~I‘ .V .1.“ .4“. "2533}: . ‘_ , . , ‘ , . 7 H, J;‘~3?l.§J1:{w~.x . ~ » ‘ . . . . . ,‘ . . Y,“ ‘ 41:39.1 \,1 “3,3475; . ' " ~ ,- . 4 i ‘ '1 3 ' . 4 ~. - ~ . , y, I: WM) . 2423‘s.: 4‘ . ' 7' '1‘ ‘ «hm ‘ 114:1" ‘ 32:? “'-";‘;'.. my E35}: ~~ ".‘V' “c n u 1-» xk. “ilk-7C1. ‘ 511.74%.“ ‘3‘. . x ”u. .Iug‘jjm; ' x , h t 6 ‘Jv .11 .u‘ . ‘..v - "$3.20“ « w, n - 1:15. 9 ' A .1 1,‘ u !_ v". . "4" ~' I ~11 I ,l. "nun-mm} “ . " - .a...\,v,, w . . - - My. 1 4, 4‘: ,qu ,4, MM“, .. 2:“! 1—, ., ,. 4 ,| (..a.»1:.,4.1 4‘5. 7‘ .;:: z": ‘ i? 5",, ,. '. ,4. ¢ I {5.141 1'. a I w . \ N’F‘I‘ " *x‘ , ’) . e g,“ .3: 3::— . ,I" A A, I 1 d 'r -, 7w! :3] ‘ . ‘ ’ . ' " ' .. " ‘ - ‘2". .f " ', . . -“"'.' ' ~ ' ' ' .Jx _ - ,, _ x A3, . . . J :W‘ ,(fi.\‘.".‘3; ' 4 ~ . ”W ,3,“ ")5“ , v ' " may.” - .2 w A. ".3 ‘n >: v \ , A m" u ., .' ‘13. um MW 4 “mum"; NM.) v ‘.’ W ‘q: ' wté‘lj’x ,. ~ 91,! V - f-II' 1 i $ u I'V . rim m; #145 'R‘J‘ )1 «rs; ' Mr" . 75; p“ #méfingwffiiuirq , :61" . gr, 5:41 ; > A, , p ah“ . r ,.r w -‘}"fi«'»ul-‘J;§w 1 WW: :3. “WC! ’5; "'5'" - 7:”3‘34 "NW”‘O‘W‘ . WV! 111.” , 1' . 1:0 ‘ _ I J 'v $355,: fr fitsamf' x3. WW. '5,” . l w ‘ « r-‘n‘ .. . ‘ n’m -' z ‘ ' A .. . ” “19w" , WK; .J. , “”4““ , h ” via" an! 4; 'r n My} «in y p In ‘ A l" - . _ ‘ , . An. '1" .‘l ‘m.t,.,f{7y.r. _ w . , v...» Mr ‘ » . N . , J _ - MW (.y . 1,. u .w’u: ‘ Q' ‘ - ‘ fl 2; . § ' ' K A: ‘ 4‘ Kim . ”l. m: {r .x' 4.4.3:: .215: ' .u‘mkm " k“Pl-{i‘fla‘umm‘f‘I-Iv‘ ""3 4 - ' Q ' V .. 4: s. ';“ N],- "'3‘“ . 4 . ‘ y» 4. 4 : . _ » . t 3 ‘ f ' ‘ r_ ‘ . AwTwCr-«M "MN #155» n, u." " '1fi‘fi‘93t5m .‘ ' ' 1* KM; W . «(Ehrw'sr'm , ‘ _ 11? Ragnar " . $21-4" ‘LIL'Pm-lusfimvi .. ”.141 W‘Luvq mam Ap‘u y._.‘ m; «mt [2H ?’$b§ a ||||llHlllllHlllHlWIHIHIIIHNIIH HWIIIHHHI LIBRAE ! i I Michigan State- 3 129390572 4525,. a, ,fim -- University This is to certify that the dissertation entitled An Attribution—Emotion—Motivation Model of Reactions to Supervisory Feedback presented by Young Yong Kim has been accepted towards fulfillment- of the requirements for .Ph.D. Communication degree in Major professor ‘ Date July 13, 1989 MS U is an A ffirmative A ction/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ,2; m 0 3 23: WT n.689195 55> J : 9 be ' ' 5; L1 9 a 0 01 wfléflflfifis MSU Is An Affirmative Action/Equal Opportunity Institution AN ATTRIBUTION-EMOTION-MOTIVATION MODEL OF REACTIONS TO SUPERVISORY FEEDBACK BY Young Yong Kim A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1989 w“; (1“. {I a.” @oAO ABSTRACT AN ATTRIBUTION-EMOTION-MOTIVATION MODEL OF REACTIONS TO SUPERVISORY FEEDBACK BY Young Yong Kim The effect of supervisory feedback has traditionally been investigated in terms of simple stimulus-response models. These models focus on how workers respond to positive or negative supervisory feedback messages. This dissertation, however, considered supervisory attributions and their effects on emotions in accounting for the effects of supervisory feedback on a poor performer's motivation. The structural equation model developed for this study included lack of ability and effort attributions, shame and helplessness, and performance improving motivation. A revised version of the model provided an improved fit to the data accounting for the variance in the data. Shame, which rwas determined by attributions, played a critical role in predicting performance improving motivation. No significant role of helplessness was found. The model is discussed in terms of its support for the proposed theories, its individual significant links, and the theoretical and practical implications for organizational studies in the future. ACKNOWLEDGEMENTS I would like to first thank my advisor, Dr. Kathy I. Miller, for providing me with invaluable feedback and inspiration. Her knowledge and skill as a researcher was remarkable. Perhaps even more important was her unobtrusive leadership encouraging my autonomy. I feel most fortunate to have had the opportunity to work with her. Dr. Michael L. Moore, Dr. Ronald C. Tamborini, and Dr. James B. Stiff, in serving on my committee, also provided me with invaluable encouragement and support. I would like to thank Dr. Moore for his critical intellectual insights about organizations; Dr. Tamborini, for his important issues and questions; and Dr. Stiff, for his profound suggestions and advices. I also want to express gratitude to Seong H. Chi, M.D. and Mrs. Ok-Jin Kim for helping me collect data from hospitals. Finally, I would like to thank my wife Wha-Joo. How much she disliked much of my "old" graduate student life, I will never really know. However, I know she has been very patient in living with an obsessional, preoccupied, and often unavailable husband. Thanks to all of you! ii TABLE OF CONTENTS Chapter 1. THEORY AND LITERATURE REVIEW . . . Traditional Models . . . . . . . . Critique of Traditional Models and New Directions . Question about the Sign of Supervisory Feedback Question about No Intermediate Psychological Process . . . . . . Question about Role-Playing . . . Proposed Model . . . . . . . . . . Supervisory Feedback as a Message Variable . Attributions and Supervisory Feedback . . . . Perceptions of Attr1butions as Causal Factors Debate in Attribution-Emotion Relationship . Emotions as Causal Effects . . . . Shame and Its Relationship with Attributions Debate in Attribution-shame Relationship . Helplessness and Its Relationship with Attributions . . . . . Performance Improving Motivation Shame and performance Improving Motivation . Helplessness and Performance Improving Motivation . . . . . . . . . . . . Research Hypotheses . . . . . . . . METHODS O O O O 0 O O O O O O O O 0 Sample and Procedure . . . . . . . Research Design and Procedure . . . Manipulation of Disciplinary Actions . Manipulation of Managerial Actions . Operationalization . . . . . . . . Perceptions of Lack of Ability . . Perceptions of Lack of Effort . . shame O O O 0 O O O O O O O O O O Helplessness . . . . . . Performance Improving Motivation . iii Chapter Analysis . . . . . . . . . . . . . Measurement Model . . . . . . . . Structural Equation Model . . . . RESULTS . . . . . . . . . . . . . Pretest . . . . . . . . . . . . . Confirmatory Factor Analysis . . . Internal Consistency . . . . . . Parallelism . . . . . . . . . . . Path Analysis . . . . . . . . . . other Analyses . . . . . . . . . . DISCUSSION . . . . . . . . . . . . Measurement Models . . . . . . . . Path Madels O O O O O O O O O O O Hypothesized and Revised Models . Relationships among Variables in the Revised Model . . . . . . other Analyses . . . . . . . . . Implications for Managerial Pract Limitations and Directions for Future Research Conclusion . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . APPENDICES Appendix A: Pretest Questionnaire Appendix B: Regular Questionnaire Appendix C: Tables . . . . . . . iv O i ce Page 51 51 58 61 . 63 71 73 73 74 81 83 86 102 106 116 Table 1. 3. 4. 10. 11. 7A. 7B. 12. 8A. 8B. 13. 14. LIST OF TABLES Page Descriptive Statistics for REALISM . . . . . . . Descriptive Statistics for LACK OF ABILITY . . . Descriptive Statistics for LACK OF EFFORT . . . . Items and factor loadings for PERCEPTION OF LACK OF ABILITY (PLA) . . . . . . . . . . . . . . Items and factor loadings for PERCEPTION OF LACK OF EFFORT (PLE) . . . . . . . . . . . . . . Items and factor loadings for HELPLESSNBSS (KL) 0 O O O O O O O O O O O O O O 0 Deviations of observed from expected correlations (internal consistency) for PLA . . . . . . . . . Deviations of observed from expected correlations (internal consistency) for PLE . . . . . . . . . Deviations of observed from expected correlations (internal consistency) for final version of BL . . . . . . . . . . . . . Items and factor loadings for SHAME (SH) . . . . Deviations of observed from expected correlations (internal consistency) for SHAME (SH) . . . . . . Deviations of observed from expected correlations (internal consistency) for final version of SHAME . . . . . . . . . . . . . Items and factor loadings for PERFORMANCE IMPROVING MOTIVATION (PIM) . . . . . . . . . . . Deviations of observed from expected correlations (internal consistency) for PIM . . . . . . . . . Deviations of observed from expected correlations (internal consistency) for PIM . . . . . . . . . Parallelism deviation for PLA . . . . . . . . . . 117 118 119 120 121 122 123 124 125 126 128 129 130 132 133 134 Table Page 15. Parallelism deviation for PLE . . . . . . . . . . 135 16. Parallelism deviation for SHAME . . . . . . . . . 136 17. Parallelism deviation for HL . . . . . . . . . . 137 18. Parallelism deviation for PIM . . . . . . . . . . 138 19. Correlation matrix among theoretical variables . . . . . . . . . . . . . . 139 vi LIST OF FIGURES Figure Page 1. Hypothesized model . . . . . . . . . . . . . . . . . 36 2. Hypothesized model with path coefficients and standard errors . . . . . . . . . . . . . . . . 65 3. Revised model with path coefficients and standard errors .-. . . . . . . . . . . . . . . . . 67 4. Revised perceptual model with path coefficients and standard errors . . . . . . . . . . 69 vii CHAPTER ONE THEORY AND LITERATURE REVIEW In a survey of Fortune 500 companies, 97% of the supervisors reported experiencing the problem of poorly performing subordinates (Stoeberl & Schniederjans, 1981). No organization can function well unless supervisors manage these poor performing subordinates. In this sense, supervisory feedback has been long viewed as an essential vehicle to manage these poor performers in the workplace. Therefore, this dissertation will examine and integrate the relevant literature and demonstrate how supervisory feedback affects a task performer's subsequent motivation in workplaces. . Numerous scholars considering leadership in workplaces have viewed supervisory feedback as central to motivation (campbell & Pritchard, 1976; Green, Fairhurst, 8 Liden, 1981; Liden, 1981; O'Reilly & Weitz, 1980: Payne & Hauty, 1955; Vroom, 1964). A great deal of theoretical and empirical attention has been also given to understanding three relationships between (1) supervisory feedback and attribution, (2) attribution and emotion, and (3) emotion and motivation, though these three areas have rarely been linked conceptually. More specifically, attribution theorists and researchers have suggested that supervisory feedback is strongly related to causal attributions for 2 performance success and failure (Green 8 Liden, 1980; Green 8 Mitchell, 1979; Kipnis, 1976; McFillen, 1978; McFillen 8 New, 1979; Mitchell, Green, 8 Wood, 1981; Weiner et a1., 1972; Weiner 8 Kukla, 1970). That is, attribution is a crucial determinant of supervisory feedback messages. Attribution theorists and researchers have further found that perceptions of causal attributions are determinants of certain emotions (Martinko 8 Gardner, 1987; McFarland 8 Ross, 1982; Weiner, Russell, 8 Lerman, 1978; 1979). Finally, the literature on emotion indicates that emotion influences motivation (Brenenstuhl, 1976; Gibb, Smith, 8 Roberts, 1955; Horowitz, 1976; Roth 8 Bootzin, 1974; Roth 8 Kubal, 1975; Wortman 8 Breham, 1975). The research reported here will trace the sequential cause and effect processes of supervisory feedback, the - perceptions of attributions, emotions and motivation. Supervisory feedback messages will be manipulated in terms of two attributions of poor performance: lack of ability and lack of effort. The perceptions of these two attributions and two emotions (i.e., shame and helplessness) are important moderator variables in the cause-effect processes. Therefore, this paper will focus on the role of attributions and emotions in explaining the effects of supervisory feedback on motivation. This chapter will begin by reviewing the literature on the relationship between supervisory feedback and motivation. Following the literature review, this chapter will consider some shortcomings of past research. Third, a new model will be proposed. That is, the variables of the 3 proposed model will be defined and their relationships will be specified in this chapter. Traditional Models There has been a considerable amount of research on the relationship between supervisory feedback and motivation (see Annett, 1969; Cusella, 1980; 1984; 1987; Deci and Ryan, 1980; Downs, Johnson, Barge, 1984; Ilgen, Fisher, 8 Taylor, 1979; Nadler, 1979 for reviews). These studies account for the supervisory feedback-motivation relationship on the basis of a simple stimulus-response model. This simple input-output model assumes that the feedback sent is the feedback received. This model concentrates on the sign or valence of feedback as the most important determinant of motivation. The sign or valence of feedback is distinguished into one of two categories, positive and negative, and may be the most important feedback characteristic influencing recipientsI subsequent responses (Ilgen et al., 1979). Cusella (1987) states: Conce tually, feedback valence refers to the perce1ved attractiveness or value of the information conveyed in the feedback message. Operationally, positive feedback messages consist of messages that connote acceptance of the behavior or comparative statements that indicate satisfactory or higher performance, while negative feedback consists of messages that refer unfavorably to the recipient's behavior or comparative statements that indicate unsatisfactory performance (p. 632). In addition, Cusella (1984) argues that there is a point midway on the sign continuum where feedback may be characterized as neither positive nor negative but neutral. 4 Typically, the terms verbal rewards, praise, and positive feedback are used interchangeably while the terms verbal warning, punishment, criticism, and negative feedback are also used interchangeably. Not surprisingly, many studies have explored the effects of positive or negative feedback on motivation. These research findings can be divided into three groups. First, many researchers suggest that positive supervisory feedback enhances motivation. Deci and Ryan (1980) discuss more than a dozen experiments variously utilizing children, college students, and adults that have investigated the effects of positive feedback on intrinsic motivation. Nearly all the studies contend that positive feedback always increases intrinsic motivation and that negative feedback decreases intrinsic motivation if it signifies that the receiver is incompetent. In his laboratory experiment using college students, Cusella (1984) found that positive feedback increases significantly the level of intrinsic motivation as opposed to neutral feedback messages. In their laboratory experiment with students, DeNisi, Randolph, and Blencoe (1982) reported that negative feedback produces a negative change in motivation. A second group of researchers suggest that negative supervisory feedback increases motivation. For example, a survey of one hundred firms showed that 44 percent used negative feedback to correct poor performance and considered the negative feedback messages effective (Miner and Brewer, 1976). McDermott and Newhams (1971) also found that performance was significantly improved after punitive 5 feedback. Franke and Karl (1971), in a statistical analysis of the original Hawthorne experiments, indicate that negative feedback seems to be the major factor in increased motivation and performance. In their review of the research, Arvey and Ivancevich (1980) report that negative feedback can be more effective than positive feedback in work places. Finally, it has been suggested that a simultaneous use of positive and negative feedback messages increases motivation. For example, DeNisi, Randolf, and Blencoe (1982) suggest that actual performance can be best facilitated by a mixture of positive and negative feedback messages. As shown in the discussion above, the results of research are not conclusive. That is, it appears that both positive and negative feedback can enhance motivation. This evidence is enough to say that past research does not explain much about the relationship between supervisory feedback and motivation, and that there are some shortcomings or limitations contributing to this state of affairs. This study will examine three shortcomings in detail, and develop a new model that addresses these limitations. Specifically, it will be shown that: (1) the sign or valence may not be a substantial characteristic of supervisory feedback in managing poor performance in organizations; (2) past research is not concerned about the intermediate psychological processes as a theoretical framework relating supervisory feedback to motivation; and (3) in past research, the use of role-playing methodology 6 has led to a serious question about external validity. This paper then will propose a new model to overcome these shortcomings. Critique of Traditional Models and New Directions Question about the Sign of Supervisory Feedback In traditional models, the sign or valence (i.e., positive or negative) of the message is the substance of feedback. In other words, past studies adopt a simple cybernetic view of feedback. According to this cybernetic perspective, organizational supervisors simply generate positive or negative feedback that are error signals designed to put the subordinates back on the right course. However, this cybernetic perspective has some limitations in explaining supervisory feedback. First, distinguishing feedback into one of two categories (i.e., positive and negative) is not realistic and practical in workplaces, though feedback can be divided into one of these two categories. In other words, actual supervisory feedback messages tend to move beyond a simple cybernetic view of feedback. Kim and Miller (in press) analyzed data obtained by a free-response format and conclude: Managers in this stud did not generate feedback messages which were s1mple "error signals" designed to put the subordinate back on the right course. Rather, the messages generated by managers contained substantial informat1on value beyond this corrective function. Managers tended to counsel and instruct rather than reward and punish. Thus, this research suggests that our view of feedback should be as much one of providing informational support to subordinates as it is of taking corrective action (p. 19). 7 Traditional studies may be useful to identify how feedback recipients respond to feedback sign, though the results were, as noted earlier, far from conclusive. However, these traditional studies limit our understanding of other potential feedback properties that may have a greater informational value than the sign. Specifically, supervisors rarely provide a simply positive or negative feedback when subordinates perform poorly their task (Beer, 1981). For example, if a supervisor says "seek help from me if you have a similar case in the future," in a poor performance situation, it is hard to say whether this message is positive or negative. Furthermore, although the message may be categorized into neutral, the most substantial informational value in this feedback seems to be related to the supervisor's intention to instruct or emphasize the value of more learning. That is, supervisors tend to analyze the factors contributing to subordinate's poor performance before choosing a means of influence to change that behavior (Banks, 1976; McGregor, 1960; Etzioni, 1968; Kipnis, 1972; Miner, 1963). Following this analysis, supervisors will concentrate on trying to eliminate causal factors of poor performance rather than simply focusing on the performance outcome (Knowlton 8 Mitchell, 1980). In this regard, considerable attention has been devoted to understanding the role of attributions as determinants of supervisory feedback. For example, Mitchell and his colleagues (Green 8 Mitchell, 1979; Mitchell, Green, 8 Wood, 1981; Mitchell 8 Wood, 1980) argue that supervisors generate a particular feedback message in response to the 8 attributions of poor performance. That is, attributions are conveyed as substantial information characterizing supervisory feedback in workplaces. From the standpoint of the subordinate, the attributional property of supervisory feedback is critical in the area of organizational performance appraisal. Subordinates are vulnerable to how supervisors account for their failure because organizational disciplinary or managerial actions vary according to the nature of attributions in the performance appraisal process regardless of the positivity or negativity of feedback (Odiorne, 1971; Wheeler, 1976). In a practical sense, therefore, the attributions supervisors make for poor performance and the perceptions of these attributions by subordinates can be a critical dimension of supervisory feedback. These attributional aspects of supervisory feedback will be discussed in detail in a later section of this chapter. Question about No Intermediate Psychological Process The ultimate determinant of communication effectiveness depends on how a recipient perceives messages. A second shortcoming of past research is that feedback recipients have received little treatment in past feedback studies (Ilgen et al., 1979). For the most part, the traditional stimulus-response models has treated feedback recipients as passive objects in the social environment--that is, as objects which react to verbal stimuli. These input-output models neglect intermediate psychological processes preceding the motivational outcomes of feedback recipients. 9 This traditional perspective has been criticized by information processing theorists and researchers. According to an information processing approach to human performance (e.g., Hammond 8 Summers, 1972; Kantowitz, 1974), individuals are active information seekers adapting to their social environment. Theorists and researchers working from this perspective suggest that feedback recipients can actively attend to the responses of supervisors concerning the failure of their performance, and furthermore they deliberately attempt to appraise the information received in order to determine their attitude or behavior (Ilgen 8 Feldman, 1983; Lazarus 8 Folkman, 1984; White, 1959). In research relating the nature of feedback to motivation, however, the intermediate psychological processes of feedback recipients have been rarely studied, or even acknowledged. Ilgen, Fisher and Taylor (1979) point out: ...a well developed set of theoretical statements has been rarely presented to relate specific characteristics of the feedback stimulus to psychological processes preceding the behavioral response in organizational settings. The result is a large body of experimental research relating one or two dimensions of feedback to a given response (or set of responses) with little concern for the intermediate psychological processes triggered by the feedback (p. 149). Ilgen et a1. (1979) call for a theoretical framework focusing on feedback recipient's psychological processes. When researchers consider these psychological processes, it is possible to understand the way in which feedback recipients process feedback information. Therefore, this study will develop a theoretical framework as to how 10 supervisory feedback may influence subordinates' motivation in workplaces. Question about Role-PlayingiMethod Finally, with but rare exception (e.g., Knowlton 8 Mitchell, 1980), past research obtained data using role- playing methodology. For example, college students were asked to pretend to be organizational workers. These role- playing methods lead to a serious external validity question concerning the degree to which the students' role-playing performance is representative of organizational workers' behavior (Bellack, Hersen, 8 Turner, 1978). Role-playing methods use contrived situations to prompt the occurrence of the target behavior. However, the contrived situations may fail to reflect organizational reality; for example, a simple puzzle-solving task is probably not representative of real organizational tasks. Therefore, this study will obtain the data from real employees presently working in organizations, and the realism of the contrived situation used for this study will be pretested. Proposed Model According to Ilgen et a1. (1979), a feedback recipient's processing of feedback includes at least three components: (1) the feedback recipient's perception of feedback (the causal factor), (2) internal responses to the perceived feedback (intermediate psychological effects), and (3) behavioral intention (consequence of the psychological responses). In this model, the effects of feedback will 11 depend upon how recipients perceive and interpret information conveyed in feedback messages. Once recipients perceive and are aware of some aspects of feedback, they respond internally. This internal response includes both cognitive and affective components. Behavioral intention is viewed as a function of these cognitive and affective effects. Based on this framework, basic elements of a proposed model will be identified and defined, and the sequential cause and effect processes of these elements will be also traced. The sequential cause and effect processes will include: (1) supervisory feedback as an antecedent, (2) the perceptions of attributions that supervisors make for poor performance (i.e., portrayed in supervisory feedback messages) as a causal factor, (3) shame and helplessness as important moderator variables, and (4) performance improving motivation as an outcome of shame and helplessness. To do these, this section will begin by explicating the nature of supervisory feedback. Second, the literature on attributions in supervisory feedback will be reviewed. Special attention will be given to ability and effort attributions. Third, the relationship between attribution and emotion will be discussed. A current debate between cognition-emotion model and emotion-cognition model will be compared. Fourth, the literature on shame and helplessness will be reviewed in relation to attributions. Finally, this paper will review the literature that links shame and helplessness with performance improving motivation. 12 Supervisory Feedback As a Message Variable Cybernetic theorist Wiener (1948; 1954) is generally credited with introducing the concept of feedback into general usage. Wiener (1954) defined feedback as "a method of controlling a system" (p. 16). He states, "...control of a machine on the basis of its actual performance rather than its expected performance is known as feedback" (1954, p. 35). Thus, feedback is information about the gap between the actual level and the reference level of a system parameter which is used to alter the gap in some way. For example, information on overspending on travel by a salesperson used to cut his/her spending in the future is feedback. The system parameter is the salesperson's travel expenditure. The reference level of the system parameter is the budgeted expenditure; the actual level is the actual travel expenditure. The gap between the reference level and the actual level is the amount of overspending. When an attempt is made to alter the gap (most probably to decrease the gap), this attempt is feedback. In the simple error correction model of the cybernetic view, feedback is seen as a necessary component of the functioning of a system, preventing entropy and enabling adaptation (Katz and Hahn, 1978; Seiler, 1967). However, individuals are not automatically self-correcting, and thus, the process of attending to, interpreting, and acting on feedback is more complicated than these simple error correction models. Because of these limitations, behavioral scientists have not consistently used the term feedback as cybernetic theorists originally defined it (Cusella, 1987: 13 Nadler, 1979). Organizational communication scholars have defined feedback as a message comprising a variety of information. Cusella (1987) argues: Organizational communication scholars may best contribute to an understanding of feedback in organizations by focusing on verbal feedback, either orally delivered messages directed face -to-face or written or mediated messages directed in a non-face-to-face manner, regarding the actions of receivers who have been performing an assigned organizational task (p. 630). Feedback messages may emanate from various sources (see Downs, Johnson, 8 Barge, 1984; Greller 8 Herold, 1975 for reviews), but 95% of feedback through performance appraisal is performed by supervisors (Latham 8 Wexley, 1982). Latham and Wexley state several reasons for this trend: First, the management hierarchy of most organizations reinforces the r1ght of the supervisor to make both evaluative and developmental decisions concerning subordinates. Second, the supervisor generally controls the magnitude and scheduling of the rewards and punishments that can be administered to subordinates. Since performance is enhanced when rewards are based on performance, it is logical that the appraisal be conducted by the person who normally administers the rewards. If this were not the case, it is likely that in many organizations the employee might view the appraisal process as having l1ttle or no importance. Third, it is commonly felt that of all the various sources of evaluation, the immediate supervisor is in the best position to observe a subordinate's behavior, and judge the relevance of that behavior to job objectives and organizational goals (p. 80). For the communication scholar, supervisory feedback is typically viewed as strategic in nature (Bogart, 1980; Katz, 1975). That is, a supervisory feedback message can be described as a tool that supervisors delivers their supervising intentions to subordinates who perform assigned organizational tasks. In this sense, attributions can be 14 conveyed through supervisory feedback, and also attributions can be perceived through supervisory feedback. Attributions and Supervisory Feedback As mentioned earlier, supervisors become aware of an incident of poor performance and attempt to understand and control its causes. That is, supervisors concentrate on trying to eliminate the causal factors of poor performance rather than simply focusing on the evaluation of performance outcome per se (Mitchell 8 Wood, 1980). Historically, attribution theorists and researchers have focused on the attributions supervisors make for subordinate's poor performance. According to Weiner et al. (1972), supervisors can utilize four attributional elements to interpret and to predict the outcome (0) of an achievement-related event. The four attributions are ability (A), effort (E), task difficulty (T), and luck (L): O = f (A, E, T, L). Attributions of ability and effort are similar in that they describe qualities of the person undertaking activities, that is, they are internal attributions. Ability refers to task skill or competence while effort refers to the level of energy expended for job performance. These two concepts are discussed in detail in Chapter two. On the other hand, task difficulty and luck are similar in that they can be considered properties of the environment, that is, they are external attributions. Ability and effort have been considered more important than other attributions in organizational performance appraisal processes. 15 Supervisory feedback through performance appraisals is a legal mandate in organizations (For reviews, see Latham 8 Wexley, 1982; Schuler, 1981). Performance appraisal is defined as a structured system of measuring, evaluating, and influencing an employee's job-related attributes, behaviors, and outcomes. Over the last 50 years, supervisors have been more and more dependent on performance appraisal systems in dealing with poor performers in order to maintain objectivity (Cascio 8 Bernardin, 1981). Locher and Teel (1977) state that 89% of the American organizations surveyed use various performance appraisals. Schuler (1981) state: Most organizations maintain their formal evaluation systems because they feel that performance evaluation is fundamental to organizational effectiveness; that if there were no formal appraisal program there would have to be an informal one; and that it is better to maintain the visibility of a formal program fairness, validity, usefulness, and reliability of the appraisal process (p. 221). Performance appraisals consider ability and effort the most critical attributions that supervisors should control because ability and effort are the most fundamental elements for job accomplishment (Kane, 1986; Porter, Lawler, Hackman, 1975). Many empirical studies have found that these two internal attributions significantly influence the generation of supervisory feedback or leadership behavior (Barrow, 1976; Goodstadt 8 Kipnis, 1970; Goodstadt 8 Hjelle, 1973; Green 8 Liden, 1980; Mitchell 8 Wood, 1980; Rothbart, 1968). Attribution scholars have been concerned with the distinction between ability and effort attributions. Weiner et al. (1972) note that ability is a potential cause of performance that is relatively stable, and one that 16 individuals have little volitional control over, at least in the short run. Effort, on the other hand, is a potential cause of performance that is relatively unstable, and one that individuals have a good deal of volitional control over (i.e., they can choose to try hard or not to try hard). Supervisors establish different supervisory goals for dealing with a lack of ability and a lack of effort. Supervisors tend to see a lack of effort as immoral and punitive because subordinates can have a great deal of volitional control over effort. In this situation, supervisors attempt to stimulate the poor performers using various punishments (Knowlton 8 Mitchell, 1980; Lanzetta 8 Hannah, 1969; Weiner 8 Kukla, 1970). In contrast, supervisors do not see lack of ability as immoral or punitive because subordinates have little volitional control over ability. In this case, supervisors attempt to foster learning by poor performers (Miner 8 Brewer, 1976). Similarly, Latham, Cummings and Mitchell (1981) state that supervisory feedback messages are developed in relation to a pool of disciplinary actions in the lack of effort situation while supervisory feedback messages are generated in relation to a pool of managerial actions in the lack of ability situation. Disciplinary actions refer to the levels of aversive events or the removal of positive events which supervisors present to stimulate poor performer's motivation. These actions are carried out in the manner specified by organizational rules or the union contract. The usual sequence of disciplinary actions is informal warning, formal 17 warning, suspension or loss of pay, and dismissal (O'Reilly 8 Weitz, 1980; Stressin, 1960; Wollenberger, 1963). In contrast, managerial actions do not invoke aversive events or sanctions. Instead, these actions focus on developing poor performer's future potential. That is, managerial actions refer to the levels of directions which supervisors provide to develop subordinate's future potential. Training and coaching are the examples of managerial actions used frequently in workplaces. Disciplinary and managerial actions will be discussed in detail in the method chapter. Perceptions of Attributions as causal factors Since supervisory feedback may indicate a personnel judgment or decision that affects the status of employees regarding retention, termination, promotion, demotion, transfer, salary increase or decrease, or admission into retraining, supervisory feedback has been of great importance to employees in workplaces (Beer, 1981). Research suggests that subordinates attend to supervisory feedback more when they perform poorly than when they perform successfully (Folkes, 1978; Gioia 8 Sims, 1986). In this sense, subordinates in workplaces are viewed as feedback seekers who adjust according to organizational demands (Ashford, 1986; Ashford 8 Cummings, 1983). When poor performance occurs, it is likely that subordinates will concentrate their attention on the {Eternal attributions that supervisors make for poor performance because they know internal attributions can be great threats to the status of employees, while external (E 18 W attributions are related to job redesign or job rotation in performance appraisal contexts (Latham 8 Wexley, 1982; Schuler, 1981). In a poor performance situation, therefore, it is likely that subordinates will attempt to assess the magnitude of two attributions (i.e., lack of ability and lack of effort) from supervisory actions conveyed in supervisory feedback. In workplaces, it is common to relate disciplinary actions to lack of effort problems and to relate managerial actions to lack of ability problems (see Miner 8 Brewer, 1976; Latham, Cummings, 8 Mitchell, 1981 for reviews). The degree of disciplinary or managerial actions used indicates the extent to which supervisors attribute poor performance to lack of ability and lack of effort. For example, the use of a formal warning is a stronger indication of lack of effort than the use of criticism, while the demand of training is a stronger indication of lack of ability than the use of blame. Therefore, it seems reasonable that the perception of lack of ability is determined in response to managerial actions while the perception of lack of effort is determined in response to disciplinary actions. Much attention has been given to investigating the relationship between the perceptions of attributions and emotions. However, there is a long debate concerning whether attribution (as a cognitive process) precedes emotion or emotion precedes attribution. This issue will be discussed and then the literature on the relationship between the perceptions of attributions and emotion will be reviewed. 19 Debate in Attribution-Emotion Relationship Cognitive appraisal theory (Bowers, 1973; Endler 8 Magnusson, 1976; Janis 8 Mann, 1977; Lazarus 8 Averill, 1972; Pervin 8 Lewis, 1978; Ruckmick, 1936; Schachter, 1959; see also many other writers in Krohne 8 Laux, 1982) contend that how individuals construe and perceive a stimulus shapes emotional responses. In this meaning-related cognitive model, emotion is not determined by a stimulus itself, but by the perceptions that an individual takes of the stimulus according to available cognitive information. The cognitive information includes evaluative issues, such as "Am I in trouble or being benefited, now or in the future, and in what ways" and so on. For example, when employees have received some information from supervisory feedback about their job failure, they appraise the extent to which the failure will jeopardize future job or career. The perceptions of threat to job or career will then determine emotional intensity. Mandler (1975) and Schachter (1966) further suggest that a nonspecific and ambiguous stimulus or event first can cause emotions, and then the emotions are followed by cognitive activity. For example, when an aversive event is given without any specific causal information, depression is first produced and then individuals become engaged in why the aversive event happened (Abramson et al., 1978; Miller, Klee, 8 Norman, 1980). However, given specific causal information about the aversive event, individuals immediately assess the nature of causal information, and 20 then the assessment determines emotional responses (Arnold, 1970; Mandler, 1975; Schachter, 1966). Thus, cognitive appraisal theory views cognitive activities as antecedents of emotions. However, this theoretical perspective has been challenged by a different treatment of this relationship which suggests that emotions precede cognitive activities (see Basowitz et al., 1955; Child 8 Waterhouse, 1953; Isen, Shalker, Clark 8 Karp, 1978; Sarason, 1972; Sarason, Mandler, 8 Craighill, 1952; Wright 8 Mischel, 1982 for reviews and analyses). More specifically, this line of research suggests that positive emotions facilitate cognitive activity while negative emotions disturb cognitive activity. As a result, emotion cannot be legitimately disregarded as an antecedent variable in the cognition-emotion relationship. By the same token, cognition cannot be legitimately disregarded as an antecedent variable in the cognition-emotion relationship. About this controversy, Lazarus and Folkman (1984) state: Clearly, the error is to treat the relationship as a one-way street. As we noted earlier, the direction observed depends on one's point of entry into the ongoing process. Thus, if the sequence seems to begin with (1) thought, followed by (2) emotion, followed by again by (3) thought, and if we begin at point (1) in the above sequence, we must discover that cognition antecedents and in a sense of probably determines emotion; however, if we start at point (2) in the sequence, we must discover that emotion determine thought. Both principles, of course, are correct This suggests that the causal direction specified should depend on the portion of the system under investigation. As a result, it is reasonable to say that 21 both perspectives can not be rejected because of their own utility. For the purpose of this study, supervisory feedback will include specific information regarding causal attributions. In line with the cognitive appraisal theory, thus, this study will investigate how the perceptions of attributions affect emotions. This relationship will be discussed in the subsequent section. Emotions As Causal Effects Achievement motivation theory (Atkinson, 1957; 1964) contends that individuals can be motivated as a result of emotional experiences which include the span between hope of success and fear of failure. Building on the work of Atkinson, attribution theorists and researchers have attempted a cognitive reinterpretation of achievement motivation theory. Weiner and his colleagues (Weiner, 1980; Weiner 8 Kukla, 1970; Weiner, Russell, 8 Lerman, 1978; Heckhousen, 1982; Russell, 1982) have considered the diverse achievement-relevant emotions that result from attributions and view emotions systematically from the standpoint of such attributions. They view emotion not as an antecedent of causal attributions but as a consequence of causal attributions for success and failure. In an achievement- related context, failure attributed to internal attributions (either a lack of ability or a lack of effort) is said to intensify two dominant emotions: (1) shame and (2) helplessness (Covington 8 Omelich, 1979; Dweck 8 Goetz, 1978; McClelland, Atkinson, Clark, 8 Lowell, 1953; Weiner et 81., 1970). 22 Shame and Its Relationship with Attributions Traditionally, shame refers to a painful emotion about the status of the self. Shame includes several meanings in conjunction with a variety of human situations. First, Freud and his colleagues (e.g., Freud, 1962; 1963; Nunberg, 1955; 8 Jacobson, 1964) describe shame as an emotion resulting from genital deficiency or body deficiencies. These theorists contend that shame is exclusively associated with genitally focused identity and also propose that morality is the motive behind shame. Freud and his colleagues state that no body part or aspect of self other than the genitals is seen as relevant to the experience of shame. This view has been criticized because of its enormously constricted understanding of shame and of human values in general. Other theorists and researchers (e.g., Alexander, 1938; Ausubel, 1955; Will, 1987) contend that shame refers to an acute feeling of inferiority that may refer to any aspect of self or to the self experienced as a unity rather than a particular body part. For example, Alexander (1938) states: The feeling is not so much not being good in a moral sense but being weak, inefficient, unable to accomplish somethin . In inferiority feelings it is not implied, as it is in guilt feeling, that the self-condemnation is the result of wrongdoing (p. 44). However, these theorists emphasize only developmental failure. That is, they see shame as resulting from the failures to move forward developmentally. In contrast, Piers and Singer (1953), Lynd (1958), and Lewis (1971) have 23 associated shame with a great variety of human life situations. For example, Piers and Singer (1953) contend that shame is related to a collapse of self-esteem resulting from the failure of social roles that individuals or social or cultural systems assume. In this perspective, shame is a temporary feeling varying from situation to situation (Lewis, 1971; Lynd, 1958). Another important aspect of shame is related to the issue as to whether shame requires an audience who knows about the failure. Mead (1937) contends that shame requires an audience. Piers and Singer (1953) argue that shame may occur in the absence of an audience because social or cultural values are internalized in the mind of actors. However, it is speculated that individuals are peculiarly sensitive, intimate, and vulnerable to the failure situations when the failure is exposed to others. Kaufman (1974) argues: ...shame is an intense fear of exposure, of having one's badness seen by others. Such exposure of self is intolerable because of the underlying sense of being irreparably and unspeakably defective wh1ch somehow separates one from the rest of humanity (p. 569). Kaufman (1985) further suggests: The interpersonal bridge is built upon certain expectations which we have come to accept and to depend on...Shame is likely whenever our most basic expectations of a significant other are exposed as wrong (p. 13). Therefore, shame is defined as an emotional response to variety of failure situations in which a collapse of self- esteem occurs. This definition includes several characteristics of shame. First, the failure occurs by 24 attributes of self. Second, shame is a temporary emotional state. Third, shame is stronger when the failure is exposed to a significant person in interpersonal contexts than when exposed to a nonsignificant person. It has been postulated that ability and effort are perceived as the dominant attributions in determining the affective consequences of success and failure. Particularly, shame has been specified as an important emotional reaction to the perceived internal attributions of failure (Atkinson, 1964; McCelland, Atkinson, Clark, 8 Lowell, 1953; Weiner, 1977). Weiner (1977), for example, contended that: Pride and shame, as well as interpersonal evaluation, are absolutely maximized when achievement outcomes are ascribed internally and are minimized when success and failure are attributed to external causes. Thus, success attributed to high ability or hard work produces more pride and external praise than success that is perceived as due to the ease of the task or good luck. In a similar manner, failure perceived as caused by low ability or lack of effort results in greater shame and external punishment than failure that is attributed to the excessive difficulty of the task or bad luck. In sum, locus of causality influences the affective or emotional consequences of achievement outcomes (p. 183). This presumption has been tested in the context of interpersonal evaluation. In this line of research, subjects were asked to pretend that they were teachers and were to provide evaluative feedback to their pupils (see Eswara, 1972; Kaplan 8 Swant, 1973; Rest, Nierenberg, Weiner, 8 Heckhausen, 1973; Weiner 8 Kukla, 1970; Weiner 8 Peter, 1973). The pupils were characterized in terms of effort, ability, and their poor performance on an examination. The data from these investigations 25 demonstrated the following three findings. First, the level of shame was higher in the low effort attribution group than in the high effort attribution group. Second, the level of shame was higher in the low ability attribution group than in the high ability attribution group. Third, the data further demonstrated that the attribution of low effort is of greater importance than the attribution of low ability in determining the intensity of shame. However, there is a debate about this third finding, though attributions of low effort and ability are both potent sources of shame. Debate in Attributions-Shame relationship There have been two possible reasons for the discrepancy between low ability and low effort attributions as determinants of shame. First, effort attributions may elicit strong moral feelings because trying to attain a socially valued goal is perceived as something that one ought to do. Secondly, effort is instrumental to changing behavior, inasmuch as effort is believed to be subject to volitional control. On the other hand, ability is perceived as nonvolitional and relatively stable and thus should be insensitive to external control attempts. Everyday observations also supports the expressed beliefs concerning the relative affective significance of ability versus effort attributions. For example, the handicapped person who struggles to complete a race and the retarded child who persist to finish an objectively easy task elicit great social approval. On the other hand, the 26 gifted athlete who refuses to practice and the bright dropout generate great social disapproval, as though it were immoral not to utilize one's capacities (see Baker, DiMarco, 8 Scott, 1975). Thus, low effort may be more severely punished and attributions to this factor should generate more shame than attributions to low ability because effort rather than ability appears to be the main determinant of interpersonal evaluation. Other scholars, however, oppose this view. For example, Covington and Beery (1976) and Nicholls (1976) posed the following question: If effort is so rewarded and attributions of effort generate so much pride (as opposed to shame), then why do so students hide their efforts and refuse to admit that they studied hard? To solve the dilemma raised by this question, Covington and Beery (1976) suggested that students strive to maintain a self-concept of high ability. Failure attributed to a lack of effort can promote the view that one has ability (Run 8 Weiner, 1973). Weiner, Russell and Lerman (1978) suggest that this conflict of attributional importance between ability and effort may be a function of the difference between actors (evaluatee) and observers (evaluators). That is, Weiner and his colleagues focus on the perception of observers (teachers), while Covington and his colleagues focus on the perception of actors (students). In other words, from the viewpoint of evaluators (supervisors), a lack of effort may be a stronger determinant of shame than a lack of ability because, as discussed earlier, a lack of effort is viewed as immoral and punitive in workplaces. In contrast, from the 27 viewpoint of actors (subordinates), a lack of ability may be a stronger determinant of shame than a lack of effort because individuals are more vulnerable to the lack of a more stable dispositional factors. However, there has been no empirical evidence to support this distinction. Therefore, the relationship between shame and attributions of lack of ability and lack of effort is at this point a questionable issue in need of empirical verification. Helplessness and Its Relationship with Attributions The concept of helplessness has its origins in laboratory work with dogs and other animals that fail to avoid electric shock, even though an avoidance response of jumping to safety is readily available (Overmeier 8 Seligman, 1967; Seligman 8 Maier, 1967). Many investigators have extended this paradigm to research with human subjects to offer the concept of helplessness as an explanation of human emotions (Seligman, 1975). Conceptually, helplessness is viewed as an emotional reaction to aversive events (see Depue 8 Monroe, 1978; Lazarus 8 Folkman, 1987 for reviews). Operationally, helplessness refers to the degree to which aversive events or outcomes cannot be controlled or overcome; that is, the extent to which aversive events or outcomes are noncontingent on a person's responses. Thus, helplessness can be described as an emotional reaction to perceived loss of control over aversive events. As described in the preceding discussion, the internal attributions (lack of ability and lack of effort) that supervisors make for poor performance can be a threatening 28 or aversive event for subordinates. Controllability (Weiner, 1979) or changeability (Anderson et al., 1983) is an important property of these causal attributions (Weiner, 1979). Thus, attribution theorists have employed the concept of helplessness as an emotional reaction to the internal attributions of failure (Abramson et al., 1978). According to Dweck and Goetz (1978), both lack of effort and lack of ability attributions produce an increased perception of helplessness. Helplessness theorists and researchers (Abramson, Seligman, 8 Teasedale, 1978; Anderson, 1983; Dweck, 1975) further contend, in concert with traditional attributional analysis of emotion (cf. Weiner, 1974; 1980; Weiner, Graham, 8 Chandler, 1982; Weiner, Russell, 8 Lerman, 1978; 1979), that when a negative outcome is thought to be a product of a person's lack of ability, the poor performer is more susceptible to helplessness than when a negative outcome is thought to be a product of a person's lack of effort. According to achievement-related attribution theorists, ability is a more stable factor than effort (Weiner et al., 1972), and subordinates can have more control over effort than ability (Weiner, 1979). Covington and Omelich (1979) further investigated the relative importance of ability and effort in conjunction with helplessness. They found that the level of helplessness was higher when failure is attributed to a lack of ability than when failure is attributed to a lack of effort. Thus, this finding leads to the prediction that poor performance attributed to lack of ability will have a stronger effect on helplessness than 29 poor performance attributed to lack of effort, though both of these relationships should be significant. Jones and Nisbett's (1972) actor-observer hypothesis indicates that there are two information sources of internal attributions for poor performance: (1) the subordinate's own perceptions of internal attributions and (2) the supervisor's. The latter refers to the perceptions of internal attributions that supervisors make for poor performance (i.e., supervisor- attribution conveyed through supervisory feedback). The former refers to the subordinate's self-attribution. As a threatening and aversive event, the perceptions of internal attributions that supervisors make for poor performance through their feedback have been viewed as having positive effects on helplessness in workplaces. Storms and McCaul (1976) and Dweck and Goetz (1978) contend that the subordinate's self- attributions also produce an increased depression. In this sense, it appears that self-attributions also affect the perception of helplessness. For example, a self-attribution ("I think I am not smart enough") can be as strong as a supervisor-attribution ("My supervisor thinks I am not smart enough") in producing the perception of helplessness. However, there is no empirical evidence to verify the effects of self-attributions on helplessness. Therefore, the relationship between helplessness and self-attributions of lack of ability and lack of effort is also an important issue in need of empirical verification. In sum, it is a considerable possibility that both self-attributions and supervisory attributions may influence 30 shame and helplessness. Therefore, the distinction between self-attributions and supervisor-attributions will be made in this study. Furthermore, helplessness have been identified as the dominant emotions produced by the experience of failure. Specifically, the literature discussed above suggests that the lack of ability and effort attributions produce an increased feelings of helplessness, but the effects of these two attributions differ in degree. For poor performers, ability rather than effort appears to be the main determinant of helplessness. That is, helplessness increases more when failure is attributed to a lack of ability than when failure is attributed to a lack of effort. Performance Improving Motivation Campbell and Pritchard (1976) contend that defining motivation is a painful intellectual endeavor because motivation is concerned with various aspects inherent in the process by which human behavior is activated. However, motivation is generally characterized by two common denominators. First, motivation refers to energetic forces within individuals that drive them to behave in certain ways and to environmental forces that often trigger these drives. Second, motivation is goal-oriented on the part of individuals; their behavior is directed toward a particular target (see Steers 8 Porter, 1987 for review). Thus, Campbell 8 Pritchard (1976) contend that it is necessary to equate motivation with a particular behavior or physical state. 31 Supervisory feedback is concerned with the improvement of poor performance on the part of subordinates. Thus, it is necessary to equate motivation with subordinate's performance improving. In this sense, performance improving motivation will be used for this study. This motivation refers to the degree of energetic forces that poor performers exert to improve their past behavior. The relationship between emotion and motivation has received much attention from motivation theorists. Tomkins (1963) suggests that emotions are viewed as motor and direction for behavior. In their study of help-giving behavior, Piliavin et al. (1969) also suggest that emotions have positive or negative effects on help-giving motivation. Therefore, two emotions, shame and helplessness, will be viewed as the predictors of performance improving motivation. Shame and Performance Improving Motivation As explored in the shame section, shame is related to a collapse of self-esteem related to the failure of social roles expected in social systems. This statement indicates several reasons explaining the relationship between shame and motivation. First, self-esteem from others and security have been long viewed as important human needs. That is, Maslow's (1943) need hierarchy theory contends that individuals are motivated to maintain or enhance self-esteem and security in the identity-threatening and job-threatening situations. Goffman (1959; 1969; 1971) focuses on the situational 32 conditions which help individuals establish and maintain temporary identities, which in turn provide a basis for regulating their behaviors. This situated identity may be a general one, consisting of only an overall evaluative impression of an individual as being good, inoffensive, and norm-abiding, or may be specifically tied to a particular role or attribute. Therefore, individuals attempt to avoid further blame and social disapproval by disassociating themselves from negative actions and outcomes (Crowne 8 Marlowe, 1967; Tedeschi, Malkis, 8 Fernandez, 1980). Therefore, shame has been viewed as a central motive in human development and interpersonal relations because esteem and security needs, and social approval dependence are strong motivators behind shame (Kaufman, 1985). Piers and Singer (1953) describe those motivators behind shame as internal sanctions. According to their internal sanction theory, individuals punish themselves to extent to which they feel shame, and attempt to perform appropriate behaviors. In line with the theoretical perspectives regarding shame, it is speculated that individuals can be motivated to improve their performance when they experience shame because of performance failure. The relationship between shame and motivation has been rarely investigated, though Covington and Omelich (1979) found a positive causal relationship between shame and performance. Therefore, it seems reasonable to suggest that shame will serve as a predictor of performance improving motivation. 33 Helplessness and Performance Improvinngotivation Traditionally, perceived helplessness has been conceptualized as a dysfunctional emotional state. For example, Wortman and Brehm (1975) point to a considerable body of evidence that helplessness is maladaptive: There is good reason to believe that exposure to uncontrollable outcomes can result in profound psychological upset. Many kinds of maladaptive behavior have been attributed to feelings of helplessness with respect to one's environment. For example, investigators have argued that the helplessness stemming from feelings of lack of control is an important factor in the development of such psychiatric disorders as depression and schizophrenia (Bateson, Jackson, Haley, 8 Weakland, 1965; Seligman, 1975). Cofer and Appley (1964), Jannis (1958), and Janis and Leventhal (1968) have maintained that feelings of helplessness interfere with the ability to respond adaptively in stressful situations. Feelings of helplessness have also been proposed as a precursor of physical disease (Engle, 1968; Schmale, 1971). Some investigators (Greene, Goldstein, 8 Moss, 1972; Richter, 1957) have suggested that the perception of inability to exert control over one's environment can even result in sudden death from coronary disease or other factors. Feelings of lack of control have also been viewed as a cause of many types of antisocial behaviors (p. 278). Wortman and Brehm's review underlines a common thread in all the findings; helplessness has a negative and maladaptive impact on human behavior. That is, the more expectation individuals have for control, the greater motivation they have for coping, while the less expectation, the less motivation. There has been widespread agreement with this view (see, for example, Bandura, 1977; Lawler, 1973; Lazarus 8 Feldman, 1984; Lewin, 1938; Tolman, 1932). For example, Lewin (1938) and Tolman (1932) have suggested that individuals have cognitive expectancies concerning outcomes that are likely to occur as a result of what they 34 do, and they have preferences among outcomes. That is, individuals behave according to the perceived degree to which a given act will lead to a desired outcome. Vroom (1964) also has posited that whether individuals have expectation for control is an important determinant of motivating them to engage in a certain behavior in work settings. Many empirical studies have demonstrated that whether expectation or no expectation for control can serve as a predictor of performance and achievement motivation. In their laboratory experiments using students as subordinates, for example, Arvey (1972), Motowidlo, Loehr, and Dunnette (1972), and Pritchard and Deleo (1973) found that the perceived expectation for control has a significant positive effect on subsequent behavior. In their field experiments, other researchers (e.g., Arvey 8 Neil, 1972; Dachler 8 Mobley, 1973; Gavin, 1970; Hackman 8 Porter, 1968; Lawler, 1968; Lawler 8 Porter, 1967; Mitchell 8 Albright, 1972; Mitchell 8 Nebeker, 1973; Schuster, Clark, 8 Rogers, 1971) found a significant correlation between motivation and perceived expectation for control. Lawler and Suttle (1973) found a positive causal relationship between these two variables in a field study. Many clinical studies also found that patients are more motivated to be health when their perceived helplessness is low than when high (See Lefcourt, 1982 for review). Thus, it appears that no expectation for control is a negative predictor of motivation to engage in an organizational behavior. As noted in the preceding section, helplessness is an 35 emotion occurring when there is no expectation for control over an aversive event or situation. Therefore, it is reasonable to suggest that helplessness has a negative effect on performance improving motivation. Research Hypotheses The literature reviewed above provides some important insights into the relationships among supervisory feedback, the perceived attributions of lack of effort and lack of ability, shame, helplessness, and performance improving motivation when employees fail to perform successfully in workplaces. These insights are incorporated into a causal model presented in Figure One. This model proposes that poor performers have perceptions about the attributions that their supervisors make in terms of lack of effort and lack of ability. Supervisory feedback serves as an exogenous variable in the model, and predicts the perceptions of these two attributions. The nature of supervisory feedback- perceptions of attributions links will be further explicated in the methods section of this dissertation. The perceptions of lack of effort and lack of ability then serve as positive predictors for shame and helplessness. Because of a debate between Covington and Beery's (1976) and Nicholls' (1976) model, it will be proposed that the perceptions of lack of ability and lack of effort will have a different effect on shame. In line with the models of Weiner et al. (1972) and Weiner (1979), this model will also 36 Figure 1: Hypothesized model SF SF: in: PLE: SH: HL: PIM: H- 2 sn PLA PIM PLE Supervisory Feedback Perception of Lack of Ability (Supervisor's) Perception of Lack of Effort (Supervisor's) Shame Helplessness Performance Improving Motivation 37 propose that the perceptions of lack of ability will have a stronger effect on helplessness than the perceptions of lack of effort. Shame and helplessness then leads to the later phases of performance improving motivation. As many helplessness theorists and researchers have suggested, however, helplessness is proposed to have a negative impact on performance improving motivation. Formal hypotheses for this model can be specified at both the macro and micro levels. There are two macro level hypotheses: H1: There is a significant difference between the proposed model and the null model of no relationships among the variables. H2: The proposed model provides an accurate representation of the relationships found in the data (i.e., a test of goodness of fit to the data). The micro level hypotheses are hypotheses of the significance of individual relationships in the model: H3: There is a significant positive relationship between the perceptions of lack of ability and shame. H4: There is a significant positive relationship between the perceptions of lack of ability and helplessness. H5: There is a significant positive relationship between the perceptions of lack of effort and shame. H6: There is a significant positive relationship between the perceptions of lack of effort and helplessness. H7: The relationship between the perceptions of lack of ability and helplessness is stronger than the 38 relationship between the perceptions of lack of effort and helplessness. H8: There is a significant positive relationship between shame and performance improving motivation. H9: There is a significant negative relationship between helplessness and performance improving motivation. In addition, two research question is posed: RQ1: Is there a significant difference between the perceptions of lack of ability and lack of effort in the prediction of shame? RQ2: What role do self-attributions play in the prediction of helplessness, shame, and performance improving motivation? CHAPTER TWO METHODS This chapter outlines the methodology and procedures used in this study. First, the subjects in the research are discussed, then the procedures and research design is described. Then, the operationalization of variables is considered. Finally, the analyses used to examine the relationships among perceptions of attributions, emotions, and performance improving motivation is presented. Sample and Procedure The proposed model was tested with a field experiment examining employees exposed to supervisory feedback messages ~ regarding poor performance. Data for this study were collected from three large general hospitals in the midwest. The sample included only the permanent staff nurses involved in direct patient care. A systematic sample of individuals from each organization was chosen to be surveyed. 407 staff nurses were chosen who are qualified to perform intravenous antibiotic therapy. They received questionnaires through mailboxes in hospitals. A cover letter explained the nature of the research and assured respondents of anonymity. Respondents returned their completed questionnaires through drop boxes prepared. Of the 407 questionnaires, 216 were returned for a response rate of approximately 53%. Of these, 207 questionnaires were complete enough to be used in analyses. 39 40 The mean job tenure was approximately 6 years, while the average time spent in the profession was 14.5 years. All respondents worked full-time and were women. Research Design and Procedure In order to assess the perceptions of lack of ability and lack of effort, it is necessary to expose employees to supervisory feedback messages regarding a poor performance incident. Thus, supervisory feedback messages must be manipulated in conjunction with lack of ability and lack of effort. The following scenario was provided for the manipulation of a poor performance incident. As a staff nurse, you were assigned to a four- year-old fresh post-op patient. One of the post-op orders was to begin intravenous antibiotic therapy immediately (first dose stat). The administration of the stat dose, however, was delayed, allowing the child's infection to worsen. The attending physician and the child's family were extremely upset. You and your su ervisor met to discuss this incident. In your discussion, you learned that your supervisor was concerned about two issues. First, she was concerned about whether the delay occurred because you weren't putting much effort into your job, and might have been socializing with coworkers. Your supervisor was also concerned about whether the delay occurred because you lacked ability to do your job, and were not able to accurately calculate the dru dosage and were not familiar with intravenous antibiotic therapy. Your supervisor talks with you to learn more about why the error occurs, then provides you with the following feedback: Supervisory feedback messages were manipulated to lead to the perceptions of lack of ability and lack of effort iattributions that supervisors make for a poor performance. Three levels of perceptions of lack of ability were crossed 41 with three levels of perceptions of lack of effort, leading to manipulating nine supervisory feedback messages. As noted earlier, supervisory feedback messages were manipulated in terms of two types of supervisory actions: (1) disciplinary and (2) managerial actions. Each subject was exposed to the situation and then received one of nine supervisory feedback messages. Manipulation of Disciplinary Actions As noted earlier, disciplinary actions refer to the levels of aversive events or the removal of positive events which supervisors present to stimulate poor performer's motivation. Among informal oral warning, formal written warning, suspension or loss of pay, and dismissal established in workplaces (O'Reilly 8 Weitz, 1980), formal written warning has been applied most often to serious cases other than illegal behaviors (Wheeler, 1976; Fisher, 1973; McAdams, 1978; Miner 8 Brewer, 1976). Formal written warning is defined as an organizational punishment that will have harmful consequences for an employee's promotion, pay increases, and their future events. The degree of harmful effects on employees may be specified. These formal warnings usually result in suspension. Thus, formal written warning will be used as the strongest disciplinary action. The following statement is a formal written warning message: "You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning. Remember you will be suspended if you have another formal 42 warning because of these attitude problems." As a disciplinary action, supervisors may criticize the attitude of employees resulting in job failure (Hirokawa 8 Miyahara, 1986; Kipnis, Schimidt, 8 Wilkinson, 1980; Wiseman 8 Schenck-Hamlin, 1981). In this sense, criticism is described as a supervisor's negative judgment or evaluation about an employee without any formal organizational punishment. Thus, criticism is a weaker disciplinary action than formal written warning. An example of criticism is: "You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." A final type of disciplinary action is attention. As an indirect aversive stimulation, attention is related to supervisory monitoring of an employee's behavior in the future (Hirokawa 8 Miyahara, 1986; Wiseman 8 Schenck-Hamlin, 1981). When supervisors may have a slight doubt about an employee's effort, though the perceptions of lack of effort is not strong, supervisors may include this attention statement in their feedback. No formal organizational punishment and no criticism are included in attention. Thus, attention will be regarded as a weak disciplinary action. An example of attention is: "I hope you administered the stat order in time. But I will be keeping an eye on you." Manipulation of Managerial Actions Managerial action here refers to the levels of directions that supervisors provide to develop subordinate's 43 future potential. The directions are concerned with learning skills and knowledge. Assigning employees to formal training has been identified as a managerial action which will be used when performance failure is attributed to a lack of skills and knowledge (Miner 8 Brewer, 1976). For poor performers, then, formal training means an obtrusive way of correcting a severe lack of ability. An example of formal training messages is: "I've enrolled you in a formal training program to help you deal with this problem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for the training program." Instead of assigning subordinates to formal training programs, supervisors may directly instruct subordinates, giving specific information necessary for the successful accomplishment of the task (Goodstadt 8 Kipnis, 1970; Instone et al., 1983) when supervisors believe that performance failure is moderately attributed to lack of ability. Instruction is provided in normal work settings where subordinates interact with supervisors. Thus, instruction is viewed as a weaker and less obtrusive managerial action. An example of this managerial action is: "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. Remember I can also help you." Finally, a third managerial action is recognition. As a reminding stimulation, recognition refers to supervisors' unobtrusive doubt about a lack of ability. In this managerial action, supervisors suggest an employee's 44 voluntary reassessment because the perception of lack of ability is not strong. Thus, it is reasonable to suggest that recognition is a relatively weak managerial action in terms of the perception of lack of ability. An example of this managerial action is: "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedures you used." In sum, supervisory feedback messages for the perception of lack of effort attribution was manipulated in terms of three disciplinary actions: formal written warning, criticism, and attention. For the perception of lack of ability attribution, supervisory feedback messages was manipulated in terms of three managerial actions: formal training, instruction, and recognition. These three managerial actions are crossed with the three disciplinary actions, yielding nine supervisory feedback messages. These nine messages are included in Appendix A. In order to check on the realism of the poor performance situation and whether supervisory feedback messages provided adequate manipulations for the perceptions of lack of ability and lack of effort attributions as designed, a pretest was performed. One hundred forty six staff nurses were asked to read the hypothetical poor performance situation (failure of intravenous antibiotic therapy) and one of the nine supervisory feedback messages manipulated, and then rate the degree to which how they agree with realism of the situation and lack of ability and effort of the supervisory feedback messages. The results of this pretest are discussed in the chapter three. The 45 questionnaires were sent to staff nurses through mailbox, filled out by staff nurses (confidentially), returned to the researcher through drop boxes. Qperationalization The proposed model requires the operationalization of (1) perceptions of lack of ability (subordinate's own and supervisor's), (2) perceptions of lack of effort (subordinate's own and supervisor's), (3) shame, (4) helplessness, and (5) performance improving motivation. The variables will be assessed with self-report measures because this study is concerned with perceptions of supervisory feedback messages. The operationalization of each of these variables is discussed below. Perceptions of Lack of Ability According to Dunnettee (1976), the term ability can be used in two ways. Ability first can indicate human aptitude encompassing more general patterns of performance. On the other hand, ability can be used as an index of a specific job performance. In other words, ability in one performance area may not fit into another performance area. For example, public speaking ability may prove to predict success in public relations, but have no relationship to auto engineering. In this study, the term ability is used in relation to a specific job performance, that is, intravenous antibiotic therapy. Ability here is used to ascertain how much is learned and how well a specific task can be performed. That is, the 46 focus is on evaluation of the acquired skill and knowledge (Super 8 Crites, 1962). Since the skill and knowledge are acquired through specialized vocational training and experiences, Super and Crites (1962) conceptualized ability as "the degree of mastery already acquired in an activity" (p. 73). Therefore, ability here can be defined as the degree to which individuals master skills and knowledge regarding intravenous antibiotic therapy. Ability was measured using the items constructed on the following six properties: (1) education, (2) understanding, (3) memory, (4) reliability, (5) swiftness, and (6) easement. Borman (1973) and Dunnettee, Groner, Holtzman, and Johnson (1972) identify these properties as measuring a single common underlying construct, knowledge utilization. Therefore, it is reasonable to treat the items using these six properties as a unidimensional measure of ability regarding intravenous antibiotic therapy. Following Steers and Porter (1987), this study distinguishes between self-attribution and supervisor- attribution. Thus, a distinction between subordinate's own perceptions and supervisor's about attributions was made. Supervisor's perceptions of lack of ability (i.e., ,supervisor-attribution) were measured by asking employees to respond to what their supervisor thinks about the attributes. Subordinate's perceptions of lack of ability were measured by asking them to respond to how they thinks of the attributes. Each of these items used a five-point response format which ranged from "strongly disagree" to 47 "strongly agree." The complete set of items used is included in Appendix B. Perceptions of Lack of Effort As noted earlier, work motivation or effort exerted refer to energetic forces expended for job accomplishment. Traditionally, work motivation or work effort exerted has been seen as an identical concept consisting of multiple factors (Landy 8 Guion, 1970). Based on a factor analysis of 16 core items developed by Herzberg et al. (1959), Friedlander (1955) found that work effort or motivation can be conceptualized in terms of three independent factors: (1) social environment, (2) intrinsic self-actualizing work, and (3) recognition through advancement. Pearce (1983) also conceptualized motivation in terms of three similar factors: (1) social, (2) intrinsic, and (3) service motivation. According to Friedlander's descriptions about these factors, Social environment refers to motivation about interpersonal and social relationships with other members in organizations. Intrinsic self-actualizing factor indicates "the development and full use of one's capacities and talents, particularly as related to the intrinsic work process itself and to the relationship of this process to the development and growth of the individual" (p. 145). The items include performing challenging assignments, the use of one's best abilities, a feeling of achievement in the work, the opportunity for freedom, and training and experience that help one's growth. Finally, recognition through advancement is concerned with the recognition of promotion 48 and salary increase. In this study, effort was conceptualized in terms of a single factor, intrinsic self-actualizing work. Thus, relevant items were adapted from the items involved in the intrinsic self-actualizing factor and then modified to fit the situation given in this study. Therefore, it is reasonable to treat the items designed as a unidimensional measure of effort. The distinction between supervisor's own perceptions and subordinate's own about attributions was also made in measuring this variable. All items were assessed using a five-point response format and are included in Appendix B. m According to Tomkins (1963) and Kaufman (1985), there are several different types of shame. Tomkins delineated four general types of shame: body shame (self is experienced as ugly), competence shame (self is experienced as incompetent), relationship shame (self is experienced as unlovable), and character shame (self is experienced as weak). In this theoretical framework, each type of shame was operationalized in terms of its source or context. Mirman (1984) and Novak (1986) tested the shame scales for unidimensionality. They suggest that the analyses generally result in a single factor solution and the items of each shame tend to measure a single common underlying construct. Thus, it is reasonable to treat the items constructed within the context of a specific type of shame as a unidimensional measure of a specific type of shame. 49 From a review of many shame measurement scales, Cook (1985) has developed a scale designed to measure the extent to which individuals experience the feeling of shame about themselves. The scale consists of childhood (23 items) and adult subscales (48 items) of shame measurement. For the measurement of shame, relevant items were adapted from the 48-item adult subscales of "Cook Shame Instrument," and the selected items were modified to fit the performance failure situation used in this study. Shame items were assessed using a five-point response format and are included in Appendix B. Helplessness As noted earlier, helplessness refers to an emotional response to perceived uncontrollability over an aversive event. Thus, helplessness measurement scales are concerned with how much individuals can (or can't) expect to control themselves or cope the situations (Peterson 8 Seliegman, 1984). In this sense, Berkman (1977) developed "Helplessness Index" including seven items. Lazarus and Folkman's (1984) "Ways of Coping" checklist consists of sixty seven items as to how individuals can (can't) cope aversive events. This instrument includes a set of helplessness-related items similar to Berkman's helplessness scale. According to Lazarus and Folkman, factor analyses show that these helplessness items can be seen as measuring a single factor. Therefore, helplessness was measured by adapting and modifying relevant items from the past measurement scales. These items were treated as a 50 unidimensional measure of helplessness. Helplessness items were also measured using a five-point response format and are included in Appendix B. Performance Improving Motivation Performance improving motivation here refers to energizing certain responses to improving a poor performance occurred in a particular time. This performance improving motivation is not concerned with overall work motivation, but only performance improvement in a particular time. However, past research on motivation has focused on measuring global nature of motivation rather than measuring specific nature. Therefore, it is important to distinguish the measurement of performance improving motivation from the global measurement of motivation. As Friedlander (1966) conceptualize, motivation here were conceptualized in terms of a single factor, intrinsic self-actualizing work. As noted earlier, performance improving motivation is concerned with a specific effort expenditure related to I'performance improvement" by supervisory feedback. Therefore, performance improving motivation was measured by constructing relevant items. Hermans's (1970), Landy and Guion's (1970), and Herzberg, Mausner, and Snyderman's (1959) items were used for this item construction. The items included in Appendix B were also assessed using a five-point response format. 51 Analysis The investigation of attribution-emotion-motivation model of supervisory feedback involved two types of analysis. These are the evaluation of measurement models and the estimation of a structural equation model. These two types of analysis will be discussed separately. Measurement Model Confirmatory factor analysis was used to confirm the factor structure of all the scales. If the confirmatory factor analysis indicates that certain items are not internally consistent with other items, a sequential elimination of items was used to reach the most parsimonious and internally consistent factor. Thus, the use of confirmatory techniques on the measurement scales can be seen as an attempt to use the highest quality scale possible, with the realization that some generalizability might be sacrificed. The measurement models were analyzed with the confirmatory factor analysis subroutine of the PACKAGE computer program (Hunter 8 Lin, 1987). The confirmatory factor analysis subroutine of PACKAGE allows the researcher to specify an a priori factor structure. The program then provides a matrix of correlations among cluster items, correlations between factors, and factor loadings, using communalities in the diagonal. The unidimensionality of the proposed factors can then be assessed using three criteria proposed by Hunter (1980): (1) Homogeneity of item content, (2) internal consistency within clusters, and (3) 52 parallelism with outside variables. These criteria were used in assessing the quality of all scales. Structural Equation Model In Chapter one, an attribution-emotion-motivation model of supervisory feedback was presented. This path model was constructed and analyzed using data from three large general hospitals. The data input to the path model was the correlations among factors identified in the confirmatory factor analysis discussed above. The path model was analyzed using the LISREL VI computer program (Joreskog 8 Sorbom, 1985). This program provides estimates of parameters in linear structural equations through the method of maximum likelihood. LISREL VI also provides a chi square goodness-of-fit test for models which are overidentified. As the proposed model is a less than fully recursive model, it is overidentified, and the fit of the model can be tested. PATH ANALYSIS subroutine of PACKAGE computer program (Hunter 8 Lim, 1987) was also used in an attempt to improve its fit to the data. The first hypothesis stated that the proposed model would be significantly different from the null model. The null model examined was one of no relationship among the theoretical variables. The hypothesis of difference between the two models was tested with a test for the significance of the difference between chi square values for each of the models. This difference between chi square values is itself distributed as chi square with degrees of freedom equal to the difference in degrees of freedom for the two models. 53 The critical value for this test was set at alpha = .05. The second hypothesis stated that the proposed model would be a good fit to the data. This was tested using the chi square goodness-of-fit test which determines whether the structure imposed by the restrictions of the model matches the structure that created the variances and covariances in the data. The critical value for this test was also set at alpha = .05. The fit of the model was also considered through the examination of residual covariances. The last set of hypotheses considered the significance of individual structural coefficients in the model. These hypotheses were tested with the t-values of the coefficients. The critical value for t was also set at alpha = .05. T-tests were used to assess differences among coefficients and between self/supervisory attributions. CHAPTER THREE RESULTS This chapter presents the results of research investigating the attribution-emotion-motivation model of supervisory feedback that were discussed in the first two chapters. First, pretest results are presented. Second, results of confirmatory factor analyses investigating measurement scales are considered. Third, the testing of the attribution-emotion-motivation path model proposed is considered, model revisions presented, and the fit of the proposed and revised models is discussed. Finally, some additional post hoc analyses are considered. Pretest In order to check whether the hypothetical poor performance situation was realistic and whether variations in the disciplinary and managerial action messages properly manipulated lack of ability and lack of effort, a pretest was performed. One hundred forty six staff nurses were asked to read the hypothetical poor performance situation (failure of intravenous antibiotic therapY) and one of the nine supervisory feedback messages (i.e., formal training/formal written warning, instruction/formal written warning, attention/formal written warning, formal training/criticism, instruction/criticism, attention/criticism, formal training/recognition, 54 55 instruction/recognition, and attention/recognition). Respondents then rated the degree to which the situation was realistic and the degree to which they believed supervisors would attribute their behavior to lack of ability and lack of effort. Five-point Likert-type scales were employed for this pretest, "5" represented "strongly agree" and "1" represented "strongly disagree." If the manipulation of lack of ability and effort was done successfully, for example, staff nurses who received the "formal training/formal written warning" supervisory feedback messages should rate lack of ability and lack of effort higher than those who received the "attention/recognition" messages (see Appendix A for example of pretest questionnaire). Descriptive statistics for the pretest checks are presented in Table 1 (realism check), Table 2 (lack of ability check) and Table 3 (lack of effort check). As Table 1 shows, realism was checked by nine staff nurse groups. The realism scores of these groups were ranged from 4.01 to 4.37. Grand mean (4.28 on five-point response scale) indicates that the job failure situation was perceived highly realistic. Overall, this realism was perceived across the nine different groups equally (F=.238, df=4/137, p <.917, eta square=.oo7). Insert Table 1 about here 56 A second two-way analysis of variance for the rated lack of ability yielded significant main effects of managerial actions (F=317.01, df=2/137, p <.0001, eta square=.82) and no main effects of disciplinary actions (F=.624, df=2/137, p <.538, eta square=.002). No interaction effect between managerial and disciplinary actions were revealed (F=.26, df=4/137, p <.903, eta square=.001). Thus, the manipulation of managerial actions seemed to be in producing perceptions of ability attributions. Insert Table 2 about here A final two-way analysis of variance was performed on the rated lack of effort. This analysis showed that disciplinary actions had a significant main effects on lack of effort attribution ratings (F=142.419, df=2/137, p <.0001, eta square=.67) while managerial actions had no main effect on lack of effort attribution ratings (F=.249, df=2/137, p <.78, eta square=.001). No interaction effect between managerial and disciplinary actions was shown. Thus, disciplinary actions seemed to manipulate lack of effort attributions successfully. Insert Table 3 about here 57 Lack of homogeneity of error variance may distort the analysis of variance when subjects are assigned randomly to treatment groups and the treatment groups contain equal numbers of subjects (see Hays, 1981; Keppel, 1982 for review). In this regard, homogeneity of variance was considered. Cochran's test of homogeneity of variance showed that variances for rated realism were homogeneous across the nine different conditions (Cochran's C=.14, p >.05) while Hartley's F-max test of homogeneity showed that variances of the rated realism were heterogeneous across the nine different conditions (Hartley's F-max=6.46, p <.05). Second, a test of homogeneity of variance revealed that variances of the rated lack of ability were homogeneous across the nine different conditions (Hartley's F-max=3.10, p >.05; Cochran's C=.216, p >.05). Finally, Cochran's test of homogeneity of variances revealed that the rated lack of effort is homogeneous across the nine different conditions (Cochran's C=.356, p >.05) while Hartley's F-max test showed that the rated lack of effort is not homogeneous (Hartley's F-max=6.89, p <.05). Cochran's test and Hartley's F-max test show different results in testing the homogeneity of error variances of the rated lack of effort and realism. The main virtures of Hartely's F-max test are its simplicity and ease of computation, but this test is less robust and less powerful than Cochran's test particularly when sample sizes of independent groups are unequal. Thus, it is reasonable to accept the results of Cochran's tests. The Cochran's homogeneity tests above indicate that the variances for the 58 rated realism, lack of ability, and lack of effort are homogeneous across the nine different conditions. Accordingly, the manipulation of the poor performance situation and the lack of ability and effort seemed to be successful. Confirmatory Factor Analyses Confirmatory factor analyses (Hunter 8 Gerbing, 1982) were performed to determine whether proposed scales for measuring perceptions of lack of ability and effort, shame, helplessness, and performance improving motivation were unidimensional. The measurement models were analyzed with the Confirmatory factor analysis subroutine of the PACKAGE computer program (Hunter 8 Lim, 1987). Hunter (1980) has suggested three criteria for assessing the unidimensionality of hypothesized scales: (1) homogeneity of item content, (2) internal consistency of items, and (3) parallelism with outside factors. These criteria were used to evaluate the proposed measurement models. For homogeneity of item content, items within each factor were derived from standard scales or modified with the goal of tapping a single concept. Thus, the items in each cluster appear to be homogeneous in content. Because internally consistent factors are a necessary prerequisite to investigations of parallelism, the subsequent discussion will deal first with the issue of internal consistency and then the issue of parallelism will be considered. Internal consistency deviations are the difference between the actual inter-item correlations and the correlations predicted from 59 the proposed factor structure (product of item factor loadings). Internal Consistency As shown in Table 4, 5 and 6, factor clusters for perception of lack of ability (PLA), perception of lack of effort (PLE) and helplessness (HL) all have relatively high and consistent factor loadings, and high alphas. Insert Table 4, 5, and 6 about here Table 4 indicates that the lowest factor loading for PLA factor is .58 for item 18 and the highest factor loading is .74 for item 20. The average factor loading and alpha for this factor is .68 and .93, respectively. Table 9 shows that PLA factor is internally consistent. Thus, this factor was accepted as hypothesized. Table 5 indicates that the six items in PLE form an acceptable factor with an alpha of .91 and an average factor loading of .62 ranging from .42 to .72. Table 10 indicates that the six items of PLE factor are internally consistent. The HL factor was also accepted as hypothesized. Table 6 shows relatively high factor loadings across the six items with an alpha of .87. Table 11 indicates that the six items of the HL factor are internally consistent. Thus, these three factors (i.e., PLA, PLE, and HL) were accepted as hypothesized. 60 Insert Table 9, 10, and 11 about here It was predicted that the 18 items of the shame would load on one internally consistent factor. Table 7A presents the items and loadings for this factor computed in confirmatory factor analysis. The deviation matrix for shame is presented in Table 7B. The deviation matrix indicates a number of deviations larger than would be expected from sampling error. There are also a number of factor loadings which are relatively small (i.e., .13 to .22). Thus, it appears that the shame factor structure is not strong or unidimensional. In an attempt to improve the factor, a series of analyses were performed in which items were eliminated sequentially until the deviations and factor loadings showed the most parsimonious and internally consistent factor structure. The final factor solution included ten of the original 18 items for the shame factor (item 29, 31, 32, 35, 37, 39, 40, 41, 42, and 44). A second confirmatory factor analysis was performed with ten items in the shame factor. The deviations and factor loadings for this final version of shame are presented in Table 12. The standard score alpha for this final version of the shame factor structure was .90. Insert Table 7A, 7B, and 12 about here 61 It was expected that the 12 items of the performance improving motivation (PIM) would be one internally consistent factor. Table 8A presents items and factor loadings produced in confirmatory factor analysis. Table 8B presents the deviation matrix for this factor. Item 59 shows a large deviation and a small factor loading. Thus, this item was eliminated and then a second confirmatory factor analysis was performed with 11 items in the PIM factor. The 11 items represented a unidimensional factor structure. The deviations for the final version of PIM are presented in Table 13. The standard score alpha for this factor was .89. In summary, confirmatory factor analyses indicated that the items of PLA, PLE, and HL factors were internally consistent as hypothesized, while ten of the 18 items in SH and 11 of the 12 items in PIM were internally consistent. Parallelism of the shame and PIM factors was examined on these reduced scales. Insert Table 8A, 8B, and 13 about here Parallelism According to Hunter (1980), the concept of parallelism suggests that all items within a factor should have homogeneous item to total correlations with outside factors. For example, if all the items in a given factor have the same item-total correlations with an outside factor, the 62 items in the first factor are said to be parallel with respect to the outside factor. In practice, deviations from this model that are smaller than what one would expect from sampling error alone reflect parallelism. Insert Table 14 through 18 about here To test for parallelism between PLA and the other factors, for example, the mean correlations between PLA items (Item 17, 18, 19, 20, 21 and 22) and each of the other factors were produced. Then, for each outside factor, the deviations between the actual PLA item-to-outside factor correlations and the mean PLA item—to-outside factor correlation were computed. The parallelism deviations are presented in Table 14 through 18. As a whole, all the factors show reasonably good parallelism. That is, 96% of deviations in the PLA factor, 83% in the PLE factor, 75% in the SH, 92% in HL, and 86% in PIM are smaller than what would be expected from sampling error. The parallelism for the shame factor is not as strong as the other factors. In the shame factor, item 42 has correlations which are consistently lower than the mean item-to-outside factor correlation. However, this item is also the least reliable in the factor (communality of .38 as compared to .52, .40, .48, .45, .54, .55, .46, .48, and .39). That is, lower parallelism of this factor can be attributed to the reliability of item 42. Thus, all the 63 factors were retained without changes for further stages of analysis. Path Analysis Because unidimensional measurement models were confirmed for all of the theoretical variables involved, correlations between factors were used for input in path analysis of a hypothesized structural equation model. The correlations were computed and corrected for attenuation due to measurement error using the CORRELATION subroutine of the PACKAGE computer program (Hunter 8 Lim, 1987). These corrected correlations are presented in Table 19. First, the hypothesized model was run using the LISREL VI computer program (Joreskog 8 Sorbom, 1985). The model was evaluated in terms of (1) the difference between the proposed model and the null model of no relationship, (2) the fit of the model to the data, and (3) the significance of individual links in the model. If the model did not fit the data well, the PATH ANALYSIS subroutine of PACKAGE computer program (Hunter 8 Lim, 1987) was used to determine if the addition of theoretically consistent links could significantly improve the fit of the model. One aspect of the zero-order correlations among theoretical variables are worth mentioning before examining the structural equation model analyzed. There are two extremely high correlations between manipulated ability of supervisory feedback and shame (.82) and between shame and performance improving motivation (.82). These could indicate that higher order factors exists and that all items 64 of two correlated factors are indicators of the high order factor. That is, if MA and SH are random versions of an identical factor, these two variables should have the same correlation with the other variables. SH and PIM should have the same correlation with the other variables if these two variables are identical. As shown in Table 19, the correlations between MA and PLA, MA and PLE, MA and HL, and MA and PIM are .66, .33, .39, and .62 while the correlations between SH and PLA, SH and PLE, SH and HL, and SH and PIM are .74, .55, .46, and .82, respectively. The comparison of these values between MA and SH indicates no evidence that these two variables are identical. To examine whether SH and PIM are random versions of an identical factor, the correlations between these two variables (SH and PIM) and other variables. This comparison also showed no evidence to doubt that these two variables are identical. For example, the correlation between SH and MA is .82 while the correlation between PIM and MA is .62. Furthermore, the content of the items between MA and SH and between SH and PIM is clearly distinct because items within each factor and manipulation were produced with the goal of tapping a single concept different from others. Therefore, it seemed reasonable to maintain these factors as conceptually separate entities in spite of the high correlations. The hypothesized model representing the theory developed in Chapter one is presented in Figure 1. This model indicates two causal indicators of performance improving motivation (PIM) -- shame (SH) and helplessness. 65 Figure 2: Hypothesized model with path coefficients and standard errors. PIM: .58 .66 .61 MA % PLA 7. ex (.053) .79 (.044) PIM .08 (.041) .68 ME 5 PLE a) HL (.051) (.066) R square for dependent variables: PLA: .44 chi square = 377.07 PLE: .46 df = 12 SH: .69 HL: .24 All coefficients are significant, p <.05 MA: Manipulation of ability in supervisory feedback Manipulation of effort in supervisory feedback 66 (HL). Shame and helplessness are caused by two supervisory attributions, that is, perception of lack of ability (PLA) and perception of lack of effort (PLE). Perceptions of lack of ability and effort are caused by manipulated ability (MA) and effort (ME) through supervisory feedback messages. This hypothesized model, with path coefficients, standard errors, squared multiple correlation coefficients, and chi square value is presented in Figure 2. All links in this model are significant (p <.05), though the link from HL to PIM is extremely low. Thus, it is clear that this link should be removed in the model. Although this model is significantly better than the null model of no relationship (chi square difference of 1066.81 with 20 degrees of freedom), it is apparent that this model is not a good fit to the data. The chi square for this model is 377.07, with 12 degrees of freedom. This value is significant, and the chi square to degrees of freedom ratio is not under five (the value suggested by Wheaton, Muthen, Alwin, 8 Summers, 1977, for a good fitting model). The goodness-of-fit index was .71 while its adjusted index was .38. Thus, modification indices and model residuals were examined and links were added to the model in an attempt to improve its fit to the data. On the other hand, path analyses of PACKAGE program also showed the same result. That is, the average squared deviation for correlations reproduced from this model was .071 and the average absolute deviation was .22. Because both of these values are relatively high, residuals were investigated to improve the model. 67 Figure 3: Revised model with path coefficients and standard errors. .66 PIM .62 .82 (.051) (.040) SH .23 .46 (.051) (.063) ME sr>PLE HL (.046) R square for dependent variables: PLA: .57 chi square = 259.30 PLE: .57 df = 12 SH: .58 HL: .21 All coefficients are PIM: .67 significant, p <.05 68 Three links were removed from the model and then three new links were added to the model. In sequence, the links removed and added were: (1) from HL to PIM (removed), (2) from SH to HL (added), (3) from PLA to HL (removed), (4) from PLE to HL (removed), (5) from MA to PLE (added), and (6) from ME to PLA (added). These additions and removals of links seemed theoretically justifiable. The added links from MA to PLE and from ME to PLA are justifiable through the halo effect theory (Dion, Berscheid, 8 Walster, 1972). According to halo effect theory, even if a supervisor gives a feedback indicating lack of ability, the feedback may affect not only perceptions of lack of ability but also perceptions of lack of effort. On the other hand, Pearlin (1980) and Wortman 8 Dintzer (1978) contend that the feeling of helplessness can be a final stage of an adaptational sequence. The removed link from HL to PIM and the addition of the link between SH and HL also seemed justifiable through Pearlin's (1980) contention that the feeling of helplessness can be a final stage of adaptation sequence. The theoretical implications of changes in the model are discussed in detail in Chapter 4. The revised model is presented in Figure 3. All of the path coefficients in this model are significant, and, as with the hypothesized model, this model is a significantly better fit than the null. Goodness-of-fit index was .79 while its adjusted index was .54. The chi square value for this revised model is much smaller than the hypothesized model -- 259.30 with 12 degrees of freedom. PATH ANALYSIS of PACKAGE showed that the average squared deviation for. 69 Figure 4: Revised perceptual model (without manipulation) with path coefficients and standard errors. PIM .62 .82 (.052) (.040) HL SH .23 .46 (.052) (.062) PLA PLE R square for dependent variables: SH: .59 chi square = 35.52 HL: .21 df = 5 PIM: .67 All coefficients are significant, p <.05 70 correlations reproduced from the revised model was .014 and the average absolute deviation was .07. Although both of values are relatively low and chi square value is significant, the chi square to degrees of freedom ratio is larger than five, indicating a poor fit of the model to the data. Thus, modification indices of LISREL VI and error indices of PATH ANALYSIS of PACKAGE were reexamined in an attempt to improve its fit to the data. This reexamination indicated that MA and ME seemed to contribute to the larger chi square value. Although the removal of MA-PLA, MA-PLE, ME-PLA, and ME-PLE can produce a better fit to data, it is conceptually important to maintain MA and ME because supervisory feedback is an essential exogenous variables in the model. However, a further revision of the model was made in order to see interesting and important insights that only perceptual model may provide for this study. This revised perceptual model includes only four variables (i.e., PLA, PLE, SH, HL, and PIM). As shown in Figure 4, the revised perceptual model provides an improved fit to the data. All of the path coefficients in this model are significant. This model is a significantly better fit than the null (chi square difference of 461.46 with four degrees of freedom). Goodness-of—fit index of this model is .94 while its adjusted index is .87. The chi square value for this model is quite low -- 35.52 with five degrees of freedom. This chi square value is significant and the chi square to degrees of freedom ratio is slightly larger than five. This 71 model is quite parsimonious which includes only 4 structural coefficients, and overall variance explained by this model is .59. PATH ANALYSIS of PACKAGE showed that the average squared deviation for correlations reproduced from this revised perceptual model was .0037 and the average absolute deviation was .037. These values seem to be quite low, indicating a good fit to the data. Thus, this model was considered as a close representation of gply the perceptual processes. Other Analyses The last set of hypotheses were examined considering the significance of individual structural coefficients in the hypothesized model. As indicated in Figure 2, all coefficients were significant when the critical value for p- values of the coefficients was set at alpha =.05. However, the link from HL to PIM indicated an extremely low coefficient (.08). On the other hand, Figure 2 indicates that the link between PLA and HL is stronger than the link between PLE and HL. To test this difference, confidence intervals of two links was compared at alpha =.05. This comparison showed an overlap between two confidence intervals, indicating that the difference between two links was not significant. Thus, the hypothesis 7 was not supported. The first research question was also examined comparing confidence intervals of the two links between PLA and SH and PLE and SH. This comparison indicated that no overlap was found between two confidence intervals, meaning that the coefficient difference between these two links was 72 significant. Finally, analysis of variance was used to examine the second research question. Two levels (high and low) of subordinate's own perceptions of lack of ability (MYPLA) and lack of effort (MYPLE) was examined in terms of shame, helplessness, and performance improving motivation. First, two levels of MYPLA differed significantly in shame (F=6.83, df=1, p <.01), but two levels of MYPLE showed no significant difference in shame (F=.003, df=1, p <.95). Second, two levels of MYPLA and MYPLE indicated no significant difference in helplessness (F=3.05, df=1, p <.08 and F=3.44, df=1, p <.07). Finally, two levels of MYPLA and MYPLE differed significantly in performance improving motivation (F=4.63, df=1, p <.03 and F=5.07, df=1, p <.03). Further analysis of variance was performed to see if there was an interaction between MYPLA and shame, and between MYPLE and shame in predicting performance improving motivation and helplessness. No interaction was found in this investigation. Finally, a T-test was performed to see the difference between subordinate's own perceptions and supervisor's perceptions of attributions. The difference between two perceptions was significant (t=20.27, df=207, p <.0001 for lack of ability and t=22.09, df=206, p <.0001 for lack of effort). CHAPTER FOUR DISCUSSION This chapter discusses the implications of analyses presented in Chapter Three. First, implications of the confirmatory factor analyses performed on the established scales are considered. Second, the hypothesized and final models are discussed and the strength of path coefficients is also compared. Finally, limitations of this study and directions for future research are discussed. Measurement Models Five theoretical variables were submitted to confirmatory factor analysis to evaluate their unidimensionality. These were perception of lack of ability (PLA), perception of lack of effort (PLE), shame (SH), helplessness (HL), and performance improving motivation (PIM). The items for these factors were adapted from highly reliable and valid scales used widely in past research. The current analysis suggests that when these factor structures are examined in depth for internal consistency and parallelism, the quality of these scales is quite high except for the shame scale. Eight items were dropped from the measurement model for shame. This deletion may indicate that respondents did not successfully interpret the items. However, this is not likely because the sample used for this study were staff 73 74 nurses with at least two years education in college. Another possible explanation for this deletion can be found in the items themselves. An intuitive look at the deleted items suggests that such items were not associated with temporary shame relevant to the failure of intravenous antibiotic therapy. Rather, the deleted items were associated with a chronic and enduring shame relevant to an event more than the failure of intravenous antibiotic therapy (Novak, 1986). That is, the deleted items seemed to go beyond the domain of shame which might be produced in the feedback situation in workplaces. For example, the item 38, "It would be difficult for me accept a compliment from others," may be 'too general' to describe the feeling of shame associated with the failure of intravenous antibiotic therapy. Thus, it is reasonable to suggest that more caution should be exerted in distinguishing between chronic and temporary feelings of shame. Path'Models Hypothesized and revised models This study extended attribution and emotion theories to the effect of supervisory feedback about poor performance in workplaces. In this study, a poor performer's motivation (PIM) was hypothesized to be a function of supervisory perceptions of attributions (PLA and PLE) revealed through supervisory feedback (MA and ME) and mediated by emotions (SH and HL). The structural equation model hypothesized in Chapter One was significantly different from the null model, but not a good fit to the data. Hence, model modifications 75 were made in an attempt to improve the model. Although a revised model was found which provided a significantly improved fit to the data, the revised model was not a good fit to the data. Thus, the revised model was considered as a giggg representation of the attribution-emotion-motivation processes. This revised model showed several striking differences from the hypothesized model. The revised perceptual model was also compared with the original model. The first difference between the proposed model and revised model is the deletion of the hypothesized link from helplessness to performance improving motivation. The revised model does not support the traditional contention that helpless people make no effort to take the steps necessary to change negative outcomes. Rather, this deletion supports some helplessness theorists'(e.g., Pearlin, 1980) contention that helplessness is a final stage variable having no subsequent effect. That is, Pearlin (1980) suggests that, when people are once involved in a helpless situation, they tend to commit themselves to being controlled by other things. For example, if a worker is in a helpless situation because a supervisor attributed his/her failure of intravenous antibiotic therapy to lack of ability or effort, the worker may commit him/herself to performing better in tasks gppgg than intravenous antibiotic therapy. However, the feeling of helplessness does not appear to hamper (or promote) subsequent motivation about a particular tasks. The second difference between the proposed model and the revised model is the addition of a direct link from 76 shame to helplessness. In the proposed model, it was not hypothesized that shame would influence helplessness; that is, shame is independent of helplessness. However, the final model indicates that shame influences helplessness. This added link between shame and helplessness is relatively strong (.46). Furthermore, the addition of this link dropped the path coefficients from PLA to HL and from PLE to HL from .23 and .17 to -.06 and .06, respectively. Apparently, shame is a strong intermediate factor for developing both performance improving motivation and helplessness. Thus, the final model can be distinguished from the hypothesized model because the final model consists of two contrasting effects of supervisory feedback: (1) PLA (PLE) --- SH --- PIM, and (2) PLA (PLE) --- SH --- HL. The third difference between the proposed model and the revised model is the addition of two direct links from MA to PLE and from ME to PLE. According to halo effect theory (Dion, Berscheid, 8 Walster, 1972), although supervisors may talk about lack of ability, the subordinates may perceive not only lack of ability but also lack of effort. By the same token, when supervisors talk about lack of effort, this supervisory comment may influence not only perceptions of lack of effort but also perceptions of lack of ability. The strength of link between MA and PLE is .33 while the strength of link between ME and PLA is .36. Finally, two similarities between these two models should be noted. In both models, the link from PLA to shame is very strong. On the other hand, the link between PLE and shame is statistically significant in the two models, but 77 weaker than the link between PLA and shame. That is, PLA is a stronger determinant of SH than PLE. Both models also show that shame is a very strong determinant of performance improving motivation. Relationships among variables in the revised model The revised perceptual model was explored further by examining three areas: the relative impact of the determinants of shame and the relative impact of shame on performance improving motivation and helplessness. First, an examination of the standardized structural coefficients in the model indicates that PLA and PLE did not have an equal impact on shame. The coefficient for the link between PLA and shame was .62 while the coefficient linking PLE and shame was .23. The comparison of these two values in terms of confidence interval indicated that the difference between two links was significant at alpha = .05. Thus, it is clear that perception of lack of ability has a much greater impact on shame than perception of lack of effort in this model (the first research question). This finding does not support Baker, DiMarco, and Scott's (1975) contention that lack of effort can be a more important determinant of shame than lack of ability because lack of effort elicits stronger immoral feeling than lack of ability. Rather, this finding supports Weiner, Russell, and Lerman's (1978) explanation for the importance of ability over effort in predicting shame. That is, from the viewpoint of subordinates, lack of ability is a stronger determinant of shame than lack of effort because individuals 78 are more vulnerable to the lack of a more stable dispositional traits. Both the revised model and perceptual model include two processes of shame: (1) PLA (PLE) --- SH --- PIM; (2) PLA (PLE) --- SH --- HL. In the first process, shame has a very strong influence on performance improving motivation (path coefficient = .82). In the second process, shame has a relatively strong impact on helplessness (path coefficient = .46), but not as strong as the influence on motivation in the first process. This model suggests that shame plays a critical role in predicting both performance improving motivation and helplessness. However, there are two possible explanations for the two processes in the model. First, SH --- PIM and SH --- HL may be viewed as a dual (either/or) process. Second, both processes may be viewed as occurring simultaneously. Thus, it is important to examine whether these two processes can be viewed as a dual (either/or) process or both processes can be viewed as occurring simultaneously. It was not hypothesized that shame would influence helplessness, but it was hypothesized that shame would be independent of helplessness. A post hoc analysis was performed to examine the relationship between shame and helplessness. A scatter plot produced by shame and helplessness showed that there is no evidence to support a negative relationship between these two variables. Rather, the scatter plot indicated a positive relationship between these two variables (r = .46). Thus, it is clear that shame has a positive impact on both performance improving 79 motivation and helplessness simultaneously. There are some possible explanations for this. As discussed in Chapter One, shame is viewed as a central motive in human development and interpersonal relations because esteem and security needs, and social approval dependence are strong motivators behind shame. That is, our discussion focuses only on the functional aspect of shame. However, several helplessness theorists and researchers have suggested that shame is also dysfunctional (Lynd, 1958: Storms 8 McCaul, 1976). For example, Storms and McCaul (1976) contend that helplessness can be a final stage of emotional exacerbation process occurring as a function of other less damaging emotions. Particularly, it was suggested that shame is one of the most important emotional variables increasing helplessness (for review, see Duval 8 Wicklund, 1972). Thus, it seems reasonable to suggest that shame can increase both performance improving motivation (functional) and helplessness (dysfunctional) simultaneously. Other Analyses In this study, the distinction between self- attributions and supervisory attributions was made. A T- test analysis indicated that the difference between self- attribution and supervisory attribution was statistically significant (t=20.27, df=1, p <.0001 for lack of ability and t=22.09, df=1, p <.0001 for lack of effort). Thus, it appears that, although subordinates believe that their supervisor thought of them as having low ability and effort, 80 the subordinates were reluctant to think of themselves as having no ability and effort. This result supports the notion of self-protecting attributions (Miller 8 Ross, 1975), that is, people avoid attributing their failure to internal factors. Past research also indicates that people tend to attribute the failure more to effort than ability. However, no significant difference was found between these two variables in this study. The second research question was concerned with the role of self-attribution in predicting shame, helplessness, and performance improving motivation. T-test analyses indicated that subordinate's own perceptions of lack of ability (MYPLA) had significant effects on shame and performance improving motivation. However, subordinate's own perceptions of lack of effort (MYPLE) showed significant effects only on performance improving motivation. Furthermore, no significant interaction were found between subordinate's own perceptions and shame in predicting helplessness and performance improving motivation. Thus, it is clear that the role of self-attributions is not as strong as the role of supervisory attributions in this model. Implications for Managerial Practice Traditionally, many researchers (see Deci 8 Ryan, 1980 for review) have emphasized the use of positive feedback to enhance workers' motivation. This tradition has induced managers in workplaces to avoid provoking negative feelings on the part of workers. That is, managers "have talked in very general terms to the subordinates, alluding only 81 vaguely to problems" (Beer, 1981). As a result, performance appraisal systems have not brought about a positive change in workers' performance in about half of the organizations surveyed (Latham 8 Wexley, 1982). However, this research suggests that managers can enhance poor performer' motivation by provoking the feeling of shame because shame is functional in workplaces. To utilize this functional aspect of shame in increasing motivation, it is suggested that managers need to stimulate workers' feeling of shame through feedback messages relevant to increasing the perceptions of lack of ability or effort. According to the results of this investigation, supervisory feedback message strategies include the use of disciplinary messages for the perceptions of lack of effort and the use of managerial messages for the perceptions of lack of ability. Limitations and Directions for Future Research This study included several limitations, all of which suggest ways in which this study could be improved and extended. This section will discuss these limitations and propose further ways in which the supervisory feedback processes could be fruitfully investigated. One way in which this research could be improved involves the measurement of shame in the model. As mentioned earlier in this chapter, chronic and enduring shame may not be relevant to a specific situation or event. The chronic and enduring shame items were developed to assess intense emotional disorders in clinical psychology 82 (Novak, 1986). Thus, this aspect should be considered by differentiating chronic and enduring shame items from the temporary shame items in constructing a measurement scale. Another way in which this research could be improved is to include the manipulation of self-attributions in the situation. In this study, most respondents responded to the questionnaire supposing that they had ability and effort to perform intravenous antibiotic therapy, though they thought that their supervisor attributed the job failure to lack of ability and effort. Future research should include situations in which respondents both 33333 and disagree with their supervisor's attributions. This would increase our knowledge about how poor performers respond to supervisory feedback when they accept supervisory attributions. Finally, this study has several reasons to limit the generalization of its major findings because the sample for this study was entirely female. First, there are some questions as to whether males and females differ in making attributions about failure. For example, Deaux and Farris (1977) contend that males are more likely to attribute failure to lack of effort, while females are more likely to attribute failure to lack of ability. This attributional difference between male and female may influence the supervisory feedback process. There is another reason to limit the generalization of the findings. The given situation for this study is concerned with a child patient. This situation may increase an emphasis on the finding regarding shame and performance improving motivation because female nurses may feel a stronger obligation to care for the 83 child patient. Thus, it seems reasonable to suggest that a caution is necessary in generalizing these research results. In addition to considering these limitations, it seems useful to consider several other issues. A first issue is concerned with the need for a better fitting model. The hypothesized model was revised in an attempt to find a better fit to the data. The revised perceptual model explained a great deal of the variance in developing performance improving motivation. However, this final model cannot be accepted with full confidence because a variety of fit indices indicated a less than ideal fit to the data. A second issue is the inclusion of other attributional and emotional variables. This study showed that attributions and shame are important intermediate variables in accounting for performance improving motivation in the supervisory feedback processes. However, Weiner, Russell, and Lerman (1978) suggest a variety of emotions as a function of several different causal attributions. For example, other causal attributions of failure include task difficulty, luck, personality (e.g., intrinsic motivation), and mood while other emotions include despair, hostility, and surprise. Thus, this research could be extended to consider these other attributional variables and emotional variables in order to understand more about the effects of attributions and emotions on motivation. Conclusion This research considered issues of importance to both organizational scholars and practioners: the role of 84 attribution and emotion in accounting for the effects of supervisory feedback on performance improving motivation. Two attributions were lack of ability and lack of effort, and two emotions were shame and helplessness. Theoretically, this research highlights two important points which was not considered in traditional supervisory feedback models. In traditional models, the positive and negative signs of the messages were the substance of supervisory feedback. However, this research suggests that perceptions of attributions have considerable effects on a poor performer's motivation through causing the feeling of shame. In this regard, attributions are a critical dimension characterizing supervisory feedback in workplaces. This research also suggests that shame plays a strong role in predicting poor performer's motivation. The feeling of shame is influenced by attributions that supervisors make for poor performance. Thus, it appears that workers may develop motivation effectively through the feeling of shame. These intermediate psychological processes were not considered in traditional cybernetic models. Another important point of this research is that shame is functional in motivating poor performers. Traditionally, shame has been viewed as dysfunctional in promoting motivation. This dysfunctional aspect of shame has constrained managers to use general supervisory feedback strategies decreasing the effectiveness of performance appraisal (Beer, 1982). However, the functional aspect of shame releases managers from the traditional rule, and lead managers to the use of specific feedback message strategies. 85 In this sense, the functional and useful for organizational However, this study also who feel shame put themselves helplessness is not as strong motivation. Therefore, if we aspect of shame is important managers. indicates that poor performers into helplessness, though as performance improving are to continue to increase our knowledge about the workings of emotions in the workplace, we must be aware of the dual effects of shame in supervisory feedback situations. APPENDICES REFERENCES Abramson, L. Y., Seligman, M. E. P., 8 Teasdale, J. P. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, g1, 49-74. Alexander, F. (1938). Remarks about the relation of inferiority feelings to guilt feelings. International Journal of Psychology, 12, 41-49. Anderson, C. A. (1983). Motivational and performance deficits in interpersonal settings: The effect of attributional style. Journal of Personality and Social Psychology, fig, 1136-1141. Anderson, C. A., Horowitz, L. M., 8 French, R. (1983). Attributional style of lonely and depressed people. Journal of Personality and Social Psychology, fig, 127-136. Annett, J. (1969). Feedback and human behavior: The effect of knowledge results, incentive and reinforcement on learning and performance. Baltimore: Penguin Books. Arnold, M. B. (Ed.). (1970). Feelings and emotion. New York: Academic Press. Arvey, R. D. (1972). Performance as a function of perceived effort-performance and performance-reward contingencies. nganizational Behavior and Human Performance, p, 423-433. Arvey, R. D., 8 Ivancevich, J. M. (1980). Punishment in organization: A review, propositions and research suggestion. Academy of Management Review, p, 123-132. Arvey, R. D., 8 Neil, C. W. (1972). Testing expectancy theory predictions using behaviorally based measures of motivational effort for engineers, Mimeograph, University of Tennessee, Knoxville. Ashford, S. J. (1986). Feedback-seeking in individual adaptation: A resource perspective. Academy of Management Journal, 22, 465-487. Ashford, S. J., 8 Cummings, L. L. (1983). Feedback as an individual resource: personal strategies of creating information. Organizational Behavior and Human Performance, pp, 370-398. Atkinson, J. W. (1957). Motivational determinants of risk- taking behavior. Psychology Review, g1, 359-372. Ausubel, D. P. (1955). Relationships between Shame and Guilt in socializing process. Psychology Review, pp, 378-390. 86 87 Bandura, A. (1977). Self-efficagy: Toyard a unifyingtheory of behavioral changg. New York: Wiley. Banks, W. C. (1976). The effects of perceived similarity upon the use of reward and punishment. Journal of Egperimental and Social Psychology, 13, 131-138. Baker, L. D., DiMarco, N., 8 Scott, W. E. (1975). Effects of supervisor's sex and level of authoritarianism on evaluation and reinforcement of blind and sighted workers. Journal of Applied Psychology, pg, 28-32. Barrow, J. C. (1976). Worker performance and task complexity as causal determinants of leader behavior style and flexibility. Journal of Applied Psychology, .21: 433-440. Basowitz, H., Persky, H., Korchin, S. J., 8 Grinker, R. R. (1955). Anxiety and Stress. New York: McGraw-Hill. Bateson, G., Jackson, D., Haley, J., 8 Weakland, J. (1965). Toward a theory of schizophrenia. Behavioral Science, p, 171-180. Beer, M. (1981). Performance appraisal: Dilemmas and possibilities. Organizational Dynamics, p, 24-36. Bellack, A. S., Hersen, M., 8 Turner, 8. M. (1978). Role -play tests for assessing social skills: Are they valid? Behavioral Therapy, 3, 448-461. Berkman, L. F. (1977). Social networks, host resistance, and mortality: A follow-up study of Alameda County residents. Unpublished Doctoral Dissertation, University of California, Berkeley. Bogart, D. H. (1980). Feedback, Feedforward, and feedwithin: Strategic information in systems. Behavioral Science, pp, 237-248. Borman, W. C. (1973). First line supervisor: Validation study. Minneapolis: Personnel Decisions, Inc. Bowers, K. (1973). Situationism in psychology: An analysis and critique. Psychological Review, pp, 307-336. Brenenstuhl, D. C. (1976). An empirical investigation of the leadership function as affected by leader style, interpersonal trust, and commitment to future interaction. Unpublished doctoral dissertation, Indiana University. Cambell, J. P., 8 Pritchard, R. D. (1976). Motivation theory in industrial and organizational psychology. In M. D. Dunnette (Ed.), Handbook of Industrial and Qgganizational Psychology (pp. 63-139). Chicago: Rand 88 McNally. Cascio, W. F., 8 Bernardin, H. J. (1981). Implications of PA litigation for personnel decisions. Personnel Psychology, g1, 211-226. Child, I. L., 8 Waterhouse, I. K. (1953). Frustration and the quality of performance: II. A theoretical statement. Psychological Review, pp, 127-139. Cofer, C. N., 8 Appley, M. A. (1964). Motivation: Theory and research. New York: Wiley. Cook, D. R. (1985). The shame instrument. Unpublished manuscript, University of Wisconsin, Menomonie, Wisconsin. Covington, M. V., 8 Beery, R. G. (1976). Self-worph and school learning. New York: Holt, Rinehart 8 Winston. Covington, M. V., 8 Omelich, C. L. (1979). Are causal attributions causal? A path analysis of the cognitive model of achievement motivation. Journal of Personality and Social Psychology, 31, 1487-1504. Cronwne, D. P., 8 Marlowe, D. (1967). The approval motive: Studies in evaluative dependence. New York: John Wiley 8 Sons, Inc. Cusella, L. P. (1980). The effects of feedback on intrinsic motivation. In D. Nimmo (Ed.), Commupication Yearbook 1 (pp. 367-387). New Brunswick, NJ: Transaction. Cusella, L. P. (1984). The effects of feedback, source, message and receiver characteristics on intrinsic motivation. CommunicationQuarterly, 33, 211-218. Cusella, L. P. (1987). Feedback, motivation, and performance. In F. M. Jablin, L. L. Putnam, K. H. Roberts, and L. W. Porter (Eds.), Handbook of Or anizatiopal Communication (pp. 624-678). Beverly ii its", on: Sage . Dachler, H. P., 8 Mobley, W. H. (1973). Construct validation of an instrumentality-expectancy-task-goal model of work motivation: Some theoretical boundary conditions. Journal of Applied Psychology, §§, 397-418. Deaux, K., 8 Farris, E. (1977). Attributing causes for one's own performance: The effects of sex, norms, and outcome. Journal of Research in Personality, 11, 59-72. Deci, E. L., 8 Ryan, R. M. (1980). The empirical exploration of intrinsic motivational processes. In L. Berkowitz (Ed.), Advances in Experimental Social 89 Psychology 1; (pp. 39-80). New York: Academic Press. DeNisi, A. 8., Randolph, W. A., 8 Blencoe, A. G. (1982). Level of source of feedback as a determinant of feedback effectiveness. Academy of Management Proceedingg, 175-179. Depue, R. A., 8 Monroe, 8. M. (1978). Learned Helplessness in the perspective of the de ressive disorders: Conceptual and definitional ssues. Journal of Abnormal Psychology, g1, 3-20. Dion, K. K., Berscheid, E., 8 Walster, E. (1972). What is beautiful is good. Journal of Personality and Social Psycholggy, 31, 285-290. Downs, C. W., Johnson, K. M., 8 Barge, J. K. (1984). Communication feedback and task performance in organizations: A review of the literature. Organizational_ Communication: Abstracts, Analysis, and Overview, 9. Beverly Hills, CA: Sage. Dunnette, M. D. (1976). Aptitudes, abilities, and skills. In M. D. Dunnette (Ed. ), Handbook of Industrial and Organizational Psychology (pp. 473- 520). New York: John Wiley and Sons. Dunnettee, M. D., Groner, D. M., Holtzman, J. 8., 8 Johnson, P. D. (1972). Job performance categories and rating scales for sales, technical, and administrative jobs. Minneapolis: Personnel Decisions, Inc. Duval, 8., 8 Wicklund, R. A. (1972). A theory of objective self awareness. New York: Academic Press. Dweck, C. S. (1975). The role of expectations and attributions in the alleviation of learned helplessness. Journal of Personality and Social Psychology, 3;, 674-685. Dweck, C. 8., 8 Goetz, T. E. (1978). Attribution and learned helplessness. In J. H. Harvey, W. Ickes, 8 R. F. Kidd (Eds. ), New Directions in Attribution Research 2 (pp. 157- -179). Hillsdale, New Jersey: Lawrence Erlbaum Association, Publishers. Endler, N. 8., Magnusson, D. (Eds.). (1976). .Interactional psychologyyand personality. New York: W1ley. Engel, G. L. (1968). A life setting conductive to illness: The giving-up-given-up complex. Animals of Internal Medicine, 22, 293-300. Eswara, H. S. (1972). Administration of reward and punishment in relation to ability, effort and performance. Journal of Social Psychology, g1, 139-140. 90 Etzioni, A. (1968). Organizational decisions and their interrelationships. In B. Indik and F. Berrien (Eds.), Peo le roups, and organizations. New York: Teachers CoIIege Press. Fisher, R. (1973). When workers are discharged: An overview. Monthly Labor Review, 33, 4-17. Folkes, V. S. (1978). Causal communication in the early stages of affiliative relationships. Un ublished doctoral dissertation, University of Cal1fornia, Los Angeles. Frank, R., 8 Karl, J. (1971). The Hawthorne experiments: First statistical interpretation. American Sociological Review, 13, 623-643. Freud, S. (1962). Three essays on sexuality. New York: Basic Books. Freud, S. (1963). On narcissism: An introduction. Standard Edition 11 (pp. 67-102). London: Hogarth Press. Friedlander, F. (1966). Motivations to work and organizational performance. Journal of Applied Psychology, g2, 143-152. Gavin, J. F. (1970). Ability, effort, and role perception as antecedents of job performance. Experimented Publication System, p, Ms. No. 190A, 1-26. Gibb, J. R., Smith, E. E., 8 Roberts, A. H. (1955). Effects of positive and negative feedback upon defensive behavior in small problem solving groups. American Psychologist, 12, 355. Giois, D. A., 8 Sims, H. P. (1986). Cognition-behavior connections: Attribution and verbal behavior in leader -subordinate interactions. nganizational Behavior and Human Decision Processes, 31, 197-229. Goffman, E. (1959). The presentation of self in everyday life. New York: Dowbleday, 1959. Goffman, E. (1969). Strategic interaction. Philadelphia: University of Pennsylvania press. Goffman, E. (1971). Relations in public. New York: Basic Books. Goodstadt, B. E., 8 Hjelle, L. A. (1973). Power to the powerless: Locus of control and the use of power. Journal of Personality and Social Psychology, 31, 190-196. Goodstadt, B. E., 8 Kipnis, D. (1970). Situational 91 influences on the use of power. Journal of Applied Psychology, g1, 201-207. Green, B. G., Fairhurst, G. T., 8 Liden, R. C. (1981). Control of ineffective performance: history and sequences. Unpublished manuscript, University of Cincinnati. Greene, W. A., Goldstein, 8., 8 Moss, A. J. (1972). Psychosocial aspects of sudden death: A preliminary report. Archives of Internal Medicine, 129, 725-731. Green, 8. G., 8 Liden, R. C. (1980). Contextual and attributional influences on control decisions. Journal of Applied Ppychology, 3;, 453-458. Green, S. G., 8 Mitchell, T. R. (1979). Attributional processes of leaders in leader-member interaction. Organizational Behavior and Human Performance, 23, 429-458. Greller, M. M., 8 Herold, P. M. (1975). Sources of feedback: A preliminary investigation. Qgganizational Behavior and Human Performance, 13, 244-256. Hackman, J. R. (1976). Group influences on individuals. In M. D. Dunnette (Ed.), Handbook of Indpstrial and Organizational Psychology. ChiEago: Rand McNaIly. Hamond, K. R., 8 Summers, D. A. (1972). Cognitive control. Psychological Review, 12, 58-67. Hays, W. L. (1981). Stapistics (Third Ed.). New York: Holt, Rinehart and Winston, Inc. Heckhausen, H. (1982). Achievement motivation and its constructs: A cognitive model. Motivation and Emotion, 2;, 283-3290 Hermans, H. J. M. (1970). A questionnaire measure of achievement motivation. Journal of Applied Psychology, 2, 353-3630 Herzber, F., Mausner, B., 8 Snyderman, B. B. (1959). The motivation to work. New York: Wiley. Hirokawa, R. Y., 8 Miyahara, A. (1986). A comparison of influence strategies utilized by managers in American and Japanese organizations. Communication Quarterly, 31, 250-265. Horowitz, M. J. (1976). Stress response syndromes. New York: Jason Aronson. Hunter, J. S. (1980). Factor analysis. In P. R. Monge and J. N. Cappella (Eds.), Multivariate analysis techniques in human communication research (pp. 229-258). New 92 York: Academic Press. Hunter, J. 8., 8 Gerbing, D. N. (1982). Unidimensional measurement, second order analysis, and causal models. Research in Organizational Behavior, 4, 267-320. Hunter, J. 8., 8 Lim, T. 8. (1987). LIMSTAT. Unpublished manuscript, Department of Communication, Michigan State University, East Lansing, Michigan. Ilgen, D. R., 8 Feldman, J. M. (1983). Performance appraisal: A process focus. In L. L. Cummings and B. M. Staw (Eds.), Research in Organizational Behavior (pp. 141-197). Greenwich, Conn.: JAI Press. Ilgen, D. R., Fisher, D. D., 8 Taylor, M. 8. (1979). Consequences of individual feedback on behavior in organ1zations. Journal of Applied Psychology, 31, 441-456. Instone, D., Major, B., 8 Bunker, B. (1983). Gender, self confidence strategies: An organizational stimulation. Journal of Personality and Social Psychology, 11, 322-333. Isen, A. M., Shalker, T. E., Clark, M., 8 Karp. L. (1978). Affect, accessibility of material in memory and behavior: A cognitive loop? Journal of Personality and Social Psychology, 33, 1-12. Jacobson, E. (1964). The self and the object world. New York: International University Press. Janis, I. L. (1958). Psychological Stress. New York: Wiley. Janis, I. L., 8 Leventhal, H. (1968). Human reactions to stress. In E. Borgotta 8 W. Lambert (Eds.), Handbook of persopality theory and research (pp. 1041-1068). Chicago, III.: Rand McNally. Janis, I. L., 8 Mann, L. (1977). Decision Making. New York: The Free Press. Jones, E. E., 8 Nisbett, R. E. (1972). The actor and the observer: Divergent perceptions of the causes of behavior. In E.. e. Jones, D. E. Kanouse, H. H. Kelly, R. E. Nisbett, 8. Valins, 8 B. Weiner (Eds.), Attribution: Pergeiving the causes of behavior. Morristown, New Jersey: General Learning Press. Joreskog, K. G., 8 Sorbom, D. (1985). Lisrel VI: Analysis of linear structural relationships by maximum likelihood an§_least sgugres methods. Chicago: International Education Services. Kane, J. S. (1986). Performance distribution assessment. 93 In Berk, R. A. (Ed.), Performance_ass§ssment: Methods and applications (pp. 237-273). The Johns Hopkins University press. Kantowitz, B. H. (1974). Human information processingg Tutorials of achievement behavior. Hillsdate, N. J.: Erlbaum. Kaplan, R. M., 8 Swant, 8. G. (1973). Reward characteristics of appraisal of achievement behavior. Reprgsentative Research in Social Psychology, 1, 11-17 0 Katz, K. (1975). Feedback in social system: Operational and systemic research on production, maintenance, control, and adaptive function. In C. A. Bennett 8 A. A. Lumsdaine (Eds.), Evaluation and Experiment: Sopg griticalissues in assessing social program. NY: Academic Press. Katz, D., 8 Kahn, R. L. (1978). The social psychology of organizations. New York: Wiley. Kaufman, G. (1974). The meaning of shame: Towards a self- affirming identity. Journal of Abnormal Psychology, pg, 568-574. Kaufman, G. (1985). Shame: The power of caring. Cambridge, r MA: Schenkman Books, Inc. Keppel, G. (1982). Design and analysis: A research handbook (Second Ed.). New Jersey: Prentice-Hall, Inc. Kim, Y. Y., 8 Miller, K. I. (in press). The effects of attribution and feedback goals on the generation of supervisory feedback message strategies. Management Communicat on Quarterly. Kipnis, D. (1972). Does power corrupt? Journal of Personality and Social Psychology, pg, 33-41. Kipnis, D. (1976). Powerholders. Chicago: University of Chicago Press. Kipnis, D., Schmidt, 8. M., 8 Wilkinson, I. (1980). Intraorganizational influence tactics: Explorations in getting one's way. Journal of Applied Psychology, gg, 440-452. Knowlton, W. A., 8 Mitchell, T. R. (1980). Effects of causal attributions on a supervisor's evaluation of subordinate performance. Journal of Applied Psychology, 55, 459-466. Krohne, H. W., 8 Laux, L. (Eds.). (1982). Achievement, stress, and anxiety. New York: Hemisphere. 94 Run, A., 8 Weiner, B. (1973). Necessary versus sufficient causal schemata for success and failure. Journal of Research in Personality, 1, 197-207. Landy, F. J., 8 Guion, R. M. (1970). Development of scales for the measurement of work motivation. Organizational Behavior and Human Performance, 5, 93-103. Lanzetta, J. T., 8 Hannah, T. E. (1969). Reinforcing behavior of "naive" trainers. gpurnal of Personality and Social Psychology, 11, 245-252. Larson, J. R. (1984). The performance feedback process: A preliminary model. Organizational Behavior and Human Performance, 11, 42-76. Latham, G. P., Cummings, L. L., 8 Mitchell, T. R. (1981). Behavioral strategies to improve productivity. Organizational Dynamics, 1, 5-23. Latham, G. P., 8 Wexley, K. N. (1982). Increasing productivity through pggformance appraisal. Massachusetts: Addison-Wesley. Lawler, E. E. III. (1968). Equity theory as a predictor of productivity and work quality. Psychological Bulletin, lg, 596-610. Lawler, E. E. III. (1973). Motivation in work organization. Belmont, CA: Brooks/Cole. Lawler, E. E. III, 8 Porter, L. W. (1967). Antecedent attitudes of effective managerial performance. Organizational Behavior and Human Performance, 1, 122-142. Lawler, E. E. III, 8 Suttle, J. L. (1973). Expectancy theory and job behavior. Organizational Behavior and Human Performance, 3, 482-503. Lazarus, R. 8., 8 Averill, J. R. (1972). Emotion and co nition: With special reference to anxiety. In C. D. Sp elberger (Ed.), Anxiety: Current trends 1n theory and research (Vol. 2). New York: Academic Press. Lazarus, R. 8., 8 Folkman, S. (1984). Stress, appraisal and coping. New York: Springer Publishing Company. Lefcourt, H. M. (1982). Locus of contro1: Current trends in trends in_theory and research_(2nd Ed.). London: Lawrence Erlbaum Associates, Publishers. Lewin, K. (1938). The conceptual representation and the measupement of psychological forces. Durham, N. C.: Duke University Press. Lewis, H. B. (1971). Shame and guilt in neurosis, New York: 95 International University Press. Liden, R. C. (1981). Contextual and behavioral factors influencing perceptions of ineffective performance and managerial responses. Unpublished doctoral dissertation, University of Cincinnati. Locher, A. H., 8 Teel, K. S. (1977). Performance appraisal -- A survey of current practices. Personnel Journal, 11, 254-247. Lynd, H. M. (1958). On shame_gnd the search for identity. New York: Bruner/Mazel, Inc. Mandler, G. (1975). Mind and emotion. New York: Wiley. Martinko, M. J., 8 Gardner, W. L. (1987). The leadership/ member attribution process. Academy of Management Review, 11, 235-249. Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 11, 370-396. McAdams, T. (1978). Dismissal: A decline in employer autonomy? Business horizons, 11, 67-72. McCelland, D. C., Atkinson, J. W., Clark, R. A., 8 Lowell, E. L. (1953). gpe achievement motive. New York: Appleton-Centry-Crofts. McDermott, T., 8 Newhams, T. (1971). Discharge -reinstatement: What happens thereafter. Industrial and Labor Relations Review, 11, 526-540. McFarland, C., 8 Ross, M. (1982). Impact of causal attributions on affective reactions to success and failure. Journal of Personality and Social Psychology, 2, 937-9460 McFillen, J. M. (1978). Supervisory power as an influence in supervisor-subordinate relations. Academy of Management Journal, 11, 419-433. McFillen, J. M., 8 New, J. R. (1979). Situationa1 determinants of supervisor attributions and behavior. Academy of Management Journal, 11, 793-809. McGregor, D. M. (1960). The human side of enterprise. New York: McGraw-Hill. Mead, M. (Ed.). (1937). Cooperation and com etition among primitive people. New York: McGraw- il Miller, I. W., Klee, S. H., 8 Norman, W. H. (1980). Depressed and nondepressed inpatients' cognitions of h othetical event, experimental tasks, and stressful l fe event. Journal of Abnormal Psychology, 11, 96 94-97. Miller, D. T., 8 Ross, M. (1975). Self-serving biases in attribution of causality: Facts or fiction? Psychological Bulletin, 82, 213-225. Miner, J. B. (1963). gpe management of ineffective performance. New York: McGraw-Hill. Miner, J., 8 Brewer, F. (1976). The management of ineffective performance. In M. Dunnette (Ed.), Handbook of Indusgrial and_Organizational Psychology (pp. 995-1030). Chicago: Ran McNally. Mirman, C. (1984). Shame and Guilt: Activators, associated unconscious dangers, and defensgg. Unpublished doctoral dissertation, Michigan State University, East Lansing, Michigan. Mitchell, T. R., 8 Albright, D. W. (1972). Expectancy theory predictions of the satisfaction, effort, performance and retention of naval aviation officers. Organizational Behavior and Human Performance, 1, 1-20. Mitchell, T. R., 8 Green, S. G. (1978). Leader responses to poor performance: An attributional analysis. Paper presented at the annual meeting of the American Psychological Association, Toronto. Mitchell, T. R., Green, S. G., 8 Wood, R. E. (1981). An attributional model of leadership and the poor performing subordinate: Development and validation. In L. L. Cummings and B. W. Staw (Eds.), Researgh in Organizational Behavior 3. Greenwich, Conn.: JAI Press. Mitchell, T. B., 8 Nebeker, D. M. (1973). Expectancy theory predictions of academic effort and performance. Journal of Applied Psychology, 11, 61-67. Mitchell, T. B., 8 Wood, R. E. (1980). Supervisor's responses to subordinate poor performance: A test of an attributional model. nganizational Behavior and Human Performance, 11, 123-138. Motowildo, S. J., Loehr, V., 8 Dunnette, M. D. (1972). The effect of goal specificity on the relationship between expectancy and task performance. Menneapolis: University of Minnesota, Technical Report No. 4008. Nadler, D. A. (1979). The effects of feedback on task group behavior: An review of the experimental research. Organizational Behavior and Human Performance, 11, 309-338. Nemeroff, W. F., 8 Consentino, J. (1979). Utilizing feedback and goal setting to increase performance 97 appraisal interview skills of managers. Academy of Management Journal, 11, 566-576. Nicholle, J. G. (1976). Effort is virtuous, but it's better to have ability. Evaluative responses to perceptions of effort and ability. Journal of Research in Personality, 12, 306-315. Novak, W. (1986). The measurement and construct pf shame. Unpublished doctoral dissertation, Michigan Stite University, East Lansing, Michigan. Nowicki, 8., 8 Strickland, B. (1973). A locus of control scale for children. gournal of Consulting and Clinical Odiorne, G. (1971). Personnel Administration by Objectives. Homewood, IL: Irwin. O'Reilly, C. A., 8 Weitz, B. A. (1980). Managing the mar inal employee: The use of warnings and dismissals. Administrative Science Quarterly, 11, 467-483. Overmeier, J. B., 8 Seligman, M. E. P. (1967). Effects of inescapable shock upon subsequent escape and avoidance responding. Journal of Comparative and Psychological Psychology, 11, 28-33. Payne, R. B., 8 Hauty, G. T. (1955). Effects of psychological feedback upon work development. Journal of Experimental Psychology, 11, 342-351. Pearce, J. L. (1983). Job attitude and motivation differences between volunteers and employees from comparable organizations. Journal of Applied Psychology, 11, 646-652. Pearlin, L. I. (1980). Life strains and psychological distress among adults. In N. J. Smelser 8 E. H. Erison (Eds.), Thepes of work and love in adulthood. Cambridge, CA: Harvard University Press. Pervin, L. A., 8 Lewis, M. (Eds.). (1978). Perspectives in interactional psychology. New York: Plenum. Peterson, C., 8 Seligman, M. E. (1984). Causal explanations as a risk factor for depression: Theory and evidence. Psychological Review, 11, 347-374. Piers, G., 8 Singer, M. B. (1953). Shame and guilt. Springfield, IL: Thomas. Piliavin, I. M., Rodin, J., 8 Piliavin, J. A. (1969). Good samaritanism: An underground phenomenon. Journal of personality and Social Psychology, 11, 289-299. Podsakoff, P. M. (1982). Determinants of a supervisor's use of rewards and punishments: A literature review and suggestions for further research. Organizational Behavior and Human Performance, 11, 58-83. Porter, L. W., Lawler, E. E., 8 Hackman, J. R. (1975). Behavior in Organizations. New York: McGraw-Hill. Pritchard, R. D., 8 DeLeo, P. J. (1973). Experimental test of the valence-instrumentality relationship in job performance. Journal of Applied Psychology, 11, 264-270. Randy, F. J., 8 Guion, R. M. (1970). Development of scale for the measurement of work motivation. Organizational Behavior and Human Performance, 1, 93-103. Rest, 8., Nierenberg, R., Weiner, B., 8 Heckhausen, H. (1973). Further evidence concerning the effects of perceptions of effort and abilit on achievement evaluation. Journal of Personality and Social Psychology, 11, 187-191. Richter, C. P. (1957). On the phenomenon of sudden death in animal and man. Psychosomatic Medicine, 11, 191-198. Roth, 8., 8 Bootzin, R. R. (1974). The effects of experimentally induced expectancies of external control: An investigation of learned helplessness. gpurnal of Personality and Social Psychology, 11, 253-264. Roth, 8., 8 Kubal, L. (1975). The effects of noncontingent reinforcement on tasks of differing importance: Facilitation and learned helplessness. Journal of Personality and Social Psychology, 1, 680-691. Rothbart, M. (1968). Effects of motivation, equity, and compliance on the use of reward and punishment. Journal Personality and Social Psychology, 1, 335-362. Ruckmick, C. A. (1936). The s chology of feeling and emotion. McGraw-HilI, New York} Russell, J. A. (1982). The causal dimension scale: A measure of how individuals perceived causes. Journal of Personality and Social Psychology, 11, 1137-1145. Sarason, I. G. (1972). Experimental approaches to test anxiety: Attention and the uses of information. In C. D. Speilberger (Ed.), Anxiety: Curgent trends in theory and research (Vol. 2). New York: Academic Press. Sarason, I. G., Mandler, G., 8 Craighill, P. C. (1952). The effect of differential instructions on anxiety and learning. Journal of Abnormal and Social Psychology, 11, 561-565. 99 Schachter, J. (1959). The _psycholo of affiliation. Standford University Press, Standford, California. Schachter, S. (1966). The interaction of cognitive and psychological determinants of emotional state. In C. D. Spielberger (Ed. ), Anxiety and behavior. New York: Academic Press. Schmale, A. H. (1971). Giving up as a final common pathway to changes in health. Psychomatic Medicine, 1, 18- 28. Schuler, R. S. (1981). Appraising employee performance. In R. 8. Schuler (Ed. ), Personnel and Human Resource Management (pp. 219- 256). Minnesota: West Publishing Co. Schuter, J. R., Clark, B., 8 Rogers, M. (1971). Testing portions of the Porter and Lawler model regarding the motivational role of pay. Journal oprplied Psychology, 11, 187-195. Seiler, J. A. (1967). S stems anal sis in organizational behavior. Homewood, III: Irwin Dorsey. Seligman, M. E. P. (1975). Helplessness. San Francisco: W. H. Freeman. Seligman, M. E. P., 8 Maier, S. F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology, 11, 1-9. Steer, R. M., 8 Porters, L. M. (1987). Motivation and work behavior. McGraw-Hill Book Company. Stoeberl, P. A., 8 Schiederjan, M. J. (1981). The ineffective subordinate: A management survey. Personnel Administrator, February, 72-76. Storms, M. D., 8 McCaul, K. D. (1976). Attribution processes and emotional exacerbation of dysfunctional behavior. In Harvey, J. H., Ickes, W. J., 8 Kidd, R. F. (Eds.), New Directipns in Aptributiop Research (vol. 1, pp. 143-164). New Jersey: Lawrence Erlbaum Association Publisher. Stressin, L. (1960). Em lo ee disciplines. Washington, D. C.: Bureau of NationaI Affairs. Super, D. E., 8 Crites, J. O. (1962). 1ppraisipg vocational fitness. New York: Harper and Row. Tedeschi, J. T., Malkis, F., 8 Fernandez, B. (1980). First impressions and reactions to threats. Human Relations, 11, 647-657. Tolman, E. C. (1932). Purposive behavior in animals and men. New York: Century. 100 Tomkins, S. S. (1963). Affect--Ima ery--Consciousness 2. New York: Springer and Co. Vallacher, R. R. (1980). An introduction to self-theorp. In D. M. Wegner 8 R. R. Vallacher (Ed.), The sglf n social psychology. Oxford: Oxford University Press. Vroom, V. H. (1964). Work and motivation. New York: John Wiley 8 Sons. Weiner, B. (Ed.). (1974). Cognitive views of human emotion. New York: Academic Press. Weiner, B. (1976). An attributional model for educational sychology. In L. Shulman (Ed.), Review of Research n Education 4. Itasca, III: Peacock. Weiner, B. (1979). A theory of motivation for some classroom experience. Journal of Educational Psychology, 11, 3-25. Weiner, B. (1980). A cognitive (attribution)-emotion-action model of motivated behavior: An anal sis of judgments of help-giving. Journal of Personal ty and Soc a1 Psychology, 11, 182-200. Weiner, B., Frieze, I., Kukla, A., Reed, 8., 8 Rosenbaum, R. (1972). Perceiving the causes of success and failure. In E. Jones, D. Kanouse, H. Kelly, R, Nisbett, S. Valims, 8 B. Weiner (Eds.), Attributign:gerceiving the pauses of_behavior. Morris-town, N. J.: General Learning Press. Weiner, B., Graham, 8., 8 Chandler, C. (1982). Pity, anger, and guilt: An attributional analysis. Personality and Social Psychological Bulletin, 1, 226-232. Weiner, B., 8 Kukla, A. (1970). An attributional analysis of achievement motivation. Journal of Personality and Social Psychology, 11, 1-20. Weiner, B., 8 Peter, N. (1973). A cognitive-developmental analysis of achievement and moral judgments. Developmental Psychology, 9, 290-309. Weiner, B., Russell, D., 8 Lerman, D. (1978). Affective consequences of causal ascriptions. In J. H. Harvey, W. Ickes, 8 R. F. Kidd (Eds.), New Diregtiops in Attribution Research 1. Hillsdale, N. J.: Erlbaum. Weiner, B., Russell, D., 8 Lerman, D. (1979). The cognition-emotion process in achievement-related contexts. Journal of Personality and Social Psychology, 11, 1211-1220. Wheaton, B., Muthen, B., Alwin, D. R., 8 Summers, A. F. 101 (1977). Assessing reliability and stability in panel models. In D. R. Heise (Ed.), Sociological Methodology 1977. San Franscisco: Jossey-Bass. Wheeler, H. (1976). Punishment theory and industrial discipline. Industrial Relations, 11, 235-243. White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Bulletin, 11, 297-333. Wiener, N. (1948). Cybernetics: On control and commupication in the animal and the machine. New York: John Wiley. Wiener, N. (1954). The human use of human beinggi Cybernetics and society. Garden City, NY: Doubleday Andhor. Will, 0. A. Jr. (1987). The sense of shame in psychosis: Random comments on shame in the psychotic experience. In D. L. Nathanson (Ed.), 1he many faces of shame (pp. 309-317). New York: The Guilford Press. Wiseman, R. C., 8 Schenck-Hamlin, W. (1981). A multidimensional scaling validation of an inductively -derived set of compliance-gaining strategies. Communication Monograph, 11, 251-270. Wollenberger, J. B. (1963). Acceptable work rules and penalties: A company guide. Personnel, 12, 23-29. Wortman, C. B., 8 Brehm, J. W. (1975). Responses to uncontrollable outcome: An integration of reactance theory and the learned helplessness model. In L. Berkowitz (Ed.), Advances_1n Exper1mental Social Psychology 8. New York: Academic Press. Wortman, C. B., 8 Dintzer, L. (1978). Is an attributional anal sis of the learned helplessness phenomenon viable? A critique of the Abramson-Seligman-Teasdale reformulation. Journal of Abnormal Psychology, 11, 75-90. Wright, J., 8 Mischel, W. (1982). Influence of affect on cognitive social learning person variables. Journal of Personality and Social Psychology, 11, 901-914. APPENDIX A Pretest questionnaire 102 Dear Respondents: Thank you for your participation in this research. I would like to request your help in filling out several questions. Please follow the instructions and respond to the questions. These questions are designed to determine how you perceive a poor performance situation and a supervisory feedback message. Your response will be held in complete confidence by the researcher at Michigan State Univers ty. I appreciate your cooperation and wish you success in your career. If you have questions, please contacting Young Y. Kim at (517) 355-5954. 103 A situation which might occur in your hospital is presented below. Please imagine that you are confronted with this situation in your role as a staff nurse. SITUATION: As a staff nurse, you are assigned to a four-year-old fresh post-op patient. One of the post-op orders was to begin intravenous antibiotic therapy immediately (first dose stat). The administration of the stat dose, however, was delayed, allowing the child's infection to worsen. The attending physician and the child's family were extremely upset. PART ONE: Thank you for your careful reading of the situation. Please think about how you feel about the situation and circle the response which best indicates how much you agree with each of the following statements. SD = Strongly Disagree D = Disagree N = Neither Agree Nor Disagree A = Agree SA = Strongly Agree 1. The situation above is associated with a staff nurse's job performance . . . . SD D N A SA 2. The situation above can actually occur to staff nurses in hospitals . . . . . SD D N A SA 3. I have seen nurses (or a nurse) who were (or was) involved in a similar situation . . . . . . . . . . . . . . . SD D N A SA 4. The situation above is a significant event O O O O O O O O O O O O O O O O 0 SD D N A SA 5. The situation is one of the frequent incidents that may occur in hospitals . SD D N A SA You and your supervisor met to discuss the incident. In your discussion, you learned that your supervisor was concerned about two issues. First, she was concerned about whether the delay occurred because you weren't putting much effort into your job, and might have been socializing with coworkers. Your supervisor was also concerned about whether the delay occurred because you lacked the ability to do your job, and were not familiar with intravenous antibiotic therapy. 104 Your supervisor talks with you to learn more about why the error occurs, then provides you with the following feedback: I've enrolled you in a formal training program to help you deal with this problem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for the problem. Let me tell you one more thing. You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems. PART TWO: In this part, I would like you to think about what your supervisor's comments to you (i.e., supervisory feedback above) indicate. Then, please circle the response which best indicates how much you agree with each of the following statements. 6. My supervisor is very opposed to my spending more time socializing than caring for the patient . . . . . . . . SD D N A SA 7. My supervisor is trying to penalize me formally because I spent more time socializing than caring for the Patient O O O O O O O O O O O O O O O O SD D N A SA 8. My supervisor is imposing me a penalty resulting in very harmful consequences in the future because I spent more time socializing than caring for the patient . . . . . . . . . . . . . . . . SD D N A SA 9. My supervisor doubts about my com etence in the intravenous antibiotic therapy . . . . . . . . . . SD D N A SA 10. My supervisor thinks that I should have formal training about the intravenous antibiotic therapy . . . . . . . . . . SD D N A SA 11. My supervisor thinks that I need to be tau ht more about the intravenous antibiotic therapy . . . . . . . . . . SD D N A SA 105 (The supervisory feedback message was replaced with the following eight different messages). "I've enrolled you in a formal training program to help you deal with this problem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for this program. Let me tell you one more thing. You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." "I've enrolled you in a formal training program to help you deal with this roblem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for the program. Let me tell you one more thing. I hope you administered the stat order in time. But I will be keeping an eye on you." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also hel you. Let me tell you one more thing. You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also help you. Let me tell you one more thing. You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also help you. Let me tell you one more thing. I hope you administered the stat order in time. But I will be keeping an eye on you." "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. Let me tell you one more thing. You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems." 106 "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. Let me tell you one more thing. You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. I also hope you administered the stat order in time. But I will be keeping an eye on you." APPENDIX B Regular questionnaire 107 Dear Respondents: Thank you for your participation in this research. When developing organizational management system, it is very important to understand the process of providing feedback to employees who have had difficulties with their jobs. This research is concerned with this issue. For the purpose of this research, therefore, a hypothetical situation of failure on the job will be presented to you. We hope you do not feel uncomfortable when you read this situation. The situation has been designed to study the issues of feedback and job performance and requires the consideration of a job failure situation. Your help in responding to questions about how you would respond if you were in this Situation will contribute to our knowledge about feedback and hospital management. Your response will be held in complete confidence by the researcher at Michigan State University. If you have questions, please contact Young Y. Kim at (517) 355-5954. Thank you very much for your assistance with this research. Sincerely, Young Y. Kim Doctoral Candidate Department of Communication Michigan State University 108 A situation which might occur in your hospital is presented below. Please imagine you are confronted with this situation in ygur role as a staff nurse. Please read the situation carefully and think about what your supervisor says to you. SITUATION: As a staff nurse, you are assigned to a four-year-old fresh post-op patient. One of the post-op orders was to begin intravenous antibiotic therapy immediately (first dose stat). The administration of the stat dose, however, was delayed, allowing the child's infection to worsen. The attending physician and the child's family were extremely upset. You and your supervisor met to discuss this incident. In your discussion, you learned that your supervisor was concerned about two issues. First, she was concerned about whether the delay occurred because you weren't putting much effort into your job, and might have been socializing with coworkers. Your supervisor was also concerned about whether the delay occurred because you lacked the ability to do your job and were not familiar with the intravenous antibiotic therapy. Your supervisor talks with you to learn more about why the error occurs, then provides you with the following feedback: "I've enrolled you in a formal training program to help you deal with this roblem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for the training program. Let me tell you one more thing. You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems." 109 (The supervisory feedback message was replaced with the following eight different messages) "I've enrolled you in a formal training program to help you deal with this problem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for this program. Let me tell you one more thing. You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." "I've enrolled you in a formal training program to help you deal with this problem. I think you should learn more about intravenous antibiotic therapy. Go to the personnel manager to register for the program. Let me tell you one more thing. I hope you administered the stat order in time. But I will be keeping an eye on you." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also hel you. Let me tell you one more thing. You were unprofessional and irresponsible because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also hel you. Let me tell you one more thing. You were unprofessional and irresponsible in caring for the patient. You should make more effort about patient care and spend less time socializing." "When you have trouble with intravenous antibiotic therapy, you should ask for some help from pharmacy and your coworkers. I can also help you. Let me tell you one more thing. I hope you administered the stat order in time. But I will be keeping an eye on you." 110 "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. Let me tell you one more thing. You were unprofessional and irrespons ble because you spent more time socializing than caring for the patient. This incident has been filed as a formal written warning because of your unprofessional and irresponsible attitude. Remember you will be suspended if you have another formal warning because of these attitude problems." "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. Let me tell ou one more thing. You were unprofessional and irrespon31ble in caring for the patient. You should make more effort about patient care and spend less time socializing." "I hope you had no problem with intravenous antibiotic therapy. But it's not a bad idea to check again on the procedure you used. I also hope you administered the stat order in time. But I will be keeping an eye on you." PART ONE: 111 above from your superV1sor. supervisor) think of yourself and please circle the response which best indicates how much you agree with each of the following statements. 10. SD = Strongly Disagree D = Disagree N = Neither Disagree Nor Agree A = Agree SA = Strongly Agree I think I am well educated about intravenous antibiotic therapy . I think I have good recall about intravenous antibiotic therapy procedures . I think I clearly understand intravenous antibiotic therapy . I think I am quick in performing intravenous antibiotic therapy . I think I am reliable at intravenous antibiotic therapy . I think it is eas for me to perform intravenous antibiotic therapy . I think I put priority on intravenous antibiotic therapy for the child in the incident over other patients . I think I kept my mind on beginning intravenous antibiotic therapy immediately . I think I was distracted by things other than beginning the intravenous antibiotic therapy immediately . I think I was interested in performing intravenous antibiotic therapy for the Child . Thank you for your careful reading of the situation and your superV1sor's comments to you. imagine that you have just received the feedback message Think about how you (not your SD SD SD SD SD SD SD SD SD SD SA SA SA SA SA SA SA SA SA SA 112 11. I think I was attentive to performing intravenous antibiotic therapy immediately for the child . . . . . . . SD D N A SA 12. I think I was eager to perform intravenous antibiotic therapy immediately for the child . . . . . . . SD D N A SA PART TWO: In this part, I would like you to think about how your supervisor (not you) thinks of you. Then, please circle the response which best indicates how much you agree with each of the following statements. 13. My supervisor thinks that I am well educated about intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 14. My supervisor thinks that I have good recall about intravenous antibiotic therapy procedures . . . . . . . . . . SD D N A SA 15. My supervisor thinks that I clearly understand intravenous antibiotic therapy O O O O O O O O O O O O O O O O SD D N A SA 16. My supervisor thinks that I am quick in performing intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 17. My supervisor thinks that I am reliable at intravenous antibiotic therapy . . . SD D N A SA 18. My supervisor thinks that it is easy for me to perform intravenous antibiotic therapy O O O O O O O O O O O O O O O O SD D N A SA 19. My supervisor thinks that I put priority on intravenous antibiotic therapy for the child in the incident over other Patients O O O O O O O O O O O O O O O SD D N A SA 20. My supervisor thinks that I kept my mind on beginning intravenous antibiotic therapy immediately . . . . . . . . . . SD D N A SA 21. _22. 23. 24. PART THREE: as to how your supervisor thinks of you. questions, I would like you to think about how you would feel if this incident had happened to you. the response which best indicates how much you agree with 113 My supervisor thinks that I was d stracted by things other than beginning intravenous antibiotic therapy immediately . . . . . . . . . . SD My supervisor thinks that I was interested in performing intravenous antibiotic therapy for the child . . . . . . . . . 8D My supervisor thinks that I was attentive to performing intravenous antibiotic therapy immediately for the child . . . SD My supervisor thinks that I was eager to erform intravenous antibiotic therapy immediately for the child . . . . . . . SD each of the following statements. In part two above, you responded to questions In the next set of Please circle 8A SA SA SA 21. 22. 23. 24. PART THREE: as to how your supervisor thinks of you. questions, I would like you to think about how you would feel if this incident had happened to you. the response which best indicates how much you agree with 113 M supervisor thinks that I was distracted by things other than beginning intravenous antibiotic therapy immediately . . . . . . . . . . SD My supervisor thinks that I was interested in performing intravenous antibiotic therapy for the child . . . . . . . . . SD My supervisor thinks that I was attentive to performing intravenous antibiotic therapy immediately for the child . . . SD My supervisor thinks that I was eager to erform intravenous antibiotic therapy immediately for the child . . . . . . . SD each of the following statements. 25. 26. 27. 28. 29. 30. 31. 32. I would wish I could go back in time and avoid that event . . . . . . . . . SD I would see myself as not being able to measure up to other people . . . . . . SD I would think my supervisor looks down on me O O O O O O O O O O O O O O O O O SD I would feel insecure about what my supervisor thought of me . . . . . . . SD I would feel empty and unfulfilled . . SD I would feel full of self doubt . . . . SD I would say to myself, "how could my supervisor really care about me?" . . . SD I would be hard for me to get close to my supervisor . . . . . . . . . . . . . SD In part two above, you responded to questions In the next set of Please circle SA SA SA SA SA SA SA SA SA SA SA SA 114 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 114 I would feel like I could sink into the ground . . . . . . . . . . . . . . It would be difficult for me to accept a compliment from others . . . . . I would fear that my faults could be revealed in front of others . . . . . . It would be hard for me to maintain eye contact with my supervisor . . . . . . I would feel mocked and laughed at by my supervisor . . . . . . . . . . . I would feel like melting away . . . . I would feel somehow left out . . . . . I would feel miserable . . . . . . . . I would feel I am not important . . . I would feel like shrinking away . . . I would feel I couldn't concentrate on what I have to do next---the next step O O O O O O O O O O O O O O O I would find it hard to maintain self confidence . . . . . . . . . . . . I would feel to unsure about what to do . . . . . . . . . . . . . . I would try to forget the whole thing O O O O O O O O O O O O O O O O O I would avoid being with my supervisor . . . . . . . . . . . . . . I would prepare myself for the worst . . . . . . . . . . . . . . . . . SD SD SD SD SD SD SD SD SD SD SD SD SD SD SD SD U U U U U 2 2 2 2 2 2 9 y 9 V 5 SA SA SA SA SA SA SA SA SA SA SA SA SA SA SA SA 115 PART FOUR: In this last part of this questionnaire, I would like to ask about how you would feel about your future performance of your job if this situation had actually occurred. Please circle the response which best indicates how much you agree with each of the following statements. 49. If my supervisor had given me this feedback, I would try to make higher demands on myself with regard to intravenous antibiotic therapy . . . SD D N A SA 50. If my supervisor had given me this feedback, I would try to set hi her standards for intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 51. If my supervisor had given me this feedback, I would feel a greater responsibility for intravenous antibiotic therapy . . . . . . . . . . SD D N A SA 52. If my supervisor had given me this feedback, I would inquire into my shortcomings with regard to intravenous antibiotic therapy . . . . . . . . . . SD D N A SA 53. If my supervisor had given me this feedback, I would be eager to perform intravenous antibiotic therapy better . SD D N A SA 54. If my supervisor had given me this feedback, I would feel challenged in performing intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 55. If my supervisor had given me this feedback, I would lose interest in intravenous antibiotic therapy . . . . SD D N A SA 56. If my supervisor had given me this feedback, I would be more attentive to intravenous antibiotic therapy . . . . SD D N A SA 57. If my supervisor had given me this feedback, I would do my best in performing intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 116 58. If my supervisor had given me this feedback, I would keep my mind more on intravenous antibiotic therapy . . . . SD D N A SA 59. If my supervisor had given me this feedback, I would become more aggressive in performing intravenous antibiotic therapy . . . . . . . . . . . . . . . . SD D N A SA 60. If my supervisor had given me this feedback, I would exert myself more for intravenous antibiotic therapy . . SD D N A SA Finally, a few questions about yourself. How long have you worked in your profession? years How long have you worked at this hospital? ears how long have you worked in your present job at this hospital? years Do you work part time or full time? Part-Time Full-Time Are you Male or Female? Male Female Do you have any additional comments? APPENDIX C Tables 117 Table 1: Descriptive Statistics for REALISM fi INS ATT M=21.24 21.87 21.81 21.63 FW SD: 2.44 1.30 1.83 1.92 N=17 15 16 48 M=21.24 21.72 20.08 21.29 CRIT SD: 1.48 .96 2.29 1.66 N=17 18 17 52 M=2l.87 20.47 21.63 21.33 REC SD=2.26 2.17 1.75 2.11 N=15 15 16 46 M=21.43 21.38 21.43 21.41 SD: 2.07 1.62 1.98 1.89 N=49 48 49 146 FT: Formal Training INST: Instruction ATT: Attention FW: Formal Written Warning CRIT: Criticism REC: Recognition 118 Table 2: Descriptive Statistics for LACK OF ABILITY PT INS’ ATT M=12.58 10.87 8.13 10.56 FW SD: .71 .74 .81 2.02 N=17 15 16 48 M=12.60 10.83 8.12 10.60 CRIT SD: .88 .99 .93 2.14 N=17 18 17 52 M=12.87 10.87 8.44 10.67 REC SD: .74 1.25 .89 2.08 N=15 15 16 46 M=12.76 10.85 8.22 10.61 SD: .78 .99 .87 2.07 N=49 48 49 146 FT: Formal Training INST: Instruction ATT: Attention FW: Formal Written Warning CRIT: Criticism REC: Recognition 119 Table 3: Descriptive Statistics for LACK OF EFFORT Fr fns Afi M=12.00 11.87 1.88 11.99 PW SD: .87 .99 .72 .85 N=17 15 16 48 M=10.00 9.61 9.76 9.79 CRIT SD: 1.17 1.50 1.20 1.29 N=17 18 17 52 M: 8.07 8.20 8.38 8.22 REC SD: .96 .94 .96 .94 N=15 15 16 46 M=10.10 9.88 10.00 9.99 SD: 1.88 1.89 1.73 1.83 N=49 48 49 146 FT: Formal Training INST: Instruction ATT: Attention FW: Formal Written Warning CRIT: Criticism REC: Recognition 120 Table 4: Items and factor loadings for PERCEPTION OF LACK OF ABILITY (PLA) Item Content Loadings Var 17 My supervisor thinks that I am well .70 educated about intravenous antibiotic therapy Var 18 My supervisor thinks that I have good .58 recall about intravenous antibiotic therapy procedures Var 19 My supervisor thinks that I clearly .69 understand intravenous antibiotic therapy Var 20 My supervisor thinks that I am quick .74 in performing intravenous antibiotic therapy Var 21 My supervisor thinks that I am reliable .73 at intravenous antibiotic therapy Var 22 My supervisor thinks that it is easy .64 for me to perform intravenous antibiotic therapy Standard Score Alpha for factor = .93 Table 5: 121 Items and factor loadings for PERCEPTION OF LACK OF EFFORT (PLE) Item Content Loadings Var Var Var Var Var Var 23 24 25 26 27 28 My supervisor thinks that I put priority on intravenous antibiotic therapy for the child in the incident over other patients My supervisor thinks that I kept my mind on beginning intravenous antibiotic therapy immediately M supervisor thinks that I was distracted by things other than beginning intravenous antibiotic therapy immediately My supervisor thinks that I was interested in performing intravenous antibiotic therapy for the child My supervisor thinks that I was attentive to performing intravenous antibiotic therapy immeidately for the child My supervisor thinks that I was eager to perform intravenous antibiotic therapy immediately for the child .65 .69 .60 .42 .72 .65 Standard Score Alpha for factor = .91 122 Table 6: Items and factor loadings for HELPLESSNESS (HL) Item Content Loadings Var 47 I would feel I couldn't concentrate .51 on what I have to do next--the next step Var 48 I would find it hard to maintain my .56 self confidence Var 49 I would feel to unsure about what to do .57 Var 50 I would try to forget the whole thing .52 Var 51 I would avoid being with my supervisor .50 Var 52 I would prepare myself for the worst .50 Standard Score Alpha for factor = .87 123 Table 9: Deviations of observed from expected correlations (internal consistency) for PLA 'Item 17 18 19 20 21 22 17 .oo 18 -.02 .oo 19 .01 .02 .oo 20 -003 .00 001 .00 21 —.03 .oo -.03 .07* .oo 22 .oo .oo .oo -.05 -.01 .oo Average correlation: .68 Standard error: .04 * No deviations greater than expected from sampling error (p <.05) 124 Table 10: Deviations of observed from expected correlations (internal consistency) for PLE Item 23 24 25 26 27 28 23 .00 24 -.02 .00 25 -.04 .05 .00 26 -003 .00 002 .00 27 .02 -.02 .00 .02 .00 28 .07 -.01 -.03 -.03 -.02 .00 Average correlation: Standard error: .04 .61 * No deviations greater than expected from sampling error (p <.05) 125 Table 11: Deviations of observed from expected correlations (internal consistency) of final version of HL Item 47 48 49 50 51 52 47 .00 48 .01 .00 49 .13* .02 .00 50 -007 -001 -001 .00 51 -.06 -.04 -.06 .11* .00 52 -.01 .03 -.07 -.01 .05 .00 —Average correlation: .53 Standard error: .05 *deviations greater than expected from sampling error (p <.05) 126 Table 7A: Items and factor loadings for SHAME (SH) Item Content Loadings Var 29 I would wish I could go back in .56 time and avoid that event Var 30* I would see myself as not being .29 able to measure up to other people Var 31 I would think my supervisor looks .42 down on me Var 32 I would feel insecure about what .44 what my supervisor thought of me Var 33* I would feel empty and unfulfilled .22 Var 34* I would feel full of self doubt .13 Var 35 I would say to myself, "how could .37 my supervisor really care about me Var 36* I would be hard for me to get close .28 Var 37 I would feel like I could sink into .48 the ground Var 38* I would be difficult for me to accept .22 a compliment from others Var 39 I would fear that my faults could be .48 revealed in front of others Var 40 I would be hard for me to maintain .46 eye contact with my supervisor 127 Table 7A (continued) Var Var Var Var Var Var 41 42 43* 44 45* 46* I would feel mocked and laughed at by my supervisor I I I I I would would would would would feel like melting away feel somehow left out feel miserable feel I am not important feel like shrinking away .44 .38 .35 .37 .22 .38 Standard Score Alpha for factor = .91 * Items removed from final version of Shame Table 78: Deviations of observed from expected correlations (internal consistency) for SHAME 128 Item 29 30 31 32 33 34 35 36 37 29 .00 30 .04 .00 31 .02 .01 .00 32 002 -005 011 .00 33 -003 01‘* -006 -002 .00 34 .05 -003 002 002 007 .00 35 -.02 -.05 .03 .02 -.11 -.05 .00 36 -.02 .02 -.01 -.09 .22* .06 -.04 .00 37 -.01 -.03 .00 .04 -.11 .00 .14* -.06 .00 38 .05 .02 -.02 -.06 .15* .09 -.09 .13* -.07 39 -.05 -.01 “.03 .02 -.07 -.06 .14* -.10 .13* 40 -008 .00 -003 -005 .00 -001 007 -005 002 41 .02 .04 “.03 .02 -.08 “.03 .07 -.11 .00 42 -001 -011 -.03 006 -005 .00 -002 -003 -002 43 .02 .06 -.04 “.08 .07 .00 -.16* .14* -.08 4‘ .05 -002 001 .00 -005 -008 002 -O09 005 45 .00 .10 .02 -.02 .03 -.07 -.05 .08 -.06 46 -.06 -.12* .06 .04 -.08 .01 .10 -.04 .06 Table 78 (continued) Item 38 39 40 41 42 43 44 45 46 38 .00 39 -.13* .00 40 -005 .06 .00 41 -.07 .04 .11 .00 42 .02 .06 -.02 .05 .00 43 .09 -.06 -.01 .05 .03 .00 44 .00 .01 .03 .01 .04 .01 .00 45 008 -004 001 -003 -007 004 -003 .00 44 -.12 .09 .00 -.05 .12 -.07 .06 .00 .00 Average correlation: .35 Standard error: * deviations greater than expected from (p <.05) .06 sampling error 129 Table 12: Deviations of observed from expected correlations (internal consistency) for final version of SHAME Item 29 31 32 35 37 39 40 41 42 44 29 .00 31 .05 .00 32 .03 .10 .00 35 -.04 .00 -.03 .00 37 -002 -001 .00 006 .00 39 -.06 -.05 -.03 .06 .06 .00 40 -.06 -.02 -.06 .03 -.01 .03 .00 41 .02 -.04 -.02 .01 -.05 -.02 .09 .00 42 .01 —.02 .04 -.06 -.05 -.05 .03 —.02 .00 44 .06 -.01 -.02 —.03 .01 -.03 .02 -.02 .03 .00 Average correlation: .46 Standard error: .06 Standard score alpha of final version of Shame = .90 No deviations greater than expected from sampling error (p <.05) 130 Table 8A: Items and factor loadings for PERFORMANCE IMPROVING MOTIVATION (PIM) Item Content Loadings Var Var Var Var Var Var Var Var 53 55 56 57 58 59* 60 If my supervisor had given me this feedback, I would try to make higher demands on myself with regard to intravenous antibiotic therapy If my supervisor had given me this feedback, I would try to set higher standards for intravenous antibiotic therapy If my supervisor had given me this feedback, I would feel a greater responsibility for intravenous antibiotic therapy If my supervisor had given me this feedback, I would inquire into my shortcomings with regard to intravenous antibiotic therapy If my supervisor had given me this feedback, I would be eager to perform intravenous antibiotic therapy better If my supervisor had given me this feedback, I would feel challenged intravenous antibiotic therapy If my supervisor had given me this feedback, I would lose interest in intravenous antibiotic therapy If my supervisor had given me this feedback, I would be more attentive to intravenous antibiotic therapy .36 .37 .43 .37 .49 .34 .02 .38 131 Table 8A (continued) Var 61 Var 62 Var 63 Var 64 If my supervisor had given me this this feedback, I would do my best in performing intravenous antibiotic therapy If my supervisor had given me this feedback, I would keep my mind more on intravenous antibiotic therap If my supervisor had given me this feedback, I would become more a gresive in performing intravenous antib otic therapy If my supervisor had given me this feedback, I would exert myself more for intravenous antibiotic therapy .36 .42 .33 .50 Standard Score Alpha for factor = .86 * Items removed from final version ofiPIM 132 Table 88: Deviations of observed from expected correlations (internal consistency) for PIM - - - - fi - Item 53 54 55 56 57 58 59 60 61 53 .00 54 .09 .00 55 .03 .12 .00 56 .03 .11 .01 .00 57 .02 .01 .04 .08 .00 58 .04 .00 -.03 .06 .06 .00 59 -.05 -.06 .01 -.01 .04 .06 .00 60 -008 -007 -002 -.05 .00 -002 -002 .00 61 -003 .00 -.04 -009 -003 006 -004 001 .00 62 -001 -003 .00 -010 -009 -013. -001 019* 009 63 004 -006 -.07 -004 -007 -007 004 001 008 64 -008 -011 -005 .00 -004 -003 00‘ 006 -001 Table SB (continued) Item 62 63 64 62 .00 63 .02 .00 64 .07 .14* .00 Average correlation: .34 Standard error: .06 * deviations greater than expected from sampling error (p <.05) 133 Table 13: Deviations of observed from expected correlations (internal consisitency) for PIM - ‘- -— _ item 53 54 55 56 57 78 60 61 62 53 .00 54 .08 .00 55 .03 .11 .00 56 .03 .10 .01 .00 57 .00 .00 .04 .08 .00 58 00‘ .00 -002 006 007 .00 60 -009 -008 -002 -005 .00 -001 -001 .00 61 -004 -001 -004 -009 -003 006 .00 .00 62 -001 -003 .00 -010 -009 -012* 019* 008 .00 63 003 -006 '00? -004 -.07 -006 001 008 002 64 -008 -011 -004 .00 -003 “.02 006 -001 007 Table 13 (continued) Item 63 64 63 .00 64 .15* .00 Average correlation: .40 Standard error: .06 Standard score alpha of final version of PIM = .89 *deviations greater than expected from sampling error (p <.05) Table 14: 134 Parallelism Deviation for PLA Outside Mean Item-Factor Deviatiens (Actual-Mean) Factor Correlation for Individual Items 17 18 19 20 21 22 PLE 039 .02 -003 .04 00‘ -005 -004 an 055 001 -004 00‘ 003 002 -004 EL .24 .05 -.07 .01 .01 -.02 .02 PIM .53 .01 -008* 00‘ 006 001 -002 *Deviations greater than (p <.05) expected from sampling error 135 Table 15: Parallelism Deviation for PLE Outside Mean Item-Factor Deviations (Actual-Mean) Factor Correlation for Individual Items 23 24 25 26 27 28 PLA 037 002 -001 005 014* 006 -001 an 0‘0 009 -001 005 -018* 002 003 H1! 020 -001 003 007 -.04 002 -004 PIM .41 .04 -.01 .02 -.11* .03 .04 *Deviations greater than expected from sampling error (p <.05) 136 Table 16: Parallelism Deviation for SHAME Outside Mean Item-Factor Deviations (Actual-Mean) Factor Correlation for Individual Items 29 31 32 35 37 39 PLA .46 .09* -.04 -.02 -.01 -.02 -.01 PLE 034 -005 -004 -001 006* -003 000 EL .28 .02 -.02 .00 -.06 .04 -.03 PIM 050 011* -003 -001 -002 -002 -001 *Deviations greater than expected from sampling error (p <.05) Table 16 (continued) Outside Mean Item-Factor Deviations (Actual-Mean) Factor Correlation for Individual Items 40 41 42 44 PLA .46 .02 .10* -.10* -.09* PLE 034 013* 000 -008. 001 EL .28 .02 .03 .04 -.07* PIM 050 -002 008* -011* 001 *Deviations greater than expected from sampling error (p <.05) Table 17: 137 Parallelism Deviation for HL Outside Mean Item-Factor Deviations (Actual-Mean) (p <.05) Factor Correlation for Individual Items 47 48 49 50 51 52 PLA .21 .00 .01 .09* -.03 -.02 -.04 PLE 019 -.00 -003 003 00‘ 005 -002 an 030 -005 003 00‘ -003 003 -005 PIM .28 -.01 .01 .08* -.02 .02 -.06 *Deviations greater than expeéted from sampling error 138 Table 18: Parallelism Deviation for PIM Outside Mean Item-Fictor Deviations (AEtual-Mean) Factor Correlation for Individual Items 53 54 55 56 57 58 PLA 0‘]. 002 -006 005 I.002 -005 -003 PLE 033 004 -005 -002 -006 002 -002 SH .46 .03 .07* -.04 -.09* -.03 .08* HI! 02‘ -001 -003 "00‘ 009* 000 ‘002 *Deviations gréEter than expected from sampling error (p <.05) Table 18 (continued) Outside Mean Item-Factor Deviations (Actual-Mean) Factor Correlation for Individual Items 60 61 62 63 64 PLA .41 .06 -.04 .00 -.02 .05 PLE .33 .00 .06 .03 -.01 .05 SH .46 .00 .06 .02 .00 .11* HL .24 .03 .06 .01 .05 .07 *Deviations greater than expected from sampling error (p <.05) 139 Table 19: Corrected and Uncorrected Correlation Matrix among Theoretical Variables* Variables MA ME PLA PIE Sn HL PIM MA 1.00 .00 .64 .31 .78 .36 .58 ME .00 1.00 .35 .65 .39 .26 .48 PLA .66 .36 1.00 .47 .67 .29 .65 PLE .33 .68 .51 1.00 .50 .26 .52 SH .82 .41 .74 .55 1.00 .41 .73 HL .39 .28 .32 .29 .46 1.00 .39 PIM .62 .51 .72 .58 .82 .45 1.00 *Lower triangle for corrected correlation and upper triangle for uncorrected correlation. MA: PLA: PLE: HL: PIM: Manipulation of Lack of Ability in Supervisory Feedback Manipulation of Lack of Effort in Supervisory Feedback Perception of Lack of ability (Supervisor's) Perception of Lack of Effort (Supervisor's) Helplessness Performance Improving motivation . ., .. 'l:l...lu.-J‘« 2,.. TE UNIV. LIBRRRIES “11111;“.MM )Qgfl‘fl.)‘§*)-