“:3. 5. 3:?! .5 i, k 3 .229... 11:29.... a... . ..2.Q.n..|‘.sl.1j:xr 1:3 1; .3. 5.3!.leuz 83:133.. .. ‘1 51 a}, an». r»; 13$"; .1. I v. a... , r 3.3;”... “La. « .53.... ~ ~ i. dun . .1: .F. 53!... a man? 1... 4:. i zlx‘.AI. . 1149.51 a. v. .5112131. i. :3. 3.5 ......:«. a...» a L: I .: 4.5.3.”? .. -. V . .V kmgflrfi . 1 .. 3.... ‘ 4:. Iznn 59.... 3:5. 3|..x333231 ‘ . as -. ‘ . 2.1 4...... 2' a. .. v}. 33.1.”?! :51. :3...- A . A.» “$3.. I. k 5.3!! 4.» .. e : A 1... ...:J1 it! £932.11. .3 II . a 322.}. A..i.....~q. .... 1. a. ‘ it... 51.. .53.?! . n... .. y... 33.2.... .. ‘0. 1 1L . ‘ov .. I: _ a... 3 : : J 5‘. of: .. 1 ”mafia”. . a: - LIBRARY t Michigan State University This is to certify that the dissertation entitled Who's to blame? An exploration of the role of blame perceptions as moderators of relationships between organizational justice perceptions and consequences presented by Michael Horvath has been accepted towards fulfillment of the requirements for PhD degree in _Es¥c.hology_ Major pro Date 1 ‘XAU/ MS U is an Affirmative Action/Equal Opportunity Institution 0.12771 4‘4f_ PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 chlFlC/DneDuepfiS—sz By Michael Horvath A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY t‘ 7- r ’7 Department of Psychology I" S. '15."; ‘ 2m] . NLZA T" i .7 TO BLAME? AN EXPLORATION OF THE ROLE OF BLAME VVBPTIONS AS MODERATORS OF RELATIONSHIPS BETWEEN . -, ' ‘ TIONAL JUSTICE PERCEPT IONS AND CONSEQUENCES By Michael Horvath AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 2001 ‘ , ' ' t ‘ ProfessorAnnMarieRyan ABSTRACT " . To BLAME? AN EXPLORATION OF THE ROLE or BLAME 0N8 AS MODERATORS OF RELATIONSHIPS BETWEEN TIONAL JUSTICE PERCEPTIONS AND CONSEQUENCES By Michael Horvath Research and theory on organizational justice has concentrated primarily on its We and its consequences, as opposed to the cognitive factors involved in making judgments of these matters in the real world. In contrast, the social psychological research on attributions discusses processes whereby causality, responsibility, and blame are assigned to individuals. An integration of the two literatures enables a more detailed understanding of how, and at whom, justice judgments are made in organizational settings. In this manuscript, a framework is proposed describing how justice judgments are made in organizations, and this framework is discussed with respect to common organizational phenomena. Based on this framework, hypotheses were generated regarding the relationship between supervisor and organizational justice perceptions and reactions to the supervisor and organization. Hypotheses were tested using undergraduate participants responding to 1) a scenario study about a promotion decision, 2) questions about their perceptions of blame, justice, and reactions to a recent parking ticket they meived. Results provided some evidence that blame attributions moderate the teletionship between perceptions of an entity’s justice, and reactions to that entity, , M some isms with the measurement characteristics of the scales used may limit unstaingihmugn sir .-i n he. lam: ihc .’-~.'Tm= w indtbtcd to w \ Martini i‘ “pig :1, i ‘ v'f i i" - f WW if?“ last. “at: With u my; V, . {it 3* 4 {av ‘ ._\ A COPyright b Y MICHAEL HORVATH 2001 1 "its. icn Can, Mbptomymuherfimmciimmuwn “in?“ “7 ' blSU' to “rag-flog“. ‘ _ ‘ . f {9... tony-51'.- 1"-".‘l.r$«4'l-3AI "' I , ' , " TL: in- t ‘v:_) it. 5-» b 1:,- EU 5“ ACKNOWLEDGEMENTS pod at sufi‘ering through all of this cheesy, ‘thank-you’ stuff, so ' x fairly terse. I owe the following people my thanks: " 'mOm-r I am indebted to my advisor, Ann Marie Ryan, for all of her mentoring W lie years. In addition to providing comments on my dissertation, she had also played a substantial role in developing my analytical and writing skills. 9mm and Other Faculty. I need to thank the rest of my committee: Dan Ilgen, Linda Jackson, and Neal Schmitt, for their helpful comments on my dissertation. They, along with other faculty such as Rick DeShon, were also instrumental in developing me as a graduate student. The Wiiq Academic Communig. I am indebted to Zinta Byme at PDI for her justice measures, and Chip Shank at MSU for his help getting the experiment on-line. I also need to thank the Graduate School at Michigan State University, from whom I received a Dissertation Completion Fellowship that enabled me to eat while finishing my dissertation. WM. First and foremost, my wife Anne needs to be acknowledged for her constant love and support throughout the process. Similar thanks go to my close and supportive fi'iends who have been instrumental in motivating me and keeping me from getting too discouraged - Christine Scheu, Aaron Schmidt, Karen Milner, Jen Carr, and Merrie Mullins. Additional thanks go to my mother, Suzanne Horvath, as well as . momenfiudband Wave people in MSU’s I/O Psychoon program, at ' ‘, mdumormmrm i. .,.32 ‘ '_‘., Vacuvewurul uai' c L “MAC. fit 1.“... A MW of Justice Theory 4 Attribution Theory 0 Heider’s (1958) Levels 11 Correspondent Inference Theory 13 Covariation Theory 14 Shaver’s (1985) Theory of Attribution of Blame ...................................... l7 Weiner’s (I985; 1986; 1993; 1995) Attribution Theory ............................ l9 Culpable Control Model 9 s Attributions in the Justice 1 " L 77 Attributions in Justice-0 e'aw‘ “ " 77 Attributions in Justice Theory 70 A Justice Attributions Framew k ‘m The (‘ ‘ “n Locus of Causality “n Self—Locus 14 External Source as Imus “46 A Series of Examples of the Framework 18 Causal Self is Blameworthy as Causal Self is Not Blameworthy 4o Causal Supervisor is Blameworthy 40 Causal Supervisor is Not Blameworthy 41 Causal Organization is Blameworthy 42 Causal Organization is Not Blameworthy ..................................... 42 Factors Affecting Attributional T‘ ' ' 43 Factors Affecting Locus ' ‘9 ‘ 45 Individual Differences 45 Outcome Severity 47 Organizational Factors 48 Rates of Turnover, Transfer and Promotion of Supervisors, and Job “ “c“ c- 48 1”,... ‘ 49 Process and Decision Control so Self-Serving Bias 50 FactorsAfi‘eefing Blame AH " <1 SigznrpIMewithI _,' ‘ " 51 W i ' .52 MW Supervisor Justifiability and Intentionality ........................ Aeeounti/Explanations s4 ‘ " - - Part oftheJustioe Attributions Framework ................................... 56 , r. V _ M 59 'ilrspcific H,r ‘L 64 68 68 , ‘ 68 Procedure 69 Measures 70 Loom of Causality 70 Intentionality and J ustifiability 71 Blame 7] Organizational India. 72 Citizenship Behavior 72 Commitment 72 Power Analysis 73 73 Manipulation Checks 73 Psychometric Characteristics of Items and Scales 74 Organization Justifiability and Intentionality .................... 75 75 Organizational Fairness 76 Supervisor Fairness 76 Supervisor and Organization Fairness Considered Simultaneously 76 Organizational OCB’s and C " * 73 Supervisor OCB’s and Commitment 78 Correlations Among Items and Scales 79 A Word of Caution Regarding the Scale: 29 Relationships with Demographic Variables 2') Age 2') Current Emplo, ‘ in Looking for a Job Rs Experience with P ‘ Rs Hypothesis Test Results 24 Hypothesis In R4 Hypothesis 1h 24 Hypothesis “’2! as Hypothesis 7b as Hypothesis 1n ,3353g in nit. o7 .‘ur uGhoieeofBlameAmong“ 101 " Committee vs. Supervisor Blame ........................ 106 Supervisor vs. Nobody Blame ............................. 106 Committee vs. Nobody Blame ............................. 1 13 Hypothesis 6 1 13 Organizational OCB’s 121 . . _ Organizational (‘ " ‘ 121 ' NB ’ Supervisor OCB’s 123 ~ Commitment to the Supervisor 126 Controlling for the Effects of Covariates no PART 2 132 Method 132 " ', ‘ 132 Procedure 133 Measures 133 Blame 133 Justice 1‘ r" 134 Reactions Measures 134 Rmrlte 1 34 Psychometric Characteristics of Items and Scales 134 Blame ‘ ' 135 Parking Employee Justice ‘ ' 135 DPPS Justice ‘ ‘ 135 Parking Employee Reactions ‘ ‘ 138 DPPS Reactions ‘ ‘ 139 Scale Descriptive Statistics and Intercorrelations Among Scales l 39 A Word of Caution Regarding the Qoalm 141 Tests of Hypothesis 6 141 Parking Employee Reactions 144 DPPS Reactions 144 Controlling for the Effects of Covariates 145 DISCUSSION 149 Source Perspectives on Justice (Hypotheses 1-4) 150 Supervisor and Committee Blame (Hypothesis 5) 152 Blame as a Moderator of Justice/Reactions Relationships (Hypothesis 6) .......... 154 ‘ " Part I 154 Part 2 1 58 \ Limitations of the Cur-rem Study ‘ I60 pm for Future Research and Application 163 m........USI.Q£. m: 5 ‘- 176 FirstPartoftheSmdy ms , offlieExpuimmtasSeenbyP-A L ‘ 192 E ~fortheSeeondPartofflieStudy 7'“ -m)l.¢\'t‘.}§u.t\ .‘. -.7-_l _L . thSLJ' 1:11... aa- ’ 3301 ii ‘ ' I , t, ,‘ ‘ tn“ , _ .11 1‘1. 5. Remix}. “[3210" 1'1 : TIN: 0. Re- Ind Sum-v : a ‘ 75bit fix-.9. Organizirv “M.- H ;, (nu-.111." 1 Table ‘1. 1"1‘: Peasant“. av .~ Tfilc 11.x P.- Mcators u: : Niel] RC'= 1 More: 011; 5-. I“ ll Resuln- v .1 ofCau * ~1-1 of 1hr. A -1; n o e n....... LIST OF TABLES r . , ; 0958) Levels of Responsibility 12 . ""1‘1ffullvlj. .' ,, 1,"!5 ». M Influencing Attributional Judgments in (‘—w ' 1' 46 18 Rem.“ Standard Deviations, Intercorrelations, and Internal Consistency - _ mes”forthe Scenario Variables Rn ' r .1'. 1 . "#‘fikesults of Hierarchical Regressions of Organizational OCB’s on National and Supervisor Jmm'ce as Table 5. Results of Hierarchical Regressions of Organizational Commitment on Olganizational and Supervisor Justin- 26 Table 6. Results of Hierarchical Regressions of Supervisor OCB’s on Organizational and Supervisor Justice 86 Table 7. Results of Hierarchical Regressions of Commitment to the Supervisor on Organizational and Supervisor Justice R7 Table 8. Results of the Regressions of the Supervisor Blame Item on Indicators of Causality and Justification 89 Table 9. Results of the Regressions of the Committee Blame Item on Indicators of Causality and Justification 00 Table 10. Results of the Logistic Regressions of Committee vs. Supervisor Blame on Indicators of Causality and Justification 102 Table 11. Results of the Logistic Regressions of Nobody vs. Supervisor Blame on Indicators of Causality and Juafifimtinn 104 Table 12. Results of the Logistic Regressions of Nobody vs. Committee Blame on Indicators of Causality and ' "" 105 Table 13. Results of the Regressions of Organizational OCB’s on Blame and Fairness trauma 1 17 1 ,‘ . - Wewoomfi Dime " .‘1‘ ' \‘jV-érttt- w-l- LlST OF FIE-11111 ‘1 of Commitment to the Supervisor on Blame and 120 1.. ...~ Wound Internal Consistency ima‘." .o Ia '26 -Reglesaons of Employee Reactions on Blame and Fairness ’8 amm‘ I '1 .. 142 0 i ’7'- » . any!" on» v- f 1M” of the Regressions of DPPS Reactions on Blame and Fairness A (Luv: 1 1r ' 143 Wale-1:1 From: 4. A 1»: W11"! 1: Figure 5 F 131m 1.. Flam“: ‘ smtcrv .. Figure 1 macaw . Flguru .-, . SU‘L'Z'VIy r' . Hgmr .1 :1 lupcn' 1212: FM ll. 31:.1 mum IQXCIE': 12.81: 1111;» 5h at 'W-Supt'rwaq. :1. ;‘ ' a ”if, I l i“. :1 mmmtmmat ecsupwuwumy 1 _ mumquedmmwnm l u The PI.- 1 m th- 5 ”‘7 \ «:1 LIST OF FIGURES . "m receiving a negative performance , ‘ 21 i6; sample of Werner s (1986) control and stability dimensions applied to a: QW’ s supervisor, afier the individual has received a negative performance 77 trust-tum . Figure 3. A decision tree describing judgments of responsibility and justice in an Wood setting 31 Figure 4. A process model describing judgments of responsibility and justice in an organizational setting 19 Figure 5. A path model representing the typical ‘source of justice’ perspective ............... 62 Figure 6. A path model representing the current framework’s perspective ....................... 63 Figure 7. Supervisor blame as a function of manipulated causality and manipulated Slipfl'ViSO!’ justification 07 Figure 8. Supervisor blame as a function of perceived supervisor causality and manipulated supervisor justifiability 93 Figure 9. Supervisor blame as a function of chosen causality and manipulated supervisor justifiability 05 Figure 10. Supervisor blame as a function of chosen causality and perceived supervisor justifiability 96 Figure 11. Supervisor blame as a function of manipulated causality and perceived supervisor intentionality 98 Plan 12. Supervisor blame as a function of perceived supervisor causality and m. 'vad mm ' intentionality 99 MUS. Supervisor blame as a function of chosen causality and perceived I 100 ' mmeommitteeoverthesupervisorasblmewomy A, 1".“ M..- . . . jufifimunm!” Elf." 1nd u .7... .I‘LL ‘11 V": *\ _/., ofchoosing the committee over the supervisor as blmneworthy .. = - 1 causality and perceived committee intentionality ............. 108 ' H lb". Wof choosing the committee over the supervisor as blameworthy 1-1’. 11¢fmanipulated causality and perceived supervisor intentionality ............. 109 M 31% of choosing the committee over the supervisor as blameworthy Wot perceived supervisor causality and perceived supervisor intentionality] 10 Figure 18. Probability of choosing the supervisor over ‘nobody’ as blameworthy as a function of manipulated causality and manipulated supervisor justification ................... 11 1 Figure 19. Probability of choosing the supervisor over ‘nobody’ as blameworthy as a function of chosen causality and manipulated supervisor justification ........................... 1 12 Figure 20. Probability of choosing the supervisor over ‘nobody’ as blameworthy as a function of manipulated causality and perceived supervisor justification l 14 Figure 21. Probability of choosing the supervisor over ‘nobody’ as blameworthy as a function of manipulated causality and perceived supervisor intentionality ..................... l 15 Figure 22. Probability of choosing the supervisor over ‘nobody’ as blameworthy as a function of perceived supervisor causality and perceived supervisor intentionality ....... 1 16 Figure 23. Organizational OCB’ s as a function of perceived committee blame and organization procedural justice ,‘ 122 Figure 24. Organizational commitment as a function of chosen blame agent and organization procedural justice ,‘ 124 Figure 25. Organizational commitment as a function of chosen blame agent and organization interpersonal justice, ,‘ 125 Figure 26. Supervisor OCB’ s as a function of perceived supervisor blame and supervisor procedural justice, , ‘ 1 27 Figure 27. Supervisor OCB’s as a function of perceived supervisor blame and supervisor informational justice perceptions 128 Figs: 28. Supervisor OCB’s as a function of perceived supervisor blame and 'suplvisor interpersonal justice perceptions '29 “Mmfn'T'L [‘4 W: 3'1. ' as a function of perceived DPPS blame and DPPS 1's! 1 '- 5: amen of chosen blame agent and DPPS TmAWmdlstt. ;-1 «Mofindxv:.t1 W ‘m 21 3...!- Sl'flfi 'man ,3]. . Mle'tm-Q ,. (Mum. ,..- . ‘ ,udyL-tf, ; .. pmuw-v . '11)“! ,‘UU '91:}; WILX‘LI'L. W11: amt: . ‘ «gramme! :11 .r . "XVI I1! Fd'|.r!\:,_ m‘ WW agent and DPPS procedural 147 I48 INTRODUCTION ' w—be treated unfairly. As we go through life, however, we ofien UWthatmakcus stopandtakenotethat ouroutcomesarenot . Q P 3 3 “baddesired, and that we anticipated and expected better. We feel .W these events, and often take action in order to rectify the situation. My,” are distressed when we observe injustice inflicted upon another individual or W of individuals. The widespread support of various civil rights legislation (e.g., Means with Disabilities Act of 1990; Civil Rights Act of 1991) and recent United Sues interventions in several international conflicts (e.g., Somalia, the Balkans) can be m'ibuted at least in part to perceptions that the current world situation was unfair to certain individuals. Similarly, in our daily work lives we have ample opportunities to judge the fairness of organizational actions (e.g., through performance appraisals, promotion and hiring decisions, terminations, etc). Traditional theories of organizational justice, as will be described below, discuss how judgments of fairness, or justice, are made, and how these judgments affect organizational outcomes, such as job performance, commitment to the organization, and deviance and withdrawal behaviors. Ultimately, the majority of the literature on mizational justice links holistic perceptions of a few dimensions of justice (i.e., disu'ibutive, procedural, interactional, and informational) to these consequences. Mar, it is possible that the existing organizational justice literature, while possessing . {Wabilitympredictseveral important omwmes,standsyettobenefitfiom ’5 isthatofathibufiom—thatiatheoriesandresearch ' . mindividuals make judgments about the causes of events, and Wm canbe held as responsible or blameworthy for a negative outcome “M 195‘; Kelley, 1967; Shaver, 1985; Weiner, 1986). This dissertation is mum as a preliminary exploration of how the two literatures may be putatively integrated. Obviously, an integration of two such broad literatures as justice and attributions could yield no end of varied theories and propositions; the implications of a full integration would involve countless facets of individuals and workplaces. For instance, one could focus on how individual differences in attributional tendencies can affect justice judgments, how the structure of the organization can affect tendencies to make certain attributions, or even how certain aspects of organizational procedures can affect self-reactions to negative outcomes through attributions. Thus, this dissertation will need to have a narrower focus. The primary focus of this dissertation involves the different potential sources of justice typically available in organizations. That is, when an employee experiences an outcome for which justice is to be evaluated, the employee may consider that instance of justice (or lack thereof) as being a result of the actions of the employee’s supervisor, coworkers, or even such an anthropomorphic construct as “upper management” or the organization as a whole. Recent justice research has indicated that the perceived source of mayhem in determining to whom reactions to injustice are directed (c.g., WZOOO). Thisrecentresearchassmnesthat allpotential sourccsof ' - .theoryin conjunction with justice, however, it can be theorized thediffeling sources of justice can be moderated by employee ‘. y i - ' , ;. of the sources is to blame for outcomes considered to be unfair. mama developed here. jaws..- fiuhypothesized in this dissertation, the relationship between justice m ad reactions to different sources of justice is moderated by perceptions of tllebllnewotthiness of each source, important implications for organizational practice follow. That is, the current line of research into source of justice (c.g., Byme & Cmpanmo, 2000) asserts that justice perceptions for each potential source of justice are made simultaneously by an individual, and that the individual reacts independently to each of these sources, directing each separate reaction to that particular source. According to contemporary thought, this process occurs for all sources in all situations, and Byme and Cropanzano have established measures of the magnitude of the relationships for two of the sources. However, if, as hypothesized in this dissertation, blame is necessary for procedural justice perceptions to relate to reactions to a source, then Byrne and Cropanzano may have underestimated the magnitude of these relationships. Their samples are comprised of individuals who blamed one source of justice, as well as individuals who blamed another source. By combining both groups of individuals into a single sample, they may not be able to measure accurately the made of the true relationship between justice and subsequent reactions. If the W alter taking the direction ofblame into account, organizations my M ’ ofblamebybeingbetterabletofocus .m at )3; ‘1 1111 ‘vhvt 51.14. .' ”Mdiswssionmdrelatedliterannecover «'1 I 3:," 1 1. ,.,.1_'t"" _ h j ‘ . outcomeshaveocemred—thatismutcomesinwhichhlm “mm, it should also be noted that positive outcomes may WW .. ibr positive outcomes, the relationship between (positive) W and positive reactions could be stronger for sources of justice that in ugardedss‘being most directly responsible for the positive outcome. However, the Mresearch will restrict its focus on negative outcomes and attributions of blame, with the suggestion that future research expand the present theories into positive outcome situations. Afier briefly reviewing the major works in the justice and attributions fields, I discuss how the application of attribution theories to justice theories can be used to make specific predictions about justice judgments and subsequent reactions. Finally, two research projects are described that initiate exploration of the veridicality of these ideas. A Brief History of Justice Theory Justice theory can be seen as a loose collection of various theories drawn from many different domains within psychology and elsewhere. Various researchers and theorists interested in issues of justice, equity, and equality in judicial, social, and organizational settings have attempted to propose explanations regarding these phmomena that pervade our everyday lives. Subsequent work has attempted to integrate and anbellish upon four basic types of justice. These types are typically conceptualized um”, procedural, interpersonal, and informational justice, although definitions, WWW of the dimmsions are disputed. 5 , 11.7%,“. ”Finalists; jgrsllcz -1 , ., . .tr IQ} u the Jesus to . imitations! fairness was stimulated by Adams' (1965) Equity ' , ‘ ,individuals are proposed to monitor their inputs to a given m hm they receive in return, in comparison to others. If an inequity is Whom then takes certain actions designed to remedy the inequity. Such “haul! include reducing one’s own inputs, getting comparison others to increase Wham, asking for more rewards, asking for comparison others’ rewards to be lowerd, or even withdrawing from the situation. This notion of the perception of inequity is traditionally associated with “distributive justice” or “outcome fairness.” Later additions to the concept of distributive justice included provisions for consequent behaviors such as sabotage or theft to make an inequitable situation equitable (c.g., Greenberg, 1993a). Another noteworthy contribution to distributive justice comes fi'cm Referent Cognitions Theory (RCT; see Folger, 1993, or Folger & Cropanzano, 1998). RCT argues that not only may individuals compare their situation to other individuals in similar situations, but that they could compare the current situation to any real or imagined situation as a referent. That is, even if all of an individual’s cohort is rewarded equally for the same amount of work, the individual may have an imagined ideal reward that does not match the current situation (c.g., from a previous job in another organization), and may thus determine that an inequity exists, taking subsequent actions to resolve the inequity. A decade later, Thibaut and Walker (1975) discussed fairness in different terms, WNW aspects of aprocedure (originally conceived of as concepts such as W} ' ‘ 1...“.1‘ " control)canalsoleadtofavorableimpressions.'lhisidea ’,>‘proced\ualjustice'or'processfairnels,"andrefaredtod1edegeeto n" “.115 can decis 2m «it. ‘3 J" : 1 "'5 V ' .; ':,:.9§anindividual’ssimafion1heydiscusswnceptsnmhasthehmeay, " .. ~f¢rmaking a decision areregarded as fair. That is, a negative «; « toothersmaynotbeseenas unfairiftheprocedures onwhichthe ‘ __ ‘ -- ~ were regarded as fair. Thibaut and Walker’s original ideas centered ”flies!” of voice as a key justice-related component of a procedure. Voice refers M an individual is able to communicate an opinion regarding a process or Mafia, what sort of outcome / reward is appropriate in that situation). Voice can heinstrumental, in that the individual is capable of having an effect on the procedure or outcome, or non-instrumental or value-expressive, where the voice will have no effect and exists merely for the individual to communicate their opinion on the matter. Beyond having voice, procedural justice was argued by Leventhal (1980) to encompass several other features. According to Leventhal, factors affecting the fairness of procedures include the following: consistency, or whether procedures are applied in consistent ways across time and across individuals; lack of bias on the part of the administrator against a particular individual; correctability, or whether an outcome can be disputed and altered if it is determined to have been erroneous; representation of all afiected groups; the accuracy of the procedures used to derive the outcome; and the ethicality of the procedures involved. These categories have been subsequently explored, and have provided the basis for several measures of procedural justice in subsequent m (c.g., Colquitt, 2001; Moorman, 1991). Rice and Moag (1986) added yet another ingredient into the justice stew, manna interpersonal treatment given an individual can also alter perceptions . .441 h .. 'l ., r“ " "" W ' " ' ‘ ‘ ‘ "‘9 “in; 7' U .1, 5;. ' .‘7- V i ‘ - . i - ,- o ’ - v l ‘ pl 1:? . ‘ . . I‘ -- , - . , WM ‘.‘ 7. .t‘ -I_‘A‘ 335? F n. «U l Cit". ;;va Ski- rs- ! (‘14 WW die individual who was either responsible for the _._ “'V'eatedtheouteorne. I men 9931)) added a fourth dimension that further split Bies and Moag's ‘ KW justice” into its informational and interactional components. Midevised a typology by crossing the distinction between distributive and W justice with the distinction between systemic (i.e., formal) and social (i.e., interpersonal) components. The social aspects concerned facets of a justice situation that could be affected by interpersonal behaviors. Interpersonal justice was conceptualized as the “distributive” component of the social aspect of the situation and involved concern shown for the individual’s outcomes, whereas the more structural component involved the information communicated regarding the decision (i.e., “informational” justice). Informational justice has increasingly become used in later studies of justice (c.g., Greenberg, 1993a; Gilliland, 1994), and a recent discussion by Bies (2001) firrther clarifies some early confusion regarding the conceptual distinctiveness of interpersonal justice from its counterparts. Although a good deal of work has been performed to distinguish the various possible components of fairness perceptions, there remains a spirited debate as to the actual utility of such distinctions. That is, relationships among these fairness components, and between these fairness components and various organizational outcomes, have been increasingly disputed. For instance, fairness theorists have recommended abandoning the distinction between procedural and distributive justice (c.g., Sweeney & McFarlin, 1993) W and interactional justice (c.g., Skarlicki & Folger, 1997). ' - - mam.-(Colq~fiu. 2001; Colquitt. Conlon. Wesson, Porter. & , threeffects of various components of justice have not WW. Many studies have found procedural justice to have effects only Wefdimibutive justice (see Brockner & Wiesenfeld, 1996; F olger & My 1998). At times, perceptions of unfairness have been found to relate to reaction totbe supervisor, to the organization, to both (c.g., Byme & Cropanzano, 2000; Cropanzano & Prehar, 1999), or to neither. Also, various components of procedural justice (i.e., Leventhal's, 1980, criteria) have been related to different outcomes in difi'erent situations (see Colquitt et al., 2001, for a review). When one considers the implications of the variety in previous justice dimensionality and consequences, it may help to consider the cognitive demands that these judgments of justice place upon the individual. Adams’ (1965) theory insisted that the individual simultaneously consider two inputs (self and a comparison other’s) and two different rewards (self and a comparison other’s). More recent theories claim that, in addition to these factors, an individual considers three other broad categories (i.e., procedural, interpersonal, and informational justice) with regard to a single outcome, some of which have distinct subcomponents (e. g., Leventhal’s, 1980, facets of procedural justice). Furthermore, while much of the justice research has assumed a single and unambiguous causal “source” (i.e., one person or entity made and communicated an outcome), in an organizational setting there are often several different possible “sources” (c.g.,- rnultiple supervisors, coworkers and subordinates in a 360 feedback system; upper “Lament theories have asserted that individuals must also evaluate multiple in. firm bar 1110? {hit T311: 57 ' Marci”). It appears as though the amount of information v vwepmcessin ordertoevaluatethefairnessofanoutcomeis ,and it seems likely that individuals may need to use some sort of W shortcut for limiting the amount of information they need to process “it!” a conclusion as to the fairness of an outcome. such a shortcut may be provided by another line of psychological research, that of WM(e.g., Heider', 1958; Kelley, 1967; Shaver, 1985; Weiner, 1985; 1986; 1993; 1995; see Fiske & Taylor, 1991 , and Kelley & Michela, 1980, for reviews). It seems intuitively likely that, instead of evaluating each potential action or cognition of each possible contributor to a decision, the individual might make a decision regarding Which person’s or entity’s behavior led to the outcome (i.e., which person or entity was “causal"). Making this kind of attribution would then allow processing of specific outcome and process-related information to occur more easily. In the next section, a brief review of the attributions literature will be provided, followed by a discussion of how it has and how it might be integrated with justice theory in order to reach a more plausible explanation of justice evaluation processes. Attribution Theory As early as the middle of the past century, psychologists had begun to speculate about the processes typical individuals use to create impressions regarding the causality of events (c.g., Heider, 1958). Heider is generally credited for introducing attribution much to social psychology. Research on attributions followed from there to a point humuswhaeitwasmcognizedasbeingmimpomtandwidely-researched 3313,” ZEN. e I I | L _' 1:: -’ Mi. ' originated as a theoretical, prescriptive pursuit. That is, Heider , notinvolvedinmalringprescriptiveconclusionsreguding N” N" ' . r. ._ -. intimatelyreach attributional conclusions. Instead, they were W how commonplace, everyday. people arrived at conclusions with! or who caused an event. irapru .élhefollowing sections will briefly review the most important and seminal m on attribution theory (Heider’s, 1958, levels of responsibility, Jones & Davis’, 1965, Correspondent Inference Theory; Kelley’s, 1967, covariation theory; as well as Alicke’s, 2000, Shaver’s, 1985, and Weiner’s, 1985, 1986, 1993, 1995, theories of attributions and blame). Although a substantial amount of research has been reviewed for this manuscript, Fiske and Taylor’s (1991) review of the literature should also be acknowledged for its comprehensive discussion of several of the most important “theoretical traditions” covered herein. It should also be noted that these theories discussed below occasionally deal with different constructs, sometimes by the same or similar names. Jones and Davis’ and Kelley’s theories primarily involve attributions of causality (i.e., whether the cause for an event is deemed to be internal or external to an individual). Heider’s, Weiner’s, Shaver’s, and Alicke’s theories also include notions of Whether an individual can be held as ‘responsible,’ ‘blameworthy,’ or ‘culpable’ for an action. Although each of these theorists tends to use each term slightly differently (see Shaver, 1985, or Weiner, 1995), there is enough conceptual convergence among them for Mmdiscussion to treat these terms as loose synonyms. Musician-- : v. hemp animated in the c the ll 1"... 51...! respu ft) (g'fi ,; N." . i :1 3W that when individuals attempt to make attributions about 1‘ " ‘ "Ingmar, that process generally revolves around two basic forces acting on I Missing evaluated: internal (i.e., motivation and ability) and external (i.e., V simsfimnl) factors. From this distinction, Heider proposed a progression of types of What could be attributed to an individual on the basis of his or her actions, depending on whether the reasons behind the individual’s actions were judged to be internally or externally caused. These levels were originally intended to represent developmental states, in parallel to those of Piaget. In other words, an individual who was more developmentally advanced would use a higher level as the criterion for assigning responsibility and blame. However, subsequent theory and research have proposed that situations can also be characterized on the basis of these levels, with differential judgments of responsibility on the basis of which components of the situation are varied (e.g., Fishbein & Ajzen, 1973; Shaver, 1985). The most basic of Heider’s (1958) levels is association (see Table 1). That is, individuals at this level will ascribe responsibility and / or blame to another individual as long as the individual is in some way associated with the outcome. As an example, alppoae that I am in the room when a glass vase is broken, an event that recently happened in my life. If my wife were at this level of attribution, I would be held responsible merely because I was in the room. z-neenmwm sopln'sticated level of attribution is that of W In 7 -' , . f ’ ;w,_'-In'mdi~idudisonlyjudgedtobemsponsibleiffininfimis ’ " manners» morn-avast; Ll. " l s ".. . r ‘ ' (as later. N" ‘ ~1in 1“ _ of introduc- , ; mu , ' a." m , 'il‘ll‘i'? . W01! 0 it“ ‘ , " An individual is responsible for the outcome if the individual is F W!” . associated with the outcome. ‘w mbility An individual is responsible for the outcome if the individual caused the outcome. Formbility An individual is responsible for the outcome if the individual could have anticipated that the outcome would occur. Intentionality An individual is responsible for the outcome if the individual intended for the outcome to occur. Justifiability An individual is not responsible for the outcome if the individual’s intentional actions were justified by the situation. some way with the outcome. For example, I would only have been judged responsible for the broken vase if it had been I who had thrown the vase across the room. The next level, foreseeabilig, adds to the criteria whether the individual could have known that the individual’s actions would have produced the eventual outcome. For instance, if the glass vase had been at the bottom of a shopping bag that I had thought to have beat completely filled with pillows, I should not have been judged as blameworthy. Hm,iflhadknownthatmywifehadgonetothe storetobuyaglassvaseand a Wistandswreasonthatlshorddhavethoughtflntavuemm 'JéN' mmmmmordertobejudged responsibleorblameworthyforan ‘0. n ”.V dfimmindividualmustfimherbepemeivedashaving I a mWI'had meant to throw the vase across the room and break it should I ”I u Mite blame by an individual at this stage of attributions. Ml}: i'. Fwy,“ the level «My, ifthere were a good reason behind the Wt: actions that would alleviate intent (i.e., coercion, extenuating circumstances), attributions of blame should be reduced. For instance, if I had thrown the vase across the room in order to disable a burglar, I should be exonerated of the offense of breaking the vase. (Unfortunately, no burglar was present in the real-life example, and my wife and I spurt some time debating the foreseeability of a vase being in the bottom of a bag of pillows. I lost.) Although Heider’s (1958) original conceptualization of these levels proposed that justifiable circumstances would completely remove blame from the individual, subsequent research (as discussed by Shaver, 1985) has documented that the extent of blame attributed to an individual in a justifiable situation is equivalent to that which would be given if the individual had been able to foresee the outcome but had not deliberately intended for the outcome to occur. t Theo Another widely-cited theory in the realm of attributions is Correspondent Inference Theory (Jones & Davis, 1965). Perhaps the most important concept proposed in thitheoryis the “analysis of uncommon effects.” If an individual had two or more mm that could have resulted in different outcomes, the individual making 1“ -- thedifi‘ereneesbetweenthedifi‘ezentontcomesinosdatohfeu ' “:7 l.. .- v ' . " t '1'." L'. . . g N. ;. . . . _. .1 N the: by.“ WW I". _;‘T Mess-leouldhaveaskedhertomovethebaghenelfiorl .51 ‘ i‘ ». an alternate path through our house. Any of these behaviors would have ‘ mm the bag. However, only one of them (tossing the bag) would 1 W damage to the bag’s contents. Therefore, according to this theory, rny “tho inferthat myintent was to damage herpurchase, and is thus more Wh‘find me responsible for the damaged item. A person’s confidence in the attributions made increases when there are fewer difi'uenees between the outcomes — that is, there are fewer reasons for the acting individual to choose that particular behavior over the others. Jones and Davis also talk about other factors affecting attributions. For instance, if the action taken was less socially desirable than other actions, intent is more likely to be inferred for that action. If the acting individual was not seen to have had choice, then judgments of responsibility . are not inferred (similar to Heider’s, 1958, justifiability dimension). If the action is seen to be in accordance with the acting individual’s social roles, intent is also less likely to be ascribed (i.e., social roles function as situational factors). Wm Kelley (1967) introduced yet another important concept into theories of attribution. Kelley’s ideas are primarily concerned with situations where more than one “nation of a possible cause / effect relationship can be observed and examined. Kelly proposes tint individuals attribute causality (and thus responsibility) based on the mm between potential causes and outcomes. Kelley discussed three " " ‘M‘mbeobmvedinordertemakeenmldmafimforan . -”mu/truthcu:2-.‘.:‘.t‘.~.fi|htll_. ugt'ivcm '1 . V -. . ,-, . , . -. . . l- *9 ‘4‘ . ‘ u, I . Y I {3‘15 “0'; ”PM! “a. . g . " = fii‘pflfi‘ufiw entity (the entitybeing another individual, a task, . ,.n , . tie dime-n observe the 511% of the event. This dimension involves than. man cause being evaluated differs fi'om the effects of other potential mum a supervisor is acting in the same manner as any other supervisor Mm whether this supervisor’s behaviors are divergent fiom what one would expect from asupervisor). If other entities produce similar outcomes, then causality is not attributed to that entity. For instance, consider a performance appraisal in an organization. An employee being disciplined for inappropriate behavior may be inclined to attribute causality to the employee’s supervisor (the entity). However, if other supervisors have in the past disciplined the employee for similar behaviors, then the employee will be less likely to see the present supervisor’s actions as distinct, and will be lea likely to see the supervisor as the cause for the negative evaluation. Another dimension is consistency over time and modalities. Action / outcome covariations that are seen to be more consistent are more likely to be attributed to the entity in question. For instance, if the employee’s supervisor consistently rates the employee as substandard during every evaluation, and across evaluation categories, then the employee would be more likely to attribute causality for the negative outcomes to the supervisor. ‘ .. Pinflly, Kelley (1967) discusses mm as another dimension upon which to _ , . _ Wm. In other words, an individual would observe how other excl; p700 (Ail Shh 5 mar: l . supcvisorfihentheemployeeismorelikelytoambute a WIRE!) discusses how interactions between observations on each of these ”may function with respect to causality attributions. Kelley conceptualizes the m eiattributions on the basis of covariation in terms of an Analysis of Variance (MONA) metaphor, although subsequent theorists have criticized this heuristic (see Shaver, 1985). Additionally, Kelley discusses two other principles that can affect an individual’s propensity to make certain attributions. The first, discounting, states that if more than one cause is available, the individual will be less likely to make a causal attribution. In other words, there needs to be reason to eliminate other possible causes in order to make a conclusion regarding a particular cause. As an example, the cause of a particular automobile accident could be attributed to 1) either driver’s attention to the road, 2) whether or not the road were wet, 3) adequate placement of traffic signs, etc. If all of those causes for the accident remain possible, I should be less willing to choose one of them as ‘ e” cause than I would if I were able to eliminate some of them (c.g., it was a sunny day; there were no obstructions to viewing signs, etc). The other principle, augmentation, considers both “facilitative” and “inhibitory” factors with respect to a cause. If there are obstacles to the potential cause’s successfully accomplishing an outcome (i.e., inhibitory factors), yet the outcome still occurs, then ~ once Will be more likely to be ascribed to this entity, as it had the ability to overcome the Wm In the traffic accident example, suppose that warning signs had been H1073 than; i a n;- ‘3 ab Pu: ,4 a i L, \I-. a, ’ .1L"' " 5H w A it; this driver's probability of getting into an accident had . a warnings). Ifthis driver nevertheless gets into an accident, then it is Yr? ‘ . " ,'.‘ ( 1 Wthedriver will be chosen as the cause of the accident. The reciprocal ‘ Waist!forthosefactorsthatcouldbeseentomakeiteasierforthepotential I "anthem accomplished the outcome. For instance, if the other driver’s vision had uncharted suddenly by a dust cloud, that driver should be less likely to be chosen as me curse of the accident. Finally, Kelley proposes that in instances in which multiple events are not available from which to derive an attribution judgment, then the individual will use causal schemas and scripts in order to create mental depictions of other occurrences. For instahce, an employee just assigned to a specific supervisor has no past history with that supervisor, so has no means to judge Kelley’s (1967) consistency dimension (i.e., there is no evidence indicating how this supervisor would have treated this employee in similar situations). In this example, Kelley would argue that the employee would call into mind the schema of a manager, and imagine how the “typical” manager would have treated the employee in this instance. The present manager’s behavior could then be compared to the schema in order to reach judgments of consistency. ’ th ttri ution of blame Shaw’s (1985) model of blame attributions appears to closely follow previous when attributions (c.g., Heider, 1958; Kelley, 1967). However, one important WM must be attributed to Shaver’s conceptualization of the attribution l . ‘ '_ , ’ " . 5.91.,- s . '1 contemporaries is his discussion ofMaekie (1965; name, 2 " 'rx‘r ' ' j ,.;- ‘ -)—. :4" y* C0151 greet hi'bui 5151. .l a! ”'3 ‘ V . 1e space to how an individual determines which of several U A . . filmy to be considered the principal cause. Shaver reintroduces WW3 about causal attribution. According to Mackie, an individual “Airtime factors (i.e., a “condition”) that contributed to the occurrence of the “landless of whether each of the factors (i.e., potential causes) could alone be m for the event to occur. Then, each of these factors is considered in terms of whether it was necessary for the event to occur. The factor that is ultimately designated as ll}; cause is the one which is the “insufficient but necessary part of a condition which is itself unnecessary but sufficient for the result.” This particular cause, known by the acronym INUS, is then evaluated in the means specified by earlier attribution theorists. As an example, the list of contributing factors to the breaking of my wife’s purchase included her leaving the bag in the middle of my path, her obscuring the view of the breakables in the bottom of the bag, and my choosing to throw the bag as a means of moving it. Many other combinations of factors could have produced a broken item (c.g., an earthquake), so this set of factors was not necessary but sufficient. My choosing to throw the bag was itself insufficient for the damage to occur (i.e., no damage would have occurred if the bag had not been there in the first place), but given the set of factors Operating, mine was necessary for damage to have occurred. Thus, I should be seen as responsible for the damage. However, it could also be noted that, if I had thrown the bag “my. wife had removed the breakable items from the bag beforehand, no damage mowed. Therefore, she could also be said to be blameworthy. t“! . - ., -: of surfers (1985) theory distinguish it from otherflreories of 'I i * '~_ murmumeraew. .: . .. J‘-.». , at M "if 1".0 . ,e..._-- "s ‘V' ' V L \_ . ‘v—Z , — w-1 - .. . . ._ . _ ." Vx . 0-, ' m , '- 1R4" “A“ . w 6153?: s _ ’s "appreciation’ that the act was morally wrong (i.e., ’ . ~ . beheld responsible ifthe individual did not have the capacity to W hem wrong). Additionally, once an individual passes the ‘appreciation’ Mum and is deemed to be ‘responsible’ for the outcome, the attributing Wde the target’s justifications or excuses (i.e., if the justification or antenna: accepted, then ‘blame’ is assigned). ’3 l ' 1 6‘ l 3‘ 1995 Attribution Theo Of all of the attribution theorists, perhaps the one that could be considered to have the most influence on 1/0 psychology is Weiner (c.g., 1985; 1986; 1993; 1995). Awording to Weiner, attributions are most often made in response to a negative or unexpected event. That is, if people are satisfied with an outcome, or if they did not anticipate a situation any different from the one they experienced, then they are less likely to question the outcome. However, negative or unexpected events are more likely to become salient to individuals, initiating a questioning process on the part of the individual to discover the cause of the event, in order to make sense of the event and avert that outcome in future similar situations. It should be noted that the phenomenon of unanticipated events initiating more sophisticated cognitive processing is not unique to attribution theory - it is also an important component of several popular versions of self- rcgulatory motivational processes (e.g., Carver, Lawrence, & Scheier, 1996; Lord &. Levy, l994). Weiner’s emphasis in his writing on attributions is concentrated on W H ‘ _ : (i.e., school situations), although he milk“ ”W“! ”PENN“, ' ~ ,onotherphenomenasuehaspublicwsimm. ‘ ‘ " ;m(&5wm,iws).ommmw ' _ r "f _ Q . ,~ 1%-, 9‘ . v_.' ._ ;. ., H. UH *- 1.1.- “will: w w. W. ‘- 2's, iyx fi' , q n V . . »'-| t ; ‘44‘ t‘ ‘n " ‘ Manchespublic assistance (c.g., Shepelak, 1987) and customer ‘ EM 1984; 1988; Folkes, Koletsky, & Graham, 1987). m g. . . mum like other attributions theorists, gives central emphasis to the W in attributions. An individual who experiences a negative or Mm usually begins a search for understanding by assigning causality to the mm to Weiner (1986), an early typology of causes included ability, effort, task diflieulty, and luck. However, this typology influenced subsequent research so as to limit the types of causes studied. He subsequently initiated research to add to the list of the most common attributions: ability, immediate effort, long-term effort, task characteristics, intrinsic motivation, teacher competence, mood, and luck. Weiner then began to differentiate between causes, starting with internal vs. external (i.e., whether people ascribe cause to themselves, or externally to some other source), ultimately adding two more dimensions of attributions: stability and control. Milli: according to Weiner (1986), pertains to whether the particular cause of an outcome can be expected to remain constant over time. That is, could the cause be a momentary fluctuation of certain factors, or is it likely that the same outcome would occur again? Examples of stable causes might be one’s intelligence, personality, or perhaps the difficulty of a particular task. Unstable causes could be factors such as momentary fluctuations in motivation, mood, illness, etc. gem refers to whether the individual would be able to control the factors leading to the outcome. Obviously, there are Rule things that a person can change, such as motivation to excel on an exam, the at “Hart‘s ta; mt lawm. napkins t bra-five Wort-no ‘ . —-- --‘ -. -— — Slrllpilil3’ with subsequent enhancements ofthis general idea (see Misti)» "9 enhatioasofthesefactorsdeterminehowanindividual hear , , .. e demexpected event. Responsibility for an event is not automatically “if” locus of causality is attributed in a particular direction. Rather, the ELI: of the stability and control of the particular event and antecedents determine .l whéther or not a person is deemed to be responsible. Weiner discusses research showing how individuals making internal, stable attributions for failure in achievement settings demonstrate more negative emotions and reactions that are inwardly-directed (c.g., self- esteem and self-efficacy). Furthermore, when individuals are determining whether another individual should be punished for a crime, they must believe that the person had control over the event, and that there were no extenuating (i.e., unstable) circumstances surrounding that event. Figures 1 and 2 represent examples of possible attributions that an 3 Mood — I haven’t Random Chance — 3 been feeling up to these things just speed lately happen sometimes 5-? i . Motivation — I could , . do better if I applied Ablllty I); am myself more not a 3° employee in :I.‘ High Low 9 ‘ | ~t..'n..'1‘*' ;‘ . CODlI'Ol l‘t‘t ' . ‘ 0* r.. “We “.th 1.: ;. . _, ass-pleatWeiler-‘a (1986) control ndstabtfltym xii 'm-mffilddr' air; Random Chance— . filling out the PA happen sometimes I. was: me - » Or 'zational lic - m; w ~ Bias - the supervisor sup‘gtisor wasfifst y IS 0‘11 to get me following the procedures the organization set out High Low Control Figure 2. An example of Weiner’s (1986) control and stability dimensions applied to an individual’s supervisor, after the individual has received a ' negative performance appraisal. employee could make about a negative performance appraisal, depending on whether the locus of causality was attributed to the self or to the supervisor. Another important variable that has been discussed in the attribution literature concerns the idea of volition or intent to commit an act (c.g., Heider, 1958). Weiner (1986) distinguishes this variable from control, noting that we can occasionally do things 4 thatwecanconu'ol but neglect to do so by purely unintentional means. For instance, I can it‘ln'n Wmtrolwhether or not I show up on time forameeting, but occasionallylamlate, WEMIKlb.“ .,‘,,,., llidnotdehberatelymtendtomaketherestoftherndrvrduaisrnthegroup .4; c. _ w, formMeasrckandAllrsonaMmmeof -dl§h “evaporated” l (strap. ‘1. A," -..-. : ‘9 535;: it“, If. ”uh: s '1‘ With ‘ ' level of reward for themselves (i.e., how much money they should - n- ‘mtalofthenwmdsweregr'eatathmacmtainmnomrtthennobody mm receive a reward. The final person (the participant in the experiment) Whimsinder of the allotted reward or could ask for an amount that would mm over the limit, thereby preventing all group members from receiving mytlting'. Measick and Allison argued that if individuals are motivated completely by Want, then the sixth person should take whatever they can get. However, the literattn'e has shown that, if the sixth individual believes that fairness has been violated (i.e., the other people took more than they should have), then the individual may choose not to take a reward if that means that the individual’s unfair partners would be effectively punished for their greed (i.e., the individual could take an amount that would exceed the group total, preventing anyone from receiving any reward). Messick and Allison showed that the sixth person’s attributions of the intent of the rest of the group members (i.e., whether or not they knew the “limit”) moderated participants’ willingness to punish other group members. That is, if there were no way for the other group members to know that their totals were approaching a limit, the sixth group member was more likely to accept whatever small gain they could have coming to them. Schultz and Wright (1985) explored outcome favorability with regard to intent and attributions of responsibility. If the outcome is positive, then intent must be attributed to that agent in order for the agent to be judged responsible for the outcome. If the minimum/e, intent is less relevant for attributing responsibility; in this case, .1 I -V'qu‘ '4‘. x " “519$“; - . , . -.-. . ‘ “ g-ae,mintensona1negrigenoe)asweuesoommrssiontmusonu W that it does not matter whether the supervisor deliberately intended to m prnmtion or whether the supervisor merely overlooked the employee as a Milt ofnegligenee - in either case, the supervisor was judged responsible for the 01m. Similarly, Fincham and Schultz (1981) studied situations where an agent’s actions led to a certain outcome via some kind of “intervening cause.” For instance, one of their vignettes describes a woman who hides a valuable possession of her roommate’s in a shampoo bottle as a prank. Before the roommate can discover the prank, a third roommate takes a shower, and the valuable item ends up down the drain. Participants in the study were asked whether the first roommate was responsible for the loss. The roommate was judged to be responsible for the loss depending on two characteristics of the intervening cause — whether or not the intervening cause was foreseeable (c.g., whether the woman knew that the third roommate was about to take a shower), and depending on the intent of the intervening cause (c.g., whether the third roommate accidentally dropped the valuable down the drain, or whether she deliberately dropped it into thedrain to teach the second roommate a lesson). Finally, Schultz, Schleifer, and Altman (1981) also found that voluntary actions W'rakendeliberately) weremorelikelytobeseenascausalandmoredesaving .3, ' -. .. 'veevent. Additionally,meyexaminedthedifl'amma _ '3 J . . ”up; -. "' »(‘.\-_'. 't' _ , 't O I I" b “ ‘ ‘A ‘ R! e? 1’" , ‘ . to ' ._ -‘ I - . -- t ‘ -_ \- . o " ‘. eonditton(i.e.,theaetionenabledanotherfaaortoprodueethe N is. attributed to an individual only ifthe action were eonsida'ed ,- “I? fir the outcome to have occurred, reflecting the lNUS condition proposed by v3 Wimimsmm 1985). e W “ - Recently, Alicke (2000) proposed a Culpable Control Model of blame assiment. Alicke refi'ames typical aspects of the blame assignment process in terms of the perceived control that individuals have over the relationship between their thoughts and the outcomes of their behavior. The most interesting contribution that this theory adds to the extant literature on attributions is the idea of ‘spontaneous evaluation’ - individuals may form an impression of blame before any information is cognitively evaluated, and this initial impression may affect subsequent processing of evidence. Although this theory is interesting and could be found to relate to the present discussion of attributions (c.g., employees may be biased against a certain supervisor), this theory came to my attention after the present research was conducted. Future theoretical developments of the present framework (see below) are expected to incorporate ideas from this theory. In summary, the psychological literature regarding causality has been diverse, but I has shown promise for predicting attributions, blame, and subsequent reactions. Chang (1997) recently classified attributions research into two camps - covariation and power. ‘W‘flmristsarguethatindividuals make attributions about thecauseofan action L mm- mil 3'” HM.- : 6.e,thefintoftwoeveutsismorelikelytobeseaiastliecauseof “ 4. !'- exploredastheyrelatetoconclusionsabouttheeausalityof WW we I he mm, in contrast, eschew the notion that individuals are merely r W ofevents, who then use some sort ofcognitive process akin to a ram likelihood algorithm in order to obtain the expected probability of a second event occurring given the fact that the first event has or has not occurred. Power theorists instead are more comfortable with the notion that individuals do not judge one event to be causal of another unless they know about or believe in a particular causal mechanism relating the two. According to Cheng (1997), “causal power... is the intuitive notion that one thing causes another by virtue of the power or energy that it exerts over another... causes are not merely followed by their effects; rather, they produce or generate their effects” (p. 368). In a power theory sense, a particular event or object is seen as generating some sort of force to influence another event, instead of merely co-occurring. For imtance, if I observe wind and then see tree leaves moving, I would judge that wind is exerting some form of force upon the leaves. Power attributions need not merely be confined to physical phenomena; an individual’s careless action can be thought to exert some form of energy or force upon another individual’s affect. Although the differences between the covariation and power camps still appear to be salient, Cheng (1997) has proposed a “Power PC” theory that attempts to integrate the two. Cheng argues basically www.mmpumormmibuaombmmwmm __ f A. wmdmmdmmmom ' h ‘ . . ui. as trimming jam, c - the minimal 3; ‘ >— , a? _ . A . i I p I .n r A ‘ "V ' _ l 7 . , ’i 3}. , "It \ A81\ ‘1 WM- . I the; , Attributions in the Justice Literature .. 4‘ tum developinga general model incorporating the role of attributions and MW in organizational justice evaluations, I will briefly discuss some of “New that has previously explored the role of attributions as they pertain to “Mons. Some of the extant studies in the I/O and social justice literature are we immediately relevant to the present research; these works will be saved until later in this manuscript In this section I merely wish to highlight some overlaps between attributions and social and organizational justice. First, I will discuss several ways in which justice research has incorporated attributions. Then, I will highlight how attributions have begun to be included in justice theory. Ambujons in Jgtice-related Research First, it has been shown that in response to a negative or unexpected event (i.e., events that could be interpreted as unfair), the locus of causality of the agent’s behavior affects reactions to that individual. In the consumer perceptions arena, Folkes (1984; Folkes et al., 1987) has published several studies examining the effects of attributions on subsequent consumer perceptions, using an internal vs. external locus of causality framework. One study reported in Folkes (1984) asked participants to recall a recent dinner at a restaurant that they did not enjoy. A second study in the same article presented participants with a vignette describing another consumer experience. In both studies, the degree to which the negative experiences were attributed to factors internal or external to them I testament owner determined subsequent reactions (c.g., intent to ask for a .l mmmwtopauomzethebusmessmthefmnue)30flt '.. M. . y'i M 0m don are ‘ Brad ...;. ttnu gill lift .. that was not delivered as expected (i.e., they paid expecting _ ,_ . v -. cementing of lesser value). Folkes et al. (1987) interviewed airline 'flighnhadbeensubjecttodelays.Theyfoundsimilarefi'ectsforthe mm (i.e., whether the cause was internal to the airline or external, such as We! weather) as it pertained to consumers’ decisions to repurchase tickets with Maine airline, as well as consumers’ intent to complain. Other studies outside the domain of consumer research have arrived at similar results. Weiner, Amirkhan, Folkes, and Verette (1987) showed that if the cause for an outcome is perceived to be external to an individual, then attitudes toward and reactions to that individual are not as negative. Bradfield and Aquino (1999) demonstrated that an internal or external attribution for the action of a causal agent related to victims’ desire for revenge or willingness to forgive, respectively. Ployhart and Ryan (1997) found that if the cause for not being admitted to a graduate program was judged to be external, then self-attributions such as self-efficacy were less likely to decline in reaction to the rejection. Heuer, Blurnenthal, Douglas, and Weinblatt (1999) found that internal vs. external attributions for the respectful treatment given to an individual moderated the relationship between respectful treatment and perceptions of procedural justice. Second, research has been performed on the antecedents of attributions, following these antecedents through to their indirect influence on fairness perceptions. Tyler and Bies«(l990) discussed how interpersonal treatment given by a causal agent could Mhmbufionsthatthe affected individualmakes aboutthecauaal agent, . lather study mdicated that making an attributional judgment might in .- , « toalleviatesomeofthedisuessresultingfi‘omtheperceptionofan I mm Lupfer, Doan, and Houston (1998) allowed some participants a I mm about and report their hypotheses regarding why a negative event 1 mad. They found that these participants reported lower subsequent levels of distress mm who had merely filled out the distress measure after the vignette, or than those Who had participated in a filler task before completing the distress measure. In summary, previous research has, to some extent, incorporated attributions into situations that could be construed as unfair, or that measured fairness perceptions directly. In these studies, the attribution an individual makes regarding the cause of the outcome has implications for subsequent perceptions and reactions. However, the studies mentioned above, as acknowledged, are not those that are thought of as belonging to the traditional justice literature; these studies have been conducted in parallel with those in the justice domain. Thus, it is also important to consider how attributions have been considered in literature that specifically addresses organizational or social justice. A ' ' J 'ce Theo Building off of RCT (Folger, 1993), F olger and Cropanzano (1998) include in their conceptualization of Fairness Theory a discussion of “could,” “should,” or “would” cognitions. That is, when a negative event is experienced, a victim asks whether the camel agent involved could have done anything differently. The victim might also ask Whethatheagent “should” have acted differently on the basis of social roles or moral ‘t‘. «!I " E'— 'r. ,‘l .' . .‘ '. e ' _ Q3". . ‘ “' m askswhetherthe victim “would” have done mytln'ngdifi‘cuitly ., _ f “‘6; W'UMFwwmmW . .~ ‘. w 1' lit 31 - t O r ‘h l0. n: ESSYIH made Q no it '31 . audit w Nate that these descriptions discussed in Fairness Theory have already .FZI that‘dte causal agent has been chosen; the victim’s attribution has already been mfletc an external source. Ftnthermore, the dimension of “could" appears to be most similar to Weiner’s (I986) control dimension of attributions (for an external locus of causality). The “should” could be best described as similar to the notion of “intent” explored primarily by Shultz (e. g., Schultz & Wright, 1985), and the “would” concept appears to be unique to Folger’s RCT. Finally, it should be noted here that Weiner’s (1995) discussion of the blame decision process also specifically includes mentions of “should” and “could” phenomena after locus of causality has been attributed to an agent, although Folger and Cropanzano appear to have developed these ideas independently and relatively simultaneously. Thus, justice theories in recent years have been moving toward incorporating attributions. However, little empirical research in the justice realm has attempted to address how attributions fit into the traditional (i.e., four-factor) justice mold. For instance, what are the implications of Folger and Cropanzano’s (1998) “could” construct for the different types of justice? There is currently no theory linking concepts of attributions to justice components. A Justice Attributions Framework A homework integrating attributions with justice judgments is presented in Figure ' wittuomesortofoutcomeapreconditionin-mcsttheoriesfora ‘ 0 'l . W" Ila-tnylki‘ .. “Search- What is the Locus of Outcome? L" Internal (to self): Is the cause stable and controllable? Am I to blame? it ; Distributive Justice Judgment Valence and Expectancy of Outcome: negative / unexpected outcomes more likely to trigger search 0 Control and Stabilityjudgrnents will affect the degree to which self- esteem and other self-reactions will be affected. 0 If the individual is judged as blameworthy, he or she will rate procedural justice as fair. 0 If the cause is evaluated as unstable, then the absence of “correctability” in the procedure may initiate an evaluation of responsibility and blame, using “absence of correctability” as a new outcome (i.e., return to the top of the decision chart). An External Source (i.e., supervisor, coworker, anthropomorphized organization): 1) Which external source is to blame? 2) Was the event foreseeable? Did the source intend for the outcome to occur? Was the 4.—. source’s behavior controllable and stable? Were the source’s actions justified? Is the source blameworthy? External Source is to Blame: Procedural Justice perceptions of the External Source will be more negative. Procedural Justice perceptions of the External Source will be related to behavioral and cognitive reactions to the External Source (c.g., lower commitment, retaliatory behaviors). Procedural Justice perceptions of any other source will not be related to behavioral or cognitive reactions to that other source. What-ion tree describing External Source is not to Blame: Procedural Justice perceptions of the External source will be more likely to be neutral or positive. Procedural Justice perceptions of the external source will not be related to behavioral or cognitive reactions to that source. If the evidence indicates that an additional source is responsible, then the blameworthiness of that source will be evaluated. If the External Source is not judged to be responsible, but the event was seen to be foreseeable, then perceptions of the Interpersonal Justice of the External Source will be related to behavioral and cognitive reactions to the External Source. lnik A Eur. ll '7."- rll... 4 'U‘V' slit. Hut (1'93: helm 1110i 0:90 its information, or status, distributed from an organizational representative to an employee. Examples of such outcomes could include, but are not limited to, decisions about hiring, promotion, transfer, bonuses, budget requests, and assignments of specific tasks. As no one theory of attributions has yet to achieve dominance, the propositions discussed below may be drawn from one of any number of theories of attribution. However, it should be also recognized that Weiner’s (1985; 1986; 1995) and Shaver’s (1985) attribution theories have had primary influence on the framework discussed below. The present framework differs from previous models in two respects. First, it allows for individuals to choose among several sources external to the individual, as opposed to an internal/external dichotomy. Additionally, the present framework incorporates the role of justice perceptions. Additionally, it should be noted that this framework does not fully integrate attributions with all components of justice. This framework focuses primarily on procedural justice, while leaving distributive, interactional, and informational justice relatively unmentioned. Based on the attributions literature, I deemed procedural justice to be the justice component that is most readily integrated with attributions. Future developments of this framework will incorporate other components of justice. One final caveat should be introduced prior to the discussion of the framework. The present framework regarding attributional, responsibility, and blame judgments operates under an assumption of an arm principle of reward allocation. That is, this framework is intended to be relevant only to situations where outcomes are expected to be commensurate with individual inputs or contributions. Events where outcomes are 32 I” 9 16» m indi. per: dish IHEIC Edit the 0 till iii) A lair as: L V. “a”. want Pics .iy‘ allocated on the basis of other rules (c.g., equality or need) are not incorporated into this framework. m When an individual is faced with an outcome, it is typically the case that the individual had an gprigrj expectation regarding the outcome, given the individual’s perceptions of past behavior and merit. Thus, the individual can at this point make a distributive justice judgment (i.e., whether the outcome itself were fair). If the outcome matched expectations, it is likely to be regarded as distributively fair. Simultaneously, the individual may initiate a causal search in order to determine responsibility and blame for the outcome. Various theorists have asserted that more cognitive effort is engaged when individuals have to deal with negative or unexpected events (c.g., Feldman, 1981; Lord & Levy, 1994), and attributions researchers have kept in line with this assertion, arguing that negative or unexpected events are more likely to trigger causal searches (c.g., Shaver, 1985; Weiner, 1995). Thus, individuals experiencing negative or unexpected events should be more likely to begin a causal search, beginning with a search for the locus of the outcome (i.e., is the cause internal to the self, or is it the result of another individual?). Locus of Causalig Most attributions researchers discuss locus of causality in terms of whether the perceived cause is internal or external to an individual (c.g., Shaver, 1985; Weiner, 1995). However, in an organizational setting, more specific distinctions may need to be made. Although it is interesting to discuss whether the cause for an organizational outcome is attributed to the employee or to a factor external to the employee, it may yet 33 SptCl' that ll org hear the be more meaningful to consider, given an external judgment, to whom or to what the employee attributes causality. Attribution theorists assert that individuals take action against causal agents that are found to be blameworthy (c.g., Heider, 1958; Shaver, 1985; Weiner, 1995). Thus, in an organizational justice situation, a wronged employee may not merely take random actions in order to alleviate an injustice - the employee may more specifically focus action against the source (c.g., a supervisor) that is deemed to be blameworthy. Although many different potential sources are available as loci of causality in organizational settings (c.g., supervisors, coworkers, subordinates, upper management, the organization as a whole, etc.), attribution theories suggest that the individual should arrive at a narrower list of causal agents. Although the list of potential loci is vast, previous attributions researchers have tended to make the greatest distinction between internal and external loci. That is, events attributed to the self have been seen to have different implications than events attributed to external sources. Although this manuscript is intended to highlight the need to consider wing}; external source is blamed for a negative event, it is still important to recognize that the traditional distinction is yet worthwhile. The cognitive processes regarding an external source will be similar regardless of which external source is considered, and these processes will be different fiom those processes engaged when an internal attribution is made. Therefore, this section will have only two subsections. In the first subsection, the cognitive blame process for internal (i.e., self-locus) causality will be discussed, followed by a parallel subsection regarding external sources. Self-Locus. Once the attribution of causality has been made, attribution theorists pr0pose that judgments of responsibility should follow. Judgments of responsibility will 34 are; Ole: min 333 Bill! r ['1 an at” CONE $33: outs,- in» it“ vary depending on which agent (i.e., self, supervisor, etc.) is judged to be causal. Within each cause, however, attribution constructs and rules such as Heider’s (1958) levels, or Weiner’s (1986) dimensions of stability and control, will be used in order to judge responsibility and blame for the event. If the causal agent were judged to be one's self, determinations of stability and control would then influence whether the self is determined to be blameworthy. See Figure l for an example of how these dimensions might influence judgments. Individuals have high control over their own mood and motivation, with mood being the less stable of the two. Less controllable self-caused aspects would include random error or ability. If the outcome is judged to be something over which one had a high level of control (or, in Heider’s terms, the outcome could have been foreseeable given the individual’s current behavior), it might be argued that the individual "could" have acted differently in order to obtain a more positive outcome. Thus, individuals should blame themselves (i.e., not the organization or any organizational representative). If the outcome is judged to be a function of one's ability, then the individual is likely to either accept the outcome or become depressed, with a corresponding decrease in self-efficacy and expectations for success in similar subsequent situations. If the individual is found to be responsible or blameworthy for the outcome, then the individual should also rate the procedures used to reach the outcomes to be fair as well, because the individual has decided that the negative outcome was warranted based on the individual’s behavior. It is also possible that the individual may deem the locus of the event to be internally caused, yet in a judgment of ultimate responsibility mt find the self to be blameworthy. For instance, an individual may perform at a superior level most of the 35 be ju in Erie. filth i'fi“: 1 Qt time, but the individual was only observed on one particular day when the individual made an error. In this case, the individual may acknowledge that the individual’s behavior was the cause of the outcome (c.g., denial of a promotion), but not attribute respgnsibiligg to the self, chalking it up to bad timing, etc. However, if the individual, although deemed the locus of causality for the outcome, is not judged to be blameworthy (c.g., the employee was observed as part of the performance evaluation on the employee’s only bad day in four months), then the individual may also be less likely to judge the procedures used to arrive at the outcome as fair. External Source as Locus. Attributions towards an External Source are proposed to be fairly similar (see Figure 2). These individuals have control over mistakes (i.e., through effort and attention) as well as over their deliberate intent. Individuals will only be judged as blameworthy if the action was under their control (i.e., at Heider’s, 1958, intentional level). However, one would expect variance in the ultimate reaction to an External Source based on the stability of the event (i.e., justifiability). That is, if an outcome is believed to be the result of an unstable factor (c.g., momentary fluctuations or lapses in attention), or coercion (i.e., due to external pressures, social roles, etc.) then the wronged individual would be more likely to forgive the harmdoer. But, if an outcome is believed to be the result of a stable, long-lasting bias on the part of the harrndoer, then it is more likely that the wronged individual would have negative perceptions of fairness regarding this source, and would react to the injustice by specifically targeting actions against that source (c.g., withdrawal of citizenship behaviors, retaliatory behaviors, etc). However, if the External Source is not found to be blameworthy, the above negative fairness perceptions and subsequent reactions will be less likely to occur, as the 36 individual has judged that the External Source does not deserve retaliation for causing the negative outcome. It is possible that the individual’s attributional search could end here. The individual may still feel negatively about the outcome, and may thus regard the distributive justice in the situation as unfair, but it is possible that no other cognitions or behavioral reactions may result. It could be the case that, through inspection of the External Source’s situation, the cause of the External Sources actions could be perceived to be external to that External Source. For instance, a supervisor may decide not to promote an employee, and might thus be judged to be the cause of the negative outcome. However, if the employee deems that the supervisor, although making the decision intentionally, were justified (cf. Heider, 1958) because (for instance) the supervisor was rigorously following procedures set up by upper management, then the supervisor may be absolved of responsibility and blame. However, the employee may then begin to evaluate the responsibility and blameworthiness of upper management, thus returning to the previous step in the decision tree shown in Figure 3. As one final point about the decision tree, it should be noted that in the case highlighted above (i.e., the External Source is justified, but another External Source is then implicated), the degree to which the first External Source is deemed to have been able to foresee the individual’s negative outcome may still affect reactions and justice judgments against that source. In the example mentioned above, if the supervisor had known that the organization’s procedures would result in an unfair outcome for the employee, the supervisor may still be perceived as sharing some of the blame. In this case, it is suggested that the degree to which the supervisor has shown fair interpersonal 37 U63. ififiCI 3563 andr me: am"? C033 i "t‘ii (i.e.,- treatment toward the employee (c.g., sincerity, honesty, sympathy) may be related to reactions to the supervisor. In summary, once an individual judges that a negative outcome has been accrued, a search for causality begins (see Figure 4). In many cases, the attributions of causality and responsibility may serve to alleviate any negative attitudes toward oneself or toward potential External Sources not judged to be responsible for the outcome. Outcomes attributed to the self will usually only incur negative reactions if the behavior was controllable or stable. Outcomes attributed to other individuals will generate negative reactions toward that individual if the individual is perceived to have been responsible (i.e., to blame) for the outcome. A Series of Examples of the Framework In order to understand the framework more completely, it may be helpful to consider a general type of outcome, and explore how differences in the situation could lead to different types of attributions, as illustrated in Figures 3 and 4. In each of the following scenarios, the employee (“Doug”) was being considered for a promotion, to be awarded on the basis of his performance, and in each case, Doug is denied the promotion. However, Doug’s ultimate reactions to this negative outcome are very different. Six scenarios will be described, crossing loci of self, supervisor, and organization with whether or not the source was blameworthy. Causal Self is Blameworthy. Perhaps Doug has a manufacturing job where job performance lends itself easily to objective measurement. When Doug’s Supervisor sits down with him at the performance review meeting, the Supervisor hands Doug a printout of Doug’s performance over the last three months, showing that Doug’s performance has 38 \AUv-aHv-aP—Uu l‘l. tilt. Iiu.-Iu7 7.- .‘hnouv. Jaundf. o-.nfl-nl~vvu-.JI av- wan-numqu-nal-V- ‘ labilka A. l .wemtom 3:39.359: an 5 322:. can b___a_m=eam2 he mafia—wean weir—3.0.. .25:— mmooea < .Q 0.5»:— mm; uofltomzm 328088 \ Al $.53 8 commtomsm 9 3388M 853. mm I 252 mm uofltoazm 353005 02 comrcomsm 2358qu 02 852. 3.5885 $833 9 cozfifimwco 6055.83 mm 8 8288M Taowfimswwco m_ I 362 mm necemmcmmho $504 I 2:850 .8935ch mo> £52 mm 358 5528 we coups—axe oémmom Allulll oz “6:83 mono—Eu 9 .28 mm Item ”tom :wdv tom 3 mcouoeom A18? 39 been man: bong or go tori a cor. gem mil I been consistently below average. Doug thinks about his performance for the past few months, and realizes that he has not devoted suflicient effort to his job. He recently bought a new boat, and has frequently gone home early in order to repair the boat and spend time on the lake. Doug realizes that he had a choice as to whether to stay at work or go out on his boat, and he made the choice to devote more effort to boating than to work. Thus, Doug sees his actions as under his control, and fairly stable, as this has been a consistent pattern over the past few months. Although Doug is disappointed with not getting the promotion, he views the decision and procedures as fair, and harbors no ill will toward the Supervisor or Organization. Causal Self is Not Blameworthy. For this scenario, assume that the printout of Doug’s performance is identical to that of the previous scenario. However, in this scenario, the reason for Doug’s lower production is not volitional, but has to do with a bad case of the flu that Doug caught. This flu kept him out of work for a full week, and continued to hamper his performance after he returned to the job. Doug realizes that, in order to be seen as a candidate for promotion, he needs to have demonstrated superior productivity. Thus, the reason for his lack of promotion had to do with actions that he was supposed to have performed. However, contracting the flu was not under his control, nor is this a factor that will persist throughout his next evaluation phase. Therefore, Doug is disappointed in this outcome, but harbors no ill feelings toward the Supervisor, Organization, or himself. Causal Supervisor is Blameworthy. Consider a situation in which Doug's performance was indeed exemplary, but Doug still does not receive a promotion. In the performance review meeting, Doug's Supervisor describes the reason for Doug's 4O “El: l‘lSli pm Am 1‘“ bin is U m outcome. Although superior performance is a prerequisite for the promotion, the Supervisor says, the Supervisor still has some discretion to decide whether the employee should be promoted. In this case, the Supervisor believes that Doug could benefit from a few more months in his current job; the Supervisor is not convinced that Doug will be able to handle the new position. Doug knows that several of his colleagues, who work under different supervisors, had performance records similar to Doug's and were promoted. In this case, Doug blames the Supervisor. Doug perceives that the supervisor had the ability to promote him, but did not, for vague and perhaps biased reasons. If one were to give Doug a questionnaire that assessed Doug's perceptions of Supervisory justice, Doug would indicate low perceptions of distributive and procedural justice, with perhaps low perceptions of interpersonal and informational justice, depending on how the Supervisor handled the meeting. However, Doug would not respond as negatively to a questionnaire assessing perceptions of the Organization's justice. According to the present framework, if Doug chooses to retaliate for this outcome, Doug will direct the retaliation to the Supervisor, and not the Organization. Causal Supervisor is Not Blameworthy. Doug's performance is again exemplary, and the Supervisor discloses that the decision to deny Doug a promotion was under the Supervisor's control. However, the Supervisor explains to Doug how the department's budget is very tight this year, and that if the Supervisor decided to promote Doug, funds would have to be diverted from other sources (c.g., the department could not then afford to renovate the decrepit employee breakroom). Doug is disappointed by this outcome, acknowledges that the Supervisor made a judgment call, but also realizes that he probably would have made the same decision if he were the supervisor (i.e., Doug sees the 41 em and the r: had air: in: L' m 'I 1 SU; t Supervisor's decision as justified). Doug does not hold negative perceptions of the Supervisor, the Organization, or himself. Causal Organization is Blameworthy. In this scenario, Doug's performance was exemplary, but this time the decision is not in the control of the Supervisor, but rests with an Organizational promotion committee based in the Organization's headquarters. When the supervisor sits down with Doug for the performance review meeting, the Supervisor hands Doug a form letter from the Organization indicating that Doug's promotion has been denied. The letter explains that, although Doug's performance was excellent, the Organization chose to limit the amount of promotions that year, keeping the money budgeted for promotion in order to invest it in the stock market. However, Doug is also aware that upper management has a score to settle with Doug’s supervisor, and believes that the purpose behind Doug’s denial of a promotion is more a result of deliberate actions against Doug’s supervisor. In this case, Doug is not thinking about whether or not the Supervisor’s role in the promotion process has been fair at all - Doug's active perceptions are directed at the organization. If Doug chooses to retaliate, he will retaliate in such a way as to harm the organization with a minimum of harm to the Supervisor. Causal Organization is Not Mmewongv. Doug's Supervisor hands him the Organization's letter. In this case, however, the letter denying Doug the promotion explains that the Organization just found itself the target of a hostile takeover, and budgetary analysts predicted that the Organization needed every dollar it could find in order to fight the acquisition. If the Organization were taken over, everyone (including Doug) would lose his or her jobs. The Organization admits that it could have given Doug the promotion and taken its chances, but that it instead chose to deny Doug the promotion 42 in a Gilli mi .t. .5“: Her 3E0} iOCL‘ in order to fight the takeover more effectively. Doug is, as always, disappointed with this outcome. However, he realizes that the Organization's decision is justified, and harbors no ill will toward the Organization, the Supervisor, or himself. Factors Affecting Attributional Decisions An important criticism of previous attribution research is its limiting of locus of causality to merely an internal or external locus. For instance, many studies employ designs that either examine participants after having experienced an outcome, or that ask participants to express what their reactions would have been for a particular outcome related to them in a story or vignette (i.e., first-person studies; e. g., Folkes, 1984, study 2; Heuer et al., 1999, study 2). Others relate to participants or allow participants to view another individual’s actions, and ask participants to make attributions about whether the locus of causality was internal or external to that individual (i.e., third-person studies; e.g., Bradfield & Aquino, 1999; Folkes, 1984, study 1; Heuer et al., 1999, study I; Lupfer et al., 1998; Skitka & Tetlock, 1993). An exception to this deficiency is Shaver’s (1985) coverage of Mackie’s (1965; in Shaver, 1985) ideas, although to date no empirical research has been reported to test whether or not individuals actually make such distinctions or are capable of using this logical process in order to reach judgments of responsibility and blame. In other words, the ways in which the locus of causality has been conceptualized and researched have always treated the causal individual involved as a given, for which internal/external locus judgments are made. In theoretical discussions where an individual may be allowed to choose among multiple possible causal people, the issue of which source has been chosen is usually glossed over. Folkes (1984) did call for research 43 lllO-it Gilt ( 35535 {In ([1 on choices among potential causal individuals, but no research has subsequently been conducted. One possible exception to this criticism might be found in Lupfer et al. (1998), who allowed participants to choose external causes such as the influence of others, climate, God, Satan, fate / destiny, or luck / chance. However, no exploration was endeavored regarding antecedents of these specific choices. In an organizational setting, the potential causes for individual outcomes are often more than the simple matter of whether the outcome is attributable to the employee or one other individual (i.e., the supervisor). Organizations are complex, and job responsibilities and career paths are intricately intertwined. In selection settings, causal influences for a rejection that come easily to mind include not only the applicant and the assessor, but also all members of a panel interview and the individuals who wrote the applicant’s letters of recommendations. In a performance appraisal setting, recent forays into 360-degree feedback imply that not only do one’s supervisors, but one’s coworkers, subordinates, and perhaps even clients have the potential to be the causal influence on the outcome. Extant attributions research is lacking with specific research as to the direct influences on which of these causes is selected by individuals who receive an outcome. The following section will discuss potential influences on attributional decisions. It will focus primarily on factors that influence locus of causality attributions (i.e., self, supervisor, organization, etc.), although influences on other attributional variables (c.g., stability) will also be covered where relevant. For the purposes of this manuscript, attributions are important in that they allow for more precise and specific ultimate indgments of justice. The reader should be reminded that the previous section discussed judgments or perthipns of justice, as opposed to objective components of situations lfiiii Will the: in: the lift: related to justice, or to perceptions of those situations (c.g., Leventhal’s, 1980, components of procedural justice). These objective components may serve to influence the attributions made, and thus influence perceptions of justice through an attributional framework, as discussed below. Similarly, other characteristics of the organization may also influence tendencies to make different attributions. One more point should be made regarding the remainder of this section. While there are obviously many different possible external sources available to an employee experiencing a negative outcome, this section will focus primarily on the distinction between making attributions between one’s supervisor and an anthropomorphized organization, for two reasons. First, these are the two entities that seem most likely to function as external attributions (i.e., supervisors have much more control over an employee’s life than do coworkers, clients, etc). Second, the subsequent empirical study will focus on teasing out this difference in light of both the present framework and recent research (c.g., Byme & Cropanzano, 2000). Thus, the three possible sources that will be given emphasis in the following section are the self, a supervisor, and an organization. Factors theorized to affect judgments of locus and blame include individual differences, the severity of the outcome, organizational factors (e. g., turnover rate), objective procedural justice characteristics, self-serving bias, and social accounts. (These factors are summarized in Table 2.) Factors Affecting Locus Judgments Individual Differences. A good deal of the literature on attributions appears to be directed across all individuals, while ignoring individual differences (c.g., Feather, 1996). In this section, I will describe several individual difference factors that could be expected 45 \ Low: Table 2 Factors Influencing Attributional Judgments in Organizations Attributional Type Locus (internal vs. external) Locus (supervisor vs. organization) Factor Locus of control Belief in a just world Religiosity Empowerment of the individual Process control Decision control Voice Self-serving bias (valence of outcome) Objective procedural justice characteristics Outcome severity Turnover rate Transfer rate Promotion rate Job rotation policies Empowerment of the supervisor Blame attributions (i.e., justifiability) Legislation / governmental policy Corporate culture Social roles Objective procedural justice characteristics 46 to r liter tiff at ti] cor: lion am like EU} in ‘1‘.“ “LN. filial to influence the direction of causality. Most of these have been drawn from previous literature, although some are proposed on purely logical grounds. Attributions are one way in which individuals attempt to make sense of world events. According to various theorists (e. g., Tversky & Kahneman, 1974), the world can at times be so complex that individuals are likely to use heuristics in order to lighten the possible cognitive load. F urnham (1992) proposed that various individual difference variables might function as heuristics to help individuals determine causality in a complex social environment. Such heuristic individual differences include locus of control and the belief in a just world (see also Fiske & Taylor, 1991; Lupfer et al., 1998; Montada, 1992). Individuals with internal loci of control or strong beliefs in a just world are more likely to make internal attributions for their outcomes, and will thus be more likely to either see the situation as fair, or to not evaluate an External Source for responsibility. A host of other individual differences, such as religiosity (c.g., Kunst, Bjork, & Tan, 2000; Leiber & Woodrick, 1997) or stages of moral development (e. g., Tsujimoto, 1979) could also be studied with respect to their influence on the attributional process and thus justice perceptions. Future theorists and researchers should be encouraged to add to this list. Outcome Severigz. Previous research has linked the severity of the negative outcome to less positive fairness perceptions (c.g., Gilliland, Benson, & Schepers, 1998). However, severity may also function in the attributions process. According to Kelley and Michela (1980) it has been demonstrated that an individual is likely to match the magnitude of cause and effect. That is, an individual is unlikely to attribute the cause of a very severe event to what is perceived as a small and powerless cause. "Therefore, it is 47 ofor dlSCl Acct" 15ml Olin fir. Ulla probable that more severe outcomes are more likely to be attributed to greater and more powerful causes. Thus, attributions for more severe outcomes may be more likely to be directed at a multinational organization than at the supervisor who lives across the street. Organizational Factors. Based on previous literature and theory, there are a host of organizational factors that could influence individuals’ tendencies to shift loci of causality, as well as attributions of stability, control, and intent. These factors are discussed below. Rates at Turnover. T ransfer and Promotion QfSupervisors, and Job Rotation. According to theories of covariation as a basis for causal judgment (c.g., Kelley, 1967), it is more likely that causality will be attributed if a possible cause is followed by an outcome a greater number of times. That is, the more fi'equently a cause/outcome has been paired previously, the more likely it would be to be found to be the cause in the next situation. In an organizational setting, it is sometimes the case that supervisors will be transferred, promoted, or leave the organization. Thus, employees who have previously been under the control of that supervisor will lose any useful information regarding causality, because that supervisor is no longer a possible cause of the next outcome. Over time, the frequency of these occurrences may limit the ability of the employee to make a covariational connection between a supervisor and any organizational outcome. Additionally, Kelley’s principle of discounting states that when there is more than one potential cause that cannot be ruled out, the locus of causality is less likely to be given to any particular cause. Since there will be less information available for supervisors and coworkers in organizations with more turnover, transfers, and promotions, individuals may be less likely to be able to pin the cause for a negative outcome on one of these 48 inf: tiff an LC li‘OL mil: the i Sill-ll m potential sources. Alternatively, supervisors may be more likely to be judged to be the cause of an outcome, as there would be less information available to eliminate them. In a way similar to that for turnover, it seems likely that an organization that has a policy of job rotation would also have employees who would be less likely to make attributions toward the supervisor, for the same reasons - employees are less likely to be able to create a perception of covariation between a particular supervisor’s behavior and an unfair outcome. Alternatively, it could be the case that there would be less information about new supervisors to allow the employee to eliminate them as potential causes, thus making them more likely to be chosen as the causal agent than the organization or the self. Empowerment. Empowerment is generally believed to have a positive effect on the individuals being empowered (c.g., Hackman & Oldham, 1976). The research has shown that empowered individuals may be more motivated, and may report more positive perceptions of job satisfaction. However, supervisors who are perceived to be empowered may be more likely to be perceived as having power over the outcome in question. For example, in a more empowered organization, a supervisor may be expected to be able to use the supervisor’s own discretion in a promotion or termination situation, instead of strictly following a set organizational policy. As supervisors who are empowered may be seen as more able to be acting as a result of their own intention (i.e., they less coerced by any organizational structure or social rules), supervisors who are empowered may thus also be seen as more likely to be the cause of the outcome. Additionally, if it the individual experiencing the negative outcome is the person who is empowered, then an internal attribution may be more likely. 49 11‘." 9;“ , . {1105.1 ttit d 1.):- it. ~‘ 79 q . V'L‘kk ., _ will! Process and Decision Control. One benefit of the current framework is that it can pinpoint specific situations in which various components of procedural justice would be most effective, or effective at all. (Recall that procedural justice perceptions are defined as perceptions that the procedures used to determine the outcome were fair.) For instance, consider Thibaut and Walker’s (1975) process and decision control. Individuals are said to have process or decision control if they have the ability to influence l_i_o_w decisions are made, or what decisions are made, respectively. In other words, the individuals can personally take actions in order to ensure a fair process or a fair outcome. Additionally, the concept of voice (individuals being allowed to give their opinions about a process or outcome, regardless of whether it actually has influence) may also influence attributions. These factors (process control, decision control, and voice) will help to determine the locus of causality for the outcome. If an individual were given process control, decision control, or voice, they will be more likely to attribute the cause of the outcome to themselves, because they helped to bring that outcome about. These components of procedural justice relate to the ppwgr an individual might have to affect one’s outcome, so the individual might thus be seen as a more likely causal agent according to power theorists (c.g., Cheng, 1997). Self-Serving Bias. Various researchers and theorists (c.g., Cohen, 1982; Greenberg, Pyszczynski, & Solomon, 1982; Pyszczynski, Greenberg, & Holt, 1985; Pyszczynski, Greenberg, & LaPrelle, 1985) have discussed the idea of a self-serving bias. That is, individuals are more likely to believe that a favorable outcome is fair, and that an unfavorable outcome is not fair. If one were to reinterpret this phenomenon in terms of attribution theory, it stands to reason that causality attributions may be related to self- 50 . '1?“ Ole-1 5 com: Mi '1 If: I r serving biases. One can assume that an equity principle is operating in most organizational situations, such that an individual expects to be rewarded on a level commensurate with the individual’s behavior. That is, the individual expects that the procedure used is based on the concept of equity. If an individual reports that a procedure is fair, it implies that the individual was responsible for the outcome obtained — the process resulted in a reward that was commensurate with the individual ’5 performance. Similarly, if the individual reports that a procedure is unfair, it implies that the individual was not responsible for the outcome. Additional evidence could be provided by research on the “fair process effect,” (c.g., van den Bos, Bruins, Wilke, & Dronkert, 1999; van den Bos, Lind, Verrnunt, & Wilke, 1997; van den Bos, Wilke, Lind, & Vermunt, 1998) which shows that individuals experiencing a negative outcome under conditions they believe to be fair will experience more negative self-reactions than those who are able to make external attributions (i.e., an unfair process). Thus, positive outcomes are more likely to be attributed to the self, whereas less positive outcomes are more likely to be attributed to an external source. Factors Affecting Blame Attributions Compliance with Legislation. Cobb, Vest, and Hills (1997) found that non- exempt personnel were more likely to make attributions to formal organizational policy, whereas personnel who were exempt from FLSA legislation, and who thus did not perceive the need for the organization to have imposed consistent policies, were more likely to make attributions to the individual’s supervisor. It is unfortunate that in the former case, however, Cobb et al. did not firrther explore attributions regarding the control that the organization had over setting those policies. Nevertheless, it stands to 51 tori l \r Ui'ill reason that any legislation perceived to have been imposed on the organization that requires employees to be treated in a certain way will have influences on how actions of the organization and the supervisor are attributed. Thus, the presence of legislation or governmental policy prescribing certain treatment of workers will make it more likely for attributions of locus of causality to be attributed to an anthropomorphized organization rather than the supervisor. However, if the organization were to be subsequently judged for blameworthiness by the employee, it is likely that blame would not be placed on the organization either, but perhaps to the government instead, thus absolving any organizational representative of blame, with subsequent favorable impressions of fairness and a lack of retaliatory behaviors directed against the organization. Q9_r_'po_rate Culture pnd Sociial Roles. Several attributions researchers (c.g., Jones & Davis, 1965) highlighted the influence of social role perceptions in attribution judgments. Furthermore, Kelley (1967) and Lurigio, Carroll, and Stalans (1994) discussed the utilization of schemas to assist with judgments of responsibility. Thus, individuals’ perceptions of both how a person is expected to act and how a person typically acts (i.e., a schema or stereotype) may be able to influence causality and blameworthiness judgments. Bodenhausen (1990; Bodenhausen & Wyer, 1985) showed how stereotypes can influence judgments of justice and intentions to react (i.e., punish). Using student samples and vignettes, Bodenhausen and Wyer found that in cases where an actor’s actions were congruent with the stereotype for the actor (i.e., racial stereotype), punishment is recommended to a greater extent than if the action were not congruent with the stereotype. Furthermore, Bodenhausen (1990) gave participants vignettes of court cases, 52 [old .1]..— a Wk Vi. . flab» . r ill 9: g 'l’.’ F.‘ ‘91 I)“ along with the verdict that arose from the case. In these cases, recommendations for punishment still were influenced along stereotypic lines, even if the participants had been told that the actor in the vignette had been found not guilty of the crime. Although it is certainly true that individual differences in stereotypes may influence attributional judgments on their own, such findings may also be generalized to organizational-level situations. In organizations with strong cultures, for instance, a stereotype may emerge for the personality of the ideal (or typical) manager, whether this be one of fairness and empowerment or one of ruthless exploitation of subordinates. Employees may use these stereotypes / schemas in making judgments of organizational actors’ personality traits, and thus intent. Alternatively, such schemas could function as a perception of social roles, thus alleviating responsibility attributions to the individual (c.g., through coercion / choice, Jones & Davis, 1965, or through augmentation, Kelley, 1967) Objective Procedural Justice Characteristics. The section above mentioned how process and decision control (Thibaut & Walker, 1975) affected locus judgments. Similarly, other aspects of procedural justice (c.g., Leventhal’s, 1980, criteria) that have been previously considered could be seen as having more specific effects and influence at different points in this model, suggesting the reasons behind the varying levels of effectiveness reported in extant research. For instance, organizational situations and processes that are more consistent will be more likely to shift decisions in the favor of greater stability, regardless of the causal agent chosen. The presence of bias suppression will only have an effect on responsibility (and thus justice) judgments if the individual or entity whose bias is suppressed is the individual or entity judged to be causal for the 53 outcome. Other criteria such as accuracy and ethicality would only be effective if an External Source were judged to be causal. Alternatively, these criteria may also increase the tendency for the chosen locus to be the self (i.e., if all aspects of the procedure are fair, then the only other possible cause for the outcome is the individual’s behavior). Dadzie (1994) explored the effect of persisting episodes of injustice using a series of vignettes describing performance appraisal situations. In this study, some measures were taken after the vignette had described repeated instances of unfair appraisals from a particular supervisor. Dadzie found that the increasing occurrence of the injustice led to more severe behavioral intentions (e. g., to file a lawsuit instead of merely to complain). Thus, an increasing lack of accuracy and consistent applications of procedures led to increased negative reactions. Such a finding is also in accord with Kelley’s (1967) covariation principle, suggesting that repeated linkages of cause and effect can further solidify causality judgments. Social Accounts / Explanations Although the above discussion tends to highlight the facets of procedures and the situation that can indirectly influence attributions and subsequently fairness perceptions, it is also often the case that organizations or supervisors can directly assist the attribution process by communicating attributional information. The work on social accounts, or explanations (c.g., Bies, 1987; Scott & Lyman; 1968; see also Folger & Cropanzano, 1998; Shaver, 1985; Weiner, 1995) highlights how explanations given for an outcome can influence perceptions of justice. However, what is less often noted is that these explanations may affect fairness perceptions as mediated through attributions; they shifi responsibility away from a certain individual or entity. For instance, Greenberg (1991) 54 liar ittq; studied the types of explanations that were given for good and poor performance appraisals, and the effectiveness of those explanations. For good performance appraisals, the most effective type of explanation was one that attributed the level of the performance evaluation (i.e., a good score) to the individual’s performance (i.e., an internal attribution), whereas the most effective explanation for a poor performance appraisal was an apology (i.e., the explanation helped to assign responsibility to the supervisor). Similarly, Wood and Mitchell (1981) found that the provision of excuses for a negative outcome (i.e., shifting causality away fiom the account-giver) led recipients of the excuse to be less willing to assign responsibility and punishment to the account-giver, whereas apologies had the opposite effect. Bobocel and Farrell (1996) found that a causal account that shifted responsibility away fiom the supervisor and toward upper management was less likely to be seen as adequate, presumably with subsequent implications for the increased perceptions of unfairness among individuals given this explanation. However, it should be noted that this study does not necessarily contrast with the present framework; it merely indicates that explanations are not to be seen as a powerful, intractable method of altering attributions. For instance, in Bobocel and Farrell’s study, it is possible that participants had already made attributions in a direction other than toward upper management. When the explanation was given implicating upper management, then, the participants could have seen this explanation as a deliberate attempt to mislead them from the truth. Future research should explore exactly when eXplanations are likely to assist in influencing and creating attributions, and when they are likely to conflict with a priori attributions in such a way as to backfire in an organizational setting. 55 In summary, the attributions framework developed for this dissertation is represented complementarily by Figures 3 and 4. When individuals are presented with a negative outcome, they begin to evaluate the reasons behind that outcome. First, a locus judgment is made, after which that particular source is evaluated in terms of responsibility and blameworthiness. The assignation of blame to an entity enables the relationship between perceptions of that entity’s justice and subsequent reactions directed at that entity. Throughout this process, the factors highlighted in the above section (c.g., individual and organizational attributes) affect the decisions that individuals will make about locus, responsibility, and blame. A Test of One Part of the Justice Attributions Framework Obviously, the integration of justice theories with attribution theories results in myriad potential applications and implied resultant phenomena. An attempt to verify all of the implications simultaneously would be a vast and costly undertaking. Therefore, the present dissertation will study only a few of the proposed implications. Specifically, I explore recent developments in the construct conceptualization of procedural justice (c.g., Byme & Cropanzano, 2000), as they might be enhanced by incorporating attributions into theory on justice judgments. Byrne and Cr0panzano (2000) note that previous research has confounded the type of justice (i.e., interpersonal and procedural) with the source of justice (i.e., the SUpervisor or some larger conceptualization of the organization as a whole). They discuss how measures of interpersonal justice appear to ask about supervisory attitudes and behaviors, whereas measures of procedural justice tend to refer to the organization itself as an actor, creator, or enforcer of policies and procedures. Byme and Cropanzano report 56 li’ri it pm its: U mgr: was r COIL"; '. ('1) r. .1 KC? | 1'! jutit ’91», lN“\l & '5 }‘ fie-A ““"1 '4‘ 'tr- ,. Mt- the development of four separate direct measures of fairness -- interpersonal and procedural justice perceptions of both the organization and the supervisor. In two studies using separate populations, they were able to demonstrate both the conceptual distinctiveness of their four factors as well as differential relationships with antecedent cognitions and behaviors. Specifically, they found that supervisory interactional justice was related to organizational citizenship behaviors (OCB's) directed at the supervisor and commitment to the supervisor, whereas organizational procedural justice was related to the same organizational level variables. Similar studies have also been conducted by previous researchers. Cropanzano and Prehar (1999) found that procedural justice perceptions were related to organizational-level consequents, whereas interactional justice was related to consequences directed at the supervisor. Both this study as well as Byrne and Cropanzano’s (2000) study show evidence that different ways of measuring fairness perceptions can result in correlations with different consequences. Masterson, Bartol, and Moye (2000) manipulated procedural and interactional justice (through the supervisor) and found that interactional justice related to both reactions toward the supervisor as well as toward the organization, whereas procedural justice did not. Previous typologies (c.g., Greenberg, 1993b) would argue that interactional justice should not have been related to organizational reactions, but procedural justice perceptions should have been. However, according to this manuscript’ 5 framework, it is possible that some individuals could have interpreted the interactional justice as coming from the organization, and some individuals interpreted the interactional justice as coming from the supervisor, leading to both relationships. 57 Pass in WA" mm more 55516 a *2: M It situr Perceptions of procedural justice might not have been attributed to either source (c.g., perhaps some sort of outside influence like a natural disaster), or the lack of the relationship could have been due to the artificiality of the study. Cawley, Keeping, and Levy’s (1998) meta-analysis of voice in performance appraisal systems found results congruous with an attributions framework. They found that the presence of opportunities for value-expressive (i.e., non-instrumental) voice was more related to satisfaction with the supervisor than with the performance appraisal system as a whole. However, instrumental voice was more related to performance appraisal system satisfaction than supervisor satisfaction. Although there are certain to be situations where the instrumentality of voice is a decision made solely by the supervisor, it is more likely that the majority of instances of instrumental voice would be those in which a level higher than one’s supervisor decrees instrumental voice as part of a formal, organization-wide procedure. Thus, instrumental voice was probably perceived to be caused by the organization instead of the supervisor, leading to a stronger relationship between the presence of instrumental voice and satisfaction with an organization—level variable (i.e., performance appraisal system satisfaction). Byme and Cropanzano's (2000) study parallels what promises to be an important deveIOpment in justice research -- that is, the empirical study of attributions and attribution theory into contemporary organizational justice research. Although the precise mechanisms of different attribution theories do differ, attribution theory historically holds that the cause for an event is typically ascribed to a particular individual or entity (c.g., Shaver, 1985; Weiner, 1995). That is, when some sort of outcome is given to an 58 mama baker: jutitt C0151: . w. M Ilbelb; W SM ((3 If) tr, 3}: individual (such as a promotion, selection decision, termination), an individual is likely to believe that one entity was primarily responsible for the event. Separating the source of justice (i.e., supervisor vs. organization) from the type of justice (i.e., interpersonal vs. procedural) has important implications. If it is important to consider the source of justice, then most contemporary scales of measuring organizational justice perceptions (c.g., Colquitt, 2001) are deficient, despite their demonstrated construct validity -- it is possible that prediction of consequent cognitions and behavior may yet nevertheless be enhanced by considering the source of justice. For instance, Colquitt found good evidence that his scales worked in predicting outcomes. However, it is possible that, had his scales and outcome measures been separated into different sources of justice, that the relationships found might actually have been stronger — Colquitt could be underestimating the effects of justice perceptions. Similarly, Byme and Cropanzano (2000) found that justice perceptions of supervisors were related to consequent attitudes toward supervisors, with identical results for organizations. However, within this study (according to attribution theories) there could have been two groups of individuals - those who ascribed locus of causality to the supervisor, and those who ascribed locus of causality to the organization. Thus, the relationships found by Byme and Cropanzano might only have existed for a portion of their participants (i.e., locus judgments may moderate the relationship between fairness perceptions and outcome variables). The Present Study The previous sections of this manuscript discussed many factors that need to be researched subsequently (c.g., organizational characteristics, individual difference 59 tam? tie to: tired so tc‘ indi Witt ;. ., , 'Libilt are: GUM Herd variables, judgments of causality, etc.). The present study examined only a small part of the content discussed above. Specifically, this study examined whether individuals are indeed able to distinguish causality, responsibility, and justice of different possible justice sources in organizational situations. In other words, the framework discussed in Figures 3 and 4 suggests that an individual will be more likely to have negative fairness perceptions and subsequent reactions against the External Source that is judged to be responsible and blameworthy. A few restrictions on this study need to be discussed. First, as this dissertation concentrates on attributions of blame (i.e., an unjust outcome has occurred), in terms of justice, manipulations of procedural justice were the only ones made. Second, whereas in an organizational situation there could be many different potential causes for a negative outcome, only the two most likely (supervisor and organization) were studied here. Third, Heider’s (1958) levels of attributions lists five different levels of responsibility (see Table 1). When one takes into account the fact that both the organization and the supervisor can vary on all five (while simultaneously varying in interpersonal justice and the order in which they are considered by the participant), the number of combinations of manipulations quickly becomes unmanageable. Thus, it was decided to restrict the manipulations to the most interesting levels of responsibility from a justice standpoint - intentionality vs. justifiability (i.e., if the supervisor were merely associated with the outcome, it is unlikely that fairness perceptions would be affected to any geat extent). So, whenever a manipulation of ‘responsibility’ is discussed below, it pertains to actions that are intentional vs. actions that are intentional but are furthermore M. In terms of Weiner’s (1986) dimensions, this study manipulated the locus of causality and the 60 “no: )B§p\ »\ Bu 1 "I c val“, control dimensions (through manipulations of responsibility), leaving the effects of the stability of the event up to future research. Once an organizational decision is communicated, justice and attribution theorists suggest that evaluations are then made. The present study was designed so that one particular source is seen as the causal source (sources will be manipulated between situations). Sources also differed in terms of justifiability across scenarios. If the causal source’s actions are considered justified (i.e., there is nothing the source could have done differently) negative judgments of the source should be attenuated, and the individual would not react as negatively to this source. However, if the source’s actions are not justified, then that source will be considered as blameworthy, and the individual will react strongly to this source, depending on the level of perceived procedural justice from this source. The alternative perspective offered by extant research is that exemplified by Byme and Cropanzano (2000), and can be found in Figure 5. That is, it is still possible that perceptions of responsibility do not moderate the relationship between fairness perceptions and consequent variables, and that individuals simultaneously relate fairness perceptions of supervisors and organizations to their respective consequent variables. In contrast, a model taking attributions into account is presented in Figure 6. This model is congruent with the fiamework developed above, but it is expressed to reflect a statistical path model, as Opposed to the flow of individuals’ thought patterns. The focal variable of the model in Figure 6 is a blame judgment - whether the organization or the supervisor is judged as blameworthy. Whether or not an entity is judged as blameworthy is an interactive function of locus (the entity must be seen as the locus) and level of 61 953995.. .351? 3.53. .335 2: wet—5333 .23.: 52. < .m 3:»:— 533:8qu B 3383a cemmtoasm 2 meozoeom 39.3. 35335 BaouaNEmma 333 35335 Damion—um 62 0332—33.— al.—c.3058...— «uouuae 05 mac—.333.— 3—55 5.:— < .e 3:»:— 3:338qu 8 meouoeom 0283. 3833:— 33333830 3:322:82: \ £3888: £58888 8 .23 Dad—mm 383396 02 m: S 38 8 meouowom 1 .g _. 3 u m befitomsm 63 i633). mitt: subs pres W “1‘1 Elf "\ l a” £0} responsibility (the entity’s actions must be intentional, yet not justified). Blame, in turn, moderates the relationship between each procedural justice perceptions of each entity and subsequent reactions against that entity — an entity must be judged as blameworthy for this relationship to exist. chific Hmtheses This study will formally hypothesize relationships that follow from both the established ‘source of justice’ literature (c.g., Byme & Cropanzano, 2000), and from the present framework. Note that, obviously, both models cannot be correct, so it is realized that some of these hypotheses will not be supported. For instance, Byme and Cropanzano’s model tacitly asserts that the relationship between supervisory procedural justice perceptions and reactions to the supervisor is always present. However, if an attribution of blame is necessary, this relationship does not exist in all situations. First, a series of hypotheses are proposed regarding the traditional ‘source of justice’ model (c.g., Cropanzano & Byme, 2000). In these hypotheses, reactions to justice will be specifically matched to the source of justice. That is, reactions to the supervisor are affected only by supervisory justice perceptions, etc. An additional note here is appropriate regarding what is meant in the present study by ‘reactions.’ There are a number of outcome variables hypothesized to be related to different fairness perceptions (c.g., OCB’s toward the supervisor; OCB’s toward the organization; trust; organizational commitment; job performance; e.g., Cropanzano & Prehar, 1999; Masterson, Lewis, Goldman, & Taylor, 2000). These studies have traditionally asserted that procedural justice Perceptions are related to reactions to the organization, whereas interactional justice perceptions are related to reactions to the supervisor (Byme & Cropanzano, 2000). the i sure: react 0ij proc sup. l hill UllC . lllt Byme and Cropanzano argue that these results were a result of the confound of the wording of the justice measures with the source of justice. They found basically that supervisory interactional justice perceptions were related to all types of supervisory reactions studied (i.e., OCB’s to the supervisor and commitment to the supervisor, in addition to job performance), whereas organizational procedural justice perceptions were related to organizational reactions (i.e., OCB’s to the organization and organizational commitment). These findings may suggest that, depending on the source, a different type of justice may relate to outcomes at that source. However, the multicollinearity between procedural and interactional justice perceptions inherent in their measures (.70 and .71 for supervisory and organizational-level perceptions respectively) may have led to a phenomenon where the authors could have specified either antecedent variable and had that variable explain all of the variance that the other variable could have. Thus, it is yet uncertain which antecedent justice perceptions lead to which consequent variables. For the purposes of this study, consequent variables will pertain to OCB’s and commitment; and in congruence with the attributions framework, procedural justice perceptions are hypothesized to be related to both consequents (i.e., affective and behavioral reactions). H1 A: Supervisory procedural justice perceptions will be positively related to supervisory OCB’s. H13: Supervisory procedural justice perceptions will not be related to organizational OCB’s, above and beyond the effect of organizational procedural justice perceptions. HM: Supervisory procedural justice perceptions will be positively related to commitment to the supervisor. 65 H23: Supervisory procedural justice perceptions will not be related to organizational commitment, above and beyond the effect of organizational procedural justice perceptions. HM: Organizational procedural justice perceptions will be positively related to organizational OCB’s. H33: Organizational procedural justice perceptions will not be related to supervisory OCB’s, above and beyond the effect of supervisory procedural justice perceptions. HM: Organizational procedural justice perceptions will be positively related to organizational commitment. Hag: Organizational procedural justice perceptions will not be related to commitment to the supervisor, above and beyond the effect of supervisory procedural justice perceptions. Next, hypotheses are proposed that follow from the attributions framework outlined in this paper. First, in accordance with the literature on responsibility and blame (c.g., Heider, 1968; Shaver, 1985), whether an action is classified as ‘justified’ as Opposed to merely ‘intentional’ influences ultimate decisions regarding blame. Actions that are justified indicate that the target individual does not deserve to be blamed for the outcome. However, in keeping with the concept of the locus of causality, this relationship should only be present when the entity in question (i.e., supervisor or organization) is seen as the causal locus. H5: Locus of causality judgments will moderate the relationship between the ‘justifiability’ of a source’s actions and the blame assigned to that 66 trim COUC ill: ' SO 'li i source, such that I) blame will not be assigned to a source that is not the causal locus, and 2) blame will be assigned to a source that is the causal locus if the action is not justified. Finally, I may now discuss the hypothesis that establishes the critical difference between the present framework and the ways in which justice has previously been conceptualized. On the basis of previous theory in the attributions literature, it is probable that perceptions of procedural justice will be more strongly related to reactions to the source of justice if that source is regarded as blameworthy in that situation. That is, the perceived fairness of procedures that are used by a supervisor will only be related to reactions against that supervisor in the event that the supervisor is seen as blameworthy for a particular outcome (or series of outcomes in a more general case). The same argument exists for the organization - the perceived fairness of an organization’s procedures will relate to organizational reactions only if the organization is seen as blameworthy. H6: For supervisory and organizational sources of justice the blame attributed to a source will moderate the relationship between perceptions of that source’s justice and reactions to that source, such that a weaker or nonexistent relationship will be found for sources not to blame for the outcome. Hypotheses 1-6 were tested using a scenario methodology involving a promotion decision in an organization (see Part 1 below). Additionally, the ‘essence’ of Hypothesis 6 was tested in a different way (see Part 2). In more general terms, Hypothesis 6 states that, for a particular negative outcome, 1) individuals distinguish between sources of 67 jut; a sou react Hp; supc ion ('5’ (£5 justice, 2) individuals assign blame to one source or another, and 3) the blame assigned to a source moderates the relationship between justice perceptions of that source and reactions to that source. Part 2 of this study tested the more general implications of Hypothesis 6. Participants were asked to respond regarding a recent parking ticket they received on campus, with respect to blame, justice, and reactions ascribed to two sources of justice in that situation (i.e., the parking employee who issued the ticket, and the Department of Police and Public Safety, who is responsible for parking regulations and enforcement on campus). These two sources of justice are not perfectly analogous to a supervisor and an organization, but they are still two distinct sources of justice, allowing for tests of a more generally-worded Hypothesis 6. PART 1 Method Participants Participants were recruited through the subject pool at a large Midwestern university, and received course credit for their participation. A total of 422 participants provided data for the present analyses. Stimuli were presented and data were collected over the World Wide Web (the psychology department’s web server housed the ASP pages presented to participants, and collected and archived the data). Participants had the option of participating from any internet-connected computer, regardless of the physical location of the computer. The average age of participants was 19.3 years, with a standard deviation of 1.29 years (three participants were omitted from the computation of average age, as they indicated their age as ‘24 or older’). The sample contained 335 White individuals (79%), 22 African-Americans (5%), l7 Asian-Americans (4%), with the 68 rem“ the 5 “fit til par. indi pro. (31 1A3 I)" ( remainder of participants indicating other races or not responding to the question at all. In the sample, 310 participants indicated that they were female (73%), 92 indicated that they were male (22%), and 20 individuals did not respond to the question (5%). Participants were asked several questions to assess their representativeness and familiarity with the stimuli. Two hundred twenty-five participants (53%) indicated holding either a full or part-time job at the time of their participation in the study. Two hundred thirty-two participants (55%) indicated that they would be looking for a job within the next three months. When that time frame was extended to nine months, 283 participants (67%) reported that they intended to be looking for a job. Only 81 individuals (19%) reported that they would not seek a job within nine months of their participation in the experiment. Participants were asked several questions regarding their experience with promotion decisions. A majority of participants (230; 55%) reported ever having been up for a promotion in a job. Of these, 115 (50%) had been denied a promotion at least once, and 224 (97%) had received a promotion at least once. Procedure This part of the experiment used a 2 (Organizational or Supervisor Locus of Causality) x 2 (Organization’s Actions were Intentional or Justified) x 2 (Supervisor’s Actions were Intentional or Justified) between-subjects design. A series of eight vignettes were created and pilot-tested to ensure that they represented their intended manipulation (vignettes can be found in Appendix A). Each vignette asked the participants to imagine that they were up for a promotion at work. All vignettes involved both a supervisor, as well as a committee comprised of members of upper management (representing the organization). In scenarios in which the supervisor made the ultimate promotion decision, 69 in: mil id: [ll the upper management committee provided a promotion recommendation to the supervisor, and the reverse was true in vignettes in which the upper management committee made the ultimate decision. The description of the promotion process stressed that the decision-making entity was completely free to disagree with the recommendation without any consequences. After participants read a scenario, they responded to a series of items assessing their perceptions of causality for the denial of promotion, as well as the intentionality, justifiability, and blame for each source involved. Following that, participants rated their perceptions of the fairness of each of the sources involved in the scenario. Finally, participants completed measures of affective and behavioral reactions to each source in the scenario. Measures The following measures were used in this experiment. A listing of the items identified by scale can be found in Appendix B. An example of the complete experiment as participants saw the stimuli and questions can be found in Appendix C. Locus of Causality. Causality for the decision (i.e., the supervisor or upper management committee makes the decision) is one of the between-subjects manipulations; it varies across scenarios. However, 1) in order to assess whether participants noticed the manipulation, and 2) to allow for situations where participants assign causality for the event in a way not intended by the manipulation (i.e., participants may attribute causality equally to more than one source), items assessing participants’ perceptions of causality for the decision were written (see Appendix B). For this variable, separate one-item measures were created to assess whether the participant were perceived as the cause of the promotion decision, whether the supervisor 70 \l‘fli supr pm \i’hi lllit' l0 were perceived as the cause, and whether the upper management committee were perceived as the cause. Additionally, another item asked participants to choose among themselves, the supervisor, and the upper management committee regarding the most direct cause of the promotion decision. Intentionalig and Justifiability. Similarly to locus of causality, whether the supervisor’s or upper management comrnittee’s actions were intentional or justified was manipulated between subjects. Again, however, items were included to assess participants’ perceptions of these manipulations, as well as to account for instances in which participants judge intentionality and justifiability based on factors other than the intended manipulation. Three-item measures were created to assess participants’ perceptions of the intentionality and justifiability of both the supervisor and the upper management committee. These items can be found in Appendix B. Elm. Again, blame was intended to be manipulated as a between-subjects variable as a interactive function of causality and justifiability (i.e., a source would be held as blameworthy only when the source were simultaneously judged as causal and intentional, yet not justified). However, participants’ perceptions of blame were also assessed. One-item measures were created to assess whether participants judged themselves to be to blame, whether the supervisor were to blame, and whether the upper management committee were to blame. Furthermore, a separate item required participants to choose between themselves, the supervisor, the committee, or ‘nobody’ as the source that should receive overall blame for the promotion decision. These items can be found in Appendix B. ltitd jllSIlt fem pnc (rel: din lllt’l p5}.- by ' Si Organizational Justice. Organizational justice measures were drawn from those used by Byme and Cropanzano (2000), which drew from several previously published justice studies. Byme and Cropanzano used 4-item measures to assess perceptions of formal procedural justice (referred to as procedural justice below), informational procedural justice (i.e., informational justice), and social/interpersonal procedural justice (referred to as interpersonal justice below). This 12-item scale (i.e., four items per dimension), was written once in a way to pertain to the supervisor, and once to pertain to the organization overall, for a total of 24 items. Byme and Cropanzano report adequate psychometric characteristics for this scale, as derived from an applied sample. Items can be found in Appendix B. Citizenship Belrflor. Measures of citizenship behavior were similar to those used by Byme and Cropanzano (2000), and are based on Williams and Anderson’s (1991) scales of OCB. Williams and Anderson wrote items designed to tap OCB’s that benefit the organization (OCBO) and OCB’s that benefit an individual (OCBI). As originally written, these items were designed to be completed by an actual employee’s supervisor. Thus, for the present experiment, these items were rewritten to reflect hypothetical behavior, or intent to commit such behavior, as reported by the individuals themselves. Items can be found in Appendix B. Commitment. Measures of commitment to the organization and to the supervisor were drawn from some items used in Allen and Meyer’s (1990) Affective Commitment Scale. The items were rewritten in the present study to reflect hypothetical commitment, and can be found in Appendix B. 72 mod gnu an] Res; 3336 que: 0m res; COT. COT. 41: as; Power Analysis According to Cohen (1988), the sample size necessary to detect a ‘medium’ moderating effect (i.e., difference in correlations) at a power of at least .80 is 140 in each group, resulting in a minimum necessary sample size of 280 participants. The present sample exceeded this requirement. Results Manipulation Checks After reading the scenario, participants responded to two questions designed to assess to what extent participants fully read and understood the scenario. The first question asked participants whether the result of the promotion decision were favorable or not (i.e., whether they received the promotion). A total of 19 participants (4.5%) responded to this question by indicating that they received the promotion, although no scenario used a positive outcome. The second question asked participants to indicate who had ultimate control over the promotion decision (i.e., whether the supervisor or the upper management committee made the decision). Of the 217 individuals whose stimulus indicated a supervisor decision, 61 (28%) indicated that the upper management committee made the decision. Of the 205 individuals who were in an upper management committee decision condition, 32 (16%) mistakenly indicated that the supervisor made the decision. The error rate for individuals in the supervisor decision condition was significantly greater than that for the upper management committee decision (x2( 1 , p = 415) =129.49, p < .01). That is, there appeared to be a trend for individuals to ascribe the decision power to the upper management committee even when the committee was not actually responsible for making the decision as described in the scenario. 73 The results of these manipulation checks indicated that participants did not uniformly perceive the scenarios as intended, although there was still good evidence that the manipulations were largely successful. As a result, analyses of the data proceeded, with the following exception: individuals who failed the first manipulation check (i.e., they indicated that they received the promotion) were dropped from all subsequent analyses, resulting in a total sample size of 402. It was suspected that these individuals did not sufficiently understand the scenarios (due either to participant motivation or other factors) to provide meaningful data in the rest of the experiment. However, individuals who failed the second manipulation check were not necessarily deleted from the analyses unless they also responded incorrectly to the first question. The fact that these individuals were retained in the analyses may limit the extent to which the manipulations themselves have significant effects on the dependent variables in this analysis. However, the theory in this dissertation argues that individuals’ perceptions of causality and justification will relate to blame, regardless of the objective facts. Thus, these individuals may still provide meaningful data if causality is analyzed in terms of participants’ perceptions and not as a result of the manipulation itself. The analyses described below used perceptions of causality as the operationalization of choice, allowing for the retention of participants who responded incorrectly to the second manipulation check variable. (When all analyses were rerun using only participants who responded correctly, very few differences were observed.) Psychometric Characteristics of Items and Sca_le_s All items were examined for their psychometric characteristics prior to their aggregation into scales (when applicable). Below, each scale will be discussed in terms of 74 its psychometric characteristics. Unless otherwise specified, all factor analyses discussed below were conducted in the ‘exploratory’ sense, using squared multiple correlations as initial estimates of communality. Promax rotations were requested for each analysis when applicable. Exploratory factor analyses were conducted using SPSS 10.0, and all other analyses were conducted using SAS 8.0. Organization Justifigbility and Intention_a_li_ty. The three items representing perceptions of the upper management committee’s justifiability formed a single factor. The internal consistency reliability of this scale was .85. The three items representing perceptions of the committee’s intentionality also formed a single factor, and had an internal consistency reliability of .87. A factor analysis of all six items together produced a two-factor solution reflecting a distinction between perceptions of intentionality and justifiability. The factors were correlated -.48 with each other. Supervisor JMility afnd Intentionality. The results of the items for the supervisor mirrored those for the organization. Supervisor justifiability and intentionality scales had internal consistency reliabilities of .81 and .83, respectively. Additionally, a factor analysis of all six items together showed a two-factor solution appropriate to the distinction between the constructs, with an intercorrelation of -.57. Additionally, all 12 justifiability and intentionality items were subjected together to a factor analysis. Three factors had eigenvalues greater than one. An interpretation of the standardized regression coefficients for these items showed that the first factor was comprised of all six supervisor variables, with negative correlations for the justifiability items and positive correlations for the intentionality items. The second and third factors were comprised of committee justifiability and intentionality items, respectively. Another 75 factor analysis forcing a four—factor solution produced a factor loading pattern corresponding to the distinct scales. Organizational Fairness. Within-scale factor analyses for procedural, informational, and interpersonal Justice all showed evidence for unidimensionality within each factor. A factor analysis of all 12 items together showed three factors with eigenvalues greater than 1.0, and a scree plot that also indicated a three-factor solution. An analysis of the factor loadings was consistent with the dimensions intended by Byme and Cropanzano (2000). The internal consistency reliabilities of the three scales were .74, .86, and .80, respectively. Supervisor Fairness. The factor analyses for the supervisor justice items showed a two-factor solution using both a scree plot analysis and a consideration of factors with eigenvalues greater than 1.0. The items from Byme and Cropanzano’s (2000) procedural and interpersonal justice scales formed the first factor, and the items from the informational justice scale formed the second. Despite the results of these analyses, it was decided to create composites for the three scales as proposed by Byme and Cropanzano. The internal consistency reliabilities of the supervisor procedural justice, informational justice, and interpersonal justice scales were .82, .91 , and .84, respectively. Supervisor and Organization Fairness Con_s_idered Simultaneouily. Additionally, a factor analysis of all 24 items (supervisor and organization) showed results indicating five factors with eigenvalues greater than one, and a scree plot showing five factors. An analysis of the five-factor solution revealed that the first factor was comprised of Informational Justice (both supervisor and organization), the second factor had items related to Interpersonal Justice (both supervisor and organization), the third and fourth 76 factors corresponded to Supervisor and Organization Procedural Justice, respectively, and the fifth factor appeared to be a factor comprised of Items 9 and 12 from both scales. Additionally, a series of confirmatory factor analyses (CFA’s) was conducted using SAS PROC CALIS. A test of the scale as proposed by Byme and Cropanzano (2000) produced fit statistics that were less than optimal 08046, p = 303) = 1187.71, p < .05; RMSEA=.090; GFI=.80; CF I=.84; TLI=.82). Allowing the error terms of items with similar wordings (i.e., the same item rewritten to refer to a different source) to covary resulted in an improvement in the fit of the model (A1202, p = 303) = 291.31, p < .05; this model had a {(234, n = 303) = 896.40, p < .05; RMSEA=.088; GFI=.82; CFI=.86; TLI=.83. A three-factor model (i.e., not distinguishing between supervisor and organization sources) produced fit statistics that appeared worse than either six-factor model 08049, p = 303) = 1227.62, p < .05; RMSEA=.114; GFI=.69; CFI=.79; TLI=.77). The only type of j ustice perception included in this study’s hypotheses was procedural justice. Thus, a series of CFA’s can be performed analyzing just these items. A CFA of the eight procedural justice items (i.e., four supervisor and four organization items) showed good evidence of fit (7809, p = 397) = 115.74, p < .05; RMSEA=.113; GFI=.94; CFI=.92; TLI=.89). Allowing the errors of items with identical wording (except for the source of justice) to covary resulted in even better fit (x20 5, p = 397) = 15.02, p > .05; RMSEA=.002; GFI=.99; CFI=1.00; TLI=1.00; Ax2(4, p = 397) = 100.72, p > .05). Conversely, a CPA modeling these eight items as a single latent factor appeared to fit the data less well, even after allowing the error terms of items with identical wordings to covary (x206, p = 397) = 80.85, p < .05; RMSEA=.139; GFI=.91; CFI=.90; TLI=.83). On the basis of both the exploratory and confirmatory factor analyses, as well as a priori 77 theory, it was decided to retain the procedural justice scales (both supervisor and organization) as originally designed by Byme and Cropanzano (2000). Organizational OCB’s and Commitment. The scales measuring participants’ reactions to the organization were analyzed both separately and together. Separately, factor analyses on each scale showed evidence of a single factor explaining the common variance among the items. The organizational OCB’s scale had an intemal consistency reliability of .78, and the organizational commitment scale had a reliability of .70. When the items reflecting organizational reactions were analyzed together, a two-factor solution emerged that reflected the two constructs. Smervisor OCB’s and Commitment. The scales measuring participants’ reactions to the supervisor were analyzed in a manner identical to that for the organizational reactions. Separately, both scales showed evidence of unidimensionality on the basis of a factor analysis. The internal consistency reliability of the commitment to the supervisor scale was .76, and the reliability of the OCB scale was .85. Furthermore, a factor analysis of all reactions items was performed (i.e., with supervisor and organizational OCB’s and commitment together). A scree plot showed evidence of either a two or four-factor solution. When a two-factor solution was requested, all of the OCB items loaded on one factor, and the commitment items loaded on the other. A factor analysis forcing a four-factor solution showed most of the commitment items (both organizational and supervisor) loading on a single factor, and the supervisor and organizational OCB items loading on the second and third factors, respectively. The fourth factor was comprised of items 2 and 4 from both the supervisor and organizational commitment scales (these items were shown to have low item-total 78 correlations in the within-scale analyses). As the items in the supervisor and organizational commitment scales had essentially the same content (i.e., the only difference in the item wordings pertained to the organization/supervisor distinction), a single factor was not surprising. Additionally, a CF A performed on all organizational and supervisor reactions measures (i.e., OCB’s and commitment) yielded fit statistics that were less than ideal (38699, p = 301) = 1019.60, p < .05; RMSEA=.072; GFI=.81; CFI=.81; TLI=.79). When the errors of similarly-worded items of the supervisor and organization commitment scales (i.e. the items that had identical wordings except for the referent) were allowed to covary, the fit of the model improved (x2091. o = 301) = 813.08, 2 < .05; RMSEA=.060; GFI=.84; CFI=.87; TLI=.85; sz(8, o = 301)=206.52, p < .05). Correlations Among Items ind Sca_les. Composite scores for each measure were formed by taking the average of the items within each scale. The means, standard deviations, and reliabilities of, and intercorrelations between, all scales as well as the single-item measures can be found in Table 3. The intercorrelations among supervisor justice scales and among organizational justice scales were fairly high, as was expected given the results of the factor analyses, and as was evidenced by Byme and Cropanzano (2000). Furthermore, the intercorrelations between specific types of justice pertaining to the supervisor and to the organization were also high (c.g., the intercorrelation between the supervisor interpersonal justice and organization interpersonal justice was high). Justice measures were largely correlated with reactions measures, with the correlations appearing to be larger within each source (i.e., the supervisor and the organization). 79 ...m3=3:3 23... 8.- mm. "m. 8. 8. 8. 8. 8.- 8.- 8. 82 8m Mme-w mm .:.- 2. 2.- 8. 8.- S. 8. 8.- 8.- 82 8.2 88 8.85820 .8 2.- :8. .. .- :8. :8- :8. .2.- 8.- 8. 8. 83 8... 8888883 . .82 .2 .3- . :2 :_N :2.- ..8. 8.- 8.- 8. 8. 8.". .8 8.39888 0 . :8 .- :2. .:.- .:. :8: :8. 8. S.- 8. 8. 88 88 8888888288888 .2 ..o2.- 8. 2H- .2. :8: :8. 8.- :8.- 8. 8. 2.,“ 8m 88388888 .2 : N. 8 .2 - :8. :2.- . 8. 2.- 8.- 8.- 82 8.8. 88 82:8388285888820 .2 :8.- 8M 8.- :8. :2.- .Hmm. :2.- 8.- 2.- .2 88 8e 82:8888885888820 2 :8.- 8 :2: :8. :3.- . R. 2.. 8.- .. 8. 82 88 88 82:8 33888 8:88.20 .3 :mm.- 3. 3.- tom. :8: .am. . S. 3..- mo. 3.. mm.m mom 8383.3 338232282235 .2 .._m.- 8.- 8.- :2. :8.- ..S. 8.- :2.- 2.- 8.. 83. 88 82:838888588888 .2 :8” 8. .:.- :8. :8.- H.8 8.- :2; :8.- 3.2 o: 8.. 82:8 388888388 .: :8 8. :8. :5..- 8. :8.- .8. : S. 8.- :2 88 88 283 88888 .2 S- 8. :8.- ..8. :8.- ..2.- :8. :2. N2 28 88 28338288 .8 8.- 8. 8.- .:. 8. 8.- :3. 8. 8 88 828.8 .8 mm. :9... :3. :mm.. :2. mo. 8.. om. 32m 3m 23335353336800 .n ma. :omr :2... :omc :2.- mc. 3. 3:8. 3m 33833930338380 .3 mm. :8... Br :2. co. 3. Nm.m mam 933353282235 .m S. S. .._m.- 30.- 3. and mom :o38u33msfommtoqnm .v :8.- 8.- 82 88 8m 3:888 82888 .m .:. 82 88 88 288888388 .m 8. .2 8m 8.2.828 ._ o w \- 3 w v .m .m ._ Om Z Z c.338> 80 .3.3w8> otmcoom 23 5-3 3333333— 283330 3525 38 303838383: mco38_>cQ 383.85 2882 m 23h .oBEnN :5 2:82“: 85 :28 3:8 83 xom 82:8 Ema-m :o :83 Ba 8530:: :05: E: 98 .8288 “Eon-h :o :83 2m 8588: 82:3 =< 63:23:: :23, 2:536 05 :m 2: 825523: 8:88:00 3:8:— ”302 Eva: 3v:- 2. we: .2; we. 8.- S. 8.- no. oo. 8. mo. mo. xamfim .. 5. 8. we. we. we. 8. mo. 8. 3. 2.. ac. 0978 on. Gown. Itoca Gbmm. 995v. OOWN. 06mm. §O@m. OCVN. 0mm. Goo—.l HG“EHmEEOU BayaOmHmNEWHO .0“ MR. :2. :mm. :mm. :8. :8. :3. :2. .vm. :Sr Mmoo Ecoamw_:awuo.o_ on. :3. :om. :3. :VM. :3. :3. .Vm. .:.- Sarcasm 2.: 9608:8500.»— 8. :8. :8. :8. :2“. :8. .3. 8.- mmuosmmtoasms. ow. :3. :6. :3. :3. .8. :mNr 82:3 83208023282830.9— 0w. 2.9%. oobm. :Nh. .mV. «LVN-u mmoemd-w _NGOmme.~O...—dm £053?“qu .m~ 3.. : _ m. :9. .3. : _ m: 82:3 35:82: 5:82:90 .3 3. :mm. .:. :Nmr 82:3 8:082:22 Ems-5:3 .2 — a. .NO. 3% ~ .: mmuemfl Ecomumgomcm uOmWP—asm .N— N». :8: 82:3 85:82: 885:5 .: 8 cm 2 M: S E 2 3 m. 2 2 S g . “yum—HOD «M 0~DNIH 81 A Word of Caution Regarding the Scales. While some of the measures discussed above exhibited evidence for good fit, others did not. The most substantial problem with the measures was the lack of clear divergence between items measuring the same construct (e.g., justice) applied toward the supervisor and the organization, especially when the items are worded identically except for the subject of the item. Although it could be argued that one would expect these constructs to be correlated, the empirical data do not allow one to distinguish between correlations due to conceptual similarity and correlations due to conceptual convergence. In other words, there is less than optimal evidence to support the idea that participants made clear distinctions between their perceptions and judgments of the supervisor vs. their perceptions of the organization. For the purposes of this dissertation, it was decided to proceed with the scales as originally intended, and analyze the hypotheses originally proposed, and interpret the analyses with the assumption that there is conceptual divergence between supervisor and organizational perceptions. However, it should be stressed here that the lack of evidence for conceptual convergence may substantially reduce the validity of the interpretations drawn in this dissertation. Relationships with Demographic Variables. As shown in Table 3, age was not correlated with any of the measures, and sex was weakly correlated with only one variable (organizational commitment; men had slightly lower commitment scores overall). Additionally, other demographic variables were correlated with other measures and manipulations as discussed below. Age. Age was related to the forced-choice item regarding causality. Specifically, the odds of choosing the upper management committee to be causal over the supervisor 82 increased with age (likelihood ratio x2 (1; o = 355) = 8.33; a < .01; B} = .023). Similarly, the odds of choosing the committee to be to blame over the supervisor increased with age (likelihood ratio x2 (1; a = 253) = 8.25; p_ < .01; 32 = .032). Current Employment. An item measuring whether participants were employed at the time of measurement was related to both organizational (E(2,368)=3.09; p < .05; Egon) and supervisory oca’s (£(2,370)=4.95; p < .05; Rz=.026). In both cases, post- hoc tests showed that individuals who reported holding a part-time job had significantly stronger intentions to commit OCB’s than did individuals who were not currently employed. Looking for a Job. The item asking participants to indicate a time frame in which they intended to look for a job was related to the one-item measure asking about the extent to which participants blamed themselves for the decision (E (3,339) = 2.91; Rz=.025). Post-hoe analyses showed that individuals who chose “in the next 6 months” for their response placed more blame on themselves than did individuals responding with other Options. Experience with Promotiog. Individuals who had indicated that they had previously been up for a promotion at work were likely to indicate stronger agreement that the upper management committee was the causal influence on the decision (H 1,3 86)=5.46; p < .05 ; 322014), as well as more likely to agree that the upper management committee were to blame (E(1,386)=4.47; p < .05; 32:.012). Responses to this question (i.e., whether they’ve been up for a promotion at work) were also related to perceptions of organization procedural justice (15(1 ,369)=6.16; p_ < .05; 32 = .016), and organization informational justice (H1, 370)=4.21; P. < -05; 32:01 1)- In both 68868, 83 individuals who had been up for a promotion at work had significantly less positive perceptions of justice. Another experience question asked whether, if the individual had been up for a promotion, they have ever been denied a promotion. This item was related to perceptions of upper management committee causality (Ij(2,383)=3.15; p < .05 ; 32-3016), and post- hoc tests showed that individuals who had ever been denied a promotion agreed more strongly that the committee was causal as compared to individuals who had never been up for a promotion. Hypothesis Test Results Hymthesis 1a. Hypothesis la specified that supervisory procedural justice perceptions would be positively related to supervisory OCB’s. As shown in Table 3, the zero-correlation between these two variables is significant and positive. This hypothesis is supported by the data. Hypothesis lb. Hypothesis lb states that the correlation between supervisor procedural justice and organizational OCB’s will be zero after controlling for the effect of organizational procedural justice. Table 4 shows the results of hierarchical regressions, testing whether supervisor procedural justice added to prediction of organizational OCB’ s over and above the variance predicted by organizational procedural justice. This hypothesis was not confirmed; supervisor procedural justice added to prediction of organizational OCB’s above that predicted by organizational procedural justice. The standardized regression weight for the supervisor justice measure is positive but small. However, the correlation between supervisor procedural justice and organizational 84 Table 4 Results of Hierarchical Regressions of Organizational OCB’s on Organizational and Supervisor Justice. Model R2 AR2 0 1r Step1 Organizational procedural justice .0321r.032" .18 Step 2 Organizational procedural justice .05 fit t’ it Supervisor procedural justice .055 .023 .20 " p < .01 OCB’s is lower after controlling for organizational procedural justice, thus showing partial mediation. Hymthesis 2a. This hypothesis mirrors Hypothesis la, but is directed at commitment to the supervisor. As shown in Table 3, a positive zero-order correlation exists between commitment to the supervisor and supervisor procedural justice, supporting this hypothesis. Hypothesis 2b. Hypothesis 2b mirrors hypothesis 1b, but deals with organizational commitment. Table 5 shows the results of hierarchical regressions exploring the predictive power of supervisor procedural justice in predicting organizational commitment over and above organizational procedural justice. Results show that, contrary to hypotheses, supervisor procedural justice added to prediction over and above organizational procedural justice. The magnitude of the relationship between supervisor justice variables and organizational commitment does decrease, however, when the effects of organizational procedural justice are partialled out. 85 Table 5 Results of Hierarchigal Regression; of Organizational Commitment on Organizational and Supervisor Justice. Model R7 ARZ ‘3 It fir Step1 Organizational procedural justice .125" .125 .36 Step 2 Organizational procedural justice .21 O. O! it Supervisor procedural justice .153 .028 .22 Wp < .01 Hymthesis 3a. As shown in Table 3, organizational procedural justice does exhibit a small, yet significant, positive relationship with organizational OCB’s. Thus, this hypothesis is supported. Hyppthesis 3b. Although a zero-order correlation exists between organizational procedural justice and supervisor OCB’s (see Table 3), this relationship does not exist afier supervisor procedural justice is partialled out (see Table 6 for the results of these analyses). Thus, Hypothesis 3b was supported. Table 6 Results of Hierarchical Regzessions of Supervisor OCB’s on Organizational and Supervisor Justice. Step 1 SuPervisor procedural justice .1 13" .1 l3" .3 3" Step 2 Supervisor procedural justice .34" Organizational procedural justice .121” .008 .02 flip < .01 86 Hypothesis 4a. As shown in Table 3, organizational procedural justice is positively related to organizational commitment, thus lending support for this hypothesis. Hyppthesis 4b. Table 7 shows the results of hierarchical regressions exploring for the increment in prediction of supervisor OCB’s by organizational procedural justice over and above what has been predicted by supervisor procedural justice. As shown, organizational procedural justice did not add to the prediction of supervisor OCB’s once the effect of supervisor procedural justice has been controlled for. Thus, the results support this hypothesis. Hymthesis 5. This hypothesis stated that the blame assigned to a particular source (c.g., the supervisor) was the result of an interaction between the causality attributed to that source and the justifiability of the source, such that a source will be held as blameworthy only if seen as causal but not justified. Such attributes (i.e., causality and justification) were manipulated between subjects as part of the design of the study. Therefore, blame (as hypothesized) was also manipulated (i.e., certain conditions were designed to have the supervisor blamed, others the organization, and in still others neither Table 7 Results of Hierarchical Regzessions of Commitment to the Supervisor on Organizational and Supervisor Justice. Wodel 32 AR2 0 —STep 1 Supervisor procedural justice .295" .295" .54" —STep 2 Supervisor procedural justice .57" t. Organizational procedural justice .300 .005 -.02 T‘p < .01 87 one was to be blamed). However, items within the surveys also assessed the degree to which participants perceived causality, justifiability, and blame. Therefore, the extent to which the hypothesized relationships match how individuals actually go about assessing blame can be assessed in a number of ways in the present experiment. Below, many of these different assessments will be examined. Blame, as a dependent variable, can be assessed using a one-item measure of supervisor blame (Question 20 in Appendix C), a one-item measure of committee blame (Question 21 in Appendix C), or an item that forced participants to choose between different possible sources of blame (Question 22 in Appendix C). Furthermore, causality and justifiability can be analyzed as a result of the manipulations, or through several variables that assessed participants’ perceptions of these manipulations. Results of the most relevant of these combinations will be reported below. All of the relationships were analyzed using a series of hierarchical regressions. First, the blame items were regressed independently on a variable assessing causality, justifiability, or intentionality. Then, the next step was run in which both independent variables were entered simultaneously. A final step was run in which the interaction between the causality and justifiability/intentionality variables was included. An overview of the results of the regressions for the supervisor blame item and the committee blame item can be found in Tables 8 and 9, respectively. Given the number of analyses run to evaluate hypotheses 5 and 6, results will be reported to distinguish between results that are found at both a .05 alpha level, as well as a more stringent criterion for alpha (.01). Supervisor Blame. See Table 8. The three Operationalizations of causality (manipulated supervisor causality, perceived supervisor causality, and a choice between 88 Co. v 3 .550 05 uo>o «830:6:— E 358805 accommcwfi E :32 33ers, 53 can .Co. v 9 28$:me o3 gammy—wee =e < S. v m .. 3. v Mr omc. EN. 3m. 3 _. of. b:m=ou=8£ hemmtomsm 328qu :o 8. :com. :52”. :53. Wow —. zoueownmi. uofltoasm 333qu Snows 338 Heme. “mam. HMS. Hog. .HeE. 85:95.: Susana. deacoeem co 865 :e5. :5. Lem. :2 _. Lam. £22225 Become... 8285 33:8 8c. tome. :mmv. :$N. :wom. 558563 uofltoasm “.3680.” uofltoaam .moo. :wmm. L .2. 1:25. Lam. 85:35:. cozoocfia Sarcasm .6 8:828 ..e8. :32. :e2. :3 _. :35. 9.82225 someone eozoaoe woo. :3 m. :wom. :SN. :98. eouaomumi Same—095 eozoouom core—age: to 8. Limoo. tmno. :35. Emma. cone—:ESWE coreomcmsm hie—25m =2£8Q couofioWE 5588mm $032“th one—e Banger one—m oBetg 03ers.» Show? 8% mm Son .8.“ m gagging co.“ 1%. bzmmzmo com «m 2235? 5:852... bzemseu 558553. was £9.80 .«o 83865 so So: ease—m Sarto :m 05 no acumen cm 05 90 333% w 2an 89 Co. v 8 850 23 88 8303008 8 $88088 8858me 8 .330.— 83838 503 38 .Co. v 3 885:me 0.8 383882 =m < #0. v m .0 8. v at moo. :wi. :33. :oo_. :wmo. 53803838 858800 3038qu ooo. :53. :wnm. :mmm. :39 838533 858800 828qu— 3808 33:8 ooo. L05. 5. 53 Lk. mo. homo. 8328388 838533 858800 .30 88:0 ooo. :08. :03. :oo _. :5. _. 8.80388 858800 8388a 83388 moo. :03. :m mm. :mmm. :35 3 838538 858800 858.com 858800 ooo. LLom. LL om. humo. hon _. 8323888 838533. 858800 .30 838083 moo. :33 :o— _. :oo_. :omo. $38388 858800 8588a coo. :mmm. :mvm. town. :95. 838533. 858800 305089" 8338388 voo. :Ko. tooo. Limo tomo. 883888 838533. 858800 8385 830801.55 830838 3.32838 08? 0.838. 083 2838 0.883 8.3 $2 83 mm. 88 8.3 M. 838533. 83 1% 338:8 83 M 2838 838533. 8.880 838533. 38 3880 .30 898305 :0 80: 0805 0058800 23 .30 8038 cm 0.3 .30 5:8»— 0 030,—. 9O the supervisor and the committee) had significant main effects on perceptions of supervisor blame. When the supervisor made the decision or was perceived as more causal, more blame was ascribed to the supervisor. Additionally, the three justification variables (manipulated supervisor justification, perceived supervisor justification, and perceived supervisor intentionality) were related to supervisor blame in terms of zero- order correlations. Supervisor blame was negatively related to perceived supervisor justifiability, so that more justification led to lower blame attributions. Supervisor intentionality perceptions were related positively to supervisor blame — more intentional actions led to greater blame (note that the manipulations of justifiability were related in the same direction as the perceptions). The interaction between manipulated causality and manipulated supervisor justification was significant, resulting in an increase in 132 of .019 (p < .01). A graph of this interaction can be found in Figure 7. As hypothesized, blame attributed to the supervisor is more likely when the supervisor was causal (i.e., the supervisor made the promotion decision), and when the supervisor was not justified. The interaction between perceived supervisor causality and manipulated supervisor justification was also significant (A32=.008; 2 < .05). A graph of this interaction can be found in Figure 8 (for all figures using continuous measures, note that the points graphed represent points one standard deviation above and below the variable’s mean). Perceptions of supervisor causality were related to blame, such that individuals with stronger supervisor causality perceptions had greater perceptions of supervisor blame, but that relationship was stronger when the supervisor was not justified in his 91 Barton: coax—5&5“:— uca bzamaao toga—5&5:— .—c 552:: a 3 0:33 homer—unam .h «.:.»:— €2£=E§ZV ceflouo .8359“ 32.5500 ‘! I -‘Ulllllll. Ili-.i4 BEE: .353 |¢I BEE 32 5.3.5 l9: ,‘ m.— T Wm m6. né aumlg Josyuadns 92 5:55;...— ..8_>._un=m 5225352: 15. 553.53 53.33:...“ 1923.8: .3 552:... a 3 can... ucnmtoaam .w 9.53 5:830 Sflteanm 83020.n— gwi 33 ’1 I k . I ‘ ‘kl‘ l- ‘Lz .l w. ‘l‘ l\ , 10‘ ‘III l,ll I‘ll i‘ l .I - ‘ mofimmslmhmmmtitomflwf? 5 ,.- Bug“... 8 Z Sarcasm + ,, ~ _ 1 V? N M V‘. m 9111913 Josyuadns 93 actions (as manipulated between subjects). As with the other interactions in this section, this result supports Hypothesis 5. For the next few analyses (i.e., ‘choice of causal agents’ in Table 8), causality is measured using an item that forced participants to choose the source that is most causal (between themselves, the supervisor, and the committee; see Question 6 in Appendix C). In total, 21 participants chose themselves to be the most causal (5%), 156 chose the supervisor (39%), and 210 chose the committee (52%). This variable was included in the next 6 analyses as a class variable in SAS PROC GLM (where it functions similarly to a set of dummy-coded variables in a regression). Chosen causality and the manipulation of supervisor justification interacted in predicting supervisor blame (A32=.030; p < .01). See Figure 9 for a graph of this interaction. The manipulation of supervisor justifiability was related to supervisor blame only when the supervisor was chosen as the causal agent, again lending support for Hypothesis 5. Chosen causality and perceived supervisor justification also interacted (A32=.Ol 9; p < .01). This interaction is graphed in Figure 10. Perceived supervisor justifiability was not related to supervisor blame for participants who saw themselves as causal, but was for individuals who saw the supervisor or committee as causal. A regression limited to individuals who chose only the supervisor or committee as causal still resulted in an interaction, indicating that the relationship between perceived supervisor justifiability and supervisor blame is stronger when the supervisor is chosen as the most causal agent. 94 5:55:35.— ..efliR-a» cogs—2.2.2: an: binge :3er we .5325.- .. 3 «Eu... 583.895 .a 953.? 603.35% £33sz BEE: BEE. 62 t t l --l l 1‘ - L— T n.— .N N Wm S .330 828500 l-I . - I. - m .338_omnol fl m ... 0 m .330 .ofitomzm IOI m . -. t ! l- m Wm v v _, We 95 55.15:...— ..8_>._2_=m ©2509-2— cna 5:253 acme—.9 «e 532:: a 2. 2.33 252995 .3 9.5!..— §_Bac_.m:_. Sarcasm vogue?— 3: in it illii.l 1 fi— r mN .350 3%!8 ill. _ .asauaumuol 0/0 D Ema—WU Emir—23m IOI 7.. game Josyuadns r mé 96 Although Heider (1958) would argue that intentionality is a prerequisite for justifiability and should thus be orthogonal, the present data suggest that the two constructs may be related in participants’ minds so that justified behavior is seen as less intentional (see Table 3). Thus, if one views high intentionality perceptions as a proxy for low justifiability, then there are several other interactions that could indicate support for Hypothesis 5. For instance, supervisor blame perceptions were affected by the interactions between manipulated causality and perceived supervisor intentionality (1322034; 2 < .01; see Figure l 1), perceived supervisor causality and perceived supervisor intentionality (1332:.016; p < .01; see Figure 12), and chosen supervisor causality and perceived supervisor intentionality (1332:0130; p < .01; see Figure 13). These three interactions all appear to reflect the same interactions as did the justifiability interactions, but in the opposite direction (i.e., blame was more likely when intentionality was high and causality was attributed to the supervisor, or not to the organization in the case of Figure 13). Committee Blame. Table 9 provides a summary of analyses for committee blame that parallel those for supervisor blame. The three operationalizations of causality (i.e., the decision manipulation, perceptions of committee causality, and the choice among causal agents) directly affected perceptions of committee blame. In the latter case, post- hoc analyses showed that individuals who saw the committee as causal had significantly greater impressions of committee blame than those who chose the supervisor as causal. Participants who chose themselves as most causal did not differ from either of the other groups. Additionally, individuals who perceived the committee as more justified or less intentional were less likely to blame the committee. 97 inmates? oU IQIg .330 .SEonzm IOI. , m.— . nN .2 outing Joswodns . Wm r m... 98 iii- - ...l .i: a n. 1 r W.N 99 3:330 _8_tun_.:.m>“3 +1- 3-. .l- .1 £336 .8Eugm 55+ 1 ’ 19.1---- iiill. lll -fl a mlfl Joswodns 5:252:25 newer—2.9m 333.3.— .Ea 5:853 :82? no 552:... a an 25.3 Salaam .m— «.:.»:— b:a:o::3:_ .ofltoaam 3385; =9: 33 , L l l l I I. w a l l l I l I 14 ll — _ m.— .335 ooEEEoUI-I imauuh_oml¢l — . N .825 .8anam IO] _ m ,l. -u-l|| -- ll ll. 11.. ll: .1 1 . Om S n .w w n m J m m On Q . 0v L'nl occurred ft of causalitj perceived. Q1 choose thc superx'iso: chosen bs by 119 p2 is categoi combinat regressio Organize measure HObOdy ' Tables I blame“: logistic callSalit Source ; Addilic palm; Unlike the analyses for supervisor blame, none of the predicted interactions occurred for committee blame. Committee blame appeared to be a result of direct effects of causality (manipulated, perceived, or chosen), and justifiability (manipulated or perceived, as well as for perceived intentionality). Choice of Blame Among Sources. Another Blame item asked participants to choose the source most to blame for the decision (between ‘nobody,’ themselves, the supervisor, or the committee; see Question 22 in Appendix C). Overall, ‘nobody’ was chosen by 101 participants (25.1%); themselves by 15 participants (3.7%), the supervisor by 119 participants (29.6%), and the committee by 142 participants (35.3%). As this item is categorical, it could not be analyzed using typical hierarchical regression. Instead, each combination of individual choices was used as a dependent variable in a series of logistic regressions. For instance, the probability of participants choosing the supervisor vs. the organization was regressed (via logistic regression) onto each combination of causality measure/manipulation and justifiability measure/manipulation. The results for the supervisor vs. committee blame analyses, the supervisor vs. nobody blame analyses, and the committee vs. nobody blame analyses can be found in Tables lO-12, respectively (note that, as very few participants indicated themselves to be blameworthy, this category was removed fiom analysis; note also that the results of the logistic regression are expressed in terms of 32 for ease of interpretation). Overall, causality variables appeared to be relevant to blame, such that participants who saw a source as causal were more likely to choose it as blameworthy over another option. Additionally, perceived justifiability and intentionality were related to blame, such that Pal‘tiCipants who saw a source as more justified or less intentional were less likely to 101 Cram—.:.HLL—r-m Col-ZILL—Cm - -. . . . AgJST-SQU .:.; «VA U~h-me~w> :Av._:~.dmu.:7.:~i \A.~::.L~:w..v cc... w.m_.< ’\Hili-£F.haullhs' lvx‘li-ll II-Jlunllllil LC; HV-m FCCL LC.— cvm 23.-30:323.. LC.“ hV~ ii I . . :- -.-..I . .. , Iii I- gamut—IQWWmeIZT 3:..wo-Nm....3W3-wévhmflifdhcsz.—u:— CC UPC-lfiu Lmvm...H)-.-|..wh~:.f. 3.an UU:.:~::3..V-~3 “LEAN-Iv... LmUV~ 3:2..334 UC~§3 .....::v..4v~ . - . . r . C~ airing.- . . .moficmafloo o—DflF 8083563. 88 ._ .o. :30 monm. :mmo. :vom. 5:803:28 SmMEomsm 338.5; ooo. :mmv. mmmv. :ono. :vom. :ozmowcmsm 883.885 333.5“— .o_o. :mnm. mom. _ _o. :vom. coca—«EEQE 8038533 883.695 .oo. :mmm. 5mm. .0 _o. :vom. 8.82888 85888 338qu .0288 So. :omm. wmm. ._No. :vom. cocooasmi 88888 33388 32888 ooo. :vom. vom. woo. :vom. zoom—8:88 838563 828800 .8 838885 :omo. :hmm. mom. :mmo. :owm. .028—2888 commtasm oozooeom So. :mom. vow. :ono. :owN. 8228563. cemmtoqsm 33083 _ .o. :mom. mom. _ _o. :owm. coon—88m8 cozaoazmsm 883.596 woo. : Sm. mom. .mzo. :owm. 3282888 35888 volcano 3:828 ooo. :oom. moom. . _ No. :owm. cosmoczmsm 88888 33883 883.685 woo. Fwom. oom. voo. towm. cog—38:58 833563 858800 Mo 83808?— .So. : Gm. mien. :mmo. :Em. 5:82.888 883895 goofing—om So. :vmfl mmwm. :ono. :EN. cosmoEHma 883.596 3388; .m _o. :53. NNN. :o. :o_~. 8283858 8380538 cemmtoasm .omo. :ovm. omm. .o 5. :EN. 8.82828 8588 oozoeom moo. :mmm. mmm. ._No. :EN. 8058563 88888 33080.; coca—somewE _oo. Lam. omm. woo. bro—N. :ouflsfiamE cocoa—mama oozm88o0 commmooQ cozowhWE 828888 «ho—£83 82a 0388? use? 033.83 033:; .5» $2 8o m Son .8.“ M :oumouzmi. 88 M @2883 .59 M 03383 cease—LEE. 5:380 8880 .8 888:8— :o oEflm 883.8 3m .m> 358800 Mo 8238 um comm 04 05 mo $3.8M 2 anh 102 A\F.V Au-U VA\ai .r.-|.\ . .‘ u . .‘ .:.. . ... r\ 2 u..\ [.[x{ .z , i . Il| . D. . . . . lull Illil. ... I I.I|I.|.I| Jr 1.; .. In. . . II | . , 83:33.81»; CCCUSLMJZm (IO—LETS?) UCC-mw U—£~w.—Lfl> DEC—C 0—5:..Lfixr 07:...2-5» .:...— nM—Q LC. Mm. £72.; LC.— NM— CU:.~U.L:7.3...:C NW—I \Au..~.w2...~0 LA: ‘Vfi U~£=.~.-~> =3.:-Um:~,l.:~. \A.~:-..7.:.,...V r 1‘ .L‘ ‘ \ I . ul |I III“ I. El[.,[{.[.[.[ I 11“- 1- -. ‘I ‘(1 . . . . . .43. .:C. v.» C~ 97.23 F .— A~A~ Co. v 3 0308? 3:858 05 088.” new 808 %lm 8 8080.88 88:8me 0 Co. v .8 05082, comamocumsm 05 08% 88 .58 NM 8 8080808 835%? a 26% 00308? 8.00:8 =0 83 .Co. v 9 “Swami—Em 0.8 83000838 :0 < S. V a : mo. v a woo. :_ 3. non. :mmo. :mom. 8.83802: 0838095 o0>_020m » moo. :omm. wan. :ono. :Cm. 80:08:03 80.38083 “003083 boo. :omm. m _ m. :o. :m _ m. 80803858 80308803 Sarcasm ooo. :o_m. o3. .Bo. :m _ n. 5283808 008888 “002080; ooo. :w_ m. w _ m. . _ No. :m — m. 82305003 008888 v0>m080m 080? 30:00 _oo. Bonn. o _ m. woo. hm _ m. cog—38:08 82.005803 008800 «c 008:0 8:080:05 8808008 $033.5, 0:20 0308?» 0:20 0308; 03083 8m «#2 you Nw Son 88 NM. 8030053388 Nm 5:828 8.8 MM 03080> 80885003. 5:00:00 ”9.258 2 29¢ 103 U“-‘UW‘I\\' I‘"!!Il.\ Ii-JIINII‘1 I C t < F-FUmwluanLav-F-m thmwlvflflcnfihfhm I I . I L-cEM-w.leth-lfwm-wwww 1......- ....m::w-~Ww.|_-.ME we wwleJno-H..M.-..-n-M.-%- -- .:--.---- -93....89-5z:th-..,.-..-x..b“. - >.-m...:...£:... 0| CCtS-JEtZZE 3:- Ix4~:~wvn-.:w.v-u3 783.30.;x: :C U~C$~M~ L»..r....>-.Jr~:m 30> >33£CZ -:v T.CC...A.A.J.~%UN~ Umuv-mm:f~ 032;: 2:32.0V‘ ~ ~ ain‘t-ath- s I :o. v 8 0500.9 50 05 0800 0:0 00>0 mm 8 8080.58 8005830 0 30% 0050800, 53 000 .Co. v 3 80058010 000 082000030 :0 < _o. v a . V a 9 .2o. :Nmm. w—m. :No_. :oom. 83838008 00030095 0030003 mo . o 5. :wow. wmw. : _ mm. zoom. 820005008. hommtasm 0030009— 080w0 30:00 :08. L R. am. 0.08. :08. 8003958 808002.. 80505 0o 86:0 :08. zam. am. :02. :80 0:28:25 805%... 8285 00038 so. :08. wow. :5. :80. 808008.. 802003 838.50 §Eaa woo. :08. 30. E08. boa. 8003255 8080003 8030.30 .8 5000800 .. _ 8. :00 0 00 _. :8 _. mo. £08835 083.83 83850 :0. 00m. 0:. Sm. :08. 808032..H 808003 838.00 8003358 .ws. “0:. 8 0 ME. ”MS. 8022358 808003 8259.0 02080 020000008 003000008 $05083 0020 03080.» 0020 030.83 03080» .5.“ wM< 00.0 M £00 08 M 80:08:03 .80 M 5:00:00 .80 M 0308? 02.005002. 8.00000 coma—womumsn 080 :m .m> @0002 .00 082000 0% 030m 0-— 05 .«o $800M : 050a. 104 v .3 030t0> 0050 05 2500 000 00>0 NM 8 8080008 80090me 0 30:0 0050t0> 500 000 AS. v 3 80050me 0.00 002000030 “”0048 . V fl ooo. :9. _. 00 fl. :3o. :3 _ . 09200000008 0058800 00300008“ 5 a. 80. :3 m. 50m. :mnm. :02. 00000508: 0058800 00300000 080w0 80000 moo. thm. m _ m. ham _. :02. 000080808 0000000000“ 00008800 00 000000 :5. :mmm. NNN. :30. :mg. film—00000008 0038800 00200000 980000 moo. :vnm. mum. :mnm. :30 000000000a 0088800 00300000 0088800 ooo. them. SN. ENE. :23. 000008808 0000056& 00080800 .00 000000000 ooo. :03. 02. :39 : Eb. a=0000008m 0088800 00200000 woo. m H m. _ 7n. mum. :0. 000000.:000m 00000800 00300000 0000:6808 :5. “M0. 8 _. HS. Us. 82888.88 808008.. 888880 82880 000000008 000000008 080030800, 0000 050:? 000—0 0E0t0> 0B0t0> 00.0 qu 00.0 M 500 00.0 M 00000—0003 000 M 00:00:00 000 M 03080, 000020003. 0:00:00 00000E003. 000 200000 .00 0.800200— 00 0805 00000800 .m> 00002 .00 0000000.. 0% 000“ 0A 05 .00 3.000% 2 030B 105 choose that source as blameworthy over another option. Several interactions were also found between causality and justification variables, and will be discussed below. Committee vs. Supervisor Blame. Please see Table 10 for a summary of the analyses in this section. First, the decision manipulation interacted with the manipulation of supervisor justification (A32=.015; p < .05). As shown in Figure 14, supervisor justification only influences (i.e., increases) the probability of blaming the committee over the supervisor when the supervisor made the decision. In a manner similar to that for the continuous supervisor blame variable, the likelihood of choosing the supervisor over the committee as the most blameworthy entity was influenced by interactions between causality and intentionality. Specifically, there were interactions between manipulated causality and perceived committee intentionality (A32=.020; p < .05; see Figure 15), manipulated causality and perceived supervisor intentionality (A32=.014; p < .05; see Figure 16), and perceived supervisor causality and intentionality (13322030; 9 < .01; see Figure 17). In all of these interactions, greater perceptions of intentionality for a source, coupled with causality for that source, led to a greater probability of blaming that source. Supervisor vs. Nobody Blame. The series of analyses described in this section involve individuals who chose the supervisor as the most blameworthy agent and the participants who chose ‘nobody’ as being to blame for the outcome. The results of these logistic regressions are summarized in Table 11. The manipulation of supervisor justification interacted with both manipulated causality (131128033; p_ < .01; see Figure 18) and chosen causality (A32=.O30; p_ < .01; see Figure 19), and the manipulation of causality also interacted with perceived supervisor 106 803005003 000300000 0000—03000. 00 . 00:00 300:... 0.: ..0>0 08:80.0 0 ~m... 020—03008 .00 08.000.— 0 00 38030803 00 0000 0 05 9.7.0000 .8 5:50:95 .3 flaw-.0 8003 020.80 82 ._ 020.509 00008800 '0'. 000.0000 00006005 [Cl _ Y Nd 0o Y 0.9 . Nd Y 06 06 0mm roswadns ‘SA mamas 3° 11014 107 0 00 3203080. aha—00020005 0058800 0030000.— 000 5:00:00 0000—05008 u0 00:000.— 0 ma hammtofiflm on: .5000 00035509 0-: wflmmacn—o MG hum—maaA—OLA— .mn 053M:— b=000008£ 000.8800 00300.00 . 80.: 33 l to \ I'illll .|_ I! I‘ll, l lilillnaluttlltlo .0 r «.0 8w m: S n , 3m 4 - z s I I 1 m. n. A 00.0000 0088800|¢| ,. 1 m0 ”0 a 00.20 50:50: m I. l .! yl Ill- l. .. 8 m 0 3 0, 06 r 0.0 108 sat—0002038 00880000 0030000.— 000 08:00:00 0000—08008 ._0 00:00:.— 0 00 88030803 00 0058800 05 ..0>0 000080000 0.: 0800000 00 8:30:95 .3 0.8»:— 80.: Nd m6 Yo 008000 0088800 .|¢| 008600 0022005 '0' . We . ed outing anymmog "SA Joswadns 3° "101d Y fie Y «.0 . ad 109 5:000:52: 000300000 10309.0.— ..00 50:00:00 00:30:00 $020000.— ._0 00:00.: 0 m0 war—030.00.: 00 005—0800 05 ..0>0 00:20:00 05 wing—.0 :0 5:30:95 .5 0.53.: .:.—000300.:— 0003003 003020: 5...: "2 a 0mm mymmog 'SA Josyuadns 1o hmqvqmd 5:00:00 083.09% 304 |¢| 0 0.0 5:00:00 08.2095 .3: + md Y 06 5c 1 «6 V o6 110 003005003 0003000..» 5000—0300:. :00 3:00:00 0000—2000:. «0 00:00.... 0 00 3203050.: 00 2:00:00. :00 000300000. 0:. ”530:0 :0 5:30:00.— .w— 0003,..— .003390050 00:00:02... 0033005 :3: 00:20.2. 52 w 023000MBflEEmWHW .0 003000 00mm>0000m+ ‘ _.o T Nd . mac r 0.0 3 9o 0. S v. ”.O , ad aumq Apoqou “SA Jadns )0 'qmd 111 __r_i_ , ._—H 00:00:30.3. 0000:0000 300—03000. 000 b:0300 030:0 :0 00:00.: 0 0.0 03203050.: 00 2:00:00. 000 0003.000» 0:. w0mm00:0 :0 ~3:30:00.— .0— 000mm,.— Gozaqézv 8:85.00 .anea BEE; 30.0.0 82 .,\1 w ‘yf VII .Vl [4‘1 . i L“ ‘\1< ll- 3" .8000 000002000 I? 3000.0 08.50005 .IOI Y —.o . N6 7 m6 Y v.0 - nd r :6 , ad ad aumg .(poqu 'SA Jospuadns 30 hmqtqmd 112 justifiability (Afiz=.01 1; p < .05; see Figure 20). Graphs of these interactions support Hypothesis 5 - the manipulation of justifiability only affected blame when the supervisor was causal. Additionally, perceived supervisor intentionality interacted with manipulated causality (13132003 1; p < .01; see Figure 21) and perceived supervisor causality (13320030; 2 < .01; see Figure 22). Intentionality is more strongly related to blame when the supervisor is causal. Committee vs. Nobgtiy Blame. As shown in Table 12, in no instance was an interaction found between causality variables and justification variables in determining the probability of blaming the committee vs. blaming nobody; only main effects existed. Hypothesis 6. This hypothesis asserted that blame perceptions / manipulations will moderate the relationships between procedural justice perceptions for a given source and reactions to that source. Just as there were many different ways of assessing causality in analyses for Hypothesis 5, there are also many ways of assessing blame for these analyses. For this hypothesis, blame was assessed using 1) a dummy-coded variable indicating manipulated blame to the supervisor, the committee, or nobody, 2) perceptions of blame directed toward one of the sources (Questions 20 and 21 in Appendix C); and 3) an item that required the participant to choose between nobody, themselves, the supervisor, and the committee as the most blameworthy entity (Question 22 in Appendix C). Analyses were conducted using a series of hierarchical regressions similar to those used to test Hypothesis 5. Summaries of the analyses for Hypothesis 6 using organizational OCB’s, organizational commitment, supervisor OCB’s, and commitment to the supervisor as dependent variables can be found in Tables 13-16, respectively. Each dependent variable’s analyses will be reported sequentially below. 113 00:03:00.0. 000300000 00209.00 000 5:00.000 0000—0030.: 3 00:00:.— 0 m0 30.030803 00 2:00:00. .006 00030000... 0:. 00:02.0 :0 3:30:95 dN 000w:— 00:00:.300_. 08.50005 00300000 003000 020.0080 101 003000 00003005 1.1 Nd ad . od swam KPOQON 'SA Joswadns 3° Kunmqma 114 l l 30.000.00.00 00030000» 00300000 :00 0500.000 00.030305 ..0 000.000.. 0 00 3.0032003 00 1:00:00. 00>0 000.0000... 0:. 00.0030 00 5:30:00.— AN 000w:— ».:000..00.0_ 08.03005 00200000 00.5300 00.00.0000 1.01 00.3000 00030005 1.1 md Y ed 0 md :d aware @0qu 'SA Joswadns 10 Manama 115 00:05:00.0. 0000000000 00300000 000 0.0.0.0000 00030000... 00200000 .0 003000.. 0 m0 3.0032003 00 2:00:00. 00.5 00030000» 0:. w0mm00:0 .0 b:30:000 .NN 0000.0 5:00:00 000300005 33 |¢l 3:00:00 00030005 :3: 1.1. 0 -Y_ .:.—000000.0— 00030000m 00300.00 .00: 33 - 0d md ed 0d ad ad aumlg .(poqu 'SA Josguodns 3o Kmyqnqoad 116 Ame. v 5 030.00> 00.03 0:. 26:0 000 .26 0M... 0. .00000000. 3005.00 0 00: 030.00> 00:00.. 0 Amo. V 3 o-Dwmud> mmoemmm 0:.—w 00:030.. fig 00>O Nam Cm 20060000. «gowmcwmm mflfi D—Dmmbw> ”Ed—m 0 Co. v 0. 030.00> 0803 0:. 0>0:0 000 .26 NM 0. 000000000. .0000...0w.0 0 00: 030.00> 00:00.. < s. v 0 .. 00. v 0 t 0:.. .08. 008. ..Nmo. NS. 0503.0 00.0:0 .08. :08. 3.0.0. :08. :08. 2003 80.668 .0 08008.00 08. :05. 0.08. :08. So. 2.3-0. 0022:0052 00.80030. 00000030. 0030.00> 000.0 030.00> 000.0 030.00> 00.. NMQ 00. NM. :.0: 00.. NM 00000.0. 00.. M 0:00.: 00.. NM 00:00:05..an 00.0.0 0030.00 > 00.00.. .000000000 00:00.00 00 000 00000300000000 0000.0 00 0.000 002003030 004.00.000.03. 0:. .0 0.300% m . 030,—. 117 88. um n Co. v 3 00000000008000 00005 00>0 00.00.0000 00 £000 050.0? 00000.. < S. v 0 : 8. v a . 080. :3 .. zom _. :mm .. :08. 0.005 ..0 00.000 woo. : .3. :3 .. :30 :98. 0000.0 000000000 ..0 0000000000. moo. :. m .. :03. :30 .08. 903-9 000030.002 000000000. 000000000. < m03000> 000.0 030.0? 000.0 0.00.00> 00. NM? 00.. NM .000 00.. 0% 3000.0. 00. NM 00003 00,. 0% 000000000000000 000005 m0.00.00> 00000.0“. 000 080.05 00 00080000000 300.000.0030 ..0 000.00% 000 ..0 9.000% 3 0.00... 118 Co. v 9 000000000000000 080.0 00>0 00.00.0000 0. 0000 0.00.00.» 00003. < 8. v 0 : 00. v 0% moo. z: .. :m. .. :m. .. woo. 0000.0 ,.0 00.000 .m .o. :02. :m. .. :m. .. :mmo. 080.0 000000000 ,.0 0000000000 08. :80. :3 0. :2 0. 08. 903-0 8002:0052 2500 000000000. 000000000. < 00.0.0.0? 000.0 0.00.00> 000.0 0.00.09 00.. 0m? 00.. NM .000 00.. NM 00000.0. 00. 0M. 080.0 00.. 0% 00000000008000 0000:. 00.00.0W> 00000.0 0 000%00003 00 0.mUO 0000,000im ..0 00200000100 05 00.00.00. .0~.I m. 0.0.0... 119 Co. v 3 00000000008000 0000.0 00>0 000000000 00 00.00 0.00.00> 00000.. < .QV0: Noo. :wom. zoom. :mom. : .0o. 0000.0. ..0 00.0.5 moo. :Nom. :03. :mam. :03. 0000.0 000300000 .0 0000000000 moo. room. imam. imam. . .o. 903-3 00.00.00.003. 0&0... 000000000. 000000000. < 00.0.0000, 000.0 0.00.00> 000.0 0.0.0.09 00.00.”? 00.. 0M. 0.000 00.. 0% 00000.0. 000%. 0000.0 00.. 0% 00000000008000 0E0.m. 00.00.00> 00000.0 0 0300005 00 0000,0096 00.0 0. 000000000000 ..0 000.0000 0w. 0.0 ..0 00.003. 0. 0.0.0... 120 Organizational OCB ’s. Summaries for the analysis of Hypothesis 6 using organizational OCB’s as a dependent variable can be found in Table 13. In terms of main effects, only perceptions of the committee’s blame was (negatively) related to intent to commit OCB’s toward the organization (32=.025; p < .01). Additionally, organization procedural justice had a positive relationship with organizational OCB’s (£20032; p< .01). Of the three ways of operationalizing blame, only one interaction was found between blame and organizational procedural justice perceptions —- that for perceptions of committee blame (A32=.013; p < .05). See Figure 23 for a graph of this interaction. The relationship between organizational justice and organizational OCB’s was evident only when committee blame perceptions were higher. This interaction is of the pattern that was predicted by Hypothesis 6. Qgggrjgational Commitment. See Table 14 for a summary of the Hypothesis 6 analyses using organizational commitment as the dependent variable. Again, three ways were used to operationalize blame. As can be seen, all operationalizations of committee blame influenced participants’ commitment to the organization, such that higher perceptions of blame, or being in a condition in which the committee were designed to be blamed, led to lower levels of commitment (the manipulation of blame was not significant at the .01 level). Perceptions of organization procedural justice were also positively related to organizational commitment (320125; 12 < .01). Of the three ways of operationalizing blame, only one of them functioned as such a moderator (the item asking participants to choose among blame sources; A32=.022; p = 121 0000000000 0000:.— .000000000 00.30.0090 000 0000.0 00.0.5500 00>.00000 .0 000000.. 0 00 0.906 .000..00.00»0O .MN 0000.0. 00000.00 .000000000 .0000... 0000000090 00:0 33 r I It 1 - i l} l - - l - l-| ill ll ll. -ll - Elllll ltlr. - m.— . N 0000:. 000.0—0000 0.3.01? Wu 0% w 0000:. 80.000000 33 IOI W. . Ill i- .. \lmu -. 0 m. CK\\1 l‘ w 0.0 m a v 0 n... 122 .0105). Figure 24 shows a graph of this relationship. The regression of organizational commitment on organizational procedural fairness perceptions was significant for individuals who chose the committee or ‘nobody’ as being to blame, but was not significant for individuals who chose the supervisor (regressions for individuals who chose the committee and who chose ‘nobody’ were not significantly different from each other). One other Blame x Justice Perceptions interaction existed for organizational commitment. The choice among blame agents operationalization interacted with organization interpersonal justice perceptions, although no hypotheses were formally prOposed for this variable (A32=.026; p < .01). In Figure 25, where this interaction is graphed, it can be seen that the relationship between justice perceptions and organizational commitment is greatest when ‘nobody’ is chosen as being to blame. Significant differences in slopes for these three groups were found between supervisor- choosers and ‘nobody’-choosers, although the differences among the other slopes approached significance. The pattern of slepes, however, suggests that the relationship is stronger for individuals who chose the committee as being to blame, as compared to individuals who chose the supervisor as the most blameworthy agent. Supervisor OCB ’s. In Table 15, one can see the summary of the analyses for the tests of Hypothesis 6 using supervisor OCB’s as the dependent variable. Of the three operationalizations of blame, only the use of perceived supervisor blame had a main effect (negative) on supervisor OCB’s. Additionally, the supervisor procedural justice perception measure exhibited a significant positive relationship with supervisor OCB’s. 123 0000000000 00000.00»? 000 .00w0 0000... 00000.0 .0 000000.. 0 00 0005.80.00 .000: 08500 050820 0580 800005000 It lllllllul ll 0000:. 00 0000.02 III 0605 0. 00030005 |¢I 25.0 2 80005500+ will; “ \_ - n.— . nN T n... r n... 0000000000 00.00.... .000.000.00w0O .VN 0000.... luaungurwog [Buopazyunflio 124 m ‘ 6.00.3020: 00.30.. 3000.09.30. 00.30.0090 .00.. .00»: 0.00... 000000 .0 00.30.... a 0... .000...0..000 .000..au.0au..O .m~ 000w... $05.0“. 3:09.092... =0..0~.=0m..0 :9: 33 V I- 1| J :‘ ‘1 ‘5. “i i\+ x". . s l. u! \ ‘ J. — T m. m mu no 2 m. .33 3070+: ‘ W. 0:85 0..0m.>..00=ml®l . m M. Nam.mlo,~.ow§emw+.% m m. 1 3 m 9 w r v _ m... 125 Of the three operationalizations of supervisor blame, only the perceived blame variable was found to moderate the relationship between supervisor procedural justice perceptions and supervisor OCB’s (13322013; 2 = .017). This interaction is represented in Figure 26. As hypothesized, the relationship between justice perceptions and OCB’s is stronger when beliefs about supervisor blame are stronger. Although not directly hypothesized, supervisor blame perceptions also moderated the relationships between supervisor OCB’s and supervisor informational justice perceptions (A32=.043; p < .01), as well as for supervisor interpersonal justice perceptions (A32=.028; p_ < .01). The first of these, graphed in Figure 27, shows that the relationship between supervisor informational justice and supervisor OCB’s is positive only when supervisor blame perceptions is high. The second relationship (see Figure 28) shows that the relationship between supervisor interpersonal justice perceptions and OCB’s grows stronger as the supervisor is blamed more. Com_m_itment to the Slgervisor. Table 16 shows a summary of the results of the analyses for commitment to the supervisor. Main effects for blame operationalizations were found for perceptions of supervisor blame, as well as for the item forcing individuals to choose between possible blame sources. See Table 3 for the results of the former analysis - in the latter case, individuals who chose the committee as being to blame had significantly higher commitment to the supervisor than did individuals who chose the supervisor as being to blame. Additionally, supervisor procedural justice was positively related to commitment to the supervisor. The hypothesized (H6) interaction between supervisor blame and supervisor procedural justice was not found for any of the operationalizations of blame. However, 126 w‘w-— _. 000.00.00000 00.3.... .00....00000. 000.200.... .000 00.0... 000.200.... 0020000.. .0 00.300. 0 00 PGUO 000.>00..0m in 000»...— mm0E.0u. 303.000.... .0000... 00030005 00: 33 it w l l i . ll I l .14 s — .- m. T N 0.00:. 00030095 33 l¢l . n m S n 0000:. 083.095 0%.: I? .m .- ._ fill. i .1 m m. m. , 3 s . v 7 m6 127 0002000000. 00.00... .000..0.000.0. 000.30....» .000 0.00... 00n.>00..00 00.0.0000.— .0 00.30... 0 00 RGUO 000300....m SN 0000.... $000.00. 30000000.... .00.>._0..:m 00:. 33 l 1 ll l .l l I \ ll Ill 1 It i illa l. — n. a N 00.05 000.3095 33+ . md S n .m 00.03. 000.3095 :w... + w. , I, iii 0 m or m .\ Ax. . mm s . v m... . . 128 0.0.0.. 300002.080. . .. .. 0 000.. 0000 00m .3 0000... 000.300.... 000 0.00... 000.000.... 0030000.. .0 00.000... 0 00 0.3.06 000.10 00000.0... 3000.00.30. 000.2005 . Illl ‘ 0:. 05 502005 30.7 W . mN 00.05 000.20....m 0.9: g V! m m 8.830 JOSEMOdnS V . 0.? 129 one interaction between blame and another dimension of justice existed. The 3-way blame manipulation (between the supervisor, the committee, and nobody) interacted with supervisor informational justice perceptions (A32=.018; p < .05). Figure 29 shows this interaction. Separate regressions run on individuals in each condition showed that the slope for participants in the supervisor blame condition was significantly greater than the slope for participants in the ‘no-blame’ condition. No other significant differences existed between slopes, although it is apparent from a visual inspection of the graph that the supervisor-blame conditions may have resulted in a stronger justice-commitment relationship than the committee-blame conditions. Controllirgj‘or the Effects of Covariates It was mentioned above that several other variables (e. g., age) were related to some of the dependent variables examined in Part 1. Thus, it is important to explore whether the significant effects reported above still exist afier controlling for these demographic correlates. Controlling for demographic correlates in Part 1 only resulted in the change in significance of two interactions. Specifically, controlling for age eliminated the significance of the interaction term in the logistic regressions of committee vs. supervisor blame (see Table 10) between the decision manipulation and the supervisor justification manipulation, and the decision manipulation and perceived supervisor intentionality. In both cases, however, the p value dropped to just over .05 (p= .0595 in the former case, and p = .0551 in the latter). Age was not significantly related to the decision manipulation, the manipulation of supervisor justifiability, perceptions of supervisor intentionality, or the interactions between the decision manipulation and the other variables. Thus, it is likely that the decrease in the significance of this interaction 130 05000.0... 000.000... .000 0000.00.50. 00.0... 00.0000... 0... .0 00:00... 0 0.0 000.000.... 0... 0. .00.... .02 Ill? 00.05 0. 000.0800 10! 00.0.0 0. 80.2000... IOI 00.: _ _ 0.000.... 0000000000 00.0....q 10904.0“ 0.. 0w. 0 .00.... 0005.0... .000..00..0.0. .00_>.00:m .m. .. Wm . 0m r m... luauuyuuuog Josyuadns 131 term may be due to a statistical artifact. Alternatively, age may play a role in the assignment of blame to the committee over the supervisor. PART 2 In addition to the scenario study reported above, a second study was appended to the first. After participants completed the measures for the first part of the dissertation, they were then instructed to respond to questions concerning the most recent parking ticket they received on campus. Participants completed measures that were parallel to the justice measures used in Part I. This time, however, the items were rewritten slightly to reflect two different sources of justice — the parking employee who actually issued participants’ tickets, and the Department of Police and Public Safety (DPPS), the organization for which parking employees work. These two entities (the parking employee and DPPS) were considered to be similar to the supervisor and the upper management committee used in the stimuli for Part I, in that they represented two different sources to which recipients of parking tickets could attribute blame. Participants were also asked items concerning whom they blamed for their parking ticket. Thus, this part of the experiment can be used as another test of Hypothesis 6 -— it is expected that blame attributions will moderate the relationship between justice perceptions and reactions to that source. Method Participants Participants were comprised of a subset of the participants for Part I, and were identified on the basis of a survey question asking whether they had ever received a parking ticket on the campus on which this experiment occurred (222 individuals, or 132 52.6% of the total sample). The mean age of participants in this part of the experiment was l9.8 years old, and 75% of the sample was female. One hundred eighty-three participants (82.4%) reported their race as White. Additionally, this sample included 11 African-Americans (5.0%), as well as 18 individuals who identified with other minorities or did not identify a race. Of this sample, 190 (85.6%) reported that they had in the past received a parking ticket that they did not feel that they deserved. F ifiy-three individuals (23.9%) reported ever having won an appeal of a parking ticket, and 55 (24.8%) reported ever having lost an appeal (roughly 55% of the sample reported never having appealed a ticket). Fifty- seven individuals (25.7%) reported knowing someone who worked for DPPS, and six individuals (2.7%) reported that they were themselves DPPS employees. Procedure Participants were instructed to think about the most recent parking ticket they received on campus (see Appendix C for a copy of the instructions). Following those instructions, participants completed measures assessing the blame they assigned to each source, justice perceptions for each source, and reactions to each source, before completing a set of demographic items. Measures Me. Similar to the blame measures used in Part 1, these blame items took on a number of forms (see Appendix D for complete item wordings, and Appendix C for the items as they were displayed on the computer screen to participants). Participants were asked three one-item measures asking the extent to which they agreed with statements that they, the parking employee, and DPPS were to blame. Participants also responded to 133 an item that required them to choose, on a 5-point scale, whether the parking employee or DPPS were more to blame (i.e., each source was at one of the ends of the continuum). Finally, participants were given an item that required them to choose between themselves, nobody, the parking employee, and DPPS as the source that should receive overall blame for the parking ticket. Justice Perceptions. Perceptions of justice used items adapted from the same scales as the scenario portion of this study did (i.e., items adapted from Byme & Cropanzano, 2000). These items were rewritten to reflect perceptions of justice toward the parking employee and to DPPS (see Appendix D). Reaction; Measures. Items reflecting reactions to both the parking employee and to DPPS were written for the current study. Mirroring the first part of the study, they were intended to reflect both attitudinal reactions (c.g., liking of DPPS), as well as potential behavioral reactions toward both the parking employee and DPPS (c.g., being fiiends with the parking employee; supporting an increase in student fees to benefit DPPS). lterns can be found in Appendix D. Results Psychometric Chmactmsmf ltemsgd Scfls For the most part, items appeared to have sufficient univariate characteristics (i.e., approximately normal distributions with an appropriate range of responses). The most common concern that existed on the univariate level is that positive skew existed for a substantial amount of the items, particularly the reactions measures. The properties for each scale and type of measurement can be found below. 134 alpine Measures. Means and standard deviations for the first four blame measures can be found in Table 17. Although the highest amount of endorsement was for self- blame, there was still a substantial amount of blame that was also ascribed to DPPS, and, to a lesser extent, the parking employee. When the parking employee was paired against DPPS for evaluations of blame (i.e., the fourth blame item in Appendix D), more participants selected DPPS as being to blame, as compared to the parking employee. Additionally, the forced-choice item regarding different sources of blame showed that 112 participants (50.5%) blamed themselves for the ticket, 54 (24.3%) blamed DPPS, 22 (9.9%) blamed the parking employee, and 14 (6.3%) blamed no one. Parking Employee Justice Measures. Factor analyses on each justice scale separately found evidence for unidimensionality in each case. Intemal consistency reliabilities for parking employee procedural justice, informational Justice, and interpersonal justice were .77, .86, and .75, respectively. A factor analysis on all 12 items together resulted in a three-factor solution according to an evaluation of the scree plot and an analysis of factors with eigenvalues greater than 1.0. An evaluation of the factor loadings showed a replication of the factors the items were originally designed to represent. DPPS Justice Measures. Similar results were found for the items representing DPPS justice perceptions. Reliabilities for DPPS procedural justice, informational justice, and interpersonal justice were .76, .87, and .78, respectively. Additionally, a series of CFA’s was performed in order to determine whether participants distinguish among perceptions of justice. A CF A modeled according to a six- factor solution (i.e., procedural, informational, and interpersonal justice perceptions 135 . moan: 8.- 3- 5.- S.- 8. 8.- .:. .2. 0..- 0... 00.. 0.0 00%“.wa o... 0... o... 8. 3.- :.- 5.- 8.- 8.- 2.. 0.0. .00 00...... :9”. ..N@. :mm. :8. :3... :3. :mvr :vmr :3. m0. oNN 03 000.800. mama-N. :wm. :3 :3. tom. :mo. .3. 2......- :3.- ..o... c... moN .8 000.800. 0000.900 w0.....0..... 00V0. comm. 90°F. :Vm. :—m. .:.—N. ooeN... eB—f :WN. O—.~ NN.N O—N mmvemflh —Nfi0mhoevwfim WQQQAV— .0. :00. :0... :0... :00. 00. :0..- -....- :.N. 00.. 00.0 0.0 086.005.08850000.0 0.. :0... :00. :00. :00. :0..- :8.- ..m... 0... 0... 0.... 0050:8088... 00.5.0 0.. :3. :00. :. :00.- ..00.. :8. 0... 00.0 000 036.0 08009258300... 00.00.... 00. :00. S.- S. .:.- :00. 00.. G... 0.0 085.00800000580208... 00300.0 00. S. :00.- ..3. :8. 40.. 00.0 20 0880:8382. 8.29:... 00.00.... :0...- ~.. .2. 3. .00 80 one... 0.0.0.2 8.2050... :3. :00.- R. 8.0 cm. 253 00...... :0..- 00.. $0 000 2.0... 8.295 00000.. .0 00.. .0...- 0.0 0003000.. 0 0 0 0 0 0. m N .00 .2 2 2.0.5. 00.00...0> .000..- w0....0.. 0... 00.. 00.0.5023. 35.0.0000 .00..0.0. .000 002.06.080.80. 003.0300 000.0005 0000.). h. 0.0.0... 136 0.00... 0. 00.0.. 00.0.0.5. w0....0.. 0... .0. 000000 000000 00.00.00. 000.0> .000... 000 .080... 0. 00.0.. mm... .0. 000000 .0008... 00.00.00. 000.0> .030. .0... 0000 00000 00? A0805 mama .0> 00.0.0.0”... 0 0.00.0.» .0.0.0n~ 000 0.00.0.7. .0... 0000 00000 003 00m 00.000 .0.0..-m 00 0000.. 0.0 00.0000... .0...0 :0 000 00.000. 8.0..-.- 00 0000.. 0.0 00.0000... 000.00.. ..< 0.000.300 0003 0.00000... 0... 0. 00 00.0.3020. 0000.0.0000 0.0.0.0. .807. .0. v m : m... V m . :00. 8.- .0..- S. 0.00 .0. .o. 0...- m..- 00< .0. 2.. am. no. 000.0000. mama .m. i. 0. 0». m0. 000.800. 0000.080 w0.....0.. .2 O. E.- 0000..0.. .0000.0...0.0. mmmfl .c. m . N . . . o . 0E0...0> ”00.0.00... 0...... 137 considered separately for each source) had results approaching good fit (x2(246; n = 202)=690.07; p < .05; RMSEA=.O95; GFI=.80; CFI=.83; TLI=.81). Allowing the errors of items with identical wordings (except for the referent) to covary further improved the fit ofthe mode] (38(234; g = 202)=600.67; p < .05; RMSEA=.088; GFI=.82; CFI=.86; TLI=.84; A1202; 1; = 202)= 89.40; p_ < .05). This model appeared to fit the data better than a similar model in which distinctions between sources were not made (x2(249; g = 202)=928.83; p < .05; RMSEA=.117; GFI=.67; CFI=.75; TLI=.72). Finally, CFA’s performed only on the procedural justice items (both parking employee and DPPS procedural justice) resulted in good fit, both before (x20 9; n = 212)=36.35; p < .05; RMSEA=.O66; GFI=.96; CFI=.97; TLI=.96) and afier(x2(15; g = 212)=21.61; p > .05; RMSEA=.O46; GFI=.97; GFI=.99; TLI=.98; sz(4; g = 212): 14.74; p < .05) the errors of items with nearly identical wordings were allowed to covary. This model appeared to fit the data better than a model that forced all eight items to load onto the same factor (x2(20; n = 212)=92.05; p < .05; RMSEA=.131; GFI=.89; CFI=.89; TLI=.85). Parking Employee Reactions Measures. Although it was originally intended that the items represent both affective and behavioral reactions to the parking employee, factor analyses and reliability analyses indicated that these items were best represented by a single factor representing overall reactions to the parking employee. Items representing these scales were combined together, resulting in a single scale with an internal consistency reliability of .79. 138 DPPS ReacticgrsLMeasures. Similarly, the items designed to represent reactions to DPPS were best represented as a single construct. Furthermore, the sixth item in this scale was dropped due to its low correlation with any of the other item (this item asked about support for a $3 fee to their tuition to support DPPS). The resulting single DPPS reactions scale had an internal consistency reliability of .77. Scale Descriptive Statistics and Intercorrelations Among Scales. Scale scores were created by taking the average of the items within each scale. Means, standard deviations, internal consistency reliabilities (when applicable), and intercorrelations among variables can be found in Table 17. As with Part 1 of the study, the correlations among the justice variables, particularly those assessing the same source, were fairly high. Additionally, the parking employee and DPPS reactions measures were correlated. Few relationships were found between traditional demographic variables and the study variables. Sex was weakly correlated with parking employee reactions, such that males had slightly more negative reactions. Additionally, several of the experiential demographic variables showed correlations with some of the measures. Individuals who reported currently holding a part-time job reported more positive perceptions of employee informational justice than did individuals who did not have a job (£22032; p_ < .05). The next few paragraphs report on participants’ reports of parking ticket experiences (see Questions 129-132 in Appendix C). Individuals who reported never having received a parking ticket that they didn’t deserve reported significantly more Positive perceptions of employee procedural justice (320132; 2 < .05), Employee interpersonal justice (320042; 9 < .05), DPPS procedural justice (320180; p_ < .05), 139 DPPS informational justice (_I_{_2=.070; p_ < .05), DPPS interpersonal justice (32:.095; Q < .05), employee reactions (320089; 2 < .05), and DPPS reactions (1120105; 9 < .05). Individuals who have either never appealed a parking ticket or who have ever appealed a parking ticket and won reported more positive perceptions than individuals who reported never having won an appeal for employee procedural justice (320046; 2 < .05), employee informational justice (32—0069; p_ < .05), employee interpersonal justice (320051; p_ < .05), DPPS procedural justice (1320090; p < .05), DPPS informational justice (320082; 9 < .05), DPPS interpersonal justice (32:.059; p < .05), employee reactions (320081; 9 < .05), and DPPS reactions (112=.057; p < .05). For the fourth experiential question about parking tickets (see Question 131 in Appendix C), individuals who had never appealed a parking ticket had more positive impressions than did individuals who had appealed parking tickets and never lost for employee informational justice (320048; p < .05). Additionally, individuals who had never appealed a parking ticket had more positive perceptions of DPPS than people who had lost an appeal (Bl-1.035; p < .05). Individuals who had never appealed a parking ticket had more positive perceptions than individuals who had appealed for DPPS informational justice (323052; 9 < .05), and individuals who had ever lost an appeal had less positive perceptions than others on perceptions of DPPS procedural justice (322068; 1; < .05). Additionally, the manipulations in Part 1 of the study were examined for their relationship with any Part 2 variables. Individuals who had been in a condition where the supervisor was to blame reported slightly more positive employee reactions (32=.O3 8; p < .05). Finally, individuals who had been in a condition where either the supervisor or 140 committee were to blame reported more positive perceptions of employee procedural justice than did individuals who had been in conditions where nobody was to blame (1320035; p < .05). A Word of Caution Regarding the Scales. Just as with Part 1 of this dissertation, there were indications in Part 2 that measures, especially similarly-worded items pertaining to the parking employee and DPPS, may not be definitively indicative of conceptual divergence. Additionally, the reactions measures for each source of justice lacked divergence to such an extent that the items were combined into one measure of reactions to each source. Just as in Part 1, it was decided to analyze the rest of the measures in this section with the assumption that conceptual divergence did indeed exist. Thus, the caution iterated in Part 1 about the validity of the interpretations of the results applies just as much to Part 2. Tests of Hypothesis 6. Hypothesis 6 was tested using a series of hierarchical regressions in a manner similar to which the same hypothesis was tested in Part 1 of this experiment. Summaries of the analyses using employee reactions and DPPS reactions as dependent variables can be found in Tables 18 and 19, respectively. Blame was operationalized using 1) perceptions of the source’s blame, 2) an item that required participants to judge the balance of blame between the parking employee and DPPS, and 3) an item that required participants to choose among blame sources the source that overall was most to blame (note that participants who chose ‘nobody’ as an option were dropped from analyses due to small sample size). 141 mm..— D—Dmtd> OCB.—m m .Amo. v m. 0...0..0> 0000.... 0... .0>0 Nm 0. 0080.00. 000me0 Co. v m. 0...0..0> .0...0 0... .0>0 NM 0. 0.00.00.00. .000m..0w.0 0>0.. 00...0...0> 50.. 00.00 00.26050 000.03 < .0. v 0 : 00. v 0t .oo. :30. MN... :0wm. :o.m. 0.00... .0 00.000 8... :8... A 08... :000. .08. 2.0... 0...... .0> 8.2.0.. .o 002.08.... .5. tom... 0.0. E000: pfmm .. 0.00... 00.20.00 .0 00000020.. 00.80.20. 00.60.80. < 00...0..0> 000.0 030.00.. 000.0 0...0...0> .0. ml... .0. NM 0.0.. .0. NM... 0008.0. .0. NM 0.00... .0.?M 0...0..0> 0.00:. 00%.. .0....0000... 00.5%.”. 000 000..0N..0l00..0.00d 0.00... 00 0000003. 00.00190”. 3400.000me 0... .0 0:003. m. 030... I42 Co. v 9 0...0...0> .050 0... .0>0 NM 0. 0.000.080. 0.02....w.0 2.0.. 0030...?» 50.. < .0. v 0 : 00. v 0 . .0.... .3... 060m... :30 :.C. 0.00... .0 00.9.0 no... .mm... 00%.... :Sm. :.oo. 00.0... 9.09 .0> 0000.080 .0 000000800 0o... lam... 0.0mm... inwm. :00 .. 0.00... m00Q .0 000000800 00.80.20. 00.80.80. 00.00..0> 000.0 0.00..0> 000.0 0.00.03 .0. «M4 .0. WI. .00.. .0. NM 0005.0. .0. PM... 0.00... 8.0% 0.00..0> 00.0.m 00.0.04. .0.:0000.0 m00Q 000 000..0N..000..0.00O 00% 00 000.8000. PEG .0 002000.30. 0... .0 0.0.00... 0. 0.0.0... 143 Wm. Summaries of parking employee reactions analyses can be found in Table 18. All blame perception measures had main effects on employee reactions (the blame measure requiring participants to choose along a continuum of DPPS and employee blame was not significant at the .01 level). Perceptions of parking employee blame had a negative effect on employee reactions. The item that balanced employee and DPPS blame was positively related to employee reactions, such that participants who believed that the employee was more to blame had more positive employee reactions than did individuals who believed that DPPS was more to blame. Finally, the forced—choice blame item was related to employee reactions, such that individuals who blamed themselves for the ticket had significantly more positive reactions than did individuals who believed that the parking employee or DPPS was to blame. Additionally, parking employee procedural justice was positively related to employee reactions. Hypothesis 6 posited an interaction between blame and employee procedural justice perceptions when employee reactions is the dependent variable. However, no such interaction existed using any of the operationalizations of blame. DPPS Reactiorg. A summary of the results of the analyses for Hypothesis 6 using DPPS reactions as the dependent variable can be found in Table 19. Again, all operationalizations of blame perceptions had main effect relationships with DPPS reactions. DPPS blame perceptions were negatively related to DPPS reactions. The item asking participants to balance blame between DPPS and the parking employee was positively related to DPPS reactions, indicating that participants who assigned a greater share of the blame to the parking employee had more positive reactions toward DPPS (see Table 17). The item asking participants to choose among possible blame targets was 144 related to DPPS reactions, such that participants who chose themselves as blameworthy for the ticket had more positive DPPS reactions than did those participants who blamed DPPS for the ticket. Additionally, DPPS procedural justice perceptions were positively related to DPPS reactions. Hypothesis 6, as applied to DPPS reactions, would assert that an interaction exists between operationalizations of DPPS blame and DPPS procedural justice perceptions. As can be seen in Table 19, only one of the three operationalizations of blame exhibited an interaction with DPPS procedural justice (i.e., the forced-choice item; 9 < .05). This interaction is graphed in Figure 30. As can be seen, the relationship between DPPS procedural justice and DPPS reactions is weaker when DPPS is held to blame, compared to when the parking employee or the participants themselves are blamed. Note that this . interaction is in the opposite form as would have been predicted by Hypothesis 6. Additionally, two other Blame x Justice Perception interactions existed for DPPS reactions. DPPS interpersonal justice perceptions interacted with perceptions of DPPS blame (A32=.Ol7; p < .01) and the forced-choice blame item (A32=.025; p < .05). The first interaction (see Figure 3]) reveals that the relationship between justice and reactions is weaker when DPPS is seen as more to blame. The second interaction (see Figure 32) displayed a pattern similar to that evidenced in Figure 31 - the relationship between justice perceptions and DPPS reactions existed only when the parking employee or participant was blamed for the ticket. Controlling for the Effects of Covariates Several demographic variables (e. g., those asking about past experiences with parking ticket appeals) were found to be related to DPPS reactions or Parking Employee 145 .0.—0.2.09.2. 00.3...- .00000003 max—G .000 .009. 00.0... 003:0 .0 00.30:. 0 0.0 000.600.. mama SM 00:“... $05.0... .05—0000.... 30:0“. mama :05 33 . .0520 20.2.0 III 0.005 0. main. |¢I 00.05 9 00.8.00... IQI. w.— Om Wm n... “09339)! deG 146 .0.—02:09.0: 00.02... .0.—0209.30. man:— 000 0.003 mam: .00.».020: .0 00.80:. 0 00 000.800.. mam: .:. 0.5m...— mm0E.0.... 300209000. mar—Q :05 33 “ wagsswfimidli $520.2 .2 9:0 IOI , , m.m WWW}! deCl 147 .0.—0.0000000 .00.. 3000005000. mama 000 .00w0 0.00.0 000000 .0 00.0000. 0 00 000.300.. max:— .NM 000»... 00. . . 3000.0“. 000000000000. mama :0: 33 .‘i Fri iii | . t . l lullllfl _. . 0900.300 II: 080E 0. main. I? _ 00.05 00 00.0.00... ICI _ Ll ill. .0. Wm . Wm .- m... suopmx saaa 148 Reactions. Controlling for these variables before rerunning the hypothesis tests resulted in very few differences in the results. When Question 129 (i.e., whether participants had received a parking ticket they did not think they deserved) was used as a covariate, the blame item asking participants to balance the blame between DPPS and the parking employee no longer significantly predicted reactions to the parking employee. These two items were significantly correlated. Thus, it could be argued that it is not blame perceptions necessarily that influence parking employee reactions, but the degree to which participants had lost a parking ticket appeal. The results of this analysis should be interpreted with caution. Similarly, controlling for the manipulation of supervisor blame from Part 1 eliminated the correlation between parking employee reactions and the item asking participants to balance blame between DPPS and the parking employee. Although the relationship between the manipulation and blame perceptions was not significant, controlling for the manipulation nevertheless resulted in the relationship between blame and parking employee reactions becoming non-significant. DISCUSSION Overall, the results of the hypotheses tests provided somewhat encouraging results, both for the antecedents of blame cognitions for a promotion decision, as well for the moderating effect of blame on the relationship between justice perceptions and workplace reactions. The blame that is placed on a particular entity sometimes appears to affect the way in which perceptions of that entity’s fairness leads to subsequent reactions. However, these results are preliminary, as they were based on a student sample and non- work situations (i.e., a scenario study and a parking violation). Additionally, although the hypothesized interactions were found more frequently than might be explained by 149 chance, the increments in prediction gained by these interactions were often small. Finally, there was a lack of clear evidence for the distinctiveness of perceptions of each source of justice, necessitating the need for caution when interpreting any of the results discussed below. In the next few sections, I will briefly recap the salient findings from this research, its limitations, and its implications for both research and practice. Source Perspectives on Justice (Hypotheses 1—4) Substantial support was provided Hypotheses 1-4 on the basis of the present analyses. Procedural justice perceptions directed to a source (i.e., the committee or the supervisor) were indeed related to reactions to that source. The organization procedural justice measure did not add to prediction of supervisor reactions (i.e., supervisor OCB’s and commitment to the supervisor) over and above the effects of supervisor justice. In other words, reactions to the supervisor could be said to be determined by supervisor- related justice perceptions only. Similarly, organization procedural justice was related to organizational reactions (i.e., organizational OCB’s and organizational commitment). However, for the organization, supervisor justice perceptions still added to the prediction of organizational reactions over and above organization justice perceptions. Afier partialling out organizational procedural justice, however, the relationship between supervisor procedural justice and organizational reactions were lower than the corresponding zero-order correlations. This fact indicates that, while supervisor procedural justice perceptions do indeed contribute to organizational reactions over organizational justice perceptions, some of the effect of supervisor justice perceptions may be mediated or duplicated by organizational justice perceptions. 150 Thus, it appears as though a spillover effect may exist when one considers workplace justice perceptions from a multiple-source perspective. This study found that perceptions of the supervisor’s justice may generalize to overall reactions toward the organization. This phenomenon may not be surprising, given that supervisors are employed by organizations, and can therefore be said to be a part of the organization. If a supervisor is to blame, then the organization is also likely seen as deserving some sort of reaction by association. However, the opposite case was not found in this study. If an organization were to blame, these perceptions did not generalize to reactions to the supervisor. One interesting implication of the current study’s design, given this phenomenon, is how to conceptualize the ‘organization’ in the minds of participants and employees. The present study used an ‘upper management committee’ to represent the organization as a whole. However, it could also be argued that the upper management committee could be seen as an entity distinct from a conceptualization of an anthropomorphized organization (see the sections below for a discussion of more evidence for this possibility). That is, the manipulation of the upper management committee’s blame may not have resulted in perceptions of the organization’s blame as much as it did perceptions of the upper management committee’s blame; any relationship between the manipulation and perceptions of organizational blame may have merely been spillover similar to that evidenced by the manipulation of supervisor blame. 1n the present study, use of ‘the upper management committee’ was seen as an appropriate way to induce general perceptions of the organization. However, there may have been other more appropriate ways to do so, although the extant scientific literature is lacking in this respect. To what 151 extent are general impressions of the organization’s justice and blame the result of the actions of an upper management committee? The CEO? The CFO? The parent corporation? The human resources department? The supervisor? What weights are given to each of these? Research on overall organizational perceptions generated on the basis of interactions with different parts of an organization is needed, and could help with research on and manipulations of general organizational fairness and blame. Supervisor and Committee Blame (Hypothesis 5) The antecedents of supervisor blame functioned fairly well. Recall that there were several ways to operationalize causality and justifiability for the supervisor - there were six different causality / justifiability combinations (plus three more if intentionality is also considered). Using the perceived supervisor blame variable as the dependent variable, four of the six combinations resulted in interactions that matched the pattern predicted by Hypothesis 5. Additionally, analyzing individuals who chose either the supervisor or ‘committee’ as the ultimate blameworthy agent for the promotion decision, one of the six combinations produced the hypothesized interaction. Furthermore, an analysis of individuals who chose either the supervisor or ‘nobody’ as being to blame showed that two of the six combinations showed the interaction as hypothesized. Overall, 38.9% of the analyses run for supervisor blame resulted in the hypothesized interactions. If one expands the operationalization of justifiability (c.g., using low perceived intentionality as a proxy for high justifiability), even more similar interactions are found. The interactions were spread fairly evenly across different ways of operationalizing causality, justifiability, etc. — no one way seemed to be more likely to produce the interaction. One substantial caveat that should be offered at this point, however, is the magnitude of the 152 interactions observed. Although the regressions of blame on causality and justifiability wuld at times result in quite substantial proportions of variance explained (c.g., 322.40 for the regression of supervisor blame on the interaction between participants’ choice of causal agent and perceived supervisor justification), the incremental variance of the interaction term could be small (ranging from 1-3% of the total variance). Thus, even though the interactions were found as hypothesized, the added benefit of the interaction in predicting blame over simple main effects may be limited. The antecedents of committee blame did not function as well. None of the 12 hypothesis tests resulted in an interaction between committee causality and committee justification for any of the operationalizations of blame. Committee blame appeared to be a result of direct main effect of both causality and justifiability. That is, greater blame was placed on the committee when it made (or was perceived to have made) the decision, and greater blame was placed on the committee when it was less justified (or was perceived as such). It is puzzling as to why the antecedents of blame would function as hypothesized for the supervisor but not for the upper management committee. One possible reason is that perhaps there is a different means / system of assessing blame in different levels of an organizational hierarchy. Individuals at an upper level of an organization may be perceived as bearing responsibility for an event even if they were not directly involved, merely due to the position in which they are placed (c.g., Heider’s, 1958, level of association). The existence of a committee (i.e., comprised of multiple individuals) may also have led participants to be less willing to place blame in the hypothesized way (i.e., 153 they may have been more willing to think the offense due to one or two particular individuals within the committee, and less willing to blame the entire committee). Finally, the lack of support for this hypothesized relationship could be due to the manipulations themselves. First, participants may not have believed certain parts of the story (the manipulation of a not-justified committee was set up through hearsay from a coworker). There may have been sufficient doubt, even given the information provided, to prevent blame fiom being assigned to the committee in the hypothesized way. Additionally, note that there was a tendency for participants to assign causality to the upper management committee, even in cases when the decision (as described in the scenario) was made by the supervisor. Thus, there may have been some error induced into the measurements that could be a result of either participant attention (i.e., they missed the manipulation) or previous participant experience (i.e., in the past, participants have learned that committees are usually more causal in promotion decisions than are supervisors; cf. Kelley, 1967). Blame as a Moderator of Justice / Reactions Relationships (Hypothesis 6) L231 Across all operationalizations of blame (i.e., as manipulated, perceived, chosen), some support was found for the ability of blame to moderate the relationship between perceptions of justice for a particular source and reactions to that source. This relationship was not always produced, however. Of the 12 tests that could have produced an interaction to support Hypothesis 6, only three did (i.e., 25% overall). Although the rate of finding the predicted interactions was low, the frequency was still high enough to support an assertion that they happened at a greater rate than could be expected as a result 154 of Type I error. alone. Additionally, in all instances in which an interaction was found, the pattern matched the predicted one - the relationship between justice perceptions and reactions to a source were greater when the source was perceived to be to blame, rather than not to blame. If this were simply a matter of Type I error, one would expect the pattern of the obtained reaction to vary from one instance to another. However, just as with the regressions for supervisor blame, the amount of incremental variance predicted by the interaction term was fairly small (ranging from 1.3 to 2.2% of the variance in reactions). Thus, the practical significance of taking blame attributions into account is still in question. In a laboratory study, it is typical to attempt to make the manipulation as clear as possible, thus allowing for a large effect size that might not be ordinarily observable in the field due to the uncontrollability of the environment and measurement difficulties. However, the present laboratory study did not produce effect sizes that were very large at all. Several reasons could exist to explain the small effect sizes. First, the hypotheses may not be true. Blame attributions may not moderate the relationship between justice and reactions; the results found might have been due to chance. Second, the effect of blame might in reality be small. Third, the nature of the study might have led to the magnitude of the effect sizes. Even though this was a laboratory study, and material was presented to participants in a manner that was intended to result in clear attributions of causality, reSponsibility, and blame, it is possible that a better laboratory experiment could have been designed. For instance, the artificiality of the scenario methodology could have decreased participants’ motivation to think carefully and respond accurately; a more involving study placing participants in actual situations may be more likely to produce 155 more substantial effects. Or, the limited amount of information provided in the scenario might have led participants to be reluctant to make attributions about supervisor or committee intentions — again, it may be useful to conduct a controlled participatory laboratory experiment in which more information about sources’ motives could be provided. This experiment should be replicated using a more realistic methodology in order to explore more fully the magnitude and existence of the hypothesized interactions. In contrast, two points provide justification for increased optimism about the moderating role of blame. First, although the overall rate of finding the predicted interactions was low, this rate included both manipulated blame as well as perceived blame. However, if one were to examine only instances of perceived blame (i.e., ratings of supervisor blame, organization blame, and choices between them), the rate increases to 37.5%; the predicted interactions were found only for perceptions, not manipulations, of blame. It was shown that the manipulation of blame did lead to perceptions of blame (especially for perceptions of supervisor blame), but the relationship was not perfect. Participants derived perceptions of blame from sources other than the manipulation. However, it appears that, no matter how participants derived their perceptions of blame, the perceptions were more likely to function as hypothesized. In organizational (i.e., ‘real-life’) settings, employees do not act directly on the basis of a manipulation; it is always the perceptions of organizational events that determine employee cognitions and reactions. Future research should continue to work toward a more complete understanding of the antecedents of these perceptions. Second, the rate of the discovery of interactions is based only on a narrow definition of the constructs being assessed. However, if one also examines other facets of 156 justice perceptions (i.e., lnforrnational and Interpersonal Justice), then other interactions also occur. Again, the patterns of these interactions when graphed do not show random relationships, but moderating relationships in the direction predicted by Hypothesis 6. It appears as though the moderating role of blame may not apply just to procedural justice, but to other forms of justice as well. The strong intercorrelations among justice variables in the present study may limit the strength of this assertion. However, in retrospect, there were reasons why the other types of justice perceptions should have been expected to function similarly to procedural justice in these scenarios. In all eight scenarios, procedural justice had been violated - the organization or the organization’s representatives did not use fair procedures to reach a decision. However, the manipulations of justifiability / intentionality in the scenarios are confounded with the kind of information provided to the participant, as well as the honesty (i.e., interpersonal justice) that the participant perceives. The manipulations designed to eliminate any justification for the committee’s or supervisor’s actions involve hidden motives on the part of the actors that become known to the participant through hearsay from third parties. Thus, informational (i.e., false information), and interpersonal (i.e., being deceived by a supposedly trusted source) justice is also manipulated in these scenarios. More intricate designs should attempt to control for these other factors. Some of the interactions that were discovered also pose some interesting topics for discussion. For instance, consider the interaction between choice of blame agent and organization procedural justice perceptions influencing organizational commitment (see Figure 24). As opposed to the operationalizations of blame that involved one source only (c.g., the supervisor was or was not to blame), the choice of blame agent 157 operationalization allows for a richer examination of the nature of the moderation. In this interaction, individuals who chose ‘nobody’ as being to blame had the same relationship between justice and organizational commitment as did individuals who blamed the committee. People who blamed the supervisor exhibit no such relationship. Why are the people who do not assign blame so similar to those who blame the committee? Hypothesis 6 was originally framed such that a source must be blamed in order for the relationship between that source’s justice perceptions and reactions to that source to exist. However, this figure appears to indicate that the phenomenon may be slightly different -— the relationship between a source’s justice perceptions and reactions to that source may exist unless another specific source is blamed. Although this finding is an interesting seed for future thought, it should be considered as only that - there were other instances in which different relationships were found (cf. Figure 25). Part; Although, when found, the interactions found in Part 1 supported Hypothesis 6, the opposite result occurred for Part 2. Reactions to the parking employee were not the result of an interaction between employee blame (however Operationalized) and employee procedural justice perceptions. Only one of the operationalizations of DPPS blame interacted with DPPS procedural justice perceptions. Furthermore, the graph of this interaction revealed that the interaction was opposite that hypothesized. When DPPS was blamed for the ticket, the relationship between DPPS justice perceptions and DPPS reactions was w_e_akgr than when DPPS was not blamed. Two possible reasons exist for this phenomenon. First, the finding could have been a result of the limitations of the dependent variable measurement. The item analyses 158 showed evidence for positive skew as a result of floor effects — participants (especially those who blamed DPPS) had extremely low reactions as measured by the items. These items were constructed for the present study, but it was not anticipated that responses to these items would be as low as were found. Thus, it is possible that the counterhypothesized findings could be a result of item calibration; if the measures had been written to allow for greater variance in responses at lower levels, the same interactions as found in Part 1 might have been found in Part 2. Alternatively, the differences between the findings in Part 1 and Part 2 could be the result of differences between the Parts of the experiment. There are four (confounded) distinctions between Parts 1 and 2: 1) differences in the quality of dependent variable measurement; 2) differences in the content (work setting vs. parking situation); 3) differences in realism (hypothetical event vs. real event); and 4) the extent to which participants interacted with each source (i.e., participants may not have ever interacted directly with DPPS or the parking employee, but were described as having interacted with the supervisor in the scenario). Thus, the relationships hypothesized may generalize only to work settings, but not settings in other aspects of life. Or, the relationships found in Part 1 resemble how people (including the present author) m that they would react, but the relationships found in Part 2 indicate how people actually react. It is important that future research explore the reasons behind the discrepancies between Parts 1 and 2. Using better calibrated and more reliable measures, one could (for instance) study both real-world workplace situations (or more realistic simulations) and hypothetical traffic violation scenarios in order to determine the precise causes. 159 Limitations of the Current Study While the preceding paragraphs discussed particular analytical findings as they applied to the limitations of the design, there are a number of additional limitations that limit the success of the present study. First, there is the limitation regarding the sample and experiment used. While the theories proposed in the present study were meant to apply to modern employees in real work settings, the desigi used fictitious work scenarios with a sample of participants that has had limited experiences with promotions. Thus, the fact that participants are not optimally familiar with the situation used in the scenario may add one additional reason for the differences between Part 1 (promotion situation) and Part 2 (parking ticket study limited to participants who had actual experience with a parking ticket), including the problem of the lack of clear empirical convergence among the scales measuring perceptions of different sources. Two of the variables in the study (organization procedural justice and organization informational justice) displayed relationships with participants’ experiences with promotions, such that individuals who have previously been up for promotions at work perceived the organization as less fair. However, none of the other variables in the study were related to experience with promotions. At best, there is only slight evidence to suggest that the relationships found may function differently when studied using a more experienced sample, although there is similarly no evidence to prove that the relationships will stay the same. As with any laboratory study using student participants, these findings need to be replicated in the field before definitive conclusions are drawn. Another limitation concerns the types of measurement used. First, consider Byme and Cropanzano’s (2000) measure of justice. This measure exhibited high 160 intercorrelations 1) among measures for the same source, and 2) of the same type of justice perceptions between sources. Byme and Cropanzano reported intercorrelations that were similar to the ones found in the present study, but also showed evidence for divergence among the constructs through CFA. However, the intercorrelations found in the pr0sent study may also be the result of the limited information provided to participants as a limitation of a scenario design. Had participants been given more information about the situation (c.g., if they were measured in their actual workplace), it is possible that more divergent impressions may have been formed. One other possibility regarding the intercorrelations among perceptions for the same source has to do with the order in which the items were presented. In this study, participants responded to items in the same order in which they were presented by Byme and Cropanzano (2000; i.e., items for each scale were scrambled within source). Had these items been presented as separate constructs (i.e., sequentially, perhaps with construct titles), empirical divergence among constructs might have increased. One other major limitation of this dissertation is the lack of clear evidence for conceptual convergence of the measures used in this study. The intercorrelations among scales were high, and factor analyses conducted on multiple scales often showed evidence for poor fit, or for better fit if the items were wllapsed into different scales. Most of the problems with convergence had to do with the same construct (c.g., interactional justice) measured for different sources of justice, especially when the items were almost identically worded. It could be argued that one could expect there to be strong correlations among some of these measures, especially with the similarity in item content. However, there is not much empirical evidence to support this assertion over a 161 counterclaim that participants do not make clear distinctions between different sources of justice. The interpretations in this dissertation were made given the assumption that divergence among perceptions of different sources does indeed exist. However, until more research can be conducted to establish better the divergence of these perceptions, the interpretations offered in this dissertation should be taken with an accompanying dose of caution. Another limitation worthy of mention has to do with the items of causality and blame. The intent of the original study was to use dummy-coded variables for the manipulations of causality and blame as the effects in the moderated regession analyses. However, due to the fact that the manipulations were not completely effective, it happened that perceptions of the manipulations were more predictive of dependent variables and functioned more as moderating variables. The perception items used in the current study were intended originally to have been mere manipulation checks, to verify that the manipulations had achieved their desired effects. Thus, they were included as one-item measures of general causality and blame for each source. It is possible that one- item measures may lack internal consistency reliability, therefore limiting the ability of the present study to detect more effects than it did; more reliable measures of causality and blame may have improved the study’s findings. Nevertheless, the measures of causality and blame still interacted in the ways predicted by Hypotheses 5 and 6 (in Part 1), and there is evidence to suggest that global perceptions may be adequately assessed with one item (Y oungblut & Casper, 1993). It would behoove future research to focus on perceptions, rather than manipulations, of causality and blame, and also to develop longer, reliable measures of the constructs. 162 Finally, although it was mentioned earlier, the floor effect in the reactions measures used in Part 2 of the study may have substantially influenced the results found, making the nature of the interaction suspect. Future endeavors applying blame and justice to new situations (c.g., parking tickets) should devote more effort prior to beginning the study to develOp reactions measures capable of capturing the firll variability in reactions across all types of participants. Directions for Future Research and Application There are a number of avenues that should be pursued in future research as a result of both the theoretical and empirical sections of this manuscript. First, of course, is a need to replicate this study using more appropriate samples and situations, and with different measurements (survey, as well as behavioral). One of the primary goals of these replications should be to establish further whether there is indeed conceptual divergence among perceptions of different sources of justice. Simultaneously, one should begin to take great care with the consistency of the type of assessment desired. Part 1 of the present study dealt with blame regarding a very specific situation (i.e., one promotion decision). However, the measures of justice and reactions were written in such a way as to pertain to overall perceptions across multiple situations. The fact that multiple situations were not available to the present participants could have limited the applicability of the scales. However, the nature of blame perceptions may change if one were to extend these results to a more applied situation. That is, if employees have had the opportunity to regard multiple events and incidents in their workplace, how is blame assigned? Is blame assigned individually for each outcome, with the hypothesized interactions present? Or, is an overall (i.e., aggregate) perception of general blame 163 acquired, which moderates the relationship between general fairness impressions and outcomes? Within the justice literature, the distinction between these two types of justice (single event vs. general overall impressions) has not been clearly explored; multiple studies have explored each type of justice. Careful attention should be applied to the type of justice events under investigation, and one should interpret blame interactions accordingly. Another direction for productive research is to study the precise nature of the blame/justice interaction in predicting reactions to sources of justice. In Part 2 of the study, the source that is blamed for the outcome is the target of more negative reactions overall than sources who are not blamed (e. g., see Figure 30). That is, blamed sources experience more negative reactions regardless of their perceived justice. However, certain interactions in Part 1 (c.g., see Figure 24) imply that a source perceived as blameworthy may- elicit more positive reactions than if that source were not perceived as blameworthy, if that source is also perceived as being high in justice (i.e., high justice may ‘atone’ for blame). One question that could be answered with future research involves when each reaction could occur. For instance, is the interaction found in Figure 24 related to constructs such as forgiveness for organizational wrongdoers (c.g., Aquino, Tripp, & Bies, 2001)? Similarly, what leads participants to display very negative reactions regardless of justice? Is it a function of the valence of the outcome, such that a critical threshold of violations is reached (cf. Gilliland et al., 1998)? Previous experience with promotions (i.e., being denied a promotion) was related to perceptions of causality and blame in Part 1 of the study, implying that organizational experiences may lead employees to create general frames from which to interpret 164 organizational events. Thus, another interesting line of study would explore the baselines of such perceptions in organizations, as was done recently by Cobb et al. (1997). Across individuals, or across organizational situations, is there a general tendency to attribute causality and blame to particular sources? The answer to this question has interesting theoretical as well as practical implications. Finally, now that there is at least initial evidence to suggest that blame may be a moderator in justice relationships, research needs to explore the antecedents of blame attributions in the workplace. The theoretical section of this dissertation proposes a samwe of many of the situational and individual difference variables that may influence tendencies to place blame on particular sources within an organization (c.g., turnover, empowerment, corporate culture, or even individual difference variables such as locus of control or belief in a just world). An understanding of these factors would give organizations useful mechanisms to understand what may be happening within their specific organization. Aside from merely studying the antecedents of blame within their organization, orgarnizations may also be able to use knowledge of the interactions between blame, justice, and reactions to improve conditions within the organizations (given, of course, that these results are replicated in more applied settings). First, the evidence provided here suggests that reactions to a supervisor are a result only of perceptions of the supervisor’s justice; justice perceptions of the rest of the organization do not add to this prediction. Thus, if an organization wants to change behaviors and commitment directed at a specific supervisor, the organization should concentrate the intervention on how that specific supervisor is perceived. If, however, the organization is concerned about 165 behaviors and commitment that is directed at the organization instead of at one particular supervisor, then the organization should work to change perceptions not only of the organization in general, but of the supervisor as well, given the spillover effect that supervisor causality and justification have for blame attributed to the organization. Changing the fairness perceptions of the organization may not result in the expected organizational reactions if employees still perceive their supervisor to be unfair. Additionally, given the pattern of the interactions found in Part 1 (c.g., see Figure 38), an organization may wish to tailor judgements of blame to a particular source depending on how perceptions of justice of each source are perceived. For instance, Figure 38 shows that, when the organization’s procedural justice is perceived as low, organizational commitment may be greater if the supervisor is blamed as compared to when the organization is blamed. Thus, an organization having problems with managing fairness perceptions of a particular source may attempt to shift blame perceptions to another source, if it is found that it would be easier to do so than to alter fairness perceptions. The scenarios used in the present study provide an example. If an organization believes that a supervisor’s policies will not be perceived as fair, they could highlight a causal influence that the supervisor does not control (c.g., a disaster), that would shift blame away from the supervisor. Although this, as well as most of the research on justice in organizations, could be used to unethically deceive employees in order to benefit the organization, it is hoped that organizations could also etlnically use this information in situations in which employees may be making incorrect attributions about causality, justifiability, and blame. 166 Several of the suggestions for managerial practice outlined above involve altering how employees perceive the causality, justifiability, and blame for a particular event. In order to more efficiently do so, orgarnizations will need an understanding of how to best communicate information to employees so as to achieve the desired impression. One way to do so is through the use of explanations, or social accounts (c.g., Bies,1987; Scott & Lyman, 1968). Research has shown that different types of explanations for organizational procedures are capable of having different effects on reactions to those procedures (c.g., Horvath, Ryan, & Stierwalt, 2000). It is necessary that the processes underlying how individuals form perceptions of causality, justifiability, blame, and justice be more fully understood in order to discover ‘levers’ by which these perceptions may be affected. Conclusion The present study’s integation of attribution and justice theories was productive. First, this research contributes to attributions research with its empirical examination of how people make causality and blame attributions among multiple external sources. Second, this research contributes to justice research by further delineating its intricacies. Early research on organizational justice simplistically related global perceptions of an organization’s justice to a variety of reactions to the organization. This irnitial conceptualization of the relationship was beneficial to both theory and practice, as it highlighted the need to consider the extent to which employees found organizational practices to be fair. More recent research, however (c.g., Byme & Cropanzano, 2000) have identified the importance of considering the type of justice by illustrating how different types of justice are related to different types of employee 133015005 (i.e., ‘0 tlne SUpervisor vs. toward the organization). This work improves upon earlier justice research 167 by increasing the fidelity with which we understand employee reactions. Given the present findings, the current research further improves upon recent work in organizational justice by going one step further. This study shows that, instead of there being simple relationships between justice perceptions for a source and reactions to that source, that this relationship is moderated by whether that source is blamed for an event. Thus, the contribution of the present study can be said to be in line with an established direction of justice research, by allowing for a more complete and precise understanding of justice relationships. However, as stressed earlier, given the issues associated with the scales used in this study, the interpretations and conclusions of this research should be used cautiously until future research can be conducted. 168 REFERENCES Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed), Advances in experimental social psychology (V 01. 2, pp. 276-299). New York: Academic Press. Alicke, M. D. (2000). Culpable control and the psychology of blame. Psychological Bulletin, 126, 556-574. Allen, M. W., & Meyer, J. P. ( 1990). The measurement and antecedents of affective, continuance, and normative commitment to the organization. Journal of Occupationpl Psycholo 63, 1-18. Aquino, K., Tripp, T. M., & Bies, R. J. (2001). How employees respond to personal offense: The effects of blame attribution, victim status, and offender status on revenge and reconciliation in the workplace. Joumal of Applied Psychology, 86, 52-59. Bies, R. J. (1987). The predicament of injustice: The management of moral outrage. In L. L. Curmnings & B. M. Staw (Eds.) Research in Organizational Behavior (Vol. 9, pp. 289-319). Greenwich, CT: JAI Press, Inc. Bies, R. J. (2001). Interactional (In)justice: The sacred and the profane. In J. Greenberg & R. Cropanzano (Eds), Advancces in Organizational Justice (pp. 89-118). Stanford, CA: Stanford University Press. Bies, R. J ., & Moag. J. F. (1986). Interactional justice: Communication criteria of fairness. In R. J. Lewicki, B. H. Sheppard, & M. H. Bazerman (Eds.) Research on negotiations in organizations (Vol. 1., pp. 43-55). Greenwich, CT: JAI Press. Bobocel, D. R., & Farrell, A. C. (1996). Sex-based promotion decisions and interactional fairness: Investi gating the influence of managerial accounts. Journal of Applied Psychology, 81, 22-35. Bodenhausen, G. V. (1990). Second-guessing the jury: Stereotypic and hindsight biases in perceptions of court cases. Joumpl of Applied So_cial Psychology, 20, 1112- 1121. Bodenhausen, G. V., & Wyer, R. S., Jr. (1985). Effects of stereotypes on decision making and information processing strategies. Journal of Personality and Social Psychology, 48, 267-282. Bradfield, M., & Aquino, K. (1999). The effects of blame attributions and offender likableness on forgiveness and revenge in the workplace. Journal of W. 607-631. 169 Brockner, J ., & Wiesenfeld, B. M. (1996). An integrative framework for explaining reactions to decisions: Interactive effects of outcomes and procedures. Psychological Bulletin, 120, 189-208. Byme, Z. S., & Cropanzano, R. (2000, April). To which source do I attribute this fairness? Differential effects of multi-foci justice on organizational work behaviors. Paper presented at the 15‘h Annual Conference of the Society for Industrial and Organizational Psychology, New Orleans, LA. Carver, C. S., Lawrence, J. W., & Scheier, M. F. (1996). A control-process perspective on the origins of affect. In L. L. Martin & A. Tesser (Eds) Striving and feeligg: Intergctiogs mamasiffectpand self-regulation (pp. 11-52). Mahwah, NJ: Lawrence Erlbaum Associates. Cawley, B. D., Keeping, L M., & Levy, P. E. (1998). Participation in the performance appraisal process and employee reactions: A meta-analytic review of field investigations. Journal of Applied Psychology, 83, 615-633. Cheng, P. W. (1997). From covariation to causation: A causal power theory. Psychological Review. 104. 367-405. Cobb, A. T., Vest, M., & Hills, F. (1997). Who delivers justice? Source perceptions of procedural fairness. Joumpl of Applied Social Psychology, 27, 1021-1040. Cohen, J. (1988). Statistical fipo wer analysis for the belgvioral sciences (2nd Ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Cohen, R. L. (1982). Perceiving justice: An attributional perspective. In J. Greenberg & R. L. Cohen (Eds), Eguig and justice in socjal bghpvior (pp. 119-160). New York: Academic Press. Colquitt, J. A. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied Psychology, 86, 386-400. Colquitt, J. A., Conlon, D. E., Wesson, M. J ., Porter, C., & Ng., K. Y. (2001). Justice at the millennium: A meta-analytic review of 25 years of organizational justice research. Journal of Applied Psychology, 86, 425-445. Cropanzano, R., & Prehar, C. A. (1999, April). Using social exchange theory to distinguish procedural from interactional justice. Paper presented at the 14‘h Annual Conference of the Society for Industrial and Organizational Psychology, Atlanta, GA. Dadzie, E. W. (1994). Behavioral consguences of continuing injustice. Unpublished doctoral dissertation, Duke University. 170 Feather, N. T. (1996). Extending the search for order in social motivation. Psychological Inquiry, 7, 223-229. Feldman, J. M. (1981). Beyond attribution theory: Cognitive processes in performance appraisal. Journal of Applied Psychology, 66, 127-148. Fincham, F. D., & Schultz, T. R. (1981). Intervening causation and the mitigation of responsibility for harm. British Journal of Social Psychology, 20, 1 13-120. Fishbein, M., & Ajzen, I. (1973). Attributions of responsibility: A theoretical note. Journal of Experimental Social Psycholo 9, 148-153. Fiske, s. T., & Taylor, s. E. (1991). Social cogpition (2“d 13.1.). New York: McGraw-Hill, Inc. Folger, R. (1993). Reactions to mistreatment at work. In K. Muminghan (Ed.), Social psychology in organizations: Advances in theory and research (pp. 161-183). Englewood Cliffs, NJ: Prentice Hall. Folger, R., & Cropanzano, R. (1998). Organizationial Justice ;and Human Resource Management. Thousand Oaks, CA: Sage Publications. Folkes, V. S. (1984). Consumer reactions to product failure: An attributional approach. Journal of Consumer Research. 10. 398-409. Folkes, V. S. (1988). Recent attribution research in consumer behavior: A review and new directions. Journal of C(msumer Researcln 14. 548-565. Folkes, V. S., Koletsky, S., & Graham, J. L. (1987). A field study of causal inferences and consumer reaction: The view fiom the airport. J oumal of Consumer Research 13, 534-539. Fumham, A. (1992). Lay explanations. In J. L. McLaughlin, M. J. Cody, & S. J. Read (Eds), Explaining ones self to others: Reason-giving in a social context (pp. 83- 104). Hillsdale, NJ: LEA. Gilliland, S. W. (1994). Effects of procedural and distributive justice on reactions to a selection system. Journal of Applied Psychology, 79, 691-701. Gilliland, S. W., Benson, L. 1H, & Schepers, D. H. (1998). A rejection threshold in justice evaluations: Effects on judgment and decision-making. Organizational Behavior and Human Decision Processes 76, 113-131. Greenberg, J. (1991). Using explanations to manage impressions of performance appraisal fairness. Employee Resmnsibilities and Rights Journal, 4, 51-60. 171 Greenberg, J. (1993a). Stealing in the name of justice: Informational and interpersonal moderators of theft reactions to underpayment inequity. Organizational Belgryior and Human Decision Processes 54, 81-103. Greenberg, J. (1993b). The social side of fairness: Interpersonal and informational classes of organizational justice. In R. Cropanzano (Ed), Justice in the workplace: Approaching fairness in human resource management (pp. 79-103). Hillsdale, NJ: Erlbaum. Greenberg, J ., Pyszczynski, T., & Solomon, S. (1982). The self-serving attributional bias: Beyond self-presentation. Journ_a1 of Experimental Social Psychology. 18, 56-67. Hackman, J . R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a theory. Organizatiogal Begyior and Hunym Performancepl6, 250-2 79. Heider, F. (1958). The psychology of intgpersonal relations. New York: John Wiley & Sons, Inc. Heuer, L., Blurnenthal, E., Douglas, A., & Weinblatt, T. (1999). A deservingness approach to respect as a relationally based fairness judgment. Personalityand Sogafl Psychology Bulletin, 25, 1279-1292. Horvath, M., Ryan, A. M., & Stierwalt, S. (2000). Explanations for selection test use, outcome favorability and self-efficacy: What influences test-taker perceptions? Organizational Behavior and Human Decision Processes, 83, 310-330. Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed), Advances in experimental social psychology (Vol. 2, pp. 219-266). New York: Academic Press. Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed), Nebraska smpgsiurn on motivation 1967 (pp. 192-23 8). Lincoln, NE: University of Nebraska Press. Kelley, H. H., & Michela, J. L. (1980). Attribution theory and research. In M. R. Rosenzweig & L. W. Porter (Eds), Annual Review of Psychology (vol. 31, pp. 457-501). Palo Alto, CA: Annual Reviews Inc. Kunst, J. L., Bjorck, J. P., & Tan, S-Y. (2000). Causal attributions for uncontrollable negative events. Journal of Psychology and Christianig, 19, 47-60. Leiber, M. J., & Woodrick, A. C. (1997). Religious beliefs, attributional styles, and adherence to correctional orientations. Criminal Justice and Behavior 24, 495-511. 172 Leventhal, G. S. (1980). What should be done with equity theory? New approaches to the study of faimess in social relationships. In K. Gergen, M. Greenberg, & R. Willis (Eds), Social exchanges: Advances in theory and research (pp. 27-55). New York: Plenum Press. Lord, R. G., & Levy, P. E. ( 1994). Moving from cognition to action: A control theory perspective. Applied Psychology: An International RevieMl 335-367. Lupfer, M. B., Doan, K., & Houston, D. A. (1998). Explaining unfair and fair outcomes: The therapeutic value of attributional analysis. British Journal of Socijal, Psychology, 37, 495-511. Lurigio, A. J ., Carroll, J. S., & Stalans, L. J. (1994). Understanding judges’ sentencing decisions: Attributions of responsibility and story construction. In L. Heath, R. S. Tindale, J. Edwards, E. J. Posavac, F. B. Bryant, E. Henderson-King, Y. Suarez- Balcazar, & J. Myers (Eds), Applications of heuristics a_ngl biases to socijal issues (pp. 91- 115). New York: Plenum Press. Masterson, S. S., Bartol, K. M., & Moye, N. A. (2000, April). Interactional and procedural Lustice: Type versus source of fairness. Paper presented at the 15m Annual Conference of the Society for Industrial and Organizational Psychology, New Orleans, LA. Masterson, S. S., Lewis, K., Goldman, B. M., & Taylor, M. S. (2000). Integrating justice and social exchange: The differing effects of fair procedures and treatment on work relationships. Academy of Management Journal, 43, 738-748. Messick, D., & Allison, S. T. (198 7) Accepting unfairness: Outcomes and attributions. Representative Research in Social Psychology, 17, 39-51. Montada, L. (1992). Attribution of responsibility for losses and perceived injustice. In L. Montada, S-H. F ilipp, & M. J. Lerner (Eds), Life crises and experiences of loss in adulthood (pp. 133-161). Hillsdale, NJ: Lawrence Erlbaum Associates. Moorman, R. H. (1991). Relationship between organizational justice and organizational citizenship behaviors: Do fairness perceptions influence employee citizenship? Journal of Applied Psychology, 76, 845-855. Ployhart, R. E., & Ryan, A. M. (1997). Toward an explanation of applicant reactions: An examination of organizational justice and attribution frameworks. Organizational Behavior and Human Decision Processes, 72, 308-335. Pyszczynski, T ., Greenberg, J ., & Holt, K. (1985). Maintaining consistency between self-serving beliefs and available data: A bias in information evaluation. Personaligy and Social Psychology Bulletin, 11, 179-190. 173 .,.I" -_.‘-’ Pyszczynski, T., Greenberg, J ., & LaPrelle, J. (1985). Social comparisons after success and failure: Biased search for information consistent with a self-serving conclusion. Journal of Experimental Social Psycholo 21, 195-211. Schultz, T. R., Schleifer, M., & Altman, I. (1981). Judgments of causation, responsibility, and punishment in cases of harm-doing. gapadian J ournpl of Behavioral Science, 13, 238-253. Schultz, T. R., & Wright, K. (1985). Concepts of negligence and intention in the assignment of moral responsibility. Canadian Joumypl of Behavioral Science. 17, 97-108. Scott, M. B., & Lyman, S. M. (1968). Accounts. American Sociolcgical Review 3, 46-62. Shaver, K. G. (1985). The attribution of blame: Causalig, responsibility, and blameworthiness. New York: Springer-Verlag. Shepelak, N. J. (1987). The role of self-explanations and self-evaluations in legitimating inequality. American Sociolog' cal Review, 52, 495-503. Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The role of distributive, procedural, and interactional justice. Journal of Applied Psychology, 82, 434-443. Skitka, L. J ., & Tetlock, P. E. (1993). Of ants and grasshoppers: The political psychology of allocating public assistance. In B. A. Mellers & J. Baron (Eds), Psychological perspectives on Justice: Theoyry and Applications (pp. 205-233). New York: Cambridge University Press. Sweeney, P. D., & McFarlin, D. B. (1993). Workers’ evaluations of the “ends” and the “means”: An examination of four models of distributive and procedural justice. Organizational Behavior and Human Research Processes, 55, 23-40. Thibaut, J. W., & Walker, L. (1975). Procedural justice: A psychological perspective. Hillsdale, NJ: Lawrence Erlbaum. . Tsujimoto, R. N. (1979). Kohlberg’s moral judgment stages: Perceptions of Injustice. The Journal of Psychology, 101, 79-82. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science 185, 1124-1131. Tyler, T. R., & Bies, R. J. (1990). Beyond formal procedures: The interpersonal context of procedural justice. In J. S. Carroll (Ed), Applied social psychology and organizational settings (pp. 77-98). Hillsdale, NJ: Lawrence Erlbaum Associates. 174 Ill -c—r van den Bos, K., Bruins, J ., Wilke, H. A. M., & Dronkert, E. (1999). Sometimes unfair procedures have nice aspects: On the psychology of the fair process effect. Journal of Personalig and Social Psychology. 77. 324-336. van den Bos, K., Lind, L. E., Vermunt, R., & Wilke, H. A. M. (1997). How do I judge my outcome when I do not know the outcome of others? The psychology of the fair process effect. Journg of Personplig ping! Social Psychology, 72, 1034-1046. van den Bos, K., Wilke, H. A. M., Lind, E. A., & Vermunt, R. (1998). Evaluating outcomes by means of the fair process effect: Evidence for different processes in fairness and satisfaction judgements. hump] of Personali and SocigPsycholo 74, 1493- 1503. Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Ifiychologjcal Review.92, 548-573. Weiner, B. (1986). An attributional theop/ of motivation pnd emotion. New York: Springer-Verlag. Weiner, B. (1993). On sin versus sickness: A theory of perceived responsibility and social motivation. American ngchologist, 48, 957-965. Weiner, B. (1995). Jrfirynents of responsibiliga A foundation for a theory of social conduct. New York: The Guilford Press. Weiner, B., Amirkhan, J ., Folkes, V. S., & Verette, J. A. (1987). An attributional analysis of excuse giving: Studies of a naive theory of emotion. Joumpl of Personalig and Social Psychology, 52, 316-324. Williams, L. J ., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17, 601-618. Wood, R. E., & Mitchell, T. R. (1981). Manager behavior in a social context: The impact of impression management on attributions and disciplinary decisions. Organizational Behavior and Human Decision Processes, 28, 356-378. Youngblut, J. M., & Casper, G. R. (1993). Focus on psychometrics: Single-item indicators in nursing research. Research in Nursing and Health, 16, 459-465. 175 APPENDIX A Vignettes 176 Vignette 1 Locus: Supervisor Responsibility of Supervisor: Intentional (not Justified) Responsibility of Organization: Intentional (not Justified) Hypothesized Blame: SUPERVISOR You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "The supervisor has the final decision regarding a promotion decision, which the supervisor will communicate to the employee at the performance review meeting. Although a committee from upper management will review the employee's performance and make a recommendation to the supervisor, this is only a suggestion to the supervisor. The supervisor is completely free to disagree with the committee's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from the upper management committee. According to the memo, the committee's recommendation is that you should be denied the promotion. A coworker tells you that he overheard a member of the upper management committee remarking that they don't like your department, so they recommended to deny you the promotion. When your supervisor sits down with you at the performance review meeting, the supervisor tells you that he also does not think that you should get the promotion, so he has decided not to promote you. After the meeting, you talk to a coworker, who tells you that your supervisor has a record of passing good people over for promotions so as not to lose good employees, and that the supervisor was overheard last week saying what a good job you've been doing. 177 Vignette 2 Locus: Supervisor Responsibility of Supervisor: Intentional (not Justified) Responsibility of Organization: Justified Hypothesized Blame: SUPERVISOR You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company’s upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization’s employee handbook, the promotion process works this way: “The supervisor has the final decision regarding a promotion decision, which the supervisor will communicate to the employee at the performance review meeting. Although a committee from upper management will review the employee’s performance and make a recommendation to the supervisor, this is only a suggestion to the supervisor. The supervisor is completely flee to disagree with the committee’s recommendation without suffering any negative consequences.” A week before your performance appraisal meeting, you receive a memo fi'om the upper management committee. According to the memo, the committee’s recommendation is that you should be denied the promotion. The memo explains that the factory of one of the company’s suppliers was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company cannot make enough products right now to pay all of its bills. The upper management committee is taking many different actions to cut costs in order to save the company. For instance, nobody in the company will be able to fly “First Class,” and the company’s annual leadership conference, usually held at a resort in Florida, has been relocated to a meeting room inside the factory itself. Thus, the committee has decided to recommend that you not receive the promotion. When your supervisor sits down with you at the performance review meeting, the supervisor tells you that he also does not think that you should get the promotion, so he has decided not to promote you. After the meeting, you talk to a coworker, who tells you that your supervisor has a record of passing good people over for promotions so as not to lose good employees, and that the supervisor was overheard last week saying what a good job you’ve been doing. 178 Vignette 3 Locus: Supervisor Responsibility of Supervisor: Justified Responsibility of Organization: Intentional (not Justified) Hypothesized Blame: NONE You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "The supervisor has the final decision regarding a promotion decision, which the supervisor will communicate to the employee at the performance review meeting. Although a committee from upper management will review the employee's performance and make a recommendation to the supervisor, this is only a suggestion to the supervisor. The supervisor is completely free to disagree with the committee's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from the upper management committee. According to the memo, the committee's recommendation is that you should be denied the promotion. A coworker tells you that he overheard a member of the upper management committee remarking that they don't like your department, so they recommended to deny you the promotion. When your supervisor sits down with you at the performance review meeting, the supervisor tells you that he has decided not to promote you. The supervisor explains that one of the factories that supplies your department with parts was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company will not be able to make enough of your product to pay its bills. Your supervisor is taking many different steps to reduce costs within the department. For instance, he has cancelled the purchase of new computer systems for the department, and he has decided not to hire any additional employees until this crisis is resolved. Thus, the supervisor has also decided not to give you the promotion. 179 Vignette 4 Locus: Supervisor Responsibility of Supervisor: Justified . Responsibility of Organization: Justified Hypothesized Blame: NONE You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "The supervisor has the final decision regarding a promotion decision, which the supervisor will communicate to the employee at the performance review meeting. Although a committee fi'om upper management will review the employee's performance and make a recommendation to the supervisor, this is only a suggestion to the supervisor. The supervisor is completely free to disagree with the committee's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from the upper management committee. According to the memo, the committee's recommendation is that you should be denied the promotion. The memo explains that the factory of one of the company's suppliers was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company cannot make enough products right now to pay all of its bills. The upper management committee is taking many different actions to cut costs in order to save the company. For instance, nobody in the company will be able to fly "First Class," and the company's annual leadership conference, usually held at a resort in Florida, has been relocated to a meeting room inside the factory itself. Thus, the corrunittee has decided to recommend that you not receive the promotion. When your supervisor sits down with you at the performance review meeting, the supervisor tells you that he has decided not to promote you. The supervisor explains that, as soon as he heard about the earthquake, he also began to take steps to reduce expenses in the department. For instance, he has cancelled the purchase of new computer systems for the department, and he has decided not to hire any additional employees until this crisis is resolved. Thus, the supervisor has also come to the conclusion that you should not get a promotion. 180 Vignette 5 Locus: Organization Responsibility of Supervisor: Intentional (not Justified) Responsibility of Organization: Intentional (not Justified) Hypothesized Blame: ORGANIZATION You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "An upper management committee has the final decision regarding a promotion decision, which the they will communicate to the employee at the performance review meeting. Although the employee's supervisor will review the employee's performance and make a recommendation to the upper management committee, this is only a suggestion to the committee. The committee is completely flee to disagree with the supervisor's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from your supervisor. According to the memo, the supervisor's recommendation is that you not receive the promotion. A coworker tells you that your supervisor has a record of passing good people over for promotions so as not to lose good employees, and that the supervisor was overheard last week saying what a good job you've been doing. When you sit down with the upper management committee at the performance review meeting, they tell you that they have decided not to promote you. After the meeting, you talk to a coworker, who tells you that he overheard a member of the upper management committee remarking that they don't like your department, so they decided to deny you the promotion. 181 Vignette 6 Locus: Organization Responsibility of Supervisor: Intentional (not Justified) Responsibility of Organization: Justified Hypothesized Blame: NONE You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "An upper management committee has the final decision regarding a promotion decision, which the they will communicate to the employee at the performance review meeting. Although the employee's supervisor will review the employee's performance and make a recommendation to the upper management committee, this is only a suggestion to the committee. The committee is completely free to disagree with the supervisor's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from your supervisor. According to the memo, the supervisor's recommendation is that you not receive the promotion. A coworker tells you that your supervisor has a record of passing good pe0p1e over for promotions so as not to lose good employees, and that the supervisor was overheard last week saying what a good job you've been doing. When you sit down with the upper management committee at the performance review meeting, they tell you that they have decided not to promote you. The committee explains that the factory of one of the company's suppliers was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company cannot make enough products right now to pay all of its bills. The upper management committee is taking many different actions to cut costs in order to save the company. For instance, nobody in the company will be able to fly "F irst Class," and the company's annual leadership conference, usually held at a resort in Florida, has been relocated to a meeting room inside the factory itself. Thus, the committee has decided not to give you the promotion. 182 Vignette 7 Locus: Organization Responsibility of Supervisor: Justified Responsibility of Organization: Intentional (not Justified) Hypothesized Blame: ORGANIZATION You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "An upper management committee has the final decision regarding a promotion decision, which the they will communicate to the employee at the performance review meeting. Although the employee's supervisor will review the employee's performance and make a recommendation to the upper management committee, this is only a suggestion to the committee. The committee is completely flee to disagree with the supervisor's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from your supervisor. According to the memo, the supervisor's recommendation is that you not receive the promotion. The supervisor explains that one of the factories that supplies your department with parts was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company will not be able to make enough of your product to pay its bills. Your supervisor is taking many different steps to reduce costs within the department. For instance, he has cancelled the purchase of new computer systems for the department, and he has decided not to hire any additional employees until this crisis is resolved. Thus, the supervisor has decided to recommend not to promote you. When you sit down with the upper management committee at the performance review meeting, they tell you that they have decided not to promote you. After the meeting, you talk to a coworker, who tells you that he overheard a member of the upper management committee remarking that they don't like your department, so they decided to deny you the promotion. 183 Vignette 8 Locus: Organization Responsibility of Supervisor: Justified Responsibility of Organization: Justified Hypothesized Blame: NONE You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company's upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization's employee handbook, the promotion process works this way: "An upper management committee has the final decision regarding a promotion decision, which the they will communicate to the employee at the performance review meeting. Although the employee's supervisor will review the employee's performance and make a recommendation to the upper management committee, this is only a suggestion to the committee. The committee is completely free to disagree with the supervisor's recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo fiom your supervisor. According to the memo, the supervisor's recommendation is that you not receive the promotion. The supervisor explains that one of the factories that supplies your department with parts was hit by an earthquake a few days ago (you remember reading about this event in the newspaper). As a result, your company will not be able to make enough of your product to pay its bills. Your supervisor is taking many different steps to reduce costs within the department. For instance, he has cancelled the purchase of new computer systems for the department, and he has decided not to hire any additional employees until this crisis is resolved. Thus, the supervisor has decided to recommend not to promote you. When you sit down with the upper management committee at the performance review meeting, the committee tells you that they have decided not to promote you. The committee explains that, as soon as they heard about the earthquake, they also began to take steps to reduce expenses in the organization. For instance, nobody in the company will be able to fly "First Class," and the company's annual leadership conference, usually held at a resort in Florida, has been relocated to a meeting room inside the factory itself. Thus, the committee has also come to the conclusion that you should not get a promotion. 184 APPENDIX B: Measures for the First Part of the Study 185 Locus of causality To what extent did your actions cause the promotion decision you received? To what extent did your supervisor’s actions cause the promotion decision you received? 0 To what extent did the upper management committee’s actions cause the promotion decision you received? 1. Not at all 2. To some extent 3. To a moderate extent 4. A great deal 5. Completely o Whose actions were the most direct cause of the promotion decision you received? 1. My own 2. My supervisor’s 3. The upper management committee’s Justificanion ofpupervisor Given the circumstances, the supervisor’s actions were justified. Given the circumstances, there was nothing that the supervisor could have done differently. o The supervisor was forced to do what he did due to situations outside his control. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree Intentionalig of supervisor o The supervisor’s recommendation about the promotion was intentional (that is, he consciously made a decision regarding his opinion). The supervisor meant to act the way he did. The supervisor’s actions were deliberate. 1. Strongly Disagree * 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 186 Justification of the upper management committee 0 Given the circumstances, the upper management committee’s actions were justified. 0 Given the circumstances, there was nothing that the upper management committee could have done differently. o The upper management committee was forced to do what it did due to situations outside its control. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree Intentionalig; of the upper management committee 0 The upper management committee’s recommendation about the promotion was intentional (that is, the upper management committee consciously made a decision regarding their opinion). The upper management committee meant to act the way it did. The upper management committee’s actions were deliberate. l. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 29 S (b I am to blame for the promotion decision. The supervisor is to blame for the promotion decision. . . The upper management committee is to blame for the promotion decrsron. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 0 Overall, who is to blame for the promotion decision? 1. Nobody 2. myself 3. the supervisor 4. the upper management committee 187 Organization gocedural justice (from Bupe & Cropanzano, 2000) Where I work, the organizations’ procedures and guidelines are very fair. The procedures the organization uses to make decisions are not fair. (Reverse- scored) I can count on my organization to have fair policies. We don't have any fair policies in my organization. (Reverse-scored) Strongly Disagree Moderately Disagree Slightly Disagree Neither Agree nor Disagree Slightly Agree Moderately Agree Strongly Agree NP‘S‘PP‘P!‘ Organization informational justice (from ngpe & Cropanzano, 2000) I am kept informed, by my organization, of why things happen the way they do. Whether the outcome is good or bad, I always feel like I am kept informed by my organization. My organization’s decisions are made out in the open so that everyone always knows what's going on. Whether right or wrong, the organization always explains decisions to me. 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Organization intflersonal justice (from Bme & Cropanzano, 2000) The organization makes it clear to me that I am a valuable employee My organization treats me with dignity and respect. I feel the organization holds me in high regard. This organization doesn’t care how I am doing. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree 188 Supervisor procedural justice (from Bygpe & Cropanzano, 2000) Where I work, my supervisor’s procedures and guidelines are very fair. The procedures my supervisor uses to make decisions are not fair. (Reverse- scored) I can count on my supervisor to have fair policies. My supervisor doesn’t have any fair policies. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Supervisor informational justice (from Bme & Cropanzano, 2000) My supervisor keeps me informed of why things happen the way they do. Whether the outcome is good or bad, I always feel like I am kept informed by my supervisor. My supervisor’s decisions are made out in the open so that everyone always knows what's going on. Whether right or wrong, my supervisor always explains decisions to me. 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Supervisor inte_rpersonal justice (fi'om Bme & Cropanzano, 2000) My supervisor makes it clear to me that I am a valuable employee. My supervisor treats me with dignity and respect. I feel my supervisor holds me in high regard. My supervisor doesn’t care how I am doing. (Reverse-scored) l. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree 189 Organizational commitment (adapted from Allen & Meyer, 1990) I would be very happy to spend the rest of my career with this organization. I would enjoy discussing this organization with people outside it. I would really feel as if this organization’s problems are my own. I think that I would be able to easily become as attached to another organization as I am to this one. I would not feel like ‘part of the family’ at this organization. I would not feel ‘emotionally attached’ to this organization. This organization would have a great deal of personal meaning for me. I would not feel a strong sense of belonging to this organization. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree Supervisogg commitment (adapted fiom Allen & Mgwerpl 99(1) I would be very happy to spend the rest of my career with this supervisor. I would enjoy discussing my work under this supervisor with people in other departments. I would really feel as if this supervisor’s problems are my own. 11 think that I would be able to easily become as attached to another supervisor’s work group as I am to this one. I would not feel like ‘part of the family’ in this supervisor’s work group. I would not feel ‘emotionally attached’ to this supervisor’s work group. Working with this supervisor would have a great deal of personal meaning for me. I Would not feel a strong sense of belonging to this supervisor’s work group 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 190 OCB’a to the organization (adapted from Williams and Anderson, 1991) I would try my hardest to not miss a single day of work at this organization. If I knew that I had to miss work, I would let the organization know as soon as possible. If nobody saw me, I would be likely to take unauthorized breaks during the day. I would be diligent about not using the organization’s phones for personal calls. I would be diligent about not using the organization’s internet access for personal business. I would not complain about working for this company. If I saw coworkers stealing from the company or vandalizing company equipment, I would not hesitate to turn them in. I would do things the way my organization wanted me to, even if I thought they should be done differently. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree OCB’s to supervisor (adapted from Williams apd Anderson. 1991) If my supervisor got sick and missed a day of work, I would try to help him catch up when he got back. If my supervisor had a heavy workload, 11 would volunteer to help do some of his work. If my supervisor needed my advice, I would be willing to stop my work and listen to his problem. If I had some knowledge that would help my supervisor out, I would tell him. I would come in to work on my day off if my supervisor needed the help. If my supervisor seemed to be stressed, I would go out of my way to encourage him. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 191 APPENDIX C Example of the Experiment as Seen by Participants 192 Psychology Subject Pool 3 Michigan State University, East Lansing, Michigan ‘-‘Ll.\n~ a av “.- n" . 'WM'J J” .43“..‘.t‘ ”L“.— 13' 1.... .I .L ** This is how the Subject will see the Information Questionnaire ** * PM After 45 Min. of inactivity you are logged out. so please u-y to answer the questionnaire in under 45 Min. MW REACTIONS To NEGATIVE OUTCOMES COVER SHEET AND INFORMED CONSENT SUMMARY OF RESEARCH: This study is an attempt to understand some factors that influence how people react to promotion decisions in the workplace. In this study you will read a brief scenario, after which you will answer a series of questions about your reactions to the scenario. Additionally, you will be asked to respond to questions about your most recent parking ticket (if you have ever received one). ESTIMATE OF TIME: Your participation in this study should take about 30 minutes. CONDITIONS OF PARTTCIPATION: Your participation in this study is completely voluntary. You may choose not to participate at all. You may refuse to answer any question. You may discontinue this experiment at any time without penalty. CONFIDENTIALITY: Your responses will be kept confidential by the experimenters. Your privacy will be protected to the maximum extent allowable by law. RISKS AND WITHDRAWAL FROM THE EXPERIMENT: There are no risks associated with this procedure aside from the recurrence of any negative emotions experienced after your most recent parking ticket. You may withdraw from this experiment at any time without penalty. CONTACT INFORMATTON: You may contact the investigators at any time if you have questions or concerns about this experiment: Ann Marie Ryan (517) 353-8855 Michael Horvath (517) 432-7069 Additionally, if you have questions or concerns regarding your rights as a human subject, you may contact the chairperson of the University Committee on Research Involving Human Subjects (UCRIHS): David E. Wright (517) 355-2180 You indicate your voluntary agreement to participate by completing and . returning (subnritting) this questionnaire. 193 UCRIHS approval (00-745) expires 21 November 2001. INSTRUCTIONS: Read the following scenario, and answer the accompanying questions. You have a manufacturing job where job performance lends itself easily to objective measurement. That is, your job is to produce objects (machine parts, technical reports, etc.) that can be directly observed and counted in order to evaluate your performance. When you first got your job, both the supervisor and a memo from the company’s upper management told you that you had to produce a certain number of products per week in order to be eligible for the promotion. It is time for your annual performance review. You know that you have been performing well, and you are looking forward to getting that promotion. According to the organization’s employee handbook, the promotion process works this way: "The supervisor has the final decision regarding a promotion decision, which the supervisor will communicate to the employee at the performance review meeting. Although a committee from upper management will review the employee’s performance and make a recommendation to the supervisor, this is only a suggestion to the supervisor. The supervisor is completely free to disagree with the committee’s recommendation without suffering any negative consequences." A week before your performance appraisal meeting, you receive a memo from the upper management committee. According to the memo, the committee’s recommendation is that you should be denied the promotion. A coworker tells you that he overheard a member of the upper management committee remarking that they don’t like your department, so they recommended to deny you the promotion. When your supervisor sits down with you at the performance review meeting, the supervisor tells you that he also does not think that you should get the promotion, so he has decided not to promote you. After the meeting, you talk to a coworker, who tells you that your supervisor has a record of passing good people over for promotions so as not to lose good employees, and that the supervisor was overheard last week saying what a good job you’ve been doing. 1. In this spenan'p, what happpnpg alum yam prpmptjpp? O I got the promotion 0 I did not get the promotion , O I choose not to respond tothis question 2. W? 194 O the supervisor O the upper management comnrittee O I choose not to respond to this question jl'p what axtgpt gig yap; agtippg cauaa m pmmptipn dgcision you received? 0 not at all 0 to some extent 0 to a moderate extent 0 a great deal 0 completely 0 I choose not to respond to this question 4. T w n ' r 0 not at all C to some extent 0 to a moderate extent 0 a great deal 0 completely 0 I choose not to respond to this question 5. w .x i r neen mmi ’ ti ausee mti IV ‘7 0 not at all 0 to some extent 0 to a moderate extent 0 a great deal 0 completely 0 I choose not to respond to this question 6. flhogg aptjpn§ wera tha mpgt dirgt pause of tha prpmptipn gapision ypp regeivgd? O my own 0 my supervisor’s O the upper management committee’s O I choose not to respond to this question e su rvi r’ actions w ' ' red. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 8. V'rr.‘ ' n .-.r‘ r 'W;-ro l'Ji .t' .0; ° ‘,‘1-.V'ccj' 195 Ht rvis ’sa in cau the r otion decision on received? i ion rly 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 9. Ihe §pmrvigpt wa§ fprped tp gp what hp did dag to §ituatipns outside his control. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 10. 2111; sugm'ap r’§ [agpmmend‘ ation a bput thg ptomotign was intentipnal tthat i§, he gonsciously mad; a degrgrpp regargjng his oprnran). 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 1 1. lb; summispt mat tp act the way he aid. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 12. We 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 13. ' ‘r .r‘ 'r u tr" .1' H.‘ rrars"tt‘r Orrnr‘v.‘ =. 01W. 0 strongly disagree 0 disagree 196 O neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 14 ‘ W ' r ana ement committee could have 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 15 :1 hp ppmr management cpmmitteg wag forced to do what it dij due to situations outside its ' scam. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question ment c m ittee’ r comm ndation ab tthe rom i n wa in nti nal l§. thg upmr management committg gdnsgidugly made a dagj§i rdp regarding thair dpjnidn). 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 17. WW‘MM 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 18. Th e nt mm' ’ ' W 1' rate. 0 strongly disagree 0 disagree 197 O neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 19. WW 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 20. is r ° blam r the r m tion ecision. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree 0 strongly agree 0 I choose not to respond to this question 21. r na ement ommittee i t lam forth r mo i n decision. 0 strongly disagree 0 disagree 0 neither agree nor disagree 0 agree. 0 strongly agree 0 I choose not to respond to this question vralw' lame rh rmi O Nobody O myself 0 the supervisor O the upper management committee 0 I choose not to respond to this question 9mm tjpna] Prmpdhm 23. . m A. 1,. u .e H,- 10W 11 s ionzForwh- 'ru'W°‘a' 'A‘A'W . _. " any _!'0 i-.r"'_.-.ur _r ‘1...“ Ht Wt. wrr o oWr ' 01 85 WI 1 O. 198 24. 25. W w ' a ’ ' ' elines are ve O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question i tl ake it ha am v luabl 1 ee. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question r ed res the r anization ses t make (1 cisions e not fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question fair. 26. i f r anizati fwh in s w e do. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 27. W 199 O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 28. Wn O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 29. M r izati a me with di nit an res ct. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question i i ’ ci ions are ad u ' n v ne alwa sknows what's 30. in n O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 31. W 200 32. 33. 34. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question r w an' ti n alw x lains decisi n to me. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question '2 tion h l m in h' r ard. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question WW 0 Strongly Disagree 0 Modemme Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 201 35. 36. 37. 38. m P R s Inappctions; ESE! dagh statement mldw, pleasa chpose the re§mnsa that best describes hpw you f l t s rvis r ' hi n ' . In ther w rds nswer th followin uestions as h w e l l e an ' he cenario. w ' r’ es i line at v fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question ' [ear to m at am valuable m 10 ee. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question r e re rvr r uses make de isi n ar no fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question rvi ' rrne fw tin ha w o. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 202 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 39. WW5 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree O I choose not to respond to this question 40. r c me i o r a w f 1 like I m ke informed O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 41. ' r wi i i r ct. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question rt“ 11‘." 0. 'l . ‘ urn 0 . -_ 'V' t -._.w-_ . ' ’ OI U 0A I 0‘10] QJL O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 203 m su rvis r. ’ 'oing 43. 44. 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question S ’ v f 'r 1i i s. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question h ' ht or wron m u rvi r alwa 5 ex lains e isions to me. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 45. Ifegl my appdryigdr hold§ [he in high regard. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 46. Warn—6.91.02 O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 204 47. 48. 49. 50. O Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question R NI N M MEN CIT N :Intrrns th ' 'str 1 sere ndt thefll win items, in s r vi . e e ke in 'nd th e e 'tem r fer the or anization as a w n ' 'c lar wor w d be v h t th res f career wi h hi r anization. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question I wppld ahjdy disgpgsing thj§ organizatipn with xdplg putside it. Q Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 1 wpuld malty fgel as if thi§ drganizatidn’§ prohlgms at; my own. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question -Ilt J: W. I. 0" 1'1' .63 i] .A. .m '. al.: -A. . in.J-r. a l i '1‘- 2112. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 205 0 his 51. 52. 53. 54. 55. O I choose not to respond to this question I w not C l 'k ’ art f il ‘ at this or ni tion. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question w o f l ’em tionall ttached’ t is or anization. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question i r ization w uld have a reat eal f rsonal meanin for me. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question W s r n n f belon in is or anization. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question V EN 'k I'U CNS: n rmsf '-‘nr_u 0-_s at ca ‘mn o h i ' o .l'wor. .1-' 1"I'0V") .‘r,s't'x'°llll‘-lt .1'1 '11 nts rvis r a izatio wh 1e. W V st f w r or. O Strongly Disagree 206 HIIWI ‘ ' on 00 -ms, 56. 57. 58. 59. 60. O Disagree O Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question w ' ' ' w r n his rvis r wi le ' ther rtments. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question w r l as if this su ervisor’s r blems are m own. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question hi ha woul be able t easil bec me as at he an th r u rvi r’s wor o as I We. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question I would not feel like ’part of the family’ in this supgrvisor’s work gtoup. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 207 61. 62. 63. 0 Agree 0 Strongly Agree 0 I choose not to respond to this question W 'w' i N woulhva telf O Strongly Disagree rsonal meanin for me. O Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question tron sense 1 n in t his u rvi or’s w rk roup O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question ANI N A IN N : w lw' ' 1 tin base on wha 0 know about this drggizatidn, Wl rde ' 'Ia w 'raiin. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question f nw Iha iwor wl " w sn sile. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 208 65. It nphmy saw ma, 1 would m likely to take unauthorized breaks during the day. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 66. I wdpld hp diligppt ahput ndt usipg tha prganization’g phona§ fpr parsonal calls. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 67. wo ili e bout not sin he or anization’ internet access for rsonal business. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 68. depld ndt ppmplaip about wdrking for this cpmpany. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question . 69. .w tWokr in You.' one: Ol'Vl-l't, ~ 0 oar :.c, rm wul not W- O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 0 n _ _ ‘ . _ . 70 Wo_ooo win u t at ,._|Wu._1'nu' t r 10.1.! Ltudne mutt. 209 71. 72. 73. 74. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question RV S RE N N : n wer the f 11 win thetical uestions based what on know b ur s rvisor. if my sampyisdr gpt sick and missed a day of work, I would hy to hglp him catch up whan he £91m O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question If my summisdt had a haavy wdrkldad, i would voluntgr tp help do same at his wprk. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question Kn]! sgmfllsdi; nagggd my adviga, i wdhld E willing td stpp my E91215 and listgn td his mm 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question w w l O Strongly Disagree 0 Disagree 210 0 Neither Agree nor Disagree 0 Agree 0 Su'ongly Agree 0 I choose not to respond to this question 75. w c ' w r ff' 5 rvi rnee ed he help. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 76. u rvi r seemed to be stres e I woul of m w t ncoura him. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question mu 3 ° er rfhixriencnc a'nto mostrecent parking tigket that ypu have received on campus. it ypu hava havgr mperved a parking ticket on m ase w r"I h en tor 5 nd i " h uestron. DPPS PROCEDURES -1 -l ,0 2 o '21 ‘1'31'1 010W ol'gq' 0A0 ‘ ,‘ ‘mnr -D‘aIS 'C [The 11!“! 00 f 1 b t e art ento P01' 1' . O Strongly Disagree ODisagree 0 Neither Agree nor Disagree 211 78. 79. 80. 81. 0 Agree 0 Strongly Agree 0 I choose not to respond to this question e ' e i rkin icket. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question The Dgpagtment df Police and Public Safety (DPPS) is to blame for the parking ticket. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question W ' l efor arkin 'k ? O DPPS is entirely to blame O DPPS is somewhat to blame 0 Neither is to blame O The parking employee is somewhat to blame O The parking employee is entirely to blame O I choose not to respond to this question v r w ' ' ' k 7 O Nobody O I am 0 DPPS O The parking employee 0 I choose not to respond to this question 82. Pl' n ° f r 9 ies ev fair. 0 Strongly Disagree 0 Moderately Disagree 212 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 83. WI. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree , O I choose not to respond to this question 84. r e res PP s a eci ion are not fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 85. inform P f wh 'n h w do. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree O Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 86. WW 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 213 88. 89. 90. 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question it l‘o outfits .1' i=9- 191.22%... 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question P ea with i nit and re c . O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree O I choose not to respond to this question P S’ cisi ns ar made out i h 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question from 1' he V n We don’t have any fair parking pdiicigs at MSU. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 214 iw'. W --l ik ke or informed kn w what’s oin n. 91. 92. 93. 94. 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question PS w l ' i ' ns me. O Strongly Disagree O Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question P old einhihrear. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Snongly Agree 0 I choose not to respond to this question DEBS dpasn’t gm ahdut my parking needs. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question ' ‘ ~ ' A. . A .A- '11 ‘13 ' ,'-. h .-. 'lt‘l u‘th t '«‘ m ' -1 Sm . 1.. ' I ' l w wr 215 10W 0U ’ roce r n ' ines v fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 95. Iha kaihg empldyeg made it glear to me that I am a valued Sthdent. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question. 96. roce res the arkin em 1 cc t ak ' i ns are not fair. 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 97. ki m l inf rrned w 0t iCkCt. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 98. W5 216 99. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question i was infprmgd hy tha patting amplpyga abdat why 1 gm my tigket. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 100. The parking amployee treated me with dignity and resppct. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 101. lb; patting amplpyae’s ratidnala f9: mahihg thg dgpisiph was ndt kept segrat. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 102. W O Strongly Disagree 217 O Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 103. WM 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 104. Ifggi that tha parking employee holds me in high regard. O Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 105. WS- 0 Strongly Disagree 0 Moderately Disagree 0 Slightly Disagree 0 Neither Agree nor Disagree 0 Slightly Agree 0 Moderately Agree 0 Strongly Agree 0 I choose not to respond to this question 218 idislilsa this pathihg amployee. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 107. Wt O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 108. We 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 109. Eatlg'ng pmplpyaps ahjpy making sthdapts’ livas misstahle. 110. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 1w! '1- .1"1010 'v.‘ .r’: 0;! 018'”! .1‘03' W'.'t" 21' " “1L 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 111. WW 0 Strongly Disagree 219 O Disagree O Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 112. Willem O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 113. ' i s n ‘ 'vesmi ale. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 114. M O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 115. I like D135. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 116. WW- 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 220 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 1 17. W5. 0 Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 118, w n v tu nt’ tuition bill o towar i r vin DPPS. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 119. I make eveg effort to park according to the regulations established by DPPS. O Strongly Disagree 0 Disagree 0 Neither Agree nor Disagree 0 Agree 0 Strongly Agree 0 I choose not to respond to this question 120. What is your age? 0 18 O 19 O 20 O 21 O 22 O 23 O 24 or older 0 I choose not to respond to this question 121. Wand? O African-American 221 122 123. 124. 125. 126. 127. 128 O Asian-American O Hispanic-American 0 Native American 0 White / Caucasian O Other 0 I choose not to respond to this question W? 0 Female 0 Male 0 I choose not to respond to this question I i ' b? 0 Full Time 0 Part Time 0 I currently do not have a job 0 I choose not to respond to this question Wi ' r O in the next 3 months 0 in the next 6 months 0 in the next 9 months 0 I do not plan to look for a job within the next 9 months 0 I choose not to respond to this question ’vhadhv uvr r otin? 0 Yes O No 0 I choose not to respond to this question v r m i isi v v i n? 0 Yes O No 0 Not applicable 0 I choose not to reSpond to this question WW? 0 Yes O No 0 Not Applicable 0 I choose not to respond to this question WW? 0 Yes 222 ONO O I choose not to respond to this question 129. Whammmwd? 0 Yes O No 0 Not Applicable 0 I choose not to respond to this question 130. WW? 0 Yes 0 No Q I have never appealed a parking ticket 0 I choose not to respond to this question 131. v v r a l d arki icket and lost? 0 Yes O No Q I have never appealed a parking ticket 0 I choose not to respond to this question 132. Do you koow aoyone who works for the Department of Eolioo and Boolic Safety on campus? 0 Yes O No 0 I am an employee of DPPS O I choose not to respond to this question R l iin 133. . l 1"“ H "'1 H tg/IWH: ' 10 0‘ 1‘ I'LL! "'1! h [-.iOft - 02 grra. :.l' ' 10"U'V ' i ' a1 'u- 0 v; 02‘ 1.010 0mg. (1' iwld 223 d§§QflE mg CXEQDI to which the omployee thought a cortain procedure (that was meant to bring 329!!! g eenain nggmg) was fair. An omployee’s porcoption of distrioutive jostico is the oxtent wi l h e tmew fair. jthis [osemh hg oonsiderable implioatioos in the workplace. It has beon shown that r e tions f fairness influen e an individual’s motivation and satisfaction. The resent study was dosignoo to oxploro tho varioos "sources" of fairness that can be blamed for unfairness in a l l ' i . me sub'ect ot escri ti n f situation that wer set so tha u l r rvisor. t were ch n so hat i ht blam th u r n m m it nd e others wer esi ned so hat on would not blame ei herf r he ootoomo. It was our goal to see whether these differonoos led you to roast difforently to the na er 'tee. u f his research mi hl O holo is and human re 0 a er bet r ea wi work lace roblem . v esti n mmentsa o tthis t d feel fre to all us: ' h l rva 4 - n M ' n .D. 3-8 5 O I have read the debriefing 134. Th al tw estions aret hel me or anize the data What is tflay’s date? 3-...-. .— __ . - .,... - 7‘ , ' . -v'-r «-—- ~m. wn-u....-m~. . s -—‘ .. “ 'e ‘ -‘ ' ‘ — 135. i ro imat ime? 224 APPENDIX D Measures for the Second Part of the Study 225 I am to blame for the parking ticket The parking employee is to blame for the parking ticket. The Department of Police and Public Safety (DPPS) is to blame for the parking ticket. I. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree Who is to blame for the parking ticket? 1. DPPS is entirely to blame 2. DPPS is somewhat to blame 3. Neither is to blame 4. The parking employee is somewhat to blame 5. The parking employee is entirely to blame Overall, who is to blame for the parking ticket? 1. Nobody 2. I am 3. DPPS 4. The parking employee DPPS procedural justice (adapted from Byr_ne & Cropanzano, 2000) DPPS’ procedures and guidelines are very fair. The procedures DPPS uses to make decisions are not fair. (Reverse-scored) I can count on DPPS to have fair policies. We don't have any fair parking policies at MSU. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree 226 DPPS informational justice (adapted from Byr_ne & Cropanzano, 2000) I am kept informed, by DPPS, of why things happen the way they do. Whether the outcome of a parking situation is good or bad, I always feel like I am kept informed by DPPS. DPPS’ decisions are made out in the open so that everyone always knows what's going on. Whether right or wrong, DPPS always explains decisions to me. 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree DPPS inte_rpersonal justice (adapted from Bme & Cropanzano, 2000) DPPS makes it clear to me that I am a valued student. DPPS treats me with dignity and respect. I feel DPPS holds me in high regard. DPPS doesn’t care about my parking needs. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Parking employee procedural justice (adapted from Bme & Cropanzano, 2000) O The parking employee’s procedures and guidelines are very fair. . The procedures the parking employee used to make decrsrons are not fair. (Reverse-scored) . . I can count on that parking employee to have fair policies. This parking employee doesn’t have any fair policies. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Parking employee informational justice (adapted from Byme & Cropanzano. 2000) The parking attendant informed me why I got my ticket. I was informed by the parking employee about why I got my ticket. The parking employee’s rationale for making the decision was not kept secret. The parking employee explained the decision to me. 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Parking employee inte_rpersonal justice (adapted from Bme & Cropanzano, 2000) The parking employee made it clear to me that I am a valued student. The parking employee treated me with dignity and respect. I feel that the parking employee holds me in high regard. The parking employee doesn’t care about the parking situation on campus. (Reverse-scored) 1. Strongly Disagree 2. Moderately Disagree 3. Slightly Disagree 4. Neither Agree nor Disagree 5. Slightly Agree 6. Moderately Agree 7. Strongly Agree Parking employee reactions I dislike this parking employee. I like the parking employee who issued me the ticket. This parking employee was on a power trip. Parking employees enjoy making students’ lives miserable. If I saw a parking employee stranded on the side of the road, I would stop and help out. I would consider becoming a parking employee. I would be friends with a parking employee. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 228 DPPS reactions DPPS enjoys making students’ lives miserable. I dislike DPPS. I like DPPS. DPPS is only out to make money. If I saw someone parking illegally on campus, I would report it to DPPS. I would support a $3 fee placed on every student’s tuition bill to go toward improving DPPS. o I make every effort to park according to the regulations established by DPPS. 1. Strongly Disagree 2. Disagree 3. Neither Agree nor Disagree 4. Agree 5. Strongly Agree 229 MlL NH, 1M“ ATf L|i5HARIFS llHlllHlllllll 1293 02177 428