3:19:36 w: 3.x :1: _. t: I: 1 .3 ‘ .r L522: . Q 3 v» a h... ; fag @& a , 2D. A.‘ IN; .n , 1‘ - 4| ‘- . kl?- . 93%.... 4.4.. 3...». . x. 3W": .8. hunk... w.“ .t 5.. h... 3.6.1: . 333".) .3... I «saw. 33 . r. 2km“, x2: 3.. 1.... kme .3, warm 0,? anng it 546632qu LIBRARY Michiga. ’ State University This is to certify that the dissertation entitled RELATIONS OF ADULT ATTACHMENT ORIENTATIONS TO EXPECTATIONS AND PERCEPTIONS OF GROUP CLIMATE AND GROUP THERAPIST presented by James W. Wyssmann -‘---.-u.-.---‘-t- gin--oOn-.-<----.--~-----.-i— has been accepted towards fulfillment of the requirements for the Ph.D. degree in Counseling, Educational Psychology and Special Education Major Professor’ 5 Signature 0 5/0}/& 3 Date MSU is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJCIRC/DateDuepss-p. 15 RELATIONS OF ADULT ATTACHMENT ORIENTATIONS TO EXPECTATIONS AND PERCEPTIONS OF GROUP CLIMATE AND GROUP THERAPIST By James W. Wyssmann A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology, and Special Education 2003 ABSTRACT RELATIONS OF ADULT ATTACHMENT ORIENTATIONS TO EXPECTATIONS AND PERCEPTIONS OF GROUP CLIMATE AND GROUP THERAPIST By James W. Wyssmann This research explored the relationships between group therapy members’ adult attachment orientations and expectations of group therapy, perceptions of group climate, and perceptions of group leaders. Participants included 98 entering members of therapy groups who completed the Experiences in Close Relationships questionnaire and Expectations About Counseling Brief Form before the group sessions started. A total of 80 members who completed four sessions of group therapy also completed the Group Climate Questionnaire Short Form and the Trainer Behavior Scale. Hypotheses linking dimensions of adult attachment with dimensions of expectations and perceptions of group climate and leaders were formulated from attachment theory. Correlation and hierarchical regression analyses revealed no significant findings. Implications of the lack of support for the research hypotheses and future directions are discussed. Acknowledgments This dissertation is dedicated to the memory of my Grandmother Mary Bargfrede, Aunt Vemita Holsten, Uncle Theodor Holsten and my big black dog—Buster McKinley the Intrepid. “If I have seen further, it is by standing upon the shoulders of giants” Sir Isaac Newton, February 5, 1675 Those giants are: My mother Ruthleen and father Robert who have been unending and lifelong inspirations and supports. 1 cannot tell you how proud I am to be your son.. Deanne, for understanding when I had to be gone to complete this thing. With you my world is complete. I Love You. My sister Mary and brother Jeff who believed in me when I’d lost the ability to. Dr. Fred Lopez, my advisor, who served as mentor, coach, drill sergeant, and fi'iend and saw this through to the end with me when you didn’t have to. Thanks. Dorothy Updike for countless hours of readings, revisions, formatting and suggestions and never tiring of my writing when I couldn’t stand it. David “Bud” Updike for always asking how far I’d gotten. Brian Krylowicz, a friend who always encouraged and grounded me during the entire process. Thanks for your wisdom. Bertram Stoffelrnayr, my mentor and friend. If I am half the psychologist and friend that you are I will be brilliant and widely loved. Carter Rees (my stats guy) and Nicole Whitlock. . .may you have years and years of love and life’s adventures. Drs. Steve Ross, Lauren Weitzman, Robert Gleave, and Andrew Turner for working with me on data collection. The many brave and resilient people who not only agreed to participate in this research, iii but took the risk to share their lives with us. My brother Gene and sisters Kathy and Amy for setting the bar high. Tawa Sina, Fred and Julie Snelson, Charlie Davis, Nancy Wright, and Cathryn Smith for walking through this life with me as friends. The guys Jon and Zach, who don’t care if I’m Dr. Wyssmann or not. Dr. Dennis Kivlighan for inspiration on how to be a great researcher and a good man. The members of my committee Dr. Ken Rice, Dr. Jean Baker, Dr. Steve Yelon, Dr. Tawa Sina and Dr. Robert Lee for their contributions and patience. Dr. David Wenger for accompanying me through the loneliest year of my life. Dr. Hope Hills for writing on one of my first research papers in my Master’s program “I think you should go for your Ph.D.” I didn’t think I could do it at the time. My classmates Judy Ferris, Eric Sauer, Karyn Boatwright, Kim Thomas and Ellie Berger for the years of laughter together. And God. . .who inspired every concept, emotion, and keystroke I poured into this project. iv TABLE OF CONTENTS LIST OF TABLES .......................................................................................................... INTRODUCTION AND CONTEXT FOR THE PRESENT STUDY ix Group Therapy and Managed Care .............................................................................. Eflectiveness of Group versus Individual Therapy ....................................................... Attachment Theory as a Framework for Understanding Important Group Processes and Outcomes ......................................................................................................... Group Therapy and Interpersonal Demands ............................................................... Main Questions of the Proposed Study ........................................................................ Summary ...................................................................................................................... Problem Statement ....................................................................................................... Literature Review ......................................................................................................... Introduction. Individual Differences and Expectations about Therapy. Impact of Individual Differences on Group Processes and Climate. Impact of Individual Differences on Perceptions of Group Leaders. Attachment Related Differences Relevant to Group Therapy ...................................... Summary. METHOD Design .......................................................................................................................... Participants .................................................................................................................. .1 .2 .4 10 12 12 14 16 l6 17 22 3O 34 42 45 45 46 Procedures ................................................................................................................... Measures ...................................................................................................................... Demographic Questionnaire (Appendix D). Experiences in Close Relationships (Brennan, Clark, & Shaver, 1998; Appendix E). Expectations About Counseling-Brief Form (Tinsley 1982; Appendix F). Trainer Behavior Scale (Bolman, 1971; Appendix G). Group Climate Questionnaire—Short Form (MacKenzie, 1983; Appendix H). Group Satisfaction Scale (Lopez & Wyssmann, 2000; Appendix I). Research Hypotheses ................................................................................................... Hypothesis 1. Hypothesis 2. Hypothesis 3. Hypothesis 4. Hypothesis 5. Hypothesis 6. Hypothesis 7. Hypothesis 8. Analyses ....................................................................................................................... RESULTS Introduction ................................................................................................................. Summary of Demographic Data for Initial Research Participants across Sites ......... vi 46 47 48 48 49 50 52 53 53 53 53 54 54 54 54 54 54 55 56 56 57 Comparison of demographic data and responses to dimensions of the independent and dependent variables across participating sites .............................................. 60 Comparison of demographic data of therapy continuers and therapy dropouts. ........ 61 Comparison of therapy continuers and therapy dropouts on dimensions of adult attachment and expectations about counseling ..................................................... 63 Relationship of Dimensions of Adult Attachment to Perceptions of Group Leader Behavior, Group Climate, and Expectations About Counseling ........................... 64 Reliability issues related to the dependent measures .................................................. 65 Hypotheses regarding the relationship of adult attachment orientations to perceptions of group leader behavior ................................................................... 66 Hypotheses regarding the relationship of adult attachment orientations to perceptions of group climate ................................................................................ 66 Hypotheses regarding the relationship of adult attachment orientations to expectations about group counseling ................................................................... 67 Post hoc analyses ......................................................................................................... 6 7 DISCUSSION 84 Introduction ................................................................................................................ 84 Research findings regarding major hypotheses of the study ....................................... 84 Reliability concerns with dependent measures ............................................................ 85 General discussion ....................................................................................................... 87 Adult attachment orientation and perceptions of group leader behavior ................... 88 Adult attachment orientations and perceptions of group climate ................................ 89 Adult attachment orientations and expectations about group therapy ........................ 94 Other findings .............................................................................................................. 97 Strengths of the study ................................................................................................. 10] Limitations ................................................................................................................. 1 0] vii Future directions ........................................................................................................ I 02 APPENDICES 110 APPENDIX A ............................................................................................................. I 11 Recruitment Script 1 11 APPENDIX B ............................................................................................................. [[3 Participant Response Form 113 Appendix C ................................................................................................................. 1 15 Participant Consent Form 1 15 APPENDIX D ............................................................................................................ 1 l 8 Demographic Questionnaire 118 APPENDIX E ............................................................................................................. 121 Experiences in Close Relationships 121 APPENDIX F ............................................................................................................. 124 Expectations About Counseling—Brief Form 124 APPENDIX G ............................................................................................................ [30 Trainer Behavior Scale 130 APPENDIX H ............................................................................................................ 133 Group Climate Questionnaire—Short Form 133 APPENDIX I .............................................................................................................. 135 Group Satisfaction Scale 135 REFERENCES 137 viii LIST OF TABLES Table 1 ......................................................................................................................... 72 Demographic Comparison of Participating Sites 72 Table 2 ......................................................................................................................... 76 Intercorrelation Matrix 76 Table 3 ......................................................................................................................... 78 Summary of Hierarchical Regression Analysis for Variables Predicting Trainer Behavior Scale Affiliativeness (N = 80) 78 Table 4 ......................................................................................................................... 79 Summary of Hierarchical Regression Analysis for Variables Predicting Engagement from the Group Climate Questionnaire—Short form (N = 80) 79 Table 5 ......................................................................................................................... 80 Summary of Hierarchical Regression Analysis for Variables Predicting Conflict from the Group Climate Questionnaire—Short form (N = 80) 80 Table 6 ......................................................................................................................... 81 Summary of Hierarchical Regression Analysis for Variables Predicting Commitment from the Expectations About Counseling—Brief Form (N = 80) 81 Table 7 ......................................................................................................................... 82 Summary of Hierarchical Regression Analysis for Variables Predicting F acilitative Conditions from the Expectations About Counseling—Brief Form (N = 80) 82 Table 8 ......................................................................................................................... 83 ix Summary of Hierarchical Regression Analysis for Variables Predicting Nurturance from the Expectations About Counseling—Brief Form (N = 80) 83 INTRODUCTION AND CONTEXT FOR THE PRESENT STUDY Group Therapy and Managed Care These are tumultuous times for health care providers in the United States. Managed care companies have been increasing competition and demanding more therapeutic effectiveness from all providers of health services. In light of the major changes shaping the workplace of health professionals, it is important for those of us who perform “talk therap)?’ to provide evidence that our approaches are cost effective when used in conjunction with or as an alternative to pharmacotherapy. Therefore, it will be crucial that clinicians consider forms of therapy other than the typical one on one counseling session. One form of therapy gaining in importance because of these managed care forces is group psychotherapy (Brabender, 2001; MacKenzie, 1995; MacKenzie, 2001; Spitz, 2001; Taylor et a1, 2001). Piper and Joyce (1996) maintained that managed care has increased the emphasis on providing short-term treatment and, since short-term group therapies are among the most efficient in terms of the use of therapist time, they expect to see more emphasis on group therapy as a primary treatment modality. Riva and Smith (1997) speculated that, due to pressure from insurance companies, consumers, and government agencies the resultant movement toward more brief treatments would make the use of group therapy more prevalent and would shape practice and research on groups. It is clear that, if managed care continues to shape psychotherapeutic treatment, group therapy is going to become more prominent in our clinical work. This will result in a major shift in the way therapists train and practice in coming years, with a resultant emphasis on group therapy and the acquisition of group facilitation skills. In fact, Mackenzie (2001) predicted that practitioners in the future would specialize in group work, their main professional identity being that of a group practitioner. He also believed that group psychotherapy utilization would increase, while individual psychotherapy might be limited to select patients. Helfrnann (1994) noted that with this increased emphasis on group treatment, there comes a responsibility to conduct more research on group therapy processes and outcomes. Horne and Rosenthal (1997) acknowledged that group research had evolved from its roots in the early twentieth century, but they concluded that there was still much to be done. The present study is intended to be part of that effort. Effectiveness of Group versus Individual Therapy Much of the impetus for the increased emphasis on group work was and continues to be due to the comparative inefficiency of individual therapy. Managed care companies have expressed concern over whether individual therapy can be justified when a therapist can see so many more clients in a group. A number of researchers have supported this assertion. For example, Smith, Glass, and Miller (1980) in their classic meta-analysis, compared the efficacy of group therapy with individual therapy by reporting an average effect size of 0.83 for group therapy and 0.87 for individual therapy. In a meta-analysis of 23 outcome studies McRoberts, Burlingame, and Hoag (1998) found no difference in outcome between the two modalities. Numerous authors such as Fuhriman and Burlingame (1994), Helfrnann (1994), MacKenzie (1994), Tillitski (1990) and Toseland and Siporin (1986) concluded that group therapy was an efficacious and efficient use of client and therapist time when compared to individual therapy. Not only are there economic benefits from the use of group therapy but there are therapeutic benefits as well. Yalom (1985) cited several advantages that group therapy offers over individual therapy. For example, in group therapy it is possible for the therapist to gain a broader sense of how the client interacts with members of the outside world extrapolated from their interactions with other group members. Also, there is the possibility for re-enactment and resolution of family-of-origin dynamics and issues with group members (siblings) vying for the attention of group leader/s (parents). Clearly there are advantages to the use of group therapy and we can expect to see more use of it in the future; however, it is important to note that it has been historically difficult to get therapists to refer clients to group therapy. Oftentimes this may have been due to the therapist’s belief that individual therapy is more effective or his/her greater comfort with conducting individual therapy. But it also may be due to client factors such as social withdrawal, discomfort with self-disclosure, or simply poor interpersonal skills. Riva and Smith (1997) cited a growing belief among clinicians that few clients cannot benefit from group participation, but that it was important to learn more about matching clients to groups in order to maximize this benefit. Delucia—Waack (1997) called for a more focused and probing inquiry into the ways individual characteristics impact group processes, particularly group formation processes. Ideally this inquiry would occur early in the group before normal maturation and therapeutic change also affect these processes. To date, no systematic theoretical framework has been applied to the study of group therapy to help clinicians and researchers understand why group members exhibit different behaviors and perceive group processes differently. Thus, there is a clear need for theory-driven rationales that help to explain these behaviors and perceptions. It was the contention of this present study that attachment theory as outlined by John Bowlby (1988) offered such a framework. This research examined how group member individual differences as seen in adult attachment orientations were related to expectations and perceptions of group processes and leaders. Attachment Theory as a Framework for Understanding Important Group Processes and Outcomes The reason that attachment theory may be especially relevant to group therapy is that it is based on the actual social interaction of a small group: caregiver and infant. Bowlby (1988) hypothesized that infants develop expectations of, and beliefs about self and other based on interaction with their primary caregiver. These expectations and beliefs, or working models, serve as a template for subsequent intimate relationships. Presumably then, these expectations and behaviors may impact interpersonal expectations and behaviors in a therapy group. For example, someone who has a positive model of self and other would be more likely than someone with a negative view of other to expect the group to be empathic and available to provide comfort and support in times of distress. These expectations are also very likely to affect the members commitment to remaining in the group, especially during the initial sessions that are marked by ambiguity and uncertainty. Bowlby (1988) argued that the process of evolution has resulted in humans becoming programmed to form enduring emotional bonds or attachments with others. These bonds offer nurturance and protection as a means of ensuring survival of the individual as well as the species as a whole. These bonds are formed between a vulnerable infant who seeks care and protection when facing new Situations and a caregiver who provides comfort and nurturance. As infants develop, they form cognitive schema or working models of self and other based on their attachment figures’ responsiveness to their solicitations for comfort and proximity. Specifically, Bowlby said that the individual draws conclusions about the availability and responsiveness of others as well as the selt’s competence and worthiness of love based on this critical early bond. He said that these conclusions about the self and other (expressed interpersonally as attachment styles) are enduring and serve as an expectational template for interpersonal relationships extending into adulthood. Ainsworth, Blehar, Waters, and Wall (1978) studied attachment using the “strange situation” observational methodology. In this methodology, infants were observed in a novel environment (a) in the presence of their parent, (b) when lefi alone, and (c) upon return of the parent. This resulted in the identification of three attachment styles. Infants with a secure attachment style have experienced their caregiver as available and responsive, acting as a secure base from which the infant can explore his/her world. These individuals incorporate a working model of the self as competent and worthy of love and others as reliable and responsive. The anxious-ambivalent or preoccupied (Bartholomew & Horowitz, 1991) attachment style is exhibited in infants whose caregivers were inconsistently responsive and available. This results in individuals who question their own competence and lovability and see others as desired, but unreliable providers of comfort and care. The development of self-confidence and willingness to explore their environment is dependent on the presence and approval of the attachment figure. The avoidant or dismissive (Bartholomew & Horowitz, 1991) infant has experienced a caregiver who was consistently unresponsive or rejecting. These persons develop working models of the self as alone in the world and unwanted, and they view others as uncaring, unsupportive and unreliable. They are believed to be individuals who have deactivated the attachment system due to expectations that yeamings and solicitations for intimacy go unrewarded. Main, Kaplan, and Cassidy (1985) proposed a fourth category known as disorganized/disoriented, composed of infants with both avoidant and anxious responses. Bartholomew and Horowitz (1991) labeled the adult version of this attachment style as m. F or the purpose of the present study it was important to distinguish between attachment style and attachment orientation. This distinction is mainly one of method of measurement. Attachment style refers to the placement of individuals into one of the four previously mentioned discrete categories based on beliefs in the availability of significant others and one’s own love worthiness. The present study utilized the concept of attachment orientation which refers to the amount of anxiety one feels in intimate relationships and also the degree of avoidance of intimate relationships. Of particular importance to group process is how adult attachment style impacts an individual’s ability to manage and regulate the many intense emotions that are elicited in the group. F uendeling (1998) noted stylistic differences in affect regulation according to adult attachment category. He observed that secure individuals’ low anxiety and low perception of threat in intimate relationships promotes using interaction with others to deal with negative emotion which, in turn, promotes intimacy. This low appraisal of threat also contributes to interpersonal risk taking and self-disclosure. In addition, secure persons tend to balance attention to their internal emotional state with other relevant material. Finally, he concluded that securely attached individuals are better able to effectively mobilize social support in response to felt stress. In contrast, anxious/ambivalent persons have a relatively high amount of interpersonal anxiety and are more likely to perceive threat in interpersonal situations. They tend to give disproportionate attention to their negative emotions and have greater access to negative, emotionally charged memories. Their use of social resources in dealing with negative affect tends to be ineffective and indirect. They are likely to engage in self-disclosure but also to repress hostility and tend not to seek out social support in times of true distress. Avoidant persons, on the other hand, tend toward repression of, and inattention to negative affect. They do not use social interaction as a means to deal with negative affect. They tend to be uncomfortable with self-disclosure due to their low expectations of others, and this inhibits open communication. The implications of these stylistic differences in affect regulation on members’ perceptions, performance, and expectations of group therapy are clear. In their groundbreaking work on the social psychology of groups, Thibaut and Kelley (1959) noted a number of group and individual processes which, intuitively, one would expect to be impacted by attachment orientation. They noted that first time entry into a therapy group is accompanied by feelings of strangeness and novelty. This could be considered analogous to Ainsworth’s Strange Situation. This novelty serves to delay group formation and is even more pronounced by an increase in perceived differences among the group members. Attachment theory would predict a wide perceptual difference among group members with differing attachment orientations since a cornerstone of one’s working models is either a sense of belonging and being valued by others or of being essentially alone in the world. It seems obvious that group members with a negative view of self and a chronic feeling of being alone would perceive themselves as being inferior to and different from their peers. If that assumption were to prove valid, it might have important implications for group formation and cohesion. Thibaut and Kelley also held that as the group relationship became established and intensified, dependency among group members increased, which could be threatening to some members. Attachment theory seems particularly relevant here, as we would expect members with low attachment security to view others as inconsistently or completely unresponsive to their needs. As a result, we would also expect members with low attachment security to be reluctant to place any of their dependency needs in another’s hands. It might reasonably be anticipated that these persons would prematurely leave the group in order to prevent this dependency from occurring. Thibaut and Kelley also noted that coalitions formed when roles were complementary. In a therapy group, coalitions are almost always destructive to group cohesion, so consideration of this effect would be important. Again, attachment theory might offer insight into how coalitions form. For example, in an instance where a person who is uncomfortable with intimacy and self disclosure (high avoidance and low anxiety) might collude with someone who seeks intimacy and is perhaps too willing to self disclose (low avoidance and high anxiety), we might anticipate that the self-discloser would do most of the sharing so that the other wouldn’t have to. Finally, Thibaut and Kelley maintained that group normative behavior was only relevant inasmuch as the norms had intrinsic value to the members. A central norm of group behavior is the seeking and maintenance of intimacy and closeness among the members, but if a member does not value these group attributes, as might be expected of persons with high avoidance and low anxiety, he/she may be more motivated to violate than conform to these norms. Attachment theory then informs, by inference, that such a person might eventually be susceptible to dropping out of the group, might become a group scapegoat, or a target of group anger. Several writers have supported the importance of attachment theory to group work. Mallinckrodt (2001) said that the study of individual difference variables such as the “Big Five” personality traits, or attachment orientations and their relationships to social behavior, or group therapeutic factors was likely to be a promising research direction. Flores (2001) believed that attachment theory had particular relevance to the treatment of addictions and the use of group therapy in treating these disorders. He argued that the process of addiction resulted in the patient developing an unhealthy attachment to the substance of abuse and, by immersing oneself in a therapy group, this attachment could be eventually and healthily displaced onto the group and its members. He poignantly noted a concept that attachment theory and group therapy share is a belief “that the essence of being human is social, not individual” (p.68). A number of researchers have used attachment theory as a means by which to better understand group processes or to plan group interventions. Smith, Murphy, and Coats (1999) in a series of studies determined that individuals possessed attachments to groups as well as to particular others. They found in a non-clinical sample that group- related attachment anxiety and avoidance could be reliably assessed and that these dimensions predicted emotions about group membership, social support, collective self- esteem, methods of conflict resolution, and time and activities shared with the group. The implication in these findings is that the group to which one belonged could become an attachment figure or secure base in and of itself. Kilmann et a1. (1999) designed an attachment-focused group intervention with 13 insecurely attached women. This intervention was a blend of typical group processing of participants’ relational histories along with psycho-education aimed at confronting dysfunctional relationship beliefs, examining relational patterns, and practicing new relationship skills. Compared to a control sample, group participants reported more secure attachment patterns, better interpersonal functioning and less endorsement of maladaptive relationship beliefs. Perhaps the most impressive aspect of the intervention was that these changes were evident at 6 months after the intervention. De Zulueta and Mark (2000) developed an approach using the secure base concept for working with seven patients with borderline personality disorder to contain splitting which is characteristic of this disorder. These patients were seen concurrently in individual and group therapy. Individual therapy was provided to establish a secure base from which patients could enter group therapy to explore new behaviors. The group therapist and six individual therapists also formed a “therapeutic matrix” by meeting regularly to monitor patient splitting. This had the unforeseen effect of providing a sense of group purpose and secure base for the therapists who were working with this difficult population. Of the seven participants, only one had not achieved an improved level of functioning after completion of the group. Group Therapy and Interpersonal Demands A key assumption of this present study was that group members’ adult attachment orientations would differentially impact their perceptions of group processes and leaders 10 as well as expectations of the group therapy process. Any person who has been a member of a group or has facilitated one can attest to the many and varied demands that are placed on the participants. Participants enter because of some form of interpersonal distress, and few are aware of the parameters and rules for group participation. In contrast to individual therapy where there is an emphasis on the therapist’s part to accept the client and try to understand the client’s world, group members (who are focused on their own distress) are often less invested in the others’ therapy experience and are likely to be less attentive, less empathic, and more confrontational. It was thought that this would be particularly true in the initial sessions of the group prior to the emergence of a group identity and coherence. It was believed that adult attachment orientation-related differences might well cause group members to be more or less sensitive to these group experiences. Participation in group therapy experiences require that members self-disclose and engage in self-examination. Yalom (1985) argued that group members who are able to share their problems and to reflect on their participation in the creation of them are more likely to experience a positive group outcome, to be more committed to the group, and to view the group more favorably. Attachment orientation-related differences in these abilities would clearly dispose group members to have different experiences of the group process and outcome. Of particular interest to the present study was how group members contributed to a process that eventually recapitulated the dynamics they experienced in their families of origin and how adult attachment orientation-related differences disposed them to view group leaders (who might be viewed as parents) and climate (might be viewed as trust in 11 the family unit) differently in the formative stages of the group. It seemed plausible that the member’s adult attachment orientations could have important implications for their ability to trust the group leader/s and other members. For example, one would reasonably argue that a person with a secure adult attachment orientation that was reinforced by positive memories of early care giving would have expected the group leader/s and members to be more available and willing to provide support than would someone with negative care-giving memories. Main Questions of the Proposed Study The main questions that were addressed by the present study were: Do individual differences in adult attachment orientations relate to differing (a) expectations of the group experience, (b) perceptions of group leaders, and (c) perceptions of group climate in the formative stages of the group? Summary To summarize, the field of group psychotherapy is not held together by a theory of psychological functioning that would help explain the relational transactions and therapeutic outcomes that issue fi'om a therapy group. The present research suggested that attachment theory would provide such a theoretical framework. The preceding discussion was intended to focus the reader’s attention on the fact that attachment theory has arrived as a prominent player in the writings of theoreticians and researchers of group processes. However, there were a multitude of group processes and variables that could have been studied relative to attachment. Therefore, a decision was made about how to delimit the present study. This research argued that group member’s expectations of the therapy 12 experience and their perceptions of group climate and leader/s were critical variables to examine in relation to adult attachment orientation. For Bowlby (1988), the internal working models of self and other comprised beliefs about, and expectations of, self and others in interpersonal relationships. These working models formed a template that the individual used to guide interpersonal behavior that would reinforce the existing working models. For example, a dismissive person with a positive view of self and negative view of other would be expected to act rather cold and superior to others which, in turn, could have caused these persons to withdraw their offers of support and comfort, thereby confirming the dismissive person’s view of self as being essentially alone and others as unavailable. This example provided support for including the examination of expectations in the present study. Specifically, expectations regarding the care the group was likely to provide, the presence of a healing environment, and commitment to the group were examined. Yalom (1985) pointed to the critical importance of a supportive group climate and empathic facilitators for an optimal group experience. Most important was the perception of support and empathy on the part of participants. He believed that the leader/s and members of a group could recapitulate the family-of-origin for group members with the potential for re-enactment and curative experiences. From an attachment perspective, a member’s entry into an adult therapy group could be considered analogous to an infant encountering a new or novel situation. In these circumstances anxiety would likely be heightened as each member of the group would be surrounded by strangers and immersed in an ambiguous environment. Bowlby (1988) proposed that these were the kinds of conditions in which internal working models 13 of self and other were likely to be activated. In these circumstances it would be especially crucial that the leader be seen as empathic and available, and the group as safe and supportive, in order for these underlying models to be examined and revised. It was suggested that the perceptions of self and other associated with different adult attachment orientations could prove to be very influential in the initial stage of group formation. One could expect securely attached participants to view the group leader more favorably and to believe in his/her availability and willingness to help, whereas preoccupied participants would likely view leaders favorably but also see them as potentially rejecting. Dismissive and fearful participants (i.e. those with high levels of attachment avoidance) would, theoretically, view the leader as uncaring and unavailable, but these participants were likely to differ in that the dismissive person was not apt to convey any need or desire for the leader’s help. Similar predictions could be made in terms of the group in general, and group climate specifically. Individuals with low attachment anxiety and, presumably, positive views of others (i.e. a secure orientation) would be likely to view the group climate more favorably as seen in higher levels of group engagement and lower levels of group avoidance of interpersonal issues than their dismissive, and fearful peers. Pistole (1997) provided additional support for these views by asserting that members’ adult attachment orientations are exhibited in the here and now interactions they exhibit in the group. Problem Statement The arrival of managed care has had an important impact on the practice and provision of psychotherapy. No longer will insurance companies pay for an unlimited number of therapy sessions. This marked change has forced many psychotherapists out of 14 practice and has caused others to reconsider the efficacy of their methods. Although this shift may be seen as negative, it also offers the field of psychotherapy an exciting opportunity to try new and lesser-used forms of treatment. One such treatment, group therapy, has long been viewed as a second-class form of treatment. It was the contention of this study that group therapy deserved greater recognition and increased research emphasis. However, it was the assertion of this study that the field of group therapy was not held together by a theory of how individual differences among group members affected their expectations of a group therapy experience as well as their perceptions of group therapy leaders and group climate. Instead, the field seemed to be largely composed of clinical wisdom derived from years of practice. This study was an initial attempt to determine if attachment theory could provide some theoretically based insights into group member expectations and perceptions. A Significant body of research on the relationship of adult attachment styles/orientations to personal characteristics (such as self disclosure and reactions to stress) that are likely to impact group processes supported an examination of adult attachment and its impact on group therapy. It was proposed that individuals with different adult attachment orientations characterized by positive or negative views of self and other would enter the group with different relational templates and expectations which would influence them to perceive aspects of the group experience differently. Specifically, it was hypothesized that adult attachment-related avoidance and anxiety in interpersonal relationships would be reflected in differences in perceptions of group 15 leaders and group climate as well as members’ assessments of their expectations of the group therapy experience. To conclude, the purpose of the study was to extend the application of attachment theory into the realm of group processes by examining adult attachment orientation- related differences in perceptions of group leaders and to perceptions of the group climate. There were several benefits of the study. First, it represented an initial empirical attempt to understand group therapy participants’ expectations and perceptions from an attachment perspective. Second, this understanding could serve to help guide practitioners in attending to attachment-related behaviors in the group setting. Finally, it offered a preliminary step into a consideration of whether attachment theory might serve as a theoretical perspective from which to view group dynamics in a therapy setting. Literature Review Introduction. This literature review provided the rationale for the completed study. For years, researchers have studied the effects of personality characteristics on individual therapy process and outcome but, curiously, the same had not been true of group therapy outcome or process. In fact, although there were numerous theories of therapy and of how therapy begets personal change, group therapy was not held together by a particular theory that helped to explain why people do the things they do in a group therapy context. Therefore, the first three sections of the literature review pointed out how critical individual differences were to expectations of therapy, and perceptions of group processes and leaders. The reader will note that none of the researchers used individual 16 differences that were based on a particular theory of personality functioning. The first section examined studies dealing with individual differences and expectations of therapy. Since there was no extant research on individual differences and expectations of group therapy, most of the research reviewed focused on analogous relationships in individual therapy. The second section reviewed studies of individual differences and their impact on group processes and, importantly, on group climate. The third section reviewed studies examining the impact of individual differences on perceptions of the therapist. The final section reviewed studies supporting the use of attachment theory as a framework for understanding group dynamics. Although the research presented in this section did not deal explicitly with attachment theory and group therapy, the reviewed literature presented attachment-related differences (such as self disclosure, stress-related coping, and discerning of facial emotion) that one would expect to have an influence on group processes. Individual Difi'erences and Expectations about Therapy. Frank (1959) cited the importance of client’s expectations for psychotherapy. He stated that the expectation of benefit in and of itself may have “enduring and profound effects” on the mental and physical welfare of the client. Wilkins (1973) disputed this and said that the assumption that expectations were of profound importance had led researchers to place unexplained therapeutic benefit into the vague category of client expectation. Wilkins argued that this had served to attenuate research into this area of unexplained therapeutic gain. Given this debate, it seemed reasonable to examine expectancies as they related to individual differences, as it was considered plausible that certain individual differences could account for a portion of this expectancy effect. For 17 the present research, the impact of adult attachment orientation-related differences on expectations for entry into a group was most important but there was a noted lack of such research. A literature search failed to uncover any empirical studies on how individual differences impacted members’ expectations of the group experience. There were, however, several studies of individual differences and their impact on expectancies of entering individual counseling. For example, Schaub and Tokar (1999) explored the relationship of expectations about counseling to the NBC five-factor model of personality. The individual difference in this study was participants’ differing personality components. Five clusters emerged from their analysis. The first consisted of high expectations for taking personal responsibility for working hard in counseling, with above average expectations for the establishment of facilitative conditions by the counselor, and with average expectancies that the counselor would be expert. Participants in this cluster exhibited low scores for neuroticism and above average scores on openness, extraversion, and agreeableness. Cluster 2 individuals exhibited slightly above average personal responsibility for success in counseling but lower expectations that the counseling would be facilitative and the counselor effective. Personality profiles for this cluster of participants showed relatively average scores on all five dimensions of the NEO. Cluster 3 individuals had high expectations about their role in counseling and very high (more than one standard deviation above the mean) expectations about the counselor’s expertise and role in facilitating therapy. These participants scored highest on the Neuroticism, Closedness, and Optimism dimensions. Cluster 4 participants showed elevated expectations about counselor expertise, average expectations about facilitative conditions, and low 18 expectations for assuming personal responsibility for counseling success. Cluster 5 participants had very low expectations for taking personal responsibility, counselor effectiveness, and the possibility for the establishment of facilitative conditions. These individuals collectively scored very low on the Optimism dimension as well as extraversion, openness, agreeableness, and conscientiousness factors. The strengths of this study included that it was based on research-tested constructs and well-validated measures. In addition, the large sample size was relatively diverse. Limitations included the use of a non-clinical sample which likely diminished generalizability, the use of self- report instruments may have increased the likelihood of a response set and, measures were standardized to eliminate gender differences so related influences were eliminated as well. Satterfield, Buelow, Lyddon, and Johnson (1995) studied outpatients seeking counseling at an outpatient clinic and found that client’s stage of/readiness for change significantly correlated with expectation factors. In this study the individual difference variable was the client stage of change which referred to how ready a person was to tackle a problematic issue in their life (i.e. an addiction). They found that the composite of the contenmlation (when a person acknowledged a problem and considered a change), ict_ig_n (when a person was actively trying to deal with the problem), and maintenance (when a person was focused on maintaining the changes that he had enacted) stages of change were significantly related to expectations that the counseling process would be facilitative to change and personal expectancies of being committed to the counseling process. In addition, a composite of the contemplation and maintenance stages accounted for additional variance in expectations for facilitative conditions, counselor expertise, and 19 nurturance. The strengths of this study included the use of well-established and validated measures, a clinical sample, and the introduction of the concept of stage of change from the behavioral medicine literature. Factors that limited the generalizability of results included the exclusive use of self-report measures and a small sample Size with an over- representation of women. Gladstein (1969) examined the counseling expectations, experience, and satisfaction of 181 secondary school participants. In this study the individual difference variable was the nature (i.e. personal, vocational, or career) of the counseling that was sought. He found that expectations were numerous and diverse, that expectations pre- and post- did not change except for persons with vocational expectations, that most expectations were met, and that participants who had only some of their expectations met did not Show less satisfaction. The strength of this study was that it assessed a large number of actual clients. A major limitation was that clients presented for a diverse range of needs including personal counseling, academic advising, and vocational information which were all included in the realm of counseling which led to diverse and numerous (17) expectations. Also, clients did not know what to expect of the counseling experience and were not informed as to the nature of counseling. Finally, expectations were assessed with a single open-ended question and the outcome measure of post-counseling perceptions was developed for this study and so, was untested. Tinsley, Brown, de St. Aubin, and Lucek (1984) examined expectancies of help providers and individuals’ tendencies to seek help. In this study the individual difference variable was tendency to seek help. Results indicated that participants had higher expectations of being motivated by a counseling or clinical psychologist or psychiatrist 20 than by college counselors. They also anticipated being more open with members of these disciplines than with college counselors. Participants also perceived differences among the disciplines according to amount of confrontation, directiveness, concreteness, and use of immediacy. Finally, the participants had significantly greater expectations of a positive outcome from a helping session with a counseling or clinical psychologist or psychiatrist than from a session with a career or college counselor. Strengths of the study included a large sample size and straightforward design with 20-30 participants per MANOVA cell per professional provider. There was also a relatively even gender split. The fact that the study was based on a young, ethnically homogeneous, and nonclinical sample limited generalizability of results. Additionally, for a portion of the sample, counseling was directed toward career issues but the expectations measure was directed toward personal counseling which could have caused confusion for some of the participants and resulted in lower expectancies for some professionals, such as career counselors, than others. There also could have been confusion among participants about how counseling psychologists, clinical psychologists, psychiatrists, advisers, and the other ranked professions differed distinctly from one another since the authors did not indicate that they had made these distinctions clear to participants. As the reader can see, the literature concerning how individual differences impact expectations is sparse. In addition, a search uncovered no articles that examined how group member individual differences, such as personality structure, impact expectations of entering a therapy group. 21 Impact of Individual Differences on Group Processes and Climate. Yalom (1988) identified numerous curative factors that were essential to group formation and progress but said virtually nothing about how individual personality factors affected group processes or group climate. However, he did identify paranoid, hypochondriacal, and sociopathic individuals as persons to be excluded from group. He stated that the inclusion of such persons in a group would disrupt the development of group cohesion and a positive working alliance or climate that were critical for group and member progress. Again, no research demonstrating the impact of adult attachment orientations on member perceptions of group processes and specifically, group climate, was located. In fact, there was a paucity of research examining the effect of individual differences on group processes. Therefore, the reader would note the first four studies presented in this section dealt with individual differences as they related to group- relevant concepts such as beliefs about social connectedness, social network development, and relationship quality. One study indirectly supported how attachment-related processes may be important to social connectedness, interpersonal behaviors and psychological distress (Lee, Draper, & Lee, 2001). These authors cited Lee and Robbins’ (1998) suggestion that parent-child attachment and adolescent experiences with peers and groups resulted in an adult’s sense of social connectedness which was a relational schema that exhibited the individual’s enduring beliefs about their own closeness with the social world. Lee et a1. (2001) concluded that the lack of a sense of social connectedness likely resulted in the development of maladaptive interpersonal behaviors such as difficulties with submissiveness and assertiveness that led to heightened psychological distress. One 22 strength of this study was its relevance to group process research through its direct linkage of social disconnectedness to less favorable therapy outcomes. The authors also used well-established and reliable measures and had a reasonably large and ethnically diverse sample. Limitations were the self-report nature of all assessments and the fact that the social connectedness scale was highly correlated with loneliness, which could have called into question the distinctiveness of this concept. In addition, the sample was non- clinical and very young (mean age 18.9 years) which constrained the generalizability of findings to other participant groups. Asendorpf and Wilpers (1998) reported significant effects of personality characteristics on social relationships, albeit not group therapy. These investigators studied the effects of the Big Five personality traits (Costa & McCrae, 1992) on social relationships and found that, after controlling for the initial correlation between personality and relationship quality, Extraversion, Agreeableness, and Conscientiousness predicted relational aspects such as number of peer relationships, conflict with these peers, and falling in love. The strengths of this research were that it was longitudinal with repeated assessments over 18 months and based on young adults experiencing a major transition (entry into a university). This study was also relevant to the present study as it assessed participant perceptions of social relationships. The generalizability of this study to American university group therapy participants may be limited due to the fact it was conducted with young, German college students using self-report questionnaires. Additionally, generalizability may be limited due to a high level of attrition with half of the originally enrolled women dropping and more than a third of the men. The sample was also disproportionately female and sex differences were not studied. The authors also 23 included no reliability data on the relationship questionnaire they used. Finally, the results might have been skewed due to students scoring more highly on the Big Five dimensions of Openness to Experience and Agreeableness than the German normative sample. Neyer and Asendorpf (2001) asserted that personality actively creates one’s social environment instead of being created by the environment. In a four-year longitudinal survey of 489 young (18-30 years old) German participants, they found that increased Extraversion predicted closeness in and, importance of, relationships; Shyness and Neuroticism were associated with increased insecurity and decreased closeness in relationships; self esteem predicted later importance, conflict, insecurity and closeness in relationships; and Agreeableness was related to later increases in closeness and importance of relationships. This study benefited from longitudinal assessment of a large, heterogeneous but nonclinical sample drawn from the German general public. It was limited in that all assessments were self-report. The quality of relationships was a single question measure and no reliability or validity data was presented on the social relationship inventory. The results also might have been influenced by the voluntary nature of the research with low control on the part of the researchers evidenced in respondents’ higher scores on Conscientiousness and Agreeableness than would be seen in a general sample. Brissette, Scheier, and Carver (2002) examined the impact of optimism on social network development, coping and adjustment. In a study of 89 college students they found that increased optimism was related to increased social support which, in turn, resulted in better psychological adjustment during a life transition. These authors also 24 conducted longitudinal assessments at entry to college and after completion of the first semester. There was a near even split of male and female participants and the measures they used were well-validated and reliable. The limitations were that the sample was nonclinical and young (all college freshmen). In addition, the sample size was rather small and all assessments were self-report. de Carufel and Piper (1988) also studied patient characteristics as predictive factors for progress in individual or group therapy. These authors did not organize patient characteristics according to any theoretical perspective but rather used predictive characteristics such as psychological mindedness, quality of object relations, defensive style, and likability. They found none of these variables to be significant predictors of greater progress in one type of therapy (group versus individual) over the other. The strengths of this research were in the use of actual patients and the fact that severity of target objectives and alleviation of these (noted to be progress) were rated by the patient, therapist and an independent assessor. However, shortcomings included the fact that the authors did not state the criteria employed for recruitment into the study but merely said participants were deemed “suitable” for psychodynamic treatment. Additionally, predictor variables were operationalized by a single therapist interview that was Likert scaled. And finally, therapy was psychoanalytically oriented which could have called into question the relevance to a typical short-term therapy group. Two studies did show clear results supporting the investigation of individual differences and their impact on group therapy. Tasca, Russell, and Busby (1994) studied 20 participants who chose between two types of group therapy and found that persons who preferred verbal and process-oriented groups were more likely to have externalizing 25 defenses such as projection and expression of hostility. These authors pointed out the particular use of projection and anger against others by these individuals. They also found patients who preferred structured and activity oriented groups tended to have internalizing defenses particularly seen in the use of repression, denial, and intellectualization. These groups were characterized by less emotional expression. The strength of this study was the use of actual patients presenting to a day treatment program however, this might also be seen as a limitation as psychopathology was likely to be more severe than would be seen in participants in the present study. Limitations included lack of random assignment and a small sample size with no statement about composition of men and women or ethnic diversity. In addition, groups were composed of differing types of pathology with a mixture of Axis I and Axis II disorders that could have affected choice of group and preferred defenses. Except for the psychological mindedness assessment that was completed by two independent raters, all other assessments were self-report. There was no statement about training of therapists except that they had five years of experience. The authors also noted that most of the participants were on medication but did not address the potential confounding impact this may have had. In a study of 139 psychiatric outpatients presenting for either interpretive or supportive time-limited group therapy due to grief issues, Piper, McCallum, Joyce, Rosie, and Ogrodniczuk (2001) found that two personality characteristics—quality of object relations and psychological mindedness—exerted a significant influence on alleviation of grief symptoms. These authors found that persons who scored highly on quality of object relations improved more in interpretive therapy and that participants low on this dimension improved more in supportive therapy. They also found that higher degrees of 26 psychological mindedness were associated with improvement in grief symptoms in both types of therapy. Strengths of this study were that it was a randomized clinical trial with actual patients. Treatment was highly controlled by use of manualized treatment and checks for adherence. In addition, the authors completed an intent to treat analysis on dropouts and found no significant differences between completers and noncompleters. Limitations of the study included a lack of reliability data on complicated grief assessments, an attrition rate of nearly a fourth, an over-representation of Caucasian women, and a marked severity of psychopathology as was seen in more than a third of the sample having Axis I and II diagnoses concurrently. The authors said their findings may only have generalized to psychodynamic therapies since this was the format used. In addition, grief was not a major presenting issue for most of the sample and was discovered by staff only after entry into treatment so a question existed as to how important an issue grief was to the participants. It was argued that individual differences in perceptual differences among members could have important predictive power in terms of therapeutic progress. Flowers (1987) studied 24 adults participating in outpatient group therapy to examine . individual differences in agreement/disagreement with what the majority of group members reported as the most important of curative factors. He found that people who improved in treatment generally agreed on the rank order of curative factors but those who did not improve exhibited significantly different rankings. In addition, he found that these individuals also differed from one another on the rank ordering of curative factors. It was also noted that two of the three participants who experienced no improvement or regression exhibited significantly less satisfaction in their experience of the group. 27 Strengths of this study included use of short-term problem solving group therapy with a clinical sample. Limitations included small sample size with no information on selection criteria. In addition, the outcome assessments may have been biased as they were completed by the therapists who were leading the groups. These outcomes assessments were also overly simplistic reflected in number of DSM diagnoses. Kivlighan and Goldfine (1991) studied members of personal grth groups to determine how the stage of group development and interpersonal attitudes affected the perceived importance of therapeutic factors. They found that more affiliative members reported cognitive factors such as universality, self-understanding, guidance and vicarious learning as being most important whereas nonaffiliative members reported behavioral factors such as altruism, learning from interpersonal actions and self- disclosure as most important. This study was conducted on students in a group therapy class who participated in a personal growth group as part of the experience although participation in the research was voluntary. Therefore, generalization to a clinical sample was not recommended aithough this sample could have been more representative of typical college student concerns than would a person seeking help at a community mental health center or psychiatric hospital. Limitations of this study included small sample size, disproportionate representation of females, inexperienced group leaders, and the fact that the data was not collected immediately after group which may have colored participants’ ratings. Kivlighan and Angelone (1992) explored the relationship between group member interpersonal problems and their perceptions of the group atmosphere or climate. They hypothesized that participants who saw themselves as too dominant relative to more 28 submissive group members would perceive the group as evidencing more submissive behaviors. They also hypothesized that group members who saw themselves as too cold relative to their too nurturant counterparts would see the group as colder. Their hypotheses were largely corroborated with dominant members viewing other group members as uncertain in their interactions and too dependent on others for support and guidance. In line with expectations was the finding that too cold members saw other members as less engaged. Contrary to expectations, colder members saw the group as having less conflict and the members as being less anxious. Nonassertive and nurturant members saw the group as more engaged, conflictual, and anxious. This study was conducted on 61 students in a group therapy class who participated in a personal growth group as part of the experience although participation in the research was voluntary. Therefore, generalization to a clinical sample was not recommended. Limitations of this study included small sample size with disproportionate representation of females. Also, group leaders were not experienced therapists but rather practicum students. Marcus and Holahan (1994) used a social relations model to analyze data on group member interpersonal perception in the early stages of group therapy. They found that perceptions of other group members issued from two sources. The first source, assimilation, was highly dependent on the person’s perceptions and could be thought of as transference. For example, a person assumes that all people are cold and distant. In consensus, the second source, a group of people see an individual as being consistently one way (i.e. a group of friends view one of their members as being overly serious). These investigators used Kiesler’s (1983) interpersonal circle as an organizing theme for their study, with the circle being organized along dimensions of control and affiliation. 29 Each of these dimensions had 2 poles with the control dimension composed of the dominant and submissive poles and the affiliation dimension composed of the hostile and friendly poles. Kiesler posited that relations along the control dimension were reciprocal, with dominance eliciting submission, and vice versa. He said that relations along the affiliation dimension were complementary with friendliness eliciting friendliness and hostility eliciting hostility. Strengths of the Marcus and Holahan study were that it was based on an actual clinical sample from a university counseling center and ratings were not just self-report but based on other member’s perceptions as well. Limitations were the small sample size and low return rate of the measure assessing perceptions of other group members. In addition, there was an over-representation of women in the study. Although the literature on adult attachment is growing there was not a single reference for the impact of adult attachment orientations on group processes and climate. There was, however, strong empirical support for the influence of other individual differences on group process and climate. The reader is again reminded that the individual differences explored in this section were not based on any psychological theory. Therein was a strength of the present study. Attachment theory does provide a theoretical framework for understanding the impact of individual attachment orientation differences on perceptions of group climate. Impact of Individual Diflerences on Perceptions of Group Leaders. A final variable of importance to group process and outcome is perception of the therapist. Yalom (1985) noted the importance of group members trusting, and feeling empathy from, the group therapist. He believed that group participants often transfer feelings about parents to group therapists much as they do in individual therapy and that 30 this can be a potentially powerful source of therapeutic material and intrapsychic problem resolution. There was little research on the impact of individual differences on perceptions of group leaders and none specifically relating to how adult attachment orientations might impact these perceptions. Although there was little literature on perceptions of group leaders, there were a number of studies examining client’s adult attachment orientations and perceptions of the working alliance in individual therapy. It is reasonable to assume that, if ratings of adult attachment are related to the working alliance in individual therapy, they may be significantly correlated with ratings of group leaders. It was argued that the transference directed toward a group leader was likely to be reflective of members’ early experience with parents, as the group setting with its other members (siblings) and co-facilitator (other parent) were likely to elicit reactions similar to those elicited in the family of origin. Mallinckrodt, Coble, and Gantt (1995) and Mallinckrodt (1991) provided support for this idea by finding significant correlations between ratings of parental bonds and therapeutic alliance in adult individual therapy. Several other authors have found significant correlations between adult attachment orientations and ratings of the therapeutic relationship. Satterfield and Lyddon (1995) conducted a study of 60 first-time therapy clients and examined the relationship of adult attachment orientation to perceptions of working alliance. They found that clients who lacked belief in the availability and dependability of others were more likely to give negative ratings to the counseling relationship in the early phases of counseling. The strengths of this study were that it was based on a clinical sample and the clientele was more ethnically diverse than in other studies cited. Limitations included a high rate of 31 attrition resulting in a small, largely female sample size. In addition, the counselors were graduate students, relatively inexperienced, and ethnically homogeneous. Finally, all measures were self report and the Working Alliance Inventory was administered after only three counseling sessions. Kivlighan, Patton, and Foote (1998) studied the effects of counselor experience and client attachment orientation on working alliance ratings across 40 counselor-client dyads. They found that clients who were comfortable with intimacy and who could trust and rely on others in time of need, perceived stronger alliances with their counselor. They also found that anxiety in interpersonal relationships was not related to perceptions of the working alliance. Among clients who were uncomfortable with intimacy, counselor experience was a significant predictor of working alliance ratings. However, for clients who were moderately to highly comfortable with interpersonal intimacy, experience was not significantly related to working alliance strength. Strengths of this study included use of a clinical sample and rating of working alliance by both the client and counselor. This study was hampered by a small sampse size with high attrition. In addition, a sizeable portion of the sample had previous therapy experience which may have impacted the working alliance ratings. Finally, counselor experience was simply coded according to number of years of graduate school plus years postgraduate which may not have reflected actual experience. Only one study examined perceptions of the group leader. Brykczynska (1990) studied how perceptions of the leader varied from individual to group therapy with 73 patients being treated for neurotic concerns. She found that patient symptom relief and personality trait change were only correlated with certain characteristics of the 32 therapeutic relationship with an individual therapist. This author suggested that perception of the therapist may be more important in individual versus group therapy. She believed that the relationships that group members had with one another were more important than their relationship with the therapist. Strengths of this study were that it was based on a clinical population with random assignment to either group or individual therapy. Limitations included no report of criteria for entry into the study and no mention of age, gender, ethnicity, or previous therapy experience of the sample. All measures were self-report and the perception scale was developed specifically for the study; hence, no reliability data on this measure were available. The study was atheoretical and the therapy was described as eclectic. Finally, the study was completed in Poland which constrained generalizability of results to American university students. Only one study was located that examined how client individual differences impacted perceptions of group leaders. Kivlighan, Marsh—Angelone, and Angelone (1994) hypothesized that group member’s perceptions of group leader’s control and affiliativeness would be related to the group member’s interpersonal problems. Results indicated general support for their hypothesized relationships with a diminishing correlation among group member variables with group leader variables over time. They found that group member control problems—a tendency to be domineering—accounted for 14% of the variance in early (first 10 sessions) ratings of group leader control but decreased to 1% in late (sessions 11-20) ratings of control. Group member problems with affiliation accounted for 18% of the variance in early ratings of group leader affiliation and 9% in late ratings. These findings suggested that projection by group members upon the group leaders regarding issues of control and affiliation diminished markedly over 33 time. In sum, these authors found more dominant group members perceived the leader to be more dominant, and overly cold members saw the leader as being less affiliative, and thus, provided support for an active process of projection that occurred in group therapy. The results of this study strongly supported examining adult attachment orientation as an individual difference variable that could affect perceptions of group leaders. Other strengths included random assignment of participants to groups as well as the use of an ethnically diverse sample. Limitations included use of a nonclinical population, self- report measures, novice group leaders, and over-representation of female participants. Attachment Related Differences Relevant to Group Therapy. In considering the placement of individuals with different adult attachment orientations into a group, it is clear that, given the great variability in self/other models of the participants, random assignment to group therapy could result in less cohesion, slower development, and less favorable outcome. In recent years there has been a burgeoning literature on adult attachment which supported an inquiry into the impacts of adult attachment characteristics on group processes (Pistole, 1997). Yalom (1985) believed that individual personality differences among group members were important to group process and outcome. He indicated that participants who were most apt to change were those who were popular and influential in the group. He noted, however, that this popularity and influence did not come about randomly but rather was conveyed to members who actively participated, self-disclosed, expressed their feelings, engaged in self-exploration, displayed interest, provided support and leadership, and were non- defensive. A number of authors investigated, from an adult attachment perspective, 34 individual variables such as the ones cited above and have found individual differences that could have important implications for individual outcome and group process. One such variable is social competence, defined as “a positive sense of self- efficacy for valued social outcomes” (Mallinckrodt, McCreary, & Robertson, 1995, p. 179). Rice, Cunningham, and Young ( 1997) and Mallinckrodt, Coble, and Gantt (1995) found significant relations between secure attachment memories and social competence. Engels, Finkenauer, Meeus, and Dekovic' (2001) found that, for adolescents in the 15-18 year old range, parental attachment was related to social skills, which further impacted relational competence with both attachment and relational competence predicting emotional adjustment. Duggan and Brennan (1994) studied social avoidance from an adult attachment perspective and found that fearful individuals’ avoidance was due to shyness, a tendency to feel uncomfortable and anxious in the presence of strangers, whereas dismissive persons’ shyness was due to low sociability, a preference for being alone. This study was based on a large, young sample of 676 undergraduate students at a major university. There was a near even split of men and women with adequate ethnic diversity. The two major limitations were the use of a forced choice measure assessing adult attachment and a non-clinical sample. Self-disclosure was another critical individual variable believed to be associated with a successful group experience. Tschuschke and Dies (1994) and Yalom (1985) found that high levels of self-disclosure characterized highly cohesive groups. Mikulincer and Nachson (1991) studied adult attachment style and its relation to self-disclosure. They found that participants who believed others would be attentive to attachment signals were more willing to respond to others’ communications. Specifically, they found that 35 secure and ambivalent participants displayed greater self-disclosure than did avoidant persons. These two groups also disclosed more information, felt better about the disclosures, and were more drawn to high disclosing partners than were avoidant persons. In addition, relative to persons with preoccupied styles, secure participants showed greater flexibility in regard to disclosures and greater ability to shift the conversational topic. This novel research with a large sample size expanded the study of adult attachment into an area which was particularly important for psychotherapy generally, and group therapy, specifically. It was conducted on an Israeli, non-clinical sample of college students so the generalizability to a clinical setting was questionable. It was also possible that culture impacted self-disclosure. In addition, although the 2 attachment measures used provided agreement on participant’s attachment status, one continuous measure was quite brief (15 questions) and the other was a categorical measure of 3 attachment patterns. A non-verbal form of communication that was considered crucial to the group therapist’s and members’ understanding of group member behavior was that of body language and specifically, facial expressions. Niedenthal, Brauer, Robin, and Innes—Ker (2002) examined adult attachment as it related to perception of emotion expressed in facial gestures. These researchers had participants watch computerized movies portraying a face with happiness, anger, or sadness expressed. Participants were to watch each movie and indicate the offset (the point at which they perceived the face to express emotional neutrality). They found that, under conditions of distress, insecure participants perceived the offset of negative facial emotions later than did their secure counterparts. In the no distress condition, fearfully attached participants perceived offset of happiness and 36 anger earlier than did secure participants, but preoccupied and dismissive participants perceived the offset of these emotions later than secures. The present study suggested that the ability to accurately discern other group members’ emotions would have had important implications for successful group participation. The Niedenthal et al. (2002) experimental study used a novel method of eliciting responses by participants by using a projected movie to determine their responses to facial stimuli. Participants were a nonclinical sample of college students, so generalizability of results to other groups was questionable. In addition, the experimenters used only a forced choice 4—category measure of adult attachment and did not assess the possible confound of emotional state prior to participation. It was asserted that another critical component of successful group adaptation was the ability to engage in problem solving when conflict and anger arose. Yalom (1985) asserted that conflict was not only inevitable, but desirable in the evolution of the group. He stated that only through the experiencing and resolution of conflict could genuine intimacy result. Once again, research pointed to the importance of adult attachment as it related to a critical individual difference, in this case, conflict resolution. Lopez et a1. (1997) found that secure participants scored significantly higher on collaborative problem solving than did their insecure counterparts. These authors reported that participant’s guilt and shame proneness mediated the relationship between attachment assessments and collaborative problem solving. Guilt proneness was positively correlated with problem solving whereas shame proneness was negatively correlated with problem solving and positively correlated with avoidance. They found participants with negative working models of self were significantly more shame prone than were participants with positive 37 working models of self. The importance of problem solving in group therapy was thought of as obvious and the impact of adult attachment orientations on these capabilities provided further justification for studying the influence of attachment on group therapy. Limitations of this nonclinical study of college undergraduates were that females were disproportionately represented in the sample which prevented an examination of the effects of gender. In addition, the assessment of problem-solving was directed toward the ways couples in romantic relationships work to solve their differences versus group problem solving. Mikulincer (1998) studied the relations of adult attachment styles to individual experiences of anger. He found that secure persons tended to be less anger—prone, had more constructive anger goals, and expected more favorable outcomes of anger episodes than did insecure persons. Secure participants also reported more adaptive responses to episodes of anger and were less likely to see others as hostile. Ambivalent persons were found to experience a lack of control during anger episodes. Avoidant persons had little awareness of their physiological responses to anger, displayed greater hostility, and tended to use escape responses to anger. Although it was based on a small sample of college students and generalizations were made cautiously, the main strength of this study was its relevance to the study of group therapy. Anger is a common, and often alienating, emotion in therapy groups and this study contributed a beginning understanding of how the expression and experience of anger might be enacted in a group. Although the author found a link between attachment and anger, there were no claims about causation or the mechanisms which might underlie this connection. In 38 addition, potential confounds such as emotional state or relevant personality traits were not assessed. In a study that was considered important and with potential implications for group therapists who wish to conduct therapy from an attachment theoretical perspective, Mikulincer et al. (2001) examined the impact on empathy and personal distress of participant’s chronic sense of attachment as well as a condition in which there was contextual priming of participants’ senses of attachment security and insecurity by showing them pictures or providing reading material that prompted various attachment- related memories. They found that induction of secure attachment-related memories resulted in more empathic reactions and lessened personal distress. They also reported an inverse relationship between attachment-related anxiety and avoidance with empathy and a positive relationship between attachment-related anxiety and personal distress. Again, these results provided research support for pursuing an understanding of attachment- related influences in group therapy. Limitations of the study included the small sample size consisting of Israeli undergraduate students. One also had to consider the possible confounding effect of chronic unrest in the Mideast and the impact this might have had on anger. The authors acknowledged that other factors such as feelings of superiority could have accounted for their findings. Yalom (1985) reported that dropping out of group therapy was associated with reports of external stress. He said that group members who were unwilling or not ready to deal with the actual immediate stresses within the group (such as expectations of self- disclosure, intimacy, conflict, and self-examination) frequently used reports of external stress as a rationalization for terminating their group experience. It was argued that 39 individuals who were more able to negotiate these stresses in a therapy group would be more likely to remain in, and experience a successful termination with the group. That adult attachment orientations could be important predictors of this ability to commit to the group was indirectly supported by Mikulincer and Florian (1995) and Mikulincer, Florian, and Weller (1993), who found that, relative to their insecure counterparts, secure participants engaged in more effective stress-related coping. Yalom (1985) also reported that some group members dropped out because of fear of emotional contagion, or that they would not be able to adequately defend themselves from feeling the intense feelings other group members could experience. A number of authors (Brennan & Shaver, 1995; F eeney & Ryan, 1994; F uendeling, 1998; Kobak & Sceery, 1988) noted that securely attached individuals are able to more effectively regulate felt affect. Fuendeling (1998) noted important stylistic differences in the ways individuals of different attachment styles deal with negative affect. He asserted that preoccupied persons deal with negative affect by continually being attuned to the proximity or lack thereof, of the attachment figure. Their hypervigilance occurred at the expense of exploratory behavior that could promote environmental mastery. Clearly then, attachment-related differences in the ability to regulate feelings would seem likely to have had important implications for group processes and outcomes. In any therapeutic relationship trust is essential to the therapy process and outcome. Mikulincer (1998) explored the relationship between adult attachment style and trust in intimate relationships. He found attachment security correlated with higher degrees of interpersonal trust toward partners, easier access to trust-related memories, higher frequency of trust episodes over a three-week period, and more constructive 4O strategies in coping with violations of trust. He also found that attainment of intimacy was a main trust-related goal for all attachment groups. Additional goals were security attainment for anxious-ambivalent persons and control attainment for avoidant individuals. If the results of this research had relevance for group work one could imagine that having more secure individuals in a group would lead to higher levels of trust and more rapid group development. Again, the small sample consisted of Israeli undergraduate students so cultural tension may have impacted on ability to trust. Also, since the study was specifically focused on attachment as it related to trust in intimate romantic relationships the results may not have generalized to more social and less intimate relationships that would be more characteristic of the relationships in a therapy group. Horowitz, Rosenberg, and Bartholomew (1993) studied the relationship between adult attachment style, interpersonal problems, and outcome in brief psychodynamic therapy. They found that persons with problems in friendly submissiveness in contrast to hostile dominance tended to be more easily treated in brief dynamic therapy. They reported that persons with dismissive adult attachment styles tended to have more problems in the area of hostile dominance. They also found that the interpersonal problems of persons with a dismissive style were related to difficulty describing people clearly. Since their style prevented them from knowing others well, dismissive individuals had difficulty constructing internal representations and descriptions of others. These authors suggested that group therapy could have been a potentially helpful treatment for persons with dismissive attachment, since they were afforded the opportunity to hear others describe significant people in their life and thus, could learn to 41 construct their own descriptions of significant others. Strengths of this study included use of an actual client sample drawn from an outpatient clinic and measures of interpersonal problems rated by both client and therapist and outcome by client, therapist, and an independent rater providing validity to client ratings. The major limitation of this study was the small sample size and focus on individual therapy. Yalom (1988) stated that group members created a social microcosm in their therapy group that was reflective of their interactions with members of the outside world, which, in turn, reflected dynamics from their family of origin. Diehl, Elnick, Bourbeau, and Labouvie—Vief (1998) studied 304 adults to determine if there was a relation between adult attachment style, personality, and descriptions of family. They found that persons with more secure attachment scored higher on personality variables such as self- confidence, psychological well-being, and social functioning. They were also more likely to describe their present families and families of origin more positively. Strengths of this study were the large, randomly selected sample. However, the participants were noted to be older and more ethnically homogeneous than would be expected in a college counseling center. Summary. The previous discussion made a case for the existence of adult attachment orientation-related differences in social behaviors, expressions, and perceptions that could have important implications for participation in a therapy group. It also argued for adult attachment orientation-related differences in expectations of group therapy as well as perceptions of group climate and leader. The absence of adult attachment research on expectations of therapy and group processes and leaders made the argument necessarily 42 difficult. However, this absence has also made the argument easier by pointing to a crucial gap in the group literature. This review held that adult attachment orientation is a critical individual difference predisposing individual group members to perceive the group and leader in unique ways. Due to the lack of research on attachment and group therapy, the argument for the present study was supported by studies examining other individual differences that impact expectations and perceptions of group. In addition, further support for examining perceptions of the group leader was derived from examining literature that was concerned with how individual differences impact perceptions in individual therapy. Although the reviewed studies provided a compelling rationale for the present research, they lacked a coherent theoretical framework for organizing their diverse findings. The present research offered an opportunity to advance an understanding of group members’ expectations and perceptions within the framework of attachment theory. The present study proposed that differences in adult attachment orientations should predict differing expectations regarding entry into group therapy as well as differing perceptions of the group climate and of the leader during the early phase of group formation. Theory supported such a proposal in that these expectations and perceptions were based on the working models of self and other that participants brought to the group. It was thus argued that differences in adult attachment orientations differentially disposed members to engage in behaviors that ultimately affected the nature and quality of their group therapy experience. For example, it was expected that adult attachment-related anxiety and avoidance would be related to socially effective behaviors in the group. Attachment theory would suggest that these behaviors were due to 43 perceptual differences, with members low in adult attachment-related avoidance being more apt to believe in the availability of others (in this case, the group and group leaders) to provide aid and support. In addition, it was anticipated that participants low in adult attachment-related avoidance would be more committed to the group and expect to receive more nurturance and support from the group leader/s. Participants low in adult attachment-related avoidance were also expected to perceive a greater desire for closeness from the group leader. Group members who possessed low adult attachment- related anxiety were expected to perceive less need for control from the group leader and lower levels of group conflict. Finally, participants low in adult attachment-related avoidance were expected to perceive less avoidance of issues as well as more engagement in the therapy process by all group members. Pistole (1997) argued that attachment theory could be a perspective from which to more fully understand group processes. She supported the exploration of expectations and perceptions and asserted that working models filter information so that present models of self and other are confirmed. She stated that attachment may be particularly relevant for group processes since the initial sessions are often marked by anxiety and ambiguity which could serve to stimulate the attachment system, thus offering the opportunity for an attachment bond to develop in the group. 44 METHOD Design The study used a passive correlational design. Group therapy participants were asked to respond to 2 questionnaires prior to the first group session. One questionnaire assessed participants’ adult attachment orientations and one their expectations of the counseling experience. In addition, participants were administered the adult attachment measure, instruments to assess perceptions of group leaders and group atmosphere or “climate”, and a brief satisfaction survey after the fourth group session. This allowed the researcher to explore differences in expectations of counseling and perceptions of group leader/s and climate among group members with different adult attachment orientations. Adult attachment orientations toward avoidance and anxiety were the key predictors for this study and were operationalized through the use of the Experiences in Close Relationships questionnaire. This measure was administered before entry into the therapy group and after the fourth group session in order to establish stability of the measure within the sample. The dependent variables for this study included: members’ perceptions of the group leader operationalized using the Trainer Behavior Scale; members’ perceptions of group climate operationalized using the Group Climate Scale-Short Form; and members’ expectations of the therapy experience operationalized using the Expectations About Counseling—Brief Form. Expectations were assessed prior to entry into the group and perceptions of climate and leader/s were assessed after the fourth group session. 45 Participants Eighty undergraduate and graduate students who presented for counseling at one junior college and four university counseling centers were recruited to be participants. The groups were general therapy and theme therapy groups (i.e., survivors of incest). Groups were open (those that added members throughout the duration of the group) and closed (those that ceased adding new members after the third session). Data was collected only on group members who were part of the beginning cohort. Recruitment strategies attempted to ensure appropriate representation of women and men and different ethnic groups. Group leaders were Master’s and Doctoral level group therapists. Permission to conduct the proposed study was received from the directors of all five counseling centers. Procedures Participants were recruited from persons who presented for counseling at one junior college and four university counseling centers in the Rocky Mountain Region and were assigned to either a general or theme therapy group. When prospective group members attended a screening session, the screener read a prepared recruitment script (Appendix A). The script stated that they were being asked to participate in a research study concerning how individuals’ views of close relationships influenced experiences they had as members of therapy groups. Participants were asked to complete two self- report questionnaires, an informed consent form, and demographic questionnaire prior to starting the group and four self-report questionnaires after the fourth group session. Participants were advised that all information they provided would be kept confidential. Individuals then responded on a form (Appendix B) whether they did or did not wish to participate. This form was submitted to the researcher. Group therapists did not know 46 which members were participating. Persons who agreed to participate were advised they would be remunerated in the amount of $10.00 after submission of all questionnaires. Participants were contacted by a research assistant and asked to arrive 20 minutes early for their first group session so they could fill out the informed consent (Appendix C), demographic questionnaire, adult attachment measure, and the expectations of counseling questionnaire. After the fourth group therapy session, members were asked to remain 30 - 40 minutes to complete adult attachment, group leader, group climate and group satisfaction questionnaires and be compensated. These assessments took place in a confidential setting to minimize participants’ concerns with repercussions from group leaders or other members. Measures were administered by the primary researcher and assistants who had no involvement with the group. All participants were advised that their answers would be kept confidential. Participants were also given the opportunity to request information about the results of the study they participated in. Participants who did not appear for the fourth group session were contacted by a research assistant to complete assessments prior to the next group session. Data from participants who ceased participation prior to their fourth group session were included in demographic comparisons as well as the analysis of adult attachment to expectations of counseling. Therapy dropouts were verified by contact with group therapists and the initial data they presented was dropped from further analysis. Measures The following measures were used to acquire demographic information, consent, adult attachment orientation, expectations of group therapy and perceptions of group leaders and group climate from research participants. 47 Demographic Questionnaire (Appendix D). Each of the participants were asked to complete a Demographic Questionnaire soliciting information about their age, gender, ethnicity, academic classification, relationship status, dating status, previous participation and satisfaction in group and individual therapy, level of perceived sociability, and nature and severity of the problem that brought them in for counseling. Experiences in Close Relationships (Brennan, Clark, & Shaver, 1998; Appendix E). The Experiences in Close Relationships scale was used to assess adult attachment orientation. The ECR is a 36-item measure that yielded continuous scores on two factor- analytically-derived subscales, anxiety and avoidance, of adult attachment. The anxiety subscale measured the amount of worry or concern respondents experienced in their intimate relationships. Avoidance assessed their degree of comfort with closeness and dependency. Examples of questions from the ECR and the corresponding subscale included: “When I’m not involved in a relationship, I feel somewhat anxious and insecure” (Anxiety); and “I am nervous when partners get too close to me” (Avoidance). The higher the score on these two continuously scored factors indicated a higher degree of worry and discomfort. The authors reported the Avoidance subscale to be positively correlated with Griffin and Bartholomew’s (1994) classifications of Fearfulness (.81) and Dismissiveness (.39). Avoidance was negatively related to Security (-.70) and Preoccupation (-.13). Brennan et al. (1998) reported Anxiety to be positively related to Griffin and Bartholomew’s (1994) Preoccupation (.73) and Fearfulness (.32) but negatively related to Security (-.46) and Dismissiveness (-.29). Fraley, Waller, and 48 Brennan (2000) reported the ECR to be reliable with Cronbach alphas for Avoidance of (.91) and Anxiety of (.94) The scores on Anxiety and Avoidance were used to predict scores on dimensions of the dependent variables. In the present study this questionnaire exhibited excellent reliability over time. Reliability analysis resulted in a Cronbach’s alpha of .90 for ECR Anxiety and .88 for ECR Avoidance prior to entry into group therapy. Cronbach’s alpha for ECR Anxiety after the fourth group session was .94 and for ECR Avoidance it was .90. Expectations About Counseling—Brief Form ( T insley I 982; Appendix F). The Expectations About Counseling—Brief Form (EAC—B) is a shorter version of the original EAC and consists of 66 items answered on a seven-point Likert scale. Response options range from one (not true) to seven (definitely true). The EAC—B provided 17 scale scores that measured expectancies toward four factors upon entry into counseling. These included Personal Commitment, Facilitative Conditions, Counselor Expertise, and Nurturance. Examples of questions from the EAC—B and the corresponding subscale included: “I expect to go to group counseling only if I have a very serious problem” (Personal Commitment); “I expect to feel safe enough with the group members and counselor/s to really say how I feel” (Facilitative Conditions); “I expect to see an experienced group counselor/s” (Counselor Expertise); and “I expect the group counselor/s to praise me when I show improvement” (Nurturance). Hayes and H.E.A. Tinsley (1989) and D. J. Tinsley, Holt, Hinson, and Tinsley (1991) have supported the factorial validity and structure of the EAC—B. Tinsley (1982) found internal consistency reliabilities ranging from .69 to .82, with median reliability of .77. Test/retest reliabilities over a two-month period of the EAC—B ranged from .47 to .87, 49 with a median reliability of .71. H.E.A. Tinsley and Westcot (1990) demonstrated construct validity for the EAC—B by finding that the items discriminated between client expectations about counseling and two similar constructs—perception and preference. For this study, scores on the Personal Commitment, Facilitative Conditions, and Nurturance factors were predicted from adult attachment-related Anxiety and Avoidance scores. For the present study, Cronbach’s alpha analysis of these three dimensions indicated acceptable reliability with (a) expectations of commitment to the group equaling .83, (b) expectations of provision of facilitative conditions by the group leader/s and members equaling .79, and (c) expectations of nurturance by the group leader equaling .73. Trainer Behavior Scale (Bolman, 1971; Appendix G). The Trainer Behavior Scale (TBS) was composed of 28 items that described group members’ perceptions of the leader’s actions. Items were rated on a five-point Likert scale ranging from one (strongly disagree) to five (strongly agree). Higher scores indicated that more of the behavior was present. Factor analysis of these items has yielded seven dimensions of leader behavior: congruence-empathy, conceptual input, conditionality, perceptiveness, openness, affection, and dominance. Examples of items and dimensions they represent included: “She or he seems to be in close touch with how members of the group are feeling” (CongruenceEmpathy); “When she or he calls attention to something, she or he gives a theoretical explanation of why it occurred” (Conceptual Input); “She or he gives the impression that she or he likes some kind of behaviors better than others” (Conditionality); “She or he misinterprets what people say” (Perceptiveness); “She or he seems to hold back from expressing her or his own reactions 50 to what is happening in the group” (Openness); “She or he shows considerable affection for most of the group members” (Affection); and “She or he exerts considerable influence over the direction the group takes” (Dominance). Bolman did not report scale reliability for the original sample, and validity for the TBS was estimated by correlating group leader dimensions with participants’ ratings of group atmosphere. For the present study, dimensions used from the Trainer Behavior Scale were group leader control and affiliativeness. The group leader control dimension exhibited poor reliability within this sample with a Cronbach alpha equal to .41. The group leader affiliativeness dimension exhibited marginal reliability with a Cronbach alpha equal to .67 within this sample. Dies (1983) suggested that instruments measuring group leader behavior measure two basic dimensions called control/technical and affiliativeness/personal. Kivlighan and Shaughnessy (1993) performed a factor analysis on TBS dimension scores and found that two dimensions they labeled as control and affiliativeness emerged. They reported that dominance and conditionality had the highest loadings on the control dimension, and affection and congruence-empathy had the highest loadings on the affiliative dimension. Kivlighan et al. (1994) reported coefficient alphas for the four dimensions of the TBS ranged from .87 for affection to .94 for congruence-empathy for a sample of college students in personal growth groups. Kivlighan et al. (1996) also reported coefficient alphas of .88 for Technical and .92 for Personal in their study of college students in personal grth groups. In the present study, for groups in which a co-facilitation model was used, participants were asked to provide a composite rating of both leaders on the TBS. 51 Group Climate Questionnaire-Short Form (MacKenzie, 1983; Appendix H). The Group Climate Questionnaire—Short Form (GCQ—S) is a 12-item measure used to assess a participant’s perception of the group’s ambience or atmosphere. Items were reported on a seven-point Likert scale indicating degree of agreement, from one (not at all) to seven (extremely). Factor analysis of the 12 items has resulted in the development of four scales entitled engagement (five items), avoidance (four items), conflict (two items) and anxiety (one item). For the purposes of this study only engagement, avoidance, and conflict were examined. These scales reflected the participant’s perceptions of what was taking place in the group. The engagement scale assessed the amount of cohesion and work orientation in the group. The avoidance scale measured the degree to which individuals relied on the other group members or leaders. The conflict scale tapped the amount of interpersonal conflict and distrust. MacKenzie (1983) reported interscale correlations of .44 for Avoidance and Engagement, .18 for Conflict and Engagement, and .30 for Conflict and Avoidance. Example GCQ—S items and the scale they represent included: ‘ The members tried to understand why they do the things they do, tried to reason it out” (Engagement); “The members avoided looking at important issues going on between themselves” (Avoidance); and “There was friction and anger between the members” (Conflict). Kivlighan et al. (1996) reported coefficient alphas for the three scales were: .94 for Engagement, .92 for Avoidance, and .88 for Conflict. Dimensions used from the Group Climate Questionnaire Short Form were group engagement, conflict and avoidance. In the present study, the group engagement dimension exhibited acceptable reliability within the sample (Cronbach alpha of .71). The group conflict dimension was marginally reliable (Cronbach alpha of .57) and the group 52 avoidance dimension exhibited unacceptable reliability (Cronbach alpha of .36) within the present sample. Group Satisfaction Scale (Lopez & Wyssmann, 2000; Appendix I). The Group Satisfaction Scale was a 14-item measure used to assess a participant’s satisfaction with their group experience. Items were reported on a five-point Likert scale indicating degree of agreement, from one (Strongly Disagree) to five (Strongly Agree). Example GSS items included: “Overall, my experiences thus far within this group have been helpful” and “I feel I am important to the functioning of the group.” This scale was developed for the purposes of this study and did not enter into the primary hypotheses. It was used to examine differences among adult attachment orientations in their level of satisfaction with the group experience. This scale exhibited good reliability with a Cronbach alpha of .86. Research Hypotheses The following research hypotheses were intended to operationalize the relationships that were expected between the key independent variables, adult attachment avoidance and anxiety, and the dependent variables. Hypothesis 1. Adult attachment-related anxiety would be negatively related to perceptions of group leader control. This hypothesis could not be tested due to unacceptable reliability of the group leader control dimension within the sample (Cronbach alpha = .41 ). Hypothesis 2. Adult attachment-related avoidance would be negatively related to perceptions of group leader affiliativeness. 53 Hypothesis 3. Adult attachment-related avoidance would be negatively related to perceptions of group engagement. Hypothesis 4. Adult attachment-related anxiety would be positively related to perceptions of group conflict. Hypothesis 5. Adult attachment-related avoidance would be positively related to perceptions of group avoidance. This hypothesis could not be tested due to unacceptable reliability of the group avoidance dimension within the sample (Cronbach alpha = .36). Hypothesis 6. Adult attachment-related avoidance would be negatively related to expectations of commitment to the group. Hypothesis 7. Adult attachment-related avoidance would be negatively related to expectations of the provision of facilitative conditions by the group leader/s and members. Hypothesis 8. Adult attachment-related avoidance would be negatively related to expectations of nurturance by the group leader. 54 Analyses A correlation analysis was used to examine the relationship between scores on the ECR and scores on the dependent measures. A Pearson—Product Moment Correlation was used to examine the relationship between the dimensions of Avoidance and Anxiety from the ECR and scores on: Conflict and Engagement from the GCQ—S; Affiliativeness of the TBS; and Personal Commitment, F acilitative Conditions and Nurturance of the EAC—B. Correlation analyses between the dimensions of Avoidance and Anxiety from the ECR and group leader control from the TBS and group avoidance from the GCQ—S were not conducted due to the poor reliability of the dependent measures within the sample. Post hoc analyses consisted of hierarchical regressions to determine if adult attachment-related anxiety, avoidance and the interaction of these variables might predict participant responses on the dependent measures that demonstrated acceptable reliability within the sample. 55 RESULTS Introduction The results presented below derive from a study designed to explore the relationship of adult attachment with expectations about group therapy, perceptions of group leaders, and perceptions of group climate for participants entering and participating in a therapy group. Assessments were completed and gathered prior to participants entering group therapy and again after the fourth group therapy session. Participants split into two groups: dropouts, those who completed the initial assessments but later dropped out of the study; and continuers, those who completed the study and final assessment. Presentation and analyses of gathered data from the study have been arranged in three major sections. The first major section is intended to familiarize the reader with the demographic background of the research participants across the participating research sites and to present significant correlations with other background variables as well as dimensions of the independent and dependent variables. Within this section are three subsections, the first presents a comparison of the five participating sites on differences regarding the demographic variables and independent and dependent measures. The second subsection presents a demographic comparison between participants who dropped out before the final round of assessments and those participants who completed the study by filling out these assessments. The third subsection examines these dropouts and continuers on dimensions of the independent variable, adult attachment, and dimensions of the expectations about counseling dependent variable. 56 The second major section presents the results of the hypotheses developed for the study examining relations of dimensions of adult attachment with dimensions of the dependent variables (a) perceptions of group leader behavior, (b) perceptions of group climate, and (c) expectations about entering group counseling. This section is further subdivided. The first subsection presents problems encountered within the sample regarding responses to several dimensions of the dependent measures and attempts made to correct these problems. The remaining subsections are arranged according to each of the dependent variables as they relate to adult attachment. The final major section presents post hoc analyses. This is also divided into subsections dealing with dimensions of the dependent variables, analyses regarding satisfaction with the group counseling experience as measured on the Group Satisfaction Scale developed for this study, and changes in attachment-related avoidance over time for the group that completed the study. Summary of Demographic Data for Initial Research Participants across Sites Table 1 presents the results ofa demographic comparison of a group of research participants’ responses to assessments of (a) adult attachment, (b) expectations of group therapy, and (c) perception in group therapy. The research was conducted at one community college and four university counseling centers in the Rocky Mountain region. Initially, 98 people expressed interest in participating in the study and completed preliminary assessments measuring (a) adult attachment and (b) expectations of group therapy. Of the original group, 80 completed the study and the assessment of (c) group therapy perceptions. 57 The group as a whole was older than what might be considered the typical college student with the mean age for these participants being 28.19 years (SD = 8.42). Age was significantly and negatively correlated with dating (r = -.37, p < .01) and marital status (r = -.32, p < .01), meaning the older participants were, the less likely they were to be dating and the more likely they were to be married. There was also a negative and significant relationship between age (r = -.25, p < .01) and expectations of facilitative conditions being provided by the group leader/s and members, and between age and expectations of nurturance by the group leader (r = -.38, p < .01). Of the initial participants, 39 (40%) were male and 59 (60%) were female. Their questionnaire responses revealed a significant correlation between gender (r = .21, p < .04) and attachment-related anxiety measured prior to their entry into the therapy group indicating that female participants experienced higher levels of attachment- related anxiety as they were entering the group. There was a notable lack of ethnic diversity with the sample being predominantly Caucasian/white, number: ng 87 (89%), with 4 (4%) Hispanic/Latino, 2 (2%) Native American, and 5 participants (5%) who classified themselves as other. Due to the small representation of ethnic minority participants in this sample, no analyses were deemed feasible comparing these groups on major variables in the study. The initial research group was comprised of 75 undergraduate students (77%) and 23 participants (23%) who classified themselves as other. Twenty-nine (30%) classified themselves as being in committed relationships, either married or partnered, and 69 (70%) were divorced, widowed or single. There were significant correlations between marital status and expectations, with single persons having greater expectations of (a) 58 commitment to the group (r = .21, p < .04), (b) facilitative conditions being provided by the group leader and members (r = .22, p < .03), and (c) nurturance by the group leader (r = .31, p < .01). Of this initial sample, 63 participants (64%) indicated they were not dating and 3 participants (3%) indicated they had never dated. Using a rating scale from 1 = Very dissatisfied to 5 = Very satisfied, the 41 participants (42%) who had previous group therapy experience rated an average satisfaction with these experiences of 3.83 (SD = .95). Satisfaction with previous group therapy experiences was significantly correlated with satisfaction with previous individual therapy experiences (r = .39, p < .01) and self-rated extraversion (r = .48,p < .01). A large majority of the original participants, (N = 85 or 87%), had previous individual therapy experience and expressed an average satisfaction score with these experiences of 3.93 (SD = .87) on a scale from 1 = Very dissatisfied to 5 = Very satisfied. Satisfaction with previous individual therapy experiences was significantly correlated with (a) expectations of being committed to the group (r = .31, p < .01), (b) expectations of facilitative conditions being provided by the group leader/s and members (r = .27, p < .01), (c) expectations of nurturance by the group leader (r = .34, p < .01), and (d) perceptions of the group leader being affiliative (r = .33, p < .01). Participants, as a group, reported being neither markedly outgoing or reserved by exhibiting an average rating of extraversion of 3.14 (SD = 1.13) with 1 = Very reserved and 5 = Very outgoing. Extraversion was significantly correlated with expectations of being committed to the group (r = .21 , p < .04). 59 This sample presented for counseling with a diverse range of complaints. Problems presented by participants of the therapy groups included (a) relationship difficulty (N = 35 or 36% of sample), (b) disability issues including depression and other psychiatric diagnoses (N = 26 or 27% of sample), and (c) family problems (N = 17 or 17% of sample). Issues concerning academic problems, being a single parent, eating disorder, and sexual abuse were presented less frequently, ranking (N = 7 or 7% of sample), (N = 6 or 6% of sample); (N = 5 or 5% of sample); and (N = 2 or 2% of sample) respectively. The group, as a whole, tended to be quite distressed as seen in average severity rating of these problems of 4.23 (SD = .82) on a scale from 1 = Not at all affecting to 5 = Very significantly affecting. All groups were interpersonal process groups devoted to different themes. These included: Recovery from Childhood; Single Parent; Men’s Issues; Women’s Issues; Disability; General Therapy; Gay Lesbian Bisexual Transgendered (GLBT); and Dream. Comparison of demographic data and responses to dimensions of the independent and dependent variables across participating sites An AN OVA procedure was used to detect differences between the participating sites on demographic variables and dimensions of the ECR and EAC—B. For the most part there were few differences between the sites; however, there were significant between group differences in age between site 1 and site 2 (mean difference = 7.32, SE = 1.78,p < .01) and site 1 and site 4 (mean difference = 7.57, SE = 2.13,p < .01). There were also significant age differences between site 2 and site 3 (mean difference = 12.80, SE = 2.65,p < .01) as well as site 3 and site 4 (mean difference = 13.05, SE = 2.89, 60 p < .01). There was also a significant difference in expectations of nurturance by the group leader at sites 3 and 4 (mean difference = 1.16, SE = .39, p < .03). Comparison of demographic data of therapy continuers and therapy dropouts. Table 2 presents a correlation matrix for all demographic, independent, and dependent variables for the 80 participants who remained in group therapy until completion of the follow up assessment after their fourth group therapy session. A total of 18 participants dropped from the study prior to attending their fourth group therapy session for an overall attrition rate of 18%. A greater proportion of men (87% of beginning men) remained in therapy than did women (78% of beginning women) but women made up a larger proportion (58%) of the final sample of 80 participants. The mean age for therapy continuers was significantly higher (M = 29.10, SD = 8.21) than for therapy dropouts (M = 24.17, SD = 8.39), t(96) = 2.29, p < .02 (two tailed). For continuers, age was significantly and negatively correlated with: dating status (r = -.34, p < .01) meaning that older participants were less likely to be in a dating relationship; marital status (r = -.26, p < .02) indicating older participants were more likely to be married; and expectations of nurturance by group leaders (r = -.28, p < .01). Mean age of dropouts was also significantly and negatively correlated with marital status (r = -.49, p < .04) and expectations of nurturance (r = -.61 , p < .01) but not with dating status. Undergraduates made up a smaller, but not statistically significant, proportion of the therapy-continuing group (74%) when compared to the dropout group (89%). A larger and statistically significant proportion of therapy continuers (71%) (mean difference = .21, SE = .12, p < .05) reported not being in a dating relationship or having never dated compared to therapy dropouts (50%). For continuers, there was a 61 significant correlation with expectations of nurturance by the group leader and marital status (r = .27, p < .02) indicating that unmarried persons expected the group leader to be more nurturant. For dropouts, being unmarried was significantly correlated with satisfaction with previous individual therapy experiences (r = .63, p < .03) and having expectations of facilitative conditions being provided by the group leader/s and members (r = .58,p < .01). A greater, and statistically significant (mean difference = .22, SE = .13, p < .05) proportion of therapy continuers (46%) reported having been in group therapy before than did dropouts (22%). Continuers also reported higher average, but not statistically significant, satisfaction levels (M = 3.86, SD = .98) on a scale from 1 = Very dissatisfied to 5 = Very satisfied with these past group experiences than did therapy dropouts (M = 3.50, SD = .58). Satisfaction with previous group experiences was significantly correlated with satisfaction with previous individual therapy experiences (r = .38, p < .02) for the 37 continuers who had both types of therapy experience. There was also a significant correlation between previous group satisfaction and self-rated extraversion (r = .47, p < .01) for the 37 continuers with previous group experience. A larger proportion of therapy continuers (91%) reported having been in individual therapy before than did dropouts (67%). In addition, continuers reported lower, but not statistically significant, average satisfaction levels (M = 3.92, SD = .89) on a scale from 1 = Very dissatisfied to 5 = Very satisfied with these past experiences in individual therapy than did therapy dropouts (M = 4.00, SD = .74). For the 73 continuers with previous individual therapy experience, satisfaction with these experiences was significantly correlated with expectations of being committed to the group (r = .31, p < .01); expectations of 62 facilitative conditions by the group leader/s and members (r = .25, p < .03); expectations of nurturance by the group leader (r = .32, p < .01); perceptions of the group leader being affiliative (r = .33, p < .01); and satisfaction with the present group experience (r = .24, p < .05). For dropouts, none of these dimensions were significantly correlated with previous satisfaction in individual therapy. Therapy continuers reported themselves to be, on average, more extraverted (M = 3.20, SD = 1.14) compared to dropouts (M = 2.89, SD = 1.08) but this difference was not statistically significant. For continuers, extraversion was significantly correlated with perceptions of the group being engaged (r = .22, p < .05) and self-rated levels of satisfaction with their group (r = .25, p < .03). No significant differences were found between continuers and dropouts regarding the type or severity of presenting problems. Comparison of therapy continuers and therapy dropouts on dimensions of adult attachment and expectations about counseling Dropouts and continuers completed the Experiences in Close Relationships (ECR) measure which was the source of the two predictor variables, ECR Anxiety and Avoidance. Both groups also completed the Expectations About Counseling—Brief Form (EAC-B) that yielded three dependent variables; EAC—B Commitment, EAC—B Facilitative Conditions, and EAC—B Nurturance. Continuers exhibited significantly lower mean attachment-related anxiety (M = 3.72, SD = 1.14), t(96) = -3.51, p < .03 (two- tailed) than dropouts (M = 4.34, SD = .53) but slightly higher, although not statistically significant, attachment related avoidance (M = 4.26, SD = .97) than dropouts (M = 4.05, SD = 1.19). On the dependent measure assessing expectations of counseling, continuers 63 scored lower, but not statistically significantly, on expected commitment to group (M = 5.40, SD = .78), and expectations of nurturance (M = 4.67, SD = .97) than did persons who dropped out (commitment M = 5.62, SD = .68; nurturance M = 5.15, SD = 1.10). Continuers scored significantly lower on expectancies of facilitative conditions (M = 5.30, SD = .83), than did persons who dropped out (M = 5.88, SD = .55), t(96)=-2.81,p< .01. Relationship of Dimensions of Adult Attachment to Perceptions of Group Leader Behavior, Group Climate, and Expectations About Counseling A complement of 80 participants, 34 males and 46 females, completed assessments before the start of their therapy group and after completion of the fourth group session. The Experiences in Close Relationships questionnaire was given prior to entry into group therapy and after the fourth session to ensure stability of self-reported attachment orientation. This measure assessed participant anxiety and avoidance in intimate relationships. Both of these attachment-related dimensions fiom the ECR demonstrated excellent reliability (Cronbach alphas > .80 for anxiety and avoidance) before participation in the group and again after the fourth session. Time 1 and Time 2 ECR anxiety scores were highly correlated (r = .79, p <.Ol) as were Time 1 and Time 2 ECR avoidance scores (r = .75, p < .01). Thus, separate composite scores were derived for the Anxiety and Avoidance dimensions of the ECR by respectively averaging the pre- and post- scores. These composite scores were then correlated with the dependent measures. The dependent measures were dimensions of the Trainer Behavior Scale, Group Climate Questionnaire—Short form, and Expectations About Counseling—Brief Form. 64 The alpha level of statistical probability for analyses of the major hypotheses of this study and seen in Table 1 was set at p < .01 however, correlations between demographic variables and dimensions of the dependent variables are reported in the text atp < .05. Reliability issues related to the dependent measures The scores for this sample on several dimensions of the Trainer Behavior Scale and Group Climate Questionnaire—Short form demonstrated inadequate reliability to justify their inclusion in the correlation and regression analyses. For example, the observed Cronbach alpha for the TBS Control dimension was .41. On the GCQ—S, the Conflict dimension obtained a Cronbach alpha of .57 which was marginally adequate. GCQ—S Avoidance obtained an alpha of .36. In an attempt to improve the reliability of these dimensions within the sample, an item by item analysis was completed to determine if exclusion from the scale would increase the reliability score. These analyses failed to increase the Cronbach alpha for these dimensions within the sample so TBS Control and GCQ—S Avoidance were dropped from the analysis and the hypotheses with these dimensions included were not tested. The marginally adequate reliability of the GCQ—S Conflict dimension did allow for retention in the statistical analysis. 65 Hypotheses regarding the relationship of adult attachment orientations to perceptions of group leader behavior Hypothesis 1 stated that attachment-related anxiety would be negatively related to perceptions of group leader control. Due to the poor reliability of scores on the Trainer Behavior Scale-Control dimension within this sample (Cronbach alpha = .41), this hypothesis could not be tested. Therefore, no conclusions can be drawn about the relationship of these two variables. Hypothesis 2 stated that attachment-related avoidance would be negatively related to perceptions of group leader affiliativeness. A Pearson product-moment correlation (r = .00, p = .50), and reported in Table 1, did not lend statistical support to this hypothesis. Hypotheses regarding the relationship of adult attachment orientations to perceptions of group climate Hypothesis 3 stated that attachment-related avoidance would be negatively related to perceptions of group engagement. The Pearson product-moment correlation analysis, shown in Table 1, did not lend statistical support (r = .07, p = .28) to this hypothesis. Hypothesis 4 asserted that attachment-related anxiety would be positively related to perceptions of group conflict. This hypothesis was also not supported (r = -.O9, p =.42). Hypothesis 5 stated that attachment-related avoidance would be positively related to perceptions of group avoidance. This hypothesis was not tested given that scores on the Group Climate Questionnaire—Short form Avoidance dimension exhibited very poor reliability (Cronbach alpha = .36) within this sample. 66 Hypotheses regarding the relationship of adult attachment orientations to expectations about group counseling Hypothesis 6 stated that attachment-related avoidance would be negatively related to expectations of commitment to the group. The Pearson product-moment correlation seen in Table l (r = .16, p = .08) did not support this hypothesis. Hypothesis 7 stated that attachment-related avoidance would be negatively related to expectations of the provision of facilitative conditions by the group leader/s and members. This hypothesis was not supported (r = .15, p = .08). Hypothesis 8 asserted that attachment-related avoidance would be negatively related to expectations of nurturance by the group leader. The correlation did not support this hypothesis (r = .07, p = .27). Post hoc analyses Since the initial analyses did not find support for hypothesized relationships between single dimensions of adult attachment with dimensions of the dependent variables, it was determined that including attachment-related anxiety and avoidance scores as well as their interaction in a hierarchical regression model predicting scores on the dependent measures might offer a more fine grained way of examining these potentially complex relationships. It is plausible that differing combinations of adult attachment anxiety and avoidance as shown in different adult attachment orientations might be predictive of (a) expectations of group therapy, (b) perceptions of group climate, and (c) group leaders. For example, we could expect that a person with a fearful orientation consisting of high 67 avoidance and anxiety (Bartholomew and Horowitz, 1991) might perceive the group leader to be less nurturant than a group member of different orientation. In conducting these analyses the first step of the regression model entered attachment-related anxiety and avoidance. The interaction of these main effects was entered at the second step. In predicting scores on the dependent measures, scores of anxiety and avoidance from the Experiences in Close Relationships scale were centered prior to creation of the interaction term as recommended by Aiken and West (1991). Due to the poor reliability of scores on the Trainer Behavior Scale—Control dimension within this sample, no analysis was conducted to predict group leader control from attachment-related anxiety and avoidance. The first post hoc hierarchical regression explored if low levels of attachment- related anxiety and avoidance, indicative of a secure attachment orientation (Bartholomew and Horowitz, 1991), would be predictive of higher scores on perceptions of group leader affiliativeness. As seen in Table 3, attachment-related anxiety, avoidance, and the interaction of these main effects did not predict higher scores on group leader affiliativeness from the Trainer Behavior Scale. The second post hoc regression explored whether the three attachment-related predictors would predict scores on perceptions of group engagement as measured by the Group Climate Questionnaire—Short Form. Table 4 shows that the main effects and interaction of attachment-related anxiety and avoidance did not predict higher scores on group engagement. A third post hoc regression investigated whether the three attachment-related predictors would predict scores on perceptions of group conflict as measured on the 68 Group Climate Questionnaire—Short form. Table 5 shows that the main effects and interaction of attachment-related anxiety and avoidance did not predict higher scores on group conflict. Due to the poor reliability of scores within the sample on the avoidance dimension of the Group Climate Questionnaire—Short Form (Cronbach alpha = .3 6) no analysis was completed for predicting group avoidance from the dimensions of adult attachment. A fourth hierarchical regression explored whether high levels of attachment- related anxiety and avoidance, indicative of a fearful attachment orientation (Bartholomew and Horowitz, 1991), would predict lower levels of commitment to the group as measured by the Expectations About Counseling—Brief Form. As seen in Table 6, this analysis indicated that attachment anxiety, avoidance and their interaction did not significantly predict expectations of commitment to the group. A fifth hierarchical regression explored similar expectations with regard to perceptions of facilitative conditions as measured by the Expectations About Counseling— Brief Form. Table 7 presents the results of this analysis and shows that attachment-related anxiety, avoidance and their interaction did not significantly predict these expectations. The final post hoc hierarchical regression explored whether low levels of adult attachment related anxiety and avoidance, indicative of a secure attachment orientation, would predict higher scores on expectations of group leader nurturance as measured by the Expectations About Counseling—Brief Form. This contention was not supported as indicated in Table 8. 69 Another post hoc analysis, using Pearson correlations, examined relationships between scores on the Group Satisfaction Scale, which was specifically developed for this study, and the dependent measures. Significant correlations were found between scores on the GSS and engagement from the GCQ—S (r = .46, p < .01); GCQ—S conflict (r = -.31, p < .01); TBS affiliativeness (r = .32, p < .01); EAC commitment (r = .27,p < .02), EAC facilitative conditions (r = .24, p < .04), and EAC nurturance (r = .26, p < .02). In addition, a paired samples t-test was performed to determine if there had been a significant change in attachment-related avoidance and anxiety from the beginning of the group to the fourth group session. This analysis indicated that attachment-related anxiety did not significantly change over this period but attachment avoidance did from M = 4.26, SD = .97 before the start of the group to M = 4.02, SD = 1.06 after the group t(79) = 3.04, p < .001, two—tailed. A discrepancy score was then generated for the sample by subtracting the attachment-related avoidance scores from after the fourth group session from the avoidance scores exhibited prior to entry into group therapy. A low discrepancy score indicated that the avoidance score had not changed a great deal between entry and the final assessment with high discrepancy scores indicating a marked change in avoidance. This discrepancy score was then correlated with demographic variables and independent variables that exhibited adequate Cronbach alphas (> .70). This analysis resulted in significant correlations between the avoidance discrepancy score and average problem severity at entry into group therapy (r = .22, p < .05) as well as satisfaction with previous group therapy experiences (r = .43, p < .01). This indicated that higher ratings of personal distress and satisfaction with past group therapy experiences at entry into group were associated with greater decreases in avoidance during the course of group therapy. 70 There were no other significant correlations between the avoidance discrepancy score with any dimensions of the group process measures. 71 Table 1 Demographic Comparison of Participating Sites Site 1 Site 2 Site 3 Site 4 Site 5 Participants per Site Number of participants 37 30 10 17 4 Male/Female 14/23 8/22 7/3 8/9 2/2 Number of dropouts 4 8 0 6 0 Age of Participants Mean 31.22 23.90 36.70 23.65 30.50 Range 18-50 18-35 29-50 18-40 20-49 Standard Deviation 9.09 4.44 6.78 5.12 13.03 Ethnicity Caucasian 33 25 10 16 3 Hispanic 1 2 0 1 0 Native American 2 0 0 0 0 Other 1 3 0 0 1 Academic Rank Freshman 12 3 0 1 0 Sophomore 14 7 4 3 1 Junior 7 7 0 3 1 Senior 2 2 0 7 1 Other 2 11 6 3 1 72 Table 1 continued Demographic Comparison of Participating Sites Site 1 Site 2 Site 3 Site 4 Site 5 Relationship Status Manied 7 3 4 5 0 Partnered 2 6 1 1 O Divorced 10 1 2 0 2 Widowed 1 0 l 0 0 Single 17 20 2 11 2 Dating Status Never dated 1 2 0 O 0 Not dating 26 15 9 11 2 Dating 1 person 9 12 1 5 l Dating>1 person 1 1 O l 1 Prior Group Experience Yes/No 15/22 13/17 6/4 6/11 1/3 Prior Group Therapy Satisfaction Mean 3.67 4.00 4.17 3.33 5.00 Standard Deviation 1.11 1.00 .41 .52 - Prior Individual Therapy Experience Yes/No 31/6 28/2 9/1 13/4 4/0 73 Table 1 continued Demographic Comparison of Participating Sites Site 1 Site 2 Site 3 Site 4 Site 5 Prior Individual Therapy Satisfaction Mean 3.90 4.11 3.56 3.69 4.50 Standard Deviation .91 .88 .88 .75 .58 Extraversion Mean 3.27 3.00 3.50 2.76 3.75 Standard Deviation 1.22 1.08 .97 1.03 1.26 Problem Type Relationship 8 13 5 6 3 Academic 5 1 1 0 0 Family 5 6 1 5 0 Single Parent 5 l 0 O 0 Disability l3 7 3 2 1 Sexual Trauma 1 1 0 0 0 Eating Disorder 0 1 0 4 0 Problem Severity Mean 4.27 4.23 4.20 4.06 4.75 Standard Deviation .77 .82 .92 .97 .50 74 Table 1 continued Demographic Comparison of Participating Sites Site 1 Site 2 Site 3 Site 4 Site 5 Type of Group RFC X Single Parent X Men’s X Women’s X X Disability X General Therapy X X X X GLBT X Dream X ACOA X Couples X Note. Prior Individual Therapy Satisfaction rated on 5 point scale (l=very dissatisfied, 5=very satisfied). Average Extraversion rated on 5 point scale (1=Very reserved, 5=Very outgoing). Problem severity rated on 5 point scale (1+Not at all affecting, 5+Very significantly affecting). RFC+Recovery From Childhood. ACOA+Adult Children of Alcoholics. 75 Table 2 Intercorrelation Matrix Variable M SD Range 1 2 3 4 5 6 7 l.Age 29.10 8.21 18-50 - .07 -.34** -.26* -.11 -.O3 -.10 2.Gender .58 .50 0-1 - .04 -.12 .09 .15 -.07 3. Dating Status .29 .46 0-1 - .12 .06 -.08 -.19 4. Marital Status .70 .46 0-1 - .03 .08 -.02 5.G-SAT 3.86 .98 1-5 - .38* .24 6. I-SAT 3.92 .89 2-5 - .24* 7. GSS 3.97 .60 2.14-5.00 - 8. Extraversion 3.20 1.14 1-5 9. P-SERV 4.27 .75 2-5 10. Anxiety 3.64 1.07 1.28-5.97 11. Avoid 4.14 .95 1.44-6.86 12. TBS-A 4.24 .47 3-5 13. GCQS-E 5.44 .86 3-7 14. EACB-C 5.40 .78 3-7 15. EACB-F 5.30 .83 3-7 16. EACB-N 4.67 .97 3-7 Note. Gender code: 0 = male, 1 = female; Dating status code: 0 = Not dating, 1 = Dating; Marital status code: 0 = Married, 1 = Not married; GSat = Satisfaction with previous group experiences; ISat = Satisfaction with previous individual therapy experiences; GSS = Present group satisfaction; Pserv = Problem Severity; TBS-A = Trainer Behavior Scale Affiliativeness; GCQS- E = Group Climate Questionnaire Short form Engagement; EACB = Expectations About Counseling Brief Form- C = Commitment/ F = Facilitative Conditions/ N = Nurturance. "' p < .05 **p<.01 76 Table 2 continued Intercorrelation Matrix Variable 8 9 10 ll 12 l3 14 15 16 1. Age .06 .00 .08 -.13 .14 -.15 -.15 -.18 -.28** 2. Gender .20 -.06 .30“ -.09 .19 .21 .07 .01 -.02 3.Dating Status -.02 .06 -.18 .01 .03 .04 -.01 -.08 -.01 4. Marital Status -.01 -.02 .19 .11 .12 -.03 .20 .14 .27* 5. G-SAT .47" .16 .17 -.12 .23 .10 .21 .01 .24 6. I-SAT .21 .16 .09 -.18 .33" .09 .31" 25* .32" 7. GSS .25* .06 .02 .Ol .32” .46" .27* .24* .26* 8. Extraversion - -. 18 -.17 -.18 .20 .22* .20 .06 .11 9. P-SERV - -.06 .09 .05 .04 .10 .05 .09 10. Anxiety - .04 .17 .03 -. 14 -.O3 .04 11. Avoid - .00 .07 .16 .16 .07 12. TBS-A .32" .12 .20 .22 13. GCQS-E - .27* .18 .14 14. EACB-C - .74" .69" 15. EACB-F - .83" 16. EACB-N - 77 Table 3 Summary of Hierarchical Regression Analysis for Variables Predicting Trainer Behavior Scale Afliliativeness (N = 80) Variable B SEB Beta Step 1 ECR Avoidance .00 .06 -.Ol ECR Anxiety .07 .05 .17 Step 2 ECR Avoidance .01 .06 .02 ECR Anxiety .09 .05 .22 Interaction .07 .05 .15 Note. N = 80. R2 = .03 for Step 1; AR2 = .02. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores fi'om the ECR were centered prior to creation of the interaction term of these main effects. 78 Table 4 Summary of Hierarchical Regression Analysis for Variables Predicting Engagement from the Group Climate Questionnaire—Short form (N = 80) Variable B SEB Beta Step 1 ECR Avoidance .06 .10 .07 ECR Anxiety .02 .09 .03 Step 2 ECR Avoidance .08 .10 .09 ECR Anxiety .06 .10 .07 Interaction . 1 1 . 10 . 14 Note. N = 80. R2 = .01 for Step 1; AR2 = .02. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores fi'om the ECR were centered prior to creation of the interaction term of these main effects. 79 Table 5 Summary of Hierarchical Regression Analysis for Variables Predicting Conflict from the Group Climate Questionnaire—Short form (N = 80) Variable B SEB Beta Step 1 ECR Avoidance .12 .12 .11 ECR Anxiety -.05 .11 -.06 Step 2 ECR Avoidance .12 .13 .11 ECR Anxiety -.06 .12 -.06 Interaction -.02 . 12 -.02 Note. N = 80. R2 = .05 for Step 1; AR2 = .00. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores from the ECR were centered prior to creation of the interaction term of these main effects. 80 Table 6 Summary of Hierarchical Regression Analysis for Variables Predicting Commitment from the Expectations About C ounseling—Brief Form (N = 80) Variable B SEB Beta Step 1 ECR Avoidance . 14 .09 .17 ECR Anxiety -.11 .08 -.15 Step 2 ECR Avoidance .14 .09 .17 ECR Anxiety -. 10 .09 -.13 Interaction .04 .09 .05 Note. N = 80. R2 = .05 for Step 1; AR2 = .00. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores from the ECR were centered prior to creation of the interaction term of these main effects. 81 Table 7 Summary of Hierarchical Regression Analysis for Variables Predicting F acilitative Conditions from the Expectations About Counseling-Brief Form (N = 80) Variable B SEB Beta Step 1 ECR Avoidance .14 .10 .16 ECR Anxiety -.03 .09 "04 Step 2 ECR Avoidance .14 .10 .16 ECR Anxiety -.03 .09 -.04 Interaction -.01 .10 -.01 Note. N = 80. R2 = .03 for Step 1; AR2 = .00. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores from the ECR were centered prior to creation of the interaction term of these main effects. 82 Table 8 Summary of Hierarchical Regression Analysis for Variables Predicting Nurturance from the Expectations About Counseling—Brief Form (N = 80) Variable B SEB Beta Step 1 ECR Avoidance .07 .12 .07 ECR Anxiety .04 .10 .04 Step 2 ECR. Avoidance .07 .12 .06 ECR Anxiety .02 .l 1 .03 Interaction -.04 .1 1 -.04 Note. N = 80. R2 = .01 for Step 1; AR2 = .00. ECR = Experiences in Close Relationships, Avoidance and Anxiety scores from the ECR were centered prior to creation of the interaction term of these main effects. 83 DISCUSSION Introduction This section begins with a summary of the findings concerning each of the major hypotheses of the study followed by discussion of the failure of two dependent measures to exhibit acceptable reliability within this sample. Next, some reasons why none of the research hypotheses received support will be discussed prior to a specific exploration of possible reasons for the lack of support unique to each hypothesis. Post-hoe findings will then be considered. Lastly, the strengths and limitations of the study as well as directions and recommendations for future research will be discussed. Research findings regarding major hypotheses of the study ' Hypothesis 1 proposed that attachment-related anxiety would be negatively related to perceptions of group leader control. This hypothesis could not be tested due to unreliability of the dependent measure (Trainer Behavior Scale—Control) within the sample. Hypotheses 2 and 3 stated that attachment-related avoidance would be negatively related to perceptions of group leader affrliativeness and to perceptions of group engagement, respectively. Neither hypothesis was supported. Hypothesis 4 anticipated that attachment-related anxiety would be positively related to perceptions of group conflict. This hypothesis was not supported. Hypothesis 5 stated that attachment-related avoidance would be positively related to perceptions of group avoidance. This hypothesis could not be tested due to unreliability of the dependent measure (Group Climate Questionnaire—Short Form Avoidance) within the sample. 84 Hypotheses 6, 7, and 8 proposed that attachment-related avoidance would be negatively related to expectations of commitment to the group, of expectations of the provision facilitative conditions by the group leader/s and members, and of expectations of nurturance by the group leader/s, respectively. These hypotheses also failed to be supported in this study. Reliability concerns with dependent measures Two of the hypotheses (numbers 1 and 5) could not be tested due to unreliability of the dependent variables (Trainer Behavior Scale Control and Group Climate Questionnaire—Short Form Avoidance) within this sample. Despite attempts to increase the reliability of these two scales by deleting items, reliability estimates remained unacceptable. The control dimension of the Trainer Behavior Scale (Hypothesis 1) exhibited poor reliability within this sample indicating this scale did not reliably assess leader control. This is in contrast to Kivlighan et al.’s (1994) finding in which the control dimension of the TBS had a coefficient alpha above .87 for a sample of college students in personal grth groups. Although the participants of the present study were in actual therapy groups versus personal growth groups (as in the Kivlighan study), it was surprising to find this dimension to be unreliable with this sample as Bolman (1971) had successfully used it with hospital outpatients. There are, however, several possible reasons for the failure of this dimension within the sample. First, the scale was very brief, being composed of just 6 questions that focused on the leader trying to exert influence on the group and his being approving/disapproving of group member behavior. This brevity may not have allowed 85 for a sufficiently broad response set by participants to group leader behavior. Second, given that the final assessment was conducted relatively early (after the fourth session) in the life of the group, it may be that members felt they simply didn’t have enough information about the leader/s to make these kinds of judgments. In addition, two subscales (conditionality and dominance) that compose the control dimension were phrased with a negative valence (i.e., “The leader is punishing. . .”). The negative connotation may have influenced some members to be reluctant to be critical of the leader at such an early stage of the group. Third, the instruction to members whose groups were co-facilitated to average the scores for both leaders may have been confusing. Fourth, some group members may also have been concerned that the leader might see their responses despite having been advised that this would not occur. Fifth, the range in age of group participants, from 18-50, may have contributed to greater variability within the sample and partially accounted for the failure of this measure. Finally, Groth—Marnat (1990) stated that boredom, fatigue, and the mood of respondents can impact responses on such measures. These factors may have been especially important since the assessment occurred after the completion of a one-and-one-half hour group therapy session. The avoidance dimension of the Group Climate Questionnaire-Short Form (Hypothesis 5) also did not reliably assess the group members’ avoidance of therapeutic issues within the sample. This measure deals with themes of avoidance, dependence, and social acceptability. In two earlier studies, Kivlighan et al. (1996) and Kivlighan and Goldfine (1991) reported coefficient alphas of .92 and .88 respectively so clearly, the results of the Cronbach alpha analysis (.36) of the present study is surprising. A possible 86 reason for the problem with this scale was its brevity (4 items). As with the TBS—Control dimension, this scale was administered early (after the fourth session) in the group formation process, so members may not have had sufficient time to ascertain the level of avoidance they perceived in the group. The early final assessment (prior to the beginning of the norrning phase) may also have impacted responses on this scale, as some members may have felt freer to express divergent opinions than others. Again, age range, mood, fatigue, and boredom may have increased the measurement error, since this assessment was also conducted after a one-and-one-half hour group therapy session. In addition, there can be no guarantee that participants did not misread the items and react differently to the assessment atmosphere. General discussion Perhaps the most striking result of the present study is that not a single research hypothesis was supported. There are a number of possible reasons for not finding a single significant result. Perhaps adult attachment orientation in intimate relationships does not provide the proper lens for viewing members’ perceptions of less intimate relationships, such as that with a group leader and other group members. Indeed, Smith, Murphy, and Coats (1999) asserted that attachment to a relationship partner and attachment to social groups are not conceptually similar. Other authors such as Baldwin et a1. (1996), Collins and Read (1994), and Pierce and Lyddon (1998, 2001) have also questioned whether romantic attachment can be used as a viable framework for viewing expectations of and perceptions in group therapy, stating that people have multiple models that change in accessibility depending on the context. It seems possible that attachment to a therapy group is distinct and secondary from other forms of attachment. Another possible reason 87 for the lack of support for the research hypotheses is that, as many participants acknowledged having previous group therapy experience, they may not have experienced entry into the group as a “strange situation.” This prior group experience may also have resulted in expectations of group counseling that were less influenced by group members’ adult attachment orientations. It also may be that the relatively brief time (4 therapy sessions) between pre- and post- assessment may not have allowed members sufficient time to gather consistent perceptions of group leaders and the group climate. In addition, the Experiences in Close Relationships scale is clearly focused on present romantic relationships. Since a large number of participants were single and not in dating or intimate relationships, they had to answer this scale in an “as if” manner, which may have detracted from the validity of responses. Adult attachment orientation and perceptions of group leader behavior Findings failed to reveal a relationship between adult attachment-related avoidance and perceptions of group leader affiliativeness (Hypothesis 2). This result is surprising in that one could argue that ratings of leader affiliativeness in a therapy group are likely to be similar to ratings of the therapeutic alliance in individual therapy. Prior research has clearly established a link between adult attachment and ratings of the therapeutic alliance in individual therapy. For example, Mallinckrodt, Coble, and Gantt (1995) reported the ability to form adult attachments was significantly associated with positive ratings of the working alliance in individual therapy. In addition, Satterfield and Lyddon (1995) found that persons with low levels of adult attachment related trust, as related to the perceived availability and dependability of others, evaluated the counseling relationship in more negative terms in the early phase of individual counseling than did 88 their more trusting peers. In light of these findings, and of Pistole’s (1997) theoretical assertion that client attachment to counselor may occur to the same extent in groups as in individual therapy, it is unexpected to find no relation between adult attachment and perceived group leader affiliativeness. In addition to the reasons outlined in the general discussion, there may be other reasons for the lack of support for this hypothesis. The study design allowed neither control over selection of therapist/s nor control of the manner in which they facilitated their group/s, so the variability in group leader/s and their leadership style may have contributed to the lack of relationship. As mentioned previously, some groups were co- facilitated by a leader and a student, so group members may have had difficulty in making judgments when they were required to submit an averaged affiliativeness score for the pair of leaders. Also, participants may have felt compelled to respond more favorably on this questionnaire due to the $10.00 reimbursement as they may have seen it as incentive to rate their leaders highly. Finally, it may be that members were more focused on their relationships with other group members rather than with the group leader/s. Adult attachment orientations and perceptions of group climate The expectation that adult attachment-related avoidance would be negatively related to perceptions of group engagement (Hypothesis 3) was not supported. Again, this outcome is surprising considering the literature cited that found links between adult attachment and phenomena one would expect to be relevant to perceptions of group engagement described by MacKenzie (1983) as “a positive working atmosphere characterized by group cohesion, self-disclosure, interpersonal understanding, and 89 interpersonal learning”. For example, one could expect social functioning and competence, recollection of family, and self-disclosure to be related to perceptions of group engagement. Diehl, Elnick, Bourbeau, and Labouvie—Vief (1998) found a positive relationship between secure adult attachment and social functioning, self-confidence, psychological health and favorableness of family description. Mallinckrodt, Coble, and Gantt (1995) detailed significant relations between capability for secure adult attachment and social competence. Engels, Finkenauer, Meeus, and Dekovic' (2001) found parental attachment, measured in terms of trust and communication with parents, positively related to social skills, relational competence and emotional adjustment with adolescents aged 15-18. Duggan and Brennan (1994) and Bartholomew (1990) reported adult attachment style to be related to social avoidance, with fearful and dismissive persons being more socially avoidant than their secure and preoccupied peers. Finally, Mikulincer and Nachson (1991) found a positive relationship between adult attachment and self disclosure. Persons with positive views of other (secure and ambivalent) evidenced more self disclosure, felt better interacting with, and were more attracted to a high disclosing partner than peers with negative view of other (avoidant). Although all of these studies supported the examination of adult attachment-related avoidance relative to perceptions of group engagement (albeit some indirectly), the simple fact is that, in the present study, there was no relationship exhibited between these dimensions. It may, however, suggest that there were factors inherent in the present study that impacted the study results. For example, the group formation process and the stage of development the group was in may have influenced interpersonal perceptions. Theoretically, the groups should 9O have been in the storming (Tuckman, 1965) phase of group formation by the time the final assessment was taken. Ifthis were the case, it is possible that the members viewed the normal conflict of this stage as indicative of alienation instead of engagement. Additional factors derive from fluctuating group membership. Since there was no control over member attendance, some participants may have missed one or more sessions prior to the final assessment. It is possible that the group formation process was impeded by these absences, thereby causing an attenuation of engagement scores. Similarly, some of the groups added members after the initial, but before the fourth, session which could also have had an unforeseen impact on the group formation process. However, the group average of perceptions of being engaged was quite high for the study as a whole (5.4 on a seven point scale). This rating that would seem to contradict this possibility. The relative elevation of group engagement may also reflect the previous group experience many of the participants had and speak to a tendency for them to self- select into group therapy. Indeed, of the final sample of 80 participants, no participant rated engagement lower than a three on the seven point scale. This range restriction may also have attenuated the expected relationship between adult attachment avoidance and perceptions of group engagement. There was also no support for the expectation that attachment-related anxiety would be positively related to perceptions of group conflict (Hypothesis 4). In this instance, the dependent measure assessing group conflict exhibited weak reliability (Cronbach alpha = .57) within the sample. Thus, this result is reported cautiously. Again, a number of authors have provided support for the investigation of adult attachment and perceptions of group conflict. For example, Horowitz, Rosenberg, and Bartholomew 91 (1993) noted a relationship between adult attachment style problems in the area of hostile dominance which presumably would lead to elevations in member ratings of group conflict. Mikulincer (1998) found that persons with a secure adult attachment style were less prone toward anger and expected more favorable outcomes from episodes of anger than did their insecure counterparts. Niedenthal, Brauer, Robin, and Innes-Ker (2002) described a positive relationship between adult attachment and perception of emotion including anger. They found that when watching a movie of a computerized face moving from a negative facial expression to one of neutrality, secure participants under stress perceived emotional neutrality sooner than did their similarly stressed insecure counterparts. These studies provided evidence for a hypothetical link between adult attachment-related anxiety and perceptions of conflict. Consequently, the lack of significant findings is puzzling. Again, there are several possibilities as to the source of this lack of support for the research hypothesis. One possibility is that the groups did not reach the storming phase. As this, theoretically, should have begun with the fourth session, there may have been very little actual conflict. Indeed, the data suggest that participants perceived very little conflict in their groups as evidenced by the majority of participants indicating “not at all” in response to questions inquiring into the presence of symptoms of conflict. The conflict dimension was Likert scaled and ranged from one (not at all) to seven (extremely) and no participant responded above a score of five for any of the questions dealing with perceptions of group conflict. The average perceived group conflict across the entire sample was 1.8. As noted with perceptions of group engagement, some groups continued to add members to the original cohort. This may have served to attenuate the group 92 development process thereby postponing the onset of the group storming phase resulting in low conflict scores. In addition, since there was no control over the enactment of group therapy by the leader/s it is possible that some leaders, especially novices, may have sought to actively diminish discord and conflict. It was also noted that over half of the therapy groups centered on a particular theme which may have served to promote cohesion among the members initially and to limit conflict since they all entered with a common problem around which they could rally. The expectation that adult attachment-related avoidance would be positively related to perceptions of group avoidance (Hypothesis 5) could not be tested due to the inadequate reliability of the avoidance dimension of the Group Climate Questionnaire— Short Form within this sample (Cronbach alpha = .36). The brevity (four items) of the scale may have impacted the ability to accurately assess perceptions of avoidance. Although this scale was to assess a one-dimensional construct of avoidance, the 4 items were not significantly correlated with one another indicating that, within this sample, the questions assessed multiple dimensions of perceptions of the group. One question directly assessed avoidance of therapeutic themes but the other three examined perceived (a) group dependence on the leader/s for direction, (b) interpersonal distance among members, and (c) group adherence to acceptable behavior. This analysis appeared to conflnn that the scale was not assessing perceived avoidance. In fact, of the four items only two were significantly correlated with one another and two were actually negatively correlated with one another. 93 Adult attachment orientations and expectations about group therapy The predicted link between adult attachment-related avoidance and expectations of commitment to the group (Hypothesis 6) was also not supported. Although there is no literature linking adult attachment orientations and expectations of being committed to therapy, some of the previously cited studies provided indirect support for looking at this hypothesized relationship. One would expect adult attachment-related abilities such as investing trust in others and effective stress-related coping and problem solving to be predictive of commitment to group therapy as persons with these abilities are apt to see the group as less threatening. As previously noted, Mikulincer (1998) described positive relations between adult attachment security and ability to trust in intimate relationships. Mikulincer and Florian (1995) and Mikulincer, Florian, and Weller (1993) found that securely attached research participants engaged in more effective stress-related coping than did their insecure peers. Lopez et al. (1997) and Pistole (1989) found significant relations between security of adult attachment and collaborative problem solving abilities. Again, there are several possibilities as to the source of this lack of significant results. One possibility is that since a large proportion of the group members had previous therapy experience, they knew ‘fiivhat they were getting into” and thus, any attachment- related effects on commitment may have been minimized. It also is possible that group screeners may not have adequately prepared participants for the group therapy experience as there was no control over the screening process at the sites. If this is true, it may be that some group members adopted an “I’ll wait and see” attitude toward continuing group participation. This study affirmed previous clinical knowledge that it is extremely 94 difficult to form and continue therapy groups, let alone a strong group therapy program, in college counseling centers as the coordination of schedules among busy students is extremely problematic. Thus, it could have been that some of the participants may have made a commitment to attendance without taking into consideration the many work and relational demands that might take priority after they began the group experience. The prediction that adult attachment-related avoidance would be negatively related to expectations of provision of facilitative conditions by the group leader/s and members (Hypothesis 7) was not supported. These facilitative conditions might reasonably be thought of as similar to the expected group climate. It seems plausible that the view one has of self and others would powerfully impact the perceptions of these facilitative or “climatic” conditions being provided by the group. Kivlighan and Angelone (1992) found that group members who perceived themselves as too dominant viewed the group climate as more avoiding and tense. Marcus and Holahan (1994) also presented support for the link between view of self and other and group member interpersonal perception. They found that participants’ view of others (similar to working model of other in an adult attachment fi'amework) was significantly related to perceptions of others’ friendliness, submissiveness, dominance, and hostility. Although this prior research supports examining the hypothesized relationship, the lack of results was disappointing. Several possibilities exist as to the source of this lack of significant findings. The average expectation of facilitative conditions was quite high for this group of participants (greater than five on a seven point scale). In fact, none of the 98 initial participants indicated “no” or “slight” (one and two respectively on the seven point Likert scale) expectations of facilitative conditions. This finding makes intuitive sense in that no 95 therapist is going to refer a client who they assume will have low expectations for the likely presence of these facilitative conditions. It also may be that the previous group therapy many participants had experienced may have been very favorable thus, leading to high expectations (and a restricted range of) facilitative conditions. Another possibility is that members may have been unduly influenced in their expectations of facilitative conditions by a referring therapist who saw group therapy as a favorable option. Indeed, clinicians are advised that “selling” the idea of group participation is desirable in recruiting new members into group. The fact that many of the group members were advised they were entering groups that were particularly devoted to the problem they presented with may also have favorably influenced their expectations of facilitative conditions. Finally, the prediction of a negative relationship between adult attachment-related avoidance and expectations of nurturance by the group leader (Hypothesis 8) was not supported. The expectation of nurturance and support is a critical component of the therapeutic relationship and, as previously noted, several authors have found relationships between individual differences, attachment and ratings of the therapeutic alliance. As cited for Hypothesis 2, Mallinckrodt, Coble, and Gantt (1995) found a positive relationship between the ability to form adult attachments and ratings of the working alliance. Satterfield and Lyddon (1995) found adult attachment-related trust related to evaluations of the counseling relationship in the early phase of individual counseling. Sauer (1999) reported adult attachment-related security to be related to ratings of the therapeutic alliance as made by client and counselor. Mallinckrodt, Gantt, and Coble (1995) also noted significant findings between adult attachment security and positive 96 ratings of the therapeutic alliance. The fact that this hypothesis was not supported is all the more surprising considering Pistole’s (1997) contention that the relationship with the group therapist is the focus for group members in the early stage of group therapy. Elsewhere, Brykczynska (1990) suggested that perception of the therapist might be more important in individual than in group therapy, and that the relationships group members have with one another are more important than their relationships with the therapist. One of the possible reasons this hypothesis was not supported is, again, the prior group experience of this cohort. It may have served to restrict the range of scores on the expectation of nurturance dimension. As with facilitative conditions, the average expectation of nurturance was elevated (mean = 4.76) for this group of participants. And similar to the facilitative conditions dimension, none of the 98 initial participants indicated “no” or “slight” (one and two respectively on the seven point Likert scale) expectations of nurturance by the group leader. Participants may also have rated the expectation of nurturance more highly based on an assumption the therapeutic relationship would be the same in tone as with an individual therapist. Other findings In this study almost one-fifth of the original participant pool dropped out of therapy prior to finishing their fourth session. This rate compares favorably with the range of attrition from group therapy (28-57%) that Yalom (1995) reports from University outpatient clinic research. It was noted that a larger proportion of dropouts from the present study, when compared to continuers, had not previously participated in group therapy. Although there was no noted relationship between drop status and commitment to group, these dropouts may have been due to inadequate preparation by 97 staff, heightened anxiety, or unrealistic assessment of the required commitment. The average age of the continuers was higher than what would be expected of a typical undergraduate pool of participants. Reasons for this include the fact that one of the sites (the community college) has an older average age of students (greater than 31 years of age) in general and more returning students who gravitated toward group therapy. It may also be that older students were more comfortable being in groups as they may have viewed group participation as similar to working within a family structure. The significant correlation between age and marital status (older students were more likely to be married) may support this. In addition, older students are further distanced from the need for individuation from family or other groups and may have been more willing to participate in non-homogeneous but cooperative groups. Surprisingly, older participants had significantly lower expectations of nurturance by group leaders and the provision of facilitative conditions by fellow group members and group leaders. Perhaps older participants saw nurturance as being a more parental function and did not readily assign this expectation to group leaders since they were more likely to be of the same or nearly same age. Older participants may have had lower expectations of facilitative conditions due to such factors as (a) believing they were likely to be the oldest member in the group and thus, having little in common with other members aside from presenting problem due to life experience, (b) having lived with their presenting problem for longer so believing less in others’ abilities and desire to help them, or (c) expecting greater self reliance due to having more time in which they were expected to take care of their own and others’ needs. 98 A large proportion of the sample was not in a committed relationship and it seems possible that they may have sought out the group experience to satisfy unmet intimacy needs. However, if this were the case, one would expect that adult attachment orientations would have been related to at least some of the dependent measures. A large majority of the initial sample (87%) had previous individual therapy experience, and, not surprisingly, satisfaction with previous individual therapy experiences was significantly correlated with satisfaction with previous group therapy experiences. Satisfaction with previous individual therapy experiences was also significantly correlated with (a) expectations of being committed to the group, (b) expectations of facilitative conditions being provided by the group members and group leader, (c) expectations of nurturance by the group leader, and (d) perceptions of the group leader being affiliative. These findings suggest that a positive relationship with an individual therapist may bias potential group participants’ expectations about how group members and leaders will be and what they expect to have provided to them. It also clearly has a positive impact on expressed commitment to being a group member. Given these findings it seems reasonable to select potential group members from a pool of participants who have reported being very satisfied and happy with past individual therapy experiences. For those with previous group therapy experience, satisfaction with these experiences was significantly correlated with self-rated extraversion indicating that participants who perceived themselves as more outgoing experienced greater satisfaction with past groups. It was also noted that participants who persisted in group therapy were significantly older than dropouts. This may reflect a simple difference in maturity, an 99 ability to prioritize due to life experience, or a lack of social support that often accompanies older students on campus. Continuers also exhibited significantly lower mean attachment-related anxiety than dropouts at the initial assessment. Although adult attachment-related anxiety was not significantly related to commitment to the group therapy experience, it does seem that this anxiety may have contributed in some way to dropping out. Though continuers exhibited lower mean attachment-related anxiety, this anxiety did not change significantly over the course of the four session time span. However, adult attachment-related avoidance did change significantly. This change in avoidance for continuers was significantly related to the severity of presenting problem, indicating that the higher the participant’s reported initial distress, the greater was the change in attachment-related avoidance. In addition, participants with previous group therapy experience exhibited greater alleviation of this avoidance than did participants with no prior group therapy experience. Although participation in group did not significantly impact adult attachment-related anxiety, it appeared that participation in group therapy helped to alleviate adult attachment-related avoidance overtime. Perhaps this avoidance decreased from participants’ experience of acceptance by diverse others as well as a caring leader. Finally, the Group Satisfaction Scale which was designed for this study was significantly correlated with the following scales: (a) GCQ—S engagement, (b) TBS affiliativeness, (c) EAC commitment, (d) EAC facilitative conditions, (e) EAC nurturance and (d) self-rated extraversion. The results of this correlation analysis indicate that participants who were highly satisfied with the present group experience were more likely to: (a) view their group as involved with one another and intimate, (b) see the 100 leader/s as desiring a bond and closeness, (c) expect to be committed to the group experience, ((1) expect the leader/s and other members to provide conditions favorable for personal growth, and (e) view the group leader/s as nurturing. Although the results of this study failed to support any of the major hypotheses, this final finding would appear to reinforce the notion that persons tend to get what they expect. If they enter a new situation with positive outcome expectations, they rate the outcome positively. Strengths of the study There are a number of strengths of this study including the fact that the sample was composed of an actual, distressed client base with a diverse set of problems. In addition, there was a broader age range to the sample than might be expected at most college counseling centers. The rationale for the study was theoretically sound, and the constructs examined are considered crucial to group therapy processes. It is also relevant to the expected rise in the use of group therapy in the future due to cost-containment measures adopted by managed care forces. Finally, although several measures performed poorly within this sample they have, in general, been usefully employed in past research. There was also an attempt to gain a longitudinal view of adult attachment as seen in pre- and post-assessments during the course of a therapy group. Limitations There are several limitations to the present research, perhaps the most far- reaching in overall effect being the lack of control over the five research sites. As a result, there is no guarantee that therapy was standardized. Since the study was conducted at five different sites, the definition of interpersonal process therapy group may have differed somewhat among the centers. The researcher depended on the site supervisors to advise 101 as to the nature of the group. There may also have been variation among leaders as to how they facilitated the interpersonal process group they were in charge of. All center directors acknowledged the difficulty the centers had in initiating therapy groups and with the ambitiousness of doing longitudinal data collection with participants who agreed to participate. In addition, the fact that all participants were college/university students may indicate the sample was unlike what might be encountered in a community setting. The study also was constrained by a modest sample size and a moderate rate of attrition of its initial participants. Additional conditions which may have influenced the results of the present study include: (a) a noted lack of ethnic diversity among participants, (b) self-report type ._ assessment instruments provided no third-party corroboration, and (c) an over- representation of participants who were not in intimate relationships which may have meant these participants responded to the adult attachment scale based on remote recall versus present, first-hand knowledge. Also, most of the groups admitted members after the initial meeting, which likely had some impacr on participants’ perceptions of group leaders and group climate. Finally, the absence of any significant findings may have been due to the fact that many participants had previous group therapy experience. Prior group experience likely minimized the uncertainty of the group formation process. Future directions The lack of significant results fiom the present study is surprising and quite disappointing, especially in light of the fact that two years were required to collect the needed data. However, this experience has proven that firture research must attend to the issues of (a) control, (b) group member selection, and (c) group member retention. 102 The issue of lack of control was possibly the most important single influence affecting this study examining the relations of adult attachment orientation to expectations of group therapy and perceptions of group leader/s and group climate. This was an ambitious project that, when pursued further, will require a sense of long-term commitment by the researcher. It would be better served by being conducted in a single site that has an established group therapy program with leaders who are all trained within the same theoretical school or in the use of a standard and manualized form of group therapy. This would ensure that all group members are getting the same “product.” The use of a single site with a program evaluation component would also obviate the need for reimbursement, and thus eliminate this potentially confounding variable, as prospective group members would understand that responding to the research instruments is part of the service they are being provided. In addition, the researcher would be well advised to have one primary person (such as a project manager) in charge of enacting the research and attending to the myriad details of member recruitment, screening, group scheduling, data collection, etc. It would also be wise to consider limiting group membership to those who attend the initial group session as the addition of members after that first meeting may impact the group development process and thus, the scores on the dependent measures. In order to establish a common starting ground for group therapy experience it would also be advisable to limit research participation to group therapy novices since entry into group therapy might be more apt to activate the attachment system than for those participants with prior group experience. This could lead to a clearer understanding of the relationship between adult attachment-related anxiety and avoidance with the dependent measures. 103 Since group formation processes depend upon consistency and continuity, regular attendance is of primary importance. In the process of group selection it is advised that screeners inform potential group members to seriously consider their schedules and the amount of time that will be required of them as a group member. This will, of course, impact the expectations of commitment scale as only members who have full awareness of the demands of group therapy will be entering into the group therapy process. In fact, it may be wise to drop commitment as an independent variable altogether considering the present study showed that dropouts scored higher on this dimension of expectations than did continuers. Therefore, one must consider how stated commitment is translated into actual group commitment and attendance. Use of a community population, such as in a community mental health center, might also provide greater continuity in group attendance and completion. Community members are not likely to be as transient as a college student population. This research could also be done with groups that are mandated into group counseling for such issues as alcohol and drug abuse, and domestic violence. If such were the case, the mandate is likely to have an impact on veracity of self report and perhaps, on perception of group climate and leader/s so the impact of this mandate would need to be taken into account. Retention of group members, once they have started the therapy process, has been a long-term problem. For future research with college students it might be advisable to shorten the length of the group session from one-and-one-half hours to 50 minutes as this length of time fits more easily into an academic schedule and is less likely to cause scheduling problems. The researcher will have to consider the impact of this shorter group length on group development processes. Another possible strategy to enhance 104 member retention (and to increase the power of group expectation and pressure) is at the end of each session to have each person commit to attend the next session in front of the entire group. The use of other questionnaires as the independent and dependent measures could also be considered. For example, the Relationship Questionnaire by Bartholomew and Horowitz (1991) would allow categorization of participants into secure, preoccupied, dismissive, and fearful adult attachment styles. This would allow examination of the dependent dimensions with an AN OVA procedure and thus, possibly point out differences between the adult attachment style categories. In addition, Smith, Murphy, and Coats (1999) have developed the Social Group Attachment Scale to assess attachment to social groups along avoidance and anxiety dimensions. The researcher would be well advised to adapt this scale to assess attachment to therapy groups, especially in light of these authors’ assertion that attachment to groups is conceptually distinct fiom attachment in intimate relationships. If this assertion is true, then use of this measure may be more likely to produce significant results with the dependent variables. In addition, it could be productive to use measures of personality such as the NEO-PI-R by Costa and McCrae (1992) or 16 PF by Cattell, Cattell, and Cattell (1993) to measure the independent variables. This could aid in discerning personality characteristics that are related to expectations and perceptions of group processes but are separate and distinct from adult attachment. Of the questionnaires used to assess the dependent variables, only the Expectations About Counseling—Brief Form (EAC—B) performed in a uniformly exemplary fashion within the sample. The lowest Cronbach alpha exhibited by a variable 105 derived from this questionnaire was .73. It was easily adapted for use with group counseling and assessed a broad range of possible expectations although the present study was restricted to commitment, facilitative conditions, and nurturance. It is strongly recommended that the EAC—B be used in subsequent research examining adult attachment and expectations of counseling, perhaps in tandem with an adapted Social Group Attachment Scale, due to its performance in the present study. If one wanted to choose an alternate measure to assess expectations of therapy, the Psychotherapy Expectancy Inventory—Revised (PEI—R) by Berzins (1971) could be used. This 30-item measure assesses members’ expectations of receiving supportive relationships, approval, advice, and an audience for discussing problems. The PEI—R was designed for use with individual therapy, as was the EAC—B in the present study, and would have to be adapted for use with group therapy. Alternate measures that could be used to assess group climate or atmosphere include the Group Attitude Scale (GAS) by Evans, Jarvis and Dawson (1986). The Group Climate Questionnaire—Short Form used in the present study poses some of its questions as though the participant is an outside observer, whereas the GAS asks for more personal information regarding the members’ desire to be an accepted participant in their therapy group. Chen and Mallinckrodt (2002) suggest that “GAS ratings may hence reflect cohesion of the group experienced on an individual level.” The GAS is only slightly longer than the GCQ—S (20 items versus 12) and so, would not add a great deal more time to the assessment procedure. Another measure that could be used is the Single Session Evaluation Questionnaire by Stiles (1980). This measure assesses members’ perceptions of group therapy sessions by following “This session was...” with 20 bipolar 106 adjectives such as “good-bad” or “deep-shallow.” It would have the added benefit of being administered after every group session thereby allowing researchers to track the progress of the individual as well as the group formation process from the members’ perspectives. Although the Trainer Behavior Scale did not perform to expectations within the present sample, it has been used productively elsewhere (see Kivlighan, Marsh— Angelone, & Angelone 1994). It would seem advisable, when this inquiry is pursued further, to develop an instrument specifically designed to assess perceived attachment to the group therapist. This would help determine the degree of correspondence between attachment to the larger group and therapist. Perhaps Mallinckrodt, Gantt, & Cobles’ (1995) Client Attachment to Therapist Scale (CATS) could be adapted for this purpose. This 36-item measure produces secure, avoidant-fearful, and preoccupied-merger subscales which could then be correlated with an adapted Social Group Attachment Scale to determine disparity between attachment to therapist and group or disproportionate dependence on the group or therapist, or vice versa. It is also possible that the Working Alliance Inventory by Horvath and Greenberg (1989) could be adapted for examining the members’ perception of the group leader. This 36—item scale would allow an examination of the relation between adult attachment and Bordin’s (1979) concepts of goals, bonds, and tasks in a group setting. In addition, this form could be used by the therapist to rate the quality of alliance with each of the group members thus pointing out discrepancies between group member and therapist on perceptions of their working relationship. An analogue design which uses a non-clinical sample, such as graduate students in a group counseling class, could also be a productive course for future research. Since 107 participants are expected to attend personal growth groups as the experiential component of such a course, a researcher could expect more consistent group attendance and lower frequency of dropout. This would allow a high degree of control and, very likely, participant participation. However, it would limit generalizability to clinical populations. Alternatively, a qualitative methodology may be more appropriate to assessment of working models, expectations, and perceptions of group processes. In this regard, the Adult Attachment Interview by George, Kaplan and Main (1984) could be used in conjunction with detailed interviews regarding expectations prior to entering group therapy and perceptions after a period of group participation. In addition, the relationship between adult attachment and expectations of group therapy could easily be conducted by the administration of questionnaires to classes of undergraduate students in a university setting. Participants could be instructed to respond to the expectations questionnaire as if they were about to enter group therapy. Again, the use of this non-clinical sample and hypothetical group setting would limit generalizability to actual clients who present for counseling. If future research is conducted in a setting such as a university counseling center where members are referred for group by an individual therapist, it would be interesting to assess parental attachment, romantic attachment, attachment to therapist (both individual and group), and attachment to the therapy group. This could provide both a way to examine the degree of concordance of working models among these many and complex relationships and a means to determine which one/s are particularly powerful in affecting expectations and group development processes. 108 In conclusion, the results of this research are disappointing in revealing no relations between adult attachment orientations in intimate relationships and: (a) expectations of group therapy, (b) perceptions of group leader/s, and (c) perceptions of group climate. However, this lack of results does suggest that Smith, Murphy, and Coats (1999) may be correct in asserting that relational attachment and attachment to groups are conceptually distinct constructs. If this i_s true, then the most important requirement for future research in examining expectations and perceptions of group therapy would be to find an assessment tool that enables us to operationalize adult attachment to therapy groups. Although this was a logistically difficult and time intensive project, it was also infinitely rewarding in seeing it through from inception to completion. However, future researchers would be well advised to (a) consider the feasibility of carrying out this research, (b) carefully coordinate efforts in a centralized facility, and (c) expect to assume a long-term commitment to this line of inquiry. 109 APPENDICES 110 APPENDIX A Recruitment Script lll You are invited to take part in research being conducted by Jim Wyssmann, a doctoral candidate in counseling psychology from Michigan State University. This research investigates how individuals’ close relationship experiences influence their group therapy experience. You will be asked to complete an informed consent, a demographic form, and 2 questionnaires prior to your entry into the group and 4 questionnaires after completion of the fourth group session. Your participation should take approximately 1 hour and you will be reimbursed in the amount of $10.00 after completion of all questionnaires alter the fourth session of group. Your group leaders will not know whether you are participating and all of your responses will be held in strict confidence. After you have left the room please indicate on the other side of this sheet whether you do or do not wish to participate. Then place it in the envelope and submit it to the counseling center receptionist. If you choose to participate please write your first name, phone number, and good times and days to call on the sheet so that the researchers can contact you. Thank you. 112 APPENDIX B Participant Response Form 113 PARTICIPANT RESPONSE FORM Yes, I AM interested in participating in the research First Name: Phone Number: Good times and days to call: No, I AM NOT interested in participating in the research 114 Appendix C Participant Consent Form 115 PARTICIPANT CONSENT FORM This research project investigates how an individual’s experiences in close relationships impacts ideas one might have about entering group therapy. Your participation should take a total of about an hour. Your responses will be maintained in confidence and not associated with your name. Your privacy will be protected to the full extent of the law. T By participating you agree to respond honestly to the research questionnaires. You are, however, free to discontinue your participation at any time. There will be no repercussions to you should you decide not to participate or withdraw prior to completion. Answering of these questionnaires is not expected to engender any risk to you although you may experience some emotional discomfort by reflecting on recollections about relationships. You also may experience a heightened awareness of your thoughts and feelings about the group after completion of questionnaires after the fourth group therapy session that may have an unknown impact on the course of your group therapy experience. By signing this form and returning questionnaires today and after completion of the fourth group therapy session, you indicate that any risks of participation in this study have been explained to your satisfaction and that you consent to take part in this research with the understanding that the methods and purpose of this research may not be fully explained to you until after the collection of data is completed. You will be compensated in the amount of $10.00 after you have completed all of the questionnaires after your fourth group therapy session. If you have questions regarding your rights as a participant in this research please feel free to contact Dr. David Wright Participant Consent Form — Continued. 116 (dewritMmedu) at (517) 355-2180 in the office of the University Committee on Research Involving Human Subjects, 246 Administration Building, Michigan State University, East Lansing, MI, 48823. Signed: Date: If you have questions about this research, you may contact Jim Wyssmann, M.A. at the Casper College Counseling Center: (307) 268-2231 or flssmann@gdmin.cc.whecn.edu. 117 APPENDIX D Demographic Questionnaire 118 DEMOGRAPHIC QUESTIONNAIRE Thank you for deciding to participate in this research project. The following questions ask about your background. Please circle the appropriate number under each of the items below or enter the correct information in the blank spaces that are provided. Age Gender (1) Male (2) Female Ethnicity r.- (1) Caucasian/White (2) Hispanic/Latino (3) Native-American (4)African—American (5)Asian-American (6)Other Academic Classification (1) Freshman (2) Sophomore (3)Junior (4)Senior (5)Other Relationship Status (1) Married (2) Partnered (3) Divorced (4) Widowed (5) Single Dating Status (1) Never dated (2) Not dating (3) Dating 1 person (4) Dating more than 1 person Previous Experience in Group Therapy (1) Yes Please specify number of previous groups Please rate your average satisfaction with past group experiences (A) Very dissatisfied (B) Dissatisfied (C) Neutral (D) Satisfied (E) Very satisfied 119 Demographic Questionnaire — Continued. Previous Experience in Group Therapy — Continued (2) No past experience in group therapy Previous Experience in Individual Therapy (1) Yes Please specify number of previous experiences in individual therapy Please rate your average satisfaction with past individual therapy experiences (A) Very dissatisfied (B) Dissatisfied (C) Neutral (D) Satisfied (E) Very satisfied (2) No past experience in individual therapy How reserved or outgoing do you believe you are? (Choose only one) (1) Very reserved (2) Reserved (3) Neutral (4) Outgoing (5) Very outgoing Briefly describe the problem that brought you in for counseling? Please rate how greatly this problem is affecting your life (Choose only one) (1) Not at all affecting (2) Somewhat affecting (3) Moderately affecting (4) Significantly affecting (5) Very significantly affecting 120 APPENDIX E Experiences in Close Relationships 121 EXPERIENCES IN CLOSE RELATIONSHIPS Instructions: The following statements concern how you feel in romantic relation- ships. We are interested in how you generally experience relationships, not just in what is happening in a current relationship. Respond to each statement by indicating how much you agree or disagree with it. Write the number in the space provided, using the following rating scale: Disagree strongly Neutral/mixed Agree strongly l 2 3 4 5 6 7 _1. I prefer not to show a partner how I feel deep down. 2. I worry about being abandoned. I am very comfortable being close to romantic partners. ____3 _4. I worry a lot about my relationships. _5 Just when my partner starts to get close to me I find myself pulling away. __6. I worry that romantic partners won’t care about me as much as I care about them. _7. I get uncomfortable when a romantic partner wants to be very close. _8. I worry a fair amount about losing my partner. _9. I don’t feel comfortable opening up to romantic partners. 10. I often wish that my partner’s feelings for me were as strong as my feelings for him/her. _1 l. I want to get close to my partner, but I keep pulling back. 12. I ofien want to merge completely with romantic partners, and this sometimes scares them away. 13. I am nervous when partners get too close to me. 14. I worry about being alone. 15. I feel comfortable sharing my private thoughts and feelings with my partner. 16. My desire to be very close sometimes scares people away. 17. I try to avoid getting too close to my partner. Experiences in Close Relationships - Continued. 122 Disagree strongly Neutral/mixed Agree strongly 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 1 2 3 4 5 6 7 I need a lot of reassurance that I am loved by my partner. I find it relatively easy to get close to my partner. Sometimes I feel that I force my partners to show more feeling, more commitment. I find it difficult to allow myself to depend on romantic partners. r- 1 do not often worry about being abandoned. I prefer not to be too close to romantic partners. If I can’t get my partner to show interest in me, I get upset or angry. I tell my partner just about everything. I find that my partner(s) don’t want to get as close as I would like. :2— I usually discuss my problems and concerns with my partner. When I’m not involved in a relationship, I feel somewhat anxious and insecure. I feel comfortable depending on romantic partners. I get frustrated when my partner is not around as much as I would like. I don’t mind asking romantic partners for comfort, advice, or help. I get frustrated if romantic partners are not available when I need them. It helps to turn to my romantic partner in times of need. When romantic partners disapprove of me, I feel really bad about myself. I turn to my partner for many things, including comfort and reassurance. I resent it when my partner spends time away from me. 123 APPENDIX F Expectations About Counseling-Brief Form 124 EXPECTATIONS ABOUT COUNSELING—BRIEF FORM We would like to know just what you think group counseling will be like. On the following pages are statements about counseling. In each instance you are to indicate what you expect counseling to be like. The rating scale we would like you to use is printed at the top of each page. Your ratings of the statements are to be recorded in the space to the left of each statement. Your responses will be kept in strictest confidence. Your answers will be combined with the answers of other group members and reported only in the form of group averages. Your participation is voluntary. When you are ready to begin, answer each question as quickly and as accurately as possible. Finish each page before going to the next. Not Slightly Somewhat Fairly Quite Very Definitely True True True True True True True 1 2 3 4 5 6 7 I EXPECT TO. .. Take psychological tests as part of the group experience. Like the group counselor(s). See a group counselor in training. Gain some experience in new ways of solving problems within the group counseling process. ___Openly express my emotions regarding myself and my problems. __ Understand the purpose of what happens in the group. _ Do assignments outside the group. __ Take responsibility for making my own decisions. Expectations about CounselingBrief Form — Continued. 125 Not Slightly Somewhat Fairly Quite Very Definitely True True True True True True True 1 2 3 4 5 6 7 I EXPECT TO. .. Talk about my present concerns. Get practice relating openly and honestly to other people in the group. Enjoy my sessions with the group members and counselor(s). Practice some of the things I need to learn about relationships in the group. Get a better understanding of myself and others. Stay in counseling for at least a few weeks, even if at first I am not sure it will help. Participate in the group for more than 3 sessions. I EXPECT TO. .. Never need counseling again. Enjoy being with the group members and counselor(s). Stay in counseling even though it may be painful and unpleasant at times. Contribute as much as I can in terms of expressing my feelings and discussing them. Participate in group counseling for only one session. I EXPECT TO. .. Go to group counseling only if I have a very serious problem. Find that group counseling will help the group counselor(s) and me identify problems on which I need to work. Become better able to help myself in the future. Find that my problem will be solved once and for all in counseling. Expectations about CounselingBrief Form — Continued. 126 Not Slightly Somewhat Fairly Quite Very Definitely True True True True True True True 1 2 3 4 5 6 7 I EXPECT TO. .. Feel safe enough with the group members and counselor(s) to really say how I feel. See an experienced group counselor(s). Find that all I need to do is answer the group members’ and counselor(s’) questions. Improve my relationships with others. Ask the group members and counselor(s) to explain what they mean whenever I do not understand something that is said. Work on my concerns outside the group. Find that the group session is not the place to bring up my personal problems. THE FOLLOWING QUESTIONS CONCERN YOUR EXPECTATIONS ABOUT THE GROUP COUNSELOR(S) I EXPECT THE COUNSELOR(S) TO. .. Explain what’s wrong. __ Help me identify and label my feelings so I can better understand them. Tell me what to do. Know how I feel even when I cannot say quite what I mean. Expectations about CounselingBrief Form — Continued. 127 Not Slightly Somewhat Fairly Quite Very Definitely True True True True True True True 1 2 3 4 5 6 7 I EXPECT THE COUNSELOR(S) TO. Know how to help me. Help me identify particular situations where I have problems. Give encouragement and reassurance. 1 EXPECT THE COUNSELOR(S) TO. .. Help me to know how I’m feeling by putting my feelings into words for me. Be a “real” person not just doing a job. Help me discover what particular aspects of my behavior are relevant to my problems. __ Inspire confidence and trust. Frequently offer advice. __ Be honest with me. __ Be someone who can be counted on. __ Be fiiendly and warm towards me. __ Help me solve my problems. __ Discuss his or her own attitudes and relate them to my problem. Give me support. Decide what treatment plan is best. Know how I’m feeling at times, without me having to speak. Do most of the talking. Expectations about CounselingBrief Form — Continued. 128 Not Slightly Somewhat Fairly Quite Very Definitely True True True True True True True 1 2 3 4 5 6 7 I EXPECT THE COUNSELOR(S) TO. .. Respect me as a person. Discuss his or her experiences and relate them to my problems. Praise me when I show improvement. I EXPECT THE COUNSELOR(S) TO. .. Make me face up to the differences between what I say and how I behave. __ Talk freely about himself or herself. _ Have no trouble getting along with people. Like me. __ Be someone I can trust. __ Like me in spite of the bad things that he or she knows about me. Make me face up to the differences between how I see myself and how I am seen by others. Be someone who is cahn and easygoing. Point out the differences between what I am and what I want to be. Just give me information. Get along well in the world. 129 APPENDIX G Trainer Behavior Scale 130 TRAINER BEHAVIOR SCALE Please indicate next to each question how you feel toward the leader of your therapy group. If you have more than one leader please use an average score to represent both. Please use the following rating scale: Strongly Disagree Neutral Strongly Agree 1 2 3 4 5 _ 1. The leader is quite comfortable and relaxed when the attention of the group focuses on him/her. _ 2. The leader's contributions interpret to the group the underlying meaning of what had been occuning. _ 3. The leader tries to get the group not to engage in activities which he/she feels will be unproductive. _ 4. The leader misinterprets what people say. __ 5. The leader adopts a professional role that makes it difficult to know him/her as a person. _ 6. The leader seems to be uninterested in some of the members. _ 7. The leader exerts considerable influence over the direction the group takes. _ 8. The leader sees things through the eyes of members of the group. __ 9. The leader when calling attention to something, gives a theoretical explanation of why it occurred. __10. The leader is punishing toward behavior which he/she feels is inhibiting the progress of the group. _11. The leader seems to have difficulty understanding what members of the group are trying to communicate. Trainer Behavior Scale - Continued. 131 Strongly Disagree Neutral Strongly Agree 1 2 3 4 5 12. The leader gives the impression that he/she likes some kind of behavior better than others. 13. The leader seems to really value the persons in the group. 14. The leader tries to minimize the extent of her/his influence on the group. 15. The leader seems to be in close touch with how members of the group are feeling. 16. The leader's contributions include a considerable amount of conceptual and theoretical content. 17. The leader contributions are characterized by open, spontaneous expressions of his/her own feelings and reactions. 18. The leader seems to be puzzled and uncertain about what is happening in the group. 19. The leader seems to hold back from expressing his/her own reactions to what is happening in the group. 20. The leader shows considerable affection for most members of the group. 21. The leader avoids influencing the group’s decision about what to do next. 22. The leader is secure and comfortable in the group. 23. The leader gives short lectures on concepts relevant to current problems in the group. 24. The things the leader says seem to be highly consistent with what he/she is feeling. 132 APPENDD( H Group Climate Questionnaire—Short Form 133 GROUP CLIMATE QUESTIONNAIRE—SHORT FORM Please indicate next to each question below your feeling about the events that transpired in the group therapy session you just completed. Please us the following rating scale: Not at all Neutral/mixed Extremely l 2 3 4 5 6 7 _ 1. The members liked and cared about each other. __ 2. The members tried to understand why they do the things they do, tried to reason it out. _ 3. The members avoided looking at important issues going on between themselves. _ 4. The members felt what was happening was important and there was a sense of participation. _ 5. The members depended on the group leader(s) for direction. _ 6. There was friction and anger between the members. _ 7. The members were distant and withdrawn from each other. _ 8. The members challenged and confronted each other in their efforts to sort things out. _ 9. The members appeared to do things the way they thought would be acceptable to the group. _10. The members distrusted and rejected each other. 11. The members revealed sensitive personal information or feelings. _12. The members appeared tense and anxious. 134 APPENDIX I Group Satisfaction Scale 135 GROUP SATISFACTION SCALE Please indicate next to each question how you feel about your experience as a member of the therapy group you are in. Please use the following rating scale: Strongly Disagree Neutral Strongly Agree 1 2 3 4 5 __ l. I feel I am benefiting from my participation in this group. _ 2. My group seems to be "stuck". _ 3. I have been learning a lot about myself in this group. _ 4. Outside stresses are making it difficult for me to continue my participation in this group. _ 5. I am satisfied with my group's progress thus far. _ 6. I feel I could be benefiting more from individual counseling rather than group counseling. _ 7. Overall, my experiences thus far within this group have been helpful. __ 8. I am unsure about continuing my participation in this group. _ 9. I feel I am important to the functioning of the group. _10. Others do not seem as involved in the group as I am. _11. I look forward to attending group therapy sessions. _12. I feel “at home” in the group. _13. I oftentimes feel discouraged after attending the group. _14. If I had it to do over I would be in this group again. 136 REFERENCES 137 ‘1' REFERENCES Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, NJ: Erlbaum. Asendorpf, J. B., & Wilpers, S. (1998). Personality effects on social relationships. Journal of Personality and Social Psychology, 74, 1531-1544. Baldwin, M. W., Keelan, J. P. K., Fehr, B., Enns, V., & Koh Rangarajoo, E. (1996). Social-cognitive conceptualization of attachment working models: Availability and accessibility effects. Journal of Personality and Social Psychology, 71, 94- 109. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A L: test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244. Bartholomew, K. (1990). Avoidance of Intimacy: An attachment perspective. Journal of Social and Personal Relationships, 7, 147-178. Berzins, J. I. (1971). Revision of the Psychotherapy Expectancy Inventory. Unpublished manuscript, University of Kentucky at Lexington. Bohnan, L. (1971). Some effects of trainers on their T-groups. The Journal of Applied Behavioral Science, 7, 309-326. Bordin, E. S. (1979). The generalizability of psychoanalytic concept of the working alliance. Psychotherapy, 16, 252-260. Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge. Brabender, V. M. (2001). The future of group psychotherapy: Expanding the conversation. International Journal of Group Psychotherapy, 51, 181-189. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp.46-76). New York, NY: The Guilford Press. Brennan, K. A., & Shaver, P. R. (1995). Dimensions of adult attachment, affect regulation, and romantic relationship functioning. Personality and Social Psychology Bulletin, 21, 267-283. 138 Brissette, I., Scheier, M. F., & Carver, C. S. (2002). The role of optimism in social network development, coping, and psychological adjustment during a life transition. Journal of Personality and Social Psychology, 82, 102-1 1 l. Brykczynska, C. (1990). Changes in the patient’s perception of his therapist in the process of group and individual psychotherapy. Psychotherapy and Psychosomatics, 53, 179-184. Cattell, R.B., Cattell, A.K., & Cattell, HE. (1993). Sixteen personality factor questionnaire (5th ed.) Charnpaigrr, E: Institutes for Personality and Ability Testing, Inc. Chen, E. C., & Mallinckrodt, B. (2002). Attachment, group attraction, and self-other agreement in interpersonal circumplex problems and perceptions of group members. Group Dynamics, 6, 311-324. Collins, N. L., & Read, S. J. (1994). Cognitive representations of attachment: The structure and function of working models. Advances in Personal Relationships, 5, 53-90. Costa, P. T., & McCrae, R. R. (1992). Revised NEO personality inventory (NEO-PI—R) and NEO five factor inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources. de Carufel, F. L., & Piper, W. E. (1988). Group psychotherapy or individual psychotherapy: Patient characteristics as predictive factors. International Journal of Group Psychotherapy, 38, 169-188. Delucia—Waack, J. L. (1997). What do we need to know about group work: A call for firture research and theory. The Journal for Specialists in Group Work, 22, 146- 148. de Zulueta, F., & Mark, P. (2000). Attachment and contained splitting: A combined approach of group and individual therapy to the treatment of patients suffering from borderline personality disorder. Group Analysis, 33, 486-500. Diehl, M., Elnick, A. B., Bourbeau, L. S., & Labouvie—Vief, G. (1998). Adult attachment styles: Their relations to family context and personality. Journal of Personality and Social Psychology, 74, 1656-1669. Dies, R. R. (1983). Clinical implications of research on leadership in short-term group psychotherapy. In R.R. Dies & K.R. MacKenzie (Eds.), Advances in group psychotherapy: Integrating research and practice (pp. 27-75). Madison, CT: International University Press. Duggan, E. S., & Brennan, K. A. (1994). Social avoidance and its relation to Bartholomew’s adult attachment typology. Journal of Social and Personal Relationships, 11, 147-153. 139 Engels, R. C. M. E., Finkenauer, C., Meeus, W., & Dekovic', M. (2001). Parental attachment and adolescents’ emotional adjustment: The associations with social skills and relational competence. Journal of Counseling Psychology, 48, 428-439. Evans, N. J ., Jarvis, P. A., & Dawson, C. (1986). The group attitude scale: A measure of attraction to group. Small Group Behavior, 1 7, 203-216. Feeney, J. A., & Ryan, S. M. (1994). Attachment style and affect regulation: Relationships with health behavior and family experiences of illness in a student sample, Health Psychology, 13, 334-345. Flores, P. J. (2001). Addiction as an attachment disorder: Implications for group therapy. International Journal of Group Psychotherapy, 51, 63-81. Flowers, J. V. (1987 ). Client outcome as a function of agreement or disagreement with the modal group perception of curative factors in short-term, structured group psychotherapy. International Journal of Group Psychotherapy, 3 7, 1 13-1 17. Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78, 350-365. Frank, J. D. (1959). The dynamics of the psychotherapeutic relationship. Psychiatry, 22, 17-39. Fuendeling, J. (1998). Affect regulation as a stylistic process within adult attachment. Journal of Social and Personal Relationships, 15, 291 -323. Fuhriman, A., & Burlingame, G. M. (1994). Group psychotherapy: Research and practice. In A. Fuhriman & G. M. Burlingame (Eds.), Handbook of group psychotherapy: An empirical and clinical synthesis (pp. 3-40). New York: Wiley. George, C., Kaplan, N., & Main, M. (1984). The Adult Attachment Interview protocol. Unpublished manuscript, University of California at Berkeley, Department of Psychology. Gladstein, G. A.(1969). Client expectations, counseling experience, and satisfaction. Journal of Counseling Psychology, 16, 476-481. Griffin, D., & Bartholomew, K. (1994). The metaphysics of measurement: The case of adult attachment. In K. Bartholomew & D. Perlman (Eds.), Advances in Personal Relationships: Vol. 5 Attachment processes in adulthood. (pp. 17-52). London: Jessica Kingsley Groth-Mamat, G. (1990). Handbook of psychological assessment (2nd ed.). New York: Wiley. 140 Hayes, T. J ., & Tinsley, H. E. A. (1989). Identification of latent dimensions of instruments that measures perceptions of and expectations about counseling. Journal of Counseling Psychology, 36, 492-500. Helfinann, B. (1994). Here is now. International Journal of Group Psychotherapy, 44, 429-436. Home, A. M., & Rosenthal, R. (1997). Research in group work: How did we get where we are? The Journal for Specialists in Group Work, 22, 228-240. Horowitz, L. M., Rosenberg, S. E., & Bartholomew, K. (1993). Interpersonal problems, attachment styles, and outcome in brief dynamic psychotherapy. Journal of Consulting and Clinical Psychology, 61, 549-560. Horvath, A. O., & Greenberg, L. S. (1989). Development and validation of the working alliance inventory. Journal of Counseling Psychology, 36, 223-233. Kiesler, D. J. (1983). The 1982 interpersonal circle: A taxonomy for complementarity in human relations. Psychological Review, 90, 185-214. Kilmann, P. R., Laughlin, J. E., Carranza, L. V., Downer, J. T., Major, S., & Parnell, M. M. (1999). Effects of an attachment-focused group preventive intervention on insecurely attached women. Group Dynamics: Theory Research and Practice, 3, 138-147. Kivlighan, D. M., Jr., & Angelone, E. O. (1992). Interpersonal problems: Variables influencing participants’ perception of group climate. Journal of Counseling Psychology, 39, 468-472. Kivlighan, D. M., Jr., & Goldfine, D. C. (1991). Endorsement of therapeutic factors as a function of stage of group development and participant interpersonal attitudes. Journal of Counseling Psychology, 39, 468-472. Kivlighan, D. M., Jr., Marsh—Angelone, M., & Angelone, E. O. (1994). Projection in group counseling: The relationship between members’ interpersonal problems and their perception of the group leader. Journal of Counseling Psychology, 44, 99- 104. Kivlighan, D. M., Jr., Multon, K. D., & Brossart, D. F. (1996). Helpful impacts in group counseling: Development of a multidimensional rating scale. Journal of Counseling Psychology, 43, 347-355. Kivlighan, D. M., Jr., Patton, M. J ., & Foote, D. (1998). Moderating effects of client attachment on the counselor experience-working alliance relationship. Journal at Counseling Psychology, 45, 274-278. Kivlighan, D. M., Jr., & Shaughnessy, P. (1993). Dimensions of group leader behavior. Unpublished manuscript, University of Missouri at Columbia. 141 Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working models, affect regulation, and representations of self and others. Child Development, 59, 135-146. Lee, R. M., & Robbins, S. B. (1998). The relationship between social connectedness and anxiety, self-esteem, and social identity. Journal of Counseling Psychology, 45, 338-345. Lee, R. M., Draper, M., & Lee, S. (2001). Social connectedness, dysfunctional interpersonal behaviors and psychological distress: Testing a mediator model. Journal of Counseling Psychology, 48, 310-318. Lopez, F. G., Gover, M. R., Leskela, J ., Sauer, E. M., Schirrner, L., & Wyssmann, J. (1994). Attachment styles, shame, guilt, and collaborative problem-solving orientations. Personal Relationships, 4, 187-199. MacKenzie, K. R. (1983). The clinical application of a group climate measure. In: R. R. Dies & K. R. MacKenzie (Eds.), Advances in group psychotherapy: Integrating research and practice (pp.159-170). New York: International Universities Press. 1'! MacKenzie, K. R. (1994). Where is here and when is now? The adaptational challenge of mental health reform for group psychotherapy. International Journal of Group Psychotherapy, 44, 407-428. MacKenzie, K. R. (1995). Eflective use of group therapy in managed care. Washington, DC: American Psychiatric Press. MacKenzie, K. R. (2001). An expectation of radical changes in the future of group psychotherapy. International Journal of Group Psychotherapy, 51, 175-180. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. Monographs of the Society for Research in Child Development, 50, 66-104. Mallinckrodt, B. (1991). Clients’ representations of childhood emotional bonds with parents, social support, and formation of the working alliance. Journal of Counseling Psychology, 38, 401-409. Mallinckrodt, B., Coble, H. L., & Gantt, D. L. (1995). Working alliance, attachment memories, and social competencies of women in brief therapy. Journal of Counseling Psychology, 42, 79-84. Mallinckrodt, B., Gantt, D. L., & Coble, H. L. (1995). Attachment patterns in the psychotherapy relationship: Development of the client attachment to therapist scale. Journal of Counseling Psychology, 42, 307-317. 142 Mallinckrodt, B., McCreary, B. A., & Robertson, A. K. (1995). Co-occurrence of eating disorders and incest: The role of attachment, family environment, and social competencies. Journal of Counseling Psychology, 42, 178-186. Mallinckrodt, B. (2001). Interpersonal processes, attachment, and development of social competencies in individual and group psychotherapy. In B. R. Sarason & S. Duck (Eds.), Personal Relationship: Implications for Clinical and Community Psychology (pp. 89-117). New York: Wiley. Marcus, D. K., & Holahan, W. (1994). Interpersonal perception in group therapy: A social relations analysis. Journal of Consulting and Clinical Psychology, 62, 776- 782. McRoberts, C., Burlingame, G. M., & Hoag, M. J. (1998). Comparative efficacy of individual and group psychotherapy: A meta-analytic perspective. Group Dynamics: Theory, Research, and Practice, 2, 101-117. Mikulincer, M. (1998). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74, 513-524. Mikulincer, M. (1998). Attachment working models and the sense of trust: An exploration of interaction goals and affect regulation. Journal of Personality and Social Psychology, 74, 1209-1224. Mikulincer, M., & Florian, V. (1995). Appraisal of and coping with a real life stressful situation: The contribution of attachment styles. Personality and Social Psychology Bulletin, 21, 406-414. Mikulincer, M., Florian, V., at Weller, A. (1993). Attachment styles, coping strategies, and post-traumatic psychological distress: The impact of the Gulf War in Israel. Journal of Personality and Social Psychology, 64, 817-826. Mikulincer, M., Gillath, O., Halevy, V., Avihou, N., Avidan, S., & Eshkoli, N. (2001). Attachment theory and reactions to others’ needs: Evidence that activation of the sense of attachment security promotes empathic responses. Journal of Personality and Social Psychology, 81, 1205-1224. Mikulincer, M., & Nachson, O. (1991). Attachment styles and patterns of self-disclosure. Journal of Personality and Social Psychology, 61, 321-331. Neyer, F. J ., & Asendorpf, J. B. (2001). Personality-relationship transaction in young adulthood. Journal of Personality and Social Psychology, 81, 1190-1204. Niedenthal, P. M., Brauer, M., Robin, L., & Innes—Ker, A. H. (2002). Adult attachment and the perception of facial expression of emotion. Journal of Personality and Social Psychology, 82, 419-433. 143 Pierce, T., & Lydon, J. (1998). Priming relational schemas: Effects of contextually activated and chronically accessible interpersonal expectations on responses to a stressful event. Journal of Personality and Social Psychology, 75, 1441-1448. Pierce, T., & Lydon, J. (2001). Global and specific relational models in the experience of social interactions. Journal of Personality and Social Psychology, 80, 613-630. Piper, W. E., & Joyce, A. S. (1996). A consideration of factors influencing the utilization of time-limited, short-term group therapy. International Journal of Group Psychotherapy, 46, 311-328. Piper, W. E., McCallum, M., Joyce, A. S., Rosie, J. S., & Ogrodniczuk, J. S. (2001). Patient personality and time-limited group psychotherapy for complicated grief. International Journal of Group Psychotherapy, 51, 525-552. Pistole, M. C. (1989). Attachment in adult romantic relationships: Style of conflict resolution and relationship satisfaction. Journal of Social and Personal Relationships, 6, 505-510. Pistole, M. C. (1997). Attachment theory: Contributions to group work. The Journal for Specialists in Group Work, 22, 7-21. Rice, K. G., Cunningham, T. J ., & Young, M. B. (1997). Attachment to parents, social competence, and emotional well being: A comparison of Black and White late adolescents. Journal of Counseling Psychology, 44, 89-101. Riva, M. T., & Smith, R. D. (1997). Looking into the future of group research: Where do we go from here? The Journal for Specialists in Group Work, 22, 265-276. Satterfield, W. A., Buelow, S. A., Lyddon, W. J ., & Johnson, J. T. (1995). Client stages of change and expectations about counseling. Journal of Counseling Psychology, 42, 476-478. Satterfield, W. A., & Lyddon, W. J. ( 1995). Client attachment and perceptions of the working alliance with counselor trainees. Journal of Counseling Psychology, 42, 187-189. Sauer, E. M. (1999). Therapist and client adult attachment: A longitudinal examination of the therapeutic working alliance (Doctoral dissertation, Michigan State University, 1999). Dissertation Abstracts International, 60, 3612. Schaub, M., & Tokar, D. M. (1999). Patterns of expectations about counseling: Relations to the five-factor model of personality. Journal of Counseling and Development, 7 7, 177-1 88. Smith, M., Glass, G., & Miller, T. (1980). The benefits of psychotherapy. Baltimore, MD: Johns Hopkins University Press. 144 Smith, E. R., Murphy, J ., & Coats, S. (1999). Attachment to groups: Theory and measurement. Journal of Personality and Social Psychology, 77, 94-110. Spitz, H. I. (2001). Group psychotherapy of substance abuse in the era of managed mental health care. International Journal of Group Psychotherapy, 51, 21-41. Stiles, W. B. (1980). Measurement of the impact of psychotherapy sessions. Journal of Consulting and Clinical Psychologi, 48, 176-185. Tasca, G. A., Russell, V., & Busby, K. (1994). Characteristics of patients who choose betweentwo types of group psychotherapy. International Journal of Group Psychotherapy, 44, 499-508. Taylor, N. T., Burlingame, G. M., Kristensen, K. B., Fuhriman, A., Johansen, J ., & Dahl, D. (2001). A survey of mental health care provider’s and managed care organization attitudes toward, familiarity with, and use of group interventions. International Journal of Group Psychotherapy, 51, 243-263. Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York: Wiley. Tillitski, C. J. (1990). A meta-analysis of estimated effect sizes for group versus individual versus control treatment. International Journal of Group Psychotherapy, 40, 215-224. Tinsley, D. J ., Holt, M. S., Hinson, J. A., & Tinsley, H. E. A. (1991). A construct validation study of the Expectations About Counseling—Brief Form: Factorial validity. Measurement and Evaluation in Counseling and Development, 24, 101- 110. Tinsley, H. E. A. (1982). Expectations about counseling. Unpublished test manual, Southern Illinois University at Carbondale, Department of Psychology. Tinsley, H. E. A., Brown, M. T., de St. Aubin, T. M., & Lucek, J. (1984). Relation between expectancies for a helping relationship and tendency to seek help from a campus help provider. Journal of Counseling Psychology, 31, 149-160. Tinsley, H. E. A., & Westcot, A. M. (1990). Analysis of the cogrritions stimulated by the items on the Expectations About Counseling—Brief Form: An analysis of construct validity. Journal of Counseling Psychology, 37, 223-226. Toseland, R. W., & Siporin, M. (1986). When to recommend group treatment: A review of the clinical and the research literature. International Journal of Group Psychotherapy, 36, 171-201. Tschuschke, V., & Dies, R. R. (1994). Intensive analysis of therapeutic factors and outcome in long-term inpatient groups. International Journal of Group Psychotherapy, 44, 185-208. 145 Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological Bulletin, 63, 384-399. Wilkins, W. (1973). Expectancy of therapeutic gain: An empirical and conceptual critique. Journal of Consulting and Clinical Psychology, 40, 69-77. Yalom, I. D. (1985). The theory and practice of group therapy (3rd ed.) New York: Basic Books. Yalom, I. D., & Vinogradov, S. (1988). Bereavement groups: Techniques and themes. International Journal of Group Psychotherapy, 38, 419-446 Yalom, ID. (1995). The theory and practice of group therapy (4’h ed.). New York: Basic Books. 146 11111'1111111111‘11111111‘111 31