“ 7" 4:; R00 MW 2 @‘WM . ~ I. , / -;. /, 1' ' DECO 81999 ‘E / s,” ‘ ‘ 4:21 ‘ «it f..\ 1 w , , \ [I’m "M basis?” SHAME AND GUILT IN RELATION TO OTHER ASPECTS OF PERSONALITY By Thomas F. Negri A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1978 ABSTRACT SHAME AND GUILT IN RELATION TO OTHER ASPECTS OF PERSONALITY By Thomas F. Negri The focus of this study was the assessment of shame and guilt in relation to: perceptual style (field independence - dependence); defensive style (repression-sensitization); sex role adherence (endorsement of sex role characteristics in self reports); attention deployment (attention to positive and negative trait feedback after experimentally manipulated success or failure, and an incidental learning task.) College student volunteers (241 males, 207 females) completed a packet containing the revised Attitude Anxiety Survey (Perlman, 1958), Bem Sex Role Inventory (Bem, 1974), Closure Flexibility Test (Thurstone and Jeffreys, 1965) and the R-S Scale (Byrne, 1961). Based on above or below median scores on the AAS, 128 subjects (64 males, 64 females) were included in one of four shame-guilt combined score groups (i.e. High Anxiety, High Shame, High Guilt, Low Anxiety) and asked to return for a second session. Equal numbers of females and males within each shame-guilt group were randomly assigned to experimentally manipulated success or Thomas F. Negri failure conditions. The design allowed for inter~group comparisons on inventory and behavioral measures. Second session subjects were given several tasks and then given the opportunity to read trait feedback ostensibly based on their inventory responses. The task consisted of an incidental learning paradigm that combined word memorizations and anagram solu- tions. Anagram solutions were used as the basis for the experimental manipulation of success and failure. The subjects were then given ten minutes to read personal trait feedback. Subjects' attendance to feedbackwas timed in ten second blocks through a one-way mirror. Subjects were then asked to complete a recall form that included traits on which they were rated (an additional measure of defensive style). Subjects were then debriefed. Reported belief in the success-failure manipulation and trait feedback was high, though both were in fact false feedback. Contrary to expectation, shame, along with guilt, was positively associated with sensitization. The pattern was particularly strong for High Shame and High Guilt subjects. Both High Shame and High Guilt groups showed elevated field independence. This was unexPected for the High Shame group. The expectation that femininity and shame, and masculinity and guilt, would be positively associated was not borne out. Femininity and guilt, and masculinity and shame, were positively associated, a reversal of expectation. Males were more available for inclusion in High Shame groups and females, in High Guilt group. Thomas F. Negri High Shame group and High Guilt group subjects were similar in their level of differentiation and their defensive style, there- fore muted differences between these groups on attention deployment, recall, and incidental learning measures, were not surprising. Statistically non-significant trends suggest a pattern of differences reflective of an interaction of sex role endorsement and self- disclosure, with shame and guilt variables. In general, some support was found for a developmental view of shame and guilt, particularly in regard to defensive style and level of differentiation. DEDICATED In deepest gratitude to Dr. Elaine Donelson ii ACKNOWLEDGMENTS The Lord often does Hiswork through people, and I want to deeply thank all those indivdiuals who provided guidance, direction, and support, throughout the writing of this dissertation and my time in East Lansing. Dr. Elaine Donelson--for being my teacher, friend, advisor, dissertation chairperson and my guide on a personal quest for wholeness. Dr. Sam Plyler, Dr. Dave Wenger, and the late Dr. Bill Kell-- for sharing with me some of their knowledge about doing psychother- apy, and Sam's putting corks in my feet, so my self wouldn't drain out any more. Dr. Dozier Thornton, Dr. Al Aniskiewicz and Dr. Sam Plyler-- for their openness to my topic choices and their participation on my doctoral committees. Dr. Judith Taylor--for her wonderfully contagious exhuber- ance about, of all things, statistics. Dr. Art Seagull--for his gracious permission to use the Psychology Clinic facilities for the running of this study. Mr. Roger Halley and the Psychology Department Business Office--for financing the purchase of some of the testing instru- ments used in this study. iii Ms. Sandy Harshman and Ms. Ann Browns-for their patience and their work for the typing of seemingly endless pages of work. My friends-Dr. Dennis Pataniczek, Liz Mbnroe-Cook, Diane Deutsch, David Bellville, Rachael Male, and countless others--without their support, I could not have done this project or grown as a person. My very special friend, the Big Bear, Robert Patterson-- who has been with me lovingly, through so much. My Parents--for their beautiful loving, patience, tolerance, and willingness to know me and grow with me. And last, as first--the beautiful and loving Lord, who lights up my Life. iv TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O 0 O O O O O OViii LIST OF APPENDICES . . . . . . . . . . . . . . .xiii INTRODU CT ION I O O O O O O O O O O O O O O O O 1 THEORETICAL BACKGROUND . . . . . . . . . . . . . . 3 Shame . . . . . . . . . . . . . . . . . . 3 Early Theories . . . . . . . . . . . . . . . 3 Psychoanalytic View of Shame . . . . . . . . . 4 Elaboration of Ego and Superego involvement in Shame . . 8 Phenomenon of Shame . . . . . . . . . . . . . 10 Self Exposure . . . . . . . . . . . . . . . 11 Recent Views on the Genesis of Shame . . . . . . . 13 Implications for Defenses . . . . . . . . . . . 16 Reflective and Non-Reflective Functions of the Brain in Shame . . . . . . . . . . . . . l6 Experiential Characteristics of Shame . . . . . . . 21 Guilt . . . . . . . . . . . . . . . . . . 26 Guilt . . . . . . . . . . . . . . . . . . 26 Phenomenology of Guilt . . . . . . . . . . . . 26 Recent Views on the Genesis of Guilt . . . . . . . 28 Experiential Characteristics of Guilt . . . . . . . 29 Definitions: Shame and Guilt . . . . . . . . . . 31 Shame and Guilt . . . . . . . . . . . . . . . 32 The Self and the Ego . . . . . . . . . . . . . 32 The Superego Construct . . . . . . . . . . . . 33 Review of Previous Research . . . . . . . 40 Psychological Differentiation and Personality Characteristics . . . . . . . . . . . . . 41 Differentiation Research Relevant to Shame and Guilt . . . . . . . . . . . . . . . . 44 Research on Defense Mechanisms Relevant to Shame and Guilt . . . . . . . . . . . . . 47 Sex Differences in Perceptual Style and Their Relationship to Shame and Guilt . . . . . . 52 STATEMENT OF RESEARCH PROBLEM . . . . . PILOT STUDY . . . . . . . . . . . Testing Instruments . . . . . . Original Attitude Anxiety Survey (AAS) Revised Attitude Anxiety Survey . . Bem Sex Role Inventory (BSRI) . . . Closure Flexibility Test (CFT) . . . Procedure . . . . . . . . . . Results . . . . . . . . . . Assessment of BSRI . . . . . . . Assessment of the AAS . . . . . . Subject Classification on the AAS . . Intercorrelations: Total Sample . . Discussion and Implications . . . . MAIN STUDY . . . . . . . . . . . Hypotheses . . . . . . . . . . Method . . . . . . . . . . . SMflams. . . . . . . . . . . Testing Instruments . . . . . . Attitude Anxiety Survey: Subject Classification Bem Sex Role Inventory . . . . . Closure Flexibility Test . . . . Repression-Sensitization Scale . Procedure . . . . . . . . . Session 1 . . . . . . . . . . Session 2 - General Procedure . . . Behavioral Task . . . . . . . Success/Failure Manipulation . Trait Feedback . . . . . . . . Timing Selective Attention to Feedback Post Feedback Period . . . . . Recall Measure . . . . . Manipulation Check . . . . Debriefing . . . . . . RESULTS . . . . . . . . . . Subjects . . . . . . . . . . Descriptive Statistics: Individual Variables (first session). . . . Attitude Anxiety Survey (AAS) . . Bem Sex Role Inventory (BSRI) . . Closure Flexibility Test (CFT) . . . Repression-Sensitization Scale (R—S) . vi Page 58 66 66 66 68 69 7O 72 73 73 74 74 78 81 84 85 9O 9O 92 92 92 93 93 94 94 95 95 98 98 101 101 102 102 102 104 104 106 106 108 112 115 First Session Inventory Measures: Hypotheses 5, 6, and 9 Hypothesis 5 Hypothesis 6 Hypothesis 9 Comparison of Main Study Results and Pilot Study Results with Regard to Pilot Study Hypotheses Second Session Inventory and Behavioral Measures: Hypotheses 10-13 Manipulation Check Attention to Positive and Negative Feedback Hypothesis 10 Hypothesis 11 Incidental Learning Hypothesis 12 Recall of Trait Feedback Hypothesis 13 Summary of Results Field Independence Defensive Style Attention Deployment . Sex Role Adherence and Shame-Guilt DISCUSSION Methodological Issues Instruments . Subject Selection . . Operationalizations of Shame and Guilt . Defensive Sensitization and Field Independence Attention to Feedback Memory of Feedback Incidental Learning Sex Role and Gender SUMMARY AND SUGGESTIONS FOR FURTHER RESEARCH . APPENDICES BIBLIOGRAPHY vii Page 115 117 119 123 129 131 133 138 139 143 146 147 150 150 . 154 154 . 154 . 155 155 . 158 . 159 . 159 . 165 . 166 175 . 182 . 193 . 200 . 202 . 207 . 211 . 368 LIST OF TABLES Pilot Study Table Page 1 Classification of Subjects, Based on Bem Norms . . . . 75 2 Attitude Anxiety Survey: Perlman and Current Study Item Homogeneity Reliability . . . . . . . . . 76 3 Criterion Groups Based on AAS Relative Score for EaCh sex 0 O O O O O O O O O O O O I I O 77 4 Number of Subjects of Each Sex within Criterion Shame- Guilt and Sex Role Endorsement Categories . . . . . 79 Main Study 5 Attitude Anxiety Survey . . . . . . . . . . . . 105 6 Number of Subject of Each Sex Available for Inclusion in Each Cell, Based on the Total Sample . . . . . . . . . . . . . . . 107 7 Attitude Anxiety Survey: Intra-test Correlations . . . 109 8 Comparison Between Main Study (Sexes, Total and Separate) and Bem (1974) Samples on Bem Sex Role Inventory . . . . . . . . . . . . . . 110 9 Median-based Sex Role Classification (Based on Spence Method) . . . . . . . . . . . . . . 113 10 Bem Sex Role Inventory: Intra-test Correlations . . . 114 11 Closure Flexibility Test: Intra-test Correlations . . . 116 12 BSRI Variables within Shame-Guilt Levels . . . . . . 118 13 Multivariate F Tests for Effects of Shame, Guilt and Sex, on Inventory Measures . . . . . . . . . 120 viii Table 14 15 16 17 18 19 20 21 22 23 24 25 26 Table of Univariates . . . . . . . . . . . . RrS Mean Scores by Shame-Guilt Level Compared to Total Group and Byrne Sample . . . . . . . . . Shame-Guilt Groups: Means for Closure Correct on Closure Flexibility Test . . . . . . . . . . AAS Intercorrelations (Main Study) . . . . . . . Manipulation Check Questions within Shame-Guilt Level, Sex and Success-Failure Condition . . . . . Mean Time (in Seconds) Spent With Trait Feedback by Shame-Guilt Level, Sex, and Success-Failure . . . . Table of Multivariate F Tests for Effects of Shame, Guilt, Sex and Condition, on Reading and Recall of Trait Feedback . . . . . . . . . . Mean Anagram Solutions for High Shame, High Guilt, and Total Group, Compared to Mendelsohn and Griswold (1966) sample) 0 o o o o o o o o o o o 0 Table of Multivariate F Tests for Effects of High Shame by High Guilt (Covariance Controlled for Baseline) on Memory-cued and Tape-cued Variables . . . . . Mean Recall for High Shame and High Guilt Groups within Success and Failure Conditions . . . . Multivariate F Tests for Task Effects within High Shame and High Guilt Groups . . . . . . . Median Sex Role and Repression-Sensitization Classi- fication within Sex and Condition for High Shame and High Guilt Groups . . . . . Median-based Sex Role, Reclassified Based on Exactly (or within .05) Meeting Criteria, within Sex and Condition for High Shame and High Guilt Groups . . . . . . . . . . . . ix Page 121 125 128 132 134 140 142 148 149 . 152 . 153 . 188 . 198 Appendix B Table 1 10 11 12 13 Intercorrelations of Shame, Guilt, and Closure FleXibility scale 0 O O O O O O O O O O Intercorrelations of Shame, Guilt and BSRI Variables . . . . . . . . . . . . . . Intercorrelations between Femininity, Masculinity and Closure Flexibility Scale . . . . . . . T-tests for Sex Differences on All Variables in Pilot StUdy O O O O O O O O O O O O 0 Relative Score . . . . . . . . . . . . T Tests for Sex Differences on AAS Relative Scores for Criterion Groups . . . . . . . . . . T-tests for Differences in Relative Score Means Between AAS Criterion Groups . . . . . . . Intercorrelations of BSRI Variables with All Other Variables for the Masculine Criterion Group . . Intercorrelations of BSRI Variables with All Other Variables for the Feminine Criterion Group . . Intercorrelations of BSRI Variables with All Other Variables for the Androgynous Criterion Group . T-tests for Sex Differences on BSRI Variables Within Androgynous Criterion Group . . . . . T-tests for Sex Differences on BSRI Variables Within Masculine Criterion Group . . . . . . AAS Data on Excluded Subjects . Appendix C Classification of Subjects by Sex, into High, Medium, and Low Shame-Guilt Levels (Self and Other Forms). Sex Role Categories (based on median cutoffs) by Shame—Guilt Intensity (Self and Other Forms) Page 271 272 274 275 280 281 282 283 285 286 287 288 290 294 295 Table 3 10 11 Mean Closure Scale Score Within Sex Role Categories and Shame-Guilt Levels (Self and Other form) . . . Number of Subjects of Each Sex Role Category Within High, Medium, and Low Closure Scale Score Groupings and Shame-Guilt Levels . . . . . . . Cases with Consistent and Inconsistent Designations Over Two Sessions . . . . . . . . . . . . Subjects Included in High, Medium, Low, Relative Score Categories Within High, Medium, Low Distribution- Based Shame-Guilt Levels . . . . . . . . . . Appendix E Three Way Analysis on BSRI, CFT, and RrS Scale, Observed Cell Means . . . . . . . . . . . Three Way Analysis on BSRI, CFT and RrS Scale, Observed Cell Standard Deviations . . . . . . . . . . Table of Multivariate F Tests for Effects of Shame, Guilt, and Sex on Inventory Measures . . . . . . Table of Univariates . . . . Tukey Pairwise Comparisons . . . . . . . . . . Four Way Analysis on Trait Feedback and Trait Recall, Observed Cell Means . . . . . . . . . . . Four Way Analysis on Trait Feedback and Trait Recall, Cell Standard Deviations . . . . . . . . . Table of Multivariate F Tests for Effects of Shame, Guilt, Sex, and Success-Failure . . . . . Multiple Analysis of Variance for Nested Task Effects (Cells Means and Standard Deviations) . F Test Results . . . . . . . . . . . Analysis of Variance on Negative Feedback for Nested Shame and Guilt Effects in Success and Failure Conditions (Cell Means and Standard Deviations) . . xi Page 296 297 298 300 325 326 327 328 330 331 333 335 336 337 338 Table Page 12 Multiple Analysis of Covariance on Memory- and Tape- Cued Variables, Controlling for Baseline (Neutral) Variable . . . . . . . . . . . . . 339 13 T-tests Computed on Main Study . . . . . . . . . . 341 Appendix F 1 Mean Recall of Trait Feedback by Shame-Guilt Level, Sex, Success-Failure Condition . . . . . . . . . 350 xii LIST OF APPENDICES Appendix Page A BASIC TESTING INSTRUMENT . . . . . . . . . . . 212 B PILOT STUDY . . . . . . . . . . . . . . . 267 C SUBJECT CLASSIFICATION . . . . . . . . . . . . 291 D FALSE FEEDBACK MATERIAL . . . . . . . . . . . 301 E MULTIVARIATE F TESTS . . . . . . . . . . . . 323 F RECALL OF TRAIT FEEDBACK . . . . . . . . . . . 349 G THEORETICAL REFINEMENT . . . . . . . . . . . . 352 xiii INTRODUCTION The phenomenology of shame and guilt have been little studied in the field of psychology. The primary interest of this paper is to further elucidate each, with particular emphasis on the development of the experience of shame. Guilt has gained some attention, largely as a result of its central place in psychoanalytic theory, and it is a term actively used in the clinical setting, though research is lacking. The study of shame has been, until recently, largely non- existent. Lewis (1971) suggests that this arises from the SERIEELEXHQE"Ehgwgggition of shame in the psychoanalytic theory. Lynd (1958) attributes the lack of concern with the tendency among researchers and theorists to subsume the shame experience under the label of guilt. She further suggests that the very nature of shame--an experience that arises from that which one wishes to hide-- effects not only the experience and expression of shame, but has effected the field of psychology overall, via an unwillingness to approach what is hiding in each of us. The ambiguity of shame and the wish to defend against exposure to others and to the self exerts a continual press to substitute more clearly defined and limited feeling states (Bassos, 1973). Thus, shame has been subsumed under guilt. Although there are similarities between shame and guilt, there seem to be qualitative differences to be explained. Recently, there has been some attention to shame and guilt as experiential phenomena, and some attempt to assess their relation to one another. There also has been some attempt to relate each to specific con- stellations of characteristics and to personal styles of functioning (Perlman, 1958; Binder, 1971; Lewis, 1971; Negri, 1974). This writer views shame and guilt organismically. They influence and are influenced by other consistent aspects of a person. It is hypothesized that constellations of characteristics predispose individuals to experience shame and/or guilt. Further, complex interactions among factors facilitate the continuance of the personal style and within it, the modal experience of shame and/or guilt. The primary focus of this paper and the research described is to assess shame and guilt as they relate to: perceptual style; as measured by field dependence-independence; defensive style, as measured by repression or sensitization responses; sex role adher- ence, as measured by the endorsement of sex role consistent or inconsistent characteristics in self descriptions; and attention to positive and negative feedback after experimentally manipulated success or failure experiences. THEORETICAL BACKGROUND $12-12 Early Theories Regarding:Shame James (1890), in discussing the "self," suggested that a person's "social self" is the result of " . . . an innate propensity to get ourselves noticed favorably, by our kind" (p. 293). Addi- tionally, he describes the "self-dissatisfied" as " . . . those good people who think that they have committed the unpardonable sin and are lost forever . . ., who crouch, and cringe and slink from notice and are unable to speak aloud or look us in the eye" (p. 293). In this vivid description of the self-dissatisfied, James aptly out- lines, in the extreme, what this writer views as some of the com- ponents of the shame experience, and characteristics of individuals prone to shame experience. James captures the need for approval and affirmation; the attribution of power to another to control one's self esteem; the subsequent needs to achieve in order to please the other, attain perfection and overcome inferiority; the self-directed hostility, denegration and the tenuousness of positive self esteem; the with- drawal, inability to speak and aversion of gaze, for fear that the individual will be made more aware of the environment's ridicule and loathing, which the individual already assumes is occuring; and most importantly, the ongoing sense that one has committed an unknown, and unpardonable sin, that cannot be stoned. Theories of shame and guilt are presented in the following pages. ‘Little research has been done in thisgarea,_thoughsome descriptive discussions do exist. It is this writer's purpose to integrate previous writings and operationalize the shame and guilt phenomena in order to build a suitable framework for research. Psychoanalytic View of Shame Lewis (1971) re-evaluated Freud's early published case studies and she tried to demonstrate the involvement of uninterpreted shame in symptom formation. She cited a Breur and Freud (1893-1895) discussion of hysteria. In the Breur and Freud presentation, they assign to mortification, a variant of shame, a central role in symptom formation. They observe: . . . that an injury suffered in silence is a mortification-— a krankung-which literally means "to make sick." When one suffers an injury, they wrote, one tries to obtain revenge, as a catharsis. Or one can right the memory of a humilia- tion (also a variant of shame) by remembering its worth. By contrast an injury suffered unsilence makes one ill (p. 201). Further they describe hysterical symptoms as: . . . strangulated affect [where] people grow senseless with [unexpressed] zinger and fright . . . The hysterical symptom is a nmemic symbol of an unbearable affect . . . Any exper- ience that elicits (listress in affect--anxiety, shame or physical pain-~can be regarded as traumatic (p. 201). Freud and Breur recognized shame and its derivatives as sources of symptom formation. Likewise, they appreciated the quali- ties of shame that prompt directing aggression to the self, and the difficulty in its expression and communication. However, these early formulations about shame-like experiences were superseded by later theoretical emphasis on guilt, and shame was relegated to a more limited, sexually based context. Freud (1896) proposed that shame arises from a self-reproach by the exposure to disapproving others of a sexual act in childhood, then (1905) expanded this theory, citing shame as nearly synonomous with disgust and self-loathing, and acting as a force to oppose voyeuristic and sexual drives generally. In further elaboration (1914), he differentiated between "a sense of shame" and the "dread of conscience," the former was linked to shame, functioning to insure narcissistic satisfaction by seeing that the ego meaured up to the ego ideal. In 1924, Freud stated, ". . . the biological necessity demands that an affective state of danger should have an affective symbol" (p. 19). Shame acts as a symbol or signal to redirect and limit sexual or pleasure oriented actions and thereby insures the avoidance of rejection. Fenichel (1945) presents shame ". . . as the specific force directed against the urethral erotic temptation" (p. 139), and counterposes it to ambition, which has as its purpose the conquering or overcoming of shame. He views shame as arising from loss of or lack of bladder control in childhood, and relates this to the custom of pillorizing as punishment. The individual's failure is exposed to other people. Thus, Fenichel equates the experience of being looked at with the experience of shame. He suggests that the shamed individual wishes to hide or avert the face to avoid the gazes of others because of a magical belief that the aversion of gaze, or closing of the eyes, will enable him not to be seen. The associative link of being shamed and being stared at is summarized as ". . . to be looked at is to be despised" (p. 139). Fenichel points out that more generally, shame acts as a motive for defense directed against exhibitionism and sceptophilia. Here again exposure and looking are coupled, and each is defended against with shame. Although he did not fully develop the idea, Fenichel suggests that shame is rooted in a primitive physiological reflex pattern. As the ego develops, this pattern is used for defensive purposes, " . . . if you do this or that, you may be looked at and despised" (p. 139). The ego uses anxiety as a signal to limit these behaviors, for fear of shame. It is also suggested that in those individuals in whom there is a damning up of conflict, signal anxiety fails to operate appropriately and overwhelming panic-like shame occurs. This writer contends that at this point the shame reaction, and the counter-energizing measures that are mobilized to deal with the shame producing stimuli, prompt imagery, thought, and actions, paralleling those outlined by Fenichel in his discussion of responses to trauma (regression, helplessness, passive dependence, and oral trend manifestations). Several major points arise in Fenichel's description of shame: the physiological reflex base; the equation of being viewed or stared at with being despised; the necessary experience of expo- sure to others; the idea that shame can arise from action, or thoughts of action of an exposing or "looking" nature. Theorists attempt to confront these points and integrate them into a broad- based view of the involvement of shame in personality development. Levin (1967) also discusses the involvement 0f shame with sexuality, suggesting that shame is linked with over-exposure and A... ..,___‘_ _____ q subsequent rejection by significant others. This fear of exposure prompts the individual to take on a "chameleon-like" quality, in which these persons develop the ability to "read" their environment and change themselves to suit and satisfy those populating their surroundings, relying on repression and facade to do so. In order for an individual to be in a position to satisfy his libido it is necessary for him to expose himself to others. However, it is also necessary that he be able to control the degree of self-exposure so as not to reach out to others indiscriminately and subject himself unnecessarily to the possibilities of rejection, . . . shame affect contri- butes to the latter goal . . . [and] also leads to caution in the face of advances made by others (p. 268). Gerhart_Piers (1953), a major contributor to the literature mm on shame, suggests that " . . . only Erikson and Alexander have ascribed to shame an importance equal to guilt in human pathology" (p. 11). He refers to Franz Alexander's paper (1938) in which the latter discusses the experience of inferiority feelings. Although Alexander attributes both guilt and inferiority feelings to tension between ego and ego ideal, it is suggested ". . . that they are fundamentally different psychological phenomena, and as a rule their dynamic effect on behavior is opposite" (Piers, p. 10). Piers adds ". . . it would seem imperative that emotions phenomenologically and dynamically so different would also differ structurally" (p. 10). Erikson's (1950) Neo-Freudian theory further develops the distinc- tion between shame and guilt. Elaboration of Ego and Superegp_lnvolvement in Shame Erikson (1950), in presenting his ego development sequence, recognizes the development of shame as a potential hazard arising with the beginning of autonomy. One of his most salient points, which later contributed to the work by Piers (1953), is the use of the phrase "a sense of shame," implying a state that pervades both surface and depth, conscious and unconscious. Also, ". . . senses are at the same time ways of experiencing, accessible to intro- spection; ways of behaving, observable by others; and unconscious inner states, determinable by tests and analyses" (p. 251). These two factors broaden the focus with which to view the shame pheno- menon, and they attribute particular perceptual, experiential, and behavioral components to shame. More specifically, Erikson describes the shame experience as coupled with dgubt andcounterposed to autonomy. He discusses the development of this stage in terms of the move toward autonomous functioning and learning to "let go." This is an extension and abstraction paralleling the general physical development of the child, and in western society parallels too the task of learning sphincter control over bowel movement. Erikson attributes impor- tance to the parenting agent, who may show respect for the child's autonomy, while acting as a vigilant protector against shame- and doubt-producing failures. This approach communicates to the child a recognition of separateness, with the availability of the parent to the child; and facilitates the child's recognition of self-worth as well as discretion. When this approach is not used, and active shaming or exposure of the child's shortcomings occurs, the child experiences a sense of badness that seems all-pervasive. Weigel (1974) suggests that active shaming by parents prompts the child to feel that he or she is different and bad because of aha-NV __ their shortcomings. This, then, causes the child to hold the self responsible, not because of the doing or not doing of a particular act, but because one ii a certain way. This gives rise to shame and self-doubt in the individual. Piers (1953) presents shame and guilt as forms of intra- psychic tension, each quality equally important in the manifestation of pathology in ego development. More specifically, he differen- tiates between shame and guilt, attributing shame to a tension between ego and ego ideal, and guilt to a tension between ego and conscience. Guilt arises when a superego boundary is touched or transgressed, and shame occurs when there is a failure to meet or reach a goal set by the ego ideal. Simply put, guilt arises from transgression; shame arises from failure. Piers views the anxiety w in shame as a fear of abondonment, and the anxietyixlguilt as the fear 0f castration. To this Levin (1967) adds that fear of rejec- tion as well as fear of abandonment lies at the core of the shame experience. Grinker (1955) discussed the individual's "shame at failures in realization of growth potential." He suggests this was largely self—reflective and was experienced as "I am not as good as he is or as they expect me to be, therefore, I hate myself" (p. 36). ~ 10 Lewis (1971) views the superego as having multiple functions- She specifically speaks of the superego as a psychic monitoring device which regulates the individual's self—evaluation in an attempt to maintain stasis. HGuilt and shame are among the states evoked bythe monitoring agency" (p. 19). The superego regulates drive, and is active as an integrative force dealing with the development of human values and ideals. Further, Lewis discusses the Freudian hypothesis that super- ego development involves an identification phenomenon. The inter- nalization of the moral code can occur because of castration threat, and is likened by Lewis to avoidance conditioning. Avoidance Sgpditioning ogcurs because of the attachment of negative consef quences to certain inpulses, so that the child comes to view the impulses as "wrong." A second route of identification is through the imitation of the beloved and respected parent, facilitating the early development of the ego ideal. This direct involvement of the child with the model involves the threat of loss of parental love, and hence, as far as modeling has occurred, the loss of self- love and the loss of esteem in the eyes of the valued other. Lewis does not view these identification patterns as mutually exclusive; each is likely experienced at one time or another by the child. Phenomenon of Shame Piers (1953) uses the word shame to imply an experience transcending the focus on comparison with external figures that is implied by "inferiority feelings," as described by Alexander. This 11 is not an attempt to negate the importance of external figures, but to allow for the inclusion and recognition of a more internal- ized tension between the ego and ego ideal. Theoreticians have also attempted to recognize something more in shame than sexuality; ". . . is not to be ashamed of loving on account of exposing shame or surrendering the body, but to be ashamed that love is not complete . . . That something inimicable in oneself keeps love from reaching completion and perfection" (Hegel, cited by Piers, 1953, p. 18). The etymology of the word ghgmg_also belies a broader con- text than its limited sexual reference suggests. The word sham; has its roots in Old English, but appears also in germanic languages. It carries the meaning of ". . . a wound or the exposure thereof." It includes a self-reference and can imply "self-degradation," a wound to one's self-esteem, and felt unworthiness in one's own eyes and in the eyes of others. There is no act of repayment or legal (codal) reference, nor is there necessarily imPlied the trans- gression of a code (Webster Dictionary, 1950). Self Exposure Of particular importance is the self-referent in shame. The self acts as the registrant of activities, attitudes, and feel- ings recognized by the person as one's own. Lewis (1971) suggests that in shame there exists more self-consciousness than in guilt. The primary focus of the experience of shame is directly about the self, and the negative evaluation of it. This occurs as a result 12 of the psychic structure that allows shame to occur. Within the structure, the occurrence of a thought, behavior, impulse, feeling or need that is shamedworthy, prompts both acute self-awareness, and an awareness of the self from the negatively evaluating, external other's viewpoint. There is a doubly intense awareness of the self. Lynd expands on this premise, describing shame as an experience ". . . that affects and is affected by the whole self . . . and that discrete acts or incidents, including those of seemingly extreme triviality have importance. In this moment of 'self' consciousness,_th§L:§elf:fistands_reyealeg" (1958, p. 49). gun—- -.fi~—...... The individual fiflififil’ifiiffiiflfit- 999.1£an-er1.4.t..1§ not... This reflects the awareness of some inappropriate occurrence and acute awareness and negative evaluation of the self (by the self, the real or imagined other, and that aspect of the self that acts in league with the other). This brings to light again a factor of major importance in the shame experience, that of exposure, particularly unexpected exposure. Erikson (1950) describes shame as an experience of complete self-consciousness, complete exposure, and awareness of being wholly visible, both for one's external being, and one's internal life and thoughts; all of one's badness is visible whether the audience is actual or imagined. Erikson describes the response to such exposures ". . as a wish to sink, then and there, into the ground" (p. 252). He also suggests that experienced shame includes rage at the self because some failure or inadequacy has been exposed to a significant other. In the present writer's view, this exposure 13 can also be to an internalized image of the other or one's own ideals. Erikson discusses the shamed individual's wish to force the shame-inducing eyes to look away; however, the individual retreats from.the scene. He further suggests that shame is induced visually as well as verbally and the individual feels small in relation to the other and to the ego ideal. Although shame often involves visual perception, this sensory modality is not the exclusive vehicle for the induction or experience of shame. Congenitally blind individuals who experience shame became aware of their exposure even though no visual imagery was available to them and they could not imagine being seen. Blank (1958) and Feldman (1962) suggest that they construct their images of being seen from non-visual (e.g., auditory, tactile, and kinesthetic) elements, and it is through these modalities that they assess that those perceiving them are doing so in a shaming manner. Recent Views on the Genesis of Shame Kell (1973) suggests a discrimination between primary and secondary shame. Kell's idea, described by Bassos (1973), is that ". . . primary shame experiences occur at the point where some previously unsuspected or neutrally evaluated part of the self becomes associated with shame for the first time. Secondary shame occurs when some aspect of the self that had been previously labeled shameful is suddenly exposed" (p. 1). Hell describes the genesis offlphe shame experience in a A“ Hug-v... ._. - developmental model within the family structure. He suggests that 14 I I "'yw { \_/ 2‘" when a need (such as for attention, care, contact, love) or expec- tation is expressed directly or indirectly (for the child this may be magical thoughts of parental foreknowledge), and this need is ignored, shame may occur for having had the need. If the individual to whom the request was made attempts to respond to the need later, the shamed individual will experience a secondary shaming and rage, based on the exposure of the primary shame. Kell suggests that the individual experiencing shame directs rage at the self for having expressed a wish, need, or attitude that could go without response from the significant other. This individual does not recognize the ignorance of the beloved or respected figure, but rather demeans the self for having M" -—fi..i__..-- HW‘..- ,, had the need, and having had the audacity to express it. The ‘~__flfgmwwi_iw individual assumes that in order to be loved and gain response, one must be like the beloved object; iguan_expression of needs goes without receiving a response, the individual experiences this as a deviation from the ego ideal, which may have been modeled on the beloved individual. In doing so, the individual sees the need as illegitimate, the self as imperfect and "notemough to be loved" and gain response. In this sense, shame, and the anxiety arising from it, relate to the terror of interpersonal disdain, and when internalized, to the fear of abondonment for having expressed any need, thereby risking the rejection of the beloved other. Kell (1973) also suggests that if one responds to the other's need late (i.e., after the initial shame reaction occurs) the lateness of the response and the initial insensitivity to the 1'," i /‘ 32.77;...“ ’7' V "\_,...d’ 15 request must be communicated. This allows for the individual's need, recognizes the shame, removes the stigma of the need and the request, and re-establishes the interpersonal bond often lost (with the loss of trust) as a result fo the shame experience. Similarly, Levin (1967) suggests that shame experience can lead to elaborate shame avoidance behaviors in the service of the inhibition of sexual and aggressive impulses. The individual must learn what to do, and what not to do, what to expose and what not to expose, if one is to avoid shame. This writer suggests that some experienced shame may also arise from the expression of non-sexual and non-aggressive impulses that are either purposely shamed or inadvertently ignored, giving the individual or child the message that those needs are shameworthy. Though some shame may be necessary for the inhibition of sexual expression, the occurrence of shame is not limited to sexual or aggressive expressions. The outcome of inadvertantly or inappro- priately shamed impulses or needs may be that the shame is inte- “my grated as a testament to one's lackof self-worth and badness for "'- h—-¢-u l s-J‘L—l. —.qI‘~ ‘1‘ having had the need. The shamefulness and subsequent lowered self- regard may become generalized and held as a non-verbal self experience. This internal conflict revolving around the self, self-worth, needs and need expression, may occur in the absence of others and be carried on as a primary mode of perceiving, exper- iencing, and behaving in adulthood. l6 Implications for Defenses Both Anna Freud (1936) and Valenstein (1961) assert that the individual's response to shame involves the repression of those thoughts, feelings and impulses which mobilize the shame affect. Hence, the experience of shame not only prompts the conscious mw-—‘---m ‘- avoidance of revealing certain thoughts, feelings, and impulses, fl. “mwv—qu liq-Hm”. ., .‘w but may . ”‘9“. s also prompt repression of them. Fenichel (1945) pointed out that the repressed impulses rising to consciousness may trigger a signal anxiety of impending shame, and reinforce the repression. This is discussed later in regard to sensitization and repression in defensive perception. In some instances, shame and repression act to limit tabooed sexual or aggressive expressions, butmlikewi§e_may also limit the genuine a expression of’previously shamed, nonrsexual, non-aggressive needs. In doing so, there is a damning up of the need system of the indi- vidual, and forced repression. The welling up of the needs over- takes the signal anxiety apparatus, prompting overwhelming "panic—like" shame, self-rage at the need expression, fears of insatiability, and withdrawal (due to secondary shame) before allow- ing the need to be fulfilled. Reflective and Non-Reflective Functions of the Brain in Shame Another view of the genesis of shame has been offered by Bassos (1973). He bases his discussion on concepts of brain later- ality. The left hemispheric cerebral function is seen as a reflec- tive capacity, enabling prediction, control, and manipulation of the 17 environment. To the right cerebral function is attributed a non- reflective capacity, which is characterized as non-verbal, spon- taneous and "unconscious." He suggests that the non-reflective mode is experienced as an "effortless being," in contrast to the "doing" quality of the reflective mode. In the reflective "doing" mode (nae experiences an objectified self, whereas in the non- reflective mode, one is "at one with" or absorbed with the exper- ience of the moment. He further suggests that both modes of consciousness are simultaneously involved in the experience of shame. Bassos differentiates between objectified and nonrobjectified relationships. The former are viewed as "I-You" car "I-It" rela- tions, where there is an objectified self, and may "not-I" objects. The reflective mode predicts and controls the distance between the objectified self and the object. Inherent in this type of relation- ship is a recognition of separateness. In the non-reflective mode, separateness is not assumed. Bassos finds "at oneness" more suitable than "relationship" here, for ". . . there is no discrimination of a self or other, no attempt to control or predict the self, the other, and there is an intense focus on that other" (p. 4). The shame experience occurs when one individual in a non- reflective state is not received and experienced in that same mode, but is met with a "finger pointing" response to some aspect of their being; the person is not aware of this aspect of themself while in the non-reflective, spontaneous state. This causes a startling self-awareness, and desperate search of the self for the cause of the finger pointing. The self becomes the focus of all attention, 18 including one's own; sees oneself as responsible, and feels oneself objectified, alone, and denied "at oneness." Bassos describes this as an undesired separation via objectification. The result is a jarring back into reflective activity, with an awareness that one has just been out of self-control. The reflective self now objecti- fies the non-reflective self. The final outcome is an inability to organize behaviors, helplessness, and disorientation, until the reflective capacity is fully restored. Bassos further suggests that shame, developmentally arising from a rejection of the child's need for "at oneness," prompts a feeling of power at being able to cause rejection by others but of powerlessness to gain their positive response. To the child's mind the rejection is the outcome of being in a non-reflective state. Essentially, one experiences one's being as rejected. In an attempt to control one's world and see to it that one's needs are met in the future, the individual (in the reflective mode) plans and behaves in ways to regain the lost love. Yet in doing so, the individual gives up being in the non-reflective mode. The outcome may be a sense of futility and impotence to affect those around one, as it is oneself who causes others' rejection, yet cannot insure others' acceptance. Menaker (1953) also suggests that the development of this devalued sEIf;image acts to preserve an.:lllusion of a loving and nurturant parent--inadequacy is felt to be inherent in the child rather than the parent. “He relates the chronically devalued self to the development of ongoing masochistic trends. Jacobsen (1954) 19 suggests that patterns of masochism are related to the cathexis of the self with too much aggression (i.e., self-directed hostility). Lewis (1971) also develops and tests the premise that the direction of expression of hotility is related to superego style. Excessively shamed individuals are said to experience themselves as less differ- entiated from their environment than individuals who experience guilt, rely more on the positive regard of others, attribute power to others to approve or disapprove of them and, in experiencing external disapproval, they direct anger not at the other, but at the self forrhaving failed the other. In psychoanalytic theory (Freud, 1930), the rechanneling of aggression into the superego arises from a blocking by the ego of the actual or appropriate aim. Thus, the experience and expression of anger about disapproval by another is not available to the person being shamed. The ego acts in league with the shaming other and directs rage at the self. This fosters the salience of the other, and negates (or reinforces the early established negation) the right of the self to have needs, to be separate and internally directed, and to ask for and gain gratification from the environment. Negri's (1974) position acknowledges the Freudian and developmental viewpoints and suggests that shame can arise from active shaming on the part of the parents (e.g., ridicule or showing contempt). The parents can indicate disappointment in or dislike of the child as a person, as well as for the child's sexual or non- sexual impulses. Additionally, shame may arise from inadvertent disregard for the child's immediate needs. This can set the stage 20 for an individual who will view the self (particularly the non- reflective or non-doing self) as unworthy of response or regard and can engender an introjected rage resulting from the perceived ' closeness and affection. parental withholding of "at oneness,‘ Subsequently, exposure of the established primary shame can give rise to secondary shame. This is because of the shame of exposure of the primary shame, the re-experiencing of the initial shame and self as unworthy, and the unexpressed rage at the parent for the lack of response to the original need. The outcome for the developing child is one of having a fragmented, nonaffirmed sense of self. Such a child would believe that being (in anon-reflective sense) must be controlled or avoided for fear of secondary exposure of the initial primary shame. This may further generalize into evaluating the self as worthless and all emotional needs as illegitimate because they have the power to cause the loss of the beloved object. Continuing primary and secondary shame experiences would lead to a devalued sense of self, and to the repression and suppres— sion of shamed needs and impulses. This fosters the salience of the early frustrating identification models, prompting more and more desperate attempts to please the parent or parent symbol and to gain the previously withheld affirmation by achievement. It also gives rise to the description by Levin (1967) of the "chameleon- like" person, whose other-directedness is both an attempt to control the self (avoiding non-reflective experience) and to control the 21 environment in hopes of avoiding further shame. There is an attempt to gain affirmation by being what the environment demands. This writer finds a useful parallel in Guntrip's (1969) description of the development of the schzoid personality in terms of early experienced deprivation and lack of affirmation. Guntrip suggests that once the schizoid pattern is established, an individual may desperately attempt to please others, in hOpes of regaining the love missed as a child. .At the same time the individual fears that "letting go" (cf., Erikson, Bassos) into true union with another will be overwhelming and devastate the brittle sense of self. Hence, an "in-out program" becomes established. The individual approaches the relationship out of need. However, they must recoil out of fear of exposure of early shamed needs, fear of emotional insatia— bility (a likely resultant of the early experienced lack of fulfill- ment of some needs), and fear of the expression of the early experienced rage at parental non-response. Insatiability of emo- tional needs is a major concern of shame-experiencing individuals and is tightly controlled, because of the early experience that the expression of a need or impulse causes pain, frustration, lack of fulfillment, shame, and ultimately, loss of the other through rejection and abandonment. Experiential Characteristics of Shame Several basic characteristics recur in the descriptions of the shame experience: unexpected exposure; involvement of the whole self; incongruity and inappropriateness; threat to trust (Lynd, 22 1958). Lewis (1971) adds that intense hostility is directed against the self. Negri (1974) found that shame-prone individuals (both sexes) demonstrate high needs for achievement with high fear of failure. He viewed this as an expression of a need to arrive at and ensure the attainment of the idealized self-view. Subsequent failure manifests itself as an acutely negative evaluation of the self, anxiety about exposure, and loss of esteem, regard, and respect in the eyes of the individual and in the fantasized evaluation by others-—shame. With the self as the focus of evaluation, we see the shame experience as sudden and total. Its basis in early interaction with significant others, and subsequent integration as a deviation from the ideal self-view, allows us to appreciate that one's entire life may act as a preparation for the continuation of such startling experiences. As dealt with here, the shame response involves an aspect of interpersonal interaction and is integrated as an expo- sure of the self to the self and the internalized other. Within the established shame system, incongruity, loss of or threat to trust, or unfhlfilled expectations, may bring to the fore acute feelings of unworthiness and failure, experienced as shame, by the individual. The oftentimes trivial nature of those occurrences is one contribution to the problem of communicating shame experiences. Lynd (1958) suggests that shame is a wordless, emotionally axlphysically preoccupying state. The difficulty in its communica- tion involves several factors. First, there is a loss of trust in the self as an accurate monitor of reality, and an acute awareness 23 of having built on a false assumption about one's self or others. Second, there may be a loss of trust for those populating the environment. The loss may be based on a fear of "being seen through" at a time when the essence of the self is felt to be exposed, or based on a realization of having misplaced one's confi- dence or of having falsely anticipated some particular response that does not occur. Third, the nature of the shame experience prompts a need to move away from others, retreating into onself, in order, r -.. _. r/I— -—‘ "a. , u p ‘to regain and solidify the self; this is because the experience is one of momentary implosion of the self via the breaking through of earlier repressed impulses which generates self—rage and disdain. Finally, the stimulus that prompted the shame may be trivial, as noted above, in which case the individual experiences not only the primary shame, but also shame for such over-response to triviality. In sum, the experiential core of shame consists of: the unexpected exposure; the acute awareness of self; the clarity of the discrepancy between what is, and what one wishes it to be. Be- haviorally, the shame experience may be marked by blushing, tremors, and sweating, which Laing (1960) describes as an "implosion of the self" in which the eyes close and the body turns inward on itself; the intent is to make the person as small as possible, as a response to the shame and as a protection from furthgrwhurt. Shame may be “MMW*- ‘1 described as a total, body felt experience, where the senses are flooded, both by autonomic over-stimulation and by shame thought and imagery. 24 Levin (1967) discusses the blushing that may accompany shame as both a communication of interest (sexually based) and a sense of uncertainty and caution about the appropriateness of the expression of that interest. The present writer agrees that shame masks a communication of interest and caution but maintains that shame is a reaction to the thought of or actual acting out of that interest or need. That is, blushing is interpersonally triggered by the other's expected negative response. The blushing in secondary shame probably arises as a reaction to the exposure of a need that pre- viously had been shamed, was repressed or defensively unexpressed, and again is breaking through the repression or seeking expression. Blushing in primary shame ariseszhareaction to actual or inadvertent induction of shame by the other. Lewis (1971), in discussing the direction of expression of hostility in shame, suggests that the other-oriented individual attributes to the idealized environmental objects and to their partial introjects, the capacity to regulate one's own self image and esteem. In doing so, the individual becomes unable to direct hostility at the object, for fear of rejection or abandonment and for fear of permanent loss of esteem. As stated by Freud, ". . . every piece of aggression whose satisfaction the subject gives up is taken over by the superego and increases the latter's aggressive- ness (against the ego)" (1930, p. 31). This writer suggests that the self is the actual recipient of hostility in shame, and as such, both the goal attainment and impulse controlling aspects of the superego direct aggression at 25 the self for having allowed overexposure, and in its wake, failure. It is in this context that both Lewis (1971) and Lynd (1958) suggest that the "righting" of shame occurs through initial withdrawal and subsequent reconstruction of the self. Shame Avoidance Levin (1967) suggests that the early experience of shame prompts the development of_slaborate avoidance techniques. Initially, signal anxiety acts to alert the individual to situations that might produce shame, but in cases where the individual experienced exces- sive shame, signal anxiety is replaced with avoidance behavior. The present writer agrees with this position and suggests that for the extremely shame-prone individual, signal anxiety has likely become overactive and diffuse and prompts the individual to avoid evaluative risk, even when the outcome would be self-enhancing praise. This can be seen in extreme in low esteem individuals who, even when they are given clear messages of liking, are ashamed of the attention given. Feldman (1962) notes that excessively shy individuals experience shame with any public attention, whether it brings glory or debasement. The present writer suggests that the early and excessive experience of shame can significantly effect the self-image. It can lead to oontinual self-debasement, self-hate, and avoidance of risk taking with the generalized other, for fear of further rejection. In doing so, repression and non-expression of needs can lead to 26 affective starvation, continued low self-image, and low self-regard, and continually intensified fears of exposure. Guilt In preparation for a discussion of the differences between shame and guilt, the development, phenomenon and experience of guilt will be discussed. Guilt Phenomenology of Guilt Fenichel (1945), in discussing the activity of the superego, points out ". . . that a constant watchman remains effective in the mind, who signals the approach of possible situations or behaviors that might result in the loss of the mother's love" (p. 102). He further suggests that the superego develops by introjection, demon- strated by the acceptance of parentally defined limits and prohi- bitions. This is due to the wish to be like the parent. The identification with parental prohibitions becomes a displacement for the intended or desired identification with the parents. Con- science is considered to be developed when the prohibitions retain their stand in the absence of the parent(s). The ego "borrows" from its strong parents the strength that enables it to suppress the Oedipus complex. In this way the resolution of the Oedipus complex brings about the marked and decisive "step within the ego" (606), which is so important for subsequent ego development and which by its organization is differentiated from its forerunner-- the superego (p. 104, 105). With the development of the superego, anxiety changes, in part, into guilt feelings. Hence the feared danger (castration or 27 loss of love) is no longer external but is internally represented by the loss of protection by the superego. This, according to Fenichel, is felt as an extremely painful decrease in self-esteem and in extreme cases, as a feeling of annihilation. A need for absolution arises. Punishment is accepted in the hope that the guilt and corresponding low self-regard will cease. Penance then acts as a substitute for castration; it is an active reparation and allows for the avoidance of the anxiety of passively waiting for something terrible to happen. Anxiety, manifested as guilt, acts as an internal signal monitoring the wrongness of one's thoughts, feelings and actions. Thus, guilt is a psychological state evoked by the con— science and the concept has been a generally used shorthand des- cription of drive regulation in the interest of social and moral concerns. Guilt seemingly arises out of a conflict between the ego and the impulse controlling aspect of the superego; it is an internal tension generated by the transgression of a barrier or boundary erected by the superego. (See later discussion of super- ego construct in shame and guilt.) The transgressions are usually, but not necessarily, expressions of aggressive or sexual impulses which are unacceptable to the early formed internalized punishing image. In more general terms, guilt occurs when an individual becomes aware of doing something, or of an impulse to do something, which violates the limits set by the conscience. Piers (1953) main- tains that the anxiety arising in guilt demonstrates a fear of 28 punishment by castration. The present writer would modify this to a more general view of fear of punishment. Recent Views on the Genesis of Guilt Guilt arises through transgression and essentially emerges in a later developmental stage than shame; the child is up and around and involved in the environment. Erikson (1950) suggests that the "sense of guilt" is a result of that period when the child must come to grips with infantile sexuality, incest taboos, and for the male child, castration fears. He suggests that the child, energized by the successful resolution of the Autonomy-Shame/Doubt stage, directs energy outward, in a more focused way than previously. ". . . attack and conquest" for the male, and ". . . In doing so, catching and making oneself endearing" for the female, become major endeavors. It is the vehemence of this energy and the vestiges of infantile sexuality that promotes the occurrence of guilt. Erikson suggests that guilt and anxiety are encouraged by increased genital awareness, as well as by the last desperate attempt to establish a sphere of unquestioned privilege with the mother, and the usual failure in this realm. It is at this point that the child moves from an exclusive involvement with the parents to one of modified, externally directed initiative. This is a resolution via compromise where the child learns to regular their wants and to deal with the external reality. In the male child, this seemingly arises from resignation to the loss of the mother, at least partially out of fear of castration by the father. Rules take on relevance 29 for the child. It is generally a time of rule setting, where the parents delineate codes of behavior and ethics and set expectations for compliance. Transgressions of commission or omission prompt punishment, and with the integration of parental values, self- regulation and guilt. Both Lewis (1971) and Piers (1953) also discuss the develop- ment of guilt as arising in the early Oedipal stage, in which punishment is meted out on the occurrence of a transgression. The experience of guilt, as stated, arises from transgressing proscrip- tive and prescriptive rules, felt responsibility for the action, and the evaluation of the act, or of the self via the act. It does not concern survival for the child (i.e., there is no fear of abandonment) and may follow adaptive responses to the trust- mistrust and shame/doubt-autonomy stages of development. Thus, it can be limited to the specific act. It is not integrated as a wordless self—denegration or questioning of the right to live, but involves an assessment of culpability and evaluation of the retri- bution. The ego is intact in guilt, while it is shattered or dissolved in shame. Experiential Characteristics of Guilt The popular meaning of EEEIE suggests concern with trans- gression, implying debt and the necessity for subsequent restitution. Specifically, guilt deals with the real world, acts done or left undone, and events for which one bears responsibility. Lewis (1971) suggests that the individual, having experienced guilt, attempts to 30 assess and determine the extent of one's own culpability, the extent of the injury caused, and the reparation and/or punishment necessary to amend the transgression. There is an objective quality to this assessment. However, "In cases of irreparable injury, guilt may lead to despair, which may in turn also evoke shame, as a reac- tion to the helplessness of the self to reverse the course of events" (p. 44). This may prompt an additional clouding of the realistic assessment of the injurious act. Likewise, the more self- contained, articulate experience of guilt may act as a defense for the more ambiguous pervasive feelings of shame. This may be demon- strated in individuals who, after mending the transgression, continue to carry the guilt ideation. One might suggest also that within western society there is an institutional recognition of guilt as a mode of experience, and appropriate ritual for its expiation. This is not the case with the shame experience. The "act" retains the central focus in the experience of guilt. This does not preclude the possibility of attaching a negative evaluation to the self, but if done, it is via the act and the self does not become the center of the experience. The self is intact and can actively pursue an objective evaluation of the situation and a rational assignment of responsibility and subsequent reparation. The preoccupation with these factors may prompt the guilt affect to subside, leaving guilt ideation and thoughts about the problem that have an :hnsoluble quality. Unlike shame, there is no further interference by acute awareness of the 31 self. The intactness of the self in guilt also allows for directing hostility against the self or outward to others in the environment. Finally, Lewis (1971) proposes that hostility against the self arising in guilt takes on the quality of righteous indignation when discharged through the righting of a transgression. In shame, the hostility directed against the self is a humiliated fury that is not expressed outwardly and is discharged ". . . only by repair of the psychic injury to the self" (p. 45). Definitions: Shame and Guilt In summary, the present writer offers definitions of shame and guilt that integrate the above mentioned theories. The basic qualitative differences between shame and guilt can be characterized as a conflict between two different aspects of the superego (ideal self and conscience, respectively). Each are involved with the self. In guilt, the impulse-limiting, transgression—oriented superego function is active with its subsequent fear of punishment, its expected need for retribution, and its negative evaluation of the self via the commission or omission of the act. Shame arises out of the ideal self and its conflict with the real self, manifested as a failure to achieve the idealized self- expectation, subsequent negative self-evaluation for being inadequate, anxiety about having exposed a shortcoming, and fear of a loss of esteem or respect in the individual's own eyes and the fantasized evaluation by others. 32 Shame and Guilt The Self and the Egg_ The definitional confusion about the terms gglf and egg (and ideal self and ego ideal) has been a continual, unresolved concern for personality theorists. Allport (1943) notes that the term egg_has been used in at least eight different senses in psycho- logical theory. Like Allport, Hartmann (1950) suggested the neces- sity of differentiating the self from the ego. Ego, as Lewis (1958) defines it and as this writer shall use the term, is essentially Freudian: ". . . the sybsystem of person- ality which organizes and controls motility and perception, tests reality, and is the seat of defenses" (p. 4). ‘Sglf is defined here by using descriptions provided by Chein (1944) and Lewis (1958). Chein distinguishes between ". . . awareness of the self and knowledge about the larger system of ongoing motives which constitute ego" (cited by Lewis, 1958, p. 4). In referring to the self, Lewis adds that it is ". . . that sense of our own identity which each of us experiences, at times consciously, and more often unconsciously, . . . the unifying tendency in the activities of the organism, . . . the registrant of activities, attitudes and feelings, recognized by the person as his own" (p. 4). Lewis (1958) demonstrates the distinction between the involvement of the ego and that of the self in her discussion of estrangement and depersonalization. She suggests that estrangement is a disturbance manifested as ". . . a strange sense of distance from the self, i.e., 'the me I know,‘ though percepts may continue 33 to be accurate, vivid and rich" (p. 5). Depersonalization exper- iences, on the other hand, suggest ego-based difficulties where ". . . inner feelings of identity are disturbed, although the kinesthetic, visual percepts are again rich and accurate" (p. 5). Further, Lewis suggests that implicit in the distinction between self and ego " . . . is the possibility of discrepancy between the self image and the reality of the ego's function- ing . . ." (p. 4). With reference to shame, she writes ". . . distortion between the self-image and ego functioning leads to under-individuation of the self, i.e., lacking in sense of separate- ness from the environment" (p. S). Conflicts can occur between the self and the ego, the self and ideal self and between the ego and the ego ideal. One might assume that the functioning egoidealxand ideal self in reality, are ideals, just as the terms indicate. They would in pure form reflect the perfect functioning of motile, perceptive, reality testing factors and subsequent high evaluation of the self, that is, ideal self. The Superego Construct As presented by Lewis (1971), the superego construct includes both "impulse controlling” and "ideal attainment" activi- ties. The superego has the goal of maintaining stasis (non-conflict) with the environment. This is mediated by the ego and self func- tions. 'Both guilt and shame occur in the service of the superego to inhibit inappropriate impulse expressions. 34 Lewis (1971) suggests that both moral and nonrmoral shame exist, but only moral guilt. By this she means that guilt is a manifestation of a sophisticated, internalized set of values about attitudes, impulses, and behaviors. Guilt arises as a result of the transgression of proscriptions and prescriptions. These are often related to, but not limited to, sexual and aggressive aims. Psychoanalytic theory maintains that the values arise from defensive identification in childhood, based on factors such as the fear of talion—like punishment in retaliation for transgressions of a taboo, and in the extreme, fear of castration or annihilation. Thus, in the development of guilt the child initially fears physical or emotional punishment, including loss of love by the parents. "If I do transgress I will be punished and lose mother's love." The child comes to live out his or her understanding of the parental values and the parentally defined (and later societally defined) behavioral manifestations of the values. This assumed knowledge in turn becomes established internally, and in the event of transgression or thoughts thereof, the internalized parental image punishes the self for the thought or act. In this way a completely internal system is established that can regulate self- regard in the absence of the actual parental image. Moral shame seems more limited to overexposure of sexual or aggressive impulses, arising from an anaclitic identification or positive modeling after a parent. The threat that exposure generates is of the loss of esteem in the eyes of the significant other and, in the extreme, rejection or abandonment by the beloved other. 35 Non-moral shame, like moral shame, also arises from ana- clitic identification and involves exposure, fear of rejection, and the physical expression of shame. Both can be viewed as inhibitors of impulses, behaviors, attitudes, and needs that are defined by the environment as taboo. Nondmoral shame, however, occurs in the absence of the narrowly defined sexual and aggressive impulses. Shame for succorance needs might be illustrative. Important in both moral and non-moral shame is the psycho- dynamic structure that differs from guilt and prompts a qualitatively different experience. Although often described as inherently involving the presence of the "shaming other" (Singer, 1953), shame, like guilt, can occur without the physical presence of an other. The other may be a partially internalized parental image. szcording to anthropologists,‘the most prevalent fear in shame is that the other might find out about the shameworthy failure. The partial internalization insures that the failure is felt to be known. It is this writer's view that shame, like guilt, requires internalized values and expectations, but that it arises in response to failures and inadequacies rather than in the transgression of limits. Guilt, as anxiety, functions to avoid transgression via sigHal anxiety, and subsequent reinforced repression of the impulse to transgress. It also punishes transgressions as they occur by cutting off narcissistic supplies proportional to the severity of the transgression, and thereby lowers esteem. Shame, as anxiety, acts to insure the attainment of parentally prescribed behaviors and values via signal anxiety and subsequent repression of the 36 impulse or need. Shame punishes failures by temporarily cutting off .all_narcissistic supplies and acutely lowering self esteem. What difference is there between moral guilt and moral and non-moral shame? Guilt is fglly internalized, allowing for the modification of esteem level to occur as a matter of course when there is a transgression. The occurance of guilt also, by defini- tion, assumes the previous attainment of some level of autonomy and a sense of separateness of the self. In the event of transgression, this felt separateness further allows for the assessment, by the self, of the extent of transgression, the amount of retribution necessary, and the extent to which self-esteem and narcissistic supplies will be lowered. In this sense the self is active in its own assessment; the self assesses the severity of transgression and extracts a suitable punishment from itself. Environmental punish- ments interact with this process but are secondary. The guilty person can actively negotiate with himself and the environment (if the environment has been impacted by the transgression), is in control of his or her punishment, and can accept or reject the appropriateness of environmental punishment. Note here, the individual who feels they have "paid their dues" for a crime though a further criminal sentence may be imposed. In cases where the self assesses the transgression as less severe than does the environment, then the self is intact, and able to direct anger at those in the environment who would punish further. There is generally a self-righteous quality in the "righting of guilt," a sense of having expiated the guilt by reparation. The self is 37 intact, and allows for the articulate experience and communication of the guilt to others, further facilitating expiation. Addition- ally, guilt is often situation specific. Shame, on the other hand, while internal, is less well articulated. The route by which the shame mechanism is established (anaclitic identification), occurs earlier than that of guilt, at a time when the self is pgt_experienced as autonomous and separate from the parent. Initially, the parent intentionally or uninten- tionally expresses shaming or labels a child's behaviors, thoughts, impulses, and needs as shameworthy. The child is given the overt message of shame, and covertly or overtly given the expectation that he or she should attain the parental values. The suppression or repression of shamed impulses and needs occurs in the context of the parental disapproval and/or disregard. When internalization occurs, its structure allows for the self to evaluate and regulate its esteem. In addition, the iunternalized powerful parental image is available to the self to re-project onto the environment; this image in turn also evaluates the self. It is the power of the early loved object, and the attribution to it of esteem regulating capa- cities that gives shame its other oriented flavor. The initial fear of the infant or child, in losing the positive regard of the parent by failing to live up to parental expectations, is rejection and abandonment. Coming as early as it does, that is, before autonomy and separateness, loss of positive regard takes on life and death proportions. The initial and subse- quent occurance of shame also bring to the fore intense either/or 38 self-assessments. The self experiences positive regard as either fully present or fully absent. Unlike guilt, in shame the capacity for quantitative and qualitative assessment of the extent of failure is absent. Either full narcissistic supplies or none at all are available. The experience of abandonment and panic.over the feared loss of narcissistic supplies occurs when none are available. The dual evaluation by the self and the internalized other keeps the self under too much scrutiny for the self to assess the situation rationally (as in guilt). Likewise, the attribution of power and the definition of parental ideals as positive, cuts off the possibility of expressing anger over felt injustice from the other. The fear that arises, particularly for the infant or young child, is that as a non-separate extension of the parent, a life sustaining part of the self would be lost or driven further away. In order to ensure survival, the individual redirects the hostility against the self, and views the self as inadequate in order to maintain the illusion of having a loving parent. At least in the early life of the child the parents define those attitudes, behaviors, impulses, and needs that are acceptable and those that are taboo. Large differences in the type and number of taboos occurs because of the variety among parents. This writer suggests that in some cases the inhibition, both by guilt and shame induction, may prompt the development of an excessively punitive superego, in both impulse controlling and ideal attainment aspects. Further, both moral and non-moral shame arise from and are a res- ponse to the self (and early expressed other) expectations of ideal 39 attainment, based on anaclitic identification. This leads not only to the continual need to achieve and attain, but to fear the failure which would generate shame, and abandonment. Such early experienced orientation to the external world for self-satisfaction encourages the continued salience of the external (the parent and parent sym- bol, later manifested as a generalized other), and thus the potential occurance of shame. While Lewis (1971) makes good use of the differentiation between moral and non-moral shame, she remains attached to the narrow, psychoanalytically defined viewpoint of the negativity of sexual and aggressive expression. This writer takes issue with these narrow delineations and sees it as beneficial to view "sexual" more broadly, as "alive." That is, a sexual organism is energized; it acts upon and reacts to the environment. Likewise, "aggression" may be viewed as the assertion of a tendency to move toward need and impulse satisfaction. Within this perspective, any a§g_many impulses, needs, behaviors, feelings and attitudes may be shamed. If shaming is extensive, there can be extensive inhibition, hampering of life and behavior choices, and neurotic symptoms. Lewis (1971) discusses the tendency of undischarged shame to develop into depressive symptoms, and undischarged guilt into obsessive symptom patterns. Perlman (1958) has also explored the qualitative differences in guilt and shame and their implications. He developed the Atti- tude Anxiety Survey (AAS) to measure the component affects associated to the experience of anxiety. Drawing from the earlier writings of 40 Alexander (1938) and Piers (1953), he assumed that shame anxiety and guilt anxiety are two aspects of anxiety. He validated the questionnaire on a sample of 64 first year medical students. The purpose of his study was to assess the extent of relationship between shame and evaluation anxiety (measured by the Mandler Sarason Questionnaire on Attitudes Toward Three Types of Testing, 1952). He hypothesized that shame anxiety and evaluation anxiety would show a positive association. For the overall sample, significant support for this hypothesis was lacking. Intensive assessment of six subjects who scored extremely high on guilt or on shame pro- vided support for the ability of Perlman's questionnaire to identify an individual's vulnerability to guilt anxiety or shame anxiety. Individuals who reported high responsiveness to shame items also expressed, in intensive interviews, strong concerns about issues involving inadequacy and failure. Individuals highly responsive to guilt items reported concerns that centered around aggression and its control. These findings are consistent with the earlier dis- cussed formulations of Alexander and Piers regarding shame and U guilt respectively. Review of Previous Research Several areas of research are relevant to the shame and guilt experiences. The present writer will discuss the constructs of psychological differentiation, defense mechanisms, incidental learning, and sex role endorsement, and present research in each 41 area. This research lays a foundation for hypotheses of the main study. Psychological:Differentiation and Personality Characteristics Lewis (1971) and Binder (1971) have both discussed the rela- tionship of shame and guilt to aspects of psychological differen- tiation, hypothesizing that shame occurs within a less differentiated, and guilt within in a more differentiated psychological structure. The differentiation construct has been characterized as "a difference between people in the ease with which they are able to maintain their orientation in space" (Lewis, 1971, p. 127). Originally formulated by Wertheimer (1938), this construct suggests the self as an "egotistical product" or a product of interaction with the "field." Witkin (1950) has suggested that the differing capacity for perceptual-cognitive discrimination is associated with relatively consistent differences in personality characteristics and styles. He uses field dependence-independence as an operation- alization of this perceptual-cognitive dimension. In turn, this construct is subsumed under the more general construct of global- analytic cognitive functioning. Witkin (1954, 1962, 1965, 1969) suggests that the extent to which a person is perceptually dependent on the field for information (field dependent) also reflects depen— dence on external information in making judgments in general (global functioning). Analytic functioning is associated with a greater degree of psychological differentiation, including a well differ- entiated self and less reliance on external information for 42 judgments. Global functioning reflects a less separate sense of self, implying that global people tend to see themselves fused with other people and with the environment generally. Witkin's construct of psychological differentiation encompasses cognitive functioning, functioning of the self, and defense mechanisms. A frequently used measure of perceptual-cognitive differ- entiation is the Rod and Frame Test (Witkin, 1949), in which the individual must discriminate the verticality of an illuminated rod within an illuminated non-vertical frame, within a darkened room. The extent to which verticality is attained is considered a measure of field independence, that is, the ability to use one's internal cues to assess the body's verticality and that of the rod. This in turn implies an independence of the body and the self from the environmental cue of the frame. The individual who is relatively inaccurate is said to have less firmly established self boundaries and, thus, is unable to separate himself from the dominant but misleading environmental cue. Lewis and Witkin (1958) indicate that extreme accuracy and inaccuracy in discriminating the rod from the frame (more generally, the figure from the ground) denotes a self that is over- differentiated or under-differentiated. Extreme differentiation allows for precise discrimination but may be associated with a rigid, wholly analytic style of functioning, with little capacity to experience "union" with the environment (see previous discussion of Bassos). Conversely, poor discrimination denotes difficulty in 43 in establishing a sense of self that is separate, autonomous, and internally directed. Witkin (1969) has found that physical and psychological separateness develop with a movement away from unity with the mothering agent, and that the development of the analytic perceptual style is contingent upon the separation. Witkin's findings, suggest that mothers of little boys later measured as global tend to lack definition in their own roles, implying some interference with the development of separateness in their child. Conversely, more psy- chologically differentiated mothers tended to encourage an indepen- dent sense of self in their child. These boys, in turn, later showed a more articulated (analytic) cognitive style. Seder (1952), Dawson (1963, 1967), and Berry (1966a & b) confirm these findings with varying populations and cultures. There is a tendency for perceptual discrimination to increase with age. As measured by the Rod and Frame Test, but Witkin, Goodenough and Karp (1967) have found that individuals who are global at age eight remain relatively global when retested at age 22, suggesting a relative stability of perceptual-cognitive style over time. Likewise, global, field dependent individuals tend to be more influenced by the judgments of others on autokinetic and Ash-group pressure tests (Sherif, 1946; Linton, 1955; Allen and Crutchfield, 1963). Sanguliano (1951) demonstrated that field dependence and suggestibility were positively correlated in a female psychiatric {natient group. Rudin and Stagner (1958) suggest that 44 field dependent individuals are more likely to be malleable in their self views and change them.when asked to describe themselves in different contexts. These findings suggest a complex interaction of perceptual- cognitive style with personality characteristics. . . . One of the main characteristics of differentiation is specialization of function; another is a clear separation of the self and not self. Integration refers to the func- tional relations between parts of a system. At any level of differentiation, integration is possible, although more complex integrations would be required at increased levels of differentiation. Adjustment of the whole personality may be thought of as reflecting its integration. It is to be expected that impairments of integration or adjustment take place at any level of personality differentiation. The form of impairment or pathology would be expected to vary with level of differentiation. (Lewis, 1971, p. 131) Diffepgntigtion Research Relevant to Shame and Guilt Lewis (1971) has discussed the relationship between psycho- logical differentiation and the experiences of shame and guilt. Consistent perceptual, cognitive and defensive styles seem to be .~,._. ~ manifested by shame experiencing and guilt experiencing individuals. ‘In her work with Witkin, et al., (1968) she assessed field indepen- dence, measured with the Rod and Frame Test, and self/body boundar— ies, measured as the articulation of body drawings (DAP). The results indicated that those individuals with highly diffuse, non- articulate self drawings with porous body boundaries tended to be more field dependent. Likewise, their verbal productions in the initial sessions of psychotherapy were heavily dominated by reports of shame oriented experiences. Shame or guilt orientation was 45 measured by the Gottscalk method of scoring implied affect (1961, 1969). Verbal productions that suggested fear of abandonment were used as operationalization of shame, based on Piers' (1953) concep- tualizations. Additionally, field independence and articulated body drawings with solid boundaries were positively correlated with guilt oriented verbal productions. Guilt affect was judged as the pre- sence of castration concerns. Lewis suggests, on the basis of these findings, that the experience of shame occurs within a self organization pattern of field dependence and porous self boundaries. The self organization associated with guilt involves field independence and solid self boundaries. The porosity of the body boundaries and the inability to extract a figure from a ground may be indicative of a self inadequately differentiated from the environment. Under- differentiation is a characteristic of theoretical descriptions of ‘the experience of shame. Conversely, excessive articulation of body and self boundaries and extreme sense of separateness is thought indicative of over-differentiation and is consistent with the theoretical descriptions of guilt. These findings also support the view that shame experiencing individuals attribute self-esteem regulating capacities to the '"“d'-I-' ..—_.fi.-__.— *7 environment, and they are consistent with the previously discussed view that shame arises from thepre—autonomous, anaclitic identifi— cation patterns of early childhood. Liekwise, the dependence on 46 external environmental cues for infarmation and affirmation, supports the theoretically described difficulty with the discharge of hos- tility. In an attempt to illuminate the relationship of shame and guilt with other consistent characteristics of individuals, Binder (1971) described the relationship of shame and guilt (as measured by a variation of the Perlman, 1958, scale of shame and guilt), with attention deployment (Mayman and Voth, 1967; Bush, 1967: Reality Close-Reality Distant Scale), self-ideal self discrepancy (Mayman, 1953: Sense of Self Inventory), the tendency to falsely present oneself in a positive way (MMPI, K scale) , and defensive cognitive-perceptual style (leveler-sharpener dimension). There was a significant correlation between the extent of awareness of self-ideal self discrepancy and shame or guilt separately, for women only. Running counter to Binder's hypothesis, women with a high tolerance for self-ideal self discrepancy tended to experience shame; women with a low tolerance for self-ideal self discrepancy tended to experience guilt. Consistent with his hypotheses shame was found to be related to a leveler cognitive and defensive style, and guilt to a sharpening style. Correlations supported the view that individuals who have a low tolerance for self-ideal self discrepancy and who have high external attention deployment, tend to falsely present themselves in a positive manner. Conversely, high tolerance for discrepancy and high internal attention deploy- ment subjects tended to present themselves in a negative manner. 47 The present writer would expect that people with a high tolerance for self-ideal self discrepancy would more likely exper- ience shame, due to their awareness of their own shortcomings. Although Binder found a positive relationship between shame and leveling cognitive style, the present writer will hypothesize that individuals who expereince intense shame will have a more heightened awareness of their environment and their self-functioning. Lewis (1971) and Binder (1971) also theorize that individuals who modally experience shame use repressive defenses, and individuals who experience guilt use sensitizing defenses. The present writer disagrees with these formulations and will test hypotheses that reflect a reversal of the above theory regarding shame. Research on Defense Mechanisms Relevant to Shame and Guilt Witkin (1950) posits differences in the cognitive-perceptual style, functioning of the self, and defensive style, as a function of the extent of differentiation. These differences can be use- fully described as the extent to which there is a reliance on exter— nal or environmental information. This is one operationalization of the global-analytic cognitive functioning construct. Two specific areas that can be used to assess this construct are differences in defensive style (a measure of defensive perception), and more generally, differences in attention deployment (a measure of the style of perception, both defensive and non-defensive). This writer has focused primarily on the repression-sensitization dimension of 48 defensive perception presumably demonstrative of global and analytic cognitive styles, respectively. The construct of repression-sensitization arises from theo- retical and research literature concerning individual differences in defense style (see Byrne, 1961; Eriksen, 1968). Repressors are characterized as individuals who tend to try to avoid recognizing and reporting anxiety inducing stimuli. Sensitizers are character- ized as defending by approaching anxiety inducing stimuli, such as by showing perceptual vigilence in perception studies. The sensiti- zer tends to use isolation of affect (Bertini, 1961; Benfari, 1966; Gardner, 1962), and intellectualization (Weiner, Carpenter and Carpenter, 1957), or, more neutrally, heightened recognition, as a defense style. Repressors typically use avoidant mechanisms, such as denial, repression, or, more neutrally, lowered perceptual reports or recall of information (Mischel, Ebbeson, and Zeiss, 1973). Repression and sensitization are not considered mutually exclusive mechanisms. It has been noted that in response to neutral words, neither group may employ defensive perception. However, a basic perceptual style which some assume tounderlie the defense style is in fact measurable with neutral stimuli (Markowitz, 1969; Klein, 1954). A sensitizing attention style may prompt an ongoing aware- ness of negative self qualities, in turn maintaining a negative self-view. The present writer proposes that this may be the template on which recurring proneness to an experience of shame may be based. 49 Studies of attention style essentially attempted to denote what is perceived in the environment. Two Operationalizations of attention style are relevant to the present research. These include the relationship of attention style with the use of sensitizing or repressing defenses and the relationship of incidental learning as a measure of attention style with defensive style. Luborsky, et al., (1965) have studied the breadth of scanning of environmental stimuli and have shown that individual differences in scanning do exist, and that as an aspect of attention deployment, they define the attentional range. Narrow scanning of the stimulus field was correlated with the perceptual defense of repression, and broad scanning was correlated with sensitization. The researchers theorized that limiting the range of what is seen reduces the proba- bility of receiving inputs that may threaten the established repres- sion or denial. Conversely, a broad scanning range facilitates the sensitizer's vigilance, providing information of conflictful situations. Incidental learning paradigms can also be used to measure individual differences in the breadth of attention deployment. Incidental learning is that learning that occurs in the absence of formal instructions to learn. Structurally, one paradigm presents an instruction for the individual to learn one set of words (focal task) and then tests the individual for knowledge of other words that were present (peripheral task/incidental task) during the focal task. The variation of this paradigm that will be used in the present study was developed by Mendelsohn and Griswold (1964). 50 Subjects are asked to memorize words on a typed word list (focal list) while a tape recording presents other words (peripheral stimuli). The subjects are tested later on an anagram task. Some of the anagram solutions were included in the focal list and some in the peripheral stimuli. The extent to which the subject is able to do the anagrams whose solutions were cued focally or peripher- ally, is the measure of incidental learning. Using this model. Mendelsohn and Griswold (1966) found that poor performance on anagram solutions is correlated with repression, as measured by the MMPI R scale (repression, denial). Differences in attention deployment of sensitizers and repressors seem to be accentuated under threat. Markowitz (1969), in a study of incidental learning, showed that sensitizers looked around more, responded with increased vigilence, and had a higher recall of incidental stimuli under a high threat condition. Re- pressors, who tend to look around little responded with even narrower attention under threat. More broadly, this suggests that manifest anxiety may facilitate effective responses or, depending on the environmental task, impair it. Sarason (1972) suggests that individuals differing in test anxiety also differ in the degree to which they attend to environmental stimuli and how they use cues in problem solving. The effects of threat on defensive perception and attention style was also assessed by Mischel, Ebbeson and Zeiss (1973). They exposed groups of sensitizers and repressors to personal feedback material labeled either as assets or as deficits and recorded time 51 spent with the feedback. Sensitizers who had been previously told they had done poorly on a task tended to spend significantly more time reading negative feedback than positive feedback. The sensi- tizers in the control condition also spent more time on their liabilities than did repressors; sensitizers spent much less time on their as sets than did repressors. The apparent consistency of these findings suggests that the sensitizing mode acts as a selective attending device that may prompt awareness of, interest in, and responsiveness to, negative feedback from the environment in the absence of clear positive environmental cues. Theories previously discussed suggest that heightened awareness of negative cues from the environment is an active factor linlthewshame experience. From the tradition of research on percep- tual defense, Dulaney (1957) points out that perceptual vigilence occurs when one's perceptual response is instrumental in the avoidance of punishment and competing responses are punished; repressive modes occur when one's perceptual response is responded to with punishment. Byrne (1964) hypothesized that the repressive orientation would arise in a restrictive family, in which the child's expression of impulses are punished; a sensitizing orien- tation would arise in a permissive family, in which the impulses were allowed expression. His own data suggest a reversal of prediction, with sensitizers reporting more restrictive home backgrounds and repressors, more permissive home backgrounds. Donelson (1973) comments that the vigilence of the sensitizer may 52 be adaptive as a response to the sternness of the parental restric- tions, whereas the repressor may not have sensed as clear a need for vigilence. Parents with whom vigilence is required may also be likely to provide shame inducing experiences for the child. Along with others, the present writer views threats to the esteem of an individual as leading to defense, of which perceptual vigilence and acute awareness of the potential threat of the environ- ment, may be part. These factors have been previously discussed as an integral part of the shame structure. This selective review of research prompts the present writer to propose that there may be a relationship between attention deploy- ment, defensive perception, and shame and guilt affects. A more elaborate review of attention and defense research is beyond the scope of the present paper, but it is available in Eriksen (1968) and Silverman (1970). The present study will combine the Mendelsohn and Griswold incidental learning paradigm and the Mischel, Ebbeson and Zeiss method and relate behaviors observed within these paradigms to repression-sensitization, as measured by the R-S scale, and to shame and guilt variables. Sex Differences in Perceptual Style and Their Relationship to Shame and Guilt Several factors prompt a discussion of the interaction of sex differences, perceptual style, and shame and guilt. The extent of separateness is theoretically relevant to the experience of shame and guilt; the non-separate individual is predicted to be prone to 53 shame, and the more separate individual to guilt. Traditionally, it has been assumed that women are less psychologically separate than are men. Lewis (1971) suggested that women experience more shame than do men, and women experience less guilt than men. Hoffman (1975) reports data suggesting that males show greater concern with detection for wrongdoing, than do females. Fear of detection can be viewed as more an aspect of the shame experience, than of the guilt experience. Thus, there seems to be an inter- action between pairs of these factors, though no empirical study has shown a relationship almong all the variables treated together. Several theoretical approaches suggest that there should be sex differences in these areas. Bakan (1966) discusses differences in agency and cnxnmunal orientations. Self-protection, self- assertion, and self-expression are attributed to agency. A sense of "at oneness" with the communal whole characterizes communion. Freud's postulated self versus species preservation tendencies are paralleled by Bakan's concepts. Similarly, the Logos and Eros princi- ples, delineated by Jung parallel the antitheses of separateness and communion. Logos attributes, marked by compartmentalization, dis- crimination, analytic thinking, independence, and self-sufficiency, are said to be characteristic of the masculine principle. The Eros attributes of subjectivity, passivity, tender mindedness, sen- sitivity and dependency are viewed as characteristic of the feminine principle. Eros binds together what Logos separates. 54 Gutmann (1965, 1970) discusses the differences in auto- centric versus allocentric ego styles. The autocentric style is marked by a sense of union and interaction with the external and non-self aspects of the environmental stimuli. The allocentric style is marked by a separateness from nonrself aspects of environ- mental stimuli and a view that that which is external to the self had a logic of its own. With more permeable ego (self) boundaries, the distinction between the self and other (environment) is not as sharply made, the self being more easily a part of the stimulus (Donelson, 1977b). Guttman (1970) notes that the work world conven- tionally inhabited by men, is an impersonal one in which firm ego boundaries between self and others are adaptive in using others in the pursuit of ones own goals, for gel: preservation. In contrast, firm ego boundaries would be disruptive in the self-extending habitat assigned to women, where the dominant task is more one of species preservation. Taken as a whole, these theoretical formulations propose differences in perceptual attention styles, and maintain to varying extents that self-extending qualities are attributable to females, while self-preservatory qualities are attributable to males. Significant sex differences can be found in some tests of perceptual-cognitive style, as represented in studies of sensory threshold, responses to strong stimulation, responses to discrete segments versus whole configurations, awareness of subtle differ- ences in configurations, attention versus inattention to interfering irrelevant stimulus cues (see Silverman, 1970). Witkin cites cross 55 cultural findings that suggest support for females being more field dependent than males during some life periods. Lewis (1971) has found perceptual-cognitive style sex differences on the Embedded Figures Test and posits that a very close relationship exists between sex differences and perceptual style. . . . the factors which generally result in difference in perceptual style, should be the same factors that result in sex differences in perceptual style. (Lewis, 1971, p. 145) Donelson's (1977a) discussion of conventional feminine socialization points out the extensiveness of the requirement of social responsiveness for females in this social structure. The emphasis on self-extending qualities exacts a toll from the female in ttre form of a lessened sense of separateness. The present writer views the Donelson discussion as adding important elaboration to Lewis' position in that it acknowledges the contribution of sex role training to sex differences, and explicitly recognizes within-sex differences based on the extent to which traditional sex role consistent behaviors are endorsed. Recent research and criticism suggests caution in generaliz- ing from sex difference findings, due to a lack of consistency across studies (see Maccoby, 1966; Maccoby and Jacklin, 1974). All findings, therefore, must be considered in this light. However, Silverman (1970) in his survey of findings suggests that sex differ- ences do exist, and variation or deviation from these (i.e., within- sex variation) occurs as a function of "moderator variables," including anxiety level, level of ego strength, defensiveness, and 56 masculinity-femininity. The present writer suggests that the number of recognized moderator variables, their vmidespread presence in this culture, their pervasive influence on personaltiy, and their overall level of importance, makes difficult the isolation of any existing biologically inherent sex differences. Of particular concern is Silverman's relegation of masculine and feminine sex roles to a "moderator" status. The present writer disagrees with this position and emphasizes the importance of the sex role training factor in the development of what heretofore have been considered inherent sex differences (e. g., in areas of perceptual-cognitive functioning, extent of separateness of self, and shame and guilt affects). Sex role modeling, regardless of biological sex, within the masculinity-femininity dimension impacts the individual's cognitive style, and prompts a questioning of assumed inherent biological sex differences. The present writer does not negate that relevant biological sex differences may exist, but points out the importance of sex roles and the socialization experiences typically related to sex differences. Barclay and Cusumanos (1965) have found that field depen- dence in boys tends to increase as a function of father absence. Bieri (1960) found that father identified women tended to be field independent. For men, the same finding was apparent though it was not significant. Milton (1957) has found that high masculinity or femininity scores inconsistent with biological sex were associated with the perceptual style and problem solving approach typically used by 57 the other sex. Vaught (1965) also shows that feminine females and males show poorer performance on the Rod and Frame Test than do masculine males and females. Regarding sex role training, there has recently been a questioning of the traditionally bipolar view of masculinity and finininityas asingle dimension. Constantinople (1969) suggests that it is more useful to view them as separate dimensions. Some indi- viduals show a mixture of traditionally stereotyped masculine and feminine qualities. Bem (1975) hypothesizes that androgyny (rela- tively equal masculine and feminine scores) allows for behavioral adaptability inhibited in either consistent (same sex) or inconsis- tent (other sex) sex role stereotyped endorsements. She demonstrated that masculine and androgynous subjects conformed less to peer pressure than feminine subjects, and that feminine and androgynous subjects showed greater nurturance than did masculine subjects. In the present writer's research, the effects of sex roles is being explored in relation to both field independence and shame and guilt experiential orientations. As Lewis (1971) discussed, females are more likely to be shame oriented and field dependent; males are more likely to experience guilt and be field independent. It is the present writer's view that sex role commitment and not necessarily inherent biological sex, is associated with the cognitive-perceptual orientations, and shame or guilt. Masculine, feminine, and androgynous subjects of each sex will be compared on field independence, sensitization, and shame or guilt affects. STATEMENT OF RESEARCH PROBLEM The previously described theory and research seem to provide reason to claim that shame is associated with field dependence, and the use of sensitizing defenses, and is more prevalent for females or males who endorse feminine sex role characteristics. Guilt is assumed to be associated with field independence, and is more pre— valent for males or females who endorse masculine sex role char- acteristics. Lewis (1971) rated the early session psychotherapy tapes (using the Gottscalk System for Evaluation of Implied Affect), administered the Witkin (1950) Embedded Figures Test and Draw a Person Test to eight out-patient psychiatric subjects. She found that diffuse ego boundaries (measured as field dependence) is related to the experience of diffuse affective states (1. e., shame). fro this she adds that: shame is typically greater for females; the anger arising in shame is discharged against the self rather than against the external environment; shame is related to the use of repressive defensive style (e. g., repression and denial). Conversely, firm ego boundaries (measured as field inde- pendence) is found to be related to the experience of more limited and bound affective states (e. g., guilt). She also suggests that: guilt is more typical for males; the anger arising in guilt is 58 59 usually discharged against the environment rather than the self; guilt is related to the use of sensitizing defensive style (e. g., intellectualization). Several factors led the present writer to re-assess the Lewis (1971) findings. One reason for reevaluation was the fact that a psychiatric sample was used in her study. The present writer felt that a nonrpsychiatric sample should be tested in order to report results that were more typical of the general population. Secondly, Lewis' sample was a small group (n - 8) of intensively studied cases, and it seemed appropriate that a larger sample should be tested in this study. The present study employed objective measures of shame and guilt affects, rather than using tape ratings, and administered objective measures of all the other variables of interest. The shame and guilt scale employed was a modified version of the Perlman (1958) Attitude Anxiety Survey (see Test Instrument Section) that was restructured to provide male, female, and self reflective forms. Although Lewis does not view shame and guilt as mutually exclusive, the focus of her research emphasizes the bipolar quality of these affects. In turn there is emphasis on the bipolarity of the differentiation construct, defensive style, and sex role endorse- ment, in relation to shame and guilt. The present Pilot Study tests some of Lewis' contentions, based on this bipolar view. Thus, hypotheses in the Pilot Study use a bipolar conceptualization of shame and guilt as they relate to differentiation (measured as field independence-dependence), and to sex role endorsement. It 60 is expected that there are relationships between shame and feminin- ity, and between guilt and masculinity; negative relationships are expected between shame and masculinity and between guilt and fimdninitfi Two final hypotheses concern the relationship of differ- entiation with sex role endorsement. However, there are problems with some bipolar concepts. In the Main Study, the method uses a technique by which subjects are grouped into classifications based on their shame and guilt scores separately. The instrument selected can obtain independent scores on each affect. Median cutoff points were established for the sample on each variable, allowing for a designation of subjects as above or below thezmedian value on both shame and/or guilt. This allowed for the separation into four groups of high or low scores on both shame and guilt. The complexity of the Main Study design was thus increased but this method acknowledged shame and guilt as separate dimensions, and allowed for the assessment of the relationship between the level of shame and guilt separately, and the individual's level of differentiation, attention style, defensive style, and sex role endorsement. The extent of differentiation and defensive style were each assessed by measures that reflected bipolar aspects of a single dimension. Masculinity and femininity were treated as separate dimensions. According to Lewis (1971), the experience of shame is related to a self structure that is diffuse (measured as field dependence and a lack of ego and body boundaries). Conversely, 61 guilt is related to a self structure that reflects strong body and ego boundaries and demonstrates field independence. These hypo- theses are tested here using a larger, non-psychiatric sample, evaluated with objective measures. It is hypothesized that: l. Shame scores (on the AAS described below) will be negatively associated with closure flexibility (field independence). 2. Guilt scores will be positively associated with closure flexibility (field independence). Based on the assumption that males and females endorse fully what has typically been described as sex role consistent behaviors, Lewis asserts that males would be more familiar with the guilt experience, and females with the shame experience, based on her earlier described theory and research. The present writer fully supports the importance of the recognition of sex role training, but questions two assumptions inherent in Lewis' position. It cannot be assumed that all males and females, because of biological sex, have endorsed traditionally sex role consistent behaviors to the same extent. Secondly, her assumptions are based on a bipolar view of masculine and feminine sex roles. This viewpoint has been reassessed by a number of researchers (Constantinpole, 1969; Bem, 1974; Spence, Helmreich and Stapp, 1975) who have found it useful to view masculinity and femininity as separate dimensions rather than polar opposites of the same dimension. This has allowed for the acknowledgment of the value of developing characteristics pre- viously seen as appropriate for only one sex. Additionally, this 62 conceptualization facilitates the consideration of within-sex differences in sex role endorsements. The present writer posits a relationship between sex role endorsement and shame and guilt affects. The self-extension quality, and capacity for union with the environment, expressed in the feminine sex role, is viewed as facilitative of the shame experience. This occurs due to a theoretically greater receptivity to environr mental information and assessments for those individuals (females and males) endorsing femininity. Receptivity to environmental assessment has been discussed earlier as an integral part of the shame experience. Conversely, the self-survival quality and required separateness of the masculine role theoretically leads to a greater familiarity with situationrspecific guilt affect, facili- tated by a heightened sense of separateness from the environment in both males and females. The research does not view masculinity and femininity or guilt and shame as bipolar, but will test two hypotheses that are correlation based and reflect a bipolar view of these variables, in order to test Lewis' theory. Additional hypotheses (5 and 6), tested only in the Main Study, allow for the evaluation of an association between sex role endorsement viewed as separate dimen— sions and the level of shame and guilt. Therefore, these hypotheses seem reasonable. They are that: 3. Shame scores will be positively associated with femininity and negatively associated with masculinity in both males and females. 63 4. Guilt scores will be positively associated with masculinity and negatively associated with femininity in both males and females. 5. For high shame subjects, feminine scores will be greater than masculine scores. 6. For high guilt subjects, masculine scores will be greater than feminine scores. Witkin (1969) presents findings of sex differences in field dependence/independence showing that females are more field dependent than males in some age spans. The present writer, with Witkin, suggests that sex role training includes sex role consistent perceptual-cognitive training. Therefore: 7. Closure flexibility scores (field independence) will be associated negatively with feminine sex role endorsement in females and in males. 8. Closure flexibility scores (field independence) will be associated positively with masculine sex role endorse- ment in males and in females. Lewis suggests that the level of differentiation, defensive style and shame and guilt affects are highly interrelated. She theorized a relationship between field dependence and the use of repressive defenses. Conversely, field independence is said to be related to the use of sensitizing defenses. Yet, Binder (1971) found that it was those subjects who were aware of (or sensitized to) a self-ideal self discrepancy who experienced shame, though on some perceptual tasks they were labeled levelers rather than Sharpeners. The term sensitization is being used here to represent the extent of awareness of the environment (1. e., broad vs. narrow attention deployment), as well as in the more limited definition as 64 a defensive perceputal style that arises in response to internal or external threat. Operationally, we will test (in the Main Study) the overall level of sensitization by measuring the extent of atten— tion deployment in personal feedback reading and feedback recall, and in an incidental learning paradigm. The narrower defense style- based definition of sensitization will be tested with the Repression- Sensitization Scale, RPS (Byrne, 1961). The.Main Study will assess selective attention deployment (as measured by time spent reading positive and negative personal feedback) for high shame and high guilt subjects, after experimen- tally manipulated success or failure experiences. It is expected that high shame subjects will demonstrate greater interest (sensi- tization) in negative feedback (measured as time spent reading feedback and recall of feedback) than in positive feedback, and that this relationship will be greater than for high guilt subjects in both success and failure conditions. The relationship between sensitization and the use of incidental environmental information will be assessed by the use of a modified form of the Mendelsohn and Griswold (1964) incidental learning paradigm. Finally, recall of positive and negative trait feedback received after success or failure experiences will be assessed with a trait recall inventory. It is hypothesized that: 9. Shame scores will be positively related to sensitization, as measured by the R-S Scale; this relationship will be stronger than for guilt scores and sensitization. 10. High shame (low guilt) subjects will spend a greater amount of time than high guilt (low shame) subjects reading nega- tive feedback, under both success and failure conditions. ll. 12. 13. 65 High guilt (low shame) subjects will not differentially attend to positive or negative feedback under success conditions. After failure, both high shame (low guilt) and high guilt (low shame) subjects will attend more to negative feedback than positive feedback; however, the high shame (low guilt) subjects will attend to negative feedback more than will high guilt (low shame) subjects. High shame (low guilt) subjects will perform better on an incidental learning task than high guilt (low shame) subjects. High shame (low guilt) subjects will have greater recall scores about feedback material after success experience than after a failure experience; this will not be true for high guilt (low shame) subjects. PILOT STUDY A Pilot Study was designed to assess the feasibility of using shame and guilt variables for subject classification and allowed for the re-evaluation of the item homogeneity and test/retest reliability of a modified version of the Attitude Anxiety Survey, AAS (Perlman, 1958) and the item homogeneity reliability of the Bem Sex Role Inventory, BSRI (Bem, 1974). A preliminary testing of hypotheses 1-4, 7 and 8 was also carried out. Subjects were 93 student volunteers enrolled in introductory and advanced psychology classes at Michigan State Univeristy. In order to assess the test-retest reliability of the AAS instrument, two sessions were scheduled. Two forms of the shame- guilt instrument were given to each subject in each session. The BSRI was also administered in the first session. The Closure Flexi- bility Test, CFT (Thurstone and Jeffrey, 1956), a test of field inde- pendence was administered in the second session. Subjects completed group forms of these inventories. Testing Instruments Original Attitude Anxiety Survey (AAS) The Attitude Anxiety Survey, AAS, was developed by Perlman (1958) to delineate two component affects associated with the exper- ience of anxiety. Based on the earlier writings of Alexander (1938) 66 67 and Piers (1953), he proposed that two affective aspects of anxiety, shame and guilt anxiety, could be isolated and assessed. The AAS is an objective paper and pencil measure of both the relative proneness to guilt and shame and the intensity level of each. The test consisted of 54 statements which the subject rates along a nine point continum "howanxious you feel most people would be were this (the situation) to happen to them." The situations were described by statements such as, "John becomes aware that he has mistreated a friend," and "Frank belches in public" (See Appendix A). Scoring yielded four separate, but non-independent socres: guilt anxiety score; shame anxiety score; total anxiety score (guilt plus shame score); difference score (difference between the number of guilt items and shame items rated above the subject's own median rating for the two7types of items combined). The median based difference score is used as the measure of relative proneness to shame and guilt. For the sake of brevity, this index will be referred to as the relative score. Perlman (1958) administered the items to a group of 64 male, first year medical students on two separate occasions. Face valid- ity of items was established by using only those items on which three training psychoanalysts and a majority of 11 clinicians agreed reflected the definitions of shame and guilt. His estimates of reliability were based on a combined internal consistency as test- retest method and yielded coefficients of .80, .83 and .35 for guilt 68 score, shame score, and relative score. Test-retest reliability of the total anxiety score was .71. Revised Attitude Anxiety Survey Several changes in the AAS seemed worth investigation. First, to facilitate responsiveness for both females and males, feminine gender names were used on a female form. Slight alterations of phrasing were made in two of the original items (items 42 and 50) for which the situations seemed unlikely to occur for a woman. Situations were matched in intensity to the male form. When it was possible, in other items (items 7, 12, 15, 24, 32, 44) situations were changed to a sex role neutral situation of equal intensity as the original (male) one. Item changes were made and agreed upon by the experimenter and project chairperson. See Appendix A for changes. Additionally, a third form was created with a subject or self-orientation. The instructions asked for the subject's own response to the situation; statements of "he" were changed to "you." The rationale for this form was that it might elicit more immediate personal affective responsiveness. It posed the risk, however, of prompting defensiveness, as well. The relative advantages of the MM -_ Self and Othef instructions were considered an empirical question. Finally, for the sake of scoring convenience, the continuum was extended from a nine point to a ten point scale, ranging from "not at all anxious" (l) to "extremely anxious" (10). 69 Ben Sex Role Inventory (BSRI) The Bem Sex Role Inventory, BSRI (Bem, 1974) was the measure of sex role endorsement. The BSRI differs from earlier sex role inventories in several important ways. First, the items were selected by male and female judges as being more socially desirable for women or for men. Most previous inventories selected items on the basis of differential endorsement by males and females and often included more desirable items scored as masculine. Second, masculinity and femininity were treated as two continua rather than as one bipolar dimension. In addition, the scale contained items which were neutral with respect to sex roles, and provided a measure of social desirability set. Bem selected twenty items for each of three scales (mascu- line, feminine, and social desirability) out of a pool of 400 personality characteristics, on the basis of ratings by 100 student judges on the appropriateness of the characteristic for men or for women. A personality characteristic was labeled masculine or feminine if it was independently judged by both male and female judges as significantly (p < .05) more desirable for one sex than for the other. Characteristics were considered neutral if they were judged by both male and female judges to be no more desirable for one sex than for the other (t - 1.2, p < .2). Ten positive and ten negative sex role neutral personality characteristics were selected; these provide a measure of socially desirable response bias. 70 Four scores were obtained: masculinity (M), femininity (F), social desirability (SD), and androgyny (A). Bem's androgyny score was originally defined in terms of t for the F-M difference distribution for each subject; the smaller the statistic for any subject, the greater the androgyny. Bem states that a simple differ- ence score (F-M) is highly correlated (r = .98) with the estimated androgyny score (1. e., the FAM mean difference score multiplied by 2.32). Bem (1975) has recently recognized the value of a different operationalization of androgyny, as will be discussed later. Bem's test-retest reliability coefficients for a four week interval were quite high: Masculinity, r a .90; Femininity, r = .90; Androgyny, r - .93; Social Desirability, r = .89. BSRI scales were not correlated with the Guilford-Zimmerman Temperment Survey (masculine-feminine scale) and had a low positive correlation with the feminine scale of the California Psychological Inventory (1962). Bem (1974) suggested that the lack of high correlation between the BSRI and these other tests indicated that the BSRI measured aspects of sex roles which were not directly tapped by the other tests. Closure Flexibility Test (CFT) The Closure Flexibility Test, CFT (Thurstone and Jeffrey, 1956) was designed to measure what Thurstone (1944) identified as the "second factor,‘ also called the reasoning function, consisting of the ability to hold a configuration in mind despite distraction and to see a given configuration (diagram, drawing, or figure) 71 which is "hidden" in a larger, more complex drawing, diagram, or figure. This Thurstone second factor arose in a large project on perception. Included in the testing materials was the Gottscaldt Figures Test (1929). The Gottscaldt Figures Test, on which the CFT is based, essentially tests two capacities. Thurstone's second factor reflected one of the capacities of the Gottscaldt figures, namely, the ability to form and to hold a perceptual closure despite distraction and to form the closure in the face of distraction. Thurstone (1944) found that the two factors delineated by Gottscaldt are relatively highly correlated (.38). The CFT evolved from this original base with modification: a time limit (Bechtold, 1947); paper and pencil format; and a requirement for subjects to check those figures in which the basic figure is embedded (Botzum, 1950). Thurstone reported a split half reliability of .78 (without modification) while Pemberton (1951) reported a corrected split half reliability of .94 on the time limited, paper and pencil form of the test. Validity was based on the earlier Gottscaldt figures and subsequent validation of the CFT forms. Thurstone suggested that the perceptual factor on which the Gottscaldt Figures Test had its highest loading (.51) was associated with his reasoning factor (.39). Yela (1949) reported a correlation of .59 between the reasoning factor and a perceptual factor that he identified as flexibility of closure. Thurstone found a correlation of .63 between inductive reasoning and flexibility. Botzum (1950) confirmed this relation- ship between reasoning factors and flexibility of closure; in a 72 second order factor analysis, closure flexibility had a loading of .64 on what he suggested was an analytic reasoning factor. Pemberton's (1951) data supported the relationship between analytic ability and the previously discussed closure factors. Procedure Ninety-nine student volunteers, 44 males and 55 females, enrolled in introductory and advanced psychology classes at Michigan State University, participated in a two-session pilot study for which they were given extra credit. The study was designed to assess the reliability of measures and the extent of correlations among AAS, BSRI, and CFT scores. Data on two males and four females was excluded from complete analysis because they did not complete the second testing session. The first session data range for these cases (Self and Other forms of the AAS and BSRI) was comparable to the overall group. Shame, guilt, total anxiety, and relative scores for the excluded group are in Appendix B. The volunteers were told that the task involved the assess- ment of attitudes toward various life experience. Subjects were identified by student number (or a self chosen number code) and sex. All subjects were offered feedback and discussion on the research study on completion of the second session. At the first session subjects were asked to complete a test packet which included the Self and Other forms of the AAS and the BSRI. One half of each 73 sex group was given the AAS in the Self-Other order, and the remain- ing half of each group was given the test in the Other-Self order. This was followed by the BSRI. At the second session, scheduled for a date between 14 and 18 days after the initial session, subjects again completed the Self and Other forms of the AAS, given in the same order as in the first session. CFT was also given to the group and timed for 11.5 minutes. Although the CFT norms are devised on the basis of ten minutes of testing time, with answers recorded in the test booklet, computer answer sheets were employed here to reduce scoring time. Therefore, 1.5 minutes were added to compensate for the methodological change. The data were analyzed to determine: AAS test-retest relia- bility and inter-item reliability of the BSRI, using Pearson product moment correlations; intercorrelations between all test sub-scores; sex differences, using t-tests, Computations were carred out by the SPSS program package (Version 6.5). Results Complete results are reported in Appendix B. Of immediate relevance are those data pertinent to the assessment of the instru- ments, subject classification and unexpected findings regarding hypotheses. Assessment of BSRI Reliability on the BSRI was comparable to the reliability found by Bem (1974). However, in the data of this study social desirability was positively correlated with feminine (r = .55; 74 p < .001) and masculine (r - .23; p < .02) sex role endorsement, while Bem reported zero order correlations for each. Compared with Bem's normative sample, more females were classed masculine than males were classed feminine. Approximately 33% of the subjects of this sample were classified androgynous based on the estimated E values set out by Bem (see Table 1). This was consistent with the 27-34% androgynous in the Bem sample. These results have implica- tions for later discussed subject classification. Assessment of the AAS A primary purpose of the pilot study was to assess the item homogeneity and test-retest reliability of the revised AAS (Self and Other forms). Present findings were comparable to the relia- bilities obtained by Perlman (1958). Subject Classification on the AAS In addition to the intercorrelation analysis for the total sample, the relative score on the Self form of the AAS was used for subject classification into high, medium and low shame or guilt groups. The intercorrelations among all scores for the high and low, guilt and shame subjects (four groups) were computed. The variability of the Self form was less than for the Other form and scores extended less far into the possible total shame range (Table 3). Mean relative scores on the Other form were lower (first ses- sion, 2.88; second session 3.36) than on the Self form (first ses- sion, 6.15; second session, 6.89) in each session. 75 Table 1 Classification of Subjects, Based on Bem Norms Bem t-based Criteria* Present Study* Females (N - 50) Males (N - 43) Feminine (t < 2.025) 12 1 Near Feminine (l < t < 2.025) 9 4 Androgynous (-l < t < l) 19 14 Near Masculine (-2.025 < t < -l) 5 13 Masculine (t < -2.025) 5 ll *Bem norms are approximated by multiplying the mean differ- ence score by 2.32 or by.t < 2.025, df - 38, p < .05. Bem (1974) reports a .98 correlation between the two methods. 76 Table 2 Attitude Anxiety Survey: Perlman and Current Study Item Homogeneity Reliability Testingilnstrument .N_ .Alphg. Standardized Item Alpha AAS (Self form, lst session) 89 .93 .93 AAS (Other form, lst session) 89 .92 .93 AAS (Self form 2nd session) 89 .83 .86 AAS (Other form 2nd session) 89 .95 .95 BSRI 89 .96 .96 Test-Retest Reliability AAS Test-Retestg(l4-18 days) Perlman r* Present Study r Self Form - Shame .73 - Guilt .75 - Total Anxiety .83 — Relative .62 Other Form - Shame .80 .76 - Guilt .83 .73 - Total Anxiety .71 .75 - Relative .35 .53 *Perlman reliability computed by a combined internal consis- tency and test-retest method. 77 Table 3 Criterion Groups Based on AAS Relative Score for Each Sex Guilt Criterion Group ‘ (Highest ZOZL Ragga N lst Session R - 9 Males . 6 (N - 18) (10 - l8) Females - 12 2nd Session R - 7 Males - 6 (N - 18) (ll - 1?) Females = 12 Shame Criterion Group (Lowest 20%) lst Session R = 8 Males = 11 (N a 18) (-4 - 3) Females - 7 2nd Session R = 9 Males a 16 (N - l8) (-5 - 3) Females = 2 In Both Sessions Males - 2 Females = 6 Males = 8 Females = 2 78 The AAS criterion groups were composed of the highest and lowest 20% (paralleling Perlman, 1958) of the Relative scores on the Self form for each session separately (see Table 3). Of the 18 subjects included in the first session guilt criterion group (Highest 20%) eight also appeared in the second session guilt criterion group (2 males, 6 females). In the shame criterion group (Lowest 202), 10 of the 18 subjects appeared in both first and second ses- sions (8 males, 2 females). There were more men than women in the shame criterion group (first session: M = 11, F . 7; second session: M = 16, F = 2). Similarly, the number of men in the shame criterion group was greater than in the guilt criterion group (first session: M.- 5; second session: M a 6). The shame group males were more androgynous, near masculine and masculine rather than feminine and near feminine (Table 4). There were a greater number of females than males in the guilt group, though no consistency was apparent in the sex role endorsements of the females of this group. This suggested a reversal of Lewis' (1971) hypothesis of greater guilt for males than females. It also contradicted the present writer's expectation of the relationship of femininity to shame and masculinity to guilt, though support was found in the correlational analysis for the overall group. Intercorrelations: Total Sample Main analysis of hypotheses consisted of intercorrelations among variables for the total sample. Complete results are included in Appendix B. For the sake of brevity, only summary findings for 79 Table 4 Number of Subjects of Each Sex Within Criterion lst Session 20% highest (10-18) 20% lowest (shame prone) (~4 - 3) 2nd Session 20% highest (11 - l7) 20% lowest (shame prone) (-5 - 3) Shame-Guilt and Sex Role Endorsement Categories (N = 18 per group) Males Females Males Females Males Females Males Females Total Males Females F = Feminine (t < 2.025) NO 11 10 NF 1 1 1.0000 NF = Near feminine (l < t < 2.025) A = Androgynous (-l < t < 1) NM = Near masculine (-2.025 < t < -l) M = Masculine (t < -2.025) UIP‘>' 25 13 12 NM 2 2 16 11 5 warez: 14 ll 3 Total 12 72 39 33 80 the total sample are reported here. Each statement of relationship refers to both the Self and Other forms of the AAS, for both ses- sions (i. e., four questionnaires), unless otherwise noted. Specific results of hypotheses are reported in Appendix B, as are individual correlations. The findings indicated an unexpected negative association between guilt and field independence rather than the expected posi- tive one. There was a trend for anxiety (combined shame and guilt scores) to increase with the decrease in the capacity to analytically interpret the field. Likewise, shame and guilt, each taken inde- pendently, both were related negatively to analytic capacity. Shame scores had the predicted positive correlation with femininity for both sexes combined, though only marginally. Guilt was non-significantly positively correlated with masculinity in males and females combined on all but the lst session, Other form. Shame was also non-significantly positively related to masculinity on all but the first session, Self form, a reversal of prediction. Counter to prediction, the obtained correlation between guilt and femininity was positive (r = .37; p < .001). Secondary support for these findings was found in the high proportion of females (73%) in the high guilt criterion group and the high proportion of masculine males (75%) in the high shame criterion group, as just noted. As predicted, closure flexibility was correlated negatively with femininity (r = -.33; p < .001). The obtained correlation between closure flexibility and masculinity was non-significantly negative (a reversal of prediction). 81 There was a general trend suggesting that feminine and masculine role endorsement in males and in females is associated with greater anxiety (guilt and shame separately and combined). Both femininity (r = .55; p < .001) and masculinity (r a .23; p < .02) were correlated with social desirability. T-tests for the total sample demonstrated a significant sex difference in the Self form, second session, on the AAS relative score; higher shame was found for males. On the remaining three AAS variables there were no significant sex differences. Signifi- cant sex differences for the total sample were demonstrated on femininity scores and androgyny scores on the BSRI, reflecting a feminine direction of deviation from perfect (0) within androgyny for females and a masculine direction of deviation from perfect (0) within androgyny for males (See Appendix B). Discussion and Implications Data from the pilot study demonstrated adequate reliability for the revised AAS instrument and had implications for subject classification in the main study. The range for pilot data showed that the relative score of the AAS Self form accentuated guilt responses and minimized shame responses. In comparison, the range on the Other form showed more balance of guilt and shame (R = 25; -13 to 11; R = 28; -15 to 13). Based on these differences and a concern for the possibility that the Self form was eliciting defensive responding, the AAS Other form was used in the main study. 82 The scoring method by which the shame and guilt criterion groups were established (Relative score) did not reflect the actual intensity of guilt or of shame separately or together. What it provided was a difference score between shame and guilt responses. Analysis of data basedcnrthe Relative score provided less useful information than did the analysis of the absolute scores for Shame, Guilt, and Total Anxiety (See Appendix B). For this reason, absolute Shame and Guilt scores were used for subject classifica- tion and main analysis in the main study. Overall, correlational trends suggested some support for the positive association of shame, femininity, and field dependence. The association of guilt and masculinity was supported, but the expectation of a positive relationship between guilt and field independence was reversed by the data. The findings, therefore, countered some of the expectations in the direction of correlation. Analyses of AAS Relative score-based criterion group data and the BSRI classification (feminine, masculine, androgynous) provided little apparently useful information. No significant sex differences were found between males and females in shame and guilt. Some trends were opposite to prediction. Masculine sex role endorsements in males and females combined was non-significantly positively correlated with guilt but also with shame. Likewise, although no significant sex differences were found with t tests, feminine sex role endorsement in males and females was positively (non-significant) correlated with guilt. Further, findings also suggested that feminine sex role endorsement 83 was positively correlated with total anxiety (combined guilt and shame scores), reflecting greater overall receptivity to shame and to guilt with femininity. Whereas the correlational analysis of the total group data supported some of the present writer's views, it must be noted that based on the Relative score, males showed a greater concern with shame issues, and females showed more concern with guilt issues, than was expected. MAIN STUDY The major interest in this experiment was to assess the relationship of shame and guilt proneness with the extent of differ- entiation and the extent of awareness of and responsiveness to the external environment. specifically, time spent reading positive or negative information about the self after positive or negative per- formance feedback was measured for high shame subjects and high guilt subjects. Defensive style (repression-sensitization), inci- dental learning, and recall of trait feedback were studied as indicators of attention deployment. The interrelationship of shame and guilt with sex role endorsement was also explored. This research design arose in part from the correlational results of the pilot study which indicated a need for a more detailed assessment of the shame and guilt affects and their relationship to psychological differentiation as manifested in attention style. Hypotheses presented here are sequentially numbered 9-13. These are followed by two additional hypotheses (5 and 6) tested as group differences in the pilot study (results reported in the pilot study) and refined and retested in the main study to specifically reflect the relationship of high shame or high guilt with sex role endorse- ment. A summary of results for the main study reflecting the hypo- theses presented and tested in the pilot study, is also included in the Results section. 84 85 Hypotheses This writer asserts that global orientation increases the awareness of the external world and in this way increases the possi- bility of personally acknowledged recognition of negative and positive inputs from the environment. We posit a relationship between a global orientation and a cognitive style that implies a sensiti- zation to environmental inputs. This hypothesis runs counter to the previously described relationship between global orientation and lowered sensitization (i. e., repression, when measured on the bipolar RrS Scale developed by Byrne, 1961). The contention is that the shame experiencing individual receives (is sensitized to) and is impacted by evaluative inputs from others in the environment, more than is the guilt experiencing individual, though Lewis (1971) posits greater sensitization for guilt experiencing individuals. Therefore, we hypothesize that: 9. Shame will be positively related to sensitization as measured by the RrS Scale; this relationship will be stronger than for guilt and sensitization. The sensitizing tendency facilitated by the shame experience is to take negative evaluation as legitimate and valid, direct hostility at the self for having failed the other, and lower self- esteem. Thus, it seems that given a choice to spend time with exter- nally provided positive or negative information, more time will be spent with negative than with positive feedback. This does not mean that the positive feedback is not valued by the shame exper- iencing individual. Rather, the contention is that the dynamics of shame include fear of exposure and loss of the other's respect, 86 which prompts a concern for negative feedback. As developed in theoretical perspective, the shame experiencer is concerned with either the complete presence or absence of narcissistic supplies; hence, any negative feedback will threaten the esteem, and engender concern and attention. It is hypothesized that after experimentally manipulated failure, subjects will spend more time reading negative personal feedback than positive feedback, and that this will be more marked than after success. In both failure and success conditions, the high shame subjects will spend more time with negative information than will subjects who score low on shame but high on guilt. Specifi- cally: 10. High shame (low guilt) subjects will spend a greater amount of time than high guilt (low shame) subjects reading negative feedback, under both success and failure conditions. High guilt (low shame) subjects will not differentially attend to positive or negative feedback under the success condition. 11. After failure, both high shame (low guilt) and high guilt (low shame) subjects will attend more to nega- tive feedback than positive feedback; however, the high shame (low guilt) subjects will attend to negative feedback more than will high guilt (low shame) subjects. The theoretical framework presented suggests that shame ex- periencing individuals are more vigilant about the environment, and its potentially threatening evaluation, than are guilt experiencing individuals. Of interest is whether the basic level of sensitiza- tion in the absence of threat differs for the two subject groups. A variation of the Mendelsohn and Griswold (1964) incidental 87 learning paradigm will be used to measure the extent of sensitiza— tion, based on the use of incidental cues in anagram solving. It is hypothesized that: 13. High shame (low guilt) subjects will perform better on an incidental learning task than high guilt (low shame) subjects. Although the "other" is given esteem regulating power, and although the individual may retain the awareness of negative feed- back, high shame individuals may choose or choose not to make an evaluative person(s) aware of their importance and the importance of negative feedback. By choosing not to share the impact of the other with the other, the high shame individual can avoid further exposure, although aware of (sensitized to) the negative feedback, by denying its impact. This arises from an initial shame reaction, and fear of exposure; that is, not only is there shame, but fear that the other will know that the self has been shamed. Given the theoretical proposition that shame arises as a result of the experience of failure, the concern with exposure (measured as recall of trait feedback) should be more pronounced for shame subjects after an experimentally manipulated failure than after a success. Guilt is theoretically related to wrongdoing rather than failure and it is expected that guilt subjects will not differ in recall as a function of the experimental manipulation of success and failure. We will test the relationship between high shame and recall by an inventory recall measure of personal traits completed after an experimentally manipulated success or failure experience. The hypo- thesis is: 88 13. High shame.(low guilt) subjects will have greater recall scores about feedback material after success experience than after failure experience; this will not be true for high guilt (low shame) subjects. The relationship of shame and guilt with sex role endorse- ment was explored in the Pilot Study. The specific pilot hypotheses (3 and 4) reflected bipolar definitions of shame and guilt and masculinity and fanininity. The hypotheses were tested by Pearson Product Moment correlations in order to assess several aspects of Lewis' (1971) theory. Although some support for the hypotheses was found, it was believed important to assess the impact of the individual levels of shame and guilt separately, as they relate to sex role endorsement. The present hypotheses (5 and 6) acknowledge shame and guilt and masculinity and femininity as separate dimen— sions. The expectation is that femininity scores will be higher than masculinity scores for high shame subjects. Conversely, high guilt subjects will have higher masculinity scores than femininity scores. These hypotheses reflect the theoretical discussion regard- ing the greater involvement of the self with the environment and greater receptiveness to environmental assessment inherent in the tradition feminine sex role and acknowledges the involvement of receptiveness to environmental assessment in the shame experience. Conversely, the lessened involvement with, and receptiveness to, the environment and to environmental assessment is theoretically linked to the masculine sex role and has been presented as conducive to the experience of guilt affect. 89 Two secondary hypotheses were tested in the main study that reflect the numbering of pilot study hypotheses. In acknowledging masculinity and femininity as separate dimensions, no hypotheses are presented regarding negative relationships between high shame and masculinity or high guilt and femininity. Hypotheses will be tested with correlational and multivariate analyses and t-tests for group differences. The hypotheses reflect (in part A) the corre- lational analysis for the total group of shame and guilt affect with sex role endorsement, and (in part B) the specific relationship of sex role to high shame and to high guilt. Thus, it is hypothe- sized: 5. Shame scores will be positively associated with femininity: a. Shame scores will show a greater positive association to femininity scores than will guilt scores. b. For high shame subjects, feminine scores will be greater than masculine scores. 6. Guilt scores will be positively associated with masculinity: a. Guilt scores will show a greater positive association to masculinity scores than will shame scores. b. For high guilt subjects, masculine scores will be greater than feminine scores. In hypotheses referring to shame-guilt classification, high shame refers to subjects scoring high (above the median) for shame but low (below the median) on guilt. High guilt refers to subjects scoring high (above the median) for guilt but low (below the median) on shame. These median classifications are based on the level of the absolute shame and absolute guilt scores. The absolute scores were employed in order to provide a more accurate and sensitive measure of the level of shame and guilt independently. Hypotheses are not 90 included concerning subjects either high (high anxiety) or low (low anxiety) on both shame and guilt. These latter classifications will be included in the correlational and multivariate analyses and reported where relevant. Of interest are the similarities and differences between these two groups and the high shame group and high guilt group. The htpotheses will be tested by using a revised form of the Perlman (1958) Attitude Anxiety Survey; Bem Sex Role Inventory; Closure Flexibility Test; Repression-Sensitization Scale; experi- mental manipulations of success and failure; incidental learning measures; and attention to and recall of personal feedback to which the subject attended. Method The present design allowed for inter-group comparison on behavioral and inventory variables among the four AAS classifica— tions: (1) high anxiety (high shame, high guilt), (2) high shame (high shame, low guilt), (3) high guilt (high guilt, low shame) and (4) low anxiety (low shame, low guilt). Subjects An equal number of male and female subjects were selected for participation in the second session on the basis of their shame and guilt scores on the AAS given in the first session. The first session consisted of larger group meetings for the purpose of filling out the inventory measures. The sujbect pool consisted of 91 individuals participating in introductory psychology courses. Sub- jects received extra credit for participation. Subjects' inclusion in one of the four cells was determined by their shame and guilt score levels. Median cutoff points were established on the subject score distributions on shame and guilt scales, independently. Subjects were further divided into experi- mental success and failure groups; placement was random within sex and shame-guilt score levels. Full details of the experimental structure are described in the Procedure section below. Pilot study subjects were classified originally in a nine cell format based on high, medium and low shame and guilt levels on the relative score. Additionally, an attempt was made to classify subjects into categories based on a combination of variables. Cate- gories were established combining high, medium and low, relative scores on the AAS with comparable classification on the Closure Flexibility Test and with masculine, feminine and androgynous groups on the BSRI, separately. The attempt to combine two groups of variables in order to define subject classifications, increased substantially the difficulty in findings appropriate subjects for the cells. This prompted two decisions: to use shame-guilt level alone as a subject classifer; to collapse the nine cell design, based on high, medium and low shame and guilt levels, down to a four cell high and low, shame and guilt design. All combinations of variables considered as potential methods of subject classifica- tion are included in the Appendix C. 92 Testingjlnstruments Attitude Anxiety Survey: Subject Classification The Other form of the revised AAS (developed by Perlman, 1958) was used to establish shame-guilt cutoffs for subject classi- fication. This form was selected because it produced a greater range of responses than the Self form for shame and guilt raw scores and for the relative score, a secondary measure of shame and guilt proneness. Appropriate high-low cutoff points were established for both the distributions of shame scores and guilt scores independently. Four combination classifications were established on this basis. Equal numbers of males and females were selected from each of the four combined shame-guilt levels. This procedure has the disadvan- tage of translating ordinal data into nominal data. However, the method is sensitive to both the overall score intensity (shame plus guilt scores) and the relative intensity of shame and guilt. Also, pilot data indicated that separate shame and guilt scores were more sensitive to interrelationships between variables than the median- based relative score, and were more reliable over time. Bem Sex Role Inventory The BSRI (Bem, 1974) described in the Test Instrument sec- tion of the Pilot Study, was used as a measure of masculine, feminine and androgynous sex role endorsement. The Pilot Study reliability assessment of the BSRI is comparable to Bem's (1974) findings. Scoring revisions were based on the Spence, Helmreich 93 and Stapp (1975) definition of androgyny as high scores on both masculine and feminine subscales, rather than simply relative equality (low or high) of scores on each. Sex role designations were based on masculine and feminine scale scores, dichotomized at the median for each scale on each sex separately. This scoring system gives four possible sex role designations: High scores on both masculine and feminine are classed androgynous. Low scores on both masculine and feminine are classed undif- ferentiated. A high score on masculine and a low score on feminine is classed masculine. A low score on masculine and a high score on feminine is classed feminine. Closure Flexibility_Test The Closure Flexibility Test (Thurstone and Jeffrey, 1956) was administered as a measure of field independence. A complete description of the test is included in the Test Instrument section of the Pilot Study. Repression-Sensitization Scale The R—S Scale (Byrne, 1961) was administered as a measure of the extent of use of repression relative to sensitization de- fense. The R-S scale consists of 156 scoreable and 26 buffer true- false items. The 182 items were drawn from Depression, Psychathenia, Hysteria, Lie, K, and F subscales of the MMPI. High scores indicate sensitization and low scores indicate repression. Byrne (1961) 94 reports an internal consistency coefficient, based on split-half reliability, of .88. Test-retest reliability is reported at .88 over a six week period. Procedure Student volunteers enrolled in introductory psychology courses at Michigan State University participated in a two session study. It was designed to assess the effects of success of failure on time spent reading positive and negative feedback, for individuals designated as high and low on guilt and shame measures. Session 1 A packet of the revised AAS, BSRI, RrS Scale and the Closure Flexibility Test was completed by each subject. Subject groups con- sisted of approximately 15 students per session. Sessions were carried out over a six month period. In addition to instructions for completing the tests (see Appendix A), a cover letter explaining the nature of the experiment suggested that on the basis of the tests, feedback would be pro- vided concerning individual personality traits. The first session was 1 1/2 hours. The letter read: Dear Student: Thank you for your agreement to participate in this experiment. Enclosed, you will find a series of Questionnaires, each with specific instructions. You are to fill them out according to the instructions. Among them you will find a test marked Closure Flexibility Test. You are to fill this out only dur- ing the period (10 minutes) that has been set aside for that purpose. Please complete the tests in the following order: 95 Closure Flexibility Test; Attitude Anxiety Survey; Bem Sex Role Inventory; RPS Scale. Purpose of This Study: It has long been of interest to this experimenter, whether responses to life experiences are related to other consistent aspects of the individual's personality. We will be processing these tests in order to assess such relationships. Additionally, it will provide you with personal trait feedback. During this session you will be asked to make a one hour appointment for the second session of this experiment. At the next meeting additional tasks will be assigned and you will be given feedback on our findings based on the ques- tionnaires you fill out today. Questions concerning the experiment will be answered at that time. Again, I appreciate your participation. Please be reminded that extra credit will be contingent on your participating in both sessions. Cordially, Equal numbers of males and females within each category of shame and guilt, previously described, were requested to partici— pate in the second session. Session 2 - General Procedure Subjects were scheduled for a one hour individual session. They were received in a reception area, given general instructions, and then taken to a room in which there was a comfortable chair. They were told that there would be a ten minute task followed by the opportunity to obtain feedback on the previous measures. Behavioral Task The task was an incidental learning paradigm used by Mendelsohn and Griswold (1964). This provides word memorization and anagram problem solving scores. False feedback about the number of 96 anagram.solutions were used as the success-failure manipulations. Actual scores were analyzed as the measure of incidental learning. The task was introduced to the subject in part, as an experiment of "memory under interference conditions." A list of 25 words was given to the subject, with the instructions that he or she would be given 10 minutes to learn the words. A second list of 25 words would be played on a tape recorder in the room during the memorization period, in order to provide the interference. The solutions for 10 of the 30 anagrams were "cued" by words on the memory list (focal cues); 10 of the 30 words were "cued" by words on the recorded list (peripheral cues); the final 10 anagrams were not "cued," as a measure of general anagram solving ability. Specific instructions were given. They were: You will be given a list of words to memorize. They can be learned in any order, and you may use any device you wish to aid you in memorization. You will have ten minutes to learn the words and later on you will be asked to write down the words you remember. There is more, however. During the time you are learning the words, another list of words will be played on the tape recorder. You are asked to concentratecnlthe list in front of you and to memorize it as well as you can while the other words are being played in the background. In addition, after your ten minutes are up, you will be asked to solve some problems. When the problems are over, then you will have a chance to write down the words you remember from the list you have memorized. Do you understand? Memoryplist (Starred words are focal cues for anagram solutions) hares cause swine teach apple* miles uncle* night tutor tower paper* erase* divan tiger* twins* throb slope brick* haste house repel chair* whale* start spice 97 Recorded Word List (Starred words are peripheral cues for anagram.solutions) drink* mouth* plant* dozen* medal cheap throw* plate cloud broth bacon* blows north stems* child* spare leapt lakes frost scarf* stead storm* snipe linen clams The completion of the ten minute memorization period was followed by an anagram problem solving tasks. Thirty single solution, five letter anagrams were chosen for equal difficulty from ana anagram list developed by Rees and Israel (1935). The anagram problem task was introduced as a "cognitive-verbal manipulation tast - an aspect of intelligence." Anagrams: Group A (focal cues) Group B (peripheral group) Group C (neutral) saree - erase idnkr - drink trewa - water ihrac chair oabnc bacon nirsa rains plepa apple otsmr storm etebs beets cikbr brick naptl plant vesno ovens uheso house metss stems fiken knife stniw twins nelin linen ensce scene greti tiger nedoz dozen sdlen lends ehlaw whale dlchi child neque queen ulenc uncle frsca scarf ytpar party eprpa paper htmou mouth macre cream After the completion of the anagram task, the subject was asked to write down those words that they memorized earlier. No feedback was given on the memorization task. At this point the pre—programmed false negative or false positive assessment of their anagram task performance was given by the experimenter. 98 Succesleailure Manipulation The behavioral task had been already programmed as a failure or success randomly, based on the subject's prior placement by sex, in shame-guilt combined level groups. Both conditions were carried out by a pretense scoring the number of correct anagram solutions in the room, with the subject present. The experimenter carried a clipboard, with a complex graph, to which he referred, having asked the subject's age., The subject was not allowed to consult the graph. For subjects in the failure condition, the experimenter looked somewhat perplexed and informed the subject that his/her scores was in the 30th percentile, based on age, for this task. For the subject in the success condition, the experimenter smiled at the subject after he consulted the graph. He told the subject that the score he/she had achieved is in the 95th percentile based on age. It was explained that these norms are based on nationwide studies of many college age people. The experimenter noted that the score was below (or above) average scoring and suggested that it could be discussed further after the subject completed the experi- ment. Trait Feedback Following the anagram task and recall, the experimenter explained that ten minutes had been set aside for the subject's review of the personality trait feedback on the previous session's questionnaires. The subject was conducted to an adjacent table, asked for their identification number assigned in the first session, 99 and told that their file would be brought in. The experimenter left and returned with an envelop on.which the student's identifi- cation number was visible on a (conveniently removable) tag. It was explained to the subject that due to the complex design of the study and the experimenter's interests, the data was processed and was output in a form that separated positive and negative findings. The purpose for this was to make the feedback manageable for the experimenter and the subject. Within the envelope were two lists of ten numbers each (positive and negative), which referred to traits in the trait man- uals. These lists were removed, placed on stationary clipboards on tlle same table but distant enough from one another to prevent simultaneous comparison (approximately 3 - 4 feet). Next to the clipboard (further from the subject) to each side, were manuals (labeled Positive and Negative) taped to the table. Each manual was developed to include a substantial amount of filler discussing traits not used, as well as the randomly interspersed traits that were used. The subjects were told that if they wanted feedback, they should read their lists, then go to the appropriate manual to find and read the trait descriptions. The room was set up with one chair and they were invited to move the chair to the trait feedback table. The subject was told that he or she would have ten minutes to read the feedback, and told that after ten minutes, the experimenter would return to discuss any questions the subject had. If they finished before the allotted 100 ten minutes, they were told to simply open the room door and that the experimenter would come in as soon as possible. One set of ten positive and ten negative traits was used for all subjects. The traits were taken from the Edwards Personal Preference Test (1953), the California Personality Inventory (1962), and the Omnibus Personality Inventory (1957). A listing of all trait dimensions and descriptive sketches was included in the manual for negative and positive traits. Trait lists and sources, trait descriptions, and listings of positive and negative traits used as feedback, are included in Appendix D. Students were given the following instructions: Dear Student: The analysis of the test battery taken in the first session has been completed. Its purpose is to give you individual feed- back on your personality characteristics. We hope that this feedback will be helpful to you. On completion of today's experiment any questions or concerns you have may be dis— cussed with the experimenter. Specific instructions: You will receive separate lists of your 10 highest positive score categories and your 10 highest negative score categories. The numbers printed on the posi- tive and negative lists are reference numbers. They corres- pond to the numbers of those traits on which you scored highly positive or negative. First, note the first reference number on which you were scored, then go to the appropriate catalogue and look up the character description for that dimension (it will also be preceded by the same number). For your convenience we have separated positive and negative character descriptions for each dimension and placed them in separate booklets. You will find booklets labeled "Positive Trait Descrptions" and "Negative Trait Descriptions" to your right and left. The trait manuals were structured to allow for the use of the same trait name followed by positive (+) or negative (-) notations, 101 depending on.which manual they appear in. Certain items drawn from the Edwards Personal Preference Test were not created as positive or negative manifestations of the trait listed, but simply provided one description for the trait. Where this was the case, the trait description was sorted along a positive-negative dimension and separated out into two descriptions. Timing;§elective Attention to Feedback An observer behind a onedway mirror recorded which manual was approached first and took a time measure on how much time was spent at each of three positions: Positive, negative, neither. Time was kept in ten second blocks for ten minutes after the experimenter left the room. Time spent not attending to feedback was scored "nothing." Training of observers consisted of explaining the pro- cedure for scoring time spent reading positive and negative feedback. Additionally, the experimenter monitored the scoring of approximately the first three subjects for each of the observers. This was carried out correctly by both observers. Post Feedback Period After the ten minute feedback period the experimenter re- entered the room and replaced the trait lists in envelopes. A brief questionnaire was given the subject with a request that she or he fill it out before discussion began. The questions (below) provided a report on recall of feedback, and were used to assess the effectiveness of the manipulations. 102 Recall Measure The subject was presented with a list of traits with the specific instruction; "For the listed traits place a "P" of an "N" next to those which appeared on your positive (P) or negative (N) trait lists." The trait list included 30 of 38 traits listed in the manuals (see Appendix D), and represented all positive and negative traits (20) from the feedback lists they received, along with 18 filler items. Manipulation Check Subjects were asked the following questions in written form: How did you feel about the anagram task feedback? How did you feel about the personal trait feedback? Did you believe the experimenter when he told you that you were above average or below average on the anagram task? (Yes/No) 4. Do you feel that the trait feedback was truthful, in rela- tion to you? (Yes/No) 5. In some cases, feedback on the performance on the anagram task was truthful, in some cases it was false. Which group did you feel you were in? (True/False) 6. Some people got true feedback on their trait lists, others got false feedback. Overall, which group do you think you were in? (True/False) Debriefing It was explained to the subject that: l. the study was not focused on intellectual achievement, or personality trait assessment; 2. the feedback, "above average," "below average" on performance task, was all false; that no national norms exist; 103 3. all trait feedback, even that which seemed to the subject to be true, was not based on the subject's responses to first session questionnaires. All subjects were exposed to the same feedback. A complete explanation of the study was given. Subjects were asked not to discuss the actual purpose of the study or its design with other people for the remainder of the time the study was going (one term). A complete explanation of the individuals' actual scores on each Inventory was given. Additionally, the subjects were given the opportunity to ask questions of the experi- menter and were told that results of the total study would be available in Spring, 1977. RESULTS The presentation of results is organized into two sections. In the first section descriptive statistics regarding Inventory measures and results pertaining to Hypotheses 5, 6, and 9 are pre— sented. The second section presents descriptive statistics concern- ing second session feedback reading and recall measures, evaluation of Hypotheses 10 through 13, and the assessment of the effectiveness of the experimental manipulation. Subjects Four hundred and forty-eight subjects (241 males, 207 females) participated in the first session of the experiment. From this group, 128 subjects (64 males, 64 females) were selected on the basis of their meeting criteria for one of the four shame-guilt combined score groupings. An equal number of males and females within each shame-guilt group were randomly assigned to success and failure conditions. Cutoff points for high and low shame and guilt were based on a median for shame (163) and guilt (195) scores separately. For practical reasons, median cutoff points for subject classification were based on 212 subjects. The medians for this sample were virtually the same as the mean for the sample (N = 128) selected for the second session. 104 Shame Guilt 105 Table 5 Attitude Anxiety Survey (N = 128) Obtained Mean S.D. Median Range 163.71 22.35 163 R=l42 (96-237) 192.84 28.33 195 R=123 (123-245) Possible Range R=260 (0-260) R=260 (0-260) 106 Subjects were not equally distributed among the four shame- guilt categories. Chi Squares computed for each sex separately across the four categories were significant (males, x2 - 98.24, p < .001; females, x2 - 35.53, p < .001). High shame-guilt (high anxiety group) and low shame-guilt (low anxiety group) subjects were most frequent for both sexes. Consistent with the findings on the pilot study, there were more high shamerlow guilt (high shame group) males than females. Conversely, low shame-high guilt (high guilt group) females were more numerous than comparable males. Descriptive Statistics: Individual Variablesg(first sessiqp) Each questionnaire and behavioral variable was analyzedtxade- termine, for the overall sample (N = 128), the group means, standard deviations, and ranges . Additionally, frequencies within variable com- binations were tallied. Pearson product moment correlations, within and across questionnaire variables were computed. Descriptive sta- tistics were obtained through the SPSS program package(Version 6.5). Attitude Anxiety Survey (AAS) The sample was more guilt than shame oriented by several indices. The mean guilt score (192.84) was significantly (.001) higher than the mean shame score (163.71) based on a pairwise t-test (n - 128, df - 127, t - 12.44, p < .001). Standard deviations of these scores were comparable (Guilt, 22.35; Shame, 28.33). The range of the shame scores (R = 142: 96 to 237) was larger than for guilt scores (R = 123: 123 to 245. No significant sex differences were found with t-tests. The sample mean for the relative score, 107 Table 6 Number of Subjects of Each Sex Available for Inclusion in Each Cell, Based on the Total Sample High Guilt Low Guilt Total High Shame male- 94 male- 33 male-127 female- 96 female- 28 female-124 Low Shame male- 17 male- 97 male-114 female- 30 female- 53 female- 83 Total male-111 male- 13 male-241 female-126 female- 81 female-207 Males, xz-a8.24, p<.001 Females, x -35.53, p<.001 108 denoting the difference between the number of guilt items and shame items above each individual's own median rating, was moderate in the direction denoting higher intensity (above the median) responses to guilt items than to shame items. Pearson product moment correlations for variables within the AAS were calculated (Table 7). Guilt scores were positively correlated (p < .001) with: shame scores (r I .47) and total AAS score (r I .81); and with the relative score (r I .30), reflecting again the guilt direction of response by this sample. Shame scores were positively correlated with the total AAS score (r I .87; p < .001) and negatively (r I -.54; p < .001) with the relative score. Bem Sex Role Inventory (BSRI) Means and ranges on Masculine (M) and Feminine (F) subscale of the BSRI were significantly different from each other (t-test: n I 128, df I 127, t I 1.85, p < .01). The sample mean on Social De- sirabiltiy (5.08) was marginally significantly greater (t-test: n I 128, df I 127, t I -.184, p < .06) than the M score (4.96) and signi- ficantly greater (t-test: n I 128, df I 127, t I -4.23, p < .001) than the F subscale score (4.78), though the ranges were all compar- able. There were marginally significant differences on M and F scores for females (t-test: n I 128, df I 127, t I -l.57, p < .12) and significant differences for males (t-test: n I 128, df I 127, t I 4.85, p < .001). The Derived Androgyny score (Bem, 1974), con- sisting of the difference between the individual's F and M meanscores multiplied by 2.32, showed a slightly negative trend, denoting Andro- gyny (-.45) with a slight masculine deviation from perfect (0). 109 Table 7 Attitude Anxiety Survey: Intra-Test Correlations §hggg_ Tthl_ Relative AAS Guilt .47* .81* .30* AAS Shame ---- .87* -.54* AAS Total ---- -.l8** *p<.001, nI128 **p<.02, nI128 110 MN.H ow.l 0H.H HN.HI we. nm.HI mo.mlmn.ol Hm.N mc.l oo>auon msmouon< mm.¢ wo.n mm.¢ mw.a No.< om.q wo.m Ho.m mm.¢ Hm.m as.a nm.q om.a oo.¢ Hm.m om.e no.e ow.e mH.m oa.a me.< -.m nm.wuom.u mm.~unm.~ om.numn.~ an. we. no. wo.m mn.¢ om.q huwafinouwmon onwsasom moaaaommz Hmfioom fimma GMQE GMQE GMQE anamoa some enemas some swamp .Q.m some muouoo>sm oHom xom son so moHQBMm Acnmav son was Aououoeom one Hmuos .moxomv Ivsum saws soozuom consumeaoo a «name moamsom mmHQZ Aqma u zv museum massauoom awe Ammm n av madmaom Aeee u av moan: Amme u zv mamamm muowsmum aom Aeo u av moamaom Ace n av moan: AmNH u zv macaw Hmuoa 111 Classification into Masculine and Feminine categories was defined by Ban (1974) as the extent of movement away from perfect Androgyny (0) in a negative or positive direction, respectively. The classifications are based on the absolute mean FIM difference score. A.more complex, median-based method of scoring (Spence, Helmreich and Stapp, 1975) provided a more refined sex role cate- gorization. This study used both the Bem and the Spence methods at different points, and reference is made to the system employed in the presentation of results. Subjects were also individually classified on the basis of the Spence scoring methods. This scoring system allowed for the classification of subjects into androgynous, feminine, masculine, and undifferentiated categories based on placement above or below median levels on M and F subscales (Table 9). Median levels were determined for each sex separately rather than for both sexes com- bined, an alternate method. The largest percentage of subjects were classified as undifferentiated (low on both F and M) (28.9%). The median based scoring requires this group to be viewed as separ- ate from the androgynous group (high on both F and M), which composed an additional 26.6% of the sample. Feminine classification had the lowest number of subjects in the sample. Within the male sample, using the Spence system, the greatest number of subjects were classified as undifferentiated. The classification of androgyny produced the next highest group, followed by masculine and feminine classifications. Females were more equally dispersed among the four classifications than were males. There were a higher number 112 of males than females in both androgynous and undifferentiated categories, and more females than males in feminine and in masculine categories. Results of a Chi square test were significant (x2 I 4.54, p < .05). Compared with the Bem (1974) normative samples, the present sample showed, for males: M scores and F scores that were slightly higher than the two Bem samples; for females, M scores were slightly higher than the Bem sample while the F scores were slightly lower. F scores of the BSRI were expectedly positively correlated (r I .40; p < .001) with the Derived Androgyny score reflecting a feminine direction of sex role endorsement within androgyny (Table 6). Conversely, the M score negatively correlated (r I -.71; p < .001) with the Derived Androgyny score, reflecting a masculine direction of sex role endorsement within androgyny. F and M scores were correlated with each other (r I .16; p < .02). Both M (r I .28; p < .001) and F (r I .16; p < .001) were positively correlated with Social Desirability. Closure Flexibility Test (CFT) The mean Closure Scale score for this sample (55.84) was within the average range (40 to 60) described by Thurstone and Jeffery (1965). Correct Closure responses were correlated posi- tively with the Closure Difference score (r I .87; p < .001) and the Closure Scale score (r I .86; p < .001), as would be expected from the earlier description of the scoring procedures. Likewise, the expected negative correlations were found between the Incorrect 113 Table 9 Median-Based Sex Role Classification (Based on Spence Method) Male Female Both Median FI4.80 Median FI4.80 Median MI5.15 Median MI4.60 s: z. 2 z r :4 Androgyny 19 29.7 15 23.4 34 26.6 Feminine 9 14.1 17 26.6 26 20.3 Masculine 14 21.9 17 26.6 31 24.2 Undifferentiated 22 34.4 15 23.4 37 28.9 All 64 100.0 64 100.0 128 100.0 x2 = 4.54; p < .05 Ben Sex Role Inventory: Femininity Masculinity Social Desirability *p < .001, nI128 114 Table 10 Masculinity .16* Social Desirability .16* .28* Intra—test Correlations Androgyny Derived .40* _.71* -.13* 115 Closure score and both Closure Difference and Closure Scale scores (r I -.44; r I -.45; p < .001). No significant sex differences were found on Correct or Incorrect Closure scores with t-tests. Repression-Sensitization Scale (RIS) The mean repression-sensitization score (66.60) was compar- able to the overall sample mean (62.44) reported by Byrne (1961). The female score average (61.22, n I 64) was almost identical with that reported by Byrne (61.80, n I 230). The male mean in the present sample (67.98, n I 64) was somewhat higher (more sensiti- zation) than in the Byrne sample (63.08, n I 394). A t-test for sex differences on the R-S score was significant (t-test: n I 128, df I 127, t I 2.13, p < .035). First Session Inventory Measures: Hypotheses 5,76,_and 9 A multivariate analysis was computed on a three way design that assessed the effects of shame level (high-low), guilt level (high-low), and sex. Dependent variables were the Bem Sex Role Inventory (BSRI), Closure Flexibility Test (CFT), and Repression- Sensitization Scale scores individually. Alphas were set to allow for multiple testing. Alphas for shame effect, guilt effects and shame by guilt interaction effects was .05. Sex effects within each of the four shame-guilt groups were tested separately at an .0125 alpha level. In testing for main effects on the specific univariates, the overall alpha was divided by the number of depen- dent variables for each scale. For shame, guilt, and shame by guilt 116 Table 11 Closure Flexibility Test: Intra-test Correlations Closure Closure Closure Incorrect Difference Scale Closure Correct .04* .87* .86* Closure Incorrect ---- .44* .45* Closure Difference. ----- .99* *p < .001, nI128 117 interaction effects alpha was set at .0125 for BSRI variables (.05/4); .0166 for CFT variables (.05/3); at .05 for R—S score (.05/l). For shame-guilt by sex effects, alpha was set at .0031 for BSRI variables (.0125/4); at .0041 for CFT variables (.0125/3); at .0125 for the RIS score (.0125/l). Additionally, t-tests were computed to assess group and pairwise differences. T-tests relevant to hypotheses are reported in the text. The remainder are included in Appendix E. Hypothesis 5 Shame scores will be positively associated with femininity. A. Shame scores will show a greater positive association with feminine scores than will guilt scores. There was no support for this hypothesis. The correlations for guilt and femininity (r I .04; p < .30) and shame and femininity (r I .04; p < .33) were equally low. B. For high shame group subjects, feminine scores will be greater than masculine scores. No support was found for higher femininity (4.78) than for mas- culinity (5.07). In fact, the obtained means pattern within the high shame group was a reversal of the predicted relationship suggesting an association between shame and masculinity rather than femininity. However, a.t test showed no significant difference between M and F scores within the high shame group with sexes com- bined (t-test: n I 32, df I 31, t I 1.50, p < .14). High shame group males showed a significantly higher M score than F score (t-test: n I 16, df I 15, t I 2.16, p < .04) and females showed a 118 Table 12 BSRI Variables Within Shame-Guilt Levels High Shame Femininity Masculinity Social Desirability Androgyny Derived Low Shame Famininity Masculinity Social Desirability Androgyny Derived All Gropps Femininity Masculinity Social Desirability Androgyny Derived High Guilt Males Females Both 4.78 5.05 4.91 5.35 4.62 4.99 4.92 5.19 5.05 -l.3l .82 -.24 4.64 4.70 4.67 5.21 4.29 4.80 4.95 5.16 5.05 -l.48 .72 -.38 4.69 4.86 4.78 5.22 4.67 4.94 4.96 5.21 5.08 -l.37 .45 -.45 Low Guilt Males Females Both 4.66 4.89 4.78 5.27 4.86 5.07 5.18 5.16 5.17 -l.9l .14 -.88 4.70 4.81 4.75 5.03 4.79 4.91 4.77 5.35 5.06 -.78 .13 -.32 119 non-significantly higher F score than M score. Neither the shame effect (p < .68) or the shame by guilt interactive effect (p < .67) was significant in the multivariate analysis on BSRI variables (Tables 13 and 14). Hypothesis 6 Guilt scores will be positively associated with masculinity. A. Guilt scores will show a greater positive association with masculine scores, than will shame scores. There was no correlational support for this hypothesis. A reversal of the predicted relationship was indicated. Shame had a non-significant positive relationship with masculinity (r I .12; p < .07), while guilt showed a non-significant negative relation- ship with masculinity (r I —.07; p < .19). B. For high guilt group subjects, masculine scores will be greater than feminine scores. M and F means for the high guilt group subjects (sexes combined) were in the predicted direction, though a.t test was not significant (See Table 12). Cell means on BSRI M and F scales for each sex separately within shame-guilt categories, showed that in every category mascu- line scores were higher for males than for females, and conversely, feminine scores were higher for females than for males (See Table 12). The results showed that there was expected support for this hypothesis for males and a reversal for females. Within the high guilt group, the significant sex effect was isolated on Masculine (F I 8.10; p < .005) and Derived Androgyny (F I 8.35; p < .004), 120 Table 13 Table of Multivariate F Tests for Effects of Shame, Guilt, and Sex on Inventory Measures Ben Sex Role Inventogy Sources Shame Guilt Sex x High Anxiety Sex x High Shame Sex x High Guilt Sex x Low Anxiety Shame x Guilt Closure Flexibility Test Sources Shame Guilt Sex x High Anxiety Sex x High Shame Sex x High Guilt Sex x Low Anxiety Shame x Guilt RepressionISenaitization Scale Sources Shame Guilt Sex x High Anxiety Sex x High Shame Sex x High Guilt Sex x Low Anxiety Shame x Guilt * significant at .0125, as required ** significant at .0166, as required ***significant at .05, as required I .5681 .1986 3.6855 2.0291 3.6933 3.2688 .5787 .8722 .7026 .8643 .3250 .2389 .1872 5.2298 3.8789 2.5878 4.5149 .5956 .0531 1.4708 1.8444 .6846 .9387 .0073 .0948 .0073 .0140 .6787 .4577 .5524 .4618 .8074 .8691 .9050 .0021 .0513 .1104 .0357 .4418 .8181 .2277 .1770 * * * ** *** 121 Table 14 Table of Univariates (1) Ben Sex Role Inventory Source Sex by High Anxiety Variable Ben Feminine Bem.Masculine Bem Social Desirability Bem.Androgyny Derived (2) Source Sex by High Guilt Variable Bem Feminine Bem Masculine Bem Social Desirability Bem Androgyny Derived (3) Source Sex by Low’Anxiety Variable Bem Feminine Bem.Masculine Bem Social Desirability Bem Androgyny Derived (4) Closure Flexibility Test Source Shame by Guilt Interaction Variable Closure Correct Closure Incorrect Closure Scale I"! 1.2229 6.4879 1.7610 7.8983 .0646 8.1087 1.1659 8.3512 .2212 .7209 8.3812 1.4597 7.9273 4.1004 2.3366 .2711 .0122 .0058 .7998 .0052 .2825 .0046 .6390 .3976 .0046 .2294 .0057 .0451 .1291 *** 122 Table 14 (con't) (5) Repression-Sensitization Scale Source Shame Variable E p_ RIS 3.8789 .0513 **** (6) Repression-Sensitization Scale Source Sex by High Anxiety Variable * - approaches significance at .0031 ** approaches significance at .0125 *** - significant at .0160 **** - approaches significance at .05 123 though neither meet the adjusted alpha level (p < .003). Likewise, within the high anxiety group, a significant sex effect was isolated on the Derived Androgyny score (F I 7.89; p < .0058), though again, the adjusted alpha level was not met (p < .0031). Differences in cell means in both instances suggested effects attributable to actual sex differences of the respondents rather than to level of guilt. These findings reflected a higher masculine score for males than females, and derived androgyny scores that reflected a deviation from 0 in the direction of masculinity for males, greater than the deviation from 0 in the direction of femininity for females. Neither the guilt effect (p < .93), nor the shame by guilt inter- action effect (p < .67) was significant on any BSRI variables (See Tables 13 & 14). Results of t-tests demonstrated significantly higher M scores than F scores for high guilt group males (t-test: n I 16; df I 15, t I 3.20, p < .006) and nonIsignificantly higher F scores than M scores for females. Hypothesis 9 Shame scores will be positively related to sensitization; this relationship will be stronger than for guilt and sensitzation. Both shame (r I .11; p < .09) and guilt (r I .16; p < .03) were positively correlated (marginally significant) with sensitiza- tion. There was no support for a stronger correlational relationship between shame and sensitization than between guilt and sensitization 124 for the overall sample. In fact, guilt had a slightly stronger correlation with sensitization. Significant support was provided by the results of the multi— variate analysis with the R98 Scale score as the dependent variable. The analysis demonstrated a significant shame effect (high-low), as predicted (F I 3.87; p < .05). The high shame groups had a higher R-S mean (67.67) than the low shame groups (61.53). The guilt main effect (p < .11), and the shame by guilt interaction effect (p < .17) were not significant (See Tables 13 & 14). Although the shame by guilt interaction was not significant the pattern of cell means is noteworthy (Table 15). Of the four cell means for shame-guilt groups, the high shame group (t-test: n I 64, df I 62, t I 2.20, p < .03), high guilt group (t-test: n I 64, df I 62, t I 2.01, p < .04), and high anxiety group (t-test: n I 64, df I 62, t I 2.20, p < .03), obtained mean R-S scores were significantly higher than the low anxiety group, though no signifi- cant differences were found between the high shame and high guilt groups. This suggested that high shame and high guilt, each aspects of high anxiety, were related to increased sensitization. The same suggestion was prompted by the interaction of sex with shame and guilt levels (See Table 13). The sex effects were not significant by the adjusted alpha (.0125), but the pattern of obtained means was potentially instructive, suggesting heighted sensitization with increased shame and/or guilt (See Table 13 & 14). For example, high guilt group and high shame group females' obtained means were higher than for the other two groups of females. 125 Table 15 RIS Mean Scores by Shame-Guilt Level Compared to Total Group h Shame Males Females Both Low Shame Males Females Both Total and Byrne Sample High Guilt 74.68 61.43 68.06 66.87 65.43 66.16 67.11 Low Guilt 69.68 64.87 67.28 60.68 53.12 56.91 62.04 Total 72.12 63.15 67.67 63.47 56.27 61.53 64.60 Bygne 63.08 61.80 62.44 126 Similarly, the high shame and high guilt group males had higher sensitization scores than low anxiety group males, though high anxiety males had the highest sensitization scores. In every shame-guilt group, male RIS scores exceeded female RIS scores, indicating higher sensitization scores for males over- all. The sensitization scores for females in all but the low anxiety group were comparable to the total sample (n I 128) and to Byrne's female sample. Within the high anxiety group, this sex difference was significant by conventional criteria (F I 4.51; p < .035), though short of the adjusted alpha (p < .0125). Thus, the suggested sensitization effect seemed most marked for males. An additional notable finding is also relevant here regard- ing the low anxiety category. For the overall sample a negative correlation was found between the BSRI Social Desirability Scale and the R-S score (r I -.35; p < .001), indicating an inverse rela- tionship between sensitization and concern for the portrayal of a socially acceptable self image. Thus, in view of the unipolar measurement of repression and sensitization defenses on the R-S scale, concern with one's social appearance and at least in part, denial of negative aspects of the self, was positively related to repression. Observed R-S scores for the low anxiety group were non- significantly lower for both males and females than for the overall sample and for other shame-guilt categories. A multivariate analysis demonstrated that for the low anxiety group, a sex effect approaching significance was found for the Social Desirability scale 127 (F I 8.38; p < .0046: adjusted alpha at .003). Greater concern was shown by low anxiety group females with presenting a socially acceptable image, than by low anxiety group males, suggesting that the repressive trend was more marked for low anxiety group females than males (See Table 15). Evidence of support for a relationship between shame—guilt and the broad conceptualization of sensitization (i.e., awareness of the environment) was available in the significant shame by guilt interaction effect for Correct Closure responses on the CFT (F I 7.92; p < .005). Although Correct Closure was not corre- lated with shame (r I .05; p < .26) or guilt (r I .01; p < .41), the Correct Closure mean scores within the high shame group (105.40) and within the high guilt group (104.80) were significantly higher than the mean scores for the high anxiety (91.62) and the low anxiety (95.22) groups (See Appendix E); means of the high shame and high guilt groups exceeded the means of the high and low anxiety groups by more than 9 points (See Table 16). The high shame and high guilt group Correct Closure mean scores were not significantly different. Incorrect Closure mean scores were margin- ally significantly different (t-test: n I 64, df I 62, t I -l.41, p < .16). The data pattern suggested that either high shame or high guilt can increase the individual's sense of separateness from the environment as measured by the ability to make figure- ground discriminations. 128 Table 16 Shame-Guilt Groups: ‘Means for Closure Correct on CFT High Guilt Low Guilt High Shame 91.62 105.40 Low'Shame 104.80 95.22 129 Comparison of Main Study Results and Pilot Study Result§_with Reggrd to Pilot Studthypotheses Some hypotheses were of major concern only in the pilot study, but relevant data were available in the main study as well. These hypotheses predicted a negative relationship between shame and correct closure, as measured by Correct Closure scores on the CFT, and a positive relationship between guilt and correct closure. Also, it was expected that shame would be positively, and guilt would be negatively, assoCiated with femininity. Conversely, guilt was predicted to be positively, and shame negatively, associated with masculinity. Finally, it was hypothesized that masculinity and correct closure would be positively associated, and femininity and correct closure would be negatively associated. The purpose of comparing results of the main and pilot studies with regard to the pilot hypotheses (1-4, 7 and 8) was to provide additional information, further enabling this writer to assess the Lewis (1971) contentions, based on bipolar conceptualiza- tions of shame and guilt, sex role endorsement and differentiation. Results of the main study were relatively consistent with those of the pilot study. However, both studies provided only marginal support for hypotheses outlined above. In fact, with minimal exceptions, even the correlations which attained conven- tional significance levels were almost trivally low. Both shame and guilt correlated negatively but non- significantly with correct closure responses. This finding supported 130 the hypothesis regarding shame and was a reversal of the predicted relationship in regard to guilt. These results were consistent with the pilot findings. Neither shame (r I .03; p < .33) nor guilt (r I .04; p < .30) scores were significantly positively correlated with femininity. In the pilot study, shame and femininity were marginally positively correlated, and guilt and femininity were significantly (.001) posi- tively correlated. The direction of correlationin the main study was consistent with that of the pilot study in each case, but was not significant. Guilt was not found to be significantly positively correlated with masculinity in either the pilot or main study re- sults, a reversal of prediction. Shame proved to be positively and non-significantly (r I .12; p < .07) related to masculinity; this was inconsistent with the pilot hypothesis. Pilot results demonstrated a non-significant correlation but also reflected a positive direction of relationship. Correct Closure responses correlated non-significantly and negatively with femininity. The relationship was expected, though of a lower degree than found in the pilot study (r I .33, p < .001). The positive non-significant correlation between masculinity and correct closure was supportive of the Pilot hypothesis, but incon- sistent with the findings in the pilot study. Eight specific questions were tested in the pilot study; four were supported, while four showed a reversal of predicted direction of relationship. Three hypotheses were supported by the results of the main study and five showed reversals. In instances 131 of reversal, three hypotheses were consistent with reversals found in the pilot study results and two were not. A summarization of correlational findings are included in Table 17. Second Session Inventory and Behavioral Measures: Hypotheses 10-13 Results for Session 2 measures concerned the assessment of the effectiveness of the experimental manipulation employed in the second session, total group results (n I 128), and specific results relevant to the testing of hypotheses 10 through 13, regarding incidental learning, trait feedback reading, and trait recall. A multivariate analysis was computed for a four way design to assess the effects of shame level, guilt level, sex, and task condition (success-failure). Dependent variables consisted of time spent reading positive and negative feedback, total correct and total incorrect trait recall, and positive minus negative correct and incorrect recall. Alpha level for this analysis was set at .05 for guilt effect, shame effect and shame by guilt interaction effect. The alpha for shame-guilt by sex effect was .0125. In testing for main effects on specific univariates, the overall alpha was divided by the number of dependent variables (6). For shame effect, guilt effect, and shame by guilt interaction effectwuon 60m 5H maan moo. v a Si - u «N. v a Sf - a 8. v a 2. u u 3. v a “Col In H $12338: Son 3. v a Sf .- u 2... v a 8. - u mm. v a 8. .- u om. v a «O. n u flagged sum o>aamamm Hmuoa 08am uafizu 133 The implications of the lack of significant results will be dis- cussed in regard to specific hypotheses and elaborated in the discussion section. Additional, t-tests were computed for group and pairwise differences. Results are reported where relevant. All t-tests are reported in Appendix E. Manipulation Check Six questions were asked of subjects at the end of the second session of the experiment to assess the effectiveness of the manipulations in the session. 1. How did you feel about the feedback on the anagram solving task? 2. How did you feel about the feedback on the trait lists? 3. Did you believe the experimenter when he told you that you were above or below average on the anagram task? 4. Did you feel that the trait list feedback was truthful in relation to you? 5. In some cases the feedback on the anagram task was truthful, in other cases it was false. Which group did you feel you were in? 6. Some people got true feedback on the trait lists, others got false feedback. Overall, which group do you feel you were in? Questions 1 and 2 were used to Obtain subject reports as unbiased as possible by suggestions of the appropriateness of reject- ing feedback. Doubts about performance or trait feedback truthful— ness were expressed primarily on the more structured questions. Two questions requesting the individual to describe how they felt in response to the feedback (task and trait) prefaced the four, more objective questions regarding believability. The sub- jective responses on anagram feedback were classified into: «In mne mum Nne Nnm Nn one one mHmz mmmoosm HI «nu one HI mnu one o onu one on «no one mzo N oNIu «Nun NI nmnu mmno H mmnu mmno mu NNnu mmno Hum Nne mum Nae mum one hum one zoom H mHnu «Hun HI MHuu cHno H HHnu OHno mu mnu «Hun onm Hue Hum Hne mum mne mum Nne onsom H nu sun 0 nnu sun 0 .mnu one HI one one Hum Hne an Hne mum mne one Nne onz ousHHmm c has nun HI on» man H onu one NI one one mam one Nam Hne «no one mum mne zoom H HHnu QHno HI oHnu mHno o NHnu NHno o MHnu MHno mum one Nam Hne mum mne Nnm Hne onsom H mnu «no HI one sun c was one HI onu mno Nam Nne one one Hnw Hn Hum Nne onz mmooosm o onu one c on» one o nnu nno H mnu one mzo .xomoeomm omHom uow muonuo ammo snowmen man so Homoeoom men .mommo meow oH .m .mumHH uHouu was so Homnemom menu uow oHeooe meow .e o>oHHoomHo\o>oHHmm wxmou snowman o>oHHmomano>mHHmm ~50» ou eumwou oH Hnmnuouu may no mwmuo>o BOHoo so o>oom mums sch uoeu 30% mos xomoeoom umHH uHmuu mam umnu Home 50% eHn .N eHou on soe3 HousmaHuoexo men m>oHHoo oom eHn .H nOHuHesoo ouoHHmmlmmouosm eno xom .Hm>mu “Haas-mamam canoe: msoaummao sumac coaumasaaamz wH mHAmH 135 H-e oH-e one one NHne NHne oHno one oHeamo Hence o mm-u None N- owns omno o omnn oono e- Hons omne HH-m HHne no on no on oHne enm eHez Hence o HH-u Hone m- oN-u omno H eons mono o- None omno .- m-m Hne one - oHnm HHne one one Hence mtoHHeo N HN-u mono N- oN-u oonn H None Hone -,m- mono mono HHno eH-e one one oHne oHne HHno one Hence mmooooo H oH-s oH-o m- eons mono o None None m- Hons omno NH-e oH-e one one HHne HHne oHnm one Hence HHeum>o H oN-o oH-o m- HN-o omna o HN-u Hone H- Nona mono one one H-e one one one one one once H NH-u .HH-o H- mHnu oHno o HH-u HHno H- eHnu oHno none one one one one one Hno one oHeaeo o mun one c was one o on one H- ens one e-o m-e H-o one one one Hno one mHez nuoHHeo H «us one H- H-u one o one one H- Hnu one o-e one one one no on one one once o o-u one N- NH-u eHno o oHne oH-o m- one HHne one m-e Hue one one one one one oHeams H on» H-o H- Hus one o one one H- one one .mueeao .o .N .mmmeao .o .H ammwao .e .H emmeeo .m .H uoz uoz soz oez Au.eooo oH oHoea 136 ego-enhancing (e.g., surprised, delighted, good-feeling), neutral (e.g., simple factual statement, no feeling expressed), ego- depleting (e.g., bad feeling, failure, self-dissapointment), and other (e.g., not classifiable in any other category). Approximately equal.:iumbers of subject responses were included in each category (about 25% for each group). The subjective responses to the question regarding the trait feedback prompted the inclusion of approximately 75% of the sub- jects in the "agreed with feedback: totally or in part" category. The remaining 25% of the subjects either disagreed with the feed- back or gave neutral, vague, or no classifiable responses. This observational assessment of the data provides support for the viability of the manipulation and points to somewhat higher levels of reported belief than was apparent in the later, more objective, questions (3-6). Belief reported about the anagram task feedback in the first structured question (Question 3) was quite high (88.3%). As expected, when given the option to reassess the truthfulness of the feedback (Question 5), reported belief decreased somewhat (76.6%). Reported belief in the trait feedback (Questions 4 and 6) was moderate overall and lower than reported belief in task feed- back. Approximately two thirds (66.4%) of the subjects initially reported believing the personal trait feedback. It was expected that false personality feedback purporting to describe the indi- vidual would be suspected more than performance feedback. It is noteworthy, though, that when given the option to reassess the 137 truthfulness of the trait feedback (Question 6), subjects (both males and females in the success and failure conditions) did Egg report decreased belief (69.5%) (See Table 18). Thus the sugges- tion of false feedback increased reported rejection of the task feedback, but subjects were more hesitant to question the validity of the personal trait feedback. Although reported belief in the experimental manipulations was less than desired, the manipulations appear sufficiently effec- tive. Further, therewere no significant correlations between task or trait feedback belief and dependent variables of concern (time spent reading trait feedback or recall of trait feedback). In addition, the average reading time and the average trait recall were not significantly different for subjects reporting belief in trait feedback versus those. reporting lack of belief (Question 4 and 6). When responses to Questions 1 and 3 about task feedback were viewed within shame-guilt level, sex, and success-failure condition, only high anxiety group males in the success condition increased reported belief (1 person) when asked if they received true or false task feedback. All other groups showed either no change or lower belief (from 0 to -5). In regard to trait feed- back, all groups but one (low anxiety group males in the success condition) showed either no change or increased belief when asked if they received true or false trait feedback. 138 Attention to Positive and Nggative Feedback Attentiveness to, or "reading" of positive or negative feedback was operationally defined as actual reading (i.e., eyes directed to the feedback page), or more general focusing on a particular manual (i.e., actual reading of the manual alternating with other activity at the manual, including gazing at it, looking forward or up while in front of the manual, and returning to reading). The designation of "Not-reading," that is, not reading either positive or negative feedback, was defined by a variety of behaviors, including pushing the chair away from the table, closing the manuals and gazing around the room, reading other material (that the subjects may have brought into the room). Subjects were given 10 minutes (600 seconds) to read person- al trait feedback. Typically, subjects (50%) attended primarily to positive and negative feedback, each being read for approxi- mately 200-400 seconds, with a minimum of time spent not reading feedback (0-200 seconds). Twenty percent (20%) spent 0-200 seconds attending to each type of feedback, while the remainder of their time (200-400 seconds) was spent not attending to feedback. An additional twenty percent (20%) primarily focused on either positive (10%) or negative (10%) feedback (400-600 seconds) and spent the remainder of their time with the other type of feedback or no feed- back. The remaining ten percent (10%) responded with still other patterns of attention. Although subjects (n I 128) generally approached the positive feedback before the negative (64.9% approached positive first), 139 subjects spent approximately 20 seconds more attending to negative feedback (259.84 seconds) than attending to positive feedback (237.18 seconds). A noteworthy difference, if minimally significant, results from a t-test between mean positive and mean negative reading times (t-test: n I 128, df I 127, t I -l.71, p < .08). Significantly (.001) more time was spent attending to positive (t-test: n I 128, df I 127, t I 7.71, p < .001) and/or negative feedback (t-test: n I 128, df I 127, t I 9.24, p < .001) than was spent not attending to any feedback (102.90 seconds). The greater attentiveness to negative feedback occurred for both males and ' females and within both the success and the failure conditions. Males attended to positive feedback more than did females. Con- versely, females spent more time attending to negative feedback than did males. Subjects in the failure condition spent less time attending to either type of feedback, than in the success condition, and correspondingly less time attending to either positive or negative feedback, compared with subjects in the success condition (Table 19). A multivariate analysis assessing the effects of shame- guilt level, success-failure condition, and sex, on the time spent attending to positive and negative feedback was not significant (Table 20). Hypothesis 10 High shame (low guilt) subjects will spend a greater amount of time than high guilt (low shame) subjects reading negative 140 oo.ooH No.oH~ Hm.eoH om.HoH om.HHN oo.mH~ muoHHem Hm.eo NH.oo~ om.so~ om.Hm NH.oHN Hm.eom mmouoem moHeamo mH.oHH ~o.nm~ ~o.m- oo.ooH om.HmN on.wem ouoHHem Ho.HmH ms.mom Ho.eo~ m~.omH Hm.oH~ Hm.e~N mooooom ooHez mamsm 304 oo.oo oo.HnN oH.eo~ oo.HHH oH.oom oo.ooH moHeame oH.mo oH.eHN o~.~e~ oH.ooH oH.meN oo.He~ noHez eoaHnaou munHHmmlmmooosm oo.HoH oo.ooN o~.oe~ oo.ooH oo.~o~ oo.ooH onoHHeo Hm.om oH.mo~ om.HH~ oe.eHH oH.eoN oo.Ho~ mmouooo eonH oaoo mmxom NH.mHH oo.omm Hm.emm m~.HmH oo.ooH mH.ooH onoHHeo oo.om m~.HH~ mH.mH~ on.wo om.HNm oo.ooH mmoouom meHeamm om.~oH Hm.em~ NH.oeN Ho.Ho No.mom om.~H~ onoHHeo ma.o~ oo.mo~ m~.Ho~ oo.ooH No.oe~ NH.mNN oomooom mmHez HonuHoz m>Humwmz m>HuHmom umnuHmz o>Humwmz o>HuHmom osoem owH: “Hens uHeua SHeua oHeuH uHHoo sou “HHoo awH: ounHHmmlmmoooom eon xmm .Ho>oH uHHoo-oaeam an noeoeoom uHeua aoHs scoom Ameaouom eHo oaHH new: mH mHan 141 n.mNH m.mNH oo.ooH NH.ma unauHoz «.mmm h.mmu H.mmm oo.oom o>Hummoz uHouH m.om~ o.mH~ 0.0HN om.mm~ m>HuHmom uHeua om.NOH om.mHH OH.~w om.ma oo.oOH on.wh 0H.wNH o~.m0H oo.oa HoaSHmz Ae.uaoov aH mHan ow.mm~ ow.m¢~ oo.o~N om.wo~ om.HmN ow.nm~ oo.omm om.nmN oh.qu m>Hummoz oHeua 0H.mmm oo.Nmm on.wem on.wmm OH.Ho~ oh.mom oo.mq~ oh.wmm oo.omm mpHonom uHeua esouo Houoa ounHHmm mmooonm oOHuHeaoo oHoHHmmlmmooonm mm mmHmamm monz How an mHmuoH mmHmaom eoHez UQGHDEOU QHSHHGhImmOUUDW ousHHmm mmmoosm e95 @800 mmem Ae.usoov mamnm 30H 142 Table 20 Table of Multivariate F Tests for Effects of Shame, Guilt, Sex, and Condition, on Reading and Recall of Trait Feedback Sources Shame Guilt High Anxiety x Success-Failure High Shame x Success-Failure High Guilt x Success Failure Low Anxiety x Success-Failure High Anxiety x Sex x Success x Failure High Shame x Sex x Success x Failure High Guilt x Sex x Success x Failure Low Anxiety x Sex x Success x Failure Shame x Guilt Interaction I .7383 .2663 1.0353 .6031 .5786 1.6844 .7987 .6139 .3785 1.1116 .6815 .7036 .5851 .4418 1.2707 .6200 .9515 .4067 .7273 .7467 .1318 .5731 .7188 .8913 .3605 .6649 .6474 .7416 .8494 .2770 143 feedback, under both success and failure conditions. High guilt (low shame) subjects will not differentially attend to positive or negative feedback under the success condition. Hypothesis 11 In the failure condition, both shame and guilt groups will attend moretx>negative feedback than to positive feedback; however, the high shame group subjects will attend to negative feedback more than will the high guilt group subjects. The general tendency in both the success and failure condi- tions to attend more to negative feedback was accentuated for the high shame (low guilt) subjects in the success condition, though not significantly. The high shame (low guilt) subjects in the success condition spent non-significantly more time (283.10 seconds) attending to negative feedback, than did high guilt subjects in the success condition (248.70 seconds), in part supporting Hypothe- sis 10. In the failure condition, however, high guilt subjects spent non-significantly more time attending to negative feedback (257.50 seconds) than did high shame subjects (243.40 seconds), a reversal of the predicted relationship (Hypothesis 10 and 11). In addition, the high guilt subjects' time with negative feedback, in the failure condition, was almost 19 seconds greater than their time with positive feedback, though a t-test does not show this to be significant. High shame subjects actually spent more time attending to positive (248.70) feedback in the failure condition 144 than to negative feedback (243.40). Though a t-test showed this to be non-significant, it was a reversal of the prediction (Hypo- thesis 11). Overall, high shame subjects attended more to positive personal trait feedback than did high guilt subjects in both success and failure conditions, though the differences were not significant. High shame subjects spent more time attending to both positive and negative feedback in the success condition, than in the failure condition. Conversely, the high shame subjects spent much less time attending to feedback of either type in the failure condition, than in the success condition. High guilt subjects in the success condition specifically, were not significantly differentially attentive to positive (254.40 seconds) and negative (248.7 seconds) feedback (t-test, n 8 l6, df = 15, t = .10, p < .91), and in fact, spent more time reading positive than negative feedback. In view of the non- significant multivariate analysis and the absence of predicted significant differences in regard to high shame group subjects' performance (Hypothesis 10), the lack of a significant difference for the high guilt group is not particularly meaningful (See Table 20). An observation is appropriate in regard to the not-reading time scores. In both success and failure conditions, high guilt subjects spent approximately the same amount of time (success, 96.88; failure, 103.70) not reading any feedback (t-test: n = 32, df = 30, t = -.17, p < .86). High shame group subjects showed a 145 considerable difference between success (39.37) and failure (107.80) conditions (66.43 seconds) on the time spent not attending to feedback (t—test: n - 32, df - 30, t - -l.74, p < .09). Their not-reading score in failurewas higher (but comparable to) not- reading scores for the high guilt group in success and failure. The high shame group not-reading score in success was considerably dif- ferent from that of the high guilt group (t—test: n - 32, df - 30, t - 1.61, p < .12). Although the multivariate analysis demonstrated non-significance for guilt, shame, and condition effects, there remains a notable change, as seen in the marginally significant feedback within the failure experience than within the success experience. This suggested avoidance for high shame group subjects in the failure condition shifted the not-reading time scores into the average time score range for the high guilt group subjects, suggesting that the high shame group subjects' avoidance behaviors in regard to negative feedback became more similar to the high guilt group subjects in the context of failure (See Table 19). The large obtained time differential favoring negative feedback by high guilt subjects in the failure condition, and greater attentiveness to positive than to negative feedback in the success condition is noteworthy. Of the four major groups, the high guilt subjects are the only group that attended more to positive than negative feedback in success, though in failure a large nega- tive feedback differential occurred in both the high and low anxiety groups as well. Conversely, the high shame group subjects were the 146 only group who attended more to positive than negative feedback in the failure condition. These findings suggest a possibly greater sensitivity to negative feedback as a function of failure, and a greater sensitivity to positive feedback as a function of success, for high guilt sub- jects. This suggestion is thought provoking although the multi- variate analysis for success and failure effects was not significant. The high shame subjects showed a greater approach to positive feed- back than high guilt subjects in both conditions. The major inter- ruption noted for high shame group subjects in this pattern of approaching positive feedback, was their reduced reading time of either type of feedback in the failure condition. Incidental Learning Overall sample (n = 128) means for correct anagram solutions on memory-cued (5.53), tape-cued (4.69), and non-cued (4.44) words were comparable to one another, as were the standard deviations. The greatest difference (1 correct response) in scores occurred between the non-cued and memory-cued means. Natural ability on the anagram solving task was highly correlated with the use of memory- cued (r 8 .75; p < .001), and tape-cued (r = .76; p < .001) indicen- tal information. Likewise, the use of memory-cued and tape-cued information was highly related (r = .73; p < .001). This counters Mendelsohn and Griswold's (1966) report of a non-significant correlation between neutral ability on the anagram task and the use of cued information. Results of t-tests show significant 147 differences between mean memory-cued scores and neutral ability (t- test: n - 128; df - 127, t - -6.66, p < .001), and between memory- cued and tape-cued scores (t-test: n - 128, df - 127, t - 4.77, p < .001); and a notable difference between neutral ability and tape-cued scores (t-test: n - 128, df - 127, t = -l.50, p < .13). There were no significant correlations between RrS scores and any anagram problem solving indices. This was true even with statis- tical controls (analysis of covariance) for neutral anagram solving ability. This suggests an absence of relationship between sensiti- zation and the use of peripheral cues in incidental learning, countering Mendelsohn and Griswold's (1966) finding of a positive correlation between repression and low incidental cue use. Hypothesis 12 High shame (low guilt) subjects will perform better on the incidental learning task than high guilt subjects. A multivariate analysis for high guilt group (n = 32) and high shame group (n = 32) subjects' use of tape- and memory-cued responses, with a control for baseline anagram solving ability shows no significant differences between the two groups (F = .08; p < .92). Correlations were high between non-cued and memory-cued (r = .73; p < .001), non-cued and tape-cued (r = .79; p < .001), and memory- cued and tape-cued (r 8 .76; p < .001) variables. Consistent with the overall sample, an analysis of covariance on incidental learning data for high shame group and high guilt group subjects demonstrated a strong relationship between baseline anagram solving scores and the use of cued information. 148 Table 21 Mean Anagram.Solutions for High Shame, High Guilt, and Total Group Compared to Mendelsohn and Griswold (1966) Sample Nonscued Memoiy-cued Tape-cued X X S.D. S.D. S.D. High Shame Group 4.18 5.56 4.81 N-32 (2.50) (2.82) (2.76) High Guilt Group 4.15 5.43 4.71 N-32 (2.70) (2.75) (3.10) Total Sample 4.44 5.53 4.69 N-128 (2.64) (2.70) (2.78) Mendelson and Griswold 3.65 5.21 3.53 Sample (N-46) * * * *data unavailable 149 Table 22 Table of Multivariate F Tests High Shame by High Guilt (Covariance Controlled for Baseline) on.Memory-Cued and Tape-Cued Variables m 1' 2 High Shame Group x .08 .92 High Guilt Group Univariates m ' Memory .16 .68 Tape .02 .88 150 Recall of Trait Feedback Ten indices of trait recall information were available from the recall questionnaire given after subjects completed reading their trait feedback. The total sample recalled more correct (5.3 of 10) positive trait feedback items than correct negative trait feedback (3.6 of 10), and attributed to themselves slightly more incorrect negative traits (7.3) than incorrect positive traits (6.7). This pattern condition, for 12 of the 16 combinations of shame-guilt, sex, and success-failure condition. A reversal of this pattern occurred for high anxiety group males in failure, high shame females in failure, and low anxiety males and females in success (Table of data included in Appendix F). Hypothesis 13 High shame (low guilt) subjects will have greater recall scores about feedback material after success experience than after failure experience; this will not be true for high guilt (low shame) subjects. There was no support for greater feedback recall for high shame subjects on nine of the ten absolute scores (Table 23). The one possibly supportive index showed that in the success condition the difference score between correct positive and negative recall was higher (1.81) than in the failure condition (1.43). Neither a multivariate analysis on the effects of success or failure condition within shame and guilt levels, nor a multivariate analysis for nested task effects, were significant (Table 24). 151 There was no significant difference between recall scores for high guilt subjects in success or failure (F - 1.13; p < .34). The absence of differences in recall was not particularly meaningful in light of the lack of significant findings in regard to high shame (Table 24). Observation of the non-significant obtained mean recall scores for high shame group subjects showed that recall of correct and incorrect, positive and negative traits was slightly higher for high shame subjects in the failure condition than in success, a reversal of the hypothesis. Likewise, recall scores for the high guilt group showed that recall of correct and incorrect, positive and incorrect negative traits was slightly higher in the failure condition than in success. Correct negative recall, for high guilt subjects in the failure condition (3.00) was slightly lower than in the success condition (3.56). Within the success condition, high shame subjects recalled more positive feedback (correct and incorrect) and recalled less negative feedback (correct and incorrect) than high guilt subjects. In the failure condition high shame subjects recalled more correct (positive and negative) feedback than high guilt subjects and less incorrect (positive and negative)feedback than high guilt subjects. Although this writer expected decreased reports of recall of feedback for high shame subjects in the failure condition, due to the expected wish to hide an awareness of negative self information from others, the high shame group subjects showed greater accuracy in recall than was expected, and demonstrated, though not 152 Table 23 Mean Recall for High Shame and High Guilt Groups Within Success and Failure (Complete Table for all Groups, See Appendix F) Total Positive Correct Incorrect Tbtal Negative Correct Incorrect Total Correct Total Incorrect Difference, Positive minus Negative Correct Incorrect Conditions High Shame Success Failure 11.56 12.00 5.18 5.37 6.37 6.62 10.31 11.12 3.37 3.39 6.93 7.18 8.56 9.31 13.37 13.37 1.81 1.43 -.56 —.56 -1.25 High Guilt Success Failure 11.00 12.50 4.93 5.18 6.06 7.31 10.87 11.00 3.56 3.00 7.31 8.00 8.56 8.18 13.37 15.25 1.37 2.18 -.68 153 Table 24 Table of Multivariate F Tests Task Effects Within High Shame Group and High Guilt Group (N-64) Sources 1 2 Groups (high shame/high guilt) .2932 .8812 Task (success/failure) .5160 .7243 Task in High Shame .3494 .8434 Task in High Guilt 1.1349 .3494 Groups x Task . 9683 . 4321 154 significantly, what we have broadly termed sensitization (i.e., an awareness of the environment), a point worth noting. Summarygof Results Results of hypothesis testing below are ordered in the se— quence in which they appear in the Discussion chapter that follows. Field Independence 1. Shame: negative association with field independence. Both pilot and main study, non-significant negative association. 2. Guilt: positive association with field independence. Pilot study, significant reversal of association. Main study, non- significant reversal of association. For High Shame and High Guilt groups: t-tests demonstrate signifi- cantly higher field independence than for High and Low Anxiety groups. F-tests demonstrates significant Shame x Guilt interaction effect on field-independence measure. Defensive Style 9. Shame: positive association with sensitization will be greater than that between guilt and sensitization. Guilt has a slightly stronger, low level, positive association with sensitization, than shame. Both Shame and Guilt positively associated with sensitization. For High Shame, High Guilt, High Anxiety groups: t-tests demon- strate significantly higher sensitization than Low Anxiety group. F-test demonstrates significant Shame effect (high-low) on RrS Scale. 12. No significant differences found between High Shame and High Guilt groups on trait recall in success and failure. Those sub- jects (i.e., High Shame males, High Guilt females) who show low trait recall in failure, are also classed as sensitizers and masculine. Masculinity for the total sample is positively associated with defensive repression on the RrS Scale. Those subjects (i.e., High Shame females) who show high trait recall in failure, are also classed sensitizers, but are additionally classed andrognyous. Androgyny for the total sample is posi- tively associated with sensitization on the R-S Scale. 155 Attention Deployment 10 & 11. No significant differences were found between High Shame and High Guilt groups on attention to feedback in success and failure, or on the incidental learning task. High Shame sub- jects, expected to attend more to negative than to positive trait feedback in success and failure, and to do so more than High Guilt subjects, only acted as expected in success, reversals occuring in failure. While the High Shame subjects showed an expected stronger pattern of response than High Guilt subjects in success, High Guilt subjects exceeded High Shame subjects on negative feedback attendance in failure. When viewed by sex, there is support for the hypotheses for High Shame males, though not statistically significant, and a rever- sal forrHigh Shame females in the failure condition. Non-attendence to trait feedback.was increased significantly in failure for the High Shame group and High Guilt females. The pattern of non-attendence suggests that these subjects were disturbed by failure. Sex Role Adherence and Shame-Guilt 1. Shame: positive association with femininity and negative asso- ciation with masculinity. Significant support was found for a positive association between shame and femininity in the pilot study. In both the pilot and the main studies, a non-significant reversal is found between shame and masculinity, suggestive of a positive relational trend. Guilt: positive association with masculinity and negative association between guilt and femininity. The relationship be- tween guilt and masculinity is non-significant and inconsistent in direction in the pilot and main studies. There is a signifi- cant positive association between guilt and femininity in the pilot study, though this positive trend is non-significant in the main study. For High Shame: feminine scores will exceed masculine scores. This was not significantly supported. Non-significant results suggested support for females, a reversal occurring for males. For High Guilt: masculine scores will exceed feminine scores. This was not significantly supported. Non-significant results suggested support for males, a reversal occuring for females. Femininity: negative association with field independence. This relationship was supported (significant) in the pilot and main (non-significant) studies. 156 8. Masculinity: positive association with field independence. This relationship was not significantly supported and results were inconsistent between the pilot and main studies. Results often countered theoretical predictions of differ- ence in shame and guilt developed by this and other researchers. Contrary to the Lewis (1971) viewpoint on shame and, in part, consistent with the present writer's expectations, there seems to be a high level of psychological differentiation and sensitization for high shame subjects, as well as for high guilt subjects. Similarly, most other hypotheses presented by the present writer were reversed in the results for shame. The expectation that femininity and shame, and masculinity and guilt, would be positively associated is not borne out. In fact, femininity and guilt, and masculinity and shame, are positively associated, a reversal of expectation. Males were more available for inclusion in high shame criterion groups. Com- prehensive sex differences on shame, guilt, and level of differen- tiation variables, hypothesized by Lewis, were not found in the present study. High shame group and high guilt group subjects are similar in their level of differentiation and defensive style, therefore muted differences between these groups on the behavioral and inven- tory measures in the second session are not surprising. There are an absence of statistically significant group differences on atten- tion to feedback, recall, and incidental learning measures. Yet the patterns of the differences, while non-significant, are noteworthy and thought-provoking in light of the theory presented earlier in this paper. These trends have implications for the interaction of 157 gender, sex role endorsement, self disclosure, and attention deploy- ment, with shame and guilt variables. DISCUSSION This study assessed the relationship of shame and guilt, separately and together, with: the field dependence-independence dimension of psychological differentiation; the extent of masculine and feminine sex role endorsement; and the repression-sensitization dimension of defensive perception. Additionally, three related areas were studied. These were: attention deployment and recall, measured as seeking out and report of trait feedback within a success-failure paradigm, and the use of incidental cues in a learning task. The research design of the pilot study used a basically bipolar view of the variables of interest, tested with paper and pencil inventory measures. This approach was elaborated and modi- fiedth1the main study and most variables were treated as separate dimensions. Both inventory and behavioral measures were employed in the main study. In several instances, hypotheses involved comparisons be- tween subjects with high shame scores (and low guilt scores) and subjects with high guilt scores (and low shame scores, Hypotheses 5, 6, 9, 13). Hypotheses 1-4, 7 and 8, predicting correlations for the total sample were primarily tested in the pilot study and were rechecked in the main study and compared to pilot results. 158 159 In the following sections, methodological issues will be addressed, and the pattern of results regarding the relationship of shame and guilt with psychological differentiation, sensitization, gender, and sex role endorsement will be discussed. Methodological issues and the unexpected reversals in results have led the present writer to provide a refined conceptualization of shame and guilt, in order to facilitate future Operationalizations of these constructs. This discussion is more fully developed in Appendix G. Briefly, the reconceptualization focuses on the developmental roots of shame and guilt, and more thoroughly examines the relationship of these affects to concurrent development of other aspects of physical, intellectual-cognitive, moral, social, and emotional growth. These refinements draw heavily on the concepts of Odier (1956), Tomkins (1963), and Seligman (1975), whose direct relevance was not readily apparent at the initiation of this project. Methodological Issues Instruments The purpose of the present research has been the study of the differences between shame and guilt, in their relationships with other aspects of personality. A range of dependent measures have been studied including behavioral measures structured to maximize failure-based shame. This writer has tested hypotheses presented by other theorists, primarily Lewis (1971), and tested additional hypotheses that seem logically related to the shame and guilt varia- bles. 160 The findings of the present study do no demonstrate substan- tial support for Lewis's psychoanalytically-based theory, and often counter the logic of her hypotheses and reverse her findings. As was presented earlier, the present writer took issue with Lewis's use of bipolar definitions of the variables of interest, and with her presentation of logically questionable comprehensive interrelationships among the inventory variables of the present study. Overall, the greatest strength of the present study is the _ _..____;.._ _.,___._._. -.._i.r recognition of shame and guilt as separable, if not independent, , .- variables. The design of the study allowed for viewing the relative contributions of each variable to personality, in turn allowing for the differentiation of four basic styles of orientation. The use of these variable scores in combination for subject selection, constituted the four groups and enabled the isolation of effects on individual groups that could not have otherwise been obtained. The present writer views this method as superior to Perlman's (1958) design employing guilt- and shame-criterion groups, selected on the basis of the relative score. Within that study, the relative score (guilt score minus shame score) had a test-retest reliability coefficient of .35, while the guilt and shame scores test-retest reliability coefficients were .80 and .83, respectively. While Perlman focuses much of his discussion on developing the idea of modal guilt and shame styles of functioning, he could not even be assured of any significant consistency over time of the measures of the shame and guilt characteristics based on the relative score. 161 Thus, not surprisingly, the results of his study showed minimally significant or nonrsignificant findings. In the present study, the test-retest reliability coeffi- cient of the relative score was .53. It was higher than that found by Perlman, but too low to warrant its being employed as a method of subject selection for a study of shame and guilt as relatively consistent aspects of personality. Additionally, as has been dis- cussed earlier, the range of shame and guilt raw scores within the high shame and the high guilt groups, based on relative scores, fluctuated widely in the pilot sample of the present study. For example, a high shame subject might have a guilt raw score greatly in excess of the shame raw score. This finding solidified the pre- sent writer's belief that a raw score determination of shame and guilt, separately and together, would more accurately tap the shame and guilt phenomena. For the same reason, subject classification was based on a median-split assessment of the guilt and shame scores separately. On the other hand, Lewis (1971), while employing highly reliable instruments, essentially studied aspects of field indepen- dence and field dependence, among them shame and guilt. The major criticism of her study regards her tendency to emphasize the shame and guilt variables as if they, and not the field independence di- mension, were the independent variables. This is a dangerous procedure. The focus of her discussion suggests that guilt, field independence, and sensitizing defensive style go hand in hand, and likewise that shame, field dependence, and repressive defensive 162 style, show a similar pattern. A.more solid foundation of informa- tion regarding shame and guilt would seemingly have arisen, if Lewis had assessed shame and guilt as independent variables. It is suggested that such an altered design would have enabled the emer- gence of a greater range of differentiation levels and defensive styles in her subjects, rather that the consistent support for her hypotheses that were reported. Lewis (1971) presents examples of denial from shame-prone subjects and intellectualization from guilt-prone subjects, and strongly suggests that a relationship between these respective variables exists. Yet there was no effort to empirically study defensive style as it relates to shame and guilt or, as was pointed out above, no effort was made to empirically study defensive style as it related to differentiation. The present study employed both Bryne's (1961) Res Scale, and a recall measure after a success-failure manipulation to assess repressive and sensitizing defensive styles. These instruments allowed for the assessment of repression and sensitization as they related to both inventory measures of shame and guilt, and behavioral responses to the actual experience of failure. The method also allowed for the emergence of strong response trends that run counter to the Lewis theory. The use of the R-S Scale along with the behavioral measure employed also allowed for the recognition of a shortcoming in the present scoring structure of the R-S Scale, that being, an inability to separate those items reflecting internal vigilence from those 163 items reflecting external vigilence, within the sensitizing style. While the basic level of sensitization can be ascertained with the inventory measure, its single score structure is not necessarily pre- dictive of behavior in the anxiety-provoking circumstances of the present study. The use of multiple tests to reflect various aspects of sensitization, defined both in terms of defensive style, and atten- tion deployment, elucidate the complexity of the shame and guilt phenomena. Finally, the use of sex role endorsement variables viewed as separate dimensions provided more useful information than simple gender-based comparisons, and led to some unexpected findings. These findings countered earlier described theories of shame and guilt, and challenge previous research findings on defensive style and differentiation dimensions, that are based on viewing sex dif- ferences. Generally, the present method seems superior to other designs for the study of shame and guilt. The design acknowledges, both in subject selection and statistical analysis, the separate contributions of shame and guilt to personality, and allows for the assessment of varying shame-guilt level combinations. The use of a behavioral measure provides further information on shame and guilt and allows for a comparison between differential shame and guilt response level on an inventory measure and in an actual failure experience. Finally, the measures employed to assess defen- sive style, differentiation level, and sex role, as a substitute 164 for simple gender-based comparisons, all provided useful informa- tion. There were an absence of statistically significant results on the behavioral measures of this study, which called to question the strength of the success-failure manipulation. As has been reported, subjects' beliefs in.anagram task and trait feedback seems high based on responses to both openrended and specific questions. There is no way to accurately estimate whether the manipulation itself was either too weak or too strong. Hypothesized differences between shame and guilt in attention deployment measures were not found. This may have been a function of some aspect of the manipu- 1ation itself. Yet the results of the manipulation check on belief in the feedback, and the earlier reported similarity between shame and guilt on sensitization and field independence measures suggests otherwise (elaborated below). Unexpected reversals found in the results of the present study also led the present writer to review the effectiveness of the Attitude Anxiety Survey (AAS) as an instrument for the assess- ment of shame and guilt. The review uncovered a significant dif- ference (t-test: n = 128, df = 127, t = 12.44, p < .001) between the mean guilt and mean shame scale scores for the total sample, which actually strengthened the findings in regard to shame and had no effect on results for guilt. The results themselves and the posthoc review of the AAS served several purposes: it suggested that significant results and non-significant trends for high shame subjects might be more powerful than is immediately apparent; it 165 suggested a methodological refinement that would employ a single median cut-off point for guilt and shame to equalize the classifi— cation criteria; it prompted a reassessment of the operationaliza- tions of shame and guilt as they are represented in the AAS; it led to a refinement of the conceptualizations of shame and guilt and their developmental concomitants that may provide for more accurate future Operationalizations of the shame and guilt constructs. Subject Selection One implication of the disparity of mean guilt and mean shame scores is the parallel disparity of the medians for those scores on which subject selection was based. For three groups (i.e., high and low anxiety, high guilt) the variation between mean shame and guilt scores has no effect. For the high shame groups, individuals were selectedcn1the basis of scores exceeding the median for shame (median - 163) and being less than the median for guilt (median - 195). While meeting the criteria for inclusion in the high shame group, in some instances raw guilt scores still exceeded raw shame scores. The particular relevance of this findings is that while shame in some cases did not exceed guilt scores, there was still a significant shame (high versus low) effect foundcn1the RrS Scale that suggested heightened sensitization due to heightened shame. Similarly, several strong though not statistically significant trends (elaborated below) were isolated for shame group subjects and take 166 on heightened significance due to the differential scoring criteria for shame and guilt. The difference between shame and guilt medians that led to the inequity in the criteria for high shame group inclusion suggests a methodological change is necessary. One possibility is to use a single median for both shame and guilt scores. Another would be to retain the median structure, but only designate as high shame, those subjects whose shame scores exceed the guilt scores. These methods would increase the difficulty in obtaining subjects scoring high enough on shame to be suitably included in the high shame group. Yet such a methodological modification might very likely allow the strong shame trends found in the present study to attain statisti- cal significance. Qperationalizations of Shame and Guilt The difference between mean shame and mean guilt scores for the total sample suggested that shame items did not elicit the intensity of response that is found in response to guilt items. One area of speculation regarding the differential level of response is that of the test form chosen for the main study. The present writer selected the male— and female-appropriate (i.e., male or female names used in situation descriptions), Other form of the AAS, that includes the instruction that the subjects assess the anxiety level "that most peeple would have in response to the situation described." This selection was made in order to avoid defensiveness in personal responding, and viewed as a projective 167 assessment of the subject's own receptiveness. A review of the differences between mean shame and mean guilt scale scores on the Self and Other forms used in the pilot study showed that the discre- pancy was minimized by the use of the Other Form. A second area of speculation involved the revisions made in the AAS instrument by the present researcher. Yet revisions in the phrasing of several of the survey items were not numerous or extensive enough to explain the disparity of scores (see Appendix A). Several explanations of differences between the mean scores arise in the reassessment of the test content validity and the accuracy of the representations of shame and guilt. It was found that where guilt or shame items may be res- ponded to with high anxiety, the affect reported for the guilt items may be due to the triggering of guilt or shame, or both. Shame items seem fairly clearly related to the failure to live up to one's ideal self image, and seem.more limited to that. They focus generally on failures to perform; failures in ettiquette; being humiliated, ridiculed, criticized for failures to live up to others' expectations of the self; or not being physically or socially desirable. These items meet the content criteria for shame as earlier outlined. Guilt items are much less clearly limited to meeting the criteria for guilt without also overlapping into the shame category. Where the shame items maintain a focus on the self, with few options to place blame or responsibility on others or outside the self (typical of guilt), guilt items are not 168 limited to their definition. Guilt items often allow for the re- flection on the whole self, which to this writer is more shame- reflective. The guilt items then, seem to include both guilt and shame affect, and may prompt greater intensity of response. Most guilt items describe activities, advertantly or inadvertantly occurring, that counter social taboos and laws, which in turn may prompt heightened response intensity due to the establishment of a social desirability response set. Items of this type include suggestions of acting out aggressively and causing harm to others. For all items, although culpability is the major focus, one can infer a failure in self-control as well, causing a blurring of guilt and shame characteristics. For example: In an emergency, when no one is around, Larry steals money from his parents' hidden penny bank. The focus on culpability is apparent, but there is also a failure of self control and ideals. The severity of the activities involved in the guilt items are questionably matched to the shame item intensity, as in physical abuse versus not being as assertive on the job as one wanted to be. The format of the test is a random interspersing of shame and guilt items. The comparative lack of severity in the shame items, within the intermixed guilt items, may have prompted comparative responding, rathertfluulindividual assessment of each situation independently. Although this may provide information on the relative receptivity to shame and guilt, it may not be an accurate assessment of the actual shame or guilt level. 169 Additionally, Lewis (1971) suggests that most people view shame as more appropriate for children than for adults and that shame may be viewed as fggly about the self." The latter phrase carries with it the suggestion of adult shame about feeling shame! This may have come into play in the responses of subjects, parti- cularly given the transitional developmental phase that most subjects are in (i.e., college students). Lynd (1958) suggests that there is a hesitancy to admit that one has experienced shame or anxiety for certain situations "of no great import." In essence, shame items may have acted as a painful reminder of earlier or presently exper- ienced shame-receptivity, and an unwillingness to acknowledge its importance. One subject's feedback suggests some support for this view. She suggested that though anonymity was insured on the AAS, she felt vulnerable in response to some items, and was unwilling to record her honest response. On the other hand, the absence of intense response may have been prompted by the absence of the actual and immediate experience of the shame situations described. One of the major aspects of shame is its unexpectedness. It is a "jarring" change from previous unselfconsciounsess, into an acutely self-conscious state. Though it was expected that the items themselves would prompt a partial trig— gering of the shame and guilt affects, this may not have occurred. The emotional distance available in the paper and pencil test and the nature of shame itself, may have protected the subjects from this acute awareness of the situations described. 170 The present writer has attempted to isolate and assess as far as possible, the possible sources of measurement error, but no definitive statement can be made at this point. In addition to attempting to explain the reasons for the differential item response in the present study, this writer suggests that the AAS operational- izations of shame and guilt reflect particular conceptualizations of those affects. These conceptualizations in turn reflect the different values placed on the experiences of shame and guilt by researchers and, more broadly, by Western society. The conceptualizations of shame and guilt reflected in the AAS which led to the differential item response reflect, in part, Freud's relegation of shame to a peripheral role in Psychoanalytic Theory, while emphasizing the importance and centrality of guilt. While Freud (1914) did eventually acknowledge the potential for the internal occurrence of shame, his disciple (e.g., Fenichel) did not. These interpreters of Freud concretized the differences between shame and guilt into a matter of internal (e.g., guilt) versus external (e.g., shame) control. Secondly, it was assumed that the internalizing of parental and group norms would allow for the occur- ence of guilt in the absence of an exposure to others of a wrong- doing. Shame, on the other hand, viewed as external, was assumed to occur only with the exposure to others of one's own failures. These views of shame and guilt were the basisfor the popularity, among anthropologists of assessing guilt-based and shame-based cultures, operationalized as internally-versus externally-controlled moral structures. An obvious shortcoming in this approach was the 171 mutual exclusivity implied. Singer (1953) takes exception to this vieWpoint in a review of the anthropolgoical studies assessing shame and guilt variables, and amply demonstrated the simpleminded— ness of the internal-external dichotomy. He reassessed the research that has been carried out and pointed out the inconsistencies in methods of assessment, difficulties in assessing shame and guilt in other cultures, and fallacious interpretations of the data involved. His reassessment of the studies led him to point to the likelihood that both shame and guilt could be internal experiences and that each could occur in the absence of exposure to others. Overall, he maintained that shame and guilt could be equally powerful exper- iences. The development and cultural emphasis on transgression, guilt, and restitution, and the diminution of the importance of personal failure, shame, and self-reconstruction actually reflects the exact opposite of the later Psychoanalytic and related anthro- pological view. While guilt is internal, it is culturally empha- sized because it involves the social realm. Shame, while internal, is diminished because it is deeply personal and may be less directly involved in the social order. The orientation around guilt is reflective of the societal need for a common set of laws, mores, and norms that are shared, supported and upheld by most group members. In this sense, the social system reinforces abaseline morality reflective of the culture as a whole. The society reinforces the development of increasing responsibility of the individual as he or she interacts with the society and internalizes the societal values. 172 Guilt is viewed as developmentally more advanced than shame, in part, for this reason. Guilt receptivity develops after the recepti- vity to shame. Yet once one accepts the idea that shame is internal and can occur in the absence of others, there is no basis on which to suggest that guilt is more advanced. Shame implies an individual morality and values which may totally or only partially coincide with the societal needs for order. More often, shame-based values transcend guilt-based values, and embrace more universal themes (Lynd, 1958). It can also be noted that, in general, guilt arises (and is presented as such in the survey items) in response to more clear-cut and dichotomous situations, related to wrongdoing. Western society begins teaching apprOpriate behavior and the ramifications of trans- gression early in the individual's life. Highly formalized, visible methods and institutions for the assessment of guilt and culpability are present in the society. This is not the case for shame or the assessment of failure. Our society does not equally acknowledge, in a formal way, shame or shame-receptivity, nor does it provide institutionalized means for its expression or alleviation. Here again we point to the articulateness and visibility of guilt in our culture and the mutedness and invisibility of shame and suggest that this may have led to differential responding to guilt and shame items. The AAS portrays the shame experience as pedestrian. Items describe situations of minimal importance, covertly implying that shame is simply about the trivial, blown out of proportion. Guilt 173 items, on the other hand, describe the transgression of strongly sanctioned behaviors, attitudes and feelings. The survey is unsuc- cessful in communicating the full, personal, and internal implica- tions of the failure to attain one's own ideals or live according to one's own values. There may be a difference between rating how anxious one would be in response to "Sue was not as assertive as she would have liked to be " (a shame item) and "Sue did not come to the aid of a rape victim for fear of getting involved." Yet both may reflect failures of personal values. Shame can be about the "seemingly trivial" and can be overpowering in its occurrence, but too, it is not limited to those seemingly small occurrences. It is limited, if at all, by the extent of development of personal ideals and values. It is in this sense that the survey does not do justice to the shame phenomenon. Shame is not primitive when compared to guilt, but is the foundation of guilt-receptivity. Underlying guilt-receptive po- tential is shame-receptive potential. Guilt-receptivity involves the capacity to function efficiently within the culture, while shame-receptivity concerns itself with personal moral, value- and ideal-based underpinnings of the personality. The present writer, along with Tomkins (1963), maintains the validity of viewing the guilt affect as a derivative and modi- fication of the more central affect of shame. By this, it is 22E meant to presume that they are the same, but that they are linked developmentally and share some common features. While not fully developing this theme, Lewis (1971) also points out the relatedness 174 of shame andéguilt. She suggests that guilt may turn to shame in those instances where a guilt generating transgression is too massive to be remedied by restitution. The sequence of events following a severe transgression seems to move from active guilt anxiety to esteem modification in response, and to a realization of the inability to make an active restitution. According to Lewis, the individual would experience an "insoluble dilemma," which might trigger shame. The present writer suggests that shame arises after the realization that scrutinizing or excising a part of the self is not enough to reestablish esteem and there is no active restitution that can remedy the guilt-producing situation. At this point, the transgression comes to implicate the whole self and with it shame is experienced. At least in part, this is due to the inability to make active restitution, and the subsequent experience of passivity and helplessness regarding the situation. To the extent that shame receptivity is developed in the individual, a regression into shame anxiety can occur. This in turn may ultimately lead to a learned helplessness condition. This sequence of events is presented in order to acknowledge both the development of shame and guilt and the potential for regression in both. Also, this writer sees the need to emphasize both the relatedness of shame and guilt and their differences. As this study demonstrates, there are commonalities between shame and guilt, but as well, there are differences in their rela— tionship to other aspects of personality and to behavioral responses to life experiences. Either shame- or guilt-receptivity can be 175 excessively developed with negative ramifications for the individual, yet the present writer maintains that shame is no more primitive or external than guilt. They are both highly complex phenomena, equally important, arising from, and appropriate in response to, different aspects of the living process. Defensive Sensitization and Field Independence Both high shame and high guilt group subjects demonstrated heightened field independence (Closure Correct, on CFT) and height- ened defensive sensitization (Sensitization, on R-S Scale). The results for guilt were eXpected and are congruent with findings pre- sented by other researchers, primarily Lewis (1971). The findings for shame were, in part unexpected. While the present writer re- tested Lewis' contention of a positive association between shame and field dependence (Hypothesis 1), her contention of a positive relationship between shamezand repression was reversed by this writer. It was hypothesized (Hypothesis 9) that high shame subjects would show heightened internal and external vigilence (i.e., sensi- tization), and that this relationship would be greater than for high guilt subjects and sensitization. In fact, both shame scores (r = .11, p < .09) and guilt scores (r = .16, p < .03) were positive- ly correlated at low levels with sensitization, for the total sample. In addition, a significant shame effect (F test: F = 3.87, p < .05) occurred for the RrS Scale. The level of shame (high-low) was associated with the extent of sensitization. No comparable signifi- cant guilt effect occurred. When group mean ReS scores were compared 176 across the four groups, t-tests showed a significantly higher sensi- tization for high anxiety, high shame, and high guilt groups, than for the low anxiety group. Similarly, both the high shame and the high guilt subjects demonstrated significantly higher mean Correct Closure score than overall high and overall low anxiety groups (based on t-tests, see Appendix E). When these two groups are viewed, it seems that high guilt (low shame) 25_high shame (low guilt), are associated positively with heightened levels of field independence. Total scale scores (i.e., differences between Correct and Incorrect Clo- sure scores) of the two groups are classified as high, compared to a medium level (based on the CFT norms) of field independence for the overall high and overall low anxiety groups. The results on the differentiation measure point to the value of the multiple group methodology employed in the present study, since the results for the high shame and the high guilt groups are submerged whentfluatotal sample is considered. Both shame and guilt scores are found to be non—significantly and negatively correlated with Closure Correct socres. These findings would suggest no apparent significant association between shame and guilt variables themselves and the extent of differentiation. The findings of heightened defensive sensitization and heightened field independence are consistent with other empirical literature linking the two variables. They each are seen by this author as an expression of analytic style within the global-analytic cognitive style construct. 177 In regard to shame, the findings of the present study are contradictory to those presented by Lewis (1971). Her findings suggested that subjects with extreme field dependent scores on the Rod and Frame Test produced more shame-oriented than guilt-oriented verbal material in initial sessions of psycho-therapy and employed more repressive than sensitizing defenses. High field independent subjects expressed more guilt- than shame-oriented verbal material and employed more sensitizing defenses in the same situation. It is worth noting that the present study does not parallel the methodology employed by Lewis, nor did it employ the same mea- surement instruments or kind of sample. It is also noted that while Lewis reports samples of repression from shame-prone subjects, she made no effort to empiricially document her findings. DeGroot (1968), employing Lewis' methodology, was not able to replicate her results, and instead found a positive association between both shame and guilt and field independence, consistent with the present study. I.ewis' theory is logical. She suggests that field indepen- dence arises as a function of the maturation of the organism, as does guilt. In this view, there is a parallel between the increas- ing separation and autonomy of the individual, and the guilt exper- ience with its more clearly defined boundaries. With regard to shame, she suggests that the (earlier) transitional stage during which shame capacity is developed is marked by incomplete separation, a lack of full autonomy, and a greater reliance on parental emotional support. Her view suggests that the shame-prone adult would retain 178 an outward orientation and a felt lack of separateness. Similarly, based on the empirical literature relating field independence with sensitizing defensive style, she suggests that guilt and sensiti- zation are positively related, and conversely, that shame and repressive defensive style are related. Yet the results of the present study, along with DeGroot's replication, do not support her contentions. In addition to pointing out general methodological inade- quacies, the present writer suggests that Lewis's choice of a psychiatric sample, maximized the likelihood of her findings, and for that population her findings may be accurate. The Lewis study was carried out on a small sample of individuals who had just entered outpatient psychotherapy threatment. The research situation and subject sample, that is, individuals who had problems they felt were severe enough to prompt the seeking of professional help may have fostered both the surfacing of field dependence and subse- quent shame. In addition, while Lewis found heightened shame- oriented productions and repressive defenses for extremely field dependent subjects, she takes the questionable liberty of reversing her findings, suggesting that shame-prone subjects can be assumed to be field dependent, and repressors. The present writer interprets the findings of heightened field independence and sensitization for high shame subjects in the present study, as a result of extensive shame experiences in childhood. A part of shame seems to be forced emotional separation leading the individual to develop a heightened separate sense of 179 self as reflected bathe high field independence scores. The reality, therefore, may be that there is a possibility of high field inde- pendence in both high guilt and high shame. The reasons for their development though may differ: the field independence in guilt being a natural function of maturation; and in shame, a premature development due to mistrust of the potentially shaming environment. This interpretation might at some level of testing allow for the surfacing of differences between shame and guilt on separateness variables. It is suggested that the inconsistency of results in the Lewis and the present studies on the differentiation dimension may relate to a difference in the level of testing. Lewis's use of the Rod and Frame Test involves kinesthetic as well as visual perceptual senses, whereas the present study use of the Closure Flexibility Test involves only the visual sense. One very real possibility is that typically, the shame-prone individual might function quite adaptively, yet on a more basic (e.g., kinesthetic) level, or when in conflict, may be less separate. It is tempting to discount the Lewis finding as biased due to her use of psychiatric sample or methodological inadequacies in that study, but such an argument might preclude the consideration of the consistency within the seemingly inconsistent findings. From the present writer's understanding of the shame phenomenon and from clinical observation, the shifting from autonomy and separateness to reliance on the external environment is a characteristic of the shame expression in adults. Lynd (1958) points out that a part of the power of shame is that it occurs abruptly and unexpectedly at 180 a time when the individual may be functioning with relative unself- consciousness. This writer suggests that the unself-conscious state uxay'parallel the field independent stance. Too, the source of shame can be internal or external. It is possible to consider the internal triggering of shame, which in turn may prompt a reduction of the boundary with the environment, in turn increasing attention to, or reliance on that which is external. An external source of shaming, if effective in activating shame would, by definition, also increase the salience of the environment. It seems inappropriate to suggest that the separateness in shame is defensive or less real than the reliance on the external environment. To the extent that the shifting is productive or detrimental to the maintenance of esteem and contact with reality, either could be defensive or adaptive. With regard to the positive association of high shame and sensitization, it is suggested that the subject who, as a young adult, is differentially oriented around heightened shame-receptivity (high shame score), as a child experienced recurrent shame. Whether because of specific shaming by the parents, failure- generated shame, or inadvertant shaming, the individual learned to mistrust the environment and the self. According to Lynd (1958), the mistrust arising from shame is caused by the child feeling betrayed by the significant other in the environment, and coming to mistrust the self for having tursted the need-frustrating other and having expected affirmation from them. In response to the break- ing of the emotional bond in shame, the individual blames the self 181 for having expected a positive response from the other. Rather than expressing the anger toward the other, and possibly further lessen- ing the likelihood of future affirmation, anger comes to be self- directed. The individual becomes mistrustful and less willing to express needs for affirmation to others. The mistrust prompts the establishment and maintenance of vigilence of external and internal events, in order to protect the self from the potential recurrence of painful experiences. Here, too, the logical relatedness of sensitization and field independence is apparent. The present finding is consistent with Bryne's (1964) finding of a positive association between punitive parental disci- plinary techniques and the level of sensitization in their off- spring. The interpretation provided by Byrne suggests that sensitization develops out of a felt need to monitor the level of potential threat in the environment in an effort to avoid punishment. Vigilence is active in guilt as well as in shame, and likely is interpretable in the same way. Yet in guilt, as has been pointed out, there may be less need for pervasive vigilence, since guilt arises in response to activities that have clearer parameters. The isolation of a significant shame effect on the R-S Scale supports this suggestion of a further need for high shame group subjects to maintain an awareness of internal and external cues. The present writer therefore views the heightened level of defensive sensitiza- tion as logically arising from the early experience of shame and guilt. 182 On both the field independence and sensitization measures in the study, the commonalities between shame and guilt seem to be expressed. In both shame and guilt there tends to be a heightened awareness of internal and external events and heightened separate- ness. A lack of statistically significant differences emerges on attention deployment, incidental learning, and trait recall masures of the second session. All three measures reflect, to varying degrees, the extent of sensitization, therefore, muted differences between the high shame and the high guilt groups are not surprising. Several interesting trends do emerge, though not statistically significant, regarding the effects of failure on the high shame and the high guilt group subjects. These findings are relevant to the overall viewpoint of this paper and will be developed below. Attention to Feedback Several hypotheses in the main study were structured to test the behavioral effects of success and failure conditions on atten— tion deployment for high shame and high guilt group subjects. This author is using the term sensitization in the present discus— sion to denote attention to internal and external events, Operation- zlied here as attention deployment, rather than in its earlier defined defensive sense. The behavioral measure of attention deployment used in this study was based on the paradigm developed by Mischel, Ebbesen and Zeiss (1973). Their data demonstrated that under a condition where no further testing was expected (paralleling the present design), subjects in the success condition attended 183 significantly more to assets (i.e., positive feedback) than to liabilities (i.e., negative feedback) and in the failure condition, significantly more to liabilities than to assets. Within the total group findings, they found that sensitizers (R—S Scale) showed significantly greater attendance to negative (than to positive) feedback in failure, and greater attendance to positive feedback (than to negative feedback) in success. As well, they found that sensitizers attended more to feedback of any type than did repres- sors. Repressor and sensitizer designations were established on the basis of above or below median RPS scores for the total group (as was done in the current study also). In order to adequately compare the present results to those of Mischel, Ebbesen and Zeiss (1973), a posthoc analysis was carried out on the interaction of attendance patterns with median- based repressor or sensitizer designations. Additionally, median— based sex role designations were reviewed for subjects, in order to assess their interaction with both attendance patterns and repressor- sensitizer designations. For the total subject sample (n = 128), a median (median = 63) was computed for the Res Scale, and subjects in the two groups of interest (i.e., high shame group, high guilt group) were classi- fied as above the median (sensitizers) or below the median (repres- sors). The total sample median, comparable to the total sample RrS Scale mean, was in the sensitizer range when compared to Byrne's (1961) norms. 184 It was expected that high shame group subjects in both success and failure conditions would attend more to negative than to positive feedback, and that this trend would be stronger than for high guilt subjects. The prediction was made because of the general view in this study of shame as involving increased vigilence of both internal and external potential threat, thereby heightening interest in and awareness of negative feedback. High guilt subjects were not predicted to differentially attend to positive and negative feedback in success; they were expected after failure to read more negative than positive feedback, but this was expected to be less strong a preference than for high shame subjects. None of the results directly relevant to the hypotheses were significant. However, there were some distinct trends, indicative of a general tendency for high shame subjects, particularly males in success, to read more negative (283.10 seconds) than positive feedback (277.50 seconds) while high guilt subjects read more posi- tive (254.40 seconds) than negative (248.70 seconds) feedback. Conversely, in failure there was a tendency for:high shame subjects, particularly females, to read more positive (248.70 seconds) than negative (243.40 seconds) feedback while high guilt subjects, parti- cularly females, read more negative (257.50 seconds) than positive (238.70 seconds) feedback. In general, the hypotheses (Hypotheses 10 and 11) are supported only in the success condition, reversals occurring in failure. When results are viewed by sexes separately, some further 185 support for the hypotheses regarding attendance patterns in failure for high shame males is provided. The high shame males in success and failure, acted as hypo- thesized, though not as strongly as expected. High shame females acted in reversal to the hypotheses in both success and failure, and exaggerated this reversal in failure. High guilt males behaved as predicted by not significantly differentially attending to positive or negative a feedback in success or failure. High guilt females, also as predicted, showed non-differential response in success. This is reversed in failure with significantly (t-test: N = 16, df = 15, t = -.51, p < .04) greater attention to negative than to positive feedback. In success, the high and low anxiety groups and the high guilt group demonstrated a relatively high "not reading" time score (average, 103.19 seconds out of 600 seconds), compared to a very low "not reading" time score by high shame group subjects (39.37 seconds). One exception to this general findings is that high guilt females in the success condition also have high overall reading times (not reading = 37.50 seconds), paralleling the high shame male and female respondents. Within the failure condition, the "not reading" scores for all groups (sexes combined and separate) are comparably high (average, 102.9 seconds). This represents a significant increase in "not reading" scores for the high shame group (t-test; N = 32, df = 30, t - -1.74, p < .09) and high guilt females (t-test: N = 16, df = 15, t = -l.64, p < .08). The high shame group and the high guilt females seem to be more avoidant of 186 feedback in general as a result of the experience of failure, though an F Test for a condition effect within groups was not significant. Thus, one general finding concerning the effects of failure seems to be an avoidance of reading feedback, most marked for the high shame group and high guilt females. Also, actual reading time varied from exaggerated attendance to positive feedback (high shame females, high guilt males), to balanced attendance to both positive and negative feedback (high shame males), to exaggerated attendance to negative feedback (high guilt females). The present study findings do not show significant differ- ences in attendance patterns based on condition. While high guilt group and high shame females read more positive than negative feedback in success, the margins only range from 2.5 to 6.25 seconds greater in the positive direction, high shame males in success show a reversal, reading slightly more (11.25 seconds) negative than positive feedback in success. Thus, while the direction of reading scores is partially consistent with the Mischel et a1. findings; the results are neither as strong nor pervasive. In failure, high shame males and high guilt females read more negative than positive feedback, consistent with Mischel et al. findings, while high shame females and high guilt males demonstrate a non-significant reversal of this pattern. Similarly, the results of the present study do not support the Mischel et al. findings regarding sensitizers and repressors. Notable for the high shame males and females and high guilt females was the variety of attendance patterns within reduced overall 187 attendance to feedback in the failure condition. Yet in all three groups there are more sensitizers than repressors. Additionally, all three groups have high numbers of andro- gynously classified subjects. For the total sample there is a small but significant positive correlation between sensitization and the derived androgyny score (r - .14, p < .05). These groups contrast with the high guilt males with their balance of sensitizers and repressors, and fewer numbers of androgynously classified subjects. All of this suggests an interaction of defensive sensitization and avoidance of feedback. It seems that in failure, for the three groups with lower overall attendance, high numbers of sensitizers, who are also androgynously classified, tend to be disturbed by the failure experience and avoid feedback. Conversely, the high guilt males, composed of a balance of repressors and sensitizers and few andro- gynously classified subjects, tend to increase their attendance to feedback in the failure condition. Regarding the overall avoidance pattern, it is suggested that avoidance is at least in part due to the fact that, in failure, attention is taken by internal activity. As is pointed out earlier and above, both the high shame and the high guilt groups demonstrate heightened sensitization scores and within the three specific groups, a majority of subjects are sensitizers. One would usually expect the sensitizers would more readily approach feed- back of either type, but the sensitizing defensive style includes heightened awareness of both internal and external potential threat. 188 Table 25 Median Sex Role and Repression-Sensitization Classification with Sex and Condition for High Shame and High Guilt Groups High Shame High Shame High Guilt High Guilt Success Males Females Males Females Androgynous 2 l l 2 Undifferentiated 4 2 3 2 Masculine 2 4 3 l Feminine 0 l l 3 Total 8 8 8 8 Sensitizers 3 3 4 3 Repressors 5 5 4 5 Failure Androgynous 3 3 2 3 Undifferentiated 1 2 3 2 Masculine 3 l l 3 Feminine 1 2 2 0 Total 8 8 8 8 Sensitizers 5 5 4 Repressors 3 3 4 3 Median for R-S Scale = 63 Medians for BSRI = Masculine Feminine Males 5.15 4.80 Females 4.60 4.80 189 It is feasible. therefore to suggest that the actual occurence of failure (e.g., threat) is associated with self-preoccupation, internal focus, and a decision to reduce attention to feedback of any type for these subjects. When the groups of sensitizers and repressors are viewed without regard for their high shame or high guilt status, the pattern of lessened attendance to feedback of any type in failure as compared to success, is again demonstrated by sensitizers. Beyond this suggested internal preoccupation-based avoidance, variations in the actual reading patterns occur. With regard to the non-significant sex and condition variation in the actual reading patterns for high shame and high guilt groups, the use of RPS designations or sex role classification provide no solid information. Subjects designated repressors or sensitizers within each group show a variety of attendance patterns, some reading more positive than negative feedback, and others, the reverse. High shame males and high guilt females (as groups) read more negative than positive feedback in failure. In reviewing the sex role classifications for these groups, the majority are androgynous, with the remainder primarily classified masculine. The present writer suggests that the presence of masculine classified subjects in increased numbers relative to feminine or undifferentiated classifications may be associated with the differential attendance to the negative feedback. Masculine characteristics such as assertiveness, self control, and rationality, representing an instrumental orientation, might be 190 associated with the need to attend to negative personal information, in order to explain the task failure to oneself. In the success condition there are greater numbers of repressors than sensitizers for the high shame subjects and high guilt females. The overall attendance to feedback is high. The actual pattern of attendance emphasizes the positive feedback more than the negative feedback, except for high shame males who show a non-significant (11.25 second) difference in the direction of more negative than positive feedback reading. It would seem viable that after a success experience, the individual's esteem may be bolstered and prompt more attention to feedback in general. The overall posi- tive direction of reading scores is consistent with the Mischel et a1. findings, though not significant. They suggest that success prompts the reading of more positive feedback than negative, and more generally, an orientation around maintaining the success-induced positive feelings. Mischel, Ebbeson and Zeiss (1968) have found similar results in another study, pointing out that success seems to prompt children to increase non-contingent self-gratification in the form of self-administered tokens. Isen (1970b Isen and Levin (1972), Berowitz and Connor (1966) have found that positive res- ponses toward others increase more after a personal success than after a personal failure. The general interpretation is that "feel- ing good" prompts activities that attempt to maintain the good feelings. The present findings, within the context of the previous studies, suggest that while the number of repressors in the high 191 shame and the high guilt female groups is high in success, the success condition has on other occasions been enough to prompt high attention and orientation around positive feedback. While the impact of success alone is important to acknowledge it must also be remembered that while high guilt males show a greater attention to positive than to negative feedback, consistent with the other groups, they have a reduced attention pattern in both success and failure. The number of sensitizers (n = 4) and repressors (n a 4) was not different across conditions, nor is there any clear trend in their sex role classifications that would explain their lowered overall attention pattern. This inconsistency in their attendance pattern therefore cannot be explained based on the variables tested. In general, when sexes within groups are viewed separately, a nonsignificant pattern of interaction between avoidance of feed- back and sensitization in failure is apparent. While the success condition in and of itself may be enough to explain the high atten- tion to feedback by these groups, there is, as well, a heightened presence of repressors. While the avoidance behavior in failure is a reversal of what would typically be expected of sensitizers, the results are bolstered by correlational support presented elsewhere in the study. Lewis (1971) has suggested that high shame subjects tend to employ a repressive defensive style and high guilts subjects, a sensitizing defensive style. A high mean sensitization level for both the high guilt and the high shame groups counters her theoreti- cal viewpoint. Although these two groups have high levels of 192 sensitization, in an actual failure experience, there is a rela- tively consistent pattern of lessened attention to, or avoidance of, feedback. The findings, in part, uphold the view that the high shame subjects are affected by failure. Failure seems related to avoidance. The present writer though points out that avoidance behavior, manifested as selective attention, does not constitute what Lewis and others describe as Psychoanalytically defined, dynam- ic, defensive repression. Nor is there support for the occurrence of repression or denial on the recall of trait feedback measure (elaborated below). Not all the groups that show an avoidance pattern in attendance "repress" the recall of the traits. What seems to be occurring is a behavioral manifestation of the inter— actioncufa high level of separateness, with the internal aspect of sensitization of a high level of separateness, with the internal aspect of sensitization and with high shame. For those individuals with a heightened shame receptivity, failure seems to activate that aspect of sensitization that is internally focused and prompts the sensitizer to pull back and be wary of the information offered after a threat to esteem has been experienced. The present writer draws on Lewis' work to understand the similarity of the high guilt females and the high shame group avoidance pattern. It has already been suggested that shame and guilt are not mutually exclusive. An overall guilt style may give way to a responsiveness to shame. The failure experience maximizes this possibility. Lewis suggests that, based on traditional sex role training, females may have a lower threshold for actual shame 193 receptivity than males. This seems a viable interpretation, particularly given the mix of shame-typical and guilt-typical behaviors demonstrated by the high guilt females in the failure con- dition. The present writer is willing to consider that actual failure, early established receptivity to shame, based in part on traditional sex role training may occur, regardless of later developments of more androgynous characteristics. As has been dis- cussed by Odier (1956), there in fact may be a carrying to adult- hood of early unchanged or only partially matured cognitive, emotional, and moral patterns that may be triggered totally or partially (elaborated in Appendix G). This may explain the overall avoidance for the three groups in failure, and account for the individual group variation in actual reading patterns as well. Memory of Feedback It was hypothesized on the basis of Lewis's theory that high shame group subjects would have lower recall of personal trait feedback after a failure than after a success (Hypothesis 13). The data did not support this expectation. In fact, high shame subjects (sexes combined), even with their decreased over— all reading time scores, show nonsignificantly greater accuracy in the recall of positive (5.37 of 10) and negative (3.93 of 10) traits in failure than in success (positive, 5.06; negative 3.37) and greater accuracy than any other group in the failure condition. The high shame subjects also tend to attribute to themselves fewer incorrect traits than the high guilt group. 194 The assessment of recall was intended to be an objective, indirect measure of the extent of repression or sensitization acti- vated by failure. The recognition format of the paper and pencil measure may have been too objective to trigger a sense of vulnera- bility and resistance to recall. This remains a methodological question. However, it does not explain the unexpected heightened capacity for recall, especially after failure, demonstrated by the high shame subjects, though it is consistent with their earlier described heightened sensitization level. This finding suggests a further undermining of support for the definitive relationship hypo- thesized by Lewis (1971) between shame and the use of repressive defenses. When viewed by sex, these nonsignificant trends are not as unified as the group differences suggest. It seems that failure is associated with an increased accuracy in recall and report of trait feedback for high shame females and a decrease for high shame males. High shame females, even with their overall lower reading time scores and their specific increase in positive feedback reading in failure, show the greatest accuracy of all groups. This is a re- versal of the hypothesis. High shame males act as expected though not to the extent expected. In failure, they read slightly more negative than positive feedback, and have the least accurate recall and high levels of incorrect self-attributions of all groups. This occurs even though their sensitization and differentiation scores are high overall and higher than comparable scores for high shame females. High guilt females tend to be less accurate than high 195 shame females, having a recall pattern less extreme but similar to the high shame males in failure. The seeming search for the reasons for failure, carried out by the high guilt females in their increased negative feedback reading within a decreased overall attention pattern, does not increase their ability to recall the feedback given. In fact, it is associated with decreased accuracy. High guilt males demonstrate a moderate level of recall and a compara- bility of scores scores across the two conditions, as expected. It would seem that the failure, in addition to being asso- ciated with decreasing the overall reading time scores for both sexes, is also associated with affect that interfered with the accurate recall of trait feedback for high shame males. This suggests that high shame males show a more long term effect from task failure than do high shame females. High shame females, on the other hand, tend to recall feedback more accurately than in success. Their recall behavior is consistent with the earlier discussed sensitizing perceptual style. High shame females, while being effected by the failure, seem to deal with it by taking a complete and accurate assessment of the personally relevant environmental information, maintain it consciously, and have it available for report. The findings is consistent with Binder's (1971) report of high shame for those females who are aware of their self-ideal self discrepency. The behavior of high guilt females reflects some effect in response to the failure experienced. As with the high shame males, there is a decrease in the accuracy of recall, and an increase in 196 the number of incorrect negative self-attributions. This also suggests some longer term effects of the failure experience. The variation in recall across the high shame male and female and high guilt female groups in the failure condition prompted a reviewing of the RPS median-based designations and the sex role classifications of the subjects of each group. The groups are primarily composed of sensitizers with highnumbers of androgynously classified subjects. Yet, the high shame female group, in addition, has only one masculine classed subject. This group has high and accurate recall of trait feedback. The high shame male and the high guilt female groups have three masculine classed subjects in each, and have low recall (see Table 25). The presence of increased numbers of masculine classified subjects in these groups may be coincidental. An effort was made to assess the sex role classifi- cations more stringently by employing an additional criteria (beyond the median structure) of including in a classification, those subjects whose scores on masculine and feminine scales were .05 or more from the median in either direction. This technique allowed for the reclassifying of some subjects whose scores on the masculine and/or feminine scales were close to the median. For example, while a subject might have initially been classified as masculine based on having an above the median masculine score and a below the median feminine score, the masculine score may have only been .02 above the median. With this refined technique, the mascu- line score would be viewed as below the more stringent requirement, and the subject would be classified as undifferentiated, based on 197 having both masculine and feminine scores that did not meet the more stringent median requirements. With this method the number of masculine classed subjects remains the same for the high guilt female group (n - 3) and increases by one for the high shame male group (n - 4, Table 26). These groups show low accuracy and high nubmers of incorrect self-attributions. There is a significant negative correlation, for the total sample, between the extent of sensitization and masculinity (r I -.14, p < .05). The implication of this correlational finding, based as it is on the bipolar R—S Scale, is that masculinity and repression are significantly and positively linked at a low level. Two explanations are relevant here. First is the reported sex differences in the literature regarding self-disclosure. Jourard (1964) suggests that males typically disclose less than females. Komarovsky (1973) points out that the sharing of personal (e.g., body and personality dimensions) or private information is lower for males. In keeping with our earlier described opposition to the assumption that, based on gender, one necessarily endorses one's traditionally gender-appropriate sex role characteristics, we employed the BSRI to allow us to separate out the gender from the sex role factors. Again, as earlier, this is a useful discrimina- tion. The masculine sex role variable does interact positively with repression. Repression may lead to a lack of self-disclosure. Pleck and Sawyer (1974) and other writers suggest that the endorse- ment of masculine sex role characteristics and lack of endorsement 198 Table 26 Median-based Sex Role, Reclassified Based on Exactly (or within .05) Meeting Criteria, within Sex and Condition for High Shame and High Guilt Groups Success High Shame High Shame High Guilt High Guilt Males Females Males Females Androgynous 1 2 2 l Undifferentiated 4 2 4 2 Masculine 3 3 l 2 Feminine 0 1 l 3 Total 8 8 8 8 Failure Androgynous 2 3 l 3 Undifferentiated 1 3 3 2 Masculine 4 l 2 3 Feminine 1 l 2 0 oo oo oo on Total 199 of feminine characteristics often reflects a personality that demonstrates a lack of self or other awareness and expression. Research findings support the relationship of high endorsement of both masculine and feminine characteristics with other indices of psychological health (e.g., Kelly, J. & Worell, J., 1977; Spence, J., Helmreich, R., & Stapp, J., 1975; Donelson, E., 1977). This writer does not here necessarily assume defensive repression as it is traditionally used to describe the making uncon— scious of previously anxiety related material, but this possibility is not negated. Minimally, it is suggested the high masculine role endorsement may minimize the expression of personal or private areas of one's life. Given the orientation around achievement in the masculine role, failure would maximize threat, and this may be responsible for the reduced accuracy of recall for the subjects in the high shame male and high guilt female groups. A second related issues arises regarding the reduced recall of subjects. The distance provided by the objective measure has been discussed, but the reduced recall may well have been prompted by the tendency for these subjects, who are largely classed mascu- line, to be less willing to disclose personal information to a male experimenter. Both Jourard (1964) and Komarovsky (1973) suggest that females tend to receive, as well as give, the most intimate self disclosures. If instead of gender, sex role is considered, it is the instrumentally oriented indivdiual (i.e., masculine) who shares with an expressively oriented individual (i.e., feminine). It is possible that within both the high shame male and the high 200 guilt female groups, the masculine classed subjects avoid reporting trait feedback, based on the sex or general style of the experi- menter. A third factor to be considered is the context of our analysis, that is, high shame and high guilt. It has been discussed that high shame subjects would be expected to have had a history of receptivity to acute and painful shame experiences. If this history is coupled with sensitization and with masculine classifi- cation, the possibility seems to increase that recall will be reduced. The writer has discussed the Lewis suggestion that females have a hightened receptivity to actual shame experiences, even though this may not be their primary orientation. This, coupled with the presence of high numbers of masculine subject classifica- tions, might explain the low recall for high guilt females. The absence of high numbers of masculine classed subjects in the high shame female group and the presence of androgynously classed sub— jects seems to be associated with heightened recall accuracy. Incidental Learnigg. It was hypothesized that high shame group subjects would perform better on an incidental learning task than high guilt group subjects (Hypothesis 12). This hypothesis was an additional vehicle for testing the present writer's view that high shame group sub- jects would have a heightened awareness of their environment, in this instance making greater use of incidental cues. Mendelsohn and Griswold (1966) found that incidental cue use was negatively 201 correlated with repression (MMPI R Scale). Conversely, the expec- tation of the present study was that the previously predicted sensitization would be positively associated with high incidental cue use. The paradigm employed paralleled the Mendelsohn and Griswold design. The results of an F test show no significant differences between the high shame and the high guilt groups on incidental cue use. The levels of sensitization and the number of sensitizer-classed subjects in the groups is comparably high. Where the absence of differences in sensitization and the consequent absence of differences in incidental cue use is consistent with the Mendelsohn and Griswold findings, results for the total sample in the present study demonstrate no significant correlations between the R-S Scale score and any of the anagram problem solving indices. For the overall sample and the specific groups of interest, corre- lations between the baseline anagram solving ability and the inci- dental cue use scores are highly significant and positive in direction. This findings counters the Mendelsohn and Griswold results of a nonsignificant association between baseline and inci- dental scores. Overall, in addition to not finding differential incidental cue use for the high shame and the high guilt groups, there is no support for a significant positive association between the level of sensitization and heightened levels of incidental cue use. No readily apparent explanation is available for the present study's lack of consistency with the Mendelsohn and Griswold 202 findings. T-tests for the differential effects on feedback reading and recall, due to condition, were not significant. One methodological issue that may have given rise to both the lack of group differences and the high interrelationship between baseline and incidental cue use level is the fact that the pre- anagram task instruction included pointing out that trait feedback would be available upon completion of the anagram task. This may have triggered an expectancy set and influenced the behavior on the anagram task. One final consideration is that high shame and high guilt, similar in their levels of sensitization, may have also led to the development of comparably high levels of baseline anagram solving skills and a comparable level of incidental cue use. Sex Role and Gender No specific predictions were made for sex differences. Few statistically significant differences between sexes arise in the present study. Lewis (1971) theorizes that females would have an increased likelihood of experiencing shame, and males, guilt. This is not supported in the present study. In fact, more females than males are included in the high guilt criterion group of the pilot study. The opposite is the case for the high shame group. This pattern of availability of males for inclusion in the high shame group and females for the high guilt group occurs again in the main study. 203 The present writer tested the masculine and feminine dimen- sions of sex role as an alternative to the Lewis expectation of gender differences with regard to shame and guilt. Although femininity and shame are marginally related at a low level, a far stronger relationship exists between shame and masculinity (in both pilot and main studies). Additionally, while support is found for a nonsignificant positive association between guilt and masculinity, a stronger relationship between guilt and femininity is found (Hypotheses 3-6). The combination of findings seriously calls to question the Lewis position regarding a predisposition to shame or guilt based on gender, with results that reverse her expectations. The correla- tional results and frequencies suggest a relationship between masculinity and shame and femininity and guilt, rather than the reversed, hypothesized direction of association. In retrospect, the heightened shame for males and masculin- ity is logical given the demanding requirements of the masculine role, one of which is the achievement of appropriate characteristics of the role. Any shortcoming can potentially prompt shame for the failure. Inherent in the masculine role is the nonexpression of failure, weakness, or sign of emotion. This puts further demands on the individual striving to achieve the masculine role and, in instances of failure, may prompt shame, withdrawal, and possibly denial, in order to avoid having the failure known to others or to the self. In fact, the presence of high numbers of subjects with masculine classifications in the high shame male group is associated 204 with reduced accuracy of personal trait recall. A similar finding arises in the high guilt female group as well, but here again, there are high numbers of masculine classed subjects and their behavior suggests a heightened responsiveness to failure that mani- fests itself in shamelike behaviors. Hoffman (1975) has also found that males tend to fear detection or exposure of their wrongdoings and failures. Again, the present writer suggests that this effect may be due to sex role factors rather than gender itself. Lewis expected females to be more field dependent and males to be more field independent. There are mixed findings regarding the effects of sex role on the extent of field independence. There is a significant relationship between inaccuracy of figure-ground discriminations and femininity, and a non-significant negative relationship between accuracy and femininity (Hypothesis 7). These findings suggest the logical positive relationship of perceptual fusion with receptive sex role traits. Masculinity is not signifi- cantly correlated with accuracyin figure-ground discriminations but there is a positive direction of association (Hypothesis 8). This finding logically associates instrumental masculine characteristics, one of which is autonomy, with field independent orientation. These findings parallel the Lewis expectations of gender differences, but as expected, are more reflective of sex role characteristics (regardless of sex) rather than of gender. A significant sex difference was found on the ReS Scale, indicating lower sensitization scores for females. The finding suggests that males have a greater need to develop a heightened 205 level of awareness of potential environmental threat. This is con- sistent with the traditional role training of males in regard to confronting anxiety-provoking issues. Regarding the association of the bipolar R—S variable and the sex role dimensions, a significant correlational finding of a low order for the total sample (sexes combined) is suggestive of a positive association between repression and masculinity (r = .14, p < .05). Where males tend to have a heightened level of sensiti- zation, the extent to which they endorse masculine sex role charac- teristics and exclude feminine endorsements may counter this development and be associated with repression. This finding is supported by the reduced recall of trait feedback by those groups with high numbers of masculine classed subjects (both sexes). An additional noteworthy association is the positive corre- lation between the sensitization and the derived androgyny scores (r = .14, p < .05). The present writer suggests that the high endorsement of both traditionally masculine and feminine character- istics is associated with the heightened awareness of, and ability to recall anxiety-provoking material. This finding is supported in the present study by the behavior of the high shame females, the majority of whom are androgynously classified. The data on the recall of feedback, as an indirect measure of repression-sensitization, is consistent with a statistically nonsignificant relationship between the level of sensitization and sex role endorsement factors. While sensitization level may be high, the presence of masculine classification tends to be related 206 to behavioral manifestations of respression or minimally, the unwill- ingness to report personal information. Conversely, androgyny tends to be associated with sensitization behaviorally manifested in heightened recall. No significant differences on gender or sex role variables are isolated for attention to, or recall of trait feedback or inci- dental learning measures for the high shame or the high guilt groups. In general, the variety of measures of shame, guilt, defen- sive sensitization, selective attention, and level of differentiation show almost no support for Lewis's expectations, while in part support is found for some of the present writer's views. SUMMARY AND SUGGESTIONS FOR FURTHER RESEARCH The present study led to a number of interesting findings regarding shame and guilt, often countering this writer's and others' assumptions. These reversals led to a refinement of the concepts of shame and guilt. Generally, support was found for a developmental viewpoint, with guilt being viewed as a refinement of the more pervasive shame affect. This conceptualization of shame and guilt is presented fully in Appendix G. Results of testing often countered hypotheses for shame and guilt. An unexpectedly high level of differentiation, along with the hypothesized high level of sensitization, was found for high shame subjects, as well as high guilt subjects. These simi- larities between the two groups also seems to explain the lack of significant differences on the sensitization-related behavioral measures. The finding of increased differentiation and sensitiza- tion provides support for this writer's view of early experienced shame as prompting the early development of both separateness and vigilence. This seemingly occurs as a result of the breaking of the child-parent emotional bond through shame, and prompting mis- trust of the significant others in the environment. The results for the present study's young adult sample suggests that these patterns are maintained as characteristic styles in adulthood. 207 208 While both the high shame and high guilt groups demonstrated heightened vigilence, trends in behavioral responses, though not statistically significant, suggest that after a failure, sensitizers who make up those groups, tend to avoid, rather than confront, per- sonal trait feedback. This writer has suggested that the reduced feedback attendance may be due to the activation of the sensitizer's vigilence and mistrust of the environment. This view is particularly viable for the high shame group since the experience of failure- based shame often prompts the searching of the self for the reasons for failure. High guilt females showed a similar pattern of avoid— ance, though they did focus on searching the environment for the reasons for the failure. Results regarding trait recall after failure suggest than an interaction of sensitization with sex role endorsement may be active. For those groups with high numbers of sensitizers, who were also classified as androgynous, recall is high while for similar groups where a large number of subjects are classified as masculine, recall is low. These findings, along with others, suggest that masculinity is positively associated with repression, on the R98 Scale, and non-report on the trait recall measure. The present results are supportive of other theoretical and empirical work that suggest that the masculine sex role may inhibit the expression of personal information. Support is also found for a strong association between differentially high endorsement of masculinity and the experience of shame. This, too, is logical, given the demanding masculine role 209 requirements, the self-expectations for achievment and the belief that the self is shameworthy in failing to live up to the masculine ideals. While no attempt is made here to causally link all the find- ings, there is an indication that adult patterns that reflect heightened receptivity to shame and guilt, and are associated with defensive sensitization and heightened separateness, may have early developmental antecedents. Several modifications of the research paradigm are suggested by the present study results. Based on an analysis of the pilot data, a decision was made to employ only the shame and guilt scales in combination for subject selection in the main study. This was the most efficient method of those considered, and was therefore employed. At that time, the sensitization variable, measured by the RrS Scale, was viewed as peripheral to the major concern of the study, and was included as a dependent measure. The writer's present view is that, given the interesting trends in results regarding the level of sensitization, some combination of the shame and guilt variables, and the R—S Scale variable, used for subject selection, may lead to additional useful information. Within such a design, subject selection might be based on the nesting of median-based repressors and sensitizers within the shame-guilt levels, sex, and conditions. While this would add complexity to the design and increase the difficulty of finding appropriate subjects, it might lead to the emergence of the effects of both internal and external vigilence on both inventory and behavioral measures. 210 The use of multiple tests of other variables would also allow for the assessment of the comparability or disparity of results across tests and groups. This is particularly appropriate for the differentiation variable. It has been earlier suggested that the measure of differentiation employed in the present study was primar- ily visual, while that employed by Lewis (1971) required both visual and kinesthetic involvement. Similarly, multiple measure of shame and guilt could potentially tap varying levels of shame and guilt. While these suggested alternatives to the design used in the present study may lead to the uncovering of statistically significant findings regarding the variables, it may be that a more effective route of study might include the intense assessment of individual cases. This more clinically oriented method might allow for the elucidation of the subtle interactions among variables hinted at in the present study. APPENDICES 211 APPENDIX A BASIC TESTING INSTRUMENTS 212 217 222 229 236 243 243 244 246 246 254 255 255 266 Appendix A Test Instruments Attitude Anxiety Survey: Directions and Revised Directions Shame and Guilt Items by Number Original AAS Items and Listed Changes Revised AAS Male Form Revised AAS Female Form Revised AAS Self Form Bem Sex Role Inventory: Items Scoring Procedure and Norms Closure Flexibility Test: Items Scoring Procedure Repression-Sensitization Scale: Items Scoring Procedure 213 214 Attitude Anxiety Survey Listed below are the original Perlman (1958) directions and items and subsequent changes made in the revised A.A.S. forms. Directions (Original) There are many situations that happen every day that make people feel anxious (nervous, tense, or uncomfortable). Here is a list of such situations that have happened to some people. You are asked to look at each of these situations and rate it according to how "disturbed" you think most people would feel were it to happen to them. Rate them along the scale which ranges from "not at all anxious" through "extremely anxious (tense)," as anxious as you can conceive a person being. The marks along the scale are only for your guidance. Do not hesitate to put your mark (X) anywhere along the scale as long as that mark reflects the strength of feeling you believe would be around by that situation. (Revised-Sex Appropriate) There are many situations that happen every day that make people feel anxious (nervous, tense, or uncomfortable). Here is a list of such situations that have happened to some people. You are asked to look at each of these situations and rate it according to how "disturbed" you think most people would feel were it to happen to them. Keep in mind that it does not matter whether or not you would react this way, but give your assessment of how most people would feel. Rate them along the scale which ranges from "not at all anxious" through "extremely anxious (tense)," as anxious as you can conceive a person being. Blacken in the block on the answer sheet. There are £g§_spaces on the answer sheet, ranging from l_(not at all anxious) to lg (extremely anxious (tense)). Mark only one box for each situation. Assess all situations. Erase all changes fully. (Revised-Self) There are many situations that happen every day that make people feel anxious (nervous, tense, uncomfortable). Here is a list of such situations which have happened to some people. You are asked to look at it and rate these situations according to how "disturbed" you think you would feel were i£_£g_happen £g_ygg, Keep in mind that it does not matter whether you think others would feel the same way, but report how ygg_would feel. Rate them along the scale which ranges from "not at all anxious" through "extremely anxious (tense)," as anxious as you can conceive yourself being. 215 Blacken in the block on the answer sheet. There are _t__e£ spaces on the answer sheet, ranging from _1_ (not at all anxious) to E (extremely anxious (tense)). Mark only one box for each situation. Assess all situations. Erase all changes fully. 216 Item # listing by shame or guilt classification Shame Items: 5, 6, 10, 12, 13, 18, 19, 20, 22, 25, 27, 28, 29, 31, 33, 34, 35, 38, 39, 41, 45, 46, 47, 49, 50, 52. Guilt Items: 3, 4, 7, 8, 9, 11, 14, 15, l6, 17, 21, 23, 24, 26, 30, 32, 36, 37, 40, 42, 43, 44, 48, 51, 53, 54. Filler Items: 1, 2. 217 .msoanoua ass we: now: oaou ou manna: so» ouuouoo I oz osmm aw on mougaoou hanopoam Hmoz .OH no» ouuocoon I oz osmm .cowuoousmxo no no mucosa neon .m .uouoom o.oco«uu man nor nausea I oz oeom ouoo>ou zaaouoovwoom comm .w :oooooao: :voadmam: .oooan omoo no: oops cu vowcono own: on vowomeo :uwz: :oodaoaa: one owe: we: owe mm: on umnu so» song I oz Ou vowoono :uwn: moon poem .ucosawum so nouu< .5 nos use I oz «sum assess an massage as: .s «Muscuo vwmuuu no oawuoum “Aoamav anemones no oawuoun :owwauu .ucouoae« no mauuoum "one 30% uao mafia no» so oauuouu: ou camcoso 06mm ma on uoo macaw Hoaow>wvea o< .n .xcoo zoned sauna: .muoouma mm; scum Avnaouo oso on tons: mzonoa maooum zoom; .vcaoum mu so» "wowmmuav mqoq I oz meow etc on tons .honowuoso so sH .c .ooouoa monsoon nouoouumwa so» maumo I oz osmm no; on sumac mesoooo o>mn .m .mmoao you ouma nuances sou , momma an on Haw: a; wouccau so» owuoao I oz oaom he so>o nonu muo>oomuv oumz .N .amouo o uoo>oua ou oaomo: on Haas on ones moon on use mow ecu no page cu no» osuoz I oz osom moamon use m.oom.w:«>auv mafia: .H Anomu>ozv «Ham momma>omv damask Avomw>omv mam: AcoeHuomu Hocwwwuo owcazo osmz I oz Amsuow «unannouaam xom a uaom co mownonu omouzav .m.<.< moma>oz 2113 :sooo mewaaoo mcamonz Ou vowcono :Eooo woumoa: no» no» 30> so» so» :usoeawoo: oo vowoozo zoswwu: so» so» so» :z~o>uuoouuo: 0o oowcoao :AHaouoouow: so» so» =esao wcozroa messes: 0o vowcozu :sooo weaned: umom I oz Sodom I oz mafia I 02 zoomm I oz oeom :ucoeawuo: co vomsono :uswwm: occaOfi I 02 zozoo I oz zone I oz :zao>ooooooo: co vomsmno znaowoou0m: :ucmsosmaosamnn: oo vowconu :nosomonaoan: zccoa I oz :smou wcufisoo mnemo_: oo vmwnoso :soou madman: meow meow meow oEom :unossmzo: Ou vowcmso :uzwau: oeom 08mm =z~o>oooucoo= on vomcoco :zfiasuoouOM: meow .onoeoco30u o :« soou wafimoa oeu mo summons ozu mu uuoooz .oo co wsazoo ow as one: o« wowawou mu on oonu uuo>oomfio com .vommouv zauoaouosw no on use» muo>oo Imav sow .cowuocwumoo was ad wcu>azum noum< .ooouoa moaoono moxwuom one nausea ma: momoa 8mm .nmeoz vac .voaaauuo o uo>o mxoocx zauoovwooo common < .mamsoauoo umucommoco was moon; one ucosawom no coca vouuo~ ma non . .ozom no: a: one: zo comzoa nozoocm uo mwowaoou ecu nous: usaou< .mmoo was so :oaamouasfi ooow a same on voawmu mm: on uoeo moon San .Oo ooxuu o>os pass: on on Hump mmocamsn m c« zaaam Ioouow mm oom uoc new on one» monaaoou soemoooamao < .omoaoo Iuooca Hmoqoms ouuxo cw amazon“ on acumen noumoq .mH .mH .NH .nH .QH .nH .NH .AH Avomfi>ozv «Hum Aoomu>ozv ouosoh Aoomw>o¢v was: “amazoozo Hmcawwoo 2153 sham sacs no owcoeo seem so» no» 30> no» no» so» now no» wouuwso :Haonuoom: sow :zosoao: cu vownueo :uooooouom: so» so» so» Show same no owsmzo seam ocean I oz Houou I oz :nosoz: ou vowcono :ome: ouooumm oz coca I 02 eczo I oz moaazzz I oz coac< I oz oooooeo :Haoaoooo: Quiz I 02 :zmznam: 0o powooau :uoouoouum: sun: I oz unuouoo I oz uwxoon I 02 :momoH one mooxoao zuouooa no zocoa oooo m.zooeao «no No omma ozo mocoam xoazo: Ou vowsono meow meow meow meow meow oaom osom VQUUHEO :HHQfiUOOH: :zmznam: Ou oomnono :uuoooouum: osom oaom .zoaos coca m.>~aaow onu momoa vow osmw vuoo o coca muow xoano .aoamumas emaaoou osom some we; on oozu moon Homo .oamw o cH .com ma: uo>o was» hausoowooo one o .owmumw ma: «0 use poo we: «saxomo oawcz .sofi no; so mmouwoum noon moans hunch .memm unaquQEa no momoa some .8“; mo sou wowxos mvnowuu ma; muoonuo>o H< .onoao sq; o>aoa ou Bozo mHHoo one unsound mu; sows zuwom mosoooo Hana .moxoomae ma: . now powwoauwuo mu voz .voaaaauo osooon 0o common nonuono momaso nedoz .Hamnuoou mezzoaa mazcz .owmuo amanam a so oafin3 madness» m.comuoa noeooco mosoaOu zHuoovaouo zoom .oooaa owansa a nu mow uuo mood moaomno .haoaou o mouasou maunouuo>vmca xomw .Nm zouuo>ozo Loom AUOQH>~MV OHQENK zoomo>ozo was: Acmeauomv Hmcqwaoo 221) =umuu: nouuaso no» no» Aonmeo zooonooua u.vnouumau«w moo» oomsou o>o£ so» Aonoaouo uoecwoua one so» ”uno nova so» so» so» 90> =...=oa uHHou usowuu <= so» so» no» show came so owsozo oaoo 90> :...Aouw3v cannon: woo»: ou oowcozu :smoo. saunas. ooaH< I oz ozone: I oz :oconwoua ow on» one macaw ammo sawewa wcaoz <= nouns I oz coso I oz ouaauo oz «cash I oz snow cowowsou ou oowcozu neocouowou ocaaaomoe coon: I oz osom ence same no owsoco mama ocuoam I oz :...ocoomae m.ooon: cu newsman :UQHM: VOUUfiEO oEom 08mm meow 06mm meow 05mm 06mm osom oaom :zzoaooa: ou oowcoso :uconauuo: osdm .zaocoofiooo one nozuoco mass; was: .ozooo omoo a so .mHoco mo» ago no xooa use 0» ooaamu on comaoa waaxsuuv souu Haw mu oaweo ma: oao monau Hoop .ocoauu Iauam was mononuouaefi mm; on uao woman cos mono» < .vcofiuu vHo so an vouosww ma zoom .zocoe wcwaooum mo ammosou m was soon .zcan a so wcoxooz «one: .usmfi: wcwcoao co head a nu cocoa was muowuou zoos: .Hooo uoouw o momooo o: H< muddy voodoo < .mofiaoco m.zuu::oo man no mono: ozu oucw us» On muouoom ouoom wowsoaao new oaowmooamou zanofioaunouswna soon mm; m; uonu ouo>oomav Hoaow>amca n< .vnowuu a one: no: an away mouwsmooou nae: .muooa man mo ucouw ow vonuouuauo mu common < .oaaooa usonuuuo whoa mo oocomoua ecu :« «Houses ocean neon .mousaamu was now: 5a: mucoumooo mafia m.aom .Nv .He .mn Nooao>uzo ozom noomu>ozv oHoEom Avomn>omv can: AcmEHuomv Hmcwwauo 22E1 so» =oooAH0£UQWV noeoos= cu ownoeo no» so» no» =ouoov o.anouw unom you oumv o Avomauou ouo ro oozed you one nor: on owsoso no» :uosuoa uo oozoao use» ocoaaoa: now now no» :zaaoso one zuuwvz ou owoozo nor Eom I oz :oooHOSUGN m.zuuom: cu owsmso ooaooo I oz mood; I oz :oonuv m.o:ouw no; how sumo o How vesno uoc ow coupon: ou ownozo mayo: I oz :uosoow no: wnuaawx: oo owooso «ooh I oz ocooouzo I oz :zamosm vow zouwv: :.cosos wosoz <: o» ownoeo meow meow meow meow meow meow 05mm 08mm 05mm :zaaoao one zuuav: ou ownmso .otoauw a «0 nomoo oeo you mamas Inoamou mauuouuvca was on one» Hoom oo ocfiwon vaooo: .ouos zoo no: mom on use: u.nvwv o: no: vHOo on nouns ow onoenmuo cm noowo map can Han zfimsoHoom moeoooo nocoos m.HH«m .zuoqoom Mono; osu axes oo: paw ooco macaw was ad oco haso oao ma on oao woman cox .wszxosm odes: madame Haou a; non: mooumom cows: mafia was he cocoon szSOHuom mo: vawso o.uon:wao= a“; uao madam snoz .oomo o oomawou ma uuoz .noooowuow mm: on memo omens powwow a moose one: . .uonoou was wcwaawx ozone moasmouv zooms mcoxmso somuoa < .vsouu o no on ozoam ow eoa .zouma nomoo a co cowoauw>nu on mo>wooou one uqsm mounuoo a nu ouo3x3o axooa on open oHoou moauoso .monooao venomous: .zouuo wowuooa mu on sons 05“» a an woodwou mu; momma one munch < zouuo>ozo zoom noooa>ozv oHoSom zoomo>uzo was: Anosauomo Hooowauo 222 ATTITUDE ANXIETY SURVEY Identification humber Sex: Male Female or Student Number DIRECTIONS There are many situations that happen every day that make people feel anxious ( nervous, tense, or uncomfortable). Here is a list of such situations that have happened to some people. You are asked to look at each of these situations and rate it according to how "disturbed" you think mostypeople would feel were it to happen to them. Keep in mind that it does not matter whether or not you would react this way, but give your assessment of how most people would feel. Rate them along the scale which ranges from "not at all anxious" through "extremely anxious' (tense)," as anxious as you can conceive a person being. Blacken in the block on the answer sheet. There are Egg_spaces on the answer sheet, ranging from l.(not at all anxious) to lQ_(extremely anxious (tense)). dark only one box for each situation. Assess all situations. Erase all changes fully. 1. While driving, Bob's begins to skid on the ice and he sees that he will be unable to prevent a crash. / / / / / / / / / / / not at all extremely anxious anxious (tense) 2. Hate discovers that even by running he will be at least ten minutes late for class. / / / / / / / / / / / not at all extremely anxious anxious (tense) 3. Dave becomes aware he has mistreated another person. / / / / / / / / / / / not at all extremely anxious anxious (tense) h. In an emergency, when no one is around, Larry steals money from his parents’ hidden penny bank. / / / / / / / / / / / not at all extremely anxious anxious (tense) 223 S. An individual finds out he is sterile or impotent: __L. l / / / / / / / I / not at all extremely , anxious anxious (tense) 6. Hal belches in public. I / / / / / / / / / / not at all extremely , anxious anxious (tense) 7. After an argument, Fred sees that he has slapped his wife and made her nose bleed: / I / / l / / / / / / not at all extremely anxious anxious (tense) 8. Fred accidentally reveals a friend's secret: / / / / / / I / / / / not at all extremely . anxious anxious (tense) 9: John cheats on an examination: / / / / / / / / / / / not _at all? extremely anxious anxious (tense) 10: Neal suddenly realizes he is unable to cope with his own problems: / / / / / / / / / / / not at all extremely anxious anxious (tense) ll: Lester begins to engage in extra-marital intercourse: / / / I / l / / / / / not at all extremely anxious anxious (tense) 12. A businessman realizes that he did not act as effectively in a business deal as he would have like to.“ / / / / / / / / / / / not at all extremely anxious anxious (tense) 13: Jim sees that he has failed to make a good impression on his boss; -! -!— I / A! .! :! ! 4!. J not at all extremely anxious anxious (tense) 224 14. Arthur hurts the feelings of another person by what he has said. / I / / / / / ./§ /. I. / not at all extremely anxious anxious (tense) 15. Bob is forced into an argument and hurts his antagonist seriously. / / / / / / / / / / / not at all extremely anxious anxious (tense) 16. A person accidentally knocks over a crippled, old woman. / / / / / / / / / / / not at all extremely anxious anxious (tense) 17. Sam loses his temper and strikes another person. a! 44 1!: —! :J J :J J !— J not at all extremely anxious anxious (tense) 18. After arriving at his destination, Tom discovers that he is improperly dressed. __j l 1 1L I l_. l j I 4!. / not at all extremely anxious anxious (tense) 19. Ben discovers that he is failing in what he is trying to do. _l / l / J 1 J l I l / not at all extremely anxious anxious (tense) 20. Robert is the manager of the losing bowling team in a tournament. Al / l J j l l l l l / not at all extremely anxious anxious (tense) 21. Jack inadvertently commits a felony. _/ l / J I I 1 l l j / not at all extremely anxious anxious (tense) 22. Charles lets off gas in a public place. 1 j / 1 l 1 I 1 I J / not at all extremely anxious anxious (tense) 225 23. Stan accidentally touches another person‘s genitals while on a subway train. / / I / / / / I / / / not at all extremely anxious anxious (tense) 24. While playing, Ralph causes another person to become crippled. I I I I I I I I I I I not at all extremely anxious anxious (tense) 25. Ned is criticized for his mistakes. w! I I —J J I J J _J J J not at all extremely anxious anxious (tense) 26. Phil becomes angry with his parents and tells them to leave him alone. —J> I o! —J 4L7 I I a! I II_ —/ not at all extremely anxious anxious (tense) 27. A1 overhears his friends making fun of him. _I_ I; All 1 In #1 I I / _/ I not at all extremely anxious anxious (tense) 28. Jack loses an important game. _I I I I I I I I I, I I not at all extremely anxious anxious (tense) 29. Jerry makes poor progress in his Job. I] I I I IL I I; II I. I / not at all extremely anxious anxious (tense) 30. While backing his car out of his garage, a man accidentally runs over his son. _I. II / I. I I pl ./ / l / not at all extremely anxious anxious (tense) 31. In a game, Carl sees that he has made some foolish mistakes. / / / / I / l / / II I not at all extremely anxious anxious (tense) 226 32. Chuck spends the last of the family's food money on lottery tickets, and loses. J. J 1 ,1 .L 1 fiI 1 1 1 1 not at all extremely anxious anxious (tense) 33. Sam's wife confronts him with his failures. _I I II I II, I I I_ II I I not at all extremely anxious anxious (tense) 34. John finds himself in the presence of more wealthy people. _JL 1 I II I I I] I II I II not at all extremely anxious anxious (TenseO 35. A person is criticized in front of his peers. I _I I, I I I _I I I I, .I not at all extremely anxious anxious (tense) 36. Mitt recognizes that he has hurt a friend. I I I / I I I I I I I not at all extremely anxious anxious (tense) 37. An individual discovers that he has been unintentionally responsible for allowing state secrets to get into the hands of his country's enemies. I I I I I / I I I I I not at all extremely anxious anxious (tense) 38. A friend tells Al he boasts a great deal. I I I I I / I I I I I not at all extremely anxious anxious (tense) 39. Buddy forgets his lines in a play on opening night. _[ I J l L I I 7] {—4 ll not at all r ' ' ’ '7 extremely anxious anxious (tense) 40. While working in a bank, Bill has a fantasy of stealing money. / I I I I I I I I I I not at all extremely anxious anxious (tense) 227 hl. Stan is ignored by an old friend. I I I I p] I I I: I, I I not at all ' ' extremely anxious anxious (tense) h2. A young man finds out he has impregnated his girlfriend. I I I I I I I I I I I not at all extremely anxious anxious (tense) h3. Joel finds out his child is ill from drinking poison he failed to put back on the top shelf. I I I I I I I I I I I not at all I - extremely anxious anxious(tense) hh. In a fight, Mike kills another man accidentally. I I I I I I I / I I I not at all extremely anxious anxious (tense) h5. A young man meets His friends at a time when he is wearing dirty and smelly clothes. I I I I I I I I I I I BBi‘EE all extremely anxious anxious (tense) h6. Charles feels that he looks awkward in a bathing suit and receives an invitation to a beach party. I I I I I I I I I I I not at all extremely anxious anxious (tense) h7. Tom is shown up as a fraud. I I I I I I I I I I I not at all extremely anxious anxious (tense) h8. A person awakens after dreaming about killing his father. I I I I I I I I I I I not at all extremely anxious anxious (tense) h9. Herb meets a friend whose name he has forgotten. I I I / I I / I I / / not at all extremely anxious anxious (tense) 228 50. Mort is refused a date. _I I I I _/ I I I II I _I not at all extremely anxious anxious (tense) 51. Norm finds out that his neighbor's child was seriously burned by the fire which started when Norm fell asleep while smoking. I I I I I I I I I I I not at all extremely anxious anxious (tense) 52. Ken finds that he is the only one in his group that did not make the honor society. I I I I I I I I I I I not at all extremely anxious anxious (tense) 53. Bill's mother becomes seriously ill the day after an argument in which he told her he didnfit want to see her any more. I I I I I I I I / / I not at all extremely anxious anxious (tense) 5h. Harold begins to feel that he was indirectly responsible for the death of a friend. —! e! I I o_! J I— I eI_ I not at all extremely anxious anxious (tense) PLEASE CHECK TO SEE THAT YOU HAVE RATED EVERY STATEMENT 229 ATTITUDE ANXIETY SURVEY Identification number or Student number _. Sex: Male Female DIRECTIONS There are many situations that happen every day that make peOple feel anxious (nervous, tense, or uncomfortable). Here is a list of such situations that have happened to some peOple. You are asked to look at each of these situations and rate it according to how "disturbed" you think most people would feel were it to happen to them. Keep in mind that it does not matter whether or not you would react this way, but give your assessment of how most people would feel. Rate them along the scale that ranges from "not at all anxious" through "extremely anxious (tense)," as anxious as you can conceive a person being. . Blacken in the block on the answer sheet. There are ten spaces on the answer sheet, ranging from.l_(not at all anxious) to IQ,(extremely anxious (tense)). Mark only one box for each situation. Assess all situation. Erase all changes carefully. 1. While driving, Norma's car skids on the ice and she sees that she will be unable to prevent a crash. / I I I I I I I / I I not at all extremely anxious anxious (tense) 2. Gloria discovers that even by running she will be at least ten minutes late for class. I I I I I I I I I I I not at all extremely anxious anxious (tense) 3. Carla becomes aware that she has mistreated another person. I I I I I I / I I / I not at all extremely anxious anxious (tense) h. In an emergency, Lois steals money from her parents' hidden penny bank. j I I I I I I j J / I not at all extremely anxious anxious (tense) 5. An individual finds out that she is sterile or frigid. I I I I I. .I II .I I l 1 not at all extremely anxious anxious (tense) 230 6. Liz belches in public. I I I I I I I I I I I not at all extremely anxious anxious (tense) 7. After an argument, Kate sees that she has slapped her husband and made his nose bleed. I I I I I I I I I I I not at all extremely anxious anxious (tehse) 8. Sheila accidentally reveals her friend's secret. I I I I I I I I I I I not at all ' extremely anxious anxious (tense) 9. Jeanette cheats on an examination. I I I I I I I I I I I not at all extremely anxious anxious(tense) 10. Loretta suddenly realizes she is unable to cope with her own problems. I I I I I I I I I I I not at all extremely anxious anxious (tense) ll. Dinny begins to engage in extra-marital intercourse. _I_ I I I I II I I I .I not at all extremely anxious anxious (tense) 12. A businesswoman realizes that she did not act as effectively in a business deal as she would have liked to. _I I I I I .III I _I l l .I not at all extremely anxious anxious (tense) 13. Judy sees that she has failed to make a good impression on her boss. I I I I I I I I I I I not at all extremely anxious anxious (tense) 1h. Lilly hurts the feelings of another person by what she has said. I ' I I I _JL II .I I II I; LI_I not at all extremely anxious anxious (tense) 231 15. Joanne is forced into an argument and hurts her antagonist seriously. I I I I I I I I I II I not at all extremely anxious anxious (tense) 16. A person accidentally knocks over a crippled, old woman. I I I I I I I I I I I not at all extremely anxious anxious (tense) 17. Sally loses her temper and strikes another person. I I I I I I I I I I I not at all extremely anxious anxious (tense) 18. After arriving at her destination, Tina discovers that she is improperly dressed. I I I I I I I I I I I not at all extremely anxious e anxious (tense) 19. Helen discovers that she is failing in what she is trying to do. I I I / I I I I I I I not at all extremely anxious anxious (tense) 20. Rose is the manager of a losing bowling team in a tdurnament. _II I I_, I l I I I I I, I not at all extremely anxious anxious (tense) 21. Jackie inadvertently commits a felony. _I_ I I I _I I I_ I I I I not at all extremely anxious anxious (tense) 22. Dorothy lets off gas in public. _II I I I I I I I I / .I not at all extremely anxious anxious (tense) 23. Mary accidentally touches another person's genitals while on a subway train. I I I I I I I I I I / not at all extremely anxious anxious (tense) 232 2h. While playing, Rita causes another person to become crippled. I I / I I I I I I I , I not at all ‘ extremely anxious anxious (tense) 2S. Anita is criticized for her mistakes. I I I I I I I I I .I / not at all extremely anxious anxious (tense) 26. Phyllis becomes angry with her parents and tells them to leave her alone. I I I I I I / I I I I not at all extremely anxious anxious (tense) 27. Lynn overhears her friends making fun of her. I I I I I I I I I I I not at all extremely anxious anxious (tense) 28. Joan loses an important game. I I I I I I I I I I I not at all extremely anxious anxious (tense) 29. Barbara makes poor progress in her Job. _II I I I I I II I II I I not at all extremely anxious anxious (tense) 30. While backing her car out of her garage, a woman accidentally runs over her son. _I_ II I I I I I I II II I not at all extremely anxious anxious (tense) 31. In a game, Carol sees that she has made some foolish mistakes. _I II I I _I JL_ I I I I I not at all extremely anxious anxious (tense) 32. Diane spends the last of the family's food money on lottery tickets, and loses. I I I I I I I I / I I not at all extremely anxious anxious (tense) 233 33. Jane's husband confronts her with her failures. I I I I I I I I I I I not at all extremely anxious anxious (tense) 3h. Elaine finds herself in the presence of more wealthy people. I I I I I I I I I I I not at all extremely anxious anxious ( tense) 35. A person is criticized in front of her peers. I I I I I I I I I I I npt at all extremely anxious anxious (tense) 36. Margo recognizes that she has hurt a friend. I I I I I I I I I / I not at all extremely anxious anxious (tense) 37. An individual discovers that she has been unintentionally responsible for allowing state secrets to get into the hands of her country's enemies. I III I II I / I I I. l I not at all extremely anxious anxious (tense) 38. A friend tells Joyce that she boasts a great deal. I ._I I I I I I I I I I not at all extremely anxious anxious (tense) 39. Claire forgets her lines in a play on opening night. I I I I I I I I_ I I I not at all extremely anxious anxious (tense) ho. While working in a bank, Gwen has a fantasy of stealing money. I / I I I I I I II I_ I not at all extremely anxious anxious (tense) hl. Karen is ignored by an old friend. I I I I_ I I I I I I I not at all extremely anxious anxious (tense) 234 h2. A young, single woman finds out she is pregnant. I I I I I. I I I I I I not at all extremely anxious anxious (tense) h3. Martha finds out her child is ill from drinking poison she failed to put back on the top shelf. I I I I I I I I I I I not at all extremely anxious anxious (tense) hh. In a fight, Alice kills another soman accidentally. / / / / x / A / / / / / not at all extremely anxious anxious (tense) us. A young woman meets her friends when she is wearing dirty, smelly clothes. I I I I I I I I I I I not at all extremely anxious anxious (tense) h6. Charlene feels that she looks awkvard in a bathing suit and receives an invitation to a beach party. Lllz ; .14 .1 x4+ not at all extremely anxious anxious (tense) h7. Toni is shown up as a fraud. -I I I I I iI gI, _I_ I II, / not at all extremely anxious anxious (tense) h8. A person awakens after dreaming about killing her father. _I_ I I I I / AI I I I I not at all extremely anxious anxious (tense) h9. Doris meets a friend whose name she has forgotten. I I I I J I L L I / I not at all extremely anxious anxious (tense) 50. Louise is not asked for a date for her groupss dance. I J I AI‘ I / 1 I I L / not at all extremely anxious anxious (tense) 235 51. Linda finds out that her heighborfs child was seriously burned by the fire which started when Linda fell asleep while smoking. I I I I I I I I . I I I not at all extremely anxious anxious (tense) 52. Elenor finds out that she is the only one in her group that did not make the honor society. II I__ I I I I I I I I I not at all extremely anxious anxious (tense) 53. Betty’s father becomes seriously ill the day after an argument in which Betty told her father that she did not want to see him any more. I I I I I _I I III I I I not at all extremely anxious anxious (tense) 5h. Pam begins to feel that she was indirectly responsible for the death of a friend. I I I I I I I I I I I not at all extremely anxious anxious (tense) PLEASE CHECK TO SEE THAT YOU HAVE RATED EACH 0? THE STATEMENTS 236 ATTITUDE mm SURVEY (Self Form) Student Number or Identification Number . - t DIRECTIONS Sex. Male Female There are many situations that happen every day that make people feel anxious (nervous, tense, uncomfortable).Here is a list of such situations which have happened to some people. You are asked to look at it and rate these situations according to how "disturbed" you think ygg_would feel 32:3 1!; 33 213mg; 33 1533. Keep in mind that it does not matter whether you think others would feel the same way, but report how Igg_would feel. Rate them along the scale which ranges from" not at all anxious" through "extremely anxious (tense)," as anxious as you can conceive yourself being. Blacken in the block on the answer sheet. There are ten spaces on the answer sheet, ranging from 1 (not at all anxious) to 10 (extremely anxious (tense)). Mark only one box for each situation. Assess .all situations. Erase all changes fully. 1. While driving, your car begins to skid on the ice and you see you will ' be unable to prevent a crash. I I I I I (I I I I I I not at all extremely anxious anxious (tense) 2. You discover that even by running, you will be at least ten minutes late for class. I I I I I I I I ' I I I not at all extremely anxious anxious (tense) 3. You become aware that you have mistreated another person. _I. I I I I .I I I_ II I not at all extremely anxious anxious (tense) h. In an emergency, when no one is around, you steal money from your parents' hidden penny bank. I _I I I I I I I I l I not at all extremely anxious anxious(tense) 5. You find out that you are: sterile or frigid (female); sterile or impotent (male). I I I I I I I I I I I not at all extremely anxious . . anxious (tense) 237 6. You belch in public. I I I I I I I I I I I not at all extremely anxious [anxious (tense) 7. After an argument, you see that you've slapped your husband's (wife's) face and made their nose bleed. I I I I I I I I I I I not at all extremely anxious anxious (tense) 8. You accidentally reveal your friend's secret. I I I I I I I I I I I not at all extremely anxious anxious (tense) 9. You cheat on an examination. J 1 1 4 J, 1 .1 .1 .I 4 4' not at all extremely anxious anxious (tense) 10. You suddenly realize you are unable to cape with your own problems. _I_ II I I I I I I I l I not at all extremely anxious anxious (tense) 11. You begin to indulge in extra-marital intercourse. II II I I I I I I I I 41 not at all extremely anxious anxious (tense) 12. You realize that you did not act as effectively in a business deal as you would have like to. I I I I I I I I I I I not at all extremely anxious anxious (tense) 13. You see that you have failed to make a good impression on your boss. I' / I l I I I I I II I not at all extremely anxious anxious (tense) 238 1%. You hurt the feelings of another person by what you have said. I I I I I I I I I I' I not at all extremely anxious ‘ anxious (tense) 15. You are forced into an argument and hurt your antagonist seriously. I I I I I I I I I I I not at all extremely anxious anxious (tense) 16. You accidentally knock over a crippled, old woman. I I I I I I I I I I I not at all extremely anxious anxious (tense) 17. You lose your temper and strike another person. ‘I I I I I I I I I I I not at all extremely anxious . anxious (tense) 18.After arriving at your destination, you discover that you are improperly dressed. I I I I I I I I I I I not at all extremely anxious anxious Ttense) 19. You discover that you are failing in what you are trying to do. _I_ I II I _I I g] I I I not at all extremely anxious anxious (tense) 20. You are the manager of a losing bowling team in a tournament. 1 L I / I I _/ 1 1 j / not at all extremely anxious anxious (tense) 21. You inadvertently commit a felony. _II I I I _I I I I I I / not at all extremely anxious anxious (tense) 22. You let off gas in public. I I, I / I I I I _I II / not at all extremely anxious ' anxious (tense) 239 23. You accidentally touch another person's genitals.while on a subway train. I I I I I I I I I I I not at all extremely anxious anxious (tense) 2h. While playing, you cause another person to become crippledu / / I I I I I I I I / not at all - extremely anxious anxious (tense) 25. You are criticized'for your mistakes.' I I I !~ III, 1 I I I I I not at all r7 extremely anxious anxious (tense) 26. You become angry with your parents and tell them to leave you alone. .L .1 .1 .1 .1 .1 .1 .1 .1 .1 not at all extremely anxious ' anxious (tense) 27. You overhear your friends making fun 03 you. I .1 .L .1 L .1 .1 .1 .1 ,1 not at all extremely anxious anxious (tense) 28. You lose an important game. 1 1 .1 .1 .1 .1 .1 .1 .1 .1 not at all extremely anxious anxious (tense) 29. You make poor progress in your Job. II I I II I II I I II I not at all - extremely anxious anxious 30. While backing your car out of your garage, you accidentally run over your scr\ I I I II _I II I I I II II_ not at all extremely anxious . anxious (tense) 31. In a game, you see that you have made some foolish mistakes. I I II II I I / I II II I not at all . extremely anxious anxious (tense) 240 32. You spend the last of your family's food money on lottery tickets, and lose I I I I I I I I L I I not at all extremely anxious anxious (tense) 33. Your husband (wife) confronts you with your failures. I I I I I I II I I I I not at all extremely anxious anxious (tense) 3h. You find yourself in the presence of more wealthy peOple. I I I I I I I I I I I not at all ' extremely anxious anxious (tense) 35. You are criticized in front of your peers. I I I I I I I I I I I not at all extremely anxious . ‘ anxious (tense) 36. You recognize that you have hurt a friend. I I I I I I I I I I I not at all extremely anxious anxious (tense) 37. You discover that you have been unintentionally responsible for allowing state secrets to get into the hands of your country's enemies. I I I I I I I I I I I not at all extremely anxious anxious (tense) 38. A friend tells you that you boast a great deal. I I I I I I I I I I I not at all ' extremely anxious anxious (tense) 39. You forget your lines in a play on opening night. I / I I I I I I / I I not at all extremely anxious anxious (tense) ho. While working in a bank, you have a fantasy of stealing money. I I / I I I I I I I I not at all extremely anxious anxious (tense) 241 hl. You are ignored by an old friend. I I I I I I I I I I I not at all extremely anxious anxious (tense) h2. You find out: you are pregnant (female); you have caused your girlfriend’s pregnancy (male). I I I I I I I I I I I not at all extremely anxious ' anxious (tense) h3. You find out your child is ill from drinking poison you failed to put back on the top shelf. I I I I I I I I I I I not at all extremely anxious anxious (tense) hh. In a fight, you kill another person accidentally. ’-/ /_1 j / / I1 / / not at all extremely anxious anxious (tense) 1&5. You meet your friends at a time when you are wearing dirty and smelly clothmfl I I I I I II III. __I l I! I not at all extremely anxious anxious (tense) h6. You feel that you look awkward in a bathing suit and receive an invitation to a beach party. I I I I I / I I I I I not at all extremely anxious anxious hT. You are shown up as a fraud. I I I I I I I I I I I not at all extremely anxious anxious (tense) h8. You awaken after dreaming about killing your father or mother. I I I I I I I I I I I not at all extremely anxious anxious (tense) 242 h9. You meet a friend whose name you have fergotten. I I / I I I I I I I I not at all extremely anxious _ anxious (tense) 50. You are not asked for (or are refused) a date for your group‘s dance. I I I I I I I I I I I not at all extremely anxious anxious (tense) 51. You find out your neighbor's child was seriously burned by the fire that started when you fell asleep while smoking. I I I I I I I I I I I not at all extremely anxious anxious (tense) 52. You find out that you are the only one in your group that did not make the honor society. 4L— I~ I I I I I II? I I :{ not at all - extreme y anxious anxious (tense) 53. Your mother (father) becomes seriously ill the day after an argument in which you told them that you didn't want to see them any more. l l I —el 1! __I, _I I I I I not at all extremely anxious anxious (tense) 5h. You begin to feel that you were indirectly responsible for the death of a friend. I / / J I x J J I J / not at all extremely anxious anxious (tense) PLEASE CHECK TO BE SURE THAT YOU BLACKENED ONE BOX FOR EACH SITUATION 243 Bem Check List. Please indicate on a seven point scale how well each of the characteristics describes you. write the number appropriate for your answer beside each item number on the answer sheet. I l I, 2 I’ 3 I’ 1s I S I' 6 I, '7 I L Never or almost Always or almost never true of me always true of me 1. Self-reliant 31. Makes decisions easily 2. Yielding . 32. Compassionate 3. Helpful 33. Sincere h. Defends own beliefs 3%. Self-sufficient 5- Cheerful 3S. Eager to soothe hurt feelings 6. Moody 36. Conceited 7. Independent 3?. Dominant 8. Shy 38. Soft spoken 9. Conscientious 39. Likeable 10. Athletic ho. Masculine ll. Affectionate hl. warm l2. Theatrical h2. Solemn 13. Assertive h3. Willing to take a stand 1h. Flatterable hh. Tender - 15. Happy hS. Friendly 16. Strong personality h6. Aggressive 17. Loyal h7. Gullible 18. Unpredictable h8. Inefficient 19. Forceful h9. Acts as a leader 20. Feminine 50. Childlike 21. Reliable 51. Adaptable 22. Analytical S2. Individualistic 23. Sympathetic 53. Does not use harsh language 2h. Jealous 5h. Unsystematic 25. Has leadership ability 55. Competitive 26. Sensitive to the S6. Loves children needs of others 57. Tactful 27. Truthful 58. Ambitious 28. Willing to take risks 59. Gentle 29. Understanding 60. Conventional 30. Secretive 244 Scoring, Psychometric Data for Bem Checklist. From, Sam, Sandra L. The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 1974, 42, 2, 155-162, April. Masculine Items: 49, 46, 58, 22, 13, 10, 55, 4, 37, 19, 25, 7, 52, 31, 40, 1, 34, 16, 43, 28. Feminine Items: 11, 5, 50, 32, 53, 35, 20, 14, 59, 47, 56, 17, 26, 8, 38, 23, 44, 29, 41, 2. Neutral Items: 51, 36, 9, 60, 45, 15, 3, 48, 24, 39, 6, 21, 30, 33, 42, 57, 12, 27, 18, 54. The masculinity and femininity scores indicate the extent to which a person endorses masculine and feminine personality characteristics as self- descriptive. Masculinity equals the mean self-rating for all endorsed masculine items, and Femininity equals the mean self-rating for all endorsed feminine items. Both can range from 1 to 7. These two scores are logically independent. That is, the structure of the test does not constrain them in any way, and they are free to vary independently. -- Simply calculate the mean response for the masculine and feminine items separately, e.g., sum the numbers you chose for your response and divide by the total number items on the feminine or masculine scale. Androgyny. Greater absolute value, more sex typed. Hi+ - F. Hi- - M. Difference Score. Femininity minus Masculinity. -- the simplest t ratio for difference. Empriically, the two indices are virtually identical (r - .98). The t- ratio can be approximated by multiplying the difference score by 2.322 -- derived empirically from normative sample to 917 students at two different colleges. Use of t-ratio rather than the simple difference score has the advantage of allowing a statement of statistical significance (i.e., a person's feminine and masculine scores are or are not statistically significant), and allows comparing different populations in terms of the percentage of significantly sex-typed individuals present within each. Social Desirability: the extent to which a person describes himself in a socially desirable direction on items that are neutral (theoretically and empirically) with respect to sex. Reverse the 10 undesirable items and then calculate the mean response across the 20 items. Both Masculinity and Femininity were found correlated with SD, but Androgyny was not correlated with SD. Sex Differences on the BSRI (all different, p .001 except the us noted) .- Stanford University Foothill Junior College Male Female Male Female Masculinity 4.97 4.57 4.96 4.55 Femininity 4.44 5.01 4.62 5.08 SD 4.91 5.08 4.88 4.89 -ns Androgyny t -l.28 1.10 - .80 1.23 Androgyny Difference -0.53 .43 - .34 .53 245 2 SS classified as: M444 F279 M117 ‘EZZ feminine 6 34 9 40 near feminine 5 20 9 8 androgynous 34 27 44 38 near masculine 19 12 17 A m 7 masculine 36 8 22 8 246 Please fill in: CLOSURE FLEXIBILITY N... Age Sex __ Date (Concealed Figures) '(Form A) Occupation 09de by: L.L. Thm'hli. PLO. Ind TJL'. Jeffrey, Ph.D. - The Psychometric Laboratory . Th0 L'niuflllv 0‘ North Carolina Directions: The row of designs below is a sample item of this test. The parts have been la- beled to make description easier. These labels do not appear in the test items. The left hand design in each row is the figure. You are to decide whether or not the figure is concealed in each of the fourdrawxngs to the right. Put a check mark (4) in the parentheses under adrawing. if it contains the figre. Puta zero (0) in the parentheses under adrawing. if it does not contain the figlre. Look at the row of designs below. ' Figure Drawings 1 Z J 4 (0) (0) M M In the row above a zero (0) has been written in the parentheses under drawing 1. The first drawing is a square but it islarger than the figure. Azero (0) has been .written under drawing ;. Although the second drawing contains a square ofexact- ly the same Size as the figure, it has been turned. Check marks (4) have been written under the third and fourth drawings since they each contain a square of exactly the same size as the figure and have not been turned. It does not matter that the figure contained in drawmgs three and four is on a different level from the figure at the left. Sample: Here is another example for practice. Try it. X00 “ OO () (i (i () You should have placed check marks (J) in the parentheses under the first and third drawings and zeros (0) in the parentheses under the second and fourth drawings. WHEN YOU GET THE SIGNAL TO BEGIN, turn the page and mark more problems of the same Kind. Work as fast and as accurately as you can, but do not guess. Wrong answers Will count against you. You are not expected to finish in the time allowed. You will have exactly ten minutes to do as much as you can. 1 I . 'i‘ .R Copying'it 056 by Thtlfl‘l U Thurstone and T. E Jeffrey mag... TUNE- -'.‘.9 6-2- 3000 M - _ . Published In industrial Raiatnna Center - The l'nivarany of Chicago .., ....... .,.,.. .... m 132) East with 'iu-ei - Chit-go. iiasflOl. 5312.11 247 f/ i l l "Ll'i DO NOT STOP. GO ON TO THE NEXT PAGE. is C? A; m ><>< i<>< >i< if?“ @ ® {/9 XX I, @ - » Do [JED 249 U0 E9 ( l i i / / 431151 L I I I ., ‘ Walk ( i i ) A U a. v T» It 9‘9 ’t/ A v A v ) l) () LLJ 1.1 mm I _ i I I l l) I) DO NOT STOP. GO ON TO THE: NEXT PAGE. 250 6% / / / a? . 4 D $‘M WWW&.. m / ; Z] ‘ W ) l) t) l) (I t) I) () D . 252 & 4@\ ® % $6 O Q a ® 3 6 1g DO NOT STOP. GO ON TO THE NEXT PAGE. b m r 2% I D I: 253- .- 000:- v 09009 fiééfiéfi” A STOP HERE -- WAIT FOR FURTHER INSTRUCTIONS. 254 Closure Flexibility Test This is a paper and pencil test which can be individually or group- administered. Forty-nine items are included. Each item consists of a figure followed by a row of four, more complex drawings. Some of these four, more complex drawings contain the given figure in its original size and orientation. Instructions are to look for the original figure in each of the complex draw— ings and to put a check (/) under each drawing which contains it and a zero (0) under each which does not. The raw score is the number of correct answers minus the number of wrong answers, which provides a correction for guessing. The maximum raw score is 196. A negative score results when the subject answers more items incorrect than correct. 255 REPRESSIDN-SENSITIZATION SCALE Student number or Code number Sex (circle) Male Female D.B.R.S. Answer true or false. Work as quickly as possible. 16. 17. 18. I have a good appetite. I wake up fresh and rested most mornings. I am easily awaken by noise. I like to read newspaper articles on crime. My hands and feet are usually warm enough. My daily life is full of things that keep me interested. I am about as able to work as I ever was. There seems to be a lump in my throat much of the time. I enjoy detective or mystery stories. I work under a great deal of tension. Once in a while I think of things too bad to talk about. I am very seldom troubled by constipation. At times I have fits of laughing and crying that I cannot control. I am troubled by attacks of nausea and vomiting. I feel that it is certainly best to keep my mouth shut when 12m in trouble. At times I feel like swearing. I find it hard to keep my mind on a task or job. I seldom worry about my health. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 256 At times I feel like smashing things. I have had periods of days, weeks, or months when I couldn't take care of things because I cOuldn't "get going”. My sleep is fitful and disturbed. Much of the time my head seems to hurt all over. I do not always tell the truth. My judgement is better than it ever was. Once a week or oftener I feel suddenly hot all over, without apparent cause. I am in just as good physical health as most of my friends. I prefer to pass by school friends, or people I know but have not seen for a long time, unless they speak to me first. I am almost never bothered by pains over the heart or in my chest. I amsa good mixer. Everything is turning out just like the prophets of the Bible said it would. I do not read every editorial in the newspaper every day. I sometimes keep on at a thing until others lose their patients with me. I wish I could be as happy as others seem to be. I think a great many people exaggerate their misfortunes in order to gain the sympathy and help of others. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 51. 52. 53. S4. 55. 257 I get angry sometimes. Most of the time I feel blue. I sometimes tease animals. I am certainly lacking in self confidence. I usually feel that life is worth while. It takes a lot of argument to convince most people of the truth. Once in a while I put off until tomorrow what I ought to do today. I think most people would lie to get ahead. I do many things that I regret afterwards (I regret things more or more often than others seem to). I go to church almost every week. I have very few quarrels with members of my family. I believe in the second coming of Christ. My hardest battles are with myself. I have little or no trouble with my muscles twitching or jumping. I don't seem to care what happens to me. Sometimes when I'm not feeling well I am cross. Much of the time I feel as if I have done something wrong or evil. I am happy most of the time. Some people are so bossy that I feel like doing the opposite of what they request, even though I know they are right. Often I feel as if there were a tight band around my head. My table manners are not quite as good at home as when I'm out in company. 258 56. I seem to be as capable and smart as most others around me. 57. Most people will use somewhat unfair means to gain profit as or an advantage rather than lose it. 58. The sight of blood neither frightens me nor makes me sick. 59. Often I can't understand why I've been so cross and grouchy. 60. I have'never vomited blood or coughed up blood. 61. I do not worry about catching diseases. 62. At times my thoughts have raced ahead faster than I could speak them. 63. If I could get into a movie without paying and be sure I was not seen I would probably do it. 64. I commonly wonder what hidden reason another person may have for doing something nice for me. 65. I believe that my home life is as pleasant as that of most peOple I know. 66. Criticism or scolding hurts me terribly. 67. My conduct is controlled largely by the customs of those about me. 68. I certainly feel useless at times. 69. At times I feel like picking a fist fight with someone. 70. I have often lost out on things because I couldn't make up my mind soon enough. 71. It makes me impatient to have people ask me advice or otherwise interrupt me when I am working on something important. 72. I would rather win than lose in a game. 75. 76. 77. 78. 79. 80. 81. 82. 83. 34.. 85. 86. 87. 88. 89. 90. 91. 92. 259 Most nights I go to sleep without thoughts or ideas bothering me. During the past few years I have been well most of the time. I have never had a fit or convulsion. I am neither gaining nor losing weight. I cry easily. I cannot understand what I read as well as I use to. I have never felt better in my life than I do now. I resent having anyone take me in so cleverly that I have had to admit that it was one on me. I do not tire quickly. I like to study and read about things that I am working at. I like to know some important people because it makes me feel important. What others think of me does not bother me. It makes me uncomfortable to put on a stunt at a party even when others are doing the same sort of things. I frequently have to fight against showing that I'm bashful. I have never had a fainting Spell. I seldom or never have dizzy spells. My memory seems to be all right. I am worried about sex matters. I find it hard to make talk when I meet new people. I am afraid of losing my mind. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. I I 260 am against giving money to beggars. frequently notice that my hands shake when I try to do somethi: can read a long while without tiring my eyes. feel weak all over much of the time. have very few'headaches. Sometimes, when embarrassed, I break out in a sweat which annoys me greatly. I I I I have had no difficulty in Keeping my balance in walking. do not have spells of hayfever or asthma. do not like everyone I know. wish I were not so shy. enjoy many different kinds of play and recreation. like to flirt. In walking I am very careful to step over sidewalk cracks. I I I frequently find myself worrying about something. gossip a little at times. hardly ever notice my heart pounding and I am seldom short of breath. I have at times stood in the way of people who were trying to do something, not because it amounted to much but because of the principle of the thing. I I get mad easily and then get over it soon. brood a great deal. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 125. 126. 127. 128. 261 I have periods of such great restlessness that I cannot sit long in a chair. I dream frequently about things that are best kept to myself. I believe that I am no more nervous than most others. I have few or no pains. Sometimes without any reason or even when things are going wrong I feel excitedly happy, "on top of the world." I can be friendly with people who do things which I consider wrong. ' Sometimes at elections I vote for men about whom I know very litfle I have difficulty in stagéng to do things. I sweat very easily even on cool days. It is safer to trust nobody. Once a week or oftener I become very excited. When in a group of people I have trouble thinking of the right things to talk about. When I leave home I do not worry about whether the door is locked or the windows closed. I do not blame a person for taking advantage of someone who lays himself Open to it. At times I am full of energy. My eyesight is as good as it has been for years. I have often felt that strangers were looking at me critically. 129. 130. 131. 132. 137. 138. 139. 140. 146. 147 I 148. 262 I drink an unusually large amount of water a day. Once in a while I laugh at a dirty joke. I am troubled by attacks of nausea and vomiting. I am always disgusted with the law when a criminal is freed through the arguments of a smart lawyer. I work under a great deal of tension. I am likely not to speak to people until they speak to me. I have periods in which I feel unusually cheerful without any special reason. 'Life is a strain for me much of the time. In school I found it very hard to talk before the class. Even when I am with people I feel lonely much of the time. I think nearly anyone would tell a lie to keep out of trouble. I am more sensitive than most other peOple. I am easily embarrassed. I worry over money and business. I almost never dream. I easily become impatient with people. I feel anxiety about something or someone almost all the time. Sometimes I become so excited that I find it hard to get to sleep I forget right away what people say to me. I usually have to stop and think before I act even in trifling matters . 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 263 Often I cross the street in order not to meet someone I see. I often feel as if things were not real. I have a habit of counting things that are not important such as bulbs on electric signs, and so forth. ~ I have strange and peculiar thoughts. I get anxious and upset when I have to make a short trip away from home. I have been afraid of things or people that I knew could not hurt me. I have no dread of going into a room by myself where other people have already gathered and are talking. I have more trouble concentrating than others seem to have. I have several times given up doing a thing because I thought too little of my ability. Bad words, often terrible words, come into my mind and I cannot get rid of them. Sometimes some unimportant thought will run through my mind and bother me for days. Almost every day something happens to frighten me. I am inclined to take things hard. I am more sensitive than most other people. At periods my mind seems to work more slowly than usual. I very seldom have spells of the blues. 165. 166. 167. 168. 169. 170. 172. 173. 175. 176. 177. 178. 264 I wish I could get over worrying about things I have said that may have 3 injured other people's feelings. People often disapoint me. I feel unable to tell anyone all about myself. My plans have frequently seemed so full of difficulties that I have had to give them up. Often, even though everything is going fine for me, I feel that I don't care about anything. I have sometimes felt that difficulties were piling up so high that I could not overcome them. I have often met people who were supposed-to be experts who were no better than I. It makes me feel like a failure when I hear of the success of someone I know well. I am apt to take disappointments so keenly that I can't put them out of my mind. At times I think I am no good at all. I worry quite a bit over possible misfortunes. I am apt to pass up something I want to dd because others feel that I am not going about it in the right way. I find it hard to set aside a task that I have undertaken, even for a.short time. 179. 181. 182; 183. 265 I have several times had a change of heart about my life work. I must admit that I have at times been worryed beyond reason over something that really did not matter. I like to let people know where I stand on things. I have a daydream life about which I do not tell other people. I have often felt guilty because I have pretended to feel more sorry about something than I really was. I feel tired a good deal of the time. I sometimes feel that I am about to go to pieces. 266 ngression-Sensitization Scale The R-S Scale, developed by Byrne (1961), arose as a modification of a comparable scale (Altrocchi, Parsons and Dickoff, 1960), which in turn was drawn from the Depression, Psychasthenia, Welsh Anxiety, Lie, K, and Hysteria subscales of the MMPI. Multiple scored and inconsistently scored items were eliminated from the R-S Scale. The R-S Scale consists of 182 items in the order in which they appear on the NMPI. There are 156 scorable and 26 buffer items. One hundred and sixteen items are scored true and forty are scored false. High scores indicate sensitization and low scores, repression. APPENDIX B PILOT STUDY 267 279 289 Appendix B Pilot Study Total Group: Results of hypotheses tested by Pearson correlations and t-tests Criterion Group: AAS, Results of Pearson correlations and t-tests Criterion Group: BSRI, Results of Pearson correlations and t-tests Subject Data on Excluded Subjects; AAS 268 269 Data from the Pilot Study was analyzed by several methods. Overall sample intercorrelations and t-tests for sex differences on all questionnaire variables were computed. Reported here are the results of specific hypothesis testing. A second analysis was carried out on the basis of the Relative score of the AAS. Self form. Criterion groups were established by taking the highest (nPlB) and lowest (nelB) 202 of the subjects based on their Relative scores. This allowed for the separate analysis of those cases labeled guilt-prone (highest, 20%) and shame-prone (lowest, 20%). In turn, Pearson correlations between the Relative score for each of these groups and remaining questionr naire variables were computed. T-tests for significant sex differences on the Relative score within the two criterion groups were computed. Finally, t-tests were computed for group differences on the AAS Relative score between criterion groups and between each criterion group (202) and the remaining sub- jects (802). These final t-tests were computed to assess the usefulness of the Relative score as a potential method of subject classification for shame and guilt subjects. A third analysis was carried out on the Pilot data by classifying subjects into masculine, feminine, and androgynous categories, based on the Bem (1974) scoring structure (t-values). Subjects with near-masculine and near-feminine classifications were excluded. Paralleling the AAS criterion analysis, inter- correlations were carried out for each of the three sex role classifications. T-tests for sex differences were computed. Additionally, AAS data for two males and four females excluded from the sample are listed here. 270 Total Group: Results of hypotheses tested by Pearson correlations and t-tests 1. Shame scores (on.AAS) will be negatively associated with closure flexibil- ity (field independence). No significant findings of support were found for this hypothesis. The direction of nonrsignificant support suggested that a minimal trend exists in the hypothesized direction , that is, that the shame score correlates negatively with Correct Closure, Closure Difference, and Closure Scale scores, all measures of field independence (Table 1). Insert Table 1 about here 2. Guilt scores (on.AAS) will be positively associated with closure flexibility (field independence). Support was found for a reversal of this hypothesis. The direction of the correlation suggested a negative association between the guilt score and field independence (see Table 1). 3. Shame scores will be positively associated with femininity and negatively associated with masculinity in both males and females. . Results demonstrated a non-significant positive association between shame and femininity. In all but one instance (Self form, session 1), correlations between shame and masculinity were positive, a reversal of prediction (Table 2). Insert Table 2 about here 4. Guilt scores will be positively associated with masculinity and negatively associated with femininity in males and females. Results demonstrated a nonrsignificant positive association between guilt and masculinity on all but one form (Other form, session 2). In all but one ital-e Self form lst session 2nd session Other form lst session 2nd session slug Self form lst session 2nd session Other form lst session 2nd session 271 Table l Intercorrelations of Shame, Guilt, and Closure Flexibility Scale Closure Correct -005, -005, p<.69 p<.52 p<.58 p<.62 p<.07 p<.48 p<.09 p<.06 Closure Incorrect .06, -001, .12, .14, p<.27 p<.19 p<.54 p<.90 p< 003 p<.37 p<.25 p<.l6 Closure Difference -007. -.02, p<.35 p<.21 p<.46 p<.82 p<.01 p<.3l p<.07 p<.04 Closure Scale -008, -.O3, -.l6, -.20, p<.32 p<.l7 p<.44 p<.7l p<.01 p<.42 p<.ll p<.04 mes. Self form lst session 2nd session Other form lst session 2nd session 9.2ng Self form lst session 2nd session Other form lst session 2nd session 272 Table 2 Intercorrelations of Shame, Guilt and Masculine p<.85 p<.35 p<.77 p<.41 p<.74 p<.21 p<.85 p<.32 BSRI Variables Feminine p<.1O p<.O9 p<.03 p<.55 p<.001 p<.012 p<.OO3 p<.08 Social Desirability .24, p<.004 .27, p<.008 .25, p<.01 .06, p<.54 .34, p<.001 .35, p<.001 .28, p<.oos .21, p<.03 529223222. .24, .07, p<.12 p<.43 p<.l7 p<.87 p<.Ol p<.34 p<.Ol p<.46 273 instance (Self form, session 1), correlations between guilt and femininity were significantly positive (see Table 2). 7. Closure flexibility scores (field independence) will be negatively asso- ciated with feminine sex role endorsement in females and males. Significant support was demonstrated for the negative association of femin- ine sex role with closure flexibility (r--.33, p < .001). Insert Table 3 about here 8. Closure flexibility scores (field independence) will be positively associated with masculine sex role endorsement in males and females. No significant support was found for this hypothesis; the obtained correla- tion was negative (see Table 3). Total Group: T-tests for Sex Differences A significant sex difference was found on the Relative score (Self form, session 2). Males demonstrated higher shame scores than did females. No sig- nificant sex differences were found on the remaining AAS forms. The significant sex difference does prompt a questioning of the Lewis (1971) expectation that females would show greater shame than males. Significant sex differences were found for the total group on the feminine scale and the androgyny scale of the BSRI. These differences reflected higher feminine scores for females than for males; and reflected a feminine direction of deviation from perfect (0) androgyny for females and a masculine direction of deviation from perfect androgyny for males. No significant sex difference was found on the masculine scale (Table 4). Insert Table 4 about here 274 Table 3 Intercorrelations between Femininity, Masculinity and Closure Flexibility Scale Closure Closure Closure Closure Correct Incorrect Difference Scale Masculine -.04, p<.64 .16, p<.ll -.l3, p<.18 -.12, p<.24 Fain-ine -033, p<.001 008’ p<043 -026, p<.01 _025, p<001 275 Table 4 T-tests for Sex Differences on All Variables in Pilot Study (males-42, females-51, df-9l) 355.76 44.98 M_ea£ Si t-value p_ flees Self, lst session 152.88 33.62 .32 .74 150.75 30.22 Other, lst session 166.16 31.02 -.14 .89 166.98 26.83 Self, 2nd session 149.16 36.75 .36 .71 146.52 33.32 Other, 2nd session 170.61 38.55 .68 .49 165.50 33.65 sea: Self, lst session 187.61 26.29 -1.69 .09 197.68 30.36 Other, lst session 183.90 24.36 -1.17 .24 189.70 23.29 Self, 2nd session 183.02 29.52 -1.09 .27 190.25 33.66 Other, 2nd session 186.64 28.50 -.25 .79 188.09 26.49 Total Anxiety Self, lst session 340.50 54.86 -.70 .48 348.43 53.38 Other, lst session 350.07 49.85 -.58 .56 Self, 2nd session Other, 2nd session Relative Score Self, lst session Other, lst session Self, 2nd session Other, 2nd session Bem Masculine Masculine Feminine Social Desirability Androgyny Closure Flexibility Correct 276 Table 4 (Continued) _11e_a_n 329.88 336.78 357.26 352.03 6.00 7.03 2.54 3.23 5.45 7.62 3.11 3.62 100.61 96.68 91.07 102.60 89.83 92.56 —8.02 5.92 54.54 50.41 64.07 58.69 62.14 55.86 4.16 3.87 5.48 4.58 4.64 4 26 5.37 4.66 11.74 12.51 10.66 13.61 8.30 9.41 14.59 16.33 16.05 16.00 t-value -054 -1024 -2. 35 -1.55 -4.47 -1047 -4030 1.24 .51 .51 .02 .62 .12 .000* .000* Incorrect Difference Scale M 277 Table 4 (Continued) M 21.66 23.33 32.88 27.05 38.35 36.50 17.84 14.96 26.99 23.53 10.24 8.73 5:19.122 -049 1.11 .27 .35 278 Criterion Grogp: AAS, Results of Pearson Correlations and t-tests Criterion groups on the AAS were established by taking the highest (guilt- prone) and lowest (shame-prone) 202 of the subjects, based on Relative scores. For each criterion group, intercorrelations between the Relative score and all other questionnaire variables were computed. T-tests for sex differences on the Relative score were carried out within each group. Additionally, t-tests were computed comparing criterion groups, and comparing each criterion group with the remainder of subjects, on the Relative score. For the sake of brevity a summarization of the criterion analysis is presented here to point out the limited value of the Relative score of the AAS. Overall, the analysis provided almost no consistency of findings over the two sessions, or between Self and Other forms. One example of this can be seen in the fact that for the shame criterion group (session 2), the AAS Relative score showed a significantly negative correlation with Shame on three of four AAS forms. Counter to expectation, the AAS Relative score also correlated signifi- cantly negatively with Guilt, though one would expect from the scoring pro— cedure that there would be a logical positive association (Table 5). No other significant correlations were found between the Relative score and other ques- tionnaire variables for any of the criterion groups (sessions 1 and 2). The erratic quality of these and other findings, the lack of consistency of findings across test forms and sessions, and the lack of any significant intercorrela- tions between the Relative score and other inventory scales, point out some of the limitations in the use of the Relative score and prompted the present writer to develop a more stable and sensitive scoring system for the AAS. A median- based scoring system using the shame and guilt scales was employed in the Main Study as a method of subject classification. 279 Insert Table 5 about here T-tests computed for sex differences on the Relative score were not sig- nificant (Table 6). T-tests between the high and low criterion groups within each session were significant, as were t-tests between high or low criterion groups compared to the remainder of the sample (Table 7). Insert Tables 6 & 7 about here Criterion Group: BSRI, Results of Pearson correlations and t-tests Three categories, masculine, feminine, and androgynous, were established using Bem (1974) norms. Subjects were included in the categories if their mean difference scores fell in the ranges outlined by Ben. Three procedures were carried out for the BSRI criterion groups. The intercorrelations of BSRI var- iables and all other questionnaire variables were computed. T-tests for sex differences within each criterion group were determined, and t—tests for sex differences on BSRI variables within each group were computed. For the sake of brevity, results are summarized here. Masculine Group (males=10,gfema1es=4) Masculinity correlated nonrsignificantly and positively with Shame. Mas- culinity was non—significantly, negatively correlated with the Relative score, reflective of a positive relationship between shame and masculinity (Table 8). Insert Table 8 about here Feminine Group (males-Olgfemales=12) Femininity was correlated positively and marginally significantly with Social Desirability (r-.63, p < .01). Femininity was not significantly correlated Table 5 Relative Score Shame Self form lst session 2nd session Other form lst session 2nd session Guilt Self form lst session 2nd session Other form lst session 2nd session Total Anxiety Self form lst session 2nd session Other form lst session 2nd session -040, p<.001 -.46, p<.001 -053, p(0001 -.55, p<.001 .06, p<.54 -014, p<.15 .08, p<.39 —00003, p<099 -.2o, p<.04 -035, p<.001 -026, p<.009 -.32, p<.002 T Tests for Sex Differences on AAS Relative Scores Guilt Criterion Self form, Session 1 Self form, Session 2 Shame Criterion Self form, Session 1 Self form, Session 2 m-S f-13 m-6 f-lZ m-ll f-7 IFIO f-2 281 Table 6 for Criterion Groups Mean 13.20 12.15 12.00 13.50 in; T Value 1.08 -1979 .06 1.18 g; 16 16 16 16 2 tail prob. .297 .093 .953 .254 282 m2 Hoe. Hoe. no. Hoo. #00. H00. H00. Hoo. doc. Hoe. Hoe. Nm. oo.oau oo.w mo.H hm.» Hm.eu mm.mH mm.OH mm.HH qo.Hm mo.m mH.MHI QHOUQ U am am am am am am cm an an «m NB mm mm mm mm mm mm mm cm «s an on an an “I soauouauu mamnmxuflasu use ewdmmlpfiz soaueuwuo uHHsu\owsmmlpfiz was soauoufiuo madam sowuouauo oasnmxowsmMIvfiz use souuouauo uafisw Au newsman .euow maomv suoom o>HumHom soauuuwuo oaaam\uaa=o new «maneuver coauuuauo uaeau\umammaeaz was soaumuauu «swam aoauouauo mauam\uwuaeueaz can soauouauo geese An newsman dShow maomv ououm o>HumHmm soauouuuu sauna use sowueuauo uaasu owssmuowz use souuouauo madam omsmmlvfix use coauouwuo uafiso IIIIV Au commode dshow «Hewu suoom 0>Huoaom soauoufiuo madam use sofiuouauu undue emanate“: was soauouauo semen amounts“: use sowuouauo uaasu AH scammed .suom maomv ouoom weausaom mesouu sowuouauu m¢< newsman mass: uuoom oedemaem ea moosouowmwa now mundane a «same Shame Self, lst session Other, 1st session Self, 2nd session Other, 2nd session Guilt Self, lst session Other, lst session Self, 2nd session Other, 2nd session Total Anxiety Self, 1st session Other, lst session Self, 2nd session Other, 2nd session Relative Score Self, 1st session Other, 1st session Self, 2nd session Other, 2nd session Ben Masculine Bem Feminine Bem Social Desirability Bem Androgyny Closure Correct Closure Incorrect Closure Difference Closure Scale 283 Table 8 Intercorrelations of BSRI Variables with All Other Variables for the Masculine Ben was: -.29, p<.15 .14, p<.30 -.08, p<.39 .28, p<.15 .02, p<.46 e002, p<.49 .17, p<.27 .10, p<.35 -.16, p<.28 .09, p<.37 -.16, p<.28 .26, p<.17 .35, p<.10 -.14, p<.31 .13, p<.32 -.44, p<.05 1.00, p<.001 .29, p<.15 .34, p<.11 -.47, p<.04 -007, p<.‘0 -.27, p<.17 .13, p<.32 .15, p<.29 Criterion Group Ben Feminine p<.39 p<.21 p<.26 p<.07 .002, p<.49 .08, .04, .05, .04, .19, .11, .34, p<.38 p<.44 p<.42 p<.43 p<.25 p<.35 p<.11 p<.003 p<.45 p<.44 p<.40 Ben Social Desirability 0‘1, p<.06 -.14, p<.31 .29, p<.15 ‘.07, p<.40 .53, p<.02 .61, p<.01 .61, p<.009 .42, p<.07 .52, p<.02 .27, p<.17 .34, p<.11 .11, p<.34 ’oOls P<.47 .57, p<.01 .03, p<.44 .30, p<.14 .34, p<.11 .39, p<.07 1.00, p<.001 .12, p<.33 -.20, p<.24 .16, p<.29 -.21, p<.23 -.20, p<.23 Ben Androgypy .31, -.29, .06, -.37, .49, .49, .53, .33, p<.13 P<015 p<.40 p<.09 p<.03 p<.03 p<.02 p<.12 p<.05 p<.35 p<.21 p<.30 p<.40 p<.01 p<.13 p<.02 p<.04 p<.003 p<.33 P<.001 p<.29 p‘.35 p<.29 p<.33 284 with shame; was positively correlated with Incorrect Closure (r-.43, p < .08); was negatively correlated with the Relative score on all test forms, reflective of positive association between femininity and shame; was non-significantly and negatively correlated with Total Anxiety; and with Guilt. The androgyny score was non-significantly and positively correlated with Guilt; with Relative score; and negatively with Shame (Table 9). Insert Table 9 about here Androgynous Group (males-l8, females-19) Social Desirability correlated significantly and positively with Masculin- ity (r-.42, p < .004); with Femininity (r-.37, p < .01); with the Relative score; and negatively with Shame; and with Closure Difference (r--.23, p < .07). Masculinity correlated positively with Femininity (r-.91, p < .001), and negatively with Correct Closure (r--.39, p < .008). Femininity correlated pos- itively with the Relative score, reflective of positive association between femininity and guilt; and with Guilt (Table 10). Insert Table 10 about here T-tests for sex differences within categories were not significant. T-tests for sex differences on BSRI scales within criterion categories demonstrated no significant differences for the masculine group. There were significant sex dif- ferences for the Androgynous group pointing to greater Social Desirability scores for females than males, and greater deviation from perfect (0) androgyny in the direction of masculinity for males than for females (Tables 11 & 12). Insert Tables 11 6 12 about here Shame Self, 1st session Other, 1st session Self, 2nd session Other, 2nd session Guilt Self, lst session Other, lst session Self, 2nd session Other, 2nd session Total Anxiety Self, lst session Other, lst session Self, 2nd session Other, 2nd session Relative Score Self, 1st session Other, lst session Self, 2nd session Other, 2nd session Bem Masculine Bem Feminine Dem Social Desirability Bem Androgyny Closure Correct Closure Incorrect Closure Difference Closure Scale 285 Table 9 Intercorrelations of BSRI Variables with All Other Variables for the Feminine Criterion Group Ben Be Social Desirability Masculine -.47, p<.06 .05, -.39, p<.10 .29, -.04, p<.44 .04, -.06, p<.41 -.10, -.05, p<.43 .08, -.34, p<.13 -.42, —.51, p<.04 .07, -.33, p<.14 -.20, -.35, p<.12 .10, -.04, p<.44 -.31, -.41, p<.09 .08, -.14, p<.33 -.28, .15, p<.31 -.15, -.32, p<.14 -.34, .30, p<.16 .008, -.25, p<.21 -.13, .47, p<.06 1.00, .63, p<.01 .76, 1.00, p<.001 .47, -.005, p<.49 -.68, -.20, p<.26 -.12, .35, p<.12 .48, -.37, p<.1l -.42, -.42, p<.08 -.42, p<.43 p<.l7 p<.44 p<.36 p<.39 p<.08 p<.40 p<.26 p<.37 p<.15 p<.39 p<.18 p<.31 p<.13 p<.49 p<.34 p<.001 p<.002 p<.06 p<.007 p<.35 p<.05 p<.08 p<.08 Bem Feminine .09, .20, -0009, .03, -.32, -.58, -.16, -.27, -.l9, -.54, -.13, -.26, -.40, -.64, -.24, -.41, .76, 1.00, .63, -.06, -.12, .43, -.38, -.39, p<.38 p<.26 p<.48 p<.45 p<.14 p<.02 p<.30 p<.19 p<.26 p<.03 p<.36 p<.19 p<.09 p<.Ol p<.22 p<.09 p<.002 p<.001 p<.01 p<.42 p<.35 p<.08 p<.10 p<.10 Ben Androgyny .02, -.21, -.05, .23, -.49, -.02, -.29, .01, -.38, -.ll, -.25, .18, -.18, -.21, -.27, -.26, -e68, -.06, p<.47 p<.24 p<.43 p<.23 p<.05 p<.46 p<.17 p<.48 p<.10 p<.36 p<.21 p<.28 p<.28 p<.25 p<.19 p<.20 p<.007 p<.42 -0005, p<049 1.00, .06, -.24, .22, .20, p<.001 p<.42 p<.21 p<.24 p<.25 Shame Self, 1st session Other, lst session 286 Table 10 Intercorrelations of BSRI Variables with All Other Variables for the Androgynous Criterion Group Self, 2nd session Other, 2nd session Guilt Self, lst session Other, lst session Self, 2nd session Other, 2nd session Total Anxiety Self, lst session Other, lst session Self, 2nd session Other, 2nd session Relative Score Self, 1st session Other, lst session Self, 2nd session Other, 2nd session Bem Masculine Bem Feminine Bem Social Desirability Bem Androgyny Closure Correct Closure Incorrect Closure Difference Closure Scale Ben Ben Ben Bem Social Desirabiligy, Masculine Feminine Androgvny -.11, p<.24 .03, p<.4l .05, p<.36 -.04, p< .39 .01, p<.45 .08, p<.30 .03, p<.42 -.l3, p<.22 -.21, p<.09 -.10, p<.26 -.O9, p<.28 .008, p<.47 -.002, p<.49 -.12, p<.22 -.20, p<.11 -.22, p<.09 .15, p<.17 .14, p<.19 .22, p<.09 .20, p<.11 .15, p<.18 .11, p<.25 .09, p<.28 .008, p<.48 .04, p<.39 .08, p<.30 .19, p<.12 .28, p<.04 .20, p<.10 .14, p<.19 .04, p<.38 -.12, p<.22 .02, p<.44 .06, p<.36 .09, p<.29 .08, p<.30 .08, p<.3l .10, p<.26 .06, p<.35 -.07, p<.32 -.09, p<.28 -.01, p<.46 .04, p<.38 .14, p<.18 .10, p<.26 .001, p<.49 -.O9, p<.29 ¢.18, p<.13 .11, p<.25 -.02, p<.44 .07, p<.32 .21, p<.10 -.l4, p<.19 -.20, p<.10 -.08, p<.31 .31, p<.03 .42, p<.005 .19, p<.11 .30, p<.03 .30, p<.03 .25, p<.06 .31, p<.02 .34, p<.01 .19, p<.12 .42, p<.004 1.00, p<.001 .91, p<.001 -.05, p<.36 .37, p<.01 ‘ .91, p<.001 1.00, p<.001 .43, p<.004 1.00, p<.001 .42, p<.004 .37, p<.Ol -.01, p<.47 -.01, p<.47 .05, p<.36 .43, p<.004 1.00, p<.001 -.17, p<.15 -.39, p<.008 -.34, p<.01 .08, p<.30 .18, p<.13 .18, p<.14 .21, p<.10 .16, p<.16 -.23, p<.07 -.36, p<.01 -.36, p<.01 -.06, p<.34 -.20, p<.10 -.31, p<.02 -.30, p<.03 -.04, p<.39 287 Table 11 T-tests for Sex Differences on BSRI Variables Within.Androgynous Criterion Group Sex Mean _SR T-score _d_f_ 2 Social Desirability m918 4.80 .46 -3.34 35 .002 f-19 5.32 .47 Androgyny m918 -.14 .21 3.03 35 .005 f-lg - e 07 e 21 Masculine m918 4.78 .51 -.93 35 .35 f-l9 4.95 .58 Feminine m918 4.67 .48 -1.90 35 .06 f-19 5.03 .66 288 Table 12 T-tests for Sex Differences on BSRI Variables Within.Masculine Criterion Group Sex Mean _S_ll T-score d_f_ Social Desirability mP10 4.96 .54 .01 12 f- 4 4.96 .16 Masculine mPIO 5.53 .28 1.36 12 f- 4 5.30 .29 Feminine mfllo 4.16 .42 -.07 12 f- 4 4.17 .20 Androgyny mle -1.37 .45 -1.06 12 f- 4 -1.12 .15 .99 .19 .94 289 Overall, the results of the BSRI criterion analysis provided no particular- ly useful additional information concerning the interaction.among variables. Subject Date on Excluded Subjects: AAS Data from excluded cases for Self and Other forms of the AAS (session 1) are comparable to the overall group (Table 13). Insert Table 13 about here 290 Table 13 AAS Data on Excluded Subjects as g..._.. Total Relative Total Relative Subject Epilg. §h§§g. Anxiety Score Guilt Shame Anxiety Score 1 169 107 276 13 161 111 4 272 4 2 123 125 248 O 151 149 300 -1 3 89 59 148 -l 125 94 219 —S 4 86 130 216 7 195 124 319 - ‘9 159 168 327 -4 162 165 327 3 6 194 193 387 5 205 197 402 4 APPENDIX C SUBJECT CLASSIFICATION 291 296 297 298 299 300 APPENDIX C Subject Classification Classification of subjects by sex, into high, medium, and low shame-guilt levels (Self and Other forms) Sex Role Categories (based on median cutoffs) by Shame- Guilt Intensity (Self and Other forms) Mean Closure Scale score within sex role categories and shame-guilt levels (Self and Other forms) Number of subjects of each sex role category within high, medium and low Closure Scale score groupings and shame- guilt levels. Cases with consistent and inconsistent designations over two sessions. Comparison between AAS relative score and distribution- based shame-guilt score designations. Subjects Included in High, Medium, Low, Relative Score Categories Within High, Medium, Low, Distribution-based Shame-Guilt Levels 292 293 Pilot data on the Self and Other forms of the AAS was used to assess sev- eral different approaches to subject classification for the Main Study. Data was viewed initially by designating high, medium and low ranges on AAS Shame and Guilt scores separately and including apprOpriate cases in each grouping. As can be seen from Table 1, difficulty arose immediately in the availability of subjects for each cell. The present writer further explored the combination of shame-guilt level classification with sex role classification (Table 2) and with both sex role classification and Closure Flexibility groupings (Tables 3 & 4). In every case difficulty arose in subject availability. A decision was made on this basis to use only shame-guilt classification to class subjects and to use a simplified, medianrbased, high-low, shame-guilt designation. Insert Tables 1-4 about here Reliability of the AAS Distributionrbased Classifications over Two Sessions A reliability check was done on the AAS (Other form, lst session) by com- paring designations of high, middle, and low shame and guilt score combinations with the same designations based on the cuts of the new distributions available from the second session (Other form). The results show that of a total sample of 96, 46 cases remain the same designation in the two sessions. Extreme cases were heavily represented as consistent over the two sessions. The cases that did fluctuate were middle range values (Table 5). Insert Table 5 about here 294 Table 1 Classification of Subjects by Sex, into High, Medium, and Low, Shame-Guilt Levels (Self and Other Forms) High,,Middle, and Low Shame and Guilt by Sex (Self form) High Guilt Middle Guilt Low Guilt High Shame males-7 males-4 males-2 female8813 females-3 females-2 ‘Mid Shame males-l males-10 males-5 females-6 females-ll females-2 Low Shame males-O males-5 males-10 females-3 females-3 females-10 High, Middle, and Low Shame and Guilt by Sex (Other form: High Guilt Middle Guilt Low Guilt High Shame males=6 males-7 males-2 females-13 females-2 females-3 Mid Shame males-4 males-7 males-3 females-8 females-9 females=3 Low Shame males-0 malest4 malestlz females-2 females=2 females-ll 295 Table 2 Sex Role Categories (based on median cutoffs) by Shame-Guilt Intensity (Self and Other forms) Self form High Guilt Middle Guilt Low Guilt High Shame N-Zl Mh7 N-6 Mb4 N-4 MPZ F-l4 F-2 F-2 ‘M F A U M. F A U ‘M F A U S 3 8 5 3 l l l O O O 4 Mid Shame N-7 Mhl N-22 Mill N-7 MhS F-6 F-ll F-2 M F A U M F A U M. F A U l 4 l l 2 4 8 8 3 l l 2 Low Shame N-3 MFG N=8 MP5 N819 Mth F-3 F-3 F-9 M F A U M. F A U M. F A U l l l O l 2 2 3 3 2 7 7 Other form High Guilt Middle Guilt Low Guilt High Shame N-18 MF6 N-9 MP7 N-S M92 F-12 F-2 F-3 M F A U M, F A U M F A U 4 3 ll 1 O 2 2 5 3 l O 1 Mid Shame N312 MP4 N-l6 Mb? N-6 Mh3 F-8 F-9 F-3 M F A U M F A U M F A U l 4 3 4 4 l 5 6 2 l l 2 Low Shame N-2 MFG N86 Mh4 N822 MP12 F-Z F-2 F-ll M F A U M F A U M F A U l 0 l O 0 4 1 1 4 2 6 10 Categories: MhMasculine FnFeminine A-Androgynous UsUndifferentiated Self form High Shame Mid Shame Low Shame Other form High Shame Mid Shame Low Shame 296 Table 3 Mean Closure Scale Score Within Sex Role Categories and Shame-Guilt High Guilt M F A U 44.2 38.3 32.8 37.2 Overall mean938.l M F A U 29 32 25 26 Overall mean-28 M F A U 13 22 29 0 Overall mean921.3 High Guilt M F A O 43.3 35.6 34.0 35.5 Overall mean-37.1 M F A U 38 36 32.3 38.2 Overall mean-36.1 M F A U 13 0 30 0 Overall mean-10.7 Levels (Self and Other form) Middle Guilt M F A 40 41 49 Overall mean-41.7 M P A 41 31 Overall mean-36.7 M F A I, 5 38 Overall meanp36.7 Middle Guilt M F A 0 32 25.5 32.2 Overall mean-22.4 M F A 40 .7 40 Overall mean-40.5 M F A 0 37.7 29 Overall mean-26.4 U 37 U 34.8 40.2 U 26.5 37.5 U U 43.6 37.8 U 39 Low Guilt M ‘ F A U 0 0 0 25.5 Overall mean-25.5 M F A' ~ U 40.6 40 38 34.5 Overall meanP38.2 M F A U 39.8 52.5 43 42.1 Overall mean-44.3 Low Guilt M F A U 35 38 0 49 Overall mean-30.5 M F A U 47.5 28 46 44.5 Overall mean941.5 M F A U 41 41 39.5 37.6 Overall mean=38.7 ll‘lllla‘l'f'u.‘ ‘1'] Number of Subjects of Each Sex Role Category 297 Table 4 Within High, Medium, and Low Closure Scale Score Groupings and Shame-Guilt Levels Self form High Shame-High Guilt High Shame-Middle Guilt High Shame-Low Guilt Middle Shame-High Guilt Middle Shame-Middle Guilt Middle Shame-Low Guilt Low Shame-High Guilt Low Shame-Middle Guilt Low Shame-Low Guilt Other form High Shame—High Guilt High Shame-Middle Guilt High Shame-Low Guilt Middle Shame-High Guilt Middle Shame-Middle Guilt Middle Shame-Low Guilt Low Shame-High Guilt Low Shame-Middle Guilt Low Shame-Low Guilt c>c>hahdc>rac3rd<3:! C’CDP‘CDP‘C3hJC><3:: OOHONNOOONF‘ owowowou—u—amr‘ 3 HMO-“OnbOOOI-‘p O G sah‘hic>hauac>haua>' Closure Scores Middle U M F A U l 3 2 6 4 0 1 l l l 3 O 0 0 1 l O 3 0 0 0 2 2 3 6 l 1 l l 1 O 0 0 0 0 1 1 2 0 2 l 2 0 3 6 Closure Scores Middle U M F A U 0 3 2 8 2 2 0 l 2 l O l l 0 0 l l 3 l 3 l l 1 l 5 0 l 0 l l O 0 O 0 0 0 0 2 0 l 4 3 l 2 7 h‘C>C>P‘C>C><>tdlv:z FIC>C>P4ho0uauoz uoouuoun0 00.0 00.00 05.0 00.00 00.0 50.00 00.0 00.00 00.0 50.00 00.0 50.00 05.0 00.00 00.0 00.00 00.0 50.00 00.0 00.00 05.0 00.00 00.0 00.00 50.0 00.00 00.0 00.00 00.0 50.00 00.0 00.00 o>0uu002 o>0u0mom uoouuou 0euoe 080(08 mo 000u0mom 0>0u0mom UUOHHOUCH UUUHHOU a000lz 0ouoh 00000 uoe moons 0Izv 00.000 00.000 50.00 00.000 05.000 00.000 50.000 05.000 00.500 00.550 00.50 00.050 00.000 00.500 00.000 50.000 00.000 00.000 00.00 00.050 00.000 50.000 05.00 00.000 00.000 00.000 00.00 00.500 50.00 00.000 00.000 00.000 0ouuaoz o>0uewuz u0cua u0eua Anecouoo :00 mxae 00400 50.000 00.500 00.000 50.000 00.000 50.000 00.000 50.000 50.000 05.050 00.000 00.000 05.000 00.000 00.050 00.000 o>0u0mom UHQHH news: 0000 vo>uo000 .00euox 00000 use 40000000 u0euh so m0o>0e=< he: u3om 0 00088 o0300mu 0000030 I 006600 «~300mu I 0000030 I w0ma Nuo0xam 3o0 00300e0 I awoooom I u0ssou ou300e0 I 0000030 I o0ma mmmMMImwmm 0030000 I 0000030 I 000800 0030000 I 0000030 I o0ma sauna 0000 mu300eu I 0000030 I o0eaom 0030000 I oooou3m I 0006 NmmmmmMImmmm 332 05.0I 00.0 00.00 00.0 05.00 un300uu I 05 00.0 05.00 00.0 05.00 000003: I o0uauu 50.0I 50. 00.00 00.0 50.00 ou300ou I 05.0 00.0 00.00 00.0 50.0 uncou3u I 0015 NmmmmmMIMmm 00.0I .50.0 00.00 50.0 00.00 «~300au I 00.0I .00.0 00.00 50.0 00.00 0000030 I 00a800 50.I 00.0 00.00 00.0 00.00 ou300u0 I 00.0I 00.0 00.00 05.0 00.00 0000030 I 0006 00030 0000 00. 50. 00.00 00.00 50.00 «H30000 I 00.I 50.0 00.00 00.0 05.00 mauuu3o I o0uauu 05.0I 00.0 00.00 00.5 50.00 «H300au I 00.I 05.0 00.00 00.0 50.0 mmuuu3o I ”015 «a; 00.0I 00.0 00.00 00.0 00.00 0030000 I 00.0I 00.0 00.00 00.00 00.00 amouu3m I 00a300 00. 00.0 00.00 00.0 00.00 ou300au I 00. 00.0 00.00 50.00 50.00 ucmuu3c I 00:8 Nuo0x=m :00: uuouuooc0 uuouuou acouuouc0 uuonuou o>0uamuz .002I.aom .002I.oom 0ouoa 0wuoa 0ouOH 3.380 SEEM: SSS. no .555. Au.=oov undo: 0000 I 300000 063 H300 333 JUODVOOK UfiQHH uO HHQUO¢ uuoauoa>un vuovauum Hauu .Hmauom uuaua uco Joanvuom uauuh no cauAHuc< an: uaom Oon wmofl NmoH HmoN wHoN #momNH H¢.H wn.~ 03.: Hm.~ n:.~ mn.ww m¢om whom O¢.H moofi 5w.m HNoOfiH mm.m mm.a ow.m mm.: 5:.m Hm.me nm.a mm.H nH.n n:.m mm.fl mo.mm mm.m mo.m am.m oH.m mm.m m:.ou om.H md.m :o.n mn.~ mo.~ :w.oHH 500M mmofl mmoN fiO-N #NoH #bomMH ma.: mn.m :o.n mm.m N.N mw.mma om.m m:.a mo.n wn.m um.a ww.mm mm.~ wn.a o~.m NM.N mm.a :H.wma .wo.H mw.fl ¢~.H mo.d om.H mw.mm om.a mm.a ou.n mo.m mm.m m:.noa o~.m Ho.~ mo.m om.a mu.m :m.w:fl mn.m mn.~ mm.n Oa.n om.N Hm.mm 0:.N wO.N mN.n m¢.m NH.N mwomnd o>wunmmz o>auwwwzo>muaaom o>uuamoa u>auumom Hauuauz noouuoucu uuouuoo aauoh uuouuoucu uuouuoo uumuh Anvcouaw :«v mafia aways s OHAQH nm.ou an.ao ma.mo :amaw nm.ooa oa.wo an.wu mm.:m mo.ou on.wm m:.u:fl mm.mmH Ho.:u mm.mw Hm.mm oa.mo~ mm.mm 0:.Hma Hm.om n¢.mm ::.mm ~o.mo mm.woa um.mm‘ nm.mma mm.mn mm.mma um.mm ma.um um.ama mm.mw :o.maa o>dunwwz o>fiuumom “aqua «anus ousaflag $00003” onsafiau mmmoOSQ ouzflfiau mmmoosm ousaaaM Emmoofim muaafiag mwmoosm onsafiau ammoofim ouzafimu mmmuofim ousflfiag mmwOOSm macaw“ mama huoflxcm 30A maaeom mama ”Hana swam oaasou mama meanm swam maneuu «Has huoflxqu swam 334 m:.n mm.n :m.n mm.o om.n om.m nm.n mm.: #6.: $0.: mo.: mo.m ma.n mm.n nn.: uo.n uuuuuoacu .quI.uom wm.m Ho.n mm.n Mn.N HM.N om.n :m.N mm.: mm.: :N.n HH.M mm.m wo.a om.N mm.m oa.~ Hm.n nm.: NmoH ONofl BB.“ ON.N Nn.N m:.N m:.a mm.m wm.: om.a :n.: mw.~ mn.m on.¢ Hm.H mm.H mu.~ mm.N ow.a mo.~ 5:.: 00.: am.H :m.n mo.m m:.n nn.: u:.~ nm.m mn.n uuouuoo uuouuouca uuouuou .muzI.mom aauOH flauoa Au.:ouv :uonnoom uunua uo “Haooa wa.m mm.a :o.m Hm.m Hm.H 0:.H ow.H ow.H mo.n mm.a :m.H :N.HV mo.H uw.~ mm.H AN.N o>wummuz HauOH ousafiwu I mmmoosm I camsmu ousafimH I ammousm I cams 2 muofixcd aoq ouzafidu I ammoosm I mawaou ousawdu I mmoousm I came uaazm swam ousafiam I mmooosm I ofluamm wuaaadm I mmooosm I oaaa oadgm swam ousaadm I mmouosm I .oawaou ousafiwu I mmooosm I made huofixqm nmax Au.nOuv udoauuu>aa unavaaum Hmou I awauoa an: usom 335 Table 8< Table of Multivariate F Tests for Effects of Shame, Guilt, Sex, and Success-Failure Sources Shame Guilt High Anxiety x Success-Failure High Shame x Success-Failure High Guilt x Success Failure Low Anxiety x Success-Failure High Anxiety x Sex x Success x Failure High Shame x Sex x Success x Failure High Guilt x Sex x Success x Failure Low Anxiety x Sex x Success x Failure Shame x Guilt Interaction E .7383 .2663 1.0353 .6031 .5786 1.6844 .7987 .6139 .3785 1.1116 .6815 .7036 .5851 .4418 1.2707 .6200 .9515 .4067 .7273 .7467 .1318 .5731 .7188 .8913 .3605 .6649 .6474 .7416 .8494 .2770 1&36 an.n ms.n mn.~ ~¢.n an.H ~m.a mm.a no.m om.N mm.H ouaaamu I co.m cs.~ Hu.m nn.n on.H N5.N om.H nn.N Ho.~ cm.~ mmouuzm I uaasm swam cc.< oa.a oc.c mo.n NN.N ne.n . n~.N mo.~ on.u ~a.n uuaawmu I HH.n mo.H nh.m <¢.N nn.H 5H.N cm.H 5N.N ne.~ an.H mmoouam I manna swam uaouuaa>uo uuauaaum Haou QQ.I on.u nN.na wH.m co.H~ oo.o co.n on.NH an.“ . mH.m ousdfimw I nN.HI hm.” mm.nH on.» hm.oH an.h cn.n oo.HH 90.0 . na.c amuuono I uaasw swam on.l ne.H nm.nn nm.o «A.AH on.h nm.n oo.~u «v.0 an.n ousanmu I on.l ~m.a nn.mH on.m HM.°H nm.c hn.n om.Hn nn.o I mH.n «mucosa I 05650 swam uuouuoqau uuouuoo uuouuouan uuouuoo 0>Huamuz u>wumwoz o>uuauoz u>wuauom o>quwnom u>fiufimom uOZIuom anal-om Hayes Hausa Hanna uoouuouun uuuuuoo Hmuoa uuouuouan uuouuou «can: Haas uu>uomao Ago-z dunes “Hana you mouse eflnzv Annouuafipun vuqvaqum van and»: «Huey auuouuu Jada vaunoz you oonauuu> mo aqumama< oaaauaaz m odnuh 337 Table 10 F Test Results Shame Task Success/Failure in Shame Success/Failure in Guilt E .29 .51 .36 1.13 338 Table 11 Analysis of Variance on Negative Feedback for Nested Shane and Guilt Effects in Success and Failure Conditions (Cell Means and Standard Deviations, N -16 cases per cell; Total N-64) Observed Cell Means Trait Negative (in seconds) High Shame - success 283.12 - failure 243.63 Low Guilt - success 248.75 - failure 257.50 Observed Cell Standard Deviations Trait Negative (in seconds) - High Shame - success 85.76 - failure 86.33 High Guilt - success 113.61 - failure ‘ 72.89 F Test Results 1 P. Shame in Success 1.15 .28 Shane in Failure .19 66 Task (Success/Failure .46 .49 339 Table 12 Multiple Analysis of Covariance on Memory- and Tape-Cued Variables, Controlling for Baseline (Neutral) variable (N-32 per cell; Total N-66) Observed Cell Means Ease Memory Taped High Shame 6.18 5.65 6.81 High Guilt 6.15 5.63 6.71 Observed Cell Standard Deviations Base Memory Taped High Shame 2.50 2.82 2.76 High Guilt 2.70 2.75 3.10 Sample Correlation Matrix Base Memory Memory .73 Taped .79 .76 Least Square Estimates of Effects - Effects 1 Variables Base Memory Taped Grand Mean 6.17 5.56 6.76 Group Mean .03 .21 .09 Standard Errors of Least Square Estimates - Effects x Variables Base Memory Taped Grand Mean .32 i .36 .36 Group Mean .65 .69 .73 Raw Regression Coefficients - Independent x Dependent Memory Taped Base .78 .89 36!) Multiple Analysis of Covariance on Memory- and Tape-Cued Variables, Con't Standardized Regression Coefficient - Independent 3 Dependent Memory Taped Base .73 .79 Matrix of Correlations with Covariate Eliminated Memory Tape .66 Least Square Estimates Adjusted for Covariates - Effects x Variables Memory Taped Grand Mean 2.25 1.02 Group Mean .19 .06 Standard Errors of Adjusted Estimates - Effects x Variables Memory Taped Grand Mean .65 .62 Group Mean .67 .66 Statistics For Regression Analysis with 1 Covariate Variable 83. Mult. R Mult. R E. p Memory .56 .73 72.02 .0001 Taped .63 .79 106.87 .0001 Multivariate F Test pa rs Groups (High Shame, High Guilt) .08 .92 361 Table 13 T-tests Computed on.Main Study Variables (asteriks note significant difference) 1. T-teats for differences in time spent reading trait feedback as a function of belief in task feedback (belief-113, disbelief-15. df-126). Mean _S_Q t-value p Positive Trait 235.57 97.66 -.51 .60 Feedback 269.33 99.72 Negative Trait 260.13 97.69 .09 .92 Feedback 257.66 66.52 Neutral Trait 106.29 121.32 .36 .73 Feedback 93.00 110.28 2. T-tests for differences in trait recall as a function of belief in trait feedback (belief-85, disbelief-63, df-126). Mean Recall Correct Positive 5.31 5.55 Incorrect Positive 6.69 6.76 Total Positive 12.01 12.32 Correct Negative 3.76 3.39 Incorrect Negative 7.09 7.88 Total Negative 10.85 11.27 Total Correct 9.09 8.95 Total Incorrect 13.71 16.62 as 5.31 5.55 6.69 6.76 12.01 12.32 3.76 3.39 7.09 7.88 10.85 11.27 9.09 8.95 13.71 16.62 t-value -.60 -.14 -.53 .98 -1.73 -1.09 .22 -1.60 3. T-test for difference in time spent with positive or negative feedback, with- in the success condition, for high shame group subjects (N-l6, df-lS). Mean Positive Trait 277-50 Feedback 513 82.76 5.221% -s15 362 Mean SD t-value p. Negative Trait 283.12 85.76 Feedback 6. T-test for difference in time spent with positive or negative feedback, within the success condition, for high guilt group subjects (N516, df-lS). Mean _S_D_ t-value p Positive Trait 256.37 133.60 .10 .91 Feedback Negative Trait 268.75 113.61 Feedback 5. T-test for difference between M and F scale scores for high shame group subjects (N-32, df-31). Mean S11 t-va lue p Bem Feminine 6.78 .70 -1.50 .16 Masculine 5.07 .93 6. T-test for difference between.M and F scale scores for high guilt group subjects (N-32, df-31). Mean §2_ t-value p_ Masculine 6.80 .90 7. T-test for difference between Shame and Guilt scale scores for total group (N3128, df-127). Mean S_D t-value p AAS Shame 192.86 22.35 12.66 .000* Guilt 163.71 28.33 8. T-test for difference between M and F scale scores for total group (N-128, df-127). Mean SD t-value p BSRI Masculine ' 6.96 .85 1.85 .06 Feminine 6.78 .68 363 9. T-test for difference between M and SD scale scores for total group (N-128, df-127). Mean _S_D_ t-value p BSRI Masculine 6.96 .85 -1.86 .06 Social Desirability 5.08 .57 10. T-test for difference between F and SD scale scores for total group (N-128, df-127). Mean _S_D_ t-value p BSRI Feminine 6.78 .68 —6.23 .OCO* Social Desirability 5.08 .57 11. T-tests for differences between time spent reading positive, negative, and neutral feedback for total group (N-128, df-127). Mean _S_D_ t-value p Positive 237.18 97.63 -1.71 .08 Negative 259.86 96.21 Positive 237.18 97.63 7.71 .000* Neutral 102.96 119.73 Negative 259.86 96.21 9.26 .000* Neutral 102.96 119.73 12. T-tests for differences between baseline, memory-cued, and tape-cued scores on anagram task for total group (N-128, df-127). Base Memory Base Taped Memory Taped 13. Mean 6.66 5.53 6.66 6.69 5.53 6.69 _S_D_ 2.66 2.70 2.66 2.78 2.70 2.78 t-value -6s66 “1.50 6.77 E. .000* .13 .000* T-tests for sex differences on Correct Closure, Incorrect Closure, R-S, Shame, and Guilt scale scores for total group (N8128, Ms66, F-66, df-127). Correct Closure Incorrect Closure RPS Shame vu::va:zru:z'u:3 Mean 99.18 99.36 17.07 18.85 67.98 61.21 163.76 163.67 g1; 22.70 24.71 14.57 11.30 19.65 16.02 28.70 28.17 t-value -s04 -.77 2.13 .02 E. .97 366 Mean _S_Q t-value p Guilt 3M 192.59 22.86 -.13 .90 F 193.09 22.00 16. T-test for difference between M and F scale score for males (N-66, df-63). Mean S_D t-va lue p BSRI Masculine 5.22 .76 6.85 .000* Feminine 6.69 .68 15. T—test for difference between M and F scale score for females (Nh66, df-63). Mean SI; t-value p BSRI Masculine 6.676 .85 -1.57 .12 Feminine 6.86 .69 16. T-test for differences between High anxiety group subjects and Low anxiety group subjects on RPS scale score (Nh32 per group, Total N-66, df-62). Mean §2_ t-value p_ High Anxiety (RPS) 68.06 16.62 2.63 .01* Low Anxiety (R—S) 56.90 19.88 17. T-tests for differences between High shame group subjects and Low anxiety group subjects on R—S and Closure Correct scale scores (N=32 per group, Total N-66 , df-GZ) . Mean §2_ t-value p_ High Shame (R—S) 67.28 17.71 2.20 .03* Low Anxiety (R—S) 56.90 19.88 High Shame 105.60 23.28 1.86 .06 (Closure Correct) Low Anxiety 95.21 20.61 (Closure Correct) 18. T-tests for differences between High guilt group subjects and Low anxiety group subjects on R-8 and Closure Correct scale scores (N-32 per group, Total N866, df-62). ‘Mggr_ SD t-value p_ High Guilt (R-S) 66.15 16.87 2.01 .06* Low Anxiety (RrS) 56.90 19.88 365 Mean §Q_ tdvalue p_ High Guilt 106.81 25.36 1.67 .10 (Closure Correct) Low’Anxiety 95.21 20.61 (Closure Correct) 19, T-tests for differences between High shame group subjects and High guilt group subjects on RPS, Closure Correct, and Closure Incorrect scale scores (N-32 per group, Total N-66, df-62). Mean _S_D t-value p High Shame (R-S) 66.15‘ 16.87 -.26 .79 High Guilt (R—S) 67.28 17.71 (Closure Correct) High Guilt 105.60 23.28 (Closure Correct) High Shame 18.09 10.67 -1.61 .16 (Closure Incorrect) High Guilt 22.50 16.16 (Closure Incorrect) 20. T-test for differences between High anxiety group subjects and High shame group subjects on Correct Closure scale score (N=32 per group, Total N-66, df-62). Mean _SD t-value p High Anxiety 91.62 23.19 -2.37 .02* (Correct Closure) High Shame 105.60 23.28 (Correct Closure) 21. T-test for differences between High anxiety group subjects and High guilt group subjects on Correct Closure scale score (N=32 per group, Total N=66, df-62). Mean _S_D_ t-value p High Anxiety 91.62 23.19 -2.17 .03* (Correct Closure) High Guilt 106.81 25.36 (Correct Closure) 366 22. T-tests for differences between High shame group subjects and High guilt group subjects in the success condition, on time spent reading positive, negative, and neutral trait feedback (N-16 per group, Total N-32, df-30). High Shame High Guilt High Shame High Guilt High Shame High Guilt (positive) (positive) (negative) (negative) (neutral) (neutral) Mean 268.75 283.12 256.37 277.50 96.87 39.37 S_D 113.61 85.76 133.60 82.76 122.86 73.07 t-value p -.97 .36 - s 59 s 56 1.61 .11 23. T-tests for differences between High shame group subjects and High guilt group subjects in the failure condition, on time spent reading positive, negative, and neutral trait feedback (N-16 per group, Total Nb32, df-30). Mean _S_D_ t-value 2 High Shame (positive) 257.50 72.89 .50 .61 High Guilt (positive) 263.63 86.33 High Shame (negative) 238.75 73.67 -.33 .76 High Guilt (negative) 268.75 96.65 High Shame (neutral) 103.75 98.78 -.10 .92 High Guilt (neutral) 107.81 139.53 26. T-test for differences between success and failure condition, High shame groups on neutral feedback reading time scores (N-16 per group, Total N-32, Neutral Feedback Mean _S_D_ t-value p High Shame (success) 39.37 73.07 -1.76 .09 High Shame (failure) 107.81 139.53 25. T-test for differences between success and failure condition, High guilt groups on neutral feedback reading time scores (N=16,per group, Total N-32, df-30). Neutral Feedback Mean _S_D t-value p High Guilt (success) 96.87 122.86 -.17 .86 High Guilt (failure) 103.75 98.78 367 26. T-test for difference between M and F scale scores for High shame subject group (N-32, df-3l). Mean _S_D_ t-value p BSRI Masculine 5.07 .93 1.50 .16 Feminine 6.78 .70 27. T-tests for difference between'M and F scale scores for High shame males (N-16, df-lS), and females (N-16, df-lS), separately. Mean SE t-value p Males (Masculine) 5.27~ .80 2.16 .06* (Feminine) 6.66 .86 Females (Masculine) 6.86 1.02 -.13 .90 (Feminine) 6.89 .51 28. T-test for difference between M and F scale scores for High guilt males (N-16, df-15), and females (N-l6, df-lS), separately. Mean §2_ t-value p_ Males (Masculine) 5.21 .66 3.20 .006* (Feminine) 6.66 .66 Females (Masculine) 6.39 .95 -1.15 .26 (Feminine) 6.70 .78 29. T-tests for differences between positive and negative feedback reading time scores,within the success condition, for High guilt group subjects (Ni16, df-lS), and High shame group subjects (N-l6, df-15), separately. Success Condition Mean §2_ t‘YflkEE 2. High Guilt (positive) 256.37 133.60 .10 .91 (negative) 268.75 113.61 High Shame (positive) 277.50 87.76 -.15 .88 (negative) 283.12 85.76 30. T-tests for differences between positive and negative feedback reading time scores, within the failure condition, for High guilt group subjects (N=l6, df=15), and High shame group subjects (N=16, df-lS), separately. Failure Condition High Guilt (positive) (negative) High Shame (positive) (negative) Mean 238.75 257.50 268.75 263.63 368 73.67 72.89 96.65 86.33 tdvalue -s69 .19 APPENDIX F RECALL OF TRAIT FEEDBACK 369 350 00.I 0.n H.0A o.m nc.I 0.H 0.na o.m 00.I «.H «.0H «.0 0m.I «.a n.0a m.m 00.I 0.“ n.nH m.0 «0.I 0.H 0.nH H.0H n«.«I n0.H n«.¢u «H.w nu. oo.« n«.ca an.o «n.HI n0. «A.mn «~.0 m«.H «0.« oo.mH «H.m oo.nI «n.« 00.0” «0.0 oo.HI n«.H m«.m« «n.» «n.I cc.« oo.n~ 00.0 an.~I on.a on.n~ n«.0 «0. «a. «u.nn «H.H~ «0.I «0.~ «0.cH «~.o n«.HI oo.« «0.n~ cm.« en.I n«.H «H.«H oo.m n«.«I n«.« n«.na oc.o an.HI co.“ an.«~ cc.HH n«. oc.« n«.0a n«.a 00. «H.« an.«a «n.~u uosuuouou uoouuou uoouuousu noouuou UUZIaom quIsom Hauoa asuoa sowuwvsou ououHsMInooooam nos .som .Ho>sa uaasoIoaonm >0 xuoovoom uwoue mo HHouox one: m«.HH n«.oa «0.Ha «0.m on.on «0.0m on.HH «H.HH «n.HH n«.o« «0.0a «0.0 «0.Hn oo.«H «0.0a «0.0H O “”2 0.089 0.5 0.m m.HH 0.0 m.m «.« 0.n «.«H «.0 c.n m.« 0.n 0.«~ a.« «.n 0.« «.m «.aa 0.0 0.0 c.« 0.m «.afl 0.0 «.n H.« «.0 n.«a 0.0 0.0 00.0 n«.n «0.0a n«.n «0.0 oo.« n«.n om.nn n«.« n«.n m«.0 «0.m «m.uH «0.0 00.0 «0.0 n«.m n«.0~ «n.0 «0.0 m«.0 0«.« «m.HH m«.« «0.0 «H.« on.n «0.0H «~.0 n«.0 n«.« n«.n «H.ma «m.« m«.n 00.5 «0.m «H.HH 00.0 «H.n n«.0 «H.n «0.«H «0.0 00.0 «0.« «H.m oo.«H oo.« 00.0 «a.» n«.« «H.HH «n.0 m«.0 n«.0 «0.n «H.HH n«.n «m.n m«.0 «n.n «0.«H oo.« «0.0 oo.« 00.0 on.HH 00.0 00.0 oo.« «0.n «0.«H n«.« «0.0 n«.0 «0.0 oc.nH n«.0 m«.0 .woz .wuz .oom .eom .moa uuouuoucn uuouuou asuoa uuouuooou uoouuoo Am«~lz Hausa “Haas use mouse 0:20 a manna moamaom nods: muoaxo< 304 usage swam usage as“: suoaxa< swam ouoaaomI amooosmIoHoaum ouaafiouI soooosmIonz Nmmmmmdlmmm susauaml ammuosmIoamamm oumaqohI mmooosmIoHsz mafiao sud: susawmmI mmsuoomIoHsamm ousaaomI soooosmIoamx flag susdfiamI amouonmlodaah unsaammI amouoszsHo: .Nmmmmmanmumm. 351 00.I «.n 0.00 0.0 o.HH m.« 0.0 H.«n «.0 0.0 00.I «.H 0.00 0.0 «.HH «.« 0.0 o.«H 0.0 «.0 «0.I «.H 0.0a 0.0 «.0H 0.0 0.0 H.«H 0.0 0.0 uoouuoosu uoouuoo uuouaoucH uoouuoo .mwz .moz .moz .mom .mom .mom usZIaom mozIoom Houoa «snob uuouuoocw uoouuou Houoa uosuuouaa uuouuou Haney asouu Houoa ouoaaom mmouuom Avosawuaouv APPENDIX C THEORETICAL REFINEMENT 352 Appendix G Theoretical Refinement The implication of the results of the present study have prompted this writer to refine the conceptualizations of shame and guilt. The finding of heightened separateness and vigilence, along with the behavioral pattern in response to failure, suggest a very different view of shame, and of individuals who are differentially responsive to shame, than is apparent in the theoretical work and empirical studies carried out by this and other writers. It is hoped that this refined conceptualization of shame and guilt will lead to more accur- ate operationalizations of these effects in the future. This reconceptualization emphasizes the developmental roots of shame and guilt and examines their relationship to concurrent development of other aspects of physical, intellectual-cognitive, moral, social, and emotional growth. These refinements initially restate Erikson's (1951) developmental theory and addi- tionally draw heavily on the concepts of Odier (1956), Tomkins (1962), and Seligman (1975), whose direct relevance was not readily apparent at the initi- ation of this project. Of particular importance in understanding the difficulty in separating shame and guilt is the early childhood developmental sequence during which these effects become established. The two affects develop as a function of differ- ent but sequential early life periods and occur with concurrent developments of other aspects of physical, intellectual-cognitive, moral, social, and emotional growth. As has been discussed earlier in the main text, shame and guilt are affects, often accompanied by anxiety. Shame has a specific focus characterized cogni- tively by acute self-awareness, where attention is paid to the individual's own shortcomings and failures. There is an active fear attached to shame, early in 353 356 life, of rejection and abandonment by significant others or a highly salient external environment, generally. Later in life this may take on the character of the self rejecting the whole self. Physical changes, due to the shame experience, may include blushing, downcast gaze, partial postural collapse, perspiring, and a general turning inward to close the body/self from.the environment's perceived or imagined negative assessment. Behaviorally, shame may prompt withdrawal (physically and/or emotionally) or a defensive, defiant shamelessness. Guilt typically arises in response to transgressions of parentally- environmentally provided values, norms, and laws that have been learned and internalized. Transgressions or thoughts of transgressing, prompt the rising of guilt anxiety, for the wrongdoing or contimplation of wrongdoing. The active fear in guilt is one of punishment. The anxiety is felt in relation to a specific offending aspect of the self. Attention is drawn to, and judgement is focused on, that aspect of self. Depending on the severity of the trans- gression, the self extracts a suitable punishment from itself in the form of lOwering self-esteem. The self is intact in this process and assessment of culpability can be, but is not necessarily, rational. If appropriate, the guilty individual may direct some of the blame for the transgression on to the external environment. The self is under scrutiny in guilt, but it is focused on the specific offending aspect of the self, rather than the whole self, as in shame. Erikson describes (1950) the receptivity to shame as a potentially nega— tive outcome that arises during that period when the child is making a trans- ition to a separate and autonomous sense of self. This period follows the earliest life stage (i.e., Trust versus Mistrust) wherein the child is fully dependent on the primary parenting agent(s). In the Autonomy versus Shame/ 355 Self-doubt stage, according to Erikson, the child moves from full dependence on the caretaker to increasingly independent and self-assured exploration of the environment. This is ideally facilitated by parental emotional support and instrumental guidance. The establishment of autonomy comes, in part, from individual mastery experiences, and these attempts at mastery may or may not meet with success. There is a risk of failure. In failure, there is a risk of shame and self-doubt. The experience of shame and the incomplete separateness of self are inextricably tied. Being so, the shame experience carries a power and weight that can be awesome to the child. A series of shame experiences early in life could, if not remedied, prompt the establishment of a generalized negative self view, consequent low self-esteem, and fear of rejection or abandonment. This in turn may prompt ongoing insecurity and anger generated by a felt lack of security that is directed toward the self. According to Erikson, the potential for guilt arises in the Initiative ver- sus Guilt stage. This period follows the Autonomy versus Shame/Self-doubt stage and is characterized by self-assured activity in the environment. Erikson views development during this stage as a compromise arising in part from the giving up of the child's exclusive relationship with the mother (i.e., separation). Attention is progressively refocused toward the environment and agemates, and away from the parent(s). This movement is facilitated by support by the par- ents for the child to participate more in life outside the home. In.western culture this period acts as a preparation for school entrance with its more extended periods of separation from the parents. Guilt also arises from the self-assured capacity for competence, when energetic activities of the child are called to question or limited. This stage is generally a rule setting per- iod, during which the child learns where and how to direct his or her energy. Transgressions of rules may prompt both physical and love withdrawal 356 punishments. The child learns more specifically what will gain or risk the loss of positive responses from.others. The capacity for understanding the impact of his or her own behavior parallels the evolution of the child's cognitive capacity and increasing differentiation. The cognitive development also explains the more contained and rational focus in the occurence of guilt. Where the excessive experience of guilt can debilitate the developing individ- ual, the impact differs from that of earlier occuring excessive shame. There is the possibility that one can either come to the Initiative versus Guilt stage with relatively adaptive resolutions to prior stages, or less adaptive resolutions. To the extent that resolutions to the Autonomy versus Shame/Self-doubt and/or Trust versus Mistrust stages are weighted in the neg- ative direction, and are not corrected early, the subsequent stages theoretic- ally become more difficult to resolve adaptively. The issues represented by the shame and mistrust labels remain salient for longer periods and may impact the overall focus of the personality. To the extent that trust and autonomy develop to adequate degrees, the child has a firmer foundation with which to deal with the relevant issues of the Initiative versus Guilt stage. According to several theorists (Tomkins, 1963; Kell, 1973; Negri, 1976, Bassos, 1973; Kaufman, 1976), the severity of the shame experience is due to a forced emotional separation from the parenting agent at a time when the parent's emotional support is most needed. Along with these others, it is the present writer's view that this separation is experiences as painful, in part, because the young child has not yet established a fully separate sense of self (i.e., autonomy). The child cannot effectively master the environment, nor does the child have the cognitive capacity to understand the reasons for the withdrawal of parental support or its temporary nature. All that the child understands is that his or her security has been severely threatened. Odier (1956) points out: 357 The infantile realistic tendencies utilized by affectivity are especially striking in the relationships which tie the small child to his parents and educators. These relationships form the first, and therefore, the decisive stage in the child's social evolution. Indeed, they dominate the whole problem of the formation, struc- ture, and preservation of the famous feelings of security. Affective realism can be considered an essential aspect of the child's life since he imagines all his joys and sorrows to be due to external causes and all his happiness and unhappiness to be dependent on the intervention of the human beings and things around him. (Odier, 1956, page 20). Lewis (1971) points out that extended infantile dependency prompts the development of anaclitic identification with the caregiving parent. This is discussed by other theorists and researchers as an aspect of the infantile attachment response. The child's incomplete development of autonomy, coupled with a positive attachment to the parent(s) may, with forced emotional sep- aration, prompt intense anxiety. This is congruent with Piers' (1953) sugges- tion that shame carries with it a fear of abandonment which, to the child, may be experienced as a threat to both emotional and physical survival. Lewis' description of the part played by anaclitic identification in shame is also consistent with Fenichel's (1965) description of responses to traumatic events which consist of a regression to oral stage functioning (i.e., helplessness and anxiety). The present writer includes early occurences of shame in the trauma category based on their quality of abrupt and severe onset. The trauma exper- ienced in shame is the anxiety arising from emotional separation from the parent and the fear that complete rejection or abandonment is occuring. The "separ- ation" quality of the anxiety parallels that occuring in actual prolonged phys- ical separation from the parents. The interpretation of the parental depriv- ation literature suggests a link between actual separation and fears of aband- onment (e.g., Bowlby,f7' ‘). The present writer suggests that shame, occuring with the loss of parental support (i.e., emotional separation), can also lead to the ongoing fear of abandonment. A number or researchers and child 358 development specialists suggest that both temporary or permanent physical deprivation can negatively affect child development. Odier (1956) adds that the same outcome is possible through parental emotional absence, while phys- ical presence is maintained. Anxiety seems to accompany early life separation whether the separation is physical or emotional. The findings suggest that prolonged physical separation prompts grieving, depression, and in some cases, diminished capacity to physically thrive or emotionally attach to human objects. These findings have parallels in relation to shame. Outcomes in regard to physical separation may have slightly less severe parallels in emotional separation, including possible depression, temporary regression to apathy, and ambivalence in emotionally attaching to human objects. The lessened severity may be due to the fact that the early infancy experience may have been need satisfying, even for the later insecure child. Odier (1956) suggests that the child's limited capacity for the cognitive separation of "being" and "doing" and overall cognitive immaturity, coupled with the inherent threat in early life emotional or physical separation, leaves the child exceedingly vunerable to insecurity-an insecurity that can be easily maintained by a very few repetitions of the separation-based anxiety. The pres- ent writer suggests that when emotional separations occur because of some "doing" aspect of the child's behavior, the child feels anxious about the whole self. This anxiety may lead to insecurity, an insecurity based on a lack of knowledge about what ways of being or behavior that will insure parental sup- port. Seligman's (1975) human and animal research on learned helplessness demr onstrates operationally the implications of the more abstract construct of insecurity. He suggests that experienced lack of control over positive or negative outcomes relevant to the self can lead to anxiety and a learned 359 helplessness condition. The organism becomes anxious initially and then learns the futility of responding. This loss of control leads to reduced response efforts, and ultimately to apathy and possible death. Consistent with Odier's view regarding the establishment of ongoing insecurity, the learned helplessness condition can be made relatively permanent with a very few repetitions of the experienced lack of control. The present writer sug- gests that the learning mechanism described by Seligman is a process tied to the shame experience where, by parental rebuff or the child's personal sense of failure, emotional separation anxiety is experienced, insecurity is estab- lished, the whole self is condemned by the child, and the child learns that he or she cannot fully control the positive regard of the significant other. If the child comes to believe that control is impossible, helplessness and apathy will likely develop. Yet, shame, accompanied by anxiety, may-in fact, function to avoid or deny the experience of helplessness (elaboration below). Failure may generate shame because parents deride the child's frustrated attempts at mastery or conformity to parental expectations, or the parent may simply not be aware of the importance of the attempted mastery for the child, but the situations are not limited to these. Shame can also be generated by the inadvertant disregard of the child's needs for instrumental guidance or emotional support, which may or may not be effectively expressed by the child. It must be pointed out that the child is just emerging from a period of full dependence on the parent, where needs were adequately met by the parents. Few (or no) expectations were set by the parents. The introduction of parental expectations, as well as the simple fallibility of parents in regard to the child's needs, may prompt the child to experience a severing of the emotional bond between parent and child when reassuring support is most needed by the child. 360 This severing may be experienced by the child whether or not the parent actually has emotionally withdrawn from.the child. The thinking capacity of the child is limited. Piaget (1968) characterizes this early period as one of infantile realism, dominated by prelogic, precausality, and magical think- ing (similar to Sullivan's parataxic stage), where subjective experience is attributed to the world at large. The child thinks he or she has been rejected and through projection, has been. In the absence of immediate reassurance, this perception takes hold as a truth, fostered by a second aspect of infantile realism, consisting of the child's viewing his or her own percepts as absolute- ly and unconditionally true. The results of the present study support this occurence. The high shame subjects tended to be highly separate and highly sensitized to potential internal and external threat. Failure by the child at this stage of exploration prompts regression to full dependence on the parent and, in the absence of their support or in response to their shaming, shame may be generated in the child. This may in turn prompt further regression to helplessness. In part, the potency of the shame experience is due to the fact that the child has little or no esteem bolstering prior history of mastery to fall back on in the face of early failures. This fact lends a heightened importance to mastery, failure, and the need for reassuring support from the highly salient parents. Seligman (1975) in positing a view of developing potential for mas- tery in organisms, suggests failure prompts an awareness of the discrepancy between what one would like to do and what one is able to do successfully. To the present writer, learned helplessness parallels the sequence of events typ— ical of the shame experience, where shame anxiety is generated in the face of failure, and regression may occur to behaviors typical of the Eriksonian depend- ency stage (i.e., Trust versus Mistrust), finally leading to a loss of will to 361 master, attach or thrive. Initial failures in the Autonomy-Shame stage very likely prompt immedi- ate regression, in the absence of parental support or guidance. As the parent communicates more expectations for the autonomous functioning of the child, or conformity to parental expectations, they may make themselves less available for a immediate reassurance of either the instrumental or emotional type. The child is confronted with the experienced failure and the absence or loss of parental support. The child feels anxiety for both reasons-the instrument- al failure being magnified by the parental loss with which it is linked. The child is not aware, particularly in early instances of breaks in the parent- child emotional bond, that the parents' love or support will return. Expres- sions of anger at the loss of the parental support may therefore be moderated. In part, there is a belief that the parent is "all good", and the self is at fault. Menacher (1953) suggests that masochism.develops when the child main- tains the positive view of the parents at the expense of his or her own self- esteem, and directs resentment at the self. These factors give rise to all the phenomenological aspects of shame described by Lynd (1958). The child: feels betrayed by his or her world (i.e., significant others); is startled by the abruptness and unexpectedness of what is experienced as a severe negative sanc- tion (i.e., loss of support); is put in the position where he or she is acutely separate, helpless, and somehow the cause of his or her own rejection; is prompted to focus attention on the self to find the cause of rejection. These occurences, in combination with the lack of complete autonomy, and the immatur- ity of the cognitive skills (i.e., infantile realism) leaves the child with two general messages: "You are inadequate", "You must please your parents in order to avoid abandonment". Levin (1967) points out that the shame experience is inextricably tied to 362 that period in which the child learns what is expected of him or her, in order to avoid rejection and maintain a continuance of supportive contact with the caregiving other. Within the transition to independent functioning there is a need for continuance of parental support. This support may become more cons tingent on conformity to parental values and performance expectations. In the extreme, support may become wholly contingent on these expectations. In this sense anxiety, specifically shame, becomes a signal for the possibility of impending rejection. The anxiety may be used, as Seligman (1975) suggests, to avoid helplessness, mobilizing and focusing energy toward the mastery of the necessary skills that will assure support or positive reinforcement. Anxiety can serve this useful and forwarddmoving purpose. Yet, as the present writer has pointed out, shame, accompanied by anxiety, occurring before complete autonomy of the self is established, can be an overly powerful experience to the child, prompting heightened separateness and vigilence. It may lead to a sense of generalized insecurity, which in turn reinforces the need to please the parent. This may lead to activities or ways of being that please the other, but are to the detriment of the self. Rogers' (1963) discussion of the development of conditional worth parallels this viewpoint where, having learned the conditions which assure support, awareness of alternatives diminishes and alternatives become. anxiety provoking when confronted. Any behavior, feeling, or attitude, may come to be considered the "cause" of rejection, for no other reason than the fact that it coincided with the onset of a perceived parental emotional withdrawal (e.g., classical condition- ing). This may also lead to compensatory changes by the child, which have no impact on the parental response, leading to further feelings of frustration, shame, futility and a maintaining of the insecurity. Seligman (1975) also notes the occurence of superstitious learning as a function of the lack of control 363 over outcomes relevant to the self. In regard to our discussion, shame can come to be a potent motivator for the continuance of certain activities and being states (i.e., paired with positive reinforcement by parent or the term- ination of parentally caused noxious stimulus) or the termination of certain activities and being states (i.e., paired with negative reinforcement by the parent or the onset of parentally caused noxious stimuli). The child, confronted by parental withdrawal can give up into helpless- ness, or can use anxiety to mobilize the self to mastery, or can interpret the experienced parental rejection as a mastery capacity. This last point though seemingly convoluted, is a very viable outcome of extensive shame experiences. It is congruent with Seligman's view of anxiety as a motivator for mastery. It is also consistent with the cognitive capacities of the young child. The child interprets parental withdrawal as rejection of the child's whole self, therefore the child's "being" is capable of causing rejection. Taking responr ibility for rejection gives the child a sense of control (i.e., mastery) over the world and some understanding of the occurrences he or she experiences. It helps the child to avoid the awareness that nothing can be done to insure the continued parental support. On the other hand, this overemphasis on taking per- sonal responsibility for others' positive response or withdrawal can lead to: a lasting sense of inadequacy; a grandiose sense of personal "badness", import- ance, and unworthiness; a lack of self-esteem; a need to continually compen- sate for a deficiency of the whole self; a maintenance of the salience of others and the need to please them; a consequent resentment and mistrust for others; an inability to express resentment because others are good and the self, bad; a directing of anger at the self for causing rejection, which second- arily keeps the organism mobilized for compensatory activities; a fear of giv- ing up responsibility for rejection by others, because it would mean giving up 366 one's only sense of control, prompt an awareness of helplessness, and release pent-up rage. Several of the differences between the impact of shame and guilt are explainable by age and developmental changes. The child becomes increasingly self-sufficient with age. Issues of physical and emotional survival result- ing from.the temporary loss of parental support (i.e., emotional separation) are less life threatening in a real sense, and the child's awareness of his or her independence fosters this further. The child also has more awareness of the fact that, except in unusual cases, the loss of parental support is temp- orary. The child survived these losses in the past without the actual occur- rence of the feared total abandonment. The child also has a more extensive and broader history of mastery that acts as a bolster against esteem-deflating failures, or losses of support. Most children at this stage learn ways to establish and maintain social contact with others, and learn how to reestab- lish contact with parents after wrongdoing. By parentally or societally taught methods of apology or restitution, the child is able to actively reestablish the positive relationship with the parent or other. As the child learns more of what is expected of him or her, there is an internalization of the values, mores, and norms, and the internal modification of esteem based on the extent to which he or she approximates or deviates from.what is expected. These age and developmental changes assume that the child has arrived at the Initiative versus Guilt stage having relatively positive resolutions of the prior stages. In the case of children who have experienced severe shaming, this developmental flow may be disturbed, and the salience of abandonment, failure- and shame-related issues may remain heightened. In this sense, shame can become a prevalent emotion, powered by early and repetitive experiences of failure, or parental shaming or rejection, which prompts ongoing insecurity, 365 and leads to an overall sense of inadequacy. Tomkins (1963) suggests that the domination of the personality by shame leads to the development of a personal shame theory, that is, a particular way of experiencing and living in the world. All percepts, feelings, behaviors, and attitudes can theoretically be filtered through the personal shame theory. The theory can prompt the interpretation of reality in a way that reflects the early history of shame. In addition, it may retrigger shame, esteem loss, and insecurity, and thereby maintain the validity of the shame theory. Tomkins (1963) distinguishes between shame and self—contempt. He suggests that shame is a primary affective motivator, and that self-contempt is a derivative of the shame affect. He characterizes self-contempt as the capacity of the self to dissect and assess the worthiness of aspects of the self, as manifested in behaviors, attitudes, and feelings. In instances where an aspect of the self either does not meet the standards of the individual or of the surrounding environment, the individual attempts to excise that aspect from the self, or to separate it out and feel contempt for that aspect of the self. This contempt is experienced as lowered self-esteem, but it is tied to the offending aspect of self rather than the whole self. Tomkins's description of self-contempt parallels what we have here called guilt. Developmentally, this formulation acknowledges the increasing differentiation of the self from the environment and the concurrently increasing cognitive differentiation of the individual. Within this more sophisticated cognitive capacity, self-contempt represents the ability to separate parts from the whole, doing from being. In wrongdoing, the modification of self-esteem is based on the severity of the transgression, rather than the complete loss of self-esteem, as in shame. How— ever, as in shame, the excessive experience of guilt, can lead to a basic style of perceiving and interpreting the world and one's place in it. Tomkins, using 366 his term for guilt (i.e., self contempt), suggests that a personal theory revolving around acute appraisals of aspects of the self can be established, focusing on the assessment of the rightness or wrongness of one's own or others' behaviors, attitudes, and feelings. Odier (1956) presents a theory that allows for the understanding of the continuity of such unadaptive, pervasive affective styles or patterns into adulthood. His viewpoint seems particularly salient in regard to shame and guilt as we have described them. Odier's theory integrates Piaget's theory of the development of mental activity with Freud's psychoanalytic theory. Essentially, his premise is that neurosis arises from the maintenance into adulthood of thinking patterns, moral dictates, and emotions, appropriate to the young child. What is a natural point in the intellectual, moral, or feeling development of the child can come to be an area of regressed intellectual, moral, or feeling activity in the adult. He points out that Piaget (1968) focused on differentiating between intellectual and moral development and explaining their interaction. Odier elaborates on the area of affective development as it relates to both. In this regard, he points out that the superego is always infantile in character. Through development, part or most of these early moral dictates evolve and mature into an adult moral conscience. Those dictates that persist unchanged do so because they remain tied to painful, repressed, emotional childhood experiences. It is these immature, regressed areas that seem to be triggered in the adult with a potency that is comparable to that originally experienced by the child. The present writer views this conceptualization as important in the understanding of why shame, and to a lesser extent, guilt, can be experi- enced to the extent they can in the adult, and how and why the individual may experience themselves as regressed and young in the process. It seems to be an ' 367 almost literal reexperience of infantile mental and feeling activities. The experience of shame in adulthood as a primary style implicates the maintenance of infantile intellectual, moral, and affective patterns. In this discussion the present writer is not suggesting that guilt is a minor effect. As Tomkins (1963) points out, self-contempt (or guilt), though not a primary effect, can be developed into a self-contempt or guilt theory and provide a structured orientation for perceiving the environment and the self. What is being stressed are the differences between the more self-contained para- meters of guilt and the more pervasive implications of shame, on cognitions, self-view, and development in general. Both are specific affects that often seem powered by anxiety and can function to direct behaviors, attitudes, and feelings, and ultimately to affect self-view and esteem. Each can take on an excessive emphasis and redirect development. In the present writer's view, it is only with this attention to the overall development of the individual that shame and guilt can be fully understood. BIBLIOGRAPHY 368 BIBLIOGRAPHY Aberle, D. F., & Naegle, K. 0. Middle class father's occupational role and attitudes toward children. American Journal of Orthqpsychiatry, 1952, 22, 366-378. Alexander, F. Remarks about the relation of inferiority feelings to guilt feelings. International Journal of Psychoanalysis, 1938, 12, 61. Allport, G. The ego in contemporary psychology. Psycholggical Review, 1963, 59, 651-678. Allen, V. L., & Crutchfield, R. S. Generalization of experimentally reinforced conformity. Journal of Abnormal and Social Psyghology, 1963, 61, 326-333. Bakan, D. Unmitigated agency and Freud's "death instinct." In Bakan, D. (Ed.), The dualiry of human existence. Chicago: Rand McNally and Co., 1966. Barclay, A., & Cusumano, D. R. Father absence, cross—sex identiry and field dependent behavior in male adolescents. Paper presented at the Annual Meeting of the American Psychologi- cal Association, 1965. Bassos, C. 0n shame. Unpublished manuscript, 1973. Bechtold, G. P. Factorial study of perceptual speed. Unpublished Ph.D. dissertation, Department of Psychology, University of Chicago, 1967. Bem, S. The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 1976, 62, 155-162. Ben, S. Sex role adaptability: One consequence of psychological androgyny. Journal of Personaliry and Social Psycholggy, 1975, 31, 636-663. Benfari, R. Defense and control: Further indications. Perceptual and Motor Skills, 1966,‘gg, 736. Berkowitz, L. & Connor, W. Success, failure and social responsibil- ity. Journal of Personality and Social Psychology, 1966, 6, 666-669. 369 370 Berry, J. W. A study Of Temne and Eskimo visual perception. Pre- liminary report. Paper read before the Edinburgh University Psychology Society, 1966. Bertini, M. The defense of isolation in the determination of dynam- ics and structure. Contributi del' Istituto di Piscologia, 1961, p. 25. Bieri, J. Parental identification, acceptance of authority and within-sex differences in cognitive behavior. Journal of Abnormal and Social Psycholggy, 1960, 69, 76-79. Binder, J. The relative proneness to shame and guilt as a dimension of character style. Dissertation, University of Michigan, 1971. Blank, H. R. Dreams of the blind. Psychoanalytic Quarterly, 1958, p. 27. Block, J. H. Conceptions of sex role: Some cross-cultural and longitudinal perspectives. American Psygholqgist, 1973, 28, 512-526. Bowlby, J. Psychopathology of anxiety: The role of affectional bonds. In Lader, M. (Ed.), Studies of Anxiety. Headley Brothers Ltd.: Asford, Kent, 1970. Botzum, W. A. A factorial study of the reasoning and closure fac- tors. Unpublished Ph.D. dissertation, Department of Psychology, University of Chicago, 1950. Breur, J., & Freud, S. Studies on hysteria. Standard Edition, 1893-1895, 1955, 2, 3-313. Bush, M. A study of reailty-close--reality-distance. Unpublished Doctoral dissertation, University of Michigan, 1968. Byrne, D. The repression-sensitization scale: Rationale, relia- bility, validity. Journal of Personaligy, 196l,_22, 336-369. Byrne, D. Repression-sensitization as a dimension of personality. In Maher, B. A. (Ed.), Progress in experimental personality research. New York: Academic Press, 1966, p. 169. Chein, I. The awareness of the self and the structure of the ego. Psychological Review, 1966, 5;, 306-316. Constantinople, A. An Eriksonian measure of personality development in college students. Develgpmental Psycholggy, 1969, l, 357-372. . 371 Dawson, J. L. M. Psychological effects of social change in a West African community. Unpublished doctor's dissertation, University of Oxford, 1963. Dawson, J. L. M. Cultural and physiological influences upon spatial- perceptual processes in West Africa. Parts I and II. Inter- national Journal of Psycholpgy, 1967, 2: 115-128; 171-185. DeGroot, . Emotional climate of an experimental situation, inter- action patterns, and field style of subjects. Unpublished Dissertation, University of Cincinnati, 1968. Donelson, E. Personality: A scientific approach. New York: Appleton-Century-Crofts/Goodyear, 1973. Donelson, E. Social responsiveness and sense of separateness. In Donelson, E., & Gullahorn, J. (Eds.) Women: AApsychological perspective. New York: Wiley, 1977. Donelson, E. Development of sex-typed behavior and self concept. In Donelson, E. & Gullahorn, J. WOmen: A_psyehological perspective. New York: Wiley, 1977. Dulany, D. E. Jr. Nonverbal communication. Psychological Bulletin, 1969, 13, 118-137. Dulaney, D. E., Jr. Avoidance learning of perceptual defense and vigilence. Journal of Abnormal and Social Psyeholggy, 1957, 55, 333-338. Edwards, A. L. Edwards' personal preference schedule. New York: Psychological Corporation, 1953. Eriksen, C. Defense mechanisms. In Borgatta, E., & Lambert, W. W. (Eds.), Handbook of personality theory and research. Chicago: Rand McNally & Company, 1968. Erikson, E. Childhood and society. New York: W. W. Norton, 1950. Feldman, S. S. Blushing, fear of blushing, and shame. Journal of the American Psychoanalytic Association, 1962, p. 10. Fenichel, O. The scoptophilic instinct and identification. Inter- national Journal of Psychoanalysis, 1937, 55, 6-36. Fenichel, O. The psychoanalytic theory of neurosis. New York: W. W. Norton, 1965. Freud, A. The ego and the mechanisms of defense. New York: Inter- national University Press, 1936. 372 Freud, S. 1896 Further remarks on the neuropsychoses of defense. Standard Edition, 1962, 5, 159-176. Freud, S. Three essays on sexuality (1905). Standard Edition, 1953, .1, 135-263. Freud, S. 1908 "Civilized" sexuality and modern nervous illness. Standard Edition, 1959, 2, 181-206. Freud, S. 1916 On narcissism, an introduction. Standard Edition, 1957, 56, 73-106. Freud, S. The economic problem of masochism. Standard Edition, 1926, p. 19. Freud, S. Civilization and its discontents. Standard Edition, 1930, p. 21. Gardner, R. W. Cognitive controls in adaptation: Research and mea- surement. In Messick, S., & Ross, J. (Eds.), Measurement in personality and cognition. New York: Wiley, 1962, p. 183. Gleser, Gottscalk, & Springer. An anxiety scale applicable to verbal samples. Archives of General Psychiatry, 1961, 5, 593-605. Goodenough, D. R., & Karp, S. A. Field dependence and intellectual functioning. Journal of Abnormal and Social Psychology, 1961, 55, 261-266. Cough, H. G. California personality inventory. Palo Alto: Con- sulting Psychological Press, 1957. Gottschaldt, K. Uber den einfluss der erfabrung auf die wahrnehmung von figuren, II (mit 22 Abbildungen in Text). Psychologische Forschung, 1929, 1g, 1-88. Gottscalk, & Glasser. The measure of psychological states through_ content analysis of verbal behavior. Berkley: University of California Press, 1969. Grinker, R. R. Growth, inertia, shame: Their therapeutic implica— tions and danger. International Journal of Psychoanalysis, 1955, p. 36. Guntrip, H. Schizoid phenomena, object relations, and the self. New York: International University Press, 1969. Gutmann, D. Women and the conception of ego strength. Merrill Palmer Quarterly, 1965, 11, 3, 229-260. 373 Gutmann, D. Female ego styles and generational conflict. Paper presented at the Midwestern Psychological Association, 3 May, 1968, Chicago. Hartmann, H. Cements on the psychoanalytic theory of the ego. The psychoanalytic study of the child. New York: International Universities Press, 1950, 5, 76-96. Heist, P. & Yonge, G. Omnibus Personality Inventory (Form F). New York: Psychological Corporation, 1962. Hoffman, M. L. Altruistic behavior and the parent-child relation- ship. Journal of Personality and Social Psychology, 1975, .55, 937-963. Isen, A. M. Success, failure, attention and reaction to others: The warm glow of success. Journal of Personality and Social Psychology, 1970, i2» 296-301 Isen, A. M. & Levin, P. The effect of feeling good on helping: Cookies and kindness. Journal of Personality and Social Psyehology, 1972, 21, 386-388. Jacobsen, E. The self and the object world. The Psychoanalytic Study of the Child, 1956, 2, 75-127. James, W. Principles of psychology, Vol. 1, New York: Henry Holt & Co., 1890. Jourard, S. The transparent self. Princeton, New Jersey: VanNostrand, 1966. Jung, C. G. The collected works of C. G. Jung. New York: Pantheon Books, 1963, 56, Bollingen Series XX. Kagan, J. Acquisition and significance of sex-typing and sex-role identity. In M. L. Hoffman & L. W. Hoffman (Eds.), Review of Child Development Research. Vol. 1, New York: Russell Sage Foundation, 1966. Kagan, J., & Moss, H. Birth to maturity: A study of_psychologica1 development. New York: John Wiley & Sons, 1962. Kaufman, G. The meaning of shame: Toward a self affirming identity. Journal of Counseling Psychology, 1976. Kell, B. Personal communication, 1973. Kelly, J., & Worell, J. New formulations of sex roles and androgyny: a critical review. Journal of Clinical and Consulting_ Psychology, 1977, 65, 6, 1101-1115. 376 Klein, G. S. Need and regulation. In Jones, M. R. (Ed.), Nebraska symposium on motivation. Lincoln: University of Nebraska Press, 1956. Kohlberg, L. A cognitive-developmental analysis of children's sex- role concepts and attitudes. In E. E. Maccoby (Ed.), The development of sex differences. Stanford, Calif.: Stanford University Press, 1966. Kamorovsky, M. Cultural contradictions and sex roles: The mascu- line case. American Journal of Sociology, 1973, 58, 876. Laing, R. D. The divided self. Chicago: Quadrangle Books, 1960. Levin, S. Some metapsychological considerations of the differentia- tion between shame and guilt. International Journal of Psychoanalysis, 1967, 65, 267-276. Lewis, H. B. Over-differentiation and under-individuation of the self. Psychoanalysis and the Psychoanalytic Review, 1958, Q, 3-24 s Lewis, H. B, Witkin, H., & Ideil, E. Affective reactions and patient-therapist interactions among more differentiated and less differentiated patients early in therapy. Journal of Nervous and Mental Disorders, 1968, r39, 193-208. Lewis, H. B. Shame and guilt in neurosis. New York: International University Press, 1971. Linton, H. Dependence on external influence: Correlates in per- ception, attitudes and judgments. Journal of Abnormal and Social Psychology, 1955, 59, 502-507. Lowen, A. Depression and the body. New York: Penguin Books, 1972. Luborsky, L., Blinder, B., & Schimek, J. Looking, recalling, and GSR as a function of defense. Journal of Abnormal Psycho- logy, 1965, 19, 270. Ludwig, D. J. Self percept and the DAP test. Journal of Projective Techniques and Personality Assessment, 1969, 55(3), 257-261. Lynd, H. M. On shame and the search for identity. New York: Harcourt, Brace and World, 1958. Maccoby, E. E. The development of sex differences. Stanford, Calif.: Stanford University Press, 1966. Maccoby, E., & Jacklin, C. N. The psychology of sex differences. Stanford, Calif.: Stanford University Press, 1976. 375 Mandler, G., & Sarason, S. B. A study of anxiety and learning. Journal of Abnormal and Social Psychology, 1952, 61, 166-173. Markowitz, A. Influence of the repression-sensitization dimension, affect value and ego threat on incidental learning. Journal of Personality and Social Psychology, 1969, ll, 376-380. May, R. The meaning,of anxiety. New York: Ronald Press, 1950. Mayman, M. A study of sense of self. Unpublished Doctoral Disser- tation, University of Kansas, 1953. Mayman, M., & Voth, H. Reality-closeness, phantasy and autokinesis: A dimension of cognitive style. Unpublished manuscript, 1967. Menaker, E. Masochism: A defense reaction of the ego. Psycho- analytic Quarterly, 1953, gt, 205-220. Mendelson, G. A., & Griswold, B. B. Differential use of incidental stimuli in problem solving as a function of creativity. Journal of Abnormal and Social Psychology, 1966, 55, 631-636. Mendelson, C. A., & Griswold, B. B. Anxiety and repression as predictors of the use of incidental cues in problem solving. Journal of Personality and Social Psychology, 1966, 6, 623-631. Mendelson, C. A., & Griswold, B. B. Assessed creative potential, vocabulary level, and sex as predictors of the use of inci- dental cues in verbal problem solving. Journal of Personal- ity and Social Psychology, 1967, 5, 353-359. Milton, G. A. The effects of sex-role identification upon problem solving skill. Journal of Abnormal and Social Psychology, 1957, 55, 208-212. Minard, J., & Mooney, W. Psychological differentiation and percep- tual defense: Studies of the separation of perception from emotion. Journal of Abnormal Psychology, 1969, 16, 131-160. Mischel, W., Coates, B., & Raskoff, A. Effects of success and failure on self-gratification. Journal of Personality and Social Psychology, 1968, 59, 381-390. Mischel, W., Ebbesen, E. E., & Zeiss, A. R. Selective attention to the self: Situational and dispositional determinants. Journal of Personality and Social Psychology, 1973. Montague, A. Touching: The human slgnificance of the skin. New York: Harper and Row, 1972. 376 Negri, T. The assessment of shame and guilt through the use of the TAT. M.A. Thesis, Michigan State University, 1976. Odier, C. Anxiety and magic thinking. New York: International University Press, 1956. Pemberton, C. L. A study of the speed and flexibility of closure factors. Unpublished Ph.D. dissertation, Department of Psychology, University of Chicago, 1951. Perlman, M. An investigation of anxiety as related to guilt and shame. Archives of Neurology and Psychiatry, 1958, 59, 752- 759. Phares, J. E., Ritchie, M. B., & Davis, W. L. Internal-external control and reactiontxathreat. Journal of Personality and Social Psychology, 1968,_lg(6), 602-605. Piaget, J. The moral judgment of a child. Glencoe, Illinois: Free Press, 1968. (First published in French, 1932.) Piers, S., 6 Singer, M. Shame and guilt: A psychoanalytic and cultural study. In R. Grinker (Ed.), American lecture series, publication No. 171. Springfield, Illinois: Charles C. Thomas, Inc., 1953. Pleck, J., & Sawyer, J. Men and masculinity. New Jersey: Prentice Hall, 1976. Rees, H. J., & Israel, H. E. An investigation of the establishment and operation of mental sets. Psychology Monographs, 1935, 65, (6, Whole No. 210). Rogers, C. Actualizing tendency in relation to "motives" and to consciousness. In Nebraska symposium on motivation. Lincoln: University of Nebraska Press, 1963. Rudin, S., & Stagner, R. Figure ground phenomena in perception of social stimuli. Journal of Psychology, 1958, 65, 213-225. Sanguliano, I. An investigetion of the relationship between per- ception of the upright in space and several factors in personality organization. Unpublished Doctoral disserta- tion, Fordham University. Sarason, I. C. Experimental approaches to test anxiety: Attention and the uses of information. In Spielberger, C. D. (Ed.), Anxiety: Current trends in theory and research. New York: Academic Press, 1972, Vol. 2. 377 Schonbar, R. Differential dream recall as a component of "life style." Journal of Consulting Psychology, 1965, 22, 668- 676. Sherif, M. The psychology of social norms. New York: Harper and Row, 1966. Silverman, J. Attention styles and sex differences. In Attention: Contemporary theory_and analysis. New York: Appleton- Century-Crofts, 1970. Spence, J. T., Helmreich, R., & Stapp, J. Ratings of self and peers on sex role attributes and their relation to self-esteem and conceptions of masculinity and femininity. Journal of Personality and Social Psychology, 1975, 5;, 29-39. Seligman, M. Helplessness: 0n depression, development and death. New York: Scribner, 1975. Thurstone, L. L. An analysis of mechanical aptitude. (Report of the Psychometric Laboratory, No. 62) Chicago: The Psychometric Laboratory, University of Chicago, Unpublished, January, 1951. Thurstone, L. L. A factorial study of perception. (Psychometric Monographs, No. 6) Chicago: University of Chicago Press, 1966. Thurstone, L. L. Mechanical aptitude 111: Analysis of gronp tests. (Report of the Psychometric Laboratory) Chicago: The Psychometric Laboratory, University of Chicago, 1969. Thurstone, L. L., & Jeffrey, T. E. Closure flexibility (concealed figures) test. North Carolina: The Psychometric Laboratory, The University of North Carolina, 1965. Tomkins, S. Affect, imsgery and consciousness (Vol. 2). New York: Springer, 1963. Valenstein, A. F. Glossary of defenses. Psychoanalytic Study of the Child, 1961, p. 16. Weigel, D. Unpublished Doctoral Dissertation, Michigan State University, 1976. Weiner, H. et a1. Etiology of duodenal ulcers: 1. Relation of specific psychological characteristics to rate of gastric secretion (serum pepsinogen). Psychosomatic Medicine, 1957, lg, 1-10. Wertheimer, M. Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Witkin, Yela, M. 378 In W. D. Ellis (Ed.), Sourcebook New York: Harcourt, Brace, & Gestalt theory. for Gestalt psychology. World, 1938. H. A. & Asch, S. E. Studies in space orientation, III and IV. Journal Experimental Psychology, l68,_5§ 603- 616, 762-782. H. A. Perception of body position and of the position of the visual field. Psychological Monograph, No. 302, 1969, 55(7), 1-66. H. A. Individual differences in the ease of perception of embedded figures. Journal of Personality, 1950, l2, 1-15. H. A., Lewis, H. B., Hertzman, M., Machover, K., Meissner, P., & Wapner, S. Personality throogh_perception. New York: Harper & Brothers, 1956. H. A., Dyk, R. B., Faterson, H. F., Goodenough, D. R., & Karp, S. Psychological differentiation. New York: John Wiley & Sons, 1962. H. A. Psychological differentiation and forms of pathology. Journal of Abnormal Psychology, 1965, 19, 317-336. H. A., & Lewis, H. B. The relation of experimentally induced pre-sleep experience to dreams. Journal of American Psychoanalytic Association, 1965,_l5, 819-869. H. A., Goodenough, D. R., & Karp, S. A. Stability of cognitive style from childhood to young adulthood. Journal of Personality and Social Psychology, 1967. H. A., Birnbaum, J., Lomonoco, S., Lehr, S., & Herman, J. Cognitive patterning in congenitally totally blind children. Child Development, 1968,_52, 767-786. H. A., Lewis, H. B., & Weil, E. Affective reactions and patient-therapist interaction among more differentiated and less differentiated patients early in therapy. Journal of Nervous and Mental Disorders, 1968,_l55, 193- 208. H. A. Social influences in the development of cognitive style. In Goslin, D. A. (Ed.), Handbook of socialization theory and research. Chicago: Rand McNally, 1969. Application of the concept of simple structure to Alexander's data. Psychometrika, 1969, 55(2), 121-135. "7'71@1'11‘111‘1111111