"Education, Faculty of"@en . "Educational and Counselling Psychology, and Special Education (ECPS), Department of"@en . "DSpace"@en . "UBCV"@en . "Ferrada, Natacha"@en . "2009-07-02T20:33:51Z"@en . "1999"@en . "Doctor of Philosophy - PhD"@en . "University of British Columbia"@en . "This investigation sought to discover, identify and describe in-session\r\nchange episodes in marital Experiential Systemic Therapy (Friesen et al., 1989)\r\nthrough a naturalistic discovery-oriented methodology. This study responds to calls\r\nmade regarding the need to study complex and intricate processes, such as\r\npsychotherapy (Greenberg, 1986), via discovery-oriented methods (Mahrer, 1988).\r\nThe observational analysis of 40 videotaped-sessions, of three couples,\r\nrevealed the presence of ten episodes of change. The actions and interactions of\r\nthe members of the therapeutic system in these episodes were analyzed via the\r\ngrounded theory method of analysis. This analysis generated a conceptual\r\nframework describing the internal structure of change.\r\nThe resulting conceptual framework consisted of a core category named\r\nsynergetic shifting. This category refers to an interactional process in which the\r\npartners with the assistance of the therapist moved away from rigid, distancing and\r\nalienating interactional patterns toward interpersonal flexibility, compassion and\r\naffiliation by working through blocks hindering the couples' engagement and\r\nintimate connection. Synergetic shifting consists of four client and three therapist\r\ncategories. Within the progressive nature of synergetic shifting, the client categories\r\nwere: (1) owning one's part in the relational conflict; (2) couple contacting: restricted\r\nand limited; (3) couple working through blocks to intimacy; and (4) couple engaging\r\ncompassionately. The therapist's actions and interactions were categorized under\r\nsupporting, transitional and shifting operations."@en . "https://circle.library.ubc.ca/rest/handle/2429/9966?expand=metadata"@en . "11506797 bytes"@en . "application/pdf"@en . "EPISODES OF CHANGE IN EXPERIENTIAL SYSTEMIC MARITAL THERAPY: A DISCOVERY-ORIENTED INVESTIGATION by NATACHA FERRADA B.A., Simon Fraser University, 1985 M.A., The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR i OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Counselling Psychology) We accept this dissertation as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April; T999 \u00C2\u00A9 NATACHA FERRADA, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, 1 agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of t / f , y c ^ j / s q ^ i The University of British Columbia Vancouver, Canada Date 44*01.0$ DE-6 (2/88) ABSTRACT This investigation sought to discover, identify and describe in-session change episodes in marital Experiential Systemic Therapy (Friesen et al., 1989) through a naturalistic discovery-oriented methodology. This study responds to calls made regarding the need to study complex and intricate processes, such as psychotherapy (Greenberg, 1986), via discovery-oriented methods (Mahrer, 1988). The observational analysis of 40 videotaped-sessions, of three couples, revealed the presence of ten episodes of change. The actions and interactions of the members of the therapeutic system in these episodes were analyzed via the grounded theory method of analysis. This analysis generated a conceptual framework describing the internal structure of change. The resulting conceptual framework consisted of a core category named synergetic shifting. This category refers to an interactional process in which the partners with the assistance of the therapist moved away from rigid, distancing and alienating interactional patterns toward interpersonal flexibility, compassion and affiliation by working through blocks hindering the couples' engagement and intimate connection. Synergetic shifting consists of four client and three therapist categories. Within the progressive nature of synergetic shifting, the client categories were: (1) owning one's part in the relational conflict; (2) couple contacting: restricted and limited; (3) couple working through blocks to intimacy; and (4) couple engaging compassionately. The therapist's actions and interactions were categorized under supporting, transitional and shifting operations. iii T A B L E OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES xi LIST OF FIGURES xii ACKNOWLEDGMENTS xiii C H A P T E R I. P U R P O S E OF T H E STUDY 1 Aim of the Investigation 4 Research Questions 5 Overview of Design and Methodology 5 Explanation of Terms 10 Experiential Systemic Marital Therapy 10 Chemical Dependence 10 Recurrent Change Episode.: 11 Conflictual Relational Marker 13 Client and Therapist Actions and Interactions 13 New Patterns of Interaction In-Session 13 Internal Structure of Change Episodes 14 Significance of the Study 14 C H A P T E R II. REVIEW OF R E L A T E D L ITERATURE 1 7 Research Strategies in the Study of Psychotherapy 17 The Meaning of Outcome 18 The Meaning of Process 19 Historical Overview of Process Research 20 Conceptual Shift in Psychotherapy Process Research 26 Focus on Events 28 Context Specific Microtheory 31 Methodological Shift in Psychotherapy Process Research 32 Discovery Oriented Psychotherapy Research 34 Description in Psychotherapy Research 34 Levels of Description 35 Observational Perspective 37 Explanation in Psychotherapy Research 38 Research Approaches 38 Empirical Pattern Identification 39 Theory Testing 40 Performance Analysis 40 Discovery Oriented Methods 43 The Process of Change In Marital Therapy 48 Quantitative Research on the Process of Change in Marital Therapy 48 Qualitative Research on the Process of Change in Marital Therapy 51 Research on the Change Process of Experiential Systemic Marital Therapy 55 Summary of the Review 59 CHAPTER III. METHODOLOGY 61 Design 62 Investigative Procedure 64 Participants 64 Subject Inclusion in TARP 64 Therapists 66 Participant-Observer 66 Researcher's assumptions on the Process of Change 67 Experiential Systemic Therapy 68 Systemic Dimension 69 Symbolic Dimension 70 Experiential Dimension 70 Guiding Principles of ExST 71 Phases of Therapy 72 Selection of Cases 73 Description of Instruments 74 The Alcohol Dependency Data Questionnaire 74 The Symptom Checklist-90-Revised 75 The Dyadic Adjustment Scale 76 The Weekly Situation Diary 77 Decision Criteria for Case Selection 77 Demographic Description of Couples 80 Couple 2010: High Improver 80 Couple 2036: Mild Improver 81 Couple 2080: Minimal Improver 81 Identification and Selection of Change Episodes 82 Criteria for Identification of Change Episodes 83 Selection of Change Episodes 84 Overview of Grounded Theory Method 85 Historical and Theoretical Background 87 Description of the Method of Analysis 90 The Analytical Process 91 Conceptual Integration 95 Criteria for Evaluating a Grounded Theory Investigation 95 CHAPTER IV. RESULTS: OVERALL THERAPY CONTEXT 98 Brief Overall Context of the Therapy Process 98 Outline of the Psychotherapeutic Process for the Three Couples 99 Couple 2010: High Improver 99 Couple 2036: Mild Improver 102 Couple 2080: Minimal Improver 104 Introduction to the Change Episodes 108 Description of the Change Episodes 111 Episode* 1 112 Episode # 2 112 Episode #3 114 Episode # 4 115 Episode # 5 116 Episode # 6 117 Episode # 7 118 Episode # 8 119 Episode #9 120 Episode # 10 121 CHAPTER V. RESULTS: CONCEPTUAL MODEL OF CHANGE 123 Conceptual Model of Change 123 Synergetic Shifting 125 Occurrence of Categories within the Change Episodes 127 Phase One: Owning One's Part in the Relational Conflict 130 Contextual Conditions Related to the Emergence of Phase One 131 Explaining Subcategories: Phase One 135 Distancing 135 Reflecting 136 Expressing 137 Acknowledging 139 Therapist's Operations: Phase One 140 Supporting Categories 141 Accepting Client's Experiences and Actions 141 Empathizing and Validating Client's Experiences 142 Shifting Operations 143 Reifying Experiences 144 Linking the Couple 145 Encouraging Self-focus and Self-connection 145 Utilizing Client's Metaphors and Language 146 Challenging Actions and Interactions 147 Therapist Perseverance 148 Phase Two: Couple Contacting: Restricted, Limited 150 Contextual Conditions Related to the Emergence of Phase Two 151 Explaining Subcategories: Phase Two 154 Distancing: Flipping/Fear/Anger 154 Conflictual Messaging: Accusing/Distancing 155 Brief, Restricted Contact 157 Therapist's Operations: Phase Two 158 Transitional Operations 158 Bridging the Couple 159 Supporting Operations 160 Shifting Operations 160 Encouraging Direct Interaction 161 Modeling Interaction 162 Linking the Couple 162 Challenging Actions and Interactions 163 Phase Three: Couple Working Through Blocks to Intimacy 164 Contextual Conditions Related to the Emergence of Phase Three 165 Explaining Subcategories: Phase Three 170 Disclosing, sharing tales of abuse, betrayal 170 Unlocking Blocks/Ghosts/Demons 172 Transforming Stories/Metaphors 173 Letting go of Blocks 174 Therapist's Operations: Phase Three 176 Supporting Operations 177 Transitional Operations 178 Shifting Operations 179 Linking the Couple 179 Encouraging Interaction 180 Reifying Client's Experiences 181 Challenging Clients 182 Utilizing Client's Language and Metaphors 182 Phase Four: Couple Engaging Compassionately 183 Contextual Conditions Related to the Emergence of Phase Four 184 Explaining Subcategories: Phase Four 189 Interacting Freely 189 Accepting, Supporting 190 Physical Expressions of Intimacy 191 Therapist's Operations: Phase Four 192 Supporting Operations 193 Transitional Operations 193 Bridging the Couple 194 Encouraging Spontaneous Expressions of Support 195 Summary 196 CHAPTER VI. DISCUSSION 197 Findings Related to the Research Questions 197 Overview of Findings Related to the Research Questions 198 Episode Completion 198 Proximal Versus Large \"O\" Outcome 200 Implications of the Study 202 Theoretical Implications 202 Clinical Implications 207 Transitional Operations 211 Links between Clients' and Therapist's Actions and Interactions 212 Frequency of Therapist's Operations 213 Integration of Findings to the Literature on Marital Change 214 Shifting 214 Emotional Domain 216 Cognitive Domain 217 Behavioral Domain 218 Experiential Change 219 Other Related Clinical Findings 219 Intrapersonal to Interpersonal Shift during Marital Change.... 220 Perseverance 220 Limitations of the Study 221 Generalizability 221 Objectivity 224 Internal Validity 225 Reliability 226 Evaluation of Grounded Theory Investigation 227 Other Methodological Considerations 228 Suggestions for Future Research 229 Conclusion 231 REFERENCES 233 Appendix A: Participant Family's Consent Form 245 Appendix B: Therapist Competency Form 247 Appendix C: Weekly Situation Diary 250 Appendix D: Illustrations: Categories and Subcategories 262 xi LIST OF TABLES Table 1: Obtained Scores on DAS, SCL-90-R and ADDS for Couple 2010 79 Table 2: Obtained Scores on DAS, SCL-90-R and ADDS for Couple 2036 79 Table 3: Obtained Scores on DAS, SCL-90-R and ADDS for Couple 2080 80 Table 4: Emergence of Change Episodes in Sessions 110 Table 5: Occurrence of Categories within the Change Episodes 128 xii LIST OF FIGURES Figure 1: An Overview of the Therapeutic System 73 Figure 2: Conceptual Model of Synergetic Shifting 124 ACKNOWLEDGMENTS I would like to express my deep appreciation for the couples who had the courage to risk and to engage in the journey of therapeutic discovery with me. I feel honoured by their willingness to let down their protective guard and to share their stories. They have been my guides throughout this learning process. I would also like to thank Dr. John Friesen, the principal investigator of the Alcohol Recovery Project (TARP), for allowing me to use the TARP data for this investigation. Without his collaboration and assistance, this study would not have been possible. I wish to thank my committee for their excellent advice, careful reading and consistent patience. I thank my supervisor Dr. John Friesen for his engaging and thoughtful ideas and comments and his commitment to scholarship. Also, I thank Dr. Beth Haverkamp for her inspirational and thoughtful insights and her consistent encouragement. As well, I thank Dr. Donald Fisher for his generous wisdom and readiness to share his methodological expertise. Finally, I thank Dr. Marv Westwood for his valuable insights and constant support. I wish to thank my father, Eduardo Ferrada for his ever-lasting trust in me and my mother Elsa Arrauz for her support and encouragement. Finally, I wish to acknowledge my son, Eddie for his love and support, and my daughter Nadia for her love and understanding. I truly appreciate your patience with me throughout the duration of this project. 1 CHAPTER I PURPOSE OF THE STUDY As an applied science of human change (Kiesler, 1971), psychotherapy may be defined, in its broadest sense, as an interpersonal process aimed at bringing about change within the affective, cognitive, behavioral, attitudinal, and relational domains of experience. Marital therapy, as a subspeciality of family therapy (Alexander, Holtzworth-Munroe, & Jameson, 1994; Gurman, Kniskern & Pinsof, 1986), also seeks to facilitate change in the various domains of experience by altering the interactions between members of the dyad so as to improve functioning to a desired state. Given that the essence of the psychotherapist's task is to facilitate desired change in clients' lives, an important psychotherapy researcher's task, then, is to investigate both the process and result of this collaborative endeavour (Lambert & Hill, 1994; Strupp, 1978). Accordingly, empirical strategies with a focus on the investigation of outcome, that is, the results, and the examination of process related to outcome have been devised. Regarding the investigation of the results, considerable attention has been devoted to the study of the effectiveness of psychotherapy following Eysenck's (1952) challenge on its ineffectual effects. Currently, however, researchers manifest relative satisfaction with the question of its effectiveness at both the individual (Lambert & Bergin, 1994) and couples/ family (Alexander et al. 1994; Gurman et al. 1986; Jacobson & Addis, 1993) levels. 2 Concomitant with these efforts, researchers have also examined facilitative aspects of the process of psychotherapy change measured at termination. Here, the focus has been on linking the domains of process and outcome by searching for key ingredients which may facilitate the process of change within various settings, client populations, therapeutic modalities and therapist characteristics. As indicated by Strupp (1978), the field has moved beyond inquiries regarding its effectiveness to questions such as: what is it that facilitates change, how is change facilitated, and how can we best link process to outcome measures. The questions of what is facilitative in psychotherapy and what are the mechanisms under which change in psychotherapy occurs continue to intrigue both researchers and clinicians alike. Indeed, in the last four decades, the area of process and process-outcome studies has burgeoned considerably (Orlinsky, Grawe, & Parks, 1994). Significant and consistent findings have been reported on the link between process-outcome particularly as it relates to the identification of variables that may predict positive and/or negative outcomes. However, despite a significant accumulation of findings, the field has failed to provide a clear understanding of the basic mechanisms of change and the essential ingredients of the process of change (Greenberg, 1986; Greenberg & Pinsof, 1986; Rice & Greenberg, 1984). Moreover, the field of psychotherapy research has been intensely concerned with prediction and evaluation and, consequently, has failed to acknowledge description and explanation as two equally valid modes of generating knowledge (Greenberg, 1986; Greenberg, 1991). Thus, some researchers argue that the field has been experiencing a paradigmatic shift (Kiesler,1983, cited in Greenberg & 3 Pinsof, 1986) as it moves from hypothesis-verification to hypothesis-generation (Hill, 1990; Mahrer,1988) research due to the limitations and inadequacies of the traditional paradigm in examining the phenomenon of therapeutic change (Orlinsky etal . , 1994). As proposed by Rice and Greenberg (1984), this new perspective assumes a methodological shift from verification to discovery and a conceptual shift from process research to change process research. The conceptual shift implies that in order to understand the mechanisms of change and how psychotherapy produces change, a focus on the change process itself may be more relevant, productive, and clinically meaningful (Greenberg, 1986 a). As well, by investigating the process of change itself, the focus of the research moves away from studying what occurs in therapy in general to a process of identification, description and explanation of change within the context in which it occurs. Congruent with this new perspective, and within their respective undertakings, both clinicians and researchers alike, acknowledge the occurrence of rare but significant and meritorious moments of change within the psychotherapeutic process (Kelman, 1969). Various authors (Elliott, 1983, 1984, 1989; Mahrer & Nadler, 1986; Mahrer, Nadler, Sterner, & White, 1989; Rice & Greenberg, 1984) have suggested that it is to these significant moments of change that the psychotherapy researcher must attend in order to meaningfully understand the change process. Yet, despite this claim and the significance placed on understanding the change process in psychotherapy, few systematic investigations within the area of both individual and marital therapy have been conducted. 4 In fact, psychotherapeutic process investigations in the area of marital therapy are meager and thus, still in their infancy (Alexander et al., 1994; Gurman et al., 1986; Pinsof, 1981, 1986, 1988, 1989). Furthermore, investigations regarding the process of change in marital therapy (Greenberg, Ford, Alden, & Johnson, 1993; Greenberg, James, & Conry, 1988; Holtzworth-Munroe, Jacobson, DeKlyen, & Whisman, 1989; Johnson & Greenberg, 1988) are fairly recent undertakings. Indeed, only a handful of studies (Dubberley-Habich, 1992; Greenberg et al. 1988; Manson, 1997; Newman, 1995; Sweetman, 1996; Wiebe, 1993) have utilized the new conceptual and methodological framework proposed by Greenberg and Pinsof (1986) for the investigation of the change process in marital therapy. Therefore, research on the process of change in marital therapy is one of the areas in urgent need of investigation particularly through discovery-oriented methodologies (Mahrer, 1988) if the field is to advance in its understanding and generation of knowledge rather than its mere prediction and/or evaluation of treatments (Jacobson and Addis, 1993). Aim of the Investigation Given the scarcity of process research in marital therapy, the lack of research tapping into its mechanisms of change, the identified suggestion to understand the change process itself, and the current need to describe and explain these processes, the purpose of this investigation was to explore the change process of Experiential Systemic Marital Therapy (ExST) (Friesen, Grigg, & Newman, 1991). Specifically, the aim was to explore, discover, identify, understand, 5 and explain the in-session change episodes of three couples treated with marital ExST through an exploratory, naturalistic, discovery-oriented line of enquiry. The study intended to generate a conceptual framework outlining the process and structure of the change episodes. The psychotherapeutic conditions, therapist/clients operations, and consequences resulting from these operations were generally expected to comprise the possible structure of the change episodes. Research Questions In order to address the purpose of the investigation, the following research questions were generated. 1) Are there observable, identifiable, and recurrent in-session change episodes in couples treated for alcohol abuse with Experiential Systemic Marital Therapy? 2) If yes, what are those observable, identifiable, and recurrent change episodes? 3) What is the internal structure of those change episodes? 4) What are the therapist and client operations that facilitate change within the episodes? Overview of Design and Methodology The kind of questions posed in this study called for a qualitative, exploratory, and discovery-oriented type of methodology. A multiple-case design study (Yin, 1989), with adherence to the canons and methods of naturalistic enquiry (Denzin & Lincoln, 1994) and grounded theory analysis (Glaser & Strauss, 1967) were 6 employed. Congruent with the naturalistic research tradition, the researcher took the role of participant-observer; that is, the researcher conducted a posteriori observation and analysis of videotaped marital therapy sessions in which she was the therapist working with the identified couples. The multiple-case-embedded design within a natural setting and the grounded theory method of analysis were believed to be the most appropriate and relevant for this investigation because they allow for the intense exploration of complex, contextual, and complicated human and social processes, such as change processes, as they occur in their natural environment. In the field of marital and family therapy research, the multiple case design has been suggested as useful for the investigation of the change process within a particular theoretical model of change (Wynne, 1988). In addition, as a method of analysis, grounded theory has been identified as a valuable and appropriate method for psychotherapy process investigations (Elliott, 1983, 1984; Mahrer, 1988; Moon, Dillon & Sprenkle, 1990; Moon & Sprenkle, 1992; Rennie, Phillips, & Quartaro, 1988). In this study, three cases (couples) representing three distinct patterns of change in marital satisfaction at post-test on the Dyadic Adjustment Scale (DAS) (Spanier, 1976) were selected for the investigation. Because this study intends to explore the presence of change episodes in these three distinct patterns of change at post-test, a multiple-case design with theoretical replication, is deemed suitable. Accordingly, the pre-test-post-test difference scores on the DAS for seven couples with whom the therapist-researcher worked were ranked-ordered. Based on the DAS scores, the couples' pattern of change were identified as high, medium and 7 minimal improvement. The high improver case revealed the largest improvement on the pre-post-test difference couples' scores on the DAS, with a combined couples' difference score of 34 points. This contrasted with a combined pre-post-test difference score of 10.8 points for the medium improver couple and with a combined pre-post-test difference score of 5.15 points for the minimally improved couple. In addition, the pre-test-post-test difference scores of the three couples on a measure of alcohol dependency, the Alcohol Dependency Data (ADD) (Raistrick, Dunbar, & Davidson, 1983), and a measure of intrapersonal distress, the Symptom Checklist-90-Revised (SCL-90-R) (Derogatis, 1983) were examined in order to evaluate the consistency of change patterns identified through the DAS scores. The data sources for this study consisted of a library of 40 videotaped marital therapy sessions, extensive information on three standarized measures, and self-report data on daily functioning provided by the couples while engaged in marital therapy. The data sources were generated by a large scale investigation, The Alcohol Recovery Project (TARP), intended to explore the effectiveness of individual and marital Experiential Systemic Therapy (ExST) (Friesen et al., 1991) with male alcoholics. The findings of that investigation showed that, while a large proportion of couples improved, some deteriorated at termination (Friesen, Conry, Grigg, &Weir, 1995; Grigg, 1994). Once the couples were selected, the 40 videotaped therapy sessions were thoroughly and systematically observed by the researcher so as to acquaint herself with the data and the therapy process. Also, since the researcher was one of the 8 therapists involved in the TARP project and therefore, familiar with ExST, her biases and assumptions regarding the change process in couples therapy were documented prior to embarking on the analysis. The episodes selected were based on criteria for selection of change episodes generated prior to the selection process. An initial selection of possible episodes was compiled in videotape form for revision and evaluation by the researcher and the study's supervisor. Once agreement on the selected episodes was obtained, transcription and the actual analysis were conducted. The grounded theory method of analysis is a qualitative, processual, contextual, and interactional method which attempts to generate theory grounded in the data (Glaser & Strauss, 1967; Strauss & Corbin, 1990). Further, it is particularly suited for investigations on processes that evolve over time, especially processes of change (Glaser, 1978), because of the processual nature of the methodology. Consequently, it allowed for the investigation of change over time by describing and explaining the connections and phases inherent in the psychotherapeutic change episode. In addition, the grounded theory method is best suited for investigations of phenomena lacking in conceptual and theoretical maturity (Morse & Field, 1995; Stern, 1980). At present, investigations on the process of change and empirically derived explanatory models on the process of change in marital therapy are largely undeveloped. Since the primary purpose of grounded theory is to generate explanatory models of human behavior, the researcher seeks to identify patterns and the relationships among these patterns (Glaser, 1978,1992). This was done through 9 the comparative method of analysis in which pieces of the data were compared with one another. The analytical process in grounded theory is both hierarchical and recursive (Morse & Field, 1995). First, the researcher is simultaneously engaged in data collection and analysis in that, as initial recurring patterns begin to emerge from the data, the researcher collects further data through theoretical sampling, rather than probabilistic sampling, in order to eventually saturate the emerging categories. Second, the method requires the systematic categorization of data while limiting theorizing until patterns in the data begin to emerge from the process of categorizing. Third, the analytical process requires that the researcher engages in data collection, coding, memoing, determining a core category and recycling back to earlier steps of the process in terms of the analysis and generation of a core category. Therefore, at every step of the process, the emerging categories and conceptual order are tested against the data through hypothetical approximations and revisions (Polkinghorne, 1991). In addition, the researcher is required to maintain an open attitude to discovering and learning what is there to be discovered about the process from the observations of clinical phenomena (Mahrer, 1988). This brief overview of design and method of analysis employed to answer the research questions and generate a general model outlining the process of change as reflected in the episodes was intended as a general introduction. Chapter three offers an extensive explication and elaboration of the design and method. 10 Explanation of Terms The purpose of the investigation, research questions and methodology presented in this chapter introduced various concepts. In the following section, a definition of terms is offered in order to facilitate and develop common understanding and familiarity. Experiential Systemic Marital Therapy (ExST) ExST was developed for the treatment of individuals, couples, and families afflicted by drug and/or alcohol misuse. This treatment model is an integrative approach tending to the multiple layers of relational experience manifested at the intrapersonal, interpersonal, and the larger context (Friesen et al. 1991). Change is facilitated and promoted by accessing and journeying into the emotional, cognitive, behavioral, and physiological realms of experience through systemic, experiential, and symbolic means. As an interpersonal process, change emerges out of the dynamic interaction between the therapist and clients within a collaborative, spontaneous, creative, goal-oriented, and present-tense stance. ExST assumes that problems evolve out of a complex network of interconnected relationships which are rigid and static and thus, problems are relational in nature. Chemical Dependence The participants for the present study were part of The Alcohol Recovery Project (TARP) investigation. As such, an important inclusion criterion for participants was the husband's dependence on alcohol while the wife was the non-alcohol abusing partner. The males in this study satisfied the DSM-III-R (1987) 11 diagnostic criteria for severe Psychoactive Substance Dependence. A severe Psychoactive Substance Dependence includes symptoms markedly impairing occupational functioning, social activities and/or relationships with others. Recurrent Change Episode To date, episodes of change in marital therapy have not been clearly defined or identified. However, a few attempts at identifying and describing episodes of change from the theoretical perspectives of Emotionally Focused Therapy (EFT, Greenberg & Johnson, 1988) and ExST have been suggested. For instance, from the theoretical stand of EFT, good moments of change are characterized by greater depth of experience and affiliative interaction in conflictual episodes. These two aspects are accompanied by a process of \"softening\" of the blamer in the relationship. From the point of view of ExST, therapeutic change is conceptualized as relational novelty. Relational novelty is the process whereby clients experience greater interpersonal flexibility as new and more integrated ways of being in relationship emerge. This process is associated with greater depth of experiencing and affiliative interaction. Greenberg (1986) operationally defined a change episode as an identifiable and distinct psychotherapeutic occurrence consisting of the client problem marker, the therapist operation(s), the client's actions and interactions and the immediate session outcome. The client problem marker is an indication or statement by a client to the therapist which suggests to the therapist that the client is in a particular problematic and/or conflictual state. The therapist operation(s) refers to the actions 12 and interactions, including interventions initiated by the therapist, in order to facilitate resolution to the problematic situation. The clients' actions and interactions refer to the ongoing responses and interactions with the therapist in relation to the therapist's interventions. The in-session outcome refers to the resolution of the problematic situation which may be manifested as cognitive reorganization, integration of conflictual tendencies, new ways of relating to self and/or others, or softening of parts of self. Regarding this investigation, a change episode is defined as the transformation of ineffective, unsatisfactory and recurrent patterns of interaction between the partners into new and different ways of being and relating with each other in the session. The change episode distinguishes itself from the rest of the therapy through four distinctive characteristics. First, observable evidence of a verbal and/or nonverbal nature by one or both members of the marital dyad indicates that the client(s) is (are) in a problematic or conflictual condition requiring intervention on the part of the therapist in order to transform those problematic and/or conflictual conditions. Second, observable evidence suggests particular actions, interactions, or interventions on the part of the therapist addressing those conflictual conditions. Third, observable evidence indicates specific and ongoing responses and/or statements on the part of the client(s) to the therapist interventions. Fourth, an observable immediate outcome is identified in the form of a new way of being and relating in the session which is more satisfying and gratifying to the partners. 13 Conflictual Relational Marker The conflictual relational marker is defined as the point at which a problem or conflict was signaled as needing attention. Specifically, a signal was displayed in which one or both members of the dyad were verbally or non-verbally challenging recurrent relational patterns that were ineffective and unsatisfactory. The conflictual marker was expressed either verbally or non-verbally through statements or messages suggesting misunderstanding, confusion, detachment, alienation, distance, defensiveness, stuckness, pain, tension, and expressions of powerful and strong emotional arousal (e.g., tears, anger). Client and Therapist Actions and Interactions The client and therapist actions and interactions refer to the ongoing dialogue and activities engaged in by the members of the therapeutic system in order to address the conflictual condition of the couple. As such, they also include particular therapist's interventions and activities intended to generate a new pattern of relating or being in relationship for the couple. New Patterns of Interaction In-Session The new ways of being and interacting of the partners refer to observable indications which displayed satisfaction, contentment, flexibility, and understanding between the spouses. Regarding the initial conflictual situation, the couple interacted in more affiliative, gratifying, and satisfactory ways. 14 Internal Structure of Change Episode The internal structure of the change episode refers to the regularities and flow of the actions and interactions of the members of the therapeutic system in the co-creation of change. As such, it comprises the psychotherapeutic conditions, the client and therapist actions and interactions and the subsequent consequences. The internal structure is a representation characterizing the salient actions and interactions of the participants as they engage in the change process. Significance of the Study The need for investigations on the process of change in marital therapy through discovery-oriented methods has been identified as of paramount importance (Jacobson & Addis, 1993; Mahrer, 1988). However, only a limited number of studies have pursued this line of inquiry. This study is important and valuable for a number of reasons. First, as a process research investigation, with a focus on in-session recurrent change episodes, it will begin to illuminate the mechanisms of change in couples therapy. Secondly, because of the naturalistic, qualitative and processual nature of the method, it is definitively discovery-oriented. Thirdly, the discoveries may have valuable and useful information for researchers and practitioners. For researchers, it may provide a conceptual framework and context for quantitative research and some explanation for sometimes unexpected quantitative results. For practitioners, it may offer clinically relevant, meaningful, and rich information and insights which may facilitate understanding of clinical situations. Fourthly, the information generated from the study may offer valuable sources for the generation of hypotheses. In addition, it will expand and refine the 15 process of change as practiced in ExST. Finally, it will satisfy a personal curiosity regarding the dual role of taking part in a naturalistic investigation which asks questions similar to the ones asked by a psychotherapist. That is, while in couples therapy, one may ask what are the patterns that this couple is engaged in, what are the most appropriate interventions given the particular conflict or issue at hand, how to intervene given the operating conditions, and at what point to intervene, in discovery-oriented research, the questions of interest are \"What patterns can I identify in the problem and how are these problems related?\" (Artinian, 1986, p. 16). Having thus far introduced the study, the following chapters of this dissertation include a comprehensive review of the relevant literature on change process research, a thorough presentation of the design and method, a full description of the results, and a discussion of the conclusions emanating from the results of this investigation. The major topics to be covered in chapter two are: (1) a relevant review of the literature on individual therapy process research, particularly as it relates to the new paradigm in psychotherapeutic process investigations; (2) a review of process research in marital and family therapy from the perspective of the new paradigm; and (3) a review of change process research from the theoretical perspective of ExST. Following the review of the literature, chapter three describes the design and methodology utilized in this investigation. This chapter will describe in detail the various steps followed for the selection of the couples, the identification and selection of the change episodes, a description of the researcher's biases regarding the change process in marital therapy, an overview of Experiential Systemic Therapy, and a description of the qualitative method of analysis. Relevant sections will include the identification of couples and change episodes, theoretical assumptions about change in marital therapy espoused by the researcher, historical and theoretical tenets of grounded theory analysis, data collection and analytical procedures, rationale for the utilization of the method, and issues related to the scientific rigor of the study. Chapter four will describe the results of the investigation. Finally, chapter five will summarize and integrate the results with the current literature on the process of change in marital therapy. In addition, chapter five will also discuss the limitations of the present investigation as well as make suggestions for future research. 17 CHAPTER II REVIEW OF RELATED LITERATURE The review of the literature addresses three related areas to the investigation. First, a thorough review of the empirical strategies utilized in the study of psychotherapy is offered in order to provide a context to the present discovery-oriented study and to attest to its importance and validity. This review presents the most salient historical aspects of psychotherapy process research. Second, a review of process research in marital therapy is offered. This review focuses on how the process of marital therapy has been studied and the findings generated from these investigations. Third, a review of the process research studies which have sprung from the theoretical perspective of Experiential Systemic Therapy is offered given their relevance for the present investigation. Of particular interest for the present study is a review of how change episodes in marital therapy have been studied, identified, defined, and explained. Research Strategies in the Study of Psychotherapy The development of research in both individual and marital psychotherapy has progressed along the lines of what is known as outcome and process research. With the advent of electronic devices such as the tape recorder, Mowrer (1953) envisioned the possibility of studying the process of psychotherapeutic conversations, suggested outcome and process research as two valid and viable empirical strategies for its investigation, and provided an initial definition of both 18 strategies. Outcome research was defined as the measurement of the nature and extent of significant aspects of change before and after treatment. Process research was defined as the microscopic analysis of the ordinary course of therapeutic events (Mowrer, 1953). Currently, the distinction between outcome and process research has been a source of debate amongst researchers due to differences in definition and interpretation of outcome and process (Greenberg, 1986; Hill & Corbett, 1993; Orlinski et al., 1994). For example, because process and outcome are closely intertwined activities (Hill & Corbett, 1993), variables are conceptualized as process in some instances, while at other times they are treated as outcome (Orlinski & Howard, 1986; Orlinski et al., 1994). Nevertheless, the field of psychotherapy research has developed these useful conceptual abstractions for the purpose of organizing the observational and measurement materials generated in the conduct of its study. The Meaning of Outcome Outcome research refers to the investigation of change that occurs as a result of the process of psychotherapy. In outcome research, the focus is on the magnitude and strength of change (Lambert & Hill, 1994). Historically, outcome has been measured by indexes of change that occur between pretherapy and posttherapy assessments. Thus, Orlinski and Howard (1978, 1986) and Orlinski et al. (1994) have argued extensively for the measurement of outcome by the evidence generated outside the client-therapist relationship or at termination. Greenberg and Pinsof (1986), on the other hand, have suggested that attention be given to immediate and intermediate outcomes. The argument here is that changes 19 made inside the therapy session, such as client insight as a consequence of therapist interpretations or softening of internal conflict following the use of a Gestalt two-chair technique, are valid and appropriate outcomes. The Meaning of Process Historically, the term \"process\" has had different meanings in psychotherapy process research. In presenting a process conception of psychotherapy, Rogers (1961) investigated the process by focusing on the stages of change taking place in aspects of the client psychological functioning such as client's manner of experiencing, relationship to problems, and ways of relating to self and others. Aspects of the therapist's behavior were not considered as part of the therapeutic process; rather the therapist offered conditions purported to be facilitative of the client's process. Similar to Rogers's conception of process, Rice and Greenberg (1984) suggested a conception of process research in which the focus of the investigation is on the process of the change event itself. By segmenting therapy according to theoretically predefined criteria of change events, the researcher scrutinizes the process and generates a conceptual model of change. The assumption inherent in this definition is that outcome is part of process in that, by focusing on the change process, one concurrently focuses on the outcome of the event. Orlinski et al. (1994) provided a similar definition of process research but a different approach for the measurement of its outcome. Thus, the authors asserted that process research refers to an empirical strategy which attempts to determine those aspects of the therapeutic process which may be helpful or harmful to clients 2 0 by linking what actually occurs in psychotherapy with the final result of the endeavour. Still another way in which researchers ( Benjamin, Foster, Roberto & Estroff, 1986; Elliott, 1984) have conceptualized process research is with reference to any or all events that may be observed in therapy sessions in terms of actions, perceptions, thoughts, feelings and interactions of clients and therapists usually occurring inside the therapy sessions. Here, one focuses on the process of the event without making any prior distinction between helpful and hindering events. According to Orlinski et al. (1994), this approach to descriptive process analysis is taken when the researcher does not have a strong a priori hypothetical notion about the sources of change in psychotherapy. Others (Hill & Corbett, 1993; Lambert & Hill, 1994) have defined process research as what occurs during psychotherapy sessions by examining therapist behaviors, client behaviors, and interactions between therapists and clients. Historical Overview of Process Research In a review on the historical development of process research methods in individual psychotherapy, Hill and Corbett (1993) observed that the field has progressed through what they identified as three paradigmatic shifts. The authors identified these as: the early naturalistic process research period from approximately 1940 to 1970, the period of analogue research during the 1970s and 1980s, and the period advancing the utilization of naturalistic, qualitative methods in the late 1980s and 1990s. 21 The early naturalistic process research period was facilitated by the technological advancement of the audio-recorder, as it offered the possibility of recording and detailed examination of therapy sessions. In 1938, Frank Robinson pioneered the first program of process research in counselling psychology by recording sessions in order to provide supervision to his students and allow students to hear themselves (Hill & Corbett, 1993). The results of sessions recorded over a 10-year span generated a first set of categories on the process of therapy. For instance, Porter (1943) generated a category system for client and counsellor behavior which encompassed categories such as silence, interpretation, reflection, and tentative analysis. At this time, Carl Rogers also initiated research on the process of psychotherapy and developed a theory to guide and explain his research. Thus, in his book Counseling and Psychotherapy (1942), he suggested that, if the therapist clarified and expressed the feelings of the client, a shift from negative to positive experience would follow. This in turn would generate awareness and insight and would thereby facilitate positive action on the part of the client. Roger's second theoretical phase began in 1951 with the publication of the book Client-Centered Therapy. Here, his theoretical position shifted in that in addition to the clarification of feelings, the therapist was required to maintain a belief in the potential and the capacity of the client to work through his/her own situation. During this time Rogers published his classical theoretical notions on the necessary and sufficient conditions of therapeutic change. 22 A prominent theoretician and clinician, Rogers valued the scientific method and was committed to the objective study of the psychotherapeutic process. Based on his theoretical notions and his regard for objectivity in the scientific enterprise, a number of researchers developed objective instruments to measure various aspects of the process of psychotherapy and to validate Rogers's theory. Thus, observational measures such as The Experiencing Scale (Klein, Mathieu-Coughlan, & Keisler, 1986), a measure of client affective involvement in therapy, and The Client Vocal Quality Scale (Rice & Kerr, 1986), a linguistic measure of client emotional involvement in therapy were generated. As well, Barrett-Lennard (1962) developed the Barrett-Lennard Relationship Inventory to evaluate the perceptions of clients and therapists in relation to Rogers's theoretical notions such as therapist empathic understanding, level of positive regard, and congruence. As researchers and theoreticians were actively engaged in the testing and validation of ideas and instruments, Eysenck (1952) shocked the psychotherapeutic research community by claiming to demonstrate the ineffectiveness of psychotherapy. Based on his review of psychotherapy research, he argued that, compared to neurotic clients who had never entered therapy, the rate of improvement for neurotic clients who had received traditional (nonbehavioral) psychotherapy was equivalent after a 2-year period. Eysenck's (1952) overriding conclusion was that \"The figures fail to support the hypothesis that psychotherapy facilitates recovery from neurotic disorder\" (p.323). Such a powerful and shocking conclusion was the major thrust behind the next 30 years in psychotherapy research. 23 Given that the efficacy of psychotherapy was in doubt, the focus of investigation moved swiftly into the area of outcome research. Thus, throughout the 1950s, 1960s and 1970s, the research focus was to demonstrate the effectiveness of psychotherapy and generate evidence for the effectiveness of particular schools of therapy. Hill and Corbett (1993) argued that researchers felt compelled to investigate the effectiveness of the endeavour prior to investigating the components or aspects of the process that facilitated its effectiveness. In so doing, researchers have amply demonstrated the effectiveness of psychotherapy (Bergin & Lambert, 1978; Lambert & Bergin, 1994; Lambert, Shapiro & Bergin, 1986). Indeed, in the latest review on the effectiveness of psychotherapy, Lambert and Bergin (1994) concluded that a variety of psychotherapies have demonstrated significant statistical and clinical effects with a diverse population of clients; that, with the exception of addictive disorders, the therapeutic effects tend to be maintained; and that interpersonal and affective factors such as trust, warmth, acceptance, and human wisdom are significant stimulators of client improvement. Thus, the overall conclusion is that psychotherapy is effective as it helps people change at a faster and more significant level compared to changes which may occur naturally within the person's environment. With a satisfactory answer to the question of the effectiveness of psychotherapy, the next question that researchers prepared to investigate was \"Under what conditions will this type of client with these particular problems be changed in what ways by specific types of therapists?\" (Bergin & Garfield, 1971, p. xi). Recently, Orlinsky et al. (1994) identified the question as: \"What is effectively 2 4 therapeutic about psychotherapy?\" (p. 270). Both questions address the process of psychotherapy and its relationship to outcome by attempting to identify and determine variables and processes particularly helpful and/or harmful to clients. In so doing, and congruent with the conceptual and methodological assumptions of outcome research on controlled observation and measurement, generalization, prediction and verification (Kiesler, 1971), investigators have sought to answer these broad process-related questions. Consequently, the progression of process research ensued within the quantitative methodological tradition by linking process and outcome. The efforts afforded to the process-outcome area of research have generated some definitive findings (Orlinski et al., 1994). In their extensive, encyclopedic review of research linking process and outcome, Orlinski et al. (1994) evaluated a total of 2350 findings published from 1950 to 1992. Most of these findings were published in the period 1986-1994. Because some of the findings reviewed appeared to be consistently replicated, the authors considered them to be well established facts in the relationship of process to outcome. Some of the definitive findings demonstrated that 11 process variables were strongly connected to outcome measured at termination. Also, significant associations were found across multiple perspectives (client, therapist, rater). These variables were identified as client suitability, therapist skill, client cooperativeness versus resistance, global therapeutic bond/group cohesion, client contribution to the bond, client interactive collaboration, client expressiveness, client affirmation of the therapist, reciprocal 25 affirmation, client openness versus defensiveness, therapeutic realizations, and treatment duration. However, despite the general findings described above, the vast accumulation of studies and findings, and the significant volume of data generated by process research, various process researchers (Greenberg, 1986; Greenberg & Pinsof, 1986; Rice & Greenberg, 1984) have expressed discontent with the conceptual and methodological approaches utilized and the subsequent application of findings. These findings are based on the assumption of homogeneity of process (Kiesler, 1971) and reported findings have little relevancy for practitioners (Hill & Corbett, 1993). Moreover, although process research attempts to understand, describe, explain, and classify the important events in psychotherapy (Orlinski et al., 1994), often, only the measurable events have been investigated (Hill, 1990). Indeed, methodologically, the focus of process research has centered on the evaluation and prediction of phenomena to the detriment of its understanding, description, and explanation (Greenberg, 1991). In addition, researchers have argued that in order to understand how psychotherapy produces change, we need to focus on the change events themselves. That is, instead of investigating the psychotherapeutic process in general by attempting to determine, evaluate, and predict possible variables which may help and/or hinder the psychotherapeutic process for clients, the focus of research needs to address the change process itself and its related dynamics so as to understand its mechanisms. For instance, a valuable finding identified by Orlinski et al. (1994) in their review was that client collaboration is 26 associated with positive outcome. However, it does not tell us how and/or when client collaboration occurs, how it is facilitated and what are the specific components or properties of client collaboration. The conceptual and methodological discontent with the predominant paradigm in psychotherapy process research has generated what Kiesler (1983, cited in Greenberg & Pinsof, 1986) named a paradigmatic shift in psychotherapy process research. Conceptual Shift in Psychotherapy Process Research The conceptual shift in psychotherapy research was originally identified by Kiesler (1983), further developed by Rice and Greenberg (1984) in what was described as events-based research, and more recently as sequentially-patterned significant change episodes in psychotherapy (Greenberg, 1991; Greenberg & Pinsof, 1986). These authors recommended a shift toward the investigation of events, particularly the process of change events, with the aim to understand how change occurs in psychotherapy and develop context specific microtheory about the mechanisms of change. The shift in conceptual understanding of process research in psychotherapy emerged out of the integration of the process and outcome traditions (Greenberg & Pinsof, 1986). As indicated above, process and outcome research were conceptualized as separate empirical strategies. Process research referred to what occurred within the spatiotemporal domain of the session and outcome research referred to client improvement or deterioration outside the session, usually at termination. Gradually, researchers realized the need to link process and outcome in order to facilitate advances in the field. Kiesler (1986) indicated that, regardless 2 7 of whether the researcher's emphasis was on process or outcome research, he/she had to include measurements of both. The linking and measurement of both process and outcome generated basically two strategies: (1) the measurement of process at some point in the therapy and outcome at termination or follow-up; and (2) the averaging of process measures throughout the course of therapy and measurement of outcome at termination (Greenberg & Pinsof, 1986; Rice & Greenberg, 1984). Greenberg (1986), Greenberg and Pinsof (1986), and Rice and Greenberg (1984) have argued extensively on the limitations of these strategies. Basically, the first strategy assumes that any event at any point in the process is consistently related to outcome. However, the complexity of the psychotherapeutic process along with other intervening events (inside or outside therapy), and the accumulation of factors that may play a significant role in outcome at termination, are important elements which are usually ignored. The second strategy assumes homogeneity in the process; that is, it assumes stability and similarity of events and experiences for the participants across the time span of therapy. Greenberg and Pinsof (1986) suggested that both strategies assume what Kiesler (1971) identified as the myths of homogeneity. Thus, rather than continue to assume that all moments in psychotherapy are interchangeable, or that the process is uniform, it is suggested that psychotherapy researchers investigate both clients' and therapists' experiences in the context of the therapeutic events in which they occur. That is, the overall amount of client experiences, therapist operations (empathic reflections, interpretations), or client and therapist interactions are not as important as the 28 context, timing, and meaning of those actions and interactions for the participants. The authors argued that what is more relevant and therapeutically significant is the pattern of variables as they occur in a particular therapeutic event rather than the simple occurrence of isolated variables significantly associated to outcome at termination. In their introduction to Patterns of Change. Rice and Greenberg (1984) succinctly argued that: A basic difficulty with most of the process research that has been done is the underlying assumption of homogeneity of process. The typical approach has been to select samples from one or more sessions for rating, and to summarize ratings across samples for a single session, or even across sessions.... Clearly, process varies overtime and different processes have different meanings in different contexts. Aggregating processes as though all process during therapy is the same involves a uniformity myth from which psychotherapy research has been suffering. All process in a psychotherapy is not the same, just as all clients, all treatments, and all therapists are not the same... Different processes occur at different times in therapy and have different meanings in different contexts (p. 10). Focus on Events In events-based research, the focus is on understanding the workings of therapy by focusing on the internal relationship among variables and the process of change itself within the in-session context rather than relating single process variables to some external variable. Thus, an important principle of this strategy is the study of process in the context in which it occurs through the analysis of identifiable and recurrent moments. Furthermore, the target units of study are important client-change events occurring within the session or concurrently outside the session (Kiesler, 1986). 29 Events based research was envisioned and generated during the late 1970s by a number of researchers from different clinical orientations. These researchers were Greenberg (1984) with a Gestalt Therapy orientation, Rice (Rice & Saperia, 1984) with a Client-Centered orientation, Mathieu-Coughlan (Mathieu-Coughlan & Klein, 1984) with an Experiential Therapy orientation, and Luborski (Luborski, Singer, Hartke, Crits-Christoph, & Cohen, 1984) with a Psychoanalytic Therapy orientation. These researchers joined their efforts and collectively collaborated on the application of this new conceptual approach to psychotherapy research. This resulted in the publication of Patterns of Change: Intensive Analysis of Psychotherapy Process by Rice and Greenberg in 1984. A second volume with a focus on events based research was The Psychotherapeutic Process: A Research Handbook by Greenberg and Pinsof (1986). This handbook was intended as a resource for psychotherapy researchers and as an update to the previous compendium on process research measures, The Process of Psychotherapy: Empirical Foundations and Systems of Analysis by Kiesler (1973). While both volumes presented a variety of observational and self report measures for psychotherapy research, Greenberg and Pinsof s (1986) volume discussed in detail significant aspects of the conceptual and methodological underpinnings of events research. According to Greenberg (1986), an event is a therapeutic episode consisting of the client problem marker, the therapist operation(s), the client performance, and the immediate in-session outcome. The client problem marker is an indication or statement by the client which suggests to the therapist that the client is in a 30 particular problematic or conflictual state. The therapist operation refers to the interventions initiated by the therapist in order to facilitate resolution of the problematic state. The client performance consists of the ongoing client responses and interactions with the therapist in relationship to the interventions. Finally, the in-session outcome refers to the resolution of the conflict which can be manifested as cognitive reorganization, integration of conflictual tendencies, experiential novelty in relationship, or softening of parts of self. As reflected above, a significant aspect of change events process research assumes that outcome can be conceptualized as measurement occurring as a result of particular interventions. In fact, Greenberg (1986), Greenberg and Pinsof (1986), and Rice and Greenberg (1984) distinguished between three different types of outcome: immediate, intermediate, and final. Immediate outcome refers to change that is clearly evident within the session as a result of particular interventions and/or interaction amongst the participants. Intermediate outcome refers to changes measured through session outcome measures, usually at the end of the session. Final outcome refers to measurement taken at the end of treatment and follow-up. Ultimately, according to the authors, a thorough understanding of therapeutic change will be obtained when outcome at these three levels is simultaneously related to each other. When focusing on significant change events the researcher is attempting to answer questions such as: 1. What client in-therapy performances, or markers, suggest themselves as problem states requiring and ready for intervention? 31 2. What therapist operations are appropriate at this markers? What therapist operations will best facilitate a process of change at this marker? 3. What client performances following the markers lead to change? What are the aspects of the client performance that seem to carry the change process, and what does the final in-therapy performance or immediate outcome look like? (Greenberg, 1986, p. 6). Proponents of events-focus research assume that, by investigating these and related questions, the researcher will be able to stay closer to what actually does occur in therapy. Also, the investigation of specific therapist operations (reflection, interpretation), specific strategic episode contexts (reprocessing a critical incident) and specific relationship contexts (therapist perceived as empathic) will facilitate and enhance knowledge of the change process. Context Specific Microtheory The claim that psychotherapy research lacks clinical theory which is context specific has also been suggested as a limitation in the advancement of the field. Greenberg (1986, 1991), Greenberg and Pinsof (1986), and Rice and Greenberg (1984) have suggested that events-based research facilitates the generation of context specific clinical microtheory. Moreover, Kiesler (1986) claimed that, in order to understand the mechanisms of change, ideally we should rely on our current theories of psychotherapy; however, because they are too general, abstract and limited in the provision of guidance for the identification and discovery of significant change-process events; a need exists to develop and discover \"miniature\" theories of change events over the course of therapy. The generation of clinical microtheory may facilitate the identification and selection of moments of change. Greenberg and Pinsof (1986) claimed that clinical 3 2 theory may guide the researcher in determining where and what to look for in empirical investigations. In fact, a number of investigations have initiated the creation of context specific microtheory from the perspective of specific theories. Rice and Saperia (1984) for instance, generated a theoretical model of client resolution of problematic reactions from the theoretical perspective of client-centered therapy. As well, Greenberg (1984) utilized task analytic procedures to understand the resolution of an intrapersonal conflict from the perspective of Gestalt therapy. Elliott (1984, 1989), on the other hand, discovered significant change events in psychotherapy from the perspective of participants rather than from theory by utilizing the Interpersonal Process Recall and Comprehensive Process Analysis. Given that theoreticians and practitioners have gradually moved away from unitheoretical stands and toward theory integration in psychotherapy, and that currently there are approximately 400 different types of psychotherapies being practiced (Lambert & Hill, 1994), it may be imperative to generate empirically derived theories based on clinical practice rather than continue to prove or disprove theories which may have little applicability and usefulness to practitioners. Methodological Shift in Psychotherapy Process Research Like the field of psychotherapy research in general, the area of process research has been predominantly concerned with how to do better research rather than how to better understand psychotherapy (Orlinski et al., 1994). This concern for method over the understanding of phenomena continues to be the focus of attention in some quarters (Greenberg, 1991). In their efforts to understand how to do better research, investigators' methodological tendencies have been toward the 3 3 favoured and/or received research paradigms. The field of psychotherapy research has been dominated by the logical-empiricist quantitative investigative tradition and a culture of \"scientism\" (Greenberg, 1991, 1986). The culture of \"scientism\" has not only dictated how phenomena should be investigated, but it has also prescribed the quantitative-logical empiricist tradition as the scientific method and, therefore, superior to other investigative traditions. Various researchers (Greenberg, 1986, 1991; Greenberg & Pinsof, 1986; Mahrer, 1988; Rice & Greenberg, 1984) have argued that while proof, verification, testing of propositions and prediction are important aspects in the development and understanding of psychotherapy; discovery, exploration, description, and explanation are also essential to the scientific understanding and generation of knowledge in psychotherapy. Furthermore, it is suggested that true prediction and testing of propositions may be significantly facilitated and enhanced if, and when, rigorous description and explanation of phenomena are engendered. Furthermore, it is argued that in order to investigate how change occurs, how it works, and what leads to change in psychotherapy; a focus on discovery, context, patterns, and meanings is more appropriate and relevant due to the nature of the psychotherapeutic process (Greenberg, 1986, 1991; Greenberg & Pinsof, 1986; Mahrer, 1988). Indeed, psychotherapy is a highly complex, contextual, multifaceted, and multivariate process in which isolating variables for the purposes of manipulation, control, and random sampling is likely to limit the advancement of the field, to generate tenuous associations, to impede investigation of the essence and richness of psychotherapy, to ignore its complexity, and ultimately to report 34 irrelevant and unsubstantive findings for practitioners. In fact, a consistent lack of relevance of research findings to practice (Hill & Corbett, 1993) has been expressed extensively by clinicians. Discovery-Oriented Psychotherapy Research Given that psychotherapy research has overly emphasized verification to the detriment of discovery, a focus on the discovery of phenomena must be the subject matter of psychotherapy research, and thus, the task of psychotherapy researchers (Greenberg 1991; Greenberg & Pinsof, 1986; Mahrer 1988). These authors advanced the proposition that the description and explanation of phenomena through the identification of events, especially significant events, is what ultimately facilitates the discovery of underlying regularities and the construction of models which account for those regularities. A focus on description and explanation of underlying regularities of what actually occurs in psychotherapy, particularly the mechanisms of change, will in actuality bring psychotherapy research back to its scientific roots (Greenberg, 1986, 1991). Description in Psychotherapy Research In psychotherapy process research, description of the phenomenon of interest, that is, client change episodes, so as to understand the mechanisms of change, encompasses a description of the event itself along with the context in which that change occurs. In addition, this entails description of actions and interactions as they occur in the situation. Thus, rather than describing an event as a single dimension, for instance, reflective empathy, the researcher must describe the event 35 multidimensionally. Greenberg (1986) provides a clearly descriptive example of a therapist engaged in an intervention: An empathic reflection focused on a client's feeling of confusion, delivered in a \"warm\" voice by a therapist leaning toward the client, following a confrontation of a discrepancy in what the client is saying and doing, delivered in a supportive manner, by a therapist shifting position (p. 714). Levels of Description There are many levels in which a psychotherapeutic event may be described. The level of description refers to both the size of the unit to be described and the nature of the unit. Researchers have proposed various levels depending on the type of research, the questions asked, and the constructions of interest. For instance, Greenberg (1986) suggested a distinction among four levels of description: content (utterance), speech act, episode, and relationship. Content refers to what is actually being talked about or explored. Speech act refers to what a person is doing to another by saying or doing something, that is, the effect which one person's message or action may have on another. For instance, some examples of speech acts are advise, direct, inform, support, interpret, and threaten. Episodes refer to a meaningful sequence of interaction which the participants view as a distinct whole and thus, episodes form a distinct unit. The relationship level refers to the qualities of the relationship that the participants ascribe to. Elliott (1991) proposed a five dimensional model for the descriptive unit of analysis in process research. First, a description of the perspective of observation or the point of view (client, therapist, nonparticipant observer) from which the therapy is described must be included. Second, a description of the focus of 36 observation, which could be the therapist, the client or the therapeutic system, needs to be presented. Third, a description of aspects of the process, namely, content, action, style or quality of the communicative act must be observed. Fourth, a description of the unit level (sentence, speaking turn, episode, stage of treatment, treatment) is to be delineated. Fifth, a description of the sequential phase, that is, context, process, and impact or what happened before, during and after the unit level must be addressed. Orlinski et al.(1994) distinguished as many as nine levels of descriptive analysis for psychotherapeutic process and outcome. These nine levels are: (1) liminal (micromomentary processes, e.g., gaze shift); (2) momentary (moment-by-moment processes, e.g., specific utterances, interactive turns); (3) situational (session processes, e.g., rupture and repair of alliance); (4) daily (session-sequential processes, e.g., intersession experiences); (5) monthly (phase/short course processes, e.g., short-term treatment episodes); (6) seasonal (medium course processes, e.g., medium term treatment episodes, 2-24 months); (7) perennial (long course processes, e.g., 2-7+ years); (8) developmental (multi-treatment processes, e.g. sequential treatment episodes); and (9) biographical (therapeutic career, e.g., total treatment history). Whatever level of analysis and nature of descriptive unit the researcher chooses for his/her investigation, the relevancy and importance of the context in the description of the event must be clearly articulated. 37 Observational perspective In addition to defining the unit of description, process researchers must also decide on the observational perspective from which the study is conducted (Elliott, 1991; Lambert & Hill, 1994; Orlinski et al., 1994). Thus, one may observe and assess the process from the perspective of the participants, that is, the client and the therapist, and from the perspective of nonparticipant observers. Usually, therapy process from the perspective of the participants has been assessed through rating scales and post-session questionnaires, while nonparticipant assessment has been conducted through observation of video-recordings of therapy sessions (Orlinski et al., 1994). Orlinski and Howard (1986) argued that the phenomenon of therapy differs according to who observes the process. Elliott and James (1989), in discussing the strengths and weaknesses of the three perspectives, suggested that this is not only a methodological consideration but that the perceptions differ and must be accounted for in the research. Hill, Helms, Tichenor, Spiegel, O'Grady, and Perry (1988) reported that clients and therapists differ in what they perceive as helpful in the therapy process. Although noting that the two perspectives offer different ways of looking at the process, Hill et al. (1988) were noncommittal in terms of choosing one perspective over the other. Furthermore, whereas originally it was assumed that nonparticipant observers had less biases about the process, it is now believed that nonparticipant observers have as many biases as clients and therapists (Lambert & Hill, 1994). These biases seem to differ according to the level of involvement in the process of therapy. 38 Explanation in Psychotherapy Research As indicated above, explanation in psychotherapy research has been greatly neglected. Although researchers have attempted to explain events by seeking to investigate associations between process variables and outcome, that is, explaining why something is likely to occur or under what conditions an event is likely to occur, explanation about the nature and properties of psychotherapeutic events has been lacking. Discovery oriented explanation in which the nature and structure of psychotherapeutic events, the operations under specific situations, and the properties and dimensions are fully interconnected will generate new knowledge and facilitate clinically relevant understanding of the mechanisms of change (Greenberg, 1986 a). This means that data require interpretation in order to have meaning and explanatory significance. Research Approaches There are four research approaches identified within this new paradigm of psychotherapy research. Greenberg (1986) identified performance analysis (Rice & Greenberg, 1984), empirical pattern identification (Gottman, Markman, & Notarious, 1977) and theory testing (Sampson & Weiss, 1986) as three approaches which highlight the investigation of intensive analysis of process in order to elucidate the internal pattern of variables rather than the aggregation of variables. An additional approach is discovery oriented (Elliott, 1984; Mahrer, 1988; Rennie, Phillips, & Quartaro, 1988). These four approaches highlight the discovery of patterns within significant and potent events in order to describe and explain the mechanisms of change. Greenberg (1986) suggested that empirical pattern identification, theory 39 testing and performance analysis are based on two underlying assumptions. These assumptions are (1) the focus of the investigation must be on recurring events, and (2) theory is utilized to guide one's observations. While empirical pattern identification and theory testing utilize quantitative methods of analysis, performance analysis and the discovery oriented methods proposed by Elliott (1984), Mahrer (1988), and Rennie, Phillips, and Quartaro (1988) are qualitative in nature. Empirical Pattern Identification Empirical pattern identification utilizes statistical methods to find probabilistic sequences by collapsing the occurrence of sequences over time in order to obtain an overall representation of what occurs in therapy. For instance, in an empirical pattern analysis investigation, Gottman, Markman and Notarious (1977) analyzed the sequential interaction patterns of clinic couples and nonclinic couples using Markow chain analysis. The authors showed that clinic and nonclinic couples differ in their patterns of communication and problem solving tasks. Greenberg (1986) argued that, although sequential analysis is useful in the identification of sequential dependencies (the probability that x occurring given that y has occurred) among a number of variables, this approach does not facilitate the isolation of the phenomena that represent the process of change because it aggregates data. As well, through this approach, the option of refining and modifying categories generated as well as discovering new categories as one moves along is impossible because dependencies are discovered only in the behaviors represented in the coding system which one originally started with. 40 Theory Testing In a theory testing approach, the researcher intensively analyzes process patterns at specific moments in therapy so as to test the explanatory power of theoretically derived hypotheses. Thus, although the option for analyzing highly significant moments of therapeutic change is possible, the purpose is to test rather than discover or generate theory. For instance, the testing of the explanatory power of two competing theories or hypotheses is a significant feature of the research approach implemented by Sampson and Weiss (1986) at the Mount Zion Psychotherapy Research Group. Sampson and Weiss (1986) utilized a method in which they inferred, through reasoning from theory, instances in which one theory predicts one finding and a competing theory predicting a different finding. Then, based on the application of objective measures and ratings of observations, the researcher determines which theory is in better agreement with the actual observations. The events studied are recurrent, relevant and readily observable by practitioners. Performance Analysis Rice and Greenberg (1984) suggested a rational performance analysis approach to the intensive analysis of process psychotherapy. Greenberg (1984) and Rice and Greenberg (1984) suggested the task analysis method as appropriate and effective for the study of complex human performance. In task analysis, the researcher attempts to understand how a problem is solved by specifying the components of competence in successful performance. Greenberg (1984) suggested five strategies for the task analysis of therapeutic events. These are the 41 description of the task, the specifications of the task environment, the rational task analysis, the empirical task analysis, and the model construction. The description of the task entails the identification and the clear and thorough selection and description of the client's task to be performed. The client's performances are regarded as indicators or \"markers\" of the usually affective tasks in which the client is engaged (Greenberg, 1984). The appearance of these indicators suggests that the client is ready to work on a particular task, to discover some solution, and to attain a particular goal. The indicator or \"marker\" is the \"when\" of the event. This may be a conflictual experience, a problematic reaction or incomplete affective task (unfinished business) which the client may need resolution. A discriminable therapeutic event is demarcated by the marker, the \"then\" of the event or the therapist intervention, and the subsequent client processing. The specification of the task environment refers to the delineation of suitable therapist's interventions in order to facilitate the resolution of the task. The rational task analysis is performed by the investigator through reflections on the task to be solved in order to discover possible strategies and components of resolution. Here, the investigator engages freely in thought experiments in order to obtain the essence of the resolution and generates a framework for understanding the client performance in the resolution of the task. The empirical task analysis refers to the description of the actual moment by moment performance of the client engaged in the resolution of the task. The empirical task is compared and analyzed in relation to the rational task analysis. Through an iterative process between the rational and empirical task analysis, the researcher generates a performance model 42 which is the best predictor of the client's therapeutic process when dealing with a particular type of task (Greenberg, 1984). Various researchers have utilized task analysis as their method of investigation. Greenberg's (1984) work on intrapsychic conflicts represented a significant effort in utilizing task analysis (Clarke & Greenberg, 1986; Greenberg, 1980, 1983, 1984; Greenberg & Clarke, 1979) as a method in the intensive analysis of the mechanisms of change and the generation of a conceptual model accounting for those processes. Rice and Saperia (1984) also utilized task analysis in order to understand the resolution of clients' problematic reactions to particular situations from the theoretical standpoint of person-centered therapy. Heatherington and Friedlander (1990) expanded the method of task analysis by applying it to the multi-person level of family therapy. Based on the analysis of taped sessions of structural family therapy, the authors suggested that task analysis can be successfully applied to the study of multiperson events in family therapy. Although rational performance models have been strongly championed by Greenberg and his associates as the mode of enquiry for psychotherapy process research, particularly the mechanisms of change, this approach has been criticised for its underlying assumptions. Packer and Addison (1989), for instance, argued that rather than guiding the domain of study toward action in context, the rational approach of task analysis attempts to reconstruct a competence model that underlies the performance. As such, the rationalist task performance analysis places greater importance on the internal structure of competence rather than the connections between actions and setting, or the aspects of context which bear 43 meaning to the actions and interactions of the participants. Besides, the character of explanation takes the form of a reconstruction of the structures underlying performance which limits understanding of human action (Packer & Addison, 1989). Discovery Oriented Methods Contrary to performance analysis, in which the researcher constructs a performance model through the derivation of events and actions from a particular clinical theory (Gestalt therapy, Person-Centered Therapy), discovery oriented methods (Elliott, 1984, 1989; Mahrer, 1988; Rennie, Phillips, & Quartaro, 1988) attempt to describe and explain what occurs within psychotherapy from a nontheoretical stance. That is, events or occurrences and experiences of participants are identified with an open attitude and open frame of reference in order to learn about the observations. In addition, psychotherapeutic events are described and explained from a pantheoretical (Hill, 1990) perspective in which the researcher utilizes multiple clinical theoretical perspectives to understand phenomena and generate hypotheses with the ultimate goal being the development of theory. Elliott (1984) utilized a discovery oriented research strategy to investigate significant change events in individual psychotherapy which combined Interpersonal Process Recall (IPR) and Comprehensive Process Analysis (CPA). This research strategy to change events is based on four underlying operating assumptions. As a purely qualitative strategy, the first assumption is that research on the psychotherapy process must start with the experiences and perceptions of the participants (clients and therapist). The observational perspective may be from 4 4 the participants themselves (phenomenological) or from trained observers (behavioral). Secondly, client and/or therapist experiences and perceptions are more useful when they are anchored in observable, specific behavioral (verbal and nonverbal) events through the interpersonal process recall. Thirdly, research on the process of psychotherapy needs to highlight significant change events, or turning points in which something changes for the client rather than continue with the usual method of random sampling and averaging across the process. The fourth operating principle assumes that since significant change events are rare, infrequent, and highly complex; they should be investigated in a thorough and comprehensive manner when they occur. Elliott's (1984) discovery oriented strategy comprises two procedures: The Interpersonal Process Recall (IPR) and the Comprehensive Process Analysis (CPA). The IPR is an interview procedure in which a taped session is separately played back to the client and therapist in order to identify significant change events by rating the helpfulness of each therapist response. Then, both therapist and client select the two most significantly helpful and hindering interventions. Another way in which the interpersonal process recall is utilized is by requesting, from both client and therapist, a qualitative description of significant change events; that is, a description and elaboration of how the significant event unfolded. Once the researcher has identified the significant event, the comprehensive process analysis is initiated. The comprehensive process analysis procedure (Elliott, 1984, 1989) consists of a definition of the interactive episode and its context, transcription of both the 45 episode and the data obtained through the IPR, measurement of process variables, that is, the assessment of the episode through a battery of quantitative or qualitative process measures, and the final integration of the data by analyzing common characteristics or themes in narrative form or in a summary table form. It is suggested that the crucial step in the analysis is the systematic description of significant change events through a battery of process measures. In one study, Elliott (1984) identified insight as a significant, recurrent and potent change event in psychotherapy. Based on the selection of four insight events, Elliott (1984) applied both procedures and generated a tentative model of insight events. In another study exploring significant change events, Elliott (1989) utilized comprehensive process analysis and defined awareness as a significant client change event. Based on the selection and analysis of eight episodes, the author generated a model linking various therapist interactions with client awareness. Another research strategy suggested within the discovery oriented methods is Mahrer's (1988) approach based on the procedures and analytical paradigm of the grounded theory method (Glaser, 1978; Glaser & Strauss, 1967). Mahrer (1988) indicated that in a discovery oriented investigation there are at least two strategies to the endeavour: (1) to provide a closer, discovery oriented look at psychotherapeutic phenomena, and (2) to discover the relations amongst psychotherapeutic conditions, operations, and consequences. The first strategy entails selecting the target of the investigation, obtaining instances of the target, obtaining an instrument such as audiotapes or videotapes of sessions for taking a 4 6 closer look, gathering the data, and making discovery-oriented sense of the data. The second strategy refers to discovering interconnections among conditions, operations, and consequences of specific and concrete events occurring within the session. Conditions refer to how the client is doing or being in the session. For instance, the client may start the session with a deep sigh, or may be deeply silent. Operations refer to the therapist intervention immediately after the client's expression, and may consist of, for example, offering an empathic reflection, intensifying the feeling or bodily felt experience, or inviting the client to stay in the present. Consequences refer to the client's actions and interactions subsequent to the therapist intervention. As suggested by Mahrer (1988), within this strategy there are three general steps: (1) specifying the question (i.e., given this consequence, what operations and under what conditions can one achieve this consequence?), (2) obtaining the data, and (3) examining the data to obtain a discovery-oriented answer. Rennie, Phillips and Quartaro (1988), arguing for the need to move beyond theory verification and into discovery, suggested and applied the grounded theory method to the investigation of the process of psychotherapy. Specifically, the authors demonstrated the analytical steps of the grounded theory method and the generation of categories from a study on the experiences of psychotherapy as recalled by former clients. The grounded theory method was suggested as a powerful tool for the generation of theory grounded in the data and for facilitating movement into the creative, inductive, and discovery oriented research needed in psychotherapy. 47 In all of the discovery-oriented strategies, Elliott's (1984, 1989) comprehensive process analysis, Mahrer's (1988) discovery oriented methods, and Rennie, Phillips, and Quartaro's (1988) application of grounded theory, attempts are made to open up new avenues for psychotherapeutic research, to integrate research, practice, and theory, and to generate rigorous, qualitative, and scientific conceptualizations from the data. Hill (1990) argued for the necessity, validity, and usefulness of engaging in observations of the therapy process through discovery oriented methods so as to generate testable hypotheses and develop empirically derived theories on the process of psychotherapy. However, she also warned researchers of the potential dangers involved in discovery oriented research. These were identified as (1) the tendency of researchers to investigate trivial aspects of process, with little relevance to clinical practice but easily measured (the use of plural words or head nods); (2) the generation of rich and dense amounts of data which may make it difficult for researchers to discern and isolate significant aspects of the process; and (3) the possible unstated biases which may affect the results of the investigation. Given that, in the last 10 years, a significant conceptual and methodological shift was called for in the area of process research in psychotherapy and that important programmatic research was initiated in the area of individual therapy, a similar call was made in the area of marital and family therapy process research, particularly with reference to the need for methodological pluralism and diversity (Greenberg & Pinsof, 1986; Jacobson & Addis, 1993; Keeney & Ray, 1992; Moon, Dillon, & Sprenkle, 1990; Moon & Sprenkle, 1992). However, marital and family 48 therapy process research has evolved at a much slower pace. In fact, researchers (Alexander etal., 1994; Gurman etal., 1986; Pinsof, 1981, 1986, 1988, 1989) suggest that marital and family therapy process research is still in its infancy. Given that the present investigation attempts to discover and understand change episodes in marital therapy from the perspective of the new research paradigm in process research psychotherapy, the next section undertakes a review of marital therapy process research. Specifically, it reviews investigations on the process of change in marital therapy from the perspective of events-based research. The Process of Change in Marital Therapy Like individual therapy process research, investigations on the process of change in marital therapy have progressed along both quantitative and qualitative lines of enquiry. Within the quantitative mode, the traditional assessment tools utilized have been self reports of participants through questionnaires and observation of process by expert raters through the utilization of coding systems which quantify various aspects of process. Within the qualitative line of enquiry, investigators have primarily utilized task analysis (Greenberg et al., 1988) in the investigation of the change process. Quantitative Research on the Process of Change in Marital Therapy Reviews (Alexander et al.,1994; Jacobson & Addis, 1983) indicate that most of the research efforts in the process domain have examined the process of change from the perspective of Emotionally Focused Marital Therapy (EFT) (Greenberg & 49 Johnson, 1988). Yet, only three studies have examined the process of change from this clinical perspective. Johnson and Greenberg (1988), noting a complete lack of research regarding the change process in marital therapy, investigated the process of change in the \"best\" sessions of EFT. The authors selected a total of six couples on the basis of extreme change scores (three low change and three high change) from a pool of 29 couples originally involved in an EFT efficacy study (Johnson & Greenberg, 1985). The selection of the six couples was based on their marital satisfaction as reflected by the Dyadic Adjustment Scale (DAS) (Spanier, 1976). Then, the session in which the couple made the most progress, as identified by the therapist, and which the couple rated as the most useful session in resolving the conflict, was selected for intensive analysis. Intensive analysis comprised the application of the Experiencing Scale (ES) (Klein, Mathieu, Keisler, & Glendlin, 1969), a measure of client involvement in therapy, and the Structural Analysis of Social Behavior (SASB) (Benjamin, 1974), a measure of interpersonal dimensions in which dialogue is categorized and analyzed. The authors concluded that, in comparison with the least successful couples, the most successful couples had a higher percentage of affiliative and autonomous responses on the SASB and a deeper level of experiencing and disclosure on the ES. As well, the blamer in the relationship moved from a position of limited participation and emotional involvement to an increased level of self-exploration, deeper experiencing, and integration. Instances of softening, defined as redefining the relationship structure by connecting with emotional experiences underlying rigid interactional patterns, were only identified in 5 0 the successful couples. As far as therapeutic interventions were concerned, the authors suggested that, for therapists working from an experiential perspective, a focus on facilitating the deepening of experience and self disclosure, particularly accessing underlying needs of the blamer when the other partner was ready to respond, were especially effective. In a more recent publication, Greenberg et al. (1993) reported on three studies investigating the in-session change processes from the perspective of EFT. The first reported study by Vaughan (1986) investigated whether couples in conflict would demonstrate greater affiliation and interdependence by the end of therapy. Based on ratings of audio recordings of interactional episodes during session two and session seven on 22 couples, it was concluded that EFT facilitated autonomous and affiliative behaviors during the latter stages of therapy. Partners were more affirming, supportive, understanding, self-disclosing and self-expressive in the latter stages of therapy rather than at the beginning. The second reported study (Alden, 1989) investigated the degree of affiliation between peak (highly productive) and poor (unproductive) sessions as perceived by the couple through a postsession questionnaire. The findings from this study suggested that the peak sessions had a greater proportion of friendly statements when compared with poor sessions, which had a greater proportion of hostile statements. As well, the peak sessions showed a deeper level of experiencing and a greater proportion of self-focused positive statements (disclosing, expressing, approaching and enjoying) than other-focused positive statements (encouraging friendly autonomy and friendly influence). 51 In the third reported study, Ford (1989) investigated the role of intimate self-disclosure in the couple. In this study, the selection of the session for investigation was based on the therapist and couples's rating of session success through a postsession questionnaire. The selection of the intimate self-disclosing event was based on the researcher's observation of videotapes of the selected sessions. The findings from this study suggested that partners are more likely to respond affiliatively after the therapist facilitates intimate self-disclosure on their part and that self-disclosing of one partner facilitates self-disclosing on the other. From the clinical-theoretical perspective of behavioral marital therapy, Jacobson et al. (1993) investigated client and therapist behaviors associated with positive outcome at termination. Clients and therapists completed process ratings after each therapy session. This study reported that from the clients' point of view, the couples that gained the most from this form of therapy were those who were actively and collaboratively engaged in the therapy as well as those who complied with the homework assignments. From the therapists' point of view, therapists rated their effectiveness in terms of facilitating collaboration in the therapeutic environment. Based on their findings, the authors concluded that behavioral marital therapy's success depends on clients' involvement and compliance. Qualitative Research on the Process of Change in Marital Therapy Very little qualitative research exists within the domain of change processes in marital therapy, in spite of its current theoretically undeveloped status (Jacobson & Addis, 1993; Wynne, 1988). However, despite this scarcity, a few researchers have initiated qualitative investigations on the process of change in marital therapy in the 52 last decade. Investigations from the theoretical points of view of ExST, EFT, and Structural family therapy are notably valuable. Greenberg et al. (1988) investigated the process of change of EFT from the point of view of 25 couples who had participated in a previous outcome study (Johnson & Greenberg, 1985). The investigators in this study performed a task analysis in which the rational model was generated from a number of clinical notions derived from the Couples Project (Johnson & Greenberg, 1985) and the empirical model generated from three procedures. First, the couples were interviewed at a 4-month follow-up about incidents that stood out as helpful and how change came about for them during these incidents. Second, these incidents were sorted and categorized according to a cue for sorting. Third, these categories were computer analyzed in order to obtain latent categories. From the analysis, five empirically derived categories were generated. These categories were: (1) expression of underlying feelings leads to change in interpersonal perception; (2) expressing feelings and needs facilitates change in the interaction; (3) acquiring understanding (intellectual and/or emotional) of relationships in addition to understanding of self and partner's dynamics; (4) taking responsibility for experience; and (5) receiving validation, particularly therapist's validation of some aspect of the partner's experience facilitated change. Similar to Greenberg et al. (1988), Wark (1994) studied the process of marital change by exploring the clients' perceptions of change immediately after the therapy session rather than retrospectively. The clients' ideas regarding their psychotherapeutic change were qualitatively analyzed by discovering similarities 53 and dissimilarities of clients' ideas. Wark's (1994) findings comprised six categories for clients' positive critical incidents (positive results, routine provided by structure, alternative perspectives, non-directive style of the therapist, directive style of the therapist, and focus on positives by the therapist) and four categories for therapist's positive critical incidents (signs of readiness for change, techniques for change, client interaction in session, and change outcome). The author also identified three clients' negative critical incidents (no follow-through with assignments, therapist imposition, and no resolution of problems) and two therapist's negative critical incidents (therapist took responsibility for change and not enough data gathering). Heatherington and Friedlander (1990) applied task analysis to two sessions of Salvador Minuchin's structural family therapy. The authors generated a clinical model of analysis by observing the two sessions and discovering a salient and dramatic moment defined as clinically significant. During these two clinically significant moments the parent and child interactions were characterized as stuck in a pursue-distance pattern of interaction.The analysis of two therapy sessions of parent-child interactions revealed a shift in interpersonal relations from a marker identified as pursue-distance impasse, to a resolution characterized by becoming unstuck from rigid and dysfunctional patterns of relating. The task required to resolve the impasse was defined as establishing family collaboration. The authors termed the change event as commitment to engage on the part of the distancer. In another study, Friedlander, Heatherington, Johnson, and Skowron (1994) significantly expanded their previous investigation and generated a conceptual model of the change process in family therapy named sustained engagement. 54 Through a qualitative inductive method of analysis, these researchers discovered therapeutic episodes where family members moved from disengagement from each other to sustained engagement while solving a task. That is, the researchers developed working definitions of the marker of the event (disengagement) and the resolution of the event (engagement) in order to distinguish successful from unsuccessful events. The model of change describing movement from disengagement to sustained engagement consisted of patterns and themes absent from the unsuccessful events. The generated model consisted of five steps the family was engaged in during the resolution of the task. These steps were described as (1) recognition of personal contribution to impasse; (2) communication or emotional disclosure about the impasse; (3) acknowledgment of other's thoughts and feelings; (4) the formation of new attributions or constructions about the impasse; and (5) recognition of potential benefits to engaging, initiated by either the therapist or a family member. Recently, Christensen, Russell, Miller, and Peterson (1998) investigated the process of change in marital therapy by qualitatively analyzing the couples's perceptions of change as expressed through open-ended interviews. These researchers utilized the grounded theory method to analyze the couples's responses. Three clusters of change were identified. These clusters of change occurred in the couple's affect (feelings about themselves, the relationship or their partner), cognition (definition of the problem or the relationship) and communication (style of relating and talking). The authors reported that change within these clusters occurred gradually and no cluster predominated over others. In addition, 55 these researchers identified five psychotherapeutic contextual factors as preconditions contributing to the perceived change. These contextual factors were categorized as safety, fairness, normalization, hope and pacing. Research on the Change Process of Experiential Systemic Therapy A number of qualitative investigations on the process of change in marital therapy from the theoretical perspective of ExST have been conducted in the last 5 years. Although some of these investigations have focused on the applications of specific techniques and their impact on the process, others have explored the process of change as conceptualized in ExST. Two investigations regarding the application of techniques and their impact on the process of change have been conducted. Dubberley-Habich (1992) employed Conversational Analysis to investigate the change process of a couple engaged in a ritualistic ceremonial transaction. The other study (Wiebe,1993) examined the application of symbolic externalization via Comprehensive Discourse Analysis (CDA). Both studies generated themes describing the actions and interactions of the participants during the application of the techniques. Process research concerning the change process in ExST is a fairly recent undertaking and valuable for the present study. Newman (1995) investigated the change process of a successful case of marital therapy and attempted to expand the theoretical notion of relational novelty (change construct) from the perspective of ExST. Based on the identification of a successful case of marital therapy, the researcher selected two episodes comprising the theoretical notion of relational novelty as advanced by ExST. These episodes were then analyzed through the 56 procedures of Comprehensive Discourse Analysis. The identification of the successful case was based on three criteria: (1) documented client satisfaction with the therapy, the cessation of alcoholic drinking, and marital satisfaction at posttest and follow-up; (2) therapist's satisfaction with her work with the couple; and (3) evidence of relational novelty in a number of sessions. The selection of the two relatipnally novel episodes for analysis was based on six criteria as defined by ExST theory. It must be noted that, although eight relationally novel episodes were identified within the 15 therapy sessions, only three satisfied the defined selection criteria for further analysis. This means that the change process as reflected by this therapeutic modality encompasses processes unaccounted for in the criteria identified. Based on the discourse analysis of the two episodes, it was discovered that intimacy and reconciliation of initial distant beliefs and practices were facilitated through the utilization of intense experiential activities and the maintenance of a collaborative atmosphere. These actions encouraged the couple to engage in empathic and self-disclosing interaction, to share their vulnerabilities, and to increase their cooperation with one another. It was also found that the therapist facilitated intimacy when accepting clients' experiences and adopting clients' language during the interactions. The researcher termed the interactional process by which the therapist and clients facilitated the creation of relational novelty as syncretic change process. Sweetman (1996) investigated the resolution of a Relational Impasse event (Rl event) in two successful cases and one unsuccessful case of marital ExST via the task analysis method. The author formulated an ExST-specific theoretical 57 model and a transtheoretical empirical model of the resolution of the relational impasse event. The transtheoretical empirical model identified five steps in couple and therapist performance. The five steps in couple performance considered transtheoretically meaningful were: (1) formation of a strong therapeutic relationship; (2) disclosure of emotional reactions to the Rl event; (3) acknowledgment of the problem; (4) individual self-disclosure regarding the impasse; and (5) insight and resolution. The five steps concerning the therapist performance were identified as: (1) establishing a therapeutic relationship; (2) validating client's emotions; (3) increasing client awareness of the problem situation; (4) facilitating client experiencing; and (5) assisting client reframing of experience. The latest investigation on the process of change in ExST by Manson (1997) examined a Significant Intrapersonal Event (SIE) in a successful and an unsuccessful case of marital therapy. The significant intrapersonal event was defined as a meaningful encounter with a previously unregarded or avoided part of self during a relationally novel episode. Here, the author employed Comprehensive Process Analysis in order to discover the context, the process, and the effects of the event in the two cases. The comparative process analysis of the two cases discovered 13 differences between the successful and the unsuccessful case. These differences were identified as: (1) process-flow; (2) disequilibrium; (3) readiness; (4) symbolic experience; (5) intelligence; (6) forgiveness experience; (7) core conflict; (8) stress level; (9) self-support; (10) spiritual support; (11) supportive 58 partner; (12) support network; and (13) individuated interpersonal experience. Client readiness was identified as primary in distinguishing between the successful and the unsuccessful case. Although a limited number of investigations on the process of change in rnarital therapy have been conducted, valuable findings have been generated. These findings revealed the discovery of categories in the cognitive and emotive realm of the couple's experience in therapy. Cognitive aspects of marital change such as defining the conflict (Christensen et al. 1998), recognition of personal contribution to the impasse (Friedlander et al. 1994), acquiring understanding -(Greenberg et al. 1988), reconciliation of discrepant beliefs (Newman, 1995) and acknowledgment of the problem (Sweetman, 1996) were identified as important processes of marital change in successful cases. As well, emotive processes identified as expressing affect about self, other and relationship (Christensen et al. 1998), disclosure of emotional reactions (Sweetman, 1996), emotional and spiritual support (Manson, 1997), emotional disclosure about the impasse (Friedlander et al. 1994) and expressing underlying feelings and needs (Greenberg et al. 1988) were found to contribute to marital change. In addition, the findings emerging from these studies have produced either step-models or identification and description of categories in their conceptualization of marital change. Studies utilizing qualitative inductive method (Friedlander et al. 1994) and task analysis (Sweetman, 1996), for instance, have generated step-models in their descriptions of marital change. Others (Christensen et al. 1998; 59 Greenberg et al. 1988; Manson, 1997; Newman, 1995; Wark, 1994) have provided distinct categories describing change in marital therapy. Summary of the Review The field of individual and marital psychotherapy process research has experienced a significant shift as researchers conceived new ways of conceptualizing the phenomena under study and developed new lines of enquiry and methods congruent with its conceptualization. As a consequence, a shift toward events-based research with a focus on significant and recurrent therapeutic change events was recommended. Methodologically, a move toward knowledge generation, namely, description and explanation in order to generate understanding of the basic mechanisms underlying therapeutic change, was suggested. Although the new technology for change process research has been available for the last ten years, few investigations have pursued its utilization. This is particularly noticeable in the area of marital therapy. As evidenced by this review, research on the process of change in marital therapy is scarce and uncommon. In addition, the review indicates a lack of empirically derived conceptual models to account for the process of change in marital therapy. However, despite this scarcity, valuable initial efforts (Friedlander et al., 1994; Greenberg etal., 1988; Manson, 1997; Newman, 1995; Sweetman, 1996) have attempted to understand the mechanisms of change in marital therapy. This study further contributes to and elaborates on the previous investigations by exploring the process of change in marital therapy through a 60 discovery-oriented line of inquiry in order to generate a model accounting for the actions and interactions of the participants as evidenced in episodes of change. 61 C H A P T E R III METHODOLOGY The selection of an investigative approach demands consideration of both the nature of the phenomenon studied and what is known about it. For the purposes of the present study, a naturalistic, discovery-oriented, qualitative line of enquiry was deemed suitable. First, the nature of marital therapeutic change is complex, multifaceted, contextual and processual, thus, amenable to a qualitative approach. Second, as evidenced in the literature review, knowledge on the process of change in marital therapy is incipient, therefore appropriate for qualitative investigations. Third, the new conceptual and methodological shift for process research attests to the usefulness of discovery-oriented methodologies, particularly when much is unknown and yet to be discovered. Thus, in pursuing a qualitative, naturalistic, discovery-oriented investigation, this study explores, identifies, describes and elucidates change episodes in marital ExST. This chapter addresses the relevant and pertinent topics related to the methodology. First, the design of the current investigation is presented. Second, study participants, selection of cases, and selection of episodes of change are discussed. Third, given the nature of the study, naturalistic participant-observer, a description of the researcher's biases about psychotherapeutic change is presented. This is followed by a brief overview of ExST as the theory of change underlying the therapy. Then, an overview of grounded theory, as the method of analysis, is introduced along with its historical background and precursors. Finally, 62 the application of the investigative procedures and data analysis utilized to answer the research questions posited are presented. Design The design for the present study consisted of a multiple case embedded design (Yin, 1989) with adherence to the canons of naturalistic enquiry (Denzin & Lincoln, 1994). The case design is suitable when the questions and aim of the investigation call for intensive and in-depth exploration of the data collection and analysis over time (Gelso, 1979). In addition, the case study is appropriate when the relevant behaviors cannot be manipulated; that is, the investigator has little control over the events studied and, thus, observation is a significant source of evidence (Cook & Campbell, 1979; Hayes, 1981; Howard, 1983). A case study design is relevant when investigating events within a real-life context (Yin, 1989). A multiple-case embedded design involves more than one case and more than one unit of analysis (Yin, 1989). In a single case study design, the focus of investigation is on the individual unit or individual entity. However, multiple case design allows the research to extend and expand the findings and relationships beyond the individual case (Yin, 1989). Multiple case designs alllow for the replication of findings by exploring, observing, and analyzing case by case or, as Hilliard (1993) suggested, the generality of the findings is addressed through case by case replication. Replication may be literal or theoretical. The logic of replication assumes that cases are carefully selected so that they either predict similar results or produce diverse results but for predictable reasons (Yin, 1989). Accordingly, multiple case 63 studies may replicate findings in a manner similar to multiple experiments, either with similar results (literal replication), or contrary results (theoretical replication). To answer the first research question regarding the presence of observable, identifiable, and recurrent in-session change episodes in couples treated with ExST, three cases (couples) producing diverse results at post-test on a measure of marital satisfaction were selected for intensive and systematic observation. Specifically, the combined pre-post-test difference couples' scores on the DAS produced a high, a medium, and a minimal improver case. As such, theoretical rather than literal replication was intended in order to detect the presence of episodes of change throughout the continium of change. In addition to the three cases representing divergent results, the design also allows for the presence of units of analysis (possible change episodes) to be discovered. These units of analysis are embedded within the cases. Yin (1989) defined a multiple-case embedded design as a research scheme consisting of more than one case, with either literal or theoretical replication, and involving more than one unit of analysis embedded within the case. In sum, the questions and aim of this investigation called for a multiple-case rather than a single-case design. That is, in order to explore the presence of episodes of change along a continuum of post-test marital improvement (high, medium and minimal), the intense and systematic observation of more than one case was required. 64 Investigative Procedure The following section describes in detail the investigative procedure. This includes an outline of The Alcohol Recovery Project, particularly regarding subject and therapist inclusion in the research project that generated the data sources for the present study. Then, a description of the methods and criteria utilized to select the cases and the change episodes are presented. The procedure is detailed under three major headings. These headings are: the data sources for the present study, identification and description of case selection, and identification and description of episode selection. Participants The participants and data for the present investigation were obtained from a large programmatic investigation, The Alcohol Recovery Project (TARP) carried out in B.C. between 1989 and 1994. TARP investigated the efficacy of individual and marital ExST with male alcoholics. A total of 150 families agreed to participate in this investigation (see Appendix A: Participant Family's Consent Form). The couples treatment group involved male alcoholics and the non-alcohol misusing female partners. The couples could attend a maximum of 15 sessions of the ExST marital therapy format over a 20-week period. The average number of sessions attended was 13.3, SD = 2.40. Subject Inclusion Criteria in TARP In order for couples to be included in TARP, the following eligibility criteria had to be satisfied. 65 (1) The couple had to involve a male alcoholic living with a non-alcohol misusing female partner. The alcoholic was to have consumed alcohol within the last 3 months to be eligible. The male alcoholic had to have a score above the critical cut-off score of 5 on the Michigan Alcoholism Screening Test (MAST)(Selzer, 1971). (2) The non-alcohol misusing female partner had to have no alcohol dependency problems in the last 5 years. (3) Both partners had to be between the ages of 21 and 65 years old. (4) The couple had to be experiencing marital distress but maintaining a commitment to the preservation of the relationship. The couple's extent of marital distress had to be significant, in that, both members in the relationship had to have obtained a score below the value of 100 on the Dyadic Adjustment Scale (Spanier, 1976). (5) The couple had to be living together in a marriage or common-law relationship for at least 1 year. (6) The couple had to be parenting at least one child living at home. (7) The couple had to be willing to participate in 15 sessions of marital therapy over a 20-week period. As couples satisfied the above criteria, they were randomly assigned to one of the therapists providing ExST-Marital. A large battery of instruments was administered in order to assess outcome. The results of this investigation indicated that while some couples improved, others deteriorated at termination (Friesen et al., 1995; Grigg, 1994). 66 Therapists The therapists participating in TARP were required to have had: (1) the equivalent of at least a Master's degree in Counselling Psychology; (2) at least 3 years of experience working with substance abusive clients; and (3) the completion of specialized training and supervision in the conduct of ExST with demonstrated competency in the provision of the therapy. The therapist in the present investigation satisfied the above required criteria. The therapist in the present study held a Master's Degree in Counselling Psychology and had 6 years of experience in the drug and alcohol counselling field at the time the therapy was conducted. In addition, the therapist obtained a score of 3.4 out of a range of 1.0 to 4.0 on the Therapist Competency Form (TCF) (Appendix B). This form screened therapist's participation in TARP. A score of 3 or above indicated competency in the provision of ExST. The role of participant-observer taken by the researcher is an important feature of this investigation. Therefore, elaboration of this role ensues given its relevancy. Participant-Observer When the aim of the investigation is to explore intricate relationships and interconnections within a complex process and naturally occurring environment, the role of the researcher as a participant observer is common and appropriate in the naturalistic paradigm. A participant-observer is in a unique position to extrapolate the nuances, understand the complexity and richness of the process, and gain the depth of understanding as events unfold within the process of psychotherapy. Besides, as suggested by Kazdin (1980), it is difficult for an external observer to 67 gain the depth of understanding of intricate and complex processes. As a participant-observer, the researcher brings with him/her a wide range of experiences and awareness that maximize rather than minimize the depth of understanding required to answer the research questions posed by this investigation. Congruent with the naturalistic paradigm (Guba & Lincoln, 1994), this study was conducted from the perspective of participant-observer. As such, the assumptions and biases of the researcher as participant-observer on the topic of investigation are presented. Researcher's Assumptions on the Process of Change As indicated above, I was one of the therapists involved in TARP. Couples seek therapy for a variety of reasons. However, couples are usually deeply distressed and conflicted by the time they seek psychotherapy. I believe conflicts may be attributed to experiences of pain, betrayal, deeply rooted disappointments, and unresolved interactional patterns. Unresolved conflictual patterns permeate the relationship over time until a structure or a patterned way of being with each other is established. These conflictual patterns may take many forms; however, I believe that problems arise when a particular way of being with one another becomes rigid, static, unbendable and inflexible. Thus, I believe the troubled couples in this study related with one another in a predictable, rigid and inflexible style which encouraged disappointment, distance and grief. In attempting to help the couples resolve conflictual patterns, I believe the therapist must be fully present in coaching the couple and collaborating in the achievement of their desires and expressed therapeutic goals. As the therapeutic 68 goals and mandate are collaboratively established and agreed upon by all members of the therapeutic system, the actual working through of the conflict is initiated. As with any form of therapy, I believe the therapist must facilitate and establish a relationship based on trust and genuineness with all involved in the system and must generate an atmosphere of support and collaboration. I believe that actual change occurs when the couple works through and transforms those conflictual patterns into more satisfying and harmoniuosly interactive ones. I concur with the ExST notion that change in the couple occurs when rigid, recursive and unsatisfying interactive patterns are transformed into flexible patterns of interaction. For me, change occurs when the therapist accesses the depth of the couples's actual experiences, perceptions and sensations as they interact with one another. I believe that change is initiated the moment when one or both members of the dyad deeply experience and feel whatever experience or feeling is relevant and/or important for the self in relationship to the other as it is occurring in the present moment. The next section explains the theoretical underpinnings of Experiential Systemic Therapy. The basic conceptual premises along with its principles and phases of therapy are outlined. Experiential Systemic Therapy Experiential Systemic Therapy is an interpersonal process in which the therapeutic system members create a new story within a systemic, experiential and symbolic framework under a set of specified principles. As an integrative form of psychotherapy (Friesen et al., 1991), ExST embraces a variety of therapeutic 6 9 techniques and a wide scope of human experience (i.e., behavioral, emotional, cognitive, and visceral). ExST is systemic in its foundation and orientation and experiential in its process. In addition, ExST conceptualizes therapy as embedded within a deep symbolic framework. Systemic Dimension The systemic foundation emerged from three sources; namely, family therapy theory, general systems theory and second-order cybernetics (Friesen et al., 1991). As such, ExST assumes that living involves a complex, multidimensional, and dynamically interactive process where systems actively interact and reciprocally affect one another within a context. In addition, ExST maintains that, as persons interact, their structures become ontogenically and structurally coupled (Maturana, 1978) with others within a consensual domain and, as a consequence, the consensual domain is also ontogenically generated. This assumes that relational experiences and patterns of interaction between the experiencer and the experienced are of fundamental importance for therapy. In fact, relationships between the experiencer and the experienced (the other) include all levels or aspects of existence and the world. ExST assumes a reciprocal interdependence in which the two (the experiencer and the experienced) mutually influence one another and fit together in order to maintain the integrity of the relationship. Although some relationships may be experienced as more important and/or significant than others, the relational matrix operates and is experienced at the intrapersonal, interpersonal and contextual level. As well, the complexity of relationships can be extended to include inanimate objects, abstract 7 0 concepts and ideas. Within this systemic foundation, ExST process and nature is experiential and symbolic. Symbolic Dimension The symbolic dimension refers to the idea that therapy is a ritual of transformation of the living stories that clients bring into the session. As a therapeutic story of transformation, therapy is framed as a culturally sanctioned ritual in which all participants bring into being, experientially and symbolically, their emotions, thoughts, behaviors, issues and meaningful aspects of self. Actions and interactions are considered symbolic representations and enactments of clients' daily lives. As well, actual symbols and metaphors are considered powerful ways to access and intensify parts of self, interpersonal relationships and presenting problems. Experiential Dimension The experiential nature of this therapy model is also an important dimension of the process of change. ExST assumes that clients learn, become aware of and change by deeply experiencing thoughts, emotions, perceptions and sensations in the here and now. It assumes that profound change occurs by experiencing rather than by mere dialogue, instruction or, by just talking about \"it.\" As a core concept in ExST, therapeutic experiencing is defined as the enhancement, elevating and intensifying of thoughts, feelings, perceptions, behaviors, patterns and interactions so as to reach new and unprecedent depths. Therapeutic experiencing increases awareness of the recursiveness of thoughts and patterns, adds new experiences, and integrates those new experiences in 71 creative ways. Thus, through therapeutic experiencing, rigidity of intrapersonal and interpersonal patterns of interaction is transformed into new patterns with a new quality. This results into new ways of doing things; new ways of experiencing and viewing the world as a deeper and unexplored potential are activated. Second-order change (Dowd & Pace, 1989) is generated as a result of activating the unknown and the unexperienced. Guiding Principles of ExST Experiential Systemic Therapy is a brief form of psychotherapy. The story of transformation, or length of therapy, may range from four to 20 sessions. Five principles are espoused as essential to its practice. First, ExST maintains a developmental framework in which clients' experiences and problems are viewed in light of the themes of the life cycle. Second, an ecological assessment is developed in order to understand the client's difficulties. This assessment includes all levels of the system in which clients are engaged, such as individual, couple, family and community context. Third, ExST adopts an active present tense therapeutic focus without negating the current influence of events from the past. A fourth principle is the collaborative stance taken by the therapist in order to provide the opportunity for client's experiencing and transformation. The therapist is considered to be a collaborator rather than an expert in the re-writing of the story of transformation. Finally, as collaboration is maintained, the therapist and client become spontaneous and creative in their interactions. Spontaneity facilitates freedom to experiment with alternative and new ways of being in the world. Guided by these 72 five principles, the therapist and the clients engage in various transactions in order to address the relational rigidity and stagnation through four therapeutic phases. Phases of Therapy The life span of the therapeutic system is viewed as a story or a play consisting of a beginning, a middle, and an end. A play starts with an introduction where the characters and the setting are introduced and the principal themes outlined. After this introduction, the stage is set for the drama to be acted and enacted until a climax is reached. This is rapidly followed by a resolution to the drama or plot in which a new narrative path is weaved, created and integrated. The end is reached as the characters are transformed by their participation in the play. Accordingly, these parts of the play or the narrative are also reflected in the therapeutic story through four phases of therapy. These four phases of therapy include: 1) Forming the therapeutic system: Establishing a context for change; 2) Perturbing patterns and sequences and expanding alternatives; 3) Integrating experiences of change: Reorientation; and 4) Disbanding the therapeutic system: Termination and acknowledgment of accomplishments. Figure 1 presents the phases of therapy as they occur within the therapeutic story. Figure 1 illustrates the four phases of therapy as they may occur over the course of therapy. These phases may overlap and aspects of each phase may be present at any given point within the overall process. 73 Phase 1 Forming Phase 2 Perturbing Phase 3 Integrating Phase 4 Disbanding Pre-therapy > > > Post-therapy Figure 1 . An overview of the therapeutic system process Selection of Cases Three couples who took part in TARP and received marital therapy from this researcher were selected for the study. The success-failure strategy (Pinsof, 1988) was utilized in determining the chosen couples. Defined as discovery-oriented, the success-failure strategy consists of rank-ordering couples on outcome measures at significant evaluation points such as midtherapy, termination and/or follow-up. Three cases, representing three distinct patterns of change at post-test, were chosen. Based on the combined pre-post-test difference couples' scores on the Dyadic Adjustment Scale (DAS) (Spanier, 1986), the couples' patterns of change were identified as high, medium and minimal improver. In addition, the pre-post-test difference scores of the three selected cases on the Alcohol Dependency Data (ADD) (Raistrick, Dunbar, & Davidson, 1983), a measure of of alcohol dependency, and the Symptom Checklist-90-Revised (SCL-90-R) (Derogatis, 1983), a measure 74 of intrapersonal distress, were examined in order to evaluate consistency of the change patterns identified through the DAS scores. In the following section, a description of the instruments utilized for the selection of cases is offered. Then, the decision criteria for selection as well as the results of the selection process is presented. Description of Instruments The three instruments utilized for couple's selection were drawn from a large battery of instruments utilized by TARP. The couples completed the battery of instruments at pre-treatment, mid-treatment, post-treatment, and follow-up. The instruments selected for the present investigation are presumed to be sensitive to change at the three levels of client functioning: alcohol dependency, intrapersonal distress, and marital adjustment (Grigg, 1994). The alcohol dependency data questionnaire. In the TARP investigation, the Alcohol Dependency Data questionnaire (ADD) (Raistrick, Dunbar, & Davidson, 1983) was used as an indicator of change in the severity of alcohol dependency. This instrument is sensitive to changes in dependency levels over time. Measures on the ADD were taken at pretest, posttest and follow-up. This instrument was designed to measure the full range of alcohol dependency including mild, moderate, and severe dependency. The ADD consists of 39 items which are rated on a 4-point Likert scale ranging from never = 0 to nearly always = 3. The maximum dependency score is 117. A score of 0 suggests no dependency, scores from 1-30 suggest mild dependency. Moderate dependency 7 5 is indicated by scores ranging from 31 to 60. Scores ranging from 61 to 117 indicate severe dependency. A 15-item shortened version was developed and compared to the 39-item version (Davidson & Raistrick, 1986). The correlation between the two versions was reported as highly significant (r = .92). Also, the split-half reliability estimates on the shortened form was high (r = .87). Investigations of the validity of the shortened form of the questionnaire (Davidson & Raistrick, 1986; Davidson, Bunting, & Raistrick, 1990) have demonstrated that dependency as measured by the instrument is a single and unidimensional phenomenon and it does measure what it purports to measure. The TARP investigation employed the 39-item version of the questionnaire. The symptom checklist-90-revised. The Symptom Checklist-90-Revised (SCL-90-R) (Derogatis, 1983) is a 90-item self-report inventory designed to measure psychological symptom patterns of disturbed clients. This instrument utilizes a 5-point Likert type scale ranging from not at all= 0 to extremely= 4. The SCL-90-R assesses nine primary symptom dimensions and three global indices of distress. The nine symptom dimensions are: (1) Somatization, (2) Obsessive-Compulsive, (3) Interpersonal Sensitivity, (4) Depression, (5) Anxiety, (6) Hostility, (7) Phobic Anxiety, (8) Paranoid Ideation, and (9) Psychoticism. The three indices of distress include: (1) Global Severity Index (GSI), (2) Positive Symptom Distress Index (PSDI), and (3) Positive Symptom Total (PST). The global indices of distress yield an overall assessment of the client's psychological dysfunction (Derogatis, 1983). 76 Measures on the SCL-90-R were taken at pretest, posttest and follow-up. Derogatis (1983) indicated that the instrument is well suited for pre and post treatment evaluations because biases after repeated administrations have been undetected. In fact, the SCL-90-R has been successfully utilized as a clinical assessment tool and treatment outcome measure in areas such as depression, sexual dysfunction, stress, substance abuse and alcoholic family investigations. The SCL-90-R's concurrent validity has been determined by contrasting the instrument scores with several scales of the MMPI (Derogatis, Rickels, & Rock, 1977) and other instruments. The reliability measures of the SCL-90-R for the 9 symptom dimensions are adequate. The internal consistency alpha levels range from .77 to .90. The test-retest correlations range from .78 to .90. The dyadic adjustment scale. The Dyadic Adjustment Scale (Spanier, 1976) is a widely used self-report inventory for measuring marital satisfaction. This is a 32-item instrument that assess four dimensions of the marital relationship. The sub-scales of the DAS include: (1) Dyadic Consensus (the degree to which the couple agrees on matters important to the relationship, 13 items), (2) Dyadic Satisfaction (degree to which the couple is satisfied with the current state of the relationship and is committed to its continuation, 10 items), (3) Affectional Expression (degree to which the couple is satisfied with the expression of affection and sexual intimacy in the marriage, 4 items), and (4) Dyadic Cohesion (degree to which the couple experiences a sense of togetherness, 7 items). In conjunction with the generation of these four sub-scale scores, the DAS also generates a total score which represents the overall marital satisfaction of the couple. The total obtainable score on the instrument ranges from 0 to 151. A total score of below 100 suggests marital dissatisfaction. Spanier (1976) reported that the instrument has an internal consistency coefficient of .96. The reliability of the sub-scales were reported as .73 for the Affectionate Expression sub-scale, .86 for the Cohesion sub-scale, .90 for the Consensus sub-scale, and .94 for the Satisfaction sub-scale. Initially, the DAS was evaluated by three judges in order to uncover the relevancy of items to dyadic relationships, the wording of the items and the extent of consistency of items with definitions of adjustment (Spanier, 1976). The weekly situation diary. The Weekly Situation Diary (see Appendix C) is a questionnaire where TARP couples rated important changes occurring during the week. The partners rated their experience of change in the areas of self, marriage, family, friendships and work on a weekly basis throughout the duration of therapy. The present study inspected the couples' Weekly Situation Diaries during the identification and selection of the change episodes in order to detect possible change occurring outside of the therapeutic environment and cross check therapeutic events. Decision Criteria for Case Selection Three cases (couples) identified as high, medium and minimal improver at post-test on the DAS were, selected. The first criterion was to select couples with whom the researcher had worked with as a therapist. This researcher delivered couples therapy to seven couples from the ExST-Marital group (N=24). The 78 second criterion was to select couples who produced diverse results at termination; that is, couples whose combined pre-post-test difference scores on the DAS showed high, medium and minimal improvement. The rationale for the selection of these three levels was to explore the presence or absence of change episodes throughout the continium of therapeutic change. Based on these criteria, three cases representing a high, medium and minimal level of improvement were selected. The high improver case, couple 2010, obtained pre-post-test difference scores on the DAS of 34 points. The medium improver case, couple 2036, revealed mild improvement, in that their combined pre-post-test difference scores was 10.8 points. Couple 2080 was deemed as minimally improved because they displayed negligible improvement at post-test. Couple 2080's combined pre-post-test difference scores was 5.15 points. Although the pre-post-test difference score on the DAS was the primary source employed in selecting the couples, data from two other standarized measures, the SCL-90-R, a measure of intrapersonal distress, and the ADDS, a measure of alcohol dependence, were examined in order to evaluate consistency in the patterns identified through the DAS scores. Table 1 shows couple 2010's obtained scores on the DAS, SCL-90-R and ADDS. As displayed on the Table, couple 2010 improved on all measures. Couple 2036, the medium improver (see Table 2), showed improvement on the intrapersonal measure. However, the husband's pre-test score on the ADDS was missing from the data-bank, therefore, 7 9 Table 1 Obtained Scores on DAS. SCL-90-R. and ADDS for Couple 2010 DAS SCL-90-R ADDS H W H W H Pre-test 66.0 65.0 64.0 80.5 27.0 Post-test 95.0 104.0 56.8 60.8 10.0 Difference Score 29.0 39.0 -7.2 -19.7 -17.0 evaluation is groundless. Nonetheless, mild improvement was evidenced in the couple's combined pre-post-test difference scores on the DAS. Couple 2080, the minimal improver case (see Table 3), displayed some intrapersonal improvement. The husband also showed improvement on his alcohol dependence post-test score. However, the couple's measure of marital satisfaction displayed a negligible level of improvement. Table 2 Obtained Scores on DAS. SCL-90-R. and ADDS for Couple 2036 DAS SCL-90-R ADDS H W H W H Pre-test 62.0 99.0 94.8 83.4 Post-test 85.0 97.6 64.4 53.3 56.0 Difference Score 23.0 -1.4 -30.4 -30.1 80 Table 3 Obtained Scores on DAS. SCL-90-R. and ADDS for Couple 2080 DAS SCL-90-R ADDS H W H W H Pre-test 89.6 53.0 84.8 79.4 56.0 Post-test 88.0 64.9 53.6 62.9 33.0 Difference Score -1.6 11.9 -31.2 -16.5 -23.0 Demographic Description of Couples Couple 2010: high improver. At the time of therapy, the couple was married and had lived together for 10 years. The wife was 36 years old and the husband was 35 years old. The wife was previously married and subsequently divorced. This was the first marriage for the husband. The couple had four children. Two children ages 7 and 9 were from their current marriage. The other two children were from her first marriage. Their ages were 14 and 16. Both husband and wife worked full-time. He worked as a machinist and she worked in the service industry. Their total family income was in the range of $40,000 to $49,000 per year. Both husband and wife were of white Anglo-Saxon origin. They had similar educational levels. The wife finished grade 12 and the husband graduated from grade 12 with some technical training as a machinist. The couple entered therapy on December, 1991 and concluded in April, 1992. They completed the 15 sessions of marital ExST with a total of 19 hours. 81 Couple 2036: mild improver. Couple 2036 was married and had lived together for the last 6 years at the time they entered therapy. They had two children, a boy and a girl, ages 4 and 2 respectively. The husband worked full-time as a labourer and the wife was a homemaker. Their family income was in the range of $ 30,000 to $ 39,000 per year. The husband had not graduated from high school. The wife had completed her Grade 12. Both were of white Anglo-Saxon origin. Couple 2036 entered therapy on April, 1992 and ended on December, 1992. Therapy sessions were interrupted due to the husband's work schedule. The couple attended a total of 11 out of 15 sessions. They were unable to complete the full treatment program; however, they completed all other protocols for inclusion in TARP. The total therapy time was 14 hours. Couple 2080: minimal improver. The couple was married and had lived together for 18 years at the time they entered therapy. They had never divorced or remarried and they had two children. They had two boys ages 9 and 10. The husband worked full-time and the wife worked part-time. Their combined family income was in the range of $40,000 to $49,000 per year. They had both graduated from high school. The husband had attained graduation from a local college. Both were of white Anglo-Saxon origin. Couple 2080 entered therapy on August, 1992 and ended on January, 1993. The couple had a total of 14 sessions of marital therapy. The wife requested three individual sessions throughout the therapy process due to difficulties in expressing herself in the presence of her husband. The total therapy time was 19 hours. 8 2 Identification and Selection of Change Episodes Given the discovery-oriented nature of this investigation, the first step was to observe in detail the therapy process of the three couples in order to uncover the presence of episodes of change. This observation comprised the following steps: 1) The 40 video-taped therapy sessions were observed by the researcher so as to acquaint herself with the data and the process of therapy. 2) The observation process consisted of the consecutive and systematic observation of couple 2010 (high improver), couple 2036 (medium improver), and couple 2080 (minimal improver). 3) With this general understanding of the overall psychotherapeutic process and events, the researcher viewed the 15 sessions of couple 2010 once more. This free-flowing observation was intended to identify preliminary and tentative episodes of change. These episodes were broadly identified based on the actions and interactions of the participants around a conflict and/or issue they were attempting to resolve. 4) Next, all the sessions were observed once more. This third viewing was sharper and more structured. The researcher took notes on the process and the content of all the therapy sessions. This observation assisted in developing a clearer focus on what emerged as definable chunks, or more specifiable and concrete events, in which the three participants were actively engaged in solving specific and concrete conflicts and/or impasses. Guided by the experiential knowledge acquired through these observations and the operational definition of a change episode articulated by Greenberg (1986), 83 criteria for the identification of change episodes were generated. These criteria were developed by the researcher and refined in consultation with the study's advisor. Criteria for Identification of Change Episodes The following criteria specify the parameters for the identification and selection of the episodes of change. 1) The presence of a relational marker involving clear, identifiable and distinct evidence of interpersonal conflict between the spouses. This problematic/conflictual relational marker was verbally identified by one and/or both members of the couple as needing resolution. The verbal statements conveyed messages which indicated misunderstanding, tension, confusion, detachment, alienation, distance, defensiveness, stuckness, and/or strong emotions. 2) Given this distinctive marker, an ongoing dialogue comprising actions and interactions engaged in by the members of the therapeutic system ensued. These actions and interactions were intented to address the conflictual emotional/relational marker. There was no deviation into other problematic relational issues; rather, the participants maintained continuity with the particular conflictual marker until resolution incurred. 3) A new in-session pattern of relating/being emerged from the interaction which clearly indicated a resolution to the conflictual marker. This new way of being and/or relating indicated a new understanding, new discovery, greater interpersonal affiliation and intimacy, and personal autonomy. 8 4 Regarding the initial conflictual marker, the couple interacted in new and more satisfactory ways with each other. Selection of Change Episodes Once the criteria were developed, the selection of the change episodes ensued. The selection consisted of the following steps: 1) All 40 therapy sessions were, again, systematically observed in detail. 2) Initial and tentative episodes of change were selected based on the selection criteria. 3) The Weekly Situation Diaries for both spouses were inspected regarding relevant therapeutic information about the daily lives of the couples. In this questionnaire, the couple described important events taking place during the week (a summary of the Weekly Situation Diaries for the three couples is available from the author upon request). 4) Information from the Weekly Situation Diaries was evaluated and contrasted with the tentative change episodes selected in order to cross-check for differences and similarities between therapeutic experiences and daily life events. 5 ) An initial selection of possible episodes of change was compiled in videotape form. 6) These tentative episodes of change were reviewed and evaluated by the researcher and the study's advisor so as to ascertain their correspondence with the selection criteria, as well as their legitimacy and credibility. 85 7) Agreement on the selected episodes was sought and obtained from the study's advisor, who is also the principal developer of ExST. The identification and selection process generated a total of ten completed episodes. Eight change episodes were selected from couple 2010 (high improver), two episodes from couple 2036 (mild improver), and no episodes from couple 2080 (minimal improver). These episodes were transcribed in order to proceed with the analysis. Grounded theory was deemed the most appropriate method of analysis to answer the questions posed by the study. In the following section, the grounded theory method is introduced. This introduction is followed by a description of the data analysis and procedures. Overview of Grounded Theory The general purpose of a grounded theory investigation is to generate theory from the data (Strauss & Corbin, 1990). Data collection and analysis are decided by theoretical sampling (Glaser, 1978; Strauss & Corbin, 1990), that is, by the concepts generated from the analysis. Data collection and analysis are determined and initiated with the first identified protocol (Glaser, 1978; Rennie, Phillips & Quartaro, 1988; Strauss & Corbin, 1990). The emphasis of the grounded theory method is on the discovery and generation of theory grounded in the data. A basic assumption is that, by grounding the emerging theory in the data, the discovered theory will be an effective and valid representation of events as they occurred at the time of the investigation (Glaser & Strauss, 1967). Strauss and Corbin (1990) noted that \"a grounded theory is one 8 6 that is inductively derived from the study of the phenomenon it represents. That is, it is discovered, developed, and provisionally verified through systematic data collection and analysis of data pertaining to that phenomenon.\" (p.23). Hence, the essence of the grounded theory approach is on the methodical and deliberate generation of theory which conceptually accounts for the unfolding of the events under study. The role of a theory is to interpret, explain and predict phenomena as they occur in the empirical world in a coherent and logical fashion (Chenitz & Swanson, 1986). As a symbolic construction of reality (Kaplan, 1964), a theory provides a systematic explanation of an event through the generation of concepts. These concepts are linked via suggested hypotheses about the relationships amongst the generated concepts. The discovery of concepts and hypotheses relevant to the phenomenon under investigation is paramount to the generation of theory. Accordingly, Glaser and Strauss (1967) asserted that the components of a grounded theory investigation are conceptual categories along with their conceptual properties as well as hypotheses that describe relationships amongst the categories and their properties. Conceptual categories and their properties are generated from data which is also used as evidence to illustrate the emergence of the concept and properties. While a category is a conceptual element of the theory; a property is a conceptual element of the category. Hypotheses are generalized suggested relations amongst categories and their properties which throughout the course of the investigation are 8 7 verified by iteratively collecting and analyzing data until saturated (Glaser & Strauss, 1967). Historical and Theoretical Background The grounded theory method was originally developed by sociologists Anselm Strauss and Barney Glaser in the early 1960s as they collaborated in a field observational study. Their shared intellectual efforts resulted in the publication of the book The Discovery of Grounded Theory in 1967. The method was generated under an intellectual climate which favoured testing of hypotheses and theoretical verification (Charmaz, 1983; Turner, 1981). Indeed, the development of the grounded theory approach was a response to a position taken at the time that qualitative research served only as an initial, preliminary and exploratory step to subsequent investigations which could later verify findings through rigorous hypotheses testing. The assumption, at the time, was that only verification via hypothesis testing could produce rigorous and scientific findings. Furthermore, Glaser and Strauss (1970) argued that there was an over-emphasis on rigorous testing of hypotheses and a de-emphasis on whether those concepts and hypotheses were relevant for the area being investigated. The grounded theory approach emerged as a method to generate concepts and theory through the utilization of a systematic set of procedures, a clearly defined process and a defined set of criteria to evaluate its theoretical product. Grounded theory evolved from the philosophical and theoretical underpinnings of Symbolic Interactionalism (Corbin & Strauss, 1990; Strauss, 1987; Strauss & 88 Corbin, 1990). Symbolic Interactionalism was initially advanced by social psychologist Herbert Mead (1934) as a theory on the development of human behavior. Mead (1934) postulated that it is through social interaction that the person achieves a sense of self and develops a mind. In addition, Mead (1934) further asserted that it is through the role taking process and in interaction with others that the person creates meaning. The meaning generated by interacting with others is what ultimately guides action and behavior and thus, brings about consequences. Blumer (1969) further developed the particulars of Symbolic Interactionalism. From Blumer's (1969) point of view, Symbolic Interactionalism assumes three central premises. These are that \"human beings act toward things on the basis of the meanings that the things have for them\" (Blumer, 1969, p.2); that the \"meaning of such a things is derived from, or arises out of the social interaction that one has with one's fellows\" (Blumer, 1969, p. 2); and that \"these meanings are handled in, and modified through, an interpretative process used by the person in dealing with the things he encounters.\" (Blumer, 1969, p. 2). Thus, meaning is created and recreated by interacting with others and with self. This conceptualization also assumes that behavior is a process in constant flux as interaction with self and others is dynamic and in continuous activity. Change and thereby process are two significant and essential elements built into Symbolic Interactionalism. The implications of the Symbolic Interactionalist perspective for the conduct of research are threefold. First, human behavior must be investigated within the context in which interaction with self and others actually occurs so as to understand 89 its total complexity, full range of variation and varying conditions. Second, in order to examine the totality of complex interactions and how meanings are created, the contextual setting needs to be analyzed and accounted for; otherwise, the understanding of meanings will be severely limited. Third, behavior needs to be understood from the point of view of the participants involved in the study. That is, the investigator is required to \"take the role of the other\" in order to understand their perceptions and learn about their world and interpretations of self in interaction (Chenitz & Swanson, 1986). This investigation fullfils these three basic tenets of the Symbolic Interactionalist perspective. From the Symbolic Interactionalist perspective, one studies behaviors through exploration and inspection (Blumer, 1969). Exploration is defined as the process of clarification of the investigation so that later interpretations and analysis are grounded in the data. Inspection, according to Blumer (1969) refers to the actual analysis of the empirical content of the study. Both exploration and inspection are utilized to examine behaviors in the naturalistic setting. Furthermore, the researcher examines behaviors at the interactional or behavioral and symbolic levels (Chenitz & Swanson, 1986). It is assumed that observation of both the behavior in the specific situation in which it occurs, at the verbal and nonverbal level, and the symbolic meaning attached to those actions by the participants are important for the analysis. In addition, there is a manifest emphasis on action and the problem situation in need of a resolution. Consequently, the methodological thrust of this perspective conceives the method in the context of problem solving situations 90 (Strauss, 1987). Finally, a relevant aspect of this perspective is that events rather than variables are believed to be the appropriate units of analysis (Strauss, 1993). An important consideration derived from the basic premises of Symbolic Interactionalism concerns the nature of the phenomenon understudy. Basically, it is presumed that the phenomenon under study is continually changing as a result of evolving conditions (Corbin & Strauss, 1990). Thus, the idea of change is embedded into the conceptualization of the grounded theory method. In fact, change, through process, is intricately linked into the analysis not only by discovering the evolving conditions of the situation but also by discovering how participants respond to those changing conditions and the consequences of their actions. This general background on the theoretical framework and historical development of grounded theory sets the context for the following section. The next section provides a specific description of the method of analysis along with its accompanying procedures and relevant concepts. Description of the Method of Analysis As a method intended for the discovery and analysis of basic social processes within the context in which they occur, the primary purpose of grounded theory is to develop conceptual theory grounded in the data. Strauss (1987) stated that the distinctive features of grounded theory are its theoretical sampling procedure, and the methodological guidelines of the comparative method. These features are intended to generate concepts that are grounded and dense. Common to all grounded theory research are basic strategies which must be followed in 91 order to generate a conceptual model. First, concepts and interrelationships amongst concepts are generated through the method of constant comparison between or within groups. Second, theory is discovered, developed, and verified through a recursive and concurrent style of data collection and analysis with the purpose of concept saturation and integration of the emerging conceptual model. Third, concepts are refined, elaborated and saturated via theoretical sampling whereby the researcher purposely samples for the emerging concepts and conceptual integration. Fourth, the systematic application of analytical methods such as the constant comparative method, theoretical sampling and simultaneous data collection and analysis gradually lead to the generation of progressively more abstract levels of analysis and integration of concepts. The Analytical Process The logic of the research process within grounded theory is conceptualized as a nonlinear activity in which several operating activities are iteratively at play (Glaser, 1978; Stern, 1980; Strauss, 1987; Turner, 1981). As the researcher collects the first pieces of data; he/she immediately initiates examination and analysis of the data by coding, categorizing, conceptualizing, and the writing down of ideas which may be subsequently utilized in the final write up of the conceptual model. Glaser (1978) succinctly outlined the essential steps of the research process by stating that: \"The steps, as now formulated, are collection of research data, open coding of the data soon after, theoretical sampling, generating many memos with as much saturation as possible and emergence of core social psychological 92 problems and processes, which then become the basis for more selective theoretical sampling, coding and memoing as the analyst focuses on the core. This goes on all at once,...\" (p. 16). Quartaro (1986), on the other hand, identified five aspects of the research process. These aspects are: (1) collection of data; (2) comparison of data; (3) integration of categories; (4) delimitations of the emerging theory; and (5) presentation of the theory. The first four steps basically constitute the constant comparative method which comprises the essence of the grounded theory method. Hutchinson (1986) succinctly delineated the analytical process by stating that: While coding and analyzing the data, the researcher looks for patterns. He or she compares incident with incident, incident with category, and, finally category with category or construct with construct. By this method, the analyst distinguishes similarities and differences of incidents. By comparing similar incidents, the basic properties of a category or construct are defined. Certain differences between incidents establish boundaries; relationships among categories are gradually clarified. Comparative analysis forces the researcher to expand or \"tease out\" the emerging category or construct by searching for its structure, ...cause, context, dimensions, consequences, and its relationship to other categories, (p. 122). Like other qualitative methods, the data for a grounded theory investigation may come from a variety of sources. Data may be collected from interviews, observations, transcripts, audiotapes, videotapes, and documents (newspapers, letters, books) (Charmaz, 1990; Strauss, 1987) or a combination of sources (Stern, 1980). Any of these sources of data can be coded in the same manner as observations and interviews (Glaser & Strauss, 1967). The data for the present analysis was collected from transcribed videotaped marital therapy sessions. The 93 ten episodes of change were entirely transcribed (complete transcripts are available from the author). During the first stage of the analysis, I collected data through open sampling so as to discover as many potentially relevant categories as possible while also focusing on the central and crucial aspects of the topic (Rennie, Phillips, & Quartaro, 1988), that is, the questions of the study. Through this process, I was open to instances offering the greatest scope in gathering the most relevant data. Strauss (1987) noted that during the initial process of data collection, the researcher may become overwhelmed by the richness and confusing nature of the data. However, as the analysis is usually initiated with the first or second fieldwork observation, the confusion is temporary as further observations will be informed by analytical questions and/or hypotheses. As the analysis progresses, instances or events are gathered on the basis of the emergent concepts. Instances, events or subjects are selected in terms of how clearly they represent the phenomenon under study (Rennie, Phillips, & Quartaro, 1988). The analysis was initiated with the first selected episode by examining statement by statement the verbal and nonverbal messages of the participants interacting with each other. This initial analytical step consisted of concurrently analyzing and questioning the events through three mediums of information. These media were listening to the audio-tape of the episode in order to capture nuances in tone of voice without body language, listening and watching to actions and interactions in video-tape form, and the reading and close examinination of the transcripts. The salient actions and interactions were analyzed and named 94 accordingly. The statements were examined and compared by maximizing and minimizing the differences and similarities present in the actions and interactions of the participants. As well, questions such as what was this relevant to, and what was actually going on, were constantly asked as the analysis continued. This breaking down and opening up of the data generated a total of 185 pages of typed material including the memos for the first episode. Overwhelmed with the substantial amount of information generated, and confused as to where to go next, I decided to analyse the next three episodes following the same format, that of opening up the data and generating provisional names for significant and salient actions and interaction of the participants. The opening up of the data generated a total of 338 pages of typed materials for the first four episodes. The high density and richness of the generated material, as well as the tediousness of the task, forced me to move onto the next step in the analysis; that is, to begin to cluster provisional names according to their obvious fit. Initial concepts were clustered under another, broader provisional concept. As the clustering of provisional concepts for the first four episodes continued, a general pattern and internal flow to the episodes began to emerge. However, because of the extensive number of concepts and the density of the data, a conceptual story of each episode was developed. The memos generated during the analysis became very useful during this process. Creating a conceptual story for each episode greatly clarified and assisted in linking concepts regarding the unfolding of events. 95 The conceptual story for each of the four episodes also assisted in the analysis of the following six episodes. Namely, the analysis of the six episodes was based on both the concepts already generated from the previous analysis, and the emergence of new concepts. The analysis consisted of comparing and contrasting concepts as well as sampling for repeating concepts. During the analysis of the seventh episode, concepts and themes began to repeat themselves; however, all episodes were fully analyzed. The repetition of concepts forced me, once more, to integrate further the concepts and categories. Conceptual Integration Through continued close examination of the emerging concepts, some became more central or salient in relation to others. Then, I started to integrate the concepts by focusing on questions such as: how does it all fit and what brings it all together? Similar to the clustering of concepts from the first analysis, here I collapsed concepts while searching for an emergent category that would explain and hold concepts together. Once I was able to identify the central category which seemed to hold the story of the episodes together, I then integrated all other concepts around the one main category by attempting to make theoretical sense and linking concepts as they fit together. Criteria for Evaluating a Grounded Theory Investigation Traditionally, there has been a tendency for researchers to evaluate their practices with criteria derived from the positivistic paradigm. From this perspective, 96 concepts such as objectivity, validity, generalizability and reliability are the foundation for evaluation. However, the manner in which qualitative researchers evaluate their procedures, practices and research findings does not adhere to these criteria as they are viewed as being ontologically inappropriate (Eisner & Peshkin, 1990; Krefting, 1990; Guba, 1990; Grumet, 1990; Wolcott, 1990). Consequently, some researchers (Eisenhart & Howe, 1992; Wolcott, 1990) have developed ways to ensure the validity of qualitative findings. Guba (1981) proposed a model for evaluating both qualitative and quantitative research based on the concept of trustworthiness. Accordingly, trustworthiness in all research must be established according to criteria based on neutrality, truth value, applicability and consistency. Within these criteria, apropos qualitative investigations, Guba (1981) generated terms comparable to the quantitative tradition. These terms are confirmability (objectivity), credibility (internal validity), transferability (generalizability), and dependability (reliability). In general, theories are appraised according to the criteria of extensiveness of explanation, parsimony, internal consistency, usefulness, and empirical validity. Although important, these criteria are less pertinent in grounded theory than in theory generated through other methods (Morse & Field, 1995). Instead, in grounded theory the degree of fit of the emergent theory with the data is more relevant and appropriate (Glaser, 1978). In fact, Glaser and Strauss (1967) and Glaser (1978) developed specific criteria for theories generated through the grounded theory approach. These criteria are referred to as fit, relevance, work and modifiability. This study followed closely the suggested evaluative criteria as the 97 analysis was performed. The last chapter of this dissertation will evaluate the findings of the study through the criteria identified. In the following two chapters, the results of the investigation are presented. The presentation of the results is divided into two chapters given the density of the findings. The next chapter, chapter four, presents the results of the observational, qualitative analysis of the three couples. This includes a brief description of the process of therapy process of the three selected cases and an introduction and description of the selected episodes of change. Then, in the following chapter, the results of the grounded theory analysis are presented. A conceptual model of marital therapy change, along with illustrations depicting its categories and subcategories is thoroughly presented. 98 CHAPTER IV RESULTS: O V E R A L L THERAPY CONTEXT This chapter provides an overview of the 40 therapy sessions of the three marital ExST couples. The provision of this contextual background endeavors to situate the findings resulting from the qualitative analysis of the ten selected change episodes. Also, this overall therapy context is deemed necessary because it identifies the circumstances surrounding the change episodes and it locates the clinical background against which the ultimate model of change emerged. In addition, this chapter describes the selected change episodes. Brief Overall Context of the Therapy Process In order to answer the first research question regarding the identification of observable, identifiable and recurrent in-session change episodes, it was necessary to observe and systematically examine the process of therapy of the three selected couples. A complete description of the therapy sessions (280 pages) of the three couples is available from the researcher. In the following pages, a description of the therapy sessions is offered in the form of an outline. This psychotherapeutic outline consists of the following elements: (1) participants' identification of areas of conflict; (2) desired state; (3) meaning of desired state; (4) symbolic representation of desired state; and (5) salient psychotherapeutic events. In addition, the selected change episodes are identified within the outline. 99 Outline of the Psychotherapeutic Process for the Three Couples Couple 2010: High Improver Participants' identification of areas of conflict Wife: Tension over husband's alcohol misuse. Lack of trust and intimacy. Unable to share and communicate with each other. Unable to identify and share feelings. Husband: Distant, he no longer loves or likes his wife. Anger, resentments toward his wife. Conflicts over alcohol. Doesn't trust or express feelings to anyone. Therapist: Blocks to intimacy: pain, frustrations, anger, resentments, alcohol, hurt, mistrust. Desired state Intimacy and closeness. Meaning of desired state Wife: Tenderness through words. Husband: Tenderness through touch. Symbolic representation of desired state Wife: Three roses representing friendship, knowledge, and love and a teddy bear representing warmth. Husband: His wedding band representing closeness. Couple: A cushion of happiness. Salient psychotherapeutic events Session 1: 1. Externalization of current state of relationship. 2. Externalization of desired state. 3 . Removal of blocks to closeness. 4. Identification of symbol of happiness (cushion). 100 Session 2: 1. Exploration of symbols of closeness. 2. Exploration of patterns of interaction: Wife: push-pull Husband: withdraw-withdraw 3. Identification of wife's first husband and husband's anger as block to closeness. 4. Identification of alcoholism as block to closeness. 5. Identification of wife's childhood sexual abuse as block to closeness. Session 3: 1. Blocks (anger, betrayal, abuse,abandonment) are evoked. 2. Push-pull and withdraw-withdraw patterns enacted. 3. Significant session for both: experienced togetherness. Episode # 1 Working with distancing pattern. Session 4: 1. Limited dialogue: wife praised husband, husband unable to praise wife. 2. Wife admitted and shared experiences of sexual abuse and impact on marriage. 3. Conflict explored: husband blaming, wife retreating in fear and shame. Session 5: 1. Exploration of their wounds. 2. Generation of symbolic names for their ghosts: wife: the \"bastard\" husband: the \"bitch\" Episode # 2 Bringing ghosts out into the open. Session 6: 1. Exploration and working through: anger. 2. Husband angry with wife: sexual abuse. 3. Wife angry with husband: lack of compassion. 4. Husband disengaged from therapy process. Episode # 3 Session 7: Episode # 4 Session 8: Engaging and committing to involvement in therapy. 1. Evoking and enacting anger. Working through feeling/experiences of anger. 1. Evaluating therapy progress: coming closer, developing togetherness and connection. 101 Session 9: Session 10: Episode # 5 Session 11: Session 12: Episode # 6 Session 13: Episode # 7 Session 14: Episode # 8 Session 15: 2. Re-enacting their distance with blocks to intimacy externalized. 3. Blocks still present, no longer in between the couple, rather cushion of happiness in between them. 4. Exploration of husband's and wife's relationship and impact of alcohol in the marriage. 1. Working through couple's relationship to alcohol. 1. Working on husband's relationship to his ghost. 2. Externalizing alcohol, the \"bitch\" and working through connections in relationship to his wife. Releasing husband's ghost. 1. Exploring husband's relationships and perceptions of women. 2. Enacting husband's distance to his wife and alcohol. 3. Exploring their dance of push-pull in the marriage. 1. Couple closer to each other: blocks to intimacy loosing their power. 2. Couple interacting with each other. 3. Discovering ways to connect as therapy terminating. Releasing wife's ghost. 1. Openness with each other: secrets are out. 2. Experiencing intimacy. Actual removal of blocks to intimacy and closeness. 1. Sharing of symbols representing letting go. wife: photograph of first wedding husband: \"bitch\" 2. Exploration of meaning of letting go of symbols. 3. Searching for ways to celebrate termination of therapy. Burning ritual: letting go of symbols. 1. Exploring ways to stay emotionally connected through daily living. 2. Reflecting on achievements made in therapy. 3. Appreciating members of therapeutic system. 102 4. Finding concrete ways to link good and new experiences in their future lives together (cushion of happiness). Couple 2036: Mild Improver Participants' identification of areas of conflict Wife: Feeling forced to engage in unwanted sexual acts. Removed and distant from husband due to his abusive and manipulative behaviours. Lack of affection and compassion from husband. Husband: Childhood sexual abusive experiences and trauma affecting him and his marriage. Experiencing intense flashbacks about sexual abuse as a child. Issues related to the expression of his sexuality in the marriage. Therapist: Issues related to childhood trauma and sexual abuse. Conflicts around expression of intimacy, compassion, caring and respect within the marriage. Distancing and abusive behaviours expressed by partners. Desired state Wife: To open lines of communication. To discover respect, understanding and openness in the relationship. Husband: To find self-acceptance and acceptance of others. Meaning of desired state Unexplored during therapy Symbolic representation of desired state Wife: A wedding photograph. Husband: Unable to define. Salient psychotherapeutic events Session 1: 1. Exploration of problematic issue for the couple. 103 2. Exploration of desired state. Session 2: Episode # 9 Session 3: 1. Exploration of husband's childhood sexual abuse issues. 2. Exploration of wife's relationship to her husband: pushed away by him. 3. Discovering and accepting different ways of expressing pain. 4. Limited and restrictive couple interaction. Working through couples distance. 1. Exploration of drinking behaviour and its impact on the marriage. 2. Couple unable to maintain interaction/communication. 3. Exploring husband's lack of trust and lack of acceptance of his wife. Session 4: Session 5: 1. Distance revisited: discrepant interest and views on their lives. 2. Intense emoting by wife as husband continues hurtful distancing behaviors. 3. Working with husband's difficulties in trusting and accepting his wife. 4. Significant issue for husband: inability to trust anyone. 1. Time lapsed between session 4 and 5 was of 6 weeks. 2. Discovering ways to express affection in non-sexual ways. 3. Exploration of impact of sexual abuse: divided self. Session 6: 1. Working on wife's fear of husband's anger. 2. Exploring meaning of wife's symbol (wedding photograph). 3. Working with husband's push-pull relating pattern in marriage and in therapy sessions. Session 7: 1. Time lapsed between sessions: 4 weeks. 2. Working with couple's distance from each other: wife's need for affection and husband's need for sex. Session 8: 1. Working with husband's push-pull pattern with therapist. 2. Husband's beginning to notice his feelings now. 3. Husband discovering he has many issues to work on. 104 Episode #10 Session 9: Session 10: Session 11: Working with husband's childhood trauma. 1. Exploration of impact of alcohol related behaviors in the marriage. 2. Revisited their therapeutic goals; three more sessions available. 3. Couple discouraged due to limited number of sessions. 1. Unable to engage in interaction with each other. 2. Continued conflict around expressions of affection and lust. 3. Wife unable to express feelings to husband. 1. Couple's last session. 2. Expressed concern about many unfinished issues individually and within the relationship. Couple 2080: Minimal Improver Participants' identification of areas of conflict Wife: In therapy only to support husband's sobriety. Fully apart and distant from husband. Unwilling to engage with husband as she had made her decision to leave the marriage prior to coming for therapy. Marriage has ended for her. Complete lack of love or respect for her husband. Husband: His family and wife are more important now that he is sober. Wants his wife's love and respect back. Desperately holding onto wife. Wishes to have another chance and show her in therapy that he is a changed man. Therapist: Wife unwilling to engage/work on couple's issues. Marriage has more or less ended. Couple completely distant and disengaged from each other. Desired state Wife: Lacks hope/desire as marriage has ended for her. Husband: To keep his family intact. To show and convince his wife that he is a new man without alcohol. 105 To have his wife beside him. Meaning of desired state Unexplored in therapy. Symbolic representation of desired state Unexplored in therapy. Salient psychotherapeutic events Session 1: 1. Exploration of therapeutic goals. 2. Enactment of distance and blocks to coming together. 3. Emotionally the wife has already left the marriage. Session 2: 1. Wife reaffirming her decision: desire to leave the marriage. 2. Exploration of therapeutic goals: Wife: out of the marriage and husband: to gain love and respect from his wife. 3. Re-enacted conflictual dance: Husband pulling his wife and wife going away. 4. Wife in therapy out of threats and force coming from her husband. 5. Husband pleading with her to come back to him. Session 3: 1. Couple engaged in role reversal; facilitating understanding for husband about wife's decision to end marriage. 2. Couple unable to engage in dialogue around their conflict. 3. Wife reaffirming her unwillingness to come closer to her husband. 4. Husband pulling her in closer to him through threats. Session 4: 1. Exploring husband's relationship to alcohol. 2. Wife refusing to engage in process around husband's issues. 3. Therapist suggested individual sessions: wife agreed but husband reticent. Session 5: 1. Session with husband. 2. Determined to have wife back at any cost. 3. Wife unwilling to relate to him at all. 4. Encouraging focus on himself. 106 Session 6: Session 7: Session 8: Session 9: Session 10: 1. Session with wife 2. She no longer loves him and is unwilling to be with him. 3. Wants out of the marriage as soon as possible but waiting until he is more stable and accepting of her decision. She expresses fears of his retaliations. 1. Conflict continues: pushing wife to make plans for a future life together. 2. Husband unwilling to listen to wife's lack of commitment. 3. Couple delaying the final ending. 1. Husband desperately trying to win wife back. 2. Couple delaying decision to separate. 1. Individual session with husband. 2. Role reversal to facilitate understanding of wife's decision to end marriage. 1. Couple in wait and see mode. 2. Role reversal: husband understands somewhat wife's unwillingness to be with him. Session 11: 1. Individual session with wife. 2. Externalizing her conflict: needs of the children to have a father and her needs to love someone else. 3. Confirmed her decision to end the marriage. Session 12: 2. Wife in direct dialogue with her husband regarding her wish to separate from him. Husband retaliates by using the children as weapons and decides to end therapy. Session 13: Session 14: Session 15: 1. Individual session with husband, wife was ill. 2. Insisting on keeping his wife at all costs. 3. Requests a referral to a counsellor that would keep his wife in the marriage. 1. Couple distant and postponing decision to end the marriage until after Christmas. 1. Postponing her decision to end marriage due to fears of retaliation until after Christmas. 2. Wife requesting one more session after the New Year. 107 As displayed in this outline, the therapy process of the three couples showed important differences and some similarities. The similarities were evident in the couples' conflict dynamics. For instance, the three couples reported and expressed considerable distance, lack of intimacy, and conflict over expressions of intimacy and contact. Another similarity in the couples' therapeutic process concerned the identification of a desired state. However, although the three couples identified a desired state, only the highly improved couple (2010) generated a conjoint desired state. The desired goals of couple 2036 (mild improver) and couple 2080 (minimal improver) were separately perceived and individually generated by the partners. In fact, the wife of couple 2080 was unable to develop a desired state for her marriage. Rather, she clearly expressed from the beginning of therapy her desire to end the union. The noted differences were observed in the couples' therapy process and the couples' response and engagement in therapy. Concerning differences in the therapeutic process, the meaning of the desired state was unexplored with couples 2036 (mild improver) and 2080 (minimal improver), but clearly delineated by couple 2010 (high improver). Also, the generation of a symbolic representation of the desired state was explored and enacted only with couples 2010 and 2036. The couples' responsiveness and engagement in the therapy process revealed many differences. Couple 2010 was fully committed to therapy, attended all sessions and followed through with the entire therapy program. Couple 2036, on the other hand, missed many sessions and was usually 108 unprepared and late for appointments. In couple 2080, the partners were emotionally distant from each other and largely uninvolved in the therapy process. In fact, this couple was in the process of separating when they began therapy. In addition, although the three couples were emotionally distant and organizationally disconnected from each other at the beginning of therapy, the style of engagement and interaction during the therapy process was also distinct. Thus, the engagement style of the partners in couple 2010 was reflective and attentive to each others' interactions as the therapy progressed. In contrast, couple 2036's style of engagement was chaotic and disorganized. Although the partners worked together on a variety of issues, their interactions were usually confusing to each other and reflected a state of pseudo-intimacy. Couple 2080 was fully disengaged and the partners were entrenched in their distancing interactions and uncommitted to each other. This brief outline of the therapy process provides the context for the episodes of change. In the following section, the ten change episodes are presented. Introduction to the Change Episodes Based on the extensive and systematic observations of the therapy sessions of the three cases (couples), ten change episodes met the a priori generated selection criteria delineated in chapter three. Of these, eight episodes emerged from couple 2010 (high improver), two episodes from couple 2036 (mild improver) and no episodes from couple 2080 (minimal improver) (see Table 4). 109 Within the overall process of the therapy sessions, the episodes showed two distinctive characteristics: (1) recurrence of process as specific chunks within the process of therapy and (2) completeness. The reoccurrence of the ten episodes emerging from couples 2010 and 2036 was manifested in the repetition of process rather than content. That is, the process of the episode was characterized by (1) a specific conflictual-relational marker identified by the couple as in need of a resolution; (2) ongoing actions and interactions on the part of the therapist and the couple intended to address the identified conflict; and (3) a final resolution to the conflict. This process repeated itself throughout the ten episodes selected. The content of the issues addressed during these episodes varied, however. Another characteristic of the episodes selected concerned their completeness. Episode completion refers to achieving a resolution, or bringing to an end the entire piece of work once a conflictual-relational marker had been identified by the couple. Incomplete episodes were present in all three couples' psychotherapeutic process. In these episodes, no resolution to the initial conflictual marker was present. The process of couple 2010 (high improver) for instance, displayed only three incomplete episodes. The process of Couple 2036 (mild improver) on the other hand, revealed 14 incomplete episodes. Couple 2080's process displayed 42 incomplete episodes. 110 Table 4 Emergence of Change Episodes in Sessions Session # Couple 2010 Couple 2036 Couple 2080 1 2 Episode # 9 3 Episode # 1 4 5 Episode # 2 6 Episode # 3 7 Episode # 4 8 Episode # 10 9 10 Episode # 5 11 12 Episode # 6 13 Episode # 7 14 Episode # 8 15 Incomplete Episodes 3 14 42 Although all change episodes fit the a priori generated criteria for episode selection, differences in process were evident. The eight change episodes from couple 2010 (high improver), for instance, depicted a complete process of therapy with a beginning, a middle, and an end. As such, they represented what might be termed the stages of therapy, particularly as defined by ExST. The two change episodes from couple 2036 (mild improver) reflected a stage of initial therapeutic work. This couple's work centered on the husband's intrapersonal issues which were considered to be extremely relevant by both members of the couple. Couple 2080 (minimal improver) showed no episodes meeting the Ill episode selection criteria. The couple was determined not to continue the marriage. The partners' interests, goals, and motives in seeking therapy were disparate. The husband's motive was to bring his wife back to the marriage, while the wife's motive was to develop strength to leave the marriage. Despite these differences, the ten episodes of change revealed a coherent internal pattern with noticeable similarities and differences. In the following section, a description of the episodes of change is presented. This section describes the individual internal structure of each episode along with its time span and the conditions under which the episode evolved. Description of the Change Episodes Observation and analysis of the change episodes revealed diverse contextual conditions prior to their emergence. That is, within the therapy sessions, the episodes evolved under different circumstances. The context to the episode depended on the stage of therapy, particularly in terms of the couples' desired goals. As such, the context of some episodes reflected the initial stage of therapy while for others it reflected the middle or the ending of therapy. Also, the internal structure of the individual episodes, as reflected by the categories and subcategories, was slightly dissimilar. In addition, the time span and location of the episodes within the sessions was also different. As such, a brief description of the conditions prior to the beginning of the change episode along with the generated categories is presented below. 112 Episode 1 This episode was selected from the third session of couple 2010. The episode lasted 29 minutes, and it appeared 27 minutes into the session. The internal structure of this episode revealed four distinctive categories which occurred over time. These categories were named (1) couple distancing, (2) spouses acknowledging/owning their own part in the distancing, (3) tentative, restrictive, limited contact between spouses, and (4) engaging with each other through acceptance and support. Prior to the beginning of this episode, the couple explored their distancing pattern. Although the wife was actively involved in the interview, the husband was removed, distant, detached and non-expressive. He displayed little affect and used brief monosyllabic statements. The wife's anger, pain and frustration and the husband's loneliness and refuge in drinking were explored. Contrary to the wife's insistence that problems in the marriage were due to her husband's distancing actions, including his drinking, the husband denied any problems. This episode began under these conditions. Episode 2 This episode emerged in the fifth session of couple 2010. The episode appeared 12 minutes into the session and lasted 34 minutes. The internal structure of the episode revealed four categories. These categories were: (1) partners acknowledging and opening up to their inner pain, (2) partners opening up to self and to each other under conditions of constrictive and limited contact, 113 (3) working through issues and releasing the wife's ghost, and (4) partners intimately engaging with each other. The events prior to the initiation of this episode revealed tension and distancing behaviors by the couple. During the previous session, the wife had shared traumatic experiences of childhood sexual abuse. She had tentatively recognized the possible impact of those experiences on herself and her marriage. The husband had also acknowledged these experiences as the most significant block to generating intimacy and closeness in their marriage. However, in comparison to his wife, the husband's assumption was that his own painful experiences had not impinged as negatively on the marriage as hers. Furthermore, the husband blamed his wife for the problems in the marriage by reflecting on the many obstacles she had brought into their relationship. Given that the husband had not shared painful or difficult issues related to himself, the therapist invited him to do so. As a result, the husband reluctantly disclosed the existence of early childhood emotional abuse, neglect and loneliness. He also shared his experiences of living with some wounds which had been inflicted in his childhood. However, he quickly pointed out that the wounds had been put away. His assumption was that pain was to be forgotten and put away. The episode started at a point in the session in which the husband was distant, subtly angry and frustrated as he minimized and denied significant emotional pain, while the wife displayed fear of her husband's feelings and distanced herself from the therapy process. 114 Episode 3 This episode emerged during session number six which occurred near the middle phase of therapy of couple # 2010. The time duration of the episode was 12 minutes and 42 seconds. The episode occurred 25 minutes into the session. The analysis of the episode revealed three distinctive categories. These categories were conceptualized as (1) husband owning his part in the relational distance, (2) partners in brief, restrictive contact with each other, and (3) couple relaxing: husband trusting/wife relieved about the husband's new openness in the therapy process. This episode began with the husband demonstrating serious reservation and reluctance to take part in the therapy process. He was emotionally uninvolved with the process of therapy. He also expressed lack of trust of others. He had made a decision early in his life to stay away from other people because of the deep hurt and betrayal suffered as a result of past relationships. As the therapist worked hard to engage the husband in the process of therapy, the wife listened attentively to him. The husband's monotonous, detached, and monosyllabic style of response did not change easily despite the therapist's continuous empathic, supportive, and respectful reflections. The therapist conveyed the message to the couple that the current problems in their marriage were equally shared by both members of the dyad. Given this situation, the therapist's focus centered on the husband's needs in the present moment and his lack of engagement in the process of therapy. 115 Episode 4 This episode emerged from the seventh session of couple 2010. The episode appeared 8 minutes into the session and lasted 52 minutes. The analysis of the episode produced four distinguishing categories. These categories were termed (1) couple contacting: pleading for understanding in fearful-distant interpersonal contact, (2) couple acknowledging the negative impact of sexual abusive experiences in the marriage, (3) couple unlocking, expressing, and feeling anger, and (4) couple expressing understanding, compassion, and acceptance. At the beginning of this episode, the therapist encouraged the couple to decide on the focus of the session. The mood between the partners was one of tension, frustration, and distancing. Further, the tone of their voices revealed frustration and impatience. Also, the wife described instances in which they had been angry and very frustrated with each other. Anger was unresolved and easily evoked in their lives. Any minor interrruption from either spouse would provoke intense outbursts of anger which would escalate and maintain the distancing pattern in the relationship. During the session, the wife acknowledged and disclosed her awareness of and concern with unexpressed anger in the relationship. This anger resulted in a significant block to their closeness as a couple. She further indicated anger toward the \"bitch\" (woman who had deeply betrayed the husband's trust) and the \"bastard\" (man who had sexually abused the wife) for the pain and suffering they had endured individually and as a couple. While the wife eagerly identified anger as a significant block to their 116 intimacy and, therefore, the need to address it in therapy, the husband minimized and denied its power and thus was reluctant to attend to his anger. He further identified anger as being his wife's issue but not necessarily his. Episode 5 This episode occurred in the tenth session of couple 2010. The episode appeared 7 minutes into the session and lasted 16 minutes and 27 seconds. The analysis of the episode generated two categories. The first category was named exposing and releasing a secret bond with a ghost from the past and the second was termed brief sharing of perceptions and feelings. Prior to the emergence of this episode, the therapist invited the couple to decide on the obstacles (represented through actual objects in the therapy room) to their marriage that they wished to work on during the session. They decided that since the husband had brought in his letter to the \"bitch\" (the woman that had created intense pain and anguish for him) they would focus on that obstacle. He shared the depth of love and tenderness he felt toward this woman and the cruelty and harshness with which she treated him while they were together. This was his first relationship with a female. He was 27 years old at the time. The consequences of that relationship left him unable to trust and love others. He had written a letter to this female which he brought into the session for the therapist to read outside of the session. The letter was placed on the chair representing the \"bitch\" (his chosen word for the object representing a significant block to intimacy with his wife). This chair was situated to his right hand side. He was sitting and facing his wife directly. The cushion of happiness (symbol of their 117 desired state, closeness) was placed on the floor between the partners. The therapist was fully present for him, addressing him in a calm and relaxed manner. The husband had not expected to read the letter himself or have it read aloud at all. In other words, he was not expecting to bring the secrets of that relationship out into the open and share those secrets with his wife. Eventually, the husband read the letter aloud. Episode 6 This episode occurred 28 minutes into session twelve of couple 2010. Its time lapse was 24 minutes and 57 seconds. The analysis of the episode generated three categories labeled (1) partners calmly engaging with each other, (2) wife releasing her block to intimacy, and (3) partners engaging each other compassionately. Prior to the beginning of this episode, the wife reflected on the many strings that were attached to her and the negative impact and influence these strings had on their marriage. The strings were alcohol, her ex-husband, the \"bastard\" and the \"bitch.\" Also, she reflected on her explosive outbursts of anger and her lack of awareness as to which string was being pulled during these explosive episodes. Further, she pondered on the string linked to her husband, that is, his relationship with the \"bitch\" regarding his lack of trust. This particular string had prevented him from coming closer to her. The episode began as the partners were interacting and sharing their feelings of hurt about the lack of trust in each other. 118 Compared to the previous ones, this episode was initiated under a different condition. Here, the husband was fully present, engaged and authentic with his wife and the therapist. Also, the wife shared and expressed freely her feelings and experiences to her husband. More specifically, she expressed her sadness and pain about her husband's lack of trust in her. The therapist stepped back and encouraged the couple to dialogue with each other at their own pace. Episode 7 This episode emerged from the thirteenth session of couple 2010. Its duration was 17 minutes and it began 38 minutes into the session. The analysis of its internal structure revealed the emergence of two categories: (1) removing the couples's blocks to closeness, and (2) the couple engaging with each other in freedom, genuineness, and intimacy. The physical seating arrangement of the members of the therapeutic system accurately reflected the atmosphere prior to the emergence of the episode. Everyone was sitting around a circle in the middle of the therapy room. The husband and wife were facing each other. On the floor, a cushion (couple's symbol of togetherness and happiness) was placed equidistantly to each other. The \"bastard's chair\" (representing the man who sexually abused the wife) was placed to the wife's left-side. The \"bottle\" (symbol of alcohol) was placed to the husband's left-side. The wife's ex-husband's chair, another block to the couple's closeness, was placed between the \"bottle\" and the \"bastard\". The \"bitch's chair\" (representing the woman that betrayed the husband) was placed a bit further back to the husband's right hand-side. The therapist was sitting 119 equidistant from the wife and the husband. All obstacles to the couple's sense of intimacy and togetherness were evoked and physically externalized at the beginning of the episode. The couple was in a state of tension and anticipation about the heightened presence of the many blocks to their togetherness. Episode 8 Episode # 8 occurred in session fourteen of couple 2010. Its time span was 21 minutes and it occurred 42 minutes into the session. The analysis of the actions and interactions generated two interrelated categories conceptualized as (1) letting go: the husband letting go of the \"bitch\" and the wife letting go of the \"bastard\", and (2) acquiring a renewed sense of self with a new perspective. Prior to the emergence of this episode, the husband revealed doubt and skepticism about the therapy process while the wife revealed anticipation. During the previous session, the therapist suggested to the couple that they reflect on and bring into the session an object symbolic of the experiences associated with the blocks they wished to let go of. The wife brought in a photograph of her wedding to her first husband. In this photograph, she was in her wedding gown standing between her ex-husband and the \"bastard\" (the man who sexually abused her). The \"bastard\" was giving a toast to the bride. The husband brought in a large piece of paper with the word \"bitch\" written on it. The purpose of the invitation was to encourage the couple to engage in a letting go ceremony. Although the wife was ready and prepared to engage in such a ceremony, the husband was doubtful and sceptical. Following some story-telling about his 120 relationship with the \"bitch\" and the suffering he had experienced with her, the therapist wondered about his readiness to let go of some of the experiences associated with her. The husband was frustrated, uptight and tense despite his engagement and full collaboration with the therapy process. The tension stemmed from doubt and some unwillingness to let go of something that was deeply interwoven into the fabric of his being. The wife listened attentively to her husband's inner struggles. She was no longer fearful of his frustrations and tensions. The therapist was fully present for the husband. Her stance was supportive and attentive. Her tone was soothing and gentle. Episode 9 This episode appeared in the second session of couple 2036. The episode lasted 16 minutes and occurred 9 minutes into the session. The analysis of the psychotherapeutic actions and interactions generated two categories. These categories were conceptualized as (1) couple distancing: wife pulling away in pain and husband pulling away in fear, and (2) couple engaging through verbal expressions of support and acceptance. Prior to the beginning of this episode, the husband and wife were blaming and criticizing each other about their lack of communication and support. The couple had discrepant views on whether and how they communicated and supported each other. Lack of connection and communication was the result of distancing actions manifested by both partners. The wife recounted numerous occasions during the week in which her husband was unable to share openly and support her. As she related these events, the husband distanced himself 121 even more by adopting a sulky and silent pose. This episode began as the wife, in a somewhat distant and detached manner, reflected on her husband's lack of care, compassion, and support for her and their immediate family. By now, the husband was unresponsive and inattentive to her story about her wish to be cared for, supported and loved by him. Episode 10 Episode # 10 emerged in session eight of couple 2036. This episode occurred 11 minutes into the session and lasted 34 minutes and 2 seconds. Analysis of the action and interactions of the members of the therapeutic system revealed three categories: (1) couple agreeing to a mutual therapeutic mandate, (2) couple working through demons, and (3) couple coming closer to each other through appreciation and understanding. Prior to the emergence of the episode, the husband expressed frustration, tension and anticipation regarding the unfolding of events in the session. The wife, on the other hand, appeared in a state of wonderment and support toward her husband. The husband had earlier been engaged in a push-pull dance with the therapist regarding the exploration of traumatic experiences from the past. He was confused and frustrated because he expected her to provide a cure for his emotional wounds while he was also afraid to explore and delve into those issues. In addition, he also expected her to provide answers to questions he had concerning possible sexual abusive experiences to which the therapist seemed unwilling to respond to. In a gentle and caring way, the therapist challenged and encouraged him to stay focused on his need to find the answers through the process he was engaged in. The wife 122 was attentive and supportive toward him but also in a state of wonderment about the confusion her husband was experiencing. This description of the ten selected episodes of change was intended as a general account of the episodes. The episodes occurred at various points within the session and their time spans varied. Episodes number one to eight revealed a regular flow of the therapeutic process, with a beginning, middle, and end. Episodes nine and ten revealed only an initial stage of the therapy. However, despite these differences, the actions and interactions of the members of the therapeutic system, as revealed in the ten episodes of change, showed a coherent structure and noticeable similarities. The following chapter presents the results from the grounded theory analysis of the ten change episodes. 123 CHAPTER V RESULTS: C O N C E P T U A L MODEL OF C H A N G E This chapter presents the findings of the qualitative, grounded theory analysis of ten change episodes selected from two marital ExST cases. The results are presented in three major sections. First, the conceptual model of change generated from the analysis is introduced. Second, the core category, along with its explaining categories and subcategories, comprising the model of change, is presented. Third, each category and subcategory is defined and exemplified with a corresponding illustration. While another selection of illustrations is provided in Appendix D, an extensive set of examples is also available from the researcher. Conceptual Model of Change The discovery-oriented observational analysis of the 40 marital ExST sessions revealed the presence often episodes of change. The grounded theory analysis of these episodes addressed the three interrelated research questions posed by this investigation; namely, the provision of a conceptual description of the internal structure of the change episodes as depicted by the clients' and therapist's actions and interactions. The analysis of the actions and interactions of the members of the therapeutic system, in the episodes of change, was articulated in a conceptual model of change entitled synergetic shifting (see Figure 2). 124 cn CD \u00C2\u00B0> f\u00E2\u0080\u0094 CD c C >, racti eely = tr essi mac racti eely Q. O essi mac racti eely 0) CL essi mac O Q. Int \" 3 < (/) x .E UJ co u \u00E2\u0080\u00A2 i CD S.I CO H\u00E2\u0080\u0094H 0) Q . COU ting TJ o o c o (A o 0) ict t : <*-(0 tz Q . o W o fl) 75 c c o o TO JS c 0) 'E $ \u00C2\u00A9 O 4-1 tn c o CO 1_