Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An exploratory study of therapist attachment and working alliance bonds McKee, Gerard Mercier 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-0123.pdf [ 4.33MB ]
Metadata
JSON: 831-1.0054091.json
JSON-LD: 831-1.0054091-ld.json
RDF/XML (Pretty): 831-1.0054091-rdf.xml
RDF/JSON: 831-1.0054091-rdf.json
Turtle: 831-1.0054091-turtle.txt
N-Triples: 831-1.0054091-rdf-ntriples.txt
Original Record: 831-1.0054091-source.json
Full Text
831-1.0054091-fulltext.txt
Citation
831-1.0054091.ris

Full Text

AN EXPLORATORY STUDY OF THERAPIST ATTACHMENT AND WORKING ALLIANCE BONDS  by GERARD MERCIER McKEE B.A., Concordia University, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Counselling Psychology) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA December 1993 © Gerard Mercier McKee, 1993  In  presenting  degree  at  this  the  thesis  in  partial  fulfilment  of  University  of  British  Columbia,  I agree  freely available for copying  of  department publication  this or of  reference  thesis by  this  for  his thesis  and study. scholarly  or  her  for  the  I further  purposes  may It  gain shall not  permission.  Department  of (  J30\ri££a-trJ6-1  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  D z / o i / t U  St*  that  agree  representatives.  financial  requirements  c/-/DC&Grzs1 /  be is  that  the  for  by  understood be allowed  advanced  Library shall  permission  granted  an  for  the that  without  make  it  extensive  head  of  my  copying  or  my  written  ii Abstract The purpose of this study was to examine whether therapists' responses to the Adult Attachment Questionnaire (AAQ, Hazan & Shaver, 1987) and the Adult Attachment Scale (AAS; Collins & Read, 1990) would be predictive of the strength of client and therapist reported Bond scores on the Working Alliance Inventory-Short Form (WAI-S; Tracey & Kokotovic, 1989). A total of 40 Lower Mainland of British Columbia therapists participated in Phase 1 of the study by completing the AAQ, the AAS and a demographic questionnaire. A total of 10 of these therapists participated in Phase 2 of the study. Each of the Phase 2 therapists recruited three clients to participate in the study. Immediately following the third session, both therapists and clients completed the WAI-S. Clients also completed the AAQ and a short demographic questionnaire. Although an uneven distribution of attachment types in the sample precluded any formal statistical analysis, visual examination of therapists' responses to the attachment measures suggests that therapists who chose the Secure AAQ description obtained higher mean scores on the AAS Close and Depend dimensions and lower AAS Anxiety dimension scores than therapists who chose the insecure AAQ descriptions. Correlation matrices were compiled to examine the strength of relationships among the standings of therapists on the AAS dimensions, on therapist ratings of the WAI-S and client ratings of the WAI-S. No significant correlations were found in relation to therapists' AAS dimension scores and therapist or client WAI-S ratings. Significant correlations were found for both client and therapist WAI-S interscale ratings and between client and therapist WAI-S ratings. A t-test was carried out to examine high and low WAI-S Bond ratings of clients in relation to therapists' WAI-S ratings. Significant differences were found between the WAI-S Bond ratings of therapists whose clients were identified as having high WAI-S Bond ratings and the WAI-S Bond ratings of therapists whose clients were identified as having low WAI-S ratings. The study found no evidence that attachment information in general predicts the strength of working alliance bonds. However, support was found for the predictive ability of the AAS Depend  Ill  dimension. The Depend dimension scores of therapists whose clients reported higher WAI-S Bond strength were significantly different from the Depend dimension scores of therapists whose clients reported lower WAI-S Bond strength.  iv  TABLE OF CONTENTS Abstract  ii  Table of Contents  iv  List of Tables  vi  Acknowledgement  vii  Introduction  1  Chapter One: Focus of Current Research Definition of Terms  3 4  Chapter Two: Review of Literature Client-Centered theory Social-Influence theory The psychodynamic perspective The Working Alliance Alliance research Definitional diversity Research groups Attachment theory Purpose of the Study  6 6 6 6 7 7 8 8 9 18  Chapter Three: Method Overview Pilot Testing Data gathering Attrition Realignment of the study Sample Therapists Clients Measures The Working Alliance Inventory (Short Form) Adult Attachment Scale Adult Attachment Questionnaire Demographic Questionnaires Therapist Demographic/Background Questionnaire Client Demographic Questionnaire Procedures Phase 1 Phase 2 Research Design Therapist Variables Client and Therapist Variables Summary  20 20 20 20 21 21 22 22 22 22 22 24 24 25 25 25 25 25 26 27 27 27 27  V  Chapter Four: Results Pilot Study Sample Characteristics Therapist demographics Therapist demographics: Phase 2 Client demographics Phase 1 Therapist attachment measures Phase 2 Therapist attachment measures AAS dimensions and gender Therapist WAI-S ratings Client WAI-S ratings Therapist and client WAI-S ratings Client attachment and client WAI-S ratings Therapist attachment and therapist/client WAI-S ratings AAS scores across phases Therapist WAI-S ratings and client WAI-S Bond ratings Therapist mean AAS scores and client WAI-S Bond ratings Therapist median AAS scores and client WAI-S Bond ratings  29 30 31 31 31 32 32 32 33 33 34 35 35 36 37 38 41 41 42 44  Chapter Five: Discussion Rationale Therapist Attachment Client Attachment AAS Dimensions and Gender Attachment Measures Working Alliance Client WAI-S ratings Therapist WAI-S ratings Therapist and client WAI-S ratings Therapist and client ratings of the WAI-S Bond component Attachment and the Working Alliance Conclusions Limitations Recommendations for Future Research  46 46 47 49 51 52 54 54 56 57 58 60 61 63 65  References  69  Appendix 1 - Initial Contact  80  Appendix 2 - Therapist Forms Phase 2  81  Appendix 3 - Client Forms  92  Appendix 4 - Therapist Demographics  98  Appendix 5 - Raw Data  101  VI  List of Tables  1  2  3  Pilot Study: Comparing Mean AAS scores and Adult Attachment Questionnaire Distribution  30  Phase 1 Therapists: Comparing Mean AAS scores and Adult Attachment Questionnaire Distribution  33  Phase 2 Therapists: Comparing Mean AAS Scores and Adult Attachment Questionnaire Distribution  34  4  Therapist WAI-S Ratings  35  5  Client WAI-S Ratings  36  6  Therapists WAI-S Ratings and Client WAI-S Ratings  37  7  Clients: Comparing Mean WAI-S Ratings and Adult Attachment Questionnaire Distribution  38  8  Therapist AAS Scores and Therapist WAI-S Ratings  39  9  Therapist AAS Scores and Client WAI-S Ratings  40  10  Comparing Mean AAS Scores Across Phases  40  11  Comparing High and Low Therapist Mean WAI-S Ratings  43  12  Comparing High and Low Therapist Mean AAS Dimension Scores  44  13  Comparing High and Low Therapist Median AAS Dimension Scores  45  vii Acknowledgement  I am grateful to the members of my advisory committee, Beth Haverkamp, Bonnie Long, and Adam Horvath, for their support and assistance.  I am also grateful to the management and staff at Coast Mountain Geological Ltd., Gary, Heather, Will, Sophie, and Joan, for their help and generosity. In particular, I owe a great deal of thanks to my good friend Chris Basil.  I am indebted to the following individuals who acted as consultants during the final phase of thesis preparation: Caara Cardell for her kindness, patience, and electronic expertise; Cheryl Garden for her assistance with statistical matters; and, Andrew Shaw for his help in improving my computer skills.  I am most indebted to my wife, Pelly Shaw, whose love, support, encouragement, and understanding have in large part made this thesis possible.  1 Introduction The finding that outcome studies have generally shown limited or no significant differences across treatments grounded in diverse psychotherapies has led several authors to suggest a commonality among the various therapeutic approaches (Frank, Hoehn-Saric, Imber, Liberman, & Stone, 1978; Goldfried, 1980, 1982; Kazdin, 1979). Other theorists (Greenberg & Pinsoff, 1987; Shapiro & Shapiro, 1983; Smith & Glass, 1977) echo these sentiments and suggest that, in general, all treatments, irrespective of their theoretical underpinnings, have some beneficial effects, but that there appears to be relatively little reliable difference in treatment efficacy. Smith, Glass, and Miller's (1980) meta-analysis of outcome data compiled from more than 500 treatment studies concluded that the type of therapeutic approach employed, whether verbal or behavioral, psychodynamic or client-centered, does not appear to result in different types or degrees of benefit. Findings such as these provide as support for Frank's (1961) suggestion that non-specific or general factors may account for many of the beneficial effects of therapy. The consistency of results supporting the non-specific position has led researchers to look for pretreatment characteristics and process variables that may account for beneficial outcome across diverse psychotherapies. The concept of a working or therapeutic alliance between client and therapist is regarded by many psychotherapy process researchers and theorists as the most promising area of study to account for the relevance of the therapeutic relationship to therapeutic outcome (Greenberg & Pinsoff, 1987; Hartley & Strupp, 1983). A growing number of studies have begun to explore and reveal the nature and significance of the alliance (Gomes-Swartz, 1978; Hartley & Strupp, 1983; Horvath, 1981; Luborsky, 1976; Marziali, 1984; O'Malley, Suh, & Strupp, 1983). It has recently been suggested (Gaston, 1990) that alliance research might benefit from the alliance construct being considered within the context of a larger theory of psychotherapeutic change processes, such as attachment theory (Bowlby, 1988). In line with this suggestion, it has been argued that the bond component of the alliance is in fact a positive attachment between the therapist  2 and client (Gelso & Carter, 1985). Although alliance formation is seen as being influenced by the psychological attributes of both the client and the therapist (Gaston, 1991), no study to date has attempted to examine the relationship between therapist pretherapy characteristics and strength of alliance (Kivlighan & Schmitz, 1992). The present study is an attempt to examine the strength of association between alliance formation and therapist attachment information.  3 Chapter One: Focus of Current Research  It has long been accepted that the relationship between the client and the counsellor is an important, if not the most important factor in counselling outcome, regardless of the practitioner's theoretical orientation (Bordin, 1979; Frank, 1961; Freud, 1913/1958; Rogers, 1957). This point is succinctly made by Yalom (1980) who states that "if any single fact has been established by psychotherapy research, it is that a positive relationship between patient and therapist is positively related to therapy outcome" (p.401). Throughout the 60's and 70's, a great deal of research examined the core conditions of empathy, congruence, and unconditional positive regard that were postulated by Rogers (1951, 1957) as necessary and sufficient for positive therapeutic change. Although Rogers' ideas stimulated a substantial body of research, the results from many studies examining the proposed core conditions have produced contradictory and inconclusive findings (Gurman, 1977; Lambert, DeJulio, & Stein, 1978). Over the last decade, researchers have begun to focus their attention on the role of the therapeutic or working alliance as a central component in psychotherapy outcomes (Bordin, 1979; Frieswyk, et al., 1986; Gelso & Carter, 1985) and a strong alliance between client and counsellor is now widely accepted as being a prerequisite for successful therapeutic outcome (Gelso & Carter). A variety of instruments have been developed to index the alliance and, despite differences in how the construct is operationalized, researchers have consistently found a positive relationship between reported strength of alliance and outcome measurement (Horowitz, Mamar, Weiss, DeWitt, & Rosenbaum, 1984; Horvath & Greenberg, 1989; Moras & Strupp, 1982; Morgan, Luborsky, CritsCristoph, Curtis, & Solomon, 1982). It has been argued that the strength of the working alliance is a function of three factors: therapist technical activity, client pretherapy characteristics, and therapist personal characteristics  4 (Moras & Strupp, 1982). In a recent study, Kivlighan and Schmitz (1992) examined the relationship between therapist technical activity and alliance formation, and found that initially weak alliances were apparently strengthened by therapists adopting a relatively more challenging, "here and now" oriented stance. Several studies have found significant relationships between ratings of client pretherapy relationships and ratings of the strength of the alliance (Kokotovic & Tracey, 1990; Moras & Strupp, 1982), and another found that the client's attachment history was significantly related to the strength of the alliance (Mallinckrodt, 1990, 1991). However, prior to the present study, no published research has examined the relationship between therapists' personal characteristics and alliance formation. Definition of Terms Working Alliance. This consists of the client's and the therapist's awareness of a set of agreements, understandings, and bonds that are arrived at during a sequence of purposive helping interactions. The following components, according to Bordin (1979), define a viable alliance, regardless of the specific theoretical approach taken by the therapist: (a)  Goal Component: The helper and the helpee have a sense of agreement about the goals of the helping process. The helpee has an awareness that these goals are relevant to him/her and feels a degree of identification with the explicit and implicit aims of the particular helping process in which he/she is engaged. The helper has some direct or indirect evidence that the goals established in the therapy relationship are explicitly or implicitly shared and accepted by the helpee.  (b)  Task Component: The helper and the helpee have a sense of mutuality (or agreement) that the tasks demanded of each of them in the helping process are reasonable and within their global capabilities (or expertise), and relevant in a direct or indirect way to the goals of the helping process upon which they have mutually agreed.  5 (c)  Bond Component: The helper and the helpee experience a sense of a bond between them. The bases upon which such a therapeutic partnership are built include sense of mutual trusting, liking, understanding, and caring.  Different therapeutic orientations and strategies make different demands on the participants in terms of each of these components. These unique demands create a unique quality for each successful alliance. Bordin (1979, 1980) has maintained, however, that all helping dyads have to achieve a basic qualitative level in each of the three areas in order to produce the alliance component necessary for a successful helping relationship. WAI-S:  The acronym "WAI-S" refers to the short form of the Working Alliance Inventory in both client (WAI-Sc) and therapist (WAI-St) forms.  WAI-Sc:  The client form of the Working Alliance Inventory short form.  WAI-St: Attachment:  The therapist form of the Working Alliance Inventory short form. An attachment may be defined as a strong and enduring bond developed within a significant relationship with a preferred individual who may be seen as more powerful and/or experienced.  6 Chapter Two: Review of Literature  Efforts to identify and study non-specific variables suggested to account for much of the outcome variance apparent in psychotherapy have historically fallen into four major theoretical categories (Horvath & Greenberg, 1989). These are Client-Centered Theory (Rogers, 1951, 1957), Social Influence Theory (Strong, 1968), Greenson's (1967) Psychodynamic Perspective, and Bordin's (1979) pantheoretical formulation of the Working Alliance. Client-Centered theory. Roger's (1951,1957) theoretical propositions, perhaps the most influential approach with regard to the search for relationship factors influencing therapeutic outcome, have generated almost two decades of research focused on the "facilitative conditions" of the therapeutic relationship. However, it seems that the conditions once proposed as "necessary and sufficient" for successful therapeutic outcome (i.e., empathy, unconditional positive regard, and congruence) appear to account for only a limited amount of outcome variance, and effects tend not to generalize across theoretical orientations (Gelso & Carter, 1985; Mitchell, Bozarth, & Krauft, 1977; Parloff, Waskow, & Wolfe, 1978). In summary, it appears that the core conditions postulated by Rogers may well be necessary to successful outcome, but are not sufficient in and of themselves. Social-Influence theory. Strong's (1968) theory focuses on the influence of expectations on therapeutic outcome. This orientation proposes that the likelihood of positive outcome is directly proportional to the extent that the client perceives his or her therapist as expert, attractive, and trustworthy. This suggests that the locus of the generic change variable resides in the client's perception of the therapist, and although the therapist is presented as influencing outcome, the dynamic interplay between participants is not given primary focus. The psychodynamic perspective. Greenson's (1967) model of the therapeutic relationship builds on: (a) Freud's (1913/1958) distinction between the friendly, non-neurotic aspects of the client's feelings for the therapist and the hostile, neurotic components of transference, and (b) the  7 work of later psychodynamic researchers (Gitleson, 1962; Sterba, 1934; Zetzel, 1956). As noted by Gelso and Carter (1985), Greenson proposes that from a psychodynamic perspective, the therapeutic relationship consists of three components: the working alliance, which is described as an alignment between the client's reasonable, observing side and the therapist's working side; the transference relationship, in which client feelings, behaviours, and attitudes formed in early relationships are displaced onto the therapist; and the real relationship, which represents the participant's genuine and realistic perceptions of and reactions to each other. With respect to the working alliance specifically, Greenson (1967) proposes that the client's motivation and ability to work in therapy in essence defines the alliance. From this perspective, the client must not only be motivated to overcome his or her problems, but must also experience a sense of helplessness and have the capacity to comprehend the therapist's insights as they relate to his or her situation. In these circumstances the therapist is said to be able to foster the working alliance by being empathetic, non-judgemental, compassionate, understanding, insightful, and by promoting therapy as a joint venture. The Working Alliance. Building on the work of Greenson (1967), Bordin (1979, 1980) reconceptualized the working alliance as a collaboration between the client and therapist while disassociating the alliance from any reliance on transference neurosis. From Bordin's point of view, the working alliance in and of itself is not therapeutic; rather, it is the collaborative nature of the therapeutic relationship that provides the client with the opportunity to make use of the therapist's skills and so work towards alleviating his or her problems (Bordin, 1980). As noted by Horvath and Greenberg (1989), Bordin presents the working alliance as "...a vehicle that enables and facilitates the specific techniques" employed within the counselling relationship and thus "...provides a framework for viewing both relationship variables and specific intervention techniques in an integrated format" (p.224) that is independent of any specific theoretical orientation. Alliance research. The current body of research on the alliance includes studies on the effect  8 of the alliance in psychodynamic, cognitive, and experiential therapy (e.g., Greenberg & Webster, 1982; Luborsky, 1976; Rounsaville et al., 1987) in relation to a variety of client complaints (e.g., Frank & Gunderson, 1990; Gomes-Swartz, 1978; Horvath & Greenberg, 1989). The impact of the alliance has been studied in relation to both brief and longer term therapy (Frank & Gunderson, 1990; Kokotovic & Tracey, 1990) and from the perspective of the client, the therapist, and that of independent observers (Marziali, 1984; Tichenor & Hill, 1989). Finally, outcome studies have assessed the quality of the alliance by tapping a variety of indices such as participants' reports and behavioral performance measures (Horvath, 1981; Luborsky, McLellan, Woody, O'Brien, & Auerbach, 1985). Definitional diversity. Although the importance of establishing a strong therapeutic or working alliance in counselling and psychotherapy is generally accepted as a prerequisite for successful outcome (Gelso & Carter, 1985), the focus of definitions used to describe this construct have varied. For example, Luborsky, Crits-Cristoph, Alexander, Margolis, and Cohen (1983) emphasize the client's experience of the therapeutic relationship, particularly the client's perception of the therapist as being helpful with respect to achieving goals. Horowitz et al., (1984) and Marziali, Marmar, and Krupnick (1981) stress the negative and positive contributions of both client and therapist; Strupp (1980) cites the interactive nature of the therapeutic relationship; and Frieswyk, Colson, and Allen (1984) focus on the collaborative capacity of the client in relation to the tasks of therapy. Research groups. Over the last 10 years, various researchers have developed instruments designed to measure the alliance from these different perspectives. The current measures of the alliance represent the attempts of five distinct research groups: the Vanderbilt Research Group (Hartley & Strupp, 1983; Moras & Strupp, 1982; O'Malley et al., 1983; Suh, Strupp, & O'Malley 1986); the Penn Research Group (Luborsky et al., 1983; Luborsky et al., 1980; Morgan et al., 1982); the Langley Porter Group (Gaston, 1990; 1991; Marmar, Gaston, Gallagher, & Thompson, 1989;  9 Marmar, Horowitz, Weiss, & Marziali, 1987; Mannar, Weiss, & Gaston, 1989; Marziali, 1984; Marziali et al, 1981); the Menninger Group (Allen, Newsom, Gabbard, & Coyne, 1984; Frieswyk et al., 1986; Frieswyk et al., 1984); and the British Columbia Group (Horvath & Greenberg, 1986; Horvath & Greenberg, 1989; Horvath & Symmonds, 1991). Recently, it appears that the definition and theoretical framework proposed by Bordin (1979, 1980) are gaining general acceptance (Tracey & Kokotovic, 1989) as being crucial to the inception of alliance research (Frieswyk et al., 1986) and instrumental in stimulating research on the alliance (Morgan et al., 1982). Bordin's framework is also considered to provide an eclectic conceptualization that may be generalized across a variety of psychotherapeutic orientations (Gaston, 1990). Attachment theory. Recently, a growing number of researchers have begun to study the processes by which individuals develop and maintain affectional bonds within close, personal relationships (Bretherton, 1985; Clark & Reis, 1988). Much of the credit for stimulating this research is due to the theoretical work of John Bowlby on infant attachment styles and his development of Attachment Theory (Bowlby, 1958, 1969, 1973, 1980). Bowlby defines an attachment as a bond developed with "some other differentiated and preferred individual, who is usually conceived as stronger and/or wiser" (Bowlby, 1977, p. 203) and defines attachment behaviour as "any form of behaviour that results in a person attaining or retaining proximity" (Bowlby, 1977, p.203) to an attachment figure. Therefore, it can be said that much of Attachment Theory is concerned with the bond that develops between the child and the parent/caregiver and the consequences this has for the child's developing self-concept and emerging perspective on social relationships. Bowlby's theoretical formulation has been described as an evolutionary-ethological approach (Ainsworth, Blehar, Waters, & Wall, 1978) in that infant attachment behaviours are seen as being fuelled by a goal-corrected behavioral system that has as its "set goal" the maintenance of proximity  10 to a nurturing caregiver and a biological function of maximizing the child's safety and security (Bowlby, 1982). Bowlby's theory grew out of observations of infants and children who were separated from their primary caregivers for varying periods of time (Bowlby, 1969). Bowlby proposed that when separated from its caregiver, an infant typically goes through a series of predictable emotional reactions. These are: Protest, which involves crying, active searching for the absent caregiver, and resistance to any soothing efforts of available others; Despair, which is described as passivity and obvious sadness; and Detachment, an active avoidance of the caregiver upon his or her return. Bowlby (1958) noted that this sequence of reactions was remarkably reliable across a variety of species and used this observation as a cornerstone in developing a theory of attachment grounded in evolutionary principals. Specifically, Bowlby suggests that the attachment process may serve an evolutionary purpose; by remaining in close proximity to the caregiver, who could provide protection from danger, infants who exhibit these specific behavioral and emotional propensities increase their likelihood of survival and successful reproduction. Attachment Theory may also be regarded as a model of personality and social development (Bowlby, 1973, 1982) wherein the child's developing personality is profoundly influenced by the nature of attachment relationships and in particular, the emotional availability and responsiveness of the primary caregiver (Bowlby, 1973). Through repeated interactions with attachment figures, the child is said to develop "working models" or internal representations of the self and others. Bowlby (1973; 1982) states that these representations consist of organized cognitions that have a direct bearing on social perception and subsequently form a template for future relationships. This template includes fundamental assumptions and beliefs concerning both the behaviour of others and the selfworth of the individual, i.e., whether the individual is worthy of care and attention. Empirical research on Bowlby's theory has focused mainly on the different attachment styles or patterns developed by infants and young children. Building on the work of Bowlby, Ainsworth  11 and her colleagues (1978) studied the relationship between degree of caregiver responsiveness, and signals made by the infant during the first year of life. Ainsworth et al. (1978) suggested that three distinct attachment styles characterized infants' expectations of their primary caregivers' availability and level of responsiveness. Ainsworth designed a research paradigm, the Strange Situation, in which the attachment processes of infants and young children were stimulated by repeated separations from their primary caregivers. Based on the results of childrens' reactions to the separations and subsequent reunions with caregivers, Ainsworth et al. (1978) were able to identify three basic attachment patterns, Secure, Anxious/Ambivalent, and Avoidant. Later research (Grossman & Grossman, 1991; Main, Kaplan, & Cassidy, 1985; Sroufe, 1983; Waters, 1978) provides support for these findings and suggests that these three patterns of attachment show a general degree of stability during the first several years of life. In line with Bowlby, Ainsworth et al. (1978) suggest that the quality of attachment is rooted in the history of interactions between infant and caregiver, and the extent to which the infant has learned to rely on the caregiver/attachment figure as a source of security. Thus, infants whose caregivers are consistently sensitive and responsive to the infant's signals and needs develop a secure attachment style, and are able to use the caregiver successfully when distressed and rely on him/her as a secure base from which explore the environment. In contrast, infants whose caregivers exhibit inconsistency in responding to the child's signals by being sometimes unavailable or unresponsive and at other times intrusive, develop an anxious/ambivalent style characterized by the child's preoccupation with the caregiver's availability to the detriment of exploratory behaviour. Caregiver behaviour characterized as rejecting or tending to rebuff the infants' attempts to gain proximity and bodily contact is suggested to result in an avoidant attachment style. In this case the child avoids contact with the caregiver and apparently suppresses attachment behaviour by focusing his/her attention on toys. In general, these infants "turn to the neutral world of things, even though displacement exploratory behaviour is devoid of the true  12 interest that is inherent in non-anxious exploration" (Ainsworth et al., 1978, pp. 319-320). The description of the three attachment styles suggested by Ainsworth et al. (1978), along with references to infants' expectations concerning their caregivers' accessibility and responsiveness, provide support for Bowlby's (1969, 1973, 1980) notion that children construct inner working models of themselves and their major social interaction partners. The construct of working models is integral to the theory of attachment, for it is through the development of these beliefs and expectations that continuity between infants' attachment experiences and later adult attachment is said to occur (Dontas, Maratos, Fafoutis, & Karengelis, 1985; Main et al., 1985; Sroufe, 1983). This point is highlighted by Bowlby who states that working models and the patterns of behaviour influenced by them are central components of personality and that "attachment behaviour (characterizes) human beings from the cradle to the grave" (1977, p. 129). Hazan and Shaver (1987) suggest that Attachment Theory might best be summarized in three propositions presented by Bowlby: The first proposition is that when an individual is confident that an attachment figure will be available to him whenever he desires it, that person will be much less prone to either intense or chronic fear than will an individual who for any reason has no such confidence. The second proposition concerns the sensitive period during which such confidence develops. It postulates that confidence in the availability of attachment figures, or lack of it, is built up slowly during the years of immaturity (infancy, childhood, and adolescence), and that whatever expectations are developed during those years tend to persist relatively unchanged throughout the rest of life. The third proposition concerns the role of actual experience. It postulates that the varied expectations of the accessibility and responsiveness of attachment figures that individuals develop during the years of immaturity are tolerably accurate reflections of the experiences those individuals have actually had" (Bowlby, 1973, p.235). The continuity component of Attachment Theory (i.e., the relative permanence of mental models) is still somewhat controversial, but does have support from longitudinal studies conducted  13 from infancy to the early elementary school years (Dontas et al., 1985; Erikson, Sroufe, & Egeland, 1985; Main et al., 1985; Sroufe, 1983; Waters, Wippman, & Sroufe, 1979). Evidence of this kind provides support for Bowlby's proposals and suggests that an individual's adult relationship style might be influenced by early attachment experiences. Some research on adult functioning, although limited, does suggest possible linkage between early attachment experiences and subsequent parenting (Ricks, 1985; Sroufe & Fleeson, 1986) and depression (Adam, 1982; Frommer & O'Shea, 1973a, 1973b). The most recently researched area linking early attachment history and later adult functioning focuses on romantic relationships. Using the three attachment types (secure, avoidant, and anxious/ambivalent) suggested by Ainsworth et al. (1978), Hazan and Shaver (1987) explored "the possibility that love is an attachment process - a biosocial process by which affectional bonds are formed between adult lovers, just as affectional bonds are formed earlier in life between human infants and their parents" (p. 511). Using a psychometrically-based relationship questionnaire designed to investigate the theoretical foundation of the Ainsworth et al. (1978) attachment types, Hazan and Shaver found that individuals identified as secure lovers reported that their relationships tended to last longer than either of the insecure types. Secure lovers also described their most significant love experience as trusting and friendly and saw themselves as supportive and accepting of their partners. In contrast, those identified as anxious/ambivalent described their significant relationships as being fraught with obsession, emotional extremes and desire for reciprocation and union. The love experiences of avoidant types were also characterized by emotional extremes, but included jealousy and fear of intimacy. Hazan and Shaver (1987) also examined differences in the respondents' attachment histories. This was achieved through two lines of questioning. Respondents were first asked to report whether they had been separated from either parent for what seemed like a long time, and whether their parents had divorced or separated. Second, respondents were asked, by means of an adjective  14 checklist, to describe how each parent had behaved toward them during childhood. The adjectives used were derived from an earlier pilot study (Hazan & Shaver, 1987) and included items such as caring, critical, responsive, and intrusive. In addition, adjectives describing the parental relationship were included as a means of tapping how the parents related to one another. These descriptors included argumentative, affectionate, and unhappy. Hazan and Shaver (1987) report that neither parental divorce nor duration of separation from parents during childhood were significant predictors of attachment type. They conclude that "the best predictors of adult attachment type were respondents' perceptions of the quality of their relationship with each parent and the parent's relationship with each other" (p. 516). In summary, compared to insecure types (anxious/ambivalent and avoidant), those identified as secure report warmer relationships both with parents and between parents. The avoidant types, on the other hand, perceived their mothers as rejecting and cold, whereas those identified as anxious/ambivalent typically describe their fathers as unfair. Other studies (Collins & Read, 1990; Feeney & Noller, 1990; Simpson, 1990) have built on the work of Hazan and Shaver (1987) and provide support for much of their findings. For example, Collins and Read, using an adult attachment style measure based on Hazan and Shaver's (1987) categorical measure, conducted a series of three studies designed to examine the correlates of adult attachment. The respondents in all three studies were undergraduate students at the University of Southern California. The first study concentrated on the development of a measure of adult attachment that would tap dimensions of attachment rather than the discrete categories utilized by Hazen and Shaver (1987). This resulting measure, the Adult Attachment Scale (Collins & Read, 1990), was then employed to identify respondents' attachment types in the latter two studies. The purpose of the second study was to examine any connections between respondents' mental representations of self, others, romantic relationships, and adult attachment styles. The third study examined the manner in which attachment style impacted upon ongoing romantic relationships. The  15 results of the second study appear to support the notion of differential attachment being linked to patterns of belief about the self and others. For example, respondents identified as having secure attachment had a greater sense of self-worth and were more socially self-confident. Secure individuals, as compared to insecure types, also reported a more positive social outlook and saw others as trustworthy, altruistic, and dependable. With respect to romantic relationships, secure types were found to be more selfless in their love style and less likely to be obsessive or manipulative. In contrast, insecure types were more likely to have negative belief systems about themselves and others, were lower in self-worth, selfconfidence, sense of control, and, as with Levy and Davis's (1988) respondents, more likely to have a love style characterized as obsessive and dependent. The respondant's attachment histories also differed according to attachment typology. Whereas secure types recalled their parents as being warm, receptive, and responsive, those respondents identified as insecurely attached had recollections of their parents as being cold and rejecting or behaviourly inconsistent. Feeney and Noller (1990) also conducted research that drew heavily on the Hazan and Shaver (1987) study. Respondents in this study were first year undergraduate students (n=374) at the University of Queensland, Australia. Using a variety of self-report measures, Feeney and Noller (1990) attempted to gauge the usefulness of attachment style as a method of predicting types of adult romantic relationships. All respondents completed measures of love and self esteem as well as modifications of Hazan and Shaver's (1987) measures of attachment history and attachment style. The resulting analysis of responses provides support for the earlier findings of Hazan and Shaver (1987), particularly the relationship between early childhood experience of parenting behaviour and later adult attachment style. Thus, respondents identified as having an avoidant attachment style were found to be more likely to select items indicative of being mistrustful and distant from others and anxious/ambivalent types were more likely to respond in a manner suggesting high levels of dependency and typically remembered their parents as being unsupportive.  16 In contrast, individuals identified as securely attached were found to express a trusting attitude toward others and remembered their early family experiences as being positive and their parents as supportive. As was the case in Hazan and Shaver's (1987) study, secure respondents were more likely to report longer lasting love relationships when compared with anxious/ambivalent types (least enduring) and avoidant types. In a study examining the impact of attachment style on romantic relationships, Simpson (1990) asked introductory psychology students and their partners to complete a series of self-report measures inquiring about their dating relationships. The study also included 13 sentences, presented on Likert-type scales, which were derived from the adult attachment measure developed by Hazan and Shaver (1987). Simpson (1990) constructed three relationship-based hypotheses extrapolated from the attachment literature. First, it was proposed that the kind of romantic relationship would be mediated by the individual's particular mental model and attachment style (Bowlby, 1973, 1980). From this perspective, securely attached individuals would be expected to be engaged in stable, supportive relationships characterized by interdependence, trust, and commitment. Anxious/ambivalent types, on the other hand, while desiring secure relationships, would be expected to find themselves confounded by general relationship insecurity and feelings of ambivalence. Last, avoidant individuals would be expected to be involved in relationships marked by emotional unavailability and low levels of trust, commitment, and interdependence. Simpson's (1990) second hypothesis was based on the premise that differing attachment styles would impact upon the tenor of romantic relationships (Bowlby, 1973). Thus, secure respondents would be expected to be engaged in relationships characterized by relatively frequent positive emotions and infrequent negative emotion. Avoidant individuals, on the other hand, who would be expected to shun intimate, committed relationships, should report the opposite set of circumstances. Last, respondents identified as being anxious/ambivalent, given their tendency toward  17 obsessiveness and over-dependency, should also report frequent negative emotion in their relationships. The third hypothesis (Simpson, 1990) predicted that levels of emotional distress following relationship dissolution should be mediated by the respondent's attachment style (Bowlby, 1973, 1977). In this case it would be expected that individuals with a tendency to eschew intimacy and commitment (avoidant types) would report the lowest levels of emotional distress following termination of a romantic relationship and anxious/ambivalent individuals, given their tendency toward obsessiveness, should experience more intense distress. Securely attached respondents were also expected to experience some distress following relationship dissolution, but it was assumed that the structure of their mental models of themselves would provide protection from excessive levels of distress. The results of individual-level analyses carried out by Simpson (1990) found support for the hypothesized relationship between attachment style and experience of romantic relationships. First, securely attached individuals were more likely to be involved in relationships characterized as being interdependent, trusting, and committed whereas the opposite was found for insecure respondents. This was particularly true of those respondents identified as having an avoidant attachment style. Second, respondents identified as securely attached were more likely than those identified as insecurely attached to report relationships in which they experienced frequent occurrences of mild and intense positive emotion and infrequent negative emotion. The opposite pattern was found for insecurely attached individuals. Last, as Simpson (1990) had hypothesized, individuals identified as having an avoidant attachment style reportedly experienced less emotional distress after relationship dissolution. This was found to be particularly true of highly avoidant men. In addition to providing support for the existence of a linkage between an individual's recall of childhood attachment experiences and later adult functioning, the reported attachment literature also supports the hypothesis that adult respondents can meaningfully classify themselves within the  18 categories of Secure, Anxious/Ambivalent, and Avoidant. Related to this is the finding that frequency rates within the three attachment classifications (e.g., Hazan & Shaver, 1987) are remarkably similar to those found in studies of infant-mother attachment (Campos, Barrett, Lamb, Goldsmith, & Stenberg, 1983). The finding of comparable frequency of attachment style between infants and adults is particularly striking given the necessary differences in the way attachment style is assessed. Due to infants' limited language acquisition and representational abilities, it is necessary to employ behavioral methods when attempting to assess their attachment styles (Kobak & Sceery, 1988). In the area of infant attachment research, the most commonly used method to assess attachment typology is the Strange Situation format (Ainsworth et al., 1978). In a review of American studies of infant-mother attachment (Campos et al., 1983) that had employed the Strange Situation as a system of classification, researchers reported attachment proportions of 62% Secure, 23% Avoidant, and 15% Anxious/Ambivalent. In studies of adult attachment using Hazan and Shaver's (1987) self-report format, frequencies of attachment type have typically followed the same distribution pattern. For example, Hazan and Shaver (1987) found proportions of 56% secure, 25% avoidant, and 19% anxious/ambivalent, Feeney and Noller (1990) report frequencies of 55% secure, 30% avoidant, and 15% anxious/ambivalent, and Mikulincer, Florian, and Tolmacz (1990), using a sample of Israeli undergraduate students, found frequencies of 56% secure, 25% avoidant, and 19% anxious/ambivalent. Other studies (Hazan & Shaver, 1990; Levy & Davis, 1988) report similar frequencies. Findings such as these not only support the notion that respondents can meaningfully classify themselves with respect to attachment style, but also provides support for the construct validity of the self-report measure developed by Hazan and Shaver (1987). Purpose of the Study A growing body of research now suggests that individuals' attachment information may be  19 directly related to the ability to foster positive relational bonds with others (Kestenbaum, Farber, & Sroufe, 1989; Koestner, Franz, & Weinberger, 1990). Given that no published research has focused on therapist attachment in relation to working alliance ratings, the present study was exploratory in nature and sought to question whether therapists' self-reported attachment information is related to differential strength of client Working Alliance Inventory Bond ratings.  20 Chapter Three: Method  Overview Pilot Testing. Pilot testing of attachment measures was carried out to determine whether the responses of individuals engaged in the field of counselling psychology would resemble previous research results and so support the feasibility of the main study. Both the Hazan and Shaver (1987) Adult Attachment Questionnaire (AAQ) and the Collins and Read (1990) Adult Attachment Scale (AAS) were administered to 14 University of British Columbia Counselling Psychology graduate students who were engaged in the clinical practicum portion of their programs. Respondents were approached at the beginning of a regular weekly debriefing meeting and asked to complete the attachment measures. All 14 individuals attending the meeting agreed to participate in the pilot study. Data gathering. A total of 48 therapists working in the Lower mainland of British Columbia, either in agency settings or in private practices, were contacted by mail during May and June of 1992 and asked: (a) to respond to the Hazan and Shaver (1987) Adult Attachment Questionnaire, the Collins and Read (1990) Adult Attachment Scale, and a demographic questionnaire, and (b) to indicate whether or not they would be willing to participate in the second phase of the study. At the time of initial contact, potential respondents were given a description of the second phase of the study. Participation was limited to therapists holding at least a Master's degree in a counsellingrelated field. The original plan of the study was to classify those therapists participating in Phase 2 according to self-reported attachment type as gauged by the Hazan and Shaver (1987) measure. It was hoped that a minimum of five therapists per attachment type (secure, avoidant, and anxious/ambivalent) would respond, and so provide adequate comparison across attachment types and working alliance scores. The response rate to the initial mail-out was encouraging with 28  21 (58%) therapists returning completed questionnaires. However, only 10 (20%) of the therapists agreed to participate in the second phase of the study and of these seven had selected the secure Hazan and Shaver attachment description (2 anxious/ambivalent, 1 avoidant). Consequently, another 18 Lower Mainland therapists were contacted in the same manner during October and November of 1992 and asked to participate in the study. Of those contacted, 12 responded by completing the questionnaires and 10 agreed to participate in the second phase of the study. However, the distribution of self-reported attachment type was again heavily weighted towards the secure description (9 secure, 1 anxious/ambivalent). Attrition. During the course of data gathering, a total of 10 (50%) therapists who had agreed to participate in the second phase either officially withdrew from the study (n=3) or failed to complete the requirements of the study (n=7). The reasons for such a high rate of attrition are uncertain. Of the three therapists who withdrew from the study, two cited personal problems for their decisions, and one reported that the majority of her clients were in crisis and therefore did not meet the requirements of the study. The remaining seven therapists were contacted on a number of occasions and variously reported that they had not had any new clients, had forgotten to request their clients' participation, or were having second thoughts about participating in the study. Although the explanations of these therapists must under the present circumstances be taken at face value, it may be that the prospect of being evaluated by their clients and then having these evaluations studied and documented accounts for the high attrition rate. Unfortunately the paucity of research in the area of therapist participation in experimental studies and the lack of information available regarding those therapists who failed to complete the present study combine to stymie any meaningful explanation of their actions. Because this attrition considerably reduced the sample size of the study, a series of analyses were planned and carried out to realign the focus of the study. Realignment of the study. Given that both the overall AAQ distribution of attachment types  22 reported by therapists who responded to Phase 1 of the study (82.5% Secure, 10% Avoidant, 7.5% Anxious/Ambivalent) and that of therapists participating in Phase 2 of the study (80% Secure, 20% Anxious/Ambivalent) did not follow the distributions found in the Pilot study (57.14% Secure, 28.57 Avoidant, 14.28% Anxious/Ambivalent) or those reported in previous research (Hazan & Shaver, 1987, 1990; Collins & Read, 1990; Simpson, 1990), it was decided to realign the study and to use therapist scores on the Collins and Read (1990) Adult Attachment Scale as the main attachment measure. Sample Therapists. A total of 66 Lower Mainland therapists working in the Lower Mainland of British Columbia either in agency settings or in private practice were contacted by mail and asked to participate in both phases of the study. Participation was restricted to therapists holding at least a Master's degree in a counselling-related field. A total of 40 therapists meeting the selection criteria responded to the initial phase of the study by completing a demographic/family background questionnaire, the Hazan and Shaver (1987) Adult Attachment Questionnaire, and the Collins and Read (1990) Adult Attachment Scale. A total of 10 therapists participated in the second phase. Clients. The client population consisted of individuals who were recruited by their therapists to participate in the study after completing two therapy sessions and immediately prior to their third therapy session. Client recruitment was further restricted to individuals who were at least 19 years old, were engaged in individual therapy, were judged by their therapists not to be in crisis, and were able to read and understand the research instructions and give informed consent without assistance. A total of 30 clients were recruited by their therapists to participate in the study. Measures The Working Alliance Inventory (Short Form) (WAI-S, Tracey & Kokotovic, 1989). The WAI-S was employed in the study and self administered by respondents immediately after the third therapy session. The Working Alliance Inventory (WAI, Horvath & Greenberg, 1986, 1989) is a 36-  23 item, self-report questionnaire, based on Bordin's (1979) conceptualization of the working alliance, that can be administered to both clients (WAIc) and therapists (WAIt). Earlier research (Horvath & Greenberg, 1986) has demonstrated adequate reliability for the WAI with overall client scores resulting in internal consistency estimates of alpha at .93 and WAI subscale alphas ranging from .85 to .88. Overall therapist scores resulted in internal consistency estimates of alpha at .87 and subscale alphas ranging from .68 to .87. Horvath and Greenberg (1986) have also provided evidence of the convergent and discriminant validity of the WAI items with support being demonstrated for the convergent validity of all three subscales and for the discriminant validity of the Task and Goal subscales. Acceptable subscale and composite reliabilities (coefficient alpha) were also demonstrated. Estimates of reliability based on Hoyt's (1941) algorithm reportedly ranged from .85 (Bond scale) to .92 (Task scale) for the client's version of the WAI, and from .82 (Goal scale) to .87 (Task scale) for the therapist's version of the instrument. However, it should be noted that the reliability of the Bond scale (.68) on the therapist's version was somewhat weaker than other findings. In developing the WAI-S, Tracey and Kokotovic (1989) factor analyzed the WAI and chose the four most powerful items from each of the 3 WAI subscales to form the WAI-S. The factor structure of the resulting instrument was reportedly found to be a hierarchical bilevel model and therefore comparable to the original measure. As with the WAI, each item on the short form consists of a 7-point, Likert-type scale. Both client and therapist forms of the WAI-S yield three 4item, first-order subscale scores (Bonds, Goals, and Tasks) and a single, second-order overall working alliance score (Tracey & Kokotovic, 1989). Tracey and Kokotovic (1989) report comparable alpha coefficients for the WAI-S of .98 with respect to client ratings and .95 for therapist ratings taken from a sample of 124 client-therapist dyads. The WAI-S was chosen for the present study in order to minimize client involvement and thereby increase the likelihood of therapist participation.  24 Adult Attachment Scale (AAS, Collins & Read, 1990). The AAS is an 18-item instrument designed to measure adult attachment style dimensions. The scale is derived from Hazan and Shaver's (1987) discrete, categorical measure and is intended to overcome some of the perceived shortcomings of the original instrument. Collins and Read (1990) conducted a series of three studies, the first of which developed and evaluated a multi-item scale based on Hazan and Shaver's (1987) discrete measure. The analysis of results revealed three underlying dimensions of the measure: Close, (six items) the degree to which an individual is comfortable with intimacy; Depend, (six items) the degree to which an individual believes others can be depended on to be available when needed; and, Anxiety, (six items) the degree to which an individual feels anxious about being unloved or abandoned. Internal consistency ratings for the Depend, Anxiety, and Close items using Cronbach's alpha (.75, .72, .69 respectively) were within acceptable limits. Analysis also revealed a moderate relationship between the Close and Depend dimensions (r=.38) and weak relations between the Anxiety and Close dimensions (r=-.08) and the Anxiety and Depend dimensions (r=-.24). Test-retest reliability correlations were carried out with a subset (n=101) of the original sample after an approximate 2 month delay. The results indicate moderate stability, and test-retest correlations for the Depend, Close, and Anxiety dimensions were .71, .68, and .52, respectively. Adult Attachment Questionnaire (AAQ, Hazan & Shaver, 1987) The AAQ was administered to identify therapist respondent's attachment style. Hazan and Shaver (1987) translated the typology developed by Ainsworth et al. (1978) into three, paragraph-long, attachment descriptions intended to be appropriate for adult populations. Respondents are instructed to select whichever of the three descriptions best portrays their feelings in close relationships and, based on their response, are identified as having a secure, avoidant, or anxious/ambivalent attachment style. The results of Hazan and Shaver's (1987) study found that the proportions of respondents identifying themselves as secure, avoidant, or anxious/ambivalent was very similar to proportions reported in studies of infant-  25 mother attachment (Campos et al., 1983). Hazan and Shaver (1987) also performed a hierarchical discriminant-function analysis in order to gauge predictability of membership in the three attachment categories from a attachment history questionnaire that was also developed for the study. From this analysis, two statistically significant functions emerged that accounted for 69.87% and 30.13%, respectively, of the between-groups variability. Hazan and Shaver (1987) report that the most reliable discriminators between secure and insecure respondents included (a) a mother perceived as respectful of the respondent (i=.44), accepting (r^.33), responsible (r^.31), confident (r=.35), and not intrusive (r=-.42); (b) a father who was caring (r=.41), humorous (r=.40), loving (r=.40), and affectionate (r=.30); and (c) a relationship between parents that was perceived as affectionate (r^.44), caring (r=32), and not unhappy (r= .34). Demographic Questionnaires Therapist Demographic/Background Questionnaire (TDBO). The TDBQ was designed to gather the following information about the therapist: professional affiliation, highest degree earned, number of years experience as a therapist/counsellor, theoretical orientation, gender, age, marital status, level of satisfaction with current relationship, person(s) with whom the therapist lived during the first 5 years of his/her life, number of siblings or half-siblings, birthplace of parents, and predominant ethnic/cultural influence in his/her family of origin. Client Demographic Questionnaire (CDQ). The CDQ was designed to gather the following information about the client: age, gender, marital status, and level of education. Procedures Phase 1. A total of 66 therapists were contacted by mail (see Appendix 1) between May and November 1992 and asked: (a) to complete and return the Hazan and Shaver (1987) Adult Attachment Questionnaire, the Collins and Read (1990) Adult Attachment Scale, and a demographic/family history questionnaire, and (b) to indicate on an accompanying form whether or  26 not they would be willing to participate in the second phase of the study. A brief description of the second phase of the study was provided at the time of initial contact. Phase 2. In the second phase of the study, therapists were asked to sign a consent form indicating their agreement to participate in the study. Therapists were also asked to recruit three clients to participate in the study. Therapists were instructed to approach, in turn, the first three clients meeting the aforementioned criteria and to request their participation in the study (see Appendix 2). Immediately following the request to participate in the study, therapists were instructed to: (a) present each client with a client information form that explained the purpose of the study and the steps to be followed by the client when completing the client questionnaires, and (b) assure clients that their responses would be kept in complete confidence (see Appendix 3). Client information forms also contained two copies of a client consent form. Therapists were instructed to ensure that those clients who agreed to take part in the study signed both copies of the consent form. Therapists were asked to sign each client consent form as a witness to the client's signature. One copy of the consent form was to be retained by the client for his/her personal records, and the second was to be retained by the therapist for collection at the end of the study. If any of the first three clients declined to participate in the study, therapists were instructed to approach the next available client(s) meeting the study's criteria and to repeat the request to participate. Therapists were requested to continue this procedure until three clients meeting the requirements of the study agreed to participate. Each therapist was provided with a set of instructions outlining the steps to be followed during administration of the study and six envelopes containing questionnaires (see Appendix 2). Three envelopes, marked THERAPIST/CLIENT #1, THERAPIST/CLIENT #2, and THERAPIST/CLIENT #3, contained a therapist questionnaire (WAI-St), and three, marked CLIENT #1, CLIENT #2, and CLIENT #3, contained client questionnaires (WAI-Sc, AAQ, and the client  27 demographic questionnaire). Therapists were instructed to complete each therapist questionnaire, in numerical order, immediately following the end of the third session with the appropriate client and to seal their responses in an envelope provided for this purpose. Therapists were asked to ensure that client respondents were provided with a private space in which to complete the client questionnaires. Client questionnaires were also to be completed immediately following the third session. Clients were instructed to seal their responses in an envelope provided for this purpose and then to return the envelopes to their therapist for safe keeping. Therapists were asked to contact the researcher by telephone immediately following the completion of all questionnaires. Both therapists and their clients were informed that they had the right to withdraw from the study at any time. Clients were assured that should they at any point decide to withdraw from the study, the quality or quantity of service they received would not be effected. These conditions were clearly articulated on relevant consent forms. Research Design Therapist Variables: Attachment. The primary measures of attachment in the study were therapists' responses to Collins and Read's (1990) Adult Attachment Scale dimensions of Close, Depend, and Anxiety. Therapists' responses to the Hazan and Shaver Adult Attachment Questionnaire were used as a secondary measure of attachment. Client and Therapist Variables: The Working Alliance. Three aspects of the working alliance were measured in the study: (a) client reported strength of working alliance bonds as measured by the Working Alliance Inventory Short Form Client Version (WAI-Sc); (b) therapist reported strength of working alliance bonds as measured by the Working Alliance Inventory Short Form Therapist Version (WAI-St); and (c) combined WAI-Sc/WAI-St results. Summary. In the present study, the strength of working alliance bonds were examined through self-report questionnaires completed by clients and therapists in the course of their  28 therapeutic engagements. The main purpose of the study was to explore the relationship between therapist self-reported attachment information and therapist and client self-reported strength of working alliance bonds formed during counselling/therapy. The overall research question of the study focused on whether therapists' self-reported attachment information was related to differential strength of client Working Alliance Inventory Bond ratings. This general question can be broken down into two specific questions: first, is therapists' self-reported attachment information related to an ability to foster working alliance bonds with clients?; and second, are therapists' scores on the Adult Attachment Scale dimensions of Close, Depend, and Anxiety related to the forming of working alliance bonds? Comparisons looked at client and therapist ratings of working alliance bonds in relation to therapists self-reported standing on the Adult Attachment Scale dimensions of Close, Depend, and Anxiety. The variables of interest were: (a)  client ratings of the strength of WAI Bonds immediately following the third therapy session.  (b)  therapist ratings of the strength of WAI Bonds immediately following the third therapy session.  (c)  therapist self-reported standings on the AAS dimensions of Close, Depend, and Anxiety.  In addition to the areas mentioned above, exploratory analyses were conducted to examine the interplay between therapist and client demographic and attachment information.  29 Chapter Four: Results  The Results portion of the study is broken down into four sections: Pilot Study; Sample Characteristics: Phase 1 and Phase 2. The first section presents the responses of counselling psychology graduate students who completed the Adult Attachment Scale (AAS) and the Adult Attachment Questionnaire (AAQ) in a preliminary study prior to the onset of the main study. The second section presents a breakdown of the demographic information collected from therapist respondents in Phase 1 of the study. The demographic information from therapists who participated in Phase 2 of the study is also presented, as is the demographic information collected from client respondents. The third section describes the responses of the 40 therapists who participated in the first stage of data collection. It presents the distribution of the therapists' responses to the Adult Attachment Questionnaire in Phase 1 of the study and compares Phase 1 therapists' selection of attachment type with their corresponding mean scores on the Adult Attachment Scale dimensions of Close, Depend, and Anxiety. The fourth section presents the results for the 10 therapists and 30 clients who participated in the second phase of the study. The distribution of therapists' responses to the Adult Attachment Questionnaire is presented and compared to therapists' selection of attachment type. The strength of relationship between therapist and client working alliance, including subscale relationships, and the interrelationship between therapist/client attachment and working alliance ratings is also reported. In addition, this section presents the means and standard deviations of Pilot Study, Phase 1, and Phase 2 therapists' responses to the Adult Attachment Scale dimensions of Close, Depend, and Anxiety. Also reported are the results of paired t-Tests examining clients' WAI-S Bond ratings in relation to therapists's ratings of WAI-S Task, Bond, Goal, and Combined WAI-S ratings. The  30  section concludes with the results of paired t-Tests examining clients' WAI-S Bond ratings in relation to therapists's scores on the Adult Attachment Scale dimensions of Close, Depend, and Anxiety. Pilot Study. The results of the pilot testing of attachment measures conducted with 14 University of British Columbia Counselling Psychology graduate students were similar to distributions found in previous research (see Table 1). Table 1 Pilot Study: Comparing Mean Adult Attachment Scale Scores and Adult Attachment Questionnaire Distribution  n  Secure 8 (57%)  Attachment Style Avoidant 4 (28%)  Anx/Amb* 2 (14%)  Close  26.62 SD2.87  17.25 SD4.19  22.00 SD5.65  Depend  22.50 SD5.18  12.25 SD3.59  16.50 SD3.53  9.12 SD 1.72 * Anxious/Ambivalent  17.75 SD 3.86  17.00 SD 1.41  Anxiety  With respect to the Hazan and Shaver (1987) Adult Attachment Questionnaire (AAQ), a total of 8 (57%) respondents selected the Secure attachment description, 4 (28%) selected the Avoidant description, and 2 (14%) selected the Anxious/Ambivalent description. The Pilot study respondents' scores on the Collins and Read (1990) Adult Attachment Scale (AAS) were higher on the AAS Close (23.29, SD 5.41) and Depend (18.86, SD 6.24) dimensions, and lower on the Anxiety (12.71, SD 4.87) than the norms reported by Collins and Read (1990), but given that the original norming was based on the responses of an introductory psychology class and the Pilot study respondents were counselling psychology graduate students involved in their clinical practicums, these differences may be attributable to differences in the samples.  31 Sample Characteristics Therapist demographics: Phase 1. Therapists ranged in age from 26 to 65 years, the mean age being 43.10 (SD 9.35). A total of 27 of the therapists were female, and 13 were male. Therapist professional affiliations were reported in the following categories: Counsellor (n=20), Family Therapist (n=l), Social Worker (n=8), Psychologist (n=ll). Therapists held degrees in the following areas: Ed.D (n=l), M.A. (n=16), M.D. (n=l), M.Ed. (n=6), M.S.W. (n=8), and Ph.D. (n=8). Therapist theoretical orientations were: Cognitive Behavioral (n=5), Eclectic (n=13), Person Centered (n=4), Psychodynamic (n=3), and Systemic (n=15). Some therapist respondents reported several theoretical orientations. In order to arrive at a single orientation per therapist and so make orientation a meaningful category of comparison, only the first descriptor entered by the therapist was used. In cases where three or more theoretical leanings were entered by the therapist, he/she was automatically categorized as eclectic. Therapists ranged in therapeutic experience as follows: 1 to 5 years (n=14), 6 to 10 years (n=8), 11 to 15 years (n=8), over 15 years (n=10). All other therapist demographic information collected is arranged in table form and presented in Appendix 4. Therapist demographics: Phase 2. A total of 10 therapists participated in Phase 2 (Female=7, Male =3). Therapists ranged in age from 29 to 62 years, the mean age being 40 (SD 10.41). All therapists participating in Phase 2 reported their professional affiliation in counsellor category. Therapists held degrees in the following areas: M.A. (n=6), M.Ed. (n=3), and Ph.D. (n=l). Therapist theoretical orientations were: Eclectic (n=, Person Centered (n=l), Psychodynamic (n=l), and Systemic (n=7). Therapists ranged in therapeutic experience as follows: 1 to 5 years (n=7), 6 to 10 years (n=2), 11 to 15 years (n=l). With respect to marital status, 3 therapists were divorced, 2 were living with partners, 3 were married, 1 was married but separated from her spouse, and 1 was single and had never been married. Almost all therapists reported that they were raised by both of their natural parents (n=9), and 1 therapist reported being raised by her mother and grandmother. With respect to the predominant ethnic/cultural influence in their family of origin, most therapists (n=7)  reported a British/Canadian influence. Of the remaining therapists, 2 reported a Jewish/Canadian influence and 1 therapist reported a Canadian influence. Client demographics. A total of 30 clients participated in the second phase of the study. Client age ranged from 19 to 56 years, the mean age being 37(SD 9.19). The client group included 17 females (57%) and 13 males (43%). Almost half of the clients (14) were single. The educational level of clients varied with 17 (57%) having completed some portion of college training; of these 8 (26%) were college graduates and 4 (13%) ad some graduate training. Of the remaining clients 7 (23%) were high school graduates, 5 (17%) had attended high school, and 1 (3%) had attended elementary school. Phase 1 Therapist attachment measures. The distribution for therapists who had identified themselves on the AAQ as having a Secure, Avoidant, or Anxious/Ambivalent attachment type, and each groups' mean AAS dimension scores appear in Table 2. Over 80% of the sample categorized themselves as Secure, with the remaining therapists almost equally represented in the Avoidant and Anxious/Ambivalent categories. The uneven distribution of attachment type in the current sample precludes any formal analysis. However, examination of means for the three types can provide some indication of the relationship between the measures. Visual examination of therapists' responses to the attachment measures suggests that therapists who chose different AAQ attachment descriptions also differed on their mean AAS dimension scores. Therapists who identified themselves as Secure on the AAQ had the highest mean scores on the AAS Close and Depend dimensions (25.27, SD 3.05 and 24.18, SD 3.30, respectively) compared to therapists who identified themselves as Avoidant (18.25, SD 3.50 and 16.50, SD 4.04) or Anxious/Ambivalent (17.00, SD 3.60 and 15.00, SD 5.19). Therapists who selected the Avoidant attachment type also obtained their highest scores on the Close and Depend dimensions, but their scores appear to be different from those produced by the Secure group and somewhat similar to the scores produced by the Anxious/Ambivalent group.  33 Therapists who selected the Anxious/Ambivalent attachment type showed the highest Anxiety dimension scores of the sample (19.33, SD 6.80) and the lowest scores on the Close and Depend dimensions.  Table 2 Phase 1 Therapists: Comparing Mean Adult Attachment Scale Scores and Adult Attachment Questionnaire Distribution  n  Attachment Style Secure Avoidant 33 (82%) 4 (10%)  Anx/Amb* 3 (7.50%)  Close  25.27 SD 3.05  18.25 SD3.50  17.00 SD3.60  Depend  24.18 SD 3.30  16.50 SD4.04  15.00 SD 5.19  Anxiety  10.69 SD 3.39  14.00 SD 2.94  19.33 SD 6.80  *Anxious/Ambivalent  Phase 2 Therapist attachment measures. The distribution for therapists who had identified themselves on the AAQ as having a Secure, Avoidant, or Anxious/Ambivalent attachment type, and each group's mean AAS dimension scores appear in Table 3. A total of 80% of the sample categorized themselves as Secure, and 20% categorized themselves as Anxious/Ambivalent. It should be noted that the Avoidant category was not represented in Phase 2 of the study. As was the case in Phase 1 of the study, the distribution of attachment type in the sample is unequal, and thus prevents formal analysis. However, visual examination of means for attachment type does provide some insight into the relationship between the measures. Therapists' responses to the attachment measures suggests that therapists who selected different AAQ attachment type descriptions also obtained different mean  34 AAS dimension scores. Therapists who identified themselves as Secure on the AAQ had higher mean scores on the AAS Close and Depend dimensions (25.87, SD 3.39 and 24.37, SD 3.11, respectively) than therapists who identified themselves as Anxious/Ambivalent (19.00, SD 1.14 and 18.00, SD 0.001). Therapists who selected the Anxious/Ambivalent attachment type showed higher Anxiety dimension scores (15.50, SD 2.12) than therapists who selected the Secure description (10.37, SD 1.99).  Table 3 Phase 2 Therapists: Comparing Mean Adult Attachment Scale Scores and Adult Attachment Questionnaire Distribution Attachment Style Secure Anx/Amb* 8 2 (80%) (20.00%) Close  25.87 SD3.39  19.00 SD 1.14  Depend  24.37 SD3.11  18.00 SD 0.001  10.37 15.50 SD 2.12 SD 1.99 * Anxious/Ambivalent Note. Therapists reporting an Avoidant type were not represented in Phase 2 of the study. Anxiety  AAS dimensions and gender. Analysis of Therapists' scores for the AAS dimensions of Close, Depend, and Anxiety was not justified due to unequal cell sizes in the study (27 Females and 13 Males). However, a visual examination of therapists' responses does not suggest the presence of any gender differences in therapist scores. The mean scores of male therapists appear to be slightly higher on the Close dimension (24.38, SD 5.06) than the mean scores of female therapists (23.74,  SD 3.84). The mean scores of female therapists on the Depend (22.89, SD 4.99) and Anxiety (11.74, SD 4.94) dimensions were slightly higher than the mean scores for male therapists (22.38, SD 4.13, and 11.54, SD 2.60 respectively), although the size of the standard deviations suggests that these small differences are most likely as a result of sampling. Therapist WAI-S ratings. A correlation matrix was constructed to examine the strength of relationship among therapist WAI-S subscale ratings. Significant correlations were found among all therapist WAI-S subscale ratings (see Table 4). Therapist Task ratings were significantly correlated with the Bond subscale (r=.78, p=.001), the Goal subscale (r=.73, p=.001), and the Combined therapist rating (r=.93, p=.001). Therapist Bond ratings were significantly correlated with the Goal subscale (r=.58, p=.001), and the Combined therapist rating (r=.88, p=.001). Therapist Goal ratings were significantly correlated with the Combined therapist rating (r=86, p=.001). WAI-S subscale intercorrelations are very high and support previous findings of WAI subscale intercorrelations (Adler, 1988; Horvath & Greenberg, 1989).  Table 4 Intercorrelation of Therapist* WAI-S Ratings WAIt WAIt Task Bond WAIt Task  WAIt Goal  WAIt Comb  1.00 p.=.001  WAIt Bond  .78 p=.001  WAIt Goal  E=001  .58 p_=001  WAIt Comb  .93 p_=.001  .88  .86  E=001  E=.001  .73  36 Client WAI-S ratings. A correlation matrix was also constructed to examine the strength of relationship among client subscale ratings. As was the case with therapist ratings, significant correlations were found among all client WAI-S subscale ratings (see Table 5).  Table 5 Intercorrelation of Client* WAI-S Ratings WAIc Task  WAIc Bond  WAIc Task WAIc .73 Bond p_=001 WAIc .82 .73 Goal E=.001 p_=001 WAIc .93 .89 Comb p=.001 p_=001 *Based on client ratings n=30.  WAIc Goal  WAIc Comb  .93 p_=001  Client Task ratings were significantly correlated with the Bond subscale (r=.73, p=001), the Goal subscale (r=.82, p=001), and the Combined client rating (r=.93, p=001). Client Bond ratings were significantly correlated with the Goal subscale (r=.73, p=001), and the Combined client rating (r=.89, p=001). Client Goal ratings were significantly correlated with the Combined client rating (r=.93, p=.001). Again, as expected, WAI-S subscale intercorrelations were very high and again support previous findings of WAI subscale intercorrelations (Adler, 1988; Horvath & Greenberg, 1989). Therapist and client WAI-S ratings. A correlation matrix was constructed to examine the strength of relationship between therapist and client self-reported strength of the WAI-St/WAI-Sc (see Table 6).  37 Table 6 Correlation of Therapist WAI-S And Client WAI-S Ratings WAIc Task  WAIc Bond  WAIc Goal  WAIc Comb  WAIt Task  .44 p_=014  .29 E=115  .38 E=.038  .40 E=.025  WAIt Bond  .57 E=001  .47 E=.008  .44 E=.015  .54 E=.002  WAIt Goal  .51 E=.004  .27 E=.141  .41 E=024  .43 E=.015  .38 WAIt .57 .45 £=.001 E=.034 Comb E=.011 Based on the ratings of Phase 2 therapists (n=10), and their clients (n=30).  .51 E=.003  Significant positive correlations were found between client and therapist ratings of the Task component (r=.44, E = 014), the Bond component (r=.47, E = 0 0 8 ) , the Goal component (r=.41, E=024), and client/therapist Combined ratings (r=.51, p=003). Significant correlations were also found between: client Task ratings and therapist ratings of Bond (r=.57, p=-001), Goal (r.51, E=004), and therapist Combined ratings (r=.57, E = 0 0 1 ) ; client Goal ratings and therapist ratings of Task (r^.40, p=025), Bond (r=.54, E=0°2), and therapist Combined ratings (r=.45, E=°15); client Combined ratings and therapist ratings of Task (r=.40, E = 0 2 5 ) , Bond (r=.54, E = 0 0 2 ) , and Goal (r=.43, E = 0 1 5 ) ; and client Bond ratings and therapist Combined ratings (r=.38, p = -034). Client attachment and client WAI-S ratings. The distribution for clients who had identified themselves on the AAQ as having a Secure, Avoidant, or Anxious/Ambivalent attachment type, and each groups mean WAI-Sc ratings appear in Table 7. A total of 10 (33%) clients selected the Secure attachment description, 17 (57%) selected the Avoidant description, and 3 (10%) chose the Anxious/Ambivalent description.  38 Table 7 Comparing Mean Client WAI-S Ratings and Adult Attachment Questionnaire Distribution Attachment Style Avoidant 17 (57%)  WAI-Sc n  Secure 10 (33%)  Task  23.20 SD2.25  23.29 SD 5.30  22.33 SD 1.52  Bond  23.40 SD3.94  23.23 SD4.03  21.66 SD 1.52  Goal  23.90 SD2.92  22.29 SD4.80  21.00 SD2.64  68.82 SD 13.59  65.00 SD 3.60  Comb  70.50 SD 7.39 *Anxious/Ambivalent  Anx/Amb* 3 (10%)  The WAI-Sc ratings of clients who were identified as either Secure or Avoidant appear to be quite similar. Clients identified as having an Anxious/Ambivalent attachment had the lowest ratings on all WAI-S components. Therapist attachment and therapist/client WAI-S ratings. Correlation matrices were compiled to examine the strength of relationship between: (a) therapist scores on the Adult Attachment Scale (AAS) dimensions of Close, Depend, and Anxiety, and therapist self-reported strength of the WAISt (see Table 8), and (b) therapist scores on the Adult Attachment Scale (AAS) dimensions of Close, Depend, and Anxiety, and client self-reported strength of the WAI-Sc (see Table 9). No significant correlations were found in relation to therapists' scores on any of the AAS dimensions and client or therapist self-reported strength of WAI-S.  39 Table 8 Correlation Between Therapist AAS* Scores and Mean Therapist WAI-S Ratings (n=10) AAS Close  AAS Depend  AAS Anxiety  WAIt Task  WAIt Bond  WAIt Goal  Close Depend  .50 p_=004  Anxiety  -.58 E=001  -.47 E=.008  Task  -.07 p-,70  -.12 E=.52  .02 p_=.90  Bond  .24 p-18  .10 E=.58  -18 E=33  .78 E=.001  Goal  .05 p=76  .07 p=.67  .01 Q=.95  .73 E=001  .58 E=.001  .08 E=64  .02 p_=.90  -.05 p=.77  .93 E=.001  .88 E=.001  Comb  .86 E=.001  Scale  Phase 2 respondents scored marginally higher on the Close (24.50, SD 4.20) and Depend (23.10, SD 3.84) dimensions than Phase 1 only respondents (Close 23.77, SD 4.28, Depend 22.60, SD 4.99) and Pilot Study respondents (Close 23.29, SD 5.41, Depend 18.86, SD 6.24). Phase 2 respondents scored somewhat lower on the Anxiety dimension (11.40, SD 2.88) than Phase 1 only respondents (11.77, SD 4.70) and Pilot Study respondents (12.71, SD 4.87). Although AAS scores across phases appeared similar for the variables of Close and Anxiety, visual inspection revealed a possible difference on the Depend dimension, with phase means of 18.86, SD 6.24 for Pilot Study respondents, 22.60, SD 4.99 for Phase 1 only respondents, and 23.10, SD 3.84 for Phase 2 respondents.  Table 9 Correlation Between Therapist AAS Scores and Mean Client WAI-S Ratings (n=30) AAS* Close  AAS Depend  AAS Anxiety  WAIc Task  WAIc Bond  WAIc Goal  Close Depend  .50 p_=.004  Anxiety  -.58 p_=001  -.47 p=.008  Task  .15 E=40  .04 E=79  -.03 E=.87  Bond  .10 p_=59  .29 p_=.ll  -20 E=.28  .73 E=.001  Goal  -.04 E=.81  .10 p=.57  .13 E=.46  .82 E=.001  .73 E=001  Comb  .07 p=68  .16 E=.39  -.02 E=.87  .93 E=-001  .89 E=.001  * Adult Attachment Scale  Table 10 Comparing Mean Therapist AAS* Scores Across Phases Pilot Study n=14  Phase Phase 1 n=30  Phase 2 n=10  Close  23.29 SD5.41  23.77 SD4.28  24.50 SD4.20  Depend  18.86 SD6.24  22.60 SD4.99  23.10 SD3.84  Anxiety  12.71 SD 4.87  11.77 SD 4.70  11.40 SD 2.88  .93 E=.00  41 AAS scores across phases. Given the skewed distribution of attachment types across Pilot Study respondents, Phase 1 only respondents (i.e., respondents who participated in Phase 1 of the study but declined to participate in Phase 2 of the study), and Phase 2 respondents, means were calculated across phases for the three dimensions of Close, Depend, and Anxiety (see Table 10). As a result of this observation, a series of t-tests comparing respondents across phases (i.e., Pilot vs. Phase 1, Pilot vs. Phase 1 only, Pilot vs. Phase 2, Phase 1 only vs. Phase 2) on Adult Attachment Scale scores were undertaken to explore the area more fully. It should be noted at this point, that the unequal cell sizes and range of standard deviations in this area require that the results be viewed as tentative and exploratory. Significant differences were found on the Depend dimension of the Adult Attachment Scale between Pilot Study respondents and all Phase 1 respondents (t=2.12, p_=048). Pilot Study respondents obtained lower mean scores on Depend than either Phase 1 or Phase 2 participants. No significant differences were found between Pilot Study respondents and Phase 1 only respondents or between Phase 1 only respondents and Phase 2 respondents on the Close and Anxiety dimensions. Given that no differences were found on the Close or Anxiety dimensions, and the findings for the Depend dimension are tentative, the Phase 2 therapists can be cautiously viewed as representative of the larger group on this variable. Therapist WAI-S ratings and client WAI-S Bond ratings. A series of paired t-tests were employed to examine client WAI-Sc Bond ratings in relation to therapist WAI-St Task, Bond, Goal, and WAI-S Combined ratings. Client WAI-Sc Bond ratings were ranked (1 to 30) and grouped as either High Bond (top 15) or Low Bond (bottom 15). A t-test was then employed to examine the mean WAI-S Bond ratings of the High (26.47, SD 1.55) and Low (19.80, SD 1.82) Bond client groupings. A significant difference was found between the WAI-S Bond ratings of clients grouped as High Bond and the ratings of clients grouped as Low Bond (t=-10.79, p=0.001) and supports further examination of the two groups.  42 The WAI-St ratings of therapists (n=10) whose clients had been grouped as High Bond were then separated from the WAI-St ratings of therapists whose clients had been grouped as Low Bond and their mean WAI-St ratings were compared. Although the ideal analysis would have employed a nested design, the sample size precluded this approach. The current analysis was conducted as a preliminary exploration of patterns that could be predicted from previous research and theory. Therapists whose clients were grouped as High Bond had higher mean ratings on all components of the WAI-St: Task 22.67, SD 2.06; Bond 24.80, SD 1.70; Goal 23.53, SD 2.42; Combined 71.00, SD 5.21, compared to the mean ratings of therapists whose clients were grouped as Low Bond: Task 21.33, SD 2.99; Bond 22.40, SD 2.85; Goal 22.27, SD 2.71; Combined 66.00, SD 7.83 (see Appendix 5 for raw scores). Significant differences were found between the mean WAI-St Bond ratings of therapists whose clients were in the High Bond group and the mean WAI-St Bond ratings of therapists whose clients were in the Low Bond group (t=-2.80, g=0.01). Significant differences were also found in relation to comparisons of the mean WAI-St Combined ratings of therapists whose clients were in the High Bond group and the mean WAI-St Combined ratings of therapists whose clients were in the Low Bond group (t=-2.06, p_=0.05). No significant differences were found on the Task or Goal components when the mean WAI-St ratings of therapists whose clients were in the High Bond group were compared to the mean WAI-St ratings of therapists whose clients were in the Low Bond group (see Table 11). Therapist mean AAS scores and client WAI-S Bond ratings. A series of paired t-tests was employed to examine client WAI-Sc Bond ratings in relation to therapist AAS scores. Client WAISc Bond ratings were ranked (1 to 30) and grouped as either High Bond (top 15) or Low Bond (bottom 15). The AAS scores of therapists whose clients had been grouped as High Bond were then separated from the AAS scores of therapists whose clients had been grouped as Low Bond and their mean AAS scores were compared.  43  Table 11 Comparing High and Low Therapist Mean Working Alliance Ratings Working Alliance WAI-S  WAI-S High n=15  WAI-S Low n=15  Task  22.67 SD2.06  21.33 SD2.99  -1.42  0.17  Bond  24.80 SD 1.15  22.40 SD4.54  -2.80  0.01*  Goal  23.53 SD 1.15  22.27 SD3.13  •1.35  0.19  Comb  71.00 SD 5.21  66.00 SD 7.83  -2.06  0.05<  Therapists whose clients were grouped as High Bond had higher AAS mean scores on both the Close dimension (25.07, SD 3.65) and the Depend dimension (24.53, SD 3.74) compared with the corresponding AAS mean scores of therapists whose clients were grouped as Low Bond (23.93, SD 4.46 and 21.67, SD 3.18, respectively). Therapists whose clients were grouped as Low Bond had higher AAS mean scores on the Anxiety dimension (12.33, SD 3.20), compared with the corresponding AAS mean Anxiety scores of therapists whose clients were grouped as High Bond (10.47, SD 1.96). A significant difference was found between the mean AAS Depend dimension scores of therapists whose clients were in the High Bond group and the mean AAS Depend dimension scores of therapists whose clients were in the Low Bond group (t=-2.26, p_=0.032). No significant differences were found on the Close or Anxiety AAS dimensions when the mean AAS scores of therapists whose clients were in the High Bond group were compared to the mean AAS scores of therapists whose clients were in the Low Bond group (see Table 12).  44 Table 12 Comparing High And Low Therapist Mean Attachment (AAS*) Scores Working Alliance WAI-S High WAI-S Low n=15 n=15  t  p_  Close  25.07 SD 3.65  23.93 SD 4.46  -0.76  0.45  Depend  24.53 SD3.74  21.67 SD3.18  -2.26  0.032*  10.47 SD 1.96 * Adult Attachment Scale  12.33 SD3.20  1.93  0.066  Anxiety  Therapist median AAS scores and client WAI-S Bond ratings. As a follow up to the significant difference found between mean therapist scores on the Depend dimension of the AAS when clients were grouped according to High and Low WAI-S Bond ratings, a Mann-Whitney nonparametric comparison of group medians was carried out. The results indicate that therapists whose clients were grouped as High Bond (n=15) had higher AAS median scores on the Close (25.00) and Depend (25.00) dimensions compared to the corresponding AAS mean scores of therapists whose clients were grouped as Low Bond (n=15) (23.00 and 21.00 respectively). Therapists whose clients were grouped as Low Bond had higher AAS median scores on the Anxiety dimension (12.00) compared to the corresponding AAS median Anxiety scores of therapists whose clients were grouped as High Bond (11.00). A significant difference was found with respect to therapist scores on the Depend dimension of the AAS and matched client WAI-Sc Bond ratings (w=184.5, p_=0.05). No significant differences were found between median therapist scores on the Close (w=214.5, p_=0.46) or Anxiety (w=270.0, p=0.12) dimensions of the AAS and matched client WAI-Sc Bond ratings (see Table 13).  45 Table 13 Comparing High And Low Therapist Median AAS* Dimension Scores Therapist Adult Attachment Scale Scores High n=15 Close  25.00  Depend  25.00  Anxiety  11.00  *Adult Attachment Scale  Low  w  n=15 23.00  214.5  0.46  21.00  184.5  0.05*  12.00  270.0  0.12  46 Chapter Five: Discussion  Discussion of the study will be broken down into four sections: Rationale; Attachment; Working Alliance; and Attachment and the Working Alliance. The first section briefly reiterates the basis for undertaking the study. The second section looks at previous research findings relating to attachment distribution patterns, discusses these findings in relation to the different attachment distributions found in the present study and presents potential explanations to account for these differences. This section also discusses gender and attachment, re-examines the overall attachment findings of the study, looks at measurement of attachment in general, and speculates about recently proposed changes to the measurement of attachment and the implications of these changes for the present study. The third section discusses the WAI-S ratings gathered in the present study in relation to previous findings. This section also looks at the differences between therapists and clients who reported higher WAI-S Bonds and therapists and clients who reported lower WAI-S Bonds, as well as presenting a possible explanation to account for these differences. The fourth section is a general review of the main points of the study and the major findings. Rationale The major focus of the study was to explore the relationship between therapist attachment style and therapist standing on the attachment dimensions of Close, Depend, and Anxiety on client and therapist reported strength of the Bond component of the Working Alliance Inventory. Previous research has demonstrated that client pretherapy relationships are predictive of alliance strength (Kokotovic & Tracey, 1990; Moras & Strupp, 1982), that clients' attachment histories are significantly related to alliance ratings (Mallinckrodt, 1990), and that attachment information may directly mediate individuals' ability to foster positive relational bonds with others (Kestenbaum et  47 al., 1989; Koestner et al., 1990). Given these findings it was reasoned that therapists' abilities to foster alliance formation would be related to their personal attachment perspectives. Therapist Attachment It was originally planned to study therapist attachment by separating therapists according to their attachment type using Hazan and Shaver's (1987) Adult Attachment Questionnaire (AAQ), compiling relevant therapist demographic information, and then examining this information in relation to clients' ratings of the Working Alliance Inventory Short Form (WAI-Sc). Previous research (Feeney & Noller, 1990; Hazan & Shaver, 1987, 1990; Levy & Davis, 1988; Mikulincer et al., 1990) has found that responses to the AAQ consistently show similar distributions of responses with approximately 50% of respondents selecting the secure attachment description and the remaining 50% of respondents being almost evenly split between the avoidant and anxious/ambivalent descriptions. Prior to the present study, a pilot study involving counselling psychology students at the University of British Columbia resulted in a similar distribution of attachment types. However, the distribution of attachment types found in the main study failed to match the expected distribution. Of the 40 therapists who responded to the AAQ, 82.5% chose the secure description, 10% chose the avoidant description, and 7.5% chose the anxious/ambivalent description. There are several possible explanations for this disparity. Most therapists may be securely attached. Research on attachment suggests that securely attached individuals are more caring, supportive, and understanding in relationships (Collins & Read, 1990; Feeney & Noller, 1990; Hazan & Shaver, 1990; Levy & Davis, 1988) and might therefore be more likely to enter careers that have a strong relational component. On the other hand, Avoidant and Anxious/ambivalent individuals have been found to have more negative and mistrustful views of the social world and of human nature in general (Collins & Read, 1990). Indeed, avoidant individuals reportedly use work situations to avoid social interaction (Hazan &  48 Shaver, 1990) and therefore would be unlikely candidates for careers in counselling. Another possible explanation is that non-secure therapists do not remain in the therapeutic field. The data from the pilot study carried out prior to the main study found that almost 43% of respondents, who were engaged in counsellor training, identified themselves as having insecure attachments. However, these respondents, although advanced in their studies, were students and, short of a longitudinal study, there is no way of ascertaining whether they will remain active in the field of counselling, whether they will be successful in building a practice, or whether they will transfer to careers in other areas. A third possible explanation is that therapists did not represent themselves accurately. The attachment descriptions used in the Hazan and Shaver questionnaire are somewhat transparent and respondents may have chosen the description that seemed the most socially desirable. This point is particularly applicable to therapist respondents who have presumably been exposed to a variety of testing procedures and test interpretation. In a similar vein, it may be assumed that therapists are at least aware of the attributes associated with mental health and perhaps even familiar with Bowlby's work in the field of attachment. From a social desirability standpoint, the Secure attachment description is quite obviously the most positive of the three options. Although an insecure attachment orientation may have been the most positive option for a given individual in an adaptive sense (e.g., in response to an intrusive or rejecting caregiver), openness, and self-confidence are more highly valued social attributes in Western culture. Given the occupation of the respondents in question, it may be the case that insecurely attached therapists felt threatened by the prospect of reporting their true feelings. Of course, the sample in the study may not be representative of the therapist population as a whole. Low response rates to the request to participate in the study required that all therapists who agreed to participate be included in the study. This increased the likelihood that self-selection bias occurred, and therefore, the possibility that insecurely attached therapists may have chosen not to  49 participate in the study cannot be discounted. As far as I am aware, the present study is the first study to examine therapists' responses to measures of attachment and as a result, further research will be required to determine whether therapists as a group are by and large securely attached. Because the distribution of categorical attachment types failed to resemble patterns found in previous research it was decided that, although the AAQ responses gathered from therapists would be reported, the study would focus on therapists' continuous scores on the Adult Attachment Scale and how these scores relate to client alliance ratings. Although the sample size is small and therefore requires that caution be used in interpreting results, findings from the study represent a preliminary exploration of the relationship between attachment and alliance bond formation that may uncover areas for future research of a more substantial nature. Client Attachment The distributions of client responses to the AAQ in the present study, as was the case with therapist respondents, did not resemble patterns found in previous research (Feeney & Noller, 1990; Hazan & Shaver, 1987, 1990; Levy & Davis, 1988; Mikulincer et al., 1990). The majority of clients (n=17) selected the Avoidant attachment description, a total of 10 selected the Secure description, and 3 chose the Anxious/Ambivalent description. This finding is surprising not only in that clients' responses to the AAQ do not match distributions regularly found in other studies, but also with respect to the high proportion of self-reportedly Avoidant clients. Previous attachment research has found that Avoidant individuals are uncomfortable being close to others, find it difficult to trust or depend on others, become uncomfortable when others get too close to them (Hazan & Shaver, 1987; Simpson, 1990) and are unwilling to make intimate selfdisclosures (Mikulincer & Nachshon, 1991). Avoidant individuals have also been found to repress or deny feelings of distress, to inhibit the display of negative emotion (Kobak & Sceery, 1988; Main et al., 1985), and to show what Parkes (1973) calls compulsive self-reliance. Given the consistency of earlier findings, it is puzzling to find the Avoidant attachment type so substantially represented in  50 the client sample. If, as previous research suggests, avoidant individuals invest so much energy in either denying or repressing negative affect, then logically we would expect to find the Avoidant attachment type under-represented in our sample. One possible explanation may lie in the activation of the attachment system. According to Bowlby (1969, 1973, 1979, 1980, 1988) the attachment behavioral system has evolved to protect infants from danger by keeping them close to a caregiver. Working or mental models of the caregiver and the self are said to result from continued interactions between the child and his/her caregiver with individual differences in attachment being accounted for by the availability of the caregiver, the caregiver's responsiveness as a safe haven, and the extent of the caregiver's reliability as a secure base. Although working models of attachment are resistant to change, environmental influences are said to calibrate the attachment system and so ensure that individual differences in attachment are due to accurate reflections of the accessibility and responsiveness of attachment figures (Bowlby, 1979, Shaver & Hazan, 1993). Recent research (Hazan, Hurt, & Marcus, 1991, cited in Shaver & Hazan, 1993) supports the position that mental models of attachment, although resistant to change, can be modified through "model disconfirming experiences" and that change is "more likely to occur in the direction of security than insecurity" (Shaver & Hazan, 1993, p.36). Therefore, it seems that initially insecure individuals can develop secure attachment styles as a result of positive environmental experiences. However, if progression towards secure attachment can be achieved through positive experiences, it seems equally likely that negative life events, such as divorce or death of a spouse, might, at least in the case of individuals who were insecurely attached at an earlier point in time, produce the opposite effect and result in a reversal of the process. With respect to the high representation of avoidantly attached individuals in the present study, it may be that the very reason(s) that have prompted them to reach out for assistance have produced a reversion to an underlying insecure organization.  51 Bowlby (1988) refers to this point in stating that "although the capacity for developmental change diminishes with age, change continues throughout the life cycle so that changes for the better or worse are always possible. It is this continuing potential for change that means that at no time of life is a person invulnerable to every possible adversity and also that at no time of life is a person impermeable to favourable influence. It is this persisting potential for change that gives opportunity for effective therapy" (Bowlby, 1988, p. 136). Although a secure attachment organization has been found to be predictive of positive features such as relationship satisfaction, and constructive conflict resolution strategies (Levy & Davis, 1988), having a secure attachment does not eliminate the likelihood of experiencing negative life events (Sroufe, 1988). Therefore, although insecurely attached individuals might be expected to be heavily represented in client populations, individuals who, under normal circumstances, are securely attached can also be expected to seek therapeutic assistance, largely in response to a crisis (Sroufe, 1988). Whether the distribution of client attachment types found in the present study are representative of clients in general is a moot point. Further research is required to examine the flexibility/stability of attachment orientation and the extent to which clients' attachment types follow or do not follow a general pattern of distribution. AAS Dimensions and Gender A visual examination of male and female therapist scores on the AAS dimensions do not suggest the presence of any gender differences. Comparisons between the scores of all male and female therapists who participated in Phase 1 of the study (n=40) show little variation. Male therapists obtained higher scores than female respondents on the Close dimension, whereas female therapists scored marginally higher than male therapists on the Depend dimension. Female therapists also obtained slightly higher scores on the Anxiety dimension than the scores obtained by the male therapists. The results of male/female comparisons on the AAS dimensions are somewhat similar to those found by Collins and Read (1990). Collins and Read (Study 1) reported that male respondents  52 scored higher than female respondents on the Close dimension of the AAS and that female respondents (Study 3) scored higher than male respondents on the AAS Anxiety dimension. However, overall, studies of attachment (e.g., Hazan & Shaver, 1987, 1993; Mikulincer et al., 1990; Simpson, 1990) have found limited, if any, gender differences. Attachment Measures The Adult Attachment Scale (AAS, Collins & Read, 1990) was created in an attempt to overcome some of the perceived shortcomings of the Hazan and Shaver (1987) Adult Attachment Questionnaire (AAQ). In designing the AAS, Collins and Read broke down the categorical descriptions which comprise the AAQ and created a dimensional measure of attachment which, although different, is directly derived from the AAQ. The strength of relationship between the two measures is apparent in comparisons of therapists' responses to both measures of attachment. In the Pilot Study, respondents who selected the Secure attachment description obtained higher AAS scores on the Close and Depend dimensions compared to respondents who selected the Avoidant and Anxious/Ambivalent descriptions. In addition, Secure respondents obtained lower scores than Avoidant and Anxious/Ambivalent respondents on the AAS Anxiety dimension. Respondents who selected the Avoidant attachment description obtained the lowest scores on Close and Depend and scored marginally higher than Anxious/Ambivalent respondents on the Anxiety dimension. With respect to Phase 1 only respondents (i.e., therapists who participated in Phase 1 of the study but declined to participate in Phase 2 of the study, n-30), therapists who selected the Secure attachment description obtained higher scores on the Close and Depend dimensions compared to therapists who selected the insecure descriptions. In addition, Secure respondents obtained lower scores than Avoidant and Anxious/Ambivalent respondents on the AAS Anxiety dimension. Therapists who selected the Anxious/Ambivalent attachment description obtained the highest scores on the Anxiety dimension, and the lowest scores on the Close and Depend dimensions. In Phase 1 (n=40) responses, a similar pattern was apparent. Therapists who selected the  53 Secure attachment description again obtained the highest scores on the Close and Depend dimensions, and the lowest scores on the Anxiety dimension. Therapists who selected the Anxious/Ambivalent attachment description again obtained the lowest Close dimension scores and the highest Anxiety scores. Therapists who selected the Avoidant attachment description obtained the lowest Depend dimension scores. The Avoidant AAQ attachment description was not represented in Phase 2 of the study thus making the results less meaningful. However, therapists who chose the Secure description obtained higher Close and Depend dimension scores and lower Anxiety scores than those therapists who selected the Anxious/Ambivalent description. In the present study, the consistent differences with respect to AAS dimension scores found among therapists with differing self-reported AAQ attachment types lends some support to the contention that individuals can meaningfully classify themselves with respect to attachment style. There are several obvious limitations inherent in the design of Hazan and Shaver's (1987) measure of attachment: first, the use of discrete categories demands that the three attachment types be regarded as mutually exclusive domains without the possibility of overlap between types; second, the discrete descriptions of attachment type include statements that refer to different aspects of relationships. For example, the Secure description includes both being able to depend on others and being comfortable with closeness and therefore may not accurately reflect the feelings of respondents; and last, the descriptions used to portray each of the attachment types are somewhat transparent and may therefore not be appropriate for experimental use with test-wise respondents. Recently, Hazan and Shaver (1993) have addressed some of these concerns by incorporating a 7point Likert into each of the AAQ descriptions thereby allowing respondents to assign a number value to each of the description and so indicate the extent of their agreement with each of the descriptions. The Collins and Read (1990) Adult Attachment Scale and its focus on analysis of attachment dimensions has very definite theoretical and practical advantages over the limited and  54 discrete analyses possible when using the Hazan and Shaver measure. However, as Collins and Read (1990) point out, it may be the case that a three-dimension attachment range is not sufficient to capture the nuances of adult attachment style. Indeed, Collins and Read (1990) found evidence to suggest that a four-cluster solution may be a more appropriate description of their sample, with the anxious group being divided in to anxious-secure and anxious-avoidant groups. Another approach to attachment measurement (Bartholomew & Horowitz, 1991) retains the Secure and Anxious/Ambivalent dimensions (although Anxious/Ambivalent is labelled Preoccupied) and divides the Avoidant dimension into Fearful (fearful of intimacy) and Dismissing (dismissing of intimacy). Recently, Shaver and Hazan (1993) state that general consensus now holds that two major dimensions underlie the three Hazan and Shaver (1987) discrete types, with one dimension ranging from high to low anxious/ambivalence and the second ranging from secure to avoidant. Although it is speculative at this point, a dimension running from secure to avoidant might account for the presence of so many avoidant clients in the present study. Theoretically, positive/negative life experiences are capable of altering the direction of attachment (Bowlby, 1988); thus, it may be that the avoidant clients in the present study recently functioned in a secure fashion but reverted to an avoidant attachment as a result of negative life events. In the same vein, a dimension running from low to high anxious/ambivalence might also account for the low numbers of anxious/ambivalent clients in the present study. Research indicates that the relationships of anxious/ambivalent individuals are characterized by low levels of interpersonal trust (Hazan & Shaver, 1987; Simpson, 1990) and that activation of the attachment system typically results in high levels of anger and frustration (Bowlby, 1988). It may be that the experience of anger coupled with a history of interpersonal mistrust decreases the likelihood of highly anxious/ambivalent individuals reaching out for assistance during periods of high stress. Working Alliance Client WAI-S ratings. Previous research (Kokotovic & Tracey, 1990; Mallinckrodt &  55 Nelson, 1991; Tichenor & Hill, 1989), found that client ratings of the alliance components were higher than ratings given by therapists. The results of the present study replicated these findings for the Bond and Goal subscales scores, however, client ratings of the Task subscale are slightly lower than therapists' ratings. In general, Client WAI-S subscale intercorrelations were high, ranging from .73 to .93 on the Task scale, from .73 to .89 on the Bond scale, from .73 to .93 on the Goal scale, and support earlier findings (Adler, 1988; Horvath & Greenberg, 1989). Research evidence suggests that clients' perception or experience of a positive working alliance is reliably related to therapeutic success (Hartley, 1984; Luborsky & Auerbach, 1985). In addition, client ratings of the working alliance have been found to be significantly related to positive therapeutic outcome (Horvath & Symonds, 1991; Tichenor & Hill, 1989). In the present study, significant client/therapist subscale correlations were found for almost all components of the WAI-S. However, although correlations between client and therapist ratings of WAI-S Bonds were significant, subscale correlations between client ratings of the WAI-S Bond component and therapist ratings of WAI-S Task and Goal components did not reach significance and thus may lend credence to the suggestion that clients view the working alliance differently than therapists (Mallinckrodt, 1991; Tryon & Kane, 1993). Research has suggested that early bonds formed with parents may function as an enduring template for subsequent relationships, particularly relationships involving social support and emotional intimacy (Mallinckrodt, 1991). Given that the working alliance also involves social support and emotional intimacy it may be that early bonds formed with parents also serve as a template for bonds formed within the working alliance. Relatedly, it has been suggested that some clients may enter therapy with an exaggerated need to develop bonds (Saunders, Howard, & Orlinsky, 1989) and as a result may be more focused on the bonding aspect of the alliance during the initial sessions. It may be, therefore, that clients' ratings of the WAI-S in the present study reflect a desire or need to develop bonds with a therapist who is regarded as a surrogate attachment  56 figure. As Tryon and Kane (1993) suggest, it may be that, for the client, the relationship developed with the therapist may be more positive than other relationships that he or she has experienced. Although the extent to which the personal characteristics of the therapist influence client ratings of the alliance will require further study, there can be little doubt that the client's perception of the therapist is directly related to the strength of the alliance. Therapist WAI-S ratings. Therapist WAI-S subscale intercorrelations were moderate to high, ranging from .73 to .98 on the Task component, from .58 to .88 on the Bond component, from .58 to .86 on the Goal component, and support previous findings of therapist WAI subscale intercorrelations (Adler, 1988; Horvath & Greenberg, 1989). Bordin (1990, cited in Mallinckrodt & Nelson, 1991) has suggested that high correlations among therapist ratings of working alliance components occur when therapists are adept at developing all three components of the alliance. Consequently, less experienced therapists would be expected to show greater variation in their ratings of the alliance components. Research supports this suggestion in that the ratings of experienced therapists have been found to more closely match client ratings of the alliance (Mallinckrodt & Nelson, 1991). Most therapists' level of experience in the present study fell into the 1 to 5 years category (n=7), and although overall therapist ratings for the Task component closely resemble previous results (e.g., Adler, 1988; Horvath & Greenberg, 1989) ratings for the Goal and Bond components are somewhat lower. Although these findings may be interpreted as supporting Bordin's (1990) position, an alternative explanation is also possible. In an examination of the working alliance and termination type (Tryon & Kane, 1993), researchers found that therapists' ratings of the strength of the alliance were predictive of the type of client termination. Tryon and Kane (1993) found that therapists' working alliance ratings with clients who would later terminate mutually (i.e., both therapist and client agree that termination is appropriate) were significantly stronger than therapists' alliance ratings with clients' who terminated unilaterally (i.e., the client ceased to attend scheduled sessions).  57 In the present study, significant differences were found between the WAI-S Bond ratings of therapists whose clients had high Bond ratings and the WAI-S Bond ratings of therapists whose clients had low Bond ratings. Although it is beyond the scope of the present study to examine whether or not terminations occurred, and if so which type of terminations were involved, it may be that the strength of reported client/therapist alliance Bond ratings and the level of agreement between the client and the therapist with respect to alliance Bonds is in some way predictive of client termination type. Given that the therapist/client bond is said to extend beyond the individuals' therapeutic roles and to encompass certain individual qualities of the relationship that form, or fail to form, between participants (Orlinsky & Howard, 1987), future longitudinal research in this area may prove to be fruitful. Therapist and client WAI-S ratings. Previous research findings with respect to the relationship between client and therapist ratings have varied somewhat. Although several studies have found substantial correlations between client and therapist ratings (Adler, 1988; Horvath & Greenberg, 1989), other researchers report ratings which, although statistically significant, are less robust (Mallinckrodt, 1991; Mallinckrodt & Nelson, 1991; Tichenor & Hill, 1989; Tryon & Kane, 1993). In the present study, despite a modest sample size, significant positive moderate to high correlations were found between: client and therapist ratings of working alliance Bonds; client Task ratings and therapist Bond ratings; and client Task ratings and therapist Goal ratings. In addition, significant positive correlations were found between client Task ratings and therapist overall Alliance ratings and between client overall Alliance ratings and therapist overall Alliance ratings. The significance of results with respect to client WAI-S intercorrelations, therapist WAI-S intercorrelations, and client/therapist WAI-S intercorrelations, especially in light of the strengths of relationship found, the size of the sample, and their similarity to previous findings using the original version of the Working Alliance Inventory, provide support and cross-validation for the validity of  58 the WAI-S and for its future use as a research tool. Therapist and client ratings of the WAI-S Bond component. In an effort to examine whether therapist WAI-S ratings were in any way predictive of client reported WAI-S Bonds, clients' ratings were separated into two groups. Client WAI-Sc Bond ratings were first ranked and grouped as either High Bond (top 15 ratings) or Low Bond (bottom 15 ratings). The WAI-St ratings of therapists whose clients had been grouped as High Bond were then separated from the WAI-St ratings of therapists whose clients had been grouped as Low Bond, and their mean WAI-St ratings were compared. The results of these groupings indicated that therapists whose clients were grouped as High Bond had higher mean ratings on all components of the WAI-St than therapists whose clients were grouped as Low Bond. Significant differences were found between the mean WAI-St Bond ratings of therapists whose clients were in the High Bond group and the mean WAI-St Bond ratings of therapists whose clients were in the Low Bond group. Weaker, although still significant, differences were also found in comparisons of the WAI-S combined ratings of therapists whose clients were in the High Bond group and the mean WAI-St combined ratings of therapists whose clients were in the Low Bond group. No significant differences were found on the Task or Goal components when the mean WAI-St ratings of therapists whose clients were in the High Bond group were compared to the mean WAI-St ratings of therapists whose clients were in the Low Bond group. The significant differences found in relation to the WAI-S combined ratings of therapists whose clients were in the High Bond group and the mean WAI-St combined ratings of therapists whose clients were in the Low Bond group is most probably attributable to the significantly higher Bond subscale ratings of these therapists in combination with their slightly higher Task and Goal subscale ratings.  59 The significant differences in the WAI-S Bond ratings of therapists whose clients were in the High Bond group and the mean WAI-St Bond ratings of therapists whose clients were in the Low Bond group are less readily explained. Although high significant correlations were found across all therapist and client WAI-S ratings, the significant differences found with respect to the High/Low Bond groupings may be interpreted to suggest that the High Bond therapist/client dyads in some way constitute a separate and distinct group compared to other therapist/client pairings. Although this interpretation rests on an assumption of homogeneity among clients associated with a particular therapists, which could not be investigated within the current design, and while the possibility that therapists in the High Bond group intentionally selected clients with whom they had already developed a positive relationship to participate in the study cannot be discounted, it seems equally possible that the High Bond dyads formed strong relational bonds either immediately, or during the three sessions prior to completing the WAI-S. The notion of an immediate sense of mutual liking between the client and therapist is not new to therapeutic literature. Luborsky (1971) characterized a positive therapeutic relationship as a "clicking" between the therapist and client and stated that this phenomenon was indicative of successful therapeutic outcome. Using friendship as an analogy, Highlen and Hill (1984) state that a positive therapeutic relationship is similar to a strong friendship in that the relationship is determined by the two individuals involved. However, they suggest that the strength of the relationship, at least in the initial stages, may be influenced by the degree to which two individuals happen to "click" upon meeting. Similarly, Gelso and Carter (1985) in their examination of the components of working alliance bonds refer to the real relationship bond as a mutual liking that contributes to "clicking" between participants. They go on to state that although the strength of bonds between the therapist and client will develop gradually, an initial bonding or clicking, although lacking the depth of later bonds, is a contributing factor to successful therapeutic outcome.  60 Although the significance of the findings in the present study relating to the High Bond therapist/client dyads is suggestive of a "clicking" between these individuals, interpretation of these results must only be viewed as speculation. Longitudinal research using larger sample sizes is required to examine this area in a thorough and empirical manner. Attachment and the Working Alliance The main research question posed in the study asked whether therapist self-reported attachment information, in general, was related to the strength of client working alliance bond ratings. No evidence was found to support this position. Visual examination of the therapist AAQ attachment types and client ratings of WAI-S components showed no pattern of relationship. However, given that 8 of the 10 therapists in Phase 2 of the study chose the secure attachment description, this is hardly surprising. A meaningful comparison of these relationships would require a larger sample of therapists and a more even distribution of attachment types. As an adjunct to the comparison between therapist AAQ type selection and client WAI-S ratings, comparisons were made between client AAQ type selection and client ratings of WAI-S components. Again no meaningful relationships were found. Although previous research (Mallinckrodt, 1991) found a significant relationship between client ratings of early parental bonds and working alliance ratings, the relationship was limited to therapist-rated working alliance. No significant correlations were found between client representations of parental bonds and client ratings of the working alliance, thus suggesting that clients and therapists may be using different bases in forming their perceptions of the working alliance. The existence of a possible link between therapist scores on the Adult Attachment Scale (AAS) dimensions of Close, Depend, and Anxiety and client reported strength of WAI-S Bonds received partial support. Although examinations of the relationship between therapist scores on the AAS and client WAI-S ratings, and between therapist scores on the AAS and therapist WAI-S  61 ratings produced no significant relationships, when clients were split into two groups based on their WAI-S Bond ratings, it appears that the therapists of clients with high WAI-S Bond ratings obtained significantly higher scores on the Depend dimension compared to the therapists of clients with low WAI-S Bond ratings. These findings seem to suggest that therapists who scored high on the Depend dimension of the AAS may in fact foster stronger working alliance bonds with their clients. The Depend dimension of the AAS can be seen as indicative not only of the extent to which individuals feel comfortable depending on others, but also of the extent to which individuals feel comfortable having others depend on them (Simpson, 1990). However, the positive association between therapists high on Depend and clients high on WAI-S Bond can be accounted for in several ways. It may be that therapists who are high on Depend are able to communicate that they are comfortable having their clients depend on them and this is conducive to stronger client bonding. Alternatively, it may be that the clients themselves are able to discern the therapists' openness to being depended on and this is conducive to higher client bonding. Last, it may be that both factors interact and result in higher levels of client bonding. Although these findings are consistent with theoretical predictions, the limited sample size in the present study demands that caution be used in interpretations of the results. Conclusions Theory (Gelso & Carter, 1985) and research (Moras & Strupp, 1982) suggest that a client's ability to form a sound working alliance is to some degree mediated by his or her history of interpersonal relationships and capacity to form productive attachments to others. Relatedly, Bordin (1979) states that a strong working alliance is established mutually between the client and the therapist. It follows, therefore, that the characteristics that each participant brings into the relationship will affect the strength of the alliance. A growing body of research suggests that children who have secure attachment histories are  62 more likely to develop a sense of trust and to identify with caregivers. By having their own emotional needs met through interaction with caregivers who respond sensitively and empathetically, they develop the capacity to relate in a like manner in later relationships. The implications of this line of research for the therapeutic field are self-evident. If the ability to be empathetic is largely or partly a result of early attachment experiences, then it follows logically that the ability to form bonds within the context of therapy might also rely to some extent on these experiences. As Silver (1984) points out, "the analyst's own personal background and previous experience are obviously antecedents to his empathic response to an analysand" (p. 299). In delineating the practitioner's role in therapy, Bowlby (1988) identifies five therapeutic tasks. The first of these is described as providing the client with a "secure base" from which to "explore and express his thoughts and feelings" (p. 140). Bowlby likens the role of the therapist at this point to that of the mother who provides a secure base from which the child may explore his or her world. Specifically, Bowlby recommends that in order for therapy to begin, the therapist must strive "to be reliable, attentive, and sympathetically responsive to his patient's explorations and, so far as he can, to see and feel the world through his patient's eyes" (p. 140). However, Bowlby also cautions that "a patient's way of construing his relationship with his therapist is not determined solely by the patient's history: it is determined no less by the way the therapist treats him. Thus the therapist must strive always to be aware of the nature of his own contribution to the relationship which, amongst other influences, is likely to reflect in one way or another what he experienced himself during his own childhood" (p. 141). In summary, a therapist's ability to form a sound working alliance is to some extent affected by the same factors that influence a clients' ability to form a sound working alliance. The present study contributes to the existing body of literature on the therapeutic relationship; the study has examined the nature of the therapist's contribution to the relationship by exploring the impact of therapist attachment characteristics on the development of the bond  63 component of the working alliance. The results of the study do not support the position that therapist attachment characteristics, in general, are related to an ability to foster working alliance bonds. However, the presence of a significant difference between the AAS Depend dimension scores of therapists whose clients scored high on Bond ratings and the AAS Depend dimension scores of therapists whose clients scored low on Bond ratings provides partial support for a connection between therapist Adult Attachment Scale scores and client ratings of working alliance bonds. It remains for future research with larger sample sizes to examine the strength and durability of this relationship. Limitations The present study was exploratory in nature and suffers from a number of weaknesses that warrant cautious interpretation of results. The reliance on self-report data is an obvious limitation of the study and introduces the potential for problems such as response bias inherent in measures of this type. The generalizability of the study's findings are limited to the sample on which it is based and other similar therapist-client dyads. Some factors may differentiate the proposed sample from other populations. First, the therapists who volunteered to participate in this study presumably have sufficient confidence in their therapeutic skills to be willing to have their client's perceptions of them made available to the researcher, while not having access to this data themselves. Second, the clients who agreed to participate in the study were required to have sufficient education to allow them to read, understand, and complete the materials used in the study, and the requisite amount of motivation to complete the necessary questionnaire material when it was presented to them. A further limitation of the study is that Attachment theory in its present form does not make clear or precise predictions about the therapeutic relationship, therefore the research questions in the present study were derived by extrapolations from existing theory and empirical research on attachment and adult love relationships that may not apply to the therapeutic relationship. Weiss  64 (1982; 1986) notes that adult attachment differs from childhood attachment in several ways. First, adult attachments are usually reciprocal, that is, partners are at different times both the provider and recipient of care. Childhood attachments, on the other hand, differ in that the infant or child seeks but does not provide care. Second, in adulthood, the attachment system is not so easily activated as it is in childhood. Generally, it is only during periods of grief or relationship dissolution that the adult exploration system is overwhelmed by the attachment system. Last, in childhood, the attachment figure is usually a parent, whereas in adulthood, the attachment figure is generally a peer and most often a sexual partner. In light of these distinctions, it can be reasonably argued that client attachment within the therapeutic relationship more closely resembles childhood attachment in that the client is the recipient of care, the client's attachment system may well have been activated by the events that precipitated entry into therapy, and, in ethical relationships, the client and therapist are not sexual partners. Therefore, given that attachments formed within a therapeutic setting may constitute a special category, it may also be that the measures of adult attachment employed in the present study are inappropriate tools with which to examine attachment within the client/therapist relationship. Measurement of attachment is still in its infancy. The Hazan and Shaver (1987) Adult Attachment Questionnaire, although widely used in studies of attachment, has very definite psychometric problems, not the least of which is the requirement that respondents opt for the discrete category that best represents their viewpoint. Therefore, the possibility of any overlap between types is negated and valuable information is lost. In order to reduce client involvement, and thereby encourage therapist participation, clients in the present study were required only to complete the Hazan and Shaver measure and not the more sensitive Collins and Read (1990) Adult Attachment Scale. As a result, the amount of client attachment information available for statistical comparison was minimal and a limitation of the study. The original design of the present study called for the involvement of equal numbers of  65 therapists in each of the three Hazan and Shaver (1987) attachment categories. Due to an unexpected distribution pattern of attachment types, the sample employed was heavily weighted in the direction of Secure attachment and therefore the extent to which meaningful comparisons of attachment type could be made was severely reduced. With respect to the comparisons that were made, the significant differences found between High and Low Bond therapist/client dyads may be seen as being based on the unproven assumption that there was homogeneity among clients associated with a particular therapist. With a larger sample size, this shortcoming could have been avoided by employing a nested design. However, the fact that we have no method of gauging the extent to which therapist factors influenced, or failed to influence, the development of client bonds demands that these differences be viewed as extremely tentative. In a similar vein, it should be noted that by employing sets of multiple t-tests to examine differences between client ratings of WAI-S Bonds, as well as differences between therapist AAS scores and therapist WAI-S ratings, the likelihood of Type 1 error was inflated. However, although the clinical relevance of these findings can only be determined by future study, the significant differences found with such limited sample sizes remains noteworthy and merits further research. As a result of these limitations, and the high rate of attrition within the study, extreme caution must be used in interpreting its results. Recommendations for Future Research The research questions examined in the present study, although derived from the relevant literature, were exploratory in nature. The body of research pertaining to attachment theory and attachment processes is constantly expanding as newer applications are tested and revised. As with any comprehensive explanation of personality development, especially one which postulates an evolutionary basis, attachment theory makes large generalizations concerning the applicability of its theoretical underpinnings. However, to date, most studies of attachment have been conducted in  66 Western cultures, and as Bretherton and Waters (1985) point out, although attachment may be universal, patterns of attachment as assessed by typical attachment measures may not be. This is also applicable to adult measurement of attachment. Whether Western ideas of what constitutes healthy adult functioning are transferable to other cultures is a moot point. Indeed, recently several ethnopsychologists have suggested that Western concepts of emotion, emotional expression, and emotional attachment may not transfer to other cultures (Kleinman & Good, 1985; Lutz, 1988). Although measurement of attachment in Western culture has demonstrated consistent results, the universality of attachment measurement is yet to be demonstrated and merits further research. Several fertile research topics have been uncovered by the present study. In their study of romantic relationships, attachment style, and attachment dimensions, Collins and Read (1990) examined the interaction and matching of attachment types. Future research examining the impact of the attachment styles of therapists and clients may well have important implications for therapeutic outcome. For example, theoretically, a dyad consisting of a securely attached therapist and a securely attached client would be expected to be a positive match, in a therapeutic sense. However, therapist/client dyads where both the therapist and the client are insecurely attached might possibly have negative implications for therapeutic outcome. Relatedly, attachment styles have been conceptualized as possessing relatively stable, traitlike qualities. Nevertheless, the attachment style that a particular individual exhibits within a given relationship may be contingent on their own relationship history in conjunction with the style possessed by the other relationship partner (Simpson, 1990). This suggests two intriguing possibilities for further research, first, that the attachment style exhibited by a therapist can effect the attachment style of a client, and second, that the attachment style exhibited by a therapist may be at least partly contingent on the attachment style of a particular client. In the present study, therapists self-reported attachment styles were heavily weighted in the direction of secure attachment. If it is indeed the case that the majority of therapists are securely  attached, research is required to examine whether their secure attachment is as a result of parental influences, model disconfirming experiences, having worked through personal issues, or other factors. Even if it is the case that, by and large, therapists are securely attached, it is unlikely, especially given the range of scores obtained by therapists on the Adult Attachment Scale, that all such therapists are securely attached to the same degree. This point underscores the general need to develop more sensitive measures of attachment. However, if the attachment styles of therapists are to be examined more closely, particular care must be given to the instruments used for this purpose. Although current self-report measures of attachment may be appropriate for use with members of the general public, the relative transparency of such measures may invalidate their use with individuals who are aware of the attributes associated with mental health. On a more general level, research examining the extent to which therapists are willing to participate as respondents in studies is sadly lacking. Although the initial therapist response rate to questionnaires in Phase 1 of the study was encouraging, Phase 2 participation was much lower. In addition, attrition in Phase 2 of the study accounted for the withdrawal of 50% of therapists who had initially agreed to participate. An examination of whether the response rates of these therapists were representative of therapists' response rates as a whole was beyond the scope of the study, however, it may be that therapists are less likely to participate in research that involves the sharing of personal information. In reviewing the body of research on the therapeutic relationship, it seems that, in general, most studies that focus on the client examine the extent to which personal characteristics either facilitate or hinder the therapeutic process, whereas studies that focus on the contribution of the therapist examine the extent to which technical activity impacts on positive outcome. This apparent imbalance, which may or may not be attributed to the disenchantment of researchers resulting from the contradictory and inconclusive findings of studies examining the core conditions postulated by Rogers (1951, 1957), can only serve to hamper progress in the field of therapeutic research. Research examining the extent to which therapists are willing to act as  68 respondents in studies, and in what manner of studies, should prove beneficial to the field. Client attachment should also prove to be a fruitful area for in depth research. For instance, the distribution of client attachment in the present study found that a majority of clients identified themselves as Avoidant, whereas attachment research suggests that Avoidant individuals tend to deny negative affect and any need for social support. Longitudinal research examining whether clients' insecure attachment styles can be moved in the direction of security would contribute to the area, as would research on clients' attachments to their therapists and whether attachments formed are predictive of therapeutic outcome. In conclusion, Pistole (1989) has suggested that the working alliance can be regarded as a specialized form of attachment bond. If this view is accepted, it seems reasonable to assume that factors, such as unconditional acceptance, emotional consistency, and availability as a secure base, which are associated with successful parent/child attachment, will also promote the development of enduring therapeutic bonds. Given that a significant number of clients may have experienced early childhood events that severely limit their capacity to form intimate adult relationships, studies that examine how therapists can foster alliances and attachments with their clients are important not only with respect to stimulating future research, but also in the training of future therapists.  69 References  Adam, K. S. (1982). Loss, suicide, and attachment. In CM. Parkes & J. Stevenson-Hinde (Eds.), The place of attachment in human behaviour (pp. 269-294). New York: Basic Books. Adler, J. V. (1988). A study of the working alliance in psychotherapy. Unpublished doctoral dissertation. University of British Columbia. Ainsworth, M. D. S., Blehar, M. C , Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the Strange Situation. Hillsdale, NJ: Erlbaum. Allen, J., Newsom, G., Gabbard, G., & Coyne, L. (1984). Scales to assess the therapeutic alliance from a psychoanalytic perspective. Bulletin of the Menninger Clinic, 48, 383-400. Bartholomew W, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology. 61, 226-244. Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy,: Theory, Research and Practice, 16. 252-260. Bordin, E. S. (1980). Of human bonds that bind or free. Paper presented at the annual meeting of the Society for Psychotherapy Research, Pacific Grove, CA. Bordin, E. S. (1990, August). Symposium discussants' remarks. In B. Mallinckrodt (Chair), Antecedents of the therapeutic working alliance. Symposium conducted at the annual meeting of the American Psychological Association. Bowlby, J. (1958). The nature of a child's tie to his mother. International Journal of Psychoanalysis. 39, 350-373. Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York: Basic Books. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York: Basic Books. Bowlby, J. (1977). The making and breaking of affectional bonds. British Journal of Psychiatry.  70 130. 201-210. Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss. New York: Basic Books. Bowlby, J. (1982). Attachment. 2nd edition of vol. 1 of Attachment and loss. New York: Basic Books. Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. New York: Basic Books. Bretherton, I. (1985). Attachment theory: Retrospect and prospect. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Monographs of the Society for Research in Child Development 50, 3-38. Bretherton, I., & Waters, E. (1985). Growing points in attachment theory and research. Monographs of the Society for Research in Child Development 50,(1-2, Serial No. 209). Campos, J. J., Barrett, K. C , Lamb, M. E., Goldsmith, H. H., & Stenberg, C. (1983). Socioemotional development. In M. M. Haith & J. J. Campos (Eds.), Handbook of child psychology: Vol. 2. Infancy and psychobiology (pp. 783-915). New York: Wiley. Clark, M. S., & Reis, H. T. (1988). Interpersonal process in close relationships. Annual Review of Psychology. 39, 609-672. Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology. 58, 644-663. Dontas, C , Maratos, O., Fafoutis, M., & Karengelis, A. (1985). Early social development in institutionally reared Greek infants: Attachment and peer interaction. Monographs of the Society for Research in Child Development 50(1 & 2), 136-146. Erikson, M. F., Sroufe, L. A., & Egeland, B. (1985). The relationship between quality of attachment and behaviour problems in preschool in a high-risk sample. Monographs of the Society for Research in Child Development. 50(1 & 2), 147-166. Feeney, J. A., & NoUer, P. (1990). Attachment style as a predictor of adult  romantic relationships.  71 Journal of Personality and Social Psychology. 58. 281-291. Frank, A. F., & Gunderson, J. G. (1990). The role of the therapeutic alliance in the treatment of schizophrenia. Archives of General Psychiatry. 47. 228-236. Frank, J. D. (1961). Persuasion and healing. New York: Johns Hopkins University Press. Frank, J. D., Hoehn-Saric, R., Imber, S. D., Liberman, B. L., & Stone, A. K. (1978). Effective ingredients of successful psychotherapy. New York: Brunner/Mazel. Freud, S. (1958). On the beginning of treatment: Further recommendations of the technique of psychoanalysis. In J. Strachey (Ed. and Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 12, pp. 122-144). London: Hogarth Press. (Original work published 1913) Frieswyk, S. H., Allen, J. G., Colson, D. B., Coyne, L., Gabbard, G. O., Horowitz, L., & Newsom, G. (1986). Therapeutic alliance: Its place as process and outcome variable in dynamic psychotherapy research. Journal of Consulting and Clinical Psychology. 54. 32-39. Frieswyk, S. H., Colson, D. B., & Allen, J. G. (1984). Conceptualizing the alliance from a psychoanalytic perspective. Psychotherapy. 27, 460-464. Frommer, E. A., & O'Shea, G. (1973a). Antenatal identification of women liable to have problems in managing their infants. British Journal of Psychiatry. 123. 149-156. Frommer, E. A., & O'Shea, G. (1973b). The importance of childhood experience in relation to problems of marriage and family building. British Journal of Psychiatry. 123. 157-160. Gaston, L. (1990). The role of the alliance in psychotherapy: Theoretical and empirical considerations. Psychotherapy. 27. 143-153. Gaston, L. (1991). Reliability and criterion-related validity of the California Psychotherapy Alliance Scales-Patient Version. Psychological Assessment: A Journal of Consulting and Clinical Psychology. 3, 68-71. Gelso, C. J., & Carter, J. A. (1985). The relationship in counseling and psychotherapy: Components,  72 consequences, and theoretical antecedents. The Counseling Psychologist, 13, 155-243. Gitleson, M. (1962). The curative functions in psychotherapy. International Journal of Psychoanalysis, 43, 194-205. Goldfried, M. R. (1980). Toward the delineation of therapeutic change principals. American Psychologist 35, 991-999. Goldfried, M. R. (Ed.). (1982). Converging themes in psychotherapy: Trends in psychodynamic, humanistic and behavioral practice. New York: Springer. Gomes-Swartz, B. (1978). Effective ingredients in psychotherapy: Prediction of outcome from process variables. Journal of Consulting and Clinical Psychology. 46. 1023-1035. Greenberg, L. S., & Pinsoff, W. M. (Eds.) (1987). The psychotherapeutic process: A research handbook. New York: Guilford Press. Greenberg, L. S., & Webster, M. C. (1982). Resolving decisional conflict by Gestalt two-chair dialogue: Relating process to outcome. Journal of Counseling Psychology, 29, 468-477. Greenson, R. R. (1967). The working alliance and the transference neuroses. Psychoanalysis Quarterly. 34, 155-181. Grossman, K. E., & Grossman, K. (1991). Attachment quality as an organizer of emotional and behavioral responses in a longitudinal perspective. In C. M. Parkes, J. Stevenson-Hinde, & P. Marris (Eds.), Attachment across the life cycle (pp. 93-114). London: Tavistock/Routledge. Gurman, A. S. (1977). The patients' perception of the therapeutic relationship. In A. S. Gurman & A. M. Razin (Eds.), Effective psychotherapy. New York: Pergamon. Hartley, D.E., & Strupp, H. H. (1983). The therapeutic alliance: Its relation to outcome in brief psychotherapy. In J. Masling (Ed.), Empirical studies in analytical theories (Vol. 1, pp. 1-37). Hillsdale, NJ: Erlbaum. Hazan, C , Hutt, M. L., & Marcus, H. (1991). Continuity and change in inner working models of attachment. Unpublished manuscript, Department of Human Development, Cornell University.  73 Hazan, C , & Shaver, P. R. (1987). Conceptualizing romantic love as an attachment process. Journal of Personality and Social Psychology, 52. 511-524. Hazan, C , & Shaver, P. R. (1990). Love and work: An attachment-theoretical perspective. Journal of Personality and Social Psychology. 59, 270-280. Highlen, P. S., & Hill, C. E. (1984). Factors affecting client change in individual counseling: Current status and theoretical speculations. In S.D. Brown & R.W. Lent (Eds.), The handbook of counseling psychology (pp. 334-395). New York: John Wiley. Horowitz, M. J., Marmar, C , Weiss, D. S., DeWitt, K., & Rosenbaum, R. (1984). Brief psychotherapy of bereavement reactions. Archives of General Psychiatry. 41. 438-448. Horvath, A. 0. (1981). An exploratory study of the concept of therapeutic alliance and its measurement. Unpublished doctoral dissertation, University of British Columbia, Vancouver, Canada. Horvath, A. 0., & Greenberg, L. S. (1986). The development of the Working Alliance Inventory. In L.S. Greenberg & W.M. Pinsoff (Eds.), The psychotherapeutic process: A research handbook (pp. 529-556). New York: Guilford. Horvath, A. O., & Greenberg, L. S. (1989). The development and validation of the Working Alliance Inventory. Journal of Counseling Psychology. 36. 223-233. Horvath, A. O., & Symonds, B. D. (1991). Relation between working alliance and outcome in psychotherapy: A meta-analysis. Journal of Counseling Psychology. 38. 139-149. Hoyt, C.L. (1941). Test reliability measured by analysis of variance. Psychometika. 5, 153-160. Kazdin, A. E. (1979). Nonspecific treatment factors in psychotherapy outcome research. Journal of Consulting and Clinical Psychology. 47, 846-851. Kestenbaum, R., Farber, E. A., & Sroufe, L. A. (1989). Individual differences in empathy among preschoolers: Relation to attachment history. In N. Eisenberg (Ed.), Empathy and related emotional responses: New directions for child development (pp. 51-64). San Francisco, CA:  74 Jossey-Bass. Kivlighan, D. M., Jr., & Schmitz, P. J. (1992). Counselor technical activity in cases with improving working alliances and continuing-poor working alliances. Journal of Counseling Psychology, 39. 32-38. Kleinman, A., & Good, B. (1985). Culture and depression: Studies in the anthropologv and cross -cultural psychiatry of affect and disorder. Berkeley: University of California Press. Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working models, affect regulation, and representations of self and others. Child Development 59, 135-146. Koestner, R., Franz, C , & Weinberger, J. (1990). The family origins of empathic concern: A 26 -year longitudinal study. Journal of Personality and Social Psychology. 58, 709-717. Kokotovic, A. M., & Tracey, T. J. (1990). Working alliance in the early phase of counseling. Journal of Counseling Psychology. 37, 16-21. Lambert, M. J., DeJulio, S. S., & Stein, D. M. (1978). Therapist interpersonal skills: Process, outcome, methodological considerations and recommendations for future research. Psychological Bulletin. 85, 467-489. Levy, M. B., & Davis, K. E. (1988). Lovestyles and attachment styles compared: Their relations to each other and to various relationship characteristics. Journal of Social and Personal Relationships. 5, 439-471. Luborsky, L. (1971). Perennial mystery of poor agreement among criteria for psychotherapy outcome. Journal of Consulting and Clinical Psychology. 37, 316-319. Luborsky, L. (1976). Helping alliances in psychotherapy. In J.L. Cleghorn (Ed.), Successful psychotherapy (ch. 2 and 3). New York: Brunner/Mazel. Luborsky, L., & Auerbach, A. (1985). The therapeutic alliance in psychodynamic psychotherapy: The research evidence and it's meaning for practice. In R.  Hales & A.  Frances (Eds.), Psychiatry update: American psychiatry association annual review. Vol. 4. New  75 York: Brunner/Mazel. Luborsky, L., Crits-Cristoph, P., Alexander, L., Margolis, M., & Cohen, M. (1983). Two helping alliance methods for predicting outcomes of psychotherapy: A counting signs vs. a global rating method. Journal of Nervous and Mental Disease. 171. 480-491. Luborsky, L., Mintz, J., Auerbach, A., Crits-Cristoph, P., Bachrach, H., Todd, T., Johnson, M., Cohen, M., & O'Brien, C. P. (1980). Predicting the outcome of psychotherapy: Findings of the Penn Psychotherapy Project. Archives of General Psychiatry. 37, 471-481. Luborsky, L., McLellan, A. T., Woody, G. E., O'Brien, C. P., & Auerbach, A. (1985). Therapist success and its determinants. Archives of General Psychiatry. 42. 602-611. Lutz, C. (1988). Unnatural emotions: Everyday sentiments in a Micronesian atoll and their challenge to Western theory. Chicago: University of Chicago Press. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Monographs of the Society for Research in Child Development. 50. 66-106. Mallinckrodt, B. (1990). Clients' history of attachment to parents and the working alliance. In B. Mallinckrodt (Chair), Antecedents of the therapeutic working alliance. Symposium presented at the 98th Annual Convention of the American Psychological Association, Boston. Mallinckrodt, B. (1991). Clients' representations of childhood emotional bonds with parents, social support, and formation of the working alliance. Journal of Counseling Psychology. 38, 401-409. Mallinckrodt, B., & Nelson, M. L. (1991). Counsellor training level and the formation of the therapeutic working alliance. Journal of Counseling Psychology. 38. 133-138. Marmar, C. R., Gaston, L., Gallagher, D., & Thompson, L. W. (1989). Alliance and outcome in late life depression. Journal of Nervous and Mental Disease. 177. 464-472. Marmar, C. R., Horowitz, M. J. Weiss, D. S., & Marziali, E. (1987). The development of the therapeutic alliance rating system. In L.S. Greenberg & W.M. Pinsoff (Eds.), The  76 psychotherapeutic process: A research handbook (pp. 367-390). New York: Guilford Press. Marmar, C. R., Weiss, D. S., & Gaston, L. (1989). Towards the validation of the California Therapeutic Alliance Rating System. Journal of Consulting and Clinical Psychology, 1, 46-52. Marziali, E. (1984). Prediction of outcome of brief psychotherapy from therapist interpretive interventions. Archives of General Psychiatry, 41. 301-304. Marziali, E., Marmar, C , & Krupnick, J. (1981). Therapeutic alliance scales: Development and relationship to therapeutic outcome. American Journal of Psychiatry. 138. 361-364. Mikulincer, M., Florian, V., Tolmacz, R. (1990). Attachment styles and fear of personal death: A case study of affect regulation. Journal of Personality and Social Psychology, 58, 273-280. Mikulincer, M., & Nachshon, O. (1991). Attachment styles and patterns of self-disclosure. Journal of Personality and Social Psychology. 61. 321-331. Mitchell, K. M., Bozarth, J. D., & Krauft, C. C. (1977). A reappraisal of the therapeutic effectiveness of accurate empathy, nonpossessive warmth, and genuineness. In A.S. Gurman & A.M. Razin (Eds.), Effective psychotherapy: A handbook of research (pp. 544-565). New York: Pergamon Press. Moras, K., & Strupp, H. H. (1982). Pretherapy interpersonal relations, patient's alliance, and outcome in brief therapy. Archives of General Psychiatry. 39. 405-409. Morgan, R., Luborsky, L., Crits-Cristoph, P., Curtis, H., & Solomon, J. (1982). Predicting the outcome of psychotherapy by the Perm Helping Alliance Rating Method. Archives of General Psychiatry. 39, 397-402. O'Malley, S. S., Suh, C. S., & Strupp, H. H. (1983). The Vanderbilt Psychotherapy Process Scale: A report on the scale development and a process-outcome study. Journal of Consulting and Clinical Psychology. 5 ! 581-586. Orlinsky, D. E., & Howard, K. I. (1986). Process and outcome in psychotherapy. In S. L. Garfield & A. E. Bergin (Eds.), Handbook of psychotherapy and behaviour change (3rd ed., pp.311-381).  77 New York: Wiley. Parkes, C. M. (1973). Factors determining the persistence of phantom pain in the amputee. Journal of Psychosomatic Research. 17. 97-108. Parloff, M. B., Waskow, I. E., & Wolfe, B. E. (1978). Research on therapist variables in relation to process and outcome. In S.L. Garfield & A.E. Bergin (Eds.), Handbook of psychotherapy and behaviour change (pp. 233-329). New York: John Wiley. Pistole, C. M. (1989). Attachment: Implications for counselors. Journal of Counseling and Development. 68, 190-193. Ricks, M. (1985). The social transmission of parental behaviour: Attachment across generations. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Monographs of the Society for Research in Child Development 50. 211-230. Rogers, C. R. (1951). Client-centered therapy. Cambridge, MA: Riverside Press. Rogers, C. R. (1957). The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology. 22. 95-103. Rounsaville, B. J., Chevron, E. S., Prusoff, B. A., Elkin, I., Imber, S., Sotsky, S., & Watkins, J. (1987). The relation between specific and general dimensions of the psychotherapy process in interpersonal psychotherapy of depression. Journal of Consulting and Clinical Psychology. 55. 379-384. Saunders, S. M., Howard, K. I., & Orlinsky, D. E. (1989). The therapeutic bond scales: Psychometric characteristics and relationship to treatment effectiveness. Journal of Consulting and Clinical Psychology. I, 323-330. Shapiro, D., & Shapiro, D. (1983). Comparative therapy outcome research: Methodological implications of meta-analysis. Journal of Consulting and Clinical Psychology. 51. 42-53. Shaver, P. R., & Hazan, C. (1993). Adult romantic attachment: Theory and evidence. In D. Perlman & W. Jones (Eds.), Advances in personal relationships: Vol. 4. England: Jessica Kingsley.  78 Silver, D. (1984). Introductory remarks. In J. Lichtenberg, M. Bomstein, & D. Silver (Eds.), Empathy (Vol.2 pp.3-7). Hillsdale, NJ: Erlbaum. Simpson, J. A. (1990). Influence of attachment styles on romantic relationships. Journal of Personality and Social Psychology. 59. 971-980. Smith, M. L., & Glass, G. V. (1977). Meta-analysis of psychotherapy outcome studies. American Psychologist. 32, 752-760. Smith, M. L., Glass, G. V., & Miller, T. I. (1980). The benefits of psychotherapy. Baltimore, MD: Johns Hopkins University Press. Sroufe, L. A. (1983). Infant-caregiver attachment and patterns of adaption in preschool: The roots of maladaption and competence. In M. Perlmuter (Ed.), Minnesota Symposium on Child Psychology (Vol. 16, pp. 41-83). Hillsdale, NJ: Erlbaum. Sroufe, L. A., & Fleeson, J. (1986). Attachment and the construction of relationships. In W. Hartup & Z. Rubin (Eds.), Relationships and development (pp. 51-71). Hillsdale, NJ: Erlbaum. Sterba, R. F. (1934). The fate of the ego in analytic therapy. International Journal of Psychoanalysis. 115. 117-126. Strong, S. R. (1968). Counseling: An interpersonal influence process. Journal of Counseling Psychology. 15, 215-224. Strupp, H. H. (1980). Success and failure in time-limited psychotherapy: A systematic comparison of two cases: Comparison 2. Archives of General Psychiatry. 37. 708-716. Suh, C. S., Strupp, H. H., & O'Malley, S. S. (1986). The Vanderbilt Process Measures: The Psychotherapy Process Scale (VPPS) and the Negative Indicators Scale. In L.S. Greenberg & W.M. Pinsoff (Eds.), The psychotherapeutic process: A research handbook (pp. 285-323). New York: Guilford. Tichenor, V., & Hill, C.E. (1989). A comparison of six measures of working alliance. Psychotherapy: Research and Practice. 26, 195-199.  79 Tracey, T. J., & Kokotovic, A. M. (1989). Factor structure of the Working Alliance Inventory. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 1, 207-210. Tryon, G., & Kane, A. S. (1993). Relationship of working alliance to mutual and unilateral termination. Journal of Counseling Psychology. 40. 33-36. Waters, E. (1978). The reliability and stability of individual differences in infant-mother attachment. Child Development. 49, 483-494. Waters, E., Wippman, J., & Sroufe, L. A. (1979). Attachment, positive affect, and competence in the peer group.: Two studies in construct validation. Child Development. 50. 821829. Weiss, R. S. (1982). Attachment in adults. In C. M. Parkes & J. Stevenson-Hinde (Eds.), The place of attachment in human behaviour (pp. 171-184). New York: Basic Books. Weiss, R. S. (1986). Continuities and transformations in social relationships from childhood to adulthood. In W. Hartup & Z. Rubin (Eds.), Relationships and development (pp. 95-110). Hillsdale, NJ: Erlbaum. Yalom, I. D. (1980). Existential psychotherapy. New York: Basic Books. Zetzel, E. R. (1956). Current concepts of transference. International Journal of Psychiatry. 37, 369-376.  80 Appendix 1 - Initial Contact THE THERAPY RESEARCH PROJECT  Dear Colleague: I am writing to request your participation in a study which I am conducting as part of my Master's thesis project. The title of the study is The Therapy Research Project and my faculty advisor is Dr. Beth Haverkamp of the Department of Counselling Psychology at the University of British Columbia. The accompanying questionnaires are a component of a larger study examining connections, similarities and dissimilarities between interpersonal and therapeutic relationships. At this stage of the study, I would appreciate your responding to the accompanying questionnaires, which should take approximately 20 minutes to complete, and indicating on the accompanying form whether or not you are willing to participate in the second phase of the study. I have included a stamped, addressed envelope to facilitate the return of the Phase 1 questionnaires. All names will be deleted from returned questionnaires and replaced with number codes. Respondents' names and corresponding number codes will be held in separate locked filing cabinets with access restricted solely to Gerard McKee and Dr. Beth Haverkamp. In the second phase of the study, therapists are requested to: (1) ask three clients to complete two brief questionnaires (12-15 mins. in total) immediately after completion of their third therapy session, and (2) to complete a brief questionnaire (5-7 mins.) with respect to each of these three clients. Should you not wish to participate in the second phase of the study, I thank you for your assistance and request that you return the completed Phase 1 questionnaires in the envelope provided. Your participation and generosity with your time and energy are very much appreciated.  Gerard McKee *For more information, please call me at 681-1725 or Dr. Beth Haverkamp at 822-5354.  81 Appendix 2 - Therapist Forms Phase 2 THE THERAPY RESEARCH PROJECT  Professional affiliation: Psychologist  Social Worker  Counsellor  Psychiatrist  Other (please specify)  Highest degree completed: B.A.  M.S.W.  B.Ed.  B.S.W.  Phd.  Ed.D  Other (please specify)  MA. M.D.  M.Ed._  Number of years experience as a therapist/counsellor: None 11 - 15  Sex: Male  1- 5  6 - 10  More  Female  Age:  Current Marital Status:  Single  Married  Living with a partner Separated  If previously married please indicate: Divorced  Divorced  Widowed  Widowed  With respect to your current relationship, please indicate your level of satisfaction: (e.g., l=dissatisfied; 5=very satisfied) I  2  3  dissatisfied  4  5  very satisfied  How would you describe your theoretical orientation?  Who did you live with during the first five years of your life? Both Natural Parents  Mother  Father  Mother and Stepfather  Father and Stepmother  Grandmother  Grandfather  Aunt  Other (please specify)  How many brothers and sisters did you have? Brothers  Sisters  Stepbrothers  Stepsisters  Where were your mother and father born? Mother  Father  What was the predominant ethnic/cultural influence in your family? (e.g., Chinese-Canadian, British-Canadian, Indo-Canadian etc.):  Uncle  83 Appendix 2 - Adult Attachment Questionnaire THE THERAPY RESEARCH PROJECT Relationships with Others  Please read each of the three self-descriptions below and then place a check-mark next to the single alternative that best describes how you feel in romantic love relationships. (Please check only one.) (Note: The terms close and intimate refer to psychological or emotional closeness, not necessarily to sexual intimacy.) 1. I am somewhat uncomfortable being close to others; I find it difficult to trust them completely, difficult to allow myself to depend on them. I am nervous when anyone gets too close, and often, love partners want me to be more intimate than I feel comfortable being.  2. I find that others are reluctant to get as close as I would like. I often worry that my partner doesn't really love me or won't want to stay with me. I want to get very close to my partner, and this sometimes scares people away.  3. I find it relatively easy to get close to others and I am comfortable depending on them. I don't often worry about being abandoned or about someone getting too close to me.  *If this questionnaire is completed, it will be assumed that consent to participate has been given.  Appendix 2 - Adult Attachment Scale THE THERAPY RESEARCH PROJECT Please read each of the following statements and rate the extent to which It describes your feelings about romantic relationships. Please think about all your relationships (past and present) and respond In terms of how you generally feel in these relationships. If you have never been Involved In a romantic relationship, answer in terms of how you think you would feel. Please use the scale below by placing a number between 1 and 5 In the space provided to the right of each statement 1Not at all characteristic of me  very characteristic of rne  1)  I find it relatively easy to get dose to others.  2)  I do not worry about being abandoned.  3)  I find It difficult to allow myself to depend on others.  4)  In relationships, I often worry that my partner does not really love me.  5)  I find that others are reluctant to get as dose as I would like.  6)  I am comfortable depending on others.  7)  I do noi worry about someone getting too dose to me.  8)  I find that people are never there when you need them.  9)  I am somewhat uncomfortable being dose to others.  10)  In relationships, 1 often worry that my partner wil not want to stay with me.  It)  I want to merge completely with another person.  12)  My desire to merge sometimes scares people away.  13)  I am comfortable having others depend on me.  14)  I know that people wil be there when I need them.  15)  I am nervous when anyone gets too dose.  16)  1 find It difficult to trust others completely.  17)  Often, partners want me to be doser than 1 feel comfortable being.  18)  I am not sure that t can always depend on others to be there when I need them.  8  85 Appendix 2 - Phase 2 Participation  THE THERAPY RESEARCH PROJECT  Please indicate below whether or not you are willing to participate in the second phase of the study.  No, I am not willing to participate in the second phase of the study. [  ]  Yes, I am willing to participate, if selected, in the second phase of the research being conducted by Gerard McKee as part of the requirements of his Master's degree program. [  Name. Business Address.  Please indicate below whether I may contact you by telephone if necessary.  No, you may not contact me by telephone. [  Yes, you may contact me by telephone. [  Telephone Number.  ]  ]  ]  86 Appendix 2 - Therapist Information Form THE THERAPY RESEARCH PROJECT  This study is designed to explore the interactions that occur in counselling/psychotherapy. Your participation is of vital importance to the project and your generosity with your time and energy is much appreciated. The procedure we are using is designed to ensure complete confidentiality. Please follow the steps outlined below: 1)  You are being provided with envelopes marked INFORMATION AND CONSENT FORMS, CLIENT FORMS and THERAPIST FORMS. Please use these envelopes according to instructions.  2)  Please read and sign the Therapist Consent Form provided in the envelope marked INFORMATION AND CONSENT FORMS.  3)  Please turn to the envelope marked THERAPIST FORMS. Read and complete the questionnaire marked Therapist Questionnaire #1. This questionnaire is to be completed before asking any client to participate in the study. The questionnaire should take approximately 20 minutes to complete. In order for the study to avoid bias, it is important that all participants follow written procedures as closely as possible. When the questionnaires have been completed, replace them in the envelope provided and retain this and all other envelopes containing  questionnaires in safe keeping until collected by the  4)  completed  researcher.  To participate in this study, a client should be at least 18 years old and engaged in individual counselling/therapy. We are confining our study to clients who have completed two (2) counselling/therapy sessions and are about to begin their third (3rd) session. Please  87 exclude from the study any client whom you believe to be in crisis.  5)  Please approach in turn the first three (3) available clients who are about to begin a third session and ask them to participate in the study. Inform the clients that participation in the study should involve 20 minutes of their time following the session. Client Information Forms are provided in the envelope marked INFORMATION AND CONSENT FORMS and should be given to each client immediately after the request to participate. Please answer any questions clients may have about the study to the best of your ability. If any client has further questions, they may call Gerard McKee (681-1725) or Dr.Beth Haverkamp (8225354).  6)  If any of the first three clients approached declines to participate in the study, please approach the next available client about to begin a third session and repeat the request to participate. This procedure should be followed until three clients meeting the requirements of the study agree to participate.  7)  When a client agrees to participate in the study, please ask him/her to read and sign the Client Consent Form provided in the envelope marked INFORMATION AND CONSENT FORMS and sign yourself as a witness. After the Client Consent Form is signed please continue the session in your usual manner. The counsellor/therapist should retain both client and counsellor/therapist consent forms for later collection.  8)  Immediately following the end of the third session, please give the client the envelope marked CLIENT QUESTIONNAIRES and allow the him/her to complete the questionnaires in private. When the questionnaires have been completed, the client is requested to replace  88 them in the envelope provided, seal the envelope and return it to the counsellor/therapist for safekeeping.  9)  During the time that the client is completing his/her questionnaires, the counsellor/therapist should read and complete the appropriate Therapist Questionnaire #2. That is, Therapist Questionnaire #2 / Client #1 should be completed in relation to the first client participating in the study; Therapist Questionnaire #2 /Client #2 should be completed in relation to the second client participating in the study; and Therapist Questionnaire #2 I Client #3 should be completed in relation to the third client participating in the study. These questions should take approximately 5-10 minutes per client to complete. When all therapist questionnaires are completed, the counsellor/therapist should replace them in the envelope provided and seal the envelope.  10) The counsellor/therapist is asked to notify Gerard McKee (681-1725) all questionnaires are completed. The envelopes will be collected of notification. If you have any questions at any time, please call me at 681-1725 or Dr. Haverkamp at 822-5259. Thank you for your cooperation,  Gerard McKee  when within 7 days  89 Appendix 2 - Therapist Consent Form THE THERAPY RESEARCH PROJECT  I hereby voluntarily consent to participate in the counselling/psychotherapy research study being conducted by Gerard McKee as part of his M.A. degree requirements. The nature of this research has been explained to me and I understand that I will be required to complete some questionnaires. I have been informed that the responses on the questionnaires will be treated anonymously and confidentially. I am aware that I am free to withdraw from this study at any time.  Signed Date  Appendix 2 - Working Alliance Inventory (Therapist Short Fonn)  THE THERAPY RESEARCH PROJECT  On ihe following pages there are sentences that describe some of the efferent ways a person might think or feel about Ms or her dent. As you read the sentences mentally Insert the name of your dent In place of In the text. Below each statement Inside there Is a seven point scale:  1 Never  2 Rarely  3 Occasionally  4 Sometimes  5 Often  8 Very Often  7 Afways  II the statement describes the way you always feel (or think) drcle the number 7; If ll never applies to you drde the number 1. Use the numbers In between to describe the variations between these extremes.  This questionnaire Is CONFIDENTIAL! nalthur your therapist nor Ihe anencv will see voiir answers. Work fast, your llrst Impressions are the ones we would like to see. (PLEASE DONT FORGET TO RESPOND TOfKEflT/TEAf.) Thank you lor your cooperation.  1 Navar  2  3.  arri I agree abotflhe stepstobe a ^ to lmprc*»his^stuaiton. 2 3 Rarefy Occaatonaly  My dent and 1 both real confidant about tie usefulness of our currant acftVtykitterapy. 1 2 9 Never Rarafy Occastonaly  Ibefove  5.  6.  7.  5 Often  6 Very Often  7 Always  ^ ; A  • Vary Often  7 Always  5 Oram  1 have doubts about what wa ara tryingtoaccompMi ki therapy. 1 2 3 Navar Raraly Occastonaly  4 Somettnes  Often  • Very Often  7 Always  1 am confidant ki my aMtytohalp 1 Navar  3 Occastonaly  4 Sometknet  5 Often  • Very Often  7 Always  3 Occastonaly  8omeomes  5 Often  8 Vary Often  7 Always  3 Occaatonaly  4 Somettmee  S Often  « Vary Often  7 Always  3 Occaatonaly  4 8ometknea  5 Often  6 Very Often  7 Always  3 Occaatonaly  4 domatknaa  5 Often  Very Often  7 Always  . . and lhavadm»rar« kJeaacfl what hauhar real I problems ara. 3 1 2 Never Rarely Occaatonaly  4 Sometimes  5 Often  6 Very Often  7 Always  oodfor We have estaMshed a good underetandhg between us of tie kind of changes ftei would be( 4 1 2 3 Never Rarely Occastonaly Sometimes  5 Often  6 Very Often  7 Always  6 Vary Often  7 Always  2 Raraly  Wo are workingtowardsmutualy agreed upon goals. 1 2 Never Rarefy  1 appreciate  12. 1 Never  4  s  as a parson. 2 Rarely  We agree on what Is Important for 1 Never  1 Never  11.  _ 4 Sernetknes  •ofnedrnee  S.  10.  7 Always  3 Occastonaly  2 Raraly  1 Never  6  S Vary Oram  fcesme. 1 Navar  4.  5 Often  fin »i allm • m  to 2 Rarely  and 1 have but a mutual true! 2 Rarely  work on.  befeves the way wa are working with herMe problem is correct. 2 3 Raraly Occastonaly  Sometimes  5 Often  e  92 Appendix 3 - Client Forms Client Information Form THE THERAPY RESEARCH PROJECT  This study is designed to gather information about relationships between counsellors/therapists and their clients. The information that is gathered in this study will enable counsellors and therapists to help their clients more effectively. To this end, the researchers involved in the study would like you to share some information about your background, perceptions about yourself, and perceptions about your relationship with your counsellor/therapist. All of your responses will be treated with absolute confidentiality. Your counsellor/therapist will not see your responses or have access to any information that you may provide. To be involved in the study, you are requested to respond as openly and honestly as possible to items contained on a questionnaire. The questionnaire will be given to you by your counsellor/therapist immediately after you have completed your third counselling/therapy session. The questionnaire should take approximately 20 minutes to complete and should be responded to in private. When completed, the questionnaires should be replaced and sealed in the envelope provided. The envelope should then be returned to your counsellor/therapist. If you have any questions, please call Gerard McKee at 681-1725 or Dr. Beth Haverkamp at 822-5259. Thank you very much for your cooperation and participation.  93 Appendix 3 - Client Consent Form THE THERAPY RESEARCH PROJECT  I hereby voluntarily consent to participate in the Therapy Research Project being conducted by Gerard McKee. The nature of this research has been explained to me and I understand that I will be asked to complete two questionnaires. I have been informed that the responses on the questionnaires will be kept private and not shared with my counsellor/therapist. If I do not wish to participate in this study, or if I decide at a later time to withdraw from the study before its completion, I understand that my decision will in no way affect the quality or the availability of the counselling/therapy I will receive.  Signed Date Witness (Counsellor/Therapist)  94 Appendix 3 - Client Questionnaire #1 THE THERAPY RESEARCH PROJECT The enclosed questionnaires are part of a research project to study how clients feel about their therapy/counselling experiences. Please try to answer all questions as completely and accurately as possible. When you finish, replace the completed questionnaires in the envelope provided and seal it. Envelopes should then be returned to your counsellor/therapist. Thank you very much for your participation.  Age:  Sex:  Male  Marital Status: Single  Female  Married  Living with a partner Separated  Divorced  Widowed  Education (check highest level completed) Elementary school (number of years completed:  years)  Some High school (number of years completed:  years)  High school graduate (12 years:  )  College (number of years completed:  years)  College graduate (number of years in program:  years)  Graduate study or professional training (number of years:  years)  95 Appendix 3 - Adult Attachment Questionnaire THE THERAPY RESEARCH PROJECT Relationships with Others Please read each of the three self-descriptions below and then place a check-mark next to the single alternative that best describes how you feel in romantic love relationships. (Please check only one.) (Note: The terms close and intimate refer to psychological or emotional closeness, not necessarily to sexual intimacy.) 1. I am somewhat uncomfortable being close to others; I find it difficult to trust them completely, difficult to allow myself to depend on them. I am nervous when anyone gets too close, and often, love partners want me to be more intimate than I feel comfortable being.  2. I find that others are reluctant to get as close as I would like. I often worry that my partner doesn't really love me or won't want to stay with me. I want to get very close to my partner, and this sometimes scares people away.  3. I find it relatively easy to get close to others and I am comfortable depending on them. I don't often worry about being abandoned or about someone getting too close to me.  *If this questionnaire is completed, it will be assumed that consent to participate has been given.  ON  Appendix 3 - Working Alliance Inventory (Client Short Form) THE THERAPY RESEARCH PROJECT  Instructions On the following pages there are sentences that describe some of the different ways a person might think or feel about his or her therapist (counsellor). As you read the sentences mentally Insert the name of your therapist (counsellor) in place of In the text. Below each statement Inside there is a seven point scale:  1 Never  2 Rarely  3 Occasionally  4 Sometimes  5 Often  6 Very Often  7 Always  If the statement describes the way you always feel (or think) circle the number 7; if it never applies to you circle the number 1. Use the numbers in bptween to describe the variations between these extremes.  This questionnaire Is CONFIDENTIAUnelther your therapist nor the agencyjvill see your answers. Work fast, your first impressions are the ones we would like to see. (PLEASE DON'T FORGET TO RESPOND TO EVERY ITEM.)  Thank you for your cooperation.  © A. 0. Horvath, 1981, 1982; Revision Tracey & Kokotowitc 1989. WANS)  1 Never Z  3.  5 Often  6 Very Often  7 Always  Sometimes  5 Often  6 Very Often  7 Always  Sometimes  5 Often  6 Very Often  7 Always  Sometimes  5 Often  6 Very Often  7 Always  Occasionally  Sometimes  5 Often  6 Very Often  7 Always  and I are working towards mutually agreed upon goals. 2 3 Rarely Occasionally  Sometimes  5 Often  6 Very Often  7 Always  Occasionally  Sometimes  5 Often  6 Very Often  7 Always  Occasionally  Sometimes  5 Often  6 Very Often  7 Always  Occasionally  Sometimes  5 Often  6 Very Often  7 Always  and I have different ideas on what my problems are. 2 3 Rarely Occasionally  Sometimes  5 Often  6 Very Often  7 Always  We ha ve established a good understanding of the kind of changes that would be good for me. 1 2 3 Never Rarely Occasionally  Sometimes  5 Often  6 Very Often  7 Always  I believe the way we are working with my problem is correct. 1 2 Never Rarely  Sometimes  5 Often  6 Very Often  7 Always  What I am doing in therapy gives me new ways of looking at my problem. 1 2 3 Never Rarely Occasionally I believe.  . likes me. 1 Never  1 Never I am confident in _ 1  Never  1 Never 7.  I feel that. 1 Never  8.  and I agree about the things I will need to do in therapy to help improve my situation. 2 3 4 Rarely Occasionally Sometimes  2 Rarely  Occasionally  does not understand what I am trying to accomplish in therapy. 2 3 Rarely Occasionally 's ability to help me. 2 Rarely  . appreciates me. 2 Rarely  We agree on what is important for me to work on. 1 2 Never Rarely and I trust one another. 1 Never  10.  1 Never 11.  12.  2 Rarely  3 Occasionally  98 Appendix 4 - Therapist Demographics THE THERAPY RESEARCH PROJECT  Phase 1 Therapists Highest Degree Attained Ed.D.  1 (2.50%)  MA.  16 (40.00%)  M.D.  1 (2.50%)  M.Ed.  6 (15.00%)  M.S.W.  8 (20.00%)  Ph.D.  8 (20.00%)  Phase 1 Therapists Marital Status Single  8 (20.00%)  Married  17 (42.50%)  Living With A Partner  5 (12.50%)  Separated  1 (2.50%)  Divorced  9 (22.50%)  Appendix 4 - Therapist Demographics  Phase 1 Therapists Professional Affiliation Counsellor  20 (50.00%)  Family Therapist  1 (2.50%)  Social Worker  8 (20.00%)  Psychologist  11 (27.50%)  Phase 1 Therapists Years Experience 1 to 5 Years  14 (35.00%)  6 to 10 Years  8 (20.00%)  11 to 15 Years  8(20.00%)  Over 15 Years  10 (25.00%)  Phase 1 Therapists Theoretical Orientation * Cognitive Behavioral  5 (12.50%)  Eclectic  13 (32.50%)  Person Centered  4 (10.00%)  Psychodynamic  3 (7.50%)  Systemic  15 (37.50%)  Appendix 4 - Therapist Demographics  Phase 1 Therapists Parents Place Of Birth MOTHER  FATHER  Canada 18 (45.00%)  Canada 17 (42.50%)  Britain 9 (22.50%)  Britain 9 (22.50%)  U.S.A. 8 (20.00%)  U.S.A. 6 (15.00%)  Eastern Europe 3 (7.50%)  Eastern Europe 6 (15.00%)  Western Europe 1 (2.50%)  Western Europe 1 (2.50%)  India 1 (2.50%)  India 1 (2.50%)  Phase 1 Therapists Predominant Family Ethnic/Cultural Influence American  1 (2.50%)  British  4 (10.00%)  British/Canadian  16 (40.00%)  Canadian  3 (7.50%)  German/American  1 (2.50%)  Indo/Canadian  1 (2.50%)  Italian/Canadian  3 (7.50%)  Jewish/American  3 (7.50%)  Jewish/Canadian  6 (15.00%)  Ukrainian/Canadian  2 (5.00%)  101 Appendix 5 - Raw Data Therapist/Client WAI-S Bond Ratings and Therapist AAS Scores Therapist Bond Rating  Therapist Mean Bond  Client Bond Rating  Client Mean Bond  Therapist AAS Scores  1  21 22 24*  22.33  17 26* 26*  23.00  Close 20 Depend 24 Anxiety 8  2  24* 21 28*  24.33  19 19 28*  22.00  Close 27 Depend 20 Anxiety 9  3  25* 22 19  22.00  20 19 22  20.33  Close 18 Depend 18 Anxiety 17  4  25* 24* 26*  25.00  25* 27* 28*  26.66  Close 25 Depend 29 Anxiety 11  5  24 24 19  22.33  26* 28* 18  24.00  Close 25 Depend 28 Anxiety 12  6  28* 23 25*  25.33  20 22 21  21.00  Close 23 Depend 21 Anxiety 14  7  23 27* 25*  25.00  28* 27* 22  25.66  Close 27 Depend 23 Anxiety 9  8  21 19 27*  22.33  22 17 28*  22.33  Close 30 Depend 25 Anxiety 11  9  24 25* 26*  25.00  26* 18 23*  22.33  Close 30 Depend 25 Anxiety 9  24* 21 27*  24.00  Close 20 Depend 18 Anxiety 14  10  25* 22.33 19 23 * Indicates Top 15 Scores in Category  Appendix 5 - Raw Data  Therapist/Client WAI-S Task Ratings Therapist WAI-S Task Rating  Therapist Mean Task  Client WAI-S Task Rating  Client Mean Task  1  22 19 23  21.33  12 25 24  20.33  2  21 18 26  21.66  21 17 28  22.00  3  27 22 19  22.66  23 21 22  22.00  4  22 24 23  23.00  25 26 28  26.33  5  21 20 19  20.00  28 20 13  20.33  6  26 20 24  23.33  22 27 22  23.66  7  23 26 23  24.00  28 27 24  26.33  8  19 16 23  19.33  24 18 28  23.33  9  24 23 22  23.00  26 24 21  23.66  10  24 21 20  21.66  26 22 23  23.66  Appendix 5 - Raw Data  Therapist/Client WAI-S Goal Ratings Therapist WAI-S Goal Rating  Therapist Mean Goal  Client WAI-S Goal Rating  Client Mean Goal  1  22 21 24  22.33  15 21 26  20.66  2  20 16 23  19.66  18 14 27  19.66  3  25 25 21  23.66  24 18 23  21.66  4  25 25 26  25.33  24 26 28  26.00  5  19 19 20  19.33  28 25 15  22.66  6  26 23 23  24.00  22 26 23  23.66  7  26 25 25  25.33  28 20 23  23.66  8  22 20 25  22.33  23 17 26  22.00  9  26 25 24  25.00  24 24 18  22.00  10  24 21 21  22.00  25 23 27  25.00  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0054091/manifest

Comment

Related Items