, .9. . ;. .2 5 “3213-2 “32,; { 2w: .W (iffy “if "H '; m: l 1" #3,?ng ~:'>1“;:"tv»531 , ‘ \ MICHIGAN SSSSSSSSSSSSSSSSSSSS THgSIS 1293 01417 3045 V» ’«n' \ "’i/' This is to certify that the dissertation entitled SHAME, GUILT, EMPATHY AND MORAL REASONING: GENDER DIFFERENCES presented by DAVID LENSKY FINKE has been accepted towards fulfillment of the requirements for PhD degree in Psychology «m M Major professor Date December 12, 1994 LIBRARY Michigan State University PLACE II RETURN BOXto removothb chockomtrom your noord. To AVOID FINES mum on or bdoro date duo. A_____—-——— SHAME, GUILE, EMPATHY AND MORAL REASONING: GENDER DIFFERENCES BY DAVID LENSKY FINKE A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1994 ABSTRACT SHAME, GUILT, EMPATHY AND MORAL REASONING: GENDER DIFFERENCES BY DAVID LENSKY FINKE Theories of moral development often emphasize either cognitive or affective aspects of empathy, while neglecting that empathy is a complex construct which has both cognitive and affective correlates. This study proposed a relationship between affective traits and moral reasoning that was mediated by empathy. 184 female and 125 male undergraduates from a large midwestern university were administered measures of gender identity, shame-proneness, guilt-proneness, affective and cognitive correlates of empathy, and moral reasoning. Links were proposed between affective traits and empathy. It was predicted that shame-proneness and guilt-proneness would mediate the relationship between empathy and mature moral reasoning. Gender differences moderated the relationship among several of the variables. For women, shame-proneness was predictive of personal distress while guilt-proneness was predictive of empathic concern, perspective taking, and fantasy. No interaction between shame-proneness and guilt- proneness was found. The relationship between guilt- proneness and mature moral reasoning was mediated by perspective taking. In addition, the associations between guilt-proneness and moral reasoning increased with higher stages of moral reasoning. For men, shame-proneness was predictive of personal distress while guilt-proneness was predictive of each of the correlates of empathy. No interaction between shame- proneness and guilt-proneness was found. The correlates of empathy did not mediate the relationship between affect and moral reasoning. Guilt-proneness was predictive of mature moral reasoning. In addition, the association between perspective taking and moral reasoning increased with higher stages of moral reasoning. Modest differences between shame and guilt were also found, supporting the theory that they are distinctly different constructs. Suggestions for future research are provided. Mom, Helene, and Winnebago iv ACKNOWLEDGMENTS I would like to take this opportunity to thank my chair Dr. Norman Abeles, not only for his assistance on my dissertation, but his guidance and patience throughout my graduate training. I would also like to thank my committee members, John Hurley, Mike Lindell, and Ellen Strommen for their assistance and feedback on this project. And, once again, Mike, Marianne, and Bruce, THANKS. TABLE OF CONTENTS List Of Tables 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O ....... O 0 ix IntrOduction. O O .0 O O O O O O O O O I O I O O O O O O O O O O O O O O O O O O O O I O O O ....... 1 Piaget’s Contributions to Moral Development.. ......... 4 Kohlberg’s Theory of Moral Development................8 Empirical Studies of Kohlberg’s Theory...............14 Carol Gilligan - An Alternative Cognitive Developmental Theory................ ........... .19 Kohlberg on Affect and Moral Development.... ......... 23 Guilt, Shame and Moral Reasoning........... ......... .26 Shame VS. GuiltOOOOOOOOOOCOOOOOOOOOOOOOOOOOOOOO0.0.0.31 The Differentiation of Shame and Guilt as Applied to Empathy and Moral Reasoning..................36 Empathy and Moral Development........................39 Hypotheses................................................46 Method....................................................51 Participants.........................................51 Measures.............................................51 Defining Issues Test............................51 Interpersonal Reactivity Index..................52 Bem Sex Role Inventory.................. ..... ...53 Personal Feelings Questionnaire.................54 Self-Conscious Affect and Attribution Inventory (SCAAI)OOOOOOOOOOOOOOOOOOO00.0.0056 vi Procedures....................... ........ ... ......... 57 Results................................. .................. 58 Discussion..................................... ........... 69 Gender Differences in Proneness to Shame and Guilt...69 Shame-proneness................... ............. 69 Guilt-proneness.................................70 Shame-proneness and guilt-proneness.............70 The Relationship Between Shame-Proneness and the Correlates of Empathy..................... ...... 71 Shame-proneness and empathy for women...........71 Shame-proneness and empathy for men........ ..... 72 Comparisons between men and women...............73 The Relationship Between Guilt-Proneness and the Correlates of Empathy...........................75 The Differentiation Between Shame and Guilt..........77 Shame, Guilt, Empathy, and Moral Reasoning...........77 Empathy and levels of moral reasoning...........79 Shame-proneness and levels of moral reasoning...82 Guilt-proneness and levels of moral reasoning...83 Summary of Findings...................... ......... ...84 Theoretical, Developmental, and Clinical Implications.......OOOOOO0.00.0000...0.0.0.0....86 Limitations of this Study............................88 Future ResearChOOOOOOO......OOOOOOOOOOOO00......00.0.89 Appendix A: List of Cronbach Alpha Values by Scale.....92 Appendix B: Means and Standard Deviations for all Heasures...‘.0.........OOOOOOOOOOOO0.000.0.0.0..93 Appendix C: Correlation Matrix of all Measures for the Total Population............................94 vii Appendix D: Correlation Matrices of all Measures for Women and Men........................ ....... 95 Appendix E: Appendix F: Appendix G: Demographic Data....... ..... .... ......... ..96 Ben Sex-Role Inventory.....................97 Self-Conscious Affect and Attribution Inventory..................... ........ 98 Appendix H: Personal Feelings Questionnaire - 2 ....... 105 Appendix I: Interpersonal Reactivity Index...... ..... .106 List of References............................ ........... 108 Table Table Table Table Table Table Table Table LIST OF TABLES Means, Standard Deviations and t-tests of Shame-proneness and guilt-proneness for women and meDOOOOOOOOOOOO......OOOOOOOOO ........ 60 Multiple regression analyses of shame- proneness and guilt-proneness with masculinity, femininity, and gender. ............ 60 Pearson correlation coefficients of shame- proneness and empathic concern, fantasy, personal distress, and perspective taking for women and men.................... ........... 62 Multiple regression analyses of empathic concern with shame-proneness, high shame- proneness, and gender identity, by gender. ...... 63 Multiple regression analyses of the affective correlates of empathy with shame-proneness, guilt-proneness, and the interaction of shame and guilt............................ ......... ..64 Pearson correlation coefficients of guilt- proneness and empathic concern, fantasy, personal distress, and perspective taking for women and men.......................... ..... 64 Multiple regression analyses of mature moral reasoning with correlates of empathy, shame- proneness, guilt-proneness, and the shame— guilt interaction...............................65 Computed trvalues contrasting the correlation coefficients between stages of moral reasoning and correlates of empathy, stages of moral reasoning and shame-proneness, and stages of moral reasoning and guilt-proneness for women...67 ix Table 9. Computed t-values contrasting the correlation coefficients between stages of moral reasoning and correlates of empathy, stages of moral reasoning and shame-proneness, and stages of moral reasoning and guilt-proneness for men.....68 INTRODUCTION Rest (1986) describes the function of morality as: "To provide basic guidelines for determining how conflicts in human interests are to be settled and for optimizing mutual benefits of people living together" (p. 1). Studies of moral reasoning are designed to investigate how and to what extent an individual weighs the different sides of a dilemma. Research in moral reasoning is not so much concerned with what behaviors are right or wrong as much as gaining an understanding of how the guidelines for those behaviors are chosen. Research on moral reasoning became significant with the work of Jean Piaget (1965/1932). Piaget was the first researcher to disregard the societal value of a given behavior and concentrate, instead, upon the rationale that motivated a given behavior. In addition, he supplied the beginnings of the first cohesive theory on moral development: that is, the first theory on how an individual's moral reasoning changes over time. He argued that development occurred as the individual progressed from egocentrism and unilateral respect to cooperation and mutual respect. 2 Lawrence Kohlberg (1964) is generally credited with providing the first comprehensive cognitive theory of moral development. This theory, utilizing some of Piaget's principles, posits that an individual progresses through six stages of development during which the individual's ability to understand the cognitive complexity of a given dilemma increases. One is able to progress through the stages as intellectual capacity develops and as one takes on different roles, which enable the person to understand the different perspectives of competing needs. Throughout each stage, the individual makes moral judgments to resolve the competing needs based upon attempts to provide justice and fairness for all involved. A major critic of Kohlberg's approach to moral development is Carol Gilligan (1982). Gilligan argued that women prefer to base their moral reasoning on an ethic of care for others, rather than Kohlberg's ethic of justice. The consistent theme running through each of these theories is the importance of the ability for the mature moral reasoner to cognitively understand other people’s situations. Being able to place one's self in another's perspective is often the definition of empathy. In fact, some studies have found a significant relationship between empathy and mature moral reasoning (Bowman & Reeves, 1987). While empathy involves both a cognitive and affective component (Brems, 1989), Piaget, Kohlberg, and Gilligan do 3 not address the impact of an affective component. In fact, studies in which both affect and moral reasoning are assessed frequently neglect to examine the potential relationships between the two concepts (e.g., Richards, Smith & Davis, 1989) even though when they are examined a relationship appears to exist. For example, LeCapitaine (1987) conducted a study in which third and fourth graders were randomly assigned to three groups of ten. Following a pretest to indicate levels of moral and emotional development, these students participated in one of three workshops. Group one focused on increasing the students' emotional development by focusing on feeling and the content of feelings. Group two focused on moral development by discussing potential moral dilemmas and possible conflict resolution strategies. Group three stressed both emotional content and conflict resolution. Following the intervention, all three groups demonstrated a significant increase in emotional development scores on the Dupont Affective Development Test (Dupont, 1976). In addition, the group whose intervention focused on both cognitive and affective components exhibited a significant increase in moral development, while the groups that focused primarily on either affective or cognitive processes did not. Such results suggest an interaction between cognition and affect in moral reasoning. 4 Further, theoretical and empirical literature provides support for a relationship among shame, guilt, and empathy (Tangney, 1991). This study examines how guilt, shame, empathy, and moral reasoning relate to one another. PIAGET’S CONTRIBUTIONS TO MORAL DEVELOPMENT Any discussion of moral development must begin with the works of Jean Piaget, as his research revolutionized the field. Prior to Piaget's seminal work on moral development (WWW. 1932). the emphasis on cognitive morality research was on using paper and pencil measures in an attempt to predict whether an individual would violate a social norm (e.g., cheating on a test). Behaviors were viewed as either "right" or ”wrong", "good" or "bad", and researchers' only concern was predicting when such behaviors would occur. Piaget (1965/1932), however, believed that the research emphasis should not be on the external behaviors, but on what preceded those behaviors. He believed that the rules for children's games were set-up to guarantee fairness and foster respect. To understand the moral development of children, he thought it was far more significant to understand what an individual believed about certain rules and how the children perceived those rules. As a result, Piaget interviewed children and used their verbalizations to 5 determine what and how they thought about specific values. From the results of these interviews, Piaget developed a cognitive approach to understanding human moral reasoning. During his interviews, Piaget noticed that the responses of most children fit one of three categories: (a) those that exhibited a unilateral respect for authority, viewed rules as constraining behavior, and exhibited egocentric concerns: (b) those who exhibited mutual respect for others and viewed rules as rational and enhancing cooperation: (c) and those who exhibited a mixture of the two previous categories. The first set of values was found in children under the age of 8 and was referred to as "heteronomous morality." Piaget argued that these children live in a world of egocentric thought, in which their primary concern is their own well-being. As a result, they are extremely dependent upon adults, especially parents, and when an adult gives the child a rule to follow, the child tends to follow it without questioning its meanings or purposes. Failure to follow the rules is met by retributive justice, or punishment. Piaget found that around the age of 11 or 12, children began operating and organizing rules based on mutual respect, rather than unilateral respect. Piaget refers to this stage as "autonomous morality". Responses which reflected autonomous morality indicated knowledge of the principles of social exchange and a willingness to engage in 6 relationships of mutual respect. Conflicts are resolved by distributive justice, insuring that fairness and equality are available for all involved. Between these two stages are children between the ages of 8 and 11 years of age. Such children gradually shift from heteronomous morality to autonomous morality. This process occurs through the internalization and generalization of the rules set down by the parents. That is, first the child accepts the rules as defined by the parents "en bloc". As the child begins to accept the rules as his or her own, the child generalizes from the rule and applies it to similar situations. As a result, the rule develops into a method by which the child approaches problems. In Piaget's (1965/1932) formulation, children were believed to progress from heteronomous moral reasoning to autonomous moral reasoning as they developed and matured. The transition point was hypothesized to occur around adolescence. The main difference between heteronomous morality and autonomous morality was not behavior, as two individuals could abide by the same law for drastically different reasons, but on the cognitive processes which went into making their decision on how to behave. Piaget believed that children's cognitive capabilities develop over time. Therefore, two children of different ages may perceive the same thing very differently. For example, a 7 young boy may share with his friend because his parent has told him to, whereas an adolescent may share because he realizes that if he shares now, the friend will probably return the gesture. Thus, the child’s values and the specific laws of morality change as one's cognitive capabilities developed. While Piaget never presented a cogent, complete theory of moral development (Wright, 1983), his contributions to the field have been of tremendous value. Rest (1979) summarized Piaget's contributions to the field as including: the definition of the theoretical construct of moral judgment: the introduction of the method by which one could study moral judgment: the identification of specific features in children's moral thinking: the conception of moral age-dependent stages: and the provision of empirical data, albeit lacking in validity and reliability, to support his theory. While Piaget does not use the word empathy as a component of autonomous morality, his description of the cooperation process which occurs during this phase of moral development captures the essence of a cognitive definition of empathy. He stated, "henceforward, he will not only discover the boundaries that separate his self from the other person, but will learn to understand the other person and be understood by him" (p. 95). 8 A theorist who expanded Piaget’s theory while using the concept of cognitive empathy is Lawrence Kohlberg (1964, 1969, 1976, 1984). Kohlberg took many of the principles proposed by Piaget and developed what is probably the most widely accepted theory of moral judgment. KOHLBERG'S THEORY OF MORAL DEVELOPMENT Kohlberg (1964) also followed the cognitive developmental approach. He provided children with stories which contained a moral dilemma. He then asked them how they would resolve the dilemma and why they would resolve it in that way. From these responses Kohlberg developed his own model of moral development. Like Piaget, Kohlberg believed that moral development depended on the individual’s cognitive-logical abilities. However, while Piaget's model consisted of two levels of morality, Kohlberg's model depicted six stages (Kohlberg, 1969). Kohlberg believed that an individual progressed from low stages to higher stages. He insisted that this advancement of stages was supported by the fact that each subsequent stage was cognitively more complex than the previous one in that it allows for more "crucial" distinctions as well as a progression of how society can or should be organized on a philosophical basis. The theme 9 running through each of these stages is making decisions based upon justice or fairness. Kohlberg (1976) described four basic moral orientations: following the rules of society (normative); concentration on the consequences of one’s actions to others (utilitarian): maintaining liberty, equality, reciprocity, and contract between persons (justice): and abiding by one's conscience (ideal self). Although an individual may use any or all four principles, Kohlberg (1976) believed that only the principle of justice is the essential core of morality: "...the core of justice is the distribution of rights and duties regulated by concepts of equality and reciprocity...Justice is the normative logic, the equilibration, of social actions and relations...0ne can act morally and question all rules, one may act morally and question the greater good, but one cannot act morally and question the need for justice"(p.40). Thus, only justice insures that conflicts will always be settled in such a way which is most beneficial to all persons involved. To evaluate his belief that justice is a theme constant in moral development, Kohlberg separated responses to interview questions into Type A responses (those which involved normative and utilitarian ideals) and Type B responses (those which involved justice and ideal-self orientations). He found that Type A responses only described a past action or predicted what a future action might be, while Type B responses were prescriptive, indicating that the behaviors were consistent with the 10 thoughts. In addition, he found that as individuals progressed through a moral stage, they first took a normative/utilitarian orientation, then matured to the justice/ideal-self orientation before passing to the next stage of moral development. Within a single stage, however, individuals never passed from a justice/ideal-self orientation to a normative/utilitarian orientation. Kohlberg concluded that, as stages increase, moral rules become more internalized and autonomous. However, Kohlberg based his determination by grouping moral orientations founded upon justice with those founded upon the image of a conscience, without addressing the impact of conscience (which would include guilt) on decision making. This will be more fully developed in a later section. Another essential feature of Kohlberg’s theory is that moral development is facilitated by the process of socialization (1984). The socialization process occurs when one's cognitive capabilities expand allowing him to take on different roles. Through these different roles, one begins to understand his environment at increasingly complex levels. This process, briefly described here, is an interactive one. Faced with an environmental demand, one applies cognitive ability to that environmental demand. The individual's cognitive ability consequently becomes restructured by the external stimuli. It is through this process that the individual develops from one stage to the 11 next. During stage progression, the higher stages reintegrate the structures found at lower stages and may use those structures for simpler conflict resolution. Kohlberg's (1969, 1976) six stages can be further grouped into three levels. Each level consists of two stages, differentiated by the distinct social perspectives which they represent and how the role of justice is interpreted. An outline of these stages, based upon social perspective and interpretations of justice, follows. From a social perspective, Stages 1 and 2 indicate emphasis on the individual's own interests. These two stages are frequently referred to collectively as the preconventional level and are found in most children under the age of nine, some adolescents, and many adolescent and adult criminal offenders. From a justice orientation, Stage 1 consists of a punishment and obedience orientation. In this stage individuals regard rules and punishments as extremely concrete and the presence of these rules and punishments are essential for the individual to perceive a behavior as inappropriate. Behavior is categorically right or wrong. Stage 2 thinking exhibits an awareness that different individuals may have divergent interests of equal importance and value which they wish to pursue, and these interests may conflict. What is morally right is relative to the particular situation. This stage of developing does not 12 provide the individual with a means of resolving the conflict. Moral reasoning in this stage involves the first appearance of perspective taking. Perspective taking is one aspect of empathy, as will be discussed in more detail later. Socially, stages 3 and 4 represent a focus on how the individual sees himself as part of a group, and therefore responsible to that group. These stages are collectively referred to as the conventional level and can be found in most adolescents and adults in all societies. At Stage 3 the most significant change in the individual's justice orientation seems to be the desire to maintain social approval and interpersonal trust. The golden rule of "Do unto others as you would have them do unto you,” is cast into this stage. The individual, still able to recognize competing yet equally fair perspectives, attempts to resolve conflict by applying generally accepted social norms. Stage 4, called "authority maintaining morality", assumes that the individual is able to step away from a situation and utilize the accepted guidelines which are in effect in every culture. This approach is also referred to as the "law and order? stage, as individuals place special emphasis on observing and maintaining the rules so as to maintain society. Such an approach allows the individual to act impartially, thereby insuring that any individual l3 interest which they wish to pursue will only occur if it is within the societal guidelines. Stages 5 and 6 indicate an emphasis on the individual’s ability to recognize himself as a member of society. Maintaining and restructuring the rules on which that society is based is necessary to help further that society. These two stages are collectively referred to as the postconventional level and are achieved by some individuals after age 20. Such individuals recognize that moral obligations can, at times, take precedent over societal rules. Stage 5 is referred to as "morality of contract, of individual rights, and of democratically accepted law." In this stage, each individual is aware that rights and values apply to all individuals. As a result, the validity of any law or privilege can be evaluated in terms of the degree to which it preserves and protects each individual’s fundamental human rights and values. Kohlberg refers to this as a "society-creating" rather than a "society- maintaining" perspective. The primary focus can be on either individual rights or social welfare, as some rights must be considered sacred by society (Thus, individual rights would supersede social welfare.), and some individual rights must be suspended for social welfare. At this level society is conceived of as being based on social cooperation and agreement. 14 Stage 6 requires equal consideration of the interests of each individual who may be affected by the impending decision. Kohlberg referred to this as the morality of individual principles of conscience, and he believed its implementation insured fairness, impartiality and reversibility. Obviously for someone to consider the interests of each individual in a given situation, empathy is necessary. Kohlberg added one more stage in one of his last reformulations (Stage 7). This stage was to reflect meta- ethical and metaphysical thinking, as well as the ability to engage in self-reflection. Kohlberg argued that the development of this stage was necessary to account for what appeared to be continued moral development "beyond the simple duality of the justice theme (Kohlberg, 1983)." Kohlberg never operationalized this stage, however. EMPIRICAL STUDIES OF KOHLBERG’S THEORY Kohlberg's theory has been tested empirically on several occasions. It has received empirical support from cross-cultural (Hau & Lew, 1989: Zeidner & Nevo, 1987: Melkama & Ikonen, 1986) and longitudinal studies (Rest, Davison, & Robbins, 1978). Some cross-cultural studies have shown that children from less industrialized cultures or lower social economic status (SES) tend to score lower on 15 moral development tests (Zeidner & Nevo, 1987; Van Voorhis, 1985). However, Kohlberg's theory accounts for such phenomenon. For example, Van Voorhis (1985) found that moral development, as assessed by the Kohlberg,interview, successfully differentiated Hispanic criminal offenders from non-Hispanic criminal offenders. However, Kohlberg’s theory posits that individuals in less-industrialized cultures or lower SES would have fewer opportunities for a variety of roles. As discussed above, role-taking is crucial to moral development in Kohlberg's theory. Rest (1986) argued that moral development is an attitude involving the interpersonal domain and is, therefore, orthogonal to personality. However, the empirical evidence is generally to the contrary. Although Walker, Rowland, and Boyes (1991) found that mature moral development did not relate significantly to sensation seeking, mature moral development has correlated with several personality characteristics as measured by the California Personality Inventory (CPI: Gough, 1957/1975) and Minnesota Multiphasic Personality Inventory (MMPI: Hathaway & McKinley, 1948/1967). Gibson (1990) found significant correlations between moral development and measures of social poise, interpersonal adequacy, and intellectual efficiency. Lifton (1985), on the other hand, attempted to establish a relationship between mature moral development and mental health. He found significant positive 16 correlations between mature moral development and social introversion, anxiety, and schizophrenia and significant negative correlations between mature moral development and psychological mindedness, sense of well-being, communality, ego-strength, and ego-control among college males. Among college females he found positive correlates with intellectual efficiency, social dominance, self-acceptance, sociability, ego-resiliency, sense of well-being, social responsibility, achievement via conformance, ego strength, and social presence, and negatively with several types of psychopathology. The scoring scheme recommended by Kohlberg (1969) has drawn some criticism. Kohlberg's scoring scheme is based upon an interview format. As a result, it is instrument- free and requires thorough training. In addition, several studies utilizing Kohlberg's moral interview have been unable to identify individuals at Stage 6 of development (e.g., Propper 8 Brown, 1986: see Kohlberg, 1984, for a review). In contrast, Rest (1986/1979) developed the Defining Issues Test (DIT). This instrument measures moral development as formulated by Kohlberg, but avoids the subjectivity and time needed by the Kohlberg interview. The measure is composed of six moral dilemmas, each describing a situation with competing social claims. The moral considerations involved in resolving the dilemma are provided by the test and the individual is asked to rank 17 them in order of importance. It requires little training, is a task of recognition, and provides a principled morality score (P-score: those responses exhibiting post-conventional reasoning) in addition to placing individual responses into specific particular stages. Unlike Kohlberg's moral interview, the DIT has also successfully identified individuals at Stage 6 of moral development (Rest, 1979). The empirical evidence on whether Kohlberg's theory can predict moral behavior is equivocal, and some have questioned the usefulness of the theory (Wright, 1983). In a review of the research comparing moral cognition and moral action, Blasi (1980) determined that moral stage does not always predict moral behavior. Blasi found that moral stage was a better predictor when comparing delinquents to nondelinquents and that higher stage individuals were less likely to change their opinions so as to meet the groups' view than lower stage individuals. However, moral stage was not found to be a predictor of honesty or conformity of behavior, as the theory would predict. Blasi argued that the failure to predict action was due to the absence of any theoretical orientation connecting moral judgment to moral action. In response to this research, Kohlberg and his associates (Kohlberg, Levine, and Hewer, 1983) claimed that moral behavior requires three steps. First, the individual must make a moral judgment about a situation: then, one must 18 claim responsibility for the situation: finally, one must act. In addition, Kohlberg pointed out that stage is a better indicator of moral behavior at the postconventional stages (stages 5 and 6). Nonetheless, Kohlbergfs theory is unable to make accurate predictions of behavior unless other variables are considered. One possible explanation for the lack of relationship between moral reasoning and behavior is that Kohlberg's theory neglects the role of affect, especially shame and guilt, on moral development. While Kohlberg maintained that his theory is consistent with Piaget’s theory, Wright (1983) argued that Piaget's theory contained a strong affective component which Kohlberg neglects. Wright believed the discrepancy lies in Kohlberg's exclusive emphasis on cognition, whereas Piaget attempted to draw connections between moral reasoning and subsequent actions. Conn (1981), in a defense of Kohlberg’s approach, maintained that Kohlberg avoided affects like guilt and shame because of the judgments that come with them. Maintaining a social perspective, Conn argued, requires empathy. Thus, Conn reasoned that empathy is at the core of moral reasoning at the advanced levels. While it appears true that the importance of empathy is implied in Kohlberg's theory, the role-taking and social perspective that Kohlberg describes are predominantly cognitive constructs. 19 Empathy is defined in The Ameziean fierieage Dictionary (1971) as, "Understanding so intimate that the feelings, thoughts and motives of one are readily comprehended by another" (p. 428). Evidence suggests that empathy is not a purely cognitive process: it also involves affects (Brems, 1989: Chlopan, McCain, Carbonell, 8 Hagen, 1985: Davis, 1983a, 1983b). The affect involved in empathy and the cognitive formulation of another's position are distinct variables as indicated by the empathy measures which distinguish them (Shelton 8 McAdams, 1990: Davis, 1983a, 1983b: Mehrabian 8 Epstein, 1972: Hogan, 1969). Thus, if empathy is a component of moral reasoning, affect must be a part of moral reasoning. Therefore, Kohlberg neglected, what is very likely, an integral aspect of moral reasoning. CAROL GILLIGAN - AN ALTERNATIVE COGNITIVE DEVELOPMENTAL THEORY One of the most outspoken critics of Kohlberg’s theory has been Carol Gilligan who has argued that Kohlberg's emphasis on justice presents an inherent gender bias. Gilligan (1982) included a care perspective in her theoretical formulations of moral development. She argued that while men may increasingly base their moral judgments on justice as they develop, women tend to base their judgments on care for the other individuals. According to 20 Gilligan’s conceptualization, the highest stage of Kohlberg’s moral reasoning that an individual utilizing the care ethic can achieve is the third stage. Thus, Gilligan concluded that women who continue to operate based on an ethic of care will be erroneously considered developmentally limited by Kohlberg's theory. Gilligan’s studies (1982) implemented Kohlberg’s method of presenting a moral dilemma and asking the participants how they would resolve the dilemma and why they chose the particular solution. The subjects in these studies were matched for age, intelligence, education, occupation and social class. Gilligan and her associates found that, upon interviewing girls and boys, they presented with extremely different responses. Boys emphasized rules in resolving moral dilemmas, while girls emphasized the impact their decision may have on others. This pattern remained fairly consistent throughout development, although adolescents and adult women tended to respond more frequently with a justice-oriented approach, than did girls. Thus, Gilligan agreed with Kohlberg that individuals become more justice oriented as they develop. However, Gilligan concluded that the socialization of girls in our society begins emphasizing care (little girls play games which emphasize turn-taking, while little boys play games which emphasize the rules), but when girls enter adolescence and must begin to compete in 21 the male-dominated, male-oriented society, changes occur in their approach as they tend to emphasize justice. Kohlberg, et a1. (1983), did not believe his theory is sexually biased and he refuted Gilligan's assertion that using his scoring system results in a biased downscoring of girls and women. In addition, he believed that the two theories are not in conflict, but that Gilligan's theory is incomplete. He maintained that a justice ethic guarantees impartiality and empathy for all involved, while a care ethic only provides empathy for people with whom the individual has a special relationship. Puka (1991) also argued that the care ethic is deficient in that it does not place emphasis on specific rules. Thus resolution of conflict has no clear guidelines. Instead, resolution becomes subjected to uncontrollable variables (i.e. individual differences from person to person and situation to situation). The end result, Puka believed, is the neglect of the society's well-being. Others, however, have argued that the care ethic represents a more true approach to morality because it accounts for the interpersonal domain (Brown 8 Tappan, 1991). Comprehensive reviews of the field (Rest, 1979: Walker, 1984) have been unable to produce sufficient quantitative support for Gilligan’s theory. Rest (1986) maintained that few studies have found a significant sex difference, and meta-analysis of the studies published indicates that sex 22 differences accounted for only 8% of the variance. In addition, Walker (1984) argued that the studies in which sex differences were found were methodologically flawed, relying too heavily on Kohlberg's earlier stage scheme or earlier scoring model. Lifton (1985), on the other hand, found that when gender rather than sex differences are examined Kohlberg's theory does appear to present a bias, with masculinity scoring significantly higher moral development, than femininity. In addition, a recent attempt to validate Gilligan's theory by operationalizing the care ethic and placing care responses next to justice responses on the DIT produced no significant gender differences (Friedman, Robinson, 8 Friedman, 1987). Gilligan, however, argued that research .has been unable to produce a significant gender difference because of the socialization process. Specifically, in the male-dominated society male-oriented rules are followed. Thus, society teaches women to view moral decisions with a "justice* approach, even though their preferred style is more society-oriented. This hypothesis received partial support from a study conducted by Gilligan and Attanucci (1988). They interviewed 34 women and 46 men and scored the interviews along a continuum ranging form pure care ethic to pure justice ethic. They found that while both men and women used the care ethic, women used it significantly more often than men. 23 Thus, Gilligan’s alternative to Kohlberg's theory, while supported equivocally by the research, is widely studied. Nonetheless, for the purposes of this study, the two theories are more similar than dissimilar. They both emphasize the importance of the cognitive constructs of empathy, while neglecting the role of affect. It is to affect that we now turn. KOHLBERG ON AFFECT AND MORAL DEVELOPMENT Kohlberg's theory acknowledges affect and maintains that the processes of affective development and cognitive development are similar. However, affect is viewed as always mediated by cognition. Kohlberg insisted "it is crucial, to distinguish between the description or expression of a feeling about a moral situation and the making of a moral judgment about it" (p.223, 1984). He stated that both violation of logic and violation of justice may arouse strong affects. But the development of motives as well as affect is largely the result of changes in cognitive patterns. He concluded that emotional arousal does not seem to be an internal determinant necessary to define moral behavior. Such an approach ignores the result of current research in the field of affect and cognition. Isen (1984) noted that, in the past, only strong negative affect was thought 24 to have an influence on cognition. However, in a review of the literature, Isen found that even mild mood states influence individuals' memory and overall reasoning. For example, Kykendall, Keating and Wagaman (1988) found that individuals rated neutral words differently depending upon their mood state, as measured independently. This finding held true for both experimentally induced mood states, as well as naturally occurring mood states. Hanlin and Zucker (1986) administered the DIT and the Paired Hands Test (PHT: Zucker and Barnett, 1977) to 97 undergraduates (75 women and 22 men). The PHT is a projective technique in which the individual is shown a picture of two hands and is invited to explain what the hands are doing (i.e. shaking, arguing, fighting). The responses are then scored along a positive- negative affect continuum. The significant relationship between the P-score on the DIT (a measure of principled judgment) and the PHT score indicates that affect, as well as cognition, plays a significant role in moral judgment at the principled level. In his exclusion of affect as a pivotal factor in moral reasoning, Kohlberg (1984) specifically identified guilt as a non-contributing variable in moral judgment. Recent empirical literature refutes that assertion. For example, Fehr, Fischer and Stamps (1987) found in a study of 98 undergraduate women that guilt correlated with mock jurors' decisions regarding severity of punishment for cases in 25 which the defendant was clearly guilty. Guilt was measured by the Mosher Guilt Scale (Mosher, 1966). In particular, as sex-guilt and hostility-guilt in the juror increased, severity of punishment for rape offenders decreased, and as sex-guilt, hostility-guilt, and morality-guilt increased, severity of sentencing for assault cases also decreased. It should be noted that moral judgment was not measured directly. Thus, while the findings suggest a link between affect and moral reasoning, one can only assume this to be true since there was no empirical measurement of moral reasoning. Additional studies have identified a significant relationship between guilt and moral judgment. For example, Propper and Brown (1986) investigated the relationship between sex-guilt and moral judgment at Stages 4 and 5 only. They found that, as the stage increased, the amount of sex- guilt decreased. They also found that family influence increased for Stage 5 individuals. The authors argued that such findings are consistent with Kohlberg's theory, as the transition from Stage 4 to 5 involves moving away from rule oriented formulations toward rational independent decisions. However, an alternative, affective interpretation of these results is that the individuals experience less conflict because they are willing to accept their family’s values. As a consequence, they experience less guilt. 26 Kohlberg's theory, however, does not allow for this interaction as he focused on moral reasoning on a cognitive level only. However, affect, especially guilt, appears to be an important factor in moral development. GUILT, SHAME, AND MORAL REASONING, Theories of moral development which account for affect find their roots in psychoanalysis. No discussion of psychoanalytic thought would be complete without reference to Sigmund Freud. Freud hinges his discussion of moral development on the development of the superego. The superego, he wrote, is the means by which a society guarantees that its laws will be followed (Freud, 1930). The concept of a superego receives full attention from Freud for the first time in IQe_Ege_eng_Ine_1g (1923). Freud conceived of the superego as being an independent intrapsychic structure which performed vital censorship functions to prohibit infantile wishes from being fulfilled or even becoming conscious. According to classical psychoanalytic theory, the superego begins formation with the resolution of the Oedipal conflict. To understand the concept of the superego and its importance to moral development, a brief digression into Freud’s writing on the Oedipus conflict is necessary. 27 Freud observed, through his own self-analysis and the subsequent treatment of others, a phenomenon commonly referred to as the Oedipus Conflict. In the process of normal development, the male child, between the ages of three and five, develops a deep sense of love and desire for his mother and intense feelings of rivalry towards his father (LaPlanche 8 Pontalis, 1973). For mythological and literary support, Freud turned to the story of QeQipee_Bex by Sophocles. In Freud’s description of the story, when Oedipus realizes that he has murdered his father and married his own mother he is overcome with guilt and self-inflicts injury to punish himself. The intense feeling of guilt and subsequent punishment indicates that Oedipus believed that he had done something wrong. If he did not have wishes to covet his mother and rid himself of his rivalrous father, Freud argued, the sense of wrongdoing would not have been so overwhelming. Freud concluded that, in normal development, the intrapsychic processes of the boy recognize that the competition with the father is a lose-lose situation. If the boy defeats the father, he will be overwrought with guilt. If the boy does not defeat the father, the father will most likely destroy the boy. The child’s best option is to give up his desires for his mother by repressing them into his unconscious and trying to emulate his father the 28 best he can. This process of emulating his father occurs by identifying with him and attempting to follow his father’s standards of behavior. The result is a superego which watches over the child's (and later the adult's) behaviors, unconsciously reminding him of the guilt pangs of wanting to destroy his father every time he violates one of those standards. As the child grows up, other cultural institutions are able to impact the developing superego, eventually resulting in a complete set of moral standards consistent with society's demands, the violation of which result in feelings of guilt. Freud’s emphasis on identification is similar to the identification process described by Piaget (1965/1932) and Gilligan (1982). John Gilligan (1976) maintained that psychoanalysis has been incomplete in addressing moral development. It has focused primarily on the affect of guilt and the resolution of the Oedipal Conflict, while neglecting the role of shame. The study of the relationship between shame and guilt can be found in the literature in a variety of contexts (e.g., Ausubel, 1955: Lewis, 1971: Nathanson, 1987: Tomkins, 1963). Early theorists such as Ausubel (1955) distinguished the two affects on a detection level. Ausubel minimized the phenomenological and developmental differences between the affects and instead defined shame as the affect experienced when committing a social transgression, while guilt is the result of transgressions of conscience. He also argued that 29 the anthropological evidence to support a distinction between guilt and shame is actually the same constructs defined differently by different cultures. However, a study (Ferguson, Stegge, 8 Dahmuis, 1991) has found that children within the same culture are able to differentiate the two affects. Goldberg (1989) reanalyzed the story of Oedipus Rex and Shakespeare's fiemle; (which Freud, 1900, also used as an exemplar of the Oedipal conflict) and compared the originals with Freud's selections and interpretations from those works. With regard to Qedipee_3ex, Goldberg described Freud's examination of the story as "perfunctory" (p.589: 1989). Goldberg explained that Freud neglected to mention that Oedipus' father, Laius, has been told by the gods that he is to be punished by bearing a child who will kill him. Consequently, Laius avoids sexual relations with Jocasta, his wife, except for the one night when Oedipus is conceived. When Oedipus is born, Laius and Jocasta try to kill him but are unknowingly unsuccessful. Thus, the first real threat in the story is the father trying to kill the son, not the other way around. In addition, in the scene in which Oedipus kills Laius, it is Laius who begins the altercation, not Oedipus. Thus, Goldberg concluded that the resulting tragedy and self-injury are not ”guilt recognition of who he was (namely, the slayer of his father and seducer 30 of his mother": p. 589), but "the shameful realization of who he is (namely, a frail and vulnerable mortal, who has been resented and unwanted by his own parents": p. 589). Goldberg argued that similar arguments can be drawn from Heml_§. First, the character Hamlet does not appear to be the maladjusted youth characterized by Freud. He was engaged in age appropriate same-gender and opposite gender relations, and there is no indication that he desired his father's demise. Also, Freud interprets Hamlet’s distress as the result of guilt, however, it could more accurately be described as shame. When the apparition of Hamlet’s father appeared to him and told him he was murdered, Hamlet instantly realized that he had been betrayed by his mother. He begins to wonder whether, since his mother plotted with his step-father to kill his father, could he be next. And even if his mother was not an active participant in his father's demise, how can Hamlet view her as able to protect him from his step-father. Again the theme of the helpless child being subjected to the powerful parent's rage becomes apparent. Goldberg insisted that Freud neglected the importance of an individual having a sense of safety and well-being and he believes these stories re-interpreted emphasize the importance for a child to feel safe and secure with his parents, especially the same gender parent. Thus, normal 31 development necessitates perception of care and respect from a parent responding to it in kind. Hultberg (1988) maintained that shame is instrumental in the role of moral development. Shame is experienced when the relationship with others is threatened. Thus, if a child does something which may disappoint the child's parents, the child will feel shame. In this way the child actively learns to uphold the moral norms so as not to be outcast and devalued. Guilt, on the other hand, is merely the affective reaction to the recognition of some wrongdoing which will result in punishment. SHAME VS. GUILT Other theories differentiate the two affects on developmental and phenomenological grounds (Nathanson, 1987: Tomkins, 1982: Lewis, 1971). James Gilligan (1976) defined shame as different from guilt in that shame involves "the feelings of inferiority, humiliation, embarrassment, inadequacy, incompetence, weakness, dishonor, disgrace, 'loss of face’: the feeling of being vulnerable to, or actually experiencing, ridicule, contempt, insult, derision, scorn, rejection or other ’narcissistic wounds': the feeling of not being able to 'take care of’ oneself" (p. 144). A frequent physical manifestation of shame is aversion to eye contact. Typically, individuals make excessive attempts to hide their own shame and tend to avoid others’ as well (Kaufman, 1989). In addition, shame results in global 32 attributions about the self as bad, and external locus of control. Reactions to shame include fear of punishment and a consequent withdrawal into the self (Lewis, 1971). By guilt, Gilligan referred to "the feeling of having committed a sin, a crime, an evil, or an injustice: the feeling of culpability: the feeling of obligation: the feeling of being dangerous or harmful to others: and the feeling of needing expiation and deserving punishment" (p. 144). Guilt results in specific attributions regarding one's behavior and internal locus of control. Reactions to guilt usually include reparative actions (Lewis, 1971). Attempts to distinguish shame and guilt empirically have produced equivocal results. This is consistent with what one would expect, however, as the phenomenological and theoretical definitions are similar. In fact, Tomkins (1963) concluded that shame is the precursor to guilt, and serves an important function in moral development. Tomkins states, "every individual is vulnerable to shame whenever he violates the social norms which he inherits by virtue of his membership in society (p.231: 1963)." Kaufman (1989), on the other hand, argued that shame and guilt are developmentally equal. Empirically, however, most studies have found that instruments designed to measure shame are more able than guilt measures to distinguish between the two affects (Kugler 8 Jones, 1992: Harder 8 Zalma, 1990). Such findings support the theory that shame is a precursor to guilt. As a consequence, the presence of shame will have an 33 effect on guilt, while the presence of guilt should have no effect on shame. Another reason shame and guilt are difficult to differentiate empirically is that the study of shame as a distinct affect is relatively new. Few well-validated measures have been created which distinguish between shame and guilt. As a result, any investigation into shame and guilt requires careful selection of the measure, as it should differentiate between the two. Recent evaluations (Harder 8 Zalma, 1991: Tangney, 1991: Harder, 1990: Harder 8 Lewis, 1987: Hobzitelle, 1987) of several available shame and guilt measures have identified three viable measures (Personal Feelings Questionnaire - 2 [PFQZ], Harder 8 Zalma, 1991: the Adapted Shame and Guilt Scale [ASGS]: Hobzitelle, 1987: and the Self-Conscious Affect and Attribution Inventory [SCAAI], Tangney, Burggraf, Hamme, 8 Domingos, 1988). The PFQZ and ASGS each have a guilt proneness scale and a shame proneness scale which exhibit strong test-retest reliabilities and tests for internal consistency. The PFQ2 consists of sixteen items which describe affective traits. Individuals rate the states using a Likert scale. The ASGS is an adapted version of the Shame/Guilt Scale (Gioella, 1981) which consists of 30 items. Individuals are asked to rate how well the adjectives describe them using a 7-point Likert scale. 34 The construct validity of the PFQZ and ASGS have been assessed in a series of tests (Harder 8 Zalma, 1991: Harder, 1990: Harder 8 Lewis, 1987: Hobzitelle, 1987). The PFQ2 and ASGS have been compared with the Binder Shame Proneness Scale (Binder, 1970), the Beall Situational Upset Scale for Shame and Guilt Proneness (Smith, 1972), the Mosher Guilt Scale (Mosher, 1966, 1968), and the Hobzitelle Adapted Shame and Guilt Scale (ASGS). Specifically, it was assumed that both guilt and shame scales would correlate significantly and positively with self-reports of self-derogation, instability of self-concept, and inadequacy of self. However, shame was expected to have a stronger relationship with these concepts. In addition, shame was expected to correlate positively with external locus of control, while shame was expected to have no relationship or a negative one. Finally, shame was expected to relate negatively to social desirability and guilt was expected to have no relationship, or a negative one. While the ASGS shame scale performed slightly better than PFQZ shame scale, both produced good construct validity. However, the construct validity of the ASGS guilt scale is questionable, while the PFQZ guilt scale performed quite well. Neither measure was subject to effects from social desirability. Tests of intra-scale correlations indicate that the ASGS has lower correlations than the PFQZ (Harder 8 Zalma, 1991). In addition, factor analysis and 35 subsequent item analysis indicated that the PFQZ scales loaded on two factors, shame and guilt (Harder 8 Zalma, 1991: Harder, 1987). However, factor analyses of the ASGS have been less consistent (Harder 8 Zalma, 1991: Hobzitelle, 1987). One weakness of the ASGS is that the adjectives used require advanced verbal ability (Tangney, 1990: Harder 8 Lewis, 1987). In addition, Tangney (1990) noted that the instructions for the ASGS require that individuals make global attributions about the self. Global attributions about the self are assumed to be characteristics of shame (Lewis, 1971). As a result, the measure may have an inherent shame bias. The PFQ2 has been criticized for it brevity. Despite the shortness of the scale, the PFQZ has performed remarkably well on tests of construct validity, and appears to be a stronger overall measure than the ASGS. A more valid criticism of the PFQZ is that well-defended individuals may be resistent to the shame scale (Harder 8 Lewis, 1987). The Self-Conscious Affect and Attribution Inventory (SCAAI) consists of 13 scenarios a college student may encounter. In a forced-choice format, individuals are asked to indicate which of the associated responses best describes how the individual would react to the scenario in daily life. The responses to the scenario indicate reactions of 36 shame, guilt, detachment and unconcern, externalization of blame, pride in one's own self, and pride in one’s own behaviors. The SCAAI is resistant to effects from social desirability (Tangney, 1990). Test-retest correlations, indices of internal consistency, inter-correlations of the subscales, and the method in which the measure was developed indicate the SCAAI is valid. However, the measure is relatively new and the construct validity of the measure has not been examined in as much detail as the PFQZ or ASGS. One apparent weakness of all three scales is their poor discriminant validity between guilt and shame, as the guilt and shame scales of each measure correlate significantly. However, one would expect shame and guilt to be difficult to differentiate. Negative affects, in general, tend to be difficult to discriminate and this is especially true for shame and guilt as they are theoretically, phenomenologically and clinically similar (Harder 8 Zalma, 1991). THE DIFFERENTIATION OF SHAME AND GUILT AS APPLIED TO EMPATHY AND MORAL REASONING J. Gilligan (1976) provided one of few attempts to apply affect, shame and guilt in particular, to moral reasoning. Gilligan believed that shame and guilt are underlying themes of Kohlberg's theory. He warned that 37 focusing on morality results in a state of perpetual conflict. By "morality", Gilligan referred to actions and thoughts which are motivated by a sense of compulsion either to act or avoid a certain act rather than to act freely of obligation. At Stage 1 of moral development an individual makes judgments to avoid punishment. Punishment in this context is experienced as total and is similar to the humiliation which is experienced during shame. At Stage 6 of moral development, an individual makes judgments based on what he thinks is necessary to create a cooperative society. Failure to make the correct judgment results in self- condemnation, which is the equivalent to guilt. The intervening stages from 1 to 6 would represent the progression from shame to guilt. Straughan (1983) argued that if J. Gilligan is correct in his assertions that the stages go from a shame ethic to a guilt ethic, this may be able to explain why Kohlberg's theory is more predictive of behavior at the postconventional level. In the development of the shame ethic retributions are avoidable, as they are administered by an outside agent. If a child decides not to clean his room he can avoid punishment as long as he can avoid his parents. As such, internalization of the rule occurs, but internalization of the consequence does not. That is, the internalized rule becomes, "It is only bad if you get 38 caught." Guilt ethic, however, involves internally driven consequences and therefore retributions for a violation are unavoidable. If the same child does not like disappointing his parents and he still does not clean his room, there is nowhere to hide. Thus, as an individual moves toward a guilt ethic, an internalization of obedience to the rules occur (not just awareness of the rule), and one is more likely to act in a way which is harmonious with that person's developmental stage. This formulation has also been supported by the research. For example, Helkama and Ikonen (1986) conducted a study of 116 preadolescents, adolescents, and young adults living in Finland. Moral maturity was measured by the Kohlberg moral interview and guilt was measured by individuals semi-projective responses to story completion items. A significant gender difference was found among the college student sample with males having a higher moral reasoning score than females. The relationship between guilt and moral stage existed in women only. Guilt especially differentiated at the conventional level. However, the use of the Kohlberg moral interview produced a truncated range of stages (postconventional N of 9). They also broke the preconventional and conventional stages into three more categories (pre 1/2, low 2/3 8 3, and high 4) and found that guilt has a significant relationship with both gender and stage, but found no stage/gender interaction. 39 Thus, not only did guilt in women increase but guilt within individuals increased as their stage level increased. In addition, the authors attempted to dichotomize guilt responses for women into those which involved guilt over omission and guilt over transgression. They found that guilt over omission produced a main effect on moral maturity along the full continuum of stages. In contrast, guilt over transgression produced a main effect on moral maturity only up to Stage 3. The limited effect of guilt over transgression may be understood if it is reframed into what we know of as shame. Thus, guilt, as internalized conflict, appears to increase as we develop through the stages, whereas shame plays a role only in the earlier stages. EMPATHY AND MORAL DEVELOPMENT According to J. Gilligan (1976), the problem with Kohlberg’s conceptualization of moral development is that the best a person can do is progress from making decisions to avoid humiliation to making decisions so as to avoid guilt. Such an approach to moral development has received some empirical support. Lifton (1985) found that moral maturity of college-aged men reflected "an almost neurotic concern for the development of a self-concept or self- identity" (p. 327) while moral development of college-aged 40 women reflected a "concern for the development of a social self or social identity" (p. 327). Gilligan (1976) proposed another option: rising above making decisions which are based upon morality. From a clinical perspective, work with a client is considered a success when that individual is able to make decisions or choose behaviors without having to experience intrapsychic conflict. The corollary for this in this discussion is what Gilligan referred to as: "the love relationship...in which the conflict between egoism (shame) and altruism (guilt) has been transcended, because through meeting one’s own needs one meets another’s and visa versa...If mental and emotional health and maturity mean anything, they seem to me to mean the ability to structure one’s relationships with others in such a way that it is through meeting ones’ own needs that one meets others’ needs as well (p.158)." This concept sounds remarkably similar to that of empathy. Gilligan identified the capacity for empathy as an integral part of moral development. Chlopan and his associates (1985) posited that there are two general approaches to the definition of empathy - one which defines it as "the ability to take another’s viewpoint" (p. 636) and another which defines it as "the ability to vicariously experience the feelings of another" (p. 636). The first definition appears to be the one most associated with the cognitive-developmental approach to moral reasoning, while the latter pursues the affective component found in guilt and shame research. 41 Chlopan, et al. (1985), called for an integrative measure which is able to assess both the cognitive and affective definitions of empathy. In addition, a recent factor analysis of several empathy measures indicates that empathy involves both an affective and cognitive component (Brems, 1989). Furthermore, it was determined that empathy requires a cognitive understanding of another’s position and a non-egocentric affective position. Another theorist who attempts to connect empathy with moral development is Hoffman. Using altruistic behavior as an indication of moral development, Hoffman (1976) believed that empathy, "the involuntary, at times forceful, experiencing of another person’s emotional state" (p.126), is at the core of moral development. Thus, Hoffman focused on the affective component. Empathy, he stated, exists in almost everyone. Hoffman asserted that empathy occurs through the conditioning of one’s own experiences and the distressful feeling of another’s similar experience. It can occur in the form of "empathic distress" which he describes as more of a self-centered recognition of one’s own past experience, or "sympathetic distress" which occurs when another person’s experience reminds us of our own so we are able to recognize that another person is in distress and respond to their needs. Obviously, sympathetic distress will more frequently result in altruistic behavior because the action which is 42 driven by empathic distress may not be beneficial to the party in need. Hoffman also recognizes that guilt, either free-standing guilt,-guilt over the sense of responsibility for another’s distress, or guilt over inaction, can lead to self-reflection and altruism. Hoffman did not differentiate between shame and guilt. However, what Hoffman calls "empathic distress" sounds remarkably similar to shame. Experiences of guilt produce an outward focus, as the self attempts to alleviate the guilt by correcting a mistake. The result is an attempt to increase awareness of others. In contrast, experiences of shame result in a withdrawal of the self in an attempt at self-preservation. Increased awareness of others is only made in attempts to avoid external punishment. As a result, one would anticipate that guilt is more strongly connected to empathy than is shame. In fact, Ryan and Connell (1989) found that affective concern for others correlated significantly with internalized locus of control, while such a relationship was not found for individuals with an external locus of control. The authors did not measure shame and guilt directly. However, shame has been associated with an external locus of control and guilt with an internal locus of control. Such findings are consistent with the results of Tangney’s (1991) work. Tangney conducted a series of studies examining the relationship between shame-proneness, 43 guilt-proneness, and empathy. In the first three studies, empathy was measured using the Feshbach and Lipian (1987) measure of empathy. This is characterized as an integrated measure of empathy. The results of these three studies suggested that shame is generally related to impaired empathic responsiveness, while guilt is related to enhanced empathic responsiveness. In the final study, Tangney measured empathy using Davis’ Interpersonal Reactivity Index (also an integrated measure of empathy: Davis, 1980). In this study she found that shame-proneness was significantly related to affective measures of empathy but was not related to cognitive measures of empathy. In contrast, guilt- proneness was related to both affective and cognitive measures of empathy. The empirical literature generally supports the connection between empathy and moral development. Thompson and Hoffman (1980) found that empathic concern for another’s emotional state correlated with an increase in concern for justice principles. In addition, they found that in subjects in which a sense of culpability was present, guilt increased with empathy, indicating that concern for others is a mediates the relationship between empathy and guilt. Shelton and McAdams (1990) conducted a study which lends at least partial support to Hoffman’s theory. In an attempt to develop a measure of morality which assessed 44 prosocial behaviors in the every day existence of adolescents, they found that the overall score of morality correlated significantly with a measure of overall empathy, as well as perspective taking (recognizing another’s point of view) and concern (caring and compassion for other), but not distress. However, the scale Shelton and McAdams developed provided no information regarding its reliability. In addition, the morality scale contained 3 subscales, all of which significantly correlated with each other, indicating overlap between the three subscales. Research using the DIT has also provided evidence for the existence of empathy at higher stages. Bowman and Reeves (1987) found a significant and positive relationship between counselors’ overall morality judgment score (P- score) on the DIT and their empathy ratings. Empathy was measured via supervisor ratings of videotaped therapy sessions and through independent judges’ ratings of counselors written responses to mock sessions. In another study using the DIT, moral development was found to correlate with role-taking only, and not empathy or logical cognition (Lee 8 Prentice, 1988). However, the sample for this study involved delinquent and non-delinquent adolescents and none of their subjects scored in the post- conventional range of moral maturity. Another study indicated that empathy may be the most important factor contributing to moral development. Curtis, 45 Billingslea, and Wilson (1988) conducted a study of 105 undergraduates to determine the relationships between socialization, empathy, authority figures and the DIT. They found that all three variables correlated significantly with moral maturity, however, a regression analysis revealed that empathy was the most important factor, accounting for 47 per cent of the variance. HYPOTHESES The review of the literature suggests that past research on moral judgment either has focused on the cognitive aspects, neglecting affect, or the affective components, neglecting cognitive processes. Empathy appears to involve both the cognitive reframing of another’s position and the sympathetic affect which one individual feels for another. Research supports both a relationship between empathy and moral reasoning, and between empathy and affective states such as guilt and shame. However, direct connections between affects such as guilt and shame, on the one hand, and moral development, on the other, seem lacking in the literature. For example, Ryan and Connell (1989) found that principled reasoning, as measured by the DIT (Rest, 1986), correlated positively with internalized locus of control, but not with external locus of control. In addition, they found that internalized locus of control correlated with affective concern for others, while externalized locus of control did not. Even though a relationship was posited between locus of control and shame and guilt (shame seems to correspond to an external locus of control and guilt to an internal one), Ryan and Connell (1989) did not attempt to measure shame and guilt, 46 47 precluding conclusions regarding moral reasoning and affect. It is hypothesized that, through the expression of empathy, one can bridge the gap between purely cognitive and purely affective approaches to moral development. The focus of the present study is the examination of the relationships among shame-proneness, guilt-proneness, selected correlates of empathy, and moral reasoning. The hypotheses are as follows: 1) Past studies have found that women tend to admit to feelings of guilt and shame more often than men (e.g., Helkama 8 Ikonen, 1986). However, other studies have found that gender identity is predictive of the differences between women and men (e.g., Tangney, 1991). While it is hypothesized that women will acknowledge more shame- proneness and guilt-proneness than men, it is also hypothesized that gender identity is a stronger predictor of shame-proneness and guilt-proneness than biological sex. 2) It is assumed that shame results in a withdrawal of the self in an attempt at self-preservation (Lewis, 1971: Nathanson, 1987: Tomkins, 1982). One would expect shame- proneness to have varying degrees of impact upon empathy‘ depending upon the level of shame-proneness and the specific correlates of empathy under examination. Specifically, a 48 linear relationship between shame-proneness and personal distress is predicted, whereas a curvilinear relationship between empathic concern and shame-proneness is anticipated. In the presence of low shame-proneness, empathic concern will increase, and in the presence of high shame-proneness, there will be no direct relationship between the two variables. No relationship between shame-proneness and the cognitive correlates of empathy (fantasy and perspective taking)is expected. 3) Guilt feelings involve acknowledgement of one’s errant behaviors (Lewis, 1971: Nathanson, 1987: Tomkins, 1982). The consequence of guilt is assumed to be an outward focus and an effort to correct this errant behavior and is presumed to foster the development of empathy. It has also been suggested that shame is a precursor to guilt. Thus, one would hypothesize that shame moderates the relationship between guilt and empathy. As a result, it is predicted that, in the presence of low shame-proneness, guilt- proneness will correlate positively with both affective correlates of empathy (Empathic Concern and Personal Distress). In the presence of high shame-proneness, guilt- proneness will not exhibit significant relationships with either Personal Distress or Empathic Concern. 49 4) It is suggested that guilt-proneness and shame-proneness influence moral reasoning while personal distress and empathic concern (affective correlates of empathy) act as a mediating variables in this relationship. However, it is also assumed that shame moderates the relationship between guilt and empathy. As a consequence, it is predicted that as guilt-proneness increases, personal distress and empathic concern and moral reasoning will increase, but only in the presence of low shame-proneness. In the presence of high shame-proneness, guilt will exert no main effect on empathy or moral reasoning. Specifically, it is hypothesized that personal distress and empathic concern will have as much impact upon mature moral reasoning as perspective taking and fantasy (the cognitive correlates of empathy). 5) It has also been suggested that empathy becomes increasingly important as moral reasoning increases. As a consequence, it is predicted that the relationship between moral reasoning and the correlates of empathy would be strongest at the postconventional level, moderate at the conventional level, and weakest at the preconventional level. 6) It has been assumed that shame-proneness, mediated by personal distress and empathic concern, enhances moral reasoning, but only through the third stage of development, 50 as defined by Kohlberg. As a result it is predicted that at the pre-conventional level, as shame-proneness increases moral reasoning will also increase. No significant relationship between shame-proneness and moral reasoning are expected beyond the third stage of development. METHOD W Participants were 312 undergraduates (125 men, 184 women, and 3 participants who did not identify their gender) recruited from the Psychology Department’s subject pool at a large midwestern university. The participants completed a series of measures in a group format. In exchange for their participation, individuals received credit toward one requirement in an introductory level psychology course. They ranged in age from 18 to 25 (M = 19.44, SD = 1.33). HQQfiQIQfi Defining Issues Test The Defining Issues Test (DIT) is a widely used cognitive measure of moral development designed by Rest (1986/1974). The measure is composed of six moral dilemmas, each describing a situation with competing social claims. The moral considerations involved in resolving the dilemma are identified and the individual is asked to rank them in order of importance. It requires little training, is a task of recognition, and provides a general morality judgment score in addition to placing the individual into a particular (Rest, 1979). Internal consistency of the DIT is generally in the high .703 and the test-retest reliability between .70 and .80 (Rest, 1986). It has proven resilient in cross 51 52 sectional studies (Martin, Shafto, 8 Vandeinse, 1977) and longitudinal studies (Rest, 1975). Factor analyses of DIT responses generally indicate that it reflects the invariant stage progression found in Kohlberg’s theory (Ma, 1985). In addition, Napier (1979) found that while individuals with prior knowledge of Kohlberg’s theory of moral development performed better on the measure than individuals with no such information, their performance was also resilient to attempts to "fake good“ regardless of the individual’s theoretical knowledge (Napier, 1979). Meta-analysis of the 22 studies with sex differences concluded that sex differences account for only about 0.5 per cent of the variance in DIT scores (Rest, 1986b). While the DIT has been found to be sensitive to the effects of education and SES status, such findings are consistent with predictions made by the theory (Rest, 1986b, 1976). Interpersonal Reactivity Index The Interpersonal Reactivity Index (IRI: Davis, 1980) is a 28-item questionnaire with a 5-point Likert scale. Respondents indicate how well each item describes them. The IRI consists of two cognitive subscales: Perspective Taking (PT) which addresses the tendency to take another’s point of view: Fantasy (FS) which addresses the individual’s ability to identify with feelings and characters in books, etc: and two affective subscales: 53 Empathic Concern (EC) which addresses the feelings of sympathy one has for other’s distress: and Personal Distress (PD) which addresses the level of self-oriented feelings. Internal consistency of these IRI scores has been reported to range from .71 to .77 (Davis, 1983a). The test- retest reliability ranges from .62 to .71. (Davis, 1983a, 1983b). Consistent with most other measures of empathy, Davis (1983b) reported significant sex differences for each subscale. The validity of the subscales has been measured (Davis, 1983a, 1983b). The interscale correlations were low to moderate (ranging from .01 to .38), and were all either positive or non-significant except for the relationship between PD and PT. The two cognitive subscales and EC have correlated positively with a general measure of empathy (Hogan Empathy Scale, 1969) while PD correlated negatively with this scale. In addition, Davis (1983a) found that a manipulation of affect significantly changed scores on the affective subscales, but not those of the cognitive subscales. The validity of these subscales has also been confirmed by factor analysis (Carey, Fox, and Spraggins, 1988). Ben Sex Role Inventory The Ben Sex Role Inventory (BSRI: Bem, 1974) is a 60-item in a 7-point Likert scale format. It contains 20 "masculinity" items, 20 "femininity" items, and 20 filler items. The scale is a widely-used, well- 54 validated measure of stereotypical gender-based attributes, independent of biological sex. Bem (1981) reported reliability coefficients of .75 and .78 for the Femininity scale and .87 and .86 for the Masculinity scale for male and female undergraduates, respectively. Measures of Guilt-proneness and Shame-proneness Personal Feelings Questionnaire - 2 The Personal Feelings Questionnaire - 2 (PFQ-Z) is a measure with both a guilt proneness and shame proneness subscale. The measure consists of ten items which describe affective states and the individual ranks how those items apply to him or her them using a 4-point Likert scale. The two week test-retest reliability of the scale ranges from .82 to .91. In addition, the Cronbach’s test for internal consistency is also high (shame r =.78 and guilt r =.72) The measure has also held up to a series of construct validity tests, despite the shortness of the measure (Harder, 1990: Harder 8 Lewis, 1987: Harder 8 Zalma, 1991). The construct validity of the measure has been compared with the Binder Shame Proneness Scale (Binder, 1970), the Beall Situational Upset Scale for Shame and Guilt Proneness (Smith, 1972), the Masher Guilt Scale (Mosher, 1966, 1968), and the Hobzitelle Adapted Shame and Guilt Scale (ASGS). Specifically, it was assumed that both guilt and shame 55 scales would correlate significantly and positively with self-reports of self-derogation, instability of self- concept, and inadequacy of self, however, shame was expected to have a stronger relationship with these concepts than guilt. In addition, shame was expected to correlate positively with external locus of control, while shame was expected to have no relationship or a negative one. Finally, shame was expected to relate negatively to social desirability and guilt was expected to have no relationship, or a negative one. The PFQ2 and its subscales exhibited superior construct validity as compared to other measures except for the ASGS shame subscale. The ASGS has also been criticized because the wording may be too advanced, however. Factor analyses indicate that the PFQ2 shame subscale accounted for 11.4% of the variance while the guilt subscale accounted for 29% of the variance (Harder 8 Zalma, 1991). In addition, Harder and Zalma reported that attempts to construct a scale post-hoc based on item validity was not able to produce a scale which was significantly different from either subscale of the PFQ2. Finally, the measure was not subject to effects from social desirability. One apparent weakness of the PFQ2 is its poor discriminant validity between guilt and shame, as the guilt and shame scales correlated substantially (ranging from .35 to .53) for the three studies. However, one would expect shame and guilt to be difficult to differentiate. Negative 56 affects, in general, tend to be difficult to discriminate and this is especially true for shame and guilt as they are theoretically, phenomenologically and clinically similar (Harder 8 Zalma, 1991). Self-Conscious Affect and Attribution Inventory (SCAAI) The SCAAI consists of 13 scenarios a college student may encounter. In a forced-choice format, individuals are asked to indicate which of the associated responses most likely describe how the individual would react to the scenario in daily life. The responses to the scenario indicate reactions of shame, guilt, detachment and unconcern, externalization of blame, pride in one’s own self, and pride in one’s own behaviors. While the overall measure appears valid, Tangney (1990) has had difficulty providing evidence for the validity of the Alpha Pride and Beta Pride subscales. These subscales are also not essential to this study. As a result, the validity evidence provided will only concern the four main subscales (shame, guilt, externalization, detachment/unconcern). Test-retest correlations range from .71 to .79 and internal consistency ranges from .46 to .82 (Harder 8 Zalma, 1991). The interrelationship among the subscales and their correlations with other measures of shame and guilt indicate the measure is valid. However, the measure is relatively 57 new and it has not been subjected to the same strenuous tests of construct validity as the PFQ2 (Harder 8 Zalma, 1991). The SCAAI is also resistent to effects from social desirability (Tangney, 1990). EEQEEQBIQS Participants were recruited via a standard departmental sign-up sheet requesting both male and female participants. All measures were completed in a group setting and in the presence of the examiner. After signing an informed consent form, the packets were distributed randomly. The order of the measures remained consistent, however, the first measure of the packet was rotated. Participants were allotted two hours to complete the packet. As they left the examining room, they were given a debriefing form which explained the purpose of the study. RESULTS Initially, 321 participants were recruited. Nine protocols were dropped from analyses as entire or major portions of responses to a measure were omitted. The remaining 312 protocols were included in the data analyses for the first three hypotheses. Illogical responses and missing values were deleted listwise. For Hypotheses Four through Six, responses to the Defining Issues Test were scrutinized and 38 (27 women and 11 men) incomplete protocols were dropped from further analyses. The remaining 274 protocols were scored as recommended by the author (Rest, 1986): but 84 of these did not pass the reliability tests proposed by Rest. Closer examination of the protocols indicates that 76 protocols failed the reliability measure, designed to insure that subjects follow the directions correctly. The directions of the DIT require the participant to rate the importance of each moral consideration from "no importance" to "great importance" and then note the four most important considerations. During the reliability check, the overall ratings of importance are compared with the respondent’s four most important considerations. Inconsistencies are then tallied. Protocols with inconsistencies in three or more dilemmas or 10 or more overall inconsistencies have fail the reliability check and are dropped from further analyses. The remaining 8 protocols which were dropped 58 59 failed Rest’s validity measure. These participants exhibited a tendency to choose solutions which, on first glance, appear acceptable, but actually are illogical. The remaining 190 protocols (115 women, 73 men) were used for all remaining data analyses. The reader is referred to the appendices for a complete listing of means, standard deviations, and correlation coefficients for subscale. The proposed hypotheses left undecided which measure of shame-proneness and guilt-proneness was to be used for data analyses. The final decision was based on reliability of the measures: both the Self-Conscious Affect and Attribution Inventory (SCAAI) subscales were found to be slightly superior to the Personal Feelings Questionnaire (PFQ2) subscales (Appendix A contains the Cronbach’s alpha coefficients for all measures). Consequently, only the shame-proneness and guilt-proneness scales of the SCAAI were used for all subsequent analyses with those variables. Hypothesis One predicted gender differences on the measures of shame-proneness and guilt-proneness. However, it also anticipated that gender identification would be the more significant predictor of shame-proneness and guilt- proneness than biological sex. As predicted, e-tests revealed that women admitted to shame-proneness and guilt- proneness significantly more than did men (Table 1). Multiple regression analyses were then conducted to test the 60 Table 1. Means, standard deviations, and e-tests of shame- proneness and guilt-proneness for women and men HQEQB MED Variables N Meeee SQ N Means SQ t-value Shame- proneness 174 2.78 .57 112 2.44 .52 5.22* Guilt- proneness 173 3.66 .51 112 3.23 .50 7.06* * p < .001. hypothesis that gender-role identification was more prominent in determining the shame- and guilt-proneness scores than biological sex. Results are presented in Table 2. The stated hypotheses were partially supported, in that Table 2. Multiple regression analyses of shame-proneness and guilt-proneness with masculinity, femininity, and gender. MEMM 3322mm Shame 275 1 Masculinity .07** .27 2 Femininity .07** .37 3 Biological Sex .03* .41 Guilt 274 l Masculinity .00 .05 2 Femininity .26** .51 3 Biological Sex .04** .54 Shame 275 1 Biological Sex .09** .30 2 Femininity .02* .34 3 Masculinity .05** .41 Guilt 274 1 Biological Sex .15** .39 2 Femininity .14** .54 3 Masculinity .00 .54 ** Q < .001. * n < .01. both a masculine gender identity and a feminine gender 61 identity were stronger predictors of shame-proneness than was biological sex. The femininity and masculinity scales each accounted for 7% of the variance in predicting shame- proneness. By contrast, biological sex only accounted for 3% of the variance. With regard to guilt-proneness, only femininity, which accounted for 26% of the variance, was a stronger predictor than biological sex, accounting for 4% of the variance. The regression analyses were then re-run with the order of the steps reversed, to insure against spurious findings. Now, biological sex became the more significant predictor of either shame-proneness or guilt-proneness (accounting for 9% and 15% of the variance, respectively). All subsequent analyses involving shame-proneness and guilt-proneness were conducted with women’s responses analyzed separately from men’s responses. Hypothesis Two predicted a linear relationship between shame-proneness and personal distress, a curvilinear relationship between shame—proneness and empathic concern, and no significant relationship between shame-proneness and the cognitive correlates of empathy (fantasy and perspective taking). Correlations were analyzed and are presented in Table 3. The stated hypotheses were partially supported. For both men and women, shame-proneness was significantly and positively related to Personal Distress scores while no relationship was found between shame-proneness and either Fantasy or Perspective Taking scores. 62 Table 3. Pearson correlation coefficients of shame- proneness and empathic concern, fantasy, personal distress, and perspective taking for women and men . ES £§ ED 21 Womeg .15* .09 .37*** .11 (N=l73) (N=173) (N=172) (N=l7l) gee .14 .01 .25** -.04 (N=108) (N=108) (N=108) (N=lll) EC Empathic Concern: FS a Fantasy: PD= Personal Distress: PT Perspective Taking. *** n < .001. ** P < .01. * p < .05. To test the predicted curvilinear relationship between shame-proneness and empathic concern, multiple regression analyses were performed. First shame-proneness was entered and then the anticipated curvilinear effect of shame- proneness. In addition, masculinity and femininity were also entered into the regression equation, respectively. The results are presented in Table 4. Contrary to prediction, no curvilinear relationship between shame- proneness and empathic concern was found. However, as noted in Table 3, a significant positive linear relationship between shame-proneness scores and Empathic Concern scores was evident for women. While the relationship between shame-proneness scores and Empathic Concern scores were not significant for men, the correlations for women and men did not significantly differ from each other. The third hypothesis predicted an interaction between 63 Table 4. Multiple regression analyses of empathic concern with shame-proneness, high shame-proneness, and gender identity, by gender. e Va . N §_t_e_p ingep, Veg, NZ-elnnge Nuie. R EC - Women 166 1 Shame-pron. .02 .14 2 Shame-pron.2 .00 .15 3 Masculinity .01 .19 4 Femininity .27** .55 EC - Men 104 1 Shame-pron. .02 .16 2 Shame-pron.2 .01 .17 3 Masculinity .02 .22 4 Femininity .32** .61 SHZEE:E§on. = Shame-proneness: Shame-pron.2== High shame- proneness: EC = Empathic Concern. ** - p < .001 shame-proneness and guilt-proneness and their predictability of empathic concern and personal distress. Multiple regression analyses were conducted for both dependent Ivariables. Guilt-proneness was the first variable entered into the regression equation with shame-proneness and the interactive effect of guilt-proneness and shame-proneness to follow. Results are presented in Table 5. The hypothesized interactive relationship between the independent variables and Personal Distress scores was not supported. The hypothesis was partially supported, however, as guilt- proneness was found to be a significant predictor of empathic concern for both women and men. Correlations were computed between guilt-proneness and each of the correlates of empathy. The results, presented in Table 6, indicate that, for men, guilt-proneness exhibits a significant relationship with each of the correlates of empathy, "M E Drc 64 Table 5. Multiple regression analyses of the affective correlates of empathy with shame-proneness, guilt- proneness, and the interaction of shame and guilt. Dsnl__yarl N Step Indsel_!a11. Iézshange Eultl_8 PD - Women 170 1 Guilt-pron. .01 .09 2 Shame-pron. .14** .38 3 ShameXGuilt .00 .38 EC - Women 171 1 Guilt-pron. .29** .54 2 Shame-pron. .01 .55 3 ShameXGuilt .00 .55 PD - Men 107 1 Guilt-pron. .04 .19 2 Shame-pron. .14** .41 3 ShameXGuilt .01 .42 EC - Men 107 1 Guilt-pron. .13** .36 2 Shame-pron. .00 .36 3 ShameXGuilt .00 .36 PD Personal Distress: EC = Empathic Concern. ** g < .001. (negative in the case of Personal Distress), while, for women, significant relationships were evident between guilt- Table 6. Pearson correlation coefficients of guilt- proneness and empathic concern, fantasy, personal distress, and perspective-taking for women and men. 39 £§ ED RI (N=172) (N=l72) (N=l71) (N=l70) Men .36*** .18* -.19* .39*** (N=108) (N=108) (N=108) (N=lll) EC = Empathic concern: FS = Fantasy: PD= Personal distress: PT = Perspective taking. *** p < .001. ** p < .01. * p < .05 proneness and empathic concern, fantasy, and perspective 65 taking. The fourth hypothesis predicted that shame-proneness and guilt-proneness would predict mature moral reasoning with the affective correlates of empathy (empathic concern and personal distress) acting as mediating variables. The relevant multiple regression analyses are reported in Table 7. The stated hypothesis was not supported. For women, Table 7. Multiple regression analyses of mature moral reasoning with correlates of empathy, shame- proneness, guilt-proneness, and the shame-guilt interaction. M 1! Sign MW Mange FL t- R MR - Women 103 1 EC .01 .ll 2 PD .01 .14 3 FS .01 .16 4 PT .07* .30 5 Guilt-pron. .00 .30 6 Shame-pron. .00 .30 7 ShameXGuilt .01 .32 MR - Men 65 1 EC .03 .17 2 PD .00 .17 3 FS .05 .28 4 PT .05 .36 5 Guilt-pron. .06* .44 6 Shame-pron. .00 .44 7 ShameXGuilt .00 .44 MR = Mature Reasoning: EC = Empathic concern: PD = Personal distress: FS a Fantasy: PT = Perspective taking. scores on the Perspective Taking scale was the only variable which significantly predicted mature moral reasoning, and for men, only guilt-proneness was a significant predictor of mature moral reasoning. Hypothesis Five predicted that the correlation between 66 moral reasoning and the measures of empathy would be strongest at the postconventional level, moderate at the conventional level, and lowest at the preconventional level. The correlations between each stage of response on the DIT and each correlate of empathy were examined for the anticipated trends. The correlations were then converted to t-values using Fischers’ transformation. The results are presented in Tables 8 and 9. There is partial support for this hypothesis. Perspective-taking was the only correlate of empathy to demonstrate the expected trend for both women and men. In addition, the men’s responses on both the Empathic Concern and Personal Distress scales (see Table 9), demonstrated a significant trend when comparing conventional responses with postconventional responses. While the association between responses to the Fantasy scale and Stage 3 responses differed significantly from the correlation between the Fantasy scale and Stage 6 responses, there is no apparent trend. As such, it is most likely that this finding of significance was by chance. Hypothesis Six proposed that shame-proneness would be a more significant predictor of moral reasoning at Stage 2 and Stage 3 than at higher stages. As with the previous hypothesis, the correlations between each stage of response on the DIT and shame-proneness were examined for the anticipated trend. The correlations were then converted to 67 Table 8. Computed t-values contrasting the correlation coefficients between stages of moral reasoning and correlates of empathy, stages of moral reasoning and shame-proneness, and stages of moral reasoning and guilt-proneness for women. fiLflQfifi . Qontresfied E; ED ES 21 figeme guilt 2 - 3 - .22 .43 - .01 - .03 1.13 .56 2 - 4 1.13 .47 .63 1.87 1.83 1.54 2 - Sa 1.41 .32 1.07 2.55* 1.52 2.27* 2 - 5b 1.03 1.23 1.33 2.98** 1.82 1.78 2 - 6 .90 1.21 .19 3.46** 1.43 .49 3 - 4 1.18 .01 .56 1.66 .54 .82 3 — 5a 1.54 - .07 1.03 2.45* .46 1.65 3 - 5b .90 .81 1.30 2.92** .74 1.20 3 - 6 1.13 .83 .20 3.56** .39 - .03 4 - 5a .41 - .08 .49 .84 .06 .86 4 - 5b - .26 .82 .77 1.29 .21 .41 4 - 6 - .08 .84 - .37 1.84 .16 - .88 5a - 5b - .82 1.06 .28 .44 .32 - .60 5a - 6 - .60 .99 - .98 .94 - .10 -1.97 5b - 6 - .22 .00 -1.33 .56 - .40 -1.49 EC = Empathic concern: PD = Personal distress: FS = Fantasy: PT = Perspective taking. N = 104 women. *-n<0050 ** -p< .01. t-values using Fischers’ transformation. The results are presented in Tables 8 and 9. The hypothesis was not supported. Ad hoc analyses were conducted to determine if guilt-proneness exhibited a stronger positive relationship ‘with higher level stage responses. The results, also presented in Tables 8 and 9, indicate that guilt has a different relationship with moral reasoning, depending upon 68 Table 9. Computed t-values contrasting the correlation coefficients between stages of moral reasoning and correlates of empathy, stages of moral reasoning and shame-proneness, and stages of moral reasoning and guilt-proneness for men. §L§Q§§ Cen;rested fig 2Q £5 £1 Shame Guiit 2 - 3 .51 1.18 1.00 .39 - .09 .14 2 - 4 -1.65 .29 1.08 - .37 .71 .32 2 - 5a - .00 .89 2.55* -l.54 .58 1.61 2 - 5b .54 .30 .88 1.10 .72 2.11* 2 - 6 1.81 1.38 1.72 3.15** .95 2.61* 3 - 4 -2.03* -1.35 .11 - .71 .75 .18 3 - 5a - .46 - .25 1.40 1.06 .62 1.90 3 - 5b .03 - .88 .10 .73 .93 2.62* 3 - 6 1.31 -2.53* - .70 2.67** 1.02 2.41* 4 - 5a 1.45 1.01 -1.21 1.65 - .12 1.43 4 - 5b 2.00 .53 - .20 1.33 .06 2.09* 4 - 6 3.86**-1.13 - .64 3.83** .24 2.39* Sa - 5b .63 - .68 -1.89 - .51 .25 .72 5a - 6 1.75 -2.24* - .76 1.49 .36 .54 5b - 6 1.31 -1.79 - .87 2.01* .17 - .08 EC = Empathic concern: PD = Personal distress: FS = Fantasy: PT = Perspective taking. N = 66 men * - n < .050 ** - p < .01. gender. For women, guilt-proneness and stage of moral reasoning exhibit the expected trend from stage 2 through stage 5a, but no further. For men, the correlation between guilt-proneness and level of response gradually increases as the stage of response increases. DISCUSSION GENDER DIFFERENCES IN PRONENESS TO SHAME AND GUILT The first hypothesis predicted that sex and gender identification differences in participants’ acknowledgement of proneness to shame and guilt. Consistent with the stated prediction and the literature (Cohn, 1991: Tangney, 1991: Helkama 8 Ikonen, 1986), women in this sample admitted to significantly more shame-proneness and guilt-proneness than did men. W Multiple regression analyses suggested that gender identification was a stronger predictor of shame-proneness than was biological sex. However, when the regression analyses were re-run with the steps reversed, biological sex was the most significant predictor of shame-proneness. Masculinity and femininity scales continued to contribute significantly to the regression equation, but the step reversal indicates a decrease in their contributions. Closer scrutiny of the correlation coefficients between masculinity, femininity, biological sex and shame-proneness indicates that biological sex exhibits the strongest relationship with shame-proneness. This suggests that, while admission of feminine characteristics and absence of masculine characteristics contribute to acknowledgement of 69 70 shame-proneness, biological sex is the stronger predictor, with women admitting to more shame-proneness than men. Gu' - r n e Traditional masculine characteristics were not significantly related to acknowledgement of guilt-proneness. This finding suggests a societal perception that the term "masculine" implies less guilt. When regression analyses on guilt-proneness were run in the predicted order, femininity accounted for significantly more of the variance than did biological sex. Again, the steps of the regression analyses were then reversed to insure against spurious findings, and biological sex and feminine characteristics accounted for roughly equal amounts of the variance. Closer analysis of the correlations between biological sex and guilt-proneness and femininity and guilt-proneness reinforced the impression that traditionally feminine characteristics was the more important factor in predicting guilt-proneness. Feminine characteristics are associated with the admission of guilt, while acceptance of masculine characteristics were not associated with admission or rejection of guilt in this sample. Wages These shame- and guilt-proneness findings suggest that researchers must remain sensitive to gender differences based on both biological sex and definitions of gender identity. That is, real differences may exist between men 71 and women with regard to their acknowledgement of these dimensions. This suggests that factors, other than the acknowledgement of a stereotypical gender identity, may be better predictors of shame- and guilt-proneness. Given the substantial differences between women and men, discussion of the results and their implications will be conducted separately for each sex, and then the differences will be discussed. THE RELATIONSHIP BETWEEN SHAME-PRONENESS AND THE CORRELATES OF EMPATHY The second hypothesis predicted a significant positive relationship between shame-proneness and personal distress, a curvilinear relationship between shame-proneness and empathic concern, and assumed no relationship would be evident between shame-proneness and either perspective taking or fantasy. - ss 0 w Consistent with the stated prediction and the literature, women’s responses yielded a positive, linear relationship between shame-proneness and personal distress. Furthermore, multiple regression analyses of shame-proneness 'with the correlates of empathy revealed that personal distress was the only variable that served as a significant jpredictor of shame-proneness. sh cu st. pl ar an Va 72 Based upon the literature of affective development (Tangney, 1991: Tomkins, 1972), it had been predicted that shame-proneness and empathic concern would exhibit a curvilinear relationship. That is, in the presence of low shame-proneness, empathic concern would increase, but in the presence of high shame-proneness, empathic concern would level off or decrease. The multiple regression analyses did not produce the expected results for women. Ad hoc analyses of the correlations between the two variables indicates that for women the two variables are related, but linearly and positively. As anticipated, no significant relationship was found between shame-proneness and either fantasy or perspective taking. Wash Men’s responses similarly indicated that the acknowledgement of shame-proneness exhibited a linear, positive relationship with personal distress. In addition, multiple regression analyses of shame-proneness with the correlates of empathy revealed that personal distress was the only variable that significantly predicted shame- proneness. Also consistent with the findings for women, no significant relationship between shame-proneness and either perspective taking or fantasy was found. In addition, the anticipated curvilinear relationship between shame-proneness and empathic concern was not established. In fact, there was no significant relationship between the two variables. 73 o ' s n a e Comparison of the findings for men with those for women suggests that shame-proneness is predictive of personal distress in both women and men. This supports the theory that shame contributes to a preoccupation with the self and one’s anguish. In addition, for both women and men no relationship was evident between shame-proneness and the measures of either fantasy or perspective taking. This is also consistent with the literature (Tangney, 1991: Davis, 1983a), as shame-proneness is considered an affective construct while fantasy and perspective taking are both cognitive constructs of empathy. Men’s responses indicated no direct relationship between shame-proneness and empathic concern, and women’s responses suggested a positive, linear relationship between the two variables. The differences appear paradoxical. However, while the correlation for the women’s responses is significantly different from zero and the men’s is not, men’s and women’s responses did not differ significantly from each other. As such, it is difficult to draw clear conclusions from the data. It is possible that either the significant finding for women is a Type II error, or the nonsignificant finding for men is a Type I error. One further speculation is for women, shame-proneness has a positive effect on affective concern for others, and that the more intense one’s shame-proneness, the more intense 74 their concern for others. However, according to the hypothesized theory and empirical evidence in this sample, as shame-proneness in women increases, the personal distress and concern with the self also increases. Perhaps as personal distress increases, it interferes with the commonly expected effects of empathic concern. Thus, as shame- proneness increases, one’s internal focus and concern for one’s own well being also increases tending to preclude active concern for others. While these findings are consistent with Davis’ (1983b) study in which manipulation of affect resulted in changes in the affective correlates of empathy and not the cognitive correlates, it also contradicts his (1983a) earlier findings that empathic concern and personal distress are inversely related. Cohn (1991) suggests that college-age women are slightly more mature developmentally than college-age men. However, it is unlikely that these results are indicative of differences in affective development as shame developing into concern for others is a milestone of early childhood. Another possible explanation lies in the finding that, for both men and women, the greatest predictor of empathic concern was admitting to traditionally feminine characteristics. In fact, for women it accounted for 27% of the variance, while accounting for 32% in men’s responses. This suggests that, for both men and women, empathic concern is considered a feminine trait, even though both women and 75 men possess it. Given that the women in this sample admitted to significantly more empathic concern than the men, one may surmise that, for women, socialization conveys the message that their role is to care for others and that that role is so pervasive as to become entrenched in their identity. College-aged men, on the other hand, are socialized to downplay their concern for others, since it implies a feminine characteristic. THE RELATIONSHIP BETWEEN GUILT-PRONENESS AND THE CORRELATES OF EMPATHY The next hypothesis predicted that an interaction between shame-proneness and guilt-proneness would significantly predict empathic concern and personal distress. No interaction was found, although ad hoc analyses of the relationship between guilt-proneness and the correlates of empathy provided additional interpretations of the results. For women, guilt-proneness was positively correlated ‘with empathic concern, fantasy, and perspective taking, and guilt proneness was a significant predictor of empathic concern, but not of personal distress. Unlike the findings for men, no significant relationship was found between guilt-proneness and personal distress. p1 ar de ii 01 76 For men, the results were very similar. As guilt- proneness increased, empathic concern, perspective taking, and fantasy also increased, while personal distress decreased. Furthermore, multiple regression analyses indicated that guilt-proneness was a significant predictor of empathic concern and was not a significant predictor of personal distress. In comparing the findings for women and men, while guilt-proneness is positively related to the correlates which facilitate overall empathy (empathic concern, perspective taking, and fantasy) for both men and women, a major difference exists. For men, guilt-proneness is inversely related to personal distress, while no significant relationship between the two variables existed for women. It appears that, with increasing degrees of guilt, men are better at suppressing personal distress. It should be emphasized that I am talking about admitting personal distress. One can speculate that men are probably less likely to admit to personal distress under the pressure of «guilt so as to preserve a "strong and impassive" stance. Women, on the other hand, may have less of a need to remain passive and stoic . 77 THE DIFFERENTIATION BETWEEN SHAME AND GUILT It should be noted that, even though it was never a stated hypothesis, these findings support the assumption of affect theory that shame and guilt are distinctly different constructs. As suggested by affect theory, shame-proneness was predictive of an internal focus and feeling sorry for oneself and guilt-proneness was predictive of being concerned for others. This finding is even more noteworthy in that the findings were different despite a high correlation between the present measures of shame and guilt. Thus, while no interaction was found, the theory that shame and guilt are related, but distinct, remains viable. It is also noteworthy that, while women and men differed significantly in the degree to which they admit to guilt- and shame-proneness, acknowledgement of these traits is associated with similar experiences for both men and women. SHAME, GUILT, EMPATHY AND MORAL REASONING The fourth hypothesis predicted that, in the presence of low shame-proneness, guilt-proneness and the affective correlates of empathy (empathic concern and personal distress) would be predictive of mature moral reasoning. For women, only perspective-taking was a significant predictor of mature moral reasoning. While this is contrary 78 to what was predicted, it is somewhat consistent with what the author of the Defining leenee Tee; (Rest, 1986) would predict. The DIT is based upon Kohlberg’s model of moral development, which clearly emphasized cognitive, rather than affective, processes. In addition, mature moral reasoning requires perspective taking. For men, guilt-proneness was the strongest predictor of mature moral reasoning, while none of the direct measures of empathy significantly predicted of mature moral reasoning. This finding appears to be paradoxical since it seems to suggest that when men exhibit mature reasoning they are also likely to possess more guilt-proneness. While this is somewhat consistent with the predicted model, the fact that this finding existed for men and net women is not consistent with the literature. Furthermore, the women in this study acknowledged greater levels of guilt than did the men. One possible explanation is that women and men may experience and utilize guilt differently. Perhaps the guilt women experience is more consistent with what society deems as appropriately "feminine", while their moral reasoning and attachment to others operates independently of guilt. If this is true, it would be consistent with James Gilligan’s (1976) theory of affect and moral reasoning, which stated that healthy development suggests moral reasoning, independent of either guilt or shame. For men, it appears the guilt they experience is more contextual and 79 interpersonal, indicating that it is more of a factor in moral reasoning. Remembering Cohn’s (1991) findings of gender—based developmental differences in college-age students, this finding may be indicative of developmental differences between women and men. As such, one would anticipate that if an older sample was studied, men’s acknowledgement of guilt-proneness would more closely approximate that of women’s. t a so Hypothesis Five predicted that the relationship between moral reasoning and the correlates of empathy would change as one progressed across levels of reasoning. The reader is reminded that the Defining ieenee Tee; (Rest, 1986) requires the participant to indicate how they would resolve a given dilemma. Each response given by the participant corresponds with a stage (2 through 6) of moral reasoning, as outlined by Kohlberg’s theory of moral development. The correlations between stage of response and empathy were compared and tested for significance. For women, as one progressed from the lowest to the highest stages, the relationship between perspective taking and moral reasoning changed. In general, the correlations were negative for Stage 2 responses and gradually increased until they were significantly positive at Stage 6. This is consistent with the literature (Rest, 1986) as one would expect perspective taking to become more crucial at the 80 higher stages. No other correlates of empathy produced similar trends for women. While men’s responses also reflected the expected trend between moral reasoning and perspective taking, the findings for empathic concern, personal distress and fantasy were more complex. The predicted relationship was evident for empathic concern when comparing Stage 4 and Stage 6 responses. However, there was also a significant difference between Stage 3 and Stage 4 responses, but in the opposite direction! This suggests that there is something unique about stage 4 responses, as compared to the rest of the DIT in that men required a lack of empathic concern to produce Stage 4 responses. One possible explanation refers back to Carol Gilligan’s (1982) belief that Stage 3 and Stage 4 responses differ in the degree to which they pull for the "care ethic". She believed that Stage 3 responses required more of a care responses than Stage 4 responses. In fact, while Stage 3 and Stage 4 both comprise the conventional level of moral reasoning, Stage 3 responses typically reflect a desire to maintain social approval and interpersonal trust and Stage 4 responses reflect an emphasis on maintaining the rules so as to maintain society. Thus, the change in importance of empathic concern in men seems to be an accurate reflection of differences between the two stages. 81 With regard to personal distress, when comparing Stage 3 responses with Stage 6 responses and Stage 5a responses with Stage 6 responses, an extreme diminishing of the relationship is evident. This may suggest that abandoning overconcern for one’s own situation is necessary for Stage 6 responses, as they reflect attempts to consider the effects action may have on each individual involved. It is puzzling, however, that Stage 4 responses are not included in this trend. One possibility is that "maintaining the rules to maintain society", which is reflected in Stage 4 responses, does not require the individual to avoid overconcern with the self. The relationship between fantasy and moral reasoning becomes stronger as Stage 2 responses are compared with Stage 5a responses. This suggests that as one increases the level of moral decision making an increase in the ability to partake in fantasy is also available. This is logical, as the ability to utilize imagination is necessary to partake in perspective taking. Comparison of the results produced by women with those of men suggests different uses of empathy in moral decision- making processes. For men, there were no consistent relationships between each of the correlates of empathy and moral reasoning, while for women such covariation was evident for perspective taking. 82 Shame-pnonenees end ieyei e: stege response The sixth hypothesis predicted differences in the correlations between shame-proneness and level of stage response. The reader is reminded that each response on the Defining Ieenee Tee; (Rest, 1986) corresponds to one of the stages outlined in Kohlberg’s theory of moral development. No significant differences were found for either women or men. This suggests that either shame is not related to moral reasoning whatsoever, or it is so far removed that it has no direct effect. While this is consistent with the literature, which typically does not posit any type of relationship between shame and moral reasoning, it contradicts the rationale presented in this investigation by suggesting that shame has no direct impact upon moral reasoning. Perhaps this should not be surprising. Even if the theory proposed in this study were totally true, one would anticipate only traces of a connection between shame- proneness and moral reasoning, as there are several steps of development between the proposed connection of shame with moral reasoning. In addition, this sample consisted of college students who are likely to have progressed to the latter phases of affective development. As a consequence, the shame-proneness they acknowledge may manifest itself in a quantitatively and qualitatively different manner from shame-proneness manifested in younger, less developed children. For example, in younger children feelings of 83 shame frequently result in acting-out behaviors, whereas college students are more able to experience their shame without necessarily exhibiting similar acting-out behaviors. Gu' - e vels asoni Ad hoc analyses of the correlations between stages of moral reasoning and guilt-proneness were also conducted. For men, the relationship between guilt-proneness and stage of moral reasoning were generally negative for Stage 2 responses and gradually increased until they were significantly positive at Stage 6. The increasing importance of the relationship between guilt-proneness and moral reasoning mirrored that of perspective taking and moral reasoning, suggesting that in the moral reasoning process guilt and perspective taking may be linked, at least for men. Women displayed a different pattern. For women, the significance of the stage of moral reasoning and guilt- proneness gradually increased until level 5a, after which the importance declined. However, the importance of perspective taking continued to increase through all the stages. This suggests that guilt and perspective taking are utilized, increasingly so, during the preconventional and conventional levels, but as one responds at the postconventional level of moral reasoning, one is able to understands others’ situations and make difficult decisions 84 without reliance of guilt reactions. This is further indication that women and men utilize guilt differently. SUMMARY OF FINDINGS Consistent with the hypothesized theory of this paper, the overall results suggest that guilt and the affective components of empathy are important in moral reasoning. For women, the strength of the relationship between guilt and moral reasoning increases as the level of moral reasoning increases, although the importance of guilt slackens at the postconventional level. For men, the strength of the relationship between moral reasoning and guilt, personal distress, and empathic concern varies depending upon the level of moral reasoning. Furthermore, the results indicate that women do, indeed, acknowledge significantly more shame and guilt than do men. However, there is overlap, as men acknowledged such feelings as well. While some of the differences are attributable to socialized expectations based upon gender identity, that is not entirely the case. That is not to say that their biological sex predisposes women to experience more affect then men, rather, the degree to which acknowledgment is tied to identification with traditional gender roles, as measured by the Ben Sex Role Inventory (1974), is not the only factor which explains the 85 differences. It is likely that other determinants of gender identity, such as socialization, influence the acknowledgement of affect. Both women and men exhibited the expected relationship between personal distress,and shame- proneness. There were also apparent differences between the sexes as to how acknowledgement of guilt interacts with their capacity for empathy and moral reasoning. For women, guilt- proneness was not a significant predictor of moral reasoning, although the relationship between guilt and moral reasoning became stronger as women endorsed higher level responses. In addition, guilt-proneness was positively related to perspective taking, fantasy and empathic concern, and perspective taking was a predictor of moral reasoning. For men, the relationship between guilt-proneness, empathy, and moral reasoning was more consistent with what one would anticipate. Guilt-proneness was predictive of moral decision making and was significantly related to each of the correlates of empathy identified by Davis (1983a). This study also yielded data supporting the view that shame and guilt are distinctly different constructs. While the two measures were highly correlated with each other, shame-proneness was a significant predictor of personal distress while guilt-proneness was a significant predictor of perspective taking for women and empathic concern for men. While this was not exactly what was predicted, it 86 follows the general theory that shame elicits a withdrawal into the self while guilt elicits a focus which turns outward. THEORETICAL, DEVELOPMENTAL, AND CLINICAL IMPLICATIONS This study has several theoretical, developmental and clinical implications for men and women. Pivotal to these implications is the fact that the gender differences produced in this study were so pervasive as to warrant separate speculations for men and women. For men, shame-proneness was associated to personal distress and guilt-proneness. The development of guilt- proneness, in turn, fosters a decrease in personal distress, and leads to empathic concern. As intellectual development begins to catch up with emotional development, the individual is able to utilize fantasy and perspective taking and apply them, along with guilt, to moral reasoning. Guilt and perspective taking are pivotal in the process of moral reasoning for men. Thus, fostering the development of guilt and encouraging young men to look at others’ perspective may facilitate an increase in their ability to make mature moral decisions. It is somewhat paradoxical that, at least for these college-age men, merely encouraging them to feel a sense of responsibility (or guilt) for their actions may be counterproductive. Findings in this study suggests that 87 society perceives guilt-proneness as a feminine characteristic. As such, late adolescent males, who are continuing to struggle with their resurging gender-identity crisis (Erikson, 1968) may have difficulty acknowledging their feelings of guilt. For women, shame may also lead to personal distress and fosters the development of guilt. However, there is also a higher degree of inter-relatedness between shame-proneness, personal distress and empathic concern in women. One speculation on the source of this finding is that socialized definitions of femininity as guilt-ridden appears influential to women’s development. That is, they are socialized to acknowledge guilt and have concern for others. As such, the development of guilt does not contribute to the reduction of personal distress, as it does with men. As cognitive development progresses, the person is able to apply guilt and perspective taking to moral reasoning. Women transcend the use of guilt in moral reasoning, however. For women, increasing mature moral decisions requires assisting them in taking some perspective on a situation prior to making a decision. Furthermore, they may experience the opposite effect of society’s definition of feminine characteristics. That is, young women may feel the onus of carrying more guilt, because of society’s definitions of gender identity. Indeed, this study, as well 88 as many others (Tangney, 1992: Cohn, 1991: Tangney, 1991) have found that women tend to admit to more feelings of shame and guilt than do men. This study has also yielded some evidence_that women and men experience guilt in a qualitatively different manner. Such a finding lends further credence to the notion that, in therapeutic settings, it is necessary to work with a client to determine what a certain thought or feeling means to that client. The therapist can not accept the meaning of a thought or feeling at face value, as it could mean very different things. The possibility for discrepancy is especially exacerbated when the therapist and client are of opposite sex and the client is an adolescent male. LIMITATIONS OF THIS STUDY This work also had significant limitations. The sample was extremely homogeneous, placing limitations on the extrapolations one can make about these results. Also, this study attempted to measure shame-proneness and guilt- proneness (traits) and then extrapolated those findings to theories of shame and guilt (states). Such extrapolations can be tenuous and future studies would benefit from using more precise measures of shame and guilt. The sample of college students was convenient and useful for an initial investigation. It is possible that no differences in the 89 main effect of shame-proneness on moral reasoning was found because most college students are not as open to feelings of shame or have developed beyond that stage in affective development. Another weakness was that the measure of gender identity used in this study was unable to sufficiently explain the difference between men and women in their acknowledgement of shame- and guilt-proneness. Also, the measure used for assessing moral development was skewed toward cognitive processes. That is, the DIT addressed cognitive concepts of moral development, as such one would anticipate a more difficult time establishing a connection between affect and moral reasoning when moral reasoning is measured by the DIT. Other measures, such as face-to-face interviews, may be more beneficial for future studies, especially if they are longitudinal in nature. FUTURE RESEARCH The idea that shame is a precursor to guilt was not tested in this study, as the measures used precluded conclusions of causality or temporal ordering. Future studies need to address this issue. This would require either longitudinal data or cross-sectional data utilizing a sample with a greater diversity of age cohort. Another 90 possibility would be to utilize a measure of moral reasoning which is not based solely on cognitive processes. While this study provided information regarding gender identity and affective traits, it did not isolate what characteristics are predictive of shame—proneness and guilt- proneness. This would be particularly interesting given the gender differences which consistently appear in studies. In addition, this study merely began the investigation of gender differences. It would be worthwhile isolating what characteristics of men and women specifically are related to utilization of shame, guilt and empathy. Again, longitudinal data or cross-sectional data with a wider range of age cohorts would be useful in determining if some of the differences found in this study are the results of maturational differences in college students or differences inherent in women and men. The Interpersonal Reactivity Index (Davis, 1983a) does not provide an overall index of empathy. Furthermore, it does not differentiate between accurate and inaccurate attempts at empathy. It would interesting to see if there is a direct relationship between guilt-proneness, shame- proneness and an overall measure of empathy. Based on the findings in this study, one would anticipate the existence of such a relationship. One would expect shame-proneness to interfere with attempts at accurate empathy, however. Furthermore, the connection between an overall measure of 91 empathy and mature moral reasoning would be worthy of investigation. It should be noted that measures of empathy which account for affective and cognitive aspects of empathy are rare. In sum, this study successfully established a connection between affect and mature moral reasoning. However, the nature of that relationship appears moderated by sex and gender issues. Furthermore, evidence suggests that, while men commonly admit to less guilt than women, that guilt may be experienced in a qualitatively different manner. While college-age men appear to utilize guilt in attempts at moral decision making, college-age women appear to be utilize guilt only to a point in moral reasoning and then rely solely on perspective taking. Finally, this study also produced results that shame and guilt are different phenomena. APPENDICES APPENDIX A List of Cronbach Alpha Values by Scale SCALE SCAAI shame-proneness SCAAI guilt-proneness PFQ2 shame-proneness PFQ2 guilt-proneness Bem masculinity Bem femininity IRI empathic concern IRI personal distress IRI fantasy IRI perspective taking FULL SAMPLE .73 .74 .67 .70 .83 .82 .78 .75 .77 .75 92 WOMEN .74 .70 .66 .68 .84 .79 .70 .71 .76 .75 MEN .67 .67 .67 .71 .84 .79 .71 .76 .72 .74 APPENDIX B Means and Standard Deviations for all Measures Scale Tetal Sample N x so Masc 304 3.53 .48 Fem 305 3.49 .48 Shame‘ 288 2 . 65 . 57 Guilt} 287 3.49 .55 Shamez 305 2 . 64 .47 Guilt? 298 2.60 .55 EC 302 3.90 .62 pp 302 2.61 .63 PS 303 3.45 .74 PT 306 3.39 .61 Stage2 190 8.42 5.24 Stage3 190 18.61 8.99 Stage4 190 29.39 10.97 StageSa 190 25.26 9.53 StageSb 190 6.16 4.73 Stage6 190 4.27 4.12 StageP 190 35.69 11.56 N 179 182 174 173 180 175 182 181 183 181 115 115 115 115 115 115 115 ERRED Mean 3.45 3.47 8.49 17.99 29.70 25.90 6.01 4.17 SD .48 .43 .57 .51 .44 .54 .50 .60 .72 .61 5.34 9.01 9.98 9.10 4.67 4.04 36.08 11.33 N 123 121 112 112 123 121 118 119 118 123 73 73 73 73 73 73 73 Men X 3.65 8.31 19.41 29.27 24.11 6.32 4.54 34.98 Masc = Masculinity: Fem = Femininity: Shame1 8 Guilt1 = Subscales of SCAAI: Shame2 8 Guiltz = Subscales of PFQ-z: EC PD 8 Personal Distress: FS = Fantasy: PT Empathic Concern: Perspective Taking. 93 SD .46 .44 .52 .50 .49 .56 .63 .67 .61 5.16 8.91 12.39 10.22 a0opuxv «009820 a0opuzv A0opuzv aoopuzv «00'020 «mu—030 Anopuzv Ann—nay “009030 o~. ~m. mm. mm.. 63.. o~.. -. 6.. .oo..:. Acosuz. .oopua. «oop.z. .oopuz. “mop-z. .mopua. so. Pp.. 00.- o~.. 0... am. we. «oopuxo Aoopuz. “oopuzv “cop-av «mores. Rampage PP. .n.- n~.. o... ~p. ~o. Asap-a. Aoopuzc Ago—nae Ame-«z. .mopuz. s¢.- mn.. 0P.- «F. 5.. Ago-.2. .oopua. Ann—.2. Anus... o~.- mo.. 60.- ’0. “oopuzv Anew-av “mop-:0 00.- a... s0.. Amer-av “mop-av 0~.- ~p.- ..on... -. goons» coo-um anon-um anon-um coo-um moo-um ~oooum ha «a 00. amnpuzv 00.- Ann—lav N0. «mop-xv 00. “mop-av ~0.- Ann—lav «0. “nowuzv n0.- a00nnzv 0p.. «p0nuav 0'. mp. “cup-av mp. aeowuzv ~0. acownav pp. acopuav 0—.- 500.030 00.- “cavity m0.- n00nuzv ow. Anon-av an. “p0nlav up. on . “No—lav m0. awnwuxv 0p. n~npnzv awupnav n0. awopuzv n0.. a~opnxv ¢~.. anuwuzv N0. “wow-:0 N0. Apomuzv 0N. Aoo~uzv b0. «advao .oLauonaa ~.oa.acco_uoo:o ace..oou .acoaaoa. .oL-ocnau stouco>c_ co.u:n_auu< u000e< n:o_oacou-».om. 32].: fists: 39...: 30.33 goo—nay 00. “00.030 00. goo—I20 n0. «009020 00. Aoopuzv «0.. nonpuzv ~0.. a~opuav a..- a00nuav «0.. goon-xv o0. aho~uxv kn. Ahounav .oluga nu. “ospuzv n0. aospuzv aw. “esp-xv or. Amh—uzv ~0.. gospuav er.. nus—uzv 59.. anuuuzu no. «mow-:0 Nn. np0~uau o0. «won-av on. aghwuzv Np. “manage «a. .u..:0 madamm H6909 you wounded: Han no xauumz scauuaouuoo U NHOZQQA< o0. “shanty n0. ask—nay s0. Ashpuzv o0. Ash-020 s0. Ahspuzv 00.. amhpuzv n..- n¢0~u20 0—. anumuzv mp. “Nunuzv 0n. anunuzv h~. ash~lxv «N. Anna-av an. ahoflnzv pm. .898 00.. “00.020 0w. 500—020 s0.. anew-av p0.. Aonpuzv m0. goo—020 «0.. Anna-av 00.. a00nnzv n9. Anew-av sN. asoNuxv up. “sow-:0 no. ano~uzv ~p. a00nazv up. n90~nzv on. amomnzv mN. In; 00.- «00.020 «0. Ann-lav 00.. Anne-xv nw.- Anew-av e0. Ann-nay n0. Ans—lav 00.. aoowuav op. “sowing N0. asowusv on.. “cow-av ~0.. anuNnxv ¢—.. “gonna. 0~.. a00~lzu 00.. Aponnzv 0~.. Ap0nlav ~0.. one: 000.00 one-um amouaum once-um coo-um nee-um ~ouoaa he mm on on .u..:o .uI-ge .uL—ao In; 00.8 94 0:0 30Hmn mum cos you muons anunzv on.- “MNIZV pp.- annuzv mp.- Anhuzv mm.- Annexe 0N.- Ann-xv No.- anhuzv s0.- annuxv me.- Anuuzv —m.. anhnzv «0- annuzv op.. Anhuzv op. Anhuzv “nu-:0 0N. cm. anhuzv n0. Ampwuzv hm. amp-uzv pm. amppuzv N0. Amp—920 ac.- Am—puzv pm.- “appuzv on.- “Nppuzv 9N. acppuav up. Amppuxv Pp. amepuxv «0.. Ampwuzv cw.- Amppuzv 0N.- Ampwuzv «N.- aNppuzv NN. «cw—nay N0.. amp-020 N0. “app-xv NN.- “app-:0 Fm.- “mus-20 0N.- “Nppuzv we. aeppuzv «a. Amppuzv v¢.- «app-30 on.- amp—axe «N.- aNppuzv up. nvauxv 00. “my—nay o~.. “menus. amwpuzv —0. 00.. “Nppuzv “Nppuzv 00.. 0N.- aeppuav aeppuav 00.. 00.- an—puzv Anppuzv «app-av anppuzv an—puzv An-puxv 00. pp. mp. «0.. n0. N0.- anppuxv amp—nay Amp—nag amp-nay Anppuxv Amppuxv NF. 00. N0.- mp. N0. 00.. Aoppuzv “099-20 gov—nay Aer—nay now—020 “opp-xv no. 00. Np.- 00. 00. N0. aN—Puzv “Np—~20 aN—puzv «N-puav «Np—axe aN—pnzv m0. '0. 00. m0. N0. N0. “sop-:0 “so-nay “no—us. n50.nzv “509.30 ahownzv n.- 0p.- 00. op. n0. 00.- A00—uxv .00-n2. A00puzv goo—nay ‘00-'20 aoopuzv 00. N0. 00. no. no. mo.- aepwuzv Aer—nay .e—puav “cw—nay “cup-20 aeppuzv 00.- 00. «9.- 00.- NN. op.- ANFpuzv “Nppnzu “Nppuzv “Nppuzv “Nppuzv “N—Pnzv m..- or.. so. nF.- N0.- NP. aoooum oooaum amoouum omoooum «ocean nausea Anhuzc ANN-ac Ape-z. o~.- on. ow. Ans-ac .~s.zv .ehuac e... on. n.. .nkuzc Amsuzc “gauge 6... vs. ...- Ansuz. «Nu-x. nos-z. 09.- 4.. «N. Ansuzo Awsuav Aouuzv ~P.- P~.- no. .nsuac a~suac Ans-ac o... no.- no. “Nana. nosuav ~w.. o..- .~._.zc “op—.3. NN.. no. Reps-av “popuzv 66.. «~. .nF—nz. .oopux. .~o.uz. no.- N... no.- Anppuzv “oopuzv «Nowuxc n..- as. on. noppuz. Amssuz. Agnpuzc NN.. —.. «0.. .~...a. Asssuz. Asap-a. o~.- ~0.- No. “sop-z. Asap-ac .mspazc up.. «4. m~. “copes. A's—ax. Annwuzv c~.- PF. 66. .e-puzv nospuav Apes-20 mo.. on. .N. .Nppuzv «oxpuzv “asp-z. ~o.- up. ~P. ~ououm ha «a .Hmcoomfio on» o>ons a—huzv N0. .Fsuav NM.- “puns. N0. Ark-xv mp. “Funny N0.- Apsuav Np. grungy N0.- «op—030 op.- «owpnxv 50. “—0.030 00.. anhpuzv NN. ahbwuav «n. awhpuzv 00. 0thuxv sn. gob—020 «p. Anhpuzv «n.- 05 .co: one :0803 you mousmmoz Ham «0 D XHszmmd nouuzv op. “05020 Nn. a0suzv Np. “03-30 00. “0&020 ON.- “05:20 0w. “05.20 «0. «0.9020 mm. amps-av 00. “uppity 00. Anhpnav nu. non—030 N0. “Nun-av em. Annpuzv mp. a00pnzv Nm. Askpuzv 00. um AFN-av 00. “psuzv 0p. AFN-:0 e0. nwsnzv m0. «rs-20 cp- Apsuzv o0.- apsnzv Na.- aoppnzv «0. “mun-:0 0p. “opp-av NN. Am—plzv n0. Ans—lav on. Anopuzv 09. 5009.30 NN. Annuity N.- “959020 "N.- .00.:0 .Hccoooep one c0503 you mucoaowuomoo sawunaouuoo. .odooonaa N.oa_occo_uao:o ou:..oou Locooaoa. .odaoonaa >aouc0>c_ co_u:b.auu< uuoee< u:o_oocou-».om_ anhuzv 00. annuzv e0. Ansnav n0. Ann's» e0. anhuzv 00.- Anhuzv 00.. «musty p0. a-N—uzv 00.- abvpuzv «a. gnaw-:0 on. ah—puzv 50. Apruav on. “our-av Np. nphpuzv Nn. Ans—nay Mp. Ann—020 FN.. .....s Aooazv mn- Auolzv NN. «00020 0». Amouzv up. «000:0 pp.- Acouxv op.- anouxv 0N.- Ap—puzv on. ~00puzv op. “00pu20 op.- «00’030 on. «oowuzv 00. a0p-uxv cw. anhwlxv 5Q. APNFIZV Ow. AnoPIZV no.u .udwau xfiuumz :oHumHGHuou “00:20 50. anouzv 0p. “00:20 «0. «mouse N0. accuse N0. “00020 00.- “00030 00.- grapes» «0.. “009-20 —0. acorns. nN. a00pnav v—. «o0puav FN. a0pwnzv No. “N—puzv 0*. .Ns..z. ON. .ocp.x. 6~.. ..sugm “pussy N0. grungy up. annuav 00. “"5030 N0.- Apsuzv NN.- a—huzv or. awhuzv n0.- ao—pnav ne- anppnxv 00. noppuzv n0.- Amer-xv cm. “op—nag «N. a0Npuav 0N. «009020 on. 500—030 k0. Amhpuzv 3. In; “Nu-20 00.. “Nsuav 0N. “Nhuzv nu. aNhuxv 00.- aNhuzv 0p. “Nu-av pp.- ANN-av o0.- ApNvuav nN- Abe-.20 m0. “opp-30 on.- ah—pnzv 0p. aonuzv n0.- «NNpuau mp.- a0-puav mp. “oppuav m“.- APN—uzv 0'. one: sou-um 95:8 smooaum .mgnum coo-um magnum ~ooaum ha «a 68 um .u__:0 .81-gm .u._:o IOu one: 95 APPENDIX E DEMOGRAPHIC DATA Please answer these questions in the space provided. 1. Your sex: 1. Female 2. Male 2. Your age: 3. Your religion: A. Catholic B. Protestant C. Baptist (circle one) D. Buddhist E. Jewish F. Muslim G. Hindu H. Shinto I. None J. Other (please specify) 4. How strongly do you identify with your stated religion? Not Very Strongly Very Strongly 1 2 3 4 5 6 7 5. Your ethnicity: A. Caucasian B. African-American C. Asian D. Native-American E. Hispanic F. Other (please specify) 96 APPENDIX F BEM SEX-ROLE INVENTORY Please indicate on a five point scale how well each of the characteristics describes you. Write the number appropriate for your answer beside each item number on the enswer snee . 1 2 3 Never or almost never true of me 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Self-reliant Yielding Helpful Defends own beliefs Cheerful Moody Independent Shy Conscientious Athletic Affectionate Theatrical Assertive Flatterable HaPPY Strong personality Loyal Unpredictable Forceful Feminine Reliable Analytical Sympathetic Jealous Has leadership ability Sensitive to the needs of others Truthful Willing to take risks Understanding Secretive 97 4 5 Always or almost always true of me 31. Makes decisions easily 32. Compassionate 33. Sincere 34. Self-sufficient 35. Eager to soothe hurt feelings 36. Conceited 37. Dominant 38. Soft spoken 39. Likeable 40. Masculine 41. Warm 42. Solemn 43. Willing to take a stand 44. Tender 45. Friendly 46. Aggressive 47. Gullible 48. Inefficient 49. Acts as a leader 50. Childlike 51. Adaptable 52. Individualistic 53. Does not use harsh language 54. Unsystematic 55. Competitive 56. Loves children 57. Tactful 58. Ambitious 59. Gentle 60. Conventional APPENDIX G SELF-CONSCIOUS AFFECT AND ATTRIBUTION INVENTORY Below are situations that people are likely to encounter in day-to-day life, followed by several common reactions to those situations. As you read each scenario, try to imagine yourself in that situation. Then indicate how likely you would be to react in each of the ways described. We ask you to rate e11 responses because people may feel or react more than one way to the same situation, or they may react different ways at different times. Please do not skip any items - rate all responses. We’ve left space for you to add an additional response, in case we left out something important. 1. A professor whom you admire asks a question in class. You raise your bad and give the wrong answer. You: a) Have the feeling that everyone 1 2 3 4 5 is looking at you. not likely very likely b) Feel annoyed with yourself 1 2 3 4 5 raising your hand and vow to study more for the next class. c) Think to yourself "You win 1 2 3 4 5 some, you lose some", and remember that everyone is sometimes wrong. d) Think to yourself that it was a 1 2 3 4 5 tricky question anyway. 98 99 2. You’ve been helping yourself to your coworker’s supply of chocolate without her knowledge. One day, she angrily tells you that she suspects someone else. She calls him "inconsiderate, as usual". You: a) Make a joke about chocoholics. 1 2 3 4 5 not likely very likely b) Apologize and replace the 1 2 3 4 5 chocolate. c) Avoid both coworkers. 1 2 3 4 5 d) Think she shouldn’t leave 1 2 3 4 5 people with such a temptation. 3. While walking down the street, you see someone of the opposite sex looking at you with interest. You: a) Are pleased that you new 1 2 3 4 5 interest in fashion is not likely very likely noticed and appreciated. b) Feel self-conscious and 1 2 3 4 5 embarrassed. c) Enjoy the attention and feel 1 2 3 4 5 more attractive. d) Figure he/she had mistaken you 1 2 3 4 5 for someone else. e) Worry that he/she might be 1 2 3 4 5 misreading your availability. 100 4. Your roommate, a good friend who rarely dates, invites you to attend a party with her/him and a new date. You go and discover that the date is not only very attractive but is flirting with you. You exchange phone numbers and later say to yourself: a) If they really wanted to 1 2 3 4 5 develop a relationship, they not likely very likely would have spent the evening alone. b) If I hadn’t exchanged phone 1 2 3 4 5 numbers someone else would have. c) I should cancel the date because 1 2 3 4 5 I could never enjoy myself under the circumstances. d) I must be a real loser to have 1 2 3 4 5 to steal my friend’s date. 5. A friend confides a personal secret to you. Later, in a casual conversation with a mutual friend, you accidentally let the secret slip. You: a) Tell yourself that your friend 1 2 3 4 5 should have realized that not likely very likely sharing information like that is risky. b) Put it out of your mind. 1 2 3 4 5 Things like this happen all the time. c) Ask yourself repeatedly what 1 2 3 4 5 kind of friend you are anyway. d) Decide to think before you speak 1 2 3 4 5 after this. 101 6. A friend asks you to do him/her a favor. Though you could reasonably go out of your way slightly to do this, you just don’t feel like doing it. So you turn him/her down. Later you tell yourself: a) Why am I so selfish? 1 2 3 4 5 not likely , very likely b) This kind of thing happens now 1 2 3 4 5, and then between friends. c) I’ll find a way to make up for 1 2 3 4 5 this. d) Some people expect too much from 1 2 3 4 5 their friends. 7. When visiting an elderly, very wealthy aunt who is in poor health, you notice a copy of her will lying open on the table. When she goes to the kitchen to get coffee, you quickly read as much as you can. Afterwards you feel: a) That anyone who leaves a will 1 2 3 4 5 in plain view expects people not likely very likely to read it. b) You feel badly for having looked 1 2 3 4 5 at her personal papers and treat her especially well for the rest of the afternoon. c) It doesn’t matter since you 1 2 3 4 5 would have found out when the will was read anyway. d) You feel embarrassed and quickly 1 2 3 4 5 leave as soon as possible. 102 8. You and a friend are jointly responsible for your club’s finances. Your friend is balancing the club checkbook. S/he finds an error made by you which makes it necessary to re-balance the checkbook. This is a long and tedious task. You say to yourself: a) Why am I such a careless person? 1 2 3 4 5 not likely very likely b) Why doesn’t the bank come up 1 2 3 4 5 with a better form for keeping track of checks. c) I’ve wasted my friend’s time. I 1 2 3 4 5 should redo it. d) It’s too bad, but anyone can 1 2 3 4 5 make a mistake. 9. You are struggling to complete a difficult physics exam for which you feel unprepared. A student next to you deliberately holds their test papers so that you can read the answers. You know that person is an "A" student in physics and you quickly copy several answers. The next day you say to yourself: a) Everyone cheats. It’s no big 1 2 3 4 5 deal. not likely very likely b) I must be a really dishonest 1 2 3 4 5 person. c) I really feel I made a mistake 1 2 3 4 5 cheating on the exams. I should have studied harder. d) It wasn’t my fault. The answers 1 2 3 4 5 were given to me. 103 10. You and your friend each submit a project to a competition. You win. You: a) Are pleased that you have such 1 2 3 4 5 talent. not likely very likely b) Think "Boy, was I lucky!" 1 2 3 4 5 c) Wish you did not have to attend I 2 3 4 5 the award ceremony. d) Are glad that you had worked so 1 2 3 4 5 hard and it had paid off. e) Worry about your best friend. 1 2 3 4 5 He/She worked hard, too. 11. While meeting your boyfriend/girlfriend’s parents for the first time, you make a comment that they don’t seem to appreciate. You realize too late that what you said could have been interpreted another way. You: a) Wonder why your boyfriend/ 1 2 3 4 5 girlfriend didn’t clarify not likely very likely what you meant. b) Let it pass, and move on to 1 2 3 4 5 another topic. c) Wish you could just disappear. 1 2 3 4 5 d) Become concerned that you may 1 2 3 4 5 have offended them and try to undo what you’d said. 104 12. You spend most of Saturday comparison shopping for a television set. You finally decide on a model, bring it home and install it. When you switch it on you find there is no sound. Highly annoyed, you return it to the store and say a few choice words to the clerk. You then: a) Tell yourself that losing your 1 2 3 4 5 temper was understandable not likely very likely under the circumstances. b) Complain to the manager that 1 2 3 4 5 clerks should make sure the sets work before selling them. c) Apologize to the clerk. It 1 2 3 4 5 wasn’t her fault. d) Leave the store as quickly as 1 2 3 4 5 possible, avoid it in the future, and hope no one heard you. 13. Your spouse, girlfriend, or boyfriend unexpectedly has been treating you more lovingly lately. You respond by: a) Feeling vaguely uneasy, and 1 2 3 4 5 embarrassed for some not likely very likely inexplicable reason. b) Wondering what you had done to 1 2 3 4 5 deserve this and wondering how you could return the favor. c) Feeling that you are a lovable 1 2 3 4 5 person. d) Thinking that you must have 1 2 3 4 5 really pleased him/her by being thoughtful the past week. e) Wondering what’s put him/her in 1 2 3 4 5 such a good mood. APPENDIX H PERSONAL FEELINGS QUESTIONNAIRE - 2 Instructions: For each of the following listed feelings, to the left of the item number, please place a number from 0-4 reflecting how common the feeling is for you. 4 - you experience the feeling continuously or almost continuously 3 - you experience the feeling frequently but not continuously 2 - you experience the feeling some of the time 1 - you 0 - you 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. experience the feeling rarely never experience the feeling Embarrassed Mild guilt Feeling ridiculous Worry about hurting or injuring someone Self-consciousness Feeling humiliated Intense guilt Feeling "stupid" Regret Feeling "childish" Feeling helpless, paralyzed Feelings of blushing Feeling you deserve criticism for what you did Feeling laughable Feeling disgusting to others Remorse 105 APPENDIX I INTERPERSONAL REACTIVITY INDEX Read each item: decide how well it describes you. Then for each item, blacken the circle corresponding to the number that shows how well the item describes you according to the following rating scale. 10. 11. 12. 13. Does not describe Describes me me well at all very well 1 2 3 4 5 I daydream and fantasize, with some regularity, about things that might happen to me. I often have tender, concerned feelings for people less fortunate than me. I sometimes find it difficult to see things from the "other guy’s" point of view. Sometimes I don’t feel very sorry for people less fortunate than me. I really get involved with the feelings of the characters in a novel. In emergency situations, I feel apprehensive and ill- at-ease. I am usually objective when I watch a movie or play, and I don’t often get completely caught up in it. I try to look at everybody’s side of a disagreement before I make a decision. When I see someone being taken advantage of, I feel kind of protective towards them. I sometimes feel helpless when I am in the middle of a very emotional situation. I sometimes try to understand my friends better by imagining how things look from their perspective. Becoming extremely involved in a good book or movie is somewhat rare for me. When I see someone get hurt, I tend to remain calm. 106 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 107 Does not describe Describes me me well at all very well 1 2 3 4 5 Other people’s misfortunes do not usually disturb me a great deal. If I’m sure I’m right about something, I don’t waste much time listening to other people’s arguments. After seeing a play or movie, I have felt as though I were one of the characters. Being in a tenses emotional situation scares me. When I see someone being treated unfairly, I sometimes don’t feel very much pity for them. I am usually pretty effective in dealing with emergencies. I am often quite touched by things that I see happen. I believe that there are two sides to every question and try to look at them both. I would describe myself as a pretty soft-hearted person. When I watch a good movie, I can very easily put myself in the place of a leading character. I tend to lose control during emergencies. When I’m upset at someone, I usually try to "put myself in his shoes" for a while. When I am reading an interesting story or novel, I imagine how 1 would feel if the events in the story were happening to me. When I see someone who badly needs help in an emergency, I go to pieces. Before criticizing somebody, I try to imagine how I would feel if I were in their place. LIST OF REFERENCES REFERENCES Ausubel, D. P. (1955). Relationships between shame and guilt in the socializing process. Deyeneiegieei Review, §2(5), 378-390. Bem, S. L. (1974). The measurement of psychological androgyny- Journal of consulting and Qliniool Boronoloox- l2(2)- 155- -162- Bem, S. L. (1981). Gender schematheory: A cognitive account Of sex tYPihg- Psycholooiool BszioW-fifi- 354-364- Binder, J. (1970). The Beiative Ezeneneee £2 Shame er Guilt as a Dimension of character Stile. Unpublished doctoral dissertation, University of Michigan. Blasi, A. (1980). Bridging moral cognition and moral action: A critical review of the literature. Dsycnoiogieei Bulletin. 55(1). 1‘45- Bowman, J. T. 8 Reeves, T. G. (1987). Moral development and empathy in counseling. counselor Eouoation and Snoorxision 29(4). 293-298- Brems, C. (1989). Dimensionality of empathy and its correlateS- The Journal of Esxoholoox- 121(4). 329-337- Brown, L. M. 8 Tappan, M. B. (1991). Commentary. finnen Desolonnont- 11. 81-87- Carey, J. C., Fox, E. A., 8 Spraggins, E. F. (1988). Replication of structure findings regarding the Interpersonal Reactivity Index. Neeengenent end Exolnotion in.£ounsolino.ano Dossloonsnt- 21. 102-105- Chlopan, B. E., McCain, M. L., Carbonell, J. L., 8 Hagen, R. L. (1985). Empathy: Review of available measures. gogrngl of Eorsonoli_x and Sooiol Esxonoloox 88(3)- Cohn, L. D. (1991). Sex differences in the course of personality development: A meta-analysis. Eeyeneiegieei Bulletin-.lofitzi- 252-266. 108 109 Conn, W. E. (1981). Affectivity in Kohlberg and Fowler. Religious Education. 19(1). 33-48. Curtis, J., Billingslea, R., and Wilson, J. P., (1988). Personality correlates of moral reasoning and attitudes toward authoritY- Psychological Resorts. as. 947-954- Davis, M. H. (1980). A multidimensional approach to individual differences in empathy. Qfihfi Qatalog e; Selected Documents in Psychology. 19. 85. Davis, M. H. (1983a). The effects of dispositional empathy on emotional reactions and helping: A multidimensional approach. Journal of Bersonalitx- 51(2). 167-184- Davis, M. H. (1983b). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journgl of Eersonalitx and Social Esxsnclocx. 11(1)- Dupont- H- (1976)- Affectiue Derelonnentt Stage and segnence. Unpublished manuscript. Eau Claire, WI: University of Wisconsin. Erikson, E. H. (1967). IQQDLISYI Xenhhi eng insis. New York: W. W. Norton Fehr, L. A., Fischer, 8. M., 8 Stamps, L. E. (1987). Mock jurors’ behavior: Sentencing as a function of jurors’ guilt- Boreholocical Reports. to. 727-731- Ferguson, T. J., Stegge, H., 8 Dahmuis, I. (1991). Children’s understanding of guilt and shame. thig Deucloonent- t2. 827-839. Freud. S- (1900)- The Interpretation of Dream.- New York: W. W. Norton 8 Co. Freud, S. (1923). The Egg eng hhe In. London: W.W. Norton 8 CO. Freud. 8- (1930). cirilization and Its Discontents. New York: W. W. Norton 8 Co. Friedman, W. J., Robinson, A. B., 8 Friedman, B. L., (1987). Sex differences in moral judgments? A test of Gilligan’s theory- Boreholoox of Homen Quarterlx- 11- 37-46. Gibson, D. R. (1990). Personality correlates of logical and sociomoral judgment. Journal of Bereonalitx and social ESYQDQIQQY- 52(61- 1296-1300- 110 Gilligan, C. (1982). In e Diffefenn Veice. Cambridge: Harvard University Press. Gilligan, C. 8 Attanucci, J. (1988). Two moral orientations: Gender differences and similarities. Neffill-Eelmer Quarterly. 31(3). 223-237- Gilligan, J. (1976). Beyond morality: Psychoanalytic reflections on shame, guilt, and love. In T. Lickona (Ed ) Moral Derelooment ans Dehariort Theorxl Researchl and Social Tsenee. New York: Holt, Rinehart and Winston, (pp. 144- 158). Gioella, P. P. (1981). The relationship of relative shame/guilt proneness to the attribution of responsibility under no shame and shame arousal (Doctoral dissertation, New York University, 1981). Dissertation Abstracts International 40. 24328. Goldberg, C. (1989). The shame of Hamlet and Oedipus. Esxchoanalxtic Eerieu..lo(4). 581-603- Gough. H- G- (1975)- Manual of the Dalifornia Psychological Tnyenhefy. Palo Alto: Consulting Psychologists Press. (Original work published 1957) Hanlin, H. R. 8 Zucker, K. B. (1986). Affect and cognition in moral judgment. Deyeheiegieei Beneffe, 1133-1134. Harder, D. H. 8 Zalma, A. (1990). Two promising shame and guilt scales: A construct validity comparison. Tenfnei of Eersonalitx Assessment..§§(3&4). 729-745- Harder, D. W. (1990). Additional construct validity evidence for the Harder Personal Feelings Questionnaire measure of shame and guilt proneness. Deyeheiegieei Beneffe, DZ. 288-290. Harder, D. W. 8 Lewis, S. J. (1987). The assessment of shame and guilt. In J.N. Butcher and C.D. Spielberger (Eds.), acuances in Bersonalitx Assessment. Volume 6- Hillsdale, New Jersey: Lawrence Erlbaum Associates (pp. 89-114). Hathaway, S. 8 McKinley, J. (1967). anneal fer the Minnesota Multiphasic Personality Inrentorx- New York: Psychological Corporation. (Original work published 1948) Han, K. 8 Lew, J. F. (1989). Moral development of Chinese students in Hons Kong- International Journal of 25.22112102- 24. 561-569- 111 Helkama, K. 8 Ikonen, M. (1986). Some correlates of maturity of moral reasoning in Finland. Dehavior Seience EDDDDIDD. 22. 110-131- Hobzitelle, W. (1987). Differentiating and measuring shame and guilt: The relation between shame and depression. In H. B. Lewis (Ed.), The Beie pf §hame in Symptom Eernenien. Hillsdale, NJ: Lawrence Erlbaum Associates (pp. 207-235). Hoffman, M. L. (1976). Empathy, role-taking, guilt, and development of altruistic motives. In T. Lickona (Ed. ), Moral Derelonment and Deha_ior1 Theorx1 Research1a and Seeiei I§§D§§1 New York: Holt, Rinehart and Winston, (pp. 124- 143). Hogan, R. (1969). Development of an empathy scale. gounnai or consulting and clinical Psychology. 13. 307-316- Hultberg, P. (1988). Shame - a hidden emotion. Joupnai pf Analytical Eszcholccr. 13. 109-126- Isen, A. M. (1984). Toward understanding the role of affect in cognition. In R. S. wyer, Jr. 8 T. K. Srull (Eds. ), handhooh of social cognition1 _ol1 III Hillsdale: Lawrence Erlbaum Ass., (pp. 179- 236). Kaufman, G. (1989). Shenei The Dene; pf Defi_g. Cambridge: Schenkman. Kohlberg, L. (1964). Development of moral character and moral ideology. In M. Hoffman (Ed.), Beyien pf ghiid Beseefeh. Volume 1, New York: Russell Sage Foundation. Kohlberg, L. (1969). Stage and sequence: the cognitive- developmental approach to socialization. In D. Goslin (Ed-i. handbook of socialization Theorx and Research- Chicago: Rand McNally, (pp. 347-480). Kohlberg, L. (1976). Moral stages and moralization: The cognitive-developmental approach. In T. Lickona (Ed.), Moral Derelooment.and Dehaxiort Theorx1 Research1 and Seeiei Teenee. New York: Holt, Rinehart and Winston, (pp. 31-53). Kohlberg. L- (1984)- Essars on.horal Deuclooment1 Yolume II1 The Psycholocx of Moral Deuelooment- New York: Harper and Row. 112 Kohlberg, L., Levine, C., 8 Hewer, A. (1983). Moral stages: A current formulation and a response to critics, contributions to Human Derelonnent. l. Kugler, K. 8 Jones, W. H., (1992). On conceptualizing and assessing guilt- Journal of Bersonalitx and Social BSYcholch. DZIZI. 318-327- Kykendall, D., Keating, J. P., 8 Wagaman, J. (1988). Assessing affective states: A new methodology for some old problems. cocnitiue Theraor and Research. 12(3). 279-294. LaPlanche, J. 8 Pontalis, J. B. (1974). The Lengnege pf Beyehpzhneiyeie. New York: W. W. Norton 8 Company. LeCapitane, J. E. (1987). The relationship between emotional development and moral development and the differential impact of three psychological interventions on children. Psycholocx in the Schools. 24. 372-378- Lee, M. 8 Prentice, N. M. (1988). Interrelations of empathy, cognition, and moral reasoning with dimensions of juvenile delinquency- Journal of Abnormal child Bsucholccx 1m). 127-139. Lewis, H. B. (1971). Shene eng Qniif in Nenfeeie. New York: International Universities Press. Lifton, P. D. (1985). Individual differences in moral development: The relation of sex, gender, and personality to morality. Journal of Personality. 53(2). 306-334. Lickona, T. (1976). Critical issues in the study of moral development and behavior. In T. Lickona (Ed. ), Nerei De__loDment and EchleDI1 IDDDIY1 Researcn1 and Social Teenee. New York: Holt, Rinehart and Winston, (pp. 3- 28). Ma, H. (1985). Consistency of stage structure in objective moral judgment across different samples. Eeyeheiegieei M! £1 987-990 . Martin, R. M., Shafto, M., 8 Vandeinse, W. (1977). The reliability, validity and design of the Defining Issues Test. Dexeloomental Esxcholodx. 13(5). 460-468- Mehrabian, A. 8 Epstein, N. (1972). A measure of emotional empathy- Journal of Personalitx. to. 525-543- 113 Morris. W. (Ed )- (1971). The American heritage Dictionary pf the Engi_eh Lengnege (rev. ed.). Boston: American Heritage Publishing Co., Inc. Mosher, D. (1966). The development and multitrait- multimethod matrix analysis of three measures of three aspects of guilt. Journal of consulting Esrcholocu. so. 25-29. Mosher, D. L. (1968). Measurement of guilt in females by self-report inventories. genfnei pf Qeneuiting end clinical rsxcholodx. 1o. 25-29- Napier, J. D. (1979). Effects of knowledge of cognitive- moral development and request to fake on Defining Issues Test P-scores. Jenfnei pf gsyehoiogy,.igi, 45- 52. Nathanson, (1987). The Neny Eeeee pf Shene. New York: W. W. Norton. Piaget. J- (1965)- The Moral Judgment of the child. New York: The Free Press. (Original work published 1932) Propper, S. 8 Brown, R. A. (1986). Moral reasoning, parental sex attitudes, and sex guilt in female college students- Archixes of Sexual Dehauior. 15(4). 331-340- Puka, B. (1991). Interpretive experiments: probing the care- justice debate in moral development. finnen Deyeiepnent, D1, 61-80. Rest, J. R. (1975). Longitudinal study of the Defining Issues Test of moral judgment: A strategy for analyzing deve10pnental change- Derelopmental Demonology. 11(6). 738-748. Rest. J- R- (1979)- Derelooment in Judging Moral Issues1 Minneapolis: University of Minnesota Press. Rest. J. R- (1986a/1974) The Defining Issues Test Manual- Minneapolis: University Press. Rest. J- R- (1986h)- Moral Derelonmenti Aduances in Research eng Theefy. New York: Praeger Publishers. Rest, J. R., Cooper, D., Coder, R., Masanz, J., 8 Anderson, D. (1974). Judging the important issues in moral dilemmas - an objective measure of development. Dexelopnental BEYQthQQY. 12(4). 491-501- 114 Rest, J. R., Davison, M. L. 8 Robbins, S. (1978). Age trends in judging moral issues: A review of cross-sectional, longitudinal, and sequential studies of the Defining Issues Test. child Derelonment. 12(2). 263-279- Richards, P. S., Smith, S. A., 8 Davis, L. F. (1989). Healthy and unhealthy forms of religiousness manifested by psychotherapy clients: An empirical investigation. Journal of Research in Rersonalitx..21. 506- 524. Ryan, R. M. 8 Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasonsd for acting in two domains. denfnei ef Defeeneiityfi Social ESYQDDIQQY 51(5). 749' 751- Sagan. E- (1988). Freud1 Romen1 and Moralitxi The Rsxcholocx ef deed end Eyii. New York: Basic Books, Inc. Shelton, C. M. 8 McAdams, D. P. (1990). In search of an everyday morality: The development of a measure. Adolescence. 25(100). 923-943- Smith. R- L- (1972). The Relatixe Rroneness to Shane or .cuilt as an Indicationo of Defensi_e Stxle Unpublished doctoral dissertation, Northwestern University. Straughan, R. (1983). From moral judgment to moral action. In H. Weinreich-Hast 8 D. Locke (Eds. ), Neteiity in the Mahino1 Thought1 Action1 and the content Chichester: John Wiley and Sons LTD, (pp.125-140). Tangney, J. P. (1990). Assessing individual differences in proneness to shame and guilt: Development of the Self- conscious Affect and Attribution Inventory. dentnei ef Rersonalitx and Social Esxcholoox. 52(1). 102-111. Tangney, J. P. (1991). Moral affect: The good, the bad, and the ugly- Journal of Rersonalitx and Social Rsxcholodx. e1(4), 598- -607. Tangney, J. P., Burggraf, S. A., Hamme, H., 8 Domingos, B. (1988, March). Assessing Individual Diffetences in Eroneness to Shane and cuilti The Selfzconscious Affect and A_tribntien Ingentefy. Poster presented at the meeting of the Eastern Psychological Association, Buffalo, NY. Thompson, R. A. and Hoffman, M. L. (1980). Empathy and the development of guilt in children. Deyeiepnentei BDYQthDQY. 15(2). 155-156. 115 Tomkins, S. (1963). Affeett Imegety end Sonseiousness. New York: Springer. Tomkins, S. (1978). Script theory: Differential magnification of affects. In Nehte_he Synpeeinn on Metiyetien, H. E. Howe, Jr. and R. Dienstbiener, (Eds. ). Lincoln: University of Nebraska Press, 201-236. Tomkins, S. (1982). Affect Theory. In Enetiene in the finnen Eeee, Paul Ekman (Ed.). Cambridge: Cambridge University Press, 353-395. Van Voorhis, P. (1985). The utility of Kohlberg’s theory of moral development among adult probationers in a restitution field setting. geneticI Social, end ceneral RSYcDoIogu Monographs. 111(1). 101-126- Walker, H. A., Rowland, G. L., 8 Boyes, M. C. (1991). Personality, personal epistemology, and moral judgment. EEDQIES. DD. 757-772- Walker, L. J. (1984). Sex differences in the development of moral reasoning: A critical review. ehiid Dexeiepnent, 55(3), 677-691. Wise, P. S. 8 Cramer, S. H. (1988). Correlates of empathy and cognitive style in early adolescence. Eeyeheiegieei Reports-o1. 179- 192. Wright, D. (1983). ’The moral judgment of the child’ revisited. In H. Weinreich-Haste 8 D. Locke (Eds. ), Moralitx in the Makingi Thought1 Action1 and the Social Qentext. Chicherster: John Wiley and Sons LTD, (pp. 141-155). Zeidner, M. and Nevo, B. (1987). The cross-cultural generalizability of moral reasoning research: Somea Israeli data. International Journal of ESYEDDIDQL 315- -330. Zucker, K. B. and Barnett, D. (1977). The Eeifed Nende Teet Mendel. Dallas, TX: McCarron-Dial Systems. "7%W’E‘W“Tfl'iflfl'flilflflfliflr