0.... .0. a... .0... Ecuai .00.. .5080 0.28.00 .....a ...... .50 0...... .003 .5... 08.0.... .0 8.. 000. \3 000000.25 .0... .2200 .30 a. 000m :3 0023.000 .000 E0000 .2228 .90 .200 0008 00—. ....00 000.0: 2.0 832.5 .90 ... 8...... .00 a8. 0... 2. 0.8.0.2 5.38 .0... 0...... .....3 00.0.2.8 0... as: 00.... .0 0...: 0.3.0.00. IJEHE .0 0.00.00 02.5... :mflllnda >0 2 .2... 00:00.. .85.... 20%| Ilaqugz <0 0 0.00.0.5. .. ......8. . 0...... 10 aka-...... I D ”Bantam .31— 3 .883— I ma: p.830 :83 5.5.5 133-: I DIG: E 3. 52o .2553 3.333 2083.35 «3.3 383.— 3580295 Borg—om 3:38:80 .3 5:2 a: 55.... a 2:5 3. 82:2 on“ .528 83:5. a: SEE £3: .333 7 858.3 now 2 9:23 .852: 3:323 . ES. .8 £26. 39:56 893.95 8% .99 .260 new £2.55 Anna: :99.“ .3 95.5.. 2.: .518: on. .536 2. 055m. 3% as? 8F. a. 5:82 5.5 6.6%; 8.6%:2 8.. 2 in 648 gm .8? .82. eggs a: 82% .8223 33888 .3 .58. 3.22 .2. was» as; -32 38.2 on“ .588 2a 8.. 386 3% £25 £3: £82, £9.33 ..35 2.. .E 648 gm 53 2.35 22.8 .33 BE... .33 £582 a 532m 282 .82 a .232 a 5320 a 89.20 :8: 2:... . 832a 8.. 8 E a 808.98 8.2.53 .ea 2:33. ... 355. .E 33 89.20%. oz 2“ E 83...; a: was» as? an”: =3. m can 32 E 58% 0.2 aouoémufl 328.3 £338 can 228. .88 88". .586 Egon 3.3.3590 .6 .552 3 .8 S 808355 88 so 333: can .380 22 8385 3.8a to? 88: 8:2 a Gem 32:38 3.325» 3 8 38 92 Bu 3 use 85% 5 . 9.2%.. 93 ”858m a. .58. «Eu 3.9.. 5.93 oz 5 82:2 8.8 38.2 8.3» as? $8: 3:32 a. gm Bo 83m :85 .9558 .50. 328 3:38:80. .8. .Esoin I: 2.29.0.— ...a 585:... use» .53 £3: 5:8 a .3303 .202 lag—an] a Asa—«gag a $— .— 52 5.53 _ as; 11 Lack of theoretical grounding These conflicting results in both attitudinal and behavioral research pertaining to part-time workers create a need to explain these variations through solid theoretical arguments (e. g., Feldman, 1990; Lee and Johnson, 1991; Miller and Terborg, 1979). Most of the past research into full- and part-time difl‘erences is purely descriptive, providing little evidence of theoretical testing. Another way to explain these contradictory results is to build theoretical arguments to explore the relationship between work status and workplace behaviors. Few studies have developed hypotheses based specifically on a theory or concept when examining full- and part-time employment. Some theories and concepts that were used post hoc to explain previous findings have included frame-of- reference and social comparison theory (Eberhardt and Shani, 1984; Feldman, 1990; Goodman, 1977; Miler and Terborg, 1979), equity theory (Adams, 1965; Feldman and Doerpinghaus, 1992), human capital theory (Becker, 1964; Lee and Johnson, 1991), partial inclusion (Katz and Kahn, 1966; Miller and Terborg, 1979), and the discrepancy model of job satisfaction (Lawler, 1973; Locke, 1969; Morrow et al., 1994). The theory that offers the most promise in developing testable hypotheses concerning differences in the type of employment relationship that develops between full- and part-time employees and their organizations is Social Exchange Theory (Blau, 1964; Lee and Johnson, 1991). L EX N E RY According to Blau (1964), social exchange should be differentiated from economic exchange within the organizational context. Ecomonic exchanges are traditionally based 12 on quid pro quo transactions: if an employee performs a certain set of behaviors for the organization, he will receive a certain level of pay in return. In other words, the parameters of the transaction are known before the behavior occurs. However, in social exchanges, one party will behave in a certain manner with the expectation that he will receive some reciprocation from the other party at somefirture date (Blau, 1964). More specifically, an employee may perform his job with extra effort, trusting that his employer will notice and feel obliged to reciprocate by a granting a promotion. The way in which the organization treats each individual employee, and the employees’ perceptions of this treatment, may explain possible variations in social exchange relationships in an organization. Employees tend to personify the organization through the actions of its agents, such as supervisors and coworkers (Levinson, 1965). Perceptions of organizational actions directed toward specific employees are subjected to individual interpretation, during which employees make inferences in an attempt to understand the behaviors. These attributions may help the employee develop a sense of how much the organization values her contributions to company efi‘orts. Also, employees may develop the perception that certain individuals are more or less valued and important to the organization than other coworkers. For example, Tsui et al., (1995) argue that there is no employer expectation of psychological involvement on the part of certain employees. These employees may perceive this lack of expectation in regard to their work efl‘ort as a sign that the organization does not support or value them. Perceived organizational support (e.g., Eisenberger, Huntington, Hutchison, and Sowa, 1986) is one way researchers have investigated the social exchange relationships 13 that exist between employees and their work organizations. This construct is considered appropriate to examine the employee’s perceptions of the organization’s commitment to the employee (Eisenberger et al., 1986). Since I am interested in the relationship that develops between difl‘erent types of employees and their work organizations, I will use perceived organizational support to explain why work status and employee work behavior might be related. PERCEIVED R ANIZATI NAL RT Definition and background Perceived organizational support is defined as the global belief held by an employee concerning the extent to which the organization values her contributions and cares about her well-being (Eisenberger et al., 1986: 501). It is primarily afi‘ected through the same factors that Blau (1964) argues peOple use in making attributions, namely the frequency, extremity, and judged sincerity of statements of praise and approval given to employees by organization representatives. It may also be influenced by any other factor that could signify how the organization views the employee, such as level of pay and job title (Eisenberger et al., 1986). Finally, Eisenberger and his colleagues (Eisenberger et al., 1986; 1990) believe that perceived organizational support will increase the affective attachment felt by an employee toward her employing organization. The construct of perceived organizational support has received significant I empirical attention within the last 10 years. Eisenberger et al., (1986) reported that perceptions of being valued and cared-for by the organization are associated with lower 14 absenteeism rates. Eisenberger, Fasblo, and Davis-LaMastro (1990) found that high levels of perceived organizational support are related to employee conscientiousness in regard to carrying out job duties, innovative suggestions to improve the organization, and employee affective commitment towards the organization. Shore and Tetrick (1991) conducted a construct validity study on perceived organizational support and found that it was distinct from both afl'ective and continuance commitment (Meyer and Allen, 1984). In other words, employees are, “. . .able to distinguish their own commitment levels to the organization from their perceptions of the organization’s commitment to them” (Shore and Tetrick, 1991: 640). Investigations have also examined perceived organizational support as it relates to work behaviors. Shore and Wayne (1993) found that afi‘ective commitment and perceived organizational support were both positively related to organizational citizenship behavior and some impression management behaviors. Also, perceived organizational support was determined to be a better predictor of these work behaviors than affective commitment (Shore and Wayne, 1993). However, Settoon, Bennett, and Liden’s (1996) findings somewhat contradict this. They compared perceived organizational support and leader- member exchange (e.g., Dansereau, Graen, and Haga, 1975) as predictors of organizational commitment and both citizenship and in-role behaviors, and found that leader-member exchange was a better predictor of both types of work behavior. More specifically, both in-role and extra-role work behavior seems to be more associated with the employee’s relationship with her supervisor than the employee’s relationship with the organization (Settoon et al., 1996). 15 To examine this contradiction more thoroughly, Wayne, Shore, and Liden (1997) conducted a thorough study investigating antecedents and consequences for both perceived organizational support and leader-member exchange. Their structural path estimates indicate that perceived organizational support was affected by developmental experiences, promotions, and organizational tenure, and predicted affective commitment, intentions to quit, and organizational citizenship behavior. Leader-member exchange was related to organizational citizenship behavior, but at a lower level than perceived organizational support. Also, the authors found that leader-member exchange has a strongly significant efl‘ect on perceived organizational support (Wayne et al., 1997). Leader-member exchange may have a large impact on perceived organizational support because the employee views his supervisor’s behaviors as an indication of how the organization values him (see arguments above, Eisenberger et al., 1986). Levinson (1965) argues that employees personify their organization through the actions of others in the organization that may act as agents. Thus, supervisors may differentially treat employees based on variety of reasons, which will directly or indirectly impact on their perceptions of how much the organization values and cares about them (Wayne et al., 1997). Work status and perceived organizational support One way in which the organization may consciously choose to treat employees differently is to increasingly externalize the organization’s workforce (Pfefi‘er and Baron, 1988). In other words, organizations purposely create a lower level of attachment between it and certain types of employees. The company may accomplish these 16 attachment difl‘erences through variations in temporal exposure to organizational processes. More specifically, less temporal attachment, such as that which stems from part-time employment, results in a more extemalized employee with lower levels of organizational attachment (Pfeffer and Baron, 1988). prart-time employees perceive that the organization has little attachment to them, then these employees will not develop an attachment to the organization, as illustrated by the positive relationship between perceived organizational support and afl‘ective commitment (Eisenberger et al., 1990). Specifically, part-time employees will not perceive that the organization values them and cares about their well-being, because the organization is not fostering an emotional bond evidenced by feelings of attachment. F ull- time employees, however, should develop greater organizational attachment because of their temporal exposure to organizational processes. Therefore, they should also develop feelings of being valued and cared-for by the organization as one of it’s important, internalized employees. This is congruent with Social Exchange Theory arguments. Full-time employees may sense that the organization believes them to be valued members, and thus feel responsibility to reciprocate by valuing the organization, via the norm of reciprocity (Gouldner, 1960). Part-time workers would not perceive the same responsibility to the company because they do not believe that the organization supports them and cares about their well-being. Thus, part-time employees do not have a need to reciprocate feelings of value and importance in regard to their work organization. Therefore, I hypothesize that: 17 H1: There will be a positive relationship between work status and perceived organizational support, such that full-time work will be associated with higher levels of perceived organizational support than part-time work. PREFERRED W TA S There may be some individuals, approximately 18% of all part-timers (Bureau of Labor Statistics, 1998\), who have to work part-time because desirable full-time positions are unavailable, or they are unqualified to fill them (Tilly, 1991). This group of part-time workers are generally known as involuntary part-time employees (Deuterrnann and Brown, 197 8; Tilly, 1991). Recent studies have found differences in voluntary and involuntary part-time workers, using a ‘work status congruency’ variable to determine if workers who were voluntarily working part-time difl‘ered in job attitudes from those who were in an ‘incongruent’ work status, or working part-time but desired fiill-time work (e.g., Tansky, Gallagher, and Wetzel, 1995; Armstrong-Stassen, Al-Ma’aitah, Cameron, and Horsburgh, 1995). Also, there has been research on the efl'ect of ‘preferred’ work status and schedule on job attitudes (Lee and Johnson, 1991; Morrow, McElroy, and Elliott, 1994). Specifically, Lee and Johnson (1991) found that permanent full-time workers had more organizational commitment, but not more job satisfaction, than permanent part-time workers when both worked a preferred schedule. The influence of preferred work status may explain why past part-time and full- time employee difference research has found conflicting results. Several early studies found significant difl‘erences among full- and part-time workers (e.g., Gannon and 18 Nothem, 1971; Logan, et al., 1978; Miller and Terborg, 1979); however, recent studies have determined that most of these relationships are actually changed by the employee’s preference for work shift or schedule, or their voluntary vs. involuntary work status (e.g., McGinnis and Morrow, 1990). For example, a part-time restaurant server may be just as satisfied as a full-time restaurant server if he voluntarily wants to work part-time because he has another job or is a student. The current study is designed to determine if preferred work status may also influence the relationship between work status and level of perceived organizational support. Based on the studies previously mentioned, preferred work status should act to change the hypothesized relationship between work status and level of perceived organizational support for part-time employees, as can be seen in Figure 2. Specifically, when part-time employees are working under their preferred work status, the positive relationship between work status and perceived organizational support should weaken. In other words, given that they are working their preferred schedules, part-time employees may develop similar levels of perceived organizational support to those experienced by full-time workers. For example, part-time employees who prefer to work part-time may view the organization’s granting of their preferred work status as an indication that the organization cares about their well-being. They may feel as if the organization is being supportive of the employees’ choices, which can increase the employees’ perceptions of being highly supported by the company. Thus, there should be less variance in level of perceived organizational support for fiill- and part-time workers when part-time employees are working under their preferred work status. 19 Under conditions in which employees are not currently working their preferred work status (e.g., part-time employee wanting to work hill-time), there should be a stronger relationship between the type of work status and the employee’s level of perceived organizational support than the relationship which is found when employees are working their desired work status. Specifically, the work status - perceived organizational support relationship should be as hypothesized in H1: full-time employees should perceive higher levels of organizational support than part-time workers. Both types of employees may blame the employer for not scheduling them to work the number of hours per week they desire, and may feel as if the employer does not believe that they are valued enough by the organization to warrant granting schedule requests. However, full-time employees may be granted other opportunities which may indicate high levels of support from the organization, namely promotions and additional training. These rewards are typically not available to part-time employees (Pfefl‘er and Baron, 1988), and thus there is no compensating mechanism to allow part-time workers to feel valued by the company. Therefore, based on the above arguments, I hypothesize that: H2: Preferred work status moderates the relationship between work status and perceived organizational support, such that non-preferred work status working conditions create a stronger relationship between work status and perceived organizational support than preferred work status working conditions. 20 High working preferred work status PERCElVED not working ORGANIZATIONAL preferred work status SUPPORT Low Part-time Full-time WORK STATUS Figure 2: Preferred work status moderation of the relationship between work status and perceived organizational support 21 CHAPTER 3: WORK BEHAVIOR AND PERCENED ORGANIZATIONAL SUPPORT Level of perceived organizational support may be related to the employee’s behavior while on-the-job. More specifically, the employee’s willingness to participate in actions that are outside of typical job requirements may be impacted by the relationship the employee perceives having with the employer. Research on Social Exchange Theory provides solid theoretical reasoning for the existence of a relationship between perceived organizational support and extra-role behavior. First, I review and discuss extra-role behavior, and specifically organizational citizenship behavior. Then, I report past findings about the relationship between perceived organizational support and organizational citizenship behavior. Finally, I develop hypotheses connecting perceived organizational support with specific types of organizational citizenship behavior. W Definition and background According to Van Dyne, Cummings, and McLean Parks (1995), positive extra-role behavior is “...behavior which benefits the organization and/or is intended to benefit the organization, which is discretionary, and which goes beyond existing role expectations” (218). They continue by stating that extra-role behavior must be intentional, voluntary, and does not necessarily result in formal reward or punishment. Finally, they conceptually distinguish several types of positive extra-role behavior, such as prosocial behavior (Brief and Motowidlo, 1986; O’Reilly and Chatman, 1986; Puffer, 1987), and organizational 22 citizenship behavior (Bateman and Organ, 1983; Organ, 1988; Smith, Organ, and Near, 1983). In other words, they argue that positive ‘extra-role behavior’ is a broad concept, one category of which is organizational citizenship behaviors. Organizational citizenship behaviors In 1977, Organ broadened the research on the satisfaction-performance link by suggesting a new type of performance construct: extra-role, citizenship-type behavior. Researchers had virtually ignored Katz and Kahn’s (1966) distinction of in-role versus supra-role behavior, despite their argument that extra-role behaviors are imperative for organizational survival. Organ (1977) argued that, while satisfaction may not have had a clear relationship with in-role performance, it probably had a much greater impact on Organizational citizenship behavior. This may be due to the discretionary nature of these behaviors, since they are not constrained by job descriptions and organizational or group norms. Organ (1988) defined organizational citizenship behavior as “...individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and that in the aggregate promotes the efi'ective functioning of the organization” (4). This construct has been divided into specific categories of behavior, probably the most widely known being the S-factor framework developed by Organ (1988) and refined by Podsakofl‘ and colleagues (Podsakofl‘, MacKenzie, Moonnan, and Fetter, 1990). However, a recent construct validity study conducted by Van Dyne, Graham, and Dienesch (1994) found theoretical and statistical evidence for five additional categories, Obedience, Loyalty, 23 Advocacy Participation, Social PartiCipation, and Functional Participation. Also, Williams and Anderson (1991) supported a 2-dimension construct, with organizational citizenship behaviors that are directed toward the organization (i.e., OCBO) and organizational citizenship behaviors that are directed at individuals (i.e., OCBI). According to Van Dyne et al., (1994), each of the above categorizations is useful, and selection of a specific set of citizenship behaviors should depend upon the research question being studied. I will discuss specific types of citizenship behavior that are relevant to my research, and build corresponding hypotheses, later in the paper. Since Organ’s (1977) seminal article, there has been a large amount of research conducted on the antecedents of organizational citizenship behaviors. Organ and Ryan (1995) completed a meta-analysis of 55 studies in an attempt to summarize the literature of organizational citizenship behavior antecedents. They found that several job attitudes (e. g., job satisfaction, organizational commitment, perceived fairness, and leader supportiveness) were basically similar in their ability to predict 5 types of citizenship - altruism, general compliance, courtesy, sportsmanship, and civic virtue. They concluded that, because these attitudinal constructs tend to be highly related, they may be tapping a common psychological factor (Organ and Ryan, 1995: 794). I propose that the common factor is connected to the level of perceived organizational support experienced by the employee. Specifically, perceived organizational support has been shown to be associated with a variety of organizational commitment and leader-member exchange factors (i.e., dyad tenure, liking, favor doing; Wayne et al., 1997). Also, Shore and Tetrick ( 1991) 24 found that measures of perceived organizational support and job satisfaction were highly related, so much so that there was some issue about the separateness of the constructs. These findings indicate the possibility that prior relationships described in the Organ and Ryan (1995) meta-analysis between the various job attitudes and organizational citizenship behaviors may‘not be as clear cut as currently believed. More specifically, perceived organizational support may be a common predictor of both job attitudes and citizenship participation, which would result in misleading findings concerning the relationships between certain job attitude variables (i.e., organizational commitment, job satisfaction) and organizational citizenship behaviors. This area lies outside of the current study, but may provide interesting firture research questions. Work status and organizational citizenship behaviors I am aware of no prior published research on the relationship between work status and participation in extra-role behaviors. However, there are several reasons why part- time employees may differ fi'om full-time employees in their willingness to engage in non- required tasks. Feldman (1990) and Rotchford and Roberts (1984) both argue that part- time employees’ nonwork commitments may be greater in number than full-time workers, and part-timers may place more of an emphasis on nonwork involvements than full-timers. For example, full-time employees may be more agreeable about attending training seminars during oE-hours than part-time workers because part-timers may have other organizational commitments as far as family, another job, and school. In other words, 25 non-work responsibilities may interfere with the part-time employee’s ability to perform certain types of organizational citizenship behavior. Also, a significant number of part-time employees are students or elderly individuals who do not intend to work for the organization on a long-term basis (N ollen and Axel, 1995). Specifically, part-time employees have increased intentions to turnover. Feldman and Doerpinghaus (1992) and Peters, Jackofsky, and Salter (1981) both found that part-time workers tend to leave organizations more frequently than full-time workers. If part-time employees intend to leave their work organizations, then they may have little reason to invest extra time and efi‘ort in the organization, thus negating the desire to participate in extra-role behaviors. Based on the above arguments, I hypothesize that: H3: There will be a positive relationship between work status and level of participation in organizational citizenship behavior, such that full-time workers will participate in more organizational citizenship behavior than part-time workers. Based on the arguments above, I predict that work status will be positively related to all types of citizenship behavior. In other words, I do not predict difl‘erential efl‘ects for the six specific behaviors examined in this study. However, since many researchers are now calling for more fine-grained analyses in regard to citizenship behaviors (e.g., Van ‘ Dyne et al., 1994), I will separately test the relationship between work status and all six behaviors, as well as the three composite variables. 26 Perceived organizational support and organizational citizenship behaviors As discussed previously (see pages 18-20), there are a few empirical studies which found a significant relationship between level of perceived organizational support and willingness to perform extra duties or put forth extra efi‘ort while on the job (e. g., Eisenberger et at, 1990; Wayne et al., 1997). These authors have relied on Social Exchange Theory (Blau, 1964) to explain their findings. Specifically, Shore and Wayne (1993) argue that, “...employee behavior that goes beyond role requirements is most likely to be elicited when the employee feels obligated to repay the organization for support received” (778). Since this relationship is strongly supported in past research, I will delay discussing the specific types of citizenship behavior until afier I have more thoroughly addressed the relationship between work status and organizational citizenship. Employee expectations may be different for firll- and part-time workers. Rousseau (1989) argues that part-time work status may “...afl‘ect how much parties expect and possibly need fi'om the (employment) relationship”, and “Individuals who work firll time...rnight well expect and receive more forms of gratification or compensation than a paycheck...”(132). In other words, a server in a restaurant who works part-time may not expect any type of reward for good performance except good tips. However, a server who works firll-time may expect a promotion to head waiter or supervisor as a reward for top performance. These expectations of certain types of rewards may have developed because of difi‘erences in support and encouragement given to the difi‘erent types of employees. Full-time servers may perceive the opportunity to move up in the company or 27 receive specialized training, as well as gain valuable inside information about various organizational events, whereas part-time servers may not receive the same sense of being a valued, informed member of the organization. It is this possible difference in perceived organizational support that may explain variations in citizenship participation among firll-time and part-time workers while arr-the job. Work status may be directly related to specific organizational citizenship behaviors that involve working extra hours or coming to work ofl‘ the clock (see prior section). However, this does not explain why firll- and part-time employees may behave differently when they are at work during scheduled hours. Perceived organizational support, and Social Exchange Theory, can provide reasons why a full-time bartender would be willing to participate in Total Quality Management training whereas a part-time bartender may not be willing to attend. The full-time employee may feel a responsibility to agree to the training, because she perceives that the opportunity for more training is an indication that the organization values her input and participation. Therefore, she feels that she is expected to participate in reciprocation for organizational support. Part-time employees seldom receive the same benefits, opportunities, or information that firll-time workers do (Pfeffer and Baron, 1988). Thus, they may perceive that the organization does not support their increased involvement in the company, nor expects any extra efi‘ort from them. Subsequently, there is nothing for them to reciprocate, or no reason for them to feel obligated to the organization. To summarize, I hypothesize that work status may be directly related to all types of organizational citizenship behavior (see H3, pg. 26), but that most of the variation in 28 citizenship participation among full- and part-time employees is due to differences in how the employees perceive their employment relationship with their work organization. According to Social Exchange Theory, employees should experience different levels of perceived responsibility towards the organization based on the organization’s actions toward the employees. Specifically, the level of organizational support perceived by employees, which may difl‘er among full- and part-time workers, may be associated with their willingness to participate in organizational citizenship behaviors. Given the strong relationship between perceived organizational support and organizational citizenship behavior, and the above arguments based on Social Exchange Theory, I hypothesize: H4: Perceived organizational support will mediate the relationship between work status and organizational citizenship behaviors. Specific citizenship behaviors There are three primary classification fiarneworks for organizational citizenship behaviors; the first, and most commonly used, was developed by Organ (1988) and supported through the work of Podsakofi‘, MacKenzie, and colleagues (e.g., Podsakofl‘ et al., 1990). The second set of categories was deve10ped through theoretical and statistical techniques by Van Dyne et al., (1994). These frameworks are surprisingly dissimilar in the naming of their various citizenship dimensions, despite their arguments about measuring the same basic construct. The Organ/Podsakofi' scheme has 5 categories which include: Courtesy, Civic Virtue, Sportsmanship, Conscientiousness, and Altruism. The Van Dyne 29 et al., (1994) fiamework uses Obedience, Loyalty, Advocacy Participation, Social Participation, and Functional Participation. Even though the categorical names appear to be quite different, upon closer evaluation, the definitions of each category significantly overlap (see Table 2). The third categorization framework was developed by Williams and Anderson (1991). It divides citizenship into two types: interpersonal-focused behaviors and organizational-focused behaviors. Interpersonal-focused behaviors are directed toward individuals who work around the focal employee performing the citizenship behavior. Helping a coworker finish a project, assisting the supervisor with paperwork, and taking over the tasks of a sick colleague are all considered interpersonal-focused citizenship behaviors that may lie outside of traditional job descriptions. Organizational-focused behaviors are intended to benefit the organization directly, such as paying extra close attention to organizational rules and procedures, and developing ideas to improve organizational profitability or reduce costs. Table 2 illustrates the definitions of the various citizenship behaviors. It also shows how these behaviors may be separated into the Williams and Anderson (1991) dichotomous fi'amework of interpersonal- and organizational-focused citizenship behaviors. Citizenship behaviors accompanied by an asterisk (') are included in the present study. I will use the Van Dyne et al., (1994) categorization fiamework for this study for several reasons. First, these items better reflect possible behaviors exhibited by service employees if they were to participate in organizational citizenship (e. g., promoting company products and services, gaining extra training to improve performance). Second, 30 Table 2: Organizational citizenship behaviors Dimension Name Definition Altruism/Helping " “Discretionary behaviors that have the effect of helping a specific other person with an organizationally relevant task or problem” (Podsakoff et al., 1990: 115); OCBI Social Participation ‘ "...noncontroversial behaviors that involve interaction with other individuals, including attending meetings, engaging in positive communications with others, and involvement in other amliative group activities” (Van Dyne et al., 1994: no); arm Courtesy Advocacy Participation ‘— “Discretionary behavior on the part of an individual aimed at preventing work-related problems with others from occurring” (Podsakoff et al., 1990: 115); OCBI "...beiiaviors targeted? other members of an organization and reflecting a willingness to be controversial, such as innovation, maintaining high standards, challenging others, and making suggestions for change” (Van Dyne et al., 1994: 780); OCBO/OCBI Functional Participation “ “...participatory contribution in which individuals focus on themselves rather than others in their organizations but yet contribute to organizational effectiveness, including performing additional work activities, self-development, and volunteering for special assignments” (Van Dyne et al., 1994: 783); OCBO Loyalty“ “...represents allegiance to an organintion and promotion of its interests” (Van Dyne et al., 1994: 780); OCBO Obedience ‘ ‘mrepresents respect for the rules and policies of an organintion and willingness to expend appropriate effort on WM OCBO Civic Virtue “Behavior on the part of an individual that indicates that he/she responsibly participates in, is involved in, or is concerned about the life of the company” (Podsakofi‘ et al., 1990: 115); OCBO Conscientiousness “Discretionary behaviors on the part of the employee that go well beyond the minimum role requirements of the organintion, in the areas ofattendance, obeying rules and regulations, taking breaks, and so forth (Podsakofi‘ et al., 1990: 115); OCBO Sfportsmanship “Willingness of the employee to tolerate less than ideal circumstances without complaining -- to “avoid complaining or petty grievances, railing against real or imagined slights, and making federal cases out of small potatoes” (Organ, 1988: 11)” (Podsakofi‘ et al., 1990: 115); OCBO Williams and Anderson (1991) classification scheme: OCBO = Organintion-focused citizenship behavior OCBI = Individual-focused citizenship behavior 31 Van Dyne et al.,’s (1994) scale items cover a broader range of specific behaviors than Podsakofl‘ et al.,’s (1990) scale items and the dimensions are based on theory. This will improve the probability of measuring certain types of citizenship behavior that are commonly overlooked. Finally, the fiamework better supports the theoretical foundation of the current study than the Organ/Podsakofi‘ factors. The Van Dyne et al., (1994) instrument is based on civic citizenship theory, in which individual behavior may affect the success of the state. As described by Graham (1991), civic citizenship is impacted by relational ties between the citizen and the political state. In an organizational setting, emwoyees’ citizenship behavior intended to benefit the organization would be afi‘ected by the perceived relational ties between employees and the company. In order to better examine difi‘erent types of specific citizenship behavior and how employees participate in them, I will also use the Altruism/Helping scale developed by Podsakoff et al., (1990). Because restaurant employees are required to be fiiendly and helpfill by the very nature of their job, it is natural to believe that they may behave in an altruistic manner toward coworkers. Social Exchange theorists would argue that this may result in a large amount of reciprocity experienced among restaurant workers, which would subsequently lead to more citizenship-type behavior. The Altruism scale is not redundant with any of the other Van Dyne et al., (1994) dimensions, and has proven to be one of the most robust measures of extra-role behavior developed (see Bateman and Organ, 1983; Organ and Ryan, 1995; Smith, Organ, and Near, 1983). Since it is a prime ' example of interpersonal-focused organizational citizenship behavior (Williams and Anderson, 1991), it will provide insight into whether or not employees difl‘erentiate 32 between both the type of extra effort they are willing to put forth and the beneficiary of their efforts. I will use Social Exchange Theory to build hypotheses for each of the six types of organizational citizenship behaviors, starting with two interpersonal-focused behaviors: Altruism and Social Participation. Advocacy Participation, which may consist of both interpersonal- and organizational-focused behaviors, will follow, and then three organizational-focused behaviors will be discussed: Functional Participation, Loyalty, and Obedience. Finally, I will develop a hypothesis concerning the distinction between interpersonal- and organizational-focused citizenship behaviors. ' Altruism is defined as “(d)iscretionary behaviors that have the effect of helping a specific other person with an organizationally relevant task or problem” (Podsakofi‘ et al., 1990: 115). Perceptions of support, through Social Exchange Theory, may increase the likelihood that employees will perform extra-role, organizational citizenship behaviors directed towards others who represent the organization, such as managers and coworkers. Perceptions of responsibilities that the employees view as reciprocal may encourage extra efi'ort in helping others achieve organizational goals, through the expectation that the person receiving help will reciprocate at some point in the future. For example, a server in a restaurant who perceives himself to be valued by the organization may help another server who is extremely busy, with the knowledge that the busy server will help the focal server at a later date. This helping behavior ensures that the customer will receive good 33 service and will return to the restaurant, so these altruistic acts also indirectly help the organization meet its goals. According to Social Exchange Theory (Blau, 1964), those individuals who do not perceive themselves to be highly supported by the organization may be less likely to act in an altruistic manner towards others in the organization, because they will not expect it to be reciprocated. They do not believe that the organization and its members consider themselves to be responsible for providing employees with any type of reciprocation for extra effort provided by the employees; therefore, these employees may feel as if they are the only ones ‘giving efl‘ort’ in the employment relationship. There have been a few studies conducted on the relationship between perceived organizational support and altruism. Shore and Wayne (1993) found empirical support for this relationship in a cross-sectional study of a large, multinational firm headquarted in the United States. Also, a study conducted by Wayne et al., (1997) resulted in a strongly significant relationship between perceived support and altruism. HSa: There will be a positive relationship between perceived organizational support and participation in altruism - organizational citizenship behaviors. Social participation consists of: “. . .noncontroversial behaviors that involve interaction with other individuals, including attending meetings, engaging in positive communications with others, and involvement in other affiliative group activities” (Van 34 Dyne et al., 1994: 780). Using Social Exchange Theory (Blau, 1964), perceptions of organizational support may be related to the employee getting more involved in the day- to-day operations of the organization by volunteering for committee assignments, accepting responsibilities that lie outside of normal working hours, or working extended hours to ensure the success of a major project. Employees who believe that the organization supports them may view these extra-role behaviors as necessary to the survival of the organization. They care about the long-term continuity of the company as a way to ensure job security. Thus, in congruence with Social Exchange arguments, their extra efi‘orts now may create expectations that their employer will provide job security in the future. These employees act as good citizens, highly involved in the running of the organization, participating in its successes and attempting to prolong the life of the company. For example, a restaurant employee may feel responsible for building positive relationships with coworkers in order to provide a fiiendly atmosphere for customers. Employees who do not perceive themselves as supported members of the organization are not highly attached to the organization, and may not care about the welfare of the company beyond the economic benefits it provides to them. They may believe that extra involvement in daily corporate activities is not their responsibility; they each have their own job, and it is the job of the manager to ensure company survival. There has been no empirical research conducted on the relationship between perceived organizational support and social participation. However, Van Dyne et al.,. (1994) found that social participation was associated with employee perceptions of a two- way covenantal employment relationship with their work organization, which they argue is 35 characterized by “...open-ended commitment, mutual trust, and shared values” between the two parties (768). Perceived organizational support can be considered the employee’s perception of the organization’s commitment toward the employee (Eisenberger et al., 1986). Thus, since perceived organizational support has such a close conceptual tie with covenantal relationships, it should be similarly associated with social participation. HSb: There will be a positive relationship between perceived organizational support and participation in social participation - organizational citizenship behaviors. Advocacy participation is defined as: “...behaviors targeted at other members of an organization and reflecting a willingness to be controversial, such as innovation, maintaining high standards, challenging others, and making suggestions for change” (Van Dyne et al., 1994: 780). Employees who perceive themselves to be highly supported by their work organization feel comfortable and confident in their work behaviors. Based on Social Exchange arguments, they have a desire to help improve the organization, acting on the belief that the organization will feel responsible to provide reciprocation to them for their actions. They may expend extra effort to develop innovations in workplace processes, develop ideas for new products or services, or voice their opinions about improvements to their supervisors, even if their opinions may be considered controversial. Their belief that the organization and its manager/agents support them allows them to 36 behave in a somewhat risky fashion, because they have confidence in their abilities, and in the organization’s desire to improve the organization with new ideas; they are not concerned with getting fired due to actions that may be labeled as insubordinate. In a service environment, perceived organizational support may result in better guest service because the employee feels fiee enough to make service improvements without concern about possible replisal for breaking policy. Individuals who do not perceive themselves to be highly supported by the organization may not feel as comfortable expressing their opinions and offering suggestions for improvement, because they have not developed the belief that the organization supports them and their ideas. Controversial opinions, and ‘putting oneself on the line’ with innovative ideas, may create perceptions that their jobs may be threatened if the ideas do not work out as planned. They are not as willing to risk their job or ' reputation, because they do not believe that the employer perceives a responsibility to them in return for their attempt to improve the organization. There have been no studies which specifically examine the relationship between perceived organizational support and advocacy participation. However, Eisenberger et al., (1990) found that perceived organizational support was associated with innovative employee behavior designed to benefit the organization. Since advocacy participation includes making innovative suggestions, this study would seem to support a relationship between perceived organizational support and advocacy participation. 37 H5c: There will be a positive relationship between perceived organizational support and participation in advocacy participation - organizational citizenship behaviors. ' '° ai n 'v r anizaional Functional participation includes: “...participatory contribution in which individuals focus on themselves rather than others in their organizations but yet contribute to organizational effectiveness, including performing additional work activities, self- development, and volunteering for special assignments” (Van Dyne et al., 1994: 783). According to Social Exchange Theory (Blau, 1964), when an employee perceives herself to be highly supported by her work organization, she may have stronger feelings of responsibility to have better than average attendance, arrive on time to work, and perform at a high level. The intrinsic attachment and trust that may accompany high levels of perceived organizational support may create a desire to perform well for one’s employer. Personal attachment to the company may also create a greater sense of personal responsibility, leading to more conscientious behavior by employees. A strong desire to see the organintion succeed, due to perceptions of support, may affect the employee’s behavior by creating an attitude of doing everything possible to help out, including working through breaks, following all rules, and taking extra safety precautions. Employees who perceive themselves to have relatively little organizational support 1 may not experience the sense of attachment and the desire to see the company succeed. Social Exchange theorists would argue that these employees have few expectations of 38 their employer, and expect the empldyer to have few expectations of them. For example, a restaurant hostess may not agree to be cross-trained as a server in order to better understand server needs as they pertain to the hostess’s job. She may resent her employer for attempting to make her put forth more effort, because she does not think the organization should expect this effort from her. She may also believe that the organization holds no responsibility to her in reciprocation for her participation in extra training. There have been no empirical tests conducted on the relationship between perceived organizational support and functional participation. However, as discussed with social participation, Van Dyne et al., (1994) found strong support for an association between covenantal relationships and filnctional participation. Since covenantal relationships and perceived organizational support both stem from the employees’ perceptions of the organization’s commitment towards its employees (Eisenberger et al., 1986; Van Dyne et al., 1994), they should have a similar relationship with possible consequences, such as functional participation. H5d: There will be a positive relationship between perceived organizational support and participation in functional participation - organizational citizenship behaviors. Ill'Vl°i! Loyalty “...represents allegiance to an organization and promotion of its interests” (Van Dyne et al., 1994: 780). Consistent with Social Exchange Theory (Blau, 1964), 39 under perceptions of high organizatibnal support, employees tend to give back support to the organization through recommending the company’s products and services to others. For example, a restaurant employee may choose to frequent the restaurant where he works whenever he and his fiiends go out to dinner. Spending your own money in your place of business is the highest form of loyalty that an employee can give. Also, employees who perceive that they are valued by the organization may be willing to overlook an unexpected, negative situation because they have a firm expectation that it will change in the not-so-distant future. They believe that the organization has not changed its perceptions of responsibility to the employees, and thus the employees do not change their perceptions of responsibility to the organization. Employees who perceive a low level of organizational support may not choose to reciprocate in regard to supporting the organization. They have little attachment to the organization, and thus do not recommend it to others and do not choose to spend their hard-earned money where they work. Also, when less than desirous circumstances occur at their workplace, they may choose to leave the organization far earlier than their highly supported counterparts. They have no motivation to help the company overcome the negative situation, because they do not believe that the organization will perceive a subsequent responsibility to them for helping alleviate the situation. Even under benign conditions, if these employees were given an opportunity outside the organization for better employment, they would probably take it. No empirical research has been conducted on the relationship between perceived organizational support and loyalty. However, using the same logic as discussed with both 40 social participation and firnctional participation, based on the study conducted by Van Dyne et al., (1994), there should be a strong association between perceived support and loyalty to one’s organization (see prior sections). Also, perceived organizational support is viewed as similar to the organization’s commitment to the employee (Eisenberger et al., 1986). Past research has shown a strong, positive relationship between organizational commitment and loyalty (e. g., “fithey and Cooper, 1989). Therefore, since perceived organizational support is somewhat similar to organizational commitment, they should have a similar relationship to loyalty behavior. H5c: There will be a positive relationship between perceived organizational support and participation in loyalty - organizational citizenship behaviors. i n 'v ' i Obedience “...represents respect for the rules and policies of an organization and willingness to expend appropriate effort on its behalf’ (Van Dyne et al., 1994: 780). Perceived organizational support may impact an employee’s willingness to follow company rules and regulations, because the employee who perceives herself to be highly supported by the organization may follow organizational policies more readily than employees who do not perceive high organizational support. In other words, the employee, in choosing to obey company rules and policies beyond the minimum level expected, selects to accept the perceived care offered to her by the organization. Ifthe organization, through its actions, creates a perception in the employee that the employee is 41 a highly valued member, the employee may feel that disobedience in regard to traditional company policies would risk the high levels of support. Similarly, and consistent with Social Exchange Theory (Blau, 1964), extra effort directed toward maintaining company policies and procedures may be taken as a sign of ownership by the employee, a sign that she not only accepts the rules, but supports them wholeheartedly. In a more macro view, this support may be an indication that the employee supports the organization as well. Employees who perceive themselves to have relatively little organizational support may not see the necessity to strictly abide by all organizational policies and procedures. For example, a server may not see the harm in providing a free dessert to a guest in order to receive a higher tip, even though it is against company policy to give food away without management approval. Social Exchange arguments would state that since these employees perceive that the organization does not view them as valued members, the employees may reciprocate by deciding not to support company rules and policies. Far fi'om suggesting that outsiders will delibrately break company policies, I simply argue that employees who perceive themselves to be supported members in the organization will respond with extra effort and attention to organizational policies and procedures. There have been two studies conducted which indirectly examine the relationship between perceived organizational support and obedience. Eisenberger et al., (1990) found a strong relationship between perceived support and conscientiousness in carrying out expected job duties. Conscientiousness in regard to job responsibilities may be considered to fall under obedience because it is appropriate effort on behalf of the organization. Shore and Wayne (1993) examined the relationship between compliance and perceived 42 organizational support, and found that the two were positively correlated. Compliance with company policies and procedures is similar to obedience, and thus they should have similar relationships with perceived support. HSf: There will be a positive relationship between perceived organizational support and participation in obedience - organizational citizenship behaviors. Organization-focused vs. individual-focused extra-role behaviors Organizational citizenship behaviors can be divided into those behaviors focused toward the organization (e.g., obedience, loyalty, firnctional participation, and advocacy participation) and those which target specific individuals (e. g., altruism/helping, social participation; Williams and Anderson, 1991). These two categories of behaviors may be related difl‘erently to perceived organizational support. Based on Social Exchange Theory and the norm of reciprocity (Gouldner, 1960), since perceived organizational support is the employee’s perception about how the organization behaves toward the employee, it should be more closely tied to employee behavior directed at the organization, as opposed to behavior directed at other workers. Employees who perceive themselves to be highly supported by the organization may see engaging in extra-role behavior as a way to reciprocate for the feelings of importance and value that the individual seems to receive from the organization. In our service examples, this means that the employee who perceives high levels of support may be more willing to follow organizational rules and policies, recommend the restaurant to his or her fiiends, and attend extra training sessions 43 than help coworkers with their sideWork (i.e., preparatory work for servicing guests). Therefore, the relationship between perceived organizational support and the various extra-role behaviors discussed in this paper should be stronger for those extra-role behaviors aimed at the organization (e. g., advocacy participation, firnctional participation, loyalty, and obedience), as opposed to those which target individuals within the organintion (e. g., altruism and social participation), due to the very nature of the perceived organizational support. H6: High levels of perceived organizational support will be more strongly associated with organization-focused organizational citizenship behaviors than with interpersonal-focused organizational citizenship behaviors. max Prior work status research has focused on work attitudes, not behaviors. Combined with the common assumption by managers that part-time employees hold different attitudes and thus behave differently than full-time workers while on-the-job, this creates a need to study these possible variations in behavior. Using Social Exchange Theory (Blau, 1964) the current study examines behavioral differences of firll- and part- time employees, focusing on extra-role behavior. Specifically, perceived organizational support, through arguments based on Social Exchange Theory (Blau, 1964), serves as a theoretical basis for explaining why part-time employees may choose to behave differently than full-time workers. Several types of organizational citizenship behavior are used as dependent variables, due to the employee’s freedom of choice regarding whether or not to engage in such actions. More clearly, work status is hypothesized to be related to the employee’s perceived level of organizational support in the work organization, which subsequently impacts the employee’s participation in organizational citizenship behavior. Preferred work status is hypothesized to moderate the relationship between work status and perceived organizational support. Figure 1 summarizes the hypothesized relationships. The next chapter will explain the proposed methods to collect and analyze the data. 45 CHAPTER 4: METHOD Overview of methodology This chapter describes the participants, data collection procedures, measured variables, and data analysis techniques that I used in this study. This research consists of a cross-sectional study design conducted in a field setting. It relies on questionnaires completed by both employees and managers, as well as archival data gathered fiom organizational records. Power analysis As recommended by Cohen and Cohen (1983), I conducted an a priori power analysis to determine the approximate sample size needed to appropriately make conclusions through the detection of significant effects. Past effect sizes in work status correlations! studies fall into a wide range (e. g., .05 to .59), probably related to the specific dependent variable included in the research. Since there are no previous studies relating work status to either perceived organizational support or organizational citizenship behavior, an appropriate estimate for the effect size is conservative, or a low- to-mid-range correlation (i.e., .10 to .20). Based on an alpha of .05 and power level of .80, which are conventional estimates (Cohen and Cohen, 1983), approximately 240 employee participants were needed to detect the expected low-to-mid-level effects (see Appendix C for equations and results). However, due to the ambiguity surrounding the expected effect sizes, I recruited a respondent pool of 350. 46 Participants Participants in the study were employees of several restaurant companies throughout the Midwest. Restaurant employees were used for three main reasons: (1) service employees have been significantly overlooked by organizational behavior researchers as possible subjects, (2) service sector positions make up an increasingly large percentage of total jobs offered in the US. economy (Deuterrnann and Brown, 1978; Nardone, 1986; Nollen and Axel, 1995), and (3) part-time positions are typically found in most service-oriented organizations (Nollen and Martin, 1978). Approximately equal numbers of full- and part-time entry level restaurant employees were used. In order to reduce extraneous variance related to difl‘erences in job level and tenure, only entry-level employees (e.g., servers, bussers, hostesses, cooks) who have worked with the organization at least one month were included in the study. The one month cut-off was chosen because this is the conventional time period allotted for probationary employment, during which each party may end the employment at-will. In addition, this time period allows employee perceptions about the employment relationship to become stable, or more concrete. Procedure The current research was a field study, conducted using restaurant workers from several different companies. Access to these employees was gained through several conversations with the General Managers of specific properties. A total of 10 managers were approached, via phone and site meetings, about the study. Of these, 6 agreed to 47 participate. In group meetings during work hours, I asked employee respondents to fill out a questionnaire containing the perceived organizational support scale (Eisenberger et al., 1986) and self-report measures of work status and preferred work status. In addition, I asked them to answer several demographic questions to help insure a representative sample. I assured subjects that their answers would be kept confidential, and would in no way be identifiable by their employer. Employees completed surveys at short company meetings and returned the questionnaires to the investigator immediately upon completion. Organizational citizenship behaviors were measured through a questionnaire that the subjects’ immediate manager completed. I hand-collected these questionnaires. Work status (i.e., full- vs. part-time) and average number of hours worked per week for each employee were also provided to the researcher by the General Manager of each property. This information came fiom oficial organizational records. Variables Mm The very nature of the firll-time vs. part-time dichotomy stems fiom the differences in the number of hours worked. Within each category of work status, however, there may be significant variation. According to US. Government guidelines (Deuterrnann and Brown, 1978; Nardone, 1986; Owen, 1978), full-time employees may work 4070 hours per week or more, whereas part-time employees may average 1-34 hours per week. For example, one would expect the difi’erence in perceived organizational support to be greater when comparing a person who works an average of 50 hours and one who works an average of 10 hours, than when comparing two people who work 25 and 30 hours, 48 respectively. Since variation within these two categories may be significant, I feel that important information may be gained by using two conceptualizations of work status -- the dichotomous full- and part-time measure and the continuous measure of average number of hours worked per week. By using both measures, I can research whether the number of hours worked on average per week may be the driving force behind full- and part-time difl‘erences, or if there may be other factors contributing to variations in behavior and attitudes, such as nonwork issues (e. g., non-work time commitments). Work status was measured in two ways: through a dichotomous classification of part-time vs. full-time, and through the average number of hours worked per week in general. Participants were asked the following two questions: 1. “What is your current work status (please check one): __ F ull-time _______Part-time ’3 2. “Please list how many hours per week on average you work: Also, company records on work status classification and average number of hours worked are kept on employees, for both legal and company reasons. These measures were compared to the self-report measures to ensure correct work status classification. One company in the study, however, chose to classify all servers as part-time employees, regardless of average number of hours worked per week. In this case, to maintain consistency with the other companies involved in the study, employees were classified as full-time if they worked 35 or more hours per week on average. Post hoc analyses were 49 conducted to determine how this change of classification may have impacted the results of the study, and are examined in the discussion section. r ° i n Perceived organizational support was measured through the survey of perceived organizational support, developed by Eisenberger et al., (1986). The scale consisted of 9 self-report items, with a 5-point Likert-type response selection ranging fiom 1 = Strongly Disagree to 5 = Strongly Agree. This scale is a shortened version of the original survey, and has been used in subsequent research (alpha = .97; Eisenberger et al., 1990). Each respondent was asked to answer the questions using their work company as the referent. W This variable, defined as the employees’ desired work status, was measured by self-report. Each participant was asked to respond to the following questions: 1. “In terms of your overall personal and financial responsibilities, how many hours per week would you prefer to work (from Morrow, McElroy, and Elliott, 1994): 2. “Given your current overall personal situation, which work status category would you prefer (please check one):” Full-time (40 or more hours/week) Part-time (1-34 hours/week) Other (35-39 hours/week) . Q . . l . . l . l l . Managers of the participants were asked to complete the S-factor scale developed by Van Dyne et al., (1994) as the primary measure of organizational citizenship behavior, 50 to rate each respondent on their propensity to engage in extra-role behavior over the last three months. Sub-scales consisted of a lO-item obedience scale, a 7-item loyalty scale, a S-item social participation scale, an 8-item advocacy participation scale, and a 4-item functional participation scale. These items were conceived based on political philosophy and civic citizenship theory, then subsequently construct validated through a several-step process. In addition, managers were asked to complete a S-item altruism scale, which is based on Organ’s concept of organizational citizenship behavior (Podsakoff et al., 1990). Prior research has demonstrated Cronbach’s alphas for each sub-scale range fi'om .68 (social participation) to .88 (altruism and obedience). The Van Dyne et al., (1994) combined five-factor scale has ranged fiom .91 to .95. As discussed previously, I used the Van Dyne et al., (1994) instrument, as opposed to the Organ/Podsakoff fiamework, in this study for several reasons: the items better reflect possible behaviors exhibited by service employees if they were to participate in organizational citizenship; the dimensions are defined more succinctly and are based on citizenship theory; and, the scale items cover a broader range of specific behaviors, which will improve the probability of measuring specific areas of citizenship behavior accurately and validly. In addition, I also used the altruism scale developed by Podsakoff et al., (1990), since it is not redundant with any of the other Van Dyne et al., (1994) dimensions, and has proven to be a popular extra-role behavior measure. This also provides insight into whether or not employees difl‘erentiate between interpersonal- and organizational- focused citizenship behaviors. 51 ngafiates I used four main covariates in the primary study. Tenure, age, race, and gender can create extraneous variance that interferes with the ability of statistical techniques to adequately find expected relationships. More specifically, tenure in the organization may impact the individual’s willingness to become attached to, or get involved in, the organization, which might influence perceptions of organizational support and citizenship behavior. Participants’ tenure was measured through self-report. It was a continuous variable, using number of months employed by the organization. Previous research has found that age, race, and gender are often associated with part-time employment (Deuterrnann and Brown, 1978; Nollen and Martin, 1978). Specifically, part-time employees tend to be either young (i.e. teenagers, students) or older (i.e., retirees), non- white, and women. Coding was as follows: 0 = males and 1 = females, 1 = white and 2 = non-white, and age was measured by number of years. All were self-reports. Table 3 summarizes the source and measurement of these variables. It includes the literature citation for the various measures, the number of items used, and the information source for the variable in the current study. Figure 3 presents an operationalized model, illustrating the variables used to measure each construct. Data Analysis A total of 10 dependent variables are examined in the primary study. Perceived organizational support (POS) was analyzed first, with the two measures of work status as independent variables. Subsequently, each organizational citizenship behavior (OCB) 52 Table 3: Measurement Summary Construct Literature Source # of items Data Source Work Status - FT IPT e.g., Miller and Terborg, 1978 1 Archival data and Self-report Work Status -- Avg. Hrs. Hom, 1978 1 Archival data and Self-report Perceived Org. Support Eisenberger et al., 1986 9 Self-report Preferred Work Status Morrow et al., 1994 2 Self-report PWS - FI’IPT PWS - Avg. Hrs. Organizational Citizenship Behaviors (3 9) Supervisor Altruism Podskoff et al., 1990 5 Social Participation Van Dyne et al., 1994 5 Advocacy Partic. Van Dyne et al., 1994 8 Functional Partic. Van Dyne et al., 1994 4 Loyalty Van Dyne et al., 1994 7 Obedience Van Dyne et al., 1994 10 Demographics (e. g., age, 3 Self-report race, gender) Organizational Tenure 1 Archival data and Self-report 53 85.355 3—884 925$ ices—Sm Eases >oeoo>v< tagger—.233...“ C em— :ocu . 6E3 Eoom Em_E=< race 8.83% moo .252 3559.25 “n PEME tome—3m 3.223590 mi 32088 :w GE , NI 2535 a, -.E mfluwuw u-uOB 38% x83 eotouoi mm _ _ mmam 53m mm? .m? 54 factor was examined as separate dependent variables, with work status and POS as independent variables. Also, I measured OCB as a general construct by combining scale scores, and also as two composites (OCB-O and OCB-I). I tested POS as a mediator of the work status - organizational citizenship behavior relationships. Finally, I conducted a moderated regression analysis, with preferred work status as a possible moderator of the work status - POS relationship. I used hierarchical multiple regression analysis (Cohen and Cohen, 1983) to test most of the relationships among the variables in the primary study. This technique is a “...method for studying the effects and the magnitudes of the efi’ects of more than one independent variable on one dependent variable using the principles of correlation and regression” (Kerlinger, 1986: 527). However, the ability for regression analysis to detect effects may be hindered by violation of important assumptions on which the technique is based. Two possible characteristics of the data set which may afl‘ect the application and interpretation of regression results are multicollinearity and heteroskedasticity. I checked the data prior to testing hypotheses for OLS assumption violations. I discuss this further in the next chapter. For hypothesis 1, the control variables (i.e., age, gender, race, tenure) were entered into the equation first, then the primary independent variable (i.e., work status) followed. Each operationalization of work status (i.e., dichotomous FT/PT vs. number of hours worked per week on average) was used in separate analyses. The dichotomous measure was coded 0 = Part-time and 1 = F ull-time, whereas the average number of hours worked was coded by the number of hours reported, in a continuous measure. Control 55 variables were entered into the equation in the first regression step, so that the first-step regression equation is: POS =flAge, Gender, Race, Tenure); POS = aAge + bGender + dRace + eTenure + c where POS = perceived organizational support, a, b, d, and e = regression coefficients of Age, Gender, Race, and Tenure respectively, and c = constant. Each work status operational variable was entered into the equation in the second step, in separate analyses (i.e., separate regression analyses were conducted for each work status variable). The adjusted equation is: POS =flAge, Gender, Race, Tenure, WS); POS = aAge + bGender + dRace + eTenure + fWS + c where WS = work status, 1' = regression coeficient of work status (standardized beta- weight). Ifthe hypothesis is supported, the regression coefficient for work status should explain variance above that accounted for by the control variables. The hypothesis is supported to the extent that f is positive and significant at a .05 level. This indicates that there is a positive association between work status and perceived organizational support. Hypothesis 2 was tested by adding a third regression step, which allowed inclusion of an interaction term in the above equation. The interaction term was created by multiplying work status by preferred work status. The dichotomous preferred work status variable was coded 0 if subjects wanted to work part-time (under 35 hours per week), and 1 if they wanted to work full-time (35 or more hours per week). Originally, the preferred work status variable had three possible categories, with an ‘Other’ category consisting of 35-39 hours per week. Since the number of subjects who preferred this option was small (i.e., 6), the variable was recoded into a dichotomy for the analyses described above. The continuous preferred work status variable simmy used the number of hours each subject desired to work on average per week. Each measure was examined separately. Both main efl‘ects were entered in the second regression step, with the interaction term entered in the third step. The modified equation is: POS =flAge, Gender, Race, Tenure, WS, PWS, WSxPWS); POS = aAge + bGender +dRace + eTenure + M S + gPWS + hWS‘PWS + c where PWS = preferred work status, g = regression coefiicient for preferred work status, and h = regression coefficient for the work status x preferred work status interaction term. The interpretation of this result was plotted using standardized beta-weights in the overall equation to determine if preferred work status changed the relationship between work status and perceived organizational support. IfAF associated with the interaction term is significant (p < .05), and the slope off is significantly reduced for those individuals working their preferred work status, than the hypothesis can be interpreted as being supported. To test hypothesis 3, nine separate equations were needed in order to test each dependent variable. Organizational citizenship behavior was examined as three composite variables by averaging across sub-scale means to receive an overall mean, an interpersonal-OCR mean, and an organization-OCB mean. Also, each specific type of behavior was examined separately. As in hypothesis 1, control variables were entered into 57 the equation in each first regression 'step, then work status was entered as the primary independent variable in each second step. The equation is: OCB ..g =f(Age, Gender, Race, Tenure, WS); OCB ..r = aAge + bGender + dRace + eTenure + W S + c The hypothesis is supported to the extent that the work status regression coefficient is positive and significant at a .05 level in each of the equations. This indicates that there is a positive association between work status and and the various types of organizational citizenship behavior. To firlly test the mediation hypothesis, each dimension and each composite organizational citizenship behavior variable was tested separately using Baron and Kenny’s (1986) procedure for testing mediation. The analyses were conducted as described for hypotheses 3. Thus, there will be a total of 9 separate mediation analyses, each consisting of three regression analyses. According to Baron and Kenny (1986), mediation is determined based on the significance levels and the relative size of the regression coefl'lcients. More specifically, the independent variable (i.e., work status) should be significantly related to both the mediating variable (i.e., perceived organizational support) and the dependent variables when examined separately. When the regression equation includes both the independent and the mediating variables predicting the dependent variable, the effect of the independent variable on the dependent variable should significantly decrease. Full mediation occurs if there is no relationship between the independent and dependent variables when the mediating variable is included in the equation. Partial mediation occurs if the relationship between the independent and 58 dependent variables decreases, but still remains significant, when the mediating variable is included in the equation. The results of Hypotheses 1 and 3 are the first two steps in testing Hypothesis 4. The third procedure is to determine if perceived organizational support is related to organizational citizenship behavior and subsequently to what extent this variable accounts for the relationship between work status and organizational citizenship. A regression analysis was conducted that included both work status and perceived organizational support in the same equation. Control variables were entered in the first block, with work status and perceived organization support entered in the second block. The equation for this hypothesis is: OCB .4 =j(Age, Gender, Race, Tenure, WS, POS); OCB ..r = aAge + bGender + dRace + eTenure + 1W S + jPOS + c where POS = perceived organizational support andj = regression coeflicient for perceived organizational support. The hypothesis is supported to the extent that H1 and H3 are supported (i.e., work status is significantly related to both perceived organizational support and citizenship behavior) and the work status regression coefficient (1') decreases in strength and significance when perceived organizational support is included in the equation (Baron and Kenny, 1986). Hypotheses Sa-f were tested by entering the three control variables first into the equation, as in Hypothesis 1. Then, in a second regression step, perceived organizational support was entered as the primary independent variable. The equations are: OCB“ =flAge, Gender, Race, Tenure, POS); 59 OCB“ = aAge + bGender + dRace + eTenure + jPOS + c The hypothesis is supported to the extent that the perceived organizational support regression coeficient is positive and significant at a .05 level for each of the separate behaviors. This indicates that there is a positive association between perceived organizational support and the various organizational citizenship behaviors. Testing Hypothesis 6 is slightly difi‘erent than the other hypotheses. First, I calculated two composite scores by averaging respective scale means. OCB-Interpersonal included altruism and social participation, and OCB-Organization encompasses advocacy participation, functional participation, loyalty, and obedience. To test the significance of the difl'erence between dependent correlations, the following equations were used and then compared to Appendix A in Cohen and Cohen, 1983: r = (rxy + ray) / 2 R=1-r2,y-r2c,-l’,c+2r,yrcyrn t= r- - +r a/iirn -1)/(n - 3)] R + 304;)? where r,,y = correlation between perceived organizational support and OCB-I, rey = correlation between perceived organizational support and OCB-O, and rm, = correlation between OCB-1 and OCB-O. In this test, the hypothesis is supported if the 1 score, as calculated above, is determined to be significant, and rey is greater than r,,. This is indicative of perceived organizational support having a stronger relationship to ' organizational-focused organizational citizenship behaviors than interpersonal-focused organizational citizenship behaviors. CHAPTER 5: RESULTS Overview of results I present the results of the study in four parts. First, I summarize the sample through a discussion of descriptive statistics, including demographic variables. Then, I describe the assessment of assumptions regarding ordinary least squares regression to determine the appropriateness of this analysis technique. Next, I introduce the reliability analyses for each of the scale variables used in the study. Finally, I describe individual analyses for each hypothesis in detail. Descriptive statistics The sample for the current study consists of 257 restaurant workers from 6 different organizations. Approximately 350 employees were eligible to participate in the study. However, several were not present during data collection due to scheduling by their supervisors. Also, approximately 20 subjects preferred not to complete surveys. Finally, two individuals terminated their employment during the time of this study, and five individuals were not rated by their supervisors, and thus had to be eliminated fiom the study. These factors resulted in a response rate of 74 %. The sample is 73% white, 75% female, with a mean age of 27. Sixty percent of the sample consists of part-time workers. Subjects have been employed for an average of 32 months with their current employer. Due to the procedures I employed, there was little missing data in the sample. However, two of the questions did prove problematic for supervisors, and were thus eliminated fi'om their individual scales. This is discussed further in the reliability analysis. 61 OLS Assumptions Prior to completing regression analyses, it is important to determine if the data violate the requisite assumptions about data set characteristics. The first investigation tested the normality of the data. Normal probability plots and histograms were developed for all major independent and dependent variables and their residuals, to analyze the shape of the data. The normal probability plots, which plot the observed values against the predicted values for a given variable, indicated that all main variables were approximately normally distributed. The data points for all main variables fell close to the ‘normal’ line, indicating a normal distribution. Scatterplots were also developed to investigate linearity and equality of variance for main variables. Standardized residuals are plotted against their predicted values. If no clear relationship, or observed pattern, between these two variables can be detected in the plot, then it is safe to assume the data is linear and variance is homogeneous. The plots for all continuous variables showed that there was generally no discemable pattern among the data points, indicating randomly distributed residuals. Thus, no curvilinear or other nonlinear firnctions exist in the data. Also, the scatterplots suggest that the data have fairly consistent error terms across values, based on the horizontal banding of the data points around 0. Even in situations where there is minor heteroskedasticity, as may be the case for perceived organizational support, Berry and Feldman (1985) argue that regression is robust enough to overcome possible heteroskedasticity problems in testing for significant relationships. 62 Multicollinearity was examined in two ways. First, the covariance between each predictor variable and the set of other predictors was obtained, to determine the proportion of variance explained by the other IV’s (Berry and Feldman, 1985). This analysis indicated that severe multicollinearity among the variables was not a problem, since the proportion of variance of an IV explained by all other predictors was small (e. g., age/tenure = .27). Second, variable tolerances and eigenvalues were examined in conjunction with the regression analyses. Tolerance values ranged from .63 for some of the control variables (age, tenure) to .97 for the focal variables (POS, Wk Status), and eigenvalues were all similar in size. The presence of multicollinearity is detected when tolerance values are lower than .75 for focal variables and there is a large range of eigenvalues. Thus, in this sample, the effects of the independent variables on the dependent variables in the regression equations could be distinguished and separately examined. Since all required assumptions about the data were satisfactorily met, regression is an appropriate analysis technique to test the hypothesized relationships. Factor analyses I conducted exploratory factor analyses to determine the discriminate validity of the citizenship subscales and composite scales. Principal components analysis was used to extract factors, with an eigenvalue requirement of 1.0. Varimax rotation procedures were used, which allowed the various citizenship factors to covary. Additionally, scree plots were also examined as a secondary determinant of number of factors. 14 factor analyses were required, each eliminating difi‘erent items, to determine the clearest result with the 63 expected factors. After the factor loadings were examined, results support a five-factor solution, including obedience, altruism, advocacy participation, loyalty, and firnctional participation. Social participation items loaded on the advocacy and firnctional participation factors, indicating that any results obtained using this subscale should be interpreted carefully. Reliability analyses I conducted analyses to examine the reliability of scale variables, which were computed by averaging the scores of each item included in the scale. Composite scales were computed by averaging the scores of each subscale included in the composite. Several of the scales showed acceptable reliability levels (>80), and thus remained unchanged when used in the regression analyses. Specifically, the 9-item perceived organizational support scale has an alpha of .86, the 5-item altruism scale has a reliability of .91, and the 8-item advocacy participation scale has an alpha of .85. The 10-item obedience scale also has an alpha of .89. Item-total statistics indicated that this scale could be improved slightly if two of the items were removed. However, since the intact scale is considered reliable at an acceptable level, and the gain in reliability was small (.01), the items were retained to maintain consistency with past research. Three of the analyses suggested that the scales be modified to improve reliability. The reliability analysis on the S-item social participation scale indicated that the original scale had low reliability (alpha = .50). Item-total correlations suggested that one of the items, “Is not involved in outside groups for benefit of organization”, be dropped from the scale. Similarly, this item had several missing responses, probably due to the lack of applicability to the jobs in this particular study. The 4-item scale had an increased reliability (alpha = .60). However, this is still below the recommended .70 of (Nunnally, 1978), indicating any results using this scale should be interpreted with caution. The 7-item loyalty scale also needed refinement according to the analyses. One of the items, “Would not encourage coworkers to invest money in organization” did not pertain to the jobs in this study, and thus resulted in a large amount of missing data. The scale reliability increased fi'om .79 to .82 after this item was removed. Finally, the 4-item firnctional participation scale had a low reliability (alpha of .66), and item-total statistics suggested improvement if one of the items was removed. This item, “Has dificulty cooperating with others on projects”, had a low item-total correlation (.29), and a low 12 (.11). The reliability level increased to .70 after this item was renioved. To assess the multiple perceptions of supervisory ratings of citizenship, dual ratings were obtained in two of the companies. Specifically, employees in these two companies were rated separately by two supervisors, ‘a’ and ‘b’. Supervisors held jobs ranging from General Manager, Dining Room Manager, Shift Supervisor, and Production Manager. Each ‘a’ supervisor had a direct relationship, or consistently worked with, the employees they rated. ‘B’ managers were two or more levels above the employees, or did not regularly work with them. These ratings resulted in two scale variables for each citizenship behavior (i.e., altruism- ‘a’ and altruism- ‘b’) for each employee. These scale variables were then correlated, to test agreement. The agreement between these scales ranged widely, fi'om .251 for the advocacy participation ratings to .722 for the obedience 65 ratings. Due to this large variation in agreement, only the ‘a’ scales were used, since ‘a’ managers worked with the employees on a regular basis and should be more familiar with their on-the-job behavior. While this does not increase the reliability of the ratings, it is more likely that the ‘a’ supervisors were more accurate in their ratings of employees than ‘b’ managers because of their increased familiarity with the employees’ work behaviors. Regressions As discussed in Chapter 4, hierarchical regression was the technique chosen to analyze hypotheses one through five. In all regressions, age, tenure, race, and gender were used as control variables. Table 4 summarizes the means, standard deviations, correlations, and coeficient alphas among the primary variables in the study. W As summarized in Table 5, results did not support a relationship between work status and perceived organizational support. The separate hierarchical regressions, each using either the dichotomous and continuous work status variables, showed that the addition of work status into the equation failed to reach significance (W S- FT/PT: AP = .60, p > .05; WS-Angrs: AP = .00, p > .05). Therefore, full-time and part-time workers do not differ in their perceptions of organizational support. W This hypothesis, which proposed that preferred work status would moderate the work status - perceived organizational support relationship, was also not supported (see Table 6). The moderated hierarchical regressions, using four combinations of dichotomous and continuous variables for both work status and preferred work status, showed that the interaction between actual and preferred work status was not significant when entered into the equation (W S-FT/PT,PWS: AF = .06, p > .05; WS-FT/PT,PHRS: 8§F_ .835": 8956 333.282 .33?— 85. ”9560 8.2 a 8:882... 8.88 . .82. a 8.88% 835.. : .2.... 882.8 as access ooze: 99v». 992.. 998. 998. 93.. 93.- «a. 3.- 992“. 993. 3.- 99mm. #6:... med 108 d— ..8. :2.. :8. :2. 2.. 8.. 8.. 8.- :2. .2. 8.- .2. 8:. 8.... .223 .2 :8. :8. :2. :2. 8. 8. :2. ..2. :2. .2. 8.. .2. 2:. 2.2 Eden .2 99g. 992.. 993. 92. 8.. 8. 993. 2. 99am. 3. ha. 99:. 680. :6 flan—.5 .m— :8. :2.. :2.. ..2. 8. 8. :2. .2. :2. .2. 8. ..2. 28. 8.” 2.28m .2 993.. 99Na. 99hw. 92. No. 8. 99cm. 92. 99:”. 9n~. nc. 99¢N. n56. :6 Emu-52 .m— Anag 99m”. 99*. 99v". 8.. 8.. 2. n? 99:”. 990—. N9. 99:. nnwn. and QGUO .2 A23 :8. :2. 8. 8. ..2. .2. :2. .2. 8. :2. 2.8. 8..” :60 .: 63 :2. 8.. 8.. .2. 8. :2. ..2. 8. :2. :8. :8 @680 .2 G3 8.- 8.- 8.- 8.- 8. 8. 8.- 8. 2.8. 8..” mo.— .8 :2. :2. :2. 8. .2.- 8.. .2. “8...: 2.: 52333.. .8 :8. :2. 8. 2.- 8. .2. N8». :2 22.33.. s :2. 8. :2.. ..2. :2. 22.... 8.2 mini-m3 e 8. 2.- 8. 8. 88. 2... Eek-m3 .n :2. 8.- :8. 82.2 3.2 88o... ... 8.- :2. 89.. 2.. 82.8 .m 8.- 28. 2.— 8.2 .2 88.2 8.2 om< .. S 2 2 a m - o n v m N _ a mauled _ at > 83225 3.5m 5.35....— weeaa Beta—oheu 3. «...-H. 67 3v: 2 8:855.» 8.9.8 . 49v: 3 8:85:32" 8.26.. ...... .22. 882.8 as access 322. Age 998. 998. 998. 998. 99$. “.08 .w~ Asa-v 998. 99—0. 99%. 998. .033..— .5 Aer-V 99mm. 993. 992. 5.5....— .2 Ana-v 992.. 99kb- .=am.>_u< .m— As-v 998. flan—don .3 :3 52:2 .2 0&8 .N— . fimUO .: $800 .2 m2 .e pines-m?— .u EELéBm .5 38323 e Ear-Wm? .n 835,—. .v .8950 .n 85— .N on... .— 2 2 2 2 2 2 «22.5» 3.5.3 83.2.25 beam hue—E...— weeEu enemas—02.5 "v 933—. 68 Table 5: Result Summary for theWork status - Perceived organizational support Regression Analysis A. Using Work status, dichotomous measure: 5152 m B__ETA AF. BE A_R3 1 Age .131 1.42 .023 .023 Gender .022 Race .110 Tenure -.071 2 WS-FT/PT - .034 .603 .024 .001 Overall P: 1.186 Adjusted R2: .004 B. Using Work status, continuous measure: 5_TEP m 1.3.514: AE .3.2 AK 1 Age .129 1.285 .021 .021 Gender .003 Race .1 1 l Tenure -.067 2 WS-Angrs. -.003 .002 .022 .001 Overall F: 1.024 Adjusted R2: .001 “" denotes significance at 95.001, " denotes significance at 115.01, "' denotes significance at p305 69 Table 6: Result Summary for the'Work status - Perceived organizational support Moderated Regression Analysis using Preferred work status A. Using dichotomous Work status and Preferred work status measures 515?; 1! 13.15.14 AE E .433 1 Age .130 1.404 .023 .023 Gender .021 Race . 111 Tenure -.072 2 WS-FI‘IPT .023 1.498 .035 .012 PWS-Ff/PT .125 3 WS x PWS -.020 .056 .036 .001 Overall P: 1.237 Adj.R2: .007 B. Using dichotomous Work status and continuous Preferred work status measures m E BETA AE 1.13 AL: 1 Age .125 .986 .016 .016 Gender 006 Race 080 Tenure -.049 2 WS-I-‘I‘IPT .015 .542 .021 .005 PWS-Preers - 076 3 WS x PretHrs .020 .005 .021 .000 Overall F: .714 Ad; 11’ - 008 C. Using continuous Work status and dichotomous Preferred work status measures SIZE 3 BEA AF. E ARE 1 Age .129 1.276 .022 .022 Gender .000 Race .113 Tenme - .068 2 WS-Ang-Irs. .078 1.730 .036 .015 PWS-FIVPT .152 3 Angrs )1 PWS - .312 1.345 .042 .006 Overall F: 1.422 Adj.R2: .012 D. Using continuous Work status and Preferred work status measures: m 12 1.3518 a I: 4&2 1 Age .123 .904 .015 .015 Gender - .015 Race .081 Tenure - .044 2 WS-Angrs .108 1.329 .027 .011 PWS-Preers -. 154 3 Angrs x Preers .478 1.661 .034 .007 Overall F: 1.137 Adja’: 004 *“ denotes significance at ps.001, " denotes significance at p301, " denotes significance at p$.05 70 AP = .01, p > .05; WS-Ang-Irs,PWS: AF = 1.35, p > .05; WS-Ang-Irs,PHRS: AF = 1.66, p > .05). Therefore, the relationship between work status and perceived organizational support does not change depending upon one’s preferred status. Hypothesis 3. This hypothesis, that there is a positive relationship between work status and organizational citizenship behavior, was partially supported (see Table 7). Three different composite citizenship variables were used as dependent variables: an overall OCB composite comprising all six behaviors; an OCB-O composite, made up of loyalty, obedience, advocacy participation, and functional participation; and an OCB-I composite, which included altruism and social participation. The two sub-composites, OCB-I and OCB-O, allow for a more thorough investigation of the work status - OCB relationship. Past authors have grouped different OCB’s together into one general composite (e.g., Konovsky and Pugh, 1994), which is also done in the current study. However, a better understanding of the construct is gained by following Williams and Anderson’s (1991) argument that citizenship behaviors should be categorized according to the beneficiary of the extra effort (i.e., interpersonal or organizational). Thus, the three composites were used to illustrate the importance of distinguishing between difi‘erent types of citizenship behavior. Results indicate that work status is significantly related to overall OCB (WS-FT/PT: AF = 3.84, p < .05, AR2= .01, R2 = .13;WS-Angrs: AF = 7.25, p < .01, AR2= .03, R2 = .14) and OCB-I (WS-FI‘IPT: AP = 6.09, p < .01, AR2= .02, R2 = .13; WS-Angrs: AF = 12.95, p < .001, AR2= 05,112 = .15). However, work status failed to reach significance when entered into the equation for OCB-O (W S-FT/PT: AF = 1.49, p > .05; WS-Angrs: AP = 2.47, p > .05). To provide a more stringent test of the 71 Table 7: Result Summary for the Work status - OCB Regression Analysis A. Overall-OCB, Work status, dichotomous measure: $1513 I! BEIA AF. .83 AK 1 Age .094 7.792m .115 .115 Gender .101 Race .016 Tenure .238" 2 WS-FT/P’l‘ .120‘ 3.841‘ .128 .014 Overall F: 7.07am Adjusted R2: .110 B. Overall-OCB, Work status, continuous measure: 51E}? 1! BBIA A15 E. AR.’ 1 Age .100 7.532'” .114 .114 Gender .101 Race .014 Tenure .233” 2 WS-Angrs .171“ 7.247” .141 .027 Overall F: 7636'” Adjusted R2: .122 C. Interpaacnal-focused OCB, Work status, dichotomous measure: SEER I! EEIA & 83 AB} 1 Age .130 7.141'” .106 .106 Gender .093 Race --009 Tenure .195“ 2 WS-FI‘IPT .151” 6.085” .128 .022 Overall F: 7.05m Adjusted R’: .11 D. Interpersonal-focused OCB, Work status, continuous measure: $1312 I! BETA AB R_2 A83 1 Age .140 6.875‘” .105 .105 Gender .090 Race -.004 Tenure .186” 2 WS-Angrs .227‘” 12.946‘“ .152 .047 Overall F: 8369'” Adjusted R’: .134 E. Orgmization—focused OCB, Work status, dichotomous measure: STEP 1! BETA All E 433 1 Age .047 7.266‘” .108 .108 Gender .102 Race -.015 Tenure .260‘” 2 WS-FI‘IPT .075 1.486 . 1 13 .005 OverallF: 6.121m Adjusted R’: .095 F. Organinticn-focused OCB, Work status, continuous measure: SEE I! BETA Ali R_’. A83 1 Age .049 7.088‘” .108 .108 Gender .103 Race -.015 Tenure .259‘” 2 WS-Angrs .101 2.474 .1 17 .009 Overall F: 6201'“ Adjusted F3: .099 W" denotes significance at 15.001, ‘3 denotes significance at p501, " denotes significance at p$.05 72 difi‘erence in strength of significance 'of the correlations between work status and both OCBO (r = .10) and OCBI (r = .22), the procedure described for hypothesis 6 was used. Significant results (p < .05) indicate that the strength of the relationship between work status and interpersonal citizenship was much stronger than that between work status and organization-focused citizenship. Positive betas thus demonstrate that firll-time employees exhibit more interpersonal and overall citizenship behavior than part-time workers. The fact that OCB-I is and OCB-O is not related to work status indicates the importance of distinguishing among different types of citizenship behavior. Thus, separate regressions were run using the six specific types of OCB as dependent variables (see Table 8). Results show that work status is a significant predictor of altruism (W S-FT/PT: AF = 5.15, p < .05, AR2= .02, R2 = .11; WS-Angrs: AF = 9.93, p <01, AR2= 04,13.2 = .12), functional participation (WS-FT/PT: AP = 10.14, p < .01, AR2= .04, R2 = .09; WS- Angrs: AF = 10.26, p < .01, AR2 = .04, R2 = .09), and social participation (WS-FT/PT: AF = 4.34, p < .05, AR2= .02, R2 = .06; WS-Ang-Irs: AF = 10.29, p < .01, AR2= 04,123 = .09). Advocacy participation had mixed significance with the two work status measures (WS-FT/PT: AF = 1.73, p > .05; WS-Angrs: AF = 5.44, p < .05, AR2= .02, R2 = .06). The results for loyalty (WS-FT/PT: AF = .004, p > .05; WS-Angrs: AP = .07, p > .05), and obedience (WS-FT/PT: AP = .38, p > .05; WS-Angrs: AF = .02, p > .05), showed that no significant link with work status exists. For the four types of behavior that were related to work status, positive betas mdicated that full-time employees were more likely to exhibit altruism, and social, advocacy, and functional participation. 73 Table 8: Result Summary for the Work Status - Individual OCB Regression Analysis A. Altruism ' $1512 I! BEIA & Ii AR”. 1 Age .094 6.888‘“ .105 .105 Gender .065 Race -.002 Tenure .242” 2 WS-Angrs .200" 9.930" .141 .036 Overall F: 7.706‘” Adjusted R’: .123 B. Advocacy Participation: $113.2 1! EEIA ti 82 A12 1 Age .077 3.579" .057 .057 Gender .005 Race -.034 Tenure .184‘ 2 WS-Angrs .153' 5.443. .079 .021 Overall F: 4.006” Adjusted R1: .059 C. Functional Participation: 5132 I! BEIA .4: 8’ fl 1 Age 041 4.356" 069 069 Gender .099 Race .102 Tenure .162‘ 2 ws-Angrs .207" 10.259" .109 .039 Overall F: 5.675‘” Adjusted R2: .039 D Loyalty. SIZE I! ma AB 83 ARE 1 Age -.018 6612"" .101 .101 Gender .068 Race .109 Tenure 276‘” 2 WS-Angrs -.017 0.067 .101 .000 Overall F: 5282'” Adjusted R’: .082 E. Obedience: 5m 1! BEIA AE E AB: 1 Age 105 11.96... 132 132 Gender .129. Race .105 Tenure .211" 2 WS-Ang-Irs .075 0.016 .132 .000 Overall F: 7.141". Adjusted R2: .114 F SocialParticipatlon. STEP. I! am 41: B.’ ARE 1 Age .169‘ 4.344" 069 .079 Gender .101 Race - 005 Tenure .073 2 WS-Angrs 208” 10.292" .108 .039 Overall F: 5671‘” Adjusted R3: .039 “” denotes sipificance at ps.001, ’" denotes significance at p501, " denotes significance at ps.05 74 W This hypothesis, which argues that perceived organizational support will mediate the relationship between work status and organizational citizenship behavior, was not supported. Since there was not a significant relationship between work status and perceived organizational support, as discussed in Hypothesis 1 above (see Table 5), there cannot be a mediation process among the variables. mm There was significant support for a positive relationship between perceived organizational support and all types of organizational citizenship behavior. The three composite variables discussed above (i.e., OCB-overall, OCB-Interpersonal, and OCB-Organizational), as well as each specific type of citizenship behavior (i.e., altruism, social participation, advocacy participation, functional participation, obedience, and loyalty), were used as dependent variables in separate regression analyses. Sources for the independent variables of perceived support and citizenship behavior were the subject and the supervisor, respectively. Results are summarized in Table 9. Specifically, POS explained 3% (OCB-I) to 5% (OCB-O, OCB-Overall) of the variance in the composite citizenship variables (OCB-I: AF = 8.48, p s .001; OCB-O: AF = 15.43, p S .001; OCB- Overall: AF = 12.69, p s .001). Additionally, POS was responsible for between 2% (e. g., advocacy participation, altnlism) and 6% (e.g., loyalty) of the variance of individual citizenship behaviors (altruism: AF = 6.40, p s .01, AR2= .02, R2 = .11; social participation: AF = 6.94, p s .01, AR2= .03, R2 = .08; advocacy participation: AF = 4.79, p s .05, AR2= .02, R2 = .06; functional participation: AF = 9.29, p s .01, AR2= .04, R2 = .08; loyalty: AF = 17.29, p s .001, AR2= .06, R2 = .15; obedience: AF = 9.98, p .<. .001, AR2= .04, R2 = .15). Thus, findings in this study are consistent with past research on the relationship between perceived organizational support and organizational citizenship. 75 Table 9: Results Summary of POS-OCB Regression Analyses A. Altruism 511512 11.! BEIA AB 83 £3 1 Age .085 7.266‘” .108 .108 Gender .072 Race -.008 Tenure .248‘” 2 POS .154” 6.402" .131 .023 Overall F: 7.224s” Adjusted R’: .113 B SocralFarticipatron 81312 I! BEIA & 83 A83 1 Age .160- 4.344" .067 067 Gender .096 Race -.009 Tenure .084 2 POS .164” 6.939" .093 .026 Overall F: 4.949“: Adjusted R1: .075 C Advocacy Participation. I! BETA AB 133 All2 1 Age .059 3.541" 056 .056 Gender -.013 Race - 037 Tenure 198” 2 POS .138‘ 4 792: 074, 018 Overall F: 3.836“ Adjusted R2 055 D Functional Participation: $117.13 I! BEIA AE Ii ARE 1 Age 037 3.967” 062 062 Gender .115 Race -.038 Tenure .168‘ 2 POS .189” 9.286” .097 .035 Overall F: 5140-” Adjusted R2 078 E. Loyalty. SE2 1! BETA AB 83 A83 1 Age -.022 6.961‘” .104 .104 Gender .077 Race 104 Tenure .280‘” 2 P08 .248‘” 17.288m .164 .060 Overall F: 9.402": Adjusted R2 146 F. Obedience: 5m 1! BEIA A}: 83 £3 1 Age .110 9.038m .130 .130 Gender .126‘ Race .110 Tenure .204" 2 POS .189” 9.983... .165 .035 Overall F: 9.497m Adjusted R2 148 76 Table 9: Results Summary of POS-OCB Regression Analyses (con’t) G. OCB-Interpersonal: BIEE I! BEIA Q B? E 1 Age .132 7.150m .106 .106 Gender .092 Race -.009 Tenure .196” 2 P08 .177“ 3.478... .137 .030 Overall F: 7593'" Adjusted R2: .119 H OCB-Organizational BIEE I! BEIA AI: R.” AK 1 Age .048 7.252m 108 108 Gender 101 Race -015 Tenure .260‘” 2 P08 .234”: 15.434... .162 .054 Overall F: 9237'” Adjusted R2: .144 L OCB—Overall: BIEE I! BETA & IL” AR.’ 1 Age .095 7.791“: .115 .115 Gender .100 Race .016 Tenure .238” 2 ms .213". 12.693... .160 .045 Overall F: 9.075": AdjustedR’: .142 “" denotes significance at p$.001, “ denotes significance at p501, "' denotes significance at p505 W This hypothesis was tested using equations from Cohen and Cohen (1983), as discussed previously in Chapter 4. Results indicate that the strength of the relationship between perceived organizational support and citizenship behaviors that are organizationally-focused is significantly stronger than that between perceived support and interpersonal citizenship behaviors. The complete equations and computations are: r” = correlation between POS and OCBI = .18 r.,y = correlation between POS and OCBO = .24 rm, = correlation between OCBI and OCBO = .85 r = (.18 + .24)/2 = .21 R = l - .0324 - .0576 - .7225 + 2(.03672) = .26094 t = [(.24 - .18)~st6 : 1.85 ] / 42—[(256)/(254)j * .26094 + .0441 * 00375 = (.06 * 21.762352) / tl .5259892 + .0001653‘ = 1.3057512 / .725365 = 1.8001298 (significant at p<.05) 78 CHAPTER 6: DISCUSSION AND FUTURE RESEARCH Overview of discussion The discussion has four sections. First, I discuss the importance of the results of the study, addressing each hypothesis in order. Next, I include both practical and theoretical implications of the study. Then, I present the strengths and weaknesses related to the present research. Finally, I suggest future research areas which would build on the findings of this dissertation. Discussion of results Based primarily on arguments made by Pfefi‘er and Baron (1988), I hypothesized that filll-time employees would perceive more organizational support than part-time employees (HI). Results of the study, however, did not support this. In this sample, there is no difi'erence in the level of perceived organizational support experienced by full-time and part-time workers. This may be due to the fact that part-time and full-time servers have worked side-by-side in the industry for several years, earning the same wage and tips, performing the same job duties, and being treated the same way by managers and supervisors. In other words, if there are no basic difi‘erences in important work characteristics like management-employee relationships, there is no reason for variation in perceptions about the organization and the way it treats employees. Another reason that part-time employees may be treated the same as lull-time workers in service industries is because the service sector is dependent on part-time employees to help them meet their labor requirements. Specifically, there are many part- 79 time workers in restaurant organizations, creating a situation where the number of part- time and full-time workers are more equal than if the industry or business did not depend on part-time employees. This may create a perception that, since the company needs these part-timers to achieve service objectives, part-time employees perceive that they are treated well when compared to full-time workers. In new industries, or industries which are just beginning to utilize part-time employment, however, part-time workers may not have these perceptions of fair and equal treatment in relation to lull-timers. Future research may want to investigate possible differences in perceived support between part- time and full-time employees in economic segments that are not dependent on part-time workers to meet their organizational labor needs. A third reason that part-time and full-time workers did not differ in regard to perceived organizational support is that promotions and developmental activities, which are two antecedents to perceived organizational support (Wayne et al., 1997), are usually unavailable to both full- and part-time servers. In this particular sample, which consisted of young workers who may be college students, the employees may not be interested in promotions, and therefore would not view a lack of this opportunity as a sign that the organization does not support them. Specifically, promotions and developmental opportunities may not be a factor in these employees’ perceptions of organizational support, and thus it seems logical that there were no differences among full- and part-time workers in regard to perceived organizational support. Organizational culture and other company characteristics may also impact on the ability to predict perceived organizational support. A post hoc analysis was conducted to 80 determine if POS differed by company. Five dummy variables, each representing one company, were created. Perceived organizational support was then regressed on first the control variables, and then on all five dummy variables in a second hierarchical block. The results showed that there were differences in POS attributable to company effects (AP = 2.52, p < .05, R2 = .04). Subsequently, regression analyses to test hypotheses were again conducted, this time including the dummy variables as control variables in the first step. Analyses indicated no change from the previously reported results. Both work status variables were still unrelated to perceived organizational support. Also, moderated regressions were used to examine if the work status - pos relationship may be significant in certain companies. However, results showed that interaction terms were not significant. Therefore, despite a relationship between company and perceived support, company effects do not appear to help explain why the work status - perceived organizational support relationship was insignificant. Hypothesis two posited that preferred work status would moderate the relationship between actual work status and perceived organizational support; however, no support was found for the proposition in this study. This could be due to the fact that, with the current sample, the majority of employees were working at or near their desired number of hours per week (69%). This afi‘ects the results in two possible ways. First, it makes it difficult to get accurate results for the moderation test, since approximately equal cell sizes could not be obtained. However, Cohen and Cohen (1983) argue that “The conditions of unequal cell fiequencies is a nuisance in the analysis of contrasts in ANOVA, but is automatically handled in (multiple regression analysis)” (208). Therefore, the results of 81 the study should be viewed as representative for the sample used. Second, if the majority of individuals were working close to their desired hours, it makes the preferred work status hypothesis a moot point. Specifically, managers of restaurant workers may already understand the benefits of granting their workers their desired work status. In post hoc conversations with managers involved in the study, several of them expressed the efi‘ort they make to accommodate employees’ nonwork schedules. Since the hypothesis was primarily focused on the fact the part-time workers would perceive more organizational support under conditions of working their preferred status, creating less of a difl‘erence between part-time and filll-time workers, and the part-time workers in the sample were already working their preferred status, then there is no need to argue that working one’s preferred work status may make a difference in the way an employee perceives their work organization. The relationship between work status and organizational citizenship behavior was the topic of hypothesis three. This hypothesis received partial support. Full-time employees exhibit more interpersonal-focused citizenship behaviors (i.e., altruism and social participation), than part-time workers. Also, filll-timers were more likely to participate in advocacy and functional extra-role tasks. However, there is no difference among the two work status groups in regard to their loyalty and obedience towards the organization, and thus the OCB-O composite also showed no significant relationship with work status. This lack of variation between the two work status groups makes sense if we use Social Exchange Theory (Blau, 1964). According to Social Exchange, since there is no 82 difference in the level of organizational support perceived by full- and part-time workers, there should be little difi‘erence in the level of loyalty and obedience to the organization from these two groups of employees. For example, two waiters - one part-time, the other full-time -- who work every Wednesday together may both decide to recommend their restaurant to their fiiends for dinner parties. The part-time waiter feels supported by the organization for employing her while she finishes her college degree. The full-time employee behaves that the organization cares about his well-being because it held his job for him while his broken leg healed. Thus, their recommendation is viewed as a way to thank the organization for its commitment to each of them. Therefore, if loyalty and obedience citizenship behavior is performed to reciprocate for support granted to the employee by the organization, then both part-time and filll-time employees can be expected to reciprocate in the same degree, if they perceive the same level of support. Interpersonal-focused citizenship behaviors, however, are directed toward the employees’ coworkers. Any reciprocation in an interpersonal situation should depend on the relationship between the individuals involved in the exchange. Since part-time employees tend to have higher levels of turnover intentions (Peters et al., 1981), they may hesitate to invest time and extra efi‘ort in developing relationships with coworkers (e. g., helping) because they view their employment in the organization as temporary. This lack of attachmentto, or involvement with, coworkers leads part-time employees to perform less interpersonal-citizenship behaviors, because they do not become parties to social exchanges like full-time workers who become attached to one another. For example, if we return to our two waiters above, although they both want to help the organization, neither 83 has any interest in helping the other.‘ The two employees do not know each other well, because she only works three shifts a week, and only one with the filll-time worker. This lack of involvement creates a barrier to social exchange, since neither party feels obligated to the other. This is evidenced by the part-time employee not helping the full-time worker by covering his shifts while his leg was healing. Another consideration is that full-time workers tend to be present at work a great deal more than part-time employees, which may impact on the extent to which part-time workers feel comfortable interacting with their coworkers. The part-time employee’s restricted presence at work may cause her to be considered a social outsider, thus preventing the part-timer fiom forming attachments to coworkers that would enable her to be a part of possible social exchanges. Also, this may impact on the part-time employee’s comfort level in voicing her opinion to her coworkers, resulting in less advocacy participation than her full-time counterparts. Finally, part-time workers may have non-work responsibilities that limit their ability to perform extra-role behaviors. Specifically, part-time employees may have constraints on their time at work due to family obligations (i.e., raising children, school). This may interfere with their ability to stay late for meetings or attend training seminars held on weekends, both which are considered functional citizenship participation. A post hoc analysis was conduCted to explore the possibility of efi‘ects based on company characteristics. One such characteristic that may impact the work status - OCB relationship is organizational size. Two dummy variables were created to represent the three categories of organizational size: under 30 employees (three organizations), 30-50 employees (two organizations), and “more than 50 employees (one organization). First, to determine if organizational size was related to citizenship behavior, the three OCB composite variables were regressed on the dummy variables, alter inclusion of the control variables. Work status was not included in this analysis. Organizational size was significantly related to all three composites (OCB-overall: AF = 5.99, p < .001, R2 = .16; OCBI: AF = 8.00, p < .001, R2 = .16; OCBO: AF = 4.84, p < .01, R2 = .17). Next, moderated regressions were completed to determine if company size impacted the work status - citizenship behavior relationship. Two interaction terms were created to account for the interaction between average hours worked per week and the two dummy variables representing size. The control variables were entered in the first step, with the WS- Avghrs and organizational size variables entered in the second block. The final block consisted of the interaction variables. In regard to overall-OCB, when the interaction terms were included in the equation (AF = 4.01, p < .05, R2 = .18), the efi‘ect of work status on overall citizenship became insignificant. In other words, the combination of company size and hours worked cancelled the main efi‘ect of work status on involvement in overall citizenship behavior. A moderated regression was also conducted for the OCB-I and OCB-O composite variables, in an attempt to filrther understand the relationships among these constructs. Again, when the interaction terms were included in the OCB-I equation (AF = 3 .32, p < .05, R2 = .18), the efi‘ect of work status on interpersonal-citizenship disappeared. The same pattern of results also appeared for the OCB-O composite (AF = 4.00, p < .05, R2 = .17). However, prior results indicated no direct effect for average hours on OCBO (see 85 table 7). Thus, when organizational 'size is included in the equation as a control variable, the relationship between work status and organizational-focused citizenship behaviors becomes significant (beta = .13, p < .05). Graphs of the interaction effects on the three composite variables are included in Figure 4. As represented, in small organizations work status makes a difi‘erence in citizenship behavior. However, in medium and large organizations, there is no relationship between these two constructs. Future research should examine the relationship between organizational size and citizenship participation more thoroughly. Hypothesis four addressed the possible mediation of the work status - organizational citizenship behavior relationship by perceived organizational support. Since a significant relationship was not found between work status and organizational support, l. Overall-OCB: Highl Small Companies Overall / OCB / Med, Lg. Companies Low High Avg. Hrs. Worked 2. OCBI: Highl / Small Companies OCBI / Med, Lg. Companies Low High Avg. Hrs. Worked 3. OCBO: Highl / Small Companies OCBO / Med, Lg. Companies Low High Avg. Hrs. Worked Figure 4: Organizational Size effects on the Work status- Organizational citizenship behavior relationship 87 mediation cannot be considered and tested. However, other samples of part-time and fill]- tirne workers may result in differences between these two groups in regard to perceived support. For example, part-time and full-time nurses may be treated differently by their their work organizations, especially in benefits packages. Part-time employees tend to receive fewer medical and retirement benefit packages than full-time workers in order to save the company money on benefit costs (Pfefl‘er & Baron, 1988). This may act as a signal for lack of concern on the part of the organization for the employees’ well-being, and thus give the perception that the organization does not support the worker. Given this situation, perceived organizational support may indeed mediate the work status- organizational citizenship relationship in other samples. Hypothesis five and each of the sub-hypotheses presented the relationship between perceived organizational support and each type of citizenship behavior, including the three composites (overall OCB, OCB-I, OCB-O). As suggested, there were significant relationships between these constructs. Regardless of work status, perceiving support from one’s work organization is directly related to the level of engagement in all kinds of citizenship behaviors, as rated by their supervisors. This is consistent with Social Exchange Theory (Blau, 1964) and the Norm of Reciprocity (Gouldner, 1960). In a social exchange relationship, such as employment, an employee will perform many different types of extra-role behavior in reciprocation for the support given by the work organization. As described above, to reciprocate for the organization’s commitment toward them, restaurant employees may attend non-mandatory meetings on Saturday mornings for training or brain-storming purposes. They may also work extra shifts if the restaurant is very busy, or work as a host or bartender when the company is short-staffed Hypothesis six examined the strength of the relationship between perceived organizational support and both interpersonal-citizenship and organization-focused citizenship. As suggested by the trend in the data, organizational support is more strongly related to citizenship behavior that is focused on the organization (correlations are .244 vs. .178, p < .01) as opposed to behavior focused on coworkers and peers. Thus, as expected fi'om Social Exchange (Blau, 1964) and Norm of Reciprocity (Gouldner, 1960) arguments, employees are more likely to perform extra-role behaviors that will directly benefit the organization in reciprocation for perceived organizational support than those behaviors that may indirectly impact, or have no impact on, organizational goals. Research and practical implications The current study has several implications for both researchers and practitioners. First, I discuss the impact on the existing body of literature in three main areas: work status, perceived organizational support, and organizational citizenship. I then address the practical implications of the study, and how managers may apply this knowledge to . increase citizenship behavior among their employees. For work status researchers, this investigation is important for three main reasons. First, it explores an important work attitude (i.e., perceived organizational support), which was shown not to difl‘er among individuals working part- or full-time. Second, in reference to many calls for behavioral research (F eldman, 1990), it provides evidence of 89 work behavioral difi‘erences among full- and part-time employees. Third, it uses Social Exchange Theory (Blau, 1964) to explain these differences through theoretical argumentation and hypothesis testing. Thus, unlike the majority of prior research efforts in this area, this study does not rely on purely descriptive data. Finally, this study illustrates the importance of conducting a more fine-grained analysis by using a continuous variable (1.e., average number of hours worked per week) as a measure of work status instead of a simple dichotomy (i.e., part- vs. filll-time). Results were consistently stronger for the average hours variable, as evidenced by the level of significance of relationships between work status and the three OCB composite variables. Specifically, when the dichotomous variable was used, relationships were significant at either the p < .05 level (Overall OCB), or the p < .01 level (OCB-I and OCB-O). However, using the continuous variable, the relationships become significant at p < .01 and p < .001, respectively. Also, the level of variance in the dependent variables (i.e., OCB composites) explained by the independent variables are consistently higher across dependent variables when using the continuous variable (e.g., .12 vs. .15). These results are not surprising, since the increased variance of the continuous variable over the dichotomous variable will naturally make the continuous variable a stronger predictor. However, the current study indicates that prior results based on a dichotomous work status variable may be misleading, and thus past studies should be re—examined with the more powerful continuous predictor. Another concern with using a dichotomous measure of work status is which classification to use, the employee’s which is based on hours worked, or the official company classification. In this sample, one company automatically classified all servers as being part-time, regardless of hours worked, because over the span of one year the hours per week would average out to be under 35. However, at the time of the survey, and for 3 months prior, most of the servers in this company had been working an hourly average that would classify them as lull-time according to US Government standards. To maintain consistency across companies, I used the US Government guidelines in classifying all workers. 1 then conducted post hoc analysis to examine the impact of using company classification guidelines on the relationships between the constructs in the study. Not surprisingly, the work status - citizenship behavior relationships were not significant when the new classification was used. However, since the company classification was fairly unusual (i.e., not many service companies use this type of automatic classification), and was very difi'erent from the average hours worked variable, my recommendation would be to use the Government classification (i.e., under 35 hours = part-time) in future research. Researchers in the area of perceived organimtional support may value this study for two main reasons. First, it provides evidence that no differences in support exist among members of difi‘erent work status groups in this particular sample. This would suggest that other antecedents found in past research, specifically promotions and developmental activities (Wayne et al., 1997), may only apply in certain work organizations or job levels, such as manufacturing companies and mid-level management positions. Typically, in restaurant companies, line employees like servers and bartenders are seldom promoted to management positions for a variety of reasons (e. g., students only 91 working temporarily). Thus, regardless of work status, restaurant workers generally do not expect to be promoted and therefore do not use promotions as a signal of support. Another reason that organizational support researchers may value the current study is that it replicates and furthers prior research concerning the relationship between perceived organizational support and citizenship behaviors. Previously, a link between altruism and perceived organizational support was found by Shore and Wayne (1993) and Wayne et al., (1997). Also, Eisenberger et al., (1990) found a relationship between perceived support and increased willingness to make innovative suggestions, which is a part of advocacy participation. Both of these prior findings were supported in the current study. This replication of findings is important because the sample used in this study (i.e., restaurant workers) is different than that used in the three prior investigations (i.e., salaried, managerial employees). Thus, there is evidence that the relationships between perceived support and both altruism and advocacy participation generalize to a wide number of employees in various types of organizations. Several other types of citizenship behaviors were also examined (i.e., loyalty, obedience, social participation, and functional participation) and shown to be related to perceived organizational support. Since there is no previous research on the link between these behaviors and perceived support, I cannot assume that these relationships would generalize to other employee populations. However, loyalty and obedience exhibited the strongest relationships with perceived organizational support of all the behaviors. Therefore, since the perceived support-altruism and -advocacy participation relationships 92 appear to generalize across employee populations, it is logical that the relationships between POS and both loyalty and obedience should also generalize. The difi'erence in strength of these various relationships is important because it indicates that employees seem to distinguish among citizenship behaviors in which they choose to engage, as illustrated in the results of hypothesis 6. A post hoc analysis was conducted to determine if there were statistically significant differences in the strength of the relationships between POS and both loyalty (an OCBO) and altruism (an OCBI). These two constructs are conceptually similar, in that they both represent behavior which benefits others (Van Dyne et al., 1994). However, loyalty and altnrism difi‘er in their beneficiaries, with loyalty behaviors designed to benefit the company and altruism focused on benefitting coworkers. Using the procedures described for hypothesis 6, 1 determined that POS was significantly more strongly linked with loyalty (r = .27) than with altruism (r = .15). This result makes it clear that employees distinguish between the kind of extra- role behaviors they are willing to exhibit at work, especially when engaging in the behaviors is a way to reciprocate for support granted to the employee by the organization. Norm of Reciprocity (Gouldner, 1960) arguments can be used to explain why restaurant workers in this sample are more likely to engage in both loyalty and obedience type of behaviors than other types of organization-focused citizenship behavior (e.g., advocacy and functional participation) in reciprocation for perceived support from their work organization For example, servers may recommend their work establishment to their fiiends to generate more business for the organization in return for the organization’s perceived commitment to the employee (i.e., a loyalty behavior). However, they may not 93 be as willing to voice controversial Opinions or suggestions (i.e., an advocacy behavior), because they may view this as being disloyal to a company that seems to care about them. Future research should examine possible trade-ofi‘s for the employee for participating in the specific types of citizenship, or to determine if the employee conducts a cost/benefit analysis of engaging in certain citizenship behaviors. Researchers investigating organizational citizenship should find this study important for two main reasons. First, the finding that citizenship behavior varies among individuals who are members of different work status groups suggests that prior results may not apply to all types of employees. For example, the commonly accepted link between organizational commitment and citizenship (e.g., O’Reilly and Chatman, 1986) may vary for part-time workers. Specifically, part-time workers have a higher intent to turnover than filll-time employees (Feldman and Doerpinghaus, 1992; Peters et al., 1981), which should correspond to lower levels of organizational commitment and lower levels of citizenship behavior. However, the current study suggests that part-time employees exhibit as much loyalty and obedience behavior as full-time workers. Thus, it would appear that the organizational commitment - citizenship relationship needs to be examined filrther for non-traditional types of workers (e.g., part-time, temporary). Second, the importance of classifying citizenship behaviors as either interpersonal- focused or organization-focused (Williams and Anderson, 1991) is illustrated in the current study through the findings regarding relationships between work status - organizational citizenship and perceived organizational support - citizenship behavior. In fact, there was a significant difference in the level of interpersonal-focused citizenship 94 behavior, as well as advocacy and functional participation, exhibited by full-time and part- time workers, but no difference in the level of loyalty and obedience. Also, all workers, regardless of work status, seem to reciprocate for organizational support through organizational-focused citizenship behavior rather than interpersonal-focused behavior. This would suggest that a re-examination of past findings is needed to determine if prior relationships are supported for all types of citizenship behavior. For example, several studies have examined antecedents to. citizenship behavior (e. g., Organ and Ryan, 1995; Smith et al., 1983). The findings of the current study, however, would suggest that the different types of citizenship behavior may have specific antecendents associated with them. In other words, until more research is completed that distinguishes between the different types of citizenship behavior, perhaps based on the beneficiary of that behavior (i.e., organization vs. interpersonal), we cannot understand the intricacies associated with these relationships. According to the results of this study, there are two primary implications for managing part-time employees. First, managers should be aware that part-time employees tend to participate in less interpersonal-focused citizenship behavior, as well as firnctional and advocacy participation. This means that they may be less helpful toward their coworkers, less socially-focused in general, less able to attend off-schedule meetings and training, and less willing to voice their opinions than full-time workers while on-the-job. If these behaviors are considered important to achieving company goals, such as a family- like culture or providing good customer service, then managers may want to rethink their hiring strategies and refrain from using part-time employees for certain jobs where OCBI 95 is important to cost or productivity factors. However, some cost considerations may outweigh other company objectives. Therefore, the manager should prioritize goals according to the organizational mission, and then make an informed decision concerning the degree to which they use a part-time workforce in the company. The second practical consideration for managers pertains to perceived support. As suggested by the research, high levels of perceived support correspond to increased effort toward all citizenship behaviors, but especially those that will directly benefit the organization, for both full-time and part-time workers. In other words, both of these types of employees are more willing to put forth extra efi‘ort in performing tasks that lie outside traditional job descriptions if the organization seems to care about worker well- being and informs employees that their contributions are acknowledged and valued. Thus, a company that develops and fosters a supportive culture will get more work from their employees, regardless of work status, without a corresponding increase in labor cost. Specific ways to increase perceived support through valuing employee contributions include giving timely, positive feedback about desired work behavior and rewarding desired behavior through appointments to special projects or through employee-of-the- month programs. Care about employee well-being may be signaled through the institution of job sharing or other non-traditional work programs for working parents, the inclusion of on-premises child care and fitness facilities, and support through benefits for health related problems such as cancer or other chronic or long-term illnesses. Strengths and weaknesses The current study has several strengths, as well as weaknesses that need to be addressed. There are three main strengths: basing the investigation on a widely-known theory, using a seldom-investigated sample, and examining and distinguishing between a variety of citizenship behaviors. Social Exchange Theory (Blau, 1964) has been used in numerous studies to provide theoretical argumentation used in hypotheses. It is important in the current research because few prior investigations of part-time and full-time worker differences adopted a theoretical, hypothesis-testing format. Past research is primarily descriptive in nature (Logan et al., 1973; Nkomo and Fields, 1994). Social Exchange has been suggested post hoc by a number of researchers (e.g., Miller and Terborg, 1979) to explain their findings, but this is the first application of this theory in the part-time literature. In addition to providing insight into possible differences regarding the employment relationships between part-time and full-time workers and their employing organizations, it also allows researchers to understand why these differences occur and why they may lead to certain behaviors while at work. The second strength of the current study is the use of service sector employees, namely restaurant workers, instead of the traditional manufacturing or oflice employees. Organizational behavior researchers have seldom examined service employees, although approximately 20 % of the US. workforce consists of these workers (N ollen and Axel, 1995). Instead, many individuals assume that all prior findings may generalize to all types of workers in all industries. However, there may be significant differences among service sector and manufacturing employees, due to basic distinctions in job tasks. Service sector 97 jobs are more intangible in nature, because they include the ambiguity of working with people (Sasser, Olsen, and Wyckofl‘, 1978). In other words, each service encounter may be different due to the unpredictable interaction between the employee and the customer (Fuchs, 1985). Since manufacturing and service-sector jobs are so different, research findings pertaining to jobs and organizational cultures in these industries may not be generalizable from manufacturing to service. Finally, the third main strength of the study is the examination of several distinct types of citizenship behavior, and the classification of those types as interpersonal- or organization-focused. This strength is best illustrated by the findings regarding the work status - citizenship behavior relationship. Ifthe study had only examined overall citizenship, then the conclusions regarding the findings would be somewhat unclear. In other words, the discovery that there are not significant differences in part-time and filll- time workers in regard to loyalty and obedience behavior would have been overlooked, and recommendations based on these findings would be misleading. Therefore, the current study illustrates the importance of distinguishing between different types of citizenship behaviors, as well as the specific target of the behavior, to understand how and why individuals choose to put forth efi‘ort toward these non-required activities. There were several weaknesses in the present study which affect both its validity and its generalizability. The most potentially damaging weakness to the validity of the study is the lack of agreement between supervisory ratings for the two companies in which dual ratings were conducted. This is important because it calls into question the accuracy of the citizenship ratings. This lack of agreement may be due to unfamiliarity with the 98 employees’ behavior, personal opiniOns and biases regarding individual employees, or misunderstanding of the items on the questionnaire. Since no questions were raised by the supervisors regarding interpretations of survey items, it would appear that the first two reasons listed are more probable causes. To eliminate the first reason (i.e., unfamiliarity), only first-line supervisors who worked directly with employees were used in statistical analyses. Thus, any remaining inaccuracies would be due to opinions and biases. Even though supervisory ratings have been used in past citizenship research (Konovsky and Pugh, 1994), peer ratings, or even 360-degree rating systems may provide a more valid measure of actual citizenship behavior (see Van Dyne and Le Pine, 1998). However, a 360-degree system will naturally result in difi’erent perspectives of citizenship behavior, and it will be challenging to determine which is correct. Researchers should consider who would be the best observer of the different types of behavior, to determine who would most accurately rate the employee on various types of citizenship. For example, a coworker may be the best observer of altruistic behavior, but a supervisor may be most familiar with the employee’s citizenship in regard to obedience. The second main weakness of the study is that, while the range of age of the workers is large (16 - 70 years), the overall sample of service workers is quite young (mean age = 23). Post hoc conversations with general managers confirmed that a large number of employees are students who are working their way through school, either as part-time or full-time employees. Therefore, their student status indicates that their current job is somewhat temporary, and may impact their perceptions of support and their willingness to participate in citizenship behaviors. In other words, the sample in the 99 current study may not be representative of other service-type industries, namely retail establishments or hospitals, in which the labor population is more established in their careers and more likely to be working in permanent filll- or part-time positions. However, prior studies using the same citizenship behaviors found a range of mean scores of 4.6 - 5.6 out of a possible 7 (Van Dyne et al., 1994), which is similar to the current sample with a range of mean scores of 3.1 - 3.7 out of a possible 5. Also, perceived organizational support scores were similar to prior research: 3.3 out of a possible 5 for the current study compared to 4.35 out of 7 (Wayne at al., 1997). A third area of concern is the fact that the sample is quite regional, and may not be generalizable to restaurant employees in larger cities. Specifically, there are more likely to be individuals who choose a career in restaurant work in larger cities than in a mid-sized Midwestern town. These restaurant employees may then be permanent workers who would be more likely to participate in citizenship behavior and have different perceptions concerning relationships with their work organizations. Finally, another concern is the low reliability of the social participation (.60) and functional participation (.70) scales. Reliability estimates above .85 are usually desirable, indicating that results based on analyses conducted with these two scales must be carefully interpreted. However, the findings regarding social and functional participation are already significant, and may therefore be considered conservative estimates of the true coefficients. In other words, correcting the reliability coefiicients for attenuation would result in similar, but stronger findings of significance. 100 Future research Future research related to the current study should occur in five primary areas: considering other possible antecendents to perceived organizational support, using other types of non-traditional workers, comparing samples of workers in different jobs, examining other relationship-oriented variables, and investigating other types of behaviors. Iwilltalkabout eachoftheseintum. The current model may be further informed by examining the antecedents of perceived organizational support found in past research, namely promotions, developmental activities, and organizational tenure (Wayne et al., 1997). Specifically, in the present study tenure was associated with a number of citizenship behaviors, as evidenced by significant regression betas (e.g., adv. part = .19, p<.05; altruism = .24, p<.01; loyalty = .27, p<.001). It would be interesting to determine if this direct effect of tenure on citizenship behaviors was in fact mediated by, or fully explained by, perceived organizational support. Also, future research should examine the likelihood that employeesjattempt to reciprocate for promotions and developmental activities granted them by the organization through the exhibition of citizenship behaviors, and whether or not these possible effects are due to perceived organizational support. The model developed in this study may apply in comparing other types of non- traditional workers with permanent, filll-time employees. Temporary and seasonal employees are two types of contingent workers that may differ fi'om permanent employees in regard to their employment relationships and work behavior. For example, Feldman, Doerpinghaus, and Turnley (1994) argue that both temporary and part-time employees 101 feel as if they are treated impersonally by the organization and its permanent members. Similarly, Krasas-Rogers (1995) found that temporary clerical workers felt alienated from coworkers, and perceived little control over their work activity. Therefore, despite the finding that there were no differences among full- and part-time workers in perceived organizational support, there may be variations in perceived support among temporary and permanent workers based on prior findings of social isolation difi‘erences. Another area of non-traditional employment is telecommuters. These employees spend only part of their work week at the ofiice, and may thus perceive a different type of relationship with their work organization than employees who are present 5 days per week, 40-50 hours per week. However, the current study suggests that telecommuters may not difi'er fi'om on-site employees in the amount of support they perceive from the organization, but may be less likely to exhibit citizenship behaviors since they may have less opportunity to do so due to their restricted time at the ofi'lce. The third main area of research would be to examine workers with different jobs. This may include comparing service to manufacturing workers, which was discussed above. It may also include examining difi‘erences within the service sector, such as comparing restaurant servers or other relatively low-paying positions like delivery personnel or daycare workers with a more professional, high-paying sample like nurses. Since nursing is considered a profession, it may attract different types of employees than restaurants. Both areas have a high percentage of part-time employees, and thus would make an interesting contrast in how various types of part-time workers view their relationship with their employers, and how that may correspond to behavioral differences. 102 To address the question of why work status and organizational citizenship behavior are related, several aspects of the employment relationship may be examined in lieu of perceived organizational support. For example, there is a modest amount of past research on part-time and full-time differences in job satisfaction and organizational commitment (Logan et al., 1973; Still, 1983). Recent findings indicate that there are no differences in these work attitudes when employees work their preferred work status (Morrow et al., 1994). However, there is no research that relates part-time and full-time workers’, or other non-traditional employees’, satisfaction or commitment with possible work behaviors. Specifically, if non-traditional workers difi‘er fi'om traditional, full-time employees in how they conceive of job satisfaction and organizational commitment, then past research regarding these two work attitudes and related behaviors may be inaccurate and ungeneralizable to non-traditional employees. Another relationshipeoriented construct that may impact behavior is perceived inclusion, or the extent to which an employee perceives oneself to be an insider in an organization. This may explain why part-time employees do not perform interpersonal-citizenship behaviors. They hesitate to form attachments, perhaps because they perceive themselves to be viewed as outsiders as compared to full-time employees. Thus, they are less likely to help coworkers and attend social work filnctions. Other possible constructs that may represent causal processes in the present model may include covenental relationships (cg, Van Dyne et al., 1994), psychological contracts (Rousseau and McLean Parks, 1993), and justice issues (e.g., Greenberg, 1987; Moorman, 1991). For example, Moonnan (1991) found that fairness perceptions affected 103 employees’ willingness to engage in citizenship behaviors. In the current study, one could argue that part-time workers may perceive certain injustices when they compare promotional and financial opportunities to those granted full-time employees. Part-time workers typically receive no or fewer benefits, and are usually not considered for promotions (Pfefi’er and Baron, 1988), and thus may judge management as treating them unfairly. This may explain why part-time workers exhibit less functional participation than full-time employees. If part-timers believe they are being treated unfairly, they may be less willing to take on extra responsibility. Finally, choosing a difi’erent type of work behavior may add understanding to the ramifications of having a good, supportive employment relationship. Deviant behaviors (Robinson and Bennett, 1995) may be one area of study. This is a relatively new area of research. However, it is attracting a lot of attention due to increased violence and theft in the workplace. Variations in aggressive, anti-social behavior such as sabotage, tardiness, theft, and violence may occur among non-traditional workers because these employees may be less attached to their work organization and thus may not feel a need to abide by company policy. Also, if these workers do not feel supported by the organization, despite their hard work, they may feel that the company owes them some form of compensation for their hard work. They may then pilfer fi'om the company, and justify it as a form of reciprocation (Gouldner, 1960) for their prior efforts which were not recognized. To filrther understand the current study, future research should also include refinements that help clarify the relationships found in this paper. For example, working one’s preferred work status may help to explain under what conditions the work status - 104 citizenship behavior relationship will‘or will not exist. I would speculate that individuals who are currently working their preferred status, or preferred number of hours per week, would be more likely to reciprocate through citizenship behavior than those employees who are not working their desired work status. Also, the perceived organizational support - citizenship behavior relationship should be examined for conditional boundaries. One possible moderator to this relationship may be perceived violation of the psychological contract between the employee and his/her work organization. Specifically, if an employee perceives a violation to their work psychological contract, he may not be willing to participate in citizenship behavior to the extent that he normally does. Thus, under conditions of psychological contract violation, there may not be a relationship between perceived organizational support and citizenship behavior. Conclusion The current study used service employees to test a model designed to explain differences in work behavior among full-time and part-time employees. Results indicated that, although there were no differences in perceived organizational support, part-time and filll-time workers distinguish between and participate in difi‘erent types of citizenship behavior. Specifically, full-time employees exhibit more overall and interpersonal-focused citizenship behavior than part-time workers, but the two work status groups do not difi‘er on their level of efi'ort toward organization-focused behavior. Finally, relationships were demonstrated between perceived organizational support and six distinct types of citizenship behavior. 105 106 APPENDIX A: MANAQERIALl§UPERVI§QRY QUEIHQNNAIRE This quesfimnaimispanofapmjectbemgconductedbyresearchersfiomtheMichigan State University Department of Management. The purpose of the project is to examine factors that influence the attitudes and behaviors of employees. Thus, this questionnaire asks you for your opinions about the behaviors of those who work for you. We also ask you to provide information aboutyourselfthatwillhelpusunderstandyourresponsestotbcquestions we ask. Pleasctake about 45 minutes to answer the questions on the following pages. lfthisquestionnaireistobeusefill,itisirnportantthatyouanswerallquestionsfranklyand honestly. Your employees will NOTseeyourquestionnaire. Inaddition, noonewillbetold about yourpersonalanswersatanytimeinthcfirulre. Yourdatawillbeincludedinasummaryresearch reportusedinacademicsettingsalongwithtbedatafiun‘rotbers. Thereportwillnotincludeany infonnafionthatwiflaflowanyonemidenfifyanyofyourindividualresponses. Although many ofthe questions might seem repetitive, we hope that you will answer each one of them. However, you havetherighttochoosenottoanswerany individual questionsthatmay makeyouuncomfortable.’1'hisisnotatest, sotherearenorightorwronganswers. Please indicate your opinions, not what others might think. Please note that completing this questionnaire signifies your voluntary agreement to participate and your voluntary agreement to release the resultingdatafortheresearchpurposes specifiedinthis letter. Thank you for your cooperation and assistance. Ifyou have any comments about this questionnaire ortheprojerx, pleasefeelfi'eetowritetherndownonthebackofthclastpage. Inaddition, ifyou would like to talk with someone about the project, please contact Christina Daly by telephoning MSU's Management Department at 517-353-5415. Pleasesignhereifyouagreetoparticipateinthisstudy: Subject signature Date Print Full Name Company GENERAL INSTRQQHQNS The following questions are designed to measure employee behaviors. Please be as honest as possible; there are no right or wrong answers. Please answer the following questions for each employee you manage. These answers are completely confidential, and will NOT be communicatedtoyoursupervisororcompanyrnanagement. Inaddition,theywillnotbeusedto influence employment records or performance ratings. Thankyou foryourcooperation. Ifyouhaveanyadditionalthoughtsorconnncnts aboutthethings thatI’veaskedyou,pleasefeelfieetowritethcmbeloworonthebackofthis survey. 107 “’ CONFIDENTIAL “ INSTRUCTIONS: Please answer the following list of questions for EACH employee; print employees’ first and last names acm the top, and place your answers in the boxes below each name. Your answers will be kept completely confidential, and will NOT be used for performance rating purposes. Indicate the extent to which you agree or disagree with each of the following statements by writing a number in the box below each name using the following responses: PRINT EMPLOYEE NAMES HERE 1 - Smut than: 2 - Disagree 3 - Neither Disagree nor Agree 4 - Agree 5 - Strongly Arm others who have heavy work loads even otherswhohaveworkrelated a makes creative to coworkers assess current Helps coworkers think for themselves Does not push superiors to perform to higher standards pursue Represents organization favorably to outsiders 80 W33 Does not defend when criticize Would awept job at competing organizations for more money urge money work if Shares ideas for new projects or improvements widely Works so personal appearance is attractive and Is notinvolved inoutsidegroupsforbenefitof organization 108 " CONFIDENTIAL " l - Stronsly Liam 2 - Disagree 3 - Neither Disagree nor Agree 4 - Agree 5 - Stmsly Am wastes at work Regardless ofcirulmstances, produces highest quality Doesnotmeetal.,ldeadlinessetbyorganintion Keeps work area clean and neat Volunteers for overtime work when needed Leaves group meetings early without valid excuse Neglects to follow organizational norms and instructions Tells others about the lousy company they work for on Frequently comes to work late instead of working Makesfunofcoworkers Loses temper when working with others . Plays mean pranks on coworkers PRINT EMPLOYEE NAMES HERE “ CONFIDENTIAL “ 109 Please answer the following questions about yourself: Age: (years) Gender (check one): Male Female Predominant Ethnic Background (check one): White/Caucasian Black/African Ameriean Chieano/Mexiean American Hispanic/Spanish American Am Indian/Native American Asian/Pacific Islander Other Pleasespecify How long have you worked for this organization: Years Months How long have you held a job with supervisor responsibility: Years Months A.......fififififitfitflttfi......OOOOOOOQQO......flttfittfittt... 110 APPENDIX B: EMPLQYEE QEEfiTIQNNAIRE This survey is part of a project conducted by researchers from the Michigan State University Department ofManagement. The purpose ofthe project is to study factors that efiect the attitudes and behaviors of employees. This survey asks you for your opinions about your own work attitudes. We also ask you to provide information about yourself that will help us understand your answerstothequestionsweask. Pleasetakeabout lOminutestoanswerthequestions onthe following pages. Hflfisquesfimnflnismbeusefid,hishnponamMyouamweraflquesfiomfimklyand honestly. Your supervisor, or any of yourcoworkerswillNOTseeyourquestionnaire. Inaddition, noonewiflbetoldaboutyourpersonalanswersatanytimeinthefuture. Yourdatawillbe includedinasummaryreport,usedonlyinacadernicsettings,alongwiththedatafromothers. The reportwillnotinclude any informationthatwillallowanyonetoidentifyyou oryouranswers. Altl'ioughmanyofthequestionsmightseemsimilar,wehopethatyouwillanswereachoneof them. However,youhavetherighttochoosenottoansweranyindividualquestionsthatmay makeyouuncanfoflable.1hisisnotatest,sodrerearenofiflorwronganswers. Please indicate your opinions, notwhatothers might think. Please be awarethatcompletingthis survey meamMywvdumaryagreempaNdpaeandnleaseymrmswemformereseamhpumoses specifiedinthisletter. Thankyou foryourcooperationandhelp. Ifyou lave anyeommentsabout this survey orthe project, please feelfreetowritethemdownonthebackofthelastpage. Inaddition, ifyou would liketotalkwithsomeoneabouttheproject, pleasecontactChristinaDalybytelephoningMSU's ManagementDepartment at 517-353-5415. Pleasesignhereifyouagreetoparticipateinthissmdy: Subject signature Date Print Full Name Company 111 “ CONFIDENTIAL “ E IN TR NS The following questions meamre employee attitudes. Please be as honest as possible; there are no right or wrong answers. These answers are completely confidential, and will NOT be used by your supervisor to influence employment records or performance ratings. Circleoneofthenumbersthatappearstotherightofeachquestion Pleasechoosethconenumberthat bestmatcheshowyoufeelaboutthequestion. Forexample, ifyouwereaskedhowmuchyouagreewith thestatement“lenjoyeatingMexicanfood”andyoufeelthatyouagree,youwouldcirclethenumberto therightofthestatementlikethis: l-Stro-slymaam 2-Disagree 3 - Neither Disagree norAgree 4-Agree S-Stmslym IenjoyeatingMexicanfood ............................ l 2 3 4 5 Thankyou foryour cooperation lfyouhaveanyadditionalthoughtsorcomments aboutthethings that I’veaskedymnpleasefcclfieemwntefliembeloworonmebackofthissmvey. IttOtta.......ttlttttttafitflfltlt....IOOOO...ttttttfittttt “CONFIDENTIAL“ INSTRUCTIONS: 'I'hinkaboutyourpastinyourworkorganization Pleaseshowtheextenttowhich you agreeordisagreewitheaehofthefollowingstaternents. Foreachstatementcircleanumbertothe rightofeachlincusingthefollowingchoices: l-Stronslymtam 2-Disagree 3 -Neither Disagree nor Agree 4-Agree S-Stmslym Theorganimtionstronglyconsidersmygoalsandvalues ...................... l 2 3 4 5 Helpisavailablefromtheorganizationwhenlhaveaproblem ................. l 2 3 4 5 Theorganizationreallycaresaboutmywell-being ........................... l 2 3 4 5 'I'heorganirationiswillingtoextenditselfinordertohelpmeperform myjobtothebestofmyability .................................. l 2 3 4 5 Evenifldidthebestjobpossible,theorganimtionwouldfailtonotice ........... l 2 3 4 5 Theorganizationcaresaboutmygeneralsatisfactionatwork.’ ................. l 2 3 4 5 Theorganizationshowsverylittleconcemforme ........................... l 2 3 4 S Theorganizationcaresaboutmyopinions ................................. l 2 3 4 5 Theorganintiontakesprideinmyaccomplishmentsatwork .................. l 2 3 4 5 t.t.O......O...‘......OOOIOOOOOOOOOOttfiflttttttttttttltt “ CONFIDENTIAL " 112 ‘* CONFIDENTIAL “ What is your current work status (check one): Part-time Full-time Please indicate the average it of hours per week you work: How long have you worked for this organization: Years Months Is your employment in this company: Temporary/Seasonal Permanent/Long-term Intermsofyouroverallpersonal situationandfinancial responsibilities, howmany hoursperweekwould you prefer to work: Given your current overall personal situation and financial responsibilities, which work status category would you prefer to work (check one): Full-time (40 or more hours/week) Part-time (1-34 hours/week) Other(35-39hours/week) ...t..........IIOI.........OOOOIOOOQOOOIDOItfittfifltfihttt INSTRUCTIONS: Thinkaboutyourpastinyourwork organimtion Pleaseshowtheextenttowhich youagreeordisagreewitheachofthefollowingstatemeuts. Foreachstatementcircleanumbertothe right of each line using the following choices: l-Stm-slylLimrss 2-Disagree 3 - Neither Disagree nor Agree 4-Agree S-Stmslym Ifeelverymuchapartofmyworkorganization ............................ 1 2 3 4 5 Myworkorganimtionmakesrnebelievethatlamincludedinit ................ l 2 3 4 5 Ifeellikelaman‘outsider’atthisorgaruzation ............................ l 2 3 4 5 Thisorganizationmakesmefeelasiflmatterhere .......................... l 2 3 4 5 Idon’t feel includedinthisorganization .................................. l 2 3 4 5 Ifeellaman‘insider’inmyworkorganimtion ............................ l 2 3 4 5 Myworkorganizationmakesmefiequentlyfeel‘lefi-out’ ..................... l 2 3 4 5 Please answerthefollowingquestioasaboutyourself: Age: (years) Gender (check one): Male Female Predominant Ethnic Background (check one): White/Caucasian Black/African American Chicano/Mexican American Hispanic/Spanish American Am Indian/Native American Asian/Pacific Islander __ Other Please mecify What is your Native Language: (please check one) English Spanish French Other Please specifiy WW 113 APPENDIX C: PQWER ANALYSIS FQR PARTIAL QQRRELATION AND REGRESSION COEFFIQIENTS (Based on Cohen and Cohen, 1983) Alpha = .05 Power = .80 Rz-estimate = .10 sriz-estimate = .03 k = number of IV ’3 f“ = population efl‘ect size of interest sr2 = partial correlation unique to each IV k3 = SCI at 1 (4’) fi= srizl 1 - R2 (equation 3.7.3; page 118) n" = (L / £2) + k +1 (equation 3.7.2; page 117) L = 7.85 (from Table 15.2, p. 527) F = .03 / (1 - .10) = .0333 n" = (7.35 / .0333) + 3 + 1 =236+4 =240 114 'REFERENCES Adams, 18. 1965. Inequity in social exchange. In L. Berkowitz (Ed), Advances in Experimental Social Psychology (Vol. 2). New York: Academic Press. Allen, R.E., Keaveny, T.J., & Jackson, J.H. 1979. A re-examination of the preferred job attributes of full-time and part-time workers, Journal of Management, 5: 213-227. Armstrong-Stassen, M., Al-Ma’aitah, R., Cameron, S.J., & Horsburgh, ME. 1995. Perceived Support and job satisfaction of full-time and part-time Canadian and Jordanian nurses. Paper presented at the annual meeting of the Academy of Management, Vancouver, British Columbia, Canada. Baron, R.M., & Kenny, DA 1986. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173-1182. Bateman, T.S., & Organ, D. 1983. Job satisfaction and the good soldier: The relationship between afl‘ect and employee citizenship. Academy of Management Journal, 26: 587-595. Becker, GS. 1964. Human capital. New York: Columbia University Press. Berry, W.D., & Feldman, S. 1985. Muldple regression in practice. Newbury Park, CA: Sage Publications, Inc. Blau, P. 1964. Exchange and power in social life. New York: Wiley. Brief, AR, & Motowidlo, SJ. 1986. Prosocial organizational behaviors. Academy of Management Review, 1 1: 710-725. Bureau of Labor Statistics. The Employment Situation News Release: Selected employment indicators, March, 1998. (Table A-4). Web site: http/lstats.bls.gov/news.release/empsit.t04.htm. Bureau of Labor Statistics. 1996. A difl‘erent look at part-time employment. Issues in Labor Statistics, April. Cohen, J., & Cohen, P. 1983. Applied mula'ple regression/correlation analysis for the behavioral sciences, 2nd Edition. Hillsdale, NJ: Erlbaum. Dansereau, F., Graen, G., & Haga, WJ. 1975. A vertial dyad linkage approach to leadership within formal organizations: A longitudinal investigation of the role making process. Organizational Behavior and Human Performance, 13: 46-78. 115 Deutermann, W.V., Jr., & Brown, SIC. 1978. Voluntary part-time workers: A growing part of the labor force. Monthly Labor Review, 101: 3-10. Dubinsky, A.J., & Skinner, S]. 1984. Job status and employees’ responses: Effects of demographic characteristics. Psychological Reports, 55: 323-3 28. Eberhardt, B.J., & Shani, AB. 1984. The effects of full-time versus part-time employment status on attitudes toward specific organizational characteristics and overall job satisfaction. Academy of Management Journal, 27: 893 -900. Eisenberger, R, Fasolo, P., & Davis-LaMastro, V. 1990. Perceived organizational support and employee diligence, commitment, and innovation. Journal of Applied Psychology, 75: 51-59. Eisenberger, R., Huntington, R, Hutchison, S., & Sowa, D. 1986. Perceived organizational support. Journal of Applied Psychology, 71: 500-507. Feldman, DC. 1990. Reconceptualizing the nature and consequences of part-time work. Academy of Management Review, 15: 103-112. Feldman, D.C., & Doerpinghaus, HI. 1992. Patterns of part-time employment. Journal of Vocational Behavior, 41: 282-294. Feldman, D.C., Doerpinghaus, H.I., & Turnley, W.H. 1994. Managing temporary workers: A permanent HRM challenge. Organizational Dynamics, 23: 49-63. Fuchs, VR 1985. An agenda for research on the service sector. In R. P. Inman (Ed), Managing the service economy: Prospecu and problems. New York: Cambridge University Press. Gannon, MJ. 1975. The management of peripheral employees. Personnel Journal, 54: 482-486. Gannon, M.J., & Nothem, J.C. 1971. A comparison of short-term and long-term part- tirne employees. Personnel Psychology, 24: 687-696. Goodman, RS. 1977. Social comparison processes in organizations. In B.M. Staw & GR Salancik (Eds), New Directions in Organizational Behavior. Chicago: St. Clair Press. Gouldner, AW. 1960. The norm of reciprocity: A preliminary statement. American Sociological Review, 25: 53-62. Greenberg, J. 1987. A taxonomy of organizational justice theories. Academy of Management Review, 12: 9-22. 116 Haring, M.J., Okun, M.A, & Stock,'W.A 1984. A quantitative synthesis of literature on work status and subjective well being. Journal of Vocational Behavior, 25: 316- 324. Hom, P.W. 1979. Efi'ects of job peripherality and personal characteristics on the job satisfaction of part-time workers. Academy of Management Journal, 22: 551-565. Jackofsky, E.F., & Peters, LR 1987 . Part-time versus full-time employment status differences: A replication and extension. Journal of Occupational Behavior, 8: 1-9. Katerberg, R., Horn, P.W., & Hulin, CL. 1979. Effects of job complexity on the reactions of part-time employees. Organizational Behavior and Human Performance, 24: 317-332. ' Katz, D., & Kahn, ILL. 1966. The social psychology of organizations. New York: Wiley. Kerlinger, EN. 1986. Foundations of behavioral research, 3rd edition. Orlando: Holt, Rinehart, and Winston, Inc. Konovsky, M.A, & Pugh, SD. 1994. Citizenship Behavior and Social Exchange. Academy of Management Journal, 37(3): 656-669. Krasas-Rogers, J. 1995. Just a temp: Experience and structure of alienation in temporary clerical employment. Work & Occupations, 22: 137-166. Lawler, BE. 1973. Motivation in work organizations. Monterey, CA: Brooks/Cole. Lee, T.W., & Johnson, DR 1991. The efl‘ects of work schedule and employment status on the organizational commitment and job satisfaction of full versus part time employees. Journal of Vocational Behavior, 38: 208-224. Levanoni, E., & Sales, CA 1990. Difl‘erences in job attitudes between full-time and part- time Canadian employees. The Journal of Social Psychology, 130(2): 231-237. Levinson, II 1965. Reciprocation: The relationship between man and organization. Administrative Science Quarterly, 9: 370-390. Locke, EA 1969. What is job satisfaction? Organizational Behavior and Human Ped'ormance, 4: 309-336. Logan, N., O’Reilly, C.A, & Roberts, K11 1973. Job satisfaction among part-time and full-time employees. Journal of Vocational Behavior, 3: 33-41. 117 Marchese, M.C., & Ryan, J. 1996. The mediating efi‘ects of autonomy on part-time/full- time employee difl‘erences. Paper presented at the annual conference of the Society of Industrial Organizational Psychology, San Diego, CA. McGinnis, S.K., & Morrow, RC. 1990. Job attitudes among full and part-time employees. Journal of Vocational Behavior, 36: 82-96. Meyer, J .P., & Allen, NJ. 1984. Testing the “side-bet” theory of organizational commitment: Some methodological considerations. Journal of Applied Psychology, 69: 372-3 78. Miller, R.E., & Terborg, JR 1979. Job attitudes of part-time and full-time employees. Journal of Applied Psychology, 64: 380-3 86. Moorman, RH. 1991. Relationship between organizational justice and organizational citizenship behaviors: Do fairness perceptions influence employee citizenship? Journal of Applied Psychology, 76: 845-855. Morrow, P.C., McElroy, J .C., & Elliott, SM. 1994. The efi‘ect of perference for work status, schedule, and shift on work-related attitudes. Journal of Vocational Behavior, 45: 202-222. Nardone, T]. 1986. Part-time workers: Who are they? Monthly Labor Review, February: 13-18. Nkomo, S.M., & Fields, OM. 1994. A field study of demographic characteristics and job attribute preferences of new part-time employees, Journal of Business and Psychology, 9(3): 365-375. Nollen, S.D., & Axel, H. 1995. Managing contingent workers: How to reap the benefits and reduce the risks. New York: AMACOM. Nollen, S.D., & Martin, J .H. 1978. Alternative work schedules, part 2: Permanent part-time employment. New York: AMACOM. Nunnally, J .C. 1978. Psychometric theory (2nd ed), New York: McGraw-Hill. O’Reilly, C., & Chatman, J. 1986. Organizational commitment and psychological attachment: The effects of compliance, identification, and internalization on prosocial behavior. Journal of Applied Psychology, 71: 492-499. Organ, D.W. 1977. A reappraisal and reinterpretation of the satisfaction-causes- perforrnance hypothesis. Academy of Management Review, 2: 46-53. 118 Organ, D.W. 1988. Organizational citizenship behavior: The good soldier syndrome. Lexington, MA: Lexington Books. Organ, D.W., & Ryan, K. 1995. A meta-analytic review of attitudinal and dispositional predictors of organizational citizenship behavior. Personnel Psychology, 48: 175- 802. Owen, J.D. 1978. Why part-time workers tend to be in low-wage jobs. Monthly Labor Review, 101(6): 11-14. Peters, L.H., Jackofsky, E.F., & Salter, JR 1981. Predicting turnover: A comparison of part-time and full-time employees. Journal of Occupational Behavior, 2: 89-98. Pfefl‘er, J ., & Baron, J. 1988. Taking the workers back out: Recent trends in the structuring of employment. Research in Organizational Behavior, 10: 257-303. Pierce, J.L., Newstrom, J .W., Dunharn, RB., & Barber, AE. 1989. Alternative work schedules. Boston: Allyn & Bacon. Podsakoff, P.M., MacKenzie, S.B., Moorman, RH., & Fetter, R 1990. Transformational leader behaviors and their efi‘ects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. Leadership Quarterly, 1(2): 107-142. Puffer, SM. 1987. Prosocial behavior, noncompliant behavior, and work performance among commission salespeople. Journal of Applied Psychology, 72: 615-621. Robinson, S.L., & Bennett, RJ. 1995. A typology of deviant workplace behaviors: A multidimensional scaling study. Academy of Management Journal, 38: 555-572. Ronen, S. 1984. Alternative work schedules. Homewood, IL: Dow Jones-Irwin. Rotchford, N.L., & Roberts, K11. 1982. Part-time workers as missing persons in organizational research. Academy of Management Review, 7(2): 228-234. Rousseau, D.M. 1989. Psychological and implied contracts in organizations. Employee Responsibilities and Rights Journal, 2: 121-139. Rousseau, D.M., & McLean Parks, J. 1993. The contracts of individuals and organizations. In L.L. Cummings & B.M. Staw (Eds), Research in Organizational Behavior, 15: 1-43. Sasser, W.E., Olsen, RP., & Wyckom DD. 1978. Management of service operations. Boston: Allyn and Bacon. 119 Settoon, R.P., Bennett, N., & Liden; RC. 1996. Social exchange in organizations: Perceived organizational support, leader-member exchange, and employee reciprocity. Journal of Applied Psychology, 81: 219-227. Shockey, M.L., & Mueller, CW. 1994. At-entry difi‘erences in part-time and hill-time employees. Journal of Business and Psychology, 8(3): 355-364. Shore, L.M., & Tetrick, LE. 1991. A construct validity study of the survey of perceived organizational support. Journal of Applied Psychology, 76: 637-643. Shore, L.M., & Wayne, 8.]. 1993. Commitment and employee behavior: Comparison of affective commitment and continuance commitment with perceived organizational support. Journal of Applied Psychology, 78: 774-780. Smith, C.A, Organ, D.W., & Near, J .P. 1983. Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68: 653-663. Stefl‘y, B.D., & Jones, J .W. 1990. Differences between full-time and part-time employees in perceived role strain and work satisfaction. Journal of Organizational Behavior, 1 1: 321-329. Still, L.V. 1983. Part-time versus firll-time salespeople: Individual attributes, . organizational commitment, and work attitudes. Journal of Retailing, 59(2): 55-79. Tilly, C. 1991. Reasons for the continuing growth of part-time employment. Monthly Labor Review, March: 10-18. Tsui, AS., Pearce, J .L., Porter, L.W., & I-Iite, J .P. 1995. Choice of employee- organization relationship: Influence of exterml and internal organizational factors. Research in Personnel and Human Resources Management, 13: 1 17-151. Van Dyne, L., Cummings, L.L., & McLean Parks, J. 1995. Extra-role behaviors: In pursuit of construct and definitional clarity (a bridge over muddied waters). Research in Organizational Behavior, 17: 215-285. Van Dyne, L., Graham, J .W., & Dienesch, RM. 1994. Organizational citizenship behavior: Construct redefinition, measurement, and validation. Academy of Management Journal, 37(4): 765-802. Van Dyne, L., & Le Pine, 1A 1998. Helping and voice extra-role behaviors: Evidence of construct and predictive validity, Academy of Management Journal, 41(1): 108-119. Vecchio, RP. 1984. Demographic and attitudinal difi‘erences between part-time and full- time employees. Journal of Occupational Behavior, 5: 213-218. 120 Wakefield, D.S., Curry, J .P., Mueller, C.W., & Price, IL. 1987. Differences in the importance of work outcomes between full-time and part-time hospital employees. Journal of Occupational Behavior, 8: 25-35. Wayne, S.J., Shore, L.M., & Liden, RC. 1997. Perceived organizational support and leader-member exchange: A social exchange perspective. Academy of Management Journal, 40(1): 82-111. Werbel, JD. 1985. The impact of primary life involvements on turnover: A comparison of part-time and hill-time employees. Journal of Occupational Behavior, 6: 251- 258. Williams, L.J., & Anderson, SE. 1991. Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17: 601-617. Withey, M.J., & Cooper, W.H. 1989. Predicting exit, voice, loyalty, and neglect. Administrative Science Quarterly, 34: 521-539. 121 "‘uuuuu