[I L I B R A R Y Michigan State University This is to certify that the thesis entitled EXPLORING THE RELATIONSHIPS BETWEEN CHILDREN'S SOCIAL BEHAVIORS AND THEIR FANTASIZED COPING ACTIVITIES presented by Nancy Katherine Nowkanech has been accepted towards fulfillment of the requirements for TA D degree in (RSIMLQQ‘? l )' (47/ .II 1’ I / I, 41’ Q41 “4 , ' Maj r professor / 3/,4 /Z7 0-7639 . 4-, _—-——-.————.—-“-_i _._ _ ABSTRACT EXPLORING THE RELATIONSHIPS BETWEEN CHILDREN'S SOCIAL BEHAVIORS AND THEIR FANTASIZED COPING ACTIVITIES By Nancy Katherine Nowkanech The present study examined how differences in children's preferred Inode of self-representation in fantasy (i.e., identification with "Provider-Protector", "Victim", or "Aggressor" roles), fantasies of parental discipline, and behavioral reactions to fantasized transgres- sions were related to Baumrind‘s (1967) dimensions of "social competence". Over two hundred children (median age range 4 years 6 months to 4 years eight months) from seven area day care centers were rated by their teachers, as "high", "medium—high", "medium-low", or "low" on five categories of classroom behavior: “self-control", "self-reliance", “approach-avoidance“, "mood“, and "peer affiliation". Factor analysis of the ratings resulted in 117 children being assigned to one of four groups: Group I ("Competent/Compliant"), Group II ("Self—Involved/ Disruptive"), Group III ("Competent/Independent"), and Group IV ("Impulsive/Withdrawn“). Children's responses to two structured fantasy tasks, a modified miniature situations test (MST) developed by the author, and Deviation Doll Play (Sears, Rau, and Alpert, l965), were recorded. The modified Nancy Katherine Wowkanech MST consisted of l8 different items, in which the child chose to identi- fy with the “Provider-Protector", "Victim", or "Aggressor" role, by acting out a given fantasy situation with a toy animal. Deviation Doll Play consisted of five typical parent-child conflict situations in which the same-sexed child doll comitted a deviant act, and the subject was instructed to "show and tell what happened next." Children's perceptions of parental discipline, symbolic agent of punishment, and behavioral reactions to fantasized transgressions were recorded. Multivariate analyses of variance yielded significant differences in coping behaviors between two groups only: 1) "Competent/Compliant" children (Group I) predominantly identified with "Provider-Protector" roles, perceived symbolic parental figures as using "Reasoning" respon- ses to deal with fantasized transgressions, and "Apologized" for deviant acts. 2) "Impulsive/Withdrawn" children (Group IV) predominantly identified with "Aggressor" roles, and diSplayed "Continued Deviant Aggressive" behavior in response to fantasized transgressions. Significant sex differences were also found: 1) Females significantly preferred to identify with the "Provider-Protector" role in fantasy, and gave significantly more "Spontaneous Fixing" responses to fantasized deviant acts, than did males. 2) Males chose the same-sexed parent as the symbolic primary agent of punishment, significantly more than did females. Nancy Katherine Nowkanech Problems with reliability and validity of the modified MST and suggestions for more appropriate item construction were also discussed. Implications for future research were given. Approved: CommitteeiCo-Ehairman Committee Co-chairman Date: EXPLORING THE RELATIONSHIPS BETWEEN CHILDREN'S SOCIAL BEHAVIORS AND THEIR FANTASIZED COPING ACTIVITIES By Nancy Katherine Nowkanech A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1977 DEDICATION To My Fa/thejt'a Meme/Lg ii ACKNOWLEDGEMENTS First I wish to thank all members of my committee for their patience over the last three years. To Gary Stollak, who served as the Co-chairman of my dissertation committee, I wish to express my profound respect for his scholarly con- tributions to all aspects of this research. More importantly, I have had the invaluable opportunity of knowing Gary over the past seven years, and he has had the most significant impact upon me, both as a person, and as a clinical psychologist. Words cannot express the deep admiration I feel for Gary as a scholar, teacher, advisor, and friend. To Larry Messe, who also served as Co-chairman of my dissertation committee, I wish to express my profound appreciation for his super- vision of all aspects of this research, which resulted in an invaluable learning experience for me. The degree of scholarly dedication and investment that Larry gave to the supervision of my research, is unparalled in my previous learning experiences. Besides my appreciation, and respect, he has my deep thanks for being one of the most kind, sup- portive, and sincere people I have ever worked with. To Bob Zucker, who served as a committee member, I appreciate his valuable contributions toward understanding this research. During the course of my dissertation, and while working with Bob at the Psychological Clinic, I have always found that interactions with him lead to increased theoretical and clinical knowledge that benefits me. To Ellen Strommen, my sincere thanks for serving on my committee, and for her scholarly suggestions and careful editing of this disserta- tion, during all of its stages. I feel that her contributions as a committee member helped make this an important learning experience for "Eh ‘ Bob Wilson deserves much credit for helping with the data analysis. I also wish to thank Kay Guimond and Joe Pizzo, two undergradu- ates in psychology courses I was teaching during the summer of l974, for serving as observer-raters. iii Finally, I wish to express my appreciation to the children, parents, and teachers of the following area day care centers, who helped make this study possible: All Kids Day Care, Capitol Child Development Center, Eastminster Day Care Center, First Nazarene Child Care, Golden Rule Nursery, and both Michigan Young World Centers. At this time, I wish to acknowledge and express my gratitude to some of the faculty members who have contributed significantly to my graduate training at Michigan State University. I feel very inadequate just listing these people; however, I trust that each knows what I accomplished or experienced with them, and whatever I do in the future certainly will reflect their impact upon me: Drs. Mary Leichty, Lucy Ferguson, Bert Karon, Albert Rabin, Art Seagull, and last, but definite- ly not least, Ms. Marsha Worby. Now for the people-behind-the-scenes! Without my friends, I would never have made it! Becky Hollingsworth deserves this Ph.D. as much as I do! Her unabounding thoughtfulness, kindness, and warmth makes me cherish her friendship. Jeanne Burnett served as the source of my inspiration. The enjoyment I derive from her friendship continually renews me. Finally, Pam and Norm Shaughnessey are my "Rocks of Gibralter”. For the past ten years, I kept hoping some of Pam's stabil- ity would rub off on me. I keep trying! My friends have been running "hotel“, "restuarant", "easy credit", and "limousine" services for some time. At least they can say it's never dull knowing me--you never know when I might drop in! Most importantly, my family deserves the credit of encouraging, supporting, and helping me to reach this goal. Their love is what I value the most. Without them, coming to Michigan State University, ten years ago, would not have been a reality. --NKW *********** iv TABLE OF CONTENTS d CHAPTER Page I. INTRODUCTION ............................................. Self-Representation in Fantasy ........................ 2 Children's Perceptions of Parental Discipline ......... 4 Superego Development .................................. 6 THEORY AND REVIEW OF THE LITERATURE ...................... 8 Identification ........................................ 8 Identification with Provider-Protector, Aggressor, and Victim Roles ................................... ll Relevant Studies on Children's Fantasy Behavior ....... l5 Measurement of Children's Choices of Self-Representa- tion in Fantasy .................................... 20 Children's Perceptions of Parental Behavior ........... 24 Parental Discipline and Moral Development in Young Children ........................................... 28 QUESTIONS AND HYPOTHESES ................................. 33 II. METHOD.. ................................................. 36 Subject Selection ..................................... 36 Formation of Groups ................................... 37 Sample Characteristics ................................ 41 Design ................................................ 4l Dependent Measures .................................... 42 Procedure-~Task A: Modified Miniature Situations Test (MST) ...................................... 42 Procedure--Task B: Deviation Doll Play ............ 45 III. RESULTS .................................................. 49 Inter-Rater Reliabilities of Dependent Measures ....... 49 Internal Consistency of Modified MST .................. 50 Category Usage ........................................ Sl Self-Representation in Fantasy (modified MST) ...... 5l Perceived Parental Response to Fantasized Trans- gression ........................................ 52 TABLE OF CONTENTS-~continued CHAPTER IV. Child's Behavioral Reaction to Fantasized Trans— gression ........................................ Agent(s) of Fantasized Punishment .................. Factor Analysis of Dependent Measures .............. Multivariate Analyses of Variance ..................... Main Effects ....................................... Interaction Effects ................................ Analysis of Variance of Victim Dependent Measure ...... Factor Analysis of Dependent Measures ................. Factor I: Competence-Incompetence ................. Factor II: Child's Extreme Behavioral Reactions to Fantasized Transgression ........................ Factors III and IV: Agent(s) of Punishment ........ Factor V: Perceived Parental Response to Fantasized Transgression ........................ Factor VI: Fixing ................................. Factor VII: Identification with the Victim Role... Factor VIII: Denial ............................... DISCUSSION ............................................... Relevance of Results for Questions and Hypotheses ..... Relevance of Results for Past Theory and Research ..... Sex Differences on Fantasy Measures... ................ Relationships Between Fantasy Measures and Teachers' Ratings of Overt Social Behavior ................... Qualification of Results and Directions for Future Research ..... .. .................................... Methodological Issues ................................. Teacher Ratings of Children's Social Behaviors ..... Formation of Groups ................................ Dependent Measures ................................. Task Administration ................................ Sample Characteristics ............................. Conclusions and Implications of Findings ........... BIBLIOGRAPHY .................................................... APPENDICES A. B. C. LETTER TO PARENTS ........................................ LETTER TO TEACHERS ....................................... BEHAVIORAL DEFINITIONS OF TEACHERS' RATING SCALES AND TEACHER RATING FORM ...................................... FACTOR ANALYSIS OF INDEPENDENT MEASURES .................. vi 133 I36 I37 I44 TABLE OF CONTENTS--continued APPENDICES Page E. SAMPLE CHARACTERISTICS ................................... 146 F MODIFIED MINIATURE SITUATIONS TEST ....................... 148 G. DEVIATION DOLL PLAY ...................................... 151 H DEFINITIONS OF SCORING CATEGORIES FOR DEVIATION DOLL PLAY 154 I. MEANS AND STANDARD DEVIATIONS OF CATEGORIES .............. 160 J. INTER-RATER RELIABILITIES 0F DEPENDENT MEASURES .......... l62 K. INTERNAL CONSISTENCY OF MODIFIED MST... .................. 154 L. FACTOR ANALYSIS OF DEPENDENT MEASURES .................... 167 M. TESTS OF SIMPLE EFFECTS .................................. 172 vii LIST OF TABLES TABLE l. Summary of Significant Multivariate Results Reflecting Significant Univariate Comparisons not Qualified by Higher- Order Effects ................. . Means and Standard Deviations . Summary of Significant Univariate Results not Qualified by Higher-Order Effects .......... of Variables that Yielded Significant Main Effects for Groups ........................ . Means and Standard Deviations of Variables that Yielded Significant Main Effects for Sex of Subject ................ Means and Standard Deviations Classified by Group and Sex... . Means and Standard Deviations Classified by Group and Sex... Means and Standard Deviations Classified by Group and Order Means and Standard Deviations within Order B, Classified by Means and Standard Deviations fied by Group, Sex, and Order Means and Standard Deviations fied by Group, Sex, and Order of the Reasoning Measure of the Ignoring Measure of the Apology Measure of Task Presentation ......... of the Aggressor Measure, Group and Sex ................ of the Denial Measure Classi- of Task Presentation ......... of the Victim Measure.Classi- of Task Presentation ......... viii Page 55 56 58 60 63 65 67 7O 72 75 CHAPTER I INTRODUCTION The purpose of this study was to examine the interrelationships between children's (ages 4-5) overt social behaviors, and their fantasy expressions of related ego and superego functions. More specifically, this study focused on exploring relationships between Baumrind's (l967) dimensions of the Child‘s maturity, competence, and independence, i.e., "self-control, self-reliance, approach vs. avoidance tendency, subjec- tive mood, and peer affiliation", and dimensions of the child's inner fantasy life, i.e., mode of self-representation, expectations of parental punishment for fantasized deviant behavior, and anticipation of his/her own behavioral reactions to fantasized transgressions. Baumrind (l967) and Baumrind and Black (l970) differentiated among three patterns of child behavior: Pattern I (children who demon- strated self-control, self-reliance, explorative behavior, and content mood), Pattern II (children who were more discontent, withdrawn, and distrustful), and Pattern III (children who lacked self—control, self- reliance, and retreated from novel situations). Baumrind (l967: l97l: 1973; l975) and Baumrind and Black (1970) studied differences in parental attitudes and behaviors, and related these characteristics to differences in Child behaviors (notably, aspects of parental warmth and control). However, the impact of parental attitudes and behaviors upon children cannot be sufficiently investigated without exploring the child's "inner representation of his/her outer world"; i.e., aspects of his/her fantasy life revealing his/her thoughts, needs, wishes, feel- ings, desires, and fears. Since this study attempted to investigate aspects of young children‘s fantasy life, the author acknowledges a psychoanalytic bias, in conceptualizing the pertinent theoretical issues, and in interpreting the results. Self-Representation in Fantasy According to Freud (1920), the ego is the organizer of the child's perceptions of the world and himself/herself. The ego's capacity for self-perception derives from its achievement of object constancy, which initially involves hallucination of the gratifying object, later re- sulting in identifying imitation, the forerunner of superego development (Fenichel, 1945). Hartmann (1964) points to the significance of early pre-oedipal identifications as providing barriers from external and internal sources of threat; Sandler (1960) sees early self-representations as a source of narcissistic gratification. Clinical investigations of fantasy behavior (Jersild et al., 1933, 1935; IsSacs, 1933; Murphy, 1962; Pitcher and Prelinger, 1963; Ames, 1966, and Singer, 1973) have indi- cated that preschool children in dramatic play or story-telling reveal wishful or evaluative aspects of the self as "good", "evil", “naughty", “scary", "independent", etc. Recently, Gould (1972) delineated three types of primary identifi- cations in preschoolers' spontaneous fantasy play behavior; Identifica- 'tion with (1) Provider-Protector, (2) Aggressor, and (3) Victim, roles. Theoretically, identification with the provider-protector reflects the child‘s early experiences with adult caretakers as nurturant, protec- tive, affectionate, etc., leading to a basic sense of trust (Erikson, 1950), or sense of personal autonomy or "competence" (White, 1959). Hartmann, Lowenstein, and Kris (1946) feel that the ego's capacity to expand its autonomous functions develops out of its ability to control or neutralize sexual/aggressive impulses; Freud's (1920) original con- cept of sublimation. Among the vicissitudes of instinctual drives that were described by Freud (1920, 1933) was the "turning of aggression upon the self", a forerunner of the superego's capacity for self—criticism. Hartmann (1964) states this reflects the ego's inability to neutralize sexual/ aggressive energy, thereby lessening the effectiveness of the ego's coping mechanisms. The ego's mode of defense against internal threats could be to transpose active strivings into passive ones, corresponding to Gould's (1972) identification with the victim roles. Gould also hypothesizes that early disruptions in attachment lead to conflicts in- volving dependency needs and individuation with the child's role as victim being a compromise between suppressed rage, and dependency motives. Boys' identification with the aggressor was first discussed by A. Freud (1936) as a defense against castration anxiety.’ Gould relates identification with the aggressor to the child's predominant concern with negative caretaking experiences; i.e., feeling abandoned, rejected, unfairly punished, etc., although similar fears may operate in the child who identifies with the victim. However, identification with the aggressor indicates a relatively immature ego, whose anticipatory stance of attack, to maintain self-preservation, renders the ego in- competent, in the sense that acting-out of aggressive impulses brings retaliation, reinforcing basic mistrust in one‘s environment (Erikson, 1950). Thus, the child's choice of self-representation in fantasy theo- retically will reflect not only his/her previous caretaking experiences, but also the ego‘s capacities for defense against aggressive/sexual impulses. These defense capabilities, in turn, influence the relative autonomy of ego functions, since they conserve psychic energy that can be used fbr active mastery and for the learning of behaviors involved in establishing psycho-social competence; i.e., “self-control", "self- reliance", and “peer affiliation“, etc. Children‘s Perceptions of Parental Discipline In addition, the differential basis of the ego's primary mode of defense, in the case of identification with "victim" vs. "aggressor", may be studied more clearly by examining the Child's anticipations of parental punishment in response to fantasized deviant behavior. The perceived mode of punishment seems to be the critical variable, e.g., identification with the victim may be related to love-withdrawal techniques; fear of loss of love may lead to increased dependency behaviors. Identification with the aggressor may be associated with fear of physical harm or power-assertive techniques, reflecting a more pervasive fear of annihilation, resulting in withdrawal/retaliatory behaviors. Induction and reasoning approaches theoretically should correspond with identification with provider-protector roles, since these methods help the Child to understand cognitively and affectively the consequences of his/her behavior, and to provide stability and predictability in the expression and control of impulses. Incongruence of children's perceptions of parental punishment with parents' actual behavior may indicate as Freud (1924), Hartmann (1964) and others sug- gest that the child‘s defensive identification or incorporation of the loved object (parent) mainly involves the projection of his/her own hostility onto the parent, thereby increasing "superego severity". Children in the present research were, according to psychoanalytic theory, in the throes of working through Oedipal conflicts, and estab- lishing identification with the parent of the same sex. Therefore, it was also interesting to examine whg_was the primary agent of fantasized punishment: same-sexed parent vs. opposite-sexed parent. Or, if the child perceived both parents as about equally punitive, did S/he per- ceive them primarily as supportive or conflicting, in the messages that they conveyed to the child. Goldin (1969) reported that children‘s perceptions of agent(s) of punishment were extremely contradictory, and suffered from a lack of precise delineation of personality char- acteristics and behaviors of the Children being studied; this investi- gation attempted to provide some clarity in that respect as well. Superego Development Gould (1972) feels that the child's early identifications with her/his primary care-taker authority figures as well as the child's relative stability of acquired impulse-controls and the affectional pulls that are implicit in the child's struggle to meet the criteria of "good" represent the bases for formulating a concept of "superego constancy". Ego psychologists, such as Hartmann, Lowenstein and Kris (1946), and Jacobsen (1964), emphasize the importance of studying pre- oedipal manifestations of superego formation, since they are integral components of the developmental processes that effect internalization of drive-deterrence, ego-ideals, self-evaluation, and perception of self-constancy. Gould (1972) states that a fundamental relationship exists between positively charged provider-protector experiences in infancy and later "superego constancy", in that the former generate an elemental capacity for empathy and remorse-anxiety in the child, helping him/her to manage early primitive destructive impulses. The positive experiences of early attachment dependency also permit the Child to integrate his/her super- ego introjections with his/her developing self-concept, thereby pro- viding the basis for post-oedipal ownership of internalized moral dicta and value. The behavior of acting-out children especially those of whom manifest identification with aggressor roles reflects the inadequate attachment-dependency experiences with early caretakers that fosters preoccupation with rage fantasies, omnipotent thinking, and feelings of absolutism of own powers and invulnerability. This preoccupation inter- feres with appropriate internalization of prohibitions concerning aggressive impulse control, along with disruption of affects associated with “superego constancy“; i.e., “guilt“, "sadness", "shame", etc. Discontinuities in, or temporary separations from, positive, nurturant, attachment experiences with primary caretakers are likely to result in a child experiencing both a devaluation of the core me-self, and recurrent feeling sequences of mad—bad-sad. These experiences, in turn, provoke variability in "superego constancy", and a defensive stance that are congruent with identification with the victim role. Thus, a review of the literature indicates that within the frame- work of psychoanalytic personality theory, the concepts of "intention- ality", “guilt“, “shame", "fear“, etc., as well as the parental ante- cedents of moral development in children have been widely discussed and studied. However, with the exception of the research by Sears, Rau, and Alpert (1965), Wurtz (1959), and Burke (1961), the preschool Child's behavioral reaction to fantasized transgression as a precursor of “superego constancy" largely has been ignored. The 4-5 year-old child typically does not possess the cognitive ability to understand “intentionality” (Piaget, 1965; Kohlberg, 1966), and generally his/her level of verbalization is not sophisticated enough to discriminate meaningfully between “fear", "shame" and “guilt“. Thus, an indirect, but potentially highly valid measure of rudimentary superego functions in the 4-5 year-old child would involve assessing his/her anticipated behavior in response to fantasized transgressions. This study thus attempts to examine systematically 4-5 year-old children's fantasy expressions that are related theoretically to ego and superego functions. More specifically, this research has, as its primary focus, this question: Would children similar to Pattern I, II, and III children as defined by Baumrind (1967; 1971; 1973), differ on such variables as: (1) choice of self-representation in fantasy, (2) perception of agent(s) and mode of punishment following fantasized deviation, and (3) behavioral reaction to fantasized transgressions? THEORY AND REVIEW OF THE LITERATURE Identification The process of identification is theorized to be a collection of mechanisms by which children develop attributes and behavior patterns similar to those of their parents and other significant social models. Freud (1924) discussed identification in terms Of the process respons- ible for the establishment of the ego-ideal, and later, the mature superego, or Conscience. Furthermore, Freud postulated two main types of identification mechanisms; anaclitic identification, and identifica- tion with the aggressor. The basis of anaclitic identification was the infant‘s intense dependency upon the mother. Later, during the first 3-4 years of life, children, especially girls, attempt to recapture the "mother's nurturance" by imitating her behaviors and reproducing her image in fantasy. Bandura and Walters (1963) refer to this simply as observational learning: learning by imitating a model who has strong reinforcing properties of nurturance and control over resources. In any case, a child by reproducing his/her mother can provide a substitute for her, when she begins withdrawing affectionate interaction and nur- turance from the child. Freud (1924) also hypothesized that in boys, anaclitic identifica- tion with the mother is later supplemented by identification with the aggressor, i.e., the potentially castrating father, subsequent to reso- lution of the Oedipal Complex. This secondary form of identification is thought to be a defensive process, in which the anaclitic identifi- cation has already engendered an internalization of the punitive and restrictive qualities of a threatening parent, which become activated by the boy's ambivalent feelings towards the father (A. Freud, 1936). Resolution of the Oedipal conflict is achieved through identification with the father. The psychoanalytic view of identification as a "basic human ten- dency towards unification with others", either serving as a means to incorporate a lost object or as a defense against loss, is also an ego mechanism which may permit the child to learn new social behaviors, and thus, may involve the development of other ego functions-~"sex typing, adult role formation, self-control, self-recrimination, prosocial forms Of aggression, guilt feelings, and resistance to temptation" (Sears, Rau, and Alpert, 1965, p. 1). Piaget (1937) uses identification in a descriptive sense, refer- ring directly to self-representations exhibited in fantasy play. Beginning around the age of 21 months, Piaget postulates that the 10 - child's identificatory strivings in fantasy play advance through several stages: 1) "direct, imitative representations of other persons or objects,"2) "transformation of the child's observations 'dissociated' from its own or other's behavior," and 3) "description by the child of its own actions in speech, before overt behavior occurs" (p. 37). Thus, Piaget sees imitation and identification as important processes of cognitive growth, especially with respect to the child's changing aware- ness of intentionality. Most relevant to the present study, Gould (1972) regards the child's ability to transform selectively what s/he observes from its earlier copying behavior as an indicator of increasing differentiation of the self system. Gould believes that the child's awareness of the self as a distinct causal agent is a necessary but not sufficient skill related to the development of a sense of "personal autonomy" or “competence“ (White, 1959). Thus, self-representations in fantasy can be interpreted as sus- taining the child‘s development of "object constancy" by freeing him/her from dependence on the actual presence of the "mothering one", to main- tain a sense of well-being and security. Also, the ability to engage in self-imitative fantasy is an indicator of the child's increased awareness of self vs. others. Differentiation of the self-system (e.g., good vs. bad, strong vs. weak, big vs. little) takes place conjointly with differentiation of internal object representations (good mother vs. bad mother, etc.). Lastly, the child's intentional impersonation of others conveys his/her predominant affective strivings to emulate and incorporate the behaviors and attributes of "significant others". 11 Identification with Provider-Protector, Aggressor, and VTCtim Reles As previously mentioned, Gould (1972) found preschool-aged chil— dren (3-5) manifesting consistent modes of self-representation in spontaneous fantasy play behavior across time; i.e., identification with provider-protector, aggressor, and victim roles. Identification with the provider-protector role, e.g., "I'm a nurse", was associated with the child's use of distance forms as a defense against threatening impulses; e.g., "I'm a nurse and I'm going to give you a needle"; with the ability to accurately distinguish between real and pretend; with a limited (benign superego) expression of self-condemnation; with a sense of entitlement and a wish to please ("superego constancy"), and an ability to express creatively elaborated aggressive themes in fantasy, using it as vehicle for catharsis and reality adaptation. On the other hand, identification with the victim or aggressor role was associated with the child's use of direct forms of self- representation in fantasy, e.g., "I shot you"; "I'm dead"; with fluctu- ating certainty (inability to distinguish between real and pretend in the face of threatening impulses, denoting anxiety-interference); with global self-condemnation (harsh superego), and with a tendency to act out aggressive impulses or regress to infantile states of helplessness, reflecting lower levels of self-esteem, personal autonomy, and inade- quacy of "superego constancy" as a drive-deterrent. Gould views the "core me-Self" as the basis from which identifi- cation and self-representation.in fantasy, derive. It is a dynamically 12 segregated system evolving out of the Child's experiences with primary caretakers which includes body-image schemas, and schemas related to expression and control of impulses. Gould sees the evolution of a person's sense of self as the "core me-self" infused with a basic cast of "good" vs. "bad", "strong" vs. "helpless", which later becomes inte- grated with representations of the self as agent, achiever, or object for another. These later self-images are "I-self" images, which may be congruent or incongruent, with the basic core "me-self". Thus, the primary mode of identification, whether provider-protector, aggressor, or victim, is mediated by early images of oneself as either “good", "bad", "unworthy", “strong", "helpless", etc. Furthermore, the mode of self-representation is also mediated by the child's use of fantasy as a defense vs. a coping mechanism. Gould points out that the instigat- ing feeling-state of defense is anxiety, rooted in a fear of being over- whelmed, losing control, being hurt or humiliated, etc., while feeling challenged to mastery is rooted in the child's sense Of being able (through his own efforts), to control, to overcome. As A. Freud (1936; 1965) notes, in defense the aim is protective and restitutive, while in "coping“, the aim is expansive, towards mastery and'achievement. Identification with provider-protector roles is associated more with "distance defenses“, which reflect the child's achievement of a structural capacity for experiencing internal conflict about aggressive impulses, and a means for containing them and actively coping with them, as well as revealing an affective orientation towards mastery of fears, wishes, and/or needs. 13 Identification with victim or aggressor roles, is associated more with "direct defenses", indicating the Child's vulnerability to dedif- ferentiation of the self—object system under the pressure of aggressive impulses and anxiety; revealing affective strivings of self-preserva- tion and protection vs. mastery. In addition, fluctuating certainty (inability to distinguish between real and pretend) is more characteristic of Children identify- ing with aggressors or victims, and reflects the inadequacy of the ego defense mechanisms to cope with aggressive imagery, which in turn dis- rupts cognitive functioning, i.e., reality testing. A heightened selectivity of aggressive perceptions of causality and a primitive mode of magical thinking persists in these children, as compared with those who identify with provider-protector roles and can accurately dis- tinguish between real and pretend, possessing the capacity to use fantasy for the cathartic and adaptive expression of aggressive impulses. Also, the tendency towards global self-condemnation in children who identify with aggressors or victims reveals their preoccupation with "talion concerns". "Talion concerns" refer to fears of "punish- ment which exact a penalty similar to the crime," i.e., "an eye for an eye“; "a tooth for a tooth"; reflecting a level of guilt-anxiety and magical thinking, where intrinsic justice is met with retaliatory reversal. Proneness to talion concerns may represent the most unde- lineated self-object expression of condemnation and justice; the most primitive form being, defensive identification with the victim. 14 Limited self-condemnation, e.g., “I am wrong in hitting the baby," associated with provider-protector identifications refers to criticism of specific acts comnitted by the child, vs. his/her self concept as a wholes e.g., "I am a terrible boy. " This capacity for more focused criticism reflects a highly differentiated self-system, and more mature. realistic perceptions of punishment and justice, steaming from a con- comitant "wish to please." and "a sense of entitlement." Thus, Gould (1972) suggests that the choice of self-representa- tion in fantasy play reflects both cognitive and emotional resources available to the child, for cartharsis, and reality-adaptation. A child‘s spontaneous fantasy is an intrinsic response of its mental growth and creativity as well as an individualized expression of its developmental dilemas, and comprehension of significant features of its reality. Past the age of three, a child who is unable to project her/himself in fantasy expression as a person or animal is signalling interferences in developing clarity of core "me-self“ delineation, as well as in the positive experiences with other people, that mediate this delineation. Thus is the case of "acting-out" children who are Preoccupied with aggressive imagery, wherein one sees the projection 0f the self as victim or aggressor, depending on the direction of the a99ression, while in identification with provider-protector roles, nurturant imagery is predominant. 15 Relevant Studies on Children's Fantasy Behavior While studies of children's fantasy play behavior have been sparse, findings derived from doll play research do reveal important information on aggressive fantasy play behavior, sterotypy of fantasy reessponse, and the influence of adult participation in the child's fan- tasy. With respect to aggression, the most documented finding is that boys are more aggressive than girls in doll play (Sears, Rau, and Alpert, 1965). However, Johnson (1951) adds that boys do exceed girls in physical aggression, but that girls show more verbal aggression. Thus, children of both sexes may identify with the role of aggressor in fantasy, while the type or kind of expressed aggression may be the variable which differentiates them. In addition, Johnson's findings Wit}! regard to age of subjects is also important. Younger children distilay contrasocial aggression, i.e., male doll hitting baby brother, whi'le older children show "prosocial aggression," i.e., depictions of Parents punishing children. These findings could be seen as reflecting the influence of superego development on the expression of aggression 1" fantasy and are supported by additional studies (Sears, 1961; Yartwmw, 1948; Hollenberg and Sperry, 1951; Siegel and Kohn, 1959; Sears, Rau, and Alpert, 1965). Another substantial finding with respect to aggression in doll PIAS'is the consistent result that children are more aggressive in the second vs. the first session of doll play (Hollenberg and Sperry, 1951; Levin and sears, 1956; Sears, 1961). Pintler (1945), using three sessions, found that the latency of the first aggressive act decreased 16 in the later sessions. The child appears to learn with time that re- straints against aggression are not operative in doll play, and thus gives vent to his/her impulses more freely. Pintler also found that the presence of an adult can facilitate the child's expression of aggression; this finding is supported by Reif and Stollak (1972). Levin and Sears (1956) suggest that doll play aggression is associated with the sex of the child, the real life agent of punishment, the nature of the child's identification with the parent, as well as the severity of punishment meted out to the child. In general, however, child-rearing antecedents of doll play aggression are unclear, due to the confounding of replication vs. wishfulfillment variables. Herron and Sutton-Smith (1971) argue that information concerning anxiety attached to expression of aggression at home, amount of aggressive behavior exhibited at home, and degree of instigation to aggression, is critical in determining whether or not aggression in doll play is re- lated to replication or wishfulfillment fantasies. Findings on sterotypy of response (i.e., the degree to which dolls have been used appropriately and in routine actions and situa- tions), indicate that girls Show more sterotyped responses than boys; fUrther, the amount of sterotyped responses decreases with more sessions (Levin and Wardwell, 1962), and high levels of experimenter interactions are associated with nonsterotyped responses (Pintler, 1945). Reif and Stollak (1972) add that "responsiveness" and "permis- siveness“ on the part of adults in a play room setting could encourage Achildren's expression of fantasy play in general, specifically fantasy 17 aggression and fantasy behaviors involving more complexity; such as, "adult role behaviors, affect expression, interpersonal awareness, and symbolic expression of conflict,“ which they suggest reflect processes of achieving "ego mastery". Furthermore, they found that children who interacted with "non-responsive" and "non-permissive” adults displayed lower levels of fantasy behavior, including more sterotyped responses, greater use of inanimate objects in fantasy expression, and displayed more disorganized motor behavior. The latter finding on disorganized behavior is supported by Freyberg (1973), and Biblow (1973). With regard to doll preferences Ammons and Ammons (1949), Lynn (1955), Emmerich (1959), and Mussen and Distler (1959), all show support for the hypothesis that young children tend to identify more with par- ents Of the same sex vs. parents of the opposite sex; same sexed parents are depicted more frequently as agents of nurturance, punishment, and protection. The work of Singer and his colleagues (Singer, 1973) sup- ports the notion that children who readily engage in fantasy play also report greater closeness or identification with their mothers, as well as greater contact with both parents. Singer found that "high-fantasy" children orient themselves to one parent, enhancing identification possibilities but added that nothing is known about the nature or mode of'identification. Neither did he find that "high" and "low“ fantasy subjects differed in reports of type and agent of punishment. He sug- ges ts that the "overprotected" child who is constantly under adult scrutiny, or smothered by adult attention, is usually dependent on adults. and does not use fantasy to express and control impulses, or 18 achieve "ego-mastery“. Thus, the gggljty_of parent-child interaction, not quantity, is important in determining the child's mode of identifi- cation and related fantasy skills. In addition, it may be more meaning- ful to differentiate children on the basis of demonstrated social behaviors; such as, "peer affiliation", vs. “high" and “low“ fantasy production, in order to explore differences in perceptions of type and parental agent of punishment. In conclusion, it seems likely that strong identification with at least one stable parent figure, usually of the same sex, who is also generally “permissive, nurturant“, vs. “cold, restrictive", is conducive to the promotion of the child's fantasy expression as a means of achiev- ing "catharsis" and/or "ego mastery". Children who are overly dependent or lack a stable attachment to one parent would have more difficulty in utilizing fantasy to express and control impulses. Their fantasies may be more involved with inanimate objects, involve less cognitive and affective complexity, and show more sterotyping of responses. Sex dif- ferences in fantasy aggression are chiefly in mode of expression, past research indicating: (1) boys tend to express physical aggression, while girls rely more on verbal aggression, (2) girls show more stero- typy of responses, (3) younger children display "contrasocial aggres- sion", while older children exhibit "prosocial aggression", and (4) younger Children also tend to identify with the same-sexed parent in doll play, portraying the same-sexed parent as the agent of punish- ment, nurturance, and protection. 19 While Singer's work (1973) did not reveal the precise nature of the child's identification with the parent in fantasy, Gould's (1972) delineation of provider-protector, aggressor, and victim self- representation provides a meaningful way to examine further preschool age children's identifications. In light of Gould's previously dis- cussed research, it seems logical to specify more clearly and opera— tionalize personality characteristics of Children who choose to identify with each role. This study attempted to assess the relationship between mode of self-representation in fantasy and dimensions of social compe- tence specified by Baumrind (1967; 1971). Baumrind found that the most “mature", “independent“, and "socially responsible children" (Pattern I) had parents who were both warm and controlling ("authoritative"). Thus, it may be hypothesized that "more mature" children would choose to identify more with provider-protector roles in fantasy, since nurturant and protective behaviors typically would be considered as “parenting behaviors“. Baumrind's "dysphoric and disaffiliated“ children (Pattern II) had parents who were detached and controlling ("authoritarian"). Depending upon how control was achieved by parents, it could be hypothesized that these children might be more inclined to identify predominantly with aggressor or victim roles in fantasy play; i.e., either modelling power-assertion or aggressive techniques of discipline exhibited by parents, or viewing themselves as victims of parental aggression. Baumrind's “immature" children (Pattern III) had parents who were non-controlling, non-demanding, i.e., "permissive“, and who were relatively warm. Possibly, "immature" children might also 20 choose to identify with aggressor roles, since they may lack internal impulse controls, due to parental laxity in enforcing limits. Measurement of Children's Choices of TSElfirepresentatTon in Fantasy Another method conducive to investigating children's fantasy behavior (other than doll play), is the Miniature Situations Test (MST) devised by Santostefano (1960, 1962a, 1962b, 1965, 1966. 1968). The MST is based on the forced choice (Rundquist, 1950) method, which presents the subject with paired stimuli designed to evoke "coping behaviors". The paired tasks require the child to "act upon, manipulate, or avoid objects presented"; e.g., “child refuses to strike glass vs. child strikes glass with a hammer." Further, Santostefano hypothesized that the tasks had sufficient "psychological impact to trigger affective reactions in subjects, revealing personality Characteristics, through stimuli the subject valued vs. rejected, and through the kinds of changes he/she was willing to impose on the objects" (1960, p. 264). Various motive expressions, e.g., "nurturance, aggression, dependence, dominance, exhibitionism, orderliness," etc., were studied. Santostefano (1962b) believes this method is especially effective with younger children, enabling them to express their "personality in action", as the situations require the Child to mobilize various defenses and adjustive reactions to handle impulses and subsequent fantasies, elicited by the stimuli. 21 Construct validity for the MST is provided by Santostefano (1962a, 1965, 1966, 1968). This test effectively discriminated between brain- damaged children and public school children; brain-damaged children who were characterized by aggression, a need to control vs. deference to the examiner and a need for structure, vs. normal children who were characterized by the opposite behaviors. Orphaned children made choices revealing: (1) oral-regressive behavior; i.e., "choosing to drink from a baby bottle vs. a cup," and (2) a sense of abasement and inferiority, i.e., "choosing to stand on a short box vs. a tall box.“ High school students and delinquents differed in expected directions on attitudes towards authority, aggression, nurturance, and self-worth. Situational techniques have also been employed previously by other researchers to study a variety of motives frustration (Reynolds, 1928); negativism (Rosenzweig and Mason, 1934); children's coping styles (Murphy, 1962); children's emotions and motivations (Miller, 1960); children's achievement motivation (Rosen and D'Andrade, 1959), and child rearing antecedents of identification behavior (Sears et al., 1965). The latter study involved the use of a staged “telephone" game, in which a child portrays different roles in a fantasy game of calling his/her mother, and then playing a child who did something wrong; finally, playing a child of the opposite sex, and then, playing his/her father calling the mother. This situational test was designed to elicit behavior from the child in interaction with his/her mother, which was rated in terms of the following variables: "capacity for adult role playing in fantasy, sex typing, and guilt over deviation.” 22 These studies further support the usefulness of the situational tech- niques. There are some limitations to the MST, or any Situational tech- nique. Briefly, the method does not allow observations of some particu- lar motive expression, or change in mode of expression, which might occur if the child were free to act upon objects in any manner he/she chose, over an extended period of time. For instance, a child may choose a nurturant behavior for the first ten minutes, but during the subsequent ten minutes Of play, he/she might express aggressive impulses towards the same Object. The method imposes a high degree of restric- . tion on the expression a motive might be given. A child might choose to express needs for nurturance by drinking from a cup vs. a baby bottle in the original MST situation, but in real life he/she might seek an embrace, verbal praise, or a piece of candy from the experi- menter. Thus, construct validity for "traits", e.g., dependency, derived from results of studies using Situational techniques is often hard to establish. Generally, earlier studies reflect inconsistency with respect to validity (Miller, 1960), although later studies (Santostefano, 1962a, 1962b, 1965) have demonstrated consistency in “traits" over time, and over several independent sources of measurement. However, the advantages of this method seem to outweigh the dis- advantages. This method provides a high constancy of external stimulus cues, unequivocally identifiable responses, yields overt instrumental behavior that can be easily quantified, and reduced the likelihood that the subject will take an extreme or neutral position, by forcing him/her 23 to discriminate. It also can be administered in a relatively brief period of time, does not involve complicated rating systems which might influence reliability of scores, and adapts itself to the limited atten- tion span of preschool aged children. Thus, the MST, as a model, guides the construction of situations which systematically vary the agent, object, direction, outcome, and intent of an interaction, to suit the unique needs of a research problem. In the present study, the MST format was used to test for differ- ence in children's preferred mode of self-representation in a structured fantasy play. Identification with provider-protector, aggressor, and victim roles were behaviorally operationalized; i.e., a child was preé sented with a simultaneous choice of pretending to "hug a Teddy Bear" (identification with the Provider-Protector role), "spank a Teddy Bear" (identification with the Aggressor role, or pretend a "Teddy Bear was spanking him/her" (identification with the Victim role). Thus, the child was forced to Choose between expressing nurturant vs. aggressive impulses; in the latter case, the child chose between directing the aggression outward, or becoming the object of aggression. The behavior— al choices contained both "contrasocial and prosocial“ aggression, and were based upon common themes in children's fantasy play, found by Jersild et a1. (1933), Jersild and Holmes (1935), Lowenfeld (1935), Pitcher and Prelinger (1963), Ames (1963), Gould (1972), and Singer (1973). 24 Children's Perceptions of Parental Behavior Superego development as viewed by Freud (1924) consisted of the child successfully introjecting an "ego ideal", and also internalizing parental prohibitions, values, and standards. Fenichel (1945) sum- marizes the psychoanalytic concept of the superego as a "source of threats and punishments, of protection, and also as a provider of reassuring love." However, while the superego is derived from intro— jecting a piece of the external world, it also depends on the instinct- ual structure of the Child; i.e., the child's previous experiences in handling and expressing his/her impulses. Thus, the "severity of the superego" (Gould's notion of "limited vs. global" self-condemnation) may correspond to the previous real "strictness, permissiveness, warmth, rejection", etc., of the parents, at the same time reflecting the child‘s unconscious affective strivings towards the parents, i.e., “hostility“ vs. "love" vs. "dependency". Thus, it is not only important to study actual parental behaviors, especially disciplinary techniques, to investigate superego development, but also the Child's perceptions of his parents as punitive, reasonable, demanding, etc. Children's perceptions incorporate elements of reality as well as the projection of their own feelings, needs, wishes, fears, and desires, which also govern behavior and self-evaluation. Therefore, it is possible by examining the child's perceptions (If parental “punishment" for fantasized transgressions to investigate qualitative differences among young childrenin "superego severity", without resolving the issue of whether fantasy behavior indicates 25 replication vs. wishfulfillment tendencies. The range of fantasized punishment is wide, as reported in the literature. For example, Issacs (1933) reported that children feared retaliation from parents for deviant behavior; these retaliations ranged in severity and included “being given a hard task to do, being scorned, laughed at or reproached, being put into prison, whipped, shut out from home and mother's love, being deprived of food, and starving, being attacked, eaten up and killed"... (p. 370). There is a substantial amount of research in which children's reports of parental behaviors have been studied on general dimensions; such as, acceptance-rejection, autonomy-control, and punishment, which reflect empirical relationships between the child's perceptions, his/her age, sex, social class, personality, and behavior (see Goldin, 1969). For example, young preschoolers show more sterotyping in view of rela- tive power, with fathers seen as more powerful and dominant; mothers are seen as benign and facilitative (Harris and Tseng, 1957; Anderson, 1937; Droppelman and Schaeffer, 1963; Hawkes, Burchinal, and Gardner, 1957). Sex differences are inconsistent, however, with Emmerich (1961) reporting that preschool girls rate their mothers as more powerful than their fathers, although with regard to punishment, all children appear ‘bo perceive the father as more punishing (Kagan, 1956; Hawkes et al., 1957; Kagan et al., 1961; Gardner, 1947). Other sex differences reveal that boys perceive both parents as less accepting and loving, more psychological controlling, demanding, and more punitive than girls do 26 (Anderson, 1937; Bach, 1945, 1946; Gardner, 1947; Droppelman et al., 1963; Seigelman, 1965). More specifically, boys perceive fathers as using more physical punishment and isolation (Seigelman, 1965) and as being more aggressive to them (Bach, 1945, 1946) than do girls. Sears et al. (1965) using deviation doll play, found that boys emphasized getting caught by authority figures and being physically punished, while girls emphasized receiving verbal punishment or isolation. Schaefer (1965) factor-analyzed children's reports of parental behavior, and found significant variables were "Acceptance vs. Rejec- tion,” “Psychological Autonomy vs. Control," "Firm vs. Lax Control," while Siegelman (1965) found similar factors that he identified as “Loving, Demanding, and Punishing." Goldin's (1969) review of the literature on these studies points out that most research used global variables; such as, “positive or negative family adjustment,“ "dominance,“ "parental intrusiveness," etc., which did not specify exact conditions that elicited parental accepting, controlling, demand- ing, or punishing behavior. Further, neither parent nor child behaviors were operationalized to any great extent, and Child personality varia- bles were largely ignored. Kallman and Stollak (1974) found that children's perceptions and fantasies of maternal discipline in problem situations most often con- sisted of responses; such as, "ordering, commanding, persuading with logic, and providing solutions," which corresponded to both college students and maternal perceptions of their responses to children in the same conflict situations (Stollak, Scholom, Kallman, and Saturansky, 1973). However, children in the Kallman and Stollak (1974) study also 27 perceived their mothers as using physical punishment, yelling, and restriction of privileges more frequently than maternal reports indi- cated. This may be due to misperception on the Child's part and/or inaccuracy of maternal reports due to guilt or fears connected with reporting that they are "bad parents“. An alternative explanation is that younger Children cannot perceive and report differences between logical arguments, orders, commands, etc. Following Gutkin's (1973) and Gutkin and Falvey's (1971) arguments, such children have difficulty in "decentering" from the initial reponse of the mother, or their own need involvement in the problem. In the Kallman and Stollak study, children's fantasies of what they wished their mothers would say in conflict situations centered around the child receiving satisfactions of his/her needs, wishes, demands, etc., with or without restrictions, suggesting that fantasy enables the child to decenter his/her attention from the typical responses used by the mother. Indeed, children fre- quently wished their mothers would "ignore" them in problem situations, possibly indicating that children need time to examine their feelings before discussing them. Schaefer (1965) in discussing the literature on children's perceptions of parental behavior stated: "A Child's perception of his/ her parents' behavior may be more related to his adjustment than is the actual behavior of the parent" (p. 413). This statement points out the SiQnificance of studying the child's developmental level and personality characteristics in relation to ego and superego functioning. Goldin (1969) further suggested that Children's reports of outcomes of parent- chilci interactions should be related to discrete behavioral groups, 28 e.g., "children who are withdrawn, fearful, etc., may have parents who threaten to hurt them if they disobey them, vs. aggressive children, who are actually physically punished for disobedience. (p. 235). In keeping with Goldin's suggestions, this study attempted to examine the relationships between fantasy expressions of "superego severity" and “constancy", as derived from children's perceptions of parental punishment for fantasized transgressions, and dimensions of social behavior (i.e., self-control, self-reliance, mood, peer affilia- tion), which are theoretically related to "social competence“ (Baumrind, 1967, 1971). Parental Discipline and Moral Development in YOung_Chderen The impact of differential types of parental discipline upon noral development in children, or "superego constancy" (Gould, 1972) cannot be viewed in isolation from the child's cognitive-developmental level. Thus, Piaget (1965) believed that children (ages 4-6) were dominated by rules imposed upon them by authority figures, viewing these "moral dicta" as "sacred and unchangeable"; being unable to dif- ferentiate between their fantasized impulses and rules imposed upon them externally. The "egocentric" child adopts parental standards and prohibitions, out of respect for authority figures. Kohlberg (1966) disagrees with this view, stating that the "pre-moral" child's compli- ance with parental injunctions is a function of the child's fear of PUHishment. With extremely young children, such as those in this study, 29 it is difficult to assess motives such as respect, fear, shame, and guilt which are associated with moral development. Thus, research on type of parental punishment appears more relevant to the scope of the present study. Reasoned discipline and authority employed only for the child's welfare and limited permissiveness seems to create children who are more accepting of parental authority, values, and rules (Pikas, 1961; Elder, 1963). Inductive or reasoning methods foster an internal orien- tation and inner locus of control for positive and negative occurrences (Hoffman and Saltzenstein, 1960, 1967; Aronfreed, 1961; Levine, 1961; Sears, Maccoby, and Levin, 1957). Reasoned discipline along with low punishment are also more likely to foster internalization of social behaviors and moral judgments, particularly if accompanied by warmth (Unger, 1960, 1961). Hoffman and Saltzstein (1960, 1967) found that children who classified their mothers as affectionate and non-power assertive were more likely to stress how the child's behavior had hurt the parent. Hoffman (1960, 1963, 1970), Baumrind (1967, 1971, 1973) also agree that a child's consideration for others is related to par- ental warmth, explanation, and reasoning in administering discipline, and clear communication about the consequences of the child's behavior 'for others. On the other hand, power-assertive techniques; such as, physical punishment, are associated with external orientation or concern for Inanishment, rather than with concern for recognition of deviant behavior, and also with retribution (Aronfreed, 1961; Heinicke, 1953; Hoffman and 3O Saltzenstein, 1960, 1967). Power-assertive techniques are further correlated with the child's inability to accept responsibility for a negative occurrence (Sears et al., 1957; Hoffman et al., 1960, 1967). They also fbund that severe punishment was negatively associated with belief in internal control of positive outcomes. Negative control; that is, criticism, severe punishment, and rejection, all serve to decrease the child's independence, self-esteem, and internal locus of control. Power-assertive techniques are only aimed at inhibiting unacceptable behavior, and do not provide alternative behaviors for the child to express his/her feelings, needs, wishes, impulses, and/or desires, and they may, therefore, elicit anger and withdrawal behavior in children (Becker, 1964). Power-assertion also provides an acting- Out model for the child to imitate, and may separate the child from the parent, preventing positive modelling and reinforcement of acceptable behaviors (Sears, Whiting, Nowlis, and Sears, 1953; Becker, 1964). Lastly, love-withdrawal techniques; such as, isolation, depriva- tion of privileges, and direct but not physical expression of anger or disapproval, also result in less acceptance of responsibility for wrong- doing in the child as well as externally motivated behavior, and in- creased dependency and insecurity (Aronfreed, 1961; Hoffman, 1960, 1963, 1970; Sears et al., 1957; Hoffman and Saltzenstein, 1960, 1967; Levine, 1961). An explanation of these findings may be that love-withdrawal intensifies the child's strivings for adult approval (Hoffman, 1963). However, as Hoffman and Saltzenstein (1960, 1967) point out the ’ effects of differential modes of punishment interact: 31 1) Parents who use induction and low power-assertion have children who tend to identify more with the parent, emulate his behavior, particu- larly those “other-oriented" in nature, and are able to incorporate the parents' values or standards successfully. Thus, one might predict that children who perceive their parents as using "reasoned discipline" in Baumrind's (1967, 1971) terms, will choose to identify more with pro- vider-protector roles in fantasy play, emulating parental concern for others, as well as being more prone to admit and atone for fantasized transgressions. Their overt social behaviors may correspond to Baumrind's Pattern I child (demonstrating self-reliance, self-control, affiliation, and explorative behaviors). 2) Parents who use both power-assertion and love-withdrawal techniques have children who are withdrawn from adults, and who tend to withhold hostility in order to avoid punishment. High power-assertion produces a lack of concern for others, and seems to generate feelings Of anger in the child, while providing him/her with a model of aggres- sive acting-out behavior. Thus, one might predict that children who perceive their parents as using mainly power-assertion, with or without love-withdrawal, will choose to identify more with aggressor roles in ‘Fantasy, modelling the parents' behaviors, and choosing to deny or hide their deviant fantasy behaviors, due to fear of punishment, and lack of concern for others. The overt social behaviors of these children may coincide more with Baumrind's (1967) Pattern II child (who is discon— tent, withdrawn, and distrustful). 32 3) Parents who mainly use love-withdrawal techniques of punishment have children who tend actively to seek nurturance from adults and withhold hostility, in order to maintain adult approval, and gratify dependency needs. Love-withdrawal modes Of punishment may socialize a child to behave in a "conventional and responsible" way, but behavior is externally-controlled, vs. internally controlled. Thus, one might pre- dict that children who perceive their parents as mostly using tactics involving loss of "parental love or approval", will choose to identify more with victim roles in fantasy, thereby redirecting hostility that originally was directed towards the parents towards themselves. At the same time, the "victims" Should display behaviors, such as crying, that are designed to elicit care-taking behaviors and nurturance from signifi- cant others in response to their fantasized transgressions. These chil- dren, in response to fantasized deviant acts, also should recognize and attempt to make retribution for them. However, their motivation may be due primarily to a fear of loss of love vs. a concern for others, or a desire to emulate adult behaviors. Their overt social behaviors may correspond to Baumrind's Pattern III children (who are lacking self- reliance, self-control,.and tend to retreat from novel situations). In summary, children's fantasies of agent(s) and mode of perceived punishment as well as behavioral reactions to fantasized transgressions, are hypothesized to reflect "superego functioning" derived from parents' actual behaviors towards the child and mediated by the child's pre- dominant affective strivings towards the parents, i.e., "affection and P051 tive emulation" vs. "hostility and negative emulation" vs. "depend- one)! and.helplessness". Again, the purpose of this study was to explore 33 relationships between children's preferred modes of self-representation in fantasy and the aforementioned aspects of superego "severity and constancy", together with the children's demonstrated level of "social competence" as perceived by their nursery school teachers. QUESTIONS AND HYPOTHESES The following questions and hypotheses were examined in the present research: 1) Are there significant differences in expressed preference for mode of self-representation in structured fantasy play behavior between children that were classified into different groups, as a function of teacher ratings of overt social behavior in the classroom? One hypothe- sis predicts that "more mature" children would tend to identify more with provider-protector roles, while children rated as "less mature" by their teachers would tend to prefer identification with victim or aggressor roles. 2) Are there significant differences between the sexes in choice of provider-protector, aggressor, or victim roles in structured fantasy play? One might expect that girls will tend to choose the provider- protector role more frequently than will boys; on the other hand, boys may tend to prefer aggressor roles. 3) Are there significant differences between groups in behavioral reactions to fantasized transgressions? One possibility is that chil- dren judged by their teachers to be "more mature" will more frequently distalay reactions of confession or apology. while children whom 34 teachers perceive as "less mature" will tend to manifest more denial or hiding responses in an attempt to avoid reSponsibility for the fanta- sized deviant acts. 4) Is there any significant interaction between group classifica— tion and sex in the child's response to a fantasized deviant act? 5) Are there significant differences between groups in perceived agent(s) of punishment for fantasized transgressions? It could be that children who are judged by their teachers as "more mature" might per- ceive both parental figures as consistently disciplining them, while children who are rated by teachers as "less mature" might perceive both parents as inconsistent, and sending conflicting, ambiguous messages. 6) Is there any significant sex difference in the choice of the parent who is perceived as the primary agent of punishment? One might expect that both sexes would tend to perceive the father figure as more powerful and punitive, given that the subjects are preschool children. However, in response to Oedipal fear, there may also be a tendency to perceive the same-sexed parent as more threatening. 7) Are there significant differences between groups in perception of mode of anticipated punishment following fantasized transgressions? Possibly children perceived by teachers as "more mature" might perceive symbolic authority figures as employing reasoning approaches more fre- quently in response to fantasized transgressions. Children rated by their teachers as "less mature" might perceive symbolic authority f1Qures as employing power-assertion or love-withdrawal techniques of discipline more frequently in response to fantasized deviant acts. 35 8) Is there a significant difference with respect to sex, in the child's perception of mode of anticipated punishment in response to fantasized transgressions? One possibility, with respect to overall sex differences, is that boys will be more likely to perceive physical punishment as the fantasized punishment; whereas, girls may perceive isolation and verbal punishment as more likely to occur. CHAPTER II METHOD Subject Selection Approximately 300 introductory letters were distributed to parents of four and five year-old children in seven day-care centers and nursery schools in the Lansing area (see Appendix A). The letter requested that parents allow their child to participate in a research study on fantasy play behavior, to be conducted at the child's day care center or school during two individual play sessions, in July and August, 1974. Meetings with the directors of the child-care program also resulted in letters being distributed to 14 teachers (see Appendix B) that requested their participation in rating subjects along five be- havioral dimensions. The dimensions were relevant to social behavior in the classroom: self-control, self-reliance, approach-avoidance, mood, and peer affiliation. Behavioral definitions of the five rating scales were also provided to the teachers (see Appendix C), and these definitions were orally clarified during individual meetings that were held with the teachers. All 14 teachers agreed to participate. Of the approximately 300 children who were asked to participate in this study, 237 children (median age range 4 yrs 6 mos-4 yrs 8 mos) 36 37 were allowed to do so by their parents. These children were rated by their teachers as being either "high", "medium-high", "medium-low", or "low" on the five social behavior scales. Since each child was only rated by his/her principal teacher, interjudge reliability of the ratings could not be assessed. Thus, assessment of the quality of the ratings of the children's social behaviors was dependent solely upon the validity of the results. Formation of Groups Two hundred of the teachers' ratings were factor-analyzed to explore if there were relationships between the five scoring categories. A principal components factor analysis (R2 was used as the estimate of commonality; eigen value g_l), with factors rotated to a varimax solu- tion yielded two factors. All five categories had substantial positive factor loadings (> .5), on Factor I. The loadings ranged from 0.86 for Category 4, Mood; to 0.66 for Category 1, Self-Control. Thus, all five measures were highly interrelated, when zero-order correlation coefficients were compared. Appendix 0, Table I, presents a summary of all factor loadings. Two categories, Self-control (which loaded 0.62) and Self- Reliance (which loaded -.54), also loaded substantially on Factor II. In addition, Approach-Avoidance loaded negatively, though not substan- tially (-.48), on Factor II. Examination of the intercorrelation matrix (see Appendix 0, Table 11) revealed that while all five categories were 38 positively associated with each other, Self-Reliance and Approach- Avoidance, had relatively low correlations with Self-Control, i.e., 0.20, and 0.24, respectively. There were thus, three somewhat separate category groupings within Factor I: (A) Self-Control, (8) Peer Affilia- tion and Mood, and (C) Self-Reliance and Approach-Avoidance. Originally, the intent was to study three groups of children similar to Baumrind's Pattern I, II, and III children. However, the factor analysis of the teachers' ratings appeared to represent a general "high" vs. "low" adjustment factor. Examination of the correlation matrix suggested that differences between Self-Control and Self-Reliance, i.e., the relative impact of "high" vs. "low" adjustment on "control" and "initiative" dimensions of social behavior, might be a more relevant question to study. Thus, a compromise was arrived at between utilizing Baumrind's categories, but identifying groups based on the results of the factor analysis. Four groups were eventually identified. Group I consisted Of children who were rated "high" or "medium-high" on Self-Control, Mood, and Peer Affiliation (control dimensions) and "medium"' on Self-Reliance and Approach- Avoidance (initiative dimensions). Children in this group were gener- ally considered highly sociable, well-liked and well-behaved by their teachers, but somewhat hesitant to approach new situations or try things on their own; hence, it was decided to term these children "Competent/Compliant". Group II consisted of children who were rated ‘ 1"Medium" category was derived from examination and combination 0f “medium-high" and "medium—low" categories. 39 "low" or "medium-low" on Self-Control, Mood, and Peer Affiliation (control dimensions) and "medium" on Self-Reliance and Approach- Avoidance" (initiative dimensions). Children in Group II were generally perceived by their teachers as highly disruptive, and difficult to work with in group situations; these children became highly invested in their own individualized play or projects and would "act out aggres- sively" only if prevented from continuing; hence, this group of children was named "Self-Involved/Disruptive". Group III consisted of children who were rated "high" or "medium-high" on Self-Reliance, Approach- Avoidance, Mood, and Peer Affiliation, and "medium" on Self-Control. These children were perceived as the "most mature" children in the classroom, since their teachers rated them "high" on four out of the five indices of social behavior. Occasionally these children would need limits set, but teachers generally commented that they were the most responsible, and self-reliant Children in the classroom; hence, the group was labelled "Competent/Independent". Group IV consisted of children who were rated "low" or "medium-low" on Self-Reliance, Approach-Avoidance, Mood, Peer Affiliation, and "medium" on Self-Control. These children were perceived as the "least mature" children in the classroom by their teachers. They were typically described as "acting out aggressively" to gain attention, becoming sullen and withdrawing when their needs were not met, and overly dependent upon adults with little peer interaction. Hence, this group of children was labelled "Impulsive/Withdrawn". 40 Groups I, II, and IV somewhat paralled Baumrind's (1967) Pattern I, II, and III children, respectively. Group III (Competent/Independent) attempted to differentiate children in Baumrind's Pattern I category who were relatively "high" on initiative-related categories, and "low" on the control dimension, from those Pattern I children who are "high" on control, and "medium" on initiative dimensions, i.e., Competent/ Compliant children (Group 1). One explanation for the failure to repli- cate Baumrind's groupings may be that her "Patterns" are "method- specific". Baumrind (1967) used the Q-sort to obtain her three patterns of social behaviors in pre-school children, while the present study used a four point scale. Future research is needed to clarify whether Baumrind's (1967) "Patterns" of children's social behaviors can be replicated or whether some re-conceptualization of the relationships between different types of social behaviors in preschool children needs to occur. Out of the initial pool of 237 children who were rated by their teachers, 136 children were selected for the study based on the criteria for classification into one of the four groups. Nineteen subjects failed to attend both sessions of testing, and/or to complete both experimental tasks with scorable protocols, and thus, were dropped from the study. Therefore, the total number of subjects whose data were analyzed was 117 (62 females and 55 males). 41 Sample Characteristics Information on the child's age and background was obtained through the registry files of each day care center/nursery school. Children ranged in age from three years ten months to five years two months. The median age range was 4 years 6 months to 4 years 8 months. The sample was fairly homogeneous with respect to birth order, approximately 70% of the children were first-born, while nearly 60% of the subjects were "only children". The majority of the sample came from intact families, while approximately 30% of the subjects came from "father-absent" homes, due to separation, divorce, or death. Generally, children came from "lower-middle". and "working-class" urban families, in which mothers were employed, usually full-time. All chil- dren were Caucasian. (Appendix E presents a more detailed summary of sample characteristics.) Design This study utilized a 4(Child-Behavior Group) x 2(Sex) x 2(0rder of Task Presentation) factorial design. There were unequal cell fre- quencies across the 16 cells of the design, since there were 7 more female than male subjects; the number of subjects ranged from 5 to 10 per cell. 42 Dependent Measures Procedure--Task A: Modified Miniature SituatTonSTTest (MST) Children's self-representations in fantasy were assessed via their responses to a series of miniature situations that used a format similar to the Miniature Situations Test developed by Santostefano (1960, 1962a, 1962b, 1965, 1966, 1968). The MST, in its original form, consisted of 42 situations that required the child to choose between two equivalent activities, e.g., "standing on a tall box vs. standing on a short box." Santostefano hypothesized that the child's choices reflected his/her "personality in action", portraying samples of his/her "coping behaviors". The child's decision to act upon physically, manipulate, or avoid Objects presented by the experimenter in the presence of an Observer-rater was hypothesized to "trigger an affective response re- vealing personality characteristics related to stimulus materials he valued or rejected," e.g., "child prefers to tear up an old newspaper" vs. "repair a torn picture." This study employed the MST format. Each child was led into a self-contained room within his/her day care center or nursery school, and a female experimenter, in the presence of two observer-raters (one nale and one female), introduced the modified MST to each child by saying: We are interested in the games that Children like to play. All children usually like to play some of these games. These games are all make—believe; I would like you to tell me which one of these three games you want to play, and then I will let you play it. Remember you can only choose one game at a time. 43 Three or four trial items were then presented to the child to check his/her comprehension of the required task. Each child was presented with 18 different situations, all of which provided him/her with three alternatives: (1) identification with the provider-protector role; (2) identification with the aggressor role; (3) identification with the victim role. The provider-protector role was operationally defined by the child choosing to display and actually emitting symbolic nurturant behaviors towards an animal toy (e.g., "S feeds monkey with baby bottle"; "§_hugs Teddy Bear"). Identification with the aggressor role was operationally defined as the child choosing to display and actually emitting symbolic aggressive behaviors towards an animal toy (e.g., "§_shoots monkey with gun"; "§_spanks Teddy Bear"). Identification with the victim role was operationally defined as the Child Choosing to be and actually becoming the object of symbolic aggression attributed to an animal toy (e.g., "§_pretends toy crocodile is biting him"; "§_pretends Teddy Bear is spanking him"). A complete list Of the 18 situations is presented in Appendix G, along with the detailed procedures that were used in the administration of this test. Animal toys were chosen as stimuli because of their "animate-like" qualities, and because of their ability to provide psychological dis- tance from reality, which was hypothesized to enable the child to Project more freely his/her nurturant and aggressive impulses. As noted previously, all three types of roles were presented to the child, simultaneously. Thus, it was necessary to use six different 44 orders of role presentation to control for order of presentation effects. More specifically, each role was presented as the first alternative six times, as the second alternative six times, and as the last alternative, six times. Each order of role presentation contained three different situations which were assumed-tO-be equivalent, and stability of the child's choices was thus assessed by the repeated measures aspect of this design. The overall internal consistency of this instrument is discussed in the Results section. Scoring the Dependent Variables. The child's Choices were re- 2 to estab- corded simultaneously by the author and two observer-raters lish inter-rater reliability. (This is discussed further in the Results section.) Children's refusals to make choices and/or subse- quent changing of responses were noted. The most common problem was children who initially wished to play all three games, but setting limits and offering the child one choice usually resolved this dilemma. Differential toy attractiveness posed another problem, with some chil- dren preferring to "play" longer with different.toys. Thus, a time limit of two minutes of activity per item was imposed. Total testing time ranged from 15-45 minutes. Changing Of responses was relatively rare, with only 8 out of 117 children apparently deciding to change provider-protector responses to aggressor responses (e.g., "child hugs Taddy Bear, and then starts choking him"; E clarifies §fs behavior, and response is scored identification with the aggressor). Children who _ 2The observer-raters were two undergraduate students in PSYChology: Kay Guimond and Joseph Pizzo. t'n 1 3.1 rep: 155'. K» . .3 [7—1. 361; '8 it‘ll ”Er 45 refused to make all 18 choices were not included in the data analyses. Thus, each child obtained a possible score ranging from zero to three for each mode of self-representation within a specific order of role presentation (i.e., the six times a role was presented as the first alternative, the Six times it was presented as the second, and the six times it was presented as the third or last alternative). Therefore, the overall total possible score for each mode of self- representation ranged from 0 to 18, when scores were summed over the assumed-to-be equivalent situations and order of role presentation. Procedure--Task 8: Deviation 0011 Play Again, each child was led into a self-contained room within his/her day care center or nursery school, and during an individual testing session, lasting 25 to 45 minutes, was introduced to structured Deviation 0011 Play. (This task was scheduled separately from Task A, and overall order of task presentation was factorialized across Child- Behavior Group and Sex.) Deviation 0011 Play consisted of the experi- menter using standard doll play equipment to present five incomplete stimulus stories in which a doll (the child doll of the same sex as the subject) committed a deviant act, something commonly disapproved of in our culture, e.g., "refusing to go to bed at bedtime". The experi- menter handed the same-sexed child doll to the subject at the end of the description of the deviant behavior, and asked §_to "Show and tell what happened next". More specifically, Deviation 0011 Play was an attempt to evoke in the child, a mild and fantasy-level feeling of having transgressed 46 and then to measure his/her behavioral reactions to the deviation, as well as his/her perceptions of the agent(s) and mode of punishment expected. The five incomplete stimulus stories (Sears, Rau, Alpert, 1965) included common deviations understood by 4 year-olds: (l) knock- ing over a table lamp; (2) spilling juice at breakfast; (3) grabbing Baby's toy; (4) stealing cookies before dinner, and (5) getting up to play after being put to bed. A complete list of instructions to the child and stimulus stories are presented in Appendix G. Scorjgg_the Dependent Variables. The categories used for scoring completions to the deviation doll play situations are modifications of those used by Sears, Rau, and Alpert (1965, pp. 324-326): I. Child's Behavioral Reaction to Fantasized Transgression (a) Qgg121;-Child denies responsibility for deviation in response to an accusation, or fails to respond when confession is solicited by an authority, or denies voluntarily, before he/she is implicated. (b) Confession--Child admits or confesses he/she was responsible for deviant act, whether or not he/she has been accused. (c) Apglggy:-Child apologizes spontaneously to authority figure for deviant act. (d) Spontaneous Fixingr-Child restores or "undoes" deviant act, without adult authority figures' knowledge. (e) fllglgg:-Child hides himself/herself and/or the evidence from authority figures. (f) Crying--Child in the story cries and/or pleads with adult not to get angry or punish him/her. 47 (9) Gets Away_With It--The deviation is never discovered or the child is never implicated. (h) Mute, Silent, Withdrawa1--Child stands/sits silently in presence of adult authority, after deviation is discovered. (i) Continued Deviant Aggression-~Child continues deviant behavior with or without adult presence, and behavior is aggressive and destruc- tive to objects, and/or self and others. II. Child's Perceptions of Parental Discipline (A) Mode of Punishment (1) Vgggglr-an authority figure Scolds the child, asks why he/She did that, tells him/her never to do it again, or threatens punishment for repetition of the deviant act. (2) Physical Punishment--an authority figure spanks or otherwise hurts the child physically. (3) Denial of Privileges--an authority figure forbids the child to engage in a favorite activity, e.g., watching T.V., or deprives him/her of "nurturant supplies"; e.g., "child cannot have supper". (4) Isolation--chi1d is forced to remain alone as punishment; authority figure is not present. (5) Reasoning--authority figure attempts to teach child why he/she should not engage in deviant act again, and/or sets limits for present or future behavior, and/or offers alternative behaviors for the future. (6) Reflection of feelings/forgiveS--authority reflects child's feelings of fear, anger, guilt, Shame, and/or forgives the child, spontaneously. 48 (7) Forced Fixing-~authority forces child to fix, restore, or "undo" deviant act with or without adult's help. (8) Ignoring--authority figure ignores the child, as punishment for deviant behavior. (8) Agent(s) of Punishment (1) mother (alone) (2) father (alone) (3) mother and father, together (agreeing, supporting each other) (4) mother and father, together (disagreeing, giving child inconsistent messages) A detailed description of these categories and examples of scorable responses (used to train raters) can be found in Appendix H. The author and two observer-raters recorded responses to the five Deviation 0011 Play stories on the basis of these 21 categories. Each story completion was scored according to all appropriate categories, but each category could be used only once per story. The final score for a Child on any given category was its frequency of usage (summed over all five story completions). Thus, the possible total score for any given category ranged from 0 to 5. (See Appendix I for a ranked order presentation of mean category usage.) Inter-rater reliability was assessed, and is presented in the Results section. CHAPTER III RESULTS Inter-Rater Reliabilities of Dependent Measures Two raters were trained in the scoring procedures during an initial pilot study that used 25 subjects. The two raters plus an expert rater, the author, then scored all modified MST responses along three categories descriptive of the child's choice of self-representa- tion in fantasy, i.e., provider-protector, aggressor, and victim roles. Deviation 0011 Play responses were scored along three dimensions. One dimension, parental punishment, was examined via eight categories: verbal, physical, denial of privileges, isolation, reasoning, reflection . of feelings/forgives, forced fixing, and ignoring. A second dimension, the child's behavioral reaction to fantasized transgression, was examined via nine categories: denial, confession, apology, spontaneous fixing, hiding, crying, gets away with it, mute, silent, withdrawal, and continued deviant aggression. Four categories described the third dimension, the perceived agent(s) of punishment: mother (alone), father (alone), mother and father together (agreeing), and mother and father together (disagreeing).1 Each rater scored responses for all categories, ¥ 1This category was dropped from data analysis due to low frequency of usage. 49 50 over all 117 subjects, during the actual administration of the tasks. Correlations between rater one (expert rater) and each of the two undergraduate raters for all response categories are presented in Appendix J, Table I. Mean reliability for each rater pair, and mean category reliability are also included. Correlations (gs) for the rater pairs across 23 categories combined were 0.83, 0.84, and 0.93, respectively. The overall mean category reliability was 0.88 with individual reliabilities ranging from 0.71 to 0.99. Internal Consistency_of Modified MST Correlation coefficients between the number of responses per item, and the total sum of responses per item category, were computed to assess internal consistency. Appendix K, Table I, presents the part-whole correlations for all 18 provider-protector items within each of 6 different orders of role presentation. The magnitudes of the correlations were mixed, ranging from Qggl (item 16) to Q;ZQ_(item 15). The mean item-whole correlation for provider-protector scores was 9:55, Appendix K, Table II, also presents the part-whole correlations for all 18 aggressor items, within each of 6 different orders of role presenta- tion. Again correlations were variable, ranging from lel_(item 16) to Q;§§_(item 15). The mean item-whole correlation for aggressor scores was 9:52, Thus, for both provider-protector and aggressor items, "Hug or Hit the Mouse" (item 15) had the highest reliability of all items; whereas, "Put the bandaid on Kitty where she got cut" or "Spank the Kitty" (item 16) had the lowest item reliability. Appendix K, Table III, 51 presents the part-whole correlations for all 18 victim items, within each of 6 different orders of role presentation. Correlations are relatively lower for victim items than for either provider-protector or aggressor items. Correlations for victim items range from Q; 5 (items 8 and 17) to Q;§2_(item 3). The mean item-whole correlation for victim scores was Qggg, Thus, all items showed somewhat low reliability with total item scores. Therefore, the internal consistency of the modified MST, is less than optimal. Generally, "feeding", "hugging", and "petting" provider-protector responses, were relatively reliable, as compared to "putting on bandaids" or "putting an animal toy in a tub of water to swim." Likewise, aggressor behaviors such as "spanking, punching, or hitting", were generally more reliable and more intense responses; "drowning" or "shooting" were the least so. Overall choice of victim responses (7'= 2.30) was relatively low, which might have contributed to the rather uniformly low reliabilities in this category. Category Usage Self-Representation in Fantasy (modified MST) The most frequently obtained response category was Category 1 (X'= 7.96), "Provider-Protector". The category with the lowest mean frequency was Category 3 (Y'= 2.30), "Victim". (Appendix I, Table I, presents ranked-order means and standard deviations for all scoring categories of the 23 dependent measures.) Total number of possible responses per category ranged from 0 to 18. 52 Perceived Parental Response to Fantasized Transgression The most frequently Obtained response was Category 5, "Verbal Punishment" (Y'= 3.07). The category with the lowest mean usage was Category 10, "Reflection of Feelings/Forgives" (i'= 1.10). Total number of possible responses per category ranged from 0 to 5. hild's Behavioral Reaction to Fantasized r TransgresSion The most frequently obtained response was Category 19, "Continued Deviant Aggression" (7'= 2.13). The category with the lowest mean usage was Category 13, "Gets Away With It" (Y'= 0.93). Total number of possible responses per category ranged from 0 to 5. Agent(s) of Fantasized Punishment The most frequently obtained response category was Category 21, "Mother (alone)" (7'= 1.97). The category with the lowest mean usage was Category 22, "Father (alone)" (7'= 1.27). (One category, "Mother and Father Together (Disagreeing)" was only used by two subjects, and thus, was not included in the data analyses.) Total number of possible responses per category ranged from 0 to 5. Factor Analysis of Dependent Measures A principal components factor analysis was conducted to explore possible relationships between the 23 dependent measures. The cate- gories generated the following 8 factors which together accounted for .70 of the cumulative proportionate variance: 53 l) Competence—Incompetence (consisting of provider—protector, perceived parental reasoning, apology by Child for fantasized trans- gression, confession by child for fantasized transgression, and per- ceived parental reflection of feelings/forgiveness measures which loaded substantially positively on Factor I, and aggressor, hiding, and crying measures which loaded substantially negatively on Factor I). 2) Child's Extreme Behavioral Reactions to Fantasized Transgres- §jgg,(consisting of two mutually exclusive scoring categories in Deviation 0011 Play, i.e., continued deviant aggression, and mute, silent withdrawal). 3 and 4)2 Agent(s) of Punishment (consisting of three mutually exclusive scoring categories in Deviation Doll Play, i.e., mother (alone), father (alone), and mother and father (together, agreeing). 5) Perceived Parental Response to Fantasized Transgression (also consisted of three mutually exclusive scoring categories in Deviation Doll Play, i.e., gets away with it, ignoring, and verbal punishment). 6) fjgjgg_(consisting of two mutually exclusive scoring cate- gories in Deviation Doll Play, i.e., Forced vs. Spontaneous Fixing). 7) Identification with the Victim Role.3 8) Qggjgl_(consisting of parental denial of privileges in response to fantasized transgression, and child's denial of fantasized deviant acts). 2Included as separate factors because categories were linearly dependent. 31bid. 54 The factor analysis of dependent measures is discussed fully later in the Results section, and is presented briefly here to acquaint the reader with the basis for multivariate analyses of variance which are discussed next. (Appendix L presents a summary of factor loadings for all 8 factors.) Multivariate Analyses of Variance The results of a factor analysis of the dependent variables were used to examine possible relations with the independent variables. The eight factors (see later presentation of the results of the factor analysis), comprising 23 dependent measures were subjected to a 4(Child- Behavior Group) x 2(Sex) x 2(0rder of Task Presentation) multivariate analysis of variance. When a multivariate §_ratio exceeded a confidence level of 0.05, univariate §_ratios for each of the variables associated with a given factor, were examined. Findings are reported for the univariate results which: (a) were significant at a E_leve1 determined by dividing the overall significance level (0.05) by the number of dependent measures which loaded substantially on a given factor, and (b) were not qualified by significant higher-order interactions. A simple effects analysis (refer to Appendix M) further investigated any significant interactions, and individual comparisons between groups were completed, where appropriate, using Newman Kuels Tests, with p_ set at the 0.05 level. The multivariate analysis revealed a large number of significant effects. Table 1 presents the eleven f_ratios for those multivariate Table 1. 55 Summary of Significant Multivariate Results Reflecting Significant Univariate Comparisons not Qualified by Higher- Order Effects Source d_f _F_ E Main Effects I. Groups a. Competence-Incompetence 24 8.58 0.0001 b. Child's Behavioral Reaction 6 13.57 0.0001 to Fantasized Transgression II. §§x_ a. Competence-Incompetence 8 7.77 0.0001 b. Agent(s) of Punishment 2 4.31 0.0160 c. Fixing 2 5.38 0.0061 Interaction Effects III. Sex x Group; a. Competence-Incompetence 24 2.24 0.0011 b. Perceived Parental Response 9 1.86 0.0594 to Fantasized Transgression IV. Order x Groups a. Competence-Incompetence 24 2.32 0.0007 V. Sex x Order x Groups a. Competence-Incompetence 24 2.00 0.0044 b. Victim 3 4.83 0.0035 c. Denial 9 2.02 0.0384 '1 11.6 (A r» JVI‘ H 56 Table 2. Summary of Significant Univariate Results not Qualified by Higher-Order Effects Source g_f_ _F_ p_ Main Effects 1. Grou s a. Provider-Protector 3 46.42 0.0001 b. Hiding 3 11.00 0.0001 c. Continued Deviant Aggression 3 16.52 0.0001 II. gig; a. Provider-Protector l 18.82 0.0001 b. Father (alone) 1 8.57 0.0044 c..Spontaneous Fixing l 8.51 0.0043 Interaction Effects III. Sex x Groups a. Reasoning 3 5.50 0.0016 b. Ignoring 3 3.83 0.0122 IV. Order x Gropps a. Apology 3 4.84 0.0035 V. Sex x Order x Groups a. Aggressor 3 3.27 0.0243 b. Victim 3 4.83 0.0035 c. Child's Denial 3 5.25 0.0021 57 comparisons that reflected at least one significant univariate compari- son that was not qualified by any higher-order effects. Table 2 (page 56) presents the twelve [_ratios for all significant univariate effects, not qualified by higher-order interactions. Presented below is a detailed statement of these findings, including where appropriate, the final results of the simple effects analyses. Main Effects Significant multivariate main effects reflected significant uni- variate comparisons for the dependent variable, Provider-Protector, which loaded substantially on Factor I, Competence-Incompetence. The univariate {_main effects were significant for identification with the provider-protector role in fantasy for groups {(3, 101) = 46.41, p< 0.0001; for sex f_(l, 101) = 18.82, p< 0.0001, and these effects were not qualified by higher order interactions. Table 3 presents the means and standard deviations of provider-protector responses to the modified MST, classified by groups, as well as the means and standard deviations of all other variables that yielded main effects for groups. Choice of identification with provider-protector roles was Sig- nificantly higher for Competent/Compliant (Group I) children vs. all other groups. There were no significant differences between Groups II, III, and IV. Competent/Compliant children showed the highest mean usage for provider-protector responses, while Impulsive/Withdrawn chil- dren displayed the lowest mean scores for provider—protector choices. Also, the main effect for sex reflects a significantly greater 58 Table 3. Means and Standard Deviations of Variables that Yielded Significant Main Effects for Groups Groups I II III IV’ Provider-Protector Mean 13.11" 6.45b 5.72b 5.60b Standard Deviation 2.42 2.10 3.00 2.50 Hiding Mean 0.813 1.40 1.08 2.68 Standard Deviation 1.09 1.31 1.02 1.20 Continued Deviant Aggression Mean 0.852" 2. 54b 1 .92b 3. 24b Standard Deviation 0.810 1.47 1.80 1.15 N 27 28 25 27 Competent/Compliant (I) = children who were rated high or medium-high on self-control, mood and peer affiliation and medium on self- reliance and approach-avoidance. Self-Involved/Disruptive (II) = children who were rated low or medium- low on self-control, mood and peer affiliation and medium on self-reliance and approach-avoidance. Competent/Independent (III) = children who were rated high or medium- high on self-reliance, approach-avoidance, mood and peer affilia- tion, and medium on self-control. Impulsive/Withdrawn (IV) = children who were rated low or medium-low on self-reliance, approach-avoidance, mood and peer affiliation, and medium on self-control. aDenotes significant level at p f 0.05. bDenotes nonsignificance. 59 provider-protector score for females (7'= 9.07) than males (Y'= 6.63). (Refer to Table 4, which presents the means and standard deviations of all variables that yielded main effects for sex of subject.) Thus, Competent/Compliant children (Group I), and females, chose to identify significantly more with provider-protector roles in fantasy, as compared to other children in this study. In addition, a significant multivariate main effect for groups revealed a significant univariate comparison for the dependent variable, Hiding, [(3, 101) = 11.00, pA=0.0001, which also loaded substantially on Factor I ngpetence-Incompetence. This significant main effect for Hiding, was not qualified by higher-order interactions, and Table 3 presents the means for Hiding responses classified by groups (refer to page 58). Newman Keuls tests yielded no significant differences between groups in the usage of Hiding in response to fantasized transgression in Deviation Doll Play. However, the pattern of means shows Group IV children (Impulsive/Withdrawn) chose Hiding behaviors most frequently, while Competent/Compliant (Group 1) children displayed the least Hiding responses. Another multivariate main effect for groups reflected a signifi- cant univariate comparison for the dependent measure, Continued Deviant Aggression, [(3, 101) = 13.57, p<:0.0001, which loaded substantially on Factor II, Child's Extreme Behavioral Reaction to Fantasized Transgres- §1gg, This significant main effect for groups, was not qualified by higher-order interactions (again, refer to Table 3, page 58, which presents the means and the standard deviations for Continued Deviant 6O Table 4. Means and Standard Deviations of Variables that Yielded Significant Multivariate Main Effects for Sex of Subject Sex of Subject MaTE Female (N = 55) (N = 62) Provider-Protector Mean 6.623b 9.071a Standard Deviation 2.81 2.74 Father (Alone) Mean 1.710c 0.836b Standard Deviation 0.96 0.58 Spontaneous Fixing_ Mean 0.677d 1.20a Standard Deviation 0.41 0.49 aDenotes significance level at p 5_0.0001 bDenotes cDenotes dDenotes nonsignificance significance level at p §_0.0043 significance level at p §_0.0044 61 Aggression in response to fantasized transgression in Deviation 0011 Play, classified by groups). There were significant differences between Group I, and all other groups, with Competent/Compliant children displaying significantly less deviant aggressive behavior in Deviation Doll Play. Self-Involved/ Disruptive children and Impulsive/Withdrawn children, obtained the highest scores for Continued Deviant Aggression, although there were no significant differences among Groups II, III, and IV. ' Also, a significant multivariate main effect for sex reflected a significant univariate comparison for the dependent measure, Father (alone), which loaded substantially on Factor III, Agent(s) of Punish- mggt, This significant main effect, £11, 101) = 8.56, p<:0.0043, for Father (alone) as agent of punishment in Deviation 0011 Play, was not qualified by higher order interaction effects. The main effect for sex reveals a greater Father (alone) score for males (I’= 1.71) than females (Y'= 0.84). Thus, males perceived the same-sexed parent (i.e., the father) as the primary agent of punishment for fantasized transgres- sion. No similar effects were found with respect to females, and their mothers. Another significant multivariate main effect for sex reflected a significant univariate comparison for the dependent variable, Spontaneous Fixing, which loaded substantially on Factor VI, fjgjgg, This significant main effect for Spontaneous Fixing in response to fantasized transgression in Deviation Doll Play, f(l, 101) = 8.51, p<:0.0044. The main effect for sex revealed a greater Spontaneous Fixing score for females (7'= 1.20) vs. males (7'= 0.68) (see Table 4, 62 page 60). Thus, females reacted to fantasized transgression by "undoing" the deviant act, without parental knowledge, significantly more than did males. Interaction Effects Sex x Child—Behavior Groups. A significant multivariate Sex x Child-Behavior Group interaction reflected a significant univariate comparison for the dependent measure, Reasoning, which substantially loaded on Factor I, Competence-Incompetence. The univariate inter- action effect for Reasoning, F(3, 101) = 5.50, p<:0.0016, was examined further via simple effects analyses (see Appendix M for a more complete presentation of these findings). These tests indicated that individual comparisons between group conditions within each sex were appropriate. (Refer to Table 5 for individual cell means classified by group and sex, for the univariate, Reasoning.) For females, Groups I, II, and III differed significantly from Group IV, with females in the former groups perceiving significantly more parental Reasoning responses to fantasized transgressions in Deviation Doll Play. Thus, Impulsive/Withdrawn females (Group IV), perceived parental figures as responding with Reasoning approaches significantly less than females in the other three groups. Mean scores for Reasoning responses were also significantly higher for females in Groups I and II, when compared to that of females in Group III. For males, a slightly different pattern emerged (again see Table 5). Competent/Compliant males (Group I), had significantly higher mean Reasoning scores than did males in Groups II, III, and IV. 63 .mmucmemee_c cacao uceuwevcmvmcoc mmpocmon .modlvd pa mmucmgmtvu 395 28:23..“ mangoes .FogucoUTepom co Eareoe ecu .cowuowpwemm some use woos .muceeeo>u -zumoeaaa .mucmwpmeiefimm co zopusawems so zoF among mam: on: cmgopmgu n A>Hv czmeugpwzuo>vmpaaEH .poeucouuepmm co Eaveoe ecu .covumppwema some ecu woos .mucmuro>o Isumoeaam .ooceWPmeiepmm co ;m_;nszwvme so goes team; mam: on: cmeupvso u AHHHV acmucmawucfiiucmomasou .mucmuwo>aigumoeaaa ecu mocm_PwLIepem co Ezwume ecu .cowumwpveem Loon new woos .FoeucooTe—mm co zopisawnme Lo zap empme mam: 0:3 cmeuprcu u AHHV m>vuaaemwaiem>Fo>=Hiw_om .mocmuwo>migumoeaaa ecu mucmPFmeieme co Espuee can cowueppweem some use woos .FoeucouneFom co :mw;-e=vems Lo saw; been; mew: on: coge__;u n AHV pcmwngoUIucmpmasou mp mF «F FF mF NF mp e— z um.o m¢.o o~._ mN.F m~.o om.— om.— oe.F cowuaw>oo vgmvceum amm—.o mmm~.o ammm.o aom.~ nomp.o mmm.m mFm.~ mmm.~ cam: uneconomm mmpmz, mmbmemm mmpmz. mmrmswu mopez moraswm warez, morasmu >H an“ HH H manage xmm new cacao an emvmvmmapu assume: newcomomm use ea mcorpee>oo ueeeceum ecu memo: .m opaee 64 There were no significant differences for males, among Groups II, III, and IV. Thus, both Competent/Compliant males and females (Group I) obtained the highest mean scores for perception of parental Reasoning in response to fantasized transgression, in Deviation 0011 Play. Both Impulsive/Withdrawn males and females (Group IV) had the lowest mean scores for perceived parental Reasoning. Another marginally significant Sex x Child-Behavior Group inter- action revealed a significant univariate effect for the dependent measure, Ignoring, which substantially loaded on Factor 'V, Perceived Parental Response to Fantasized Transgression. The univariate inter- action effect of Ignoring, [(3, 101) = 3.83, p<:0.020, was examined fur- ther via simple effects tests. (Refer to Appendix M for a more complete presentation of these analyses.) Simple effects tests of group condi- tions within each sex yielded no further significant results, however. The pattern of means (refer to Table 6, on the following page) suggest no consistent trends, although for males, Competent/Compliant and Competent/Independent children had the lowest mean scores for perceived parental Ignoring in response to fantasized transgression. Order x Child-Behavior Groups. A significant multivariate Order x Child-Behavior Group interaction revealed a significant univariate effect for the dependent measure, Apology, which substantially loaded on Factor I, Competence-Incompetence. The univariate interaction effect for Apology, 5(3, 101) = 4.84, p<:0.0035, was examined further via simple effects tests. (See Appendix L for elaboration of these analy- ses.) These tests indicated that individual comparisons between group lilll“!.‘.lllllu ‘Illllll‘iltl‘u‘h' ‘0-.. hU-‘TI‘: .a.‘ RAIN-\Ieu..\r 65 .moososoeewu ososm usoowevsowm oz .Fososooiepom so s:_uos uso .so_oow__emo some use uooe .oosouvo>o -sooossso .oosowposiepom so sopusspuos so zop uooos use: as: sosupwso u A>Hv szosuso_zlo>wmpsssH .Fosusooaepom so aswuos uso .sovoowpweeo some use uooe .oosouwo>oisooososo .oosoWPosie—om so sowsisswuoe so goes uouos use; osz sosupwso n AHHHV osousooousaiusoooosou .oosouwo>olsooosooo uso ooso_~os-epom so Eswuos uso .sowooWmeeo smog uso uooe .Fosusooiepom so 3o~153Puos so 3o~ uouos osoz osz sosupwso u AHHV o>woo=smwaiuo>po>sHumPom .oosou_o>oisooosooo uso oosowposiepom so anuos uso .sovoowaeeo some uso uoos .Fososooimpom so sm_susuwuos so saws uopos use: osz sosupwso n AHV osowpssooiusouoosou mp mF op PF up ~_ mp up 2 om.~ ou._ m~.F oo.~ om.~ om.F op.p om.o sowpow>oo usousoom mn.F om.o mm.o Fm._ mm.~ ~_._ oo.F Pu.o sooz mswsosmH mopoz. mowoeoe moeoz mohoeom more: mopoeou mopozl, morosoe 1>H awe Ten .H mmzosw xom use goose an uovewmmopu assume: osvsosmH ms» mo msowuow>oo usousoum uso msooz .o o—oop 66 conditions within each order of task presentation were appropriate. Within Order A (modified MST followed by Deviation Doll Play), children in Group I made significantly more Apology responses to fan- tasized transgressions than did all other children in the study (see Table 7 for individual cell means classified by order of task presenta- tion and group). Competent/Independent Children (Group III) also had significantly higher mean scores for Apology, than did Impulsive] Withdrawn children (Group IV). The pattern of means suggests a decreas- ing trend in use of Apology responses to fantasized transgressions within Order A, among groups. Specifically, Competent/Compliant chil- dren had the highest mean scores for Apology, followed by Competent/ Independent children and Self-Involved/Disruptive children, with Impulsive/Withdrawn children showing the lowest mean usage of Apology. Within Order 8 (Deviation Doll Play followed by the modified MST), simple effects tests of group conditions yielded non-significant results (see Appendix M for more details regarding this analysis). The pattern of means, however, is similar to the pattern obtained with- in Order A. Competent/Compliant (Group 1) children obtained the highest mean scores for Apology responses. However, within Order 8, Competent/ Independent (Group III) children displayed the lowest mean usage of Apology. Thus, in both orders of task presentation, Competent/Compliant children used Apology in response to fantasized transgressions more than did all other children in the study. However, differences between groups were only significant for children within Order A. 67 .oosoowsvsmwmsos mouosoo .. o .mo.o v s an oosoowewsmvm moposooo .emz uowevuos xs uo:o__oe mops Fpoo sowoow>oo n m souso .xops Fpos soroow>oo as uo:oppoe hm: uoeewuos n < souso .Fospsooue—om so Esvuos uso .sowuowpwseo some use uoos .oosouwo>o Isooossoo .oosowposumpom so :opue:_uos so :o— uooos oso: oz: sosup_so n A>HV s:osusuv3-o>_m_soEH .Fososooiepom so sswuos uso .sowuowpveeo some use uoos .oosouwo>oisooossoo .oosowposiepom so sowsussmuos so sows uouos oso: os: sosupwso n AHHHV osousosousguosoooosou .oosouwo>otsooossoo uso oosowposueFom so Eswuos uso .sowuovasso soon .uso uooe .Posusoo1epom so :oF123Puos so :op uouos oso: os: sosupwso n AHHV o>wos=smwoiuo>Po>sHieFom .oosoueo>oisooossoo uso oosowposie_om so sawuos uso sowooWFwseo some uso uooe .Pososoouepom so sowsisspuos so sows uopos use: on: sosuFVso n AHV oso__osoouosouossou mp mp Np mp op o_ up mp z o~._ mm.o mm.o o~.~ om.o Fm.— on.P o~.P sowuow>oo usousopm mm._ oopm.o mm.— omm.p so.F suo.p no.~ ooo.m sow: xmopom< imn smuso < sousb m souso < souso m souso < souso .misousP < soWso >~ The“ MP H. wmoosu sowuousomoss some mo souso uso msooso so uowmwmmopo osomooz :mopos< «so mo msowoo~>oo usousopm uso msooz .u o_soe 68 Sex x Order x Child-Behavior Groups. A significant multivariate Sex x Order x Child-Behavior Group interaction reflected a marginally significant univariate effect for the dependent variable, identification with the Aggressor, which loaded substantially on Factor I, Competence- Incompetence. The univariate interaction effect for Aggressor, [(3, 101) = 3.27, p<=0.03, was examined further via simple effects tests (see Appendix M for a more complete presentation of these analy- ses). These tests indicated that individual comparisons between group conditions were appropriate, but not within conditions of sex, for Order A. The combined mean scores for groups (within Order A) are as follows: Group I (if = 8.95"); Group II (T = 14.96b); Group III (Y'= 15.86b), and Group IV (7'= 24.37a). Children in Groups I, II, and III, had significantly lower mean scores for identification with the Aggressor role in fantasy on the modified MST, than did children in Group IV. Impulsive/Withdrawn (Group IV) Children showed the highest mean usage of the Aggressor category. Conversely, Competent/Compliant (Group I) children had the lowest mean scores for Aggressor identification, significantly lower than did children in Groups II and III. Thus, within Order A (modified MST followed by Deviation 0011 Play), Competent/Compliant children (Group I) chose to identify with the Aggressor role in fantasy on the modified MST, significantly less than did all other children in the study. Impulsive/Withdrawn children aDenotes significance at p:0.05. bDenotes nonsignificance. 69 (Group IV) chose to identify with Aggressor role in fantasy, signifi- cantly more than did all other children in the study. There were no significant sex differences among children in preference for identifi- cation with the Aggressor, within Order A. Within Order 8 (Deviation 0011 Play followed by the modified MST), simple effects analyses indicated that individual comparisons between group conditions for females and males, separately, were appro- priate. Individual cell means for Aggressor responses classified by sex and child-behavior group within Order 8, are presented in Table 8, on the following page. For females, within Order B, Groups I and III had significantly lower mean scores for identification with the Aggressor, than did females in Groups II and IV. Thus, Competent/Compliant and Competent/ Independent females, within Order 8 (Groups I and III), chose to identify with the Aggressor role in fantasy on the modified MST sig- nificantly less than did Self-Involved/Disruptive or Impulsive/Withdrawn females (Groups II and IV). For males, within Order 8, the pattern of results was somewhat similar to those obtained within Order A (again, refer to Table 8 for a presentation of individual cell means). Competent/Compliant males (Group I) chose to identify with the Aggressor role in fantasy, sig- nificantly less than did males in all other groups. There were no significant differences between males in Groups II, III, and IV. Im summary, both Competent/Compliant (Group I) males and females across both orders of task presentation, obtained the lowest mean >2 oops—LPWVCPU '— Lao—GLC {—5,sz CL:....C~...E LCMVOL~S~J< NIH tea M..Rv%¢€h>avfl~ tie-ACOE-‘qu. 3:7 ...l:n~.i§< .hw hexaneh 7O .uosuowevsovmsos muuosuoo .mo.o.w s on uosuowstmwm muoosuou .ems uosssuoe sa uozopsos sass s_os eosoossos n m Loose .FososooieFum so Esvuue usu .sopooppwsso suus uso uoos .uosuuwo>u -sooososo .uosuWPusispum so :opiss_uus so :op uuuos usu: os: susquso u A>Hv s:ususp_3iu>_mpsssH .Fosusoois_um so sswuus uso .sowuoPFPeso suus uso uoos .uosuuwo>uisooosssu .uoso_pusie_um so smvsisswuus so sows uuuos usu: os: susupwso u AHHHV usuususuusmiusupuosou .uosuuvo>o-souososo usu uosowpusispum so aswuus usu .sowoow_?ssu suuo use uooe .Posusooispum so :opiE=suus so :op uuous usu: os: susquso u AHHV u>spo=smwoiuu>—o>s~is_um .uosuuwo>oisooososo uso uosoWFusispum so Es_uus usu sowpowpwseu suuo use uoos .Fospsooispum so sowsiE=Fqu so sows uuuos usu: os: susuFVso u AHV usuwpseoUTosuuuosou m m n m m m n n z o~.m uu.~ mo.~ NP.~ mm._ o~.m ~N.N PF.P sowpuw>uo usousoum soo.PF sou.“ opm.m ooo.m omu.op smm.m uum.~ upn.m sou: sommusmm< mupuz, mupusus mupuz murueus more: muroeuu mupuz mubusus >e HHfi HH H mnsosw xum usu ozosu mo uurmwmmopu m suuso sssov: uszmuuz sommusmm< use mo msovuow>uo usousoum usu msuu: .m u—ouh 71 scores for identification with the Aggressor role on the modified MST. Both Impulsive/Withdrawn (Group IV) males and females obtained the high- est mean scores for identification with the Aggressor role on the modi- fied MST, within Order A only. Within Order 8, there were sex differ- ences in the highest mean usage of Aggressor responses among groups, i.e., Self—Involved/Disruptive (Group II) females showed the highest mean scores for the Aggressor category, while males in Group IV had the highest mean usage of Aggressor responses. Also, a Significant multivariate Sex x Order x Child-Behavior Group interaction reflected a significant univariate comparison for the dependent variable, Child's Denial of Fantasized Transgression, which loaded substantially on Factor VIII, Dggjgl, The univariate inter- action effect for Child's Denial of Fantasized Transgression, {(3, 101) = 5.25, p<:0.0021, was examined further via simple effects analyses (refer to Appendix M for a complete presentation of these results). These tests indicated that within Order A (modified MST followed by Deviation Doll Play), individual comparisons of group conditions for males and females, separately, were appropriate (see Table 9). For females, within Order A, Groups I and II were significantly different from Group IV. However, there were no significant differences between Groups I, II, and III, and Group III was not significantly dif- ferent from Group IV. The pattern of means does reveal that the lowest mean usage of Denial in response to fantasized transgression, occurred for Self-Involved/Disruptive females (Group II), while the highest mean scores for Denial, were obtained by Competent/Compliant (Group I) and Impulsive/Withdrawn (Group IV) females. 72 .muosusuesvu soosu usoo—e—soemsos muuosuoo .mo.o w s «o —o>up ousuuve_sm—m muoosuoo .5: 8:32. .3 35:8 so: :8 5:258 n u souso 43:. :8 852:3 3 85:8 5: uostue. .. a toss .Fosusooiepum so sspuus use .so_oo_ppemu suuo use uooe .uosuuvo>oisooossoo .uosu_Fus-spum so :opissvuus so :oF uuuus usu: os: susupvso u A>HV s:ususueziu>*mpoos~ .pospsooisFum so sawuue use .so_uqu—uwo suuo uso uooe .uosouvo>oisooossou .uosuvpusis—um so smvsusovuue so sm_s uuuus usu: os: susup_so n AHHHV usuusuouusuiusupuoEoo .uusou¢o>oisouososu uso uosowpusimpum so Eavuus use .sowuo_pvsmo sous use uooe .Fosusooie_um so :opisovuus so :op uuuos usu: os: susup—su n Anny u>vuozsmpoiuu>~o>smiwpum .uusuu_o>oisooososu usu uosuPFusim—um so asvuus usu sowuuppseso suus use uoos .Fosusoois_um so smvsissvuue so so_s uuuos usu: os: susupvso a Any usuvposoUTusuuuoeou m m op a s m n o m u m m a s o n z mu.— mm.P Fu.— ow.p m¢.o em.~ on.o an.p os.o vo.p -.p mm.o m—.— ou.o o—.F m¢.o sowuu_>uo usuusoum umm.~ n¢¢.~ noo.— unm.~ nm~.o nov.~ unm.o nom.~ omn.o nup.~ nmm.~ oom.o ame.~ op~.o umm.o umm.~ sou: u_o: u_oeuu u_o= u_usuu u_oz u_ueuu u_o= u_oeum u_m= u_osuu u_om u_usun u_m= u_oeuu u_m= u_ueuu .mlsuumo. < suuso .m suuso z suuso _misuusc. < suuso .m souso in smmso 1>~ “an .HH Hi ammoso 85353.... 5.2: so .595 use £3 .52.... a 8:328 23...: 3:8 as so 532:8 28:5 use 28: .o oz: 73 For males, within Order A, results were different and more con- sistent. Group I was significantly different from Group II, with Group II also significantly different from Group III. Males in Groups III and I had the lowest mean usage of the category Denial in response to fantasized deviant acts, while males in Groups II and IV obtained the highest mean usage of Denial responses. Thus, Self-Involved/Disruptive and Impulsive/withdrawn males expressed the highest mean scores for Denial responses to fantasized transgression in Deviation Doll Play. Nithin Order 8 (Deviation Doll Play followed by the modified MST), simple effects analyses reflected a nonsignificant Sex x Child-BehaVior Group interaction for females, with a marginally significant simple group effect for males (refer to Table 9, page 72, for individual cell means). Groups I, II, and III were significantly different from Group IV. Males in Group IV had the highest mean usage of Denial re- sponses, while males in Group III had the lowest mean usage of Denial in response to fantasized deviant acts, similar to the pattern found for males, within Order A. Thus, Self-Involved/Disruptive and Impulsive/Withdrawn males (Groups II and IV) used Denial significantly more in response to fantasized transgressions, within both orders of task presentation, than did all other males in the study. As stated previously, no consistent pattern was found for female subjects in the use of Denial responses. 74 Analysis of Variance of Victim Dependent Measure A separate analysis of variance revealed a significant Sex x Order x Child-Behavior Group interaction for the dependent measure, identification with the Victim4 role, _F_(3, 101) = 4.82, p<0.0035, and this univariate interaction effect was explored via simple effects tests (refer to Appendix M for a presentation of these findings). Hithin Order A (modified MST followed by Deviation Doll Play), these tests yielded nonsignificant results, and no further comparisons between groups, within each condition of sex, were conducted. However, within Order 8 (Deviation Doll Play followed by the modified MST), simple effects tests indicated that individual compari- sons between groups for females and males, separately, were appropriate (see Table l0, on the following page). For females, within Order B, Group III was significantly higher than all other groups in mean usage of the category, Victim (7'= 10). There were no significant differences between Groups I, II, and IV. For males, within Order 8, individual comparisons between groups yielded nonsignificant results. The mean scores for Victim responses within males and females under both orders of task presentation are not markedly different (see Table l0), and are extremely low, with the exception of Competent/Independent (Group III) females. Thus, Competent/Independent females who experienced Deviation Doll Play first (where they were often the "victims” of fantasized 4A separate analysis of variance was conducted because the scoring categories were linearly dependent. 75 .ooseovmvso—msos mousseso .mo.o.w s we ooseo.m—sm*m mouosooe .hmz um—mpuoe an uozoppom zaps ppoo soruev>mo n m souso .xeps ppoo sowuew>oo x5 uoxop—om hm: uovvvuoe u < smusc .Fosusoulspem so Serums use .sopue—pwsse sees use uooe .uuseuvo>e-soeossoe .eosepposls—om so sopue=suos so sop uopes use: osz sosupsso a A>HV szesususzuo>.m—oosu .—osusooumpom so Serums use .sovuewpwwse some use uoos .ooseu¢o>eusoeossoe .oosev—osuspom so sassusowuoe so sows ueues use; osz sosupsso a “many usousosoussuesmpooEou .ooseuwo>e-soeossse use mosespesuspom so assume use .soreevppsme some use uoos .posusoo-s—om so zo_-s=_uoe so sop uoees use: osz sosuprsu a Aunv o>susssm.s-uo>~o>ss-s_om .ooseu*o>eusoeossoe use moses—esus—om so Ensues use so.ueVFssme sees use uooe .posusoo-spom so sass-sosuoe so sass uoues use: os: sosup_so u AHV usewpssouuusouoosou a a op o u m a o m o m m n u o s z mm.~ om.~ om.F om._ mn.p mm.m no.~ om.p m~.~ F~.~ um.m m~.o em." mu.— o—.~ oo.— sowues>oo useuseum ms.p so.~ oe.p ~_.F oo.~ eoo.op ~e.~ mm.m m~.m sum.p mm.~ oo.m mp.~ sm~.— mw._ oo.~ sew: o_m: o_esou o_m: «_eeou o_m: o_esou o_ez o_e5ms o_m: o_eeou o_e: o_esou u_ez o_eeou o_mz e_eeeg .mlsouso < souso .mimumso s souso .mlsouso s sommo m souso <.souso >H HHH HH H moaosw sosueusomoss see» so souso use .xom .sooso so uopmvmmepu «someoz evuo_> use “to esopues>oo useuseam use «see: .o_ o—oeh 76 parental punishment) chose to identify with the Victim role on the modified MST significantly more than did all other children in the study. Factor Analysis of Dependent Measures Relations between the 23 dependent measures were explored via a principal components factor analysis (R2 was used as the estimate of Communality; eigen value 3_l); factors were rotated to a varimax solu- tion. The categories generated 8 somewhat orthogonal independent factors which accounted for .70 of the cumulative, proportionate variance. Appendix L presents a summary of factor loadings for all 8 factors. Factor I: Competence-Incompetence The strongest factor (O.l953 of variance) is comprised of 8 cate- gories which form a conceptual unit (refer to Appendix I, Table I, for their mean frequencies of usage). For Factor I, the highest loading, 0.86, was obtained for Category I, "Provider-Protector". Other cate- gories which loaded substantially and positively on Factor I are: (9) Perceived Parental Reasoning in response to fantasized transgression (0.75); (l4) Apology by child for fantasized transgression (0.69); (15) Confession by child for fantasized transgression (0.59), and (10) Perceived Parental Reflection of Feelings/Forgives, in response to fantasized transgression (0.57). Categories which loaded substantially and negatively on Factor I are: (2) "Aggressor" (-.Bl); (l7) Hiding by 77 child in response to fantasized transgression (-.55), and (18) Crying by child in response to fantasized deviant acts (-.54). Thus, children who preferred to identify with Provider-Protector roles in fantasy on the modified MST, also tended to perceive parental‘ responses to fantasized transgression in Deviation Doll Play as incor- porating either reflection of feelings/forgiveness, and/or reasoning responses (i.e., giving explanation, setting limits, offering alterna- tive behaviors). These children also tended to react with Apology behaviors (i.e., "I'm sorry I spilled the milk, Mommy"), and Confession behaviors (“I did it, Daddy"), either spontaneously or when confronted by parental figures, for fantasized deviant acts. Thus, children who preferred to behave in a nurturant and protective manner in structured fantasy play, also demonstrated perceptions of parental figures as "empathic" and/or “forgiving" and/or "rational and controlling". These children also demonstrated behaviors in structured fantasy play con- sistent with "superego strength" (Gould, l972). Given theoretical formulations of effective ego and superego functioning, such children might be termed “competent". Conversely, children who preferred to identify with "Aggressor" roles in fantasy on the modified MST, also tended to react with Hiding and Crying behaviors in response to fantasized transgressions in Deviation Doll Play, e.g., "child doll runs away from home, crying that he will never come back." Thus, children whose preferred mode of self- representation in structured fantasy play was identification with the "Aggressor", also tended to demonstrate avoidance and helpless behaviors 78 in response to fantasized transgressions. This might be viewed as reflecting "ineffective" or "defensive" ego and superego functioning. Thus, such children might be termed "incompetent". Factor II: Child's Extreme Behavioral Reactions to‘Fantasized Transgression The second factor which accounted for only 0.0718 of the total variance, was comprised of 2 categories (again refer to Appendix I, Table I, for a presentation of mean frequencies and ranked order with which these responses were emitted). The highest loading for Factor II was Category 20, "Mute, Silent, Withdrawal“ (0.82), which typified the child doll standing silently, seemingly immobilized, after parental discovery of the fantasized transgression. Category l9, “Continued Deviant Aggression“, loaded -.66 on Factor II, and represented a wide range of aggressive, destructive behaviors (e.g., “killing all people in doll house," “throwing out all doll furniture," "dumping water over house to simulate a fantasy flood,“ "picking up doll house and turning it over to symbolize a tornado," "putting people in the oven or refrig- erator to burn or freeze them," and "committing suicide by shooting self, or drinking poison"). Obviously, these two categories represent polar opposites in terms of impulse control in response to fantasized transgression, i.e., total inhibition vs. impulsive, destructive aggression directed towards objects, and self, and/or others. 79 Factors III and 1V5: Agent(s) of Punishment Factors III and IV (which account for 0.0731 and 0.0786 of the variance, respectively) comprised three mutually exclusive categories, i.e., Category 21, Mother (alone), Category 22, Father (alone), and Category 22, Mother and Father (together, agreeing) (refer to Appendix I, Table I, for ranked order of mean category usage). The highest load- ing for Factor III was Category 22, Father (alone) (0.92), as primary agent of punishment following fantasized transgression in Deviation Doll Play. Category 21, Mother (alone) loaded -.53 on Factor III, and 0.76 on Factor IV, which also included Category 23, Mother and Father (together, agreeing), loading -.88 on Factor IV. From Factor III, it appears that choice of "mother" or “father" (alone) as agents of fan- tasized punishment were negatively associated with each other; whereas, given Factor IV, choice of both parents, each supporting one another, was negatively associated with choice of "mother" (alone) as the primary agent of punishment. Since these categories were mutually exclusive within a given Deviation Doll Play episode, the data suggest that if children chose one parent (alone) as the agent of punishment. they were less likely to choose the other parent in subsequent Deviation 0011 Play episodes. More specifically, if children chose both parents (who were consistent in their responses to fantasized transgression), they were less likely to choose mother (alone) as the primary agent of punishment. 5Included as separate multivariate analyses because the scoring categories were linearly dependent. 80 Factor V: Perceived Parental Response to Fantasizéd'Transgression The fifth factor which accounted for 0.0894 of the total variance was comprised of three categories (refer to Appendix I, Table I, for a presentation of mean frequencies of these responses). The highest loading for Factor V was Category l3, Gets Away with It (0.76), followed by Category 11, Ignoring (0.68), and Category 5, Verbal Punish- ment {-.66). Quite obviously this factor indicates that children who perceived parental figures as Ignoring vs. Verbally Punishing them (i.e., "yelling", “screaming") for fantasized transgressions in Devia- tion Doll Play, also were more likely to report that the child doll "gets away with it"; i.e., the deviation is never discovered, and, in fact, the child escapes punishment, reflecting the child's use of denial in fantasy. Factor VI: Fixing, The sixth factor which accounted for only 0.0631 of the total variance, was comprised of two categories. Category 16, Spontaneous Fixing, which loaded 0.72 on Factor VI, represents the child's attempts to "undo" or "restore" the results of his deviant behavior, e.g., “child spills glass of milk, wipes it up with towel, and pours himself another glass, without parental knowledge,“ Category 8, Forced Fixing, which loaded -.69 on the same factor, represents parental efforts to order and/or help the child make restitution for the consequences of his deviant acts, e.g., "child spills glass of milk, parent tells child to clean it up, and/or helps the child clean it up." Forced Fixing is 81 really a deceptive category, since it includes both parents verbal commands, often delivered in a harsh and threatening manner, and attempts to help the child comply with parental sanctions by modeling "fixing“ behavior for him/her. Thus, children who preferred to spon- taneously restore or "undo" the consequences of fantasized transgres- sions (either due to defensive "undoing" strategies or representing more advanced superego functioning (i.e., "a wish to please the parent") were less likely to perceive parental figures as either modeling "fixing" behaviors and/or directing them to perform such behaviors. Factor VII: Identification with the Victim Role The seventh factor which accounted for only 0.0655 of the total variance, was comprised of one category, 3, identification with the Victim role, on the modified MST, loading —.77. Victim items on the above test required the child generally to be the object of "oral- sadistic aggression" from an animal toy, e.g., "pretend a crocodile bites you." The next highest factor loading (non-significant) was for Category 6, Isolation (parental response to fantasized transgression) -.41. This might suggest a trend implying that children who preferred to identify with the Victim role in fantasy on the modified MST might also tend to perceive parental figures as using love-withdrawal tech- niques of punishment (i.e., "go to bed, sit in the corner") in response to fantasized transgression in Deviation Doll Play. However, this conclusion is purely speculative, given the low factor loading of Category 6, Isolation, and the relatively low mean usage of the Victim category, as well as the low percentage of total variance for which Factor VII accounted. 82 Factor VIII: Denial The last factor obtained from factor analyzing the dependent measures accounted for only 0.0694 of the total variance, and was com- prised of 2 categories. Category 7, Parental Denial of Privileges, had the highest factor loading, 0.77, followed by Category 12, Child's Denial of Fantasized Transgression, 0.61. This factor suggests that children who perceived parental responses to fantasized deviant acts as "withholding", or "depriving" (e.g., "you can't watch T.V.," or "no more juice for you"), also tended to deny (e.g., "It wasn't me who spilled the juice," or "I don't know how the lamp fell down"), responsibility for behavior that conflicted with typical parental sanc- tions. Thus, one might speculate that "low-withdrawal" forms of punishment, specifically, “Denial of Privileges", may interfere with development of internalized superego functioning in children, and may be associated with children's use of denial as a defense mechanism. CHAPTER IV DISCUSSION Relevance of Results for Questions and Hypotheses The first question pertained to what, if any, significant differ- ences existed among the groups of children with regard to preference for self-representation in structured fantasy play. Competent/Compliant children (Group I) as compared to children in other groups identified significantly more with "Provider-Protector" roles. However, one might have expected Competent/Independent children (Group III), who were rated highly on four out of the five social behaviors, also to manifest a preference for identification with the "provider-protector role" in fantasy. Since there were no significant differences among children in Groups II, III, and IV in mean usage of "provider-protector" responses, the data are equivocal with respect to confirming the hypo- thesis that more "mature" or "socially competent“ children will prefer identification with nurturant and protective roles in fantasy. Impulsive/Withdrawn children (Group IV) had significantly higher mean scores for identification with the "Aggressor" role in structured fantasy, than did all other children in the study (but only within Order A). For females with Order 8, both Groups II and IV had signifi- cantly higher mean scores for choice of "Aggressor" identification, 83 84 when compared to females in Groups I and III. For males within Order 8, there were no significant differences among Groups II, III and IV, but Competent/Compliant males (Group I) had significantly lower scores for "Aggressor“ behaviors, than did males in any other groups. Thus, within Order A (modified MST followed by Deviation 0011 Play), the data do seem partially to confirm the hypothesis that children who are rated as “less mature" by their teachers would tend to prefer identifi- cation with aggressor roles in structured fantasy play. Within Order 8 (Deviation 0011 Play followed by the modified MST), results are less clear and sex of subject significantly interacted with order of task presentation to confound results pertaining to preference for the “Aggressor" role in fantasy. The data for females follows the results obtained with respect to Order A; i.e., "less mature" tended to prefer identification with the Aggressor role. The results for males are more mixed. However, it is important to note that Competent/Compliant males (Group I) had the lowest mean usage of the "Aggressor" category. The expectation that Group II and IV children, i.e., "less mature“ children would also manifest a preference for the "Victim“ role in structured fantasy play was not confirmed. The basis for this specula- tion stemmed from the hypothesis that children who were rated "low" on initiative dimensions and medium on control dimensions or "low" on most of the social behavior dimensions would tend to withdraw from acting out aggressive impulses, and instead, might direct them internal- ly, towards the self. The only significant group difference for identi- fication with "Victim" roles occurred for Group III females, within Order 8, whose mean scores for "Victim" responses were significantly mm c fact 1 small withir justii transg course female ing ag 0f the 0f "Pr fantas We 5 there ":19er fBt‘enc 85 higher than those obtained by all other children in this study, i.e., §'= 10. This result is difficult to evaluate, given the relatively low mean category usage for “Victim" responses, i.e., §'= 2.30, and the fact that this was a third-order interaction effect, with relatively small cell frequencies. One speculation for this finding, is that within Order 8, these females experienced themselves as the "victims“, justifiably or unjustifiably, of parental punishment for fantasized transgressions, and perhaps, given current political and social dis- course on the development of the "feminine" role, these "mature" females are really passive and compliant in the classroom, internaliz- ing aggression felt towards others, an important step in the evolution of the superego (A. Freud, 1936). The second question focused on sex differences between choice of "Provider-Protector", “Aggressor", or "Victim" roles in structured fantasy play. Females did tend to choose the “Provider-Protector“ role significantly more frequently than compared to males. However, there were no significant sex differences in identification with the “Aggressor" or "Victim“ role, in this study. The third and fourth questions pertained to what, if any, dif- ferences existed between groups in behavioral reactions to fantasized transgressions. Competent/Compliant children (Group I) more frequently displayed reactions of “Apology", within Order A only, although group means are in the expected direction, i.e., Group I (highest) followed by Groups III, II, and IV, within Order B, as well. Again Groups II and III did not significantly differ with respect to use of “Apology", 86 although Group IV was significantly lower than all other groups. There were no significant group differences with respect to the use of Confession following fantasized transgression in Deviation Doll Play. Thus, there is some, although tentative, support for the hypothesis that children perceived by their teachers as "mature", would tend to use prosocial behavior, i.e., I'Apology", to deal with fantasized deviant acts. The expectation that less “mature" children (Groups II and IV) would tend to use avoidance behavior; such as, Hiding, or Denial, in response to fantasized transgression was not confirmed by the data. However, even though there were no significant differences between groups in Hiding responses, the means were in the expected direction, i.e., Group IV (highest), followed by Groups II, III and 1. Regarding the use of Denial in response to fantasized deviant behavior, there was a significant third-order interaction effect between group, sex, and order of task presentation which revealed no consistent, meaningful, and interpretable pattern of results. However, for females, within Order 8 only, Group I girls obtained significantly lower scores for Denial than did the females in the other groups. The only other significant difference among groups in behavioral reactions to fantasized transgressions, was for the category, “Continued Deviant Aggression". Impulsive/Withdrawn children (Group IV) displayed significantly more aggressive and destructive behaviors following fantasized deviant acts as compared to Competent/Compliant children (Group 1). Again, children in Groups II and III did not 87 significantly differ from children in Group 1, although mean scores for Groups II and IV were higher than those for Groups I and III. Finally, there was a significant main effect for sex for the variable, "Spontaneous Fixing". Females preferred to cope with fantasized transgressions by "undoing“ or "restoring" the consequences of deviant behavior significantly more than did males. The fifth question focused on group differences in perceived agent(s) of fantasized punishment. The expectation that "less mature" children (Groups II and IV) might perceive symbolic parental figures as inconsistent with respect to punishment, was not supported. In fact, category 24, "Mother and Father, together (disagreeing)", was only used by two subjects in this study, and thus, was not included in data analysis. There were no group differences of any kind, in perception of agent(s) of fantasized punishment in Deviation Doll Play. The sixth question explored whether or not sex differences related to which parental figure was perceived as the symbolic primary agent of fantasized punishment. Males perceived the father figure in Devia— tion Doll Play as the primary agent of punishment significantly more than did females. No similar effect for the same-sexed parent being perceived as agent of fantasized punishment was found for females. The seventh question referred to what differences might exist between groups in perception of mode of fantasized punishment. The expectation that "more mature" children might perceive symbolic parental figures as employing "Reasoning" approaches more frequently in response to fantasized transgression was only partially confirmed. iii an; Both Competent/Compliant males and females (Group I) perceived signifi- cantly more “Reasoning" responses to fantasized transgression in Deviation Doll Play, than did Impulsive/Withdrawn children. However, there were no significant differences between Groups II, III, and IV for males, and Group II females used "Reasoning" responses nearly as frequently as Group I females did. There were also no significant group differences with respect to any other categories of perceived mode of punishment following fanta- sized transgressions in Deviation 0011 Play. Thus, the expectation that "less mature" children might perceive symbolic parental figures as more frequently employing "power-assertion" or "love-withdrawal" techniques of discipline in Deviation Doll Play, was not supported. There were no sex differences either in the perception of mode of fantasized punishment in Deviation Doll Play. Relevance of Results for Past Theory and Research Gould (1972) theorized that a child's mode of self-representation in fantasy conveys his/her predominant affective strivings to emulate and incorporate the behaviors of "significant others". In the case of "Provider-Protectors", it appears that a preoccupation with nurturant as opposed to aggressive imagery tends to be associated with perceptions of parents in fantasy as demonstrating aspects of "warmth and control". Baumrind (1975) suggests that these aspects of parental discipline in reality, called "authoritative parenting" serves to promote the growth of social competence in children. They are also the main bases fOr 89 effective modeling of all types of parental behavior by children (Bandura a Halters, 1963), according to Social Learning Theory. This study did not investigate actual parental behaviors ex- pressed towards the children studied, and thus, these fantasies of parental discipline techniques may represent positive emulation or superego introjections of actual parental behavior (i.e., replication fantasies), or unsatisfied dependency longings (i.e., wishfulfillment fantasies). However, they would certainly be considered "healthy" wishfulfillment fantasies, or positive "superego-ideals“ by Gould (1972), Hoffman (1975), and Baumrind (1975). In addition, as Gould points out: . . it is clear that which motives will be considered primary instigators and which secondary accompaniments or consequences to given fantasy expression, is often a matter of type or level of analysis, or sometimes of interpretation bias ... the child's internal well-springs and external world experiences inter- mingle in various ways in fantasy expression to the enrichment of both sources of knowledge.... The two worlds of reality and imagination need never be as far apart as is often implied (1972. p. 399). Thus, future research is needed to substantiate Gould's hypothesis that children who identify predominantly with provider-protector roles in fantasy have experienced a favorable balance of positive, nurturing caretaking vs. negative, rejecting, caretaking experiences. However, the data do lend credence to her ideas of "superego constancy". "Provider-protectors", in this study, were significantly more inclined to cope with their own deviant behavior in fantasy by employing prosocial behaviors designed to further parent-child inter- action. Their coping strategies (i.e., "Apology" and "Confession“) to 90 meet fantasized parental socialization demands could be interpreted as Gould's "wish to please" symbolic parents, and thereby reduce fear or anxiety. The "wish to please" the parent is the foundation for "superego constancy" according to Gould, and is related to a "benevolent superego", or "limited vs. global self-condemnation" in the child. Since “Provider-Protectors" in this study “wished" to perceive parent figures as "warm", one might speculate that this derived from a more "benevolent" vs. "harsh" system of superego introjections. In addition, "Reasoning responses" (with low frequency, although significant) were attributed to symbolic parental discipline by “Provider—Protectors“. These aspects of "rational-inductive" discipline according to Hoffman (1975) are associated with effective moral internal- izat‘l on in children. Future research is needed to investigate whether Dre-oedipal fantasies of parents as "effective disciplinary agents" or "Superego constancy" does pave the way to post-oedipal ownership of mora'l dicta, or moral internalization. Identification with the Aggressor (A. Freud, 1946; Gould, 1972), theoretically reflects a relatively imnature ego, whose anticipatory Stance of attack to maintain self-preservation, renders the ego "incompetent", such that acting-out of aggressive impulses brings re- taliation, reinforcing basic mistrust in the environment (Erickson, 1950) - In this study, children who predominantly preferred to identify with the "Aggressor“ role in fantasy, also were significantly more 1'““filed to withdraw or become emotionally upset, i.e., act out crying episodes in which the child doll became inmobilized in response to its 91 own deviant behavior. The tendency for unorganized affect to overwhelm cognitive activity in evolving more effective c0ping strategies, and/or the need to engage in avoidance behavior suggests that preoccupation with aggressive imagery did not result in catharsis for these children. This is an important clinical distinction, since all children in this study chose to identify with the "Aggressor" role in fantasy, at least once. The differences between the children who opted for predominant identification with the "Provider-Protector" role vs. the "Aggressor" r01 3 in fantasy were twofold: 1) "Provider-Protectors" evidenced a wish to avoid punishment or reduce anxiety by struggling to please the symbolic parent figure, by employing prosocial coping behavior strategies, that utilized cognitive schemas to organize or control affect, and achieve conflict resolution/cartharsis. "Aggressors" tended to display unorganized affect expression and avoidance behaviors which terminated the fantasy, preventing conflict resolution/cartharsis, Possi bly due to experiencing "fluctuating certainty" (Gould, 1972), i.e. , "inability to distinguish between reality and pretend." 2) "Provider-Protectors" portrayed fantasized parents as "warm", i.e., "f0r91ving" and in "control", i.e., setting limits and providing reasons for the limits, in addition to using typical verbal and physical pu"lshlflent techniques, i.e., "yelling" and "spanking". Most "Aggressors" fantas‘Ies of parental discipline gull involved verbal and physical PWlShment responses, which were comnon to all children in the study. It ‘5 the absence of alternative cognitive schemas for dealing with fa"“1351 zed parent-child conflicts, or the affective interference with 92 the availability to call forth such schemas, which made the "Aggressor" fantasies qualitatively different from those of the "Provider- Protectors". It is not that "Aggressors" portrayed "cruel or unusual" punishment at the hands of symbolic parent figures. Rather, it is the lack of creatively elaborated fantasy, and perhaps the inability to engage in such fantasy due to "fluctuating certainty", which supports Gould's idea that these children do not possess the differentiated ego capacities for dealing with "drive-deterrence" and achieving “superego con stancy" . Sex Differences on Fantasy Measures As stated previously, females chose to identify with the "Provider-Protector" role in fantasy significantly more than did males. This finding supports traditional views of ”feminine personality trai ts", i.e., tendencies toward nurturant, affiliative, affectionate. behavior, consistent with anaclitic identification. However, there were no sex differences in preference for identification with the "Aggressor" role. Sample characteristics might account for the lack of Slgni ficant sex differences. The sample was composed mainly of lower middle or working-class children from urban families, with both parents usual ”Ly employed. These working mothers may have provided more "assertive or aggressive" role models, and made more demands upon their female children than do traditional middle-class mothers. Thus, females in 1“”53 study readily identified with the "Aggressor" role in fantasy although obviously less frequently than they did with the Provider- Pmtattor" role. 93 Females also chose "Spontaneous Fixing" in response to fantasized deviant behavior significantly more than did males. Hatfield, Ferguson, and Alpert (1964), who used Deviation Doll Play in studying mother-child interaction and the socialization process, found that "fixing responses were associated with maternal permissiveness, but also with some demands for achievement. The pressure for achievement occurred in a constella- tion with maternal ease and reward for good performance“ (p. 399). The use of Confession responses to deviation also occurred within the same constellation, and the authors implied that patterns of maternal ease and control (similar to Baumrind's "authoritative parenting") were con- ducive to well-socialized reactions; such as, making amends for trans- gressions. Also, Rebelsky, Grinder, and Allinsmith (1963) found that the use of confession to fantasized transgression predicted resistance to overt temptation for girls, only. Therefore, in this study, Factor I, Competence-Incompetence, which includes identification with the Provider-Protector role in fantasy, Apology and Confession Behaviors to fantasized transgression, along with ”Spontaneous Fixing“, used by females, may reflect a "coping style" that corresponds to appropriate sex-typed behavior that is consistent with females who have experienced "authoritative mothers" (Baumrind, 1975). Finally, males chose the father as the primary agent of fanta- sized punishment significantly more than did females in Deviation 0011 Play, possibly indicating that these males had completed the develop- mental task of identification with the father. There was no such trend for females. suggesting that they may have still been in the transition 94 stage from father identification to mother identification, and the process was not complete (Hatfield, Ferguson, and Alpert, 1964). Relationships Between Fantasy Measures and Teachers' Ratings ovaert SocialiBehavior The relationships between children's preferred mode of self- representation in fantasy, their behavioral responses to fantasized transgressions, and their fantasies of parental discipline, to Baumrind's dimensions of “maturity", i.e., self-control, self-reliance, approach tendencies, positive mood, and peer affiliation, were predict- ably more complex and difficult to assess in this study. As discussed previously, this study initially tried to parallel Baumrind's (1975) delineation of Pattern I, II, and III children, in the formation of groups. Methodological differences, however, appeared to prevent an exact replication of her findings; instead of three groups, the present research found a general factor of “high" vs. "low" adjustment that seemed to underly the teachers' ratings. Correlation coefficients for the five indices of social behavior were examined, and thus, four groups were identified, and these groups did differ somewhat in their responses to the structured fantasy tasks. The clearest pattern of differences occurred with respect to Groups I (Competent/Compliant children) and IV (Impulsive/Withdrawn children). Groups II (Self- Involved/Disruptive children) and III (Competent/Independent children) rendered data which revealed no consistent, interpretable, or meaning- ful response patterns. 95 The question arises whether "Competent/Compliant" (Group I) . would be clinically described as actually “passive-dependent" children who are overly concerned with the loss of affection, and who demonstrate affiliative behavior of compliance towards adults to satisfy pressing needs for nurturance. The preference for identification with the provider-protector role in fantasy then would also be a consequence of the internal press to satisfy dependency longings. Since females sig- nificantly preferred the provider-protector role, the argument could be made that "Competent/Compliant" children were children whose per- sonality resembled traditional feminine traits of "passivity and dependency". This argument has validity since teachers typically give high ratings to children whose behavior in the classroom reflects obedience, conformity and passivity vs. assertiveness, curiosity, and self-reliance. Hartup and Keller (1960) studied overt nurturant behavior in preschoolers and found that it was associated with seeking help and physical affection from adults. They concluded that children who gave "positive attention", affection, approval, etc., to others, were engaged in direct active attempts to gain adults attention and approval. However, these children generally appeared "outgoing" and "energetic" in their social interaction, and seemed to demonstrate a high activity level. Also, "nurturance" in their study, correlated most negatively with ”being near" behavior, i.e., simply maintaining physical proximity to the adult, behavior which seemed to signify passive indirect attempts to satisfy needs for dependency or succorance. Therefore, Hartup and 96 Keller (1960) suggested that the child's efforts to nurture others need not be interpreted as "passive dependency", but that "active outgoing styles" of achieving attention and approval were related to a general level of "sociability“. Hhile Hartup and Keller's study was not concerned with fantasy behavior, the behaviors defined as "nurturant" were similar to ones used by "provider-protectors" in the present study. Therefore, the argument could be made that preference for identification with the provider-protector role in fantasy represented children's attempts at “active ego mastery" of nurturance needs, and that preference for this "coping style“ in fantasy could reflect "covert rehearsal" (Maccoby, 1959) for achieving the same prosocial behavior in reality, reinforcing the teachers' perceptions of these children as "mature“. Indeed, Kagan (1955) found that a direct and positive relationship existed between overt and fantasy aggression when a) the fantasy stimulus suggested aggressive content, and b) the overt and fantasy behaviors were similar in mode of expression and desired goal. Perhaps the same case can be made for nurturance. Specifically, since the modified MST definitely suggested “nurturant behaviors" in provider-protector items, and the category of “peer affiliation" also reflected prosocial behavior towards others, perhaps for Competent/Compliant children (Group I), there was a direct relationship in coping with needs for nurturance in fantasy and overt social behavior. Unfortunately the data remain equivocal with respect to this conclusion, since Competent/Independent Children (Group III) did not 97 manifest any clear pattern of responses to the structured fantasy tasks. Hithin Order A (modified MST followed by Deviation 0011 Play) Group III children (considered most "highly-adjusted" by teachers) did not evi- dence a predominant preference for mode of self-representation on the modified MST. One tenuous speculation is derived from Gould‘s (1972) naturalistic observations of fantasy play in preschoolers over a seven year period. Occasionally Gould observed children who seemed to be able to utilize all three modes of identification, i.e., provider- protector, victim, and aggressor, equally well. An example would be a child who could become a "nurturant bee providing honey when secure and happy", a "stinging bee when threatened", and a “hungry bee when there was no honey“. Gould (1972) concluded that this ability to use self- representations flexibly reflected a high degree of integration of self- perceptions, and indicated a higher degree of cognitive-affective func- tioning than even the "provider-protector" children. Perhaps for Competent/Independent children, flexibility and integration of self-per- ceptions was the question, not predominant mode of self-representation. Interestingly enough, within Order 8, Competent/Independent females (Group III) chose to identify with the victim role over ten times more than any other children in the study. (It is the author's conclusion administering the modified MST that most children were frightened by the victim items, since in Gould's terms, they offered no possibility for "distancing“ the fantasy content, and may have provoked “fluctuating certainty", e.g., "can that snake actually bite me?") Perhaps, these girls were more "curious“ or tantalized by victim items, 98 and had adequate ego coping mechanisms enabling them to "risk" finding out whether the fantasy stimulus was actually harmful. Perhaps this is why the teachers rated them relatively higher on "self—reliance" and "approach—avoidance" than the "provider-protectors" (Group I) children. In addition, A. Freud (1936) and Gould (1972) describe identification with the victim role as a transitory state in superego development, necessary to the later emergence of self-criticism. Since these females experienced Deviation 0011 Play prior to the modified MST, perhaps the intensity of feelings aroused in them by the same-sexed doll transgres-. sing and being punished in a typical conflict situation carried over into the next fantasy task. That is, these girls could not "distance" themselves from the previous fantasy, and aggression felt towards symbolic parent figures in response to punishment was transformed into indirect aggression directed at the self. If these females also coped with aggressive impulses through internalization in the nursery school, then teachers would possibly see these children as "mature“, since such children would tend not to be "disruptive“. Finally, Competent/Compliant children (Group I) showed advanced "superego coping styles", in addition to the prosocial "coping style" mentioned with respect to needs for nurturance. The use of Apology in response to deviant behavior in fantasy involved statements of personal feelings/wishes by the child, i.e., "I'm sorry"; "I wish I hadn't done it“, usually followed by "forgiveness" (albeit wishfulfillment) by the symbolic parent figure. This sequence was usually preceded by adult behavior, incorporating elements of rational-inductive discipline, 99 i.e., "reasoning", "you should not take the baby's toy because then he won't be able to find it or play with it.“ (Admittedly, this could be the child's "wish" as well as "reason for not performing the behavior" attributed to the parent figure.) However, as Gould noted, many children display "the voice of the superego" in their fantasy play. In these fantasies, the "voice of the superego" was "bossy" and "scolding" at first, i.e., "yelling" responses if verbal and physical punishment were used, but noticeably "firm“, not "gleeful" when "reasoning" responses were given, and finally “warm", if "forgiveness" was rendered, subsequent to the child‘s "contrite" apology response. These fantasies also generally ended with a positive interaction occur- ring between parent and child figures. At other times these children might heave a sign, as if it was a struggle, and they were glad to be finished. Thus, the moods expressed by Competent/Compliant children (Group I) after Deviation 0011 Play stories seem to be comparable to ones found by Gershowitz (1974) accompanying replication fantasies, i.e., enjoyment or fatigue, denoting possible catharsis. It should be noted that Gershowitz found that these affects accompanying replication fantasies in children who played with trained undergraduates, communi- cating with children in a "sensitive, involved and accepting manner". The experimenter and her assistants were trying to provide the same type of atmosphere, in the present research. In marked contrast, Impulsive/Withdrawn children (Group IV) chose predominantly to identify with the aggressor role in fantasy, and teachers perceived them as "less mature“ in the classroom. 100 Clinicians, however, are familiar with the propensity of identification with the aggressor role in fantasy play as signifying constructive carthartic attempts to master aggressive impulses. The difference again, is in the lack of superego functioning, the lack of acquired impulse controls, which corresponds to the significant use of Continued Deviant Behavior responses to Deviation Doll Play which Impulsive] Hithdrawn children gave. The content of these responses were highly specific, predictable, and sterotyped. They mostly contained object- level references, i.e., "wind smashes all the furniture in the house after the lamp falls", and reflected uncontrollable impulsivity, i.e., "wind picks up the whole house and dumps it upside down". Gondor (1964) suggested that fantasies containing mainly object references reflected relatively immature ego coping mechanisms. In addition, accompanying affects were usually anger and/or sadness, which Gershowitz (1974) also found to follow object level fantasies. More importantly, Group IV children seemed particularly vulnerable to "fluctuating certainty", i.e., after turning over the doll house, many children attempted to overturn pieces of actual furniture in the testing rooms! Clearly these children were showing impulse control problems and might be labelled as "acting-out" aggressive children. These children could not use fantasy to achieve catharsis, and their fantasies involved cogni- tion of a "magical omnipotent“ wishfulfillment quality, which accompan- ies preoccupation with aggressive impulses vs. aggressive imagery (Gould, 1972). Thus, it is not surprising that they primarily chose to identify with the aggressor role. lOl However, identification with the aggressor role on the modified MST was also associated, based on the factor analysis, with withdrawal and avoidance responses, i.e., "Hiding" and "Crying". Neither response was significant for groups in the multivariate analyses but "Hiding" responses were most frequent in Group IV, or Impulsive/Withdrawn chil- dren. Children who used "Hiding" and "Crying" responses typically gave fantasies where the child doll hid, the deviation was then discovered, the child doll then stood immobilized while verbal and/or physical punishment was administered, and the fantasy finally ended with an accompanying affect of sadness. Other children went through a repeti- tive ritual of the child doll hiding and running away, being caught, running away again, etc. These children were qualitatively different from the majority of Group IV children, using withdrawal and avoidance behavior to cope with aggressive impulses. Unfortunately, the impreci- sion of the author‘s definition of the category "Self-Control" probably confused the accuracy of the teachers‘ ratings, since teachers were instructed to rate children "low" on this variable if they exhibited both withdrawn and impulsive, aggressive behavior. This error most probably confounded the differentiation of "acting-out aggressive" children, from children whom Murphy (1962) described as effectively using withdrawal behavior to cope with "overstimulation". This poten- tial error also might have contributed to the inconsistent results obtained with respect to Self-Involved/Disruptive children (Group II), who also would be considered as “less mature" children, based on the teacher‘s ratings. In addition, third order interactions between order 102 of task presentation, sex, and identification with the aggressor, also obscured attempts to clarify patterns of coping styles related to the management of aggressive impulses in both the groups rated “less mature". anlification of Results and Directions for Future Research Methodological Issues Teacher Ratings of Children's Social Behaviors The most obvious problem with the teacher ratings is that reli- ability could not be assessed, given that each child was rated by only one teacher. In fairness to the teachers, all appeared to know the children well, and were the "primary" or "head" teachers in their class- rooms. However, Baumrind's (l967) procedure of employing an observer- rater in the classroom for two weeks to verify reliability of teacher ratings is a better alternative for future research. In addition, Baumrind‘s ratings were based on Q-sort techniques, rather than on the four point scales (e.g., high, medium-high, medium-low, low), that were used in this study. Another serious problem with the teacher's ratings lies in the behavioral definitions of specific dimensions. For instance, on scale I, self-control, teachers were asked to rate a child "high" on this dimension, if he/she typically expressed negative feelings openly and directly, when faced with a frustrating situation. Many teachers rou- tinely do not regard these behaviors as indicative of children who manifest "high" self-control, although clinicians might be more in- clined to make that discrimination. Likewise, teachers were instructed 103 to rate a child "low" on self-control, if he/she typically either became passive or sullen when confronted with a frustrating situation or acted out aggressively. This does not discriminate impulsive children from withdrawn children, with respect to self-control, but this discrimi- nation is regarded as an important clinical distinction. Formation of Gropps Originally, the intent was to study three groups of children based on Baumrind's (l967) delineation of Pattern I, II, and III children. However, in examining her data, it became clear that one possible important pattern of child social behavior (i.e., self-controlled but lacking self-reliance) was not investigated. Thus, a decision was made to study four groups, i.e., Group I, consisting of children who were rated high on control, but medium on initiative dimensions, Group II consisting of children who were rated low on control but medium on initiative dimensions; Group III consisting of children who were rated high on initiative but medium on control dimensions, and Group IV con- sisting of children who were rated low on initiative and medium on control dimensions. In this manner, it was expected that the relative influence of self-control vs. self-reliance and approach-avoidance tendencies upon the child's social behaviors might be examined. However, statistically, there is little support for this discrimination among groups. Given the factor analysis of the teachers' ratings, described in the Methods section, it is apparent that all five behavioral scales, i.e., self-control, self-reliance, approach-avoidance, mood, and peer affiliation substantially loaded positively on one factor, 104 again arguing for a global "high" vs. "low" adjustment distinction between groups. It is also true that the factor analysis yielded a second factor, comprised of self—control, which loaded substantially positively and self-reliance, which loaded substantially negatively, on that factor. But both self-reliance and approach-avoidance (initiative dimensions) correlated positively, though nonsignificantly with self-control. Thus, there is only minimal statistical support for classifying the total sample into four groups. In addition, by utilizing these specific four groups, comparison with Baumrind's Pattern I, II, and III children is rather difficult. Initially, Baumrind (1967) reported that Pattern I children were rated as "high" on self-control, self-reliance, approach-avoidance, and mood. This does not correspond, to any of the four groups exactly, but most closely parallels children in Group III (who were rated “high/medium- high" on self-reliance, approach-avoidance, mood, and peer affiliation and “medium“ on self-control). Baumrind's Pattern II children were rated "low" on peer affiliation and mood dimensions, but did not rank "high" on approach-avoidance. This seems somewhat similar to the classifica- tion of children in this study, that were identified as Groups II and IV, since both these groups contained children who were rated "10W/ medium-low" on mood, and peer affiliation but either ”medium" or "low/ medium-low" on approach-avoidance. Baumrind's Pattern II children were rated "low“ on self-reliance, self-control, and approach-avoidance. This pattern does not correspond to any of the groups in this study. 105 In addition, the validity of the group distinctions and corre- spondance with Baumrind's delineation of Pattern I, II, and III children is further obscured by the attempt to hold one dimension constant (i.e., initiative vs. control) in forming the groups. This attempt involved a subjective "eye-balling“ of the teacher's ”medium-high" and "medium-low” ratings, to arrive at a designation of a "medium" rating for three behavioral scales (i.e., self-control, self-reliance, and approach-avoidance). This "method" was purely subjective on the author's part, and involved “examining" the individual checkmarks to attempt to "discriminate" children rated as merely "medium“, from those rated as "medium-high" tending towards "high", or "medium-low" tending towards “low". This intuitive approach to subject classification further leads one to seriously question the validity and meaningfulness of the four groups. Finally, while a goal of this study was to examine the differen- tial influence of control vs. initiative dimensions of social behaviors on fantasy behavior theoretically related to ego and superego functions, the groups were not adequately formed to explore this question. For example, both Groups I and III contained children rated “high" or “medium" on control and initiative dimensions, while Group II consisted of children rated “low" on control, and "medium" on initiative dimen- sions, and Group IV contained children rated "low" on initiative and "medium" on control dimensions. A fifth group rated "low" on both control and initiative dimensions, makes theoretical sense, and would have been comparable to Baumrind's (1967; 1970, and 1973) Pattern III 106 children. Obviously the best classification of groups would have in- volved four groups, which systematically varied each dimension (e.g., control), at both "high" and “low" levels, with both levels of the remaining dimension (e.g., initiative). Then comparisons with Baumrind's classification of Pattern I, II, and III children, as well as compari— sons between initiative and control dimensions of children's social behaviors with fantasy measures, might be examined more adequately. Thus, the four groups, utilized here appear not to be statistically valid, nor are they methodologically, or theoretically, precise. They do not permit direct comparisons with Baumrind's data, nor do they permit a direct comparison of the relative impact of initiative, and control dimensions of children's social behaviors upon fantasy behavior. Again the combined difficulties with teachers' ratings, and group formation make it difficult to draw any conclusions from the data regarding the "meaning" of group differences. The most reasonable assumption is that subjects might be differentiated as either "high" or “low" on a factor of “general adjustment". Dependent Measures Task A--Modified MST. The relatively poor internal consistency of the modified Miniature Situations Tests used in this study, and designed by the author to be used in this study, has previously been discussed in the Results section. Generally, the problems enumerated before, consisted of: (1) an inadequate consideration of children's toy preferences and their reactions (e.g., fears) towards specific animal toys (e.g., "snakes"), (2) assuming equivalence of items which 107 implied different levels of psycho-sexual imagery (i.e., "oral" vs. "anal“), and (3) assuming equivalence of items which required different levels of response intensity (i.e., "spanking vs. drowning" an animal toy). In addition, victim items were particularly difficult for the author to design and most contained "oral-sadistic" aggressive content with "frightening toys", e.g., "pretend a toy snake bites you". This may be one reason why these items had the lowest overall mean category usage (§‘= 2.30) of items on the modified MST. The order of presenta- tion of items, which was counterbalanced, may have had some effect on choice of victim responses. One victim item had the highest part-whole correlation when presented last, while the same item obtained the lowest part-whole correlation, when presented first. As Gould (1972) points out, children who typically identify with aggressor and/or victim roles in fantasy are more prone to use "direct" vs. "distance" defenses (e.g., "I shot you" vs. "I'm a cowboy and I shot you, an Indian“). Children who employ direct defenses against threatening impulses are more prone to experience “fluctuating certainty" (i.e., inability to distinguish reality from pretend), and frequently stop fantasy play behavior when this occurs. From the actual administration of the modified MST, it appeared that children subjectively experienced most “fear“ when con- fronted with victim items, and thus, this method may not be the best experimental approach to assess children's preferences for identifica- tion with victim roles in fantasy. In addition, the entire set of modified MST items consisted of fantasy material presented to the subject in a "direct" vs. "distance" fashion. For instance, a child was asked whether he/she wished to 108 "hug a Teddy Bear" vs. whether he/she wanted to play the "Grown-up Bear" and hug the "Baby Bear". Since Gould (1972) observed that most provider- protector children utilized distance defenses, the latter example might be more relevant in examining the children's preferred modes of self- representation in fantasy. Also, given “creatively elaborated" aggres- sive and victim items, one could more accurately guard against building in a response set, which promotes "fluctuating certainty" in young children, and may inhibit them from choosing victim items in particular. Also, if the items were designed to give "psychological distance“ to the child from his impulses, one could observe whether or not some chil- dren gave up distance defenses and employed “direct defenses" after choosing an aggressor or victim role, as Gould hypothesized. . Another major drawback to the use of the modified MST is that it is a forced choice technique. While only eight children changed their initial responses from provider-protector choices to aggressor choices, the question of what would have evolved over time is unclear. Often scoring choices, necessitated making judgments as to whether the child was really expressing nurturant or aggressive behaviors towards the animal toys. For example, one child started hugging a Teddy Bear, but the aggressive impulse did eventually predominate, and she ended up choking the Teddy Bear. Thus, another criticism of the forced choice method is that behaviors which are designed to reflect certain internal motive states, i.e., "spanking" as aggression, are not behaviors which the child would naturally choose to express his/her motives, needs, wishes, feelings, 109 and/or impulses. Thus, one cannot generalize from self-representations in fantasy on the modified MST to self-representations in spontaneous fantasy play behavior. A study on the differential frequency of identi- fication with provider-protector, aggressor, and victim behaviors in spontaneous unstructured fantasy play observed in a naturalistic setting, e.g., day-care center; such as, the one conducted by Gould (1972), would yield more information and would be more applicable for generalization purposes, as well as being methodologically more sound. Task B--Deviation 0011 Play. Obvious difficulties with this task were in assuming the stimulus stories had equal relevance for all sub- jects, and that the order of presentation of individual stories had no effect upon the child‘s responses. Conceivably, some children may identify more with, or be more frightened of, "knocking over a lamp" than "eating cookies before supper". However, the five stories utilized here, had previously been employed by Sears, Rau, and Alpert (1965) and Wurtz (1961). Haworth (1964) refers to them as representing typical situations which a number of doll play researchers have used in the past (i.e., Ammons and Ammons, 1949; Korner, 1949; Lynn, 1959; Moore and Ucko, 1961; Stamp, 1954; Stolz et al., 1954; Winstel, 1951). All subjects were administered the stories in the same order, and subjec- tively, at least, the author did not detect any consistent response trends which seemed to be based on the sequence of stories. Most children readily responded to the doll play situations, but there were predictable difficulties with children preferring to "act out" rather than verbalize what was happening. Following Amnons 110 and Ammons (1952) suggestions, specific structured questions; such as, “what are the dolls doing" vs. "saying" or "feeling" were asked to secure the maximum amount of information. However, limits on the termi- nation of stories were difficult to establish, without significantly influencing the production of fantasy material. Generally, a five- minute time limit was employed, per story completion. However, a child might start out with a typical scene of verbal punishment and subse— quent love-withdrawal (i.e., being yelled at and then sent to bed), but if allowed to continue beyond the limit, or spontaneously, he/she might change the entire nature of the story's outcome by stating, “the next day, the child killed his mother“, through some "magical“, "omnipotent" imagery. Thus, Deviation 0011 Play, like any projective technique, raises the issue of how to arrive at the appropriate point where fantasy production is stopped, without losing valuable clinical data, that might alter interpretations of the subject's behavior. Also, there were problems with the appropriateness of scoring categories for Deviation Doll Play. There were no significant differ- ences among children in perceptions of parental punishment following fantasized transgressions, with the exception of the categor -- Reasoning (§'- 1.29). All categories, including Reasoning, were infre- quently used (i.e., highest mean category usage was for Verbal Punish- ment, §'= 3.07, and Physical Punishment, §'= 2.79). Most children typically reacted to Deviation 0011 Play by portraying the parent figure as disciplining them along one of these examples: 1) Why did you do that, you bad boy!"; 2) "Shut up, and go to bed"; 3) "I'm going 111 to spank you real hard!“ Since these children were so young, it is questionable whether they possessed enough cognitive skills to perceive sophisticated differences among modes of parental punishment. Furthermore, even if they did perceive differences, at least with respect to verbal, physical, and reasoning punishment techniques, it is also possible, as Gutkin (1974) pointed out, that they cannot "de-center“ from the adult's initial response, which may be a verbal reprimand, or physical response, or deprivation of privileges. Thus, scoring Devia- tion Doll Play categories for perception of mode of punishment antici- pated may not be relevant for this age group. The categories for scoring children's behavioral reactions to fantasized transgressions were also used infrequently by most subjects (i.e., highest mean category usage was for Continued Deviant Aggression, :‘= 2.13). There were two other categories, i.e., Apology, and Spon- taneous Fixing, which did reflect differences among children, but they also had relatively low mean category usage (i.e., §'= 1.47, and §'= 1.96, respectively). Again, children rarely gave more than one response per story, with very few children perceiving that the "child gets away with it". The categories for scoring perception of agent(s) of fantasized punishment were used with relatively equal frequencies (refer to Appendix I, Table I), with the exception of category 24--“Mother and Father, together (disagreeing)". This latter category was used by only two subjects; hence, dropped from data analysis. Again, the content of stories was not elaborate with respect to verbal interactions between 112 child and parent figures, and perhaps, this last category subsumes a level of cognitive complexity more advanced than the average pre-school child possesses. However, the stories were complex when looked at from different clinical perspectives. Some involved very intricate plots, with intro- duction of numerous new elements into the story (i.e., new characters; such as, a "burglar:), often representing defenses of denial, projec- tion and displacement, as well as repetition compulsion (e.g., "an axe- murderer chopping up everybody slowly and eating parts of their bodies one by one"). Analytic therapists might interpret this story comple- tion as an example of “oral-sadistic incorporation". Qualitative differences along the lines of Erikson's (1940) theory of play disrup— tion versus play satiation were also subjectively noted by the author. More specifically, children's stories appeared qualitatively distinct with regard to the degree of overwhelming anxiety experienced, and/or the degree to which "fluctuating certainty" appeared to influence term- ination of the story (e.g., one subject took a toy gun and announced he was “shooting himself, since he was a bad boy“). This exemplifies Erikson's notion of play disruption. Other children appeared to be more capable of mastering their anxieties so that their play seemed to be more representative of adequate conflict resolution and coping mechanisms (e.g., one subject went through an elaborate presentation of "how to bake cookies to replace ones which the child doll had taken", and ended her story with a festive "dinner party" where she received a prize from the "Galloping Gourmet on T.V." This latter example corre- sponds more to Erikson's notion of "play situation". 113 Thus, scoring categories for Deviation 0011 Play which might have been more sensitive to differential defense modes, e.g., corresponding to analytic conceptualization of stages of development or which were more sensitive to the aims of the child's fantasy behaviors (i.e., “protective and restitutive" vs. "expansive and oriented towards mastery or achievement") (A. Freud, 1936), would be more appropriate to use in future research. Also, Gould's (1972) hypothesis concerning the use of "direct" vs. “distance" defenses might be more directly explored, as well as the frequency and conditions evoking "fluctuating certainty", in scoring Deviation Doll Play responses. For example, within the cate- gory, Continued Deviant Aggression, which did differentiate among groups, there were "direct defenses", i.e., "child shoots himself" vs. “distance defenses" i.e., "the child doll turned into a vampire bat that killed the family“. In this example, one also sees the apparent difference which the modified MST tried to measure--identification with the Victim vs. the Aggressor role in fantasy. If the Deviation 0011 Play stories had been scored for differences in self-representation some validity for the modified MST might have been demonstrated. Indeed, this is also a suggestion for future research. In addition, responses to the category, Continued Deviant Aggres- sion, incorporated aggressive behavior not only towards the self or child doll and other animate figures, but towards inanimate objects as well; such as, "child doll throws all of the furniture out of the doll house." Gondor (1964) previously suggested that fantasy behavior could be differentiated in terms of degree of reference--from inanimate to 114 animate objects. Thus, another useful way to score the stories might be in terms of whether aggressive impulses are directed at an object, the self, or others. Of course, on these specific stimulus stories the subject may have a response set, given one deviant act involves break- ing an inanimate object (lamp), and another involves an interpersonal situation (grabbing baby's toy). Also, the stories could be scored with respect to the degree of sterotypy present, i.e., "child picks up the lamp and puts it back on the table" vs. complexity or transcendence--"it is a war, and an imaginary spy who was signalling with the lamp, knocked it over". (These are four and five year old children!) In summary, it is doubtful that the Deviation Doll Play story completions lacked "cognitive complexity"; rather, it is more likely that the scoring categories were inappropriate to examine the rich clinical data available therein. A more adequate approach would have been to explore differential modes of defenses, the presence/absence of fluctuating certainty, the degree of sterotypy vs. transcendence, the degree of reference (animate vs. inanimate) of continued deviant be- havior, and the child's "wish to please" parental figures, and/or symbolic "sense of entitlement" (pre-cursors of superego development, hypothesized by Gould, 1972). Scoring the stories in this manner might also provide information pertinent to theoretical formulations of identification with provider-protector, aggressor, and victim roles, and would clarify information obtained by forced choice techniques; such as, the modified MST. 115 Task Administration As mentioned previously in the Methods section, the author administered both tasks while two observer-raters recorded the child's responses, along with the experimenter (E). A more reliable alterna- tive method would have been for the experimenter to be solely concerned with the administration of tasks, and all scoring left up to the trained observer-raters, or else ideally, done by trained raters watching video- tapes of E and the child interacting. The presence of a female experi- menter and two additional strangers observing and recording responses probably influenced children's responses. Several investigators (Pintler, 1945, Siegel, 1957, Siegel and Kohn, 1959) have found that "high levels" of experimenter interaction or "adult permissiveness" and "presence" during doll play sessions increased fantasy aggression. This may have had an effect upon children's choices of identification with the aggressor on the modified MST and/or on the display of Continued Deviant Aggression in Deviation Doll Play. Likewise, the design utilized two half-hour play sessions, and several studies commonly report an increase in aggressive responses from the first to subsequent play sessions (Amons and Amnons, 1953; Bach, 1945; Gewirtz, 1950; Hollenberg and Sperry, 1956; Phillips, 1945; Sears, 1951). Order of task presentation did significantly interact with sex and group classification in the case of choice of identifica- tion with aggressor and victim roles, and in the child's use of Denial following fantasized transgressions. Since the cell means were rela- tively small, i.e., five to t§g_children per group, and difficulties 116 with ascertaining the meaningfulness of group differences has already been mentioned, the patterns of differences for these variables are not, surprisingly, unclear and inconsistent. Thus, it is virtually impossible to determine what impact order of task presentation may have had on the three-way interaction effects for identification with the aggressor, victim, and children's denial behavior. Also, the finding that there is a decrease in the latency for the first aggressive response from the initial to subsequent play sessions (Pintler, 1945; Yarrow, 1946), especially for boys (Sears, 1951), prob- ably has meaning for the differences in choice of identification with the aggressor obtained within Order B, where the modified MST com- prised the second play session. Again, three-way interaction effects are extremely difficult to interpret, and given the inconsistency of results, comments concerning the relative impact of order of task presentation would be purely speculative. Sample Characteristics 59g, This study originally intended to focus upon children (ages four-five), since they were conceptually in the midst of working through “Oedipal identification conflicts", and internalizing parental prohibitions or standards in accordance with superego development. However, the sample of children actually tested ranged in age from three years ten months to five years two months, with a median age range of four years six months through four years eight months. This wide age span undoubtedly affected the level of overall variance in the experiment, and qualifies the significance of the results. 117 Gould (1972) found that children (ages 3-5) display cognitive growth trends in fantasy play similar to those evident in reality-directed thought. Both generally show a progressive increase in complexity and coherence of ideas and language, and in uses of reality information. Specifically, Gould observed that three year-old children would swing more fluidly between fantasy and reality (i.e., experience fluctuating certainty), more than four and five-year old children did. Further, three year-olds, while interested in assimilating new information, seemed to be more preoccupied with accommodating new skills, or knowl- edge into familiar schemas. Thus, their fantasy play may be more replicative than wishfulfillment. Three year-old children were also less divergent, manifested less differentiated self-representations, and more sterotypy in fantasy play. While their fantasies may be more "egocentric". Gould views this as not just an indicator of the child's urgency for immediate wishfulfillment, but as reflecting a necessary cognitive step in organizing inner and outer experiences in a dichotomous fashion. Four year-old children, in Gould's study, displayed the most aggressive and destructive fantasy play behavior of the three age groups. Although fantasy themes were unified around plots of death and violence, she observed rapid shifts in manifest content, action, and characters, and noted progressive changes in the level of psycho-sexual imagery similar to psychoanalytic theoretical formulations of stages of development. Five year-old children appeared more expansive with regard to differentiation of themes, content, and characters, and 118 fantasy tended to move away from family themes or omnipotent wishful- fillment themes, i.e., "Supermen stories", to encompass a widening social world. Conflict of wishes and needs with significant others appeared to reflect cognitive and social advances, i.e., ”reciprocity" vs. “egocentricism”. Thus, developmental differences in children's cognitive and social skills probably confounded the results of this study. Since the majority of children (approximately 80%) were four year-old children, aggression in Deviation Doll Play and on the modified MST, might have been influenced by this age group's preoccupation with themes of death and destruction. Deviation 0011 Play stories probably reflect many developmental shifts in levels of defense mechanisms, and degrees of sterotypy vs. transcendence, and/or divergence of thought. This may have affected children's anticipation of parental punishment, and their behavioral responses to fantasized transgressions. Subjects who ranged in age from three years ten months to three years eleven months were probably more prone to experiencing fluctuating certainty on the modified MST, and less likely to use "distance" vs. “direct“ defenses in Deviation 0011 Play, when compared with older chil- dren in the study. Thus, in summary, interpretation of significant findings must be qualified by the wide variability in age of subjects, and future studies should attempt to control for developmental differ— ences attributable to age, I.Q. , and verbal skills. Bagkggound Variables. Other information obtained on subjects, was number and age of siblings relative to the subject (i.e., younger 119 or older), parents' marital status, and level of occupation. The sample was fairly homogeneous with respect to birth order (approximately 70% of subjects were first-born, and approximately 60% of the subjects were an "only“ child). Singer (1972) reported that "only" or "first-born" children (especially first-born children with younger siblings), tended to possess more of a predisposition to imaginative play than children with several siblings, particularly older siblings. He concluded that possibly children who had fewer siblings had more opportunities for interaction and identification with adults, and more privacy available for practice of fantasy. This may have contributed to the findings of the current study, given the ease with which the subjects readily en- gaged in the fantasy tasks, but the sample homogeneity limits the generalizability of the results. The sample was also fairly homogeneous with respect to parents marital status, approximately 70% of the subjects came from "intact" families, while 30% of the sample came from "father-absent" homes (due to separation, divorce, or death). However, the fact that almost one- third of the sample came from “one-parent" families, may have had some impact on the findings. For instance, Bach (1945) found that children from father-absent homes showed less anticipation of punishment for fantasized transgressions in doll play, while children from intact families portrayed the father (alone) as aggressing more against the child doll. This finding is similar to the sex effect for Father Alone, as primary agent of fantasized punishment, that was found in the present study. 120 Several researchers, notably Singer et al. (1956) and Singer (1973), suggest that cultural background, socio-economic level, intelligence and verbal facility are related to fantasy development. Specifically, Singer (1966) suggests that "middle-class" children whose parents are educated, verbal, more tolerant of imaginative play, are more likely than lower-class children to learn to fantasize and play imaginatively. Korchin, Mitchell and Meltzoff (1950) reported that ”middle-class" subjects produced significantly longer fantasy stories than "lower-socio-economic class" subjects. Pines (1969) found that "middle-class“ pre-school children engage in role-playing behavior five times more frequently than lower class children of the same age. The study by Pines also revealed that children of overwhelmed and overworked mothers (usually "lower-class" mothers), generally did not spend much time in make-believe and pretending activities with their children. This finding is further supported by Smilansky (1973) and Freyberg (1973) who found that "lower-class" children engaged in little verbal labeling or communication about their play activities and manifested a high degree of "aggression“, "bossiness", "hyperactivity", and “overexcitement”, as well as "extreme passivity". In the current study the only information obtained, relevant to subjects socio-economic class, is parents' level of occupation. The majority of mothers (approximately 70%), were working mothers who placed their children in day care centers out of necessity. Only approximately 15% of the mothers could be classified as full time "housewives". While mothers generally were employed in Level III occupations, i.e., 121 "administrative, technical, or clerical jobs”, fathers were generally employed in Level IV occupations, i.e., "semi-skilled or skilled manual labor" (Hollingshead and Redlich, 1958). Parents employed in profes- sional occupations comprised only approximately 20% of the sample, and less than 10% were full-time college students. Thus, the majority of children came from intact "lower-middle" or "working class" urban families, with both parents employed full-time. All children were Caucasian. Given the lack of sample heterogeneity, generalization of findings is limited. Also, information regarding the assessment of childrens' intellectual level of functioning and verbal facility as well as additional socio-economic indices (e.g., parents' level of education) were not obtained, further qualifying the results, particu- larly with respect to fantasy measures. Conclusions and Implications of Findings The implications of this study are twofold: 1) Dimensions of the child's inner experiencing or fantasy life are an important area of continuing research, with regard to the development of personal and social competence. Although a child's social behaviors are objective events in the real world, they effect and reflect his/her ego and superego capacities only to the extent and form in which the child perceives them. It is equally relevant to ascertain which are the distinguishing features of the self and its perceived world, as it is to establish the nature of the actual environ- ment to which the child is exposed. Through the study of children's 122 coping styles in fantasy, we may derive more information relevant to how constant tendencies within the child effect person perception, and behavior in interpersonal interaction. This in no way contradicts the view of behaviorists or social learning theorists that one's behavior is governed by situational antecedents and consequences. What typical studies of parent-child interaction leave out, however, is that in most situations there are "options". One child may "opt" to behave in accordance with one set of antecedents and consequences, while another child chooses a “different set". Therefore, research which only attempts to assess parent-child behavior neglects the needs, wishes, feelings, thoughts, or "childish" perceptions, which predispose the child to exercise his/her “behavioral options". As Maccoby (1959) suggests, a child probably learns to produce adult-role responses more dependably to internal cues rather than to external ones. Most adult-role learning takes place through "covert rehearsal" in fantasy play. Actions which are part of adult-role behavior (i.e., disciplining a child), often remain latent until a situation arises in which they are appropriate. Thus, a child may learn to reproduce the responses of adults while failing to learn the cues which guided the response, resulting in inappropriate "adult-like" behavior directed towards peers and others. Further, covert role- playing may not be only a means for learning "adult" or "socially competent" behavior, but of learning others' reactions toward the self. 2) Given the child's inner representational world is important in influencing behavioral outcomes, continued research is needed to explore ways in which adults can be trained to be sensitive, empathic, involved 123 in, and accepting of, children's fantasy play behavior. For example, Reif and Stollak (1972) found that “children who played with undergradu- ates trained in 'sensitive and empathic' communication skills, exhibited "higher levels" of fantasy which contained references to intrapersonal and interpersonal situations, when compared with children who played with untrained college students“ (pp. 96-97). Similarly, Liberman, Stollak and Denner (1971) found that parental qualities of empathy, genuineness, and non-possessive warmth correlated significantly with children's behaviors assumed indicative of psychosocial competence. Specifically, Liberman et al. found that “empathic behavior on the parent's part was significantly correlated with the child's statements about his/her own behavior in fantasy play, and reflection of feelings was also associated with the child's statements of interpersonal aware- ness in fantasy play, open expression of aggressive themes, and the child's dominance. In general, the parent's overall participation in the child's fantasy play was also significantly correlated with the child's interpersonal awareness in fantasy, his/her statements about him/herself in a real context, and his/her dominance in the parent- child relationship“ (p. 7). Thus, comprehensive studies are needed which evaluate different types of effective child social behaviors, i.e., "social responsibility", "altruism", "empathy”, and ineffective social behaviors, i.e., "shyness", “impulsivity”, "dominance and aggression towards peers", and identify "patterns of inner experiencing or coping styles“. Further, the “inner life“ or patterns of.coping with conflict, and issues of intimacy and sharing, of parents and marital partners should be 124 explored and integrated with typical approaches to studying family and marital interaction. In this manner, one can study and possibly identi- fy how coping styles in children may be related to parental personality traits, and how the child selectively chooses what personality char- acteristics and/or behaviors he/she wishes to emulate or incorporate into his/her own self-system. BIBLIOGRAPHY BIBLIOGRAPHY Ames, L. B. Children's stories. Genetic Psychology Mono raphs, 1966, L}. 337-396. A Amons, R. a Amnons, H. Parent preferences in young children's doll play interviews. Journal of Abnormal and Social Psycholggy, 19499 _4_4_9 490-505.—-r Anderson, H. Domination and integration in the social behavior of young children in an experimental play situation. Genetic Psycho- lggical Monogpaphs, 1937, l_9_, 341-408. Aronfreed, J. The nature, variety, and social patterning of moral responses to transgression. Journal of Abnormal and Social Psychology, 1961, 63, 223-240. Bach, G. Young children's lay fantasies. Psychological Monogpaphs, 1945, _5_9_ (Whole No. 2), 3-69. Bach, G. Father-fantasies and father-typing in father-separated children. Child Development, 1946, l_7_, 63-80. Bandura, A., Walters, R. H. Social Learnin and Personality Develop- ment, New York: Holt, Rihehart 8’ nston, 1963. Baumrind, D. Childcare practices anteceding three patterns of preschool behavior. Genetic Psychological Monographs, 1967, _6_7, 43-88. Baumrind, 0., Black, A. Socialization practices associated with dimen- sions of competence in pre-school boys and girls. Child Develop- ment, 1970. _3g, 291-323. Baum-1nd, 0. Current patterns of parental authority. Developmental Psychology Monographs, 1971, 4, No. l. Baum"; nd, 0. The Development of Instrumental Competence Through Socialization. In Minnesota Symposia on Child Psycholggy, Minn.: University of Minnesota Press, 1973.1, 3-46. Baum“ nd, 0. Early Socialization and the Discipline Controversy. Morristown, New Jersey: General Learning Press, 1975. 125 126 Becker, W. Consequences of different kinds of parental discipline. In M. L. Hoffman and L. W. Hoffman (Eds.), Review of Child Develppment Research, New York: Russell Sage, 1967. Biblow, E. Imaginative play and the control of aggressive behavior. In J. Singer, The Child's World of Make-Believe, New York: Academic Press, 1973, 104-128. Conscience development in nursery school children as re- Burke, M. Unpublished M. A. Thesis, Stanford vealed through fantasy. University Press, 1961. Antecedents of resistance to Burton, R, Maccoby, E., Allinsmith, W. Child Development, 1961, temptation in four year old children. §§, 689-710. Boys' and girls' reports of maternal and Droppelman, L., Schaefer, E. paternal behavior. Journal of Abnormal and Social Psychology, 1963. 91,. 648-654. vaikson, E. Toys and reasons. In Childhood and Society, New York: Norton, 1950, 182-218. Elder, G. Parental power legitimatization and its effect on the adolescent. Sociometry. 1963, gs, 50-55. Enlnerich, W. Young children's discriminations of parent and child roles, Child Development, 1959, 30, 403-419. Enmerich, W. Family role concepts of children ages six to ten. Child Development, 1961, 32, 609-624. Fenichel, O. Psychoanalytic Theory of Neurosis, New York: Norton, 945. Freud . A. The E o and the Mechanisms of Defense, London: Hogarth Press, 193% FrWd. A. Normalit and Patholo in Childhood, New York: Inter- national Universities Fress, 1965. Freud, s, Dissolution of Oedipus Complex. Standard Edition, Vol. _l_9_, 'l 924. Freud, 3, Beyond the Pleasure Principle. 115320. Freud. S. New Introductory Lectures on Psychoanalysis, New York: Fitzrton. 1555- Standard Edition, Vol. 1g, 127 Freyberg, J. Increasing the imaginative play of urban disadvantaged kindergarten children through systematic training. In J. Singer, The Child's World of Make-Believe, New York: Academic Press, lg, 39 129-134. Gardner, L. An analysis of children's attitudes towards fathers. Journal of Genetic Psychology, 1947, 19, 3-28. Gershowitz, M. Fantasy behaviors of clinic-referred children in play encounters with college undergraduates. Unpublished doctoral dissertation, M.S.U., 1974. Gondor, L. Use of fantasy communications in psychotherapy. In M. R. Haworth (Ed.), Child Psychotherapy. New York: Basic Books, 1964. Goldin, P. A review of children's reports of parent behavior. Psychological Bulletin, 1969, 21, 222-236. Gould, R. Child Studies Thrgugh Fantasy, New York: Quadrangle Books, 1972. Gutkin, D., Falvey, M. Does differential modeling and reinforcement by mothers account for the development of moral intentionality on their children? Unpublished manuscript, Michigan State Univer- sity, 1971. HarTVis, D., Tseng, S. Children's attitudes toward peers and parents as revealed by sentence completions. Child Development, 1957, gs, 401-411. Hartmann, H. Essays on Ego Psycholggy. New York: International Uni - versities ress, 4. Hartmann, 11., Kris, E., Lowenstein, R. M. Conments on the formation of psychic structures. The Ps choanal tic Stud of the Child. Now'York: International Universities Press, 1946, 2, 11-38. Hartup. W. W., Keller, E. 0., Nurturance in preschool children and its I~elation to dependency. Child Development, 1960, s1, 681-689. Hatfield, J. S., Ferguson, L. R., Alpert, R. Mother-child interaction ilnd the socialization process. Child Develppment, 1964, §§_ £365-414. "Wkes s Burchinal, Gardner. Measurement of pre-adolescents views of family control of behavior. Child Develppment, 1957, 28, 387-392. Hawoyut:h M. R. (Ed.). Child Psychotherapy. New York: Basic Books. 964. 128 Heinicke, 0. Some antecedents and correlations of guilt and fear in young boys. Unpublished doctoral dissertation, Harvard Univer- sity. 1953. Herron, R. H., Sutton-Smith, B. Child's Play, New York: Wiley, 1971. Hoffman, M. L. Power-assertion by the parent and its impact on the child. Child Development, 1960, §_l_, 129-143. Hoffman, M. 1.. Child-rearing practices and moral development generali- zation from empirical research. Child Development, 1963, 3_4, 295- 318. Hoffman, M. L., Saltzstein, H. 0. Parental discipline and the child's moral development. Mrnal of Personality and Social Psychology, 1967. g, 45-57. Hoffman, M. L. Parent discipline and the child's consideration for others. Child Development, 1963, 3_4, 573-588. Hoffman, M. L. Moral development. In P. Mussen (Ed.). Carmichael's Manual of Child PsycholOJL New York: Wiley, 1970. Hoffman, M. L. Conscience, personality, and socialization techniques, Human Develppment, 1970, 13, 90-126. Hoffman, M. L. Developmental synthesis of affect and cognition and its implications for altruistic motivation. Developmental Psychology, 1975. 11, 607-622. Hoffman, M. L. Moral internalization, parental power, and the nature of parent-child interaction. Developmental Psycholgy, 1975, 1;, 228-239 . Hoffman, M. L., Saltzstein, H. 0. Parent practices and the child's moral orientation. Paper read at the Amer. Psychol. Assn., Chicago, Sept., 1960. Hollenberg, E., Sperry, M. Some antecedents of aggression and effects of frustration in doll play. Journal of Personality, 1951, _1, 32-43. Hollingshead, A. B., Redlich, F. C. Social Class and Mental Illness. New York: Wiley, 1958. IWacs, S. Social Develogment in Younnghildren. New York: Harcourt, Brace, an o , JacobSen, E. The Self and the Ob'ect World, Part II. New York: International Universities ress, 1964. 129 Jersild, A. T., Holmes, R. 8. Children's fears. Mongggaphs of Social Research in Child Development, 1935, No. 20. Jersild, A., Markey, F., Jersild, C. Children's fears, dreams, wishes, daydreams, likes, dislikes, pleasant, and unpleasant memories. Child Development Monogpaphs, 1933, No. 12. Attitudes of children toward authority as projected in Johnson, E. Unpublished doctoral dissertation, their dolls at two age levels. Harvard University, 1951. Kagan, J. The child's perception of the parent. and Social Psychology, 1956, 53, 257-258. Kagan, J., Hoskin, 8., Watson, S. Child's symbolic conceptualization of parents. Child Development, 1961, 32, 625-636. Journal of Abnormal Kallman, J., Stollak, G. E. Maternal behavior toward children in need arousing situations. Paper presented at the Midwestern Psycho- logical Assn., Chicago, 1974. Klinger. E. Strgcture and Functions of Fantasy, New York: Wiley, 1971. lOJlberg. L. Development of moral character and moral ideology. In M. L. Hoffman and L. W. Hoffman (Eds.), Review of Child Develop- ment Research Vol. 1 . New York: Russell Sage Foundation, Liberman, M., Stollak, G., Denner, 8. Assessment of family play inter- action. Paper presented at 1971 meeting of Midwestern Psychology Association. Detroit, Michigan. Levin. H. Permissive child-rearing and adult role behavior. In D. E. Delaney, R. L. DeValois, D. C. Beardsley, M. R. Winterbottom (Eds.), Contributions to Modern Ps cholo . New York: Oxfbrd University Press, 1956, 307-312. Le"1". H., Sears, R. R. Identification with parents as a determinant of doll play aggression. Child Development, 1956, _2_7_, 135-153. Le"?! . H., Wardwill, E. The research uses of doll play. Psychologjcal Bulletin. 1962, _5_9_ (1), 27-56. Lev-Y - D. Use of play techniques as experimental procedure. slournal of Orthopsychiatry, 1933, 3, 266-277. American Lowenfeld, M. Play in Childhood. London: Gollancz. 1935. Lynn: R. Development of validation of a structured doll play test for (:hildren. Quarterly Bulletin of Indiana University Medical Center, 'I 955. 130 Maccoby, E. E. Role-taking in childhood and its consequences for social learning. Child Development, 1959, so, 239-252. Melville, 0., Lynn, R. Structured doll play test as a projective technique for use with children. Journal of Projective Tech- nigues, 1959, 23, 335-344. Miller, 0. Motivation and affect. In P. Mussen (Ed.), Handbook of gesearch Methods in Child Development. New York: Wiley, 1960, 68-768. , MurphyI, L. B. The Widenigg World of Childhood, New York: Basic Books, 962. Mussen, R., Distler, L. Child-rearing antecedents of masculine identi- fication in kindergarten boys. Journal of Abnormal and Social Psychology, 1960, _52, 350-356. Piaget, J. The Construction of Reality in Childhood. New York: Basic Books, 1937. Piaget, J. Pla Dreams, and Imitation in Childhood. New York: Norton, 6T. Piageti .6). The Moral Development of the Child. New York: Free Press, 9 5. Pikas, A. Children's attitudes toward rational vs. inhibiting parental authority. _Jggrnal of Abnormal and Social Psychology, 1961, §_2_ 315-321 . Pines, M. Why some three-year-olds get A's--and some get C's. New York Times Magazine, July 6, 1969. Pint] er, M. 0011 play as a function of experimenter-child interaction and initial organization of materials. Child Development, 1945, 1;. 145-166. Pitcher, E., Prelinger, E. Children Tell Stories: An Analysis of Fantasy. New York: Internafionil Universities Press, 1963'.— Rabban, M. Sex role identification in young children in two diverse ssocial groups. Genetic Psychological Monographs, 1950, 42, 81-158. Bebe] Sky, F., Allinsmith, W., Grinder, R. E. Resistance to temptation and sex differences in children's use of fantasy confession. _thld Development, 1963, _3_4_, 955-962. Reif: T., Stollak, G. E. Sensitivity to Young ChilgLen: Training and l_ts Effects. 'East Lansing: MTSO. Press, 1972. 131 Reynolds, M. Ne ativism of Preschool Children. New York: Teachers College, 0 umbia UniVerSity, 1929, No. 288. Rosen, B., D'Andrade, R. The psychological origins of achievement motivation. Sociometry, 1959, 22, 185-218. An experimental study of memory in relation Rosenzweig, S., Mason, G. British Journal of Psychiatry, to the theory of repression. 1934, 25, 247-265. Rundquist, E. Personality tests and predictions. In D. Fryer and E. Henry (Eds.) Handbook of A lied Ps cholo . New York: Holt, Rinehart a Winston, 1959, 192-192. Sandler, L. W. Issues in early mother-child interaction. Journal of American Academy of Child Psychiatry, 1962, 1, 141-1 . Sandler, J. On the concept of the superego. Esychoanslytic Stugy of the Child. V01. 15. 1960. Sandler, J., Nagera, H., Holder, A., Meers, D. The ego ideal and the ideal self. Psyghoanalytic Study of the Child, Vol. 18, 1963. Santostefano, S. An exploration of performance measures of personality. Journal of Clinical Psychology, 1960, 1s, 373-377. Santostefano, S. Miniature situations as a way of interviewing children. Merrill-Palmer Quarterly, Behavior Development, 1962(a), 8, 261- 269. Santostefano. S. Performance testing of personality. Merrill-Palmer Quarterly,,Behavior Development, 1962(b), 8, 89-97. sa"tostefano, 5. Construct validity of the miniature situations test: I, The performance of public school, orphaned and brain damaged children. Journal of Clinical Psychology, 1965, 21, 418-421. Si""11<>-*‘>i.‘.efano, S. A developmental assessment of motives in children 'From the viewpoints of organismic theory and psychoanalytic ego lasychology. Paper read at the Annual Meeting of Amer. Psychol. lesoc., 1966. santostefano, S. Situational testing in personality assessment. In I). Sills (Ed.) International Encyclopedia of the Social Sciences, ‘I968. Schaefer, E. A configurational analysis of children's reports of [Jarent behavior. Journal of ConsultingAPsychology, 1965, 22, 5552-557. Sear Sea Sea 132 Sears, R. A theoretical framework for personality and social behavior. American Psychologist, 1951, 3, 476-483. Sears, R. R., Rau, L., Alpert, R. Identification and Childrearin . Stanford, California: Stanfora University Press, 1965. Sears, R. R., Maccoby, E. E., Levin, H. Patterns of Childrearipg, Evanston, Illinois: Row, Peterson and"C61, 1957. Sears, R. R., Whiting, J. W. M., Nowlis, V., Sears, P. Some child- rearing antecedents of aggression and dependency in young chil- dren. Genetic Psychology Monographs, 1953, 32, 135-234. Seigel, A., Kohn, L. Permissiveness, permission, and aggression: the effect of adult presence or absence on aggression in children's play. Child Development, 1959, 33, 131-141. Evaluation of Bronfenrenner's questionnaire for children Seigelman, M. Child Development, 1965, 33, concerning parental behavior. 164-174. .Singer, J. L. The Child's World of Make-Believe. New York: Academic Press, 1979?‘ Shrilansky, S. The Effects of Sociodramatic Pla on Disadvanta ed Preschool Children. New York: Wiley, 1958. Stollak, G. E., Scholom, A., Kallman, J., Saturansky, C. Insensitivity responses of undergraduates to children in problem to children: situations. Journal of Abnormal Child Psychology, 1973, 1, 169-180. Ungep, S. M. On the development of the guilt response system. Unpublished doctoral dissertation, Cornell University, 1960. ”096?. S. M. Antecedents of personality in guilt responsitivity. Psychological Reports, 1962, 13, 357-358. White . R. Motivation reconsidered: The concept of competence. Psychological Review, 1959, 33, 297-333 Wurtz s K. R. The expression of guilt in fantasy and reality. Journal 53f Genetic Psychology, 1959, 95, 227-238. Yarrow, L. J. The effect of antecedent frustration on projective play. Psychological Monoggaphs, 1948, 32, No. 293. Yarrow, L. J., Campbell, J. 0., Burton, R. V. Child-Rearin : An QInquiry into Research and Methods. San Francisco: ossey-Bass, Inc., 1968. APPENDICES APPENDIX A LETTER TO PARENTS 133 134 APPENDIX A LETTER TO PARENTS July 1, 1974 Dear Parent: I am doing a research study on elements of imaginative play in young children. Included in this study, is an examination of different behaviors children ages (4-5) typically choose to express their thoughts, wishes, and/or feelings in structured imaginative play. This study is also concerned with what perceptions children have about typical daily routines, e.g., "bedtime“, and their perceptions of typical encounters with adults, as they are learning appropriate social behaviors. I am asking that you permit me to engage in two 15-30 minute individual play sessions with your child in the next few weeks at his/ her nursery school or day care center. During these sessions, your child will be involved in structured imaginative play with animal toys, and also with doll play materials. I and two of my colleagues will observe the child's imaginative play. I am also requesting your permission to obtain teachers' percep- tions of how children typically play and interact in the school/center. If you object to your child participating in this study, please send a note to s er teacher. I have obtained permission from the co-ordinator and/or head teacher of this program, to conduct my study, which is serving as my dissertation. Children in several area day care centers/nursery schools are participating in this study. The co-opera- tion of parents has been more than generous, and in anticipating your co-operation, I thank you in advance. A summary of the results will be furnished to each day care center and nursery school, when the data has been analyzed. If you have any questions, please feel free to call me at the numbers listed below.. Thank you again. Sincerely yours, MW LOW Nancy K. Wowkanech, M. A. Department of Psychology ' Office phone 355-9564 Michigan State University Home phone 322-2773 East Lansing, Michigan APPENDIX B LETTER TO TEACHERS 135 136 APPENDIX B LETTER TO TEACHERS July 1, 1974 Dear Teacher: I am doing a research study serving as my dissertation, on ele- ments of imaginative play in young children. Included in this study is an examination of different behaviors children (ages 4-5) typically choose to express their thoughts, wishes and/or feelings, in structured imaginative play. This study is also concerned with what perceptions children have about typical daily routines, i.e., "bedtime", and of typical encounters with adults, as they learn appropriate social be- haviors. I am interested, as are all teachers, in pre-school education, and hope that my results will add knowledge to program design at this eve . The program co-ordinator and/or head teacher at your center/school has given permission for me to engage in two, Brief, 15-30 minute individual play sessions with each child partici- pating in the study. Parents have also been informed, via a similar letter. and will notify you by note if they do ggs_wish their child to participate. I am also interested in teachers' perceptions of the child's typical behavior in the classroom, as this will add further information relative to differences in imaginative play styles. Thus, I ask your co-operation in filling out the attached rating scale. You are re- quested to rate each child as being either Low, Medium-Low, Medium-Hig3_ or Hi h, on each of five scales, i.e., Self-Control, Self-Reliance, Approac -Avoidance, Mood, and Peer Affiliation. “A scoring sheet and descFiption of typical child behaviors in each category is provided for your convenience. Each child will be assigned a number, provided on the top of the scoring sheet, and thus, no names will be used, ensuring confidentiality. Please consider these ratings as reflecting ”consist- ent, day-to-day" behaviors of children. Most children display all of these behaviors some of the time, thus ratings are not designed—With the purpose of evaluating the child as "normal" or "abnormal", but as indicators of his/her particular style of classroom behavior. Teachers from several area day care centers and nursery schools are co-operating in this study. and in anticipation of your co-operation, I thank you in advance. I look forward to meeting with you, and discussing the study and the ratings. Results will be furnished to each center/school when analysis of data is completed. In the meantime, if you have any questions, please feel free to call me at the numbers listed below. Thank you again. Sincerely yours, ‘jfiflbtlrudéanmooJA_—-s’ Nancy . Wowkanech, M.A. Office phone 355-9564 Department of Psychology Home phone 332-2772 Michigan State University APPENDIX C BEHAVIORAL DEFINITIONS OF TEACHERS' RATING SCALES AND TEACHER RATING FORM 137 138 APPENDIX C Behavioral Definitions of Teachers' Rating Scales Instructions: The following is a list of five scales: Self-Control, Self-Reliance, Approach-Avoidance Tendency, Subjective Mood, and Peer Affiliation. Your task is to rate each child on each dimension, as being either High, Medium-High, Medium-Low, or Low, based on which cate- gory describes the behaviors he/she exhibits most frequently in your . classroom, on a daily basis. The High and Low endpoints of the rating scales are behaviorally described below. However, each dimension is to be regarded as a continuum consisting of the two endpoints, with mid- points of Medium-High and Medium-Low. Please bear in mind that 311_ children display mgs§_of these behaviors, sgmg_of the time. The ratings should portray what you observe as the child's "typical behavior" in the classroom. Thank you. Scale I--Self-Control 3193, This child typically expresses negative feelings openly and directly when faced with a frustrating situation; obeys school rules that conflict with his wishes, needs, and actions, even though he/she may not like them; possesses ability to sustain a work effort; reason- able attention span; capacity to wait his/her turn in play with other <:hildren or in use of washroom facilities or at snack-time or lunch; ability to restrain those expressions of excitement or anger that would be disruptive or destructive to his/her peer group; absence of 139 explosive emotional expression or swings between rigid vs. lax control over impulses. _L_o_w_ This child typically becomes passive, or sullen when confronted with a frustrating situation, and his/her performance deteriorates; breaks school rules which conflict with his/her needs, wishes, and actions; limited concentration and attention span; disrupts other chil- dren who are working or playing; cannot wait his/her turn, will act out aggressively or withdrawn; may "explode" given the slightest bit of frustration; his/her degree of self-control varies greatly between withdrawal and acting-out of impulses. Scale II--Self-Reliance fllgg Easily separates from parents when arriving at school or day care center; seeks help from teacher when faced with difficult task, but 333_ and pggjg§s_to work or play independently; pleasure expressed in learn- ing how to master new tasks; leadership interest and ability; interest expressed in making decisions and choices which affect him; displays comitmentxto his/her own activities, and will not stop out of need fer companionship when engrossed in something; may question adult's authority when he/she feels adult is being unreasonable. L9! Difficulty in separating from parents when arriving at school or day care center, may cry, whine, or get angry; dependent on teacher 140 and/or on other children in play and work activities, afraid to be alone, avoids learning situations, gives up easily when encountering frustration; is overly concerned with "winning and losing" in playing games, and would rather compete against other children vs. enjoy his/her own performance; "follower vs. leader" in interactions with peers; usually submissive to adult authority figures. Scale III--Approach-Avoidance Tendenoy (£531 This child is usually wide awake, fairly active, curious, alert, easily excited to active participation in a group; is emotionally expressive, particularly of positive emotions (e.g., joy, delight, happiness), forcefu11y goes after what he/she wants (without hurting others), resists domination by other children; will use reason and Persuasion to get what he/she wants from teachers and peers; manifests a sense of self-assuredness, i.e., "I can do that"; likes to explore novel objects and situations. 3333 This child appears apathetic, slow-moving; is emotionally bland, 91Ving little clue as to the extent of pleasure he/she derives from his/her activities; only makes feeble attempts to get what he/she wants and is easily put off, submitting to the demands of other children; hesitant about expressing his/her needs, wishes, etc., and is evasive When asked what he/she wants; shy around other children, hesitates to 3010 a group; will withdraw from novel objects and situations; manifests 141 an "I can't do that" attitude; maintains close physical contact with teacher. Scale IV--Subjective Mood (Buoyant vs. Dysphoric) High This child usually manifests feelings of pleasure, spontaneity, zest, happiness, etc.; “good-natured"; excited and involved in school activities; contented; secure. L21 This child appears anxious, hostile, punitive, fearful, bored, or subdued; low involvement in school activities; irritable towards peers, and teachers. Scale V--Peer Affiliation High This child expresses affection and trust towards peers; cooperates with peers in group activities; initiates play with other children; expresses sympathy towards other children, e.g., will help another child who is hurt, or might offer one of his/her toys to a child who is dis- tressed; discourages wrongdoing of other children, and dissociates him- self from others' wrongdoing; hits other children only in self-defense; his/her behavior shows consideration of possible harmful consequences of his actions on others. 142 33! This child mistrusts other children, expects to be treated by them in a hostile manner; is mainly aggressive towards other children; without concern for possible harmful consequences of his/her behavior; prefers to play alone, usually teases or ridicules another child who is hurt or crying; incites or supports wrongdoing by other children; “tattles” on other children; frequently hits smaller/weaker children; is bossy and tries to dominate other children; frequently knocks over or destroys work/play activities of other children; aggression towards other children is often unprovoked. 143 Teacher Rating:Form Name of Day Care Center/Nursery School Teacher's Name Child's Number Child's Sex Child's Age Instructions: Please place a check (7) in the column which best describes the child's typical behavior in the classroom. SCALES LOH MED.-LOW MED.-HIGH HIGH I. Self-Control II. Self-Reliance III. Approach-Avoidance IV. Peer Affiliation APPENDIX D FACTOR ANALYSIS OF INDEPENDENT MEASURES 144 145 APPENDIX 0 Table I. Summary of Factor Loadings on Independent Measures Rotated Factor Loadings Factor I Factor II Comm. 1) Self-Control 0.6642* 0.6185* 0.8283 2) Self-Reliance 0.7442* -.4789 0.7832 3) Approach-Avoidance 0.7152* 0.5383* 0.8012 4) Mood 0.8581* 0.1076 0.7479 5) Peer Affiliation 0.8077* 0.2949 0.7394 HI LOAD. 0.8581 0.6185 PROP. VAR. 0.5790 0.2000 CUM. P. V. 0.5790 0.7791 Table II. Correlation Matrix of Independent Measures Self- Self- Approach Peer control reliance avoidance Mood Affiliation Self-control 1.0000 Self-reliance 0.2031 1.0000 Approach-Avoidance 0.2465 0.6239* 1.0000 Mood 0.5449* 0.5011* 0.5303* 1.0000 Peer affiliation 0.5727* 0.4039 0.4277 0.6331* 1.0000 APPENDIX E SAMPLE CHARACTERISTICS 146 147 APPENDIX E SAMPLE CHARACTERISTICS Agg,_Distribution by Sex Males Females Total n 1 3 years 10 mos.-3 years 11 mos. 2 5 7 2 4 years 0 mos.-4 years 2 mos. 3 6 9 3 4 years 3 mos.-4 years 5 mos. 13 10 23 4 4 years 6 mos.-4 years 8 mos. 14 20 34 5 4 years 9 mos.-4 years 11 mos. 13 15 28 6 5 years 0 mos.-5 years 2 mos. 10 6 16 Number of Siblings and Birth Order Total n 1 Only child 73 2 One sibling (younger) 11 3 One sibling older) 20 4 Two siblings younger) 3 5 Two siblings older) 2 6 Two siblings middle child) 4 7 Three siblings or more 4 Parents' Marital Status Total n 1 Married 77 2 Divorced 22 3 Separated 5 4} Single Parent 3 5 One or Both Parents Deceased 3 6 Remarriage (step-parent) 7 Parents' Level of Occupation1 Mother (n) Father (n) I Professional 12 22 II Business or Managerial 23 16 III Administrative, Technical, or Clerical 42 4 IV Semi-skilled or Skilled Manual Worker 3 38 V Unskilled Manual Worker 0 18 VI Housewife 19 0 VII College Student (full-time) 16 15 1Adapted from Hollingshead and Redlich (1958). APPENDIX F MODIFIED MINIATURE SITUATIONS TEST 148 149 APPENDIX F Modified Miniature Situations Test Instructions: An experimenter introduces the modified MST to each child by saying: “We are interested in the games that children like to play. All children usually like to play some of these games. These games are all make-believe; I would like you to tell me which one of these three games you want to play, and then I will let you play it. Remember you can only choose one game at a time. Just tell me which game you want to play." Three or four trial items were then presented to the child to check his/her comprehension of the required task. The following trial items were presented, with the instructions, repeated: "Remember, you can only choose one game to play at a time. Now which of these games do you want to play?“ a) “drink from a baby bottle vs. drink from a cup vs. sip through b; asighgwoh a chair vs. sit on a chair vs. crawl under a chair"; c "put on clown mask vs. put on princess/prince mask vs. put on witch/devil mask“; d) "build fort with blocks vs. knock down a fort vs. watch E build a fort“. Modified MST Items Order I--Identification with Provider-Protector vs. Eggpssor vs. Victim l. Hug Teddy Bear vs. Punch Teddy Bear vs. Pretend Teddy Bear bit you. 2. Pet Doggie vs. Pull Doggie's tail vs. Pretend Doggie bit you. 3. Pet the crocodile vs. Step on the crocodile vs. Pretend the crocodile bit you. 150 Order II--Identification with Provider-Protector vs. Victim vs.Aggressor 4. Feed Teddy Bear with baby bottle vs. Pretend Teddy Bear dumped milk over you vs. Hit the Teddy Bear with baby bottle. 5. Give Doggie a bone to eat vs. Pretend Doggie goes pee-pee over you vs. Hit Doggie with the bone. 6. Give Kitty a saucer of milk to drink vs. Pretend Kitty bit your finger vs. Hit Kitty with the saucer. Order III--Identification with Aggressor vs. c m vs. ov er- rotector 7. Stop on the snake vs. Pretend the snake bit you vs. Pet the snake. 8. Hit bumblebee with fly swatter vs. Pretend bumblebee stings you vs. Give bumblebee honey to eat. 9. Drown Kitty in tub of water vs. Pretend Kitty scratched you vs. Give Kitty a drink of water. Order IV--Identification with Aggressor vs. Proviaer-Protector vs.’Victim 10. Spank Teddy Bear vs. Hug Teddy Bear vs. Pretend Teddy Bear spanks you. ll. Shoot monkey with toy gun vs. Hug monkey vs. Pretend monkey shot you with toy gun. 12. Growl and roar at the tiger vs. Pet the tiger vs. Pretend the tiger is growling and roaring at you. Order V--Identification with Victim vs. Aggressor vs.“Provider-Protector l3. Pretend elephant hit you with his trunk vs. Shoot elephant with toy rifle vs. Feed elephant some grapes. l4. Pretend frog bit you vs. Hit frog vs. Hug Frog. 15. Pretend mouse bit you vs. Hit mouse vs. Hug Mouse. Order VI--Identification with Victim vs. Proviaer-Protector vs. Aggressor 16. Put bandaid on your finger where Kitty bit you vs. Put bandaid on Kitty where she has a cut vs. Spank Kitty. 17. Pretend crocodile bit you vs. Put crocodile in tub of water to swim vs. Drown crocodile in tub of water. 18. Pretend bunny shot you with toy gun vs. Hug bunny vs. Shoot bunny with toy gun. APPENDIX G DEVIATION DOLL PLAY 151 152 APPENDIX G DEVIATION DOLL PLAY] Instructions and Procedure: This method consisted of showing the child an open-topped, one-story doll house, and a standard family of four dolls, e.g., mother, father, boy/girl, and baby. E sat on the floor beside the dolls with S. E said: "See this doll house? It has rooms just like your house does." Pointing to each room in turn, E continued: "Here is the living room, kitchen, etc. Here are the people who live in the house; the Daddy, the Mommy, the boy/girl, and the baby. I'm going to tell you the beginning of some stories and you can show me and tell me how the stories end." The child doll of the same sex as S is kept in hand, and active, during the story, and then is handed to S at the end of E's description of the deviant behavior. E then asks S to sflgw and ggll_him what happens next. Stimulus Stories (told here for a boy; pronouns related to active doll are reversed for girl) 1. Table lamp; Mother and Father in kitchen; child in living room. “This boy is playing in the living room, jumping on the sofa. Daddy and Mommy said it was okay to play here, but to be careful. The little boy is having a lot of fun, but is not careful and he knocks the lamp off the table. Now you take the doll and show me and tell me how the rest of the story goes." 1Adapted from Sears, Rau, and Alpert (1965), pp. 323-324. 153 . Spilling juice; all of dolls in kitchen. "This boy is in the kitchen eating breakfast with the rest of his family. He doesn't like his juice, but he knows he must drink it. He is so mad because he has to drink his juice, that he knocks over the glass and spills it. Now you take the doll and show me and tell me how the rest of the story goes." . Grabbing baby's toy; all of dolls in living room. “Mommy and Daddy and the little boy are all watching baby play with his/her favorite toy on the floor. Mommy and Daddy go out to the kitchen for a minute and tell the little boy to watch the baby. Well, the little boy watched the baby playing, and decided he wanted the baby's toy, so he grabbed it away from the baby, and the baby started crying. Now you take the doll and show and tell me what happened next." . Bedtime disobedience; Mommy and Daddy in living room, child and baby in bedroom, baby in cradle. “It's getting very late, and the children are playing in their room. Mommy and Daddy come in and tell them it's bedtime, and put the children to bed. Now, this little boy knew he was sup- posed to be asleep, but he wanted to play. So he got out of bed and started playing again. Now you take the doll and show and tell me what happened next." . Stealing cookies; Daddy in living room; Mommy in kitchen, then living room; baby in bedroom; child in kitchen. "This little boy is in the kitchen watching his Mommy make cookies. She puts the cookies on the table and tells the boy not to eat any; he can have some after supper. But when the Mommy goes into the living room, the boy reaches over and takes some cookies. Now you take the doll and show and tell me what happens next." APPENDIX H DEFINITIONS OF SCORING CATEGORIES FOR DEVIATION DOLL PLAY 154 155 APPENDIX H Definitions of Scorin Cate ories Used in DeViation 9011 Play_ 1. Child's Perceptions of Parental Discipline A. Mode of Punishment 1) Verbal--an authority scolds the child, asks why he/she did that, tells child never to do it again or threatens punishment for repetition of deviant act. Examples: “You bad girl, why did you break that?" “Come here right this minute--I told you never to play on the couch." "I'm going to whip you if you do that again." "What do you think--I'm made of money?" "You are going to grow up to be a thief!" 2) Physical--authority spanks or otherwise hurts the child physically. Examples: Child takes mother/father doll and non- verbally portrays spanking. Child verbally reports spanking occurred with- out physically demonstrating it, e.g., "Mommy spanked the boy cause he was bad." Child verbally and/or physically displays aggression directed at the child doll by parental dolls, i.e., "Daddy threw the glass at the little girl“; "parents kill the child." 3) Denial of Privileges--authority forbids the child to engage in a favorite activity, i.e., playing a game, watching T.V. Examples: "No baseball for you tonight." “Tomorrow you will have to stay in all day." “You can't have supper because you ate all the cookies up." "Because you were so sloppy, you have to stay acme, and can't come with us to your cousin's ouse." 4) Isolation--chi1d is forced to remain alone for a time as punish- ment, authority figure is not present. Examples: “Shut up and go to bed." "Sit on the couch and don't move." Child is locked in bathroom/bedroom by authority figure. "Go outside and don't come in this house, till I say so." 156 5) Reasoning-~authority attempts to teach child why he/she should not engage in deviant acts again, and/or offer alterna- tive behaviors to satisfy child's needs, wishes, feelings, and/or desires, and/or sets limits. Examples: "If you eat up all the cookies, there will be none left for dessert, and you may not be able to eat your supper." "You should not play with baby's toy because it is his, and you do not like it when he plays with your toys." "The living room is not for jumping and run- ning; if you want to play tag, go outside in the future." "Next time you cannot get to sleep, call me and we'll talk, play a game, or I'll tell you a story. After ten minutes of that, then you must go to sleep." "You may have one cookie now, and then you can have the rest after supper." "Next time you want to reach a glass which is far away, ask me, and I'll give it to you." 6) Reflection of child's feelings/forgiveness--authority reflects child's feelings (i.e., fear, shame, anger, guilt, etc.), needs, wishes, and/or forgives the child spontaneously. Examples: "zoukwere really frightened when the lamp ro e." "You were angry at baby, so you teased him by taking his toy." "The dark is scary for you." "That's okay. I know you didn't mean to break anything." "It was just an accident." "You were so excited, that you forgot it was bedtime." 7) Forced Fixing--authority forces child to fix, restore, or "undo" deviant act, with or without adult's help. Examples: "Now I want you to help me clean up this mess. Here's a rag to wipe up the juice with." “Give that toy back to baby." "I'm going to make some more cookies. Get the bowl out." "Put all your toys away, and get into bed.“ "Pick up the lamp, and put it on the table." 157 B) Ignoring--authority discovers child's deviant act, and ignores child, while fixing, restoring, or "undoing" the conse- quences of that act. Examples: "Mommy won't talk to the little boy anymore.“ "She just goes and makes more cookies." "Daddy sees the little girl playing with her toys instead of sleeping, so he goes and gurns off the light, and closes the bedroom oor." "Daddy picks up the lamp, and starts reading the paper." "The little boy asks the Mommy if he could go figt, she doesn't answer cause she's mad at m. ll "Mommy wipes up the milk, packs her suitcase, and goes away from the family, cause the little girl was bad." . "Daddy picks up the baby's toys and starts play- ing with the baby" (other child is ignored). . Agent(s) of Punishment 1) Mother (alone)--father doll does not participate in discipline in any way. 2) Father (alone)--mother doll does not participate in discipline in any way. 3) Mother and Father (together)--supporting one another, i.e., sending child consistent messages. Examples: "Mommy yells, calls in Daddy to give child a spanking, or vice versa. Mothgr and Father both reprimand or spank chi . Mother spanks child, Father sends child to bed, or vice versa. Both parents force child to fix, restore, or undo consequences of deviant act. 4) Mother and Father (together)--acting in opposition to one another, i.e., sending the child inconsistent messages. Examples: “Mother yells at child, Father yells at mother for "picking" on the child, or vice versa. "Father tells child to go to bed, while Mother tells child he has to wipe up the juice at once.“ "Mother blames child for breaking lamp, while Father laughs, and says it was an accident." “Mother asks Father to spank child, Father refuses and says no one is to bother him." II. 158 "One parent ignores the other." “One parent leaves/withdraws from the situa- tion.” Child's Behavioral Reaction to Fantasized Transgression l) Denial--chi1d denies responsibility for deviation in response to accusation, or by failing to respond when confession is solicited by authority, or denies before he/she is impli- cated. Examples: "Did you break the lamp?" "It wasn't me." "I saw you steal those cookies." I didn't do nothing." "Who spilled the juice?" "I don't know." "I didn't do it, baby did it." "A burglar broke into the house and stole baby's toy." "A big wind blew in and knocked down the lamp." 2) Child Gets Away With It--the deviation is never discovered, or the child is never implicated. 3) Confession--Child admits or confesses he/she was responsible for deviant act, whether or not he/she has been accused. Examples: "1 ate up all the cookies." “I knocked over the lamp, when I was playing." "I can't sleep--I'm afraid." "I wanted Baby's toy, so I took it when she wasn't looking." 4) Apology--chi1d apologizes spontaneously for deviant act. Examples: "I'm sorry, that baby is crying for her toy." "I'm sorry, I won't play in the living room anymore." "I didn't mean to break the lamp, I'm sorry, I know you really liked it." "I'm sorry I ate up everyone else's cookies." 5) Spontaneous Fixing--child restores or "undoes" deviant act; without parental knowledge. Examples: "Child puts lamp back on table, without anyone seeing her/him. Child goes out and boys new lamp. Child bakes more cookies/goes out and buys more cookies. Child gives baby back the toy, or child buys toy for baby, or asks a friend to give him a toy for baby. 159 6) Hiding--child hides himself/herself or the evidence from author- ity figures. Examples: Child hides empty cookies plate in closet/ refrigerator. Child climbs out bedroom window and runs away from home. Child hides lamp under the couch. - Child and baby hide under bed and play with baby's toy. Child hides baby's toy, or throws it into garbage. Child hides under his bed. Child locks himself in the bathroom. 7) Crying--child in the story cries and/or pleads with the adult not to get angry or punish him/her. Examples: "Don't hit me.“ "Don't get mad." “I'm sorry, don't leave me." "I won't do it again, please can I watch T.V.?" "Please let me go out, I won't do it again." 8) Mute, Silent, Withdrawal--child sits/stands silently while parent disciplines him/her. Child says nothing in response to adult questions. In this case, the deviation is always discovered, and the child is always punished, and com- plies, with punishment. 9) Continued Deviant Aggression--child continues deviant behavior, becoming more destructive, and aggression may be directed at objects, people, and/or the self. Examples: Child breaks lamp and turns over all furniture in the living room. Child kills parents/baby/self/Monster kills parents, etc. Child eats up all cookies, and everything else in the refrigerator. Alligator eats up all the people. Child makes more cookies, with a witch's poison. Child steals baby's toy, becomes a burglar, and winds up in jail. Flood comes and drowns everybody in the house. APPENDIX I MEANS AND STANDARD DEVIATIONS 0F CATEGORIES 160 161 APPENDIX I Table I. Rank Order of Mean Category Usage and Standard Deviation Category Mean S.D. l. Provider-Protector 7.96 2.50 2. Aggressor 7.52 2.40 5. Verbal Punishment 3.07 2.06 4. Physical Punishment 2.79 1.36 3. Victim 2.30 2.23 19. Continued Deviant Aggression 2.13 1.26 6. Isolation 2.06 1.55 21. Mother (alone) 1.97 1.87 16. Spontaneous Fixing 1.96 0.90 23. Mother and Father (together, agreeing) 1.76 1.64 20. Mute, Silent, Withdrawal 1.71 1.41 B. Forced Fixing 1.67 1.32 18. Crying 1.66 1.29 15. Confession 1.52 1.30 14. Apology 1.47 1.12 12. Child's Denial 1.47 1.20 7. Denial of Privileges 1.40 1.30 9. Reasoning 1.29 0.92 22. Father (alone) 1.27 0.47 11. Ignoring 1.19 1.06 10. Reflection of Feelings/Forgiveness 1.10 1.03 13. Gets Away With It 0.93 0.58 APPENDIX J INTER-RATER RELIABILITIES 0F DEPENDENT MEASURES 162 163 Table I. Inter-Rater Reliability for Scoring of Dependent Measures Mean Category Dependent Measures RlRZ R1R3 R2R3 Reliability Self-Ropresentation in Fantasy 1. Provider-Protector 0.9955 0.9938 0.9888 0.9927 2. Aggressor 0.9910 0.9904 0.9657 0.9823 3. Victim 0.9841 0.9837 0.9632 0.9778 Perception of Parental Response to Fantasized Transgression 4. Physical Punishment 0.9691 0.8901 0.9414 0.9335 5. Verbal Punishment 0.8834 0.8290 0.9319 0.8814 6. Isolation 0.7910 0.7519 0.9105 0.8187 7. Denial of Privileges 0.6597 0.5987 0.8976 0.7186 8. Forced Fixing 0.7886 0.7670 0.8712 0.8089 9. Reasoning 0.8825 0.7810 0.8932 0.8522 10. Reflection of Feelings/ 0.9090 0.8696 0.9453 0.9079 Forgiveness 11. Ignoring 0.9276 0.8888 0.9230 0.9131 Child's Behavioral Reaction to Fantasized Transggession 12. Denial 0.9341 0.9226 0.9260 0.9309 13. Gets Away With It 0.8265 0.7695 0.9228 0.8396 14. Apology 0.9122 0.8525 0.9365 0.9004 15. Confession 0.8595 0.8430 0.9627 0.8884 16. Spontaneous Fixing 0.8102 0.7851 0.8914 0.8189 17. Hiding 0.8030 0.8322 0.8759 0.8360 18. Crying 0.8879 0.8571 0.9729 0.9059 19. Continued Deviant Aggression 0.8864 0.8615 0.9203 0.8894 20. Mute, Silent, Withdrawal 0.8165 0.7645 0.9336 0.8381 Agent(s) of Punishment 21. Mother (alone) 0.9006 0.9020 0.9201 0.9075 22. Father (alone) 0.8215 0.8019 0.9596 0.8610 23. Mother and Father (together, 0.9814 0.9831 0.9579 0.9743 agreeing) Total Average Mean Reliability 0.8396 0.8486 0.9313 0.8813 R1 = author; R2 = Kay Guimond; R3 = Joseph Pizzo APPENDIX K INTERNAL CONSISTENCY 0F MODIFIED MST 164 165 APPENDIX K SUMMARY OF INTERNAL CONSISTENCY 0F ITEMS ON MODIFIED MST Item f—TOFder Item-Hfiole Number Number Variable (Mode of Identification) Correlation TABLE I l I Provider-Protector 0.4472 2 I Provider-Protector 0.5238 3 I Provider-Protector 0.5342 4 II Provider-Protector 0.3878 5 II Provider-Protector 0.4481 6 II Provider-Protector 0.5975 7 III Provider-Protector 0.3865 8 III Provider-Protector 0.3502 9 III Provider-Protector 0.2958 10 IV Provider-Protector 0.5302 11 IV Provider-Protector 0.4179 12 IV Provider-Protector 0.4004 13 V Provider-Protector 0.4809 14 V Provider-Protector 0.5034 15 V Provider-Protector 0.7021 16 VI Provider-Protector 0.2135 17 VI Provider-Protector 0.2375 18 VI Provider-Protector 0.5154 Mean 0.44 TABLE II 1 I Aggressor 0.5311 2 I Aggressor 0.5774 3 I Aggressor 0.4217 4 II Aggressor 0.3373 5 II Aggressor 0.5589 6 II Aggressor 0.5022 7 III Aggressor 0.4017 8 III Aggressor 0.3258 9 III Aggressor 0.3335 10 IV Aggressor 0.5705 11 IV Aggressor 0.3035 12 IV Aggressor 0.3322 13 V Aggressor 0.3582 14 V Aggressor 0.5031 15 V Aggressor 0.6333 16 VI Aggressor 0.1134 17 VI Aggressor 0.2994 18 VI Aggressor 0.4940 Mean 0.42 continued 166 APPENDIX K--continued Item Order Item:Who1e Ngmber Number Variab1e (Mode of Identification) Corre1ation TABLE III 1 I Victim 0.4820 2 I Victim 0.3798 3 I Victim 0.5936 4 11 Victim 0.4113 5 II Victim 0.2405 6 11 Victim 0.4655 7 III Victim 0.4427 8 III Victim 0.1503 9 III Victim 0.2492 10 IV Victim 0.4947 11 IV Victim 0.2959 12 IV Victim 0.4127 13 V Victim 0.3199 14 V Victim 0.4481 15 V Victim 0.3835 16 VI Victim 0.3320 17 VI Victim 0.1503 18 VI Victim 0.2250 Mean 0.36 52!; Order I = Provider-Protector vs. Aggressor vs. Victim Order II a Provider-Protector vs. Victim vs. Aggressor Order III a Aggressor vs. Victim vs. Provider-Protector Order IV = Aggressor vs. Provider-Protector vs. Victim Order V = Victim vs. Aggressor vs. Provider-Protector Order VI . Victim vs. Provider-Protector vs. Aggressor APPENDIX L FACTOR ANALYSIS OF DEPENDENT MEASURES 167 168 APPENDIX L Summary of Factors Obtained from Dependent Measures Factor I--Competence-Incompetence Var. 1. Provider-Protector Var. 9. Reasoning Var. 14. ApoIogy Var. 15. Confession Var. 10. Ref1ection of Fee1ings/Forgiveness Var. 2. Aggressor Var. 17. Hiding Var. 18. Crying Highest Loading Prop. Var. Cum. P.V. Factor II--Chi1d's Behaviora1 Reaction to Fantaiiied‘Transgression Var. 19. Continued Deviant Aggression Var. 20. Mute, Si1ent, Hithdrawa1 Highest Loading Prop. Var. Cum. P.V. Factor III--Agent of Punishment Var. 22. Father (a10ne) Var. 21. Mother (a10ne) Highest Loading Prop. Var. Cum. P.V. Factor IV--Agent of Punishment Var. 21. Mother (a10ne) Var. 23. Mother and Father (together, agreeing) Highest Loading Prop. Var. Cum. P.V. Rotated Factor Had—1m. 0.86 0.75 0.69 0.59 0.57 -.81 -.55 -.54 0.86 0.1953 0.1953 -.66 0.82 0.82 0.0718 0.2671 0.92 -.53 0.92 0.0731 0.3403 0.76 -.88 .88 0.0786 0.4189 continued 169 APPENDIX L--continued Factor V--Perceived Parental Response to FantasizedTTransgression Var. ll. Ignoring Var. l3. Gets Away With It Var. 5. Verbal Punishment Highest Loading Prop. Var. Cum. P.V. Factor VI--Fix1ng_ Var. 16. Spontaneous Fixing Var. 18. Forced Fixing Highest Loading Prop. Var. Cum. P.V. Factor VIP-Victim Var. 3. Victim Prop. Var. Cum. P.A. Factor VIII--Denial Var. 7. Parental Denial of Privileges Var. 12. Child's Denial of Fantasized Transgression Var. l7. Hiding Highest Loading Prop. Var. Cum. P.V. Rotated Factor Loadings 0.68 0.76 -.66 0.76 0.0894 0.5083 0.72 -.69 0.72 0.0631 0.5713 -.77 0.0655 0.6386 0.77 -.61 0.77 0.0694 0.7062 170 umzcwucoo nmmp.a Nmoo.o mwpo.u mwno.o ¢Nm~.o mnmo.o «mmom.o m¢m¢.u .mp .gm> Nwm—.u mmNN.u mmoo.u mpep.o mmmo.n Nuom.o mnvm.n «mo¢m.- .mp .Ln> omm¢.u PRO—.1 mono.n mmpm.o mmmo.s N~o~.o Nm¢—.u «onmm.o .np .em> N¢No.u omm—.u apmms.o mmvc.u Pump.o memo.n cmno.u mm¢~.o .op .em> Nemp.o mmFm.o mfipo.u FmoN.u mmpm.u omop.u ommm.u «mem.o .mp .gm> com—.0 Pmmm.o mmom.a mmeo.u mem.u Nomo.u mmo_.u «mamm.o .¢~ .Lm> mmeo.o ammo.o mpmm.u «mpmn.u mmmo.o w~m~.o ommo.u m¢_P.o .mp .Lm> «umpo.u omm~.1 mew—.1 memo.o —m~o.o we—o.o me~.u m—mm.u .NF .Lm> vao.u momp.n wmmp.o «mmmm.o mopp.u NPmo.o memo.u m¢¢~.u .—F .em> mmno.u mmn_.o mmmm.o nmmo.u o¢-.u ~mm~.u mmpp.u «unnm.o .o_ .sm> mmmp.o ano.o «_om.u mmpo.o uo¢~.o “pep.s mpeo.o «mom~.o .m .Lw> mpoo.o noop.u «comm.u oe~o.u mpmo.o mmom.u mm_o.o m¢m~.o .m .Lm> «mmnn.o ¢m¢N.u mmp_.u appo.o Numo.o mmmo.o mmpp.u uuuo.u .u .Lw> ~¢~e.o mmpv.u pmmo.o Nmm¢.u mmmo.o ommo.: opm~.o omno.u .o .gm> o-~.o mmmp.a pneo.u «mmoo.u oomo.o m¢-.o ammo.o newm.o m .gm> mmuo.s mom—.1 eepv.u Num¢.u vamp.o o-~.o Pmmm.u mmn~.u .v .Lm> meo.o ammum.u mpFo.o mmeo.o om¢~.u Nm_~.u woNo.n meF.u .m .La> mo~o.o memm.o «mmo.u memo.o ammo.u Nmm_.n n¢mo.u «onpm.o .N .Lm> mmo~.u mnoo.o wmeo.o ommo.u mmm—.o mmo_.o pmso.o «mmmm.o .P .Lm> HHH> Louomm HH> Lopomm H> Louumm > Lopumm >H Louumm HHH Lepumm HH Lopumu .mlmmmmmm «3:83 .583 .3332. mmm=m¢ Fmoo.o .FF .Lm> ommm.o .op .Lm> o¢om.o .mw .Lm> mmmm.o .NN .Lm> NNFm.o ._N .eu> mmmo.o .m .Lm> mmm~.o .om .L~> oumo.o .m .L~> N¢_~.o .mp .L~> ouse.o .N .La> opmm.o .m_ .La> mmme.o .m .La> meo.o .NF .La> owmm.o .m .La> Naom.o .02 .L~> mmuo.o .e .La> «Amo.o .m_ .La> mmmo.o m .2a> Nam~.o .¢_ .La> mmom.o .N .ea> «mmo.o .m. .e~> eemm.o ._ .Lm> mat: FEB-=8 ~oo~.o mmme.o «_Nm.o mmom.o mm_¢.o moem.o P~m~.o mma~.o .>.a.s=u emmo.o mmeo.o _mmo.o «mmo.o ammo.o _mxo.o _mho.o mmm~.o.ea>.aoea mu-.o mm-.- _m-.o m~m~.o Pmmm.- mowa.o wa~m.o mmom.o uaos.e: Nemo.- wmm_.- mmoo.- oemo.o .Pmmm.- Pmem.- ammo.o memo.- .mm .L~> mm_o.o o¢o_.o oomo.o homo.o memo.o .moum.o memo.o NPmo.- .NN .Lu> mmeo.o Famo.o mm_o.- mmmo.- .mom~.o .Nmmm.- New..- memo.o ._~ .L~> mm~0.- memo.- m_op.- m.mo.- “moo.- osm_.o .maum.o mes..- .om .e~> HHH> League HH> cabana ~> League > Logged >H Logged HHH Logged HH Logan; _ League #533 L33... .8333. uo==_aeau--4 xHazmaa< APPENDIX M TESTS OF SIMPLE EFFECTS 172 173 APPENDIX M TESTS OF SIMPLE EFFECTS 1. Sex x Groups The multivariate ana1ysis of variance reflected a significant (p‘=0.0021) Sex x Group interaction which was associated with a signifi- cant univariate effect for Reasoning (p<:0.0016). Tests of simple effects of the group conditions, within each condition of sex, were performed to explore this interaction. Looking within females first, a simple main effect was found (F = 15.79). The reader is referred to the main body of the text where individual comparisons between groups are presented. Looking within males next, a significant simple main effect also was obtained (F = 14.17) for groups. The reader is referred to the text for further comparisons between groups. Secondly, the multivariate analysis of variance revealed a marginally significant (p<:0.0594) Sex x Group interaction which was associated with a significant univariate effect for Ignoring (p<:0.0122). Tests of simple effects of the group conditions within each condition of sex were performed to explore this interaction. Results for females (F = 2.18) and males (F = 2.68) were nonsignificant. II. Order x Group The multivariate analysis of variance revealed a significant (p<:0.0007) Order x Group interaction which was associated with a significant univariate effect for Apology (p<:0.0035). Tests of simple 174 effects of the group conditions. within each order, were performed to explore this interaction. Looking within Order A first, a simple main effect was found for groups (F - 12.44). The reader is referred to the main body of the text where individual comparisons between groups are presented. Looking within Order 8 next, results yielded a nonsignifi- cant (F = 2.694) simple main effect for groups. III. Sex x Order x Group The multivariate analysis of variance reflected a significant (p<:0.0044) Sex Order x Group interaction which was associated with a marginally significant univariate effect for identification with the Aggressor (p<:0.0243). The data was examined via Sex x Group simple anovas within each order. Looking within Order A first, the Sex x Group simple interaction effect was nonsignificant (F = 0.65096). Therefore, no further analyses were appropriate for males and females. However, a significant simple main effect for groups was found (F = 22.88). The reader is referred to the main body of the text where individual comparisons between groups are presented. Hithin Order 8 a marginally significant simple interaction effect was found (F = 4.92). Tests of simple effects of the group conditions within each condition of sex were performed to explore this interaction. Looking within females in Order 8 a significant simple main effect was found for groups (F = 12.353). Individual comparisons between groups are presented in the main body of the text. Next, for males within Order B, a significant simple main effect for groups was also obtained (F = 17.594). Again, these results are elaborated in the main body of the text. 175 In addition, analysis of variance further revealed a significant (p<:0.0035) Sex x Order x Group interaction, associated with the uni- variate effect for identification with the Victim. The data were examined via Sex x Group simple anovas within each order. Looking within Order A first, results yielded a nonsignificant simple inter- action effect for Sex x Group (F = 0.3155). Thus, no further analyses with respect to males and females were performed. The simple main effect for groups was also nonsignificant (F = 2.03). However, within Order B a significant simple interaction effect for Sex x Group was obtained (F = 13.53). Tests of simple effects of the group conditions within each condition of sex were performed to examine this interaction. For females within Order B, a significant simple main effect for groups was found (F . 19.246). The reader is referred to the main body of the text where individual comparisons between groups are presented for females. For males within Order 8. results reflected a nonsignificant (F a 0.998) simple main effect for groups and, thus, no further analyses were appropriate for males. Finally, the multivariate analysis of variance revealed a sig- nificant (ps:0.0021) Sex x Order x Group interaction which was associ- ated with a marginally significant univariate effect for Denial (p<:0.0384). The data were examined via Sex x Group simple anovas within each Order. Looking within Order A first, a marginally signifi- cant Sex x Group interaction was found (F = 4.90). Tests of simple effects of the group conditions within each condition of sex were per- formed to explore this interaction. For females within Order A. 176 a marginally significant simple main effect for groups was found (F a 4.53). Individual comparisons between groups for females are presented in the main body of the text. For males within Order A, a marginally significant simple main effect for groups was also obtained (F a 4.06). Again, the reader is referred to the main body of the text for individual group comparisons across males. Looking within Order 8 next, a marginally significant simple Sex x Group interaction effect was found (F = 3.66). Tests of simple effects of the group conditions within each condition of sex were per- formed to explore this interaction. For females within Order 8, results were nonsignificant (F - 2.56); thus, no further analyses were carried out for females. For males within Order 8. a marginally sigp nificant simple main effect for groups was obtained (F - 4.90). The reader is referred to the main body of the text where individual group comparisons for males are presented.