* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Safran, 1990 - The Safran Lab
Attachment therapy wikipedia , lookup
Attachment in children wikipedia , lookup
History of attachment theory wikipedia , lookup
Maternal deprivation wikipedia , lookup
Caring in intimate relationships wikipedia , lookup
Emotionally focused therapy wikipedia , lookup
Human bonding wikipedia , lookup
ABSTRACT This study addresses the impact of therapists’ personalities on psychotherapy process within the context of relational psychodynamic theory. The differential role of therapists’ interpersonal schemas of their mothers verses their fathers were assessed in relation to their ability to form a working alliance, and work through alliance ruptures. The role of the father tends to be less understood than that of the mother, with limited research addressing how therapists’ relationships with their fathers may impact their relational schemas and consequent impact on psychotherapeutic process. This study draws on findings from the child-development literature suggesting that fathers are important in the formation of conflict resolutions skills, and considers therapists’ interpersonal schemas of their fathers in relation to rupture resolution. Therapists interpersonal schemas of their mothers and fathers are assessed using the interpersonal schema questionnaire (N = 96), and then calculated along four different kinds of situations; Hostile, friendly, submissive and dominant. Interpersonal schemas along these four dimensions for mothers and fathers are then assessed in relation to therapists and patients reports of working alliance (Working Alliance Inventory), rupture resolution (Post-Session-Questionnaire), and session smoothness and depth of experience (Session Evaluation Questionnaire) across the first five psychotherapy sessions. Therapist introjects using the Intrex questionnaire were also assessed in relation to working and alliance and rupture resolution variables as a point of comparison to the interpersonal schema questionnaire. All therapists and patients were seen as part of the Brief Psychotherapy Research Program. Each therapist was seen by multiple and varying numbers of patients (ranging from 1-8, with a total of 186 patients in the overall sample). Multilevel modeling was employed to account for this “nesting,” while assessing the relationship between independent and dependent variables. Findings suggested that therapists’ interpersonal schemas of their fathers and mothers in friendly and dominant situations are predictive of rupture resolution. However, only the father was predictive of both patient and therapist ratings, as well as perceived smoothness and depth of experience during rupture sessions. Findings also suggested that therapists’ affiliative introjects are predictive of rupture resolution. Theoretical and clinical implications are discussed. THERAPISTS’ CONTRIBUTION TO THE WORKING ALLIANCE AND RUPTURE RESOLUTION: THE DIFFERENTIAL ROLE OF THERAPISTS’ INTERPERSONAL SCHEMAS OF THEIR MOTHERS AND FATHERS IN PSYCHOTHERAPY by Melanie Fox-Borisoff Dissertation Submitted to the New School for Social Research of New School University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy November 2009 Dissertation Committee: Jeremy D. Safran, Ph.D. J. Christopher Muran, Ph.D. Howard Steele, Ph.D. Claudia, Barrachi Ph.D. Table of Contents List of Tables ..................................................................................................................... iv Introduction ..........................................................................................................................1 CHAPTER 1: INTRODUCTION AND THEORETICAL BACKGROUND.....................3 The Role of the Therapist in Psychotherapy Process: The Working Alliance, Ruptures, and the Internalization of Self-Other Interactions ...........................6 The Working Alliance and the Therapist’s Contribution to Psychotherapy Process .........................................................3 A Two-Person Psychology: Me-You Representations ................................13 Sullivan’s Personifications of Self and Other: An Introduction Two Person Psychology.......................................................13 Internal Working Models .................................................................17 Representations of Interactions that have been Internalized............18 Interpersonal Schema Theory ..........................................................19 The Interpersonal Circumplex: A Model for Psychotherapy Research Within the Framework of A Two-Person Psychology...........22 Structural Analysis of Social Behavior: The Intrex .........................23 Kiesler’s Interpersonal Transaction Cycles and Circumplex Model ...................................................................................25 Assessing Therapists’ Me-You Representations .........................................29 The Interpersonal Schema Questionnaire ........................................29 Reliability and Validity of the ISQ ......................................32 Relevant Studies Applied to Psychotherapy Process .......................33 The Role of the Father and Interpersonal Schemas .........................36 CHAPTER II: CHILD DEVELOPMENT, THE ROLE OF THE FATHER, AND CONFLICT RESOLUTION ....................................................................................38 Child-Father Attachment Style and Children’s Social Competence............38 Father-Child Conflict and Conflict Resolution with Peers ..........................51 i Affect-Regulation and Conflict Resolution: The Role of the Father ...........58 Summary of Father vs. Mother Influences in Child Development ..............63 CHAPTER III: STATEMENT OF PURPOSE AND HYPOTHESES................................64 Statement of Purpose ...................................................................................64 Hypotheses ...................................................................................................66 CHAPTER IV: METHOD ...................................................................................................67 Overview ......................................................................................................67 The Brief Psychotherapy Research Program ...............................................69 Participants ...................................................................................................71 Therapists .........................................................................................71 Patients .............................................................................................72 Measures of Therapist Internal Process (Independent Variables) ...............73 Interpersonal Schema Questionnaire (ISQ) .....................................73 Intrex Questionnaire.........................................................................74 Measures of Psychotherapy Process (Dependent Variables) .......................75 Working Alliance Inventory (WAI).................................................75 Post Session Questionnaire and Rupture Section ............................76 Session Evaluation Questionnaire (SEQ) ........................................76 Procedures ....................................................................................................77 Pilot Study........................................................................................77 Current Study ...................................................................................83 CHAPTER V: ANALYSES ................................................................................................82 CHAPTER VI: RESULTS ...................................................................................................87 Descriptives of Ratings ................................................................................87 Multilevel Modeling ....................................................................................88 Hypothesis I: Therapist ISQ, Working Alliance and Rupture Resolution. ......................................................................................88 Hypothesis II: Therapist ISQ and Session Smoothness and Depth .97 Hypothesis III: Therapists’ Introjects .............................................101 ii Random Effects ...........................................................................................103 Summary of Regression Coefficients .........................................................104 CHAPTER VII: DISCUSSION ..........................................................................................105 Rupture Resolution .....................................................................................105 Working Alliance ........................................................................................108 Therapist Introjects and Rupture Resolution ..............................................109 Limitations and Future Directions ..............................................................110 Conclusion ..................................................................................................113 REFERENCES ...................................................................................................................116 APPENDIX A .....................................................................................................................133 APPENDIX B .....................................................................................................................136 APPENDIX C .....................................................................................................................138 iii LIST OF TABLES Table 1: Preliminary Results for Therapist ISQ and Rupture Resolution ..................79 Table 2: Preliminary Results for Therapist ISQ and Working Alliance .....................80 Table 3: Summary of Dependent Variable Ratings ....................................................87 Table 4: Summary of Therapist-ISQ Quadrant Ratings for Desirability ....................88 Table 5: ISQ-Mother in Friendly Situations and Therapist/Patient Rupture Resolution ....................................................................................................90 Table 6: ISQ Mother in Dominant Situations and Therapist/Patient Rupture Resolution ....................................................................................................92 Table 7: ISQ Father in Friendly Situations and Therapist/Patient Rupture Resolution ....................................................................................................94 Table 8: ISQ Father in Dominant Situations and Therapist/Patient Rupture Resolution ....................................................................................................96 Table 9: ISQ-Father in Dominant Situations and Perceived Session Smoothness ....98 Table 10: ISQ-Father in Friendly Situations and Depth of Experience in Session .....100 Table 11: Therapists’ Affiliative Introjects and Rupture Resolution ..........................102 Table 12: Summary of Regression Coefficients for all Main Effects of ISQ/Intrex…104 iv Introduction Psychotherapy research has for the most part devoted it’s time to assessing patient characteristics and patient contributions to psychotherapy process. However, there is a burgeoning interest in how the therapist’s personality impacts psychotherapy process. Although these studies are still limited, there is evidence supporting further research in this area and that therapist’s interpersonal histories may impact their effectiveness as psychotherapists. In particular, preliminary studies have demonstrated on a couple of occasions that therapists’ relationships with their fathers are important determinates of psychotherapy therapy process and outcome (Nelson, 2000; Fox-Borisoff, unpublished MA thesis). In particular, therapists’ interpersonal schemas (expectations in social interactions) with their fathers may be important predictors of their ability to resolve ruptures (moments of tension or deterioration of therapist-patient communication) with patients during therapy, while the relationship with the mother may be more important for the formation of the therapeutic alliance. Convergent with these findings, recent research within the child development domain is suggesting that fathers may play an important and distinct role in the formation of mental structures that help maintain healthy relationships. For example, research is suggesting that fathers’ are particularly important for the formation of social competency skills, particularly conflict resolution, while mothers may be more important to the development of feelings and skills such as empathy that aid in the initial formation of relationships. The first chapter reviews the history of the working alliance concept and the role of the therapist, interpersonal theory, and reviews a few studies addressing how therapists’ own interpersonal schemas of their 1 fathers’ verses their mothers’ affect psychotherapy process. In the second chapter, a review of the child development literature on social competency formation supports and hopefully deepens the understanding of the relationship between fathers verses mothers, interpersonal schema formation, and subsequent abilities to establish trusting relationships and confront conflict in everyday life as well as in psychotherapy. 2 Chapter I The Role of the Therapist in Psychotherapy Process: The Working Alliance, Ruptures, and Internalization of Self-Other Interactions The Working Alliance and the Therapist’s Contribution to Psychotherapy Process The concept of the therapeutic alliance can be traced back to Freud (1912) who suggested that the analyst maintain “serious interest” and “sympathetic understanding” of their patient so that the patient’s healthy self can form a positive relationship with their analyst. Freud initially hypothesized that this accepting stance of the therapist, and the patient’s attachment to therapist would result in a “positive transference” that needs to be interpreted. For example, Freud believed that the patient would begin to associate their positive relationship with their therapist to other affectionate people in their life, and although beneficial, this would not reflect the real relationship between the patient and the therapist. However, in his later writing, Freud seemed to modify his view of the “positive transference” and suggested that a positive attachment to the therapist could be based in reality and may in fact facilitate the healing process. Despite Freud’s statement that the therapist should maintain a position of sympathetic understanding towards the patient, Freud also warned against countertransference and ardently advised analysts to go through their own analysis so that they could maintain neutrality with the patient, and essentially act as a mirror capable of reflecting the patient’s neurosis. This contradiction has led theorists to argue what the actual extent of the therapist’s contribution to the formation of an alliance is and should 3 be, verses how much of this alliance is as Freud originally proposed, a “positive transference” resulting from the patient’s personality (e.g., Brenner, 1979). Zetzel (1956) introduced the term “therapeutic alliance” in describing the relationship between the patient and therapist, and is generally credited as being the first to map out this concept. She viewed the therapeutic alliance as distinct from transference neurosis and believed that it was a recapitulation of the infant-mother relationship, whereas the patient turns to the therapist for help like an infant turns to the mother for help. Zetzel proposed that the alliance differs from transference neurosis in that the alliance is based upon a patient’s solid object relations from infancy, and consists of the non-neurotic component of the patient-therapist relationship. She argued that a healthy analysis consists of the patient oscillating between moments dominated by the working alliance and moments dominated by the transference, and that the alliance allows the patient to step back and reflect upon differences between relationships with others, verses the real nature of the relationship with the therapist. Overall, Zetzel emphasizes the impact of the patient’s personality for forming a patient-therapist relationship, rather than how the therapist’s qualities may contribute to this process. Zetzel acknowledges the role of the therapist in the formation of the alliance, for example, she proposes that the analyst be like a good mother creating a supportive environment that fosters underlying trusts, and is capable of making “intuitive adaptive responses”. However, she does not consider that the therapist’s object relation development may also be important for a real relationship to occur. 4 Greenson (1965) was the first to use the term “working alliance”. He extended Zetzel’s work in suggesting that the transference neurosis and the working alliance were separate concepts, both requiring equal attention. Greenson argued that a positive working alliance needs to be present in order for transference neurosis to be analyzed. He stated that “for a working alliance to take place, the patient must have the capacity to form object relations since all transference reactions are a special variety of them” and that the patient’s contribution to the working alliance depends on adequate ego strength. For example, he stated that working alliance depends on “his [the patient’s] capacity to maintain contact with the reality of the analytic situation and also his willingness to risk regressing into his fantasy world. It is the oscillation between these two positions that is essential for analytic work.” However, Greenson also considered the therapist’s contribution to the formation and maintenance of the working alliance. For example, he argued that analysts often take Freud’s suggestion of maintaining neutrality to an extreme and consequently come across as cold, authoritarian and rigid, smug and aloof to their patients. Greenson argued that in order for an alliance to form, analysts need to work in a manner that is “realistic and reasonable” and to maintain respect for patients as humans, and show “consistent concern for the rights of the patient throughout the analysis.” In contrast to those who proposed that the working alliance is a distinct construct, several theorists (e.g., Brenner,1979; Curtis, 1979; and Hanly,1992) criticized the above ego psychologists and felt that it is purposeless, and at times counterproductive to consider an alliance as separate from the transference. For example, Brenner (1979) felt that the alliance concept may lead analysts to leave important material unexplored and 5 ultimately limit the therapeutic process. Similarly, Hanly (1992) suggested that the alliance concept may result in an overvaluation of rational therapeutic processes, while leaving unconscious material unattended too (In Safran & Muran, 2000). Overall, these theorists believed that the alliance concept may mask resistances and unconscious conflicts such as a wish for the therapist to be like a parental figure, or an attempt to maintain equal ground with the therapist. Athough the concept (and the controversy over) the therapeutic working alliance has generally arisen out of the psychodymanic school of thought, other areas of psychology have hinted at similar ideas, and place an increased emphasis on the role of the therapist. For example, interpersonalists such as Harry Stack Sullivan (1953), and Clara Thompson (1964) emphasized the importance of the real relationship between the patient and the therapist. In addition, by assuming a two-person psychology where the therapist is embedded in the therapeutic process, interpersonalists were in a way expressing an implicit concern for a working alliance without mentioning the term itself. Relational psychoanalysis forgoes the concept of real verses unreal relations and argues that there are multiple truths at any given time and based on social construction, thus the idea that there are real verses unreal aspects of the therapeutic relationship is a “meaningless concept” (Safran and Muran, 2000). According to Safran and Muran, the therapeutic emphasis in relational psychoanalysis is more on mutuality, enactments and spontaneity than it is on neutrality and objectivity. They suggest that for this reason the working alliance concept is not as necessary to relational psychotherapists, however, they 6 also suggest that is essentially what underlies mutuality between therapist and patient, and that most analysts take it for granted as a primary medium for change. Zetzel and Greenson acknowledged the need for the therapist to be supportive and respectful of their patients in order for an alliance to form. However, they assumed that given this even, and as some would argue maternal-like stance (Zetzel, 1966; Friedman, 1969, Sandler and Colleagues, 1969) the patient’s ego capacities would ultimately determine the outcome of the alliance. Interpesonalists, and relational analysts place more emphasis on the role of the therapist, however, traditionally they have only considered the working alliance insomuch as an implicit part of the change process. In contrast, Edward Bordin’s (1979) conceptualization of the working alliance offered a pantheoretical application of the concept and has been responsible in part for renewed interest in the topic, as well as a closer look at the contribution of the therapist in terms of the alliance across psychotherapy modalities. Bordin’s broad definition builds on Zetzel’s and Greenson’s earlier work in further clarifying the distinction between the working alliance and the patient’s unconscious projections, and proposed that the working alliance is essential for therapeutic change regardless of the modality. In addition, his formulation emphasized the positive collaboration between the patient and the therapist. He identified three components of the alliance: tasks, bonds, and goals. In defining tasks, Bordin expanded on Menninger’s (1958) concept of a contract between patient and therapist. Bordin defined the tasks of therapy as the specific activities or behaviors that both the patient and therapist engage in. The specific tasks may differ according to the therapeutic modality. For example, in gestalt therapy the patient is asked 7 to attend to action, in psychoanalysis the patient is asked to fee associate. The specific tasks of the therapist may also differ depending on the modality. For example, relational therapy may ask for more self-disclosure, whereas humanistic therapy may require more empathic support. In order for the alliance to be strong, both parties must view the therapeutic tasks as efficacious. The goals of therapy are generally agreed upon desired outcomes. Again, the goals may differ depending on the therapeutic modality whereas cognitive-behavioral therapy may have a goal to reduce a specific behavioral symptom, and psychoanalysis may aim to resolve underlying conflicts. The bond includes elements of mutual trust and acceptance and confidence and reflects the overall quality of the patient-therapist relationship. Overall, Bordin’s broad conceptualization of the alliance offers a flexible definition that can be applied across modalities and across individual patient and therapist characteristics. Bordin acknowledged differences between therapy types and that each one varies in emphasis on tasks, goals, and bond. He suggested that the extent to which the therapist and patient agreed upon these emphases would determine the strength of the alliance. In addition, Bordin recognized the importance of interpersonal negotiation between the patient and the therapist for the formation of the alliance, and that each patient and therapist has varying styles and needs. In addition to Bordin’s conceptualization of the alliance bringing renewed attention to the topic, psychotherapy research also began suggesting that psychotherapy outcome does not significantly differ amongst various modalities (Luborsky, Singer & Luborsky, 1975; Stiles, Shapiro & Elliot, 1986). Researchers took this to mean that there may be underlying factors, such as the alliance, common to all forms of therapy that are 8 responsible for change (Horvath & Luborsky, 1993). Given the importance of the alliance for psychotherapeutic change, breaches in the alliance become a major concern for the efficacy of therapy (Safran & Muran, 20000). Binder and Strupp (1997) conducted a meta-analysis of negative process in psychotherapy and concluded that negative processes were unavoidable in forms of psychotherapy, and that positive outcomes rely on the therapist’s ability to effectively handle these alliance ruptures (In Ackerman & Hilsenroth, 2001). However, much of the research that addresses this issue focuses on only the patient’s personality, thus reflecting previous thinking that the quality of the alliance is ultimately an intra-psychic process, (rather than a transactional process) dependent on the patient’s personality. Several reasons have been given for this one-sided focus, for example, there are simply many more patients than therapists making it much more efficient to look at patient variables, and there are logistical difficulties with collecting confidential data from therapists (Najavits & Strupp, 1994). In addition to these more practical reasons though, Kiesler (1996) suggests that therapist’s are often not studied due to an assumption that therapists are all conducting “psychotherapy”, and are thus more or less interchangeable in research (Kiesler, 1996). In addition, Strupp (1982) suggests that many researchers may mistakenly believe that therapists are essentially superior to patients in their adaptive functioning. Nonetheless, there have been some seminal studies that have addressed how therapists’ characteristics impact the therapeutic alliance. In order to examine how therapists respond to in-session negative process, Strupp (1960) asked experienced therapists to respond to therapy situations that were initially 9 presented in written form, and then later in a film. Strupp found that therapists were immediately less capable of maintaining adequate levels of empathy when the patient was portrayed as hostile, angry or provocative. Strupp and Williams (1960) also found that therapists expressed more positive attitudes towards patients who appeared highly motivated for treatment. Strupp pointed out that these types of immediate reactions to patients could have serious consequences for the therapeutic relationship. Suh, O’Malley and Strupp (1986) assessed changes in the working alliance over the course of therapy. They found that the fate of the working alliance could be predicted in the first few sessions of therapy by therapists’ negative responses to their patients’ provocative behavior. Strupp (1980) conducted qualitative analyses of therapist-patient dyads where each therapist was assessed in one good outcome case, and one bad outcome case. He found that therapists displayed a preference for one patient over the other, and that poor outcome cases consisted of more hostile communications, a steady decline in patient involvement, a lack of therapist focus on important content, less warmth and friendliness displayed by the therapist, and overall interpersonal disengagement from both parties. These findings contributed to Strupps reconstructed view that therapists need to do more than provide an “average acceptable interpersonal climate” and that the formation of the working alliance is a collaborative interpersonal process, rather than an intra-psychic process primarily determined by the patient’s personality. Strupp stated: “…major deterrents to the foundation of a good working alliance are not only the patient’s characterological distortions and maladaptive defenses but-at least equally important-the therapist’s personal reactions.” Several studies expanded on these findings and assessed 10 interpersonal processes in psychotherapy, particularly how therapists often risk being pulled into patient’s unconscious and often hostile patterns as co-participants (Henry, Schacht and Strupp, 1986; Klee, Abeles & Muller, 1990; Kiesler and Watkins, 1989; Tasca and McMuller, 1992). Other studies have looked at therapists’ in-session characteristics and how they may contribute to the therapeutic alliance. Hartley and Strupp (1983) identified particular therapist behaviors that negatively impact the working alliance. For example, they found that therapists who foster dependency, and make irrelevant interpretations are more likely to have negative working alliances as reported by therapists. Marmar, Weiss and Gaston assessed therapists using the California Therapeutic Alliance Rating System and found that therapists who were rigid, self-focused and critical were perceived by patients as being less empathic, and as provoking more hostile resistance. Eaton, Abeles and Gutfreund (1993) coded therapist-patient dyads in insight oriented psychotherapy using the Therapeutic Alliance Rating System and the Vanderbilt Negative Indicator Scale. They found that therapists with poor working alliances across all stages of therapy (beginning, middle and end) tended to be exploitive, critical, moralistic, defensive, and lacking in warmth, respect and confidence. Additional studies have suggested that therapists who are not confident in their ability to help patients, tense or uncertain are less likely to form a positive working alliance (Sexton, Hembre & Kvarme, 1006; Eaton, Abeles & Gutfreund, 1993). Overall, psychotherapy research has made a firm argument for the importance of the working alliance for therapeutic change, and that difficulties in the alliance are 11 inevitable in the course of treatment, regardless of the theoretical approach (Binder and Strupp, 1997; Safran and Muran, 2000). The studies mentioned thus far primarily address therapist characteristics that impact the working alliance, however, there is also a body of research suggesting that therapists ability to handle breaches in the alliance (ruptures) once they have occurred has strong implications for the overall success of the therapy. For example, Castonguay, Goldfried, Wiser, Raue and Hayes (1996) assessed therapists’ use of cognitive strategies with depressed patient in relation to the working alliance in the first half of treatment. They found that therapists’ rigid adherence to cognitive treatment strategies inhibited the resolution of alliance ruptures, thus suggesting the importance of therapists need to be flexible and spontaneous. In addition, Safran and Muran (1996; 2000) have defined alliance ruptures as moments of conflict, tension or deteriorations in the working alliance, and have a proposed a therapeutic model for detecting and resolving such ruptures. Safran and Muran view ruptures as an expected part of therapy, and see them as a window of opportunity for core therapeutic change to take place. They suggest that ruptures occur when therapists unwittingly engage in patients’ maladaptive interpersonal cycles. During these cycles, patients pulls for behaviors in another in order to confirm expectations (even if these expectations are negative), and thus help avoid anxiety and abandonment in social interactions. From this perspective, moments of tension can be viewed as “windows into patient’s relational schemas”. They suggest that identifying and working through these ruptures with an acceptance of one’s (the therapist’s) own contribution to the impasse, and techniques 12 such as metacommunication by means of non-judgmental self-disclosure, mindfulness, and a focus on the here and now, may lead to changes in patient’s core beliefs. They state that while therapists are human, and therefore likely to respond to patients threats and attacks with hostility, they should stay mindful and aware of these difficult feelings, and be willing to “stick with the patient and to work toward understanding what is going on between them in the face of whatever difficult feelings emerge for both of them.” In this way patients are provided an opportunity to feel a real connection to another person as both parties explore painful feelings together, while at the same time learning how to negotiate their own needs and emotions in a more effective way. In sum, Safran and Muran’s model assumes a two person psychology where the therapist’s participation is equally central to maintaining an effective working alliance: A therapist who responds to a hostile patient with counter-hostility confirms the patient’s view of others as hostile and obstructs the development of a good therapeutic alliance. The therapist who responds to a withdrawn patient by distancing confirms the patient’s view of others as emotionally unavailable, thereby perpetuating a vicious cycle. (Safran and Muran, 2000, p.448) A Two Person Psychology: Me-You Representations Sullivan’s Personifications of Self and Other: An Introduction to Two-Person Psychology Harry Stack Sullivan, often considered the father of interpersonal psychology, broke away from the more traditional Freudian concept of a self-contained, intra-psychic mind. He opened the door for many subsequent theories of mind that acknowledge the 13 importance of social interactions. In Sullivan’s view, the individual cannot be understood outside of an interpersonal context, and interpersonal interactions shape people’s perceptions of themselves and others. A central concept of Sullivan’s concerns mental structures or personifications that are acquired through time and guide interpersonal perception. Sullivan (1953) postulated two different types of personifactions: those that apply to the perception of the self, and those that apply to the perception of others. As for self-personifications, Sullivan believed that people come to understand themselves in terms of characteristics that belong to a good me, bad me or not me. He hypothesized that the good me results from childhood experiences and characteristics that were rewarded and highly regarded by caretakers, while aspects of the bad me and the not me consists of characteristics that were not accepted or provoked negative responses from caretakers. The distinction between the bad me and the not me is the degree of anxiety that is associated with early social interactions. For example, psychological experiences and behaviors that produced moderate levels of anxiety for the child are still regarded by the individual as pertaining to the Self, however, they are negatively viewed and become associated with the bad self. However, early social interactions that evoked extreme levels of anxiety by significant others become associated with the not me remaining unacknowledged by the individual, and becoming personified at only a rudimentary level. According to Sullivan this anxiety results not only from punishment, but can be transmitted directly from the care-taker to the infant. For example, if a mother felt anxiety while breast feeding, the infant also comes to feel anxious and therefore develops a bad me or in extreme cases not me 14 personification. Conversely, if the mother feels secure, the infant internalizes feelings of safety and security which result in a good me personification. Sullivan (1956) expanded on his theory by stating that the good me (as it is associated with feelings of security) is directly related to self-esteem, while the bad me and the not me, as they are associated with anxiety, are inversely related to self-esteem. In an effort to maintain a feeling of security and protect ones esteem, the individual will engage in what Sullivan (1953) termed security operations that protect one from anxietyprovoking experiences. An important security operation is selective inattention, which consists of a process of selectively unattending to information with which one does not want to be associated with. Social cognition research has also lent support to Sullivan’s concept of personifications of others. For example, research in the field of social cognition has suggested that people actively seek social feedback that confirms their selfconceptualizations in preference to information that does not (Swann & Reade, 1981). In a similar vein, Sullivan suggested that in an effort to avoid anxiety and maintain one’s self-esteem, people actively elicit self-confirmatory information, and avoid nonconfirmatory feedback (feedback that does not confirm one’s personified good me). In addition to self-personifications, Sullivan’s theory concerns the way that people form expectations (or personifications) regarding others. Sullivan believed that humans organize information about people in terms of expectations that have formed from past relationships. These expectations provide a structure that guides and facilitates the processing of new social information. However, they may also impose rigid ways of interpreting new information. As a result, people may distort their perception of others. 15 For example, someone who had a rejecting mother may come to expect and therefore view all women as rejecting. Sullivan (1954) referred to this phenomenon as a parataxic distortion. Overall, Sullivan (1940) believed that self understanding is ultimately an interpersonal phenomenon, made up of reflected appraisals. Reflected appraisals consist of a feedback process that is based on how the care-taker communicates and responds to the infant’s needs, and results in self-worth contingencies (rules for how one should behave in order to feel loveable). As a consequence of these reflected appraisals, the infant develops a perceived understanding of what aspects of interpersonal behavior will maximize the possibility of human relatedness and self-esteem. More specifically, in an effort to remain securely related to others, Sullivan postulated that people acquire me-you patterns of representations that consists of the self and other interacting in a complementary fashion. For example, an aspect of an individuals’ good-me personification may consist of a belief that he or she must act submissive in social situations. At the same time, the individual may personify the “other” in a complementary dominant or controlling way. In this case the “me” is associated with submissiveness and the “other” is associated with dominance. These representations provide a sense of safety as they predict and guide future social interactions in a way that the individual believes will preserve interpersonal relationships. In summary, self/other personifications for Sullivan are not two different schemas, rather, these representations operate together in an interpersonal representation of self and other, where the self pulls for others to reciprocally respond. 16 In Sullivan’s formulation of a two-person, or relational psychology, the therapists’ personality is a central component of the therapeutic process. For example, Sullivan suggested that therapists’ me-you representations and patients’ me-you representations inevitably interact, and are continually guiding one another in subsequent interactions. Unlike a one-person psychology where the role of therapists is typically perceived as only observing patients’ behaviors, Sullivan coined the term participant-observer, which recognizes therapists as being simultaneous participants and observers in all therapy interactions, meaning, therapists engage in the patient’s interpersonal patterns, yet maintain an objective stance so as to monitor his or her own feelings and response tendencies. Internal Working Models Like Sullivan’s interpersonal theory, John Bowlby’s theory of attachment argues that humans are relational by nature, and the definition of the self forms through interactions with caretakers. In a similar fashion to Sullivan’s concept of personifications, Bowlby (1969) proposed the concept of internal working models (IWMs). IWM’s, like personifications are formed through early attachment experiences, and become rough drafts or cognitive-affective schemas of self-other experiences that play a central role in maintaining proximity with others by allowing the infant to predict self-other interactional contingencies. For example, if an infant experiences abandonment when expressing feelings of anger, the infant will learn to inhibit this feeling in order to maintain proximity with his or her caregiver (Safran, 1990). The 17 infant will then incorporate the message ‘expressing anger leads to being alone’ into an IWM and continue to inhibit this feeling in future relationships. Both Sullivan and Bowlby postulate that connection and proximity to others are important for survival and that any cue that poses a potential threat to this interpersonal relationship will result in anxiety for the individual. Sullivan suggested in his theorem of tenderness that this need for connection is shared by the infant’s complete dependent state and the mothers desire to provide such care. However, he was not exactly sure how much of this connection was based on the infants needs (Cortina, 2001). Bowlby goes one step further in explaining this need for connection in arguing that humans are genetically pre-disposed to a wired-in biological form of feedback that evokes anxiety in the face of separation, and provides a template to guide proximity seeking actions. Representations of Interactions that Have Been Internalized Stern (1985) articulates a related concept that he terms Representations of Interactions that have been Generalized (RIGs). He argues that infants develop prototypical representations of self-other interactions through a series of events with attachment figures. He suggests that overtime similar episodes are averaged in memory. In the process, details relevant to the individual events are abstracted and stored in prototypical mental structures. For example, an infant may average a series of episodes at the mother’s breast while abstracting critical details. Consequently, the infant forms a prototype of this type of interaction (Safran, Segal, Hill & Whiffen, 2001). 18 Stern argues that his concept of RIGs are similar to Bowlby’s concept of IWM’s, only more specific. An IWM can be viewed as an aggregate of RIGs that are averaged together to form a more generalized abstract representation of interpersonal interactions. For example, the prototype that has formed of the infant at the mother’s breast may be averaged with other types of interactions that occur with the mother to form a more generalized internal working model of self-mother relationship. Theoretically, and as supported by research (Bretherton, 1985) people would develop several different working models for different types of people in their lives (e.g., a prototype for mother, a prototype of lover). Continuing with this line of thinking, it can be reasoned that IWMs are themselves then averaged again forming an even more abstract interpersonal schema (e.g. a representation of self with controlling other or self with submissive other). Interpersonal Schema Theory Interpersonal schema theory (Safran, 1990) emerges out of Sullivan’s interpersonal psychology and is similar in nature to Bowlby’s (1969) internal working models, in that it forms through early interactions with caretakers, and serves as an adaptive mechanism allowing for the prediction of social patterns. Interpersonal schemas contain generalized representations of self-other interactions, as opposed to isolated representations of self and other, and are therefore relational in nature. They can be thought of as programs for maintaining relatedness that include beliefs in the format that if I behave in x way, the other person will do y, for example, “if I become angry, others will become distant”. Like Bowlby’s attachment theory, interpersonal schema theory 19 suggests that people have a biologically wired-in propensity for feeling connected to others. Through experience the schema forms more precise goals, action plans and ifthen contingencies (e.g. if I act controlling then the other will act submissive) relevant to maintaining this relatedness. Thus, the interpersonal schema can be thought of as a program for maintaining relatedness in which the primary goal (to maintain closeness) is biologically wired in, and the principles employed to attain this goal are learned through early interpersonal relationships. Although these interpersonal schemas are initially formed in the context of attachment relationships, they become internalized and are later applied to more general interpersonal situations and continue to play a central role in shaping thoughts, feelings, and behaviors in social interactions (Main et al., 1985). More specifically, interpersonal schemas are an elaboration of Bowlby’s IWMs that are represented at different levels of abstraction and function in a hierarchical fashion. The lowest or most specific representations of the interpersonal schema (which would be similar in nature to RIGs) contain images and episodic memories of interpersonal situations (such as episodes at the mothers breast), and expressive motor and autonomic responses evoked by those events. Higher level representations, on the other hand, contain rules, strategies and beliefs (concerning the self and others) that have been abstracted from early experiences and function to maintain relatedness. For example, a young child who was ignored when he or she made mistakes, may come to believe that he or she needs to “be perfect at everything I do” in order to maintain close relationships with others. Similar to RIGs, people are thought to develop various interpersonal schemas for different attachment figures. For example, an interpersonal 20 schema with the mother is formed through interactions with the mother and therefore may contain different beliefs and strategies than an interpersonal schema relating to the father. Moreover, interpersonal schema theory like attachment theory suggests that early interpersonal experiences with care-takers have an emotional component providing survival relevant information of the self in interaction with the environment. These emotions are rooted in expressive motor-responses and remain in the form of unarticulated meaning (as opposed to symbolic meaning). Thus, interpersonal schemas can be thought of as cognitive-affective schemas that are encoded and represented at least partially in the form of implicit, procedural knowledge that is experienced at an unarticulated bodily felt level (Leventhal, 1984). When new interpersonal encounters occur, cognitive and affective components of an existing schema are triggered, experienced and appraised, and any new information is assimilated into the already existing structure. Thus, these cognitive-affective schemas are continuously elaborated upon in ways to maximize continuing relatedness. Under normal circumstances, interpersonal schemas are formed through mutual reciprocity between the child and caretaker, where the child’s emotional expressions pull for complementary efforts in the mother to attune to and satisfy her child’s needs (Bowlby, 1969; Tronick, 1989). However, in a maladaptive early environment, this healthy reciprocal pattern may be disrupted and the child learns that to be interpersonally engaged and to ensure social connection means to be neglected, controlled, or somehow mistreated. Thus, a maladaptive interpersonal schema may form that continues into 21 adulthood, where for instance the individual may expect to be controlled by others, and act in an overly submissive manner that confirms this neglect. According to Safran & Muran, (1996) pathogenic interpersonal patterns (i.e., schemas) may play themselves out in the context of therapy resulting in moments of tension or ruptures in the therapeutic relationship. During these strong interpersonal processes, therapists’ own expectations may become a major contributing force, and ultimately determine the outcome of the rupture. As mentioned above, Safran & Muran (2000) suggest that working through ruptures with patients may play an important role in changing maladaptive interpersonal expectations. Thus, the therapists’ own object relations become an important mediator for change, and therefore an important area of research. The Interpersonal Circumplex: A Model for Psychotherapy Research Within the Framework of a Two-Person Psychology Influenced from Sullivan’s theory regarding interpersonal behavior, particularly the notion that people engage in repeated patterns of behavior with others, coupled with an interest in defining and treating psychopathology within the context of maladaptive interpersonal cycles, interpersonalists began conceptualizing personality as falling within a circular dimension. The work of the Kaiser Group and later Timothy Leary’s (1957) book Interpersonal Diagnosis of Personality brought the concept of an interpersonal circumplex to fruition. The interpersonal circumplex was originally outlined by Leary (1957) in an effort to capture his view of personality which consists of predictable and reciprocal patterns in relationship functioning. Expanding on Sullivan’s theory that interpersonal behaviors exert a pull on others to respond in a complementary fashion, 22 Leary proposed that interpersonal acts could best be conceptualized as falling somewhere on a circle. For example, Leary’s interpersonal circumplex plots individual interpersonal acts by Cartesian coordinates within a two dimensional space, where control equals the yaxis reflecting a continuum of dominant to submissive behaviors and affiliation equals the x-axis reflecting a continuum of hostile to friendly behaviors. Thus, social behaviors can be plotted on the interpersonal circumplex according to their overall degree of affiliation and control. In addition, traits are scattered around the circumplex in a complementary fashion so that the control axis pulls for ‘opposite responses’ and the affliation axis pulls for ‘similar responses’ (e.g., hostility begets hostility, control begets submissiveness). In sum, Leary’s circumplex model allows for measurement and prediction of interpersonal functioning according to what he called the “principle of interpersonal relations”. As the interpersonal circumplex systematically defines interpersonal behaviors, it lends itself well to research based on interpersonal theory (Wiggens, 1982). Consequently, the interpersonal circumplex model has been revised in several different ways amongst various researchers, and studies have burgeoned that use various versions of the circumplex to assess personality, psychopathology and psychotherapy process and outcome (Benjamin, 1974; Coady & Marziali, 1994; Henry, Schact & Strupp, 1986; Najavits & Strupp, 1994). Structural Analysis of Social Behavior: The Intrex Benjamin’s Structural Analysis of Social Behavior (SASB; Benjamin, 1974) is one example of an influential development of the interpersonal circumplex model. The 23 SASB model differs from the traditional circumplex in that it measures social behavior on three dimensions or surfaces, as opposed to just one. The first two planes represent interpersonal processes, whereas the first plane is organized around a focus on the ‘other’ and describes prototypically transitive behaviors directed by the ‘other’ person, and the second plane is organized around a focus on the Self and describes reactions to the ‘other’s’ behaviors. For example, if the ‘other’ person endorses freedom, then the Self that is interacting with that other will feel autonomous and behave in ways that reflect “freely coming and going”. The third plane represents actions turned inwards and reflects Sullivan’s principle of introjects, or self-representations resulting from interactions with others. For example, if the other person behaves in an attacking manner, the self that is interacting with that other person will develop an introject which is self-attacking in nature. Conversely, if the other person behaves lovingly, the self that is interacting with that other person will have an introject that is actively self-loving. Unlike the first two planes, this plane reflects beliefs about the Self and as Benjamin (1974) states, “…is important for connecting self-concept to the social milieu”. Each of the three dimensions are built upon an x and y axis and are broken into 36 interpersonal behaviors reflecting degrees of affiliation and autonomy. The 36 behaviors are typically broken down into octants, with each octant consisting of 4-5 items. While Benjamin’s x-axis is similar to previous circumplex models in that it represents a continuum of affliation (Love to Hate), Benjamin re-structured the control (vertical) axis so that control and submit are on the same side of the pole (at the bottom) and are opposite to emancipate and separate respectively. In making this change, Benjamin created an axis that represents 24 interdependence, or interpersonal control with varying degrees of independence and differentiation represented at the top half of the circumplex, and levels of interpersonal enmeshment at the bottom half. The SASB model employs both a coding system (Benjamin and Cushing, 2000) for measuring overt behaviors, as well as a self-report Intrex Questionnaire. The coding method can be used for all three surfaces (Self, Other, Introject), however it is typically used in research to describe individuals’ behaviors along the first two interpersonal dimensions, and consists of having trained coders assign two numbers to each thought unit (a speech act containing one idea). The first number represents the surface (either Self or Other) of the thought unit, and the second number represents which octant of the circumplex the thought unit falls in. The Intrex Questionnaire has several versions (a short, medium and long form) pertaining to the three axis. Generally, the self-report intrex is used for assessing individuals introjects (the third plane), and for the assessment of representations of others behaviors, such as mother, father and significant other (focus on other). However, the intrex can be arranged in wording so that items represent the Self in relation to others (focus on Self) as well. Each version requires the person to rate the subject of focus at best and at worst. Kiesler’s Interpersonal Transaction Cycles and Circumplex Model Kiesler (1979, 1983), another key contributor to interpersonal theory, developed a version of the interpersonal circumplex based on his earlier work concerning communication in psychopathology and in psychotherapy, or more specifically, the 25 process of “ongoing dyadic interpersonal transactions”. He described the communicative process between two people as consisting of two principle constructs: One being an “evoking message” -a term coined by Beier (1966)- that is delivered from the “encoder” (the person sending the message), and the second being what Kiesler himself called an “impact message” (the message that is received and decoded by the other person in the transaction). He described the “evoking message” as consisting of overt verbal and nonverbal behaviors emitted by the “encoder”. The impact message, on the other hand, is thought to consist of the covert cognitive and affective reactions that are unconsciously experienced by the person on the receiving end. Kiesler argued that impact messages leads to reciprocal overt responses from the receiving person that confirm the encoders’ initial expectations of social interactions. In a similar vein to the interpersonal schema, he hypothesized that this interpersonal transaction cycle is mostly an automatic process consisting of a cognitive and affective component that remains out of awareness unless attended too. Kiesler argued that feelings of self-worth, anxiety and depression result when people, outside their level of awareness, send unaccountable impact messages that elicit aversive responses from others and therefore engages others in maladaptive communicative processes. For example, a person who is depressed may send out an impact message that he is disinterested. The person receiving the impact message may respond with similar dislike, leading to an unhappy interaction confirming the depressed person’s initial pessimistic expectations. In terms of psychotherapy, Kiesler recognized that therapists are not immune to patients’ evoking messages, and emphasized the importance of therapists’ abilities to 26 attend to their own emotional and cognitive reactions. He suggested that therapists can challenge maladaptive interpersonal transaction cycles by maintaining close awareness of covert and overt reactions elicited by the patient’s verbal and non-verbal behaviors. For example, a therapist may notice feeling patronized by a patient. Upon this recognition, the therapist may link this feeling to particular expressive-motor mannerisms on the patient’s part, such as arrogant tilts of the head and a condescending tone of voice (Safran, 1998). Once identified, these markers provide an entry way into exploring and ultimately understanding maladaptive interpersonal schemas. Kiesler used the interpersonal circumplex as a base to operationalize his communication theory. He formulated his own 1982 version in an effort to revise the interpersonal circumplex in a way that would capture the overt behaviors involved in his theory of interpersonal transactions. Previous versions of the circumplex consisted, for the most part, of ambiguous or static descriptions of personality styles. Conversely, Kiesler used transitive verbs and unambiguous adjectives that describe overt behaviors as much as possible in order to clarify the boundaries of each circumplex segment, and provide accurate behavioral descriptions (descriptions that are more interactional in nature than previous circumplex descriptions). Thus, he formulated an action-based circumplex conducive to evaluating interpersonal behavioral transactions in relationships, as opposed to evaluating interpersonal styles outside the relational context. In addition, Kiesler’s 1982 version of the circumplex more fully integrates complementarity theory (Carson, 1969; Foa, 1961; Leary, 1957; Wiggins, 1982). A primary postulate of interpersonal theory is that behaviors pull for particular behavioral 27 responses from others that confirm expectations and therefore perpetuate one’s selfdefinition (Sullivan 1953; Leary, 1957; Benjamin, 1974). According to complementarity theory, behaviors are mapped onto the interpersonal circle according to either the complementarity law of correspondence, or the complementarity law of reciprocity. Correspondence is represented on the affiliation (horizontal) axis and behaviors pull for like behaviors, whereas hostility begets hostility and friendliness begets friendliness. Reciprocity is reflected on the control (vertical) axis and behaviors pull for opposite behaviors, whereas dominance begets submissiveness and vice versa. Previous theorists, while recognizing complementarity, failed to explicitly and/or thoroughly systematize its inherent rules within the context of a circular model. To this end, Kiesler advanced the notion of complementarity by organizing each behavior around the circumplex in a bipolar fashion. For example, his circumplex model consists of eight axis and opposite behaviors residing at each end of the pole, equaling a total of 16 behavioral situations varying in degrees of control and affilitiation. Kiesler thoroughly articulated semantic opposites for each interpersonal transaction style. In doing so, he provided a foundation for systematically placing behaviors around the circumplex so that all coordinates corresponding to each other on the horizontal plane pull for corresponding (“like”) reactions, and all corresponding to one another on the vertical plane pull for reciprocal (“opposite”) reactions. Thus, one can more precisely map out, for example, how a person who is controlling and hostile will pull a submissive and hostile response from the other person. In summary, Kiesler advanced the circumplex model into an even more relational realm, operationalized the interactive processes taking place in interpersonal cycles, and 28 provided a base for systematically predicting behaviors according to theoretically coherent laws. Assessing Therapists’ Me-You Represenstations Interpersonal Schema Questionnaire In order to assess interpersonal schemas, Hill & Safran (1994) developed the interpersonal schema questionnaire (ISQ). The ISQ is designed to assess expectations about how three important people (mother, father and significant other/close friend) will respond in various interpersonal situations. Kieslers’ version was chosen as a basis for the development of the ISQ because it samples the full range of interpersonal behaviors in a rigorous and theoretically guided fashion (Hill & Safran, 1994). His version was also chosen because of his use of transitive verbs and unambiguous adjective in describing overt behavior. In other words, Kiesler’s action-based circumplex model was conducive to evaluating behavioral transactions in relationships, as opposed to interpersonal styles outside of a relational context. This use of overt behavioral descriptions made it easier for Hill and Safran to develop a self-report measure in which individuals could imagine themselves in actual interpersonal interactions (Hill & Safran, 1994). (Refer to Appendix A for examples of Kiesler’s circumplex model). The ISQ consists of 16 interpersonal situations, which are anchored along Kiesler’s interpersonal circumplex model. For each situation, respondents are asked to imagine themselves in the interaction in a way that matches the behavior description from the 29 interpersonal Circle. For example, a controlling situation would entail the respondent imagining himself behaving in a controlling way towards his mother or father. The respondent would then be asked to chose from 8 different kinds of behavioral responses that they might expect from mother or father. Each of these 8 responses contains varying degrees of control and affiliation. Through the act of imagining these situations and their own behaviors with these three different people, memories and expectations that are associated with the specific close relationship are elicited. Consequently, response ratings should represent the generalized rules, goals, contingencies or strategies that comprise the respondent’s interpersonal schema for each of these types of relationships. The 8 responses that they chose from are based on Kiesler’s octant version of the circumplex, which is formed by collapsing adjacent segments of the 16 scenario version. The reason for doing this was simply to make the response process more manageable for those filling out the questionnaire. For each situation, respondents are also asked to rate the desirability of the expected response from other on a scale of 1-7. 16 Situations: Controlling, self assured, spontaneous, outgoing, compliant, warm, trusting, respectful, docile, dependent, inhibited, aloof, hostile, cold, mistrusting, competitive 8 Responses: assured/ dominant (PA); competive/mistrusting (BC); cold/hostile (HD); detached/inhibited (FG); unassured/submissive (HS); deferenet/trusting (JK); warm/friendly (FS); sociable/exhibitionistic (NO) Example: Hostile situation= Imagine yourself feeling very angry and argumentative towards ____. Respondent now chooses how the important other would respond in this situation from the examples above, then the respondent rates the desirability of that response. 30 The ISQ yields three key indices scores for mother, father and significant other: affiliation and control (with mean scores ranging from –1 to +1) and desirability of the expected outcomes (ranging from 1-7). On the affiliation axis (x-axis) +1 equals friendly, -1 equals hostility and 0 equals neutrality. On the control axis (y-axis) +1 equals dominant, -1 equals submissive and 0 is equal to neutrality. For example, an anticipated response such as, “Would be warm and friendly” is almost purely affiliative in nature and therefore would receive a score of +1 on the affiliation (x) axis. However, this same response is rather neutral with respect to beliefs regarding control, and would therefore receive a score of 0 on the control (y) axis. However, subjects anticipated responses often consist of some combination of affiliation and control. For example, the response, “Would respect me, or trust me” is considered partially submissive and partially friendly, and would be scored -.5 for control and +.5 for affiliation. The individual responses on the ISQ as well as the desirability scale can be combined in many different ways to yield different types of scores. One can look at someone’s total level of expected control and affiliation across all 16 situations, for mother, father, and significant other separately, or combined. Similarly, it is possible look at someone’s overall desirability of expected outcomes across all situations. However, as some situations may be more predictive of someone’s interpersonal schema, it is worthwhile to break down the 16 situations into different subgroups- or subscales. For example, subscales are often broken down into the four quadrants of the circumplex: *The quadrants are made up of the following situations: Hostile: hostile, detached, mistrusting, inhibited, cold 31 Friendly: friendly, sociable, trusting, exhibitionist, warm Submissive: submissive, deferent, inhibited, trusting, unassured Dominant: dominant, competitive, mistrusting, exhibitionist, assured In this way, researchers can examine interpersonal expectations in the various kinds of situations that reflect the various behavioral descriptions on the circumplex. Reliability and Validity of the ISQ Hill and Safran (1994) conducted a study with 344 university students in order to assess the reliability and validity of the ISQ. Through correlating subjects’ control and affiliation scores Hill and Safran confirmed that these two indices were orthogonal, or uncorrelated (r=.07). This supports the validity of the measure as these two axes are inherently distinct and unrelated properties. For example, the level of affiliation that one expects from others is un-related to the level of control that that same person anticipates from others. Hill and Safran assessed the internal consistency of the measure through calculating Cronbach’s alpha. High internal consistency was found for both the affiliation and desirability indices (αs = .81, .90 respectively). However, the control dimension did not demonstrate high internal consistency (αs = .62). Consistent with these findings, Hill and Safran (1994), found that test-retest correlation coefficients for 34 students (of the original 344) who re-rated the ISQ after a 4-week period were high for affiliation and desirability, but not for control (rs = .88, .87, .44, respectively). These findings suggest that interpersonal schemas in normal populations are relatively defined for affiliative interactions, but not for situations pertaining to control. Hill and Safran (1994) point out that the lack of consistency for the control index may reflect a tendency 32 for psychologically healthy individuals to be flexible in their anticipation of these types of interactions. For example, individuals with solid interpersonal skills may form expectations that are dependent on the given situation, rather than employ one interpersonal strategy across all of these kinds of situations. Given this reasoning, the control index may prove to be more meaningful (and consistent) with a more depressed sample. Hill and Safran’s (1994) results of overall good validity along with poor reliability on the control index were replicated in two Turkish studies. In a cross cultural validation study of the ISQ in Turkey, Boyacioglu and Savasir (1995) assessed the content validity of the measure through factor analysis and found consistent and strong factorial pattern amongst the represented interpersonal situations. In addition, 93 university students independently judged and gave high ratings of suitability for each dimension of the scale in congruence with its situational category. Boyacioglu and Savasir found good test retest reliability with correlations ranging from .66 to .88 on the desirability and situational sub-scales (e.g., friendly, hostile, dominant, submissive). However, Soygut and Savasir (2001) conducted a follow-up reliability study of 378 Turkish students and found lower internal consistency across all of the situational subscales (friendly, hostile, dominant, submissive), particularly the control (dominant) subscale (αs = .62, .61, .40, .63 respectively) while the desirability index continued to demonstrate high consistency for all situations (α = .90). Relevant Studies applied to Psychotherapy Process 33 The study of the impact of therapists’ defensive processes, or interpersonal schemas on the therapeutic alliance as well as in-session therapeutic processes is relatively rare. However, there has been some seminal work regarding this topic. Henry, Schact and Strupp (1990) examined therapists’ internalized self-representations (introjects) on interpersonal process and differential psychotherapy outcome. Before treatment both therapists’ and patients’ introjects were assessed by Lorna Benjamin Smiths (1988) Intrex questionnaire, which assesses self representations within an interpersonal context. Henry et al then separated good and bad outcome cases based on changes in patients’ self representations as evidenced by the Intrex. They found that (i) poor outcome cases were related to interpersonal behaviors by therapists that confirmed negative patient introjects; (ii) the number of therapists’ statements that were hostile or controlling correlated with the number of self blaming statements by the patients, and (iii) therapists’ with hostile introjects were more likely to engage in interpersonal processes that have been associated with poor outcome. Henry & Strupp (1994) later concluded that these findings may suggest a link between therapists’ early relationships with their parents, the formation of the therapists’ adult introjects, and the consequent “vulnerability to counter-therapeutic techniques.” A previous study by Nelson (2000) also provides preliminary research on this topic. In addition, she looks specifically at therapists’ interpersonal schemas and how they play themselves out in the context of alliance ruptures. Nelson evaluated the interpersonal expectations of therapists using the ISQ. She then examined the relationship between the ISQ ratings with (i) SASB ratings of in-session behavior during 34 rupture segments and (ii) treatment outcome. Nelson found that therapists who expected to elicit hostile responses from their fathers, as measured by the ISQ, also tended to engage in hostile behaviors with their patients, as rated by the SASB. Interestingly, this hostility did not predict negative outcome. This may have to do with the fact that sessions were rated based on rupture segments, and may indicate a relationship between hostility and techniques employed when working through ruptures. Therapists’ expectations of affiliation with their mothers as well as with significant others in their lives did not predict their behavior with their patients. In a more recent study, Fox-Borisoff (unpublished MA thesis) also found that the mother and father may contribute distinct characteristics to the formation of interpersonal schemas. For example, Fox-Borisoff assessed therapists’ interpersonal schemas using the desirability index on the ISQ. Desirability for each therapist was broken into the 4 quadrants or subscales as described above (hostile, friendly, submissive and dominant) in order to see if desirability of particular kinds of situations predicts in-session processes as well as therapists’ ability to form a positive working alliance. Desirability for these quadrants was assessed for both mother and father independently. Little discrepancy was found between the quadrants themselves- desirability in affiliative situations were not particularly more or less predictive of working alliance or the resolution of ruptures than controlling situations. This was true for both patient and therapist perspectives. However, there was evidence suggesting a difference between the types of interpersonal behaviors that may arise from the relationship with the mother verses the father. In particular, the findings suggest that therapists who have positive interpersonal schemas 35 with their fathers (i.e. they expect desirable outcomes, regardless of the situations, when interacting with their fathers’), are more likely to resolve ruptures. However, a positive interpersonal schema with ones mother seems to predict therapists’ ability to establish a positive working alliance. In summary, the quality of one’s relationship with their father seems to predict interpersonal techniques necessary for working through conflict, whereas the quality of one’s relationship with their mother seems to predict interpersonal techniques necessary for establishing a positive working alliance. The Role of the Father and Interpersonal Schemas Both Nelson and Fox-Borisoff’s studies suggest an important role for the father in interpersonal schema formation. Previous research regarding the relationship with the therapist and his or her father and how that affects the therapeutic process in particular is limited. However, Mallinckrodt, Coble & Gant (1995) conducted a relevant study examining the role of early attachment experiences on adults’ internal representations of social competency, and how these beliefs in turn, affect patients abilities for establishing therapeutic working alliances. They used the adult attachment scale to assess parental bonds (i.e., expected degree of parental warmth and emotional responsiveness versus expected degree of parental control), and the self-efficacy scale to assess social competency. They looked at both of these scales in relation to the working alliance in 76 adult females and found that parental bonds with fathers exclusively predicted social competency. They also found that parental bonds with fathers negatively predicted working alliance (patients with the poorest working alliances were those most likely to 36 characterize their fathers as over-involved, controlling, and emotionally over-bearing), whereas strong maternal bonds positively predicted working alliance. Interestingly, this negative correlation between fathers and working alliance was stronger than the positive correlation they found between maternal bonds and working alliance. Similar in nature to Fox-Borisoff’s findings, these finding suggest that fathers may be particularly important in the formation of social skills, and that there may be a relationship between paternal attachment style and therapeutic processes. There are a few other studies that have touched upon the relationship between the father and the formation of interpersonal schemas in general. For example, Hill & Safran (1994), in their validation study of the ISQ (1994) found that university students expected more hostile responses from their fathers and more friendly responses from their mothers. In another study, Boyacioglu and Savasir (1995) categorized subjects into low, medium and high depressive symptomatology groups and than assessed their interpersonal expectations of the mother and father by using the ISQ. They found that individuals in the ‘high’ group expected more hostile responses from their fathers in hostile situations than subjects in the low and medium symptomatology groups. This finding suggests a relationship between fathers’ hostility in conflict situations and depression in children. However, the nature of this relationship is unclear. For example, while it may be that hostile fathers cause depression, depressed individuals may simply be more likely to perceive non-threatening behaviors as hostile. Regardless of the nature of this relationship, the lack of findings with the mother variable against the relatively strong 37 findings with the father variable, suggest that fathers play a particularly important role on the formation of interpersonal schemas related to depressive symptoms. Chapter II Child Development, the Role of the Father, and Conflict Resolution Strategies Research regarding child development may point towards a difference in the roles that mothers verses fathers play in the formation of interpersonal schemas. In particular, and consistent with Fox-Borisoff’s findings stated above (that a therapist’s relationship with his or her father predicts that persons ability to resolve ruptures, whereas a therapist’s relationship with his or her mother predicts that personal ability to establish a working alliance) the child development literature reveals an interesting trend in that fathers seem to strongly affect a child’s ability for social behaviors, such as displays of confidence in negotiating interpersonal situations, including conflict, whereas mothers seem to be more relevant to children’s abilities to form trusting, warm and supportive relationships. In addition, the child development literature supports this hypothesis across a variety of theoretical frameworks in that it includes attachment theory, social information processing theory, and social learning theory. Child-Father Attachment Style and Children’s Social Competence According to attachment theory, infants develop internal cognitive-affective representations of interactions with caregivers, or internal working models (IWMs). 38 IWMs play a central role in maintaining proximity with others by allowing the infant to predict self-other interactional contingencies, or rough drafts of self-other experiences that help guide and plan future social interactions (Bowlby, 1969). For example, if an infant experiences abandonment when expressing feelings of anger, the infant will learn to inhibit this feeling in order to maintain proximity with his or her caregiver (Safran, 1990). The infant will then incorporate the message ‘expressing anger leads to being alone’ into an IWM and continue to inhibit this feeling in future relationships. Overall, Bowlby’s (1969) research suggests that people develop a predominant attachment style through these early formative experiences that determine the quality of future significant relationships and feelings of self-worth. For example, a secure working model of a person’s attachment relationship with the mother is thought to have positive expectations regarding the mother’s availability and responsiveness, along with a subsequent sense of being loveable and high self-regard. These individuals will come to expect positive responses from others, and may be more likely to reach out and form healthy relationships. Conversely, a child with an insecure attachment style is assumed to have a working model of the parent as rejecting or unresponsive, resulting in feelings of low self-worth or a belief that he or she is unlovable. These children may come to anticipate rejection and rely on either aggression or withdrawal (Ainsworth et al., 1978). Child development researchers have often considered attachment theory to be one of the most influential theories in explaining the relationship between social and emotional development and parental style (e.g., Hamilton & Howes; Sroufe, 1983). In the past, child development researchers and attachment theorists (Bowlby, 1969, Ainsworth, 1973) 39 Main, Kaplan & Cassidy, 1985; Cohn, 1999) have typically focused on the mother as the primary caregiver, and how this relationship shapes emotional and behavioral components of future relationships. However, findings suggesting that children develop a distinct attachment style to fathers at 18 months of age (Schaffer and Emerson, 1964; Lamb, 1997), has led to an increased interest in the independent influences of father-child attachment on healthy development. In addition, past research has often considered the effects of father-child attachment to be contingent upon mother-child attachment quality, or to merely create an “averaging effect” with mother-child attachment. However, more recent findings have suggested that invalid methodology may have contributed to past limited findings regarding the importance of the father. Thus, there has been a reconsideration of the role of father-child attachment. For example, previous research has typically relied on the same methodological tools for assessing father-child attachment that is used for mother-child attachment research. Research has suggested that fatherchild attachment has different consequences on development than that of mother-child attachment, thus, assessing these two different attachment relationships in the same way may be miss-leading. For example, the "strange situation” is often used to assess motherchild attachment, and measures a child’s response in a new environment under a mildly distressing situation (the mother leaves her child alone). Children who are securely attached will curiously explore the environment and happily welcome the mother back when she returns. This secure attachment is thought to be the result of adequate tender loving care of mothers (Grossman et al., 2002). Conversely, Grossman et al (2005) found that measures that assess sensitivity in terms of emotional support and gentle challenges 40 in play situations better capture father-child attachment quality. The idea that play time offers a better medium for assessing children’s attachments to their fathers is consistent with previous research suggesting that fathers widen children’s experience and help them adapt to new situations beyond the family environment (In Grossman et al., 2002: Hewlett, 1992; Harkness and Super, 1992). For example, through their active play and advice giving, fathers are thought to provide their children with a sense of safety in new and challenging experiences (Murphy, 1977; Parke, 1995), and as Lamb (1975) suggested, linking their children to the outside world. In sum, a secure attachment to the father may reflect a father’s ability to sensitively challenge and support the slightly older toddler’s exploration of the social environment and new experiences, as opposed to reflecting the kind of tender loving care that a mother expresses and that contributes to feelings of safety upon separation in infancy. Overall, attachment research is currently suggesting that the quality of fatherchild attachment has its own unique value in child development (Verschueren & Marcoen, 1999). Several of these studies use alternatives to the “strange situation” for assessing father-child attachment (however, it is unclear if this was done because of the question of predictive validity with the “strange situation” in terms of father-attachment, or simply because of the practicality of these other measures). A pattern is beginning to emerge in these studies suggesting that father-child attachment relationships are more relevant to children’s socioemotional development including social-efficacy and conflict resolution skills in particular, while mother-child attachment relationships are more relevant to issues of inward emotional experiences, warmth, trust and empathy (the same 41 characteristics necessary for a strong working alliance). The following studies address these differences: Verschueren and Marcoen (1999), argued against the idea that father-child attachment merely offers an “averaging” affect (Bretherton, 1991) in predicting selfesteem and social competence. They hypothesized that working models for mothers and fathers may remain partially unintegrated and influence different aspects and relationships in children’s lives. They conducted a study consisting of forty girls and forty boys between fifty-five and seventy-seven months of age to assess the differential and combined predictive power of children’s internal working models for mother and father. Like Main & Westin (1981) they found that children with secure working models of the father and the mother (as measured by the attachment story completion task based on Bretherton et al., 1990; and Cassidy, 1988) fared better than those children who only had a secure attachment to the mother. However, they also found that the relative predictive power of mother-child and father-child attachments differed depending on the domain of child functioning that was assessed. For example, they found that the motherchild attachment style was predictive of self-representation including feelings of selfworth (as measured by the puppet interview developed by Cassidy, (1988) and Verschuruen et al., (1996), whereas father-child attachment style was predictive of peer social competence and behavioral displays of self-confidence {as measured by kindergarten teachers’ ratings on the pre-school social and behavior questionnaire (Tremblay, Vitaro, Gagnon, Piche & Royer, 1992) and a popularity and peer acceptance six point scale developed for the study}. In particular, father-child attachment style was a 42 uniquely powerful predictor of children’s anxious and avoidant behaviors. For example, children with secure attachments to their fathers demonstrated less anxious and withdrawn behaviors, better social adjustment and more social competence with peers than children with insecure (ambivalent, hostile/bizarre) attachments to their fathers. Overall, Verschueren and Marcoen (1999) demonstrated that internal working models of fathers uniquely influence outward displays of social behavior including avoidant and anxious behaviors, whereas internal working models of mothers uniquely influence inward feelings of self-representation, including feelings of self-esteem, perceived competence and likelihood of peer acceptance. Steele and Steele (2005) found similar results to Verschueren and Marcoen (1999) with regards to the unique contributions of mother-child and father-child attachment style, where the former is relevant to the understanding of inward feelings, and the latter to the negotiation of outward displays of emotion. Steele and Steele (2005) assessed the attachment styles of expectant mothers and fathers arising from their own childhoods (through conducting the Adult Attachment Interview (AAI) (Main, Goldwyn & Hesse, 2003)), and then related these findings to the quality of parent-child attachment of such parents’ newborn children at infancy. They assessed the infant’s attachment style to the mother at one year, and the infant’s attachment style to the father at eighteen months. For both of these assessments they used the “strange situation.” They then conducted longitudinal studies of eighty-six children to track and assess emotional development at five, six and eleven years of age. When the children were five, they looked at their abilities to understand emotional conflict in the nuclear family by presenting eleven 43 different story stems from the MacArthur Story Stem Battery (Emde, Wolf, & Oppenheimer, 2003) to the children. The story stems used dolls to depict domestic emotional conflict, as well as story beginnings requiring the child to “show me and tell me what happens next”. When the children were six years old, Steele and Steele assessed their abilities to understand others’ emotional reactions of others. They administered a series of cartoons depicting social interactions involving emotional conflict. The children were asked to report how the characters were feeling, and were encouraged to provide narratives that explained their reasoning. At eleven years of age, Steele & Steele administered the “friends and family interview” (Steele & Steele, 2004) to further assess emotional and social development. This interview asked questions regarding feelings that arise within family and peer relationships. With regards to mother-child attachment, Steele and Steele’s results suggest that both the mothers’ own attachment style measured prior to giving birth, and infant-mother attachment quality at one year of age are independently important for healthy emotional development. For example, they found that mothers’ AAI’s during pregnancy were predictive of their infants’ attachment styles. More specifically, expectant mothers who spoke coherently about their own childhood experiences (whether good or bad) had children who were securely attached to them at one year of age. They also found that mother’s with coherently organized (autonomous-secure) AAIs from before giving birth had five year old children who were more likely to reference the limit-setting statements of an authoritative parent when exposed to emotionally conflictual scenarios. These limit-setting statements are viewed as positive methods for containing overwhelming 44 emotions in children, and thus their internalization it thought to be part of healthy emotional development (Steele & Steele, 2005). With regards to their follow-up assessments at six years of age, Steele and Steele found that both attachment to mothers at one year of age, and mothers’ prenatal AAI’s were predictors of children’s abilities to understand mixed emotions, and the emotional reactions of others. Taken together, Steele and Steele interpret these findings to suggest that maternal attachment style prior to giving birth, and mother-child attachment at infancy are important determinates for a child’s ability to regulate and understand their own emotions, and other’s emotions. This early ability to understand their own, and others’ mixed emotional states may be the bedrock for what is later referred to as empathy. At the eleven year follow-up, Steele and Steele found that “truthfulness” was the most reliable aspect between expectant mothers’ AAI responses and children’s responses on the friends and family interview. For example, mothers with secure AAI responses from years ago were more likely to have children who provided honest accounts of their views of themselves and others. Truthfulness seems to require a certain degree of personal acceptance and emotional awareness. Thus, it is not surprising that mothers who are secure in their own right, would somehow influence their children’s sense of self-acceptance and honesty. Interestingly, Steele and Steele also found that fathers’ secure AAIs were predictive of their sons’ truthfulness on the friends and family interview, however, not their daughters’ (mothers predicted both genders). Steele and Steele turn to Chodorow (1978) to explain this discrepancy. For example, they suggest that boys require the immediate involvement of both the mother and the father for 45 forming a coherent and secure self-identity. However, daughters can continue to turn to the mother as a source for forming their own self-understanding and therefore do not involve the father to any significant level. Despite the inconsistent effects found with regards to father attachment style prior to giving birth, and father-child attachment style at eighteen months, the eleven year follow-up yielded some strong and unique findings with regards to fathers’ own attachment styles and their children’s socioemotional development (they still failed to find any significant effects for father-child attachment at eighteen months, which may be due to the fact that they used the “strange situation” to assess the child’s attachment to the father, which as mentioned above may not be a valid measure for father-child attachment style). Steele and Steele (2005) found that father’s AAI responses before their children were born were uniquely predictive of both their sons and daughters ability to resolve and negotiate social disagreements. For example, fathers with secure-autonomous AAI’s from years before had eleven year old children who were more likely to provide coherent and resourceful accounts of how to resolve social conflicts (as presented by the Friends and Family measure). Steele and Steele suggest that these fathers may have played an important role in helping their children handle emotions that come up in social relationships, beyond the sphere of the primary mother-child relationship. Steele and Steele (2005) also found that fathers who provided insecure AAI responses many years before were more likely to have children with behavioral problems such as conduct disorder, hyperactivity and difficulties in peer relations (as reported by the children). Overall, Steele and Steele interpret their findings to suggest that mothers may uniquely 46 contribute to children’s understanding and resolution of emotional conflict within themselves (which may lead to such qualities as empathic understanding and truthful accounts of their feelings), whereas fathers may uniquely contribute to children’s ability to maintain socially appropriate behaviors with others, and to understand and resolve interpersonal emotional conflict. Coleman (2003) examined the relationship between children’s social selfefficacy, peer victimization, and quality of peer relationships with the level of secure attachment to their mothers and fathers during middle childhood. In addition, she examined social self-efficacy as a moderator linking parental attachment quality to peer attachment quality. Interestingly, Coleman does not consider internal working models as a mediator between parental attachment and future attachment relationships. For example, she highlights a definitional congruence between internal working models, and social self-efficacy beliefs, which are defined by Bandura (1997) as consisting of (i) knowledge of appropriate social behaviors, (ii) confidence in one’s ability to respond competently in interpersonal situations and (iii) beliefs that the responses from others in one’s social environment will be supportive of one’s effort to engage in the interaction. In essence, she argues that these two constructs are comparable in that they both involve relatively stable, yet potentially modifiable internal representations, that develop from past relationships with caretakers, and that likely influence affective and behavioral responses to social interactions. In making this connection, Coleman provides a theoretical grounding for comparing findings from social learning theory and attachment 47 theory, and provides a basis for her own work in examining social self-efficacy within the context of attachment theory. Coleman examined sixty-seven fifth and sixth graders (thirty-one girls and thirtysix boys) at a small public school in the Southeastern US. She used the Inventory of Parent and Peer Attachment (IPPA), developed by Armsten and Greenberg (1987) to assess children’s attachments to their mothers, fathers and peers. More specifically, this measure asks questions pertaining to children’s self-perceptions of mutual trust, communication quality, acceptance, and the degree of anger and alienation within the context of their peer relationships. In order to assess interpersonal functioning, Coleman administered the social self-efficacy subscale of the Children’s Self-Efficacy Scale (CSES) developed by Bandura et al (1996). An example item from this subscale includes “How well can you make and keep friends?” Coleman’s findings suggest that only mother-child attachment predicts the quality of peer attachment, whereas only father-child attachment predicts social self-efficacy. For example, she found a significant relationship between children with secure attachments to their fathers and reports of high levels of social self-efficacy with peers. In contrast, attachment to the mother was not predictive of children’s social self-efficacy, however, it was significantly related to the quality (e.g., warmth, trust) of children’s attachments with their peers. She also found evidence suggesting that mothers and fathers effects on children exist across genders. For example, she did not find any interactions between child gender and parent-child attachment relevant to peerattachment, or social self-efficacy. Coleman interprets her findings to suggest that 48 attachment related behaviors to the mothers and fathers are relevant to children’s likelihood of having successful relationships however, they are related in different ways. She states that mother-child attachment relationships may be uniquely relevant to children’s ability to show sensitivity and concern for others, and to be able to display emotions that merit trust and commitment in close relationships, whereas father-child attachment relationships may be uniquely relevant to children’s feelings of worth and competency in social situations. While Colman’s study does not look at conflict resolution per se, it is easy to imagine that elements of social self-efficacy are required for such behaviors. In addition, many child development researchers argue that conflict resolution is the most encompassing standard for measuring social competency (McDowell, Parke and Spitzer, 2002). For example, if one does not hold the belief that they are effective in the interpersonal domain, than it is most likely the case that that person would also believe they are ineffective in threatening social situations, such as conflict. Kern and Barth (1995) found that a secure attachment to the mother was related to boys’ and girls’ abilities to engage in play, whereas secure attachment to the father was related to particular ways of handling friendships once they were established. For example, children with secure attachments to their fathers displayed friendly cooperative behavior including using more directives and making more suggestions and positive responses. Once again, this study suggests a distinction between the contributory roles of the mother and the father, where the mother is involved with factors pertaining to a 49 child’s ability to establish a relationship, and a father is involved in factors that may pertain to conflict resolution and social negotiation. Lieberman, Doyle and Markiewicz (1999) examined the effects of middle childhood attachment security (including dependency and parental availability) to both mother and father on concurrent peer relations with 274 elementary school students and 267 high school aged students. They administered the friendship quality questionnaire (Bukowski et al., 1994) to assess the degree of security, closeness, companionship, help and conflict between peers, and the Kerns Security Scale (Kerns et al., 1996) to assess the degree of perceived parental availability and dependency (e.g., the degree to which children believe they can turn to their parents in stressful situations). They found that attachment to the father on the availability dimension in particular was the strongest predictor of peer conflict. The father’s availability seemed to be a critical factor in middle-childhood conflict with peers, whereas the more available the father (as perceived by the child) the less conflict that occurred between the child and his or her peers (again as perceived by the child). Lieberman, Doyle and Markiewicz explain this relationship by suggesting that available fathers may engage in the kinds of interactions (such as playful activities) that provide opportunities for teaching their children conflict resolution strategies. Lieberman, Doyle and Markiewicz also found that child-mother and childfather secure attachment predicted better quality peer relationships, including higher reports of closeness, help and security. However, they found that the relationship of child-mother dependency to child positive friendship qualities uniquely varied as a function of maternal availability. For example, children who perceived their mothers to 50 be available when needed, and have developed a healthy degree of autonomy (i.e., they are not dependent on their mothers) reported greater closeness, support and security in their friendships. This may suggest that mothers are particularly relevant to their children’s experience of an individual-self, which is relevant in establishing secure, trusting relationships. Overall, these findings suggest that mother verses father attachment security may play distinct roles in the formation of children’s social competence whereas (i) the father’s availability seems to be more involved with factors regarding conflict and conflict resolution and (ii) mothers seem to be relevant to children’s experience of an autonomous, individual-self capable of forming healthy adolescent-peer relationship. Father-Child Conflict and Conflict Resolution with Peers The above studies suggest that secure paternal attachment is an important predictor of a child’s peer relationships, and that secure attachment to the father, a father’s own secure attachment style (as seen in Steele and Steele, 2005) along with father availability may be particularly important in the development of conflict resolution strategies. In terms of the nature of the pathways that explain this relationship, the above literature points generally to internal working models and to a lesser extent social self-efficacy beliefs (i.e., Coleman) as possible mediating factors that may account for the transfer of social behaviors from parent to child. However, attachment literature in general focuses more on how the quality of the father-child attachment relates to child social competence, as opposed to what behaviors and attitudes transpire during actual interactions between the father and the child (Parke et al., 2002). Several studies within the child development 51 literature focus specifically on conflict resolution situations between the father and the child, and more precisely suggest that the strategies that fathers employ in these particular interactions may be most influential in the transfer of information, and formation of cognitive structures that guide conflict resolution and social competence skills in children. For example, McDowell, Park & Spitzer (2002) assessed mediating components such as goals and strategies that account for the transfer of behaviors from familial contexts to peer situations. More specifically, they looked at the relationship between parental goals and strategies with children’s goals and strategies relevant to social competence and conflict resolution. They administered an open ended vignette of a conflict situation to ninety-seven kindergarten children and their mothers and fathers. The vignettes were then coded for goals and strategies. Goals (e.g., the desired outcome for a social encounter) and strategies (e.g., the actions taken to achieve the goal) were defined in terms of social information processing theory (Crick and Dodge, 1994) and formed into the following two composites: (i) Relational-Prosocial which was comprised of prosocial and relational strategies thought to minimize conflict and (ii) instrumental –confrontational which was comprised of avoidant, confrontational, and instrumental components that are self-serving and facilitate “getting one’s way.” In order to assess children’s social competence, teachers were asked to fill out a twelve item children’s competence questionnaire developed by Cassidy and Asher (1992), and children were asked a variety of questions regarding their peers’ social behaviors (i.e., friendliness, exclusiveness vs. inclusiveness with others, aggressiveness, willingness to 52 help and share). Their findings suggested that kindergartners’ pro-social/instrumental goals and strategies in conflict situations, closely resembled the interpersonal goals and strategies that father’s employ in conflict situations. Likewise, they also found that fathers’ goals and strategies were better predictors then mothers’ goals and strategies for children’ social competence. According to McDowell, Parke and Spitzer fathers’ social representations may be important correlates of children’s own social information processing. However, McDowell, Parke and Spitzer point out that the pathways through which fathers’ cognitive representations effect children’s social competence remains unclear. For example, they conducted a mediational analysis to assess the extent to which children’s goals and strategies (cognitive representations) that are acquired through interactions with their fathers’ mediate such children’s social competence. Contrary to their predictions, they found only minimal evidence suggesting that children’s strategies may mediate this relationship, and no evidence that children’s goals are a mediating variable. They interpret this lack of findings as a possible indication that children’s social competence skills are either non-mediated, or mediated from father to child through non-cognitive pathways such as emotional-regulatory mechanisms, reinforcement, teaching, or modeling. McDowell, Parke and Spitzer (2002) explain the salience of the father in predicting children’s social competence by suggesting that perhaps mothers are more relevant to social domains related to pro-social issues such as making friends, whereas fathers are more influential in conflict laden situations. In addition, they highlight that this difference between mothers and fathers is consistent with previous research that has 53 suggested that mothers may be more relevant to children’s internalizing behaviors, whereas fathers may be more relevant to children’s externalizing behaviors (Cowan, Cowan, Schulz and Heming, 1994). Another study by McDowell, Parke and Wang (2003) assessed the relationship between parental advice giving in conflict situations and middle childhood social competence. The authors arranged triadic advice giving sessions where mothers, fathers, teachers and their third grader discussed how to handle various problems that their child had when socializing with peers. Fathers’ approaches to advice giving in these sessions were the strongest concurrent and one year later predictor of child social competence. Interestingly parental style, which McDowell, Parke and Wang define as the degree of control and warmth expressed by the parent during the advice giving interaction, appeared to be a stronger predictor of children’s social competency than the actual feasibility of the solution provided by the parent. In particular, fathers’ controlling styles in conflict situations predicted negative social competency ratings by both teachers and peers. Thus, father’s behaviors in conflict situations are particularly salient predictors of children’s social behavior. McDowell, Parke and Wang explain these findings by suggesting that perhaps mothers are more likely than fathers to provide advice in everyday settings, thus when a father provides advice in particularly problematic situations it is that much more salient to the child. McDowell, Parke and Wang (2003) found evidence suggesting that a controlling and cold approach to advice giving may also result in feelings of depression and loneliness in children. More specifically, a controlling style by both the mother and the 54 father predicted depressive symptoms and loneliness. However, only a mother’s lack of warmth predicted childhood loneliness. Assuming loneliness results from difficulties forming close friendships, this latter finding suggests that a mother’s warmth is an important predictor of a child’s likelihood of establishing close bonds with peers. Again, this supports the notion that mothers are particularly relevant to children’s abilities to establish close relationships. There is also evidence that father-child interactions are more predictive of child social competence and aggression than mother-child interactions (Parke et al 1989). For example, Pakaslahti et al (1996) examined parents’ social problem solving strategies with their sons (daughters were not included in the study) and found that only the father’s (and not the mother’s) resolution strategies were related to aggressiveness in their sons. For example, parents who were passive and cold, diverted authority or were punitive towards their sons in dealing with problematic situations had sons who were more aggressive. However, a mother’s attempt to be active in problem solving with their sons did not help their children’s pro-social behavior with peers. Only a father’s active involvement, including strategies such as helping the boy to come up with conflict resolution plans, engaging in discussion, providing advice, and in particular encouraging their sons to consider problems from multiple perspectives helped to enhance their sons’ social capabilities (McDowell et al., 2002; Parke et al., 1989; Pakaslahti et al 1996). Consistent with social information processing theory, they suggest that parents transmit their own problem solving strategies to their children, which in turn, affects elements of children’s behaviors. 55 Reese, Weber and Marchand (2001) examined conflict resolution strategies between mother - adolescent and father – adolescent dyads with late adolescents’ romantic relationships. They administered the Managing Affect and Differences Scale developed by Arellano and Markman (1995) and independently assessed male and female late adolescents for each of these two dyads. Reese, Weber and Marchand predicted that adolescents would identify with and therefore engage in resolution strategies most similar to their same sex parent. They supported this hypothesis through social learning theory, which predicts that children are more likely to imitate the behaviors’ of role models with whom they share important characteristics such as gender (Bandura, 1989). However, their findings did not completely coincide with this prediction. Reese-Weber and Marchand found that positive and negative conflict resolution strategies with the father predicted both male and female adolescents’ resolution strategies with romantic partners (albeit the degree of this relationship was weaker for females); whereas strategies employed in mother-adolescent dyads were only related to female resolution strategies with romantic partners. For example, females who reported more positive conflict resolution behaviors with their mother and their father were more likely to report positive resolution strategies with their romantic partners, whereas females who experienced more negative resolutions behaviors with both the mother and the father, were more likely to experience negative conflict resolution behaviors with romantic partners. For males, only reports of positive and negative conflict resolution behaviors with the father (and not the mother) were related to their reports of conflict resolution with partners. In sum, interactions with the father had predictive value across genders, whereas the predictive 56 value of conflict situations with the mother was limited to females. These finding suggests that the role of the father in conflict resolution strategies may involve more than imitation of same sex parents. These findings may be interpreted in light of research suggesting that conflict with fathers occurs less frequently than conflict mothers (Allison, B, 2000, Furman & Buhrmester, 1992, Laursen, 1995, Laursen and Collins, 1994, Montenmayer, 1982). Thus, when conflict with the father occurs it may be emotionally arousing to daughters as well as sons and therefore more memorable, and influential for both genders. Reese Weber and Marchand’s slightly stronger findings with father-son conflict may indicate some combined effect of imitation of same sex parent along with the powerful nature of father’s involvement in conflict situations. Ricaud-Droisy and Zaouche-Gaudron (2003) conducted a study in Toulouse, France and found a positive relationship between fathers who are highly involved and play a distinct role (from the mother’s) in raising their children with those children’s abilities to use productive conflict resolution skills with peers. Ricaud-Droisy and Zaouche-Gaudron described these nursery school aged children as working towards mutual understanding through seeking dialogue and verbal negotiation in conflictual interactions. They also suggest that these children seem to be in a position to explain to peers their own perspectives as well as be able to listen to others’ points of view. In contrast, Ricaud-Droisy and Zaouche-Gaudron also found that children with fathers who regard their role as less implicated and less differentiated from that of the mother’s tend to engage in more aggressive conflict with their peers. In a third condition, they found that children with father’s who do regard their role as having implications on the child’s 57 social development, however, as not being distinct from the mother’s role are likely to abandon conflict and seek refuge with an adult. Overall these findings suggest that it is important for the father to not only be involved, but contribute a unique quality in raising the child. Ricaud-Droisy and Zaouche-Gaudron performed an ascending hierarchical method and found that dealings with authority, respect of rules, and firmness in times of conflict (e.g., when the child requires structure and rule-enforcement) seem to be some of the primary characteristics of this distinct role. Finally, Daniel T. L. Shek and Hing Keung Ma (2001) conducted a longitudinal cross cultural study amongst Chinese female and male adolescents and found consistent results to those stated above. They administered the Conflict Behavior Questionnaire (Robin & Foster, 1989) to assess parent-adolescent conflict, and the Adolescent Behavior Questionnaire (Ma et al., 1996) to assess adolescent social behavior at various intervals to students in grades seven through eleven. Their results suggested that over time fatheradolescent conflict uniquely predicted female and male adolescent pro-social behaviors, and was overall a better predictor of adolescent anti- social behaviors. These findings suggest that fathers continue to have an influence on children’s conflict resolution strategies well into late teenage years. Affect-Regulation and Conflict Resolution: The Role of the Father Both the attachment literature and social information processing literature have suggested that affect regulation may play a mediating role between father characteristics and child social competency. Several studies within the attachment literature suggest that 58 parental emotional availability and emotional sensitivity mediates the formation of internal working models which in turn influences children’s abilities to regulate their own emotions leading to feelings of self worth and quality of interpersonal functioning (Wei, Tsun-Yao Ku and Zakalik, 2005). For example, caretakers who are inconsistent in their emotional availability and are insufficient in handling their children’s distressing emotions, result in children developing an insecure attachment style linked to emotional distress, nervousness and negative affect, interpersonal difficulties including hostility towards others, and increased feelings of loneliness. In contrast, caretakers who are emotionally available and sufficient in handling their children’s affect have children who develop secure attachments linked to positive self-regard and interpersonal successes (Wei et al., 2005). In addition, Bowlby (1982) proposed that intense emotional expression is part of the process by which internal working models form. Steele & Steele (2005) go on to hint at the importance for healthy affect regulation strategies with regards to fathers when they hypothesize that a father’s ability to handle his child’s emotions during social interactions predicts the quality of the father-child attachment style. Several researchers within the social information processing domain have acknowledged the relationship between children’s social competence and affect regulation (Barth & Parke, 1995, Carson & Parke, 1996). Researchers suggest that a child’s ability to meet the demands of a social situation requires adequate response and reading of other’s emotions, as well as the ability to regulate one’s own emotions (Eisenberg et al., 1993; Roberts and Strayer, 1987). Parents are thought to transmit emotion regulation skills to children through children’s modeling of emotion regulation, 59 parental control verses acceptance of their children’s emotions, or through parental coaching in distressing and conflict situations (Eisenberg et al, 1998; Gottman, Katz & Hooven, 1997; Isley et al., 1999; McDowell & Parke, 2000; McDowell & Parke, 2005; Roberts, 1999). With regards to the father, Parke (2000) directly suggests that affect regulation may mediate the relationship between fathers’ goals and strategies in conflict situations and children’s social competence. Although Parke (2000) does not offer an explanation for this, perhaps it is because conflict situations occur so infrequently with fathers that when these emotionally-laden situations do occur, they become a medium for transmitting meaningful emotional experiences to the child, which then influence the child’s affect regulatory abilities, which in turn, influences the child’s social competence. In addition, Parke (2000) and Roberts (1999) highlight that (i) the quality of emotions displayed by fathers and (ii) how fathers respond to their children’s distressing emotions are important predictors of children’s social competence and conflict resolution skills. These findings are consistent with general theories of psychology which often recognize the importance of parental emotional availability and emotional sensitivity for healthy child development. However, many of these theories primarily address the influence of the mother’s affective response strategies and emotional availability on child development. For example, Tronick (1998) emphasizes the importance of a mother’s ability to adjust or “attune” to her child’s emotions, through a series of affective misscoordination and repairs. Tronick (1998) suggests that the process of oscillating back and forth between states of affective miss-coordination and repair helps the infant develop a 60 belief that the caretaker is available even in moments of disconnect, resulting in a sense of safety and ability to withstand difficult moments. In addition, Stern (1985) argues that the process of affect attunement plays a central role in a person’s ability to communicate his or her emotional experience. He argues that through the mother’s affective attunement children come to experience their emotions at a bodily felt level, and because they experience emotions at a bodily felt level, as opposed to being cut off from experience, they are able to appropriately express their emotions and communicate with others what they are feeling (Safran, 1990). Shore (2001, 2003) additionally posits that a mother’s adequate emotional stimulation (that is not over-stimulating or understimulating) facilitates the maturation of the orbitofrontal system, which plays an essential role in the regulation of emotion. Self-psychologists such as Kohut (1971, 1977) suggest that a mother’s ability to empathically respond to her child’s emotions is necessary for the child’s development of Self. Kohut suggests that the Self acquires psychological structure through the process of a mother’s attunement and a mother’s empathic failures, where the gap between these two is sufficiently small and the infant can take over meeting his or her own needs. Overall, these theories suggest that the mother’s emotional availability and sensitivity is relevant to her child’s development of a cohesive Self, and her child’s ability to handle his or her own inner emotional experiences in the absence of the mother. In contrast, recent studies are confirming that a father’s emotional availability and sensitivity may become particularly relevant to child development after infancy, when social interactions become a central experience to the child. Congruent with the findings 61 from the attachment literature and the social-information processing literature, current research is suggesting that a father’s ability to handle his own emotions and his child’s emotions may be particularly relevant to the child’s ability to handle outward displays of emotion, conflict resolution skills, and social competence: Roberts (1999) asked mothers and fathers to complete the Child Rearing Practices Q-Sort (Block, 1995) and found that fathers who reported using problem resolving techniques in response to children’s distressing emotions during conflicts, had children who used more resourceful and flexible behaviors in social interactions (as measured by pre-school teachers’ ratings on the Ego Strength Questionnaire developed by Waters, Wippman and Sroufe, 1979). It is possible that paternal problem solving techniques provide the child with a tool for containing overwhelming emotions through providing structure and concrete guidance. In addition, Parke & O’Neil (1997) found that father’s who reported responding to their children’s expressions of negative affect with distress, were more likely to have children who used anger and other negative emotions to deal with distressing events. Conversely, Fathers who reported using emotion- and problemfocused reactions to negative emotions had children who were more likely to use reasoning strategies for dealing with distressing situations. These children were described as less aggressive and more cooperative by teachers. Similarly, O’neil and Parke (2000) found that fathers who use anger and criticism in social interactions had children who had more difficulty controlling negative emotions as well a children who resorted to avoidance of negative emotions all together. Both the avoidant coping strategy as well as the difficulties controlling emotions resulted in social conflict behaviors. 62 Gottman, Katz and Hooven (1997) also found that fathers’ assistance, acceptance and ability to withstand their children’s anger and sadness at age three were related to positive peer relationships at age six. Gottman, Katz and Hooven also examined the affect of mothers’ responses to children’s negative emotions. Overall, mothers’ responses were less predictive of their children’s future social behavior. Similarly, Roberts (1994) found that fathers’ who are accepting and comforting of their children’s negative emotions have children with more positive peer relationships. McDowell and Parke (2005) found that fathers’ positive affect and control of emotional expression were related to children’s behavior in the disappointing situations. Finally, Carson and Parke (1996) found that pre-school children whose fathers engaged in reciprocal exchanges of negative feelings were rated by their teachers as engaging aggressive and low in sharing. In addition, they found that the more anger fathers expressed in verbal exchanges during an emotionally arousing game, the more likely their children would be rejected by their peers. Summary of Father vs. Mother Influences in Child Development Fathers seem to play a critical role in children’s (both girls and boys) development of social competence and conflict resolution skills. In addition, there is some evidence that these findings are cross cultural (Shek & Ma, 2001). Overall, mothers’ influences are seen early on in infancy, and continue to shape children’s feelings of security, warmth and trust in the world. These feeling are most likely important for establishing a solid sense of self, lasting relationships, and being able to empathically connect to others. Although fathers may be important at infancy, their 63 influence on children’s development becomes apparent around pre-school age, where social competence and conflict resolution become central to children’s experience. Father characteristics such as, attachment style, ability to form an attachment with his child, how he handles his own emotions and his children’s emotions, and quality of conflict resolution strategies in conflict situations all seem to influence the child’s social competence and conflict resolution skills. Social competence, in turn fosters positive self-regard and self-esteem. Thus, both the mother and father seem to affect feelings of self-worth, however, they may affect different components of this construct (albeit that are constantly interacting with one another and highly interconnected), whereas the mother influences children’s innermost feelings of love and security, and the father influences children’s feelings of self-efficacy and competency in social environments. The former is required to establish a relationship in the first place (to make and desire connection), whereas the second is required for maintaining the relationship by resolving interpersonal struggles as they occur. Chapter III Statement of Purpose and Hypotheses Statement of Purpose Psychotherapy researchers often acknowledge that therapists’ personalities play an important role in determining the quality of the therapeutic alliance, and that tension in the alliance is unavoidable across all therapeutic modalities. Researchers also recognize the importance of the alliance for therapeutic change to occur, and that ruptures in the 64 alliance may provide windows into patients’ maladaptive interpersonal patterns thereby providing an opportunity for important therapeutic work. However, as Safran and Muran (2000) suggest, therapists’ own interpersonal schemas may play themselves out in the context of strong interpersonal cycles with patients. For example, therapists expecting hostility and ultimately abandonment in the face of conflict may avoid confronting and working through ruptures with their patients. The current study expands on FoxBorisoff’s pilot study and further examines how the therapist’s personality impacts psychotherapy process, particularly the formation of the working alliance and their ability to work through breaches (ruptures) in the alliance. In addition to expanding upon the psychotherapy process research literature, this study utilizes findings from child development research as a base for understanding how therapist’ past attachment experiences may influence their current abilities to form positive working relationships and effectively resolve conflict with their patients. For example, the child development literature suggests that early experiences with the mother may influence a child’s ability to form trusting, and empathic relationships (such as are required characteristics to form a working alliance), whereas childhood experiences with the father may be more relevant to conflict resolution skills such as feelings of confidence in social interactions, and ability to compromise, collaborate and accept responsibility for one’s own behaviors in relationships (such as are required characteristics to resolve ruptures with patients). Thus, in addition to contributing to psychotherapy process research, this study highlights the emerging role of the father as an important and unique predictor of personality formation. An additional aim of this study uses the intrex questionnaire to assess the 65 effects of therapists’ introjects on psychotherapy process, and is to an extent meant to provide a point of comparison concerning the usefulness of the ISQ as a measure of psychotherapy process. Finally, an independent study is included as an annex that takes advantage of the large quantity of ISQ data available by examining its psychometric properties. This separate study is not included in the primary hypothesis, or analyses sections and is intended to inform future uses of the ISQ. Hypotheses Given the findings from Fox-Borisoff’s study along with the child development literature the primary hypothesis of this study is that: Therapists’ interpersonal schemas of their mothers will contribute to those therapists’ abilities to establish a working alliance as perceived by both the patient and the therapist, whereas therapists’ interpersonal schemas of their fathers will contribute to those therapists likelihood of confronting and resolving ruptures in the alliance when they occur, as perceived by both the patient and the therapist. The second hypothesis addresses the relationship between therapists’ interpersonal schemas with the quality of rupture resolution: Therapists’ interpersonal schemas of their fathers will be a more valuable predictor than that of mothers with regards to the degree of smoothness and depth of experience perceived by patients and therapists during rupture sessions. 66 The third hypothesis attempts to shed a deeper understanding on how therapists’ personalities affect psychotherapeutic process, and how the interpersonal schema questionnaire compares to an alternative schema construct (introjects) in its contribution to psychotherapy process variables: Therapists’ introjects of themselves in worst case scenarios will predict those therapists’ abilities to establish a working alliance and resolve ruptures with patients. CHAPTER IV Method Overview This study is based on relational theory of psychoanalysis, including Safran and Muran’s concept of the working alliance as consisting of an on-going negotiation between both the therapist and the patient. The study addresses how therapists’ internal processes contribute to psychotherapy process, and how their relationship with their mothers and fathers may independently contribute to this relationship. Therapists’ internal processes are assessed through the interpersonal schema construct, and the introject construct. Interpersonal schemas with their mothers and fathers are assessed in terms of the desirability index of the ISQ which assesses the degree to which therapists’ expect desirable outcomes with their mothers and fathers in various kinds of social situations, Therapist introjects are measured by the Intrex Questionnaire according to 67 how therapists view themselves in their worst case scenarios. Psychotherapy process is measured according to therapist and patient ratings of the working alliance and various aspects of rupture resolution across the first five sessions of therapy. Therapist-Patient dyads are selected based on the completion of therapists’ ISQ and Intrex Questionnaires. The first five Post-Session Questionnaires (PSQs), which contains a 12-item version of the Working Alliance Inventory (WAI) and sections regarding the occurrence, degree and quality of rupture resolution are collected for each patient and each therapist. The raw scores for the WAI are averaged to attain a total working alliance score, single-item data is collected reporting the occurrence of ruptures, and the degree to which ruptures are resolved as perceived by both the therapist and the patient, and data from a 10-item questionnaire is averaged into two separate scores reflecting the degree of smoothness and depth of experience perceived by patients and therapists during rupture sessions. The pilot study consists of only those therapists who had seen three or more patients. Three patients were randomly selected if the therapist had seen more than this required number. Correlations were conducted to examine the relationship between therapists’ interpersonal schemas of their mothers and fathers in hostile, friendly, controlling and submissive situations with those therapists’ in-session working alliance ratings and ruptures resolution scores as perceived by both the therapist and the patient (note: the smoothness and depth of experience questionnaire was not used in this study). Findings suggested that therapists’ interpersonal schemas of their mothers (they anticipated desirable responses from their mothers across social situations) predicted therapists’ likelihood of establishing a working alliance, whereas therapists’ interpersonal 68 schemas of their fathers (they anticipated desirable responses from their fathers across social situations) predicted their likelihood of resolving ruptures with patients. In the current study, the sample is expanded to include 96 therapists and all the patients they have seen at BPRP. In addition, the 10-item smoothness and depth of experience questionnaire is added as a dependent variable in order to provide a more meaningful and comprehensive measurement of rupture resolution. More advanced and sensitive statistics are also applied, and finally, the therapists’ introjects (as measured by the Intrex Questionnaire) are examined as an independent variable affecting the therapists’ likelihood of establishing a working alliance, and resolving ruptures with patients. Brief Psychotherapy Research Program The data from this study is collected through the Brief Psychotherapy Research Program (BPRP) at Beth Israel Medical Center in New York City, co-directed by Chris Muran and Jeremy Safran. The research program began in the early 1980s to study psychotherapy process and outcome, with a particular emphasis on alliance ruptures and repair. The BPRP provides 30 sessions of low-fee psychotherapy to individuals in the community. Patients are primarily recruited through advertisements in local news papers. Patients are first screened by trained research assistants for suitability to the project over the phone, and must be between the ages of 18 and 65, have been out of previous psychotherapy for at least three months, demonstrate cognitive coherence, and be able to report one close relationship. Patients must report no indication of suicidal or homicidal ideation, impulse disorder or history of mania. If patients meet these criteria they are 69 brought in for two meetings where a trained clinical research assistant conducts the Structured Clinical Interview for the DSM-IV (SCID) (Spitzer, Williams, & Gibbon, 1987, 1994). Based on this interview, the patient is diagnosed along Axis I and Axis II of the DSM-IV. Patients are generally accepted into the program if they demonstrate suitability for time-limited therapy. Many of the patients included in the study meet criteria for Axis I anxiety and mood disorders, and Axis II cluster C personality disorders such as avoidant, dependent and obsessive compulsive personality disorder. Patients are excluded from the program if they meet diagnostic criteria for the following Axis I disorders: (i) active or recent substance use, or dependence (ii) Bipolar I or II disorder (iv) any psychotic spectrum disorder, (vi) active or recent eating disorder such as anorexia nervosa and bulimia nervosa (vii) Post Traumatic Stress Disorder, or any Axis II Cluster A diagnoses of Paranoid, Schizoid and Schizotypl Personality Disorders, and Cluster B diagnoses of Antisocial, Narcissistic and Borderline Personality Disorders. In addition, patients are excluded from the study if they meet a DSM-IV-R Axis III medical condition, neurological impairment or significant intellectual deficit, or if they are in need of crisis intervention, have had previous suicidal attempts or psychiatric hospitalizations, and are currently prescribed psychotropic medications requiring further stabilization, or that have not been stabilized for at least three months. Patients participating in the program have at some point during the history of the program been randomly assigned to one of the following manualized treatment modalities: Brief Relational Therapy (Safran & Muran, 2000), Cognitive Behavioral Therapy (Turner & Muran, 1992), an experimental conditioning combining Cognitive 70 Behavioral Therapy with Brief Relational Therapy, Brief Adaptive Therapy or Short-term Psychodynamic Psychotherapy (Winston & Pollack, 1991). Over the years therapists have consisted of psychology interns and externs enrolled in a Clinical Psychology Doctoral program, third and fourth year psychiatry residents, and licensed staff supervising psychologists at Beth Israel Medical Center. All therapists receive at least one hour of group supervision per week in their assigned treatment modality and see only one patient through the program at a time. At intake each therapist is required to fill out a confidential self-report packet that includes the interpersonal schema questionnaire and the intrex questionnaire. Therapists and Patients are required to independently fill out a post-session questionnaire at the end of each session that includes a 12-item working alliance inventory, a section regarding the occurrence and resolution of ruptures, and a 10-item questionnaire reflecting the degree of smoothness and depth of experience perceived by patients and therapists during rupture sessions. All self-report measures are returned to the treatment team and kept strictly confidential. Participants Therapists All therapists included in the study were participants in the Brief Psychotherapy Research Program at Beth Israel Medical Center, and completed 30 sessions of psychotherapy. Each therapist completed the interpersonal schema questionnaire and intrex questionnaire once in the beginning of treatment. Therapists that have successfully filled out post-session questionnaires after the first five sessions of therapy, and the ISQ 71 and Intrex at intake are selected for the study. There are a total of 96 therapists that meet this criteria. The pilot study includes 24 (10 men and 14 women) therapists who saw three or more patients, and ranged in age from 25-50 with a mean age of 38. The current sample consists of the original 24 therapists plus 72 additional therapists who have filled out the appropriate questionnaires. This totals 96 (37 men and 59 women) therapists with an age range of 25-67 with a mean age of 37. Slight variations in age and gender ratios occur between multi-level models due to missing data and cases excluded from analysis due to absence of ruptures. Patients All patients included in the study were participants at the Brief Psychotherapy Research Program at Beth Israel Medical Center, and completed 30 sessions of psychotherapy. Patient selection was based on the availability of their therapists’ data (patients were included if their therapist filled out an ISQ and Intrex Questionnaire in the beginning of treatment). In addition, all patients as well as their therapists’ had to have successfully completed the first five post session questionnaires to be included in the study. The pilot study totals 72 patients (three patients for each of the 24 therapists) including 28 men and 43 women (with one gender report missing) ranging in age from 24 to 64 with a mean age of 42, while the current study includes all patients that were seen by all 96 therapists totaling 186 patients (114 women and 70 men with two missing gender values) ranging in age from 23-64 and with a mean age of 40 (with two age values 72 missing). Slight variations in age and gender ratios occur between multi-level models due to missing data and cases excluded from analysis due to absence of ruptures. Measures of Therapist Internal Process (Independent Variables) Interpersonal Schema Questionnaire (ISQ) Therapists’ interpersonal schemas of their mothers and fathers are assessed independently using the ISQ, and in terms of their expected responses and the expected degree of desirability of those responses from each parent across 16 different situations reflecting the sixteen segments of Kiesler’s interpersonal circumplex. For example, therapists are asked to imagine themselves engaging in an activity or situation such as acting very competitive to win a tennis match, collaborating on a project and taking the lead on making decisions, or as being reserved and non-sociable with each of their parents. They are then asked to select a response reflecting the expected behavior from the given parent (e.g. I expect my mother would be friendly and warm; I expect my mother would take charge and try to influence me). Each response that therapists’ can choose from represents one of the 8 octants of Kiesler’s interpersonal circumplex and has varying degrees of interpersonal affiliation and control, thus yielding two scores: response-control and response-affiliation. After selecting a response, therapists are asked to rate the desirability of the expected response in the given situation on scale from 1-7 (with 7 being the highest.) The current study uses this desirability index to assess interpersonal schemas. This decision was made due to previous studies suggesting this index’s high reliability (Soygut and Savasir, 2001). In addition, preliminary factor 73 analysis we conducted also suggested that the control index may be less reliable, and the affiliation index may be somewhat less reliable, and highly correlate with the desirability index for both mother (r=.91, p < .001) and father (r=.83, p < .001), thus being redundant to assess both indices. (See appendix C for further discussion and CFA tables). Intrex Questionnaire This study uses a 36-item version of the Intrex Questionnaire and corresponds to the third surface of the SASB circumplex model, which measures introjects (internalized self-representations). Therapists complete this questionnaire once at intake and are instructed to think of a specific time when they were at “their best” and a time when they were at “their worst”. They then rate themselves on a scale of 0-100 (in intervals of 10) two separate times, once in response to how they would describe themselves “at best”, and again how they would describe themselves “at worst.” Example items included “I tenderly, lovingly cherish myself” and “to make sure I do things right, I tightly control and watch over myself”. The measure yields two scores reflecting an “introject at best” and an “introject at worst” which are scored along five dimensions. The first three include self-attack, self-control and self-conflict and are primarily assessed with relatively small data sets employing non-parametric statistics. The latter two dimensions are self-autonomy and self-affiliation, and are easier to analyze with relatively large samples sizes and multivariate statistics (Pincus and colleagues, 1998). These two dimensions were solely used in this study due to the large sample size being assessed and the need for multivariate statistics. In addition, this study only uses the self-autonomy 74 and self-affiliation dimensions calculated from “at worst” scores. The “introject at best” scores were left out due to the nature of the questions yielding idealized scores that are restricted in range (Henry et al, 1990). Measures of Psychotherapy Process (Dependent Variables) Working Alliance Inventory (WAI) The WAI (Horvath & Greenberg, 1986) is a 36 item scale used to assess the quality of the therapeutic alliance. The current study used an abbreviated 12-item scale developed by Tracey and Kokotovic (1989) through factor analysis. The 12 item scale, like the original, assesses the three aspects of the working alliance defined by Bordin (1979): (i) The bond between the therapist and patient, (ii) the level of agreement between the therapist and patient on the goals of therapy and (iii) the level of agreement between the therapist and patient on the tasks of therapy. This abbreviated version uses the four items for each of the three subscales that had the highest factor loadings. In addition, this version of the WAI yields a total score which is an average across all 12 items. The therapists’ and the patients’ total WAI score for sessions 1-5 are used in this study to assess psychotherapy process in relation to therapists’ interpersonal schemas and introjects. The WAI is rated on a likert scale of 1-7 (with seven being the highest rating) and consists of items such as: “I think my therapist likes me” (Bond), “My therapist and I agree on important changes” (Goal), and “Therapy gives me a new way of looking at things” (Task). 75 Post-Session Questionnaire and Rupture Section The Post-Session Questionnaire described above includes a brief section developed by Muran, Safran, Samstag & Winston (1999) regarding the occurrence and resolution of ruptures. This section was designed as a means to identify patterns in ruptures and rupture resolution over the course of treatment, and asks patients and therapists the following questions regarding the most recent session: 1) Did you experience any problem or tension in your relationship during the session? (subject responds yes or no). 2) If so, about where in the session did this problem begin? (subject selects from beginning, middle or end). 3) Please rate the highest degree of tension you felt during the session as a result of this problem? 4) To what extent was this problem addressed in this session? 5) To what degree do you feel this problem was resolved by the end of the session? This study used question number 5 to determine the degree to which both the patient and therapist perceived ruptures to be resolved by the end of the corresponding session. Session Evaluation Questionnaire (SEQ) The SEQ (Stiles Gordon & Lani, 2002) includes 21 items that are rated on a 7 point, bi-polar scale. The questions were distinguished via factor analysis and divided into two sections. The first section is the “session evaluation” and includes 11 adjective dyads that assess two core therapeutic experiences; 1) valuable and powerful verses weak 76 and worthless; 2) relaxed and comfortable verses weak and distressing. The second section includes 10 items that assess two dimensions of post-session mood; 1) positivity and 2) arousal. Both sections of the SEQ are applicable to therapists’ and patients’ experiences as suggested by factor analysis (Stiles, 1908; Stiles, Reynolds, et al., 1994). The current study only uses the “session evaluation” section. Sample items include: Rough 1 2 3 4 5 6 7 Smooth Shallow 1 2 3 4 5 6 7 Deep Respondents are asked to circle the appropriate number along this spectrum that “best describes the session.” The 11 dyads are averaged to yield two variable ratings: Smoothness and Depth of experience, with higher ratings representing greater smoothness and greater depth (1 being the lowest and 7 being the highest). Both of these variables have demonstrated strong internal consistency (i.e., .93 for smoothness and .90 for depth). Procedures Pilot Study: Goals and Procedures of Part-I of study Twenty-four therapists who completed the ISQ at intake and who had seen at least 3 separate patients were selected. All therapists and patients in the program fill out a required post session questionnaire after each session that includes the Working Alliance Inventory (Horvath & Greenberg, 1989) and the section regarding ruptures described above. The first 5 post session questionnaires for both the patient and the therapist were 77 collected and scored. The WAI was used to assess therapists’ ability to establish a positive working alliance with their patients, and the item reflecting the degree to which ruptures were resolved was used to assess therapists’ likelihood of resolving ruptures with patients. Each therapist was measured across the first 5 sessions and for three separate patient cases in order to account for the patient variable. The patient and therapist scores were collapsed separately and over 15 sessions from all three therapistpatient dyads to create averages. Thus, in the end each therapist was measured by one average WAI score and one average Rupture Resolve score representing their 3 patients’ perspectives and the therapist’s own perspectives with each patient in the first 5 sessions of therapy. Therapists’ interpersonal schemas were assessed using the desirability index on the ISQ. Desirability ratings for each therapist were broken into the 4 circumplex quadrants (hostile, friendly, submissive and dominant) in order to see if desirability of particular kinds of situations predicts in-session processes as well as their ability to form a positive working alliance. For example, does desirability of expected responses from the mother in friendly situations predict in-session processes more or less than expected responses from the mother in hostile situations? Desirability for these quadrants was assessed for both mother and father independently. Pearsons r correlations were then run between therapists’ average desirability scores in hostile, friendly, submissive and dominant situations with both therapist and patient reports of the WAI and resolve item, and with mother and father separately. The findings suggested that therapists’ who expected desirable outcomes with their fathers were more likely to resolve ruptures with their patients (refer to table 1), whereas therapists’ who expected desirable responses 78 from the mothers were more likely to establish a working alliance (refer to table 2). These findings were consistent across situations, and for the therapist and patient perspectives. Table 1: Preliminary Results for Therapist ISQ and Rupture Resolution Desirability of Social Outcomes with Mother and Rupture Resolution Patient Resolve (r) Therapist Resolve (r) Hostile Situations (Quad1) .155 .481 .341 .111 Friendly Situations (Quad2) .317 .141 .320 .137 Submissive Situations (Quad3) .286 .195 .211 .333 .364 .087 .413* .050 * Dominant Situations (Quad4) Desirability of Social Outcomes with Father and Rupture Resolution Patient Resolve (r) Therapist Resolve (r) Hostile Situations (Quad1) .486* .019 .353 .098 Friendly Situations (Quad2) .437* .037 .377 .076 Submissive Situations (Quad3) .420* .046 .164 .456 .487* .019 .401 .058 * pDominant (Quad4) Situations < *p < .05 79 Table 2: Preliminary Results for Therapist ISQ and Working Alliance Desirability of Social Outcomes with Mother and Working Alliance Quadrant Patient WAI (r) Therapist WAI (r) Hostile Situations (Quad1) .282 .181 .154 .472 Friendly Situations (Quad2) .492* .015 .600* .002 Submissive Situations (Quad3) .364 .080 .281 .184 Dominant Situations (Quad4) .442* .031 .508* .011 Desirability of Social Outcomes with Father and Working Alliance Quadrant Patient WAI (r) Therapist WAI (r) Hostile Situations (Quad1) .217 .308 .129 .549 Friendly Situations (Quad2) .318 .229 .299 .155 Submissive Situations (Quad3) .180 .400 .180 .400 Dominant Situations (Quad4) .341 .102 .341 .102 *p < .05 80 Current Study: Goals and Procedures The current study expands on the pilot study findings by continuing to look at the relationship between therapists’ anticipated desirability of social outcomes with their mothers verses their fathers in relation to psychotherapy process, however, several limitations are addressed: First, the pilot study consisted of a small sample size, thereby making it necessary to assess only those therapists who had seen at least three patients. For example, the pilot study aims to control for any impact the patient’s personality may have on the therapeutic process by averaging three separate patients for each therapist (with the idea that no three patients will have the same personality, and therefore any significant relationship found between therapists’ interpersonal schemas and therapeutic process will be due to the therapist and not the patient). In contrast, the current study increases the sample size to look at 96 therapists’ in conjunction with all the patients they had seen through the BPRP. (Therapists ranged from seeing between 1-8 patients). Such a large sample size ensures adequate patient variability, making it reasonable to attribute significant findings to the therapist. This enables therapists’ to be included in the study even if they had seen only one or two patients. Thus, interpersonal schema data was collected from intake and added for an additional 72 therapists. Consistent with the MA thesis, the desirability index of the ISQ was used to represent measures of interpersonal schemas, thus averages for the level of anticipated desirability of social outcomes with mothers and fathers were calculated for each quadrant (reflecting hostile, friendly, submissive, and dominant situations), and corresponding working alliance and rupture resolution data for the first five sessions from both the therapist and their respective 81 patient(s) was also collected. In order to simplify the complexity of analyzing and interpreting data in individual sessions for several patients per therapist, we averaged across the first five sessions for within patient and within therapist data for each dependent variable. A second limitation of the pilot study is that the primary rupture resolution measure reported by therapists and patients is a single item response and may therefore have limited validity. The current study addresses this limitation by including the SEQ which measures the degree of perceived smoothness and depth of experience in therapy sessions. This questionnaire was assessed for sessions where ruptures were reported and in relation to therapists’ interpersonal schemas of their mothers and their fathers. The goal of including this measure was to obtain a richer understanding of the relationship between interpersonal schemas and rupture resolution. Thirdly, the current study includes the therapist’s introject as an independent variable effecting therapists’ ability to establish a working alliance, and resolve ruptures with patients. Therapists’ introjects are measured by the Intrex Questionnaire which is completed once at intake. In the current study, therapists’ introjects are measured according to the self considered at worst along a self-autonomy dimension and selfaffiliation dimension These two dimensions are assessed for their predictive value of the working alliance and rupture resolution ratings across the first five sessions and as perceived by both the patient and the therapist. Fourthly, the pilot study used preliminary statistics such as pearson r correlations to assess the data. However, three patients were matched to every one therapist. Thus, 82 therapists’ independent variables were essentially measured times three thereby inflating the correlations. The pilot study made some attempt to account for this by averaging scores across all three patients. However, this process limits the depth and specificity of the findings. The current study controls for nested data while maintaining individual ratings and accounting for variance between patients by employing multi-level modeling that can accurately control for several patients being paired with only one therapist. CHAPTER V Analyses Intrex variables (i.e., therapist affiliation and autonomy at worst) and ISQ variables (i.e., therapist desirability of anticipated social outcomes with the mother and the father in hostile, friendly, submissive and dominant situations) were assessed in relation to therapist and patient ratings of psychotherapy process using multilevel modeling (Snijders & Bosker, 1992) in the statistical program “R” (R Core Development Team, 2009). This method allows for the analysis of “nested” data, in this case multiple patients being seen by the same therapist. In other words, using multilevel modeling enabled us to account for the nesting in the data created by therapists being rated by their patients as well as themselves in multiple psychotherapy dyads. This allowed for analysis at the individual level, appropriately capturing the similarities that may exist among the psychotherapy process ratings for multiple patients seen by a given therapist. Multi-level modeling is also referred to as hierarchical modeling, or random coefficient modeling, and is employed to address “individual-level data that is collected within 83 groups” (Ciarleglio & Makuch, 2007). Ordinary least squares (OLS) regression assumes independence of observations, and would not account for the similarity (or nonindependence) of ratings of patients seeing the same therapist. In this study, ratings of patients who are seen by the same therapist are expected to be more similar than ratings of patients with different therapists as a result of therapist consistencies across patients and time that impact ratings of psychotherapy process. Multi-level modeling does not assume “independence” of observations, and takes into account the hierarchical structuring of the data (patient and therapist ratings are nested within one unique therapist). In this study, multilevel models with random-intercepts are used where patient and therapist ratings of psychotherapy process (level-one) are nested within therapists (level-two). Patient gender and age are controlled in the level-one portion of the model while therapist gender and age are controlled in the level-two portion of the model. For a more in depth description of multi-level modeling see Ciarleglio and Makuch (2007). Before proceeding to multi-level modeling to examine Intrex and ISQ variable effects, therapist and patient ratings were averaged across the first five sessions creating one observation per dependent variable for individual therapists and patients. This was done in order to simplify the multilevel modeling and summarize the earliest sessions. In other words, while the data were structured for analysis to take into account the nesting of therapist and patient ratings under unique therapists, they were not structured to take into account the individual “session” as this would have been unwieldy. We also calculated Cronbach’s alpha coefficients before proceeding to multi-level modeling in order to determine the degree of agreement between therapists and patients 84 on dependent variable ratings. Results suggested low agreement on all measures including rupture resolution ( = .341) working alliance ( = .538) smoothness ( = .344) and depth ( = .435), indicating the need to separately assess patients’ and therapists’ ratings in relation to therapist interpersonal schemas and introjects. The dataset for multi-level model analysis was structured to include a “rater” variable which had values of either “therapist” or “patient.” This rater variable enabled patient and therapist ratings to remain separate yet be analyzed in the same multilevel statistical model. The interaction between an Intrex or ISQ variable and the rater variable was examined to determine whether the association of the Intrex or ISQ variable with a psychotherapy process variable was different for therapist and patient ratings. If an interaction effect with the rater variable was significant (or approaching significance at the p < .05 level), this suggested that associations were not the same for therapist and patient ratings of the process variable. When even marginally significant interaction effects involving the rater variable were observed, the simple main effects of the Intrex or ISQ variable were estimated (simple main effects were estimated by specifying appropriate contrasts of the coefficients from the multilevel model with the interaction term). Thus, when a significant interaction was observed, separate regression coefficients characterized the association of the Intrex or ISQ variable with 1) the therapist rating of the process variable and 2) the patient rating of the process variable. When the rater variable and the Intrex or ISQ variable did not have a significant interaction effect, the interaction term was dropped but the main effect for the rater variable was retained. In these cases, the regression coefficient for the Intrex or ISQ variable indicated its 85 association with a combined rating of the process variable by both the therapist and patient. In other words, when the rater variable was included only as a main effect, the multilevel model essentially averaged therapist and patient ratings of the process variable, but adjusted for the mean difference between ratings by therapists and patients. 86 Chapter VI Results Descriptives of Ratings Table 3: Summary of Dependent Variable Ratings Therapist Mean SD Median Min Max n Alliance 4.68 0.67 4.74 2.12 6.13 (First 5 session Avg) Rupture Resolution 2.79 1.92 2.77 1.00 5.00 (First 5 session Avg) Depth 4.79 0.53 4.85 3.60 5.85 (First 5 session Avg) Smooth 4.20 0.54 4.20 2.65 5.30 (First 5 session Avg) n = number of observations after averaging first five sessions 186 150 61 61 Patient Mean SD Median Min Max n Alliance 4.95 0.78 5.00 2.33 6.88 (First 5 session Avg) Rupture Resolution 3.09 1.02 3.00 1.00 5.00 (First 5 session Avg) Depth 4.86 0.60 4.95 3.35 5.85 (First 5 session Avg) Smooth 4.13 0.62 4.00 3.00 5.62 (First 5 session Avg) n = number of observations after averaging first five sessions 177 103 45 45 Note: Slight variation in the number of process ratings exists from model to model due to missing data. 87 Table 4: Summary of Therapist-ISQ Quadrant Ratings for Desirability Mother Mean SD Median Min Max n Hostile 3.65 1.52 3.40 1.00 6.80 95 Friendly 5.86 0.99 6.10 1.40 7.00 96 Submissive 4.71 1.33 5.00 1.60 7.00 94 Dominant 4.80 1.32 5.00 1.20 7.00 95 Median Min Max n Father Mean SD Hostile 3.89 1.36 3.80 1.00 7.00 95 Friendly 5.51 1.27 5.80 1.20 7.00 95 Submissive 4.58 1.34 4.60 1.00 7.00 94 Dominant 4.70 1.35 5.00 1.00 7.00 95 Note: Slight variation exists from model to model if a therapist(s) is excluded from the analysis due to missing data, or there are no reported ruptures associated with their cases. Multilevel Modeling Hypothesis I: Therapist ISQ, Working Alliance and Rupture Resolution a) Working Alliance: Therapists' interpersonal schemas of their mothers and fathers (in hostile, friendly, submissive, and dominant situations) were assessed in relation to working alliance as perceived by patients and therapists. There were no significant findings (p < .05) for any of these relationships (refer to Table 12). However, patients had significantly higher ratings of working alliance than therapists (p = .0001). 88 b) Rupture Resolution: Therapists' interpersonal schemas of their mothers and fathers (in hostile, friendly, submissive, and dominant situations) were assessed in relation to rupture resolution as perceived by patients and therapists. On average, patients tended to report higher rupture resolution than therapists (p = .0045). As illustrated in tables 5 - 8, significant relationships were found in friendly and dominant situations for both mother and fathers: (see table below) 89 Mother Findings Table 5: ISQ-Mother in Friendly Situations and Therapist/Patient Rupture Resolution FIXED EFFECTS PARAMETER Intercept Coefficient PT TH 3.09 1.43 SE PT 0.74 df TH 0.55 PT TH 151 151 t PT 4.12 p TH 2.60 PT 0.00 TH 0.01 LEVEL 1 (PATIENT) Gender -0.003 0.129 151 -0.128 0.977 Age -0.010 0.006 151 -1.577 0.116 1.660 0.805 151 2.061 *0.041 Gender 0.110 0.156 88 0.704 0.438 Age 0.015 0.009 151 1.529 0.128 -0.232 0.133 151 -1.734 0.084 Rater Effect LEVEL 2 (THERAPIST) INTERACTION ISQ MQuad2 PT TH PT -0.01 0.22 0.11 TH PT TH 0.09 151 151 PT -0.11 TH 2.39 PT 0.910 TH *0.018 RANDOM EFFECTS Standard Deviation INTERCEPT 0.417 RESIDUAL 0.896 ISQ MQuad2 = Therapists anticipated desirability from mothers in friendly situations. Male = 1; Female = 0. Rater Effect and Interaction statistics in this table reflect coding scheme of Th = 0 and Pt = 1. *p < .05 This model (table 5) consisted of 90 therapists, with 247 observations of rupture resolution (combined patient and therapist; each observation represents either a patient or therapist average rating over the first five sessions). Patients rated rupture resolution higher than therapists, t(151) = 2.061, p = .041. However, the marginally significant 90 interaction between the rater variable and ratings of mother desirability in friendly situations, t(151) = -1.734, p = .084, suggests therapists and patients ratings were discrepantly associated with the predictor variable. Due to this interaction, the patient and therapist rupture resolution effects were estimated separately. After controlling for gender and age, the results suggest that therapists' expectations of social outcomes with their mothers in friendly situations are positively related to therapists' reports of rupture resolution, t(151) = 2.387, p = .018. No significant effect was found on patient ratings of rupture resolution, t(151) = -.112, p = .911. 91 Table 6: ISQ-Mother in Dominant Situations and Therapist/Patient Rupture Resolution FIXED EFFECTS PARAMETER Intercept Coefficient PT TH 3.31 1.96 SE PT 0.69 df TH 0.35 t PT TH 150 150 PT 6.68 p TH 5.59 PT 0.00 TH 0.01 LEVEL 1 (PATIENT) Gender -0.013 0.130 150 -0.105 0.916 Age -0.010 0.006 150 -1.695 0.092 1.174 0.512 150 2.290 *0.023 Gender 0.122 0.156 87 0.781 0.436 Age 0.012 0.009 150 1.32 0.188 0.098 150 -1.810 0.072 Rater Effect LEVEL 2 (THERAPIST) INTERACTION ISQ MQuad4 -0.17 PT TH PT 0.02 0.16 0.09 TH PT TH 0.07 150 150 PT -0.25 TH 2.28 PT 0.801 TH *0.024 RANDOM EFFECTS Standard Deviation INTERCEPT 0.405 RESIDUAL 0.904 ISQ MQuad4 = Therapists anticipated desirability from mothers in dominant situations. Male = 1; Female = 0. Rater Effect and Interaction statistics in this table reflect coding scheme of Th = 0 and Pt = 1. *p < .05 This model (table 6) consisted of 89 therapists and 245 observations of rupture resolution (combined patient and therapist; each observation represents either a patient or therapist average rating over first five sessions). Patients rated rupture resolution higher than therapists, t(150) = 2.290, p = .023. However, the marginally significant interaction between the rater variable and ratings of mother desirability in dominant situations, t(150) = -1.81, p = .072 suggests therapists and patients ratings of rupture resolution were 92 discrepantly associated with the predictor variable. Due to this interaction, the patient and therapist rupture resolution effects were estimated separately. After controlling for gender and age, therapists' expectations of social outcomes with their mothers in dominant situations were positively related to therapists' reports of rupture resolution, t(150) = 2.279, p = .024. No significant effect was found on patient ratings of rupture resolution, t(150) = -.246, p = .805. 93 Father Findings Table 7: ISQ-Father in Friendly Situations with Therapist/Patient Rupture Resolution FIXED EFFECTS PARAMETER Intercept Coefficient PT TH 2.80 1.61 SE PT 0.48 df TH 0.43 t PT TH 151 151 PT 5.86 p TH 3.80 PT 0.00 TH 0.01 LEVEL 1 (PATIENT) Gender -0.027 0.130 151 -0.213 0.831 Age -0.011 0.006 151 -1.725 0.086 Rater Effect 1.186 0.540 151 2.195 0.091 LEVEL 2 (THERAPIST) Gender 0.135 0.154 87 0.872 0.385 Age 0.018 0.010 151 1.817 *0.029 -0.161 0.094 151 -1.698 0.066 INTERACTION ISQ FQuad2 PT TH 0.16 0.09 PT 0.07 TH PT TH 0.07 151 151 PT 1.70 TH 2.70 PT 0.81 TH **0.007 RANDOM EFFECTS Standard Deviation INTERCEPT 0.404 RESIDUAL 0.899 ISQ FQuad2 = Therapists anticipated desirability from fathers in friendly situations. Male = 1; Female = 0. Rater Effect and Interaction statistics in this table reflect coding scheme of Th = 0 and Pt = 1. *p < .05; **p < .01 This model (table 7) consisted of 89 therapists and 246 observations of rupture resolution. Patients rated rupture resolution higher than therapists, t(151) = 2.195, p = .029. However, the marginally significant interaction between the rater variable and ratings of father desirability in friendly situations, t(151) = -1.698, p = .091 suggests therapists’ and 94 patients’ ratings of rupture resolution were discrepantly associated with the predictor variable. Due to this interaction, patient and therapist rupture resolution effects were estimated separately. After controlling for gender and age, the results suggest that therapists' expectations of social outcomes with their fathers in friendly situations were positively related to therapists' reports of rupture resolution, t(151) = 2.696, p = .007. No significant effect was found on patient rupture resolution ratings, t(151) = 1.698, p = .805. 95 Table 8: ISQ-Father in Dominant Situations with Therapist/Patient Rupture Resolution FIXED EFFECTS PARAMETER Coefficient SE df 2.088 .272 141 Gender -0.031 0.129 141 -0.245 0.806 Age -0.011 0.006 141 -1.83 0.069 Rater Effect 0.346 0.122 141 2.823 **0.005 Gender 0.180 0.142 87 0.271 0.206 Age 0.018 0.009 141 1.998 *0.047 ISQ FQuad4 0.119 0.053 141 2.240 *0.026 Intercept t p 7.653 0.000 LEVEL 1(PATIENT) LEVEL 2 (THERAPIST) RANDOM EFFECTS INTERCEPT Standard Deviation 0.296 RESIDUAL 0.905 ISQ FQuad4 = Therapists anticipated desirability from fathers in dominant situations. Male = 1; Female = 0. Rater Effect statistics table reflect coding scheme of Th = 0 and Pt = 1. Intercept and ISQ Quad2 statistics reflect effects on combined patient and therapist ratings. *p < .05 This model (table 8) consisted of 89 therapists and 235 observations of rupture resolution. Patients rated rupture resolution higher than therapists, t(141) = 2.823, p = .005. This model also found a significant effect for therapist age, suggesting that as therapists get older, they are more likely to perceive rupture resolution with their patients, t(141) = 1.99, p < .05. The interaction variable “rater * FQuad4” was tested however dropped from the statistical model as it was not even marginally significant. Thus, after controlling for gender and age, combined patient and therapist reports of rupture were 96 assessed using the rater variable. The results suggest that therapists' expectations of social outcomes with their fathers in dominant situations are positively related to therapists' and patients’ reports of rupture resolution, t(141) = 2.40, p = .026. HYPOTHESIS II : Therapist ISQ and Session smoothness and Depth a) Smoothness: Therapists' interpersonal schemas of their mothers and fathers (in hostile, friendly, submissive, and dominant situations) were assessed in relation to the degree of smoothness perceived by therapists and patients in rupture sessions. There were no significant findings for the mother (p < .05, refer to table 12). As illustrated in table 9, a significant relationship was found for the father in dominant situations. (see table on following page) 97 Father Findings Table 9: ISQ-Father in Dominant Situations and Perceived Session Smoothness FIXED EFFECTS PARAMETER Intercept Coefficient PT 4.80 TH 4.04 SE PT 0.34 df TH 0.29 PT 46 t TH PT 46 13.56 p TH 13.89 PT 0.00 TH 0.00 LEVEL 1 (PATIENT) Gender 0.040 0.12 46 0.336 0.737 Age 0.003 0.006 46 0.489 0.626 Rater Effect 0.768 0.456 46 1.682 0.099 Gender -0.085 0.120 45 -0.705 0.484 Age 0.007 0.006 46 1.140 0.260 -0.176 0.094 46 -10.876 0.066 LEVEL 2 (THERAPIST) INTERACTION ISQ FQuad4 PT TH -0.15 0.03 PT 0.07 TH PT TH 0.06 46 46 PT -2.03 TH 0.49 PT *0.048 TH 0.62 RANDOM EFFECTS Standard Deviation INTERCEPT 2.503 RESIDUAL 0.572 ISQ FQuad4 = Therapists anticipated desirability from fathers in dominant situations. Male = 1; Female = 0. Rater Effect and Interaction statistics in this table reflect coding scheme of Th = 0 and Pt = 1. *p < .05 This model (table 9) consisted of 47 therapists and 99 observations of session smoothness. There was an approaching marginally significant interaction between the rater variable and ratings of father desirability in dominant situations, t(46) = -10.88, p = .066, suggesting that therapist and patient ratings of session smoothness were 98 discrepantly associated with the predictor variable. Due to this interaction, the simple main effects of patient and therapist session smoothness were estimated separately. After controlling for gender and age, the results suggest that therapists' expectations of social outcomes with their fathers in dominant situations are negatively related to patients' reports of in-session smoothness, t(46) = -2.029, p = .048. No significant effect was found on therapists’ ratings of session smoothness, t(46) = .494, p = .623. b) Depth: Therapists' interpersonal schemas of their mothers and fathers (in hostile, friendly, submissive, and dominant situations) were assessed in relation to the degree of depth experienced in sessions as perceived by therapists and patients in rupture sessions. There were no significant findings for the mother (p < .05, refer to table 12). As illustrated in table 10, a marginally significant relationship was found for the father in friendly situations: (See table below) 99 Father Findings Table 10: ISQ-Father in Friendly Situations and Depth of Experience in Session FIXED EFFECTS PARAMETER Intercept Coefficient 4.285 SE df 0.298 50 t p 14.361 0.000 LEVEL 1(PATIENT) Gender -0.010 0.113 50 -0.094 0.925 Age 0.001 0.006 50 0.224 0.823 Rater Effect 0.063 0.107 50 0.583 0.561 Gender -0.017 0.125 46 -0.139 0.889 Age -0.011 0.007 50 -1.614 0.112 ISQ FQuad2 0.100 0.051 50 1.959 *0.055 LEVEL 2 (THERAPIST) RANDOM EFFECTS INTERCEPT Standard Deviation 0.296 RESIDUAL 0.905 ISQ FQuad2 = Therapists anticipated desirability from fathers in friendly situations. Male = 1; Female = 0. Rater Effect statistics in this table reflect coding scheme of Th = 0 and Pt = 1. Intercept and ISQ Quad2 statistics reflect effects on combined patient and therapist ratings. *p = .055. This model (table 10) consisted of 48 therapists and 103 reports of session depth. Therapist and patient ratings were assessed together because father friendliness did not interact with rater (the interaction variable “rater * FQuad4” was not even marginally significant, and thus dropped from the statistical model). After controlling for gender and age, the results suggest that therapists' expectations of social outcomes with their fathers 100 in friendly situations are related to therapists’ and patients' reports of in-session depth of experience. This finding was marginally significant; t(50) = 1.959, p = 0.055. Hypothesis III : Therapists’ Introjects a) Therapists' introjects' as autonomous-self were assessed in relation to the four dependent variables (working alliance, rupture resolution, smoothness, and depth) for therapist and patient perspectives. No significant relationships were found (p < .05, refer to table 12). b) Therapists' introjects' as affiliative-self were assessed in relation to the four dependent variables (working alliance, rupture resolution, smoothness, and depth) for therapist and patient perspectives. As illustrated in table 11, a significant relationship was found between therapists' introjects of affiliative-self and therapists’ reports of rupture resolution. (see table on following page) 101 Table 11: Therapists’ Affiliative Introjects and Rupture Resolution FIXED EFFECTS PARAMETER Intercept Coefficient PT TH 3.00 2.75 SE PT 0.15 df TH 0.13 t PT TH PT 100 100 19.50 p TH 13.89 PT 20.80 TH 0.00 LEVEL 1 (PATIENT) Gender -0.145 0.157 100 -0.920 0.359 Age -0.008 0.007 100 -1.069 0.287 0.253 0.145 100 1.744 0.084 Gender 0.205 0.186 73 1.101 0.274 Age 0.003 0.013 100 0.239 0.811 -0.005 0.001 100 2.085 **0.005 Rater Effect LEVEL 2 (THERAPIST) INTERACTION Introject-Aff PT TH PT -0.00 0.00 0.00 TH 0.00 PT TH PT 0.00 0.00 -0.704 TH 3.01 PT 0.48 TH **0.00 RANDOM EFFECTS Standard Deviation INTERCEPT .424 RESIDUAL .901 Introject-Aff = Therapists affiliative introjects of self. Male = 1; Female = 0. Rater Effect and Interaction statistics in this table reflect coding scheme of Th = 0 and Pt = 1. **p < .01 Only 81 therapists from the total sample had available Intrex data. Due to additional missing data, this model (table 11) consisted of 74 therapists and 181 observations of rupture resolution (combined patient and therapist; each observation represents either a patient or therapist average rating over first five sessions). There was a significant interaction between the rater variable and therapist affiliative introject ratings, t(100) = 102 2.085, p = .005, suggesting that therapists’ and patients’ ratings of session smoothness were discrepantly associated with the predictor variable. Due to this interaction, the simple main effects of patient and therapist rupture resolution were estimated separately. After controlling for gender and age, therapists’ affiliative beliefs towards themselves were associated with therapists’ ratings of rupture resolution, t(100) = 3.01, p = .003. No significant association was found between therapists’ affiliative beliefs and patients’ ratings of session smoothness, t(100) = -.704, p = .483. Random Effects The random effects represent the random variation in outcome (randomly varying intercepts). In other words, the standard deviations reflect the variation in how therapists and patients rate dependent variables after removing the fixed effects of interpersonal schemas or introjects. The intercept standard deviation can be interpreted as the degree to which therapy process is rated consistently across patients treated by the same therapist, and the residual standard deviation reflects the variability within a single therapist in ratings of different patients. As illustrated in tables 5-11, the residual standard deviation is consistently higher than the intercept standard deviation. Thus, therapists are not responding in the same way to each patient. This suggests that variance not attributable to therapists’ interpersonal schemas or introjects is primarily accounted for by individual differences in patients, and how therapists and patients are likely to rate aspects of rupture resolution in response to those differences. Overall, this finding helps to confirm that outside the effects of interpersonal schemas and introjects, therapists are 103 responding to the individual differences of their patients rather than rating therapeutic processes in the same way regardless of who they are treating, and different patients are rating therapeutic processes differently despite having the same therapist. Table 12: Summary of Regression Coefficients for all Main Effects of ISQ/Intrex Mother ISQ Hostile Friendly Submissive Dominant Father ISQ Hostile Friendly Submissive Dominant Alliance RR TH .028 (.424) -.003 (.948) .067 (.086) .016 (.682) .111 (.076) .218 *(.018) .068 (.195) .156 *(.024) Alliance .028 (.458) .075 (.071) .033 (.402) .055 (.155) -.097 (.178) -.013 (.910) -.082 (.316) -.021 (.805) RR TH Depth Smooth -.007 (.870) .081 (.269) .044 (.404) -.006 (.900) -.002 (.945) .019 (.777) 0.27 (.569) -.017 (.735) PT Depth PT .045 (.417) .199 .037 **(.007) (.652) .040 (.473) .119 *(.026) RR Depth Smooth TH PT Affiliation .000 .004 -.001 .000 .000 (.556) **(.003) (.483) (.624) (.729) Autonomy -.000 .001 -.005 .000 -.000 (.539) (.574) (.063) (.399) (.668) Signficance = * p < .05; ** p < .01 Note: cells split with TH/PT represent models with significant interactions and thus patient and therapist were assessed separately. Cells with one rating reflect models that combined patient and therapist effects. Intrex Alliance .012 (.825) .100 *(.055) .035 (.552) .046 (.377) Smooth TH PT -.034 (.521) -.008 (.869) -.042 (.447) .030 -.146 (.623) *(.048) 104 Chapter VII Discussion Rupture Resolution Hypothesis I was at least partially supported, suggesting in multiple situations that therapists' relationships with their fathers, and beliefs and expectations surrounding those relationships are meaningfully related to therapists' rupture resolution in psychotherapy, regardless of patient or therapist gender. Specifically, the findings suggest that the more therapists anticipate desirable outcomes in friendly and dominant social interactions with their fathers (in response to their own friendly and controlling behavior), the more likely they are to resolve ruptures with their patients. These relationships were also observed between therapists' anticipated social outcomes with their mothers and rupture resolution; however, anticipated outcomes with mothers only predicted therapist and not patient ratings of rupture resolution. The relationship with the father in dominant situations, on the other hand, predicted both therapists' and patients' perceptions of rupture resolution. This suggests that the relationship with the father is a more powerful predictor of rupture resolution than that of the mother, as well as that interpersonal schemas in dominant situations predict more accurate representations of psychotherapy process. Therapists’ interpersonal expectations of fathers and mothers in friendly situations only predicted their own ratings of rupture resolution. Perhaps therapists’ friendly tendencies account for their likelihood of perceiving rupture resolution, whereas these same friendly tendencies conceal unresolved tension that continues to be experienced by their patients. 105 Hypothesis II was also supported in suggesting that the relationship with the father is important to not only the likelihood and degree of rupture resolution, but to the quality of rupture sessions including perceived smoothness and depth, while the relationship with the mother did not predict these variables. For example, therapists who anticipated more desirable social outcomes with their fathers in friendly situations were more likely to have rupture sessions where both patients and therapists experienced increased levels of emotional depth. Thus, therapists’ friendly interactions with their fathers may be relevant to effective conflict resolution skills, or at least to the approach taken by therapists during difficult and tense moments with patients. Perhaps therapists who imagine positive reactions to their own friendly behaviors with fathers feel more comfortable with friendliness in general, thus creating a feeling of security with patients that enables a more in depth exploration of interpersonally challenging moments. Therapists’ interpersonal schemas of their fathers’ in dominant situations were negatively related to patients’ perception of smoothness in rupture sessions. This suggests that therapists who expect desirable outcomes in response to their own controlling behavior with their fathers have patients who experience more ‘roughness’ in working through rupture sessions. It is uncertain as to whether or not this “roughness” is related to positive or negative outcome; however, it is possible that this roughness is an important characteristic of resolving conflicts, and therapists who expect positive outcomes in controlling situations may be more comfortable with control, less likely to avoid moments of tension and more likely to confront uncomfortable interactions with their patients. On the other hand, therapists’ interpersonal schemas of their fathers in 106 dominant situations only predicted patient ratings of ‘roughness,’ and may be indicative of a therapists’ controlling tendencies that are overlooked by therapists but are experienced by patients as tense or uncomfortable moments. Nonetheless, the overall combined findings from hypotheses I and II suggest that therapists’ interpersonal schemas with their fathers are meaningful predictors of rupture resolution in psychotherapy. There is also limited evidence suggesting that age may affect rupture resolution (as described in table 8), and that as therapists get older they are more likely to perceive rupture resolution. Older therapists are more likely to have more experience, and thus may be more effective at resolving ruptures with patients. Dominant and friendly situations were consistently the two interpersonal situations that yielded significant results for both the degree of rupture resolution, and for the quality of rupture sessions (while hostile and submissive situations were not significantly related to any outcome variables). Although not all four situations were meaningful predictors of rupture variables, each of the two interpersonal dimensions characterized on Kiesler’s interpersonal circumplex model of behavior were represented as significant predictors: Dominant on the Dominant-Submissive scale and Friendly on the Friendly-Hostile scale. This suggests that both the control axis, and the affiliative axis are important, and that experiences with fathers in controlling and affiliative situations are important in developing beliefs and approaches to conflict situations. These findings are consistent with studies from the child-development literature suggesting that internalized experiences with their fathers are important in the 107 development of conflict resolution skills (refer to chapter 2). This relationship may have to do with an emerging relevance of the father at preschool age when toddlers begin to negotiate their social world (Steele and Steele, 2005), and potential characteristics relevant to how fathers may engage in play, and how they approach conflict related situations (McDowell et al., 2002; Parke et al., 1989; Pakaslahti et al 1996; Roberts, 1999). Perhaps fathers are particularly relevant in discipline situations which involve interpersonal conflict, and thus the relationship with the father is more deeply or meaningfully internalized with relation to difficult social situations. Further research is required to more fully understand this relationship. This study in particular suggests that this relationship is translated into the therapy room and becomes important to the therapeutic process and rupture resolution. This study may more precisely suggest that therapists who anticipate overall desirable outcomes with their fathers in response to their own friendliness have healthy perspectives on social relationships, or feel more comfortable with their own friendly behaviors, and are therefore comfortable taking interpersonal risks with their patients to resolve breaches in the therapeutic alliance, at least from the therapist’s own perspectives. Therapists who have had positive outcomes with their fathers in response to their own controlling behaviors may know how to effectively negotiate more difficult situations, and ultimately feel more comfortable with asserting their own feelings and perceptions of breaches in the therapeutic alliance. Working Alliance 108 Hypothesis I predicted that the working alliance would be more related to therapists' relationships with their mothers than that of their fathers. The findings did not suggest any meaningful relationship between alliance and either mother or father schemas. Nonetheless, it is worth noting that in a couple of situations the results approached significance. For example, as illustrated in table 12, anticipated responses with mothers in submissive situations and anticipated responses with fathers in friendly situations approached significance. Although not quite statistically significant in this study, further assessment to understand potential relationships between therapists' interpersonal schemas of their mothers and/or fathers and the working alliance they form with patients may be worthwhile. Therapists’ Introjects and Rupture Resolution Hypothesis III was supported in that there was a positive relationship between therapists’ affiliative introjects (intrapersonal representations) and therapists’ perception of rupture resolution. In other words, therapists who are able to maintain friendly beliefs about themselves even in worst case situations are more likely to experience themselves as confronting and resolving ruptures with patients. This finding may suggest that therapists who view and approach themselves in a friendly manner (even when they are feeling at their worst and there is more potential for self-blame, self-control and inward directed hostility) have healthy self-concepts and are therefore more comfortable and perhaps confident in threatening interpersonal situations. This finding provides support to the overall concept that internalized representations of self, through intra-psychic and 109 interpersonal processes are meaningful in relation to rupture resolution, and that further research in this domain is a meaningful endeavor. In addition, the Intrex findings in concurrence with the ISQ findings demonstrate that the relationship between internalized representations of friendliness and rupture resolution is suggested in two at least partially separate constructs (interpersonal schemas and introjects). Perhaps friendliness is a more generalized characteristic that can be more easily captured in similar ways across measures, whereas beliefs of control are more nuanced in representation and measurement. Limitations and Future Directions A methodological limitation of this study includes the use of a single-item measure to assess rupture resolution. This was to some extent addressed through adding additional measures to capture the smoothness and depth of experience in rupture sessions; however, only the single-item rating was used as a measure of the actual extent of resolution experienced by patients and therapists. Psychometrists have traditionally discouraged the use of single-item measures citing concerns as to the ability of one item to correlate with a construct (Nunnally & Bernstein, 1994). However, single-item measures of moderately complex constructs have also been argued to have better face validity and be less burdensome for participants, thus reducing the risk of annoyance and unreliable responding (Wanous, Reichers, and Hudy, 1997). In addition, studies have demonstrated reliability, convergent validity and discriminant validity for single-item measures that assess such variables as symptom severity, psychosocial functioning and 110 quality of life in depressed patients (Robins, Hendin and Trzesniewski; 2001 & Zimmerman et al., 2006). Nonetheless, future research benefit from additional measures that can capture the nuances of rupture resolution processes. Examining video-tapes of sessions reporting high resolution may be one way of identifying more precisely what processes are occurring, and if any of these processes are more robustly related to therapists’ interpersonal schemas. Another methodological limitation in this study is that the independent and dependent variables are self-report measures. Findings from previous studies have indicated that reliance on self-report measures may underestimate the actual frequency of ruptures (Sommerfeld, Orbach, Zim & Mikulincer, 2008; Eubanks-Carter, Muran, Safran, & Mitchell, 2008). Therapists may be reluctant to respond openly about their relationships with their parents. Novice therapists in particular may have reservations regarding reporting a rupture, as they may feel that this could be interpreted as a lack of therapeutic skill on their part. Patients as well may be reluctant to recognize a problem in the therapeutic relationship, and may be less attuned than therapists to the therapeutic process and subtle ruptures. This may explain in part therapists’ tendency to report more ruptures than patients (150 verses 103 respectively for the overall sample). Samstag and colleagues (Samstag et al., 1998) conducted a pilot study within the brief psychotherapy research program and found patient failure to complete post-session questionnaires was a better predictor of drop-out than working alliance ratings. Assuming patients who dropout often experience ruptures in the alliance, this finding may suggest that ruptures perceived by patients were not always reported, potentially weakening the overall power 111 of the study. In addition, when patients did report ruptures they tended to report higher resolution than therapists, perhaps feeling obligated to continue to view their therapist in a positive light. This study only used data from the first five sessions of a thirty-session psychotherapy protocol. This was done in part for manageability of such a large database, but was also supported theoretically in that the first several sessions are often considered to be particularly important in predicting the course of therapy (Suh, O’Malley and Strupp, 1986; Horvath & Symonds, 1991). Nonetheless, limiting the study to these sessions may miss many ruptures that evolve over time, ebbs and flows in the therapeutic alliance, and patterns of relationships between interpersonal schemas and rupture resolution. To address several of these methodological limitations, a follow-up study may use a smaller sample size but include more sessions, and obtain observerbased rupture ratings through video-taped sessions or transcripts. Although the goals of this study were to examine therapeutic processes, follow-up studies may assess how therapists’ interpersonal schemas and rupture resolution affect psychotherapy outcomes. Other limitations of this study include the lack of cultural and ethnic diversity represented in the sample, and the fact that all subjects were participants in a research study in a hospital setting, with a focus on cluster C personality disorders. Variables related to Cluster C personality disorders, such as avoidant or dependent tendencies, as well as aspects of the research program, including video-taped sessions, time-limited therapy and extensive supervision, may limit the generalizability of the findings. 112 This study did not control for factors of father availability and whether or not and to what extent a father was present throughout a therapist’s childhood. Factors such as travel, degree of occupational intensity, divorce and death may significantly affect the relationship between interpersonal schemas of fathers and conflict resolution skills. In addition, the father as a primary or sole care-taker and the availability or absence of the mother may also significantly affect this relationship. In addition to considering these factors related to availability, future directions may include looking more closely at approaches to fathering. For example, a worthwhile endeavor may include assessing more qualitatively or elaborately -- through the use of detailed self-report measures, interviewing, or observer-based ratings -- how and what fathers do when responding to their children’s friendly and controlling behaviors, perceived outcomes of these interactions, and how these variables relate to their children’s conflict resolution skills in social settings. Follow-up studies may also asses the relationship between introjects and interpersonal schemas. Our assessment of these two variables provided evidence suggesting that constructs in relation to interpersonal and intrapersonal expectations of friendliness are predictive of rupture resolution. Future studies may assess the relationship between these two variables, including construct validity, potential interactions, or moderating or mediating effects. Conclusion 113 This study contributes to child-development literature by contributing supportive data suggesting that the father, more so than the mother, may play a critical role in the development of conflict resolution skills. More so, this relationship is translated into the therapy room, affecting psychotherapeutic process. Thus, this study contributes to the child-development literature, as well as to psychotherapy process literature suggesting that therapists’ personalities are meaningful predictors of psychotherapy process, highlighting the importance for therapists to evaluate their own internal processes and how they react to and approach breaches in the therapeutic alliance. As suggested in previous research, rupture resolution is an important component of psychotherapeutic change, and the failure of therapists to recognize and address impasses with their patients may preclude opportunity for core schematic change to occur. In addition, the ability of therapists to recognize and resolve ruptures with their patients may decrease missed sessions, and prevent patient drop-out (Muran, J.C., et al., in press). The findings from this study may be considered relevant to child-development literature and potentially as information for approaches to parenting. For example, fathers may benefit from an increased awareness, and understanding that the way they respond to their children’s behaviors (e.g., friendliness and controlling behaviors) are important determinants of their children’s future beliefs surrounding conflicts, and conflict resolution skills. These findings may also have important implications for training psychotherapists. For example, these findings may be applied to therapist supervision, helping therapists to understand their own beliefs and approaches to conflict, how these beliefs may be inhibiting or even contributing to maladaptive patterns in the alliance, and ultimately 114 helping them to become comfortable addressing difficult interpersonal moments with patients. 115 References Ackerman, S.J., & Hilsenroth, M.J. (2001). A review of therapist characteristics and techniques negatively impacting the therapeutic alliance. Psychotherapy, 38(2), 171-185. Ainsworth, M.D.S. (1973). Development of infant-mother attachment. In B.M. Caldwell and H.Ricciuti (Eds.) Review of child developmental research (Vol. 3, pp. 1-94). Chicago: University of Chicago Press. Ainsworth, M.D.S., Bhehar, M.C., Waters, E. & Wall, S. (1978). Patterns of attachment: a psychological study of the strange situation. Hillsdale, NJ: Erlbaum. Allison, B. (2000). Parent-adolescent conflict in early adolescents: Research and implications fro middle school programs. Journal of Family and Consumer Sciences, Vol. 18(2), 1-6. American Psychiatric Association (1994). Eating disorders. In Diagnostic and statistical manual of mental disorders (DSM-IV) (4th ed., pp. 539-550). Arellano, C. & Markman, H. (1995). The Managing Affect and Differences Scale (MADS: A self-report measure assessing conflict management in couples). Journal of Family Psychology, 9, 319-334. Armsden, G.C. & Greenberg, M.T. (1987). The inventory of parent and peer attachment: Invidivual differences and their relationships to psychological well-being in adolescence. Journal of Youth and Adolescence, 16, 427-454. Bandura, A. (1997). Self Efficacy: The exercise of Control. New York: W.H. Freeman. Bandura, A., Babaranelli, C., Caprara, G.V.& Pastorelli, C. (1996). Multifaceted impact of self-efficacy in academic functioning. Child Development, 67, 1206-1222. Beardslee, W.R., Keller, M.B., & Kellerman, G.L. (1985). Children of parents with affective disorders. International Journal of Family Psychiatry, 6, 283-299. 116 Beier, E. G. (1966). The silent language of psychotherapy. Chicago: Aldine. Benjamin, L. S. (1974). Structural analysis of social behavior. Psychological Review, 81, 392-425. Benjamin, L.S., (1988). Short Form Intrex Questionnaire. Madison, WI: Intrex Interpersonal Institute. Benjamin, L.S. & Cushing G. (2000). SASB coding manual. Salt Lake City: University of Utah. Billings, A.G. & Moos, R.H. (1983). Comparisons of children of depressed and nondepressed parents: A social-environmental perspective. Journal of Abnormal Child Psychology, 11, 463-486. Binder, J.L., & Strupp, H.H. (1997). “Negative process”: A recurrently discovered and underestimated facet of therapeutic process and outcome in the individual psychotherapy of adults. Clinical Psychology: Science and Practice, 4, 121-139. Block, J. (1995). The child-rearing practices report (CRPR): A set of Q items for the description of parental socialization attitudes and values. Unpublished manuscript, Institute of Human Development, University of California, Berkeley. Bordin, E. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy : Theory, Research , and Practice, 16, 252-260. Bowlby, J. (1969). Attachment and Loss. New York: Basic. Bowlby, J. (1982). Attachment and loss. Vol. 1, Attachment. New York: Basic Books. Boyacioglu, G., & Savasir, I. (1995). The standardization, validity and reliability study of the Interpersonal Schema Questionnaire (ISQ) for Turkish university students. Turkish Journal of Psychotherapy, 35, 40-58. Brenner, C. (1979). Working Alliance, therapeutic alliance, and transference. The Journal of the American Psychoanalytic Association, 27, 136-158. 117 Bretherton, I. (1991). Pouring new wine into old bottles: The social self as internal working models. In M.R. Gunnar and L.A. Sroufe (Eds.), Self processes and development: The Minnesota symposia on child psychology, 26, 1-41. Hillsdale, NJ: Erlbaum. Bretherton, I., Ridgeway, D. & Cassidy, J. (1990). Assessing internal working models of attachment relationships: internal working models of attachment relationships: An attachment story completion task for three year olds. In M.T. Greenberg, K. Cicchetti, and E.M. Cummings (Eds.). Attachment in the preschool years: Theory, research and intervention (pp. 273-308). Chicago: University of Chicago Press. Bukowski, W., Hoza, B. & Boivin, M. (1993). Measuring friendship quality during preand early adolescence: The development and psychometrics of the friendship qualities scale. Journal of Social and Personal Relationships, 11, 471-485. Carnelley, K.B., Pietromonaco, P.R. & Jaffe, K. (1994). Depression, working models of others, and the relationship functioning. Journal of Personality and Social Psychology. Carson, J.L. & Parke, R.D. (1996). Reciprocal negative affect in parent-child interactions and children’s peer competence. Child Development, 67(5), 2217-2226. Carson, J.L. & Parke, R.D. (1994). Reciprocal negative affect in parent-child interactions and children’s peer competency. Child Development, 67, 2217-2226. Carson, R. C. (1969). Interaction concepts of personality. Chicago: Aldine. Castonguay, L.G., Goldfried, M.R., Wiser, S, Raue, P.J. & Hayes, A.M. (1996). Prediting the effect of cognitive therapy for depression: A study of unique and common factors. Journal of cousnluting and Clinical Psychology, 64(3), 497-504. Cassidy, J. (1988). Child-mother attachment and the self in six year olds. Child Development, 59, 121-134. 118 Cassidy, J., & Asher, S.R. (1992). Lonliness and peer relations in young children. Child Development, 63, 350-365. Chodorow, N. (1978). Mothering, object-relations, and the female oedipal configuration. Feminist Studies, 4(1), 137-158 Coady, N.F. & Marziali, E. (1994). The association between global and specific measures of the therapeutic relationship. Pyschotherapy, 31, 17-27. Ciarleglia, M.M. & Makuch, R.W. (2007). Hierarchical linear modeling: An overview. Child & Neglect, 31, 91-98. Cohn, D. (1990). Child-mother attachment of six year-olds and social competence at school. Child Development, 61, 152-162. Coleman, P.K. (2003). Perceptions of parent-child attachment, social self efficacy, and peer relationships in middle childhood. Infant and Child Development, 12, 351368. Cortina, M. (2001). Sullivan’s contributions to understanding personality development in light of attachment theory and contemporary models of the mind. Contemporary Psychoanalysis, 37, 193-237. Cowen, P.A., Cowan, C.P., Schulz, M.S., & Heming, G. (1994). Pre-birth to preschool family factors in children’s adaptation to kindergarten. In R.D. Parke & S.G. Kellman (Eds.). Exploring family relationships with other social contexts. Hillsdale, NJ: Erlbaum. Crick, N.R., Dodge, K.A., (1994). A review and reformulation of social information processing mechanisms in children’s social adjustment. Psychological Bulletin, 115, 74-101. 119 Curtis, H.C. (1979). The concept of the therapeutic alliance: Implications for the “widening of scope”. Journal of the American Psychoanalytic Association, 27, 159-192. Downy, G. & Coyne, J.C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin, 108, 50-76. Eaton, T.T., Abeles, N., & Gutfreund, M.J. (1993). Negative indicators, therapeutic alliance, and therapy outcome. Psychotherapy Research, 3(2), 115-123. Eisenberg, N., Fabes, R., Bernzweig, J., Karbon, M., Poulin, R., & Hanish, L. (1993). The relations of emotionality and regulation to preschoolers’ social skills and sociometric status. Child Development, 64, 1418-1438. Eisenberg, N., Cumberland, A., & Spinrad, T.L. (1998). Parental socialization of emotion. Psychological Inquiry, 9, 241-273 Emde, R.N., Wolf, D.P., & Oppenheim, D. (Eds.) (2003). Revealing the inner worlds of young children. Oxford, U.K.: Oxford University Press. Englisch, H., Zimmerman, P. (2002). The uniqueness of the child-father attachment r relationship: Father’s sensitive and challenging play as a pivotal variable in a 16 year longitudinal study. Social Development, 11, 307-331. Oxford: Blackwell Publishers. Foa, U.G. (1961). Convergences in the analysis of the structure of interpersonal behavior. Psychological Review, 68, 341-353. Fox-Boriosff, M. (2004). Therapists’ interpersonal schemas, the working alliance, and the resolution of ruptures. Paper Presented at the International Society for Psychotherapy Research. Rome, Italy. Fox, N.A., Kimmerly, N.L., & Schaffer, W.D. (1991). Attachment to mother/attachment to father: A meta-analysis. Child Development, 62, 210-225. 120 Friedman, L. (1969). The therapeutic alliance. International Journal of Psychoanalysis, 50, 139-153. Freud, S. (1958). The dynamics of transference. In J. Starchy (Ed.), The standard edition of the complete psychological works of Sigmund Freud (pp. 122-144). London, Hogarth Press. (Original work published 1912). Furman, W. & Buhrmester, D. (1992). Age and sex differences in perceptions of networks o personal relationships. Child Development, 63, 103-115. Gecas, V. & Schwalbe, M.L. (1986). Parental behavior and adolescent self-esteem. Journal of Marriage and the Family, 48, 37-46. Gottman, J.M., Katz, L.F., Hooven, C. (1997). Meta-emotion: How families communicate emotionally. Mahwah: NJL, Lawrence Erlbaum Associates, Inc. Grossman, K., Grossman, K.E., Fremmer-Bombik, E., Heinz Kindler, H., Scheurer Greenson, R.R. (1965). The working alliance and the transference neuroses. Psychoanalysis Quarterly, 34, 155-181. Hanly, C. (1992). Reflections on the place of the therapeutic alliance in psychoanalysis. International Journal of Psychoanalysis, 75, 457-467. Harkness, S. & Super, C. (1992). The cultural foundations of fathers’ roles: Evidence from Kenya and the United States. In B. Hewlitt (Ed.), Father-child relations: Cultural and biosocial contexts. (pgs. 191-211). New York: Aldine de Gruyter. Hartley, D.E., & Strupp, H.H. (1983). The therapeutic alliance: Its relationship to outcome in brief psychotherapy. In J. Masling (Ed.), Empirical studies of psychoanalytical theories, (Vol. 1, pp. 1-37). Hillsdale, NJ: Analytic Press. Henry, W.P., Schacht, T.E. & Strupp, H.H. (1986). Structural analysis of social behavior: Application to a study of interpersonal process in differential psychotherapy outcome. Journal of Consulting and Clinical Psychology, 54(1), 27-31. 121 Henry, W.P., Schacht, T.E., & Strupp, H.H. (1990). Patient and therapist introject, interpersonal process, and differential psychotherapy outcome. Journal of Consulting and Clinical Psychology, 58, 768-774. Henry, W.P., & Strupp, H.H. (1994). The therapeutic alliance as an interpersonal process. In A.O. Horvath and L.S. Greenberg (Eds.). The working alliance: Theory, Practice and Research. New York: Wiley and sons. Hewlett, B. (Ed.). (1992). Father-child relations: Cultural and biosocial contexts. New York: Aldine de Gruyter. Hill, C., & Safran, J. (1994). Assessing interpersonal schemas: Anticipated responses of significant others. Journal of Social and Clinical Psychology, 13, 366-379. Horvath, A.O., & Luborsky, L. (1993). The role of the therapeutic alliance in psychotherapy. Journal of Consulting and Clinical Psychology, 61, 561-573. Horvath, A.O., & Symonds, B.D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139-149. Isley, S.L., O’Neil, R., Clatfelter, D. & Parke, R.D. (1999). Parent and child expressed affect and children’s social competence: Modeling direct and indirect pathways. Developmental Psychology, 35, 547-560. Jacob, T. & Johnson, S.L. (2001). Sequential Interactions in the parent-child communications of depressed fathers and depressed mothers. Journal of Family Psychology, 15 (1), 38-52. Kern, K.A. & Barth, J. (1995). Parent-child attachment and physical play: Convergence across components of parent-child relationships and their relationships competence. Journal of Social and Personal Relationships, 12, 243-260. 122 Kerns, K.A., Klepac, L. & Cole, A.K., (1996). Peer relationships and pre-adolescents’ perceptions of security in the mother-child relationship. Developmental Psychology, 32. 457-466. Kiesler, D.J., & Watkins, L.M. (1989). Interpersonal complementarity and the therapeutic alliance: A stdy of relationship in psychotherapy. Psychotherapy, 26, 183-196. Kiesler, D. (1996). Some myths of psychotherapy research and the search for a paradigm. Psychological Bulletin, 65, 110-136. Klee, M.R., Abeles, N., & Muller, R.T. (1990). Therapeutic alliance: Early indicators, course, and outcome. Psychotherapy, 27, 166-174. Kohut, H. (1971). The analysis of the self. New York: International Universities Press. Kohut, H. (1977). The restoration of the self. New York: International Universities Press Ladd, G.W. & Ladd, B.K. (1988). Parenting behaviors and parent-child relationships: Correlates of peer victimization in kindergarten. Developmental Psychology, 34, 1450-1458. Lamb, M.E. (1975) Fathers: Forgotten contributors to child development. Human Development, 5, 131-141. Laursen, B. (1995). Conflict and social interaction in adolescent relationships. Journal of Research on Adolescence, 5, 55-70. Laursen, B., Coy, K.C., & Collins, W.A. (1998). Reconsidering changes in parent-child conflict across adolescence: A meta-analysis. Child Development, 69, 817-832. Leary, T. (1957). Interpersonal diagnosis of personality. New York: Ronald Press. Leventhal, H. (1984). A perceptual-motor theory of emotion. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 117-182). New York: Academic. Lieberman, M., Doyle, A.B. & Markiewicz, D. (1999). Developmental patterns in security of attachment to mother and fathering late childhood and early 123 adolescence: Associations with peer relations. Child Development, 70(1), 202213. Luborsky, L., Singer, B., & Luborsky, L. (1975). Comparative studies of psychotherapies; “Is it true that everybody has won and all must have prizes?” Archives of General Psychiatry, 32, 995-1008, Lynn, D. B., & Sawyer, W. L. (1959). The effects of father absence on Norwegian boys and girls. Journal of Abnormal and Social Psychology, 59, 258–262. Ma, H.K., Shek, D.T.L., Cheung, P.C. & Lee, R.Y.P. (1996). The relation of prosocial and antisocial behavior to personality disorder and peer relationships of Hong Kong Chinese adolescents. Journal of Genetic Psychology, 157, 255-356. Main, M., Goldwyn, R. & Hesse, E. (2003). Adult attachment interview scoring and classification system (Version 7.2). Unpublished manuscript, Department of Psychology, University of California, Berkeley. Main, M., Kaplan, N. & Cassidy, J. (1985). Security in infancy, childhood and adulthood: A move to the level of representation. In Bretherton and E. Waters (Eds.). Growing points in attachment theory and research (pp. 66-106). Monographs of the Society for Research in Child Development, 50(1-2, Serial No. 209). Main, M. & Westin, D.R. (1981). The quality of the toddler’s relationship to mother and father: Related conflict behavior and the readiness to establish new relationships. Child Development, 52, 932-940. Mallinckrodt, B., Coble, H. & Gantt, D. (1995). Working Alliance, Attachment Memories, and Social Competencies of Women in Brief Therapy. Journal of Counseling Psychology. Vol 42 (1),79-84 124 Marchand, J.M. & Hock, E. (2000). Avoidance and attacking conflict resolution strategies among married couples: Relations to depressive symptoms and marital satisfaction. Family Relationships, 49, 201-206. Marmar, CR., Weiss, D.S., & Gaston, L. (1989). Toward the validation of the California therapeutic alliance rating system. Journal of Consulting and Clinical Psychology, 1, 46-52. Marsh, H. W., & Bailey, M. (1991). Confirmatory factor analyses of multitraitmultimethod data: A comparison of alternative models. Applied Psychological Measurement, 15, 47–70. McDowell, D.J. & Parke, R.D. (2003). Differences between mothers’ and father’ advice giving style and content: Relations with social competence and psychological functioning in middle childhood. Merril-Palmer Quarterly, 49(1), 55-76. McDowell, D.J., Parke, R.D. & Spitzer, S. (2002). Parent and child cognitive representations of social situations and children’s social competence. Blackwell Publishers, MA. McDowell, D.J. & Parke, R.D. (2000). Differential knowledge of display rules for positive and negative emotions: Influences from parents, influences on peers. Social Development, 9, 415-432. McDowell, D.J. & Parke, R.D. (2005). Parental control and affect as predictors of children’s display rule use and socail competence with peers. Child Development, 14 (3), 440-457. Mitchell, S.A., & Aron, L. (1999). Relational psychoanalysis: The emergence of a tradition. Hillsdale, NJ: Analytic Press. 125 Montenmayer, R. (1982). The relationship between parent-adolescent conflict and the amount of time adolescents spend alone with parents and peers. Child Development, 53, 1512-1519. Mott, F. L. (1994). Sons, daughters and fathers' absence: Differentials in father-leaving probabilities and in home environments. Journal of Family Issues, 15, 97–128. Muran, J.C., Safran, J.D., Gorman. B.S., Samstag, L.W., Eubanks-Carter, C. & Winston, A. (in press). Therapeutic alliance ruptures and their resolution in the early phase of three brief psychotherapies for personality disorders. Psychotherapy: Theory, Research, Practice, Training. Murphy, L.B. (1997). Fathers. Zero to Three, 18 (1), 9. Muthén, L.K., & Muthén, B.O. (1998-2007). Mplus user’s guide, 5th Edition. Los Angeles, CA: Muthén & Muthén. Najavits, L.M., & Strupp, H.H. (1994). Differences in the effectiveness of psychodynamic psychotherapies: A process-outcome study. Psychotherapy, 31(1), 114-123. Nelson, G.M. & Beach, S.R. (1990). Sequential interaction in depression: Effects of depression behavior on spousal aggression. Behavior Therapy. 21, 167-182. Nelson, L. (2002). Predicting therapist hostility: Therapist introjects, measured by the ISQ, and ability to resist hostile process in-session. Unpublished Dissertation. O'Neil, R., & Parke, R. D. (2000). Family-peer relationships: The role of emotion regulation, cognitive understanding, and attentional processes as mediating 126 processes. In K. Kerns, J. Contreras, & A. Neal-Barnett (Eds.), Family and peers: Linking two social worlds (pp. 195–225). Westport, Connecticut: Praeger. Pakaslahti, L., Asplund-Peltola, R.T. & Keltikangas-Jarvinen, L. (1996). Parents’ social problem-solving strategies in families with aggressive and non-aggressive boys. Aggressive Behavior, 22, 345-356. Parke, R.D., Burks, V.M., Carrson, J.L., Neville, B. & Boyum, L.A. (1994). Family-peer relationships: A tripartite model. In R.D. Parke & S.G. Kellam (Eds.), Exploring family relationships with other social contexts (pp. 115-145). Hillsdale, NJ: Lawrence Erlbaum. Parke, R.D. (1995). Fathers and families. In M.H. Bornstein (Ed.), Handbook of parenting, Vol. 3, Status and social conditions of parenting (pp. 24-64). Hillsdale, NJ: L.Erlbaum. Parke, R.D., Ladd, G.W. (1992). Family-peer relationships: Modes of linkage. Hillsdale, NJ. Lawrence Erlbaum. Parke, R.D., MaDonald, K.B., Burks, V.M., Bhavanagri, N., Barth, J.M. & Beital, A. (1989). Family and peer systems: In search of the linkages. In Kreppner K. & Lerner R.M. (Eds.), Family Systems and Lifespan Development. New Jersey: Lawrence Erlbaum Associates. (pp. 65-104). Pincus, A.L., Newes, S.L., Newes, K.A., Ruiz, M.A. (1998). A comparison of three indexes to assess the dimensions of structural analysis of social behavior. Journal of Personality Assessment, 70(1), 145-170. Reese-Weber, M. & Marchand, J.F. (2002). Family and individual predictors of late adolescents’ romantic relationships. Journal of Youth and Adolescence, 31(3), 197-206. 127 Ricaud-Droisy, H. & Zaouche-Gaudron, C. (2003). Interpersonal conflict resolution strategies in children: A father-child co-construction. European Journal of Psychology of Communication, 18 (2), 157-169. Roberts, W. (June, 1994). The socialization of emotional expression: Relations with competence in preschool. Paper presented at the meeting of the Canadian Psychological Association, Penticton, British Columbia. Roberts, W. (1999). The socialization of emotional expression: Relations with prosocial behavior and competence in five samples. Canadian Journal of Behavioral Science, 31(2), 77-85. Roberts, W. & Strayer, J. (1996). Parents’ responses to the emotional distress of their children: Relations with children’s competence. Developmental Psychology, 23, 415-422. Roberts, W. and Strayer, J. (1987). Parents’ responses to the emotional distress of their children: Relations with children’s competence. Developmental Psychology, 23, 415-422. Robin, A.L. & Foster, S.L. (1989). Negotiating parent-adolescent parent-adolescent conflict. New York: The Guilford Press. Saarni, C. (1979). Children’s understanding of display rules for expressive behavior. Developmental Psychology, 15, 424-429. Safran, J.D. (1990). Towards a refinement of cognitive therapy in light of interpersonal theory. Clinical Psychology Review, 10, 87-105 Safran, J. D., Segal, Z. V., Hill, C., & Whiffen, V. (1990). Refining strategies for research on self-representations in emotional disorders. Cognitive Therapy and Research, 14, 143-160. 128 Safran, J. D. (1998). Widening the scope of cognitive therapy: The therapeutic relationship, emotion, and the process of change. Norvale, New Jersey: Jason Aronson. Safran, J.D. & Muran J.C. (2000). Resolving therapuetic alliance ruptures: Diversity and integration. In Session: Psychotherapy in Practice, 56(2), 233-243. Samstag, L.W., Batchelder, S.T., Muran, J.C., Safran, J.D., & Winston, A. (1998). Early identification of treatment failures in short-term psychotherapy: An assessment of therapeutic alliance and interpersonal behavior. Journal of Psychotherapy Practice & Research, 7, 126-143. Sandler, J. Holder, A., Kawenoka, M., Kennedy, H., & Neurath, L. (1969). Notes on some theoretical and clinical aspects of transference. International Journal of Psychoanalysis, 3, 43-47. Schmaling, K.B. & Jacobson, N.S. (1990). Marital interaction and depression. Journal of Abnormal Psychology, 99, 229-236. Schore, A.N. (2001). The effects of early relational trauma on right brain development, affect regulation, and infant mental health. Infant Mental Health Journal, 22, 201269. Schore, A. N. (2003). Affect regulation and the repair of the self. New York: Norton. Sexton, H.C., Hembre, K., & Kvarme, G. (1996). The interaction of the alliance and therapy microprocess: A sequential analysis. Journal of Consulting and Clinical Psychology, 64(3), 471-480. 129 Shek., D. & Ma, H. K. (2001). Parent-adolescent conflict and adolescent antisocial and prosocial behavior: A longitudinal study in a Chinese context. Adolescence, 36(143), 545-555. Soygut, G., & Savasir, I. (2001). The relationship between interpersonal schemas and depressive symptomatology. Journal of Counseling Psychology, 48, 359-364. Sroufe, A. (1979). Socioemotional development. In J. Osofsky (Ed.), Handbook of infant development, (pp. 462-516). New York: Wiley. Steele, H. & Steele, M. (2005). Understanding and resolving emotional conflict: The London parent-child project. In K.E. Grossmann and K. Grossmann (Eds.), Attachment from Infancy To Adulthood. (pp.137-163). New York: Guilford Press. Steele, H., Steel, M. (2004). The construct of coherence as an indicator of attachment security in middle childhood: The friends and family interview. In K.A. Kerns and R.A. Richardson (Eds.), Attachment in middle childhood (pp. 137-160). New York: Guilford Press. Stern, D. N. (1985). The interpersonal world of the infant. New York: Basic Books. Stile, W.B., Shaprio, D., & Elliot, R. (1986). Are all psychotherapies equivalent? American Psychologist, 41, 165-180. Stolz, L. M. (1954). Father relations of war-born children. Stanford, CA: Stanford University Press. Strupp, H.H. (1960). Nature of psychotherapist’s contribution to treatment. Archives of General Psychiatry, 37, 595-603. Strupp, H.H. & Williams, J.V. (1960). Some determinants of clinical evaluations of different psychiatrists. Archives of General Psychiatry, 2, 434-440. Strupp, H.H. (1980). Success and failure in time-limited psychotherapy. Archives of General Psychiatry, 37, 595-603. 130 Strupp, H.H. (1982). The outcome problem in psychotherapy: Contemporary perspectives. H. Havery and M.M. Peeks (Eds.) Psychotherapy research and behavior change (pp.39-71) Washington, D.C. American Psychological Association. Suess, G.J., Grossman & K.E., Sroufe, L.A. (1992). Effects of infant attachment to mother and father on quality of adaptation in preschool: From dyadic to individual organization of self. International Journal of Behavioral Development, 15(1), 4365. Suh, C.S., O’Malley, S., & Strupp, H.H. (1986). The Vanderbilt psychotherapy process scale. In L. Greenberg & W. Pinsof, (Eds.), The Psychotherapeutic process: A Research Handbook. Sullivan, H. S. (1940). Conceptions of modern psychiatry. New York: Norton. Sullivan, H.S. (1940). Conceptions of modern psychiatry. New York: Norton Sullivan, H.S. (1953). The interpersonal theory of psychiatry. New York: Horton. Sommerfeld, E., Orbach, I., Zim, S., & Mikulincer, M. (2008). An in-session exploration of ruptures in working alliance and their associations with clients’ core conflictual relationship themes, alliance-related discourse, and clients’ postsession evaluations. Psychotherapy Research, 18, 377-388. Swann, W., Jr., & Reade, S. J. (1981). Acquiring self-knowledge: The search for feedback that fits. Journal of Personality and Social Psychology, 41, 1119-1128. Tasca, G.A. & McMullen, L.M. (1992). Interpersonal complementarity and antitheses within a stage model of psychotherapy. Psychotherapy, 29, 515-523. Thompson, C. (1964). Interpersonal Psychoanalysis (ed. M. Green). New York: Basic Books. 131 Tremblay, R.E., Vitaro, F., Gagnon, C., Piche & C., Royer, N. (1992). A prosocial scale for the preschool behavior questionnaire: Concurrent and predictive correlates. International Journal of Behavior Development, 15, 227-245. Tronick, E. Z. (1998). Dyadically expanded states of consciousness and the process of therapeutic change. Infant Mental Health Journal, 19, 290-299. Verschueren, K. & Alfons, M. (1999). Representations of self and socioemotional competence in kindergartners: Differential and combined effects of attachment to mother and to father. Child Development, 70(1), 183-201. Wei, D.L., Vogel, T.K.&f Zakalik, R. (2005). Adult attachment, affect regulation, negative mood, and interpersonal problems: The mediating roles of emotional reactivity and emotional cutoff. Journal of Counseling Psychology, 52, 14-24. Wiggins, J. S. (1982). Circumplex models of interpersonal behavior in clinical psychology. In P. C. Kendall & J.N. Butcher (Eds.), Handbook of research methods (pp. 183-221). New York: Wiley Interscience. Zetzel, E.R. (1956). Current concepts of transference. International Journal of Psychoanalysis, 37, 369-376. Zetzel, E.R. (1966). The analytic situation. In R.E. Litman (ed.), Psychoanalysis in America (pp.86-106). New York: International University Press. 132 Appendix A: Interpersonal Circumplex Figure A1: The 1982 Interpersonal Circumplex: Depicting 16 Categories of Interpersonal Behavior Figure A2: The 1982 Interpersonal Circumplex: Depicting 2 Levels of Interpersonal Behavior 133 Figure A1 The 1982 interpersonal circumplex circle: Depicting 16 categories of interpersonal behavior. Used with permission from D. Kiesler, retrieved August 8, 2008 from http://www.vcu.edu/sitar/1982basiccircle.jpg 134 Figure A2 Figure A2. The 1982 interpersonal circumplex circle: Depicting 2 levels of interpersonal behavior. Used with permission from D. Kiesler, retrieved August 8, 2008 from http://www.vcu.edu/sitar/1982circ.jpg 135 Appendix B: Contents of ISQ SITUATIONS (1982 Interpersonal Circle letters in parentheses) 1 (A) Imagine that you and your _______ are collaborating on something. You have more knowledge and expertise in this area than your _______, so you take the lead in making decisions. 2 (E) Imagine yourself feeling angry and argumentative towards your _______. 3 (I) Imagine yourself feeling weak or passive and wanting your _______ to take the lead. 4 (M) Imagine yourself being friendly and helpful with your _______. 5 (B) Imagine yourself in a game (tennis, scrabble, etc.) with your _______. You act very competitive and work hard to win the game. 6 (F) Imagine yourself being preoccupied with your own thoughts and detached with your _______. 7 (J) Imagine yourself in an unmotivated or lazy mood where you feel like just going along with whatever your _______ is doing. 8 (N) Imagine yourself expressing genuine interest and concern for your _______. 9 (C) Imagine a situation where you feel that your _______ has disappointed you. 10 (G) Imagine yourself in a serious mood where you are reserved and not sociable with your _______. 11 (K) Imagine yourself confiding in your _______ something that is important to you. 12 (O) Imagine feeling uninhibited and spontaneous with your _______. 13 (D) Imagine that you have had a terrible day and are feeling peeved off with the whole world. You are definitely not feeling affectionate or cordial toward anyone. 14 (H) Imagine feeling not very confident or sure of yourself and feeling dependent on your _______. 15 (L) Imagine yourself feeling warm and affectionate towards your _______. 16 (P) Imagine yourself acting independently and confidently about something you have never done before, and not feeling that you need assistance from _______. RESPONSES (1982 Interpersonal Circle letters in parentheses) A B C D E F G H (PA) (BC) (DE) (FG) (HI) (JK) (LM) (NO) Would take charge, or try to influence me. Would be disappointed, resentful, or critical. Would be impatient, or quarrelsome. Would be distant, or unresponsive. Would go along with me, or act unsure. Would respect me, or trust me. Would be warm, or friendly. Would show interest, or let me know what he/she thinks. 136 CODING OF THE ISQ Responses are recoded to reflect the amount of control and affiliation each represents. ISQ RESPONSE OCTANT A B C D E F G H Controlling Mistrustful Hostile Distant Submissive Trusting Friendly Interested CONTROL .875 .625 .125 -.375 -.875 -.625 -.125 .375 137 AFFILIATION .125 -.375 -.875 -.625 -.125 .375 .875 .625 Appendix C Table 13: Goodness of Fit Statistics of Models for Father and Mother ISQ Index Scores Table 14: Factor Loadings for the Correlated Uniqueness Confirmatory Factor Analysis 138 Preliminary Analysis: ISQ factor Analysis The ISQ consists of 16 situations which are often collapsed into four quadrants, or types of situations (hostile, friendly, submissive and dominant) and rated along three dimensions (desirability, affiliation and control) for mother and father, thus yielding 24 independent variables. Preliminary analyses were conducted to address the unwieldy task of analyzing the effects of this many variables, to determine ways to reasonably reduce ISQ variables, and better understand psychometric properties of mother and father factors, the three ISQ indices, and 16 different situations. Two confirmatory factor analyses were conducted in the statistical program Mplus (Muthén & Muthén, 19982007) to assess goodness of fit of a correlated uniqueness model (Marsh & Bailey, 1991). In this correlated uniqueness model, ratings of desirability in hostile, friendly, submissive and dominant situations all loaded on a desirability factor. Similarly, ratings of affiliation and control across situations loaded of affiliation and control factors, respectively. Correlations among observed variables pertaining to one situation were freely estimated (e.g., correlations among desirability, affiliation, and control in hostile situations were freely estimated). Analyses yielded mixed results for overall goodness of fit for mother and father, with better results for the father (refer to table 13). The chi square statistic is significant for both mother (x² = 92.75, p < .01) and father (x² = 71.36, p < .01) indicating a statistically significant discrepancy between the observed and model-implied covariance matrices; however, the CFI suggests marginally adequate fit for the mother 139 (CFI = 0.928) and adequate fit for the father (CFI = 0.963). The RMSEA statistic for father is also reasonable (RMSEA = 0.093). As illustrated in table 14, freely estimated factor loadings for both the mother and the father were quite high. For example, all four quadrants had strong positive loadings on their respective factors (desirability, affiliation, control) for both mother and father. In addition, coefficient alpha was calculated for each dimension across all 16 situations, confirming that each rating had adequate internal consistency (Table 14). Desirability and affiliation factors were highly correlated for both mother (r=.91, p < .001) and father (r=.83, p < .001) regardless of the kind of situation they were being measured in. For example, therapists who anticipated high degrees of affiliation from their mothers or fathers in hostile, friendly, submissive or dominant situations were likely to rate the response as highly desirable. This finding suggests that desirability and affiliation may be redundant and that it is superfluous to assess both of these dimensions. Given the overlap of desirability and affiliation and because desirability was more reliable than control, desirability ratings were the focus of this study. Although the overall goodness of fit statistics for the correlated uniqueness model were not strong, given the small sample size, CFI, RMSEA, and coefficient alpha statistics (Table 13), no attempt was made to undertake a data-driven search for a more complex model with better fit. On the whole, these findings suggest the reliability of desirability, affiliation, and control items is adequate for research purposes, but it may be useful to improve the measurement of interpersonal schemas, particularly with the mother, in future work. 140 Confirmatory Factor Analysis of the ISQ Table 13: Goodness of Fit Statistics of Models for Father and Mother ISQ Index Scores (N = 96) = Model 2 2(df) p CFI RMSEA 90%CI Mothers 92.75 39 0.00 .928 .120 .089-.151 Fathers 71.36 39 0.001 .963 .093 .058-.127 Table 14: Factor Loadings for the Correlated Uniqueness Confirmatory Factor Analysis Model(N = 96) Mother Estimate SE Est. / SE p Hostile 0.825 0.046 17.912 0.00 Friendly 0.720 0.058 12.494 0.00 Submissive 0.870 0.040 21.606 0.00 Dominant 0.813 0.049 16.513 0.00 Hostile 0.750 0.062 12.072 0.00 Friendly 0.550 0.082 6.689 0.00 Submissive 0.677 0.070 9.706 0.00 Dominant 0.755 0.063 12.012 0.00 Hostile 0.664 0.135 4.931 0.00 Friendly 0.419 0.155 2.700 0.00 Submissive 0.683 0.129 5.284 0.00 Desirability ( = .880) Affiliation ( = .703) Control ( = .636) 141 Dominant 0.507 0.100 5.088 0.00 Estimate SE Est. / SE p Hostile 0.856 0.034 25.342 0.00 Friendly 0.713 0.052 13.803 0.00 Submissive 0.869 0.032 27.336 0.00 Dominant 0.883 0.033 26.921 0.00 Hostile 0.799 0.049 16.231 0.00 Friendly 0.586 0.067 8.774 0.00 Submissive 0.717 0.055 13.142 0.00 Dominant 0.834 0.054 16.549 0.00 Hostile 0.658 0.074 8.853 0.00 Friendly 0.468 0.089 5.289 0.00 Submissive 0.668 0.084 7.924 0.00 Dominant 0.734 0.071 10.335 0.00 Father Desirability ( = .903) Affiliation ( = .813) Control ( = .735) Note: Cronbach’s alpha ( ) represents reliability of 16 ISQ situations/items for the http://www.theonion.com/content/news_briefs/ahmad_bradshaw_still_hadgiven index. 142