UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Therapeutic alliance and attachment in play therapy Heller, Deborah C. 1996

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

Item Metadata


831-ubc_1996-0510.pdf [ 4.81MB ]
JSON: 831-1.0054051.json
JSON-LD: 831-1.0054051-ld.json
RDF/XML (Pretty): 831-1.0054051-rdf.xml
RDF/JSON: 831-1.0054051-rdf.json
Turtle: 831-1.0054051-turtle.txt
N-Triples: 831-1.0054051-rdf-ntriples.txt
Original Record: 831-1.0054051-source.json
Full Text

Full Text

THERAPEUTIC ALLIANCE AND ATTACHMENT IN PLAY THERAPY by DEBORAH C. HELLER B.Sc.(N.), McGill University, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS 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 September, 1996 © Deborah C. Heller, 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) 11 Abstract This study examined how the evolution of therapeutic alliance during child-centered play therapy varied between two children with differing patterns of attachment. The two children recruited, both 5 years-old, were a female, classified as insecure-defended (subclassification A3 - caretaking) and a male, classified as insecure-coercive (subclassification C2 - disarming). The children's attachment patterns were determined by two independent raters, using Crittenden's (1992a) classification of quality of attachment of preschool-aged children. Each child received 10 weekly non-directive play therapy sessions, which were videotaped, with the investigator as therapist. Participant-observation was used for data collection, and open coding and pattern-matching was used for data analysis. Pattern differences that emerged in the children's categories of play and categories of the therapeutic relationship were related to changes in therapeutic alliance and in attachment strategies for managing distress. Over time, the A3 child responded more to the therapist's responsive behaviours to her expressed needs and feelings; and engaged in relatively more self expression in her play. Over time, the C2 child responded more to the therapist's mere availability to the child; and tended to engage in relatively more self agency (self direction) in his play. Further play therapy research is needed to confirm the relationship between attachment and therapeutic alliance with children. iii Table of Contents Abstract ii Table of Contents iii List of Tables v Acknowledgements vi Chapter One: Introduction 1 Rationale 2 Research Approach 5 Purpose 7 Chapter Two: Literature Review 8 Therapeutic Alliance 10 Attachment 19 Play Therapy 30 Chapter Three: Method 41 Research Approach 41 Selection of Subjects 41 Subjects' Attachment Classifications 42 Play Therapy Procedure 44 Observation Methods and Data Collection 45 Analysis 46 Chapter Four: Results 49 Summary of Sessions: Annie 49 Summary of Sessions: Ryan 62 Chapter Five: Analysis 76 Categories of Play 78 Categories of Therapeutic Relationship 91 Pattern Matching 99 Changes in Therapeutic Alliance 101 Portraits 111 iv Chapter Six: Discussion 116 Limitations 118 Implications for Theory 120 Implications for Practice 122 Implications for Future Research 123 Summary 125 References 127 V List o f Tables Table 1. Narrative Themes 85 Table 2. Categories of play 89 Table 3. Categories of therapeutic relationship 97 vi Acknowledgements Particular thanks go to Dr. Larry Cochran for his encouragement to pursue investigation of "the dance of relationship" I saw occurring in play therapy; as well as for his valuable assistance in conceptualizing the phenomena that emerged. I would like to acknowledge Dr. John Allan's contribution in introducing me to attachment theory, particularly as it applies to young children. Acknowledgement goes to my coworkers for their encouragement and for sparing me at work while I completed my work on this study. Many family members and friends have extended support and encouragement throughout the process but a few particular thanks are in order. First and foremost, thanks go to my husband Wolf for his endless patience and practical support in typing, printing and bearing with the numerous late nights over the past few years. I am also grateful for my mother's support and assistance in translating needed research material. Particular thanks goes to my friend Ellen; for her encouragement through the dry spells and for the "chat breaks" during the computer marathons. Thanks go to Loony as well (a four-legged friend), for sitting up with me through the wee hours of many mornings. Finally, I am most grateful for the spiritual strength and endurance the Lord has provided to complete the task at hand. 1 Chapter One: Introduction In clinical populations, the key to children's presenting problems often lies with the quality of their attachments to primary caregivers. Bowlby (1988) describes attachment as the tie between a child and its primary caregiver, with whom he or she seeks proximity as a source of security and comfort. Depending upon how consistently available and responsive the caregiver is, the child develops different patterns of attachment behavior, or attachment styles. According to Main and Cassidy (1988), styles of attachment tend to persist through later developmental stages, forming a working model of how to relate to others later in life. Needless to say, if a child's relationship with its attachment figure was an insecure one, negatively affecting his/her relational working model, this could give rise to serious relational and behavioral problems and the need for therapy. Play therapy has been shown to be an effective treatment modality with many diagnostic problems in children (Landreth, 1991). Relationship or child-centered play therapy focuses on the immediate, lived relationship between child and therapist (Landreth, 1991; Moustakas, 1959). It therefore can provide the means through which children can learn an alternate, more secure model of how to relate to the world and others. The development of a safe and supportive 2 relationship between therapist and child (that is, of a therapeutic alliance) is a means to explore a secure style of attachment to others. There might be characteristic changes that develop in this relationship over the course of therapy, that facilitate changes in attachment style. How does the evolution of therapeutic alliance during play therapy vary between children with differing attachment styles? This is the question that will be addressed by the present study. Rationale Both therapeutic alliance and attachment style have been neglected in the literature as variables in play therapy research. According to Bowlby (1988) the clinical goal with insecurely attached children is to recreate the experience of a secure base that they lack (i.e. of an available and responsive caregiver). The therapeutic relationship is the sole modality by which this experience of a secure base can be created during therapy. Since insecure attachment is a common clinical feature when counselling children, it would be important to examine how it is manifested and evolves in the therapeutic relationship. In the literature, client-centered therapists and child-centered therapists have looked at aspects of the therapeutic relationship as necessary conditions for effective therapy but not as variables of the therapeutic process itself. If this relationship is the very modality by which a child could develop a secure attachment style, it would be critical to use therapeutic alliance as a research 3 variable. Frieswick et al. (1986) advocate using therapeutic alliance as a process and outcome variable in psychotherapy research. They define therapeutic alliance as the client's collaboration with the therapeutic process and tasks. They distinguish it from therapeutic technique or the client's experience of the relationship (e.g. transference). However, for the purpose of this study, therapeutic alliance will be defined as mutually responsive affiliative behaviours between therapist and client. Since play therapy is a kind of 'dance of relationship' that frequently consists of the therapist following the lead of the child, measuring only the degree of the child's cooperation with therapist-initiated interventions might well miss half of this subtle relational dance. In this study, the child's approach and retreat from the therapist will be looked at as well, as part of therapeutic alliance. This view of the therapeutic relationship as a bidirectional dance is echoed by Short and Boon (1990) who see counselling as a dialectical process. They point out that counselling theory and practise has largely relied upon a dualistic, subject-object perspective where the direction of influence is largely one way, from therapist to client. However they maintain that the reality of the therapeutic relationship is more non-dualistic and bidirectional. Action and reaction of both therapist and client flow into one another and are often difficult to tease apart. They propose research that relies on analysis of the process of therapy in order to 4 illuminate this bidirectional quality of therapeutic alliance. Such research would be very clinically relevant. One study that was done to analyse the process (as opposed to solely the outcome, seen in most previous research) of play therapy was done by Faust and Burns (1991). Interactional behaviours of both the child and therapist were coded at different points in therapy, to track its progress and not just to focus on outcome. There is a dearth of research on therapeutic alliance and on attachment in children beyond infancy. It is over two decades since Bowlby (1969) presented his theory of attachment relationships between infants and their caregivers. Recently, data is emerging pertaining to attachment in preschool children. This is largely a result of impetus from the MacArthur Working Group on the Study of Attachment in the Transition Period formed in 1983 and chaired by Mark Greenberg. Further clinical research on attachment in young children could prove to be a powerful guide to the therapeutic process. In fact, Speltz (1990) presents a therapeutic model in treating conduct disorders in preschoolers that is based on attachment theory. Part of the therapy involves coaching parents through child-directed play sessions with their children. The therapeutic potential for such an approach would be strengthened if it was understood how children with each style of insecure attachment differ in how they use play interactions to potentially develop a more secure attachment style. This study will explore such differences. 5 Research Approach Assessment of attachment style in young children is still in its infancy. In 1978 Ainsworth developed the Strange Situation procedure for assessing attachment style in infants (Ainsworth, Bleher, Waters & Wall, 1978). In this procedure, the baby's responses to two brief separations from and reunion with its parent are analysed to classify the child as securely or insecurely attached. Secure infants show moderate distress on separation, and greet their parents and quickly settle on reunion. Insecurely attached children are classified as either avoidant (no distress on separation and avoiding or ignoring the parent on reunion) or ambivalent/resistant (high distress on separation but resistance and difficulty settling on reunion). Crittenden (1992a) has modified the Strange Situation Procedure for use with young children. In the present study, Crittenden's classification system was used to identify two five to six year old children, with each of the above two insecure attachment styles. During this procedure, the child was videotaped during three successive brief separations from the parent, during which a stranger was introduced who played with the child, as directed by the child. The child's attachment behaviours during separation and reunion were then analysed and classified. Once classified, the children had ten weekly play therapy sessions with the investigator as play therapist. Each session lasted approximately 45 minutes. The play therapy 6 sessions involved the child being introduced to a variety of play and art materials, including puppets, dolls, drawing and painting materials, etc. The therapist-investigator maintained a primarily non-directive approach, in that the children were free to participate in or to refuse whichever play activities they chose. It is expected that through the developing relationship with the therapist and the self-exploration and expression in play that the child naturally presents and resolves issues of concern. These sessions were videotaped by a research assistant. Field notes were taken following each session in order to record the investigator's impressions and observations during the session. Participant-observation was used for data collection and a revelatory/exploratory case study approach was used in analysis (Yin, 1989). Open coding (Strauss & Corbin, 1990) was used in videotape and field note analysis since it was not known what would emerge in sessions pertaining to therapeutic alliance and attachment. Attachment behaviour classification served as a guide in tracking the children's affiliative behaviours towards the therapist. Any changes in attachment behaviour during the course of therapy were noted as well. Following open coding, the development of therapeutic alliance with each of the children was compared and contrasted for possible relation to their respective attachment styles. Emerging patterns were noted and documented. Given that this is an exploratory study, a holistic, qualitative research approach is 7 important in order to get close to and participate in the clinical phenomena under investigation (Wolcott, 1990). Rich description of the phenomena is needed in order to generate hypotheses leading to further research in the field of play therapy. Purpose This preliminary research into the clinical relationship between attachment style and therapeutic alliance in play therapy would prove to be an impetus for further research into the clinical applications of attachment theory to therapeutic approaches with young children. The therapist's use of self and the therapeutic relationship is an area long neglected in the literature and is of particular relevance in application to how to assist children in reforming a secure base from which to develop a secure attachment with their caregivers and others. The field of counselling psychology would gain by a greater clarification of the therapeutic phenomena that emerge in the process of play therapy. In summary, the purpose of this study is to explore the nature of the relationship between therapeutic alliance and attachment behaviour as it emerges phenomenologically throughout the process of play therapy; and to generate hypotheses towards further much needed play therapy research, thus contributing to the field and informing clinical practise. 8 Chapter Two: Literature Review In the following chapter, an overview will be provided of each of the key constructs relevant to the current study, particularly as they relate to play therapy. Specifically, therapeutic alliance and attachment will be defined and the author will provide a brief overview of the development of the constructs, clinical applications, relevant research, development of assessment instruments and relevance to the field of play therapy. There is a severe lack of both theory and research in the literature on these two constructs, particularly in relation to child populations. The current study will limit itself to addressing the therapeutic relationship in individual therapies (as opposed to group therapies or day treatment programs); the therapeutic modality to which therapeutic alliance is usually applied in the literature. Individual child therapy is usually applicable to school-aged to adolescent children. There is a particular lack of theoretical and research focus on these constructs as they apply to this age group. Specifically, the literature on therapeutic alliance applies almost exclusively to adult populations. In the area of attachment, there has historically been much more focus on the attachment patterns of infants and adults than on older children. It is hoped that the current study will generate further interest and hypotheses for exploration in further 9 research with this age group. Play therapy, as a treatment modality with children, is also lacking in research, probably given that it lends itself to a more time-consuming, qualitative research approach. While play therapy is highly contextually embedded and multidimensional (making it a particularly challenging form of therapy to investigate), this makes it all the more important to identify and inter-relate the essential therapeutic factors and processes. For the above reasons, a participant-observer, exploratory and qualitative approach is appropriate to the current study. Therapeutic alliance and attachment theory are both very relevant to the field of play therapy. In therapy with children, due to their cognitive and developmental limitations, the therapist must rely less on verbal abstraction, interpretation and reflection on past events, and more on use of self in the lived experience of relating to the child. This lived therapeutic relationship (or therapeutic alliance) becomes a primary focus in counselling practise with children. As will be shown in the review of the literature, a large number of children referred for therapy are impaired in their ability to attach securely to their caregivers, impacting on their ability to relate to others, including the therapist. An application of attachment theory to play therapy processes could further clarify how the therapist's relational style impacts on children with different attachment styles, and vice versa. 10 The author has developed a keen interest in investigating these relationships and processes as a result of clinical experiences in play therapy with young children. Both use of self in the play therapy relationship, and children's different attachment styles as played out in relating to the author as therapist, seemed to be powerful factors at work in the therapeutic process. It became intuitively clear to the author that a bidirectional process was occurring, whereby each child, depending on his/her attachment pattern, seemed to both require and elicit a unique approach in responding to the child. Without the benefit of a research approach to analyse the therapeutic process for possible differences between children, these theoretical questions or hunches would remain at the intuitive level. Play therapy is viewed by many clinicians as primarily an art rather than a science and is therefore very dependent on intuitive awareness in the therapeutic process. However, in order to lend more power and efficacy to play therapy practise, the art of the therapeutic 'dance of relationship' will have to yield to the science of scrutiny and analysis afforded by research. Therapeutic Alliance Definition. The definition of therapeutic alliance is controversial, and has been operationalized differently, depending on the theoretical and clinical perspective 11 of researchers. The construct is psychoanalytic in origin, and was first introduced by Zetzel (1956). She describes it as a manifestation of transference, wherein the client identifies with the therapist and collaborates in the treatment process. She advocated that the therapist take a more empathic and supportive role than the traditionally more objective, interpretive stance of the psychoanalyst. Since then, client-centered theory has further contributed to this relational component of therapeutic alliance. Bordin (1979) introduced a second, more functional component of therapeutic alliance, as it is commonly conceptualized; that of the working alliance. He described it as including three features: (a) client and therapist agreeing on the goals of therapy, (b) each having clear, assigned therapeutic tasks and (c) developing a relational bond. Furthermore, he proposed that while different goals, tasks and bonds would be required in different types of psychotherapy, the overall strength of alliance more greatly determined effectiveness than the type of therapy chosen.. Hartley (1985) identified a point of consensus among differing theorists, in that most recognize the two primary components defining therapeutic alliance as being a real or genuine relationship and the working alliance. Not only is there disagreement on wether to emphasize the relational or the more technical component, but researchers differ as to which is more responsible 12 for the development of therapeutic alliance; client characteristics, therapist characteristics or the interplay between the two as a system. Frieswyk et al. (1986) summarize this ongoing controversy in their research review and call for a definition that exclusively focuses on the client's role. The perspective that will be adopted in this study is that of Short and Boon (1990), who advocate focusing on the bidirectional and dialectical nature of the counselling relationship. As mentioned in the introduction, therapeutic alliance will here be defined as the development of mutually responsive and affiliative behaviours between client and therapist. The bias here is more to the relational than to the functional aspects of alliance. Clinical applications. As mentioned above, Bordin (1979) proposed that the strength of therapeutic alliance has a more powerful effect on therapeutic outcome than the choice of therapy style. His initial work on alliance spearheaded much needed research in the area, which has since largely confirmed his assertions. In their research reviews Frieswyk et al. (1986) of the Menninger Foundation and Hartley (1985) summarize the predominantly supportive findings that the degree of therapeutic alliance is positively correlated with positive therapeutic outcome. Unfortunately, none of these studies involved child clients. Nonetheless, this is a construct that is very clinically relevant, for both adult and child populations. 13 Given that disruptive or dysfunctional relationships in childhood can affect the capacity for developing helping relationships later in life; according to Frieswick et al. (1986), "It is of major research and clinical interest to discover how to establish a therapeutic alliance under such conditions and to determine whether the development of such an alliance can modify the remnants of those early experiences and alter the patient's posttreatment ability to establish relationships." (p.33) This also speaks to the potential for a strong therapeutic alliance to treat and possibly alter a client's insecure attachment style. If Bordin's (1979) proposition that the degree of fit between client or therapist characteristics and the type of alliance called for by the type of therapy, affects the strength of alliance that develops is true, then understanding and fitting the components of therapeutic alliance to appropriate clients would help strengthen the potency of the therapeutic relationship. A more complex question, but one that applies to the current study is how the interactional process between client and therapist (rather than just their individual characteristics) impacts on the development of alliance. Short and Boon (1990), call for further research addressing this question, as well as the need to integrate an objective, research approach in counselling psychology with the more subjective "craftsman" clinical approach. 14 Relevant research. Hartley (1985) mentions that much of psychotherapy process research involves variables that overlap with those related to therapeutic alliance, but that until recently there have been few studies focusing exclusively on alliance. A recent development is research on the development of alliance assessment scales, and on testing their dimensionality, reliability and validity. There is less research dealing with it as a process variable (of more interest to this study) than as an outcome variable. Hartley calls for the use of a circumplex model in such interactional process research, in that by arranging behaviours or attitudes of the client and therapist on a surface divided by two axes, it allows for analysis of multifactorial relationships. One such process research study using this assessment model compared the therapeutic relational processes in two child therapy cases which used different types of therapy (Hellendoorn, van Eijk, Vollaard & Harinck, 1989). They point out that in previous related research, therapy style has been seen as a therapist variable and not a function of therapist-client interaction. In their study, 4 therapists using 'imagery interaction' play therapy and 5 client-centered therapists conducted 11 to 12 therapy sessions each with two to three 6 to 10-year-old clients. Using a circumplex model of multifactorial analysis, they looked at how child behaviours, therapist behaviours (both rated by videotape microanalysis of 5 second intervals) and child-therapist interaction during two sessions related to each other and to the type of therapy used. Clients were matched for sex, age arid social rank. The assessment scale used withstood tests of validity and interrater reliability was significant. A number of differences were found in therapist behaviours between the two therapy styles. Client-centered therapists tended to verbalize and silently attend more to the child and were involved in less active play. The play therapists on the other hand, engaged in more imaginary and preparatory play. Differences in the children's behaviours tended to parallel those of the therapists but were less marked than between the two groups of therapists. In addition, children in client-centered therapy tended towards more functional than imaginative play. In terms of interactional differences between the two types of dyads, the client-centered therapists and children tended to more complementarity; i.e. one playing or talking while the other watched or listened. On the other hand, the play therapists and children tended to more active play together. Both types of therapy interactions displayed symmetry and continuity (as either therapist or child would change to another behaviour, the other tended to adopt the same or complementary behaviour). Not only did therapists affect the children, but children affected the therapeutic activities of their therapists. Symmetry was somewhat stronger in play therapy than in client-centered therapy. 16 These results would tend to confirm Bordin's (1979) stance that the type of therapy has implications for the particular qualities of the therapeutic alliance and fit between both client and therapist and the therapy style. Clearly, therapeutic alliance takes on a different flavour in the two types of therapeutic relationship studied; regarding the bidirectional flow of alliance, as well as its goals and tasks. In fact, the authors propose that clinicians take these results into consideration in choosing appropriate therapy styles for children, depending upon their diagnostic or personal characteristics. This study powerfully illustrates the value of researching the bidirectional nature of the therapeutic relationship, as advocated by Short and Boon (1990). Development of assessment instruments. A brief discussion will follow of some trends in the development and research on therapeutic alliance scales that appear in the literature. A discussion of scales that pertain to play therapy will be left to the section on play therapy, given that these are not specifically identified as therapeutic alliance scales. In 1985, Hartley identified future directions for research in scale development. While she found most showed fairly good levels of internal consistency and reliability; cross- validity with use by groups other than the developers of the scales was found lacking. Another difficulty mentioned was that most of the tools randomly select a few brief segments of sessions out of the whole course of 17 therapy for analysis, presumably assuming that therapeutic alliance is fairly consistent across time. Finally, most of the scales use a multidimensional approach, recognizing the complex nature of alliance. However scale scores usually collapse much of the detail, possibly concealing some of the different interactional ways alliance can be expressed. The following two studies illustrate some of these strengths and weaknesses. Marziali, Marmar and Krupnick (1981) conducted an outcome study using their own developed alliance scale which scored how therapist and client individually experienced and related to the other. These scores were then related to independent measures of therapy outcome. Only client scores were significantly related to outcome. Specifically, those who made positive contributions to therapeutic alliance, had positive outcomes and those making negative alliance contributions had poor outcomes. A weakness of this study was that it failed to address the possible interaction between client and therapist contributions to alliance. Another study by Horvath and Greenberg (1989) was conducted to assess the validity and reliability of their Working Alliance Inventory. Summarizing the results of three studies as well, they found preliminary confirmation for the instrument's reliability and validity. Three subscales based on Bordin's (1979) three components of alliance (goals, tasks and bonds) were highly correlated, casting doubt on the distinctness of the 18 components measured. Despite this apparent weakness of the scale, it does contribute to an understanding of possible different aspects of therapeutic alliance. Relevance to play therapy. Play therapy is a form of child therapy that relies heavily on the interactional processes between child and therapist. As noted above, Hellendoorn et al. (1989) found a greater bidirectional symmetry in imaginary play therapy than in the more verbally focused client-centered child therapy. While the author's bias is to emphasize this interactional dimension of therapeutic alliance in play therapy, further process research on alliance in play therapy will shed light on whether child characteristics, therapist characteristics or interactional characteristics of therapeutic alliance predominate in this therapeutic mode. A possible child characteristic relevant both to the field of play therapy and to this study is that of attachment style. While Frieswick et al. (1986) applied this to adult populations, there is a particular need to discover how to tailor development of therapeutic alliance with children with respect to their attachment styles, as well as what aspects of alliance might impact on their future relationship patterns. One study that investigated these clinical relationships was conducted by Mahon in 1994 (see review in the section on play therapy). Outcome research on alliance in play therapy is also needed to confirm the positive outcome - alliance correlation found in the literature, as applied to child 19 populations. An application of Bordin's(1979) concepts of goals and tasks in the working alliance would give direction to the clinical practise of play therapy, as greater clarity is needed in the field regarding goals and outcome of therapy (Landreth, 1991). Finally, the field of play therapy has lagged behind in the development of relevant alliance assessment tools. While many of the existing scales were developed to be applicable across therapeutic modalities, most do not translate well for use in play therapy. Other weaknesses of existing trends in assessment, as they apply to play therapy include: (a) analysis of brief segments of therapy can miss the complex and changing nature of alliance in play therapy, (b) important nuances of child - therapist interactions may be missed by composite scale scores and (c) while multifactorial scales capture many of the components of alliance, they are too cumbersome and time-consuming for clinical application for play therapy assessment. The challenge remains for the development of elegant yet comprehensive assessment tools that capture the subtleties of the play therapy process. Two assessment tools that focus on child -therapist interactions (Carmichael, 1993; Faust & Burns, 1991) will be reviewed in the section on play therapy. 20 Attachment Background and definition. Drawing on anthropological and non-primate research on instinctive, behaviour, Bowlby (1969) developed his theory of attachment. He initially focused on the nature of the child's tie to its mother, pointing to the biological and neurological bases of attachment behaviour, originally being seen as mainly to meet needs for food and protection. Summarizing these historical developments, Bowlby (1988) proposed what he saw as the primary function of attachment behaviour; that of maintaining access to an attachment figure as a secure base from which to explore. He says the attachment figure "is some other clearly identified individual who is conceived as better able to cope with the world..., is available and responsive [and] gives him [the child] a strong and pervasive feeling of security...." (1988, p27). The characteristics of availability and responsiveness of the attachment figure are widely recognized in the literature as those which promote secure attachment in children (Ainsworth, Blehar, Waters & Wall, 1978; Bowlby, 1988; Crittenden, 1992c; Main, Kaplan & Cassidy, 1985). The initial work on attachment focused exclusively on infants. Later, Bowlby (1988) proposed that attachment could provide children of all ages with an 'internal working model' of how to go about maintaining secure attachment. Main et al. (1985) and later Crittenden (1992b) expanded this model for 21 application to older children who are able to draw on the levels of representation and language to access their models of attachment to help regulate their behaviour in relating to others, especially attachment figures. These theorists have proposed that these working models endure, leading to a recent trend to examine the expression of these internal models seen in attachment styles in later developmental stages. Attachment styles. The hallmark of attachment behaviour, according to Bowlby (1969) is to maintain the attachment figure's proximity, usually by crying or calling for her on separation, smiling or using other engaging behaviours on her return and following or clinging to her. These infant behaviours become increasingly more complex throughout development. The degree of these behaviours on separation or reunion with the attachment figure is what distinguishes differing patterns or styles of attachment. Ainsworth et al. (1978) developed a laboratory procedure called the Ainsworth Strange Situation procedure that has since been used as the research standard for determining attachment styles, particularly of infants. The attachment patterns are elicited through a series of separations usually from the mother, after which the child plays with a benign stranger and is subsequently reunited with its mother. Videotape analysis of the child's behaviours reveal him or her to have one of three main patterns. Secure attachment is shown by 22 moderate distress on separation from the mother, warmly greeting her and quickly resettling to play on reunion. This is due to confidence in an available and responsive caregiver, allowing for bold exploration of the child's world. The second two styles are patterns of insecure attachment. Insecure-avoidant children show no distress on separation from the parent and ignore or avoid the parent on reunion, stemming from a lack of confidence in receiving a helpful parental response when care is needed. Insecure-ambivalent (or insecure-resistant) attachment is characterized by high distress on separation from the mother, but clinging, resistance to comfort and difficulty settling to play on reunion. While the insecure-ambivalent child has not given up on seeking care from the parent when needed (as has the insecure-avoidant child), he or she is anxious and uncertain if care will be provided. In fact, attachment figures for these children tend to be inconsistent in their ability to be available and responsive when needed. Mothers of insecure-avoidant children tend to be consistently rejecting of or withdrawn from their children. It is important to remember that attachments also develop with and are expressed to attachment figures other than mothers (Ainsworth, 1989; Bowlby, 1969; Crittenden, 1992c). This points to the importance of addressing attachment beyond infancy (the Strange Situation originally being used exclusively with infants). 23 Assessment beyond infancy. Crittenden (1992a), Main et al. (1985) and Schneider-Rosen (1990) later adapted the Ainsworth Strange Situation for use with older children. The author will focus on the classification system of Crittenden (1992a), as yet unpublished, for use with preschool-aged children, as it was used for the subjects in the current study. Crittenden found that the criteria and attachment behaviours seen in preschoolers differed with those of infancy, due to them having more developmentally complex semantic (with the development of language) and procedural models of self and attachment figures. Their models can incorporate more flexible and conditional goals for meeting their need for security than just maintaining the immediate proximity of the attachment figure. For example, they may now use self-calming procedures, the use of alternate caregivers and reference to predictable landmarks (e.g. after naptime). Bowlby (1969) said that these new capacities, such as to negotiate shared plans of action indicated the development of goal-corrected partnerships. Crittenden has also developed subclassifications of each of the three styles, the details of which may be found in the literature (1992c). Descriptions of the three styles follow with the corresponding Ainsworth letter classification identified (adopted by most classification systems). Attachment styles other than these three have been developed by different researchers but have not consistently been shown to be 24 valid and reliable, and are therefore not included for description here. 1. Secure patterns (category B): In attachment relationships, these children show their feelings without distortion and take responsibility for regulating them. They are able to resolve discomfort and display a trust in their attachment figure's availability and willingness to communicate with them. They confidently explore their environment and affiliative relationships. The subclassifications differ in the degree to which children seek proximity with and reassurance from their attachment figures. 2. Defended patterns (category A - insecure-avoidant): These children either avoid close contact with interfering and rejecting attachment figures, are care giving with those who are withdrawn and unresponsive or compliant and vigilant towards hostile attachment figures, depending on the subclassification. The goal of all subtypes is to avoid direct confrontation with and loss of access to their attachment figure. 3. Coercive patterns (category C - insecure-ambivalent): Depending on the subclassification, these children develop characteristic strategies to force the attachment figure to meet their wishes on the child's own terms. They can present as either threatening, disarming, punitive or helpless in relation to their caregivers. Ainsworth (1990) reviews the literature regarding theories for categorizing and assessment procedures for attachment beyond infancy. She calls for the 25 development of new assessment procedures because most are based on the Strange Situation and different types of tools need to be developed that are appropriate to later developmental stages. For example, she points out that the capacity of the Strange Situation to evoke necessary stress for eliciting attachment behaviours diminishes rapidly beyond the preschool years. She also recommends developing instruments that can be used in a natural environment and can track stability of attachment over time. Clinical applications. There is mounting evidence in the literature on the correlation of disturbed attachment with behavioural, emotional and relational difficulties in childhood and later in life. Neufeld (1995) points out that a common disruption to childhood attachment is that of divorce, and summarizes research evidence of multiple sequelae of divorce with children, which tend to increase with time. These include various externalizing symptoms, school and sexual difficulties. Clinical symptoms that have been found to be correlated with insecure attachment in children include increased externalizing behaviour problems in preschoolers (Greenberg, Speltz, Deklyen & Endriga, 1991), lower self-esteem of 6-year-olds (Cassidy, 1988), and higher rates of adolescent conduct disorder (Moretti, Holland & Peterson, 1994). The impact of the child's social environment on security of attachment is also highlighted by a study by Cicchetti and Barnett 26 (1991), who found that preschoolers who were maltreated had consistently higher ratings of insecure attachment over an 18 month period, as compared to their demographically matched nonmaltreated peers. Crittenden, Partridge and Claussen (1991) pointed to the impact of family functioning by finding that dysfunctional parents tended to have complementary attachment styles (i.e. type A with C) and their impact on the child was a transformation to the opposite attachment style in the child. Balanced, more securely attached families displayed more matching in attachment styles. This points to the possibility of cross-generational transmission of attachment styles (for good or ill). Given the evidence for attachment styles enduring over time (Bowlby, 1988; Cassidy, 1992c; Main, Kaplan & Cassidy, 1985), the clinical impact of insecure attachment can be far reaching. Main and Cassidy (1988) found attachment classifications at age 6 to be stable since infancy. The application of attachment theory to therapy and treatment programs is increasing in recent years. Success has been found in the use of attachment-based treatment models with sexually abused boys (Friedrich, 1995), conduct-disordered teens (Moretti et al., 1994), children of divorce (Neufeld, 1995), developmentally delayed children (Mahon, 1994) and traumatized children (James, 1989). Relevance to play therapy. The underlying processes affecting the development and evolution of 27 children's attachment styles point to the suitability of play therapy as a therapeutic modality and to bring about change in the child's internal working model of relationships. First of all, their is a belief amongst attachment theorists that despite the relative durability of attachment styles, they are not carved in stone and are amenable to change through corrective experiences (Crittenden, 1992c; Main et al., 1985; Weitzman & Cook, 1986). There is also agreement that children attach to significant individuals other than their mothers, as well as later in life (Ainsworth, 1989; Bowlby, 1969). This points to the potential impact of the therapeutic relationship on attachment, particularly since, as has already been pointed out in the preceding literature review on therapeutic alliance, attachment formation between client and therapist is considered a key component of alliance. Since the process of formation of insecure attachment in children involves the restriction of access to relational information or experience discrepant with their internal guiding models ("defensive exclusion" per Bowlby (1988)), and is a result of concrete experience rather than abstract processes; the treatment approach recommended should be experientially focused, promote reality-based perceptions of self and others and increased communication of information about the lived therapeutic relationship. Thereby a more healthy attachment model could be adopted by the child at the unconscious level at which it operates. Play therapy is well suited to meet all of these treatment criteria. It is 28 experientially and relationally focused, operating more at the unconscious or . preconscious level rather than a more abstract verbal level. Child-centered play therapy in particular (to be outlined in the following section), has as its goal and tasks the experiencing and activation of the real self through experiencing one's own feelings and experiences, in the context of the therapeutic relationship. Brody (1978) developed a play therapy treatment program for developmentally delayed children that is based on an attachment model. Through offering corrective experiences with a therapist, who models the availability and responsiveness of a healthy attachment figure, children learn and slowly begin to operate on an alternate and more secure model of attachment behaviours. Relevant research One study that relates attachment to the development of therapeutic alliance was conducted by White, in 1994. A group of 29 adult clients underwent a course of individual therapy at two university counselling clinics with 17 therapists-in-training. Before the initiation of therapy, the clients' attachment styles were classified. After the 3rd and 6th sessions, the clients and therapists independently rated their perceptions of the level of therapeutic alliance, using a self report questionnaire. Her results indicated that insecure-ambivalent clients tended to initially rate alliance high, followed by a drop in alliance ratings by the 6th session. There was an opposite trend among insecure-avoidant clients; that is they 29 tended to give initially low followed by higher alliance ratings. Despite these trends, the only significant difference in therapeutic alliance relative to attachment styles was that among avoidant clients; the degree of avoidance being predictive of lower alliance scores after the 3rd session only. Other relevant findings were (a) that for all clients, low initial alliance scores significantly increased and high initial scores significantly decreased following the 6th session, (b) there was no significant difference between client and therapist ratings and (c) there was no significant difference between client ratings of alliance dependant on therapist's experience level. While it cannot be assumed that similar trends would be found with a child client population, these results raise possible implications for application to child therapy. Possibly, insecure-avoidant children would bemore resistant to establishing therapeutic alliance and insecure-ambivalent children might initially overidealize the therapeutic relationship. If the degree of therapeutic experience has a similar null effect on children's perceptions of the therapeutic alliance, then this would support a greater emphasis on relationship per se and the therapeutic use of self rather than on expertise in techniques in play therapy. Further play therapy research may shed light on these hypotheses. 30 Play Therapy Definition. Landreth (1991) provides an excellent overview of the field of play therapy, including its historical development. He describes play as "...a voluntary, intrinsically motivated activity involving flexibility of choice in determining how the item is used." (p. 13). Its functions include providing an outlet for active and creative energy, practise and development of skills, symbolic manipulation of the child's environment over which he/she often has little control (thereby allowing for integration and mastery of experiences), and finally, giving concrete form and expression to the child's feelings and inner experience. Axline (1969) has said that play is the child's natural medium of communication. Toys can be seen as the child's words and play as the child's language. Landreth defines play therapy as: "...a dynamic interpersonal relationship between a child and a therapist trained in play therapy procedures who provides selected play materials and facilitates the development of a safe relationship for the child to fully express and explore self (feelings, thoughts, experiences and behaviours) through the child's natural medium of play." (p. 14) There are several rationales for using play therapy with children. Play therapy has been found to be an effective form of therapy for many diagnostic categories (Landreth, 1991). Since it has already been shown that therapeutic 31 alliance is key to the process and outcome of treatment, and since attachment with the therapist is key to developing therapeutic alliance, it follows that the therapist must express availability and therapeutic responsiveness to the child. The therapist can only be responsive to the child in a developmentally appropriate way by using the child's own primary mode of communication; which is play (Axline, 1969). Finally, the child's world is one of action and direct experience rather than one of abstraction, therefore in order for the therapist to gain access to the child's inner world, he or she must enter it through joining with the child in active play. Background and approaches. Landreth (1991) provides a historical overview of the different developments in the field, as follows. Until the turn of the century, children's maladjustment was usually attributed to deficits in their education or training. The psychoanalytic tradition introduced the idea that children's problems could be emotional in nature and that therapy could be indicated. Play therapy was initially analytic in nature, used as a means to gain access to the child's unconscious and provide interpretation. This approach is still used by psychoanalytic sand play therapists. Release play therapy was the second development, which provided a structured play approach that encouraged the child to reenact and play out stressful situations believed to be at the root of the child's problem. It is more prescriptive and directive in nature and is still incorporated into what is called 32 focused or directive play therapy. The advent of relationship therapy in the '30's lead to the development of relationship play therapy, Moustakas (1959) being one of its main proponents. The focus of treatment moved away from the child's history or unconscious processes and onto the present, lived therapeutic relationship. The approach was a departure in terms of being nondirective, noninterpretive, and based on a belief in the child's own capacity for constructive change and growth. An outgrowth of this approach was nondirective play therapy, developed by Axline (1969) and based on client-centered therapeutic style. It is currently also referred to as child-centered therapy (Landreth, 1991). It is similar in focus and approach to relationship play therapy, but focusing even more on the child's responsibility for his/her own growth. As this is the approach used in the current study, it will be described in more detail below. The two main styles of play therapy in use today are nondirective or child-centered and directive or focused play therapy. Focused play therapy is based on different assumptions about children; that they frequently do not have the cognitive and emotional skills to effectively gain mastery of their difficulties without the guidance of the therapist. It is also favoured by advocates of a recent move towards brief therapy, as a nondirective approach can lead to long-term therapy. The therapist's role involves activation of unconscious processes and challenging children's defence mechanisms through structured play and other 33 approaches, such as cognitive-behavioural strategies, use of metaphors, bibliotherapy, art therapy and role play. James (1989) uses a focused approach with traumatized children, coaching children in identifying and sorting out feelings related to the trauma, and providing education and perspective through corrective messages. Rasmussen and Cunningham (1995) call for the integration of both focused and child-centered techniques, depending on the child's characteristics and presenting problems, the therapeutic needs and stages of the therapeutic process. Some new trends in play therapy include group play therapy and filial therapy (training parents in the use of play therapy principles with their children). Child-centered play therapy. This style of play therapy is not dictated by technique but is mainly about following the child's lead and being fully present with the child in the therapeutic relationship. Rather than use of interpretation or direction, the main therapeutic technique used is reflection of the child's feelings embedded in the play therapy process (Axline, 1969; Landreth, 1991; Moustakas, 1959). It is based on the assumptions that the child is capable of solving his/her own problems and that it is not developmentally appropriate to focus on verbal direction of the child (Axline, 1969). It is essential to the therapeutic process and relationship that the therapist convey his/her confidence in this tendency to self-actualization to the child 34 (Landreth, 1991). The goals of child-centered therapy include the child acheiving increased self responsibility and self determination, increased self awareness and self esteem and an increased ability to constructively problem-solve on one's own (Axline, 1969; Landreth, 1991; Moustakas, 1959). The following eight basic principles (summarized for brevity) for guiding treatment were developed by Axline and have been adopted by other child-centered play therapists. 1. The therapist uses a warm, friendly approach with the child. 2. The therapist must accept the child exactly as he/she is. 3. The therapist is permissive of the child fully expressing all feelings. 4. The therapist reflects these feelings to the child, who gains insight. 5. Responsibility to make choices and change problems is the child's. 6. The child leads the way; the therapist follows and does not direct. 7. The therapist allows therapy to proceed at its own pace. 8. The only limits are reality-based and to maintain child responsibility. Attachment and therapeutic alliance in play therapy. Given that child-centered play therapy is based primarily on the therapeutic relationship, attachment and therapeutic alliance are particularly relevant to this type of play therapy. In fact, Frieswyck et al. (1986) emphasize that the therapist's open and honest use of self is widely recognized as being central to the development of therapeutic alliance. The therapist's genuine use of self is a 35 hallmark of the child-centered approach (Landreth, 1991). The fit between clients' differing attachment styles and degrees of dependency and therapeutic style in the alliance is also relevant to therapy with children. There are indications that the therapeutic bond is particularly relevant to the ambivalently attached or overtly dependent client, but that while the bond between therapist and the avoidant or counter-dependent client might initially interfere with alliance, it is of increasing importance in later therapeutic stages (Bordin, 1979; White, 1994). If these trends are found with children in play therapy, it could be that the child-centered therapist would need to adapt the use of the eight principles to the child's attachment style. For example, it might be necessary to initially be less reflective of feeling with an avoidant child, who might interpret this as overly close and intrusive, only being open to an immediate and warm relationship later in therapy. Ambivalent children, anxious for an available relationship that meets their needs on their own terms, would more likely welcome the therapists unconditional acceptance and responsiveness, but there could be a danger of overidealizing the relationship. Carter (1992) described how the early attachment patterns of sexually abused 6 to 9-year-olds in play therapy affected the strength and length of time needed to develop therapeutic alliance. She also found that play therapy afforded a means to mediate the anxiety provoking aspects of relating to a caring and 36 available adult, for insecurely attached children. Their habitual reliance on significant toys to provide for their attachment needs acted as a buffer, and allowed them to approach the therapeutic relationship at their own pace. The therapist needs to remember the reason for the rejecting behaviour of avoidant children and still remain warm and available. Conversely, it may be necessary to limit the ambivalent child's overdependence and clinging to the relationship with the therapist, keeping the child responsible for experiencing his/her own feelings (James, 1989). Relevant research. While outcome and particularly process research on play therapy is particularly lacking (Faust & Burns, 1991), more needed research is beginning to occur. Related to the focus of this study, there is a need for research into the therapeutic processes conducive to therapeutic alliance in play therapy, and how these interact with attachment patterns of children and therapists. Carmichael (1993) developed an interaction matrix (the Carmichael Therapist/Client Interaction Matrix) for clinical use in analysis of the relationship between therapist verbal responses and children's behaviours in play therapy. Therapist response categories were rank ordered from least to most desirable according to child-centered therapeutic standards. Child behaviour categories were rank ordered according to desirable therapeutic outcome. Videotape 37 microanalysis was done of five to seven sessions for each of 5 children. Significant interactions were found, however most of the interactions fell in a category that was found to be neither conducive nor detrimental to the therapeutic process (therapist tracking responses/child rapport behaviours). In general, therapist response desirability was significantly positively correlated with desirable outcome child behaviours. Given that this was only a preliminary study on a very small sample, no conclusions can be yet drawn about the validity or reliability of this instrument. One weakness is that it only tracks the therapist's verbal responses, ignoring the more active play responses (particularly salient to play therapy). Faust and Burns (1991) reported their development of a clinical process and outcome scale that tracks change and interactions in child and therapist behaviours. They presented two case examples of implementation of their Nova Assessment of Psychotherapy scale, one using non-directive child-centered play therapy with a 5-year-old boy and one using directive play therapy with a 6-year-old girl. Child behaviour categories were "verbal" and "nonverbal" and therapist categories were "facilitating" and "channelling". The child-centered therapist showed significantly more reflection while the directive therapist made significantly more interpretations. The child in nondirective play therapy showed an outcome of a significant reduction in conversational language. Given that both 38 children had speech and language difficulties, this might not be desirable in this case, however it would be useful for further research to investigate how the degree of verbal vs. active child behaviours in play therapy is both related to positive therapeutic outcome and to the type of play therapy used. One final play therapy study of interest looked at therapist and child attachment behaviours for 7 children enrolled in a play therapy outpatient treatment program. Significant changes from insecure to more secure attachment behaviours towards their therapists were found for 5 of the 7 children. It was also found that the therapists' behaviours to the children were the same types of interaction behaviours seen in attachment figures of securely attached children (i.e. available and responsive). These are encouraging results for further research, despite the limited generalizability of these findings. It is hoped that child-centered play therapy is a treatment modality that encourages healthy and secure attachment modelling by the therapist, as well as a shift to more secure attachment in children treated. In summary, the constructs of attachment and therapeutic alliance both have an important impact on the process of play therapy. Therapeutic alliance is comprised of how client and therapist cooperate on therapeutic goals and tasks, as well as the development of a real relationship between the two. Attachment is the 39 process by which a person bonds to and seeks out a significant other seen as more competent than self, as a source of protection, security and a base from which to explore; and can be seen as a component of therapeutic alliance. While a child's primary attachment figure is usually its mother, in the context of a safe and trusting therapeutic relationship, the child can attach to the therapist. Play therapy (and in particular child-centered play therapy) is particularly conducive to a child developing attachment and a therapeutic alliance with the therapist. It is expected that how the therapeutic alliance develops in play therapy will partly be a function of the relational expectations and models the child has developed relative to his/her particular style of attachment. This study is intended to make a preliminary exploration of these interactions between attachment and therapeutic alliance in play therapy. Because this study will examine previously uninvestigated interrelationships, it will be exploratory and qualitative in nature, without fixed hypotheses. Two children of differing insecure attachment styles will each undergo 10 sessions of child-centered play therapy with the author. Participant-observation will be used, to allow for a close at hand examination of the therapeutic phenomenon, in a natural therapeutic setting. Open coding will be used for data analysis, to allow for identification of emerging patterns and relationships between constructs. 40 While no firm hypotheses are offered, the author has several hunches about possible patterns that might emerge. While avoidant or defended children are motivated to avoid confrontation with attachment figures, yet maintain their presence, the child in the study classified as defended might display more vulnerability and tentativeness in initially establishing the therapeutic alliance. The child classified as coercive (analogous to the ambivalent attachment pattern) might be more initially engaging in the therapeutic relationship, possibly even showing dependency in play with the investigator. It is expected that these patterns would shift at later therapeutic stages, however these shifts may not occur in the present study due to the limitation to 10 sessions. If later stages were reached, it would be expected that some secure-attachment behaviours might eventually emerge, following more relational engagement with the defended child and a possible shift to resistance and distancing by the coercive child, as anxieties in both children are triggered through the authentic, lived relationship with the therapist. Chapter Three: Methodology 41 Research Approach The current study is qualitative in nature, employing an exploratory case study approach. This allows for primary exploration of phenomena in their real life context, when the two are not easily separated (Yin, 1989). The use of participant-observation by the author (also described by Yin) allows for access to therapeutic phenomena that would not otherwise be available to investigation. It also allows for direct manipulation of key events and situations. The use of multiple case study was chosen to compare the phenomena under investigation between children differing on the dimension of attachment style. Selection of Subjects The two children involved were recruited from the 4 to 6-year-old subjects used in an attachment study in progress (Head, 1996). Recruitment letters were sent to parents of all the children, indicating that two children satisfying the classification criteria would be selected. Tim Head had used Crittenden's (1992a) attachment classification procedure, which involved assessing children's behaviours during three successive separations and reunions with mothers, the first two of which involved play with a stranger, and the last involving solitary play. Letters of parental consent were obtained and summaries of the play therapy 42 process were offered to the parents. Videotape ratings and classifications were completed independently by Head and Crittenden (with complete agreement in classifications for both children) and appropriate selection made by Head, withholding the attachment classification from the author. Two children were chosen who displayed moderate degrees of insecure attachment behaviours, one a girl classified as an exemplar of the defended pattern and one a boy classified as an exemplar of the coercive pattern. The intent was for the author to be blind to subjects' attachment styles until the completion of analysis, however Head inadvertently revealed the childrens' attachment styles prematurely (after initial analysis of individual childrens' data but prior to comparison of their emerging phenomenological patterns). Pseudonyms were used for the children's names in all field notes, coding notes and throughout the body of this paper, to protect confidentiality. The girl will be referred to as Annie and the boy as Ryan. Following the completion of the study, these notes were destroyed and videotapes erased. Subjects' Attachment Classifications One subject was a girl, aged 5.2 years at the initiation of play therapy and classified with a defended (subtype A3 - caretaking) insecure attachment style. The raters both determined this to be the primary pattern, but found some limited evidence for subtype A1-2 (inhibited) attachment behaviours as a secondary 43 pattern. She completed 10 consecutive play therapy sessions without interruption. According to Crittenden (1992b), children with this classification experience heightened and constant distress due to their caregivers' withdrawal and rejection. This leads to them limiting their own and others' access to their feelings as well as increased ability at picking up on the needs of others. This leads to behaviour characteristics towards attachment figures including; a brittle or false brightness, nurturance, keeping the self or other busy to maintain proximity but avoid genuine involvement, and initiating or tolerating closeness in order to please. The second subject was a boy, aged 5.3 years at the initiation of play therapy and classified as having a coercive (subtype C2 - disarming) insecure attachment style. There were several interruptions in his course of play therapy. After the first session, he contracted pneumonia and could not continue for 1 1/2 months. Once recovered, we reinitiated the full course of 10 sessions, at which time he was 5.4 years of age. There were two-week gaps between sessions 1 and 2, and sessions 6 and 7, due to the family's scheduling conflicts. According to Crittenden (1992b), children with this attachment classification maintain a heightened awareness of their feelings and an egocentric perspective in order to coerce their caregivers to meet their needs through coy and disarming behaviour. They are thus limited in their capacity to take others' perspectives. Behaviours towards attachment figures include shyness (e.g. head down, thumb-sucking, 44 pulling on clothes), flirtation (e.g. coy looks, whispered requests, babyish voice), and disarming behaviour (e.g. proffering smiles, gifts or over tenderness). The author-therapist's adult attachment style was classified as secure/autonomous (subtype Fl - some setting aside of attachment) by Tim Head, using Main and Goldwyn's (1990) Adult Attachment Classification System, which uses an interview format. Play Therapy Procedure A course of 10 weekly play therapy sessions, of roughly 45 minutes duration each were conducted in the common room of a university based child development centre in Vancouver, B.C. The author was a participant-observer in the role of play therapist and had exclusive access to the "therapy room" when sessions were in progress. A nondirective, child-centred style of play therapy was used, where free choice in the use of play materials was offered to the child, and no direction or limits were given to the child except as required to maintain safety, conform to limitations of the setting or to support the child's initiative (e.g. to offer help with a toy after the child's own unsuccessful attempts). The course and process of therapy was explained to the parents and structural information was given to the children at appropriate points in therapy regarding timing, duration, limits, expectations and termination of therapy. The children were initially introduced to the play materials which included dress-up 45 materials, toy phones, baby bottles and toy swords (for dramatic role play); building blocks and play dough materials (for constructive play); a Bobo doll for aggressive cathartic play; drawing and painting materials for creative, expressive play; and a sand tray with categorized miniatures for symbolic sand play. The author-therapist's background in play therapy included a 7-month field placement in an elementary school, doing play therapy with 5 to 8-year-olds. A similar 7-month practicuum experience occurred in two mental health centres with 4 to 9-year-olds. These both occurred as training requirements in a university Masters program in counselling psychology. Two trained assistants (psychology students, one of whom was Tim Head) were used to operate the video camera and were present in a corner of the room at the start of each session. A third lay assistant was used in one session due to the unavailability of the regular assistant. It was intended to use one camera operator for the entire study, but a replacement needed to be trained due to the study going beyond the anticipated deadline (one child was ill for 11/2 months). Observation Methods and Data Collection The author/therapist maintained field notes shortly following each session, to record felt impressions of the therapeutic alliance and recollections of the relational process occurring in sessions. These were subjective and personal in nature, including not only signs of alliance noticed in the child but responses 46 evoked in the author/therapist. Primary data collection was through videotaping of sessions. This permitted full immersion of the author in the participant role during sessions, without sacrificing more detailed observation that occurred subsequently in videotape analysis. Video camera operators were trained to keep focus on both the therapist and the child when possible, to capture their interrelating as well as individual behaviours. During less active sequences, they were directed to close in on therapist and child, to more accurately record changes in expressed affect. Videotape records were transcribed by the author for both verbal and visual (or action) content of sessions. Detailed observation of videotape records was made for the unfolding relational patterns of each child. Analysis Preliminary analysis used in this study involved open coding, or "the process of breaking down, examining, comparing, conceptualizing and categorizing data." (Strauss & Corbin, 1990, p. 61). In other words, the author examined the data for identifiable phenomena or concepts, which were then grouped into categories and interrelated. This then lead to the identification of patterns among the concepts, related to the research question being posed for study. Specifically, the author made a preliminary examination of the data through 47 immersing herself in the sessions; by detailed examination of videotapes, field notes and recollections of the sessions. As a result of observing the data, criteria for breaking down and describing sessions emerged; including the specific narrative themes in the children's play, the emotions they expressed and how, the frequency and ways they made contact with the author-therapist, and whether they assigned a role to the author-therapist in interactive play. Perceptions of the developing child-therapist bond were seen to be described in field note recollections of the sessions. These phenomena were all described and documented in coding notes. Following this preliminary conceptualization of data, the literature was reviewed on attachment classification (Crittenden, 1992b; Schneider-Rosen, 1990) and on therapeutic alliance scales (Benjamin, 1974; Hellendoorn et al., 1989) in order to sensitize the author to possible categories related to the question under investigation. This was done to provide guidance (but not to dictate) the categorization of data and thereby avoid getting bogged down in phenomena unrelated to the research question. The author then returned to an examination of the preliminary data (videotape records and field notes) and coding notes in order to look for emergent ways of categorizing phenomena seen and described so far in the children's play. Some of the categories from the literature fit the data (e.g. expression of affect), some were too broad (e.g. changes in attachment behaviour), 48 and some were beyond the scope of analysis in this study (e.g. synchronicity between therapist and client). Through repeatedly examining and comparing data, groupings of data and categories emerged and were recorded. The author then reflected on these categories for their adequacy to fit the data and for their redundency. This lead to several categories being eliminated, collapsed into others or revised. Once categories, concepts and the dimensions along which they varied were identified with each child, they were compared and contrasted with each other between children as well. Yin (1989) describes this process as pattern-matching, which can lead to the generation of hypotheses and ideas for further study; these being a few of the main rationales for the current study. Through pattern-matching, higher order categories for comparison of patterns were identified. Coding notes were used to record each successive level of categorizing and regrouping the data, as they emerged. Finally, the author examined the data and the various coding notes for similarities and differences in the children's themes of therapeutic alliance (or ways of relating to the author-therapist), further relating these differences to their different attachment styles. This final step in analysis was hoped to shed light on the question under investigation of how the development of therapeutic alliance in play therapy differs between children with different attachment styles. 49 Chapter Four: Results As both author and participant-observer, for the ease of readability and narration of the children's process of play, I will assume the first person throughout discussion of results and analysis in the following two chapters. First will follow summaries of each session for both children; including a description of the relational process and content, followed by my main impressions of therapeutic alliance and the main narrative themes that emerged. The summary descriptions are drawn from videotape records and transcriptions, and my impressions are drawn from field notes taken after each session. This will constitute the main body of data used in analysis in the next chapter. Summary of Sessions: Annie Session 1. Annie was introduced to the play materials, which she was very tentative in exploring, settling on the sand tray materials. At the very end she became slightly more bold and played with the art materials. She tended to only make direct eye contact with me if there was a toy physically between us, otherwise avoiding eye contact. It took her a while to settle to a sand tray narrative, initially asking me to identify and find the various miniatures, offering overly bright or incongruent 50 smiles. She slowly started to create her own story after I redirected her to explore on her own. She alternated between taking the first person and narrating the sand tray drama as an observer. Her narration was initially in a very subdued voice, while I offered tentative reflection of the emotional content (e.g. commenting with a scared voice about the octopus eating baby animals). During these reflections, her glances to me were initially furtive and later more long and intense, followed by her initiating tentative emotional content and a more animated tone of voice, as she narrated her play. She became very animated at the end of the session following me summarizing her sand tray story and complying with her request to write it down. The narrative themes included: separations and reunions between baby animals and their mothers, who also sought out places of safety to hide together; being contained as one watches others on the outside; dangerous, "bad" animals chasing and eating the babies; and rescuing or protecting-animals fighting off the attackers and trying to rescue the babies. My initial impressions were that she seemed overly eager to please me and her smiles seemed artificial. Her initial sand play seemed too nice and the relationship felt disjointed. In the second half, as she began to express more conflict, aggression and abandonment, her affect became somewhat more congruent. At this point, her frequent, tentative looks felt as if she was watching to see if I would give voice to the embedded feelings in her 51 play. By the end it felt as if their was a beginning therapeutic relationship. , Session 2. Annie enthusiastically and eagerly started the session (in contrast to her meek, tentative and restricted approach before). She initiated her own play, rather than eliciting my suggestions. In fact, on the occasions I did give feedback or suggestions, she corrected me or made a different choice. She explored the rest of the play room and decided to make a "zoo" out of building blocks and puppet animals. Her narrative and her tone of voice began to congruently reflect feelings in her play, rather than just looking to me to do so. Her glances to me were now to engage me in play rather than just looking for my reflections. For example, she now assigned me a role, as the "zooman", to protect and nurture the animals. She watched me carefully as I carried out my role and reengaged me if I slacked off. Using a baby tone of voice, proximity and eye contact, she made frequent bids for me to take care of the animals. While she was more congruent and assertive in her play, the few times when I ventured to reflect her own or my feelings (e.g. "You look happy today" or "I'm glad you had fun last time") she became tentative and quiet again, though seeming pleased. On one instance she was aggressive to me while we were in role (her bad guy attacking the zoo-man). The main narrative themes included: seeking protection and containment (e.g. in zoo cages) versus feeling isolated and lonely; alternating between 52 rest/sleep/safety and danger/attack; seeking nurturing versus independence; the empowerment of victims; and mothers and babies as victims together. My felt impressions were that her excitement was more congruent and less falsely bright. She showed more range of affect and assertiveness but was hesitant with new behaviours such as expressing negative affect or directing me. She was physically closer a few times, but this seemed guarded and without warmth. Session 3. Once again, Annie was definite in making her choice to build an "aquarium" using the blocks and animal puppets. Rather than her previous indecision at points of transition, she moved fairly freely between episodes and covered more of the playroom. She moved several times from the "aquarium", to sand tray, to dress up, and finally to paints. She immersed herself more intensely in character, for example making lots of animal noises or using a sing-song voice. She also sought less proximity to me, but she was also more active. However, for the first time she initiated direct contact with me through the toys (for example, an eagle puppet landed on my head several times). I was assigned a caretaker/protector role again as the "whale master" at the aquarium. She had lots of questions about whether I provide and care for the toys, as well as if other children came to play. For the first time my structuring went beyond modelling or telling what the play options were to more limit setting (e.g. needing to wear a paint smock). "53 The narrative themes seen were: observing/being observed from within an enclosure; seeking freedom from constraint; the empowered victim again; abandonment or neglect by mother versus care from a surrogate; and the sharing of secrets. I neglected to take field notes of felt impressions following this session. Session 4. Annie's play showed less verbal narrative and more animal noises, tactile enjoyment of the materials, and several direct requests for caretaking or nurture. While I fulfilled some of the requests (care of a cut, giving a playdough recipe), I needed to set limits with others (not permitted to take home a toy bear). Once again there were frequent transitions between playdough, playing "Christmas time" with presents being shared, puppet play, sword play, sand play and playing school. She took most of the initiative in her play choices, but looked to me to give feedback on both her messy and creative play, through making eye contact, particularly during playdough play. I was assigned roles in more of her play, not only as caregiver/protector but this time as aggressor as well (in the role of a witch). I also acted as coach in her exploratory play, when she looked or handed me materials she was struggling with, and I helped her in her discoveries and creations. The narrative themes in her play included: ambivalence about powerful 54 figures (friendly and helpful, then dangerous and neglectful); ambivalence about helper figures (rescuing others, then hiding from danger themselves); aggressive themes; and expressions of physical and emotional neediness. My impressions of this session were that Annie was becoming less overpleasing and docile and more direct and assertive about her wants, both in the play process and directly from me. At the same time I sensed more turmoil and conflict expressed in her play. Her play seemed more emotionally primitive. I was aware of her trying to get close to me through the play and the toys, but stopping short of relaxing in a truly trusting relationship. There was no directly interactive play yet. When my help was needed, she would ask for it indirectly or passively (e.g. stating the problem and then looking intently at me). Session 5. While Annie continued to play out episodes of attack on the weak or baby characters, the majority of her play involved characters getting nurtured in trusting relationships. She played out baby animals suckling from their mothers and being fed. She sought out more direct relating and eye contact with me, starting the session with a long eye to eye chat about her Christmas presents. At the end of session she directly requested if I would ever give her a present. She maintained intense focus on me while we mutually made and exchanged Christmas cards instead. She initiated direct interactional play with me for the first time, having us 55 play out going camping together (sharing food, setting up a tent and sleeping together), rather than playing our roles through toys. During this intimate play segment, her physical and eye contact with me was more limited, however. The narrative themes included: attack by aggressors now due to them not having enough to eat rather than for hurting; nurturing themes predominating; needy, weak and neglectful mothers (e.g. a mother giraffe sleeping while her baby was attacked and eaten); and a need to be vigilant yet still ignorant about danger. My impressions of the sessions were that Annie had formed a basic level of trust with me. I enjoyed the session, and the previous halting and tentative quality to previous sessions was now largely gone. My previous sense of being manipulated to meet her needs through indirect strategies was replaced by a sense of her starting to directly communicate her needs in our relationship. There was a beginning sense of her authentic self emerging. Session 6. Annie displayed some indecision at the start of the session, looking about the playroom but making no bids for my suggestions. She finally said that we would play "petshop" by building one with wooden blocks. I was assigned the role of petshop owner and she played the animals who awaited new homes. The majority of this session involved her playing out animals getting chosen or passed over and their ambivalence once chosen to go to a new home. I reflected many of 56 the feelings but they were quite overtly enacted and she directly expressed a number of feelings herself both verbally and with congruent affect (fear, loneliness, sadness, excitement and happiness). During intense active segments of play, she made little eye contact, but when there was a lull in the action, made a lot of eye contact and sat very close to me much of the time. She inadvertently touched me several times. Our tone and quality of voice matched much of the time. She chose to finish off with painting different objects that she asked me to identify by guessing, turning it into a sort of game. During this segment she became very animated, and laughing with my guesses. The narrative themes revolved around the main story line of baby animals awaiting a new home and their ambivalent feelings about it, including: wanting a new home where they would be fed, yet fear that their owners might not be kind or care for them; wanting to be picked and sad when passed over by customers; wanting the safety of their cages versus wanting to break free; and babies losing their mothers when their cages were broken, being rescued by surrogate mothers. My impressions of the session were that she was making more eye contact and there was a strong sense of mutual engagement and her vulnerability as powerful issues such as abandonment surfaced in the play. There was a sense of natural flow to the process rather than the self-consciousness seen previously with intimacy in the relationship. 57 Session 7. The session began with a long period of gentle, mutual teasing, conversation and sharing about her Christmas at home, including questions about the other child in the study, who was her friend. She chose to play "pet shop" again, skipping and jumping as she gathered the needed toys. This time the pets shared a greater enjoyment of friendship and many of the customers came in as happy warm families. A sense of loss when leaving their pet shop friends and a sense of renewed hope each day that remaining animals would get chosen were expressed freely. Many feelings were directly verbalized in the drama, and acted out. For example, she took on a whiny, plaintive tone of voice in character, when the animals were abandoned. Her eye contact was slightly less but she was also very engrossed in play. She sat quite close to me through much of the session. The energy level varied with the narrative content and flowed in a smooth manner through the highs and lows. Narrative themes were very similar to last session and new ones included: family members being reunited; a sense of rhythmicity (sadness at night/ renewed hope each morning, pet shop closing and opening, sleeping and waking, etc.); animals hiding for safety from devouring "bad guys"; and magical helpers coming to the rescue. My impressions were of a natural and warm flow to the therapeutic relationship and process. She seemed spontaneous and animated, but no longer in 58 a placating way. She seemed quite affectionate with me, though without much direct physical contact. Her use of eye contact seemed to change in that she would pause mid flow and take long searching glances at me, as if trying to size up something about our relationship. Session 8. Annie came galloping into the session and after brief and animated chatting, including mentioning Ryan, she decided to play "forest animals"; about the dangers for animals and people of getting bit or killed. She played the role of an attacking mosquito and penguin, vigorously expressing aggression to me in role as well as to other animals. I was assigned to be the "zooman" protector. As she was very animated and expressive of feeling, I backed off in my energy level. However, as she continued to display a greater range of expression, my energy and affect level came back to match hers later in the session. At the end, termination planning got started, at which she appeared upset (silent, frozen face, trembling jaw) but avoided the discussion, followed by "You know, if a mosquito bit you on the nose, you would die!". She earlier had indirectly expressed physical affection, hugging a stuffed bear repeatedly while looking at me intently. Despite little eye contact, she maintained close proximity to me through most of the session. At the end she did a painting of boxes, asking me to guess what was in them. They were pictures of forest animals (used in dramatic play) for her to 59 remember them by. Narrative themes were: babies being abandoned and looking for but finding the wrong mother (a repeating theme); aggression (biting) being used to both kill and bring back to life; aggressors hurting others only to feed their young; and the pervasiveness of danger in the forest. My impressions were that there continued to be a felt sense of a growing spontaneous and natural therapeutic relationship. The greater range of expression and spontaneity extended to her comfort in expressing what seemed to be anger and aggression to me (albeit veiled in the play). However my impression was that she, was very vulnerable in discussing termination and for the first time chose to return to the waiting room to her mother without me accompanying her. Her comment about a mosquito biting me seemed to conceal her anger and fears of the loss of the therapeutic relationship. Session 9. At the start of the session we discussed upcoming termination, at which Annie looked teary, and following my disclosure of expecting to miss her, she repeatedly expressed the same with sadness, and requested a picture to remember me by. After talking about rituals for ending she decided to play "birthday party" to which all the characters used in therapy were invited, and her mood brightened again. The session was largely devoted to the birthday party theme and the feelings related to one animal that was excluded and subsequently took violent 60 revenge on the others, but finally was accepted back into the group. I was assigned protector and participant roles. Annie maintained close proximity to me (at times very close) but alternated between stretches of concentrating on the unfolding drama with little eye contact and periods of making eye contact following emotional segments. This was done with directness and repeated after I reflected her play. She continued to directly express a wide range of feelings and there was a close match in energy and feeling tone between us. At the end of the session, during more talk of our last session and while painting "presents" that contained playroom toys, she expressed the wish to visit me. Narrative play themes include: competition versus cooperation between friends; isolation and rejection versus inclusion; finding safety and strength in numbers; and celebrating connections between family and friends. My felt sense of the session was that a special relationship had been forged, where Annie had both discovered and learned to communicate to me the range of her feelings; from aggression and competition to joy and generosity. I had the impression that her eye contact was an invitation (now confident rather than tentative) for me to reflect the feelings and themes in her play. I felt as if I had received a gift in seeing her real self begin to unfold, and shared her frustration at the relationship ending at this point in therapy. 61 Session 10. The session began with acknowledgement of termination and an exchange of gifts. Annie then decided to play "school" and I was assigned the role of teacher. All the animals were assembled to learn about various "creepy" animals, then went to the forest to find food. The animals then returned to their cages to go back to the pet shop. We finished with parting rituals and she shared her wish for continued contact. This session was characterized by movement and rich variety in the feelings expressed, energy level and closeness or distance, easily flowing from one segment to the other. There was a match in our energy levels and response to each other's play. She energetically corrected me when I misunderstood her and gave me several directions about how she wanted me and the play to proceed. The narrative themes were again similar to last session and also included: containment for protection now becoming entrapment; loss of freedom and home; fear of returning to a restricted life; and seeking ways to hold on to a new freedom. My impressions of this last session were strong. I felt we terminated well but that her sadness and sense of loss were almost too much for. At the same time, she seemed very assertive and direct in expressing her wants to me. The termination felt very premature, happening in the midst of the working stage. Despite this it felt that she was much more self expressive by this last session and 62 that an attachment had been formed between us. Summary of Sessions: Ryan Session 1. Ryan's very first session occurred 11/2 months before the full course of play therapy, due to him being home ill. Both sessions are documented, as both were beginnings for him. In his first introduction to the playroom, he gradually engaged in aggressive play with the Bobo doll and offered single word responses to my reflections of his play. He indirectly invited me to join him in sword play by the use of coy glances and gestures, responding shyly when I reflected his wish. He frequently stopped to see if I would continue playing and responded with delight when I did. While his solitary play was vigorous and energetic, he was initially tentative in interactive play, becoming more animated as he saw me actively respond to his initiative. He then chose sand play, during which he was quite verbal. Again, I played the role of observer/commentator and he did not engage me to actively join him. He did not attempt to extend the play time. Although somewhat shy, he seemed fairly relaxed and natural in his play. After his illness and upon return, he settled quicker to the play process and was more verbal. After initial avoidance, he made relaxed and occasional eye contact with me though he did not direct me towards interactive play. While he 63 stood back at first, he later moved directly closer to me. He started again with the Bobo doll, then spent most of the session in sand play, followed by exploration and use of the play dough and markers. He was physically active during sand play, using me as a helper or coach, though making little eye contact. His voice was whispery and subdued and at times sing songy. He asked a lot of structure questions about the toys and how the materials could be used, as well as expressing curiosity about Annie's play time. I needed to impose some limits on his aggressive Bobo play, which he somewhat resisted, but complied with. The narrative and play themes in his second start at therapy included: him assuming the aggressor role, killing and supplanting his attackers; competition for power and dominance; him taking a protective role towards "special" babies; and babies successfully seeking protection from strong parent figures. My impressions of the two sessions were that he seemed fairly relaxed and natural in his play, though initially shy and somewhat reluctant to engage with me in play. It struck me that he seemed to use eye contact to make bids for me to stay engaged and responsive to his play, but that he was hesitant to have interactive play. Session 2. Ryan missed a week due to a medical appointment. He was very eager to start the session but started off tentatively watchful with aggressive Bobo play. 64 He made frequent eye contact and I responded by giving my attention and reflection of his play. He responding with broad smiles as I did so. With my encouraging reflection of his play, he became more energetic and showed more emotion in his play. He needed limit setting again, when getting overly aggressive with the Bobo (damaging it). Once again he subtlety resisted but complied. The rest of the session was spent in solitary sand play, asking me to assist (e.g. find toys) as I provided commentary. He moved about a lot, frequently sitting quite close but with limited eye contact, and making inadvertent body contact. His voice was frequently subdued and whispery, but often with my lead of reflecting the energy and emotion of the narrative, his voice became more animated. This time, he tried to stall the end of the session, asking to continue his play, but then charged out of the playroom ahead of me. The narrative themes included: aggression and competitions between father and son characters; babies successfully seeking safety with parents; helper/guards protecting babies and attacking powerful animals; striving for competence and recognition with denial of risks or failure; and him assuming the role of dominant aggressor humiliating powerless opponents. My impressions were that his themes of competition were fuelled by expressions of revenge and feelings of powerlessness. He also seemed to be tentatively sharing more of his inner conflicts and feelings in the play, despite not verbalizing them. His silence at / 65 times felt as if he expected me to provide the narrative and reflection of content. Session 3. As of this session, a different video camera operator was used to tape the remaining sessions.. Ryan made more invitations for my involvement in interactive play, for helping, reflecting or attending to his play, though most of his strategies remained indirect. For example he showed me the toys or narrative action, alternated between asking for my help and solving problems independently, invited me to join him through eye contact and responded enthusiastically when I reflected or joined his play. He required limit setting again to maintain physical safety and asked once again about Annie (who had just terminated). He played closely with me in several forms of dramatic, aggressive play, responding with delight as he repeatedly scared, defeated or overpowered me (in dress up, puppet play and Bobo play). In sand play, his voice was louder and more animated than before, and I toned down my own voice, allowing him to take the lead. At the end of the session, he resisted ending and walked out with me rather than running on ahead. The narrative themes were: facing repeated risks and dangers in trying new things and persevering; running out of power; denial of failure; people and things getting damaged and fixed; ambivalence about seeking help when in danger versus maintaining independence; and returning home for safety and rest. My 66 impressions were mixed and may have been coloured by termination with Annie and his energy still being affected by illness. I sensed a greater engagement on his part in the therapeutic relationship yet his energy seemed to wane at times and I sensed myself struggling to engage with him at times. Session 4. Ryan engaged me in interactive play, made frequent direct requests for help and for the first time invited me to take a challenging aggressive role with him in sword play. He frequently tested and explored the limits of play I had set around safe use of the Bobo. There was frequent mutual gentle teasing and laughter. He became more assertive (but not aggressive), refusing my suggestions or help a number of times, in favour of figuring out how to use play materials on his own. He would confidently look to me with a smile, after successfully meeting a difficulty without my help. After interactional sword play, he spent the remainder of the session in more exploratory play and sand play. He focused on exploring and experimenting with the sand tray miniatures, as well as creating a narrative drama with them. For the first time he directly expressed characters' feelings (fear). There was a mutual sharing of reflection and elaboration of the play themes, during which he was very engrossed and made only occasional eye contact. He moved about the area freely and often moved close to me. At the end of the session he did not resist finishing and we walked out together. 67 The narrative themes included: shared aggression (as opposed to only him being in the dominant role); danger/risk versus nurture/healing; denial of fear or failure when risks were taken; war with clearly identified teams; parent/child competition; and retreat to home for play and strengthening after defeating attackers. My felt impressions in the session were mixed again; both sensing Ryan enjoying the therapeutic relationship as he drew closer to me, and sensing him holding back in both the relationship and in self expression. None the less, I perceived a relaxed atmosphere between us. He seemed to look to me for praise of his discoveries and play skills. Session 5. He used most of the toys and media (sword play, sand play, Bobo, puppets, dress up and painting), confidently announcing each time he wanted to change his play. He maintained close proximity frequently, but only made frequent eye contact during aggressive interactive play. During these episodes he required some safety limit setting again as he used the whole room beyond the defined play area for the first time. He was more independent during sand play as he sought out toys without my help, focusing on manipulating them and gaining mastery. Only later did he make direct requests for help. His play was more expressive of emotion, showing congruence of tone, voice and affect; and for the first time he expressed feelings verbally through the characters (primarily fear and loss). Later 68 he responded to my probes as to what play characters were feeling (fear, frustration, anger, embarrassment). For the first time, he purposely touched me indirectly through the toys; expressing aggression (an eagle bit me) and affection (a puppy kissed me). He finished off with sequential drawing and painting of increasingly complex and detailed boxes, and a self portrait. He left the room ahead of me after having little lead time from me this time. The narrative themes included: nurture and comfort; growing reliance on helpers providing safety/healing in the midst of danger; fears of loss of power to dominating others; and some acceptance of failure with renewed effort to master skills. My impressions of the session were that he had greater emotional spontaneity and congruence, and a renewed vigour and presence in the relationship (as opposed to his initial whispering tentativeness). His engaging with me continued to be limited to the play context however, otherwise avoiding my presence or eye contact. In the context of play however, I sensed a warm therapeutic relationship. Session 6. This session was held on the morning after the regularly scheduled time, due to mother's work conflict. The session got off to a strained start, with Ryan being withdrawn and quiet, as it was early in the morning and there was a different camera operator filling in. Once he had gathered his usual key toys about him he 69 quickly engaged with me and became very animated. This session was characterized by his use of the entire room, beyond the play area, as well as testing the limits again in aggressive play. He waited for me to establish where the limits lay, without resistance. I also cheered him on as he explored new skills and risks. He did some brief sand play followed by a new use of interactive sword play. He invited me into a close and intimate stance towards him, followed by him assuming a dominant position, as he lead and defeated me while I dramatized his conquests. He really 'hammed it up', stomping around the room in delight and boasting over me. He initiated a long 'Simon says'-type game of me imitating his sword manoevers. He also invited indirect exploratory touching with the swords. His last long sequence involved building a house for cars on the floor, from which they would venture out in learning new car tricks, and return to for repairs after crashing up. He repeated this cycle over and over, using the whole room and even briefly venturing out the door (for which more limit setting was needed). The house had a room for each car and easy access both in and out. He finished with exploratory play and play dough. New narrative themes in this session were: alternating between testing out risks/developing new skills and returning to home base for rest, repair and . strengthening; empowerment without dominating another; seeking safety and protection at home (but now specifically with mother); and a new control of 70 aggression (sheriff figure) versus resisting control as dominance. My impressions were that while Ryan was more expressive and open in relation with me, he was less aggressive. He focused on home, family and safety in his play themes. In his car sequence, if he got too far from me, he seemed to stop to check for my attention and affirmation of his acrobatic moves. There was a genuine sense of mutual warmth between us as he seemed to seek out connection with me. Session 7. There was a week missed between sessions 6 and 7, due to mother's work conflict. He started with a non-verbal invitation to sword play, using at the most code words, assuming I would know my assigned role in aggressive and imitative play (similar to last session). He took on a boastful and exultant tone again as I cheered him on or dramatized my losses. He then shifted to a unique form of sand play that started out with him aggressively burying my sword in "garbage". This quickly shifted to a cooperative game of hide and seek to find and repair "broken" swords. His play slowly changed to a sequence where he directed us to fill up each other's swords with "warm sand", to warm them up from being buried in "cold snow". There was great animation and tenderness in his voice, range of affect and tender, long eye contact alternating with shy periods avoiding eye contact. He finished the session with a request for exploratory finger painting, for the first time. He frequently requested my help with paints and cleaning his 71 hands, versus doing it himself. He asked many times about Annie. Narrative and play themes were of: reparation; some initial dominating aggression; a sharing of healing, nurture and emotional warmth; tactile exploration and enjoyment; and exploring his neediness versus need fulfilment. My impressions of the session were that a shift had occurred from him using me to explore his need to move against or away from me, to moving with or towards me. I felt a new depth of tenderness and emotional bond between us. There was a genuine presence in the moment and spontaneous experiencing of the relationship. He seemed to not only seek out nurture from me symbolically, but offered to share of himself symbolically as well. At the end of the session, he also seemed to pull away less, chatting casually. Session 8. Ryan's play moved freely again among most of the play materials (sword play, sand play, painting, play dough, Bobo and cars), no longer announcing a change but just going to the new activity. He was energetic, assertive and expressive, showing a great range of affect and body posture. He alternated between being almost bossy with me to almost passively accepting my help at times (e.g. my encouragement to use a mat to brace himself as he slid on his knees to catch cars). His sing-song whispery voice was gone, however in this session his voice became subdued and quiet while playing out the cars destroying the 72 house he had built for them. At other times, he was extremely active and animated. He initiated ritualized sequences of play by handing me my toys as he took his first moves, then following each other as we played out complementary roles (for example, mutually filling each other's sword covers with sand, testing each other's swords, etc.). He did less aggressive play and he was more verbal, expressing the underlying power themes. He finished with painting boxes inside boxes, where I was asked to guess the contents; painting off the borders of the paper and onto the paper table cover. His narrative themes had changed in that they portrayed both the aggression, effort to master new skills and seek protection seen earlier, as well as a direct playing out of fears of failure, abandonment, being shamed and loss of home and nurture (not seen before). His characters portrayed these fears directly, along with a determination to try again. There was a greater sense of emerging danger and conflict between characters (that was denied before). My impression was of him continuing to more directly express his emotions and turmoils about the safety of relationships. He portrayed a relaxed, easy confidence in our relationship, in our established ways of relating and that I would follow his lead, yet maintain safe limits. There was the continued sense of warmth from last session and previously denied negative affect starting to surface. When I mentioned upcoming termination, he seemed upset (staring, silence, averted gaze). 73 Session 9. At the start of the session, he acknowledged my reminder of termination, but he avoided eye contact, declining to plan or talk about it. Once again, he made his play choices without announcement, choosing a number of interactive play rituals establishing contact and joining with me in play ("filling" each other's swords, imitative play). For the first time he declared mutual ownership over toys (saying "Our swords!" (emphasis his) in sand play) and offered help to me while we were in role. He also directly praised himself. He demonstrated creative and unique use of play materials; for example, paint brushes becoming sea animals, "painting" with water, and using a marker for a rolling pin. He buried "treasures" inside playdough characters, which he found and then invited me to find. His dress up play was now active, immersing himself in animal roles. In his Bobo play, as usual, Bobo was dominated and humiliated, but for the first time, he had Bobo get angry and fight back to defend himself. He required limit setting again and resisted ending the session, for example running out the door half way down the hall to retrieve a car, repeatedly opening and closing the door himself. The narrative themes included: symbolically discovering an inner "treasure"; acknowledgement of neediness; ambivalence in trusting mother figures; aggression versus nurture between family members; the start of self protection; and the needs for exploration and expansion of boundaries. My 74 impressions included an awareness of myself holding back reflection of his play until he expressed himself, as he was doing this freely and spontaneously on his own. There was a synchronous and natural flow to our relationship, which spilled out of the play setting fluidly for the first time. He seemed to be exploring himself, his range of power, affect and choice with little encouragement needed from me. Session 10. When I brought up this being our last session, Ryan responded with "No!" and an averted gaze. For the rest of the session, when I brought it up he avoided eye contact, had a subdued voice and his subsequent play was tentative and used his old sing-songy voice until he immersed himself in the play again and his exuberant expressiveness returned. While he enjoyed our parting rituals and gift giving, he was not able to talk much about saying goodbye. His initial play, as with the last few sessions, included all our established play rituals. He spent much of the session in sand play and finished off with mock games of football and baseball, where of course he was the winner as I cheered him on. His sand play was organized, involving bases for the two sides in a war, babies exploring and playing once the war was safely over, and races that were a bit chaotic and dangerous becoming safer once a policeman established and maintained the ground rules. 75 The narrative themes were: limits and boundaries providing safety versus previous dominance and overcontrol; competition as win/win versus win/lose; expansion of flexible boundaries with growth; trust in caregivers to maintain safe boundaries; and safe bases for protection and places from which to explore and play. My final impressions were mixed; including delight in Ryan's beginning self-confidence and self-awareness, as well as sadness at needing to end therapy prematurely and his resistance at ending. I regretted his inability to directly express and say goodbye to the therapeutic alliance that had been established between us. He seemed very uncomfortable with his vulnerable and upset feelings that surfaced in the termination process. Chapter Five: Analysis 76 Through the use of open coding in preliminary analysis of the children's play therapy process, several concepts emerged relative to the development of therapeutic alliance. These fell within either categories of play (types of play, narrative themes, boundaries of play, and expression of feelings) or categories of therapeutic relationship (contact with therapist, context of relationship, assigned roles, and decisions in play). As mentioned earlier in the description of methodology used in this study, the process by which these concepts and categories emerged involved first immersing myself in the sessions, through review of videotapes, field notes and recollections. I then noted and described the specific narrative themes, emotions expressed and contact made in each session. After review of the literature to sensitize myself to possible categories relative to therapeutic alliance and attachment, I returned to a review of the data and my descriptions to identify emerging concepts and how these could be grouped together into categories (whether they were found in the literature or not). This involved me identifying the characteristics and similarities or differences in phenomena I noticed in the play sessions, and how these compared between sessions and between the two children. For example, as I closely examined the data, I noticed that the children 77 had recognizable ways of using the space in the playroom. They also responded in unique ways to the limits imposed during sessions. It occurred to me that these were two concepts that could be grouped together in a category called 'boundaries of play'. While this category was not specifically found in the literature, it seems to relate to the ability to explore and exceed one's limitations versus the reliance on outside structure and guidance (both being areas related to the literature on therapeutic alliance and attachment). The emergent categories are defined below, and the different conceptual patterns that emerged in each child's play described. Through the use of pattern-matching, these patterns were then conipared and contrasted, revealing themes related to changes in therapeutic alliance. These different themes are subsequently presented. The detailed descriptions of categories, patterns and themes of therapeutic alliance will finally be synthesized in a portrait of each child, highlighting their differences in evolving relational patterns, relative to the question guiding this exploratory study. Given that this is an exploratory study and that the question for investigation regards the relationship between attachment patterns and therapeutic alliance; my analysis is based on an assumption that differences seen in the children's relational patterns in therapy is related to their differing attachment patterns. I recognize however, that other intervening variables may be significantly related to the development of therapeutic alliance 78 as well. Categories of Play Type of play. The children's play differed in what they played with and how they played. These differences were grouped into different types of play, for purposes of comparison. Descriptions of these types, with examples seen in the children's play include: 1. Dramatic play: This play was characterized by toys being used as characters in a drama, with a story line and narration provided. Examples include using miniatures to create a story in the sand tray and using puppets as characters in a story. 2. Aggressive play: While there was aggression seen in other types of play as well, there were play sequences where the sole activity was the acting out of aggression, such as simple sword play and aggressing on the Bobo doll. 3. Creative play: This type of play was characterized by the child making something, as opposed to manipulating or using a toy, for example making something out of playdough or painting. 4. Exploratory play: This involved the child exclusively exploring unfamiliar areas of the playroom or toys rather than using the materials in 79 interaction with self or other toys (such as investigating a sword and asking questions about it rather than using it in a sword fight). 5. Practise play: This involved the child repeatedly demonstrating and refining a particular skill in play, for example, repeatedly hitting a ball in a mock baseball game, and cars practising spinning manoevers. 6. Role play: This involved taking on a role or character one's self, rather than assigning the role to a toy as part of a drama. Dressing up and pretending to be an animal, without an ongoing story would be an example. 7. Interactive play: In these play sequences, the sole activity is the child interacting with me through play, without enacting a story, acting out aggression or making something, for example, the child inviting me to imitate him/her or to mutually play with the sand in a cooperative way. Annie's types of play were primarily dramatic and creative. She chose dramatic play in every session. Initially she played out dramas in the sand tray, later tending to use puppets and miniatures to create her stories in a defined space on the floor. Her creative play was spaced out every one or two sessions throughout the course of play therapy. While she tried out the playdough, she focused on painting, frequently involving me by having me guess what she was drawing, or painting with her. Practise play was not done at all and she chose the other types one to three sessions each, all in the first half of therapy. Her 80 aggressive play involved use of the sword, however this shifted quickly from pure aggressive play to the sword play being embedded in dramatic play with puppets. Her role play involved dressing up and pretending to be animals and her interactive play occurred on one occasion (having us play "going camping", setting up tent together, sharing food and getting into our sleeping bags together). Within each session, her dramatic play tended to last longer than other types of play chosen. Ryan's play therapy sessions consisted predominantly of aggressive play (in every session), dramatic play (in all but one session) and practise play (in all but two sessions). His dramatic play was primarily sand play, but in later sessions he also included stories played out on the floor with miniatures and puppets. His aggressive play consisted of Bobo play (him kicking, throwing, punching, flipping and bending the Bobo doll) and simple sword play (him the aggressor (except for one session) in initiating a sword fight with me). His practise play included practising and refining sword techniques (without aggression), and using toy characters to practise different skills (horses jumping higher and higher hurdles, skiers trying new jumps and cars practising high speed spins). His interactive play (playing together with the sand, and "Simon says" type imitative sword play) and creative play (using paint, playdough and drawing materials) occurred mostly in the second half of play therapy, while his role play was 81 scattered throughout and exploration of toys was at the beginning. Narrative themes. The children's dramatic play differed in the narrative themes embedded in their story lines provided to narrate their play. The types of themes are as follows: (a) safety themes about protection, rescue, home and care or nurture; (b) danger themes about the actions and responses of attackers and victims; (c) relational themes about parents, peers, separation/reunion, and belonging/isolation; (d) mastery themes about competition, competence and failure; and (e) exploration themes about exceeding boundaries and maintenance of boundaries. • 1. Safety themes: Annie's safety themes were as follows. Protection and hiding was frequently sought out either from my character or by staying in an enclosure (an animal cage or pen) or hiding under sand. Enclosures took on a different significance in later play however (entrapment or isolation). Rescue of baby animals from danger was provided by strong helper figures, but these were not parents (e.g. a zoo keeper or aquarium keeper). Care and nurture was frequently sought out (for example, a baby animal plaintively crying after losing its mother "Who will take care of me!?"), at times successfully by surrogate figures (an eagle babysitter or an animal foster mother). Rarely was care provided for by mothers or fathers. Themes regarding home displayed ambivalence; such 82 as in hoping to move from a petshop to a new home but not being sure if care or neglect would result, and if the old home would be missed. Ryan's safety themes showed protection consistently being sought and received from parents (except for ambivalence towards protection from mother in one later session) and rescue from danger usually coming from peers (usually as a team or army in a war). Examples include a baby horse being protected by its father during a war, while the men in their army successfully fought off the bad guys. Bids for care and nurture by characters in the stories didn't occur until later in therapy and in the last few sessions, the fear of losing nurture surfaced. Themes about home indicated home being a place to retreat to after danger is gone, to rest, strengthen and go out from to play and test out new skills. An example is; after a war, a baby deer had a sleep at home then spent a long time playing with friends outside in safety. 2. Danger themes: Annie played out attackers usually coming one at a time to steal or kill baby animals. Later the attackers' goal was to kill for enough food for themselves or their young. Danger was often pervasive, e.g. "...everywhere in the forest". The victims were usually baby animals and sometimes their mothers, who, until later play sessions, were unable to protect themselves and relied on outside help. Ryans danger themes were often in the context of a war, where the "bad guys" were trying to steal or destroy his side's "base". One to one attack was infrequent. The response of the victims was to kill and/or supplant attackers, 83 retreat to home and to persevere in the face of danger. In later sessions there was more use of helpers when in danger. 3. Relational themes: Annie frequently included mothers in her narratives, and portrayed them as abandoning, weak, neglectful and as victims. When looking for her, a baby animal would frequently mistake another animal for its mother, who would then often offer surrogate care. Characters were often isolated from their friends, either due to rejection (such as one dinosaur being excluded from a party by the others) or constraint (being locked in separate cages). In the last few sessions, reunions between friends and family members surfaced in the narratives. In Ryan's narrative play, parent figures were mostly sources of safety, competed with their children, and in later sessions were inconsistent in their nurture and at times aggressive to their children. Friends or peers were portrayed as competitors in testing skills and as team mates in fighting aggressors. There was little evidence of separations from home or family until the last sessions (for example a car's "home" being destroyed and broken apart). 4. Mastery themes: This theme only occurred in Annie's second to last session, when friends competed in various party games at a big reunion and celebration. Competing and attempting to win over others initially led to rejection and isolation, followed by later inclusion and acceptance. Annie's characters were initially admitted to but were passive in the face of failure (for example an 84 excluded dinosaur having to wait for permission to try again by the others). In Ryan's narrative play, mastery themes occurred in almost every session. Competition occurred frequently, such as in races, athletic competitions, and displays of skill to see who was better. The outcome was usually either winning power and dominance or losing (humiliation and defeat). The winner not only boasted about successes but emphasized the weaknesses of the loser. Until the last sessions, main characters tended to deny failures when they wouldn't quite reach the mark. By the end of therapy though, there were some instances of failures being admitted, followed by renewed effort. There were some instances as well of different characters sharing turns at winning and not gloating over each other's losses. 5. Exploration themes: Annie's narrative play depicted more maintenance of boundaries in earlier and middle sessions. Examples are where baby animals were kept in cages as their home and caretakers needed to maintain and repair the cages or doors after bad guys tried to break in. Safety took priority over freedom until about halfway through therapy. Freedom from the constraint of cages and seeking new, less restricting homes (e.g. with a family, living in the forest) emerged. However, strong conflict between fear of the dangers that come with freedom and longing for the safety that goes with constraint emerged as well. For example, in the last session, animals longed to keep their forest home despite its 85 dangers, but missed their old friends at the petshop, to which they had to return. In Ryan's narrative play, home was not portrayed as a place of containment but a place from which to freely explore and play. However, other boundaries emerged, as challenges to overcome. For example, physical boundaries to the playroom were exceeded by characters who tested new abilities, a baby horse jumped over the fence of a corral to run and play. Boundaries or rules were mainly portrayed as restrictive, but at the end of therapy they were also for safety ( for example, a car race becoming safe and orderly once a policeman jailed the reckless drivers and established road rules). Table 1 Narrative Themes Annie Ryan Safety rescue/protection by solitary strangers rescue/protection by groups, peers, family themes home as containment, isolation; later as hope for family home as base for exploration, rest; later fear of loss many bids for surrogate care and nurture consistent care at first; bids for nurture and care at end 86 Annie Ryan Danger themes Relational themes Mastery themes personal attack to steal, kill; later for food 1st victims powerless; later empowered; some hopelessness, pervasive danger, use of helpers mothers: victim, neglect, abandon, surrogates; peers: reject, isolated; later community, accept; separation/reunion, belonging/isolation a lot failure admitted in all; hope for success, competence, compete only later attack on team "base" for power, dominance 1 st victims empowered to keep fighting, regain dominance; later use more helpers mothers: little used, later some neglect; fathers: competitor, safety, protection; peers: compete win/lose later win/win compete 1st for revenge, power; later for personal competence; practise skills; failure 1st denied, some admission at end Exploration themes 1 st maintain boundaries, containment as protection; later hope/fear to exceed boundaries, fear of/hope for freedom 1 st exceeds boundaries, seen as domination, restriction; at end maintains boundaries, seen as safety Note. Narrative themes are presented separately from the remaining categories of play, to increase clarity of presentation. Boundaries of play. This category pertains to (a) the variety of their types of play; (b) their use of 87 space in the playroom; (c) their questions about the structure of play; and (d) their response to the play therapy limits imposed. Annie's initial play sessions showed more variety in the number of types of play and play areas used. Her play became more focused later in therapy, tending to follow a single dramatic line through most of each session. Her use of space involved staying within the defined play area in the playroom, not venturing beyond the play setup. However she was more spontaneous in her movements about the playroom in later sessions, skipping and jumping about as she gathered toys for play. Her questions about structure (e.g. how toys work, what they are) were confined to the initial sessions. She complied with limits set about specific play activities (such as wearing a smock while painting), but frequently tested the limits regarding the limits of the therapeutic relationship (no extension of number or length of sessions, or visits). Ryan initially used less types of play and play areas, later showing more variety in his play. He also physically expanded the boundaries of his play over time, as he gradually played outside the defined play setup in the rest of the room, and eventually venturing out the door and attempting to go down the hall (in his play). His use of materials also progressed beyond the typical boundaries of use, for example painting off the paper and turning functional toys into symbolic ones (a paint brush becoming a baby shark). He asked few structure type questions 88 about toy identities, this being limited to the beginning of play therapy. He generally complied (after more initial testing) with the length of sessions and never tested the limits of the therapeutic relationship. He frequently challenged limits set on aggressive play and on staying inside the room. Expression of feelings in play. The way the children expressed their feelings in play therapy varied in terms of several concepts which include: (a) the mode of expression; affect (facial expression), verbal expression and narrative expression (tone of voice in narrating emotional story segments); (b) the range of feelings expressed; positive (happy, excited, hopeful, etc.) and negative (angry, sad, afraid, etc.); (c) the intensity of feelings (how powerful or subdued the expression); and (d) the spontaneity of expression (flowing smoothly with the related play content or seeming disjointed from it). Annie's expression of feeling increased in general across sessions. By half to two thirds through the course of therapy she had almost reached her full expressiveness in all modes. She started out with only implicit feelings in her narrative play, but with little to no expression in affect, verbal expression or narrative tone. She was strongest initially in narrative expression, with verbal expression being the last to increase. Her affect was initially incongruent with the other two forms of expression (for example, a falsely bright smile while she 89 described a baby being hurt in a whiny voice). There was congruence despite low intensity by the 2nd or 3rd session. In terms of changes in her range of expression, she initially expressed more positive than negative feelings, but by the middle of therapy was portraying strong feelings of sadness, loneliness and fear. Anger was the last feeling to be expressed. Intensity and spontaneity gradually increased throughout sessions. Ryan's expressiveness also increased across sessions, but his range and modes of expression were slower to increase. His use of affect was initially incongruent with other forms of expression but was congruent by the 2nd or 3rd session as well. It took until the last two or three sessions before there was full expression in all modes, with verbal expression being the slowest to develop. There was also a gradual increase in range, however there were more negative feelings expressed early in therapy, with positive expression still being a bit less by the end of therapy. Intensity and spontaneity of expression were fairly quick to develop. Table 2 Categories of Play Annie Ryan Type of play 1 st- dramatic 1 st- aggressive 2nd- creative 2nd- dramatic 90 Annie Ryan less- other types in 1st 1/2 of sessions 3rd- practice none- practice less- creative, interactive in 2nd 1/2 of sessions Boundaries of play variety: 1st high then low variety: 1 st low then high use of space: 1st high then low use of space: 1st low then high structure ?'s: early only structure ?'s: early only limits: complied in play, tested in relationship limits: tested in play, complied in relationship Expression of feelings low at 1 st, high at end low at 1st, high at end intensity and spontaneity gradually increase intensity and spontaneity gradually increase more positive than negative at 1 st, anger last more negative than positive at 1st; sad, afraid later amount of expression quickly increases amount of expression increases more slowly 91 Categories of Therapeutic Relationship Contact with the therapist. This category pertains to the ways the children approached or attempted to make contact with me during their course of play therapy. These ways include (a) eye contact, varying according to duration (brief or prolonged) and directness (face to face or sideways glances); (b) seeking proximity (or physical closeness), varying in directness (with or without an object between us) and in degree of proximity; and (c) physical contact, varying in directness (with or without the use of an object) and in intention (on purpose or inadvertently). Annie's use of eye contact in the first session or two was brief and indirect (i.e. furtive glances, often with an overbright smile). Later on, during emotionally charged segments of dramatic play and when it was my turn to respond (in role), she frequently made direct, prolonged eye contact. This became particularly intense towards the middle of therapy, but towards the end during intense segments of play, she made little eye contact at all. Her use of proximity to make contact with me steadily and slowly progressed in closeness through the course of therapy and did not differ much relative to the type of play. Proximity tended to be indirect (i.e. with a toy between us) early on, becoming more direct later. She made physical contact with me a few times throughout therapy, twice indirectly (with a toy) and twice directly, all intentionally. 9 2 Ryan made prolonged and direct eye contact with me throughout the course of therapy, but almost exclusively during aggressive play. Later on in therapy, he also made direct, brief but frequent eye contact during practise play, after he had successfully completed manoeuvers or feats with the toys. Otherwise, his eye contact remained indirect and brief. During dramatic play in his first two sessions, he made almost no eye contact. His proximity seeking progressed in closeness from session to session, but again tended to be only during aggressive play. There was some marked proximity towards the end of therapy during interactive play (tender segments of him directing us to pour sand on and in each other's sword covers, to "get warm"). Indirect physical contact occurred fairly frequently during the second half of therapy, through the use of sand or toys and was intentional. Direct contact was made only once and this was inadvertent. Context of relationship. The context of the relationship has to do with the degree to which the child brought other domains of his/her life into the therapeutic relationship. This would include the child talking about family, friends or school, engaging me in some way outside of the therapy room, bringing personal belongings into play therapy or direct requests for contact or relating beyond the agreed upon play therapy sessions. Annie chatted with me on several occasions about aspects of her personal 93 life; including oblique references to her relationship with her mother, frequent references to Ryan (a friend of hers) and events at school or home. She walked out of the playroom with me at the end of sessions, consistently engaging me in conversation on the way to meet her mother in the waiting room (except once, after the first session in which I mentioned termination). She brought a toy from home to most sessions, which she would include in her play. She also made several requests to extend the relationship beyond the therapeutic context (such as requests for gifts, visits to her home or continued sessions). Ryan also made frequent references to Annie (his friend) and talked somewhat less about his personal life (two references to family members). It was only in the second half of therapy that he started to walk with me back to the waiting room after sessions, and only at the very end did he chat easily with me along the way. He tended initially to bolt out of the room ahead of me as soon as I announced the end of the session. He only brought a toy from home once, which was incorporated into play. Finally, he made no requests to extend the relationship beyond the therapeutic context. Assigned roles. The children tended to assign me roles in their play, to varying degrees and in different ways. Descriptions and examples of the assigned roles that emerged through open coding follow. 94 1. Participant. This role was assigned when a child invited me to join him/her in a mutual play activity as myself (i.e. not in character or to serve a particular function), for example pretending to go camping as ourselves, playing together with sand or painting together. 2. Protector. This role was usually assigned as part of dramatic play, where my character (either through a toy or myself) was given the job of protecting or in some way taking care of others in danger. For example, I was a caretaker at a zoo where the animals were being attacked. 3. Helper. This was a frequent role during creative or exploratory play, when the child-made frequent requests for my help or coaching in the use of play materials or in how to proceed in play. For example, asking for guidance and suggestions while experimenting with finger paints. 4. Aggressor. This role was assigned in dramatic play, when I was directed to play the bad guy or attacker, for example a witch trying to steal babies. It also occurred in aggressive play when I was told to initiate attack rather than just participate or respond to the child's initiative. 5. Cheerleader. This role was assigned mostly during practise play, where the child looked to me to comment on and praise accomplishments in play; for example, looking to me following every successful leap over a hurdle in play, asking for my comments. 95 6. Reflector. While reflection is a technique employed through most of child-centered play therapy, I include it as an assigned role, because on a number of occasions the children specifically requested me to reflect their play. Examples of this are requests to write down the narrative of their dramatic play and to engage in a long sequence of imitation of their play. Annie predominantly assigned me the role of protector, in dramatic play. She very clearly told me my title ("zooman", "whalemaster", "aquarium man") and handed me my appropriate toy character. She directed me either verbally or by eye contact at the appropriate dramatic moments, to protect, nurture and care for baby animals (usually in danger). I was assigned this role on all but two occasions. In the second half of therapy, she also used me occasionally as a participant, once in playing at camping and several times joining her in creative play. I was assigned an aggressor role in dramatic play once, as a witch. She occasionally used me as a helper or coach in creative and exploratory play. I was never given the role of cheerleader and I was asked to act as scribe for her narration of dramatic play in the first session (reflector). Ryan predominantly assigned me roles as participant and as cheerleader; and moderately as aggressor (in the second half of therapy). The role of participant was assigned in two interactive play rituals he established. One involved us playing with sand together in the sand tray and one involved mutual exploration 96 of the other in the context of nonaggressive sword play. He used me as cheerleader by looking to me or asking for feedback during practise play. I acted as aggressor during aggressive sword play. He occasionally made bids for my help or coaching during exploratory and creative play, and a few times (towards the end) used me as reflector of his initiatives in imitative sword play. Decisions in play. This category refers to how the children went about deciding on what play materials to use and how to use them. They varied in the degree to which they sought my guidance (asking for help, permission or suggestions) or chose on their own (for example not asking for guidance in decisions, or asking for and then ignoring or challenging my suggestions). Annie displayed moderate indecision in making play choices early on, later showing more decisiveness and frequently loudly announcing what the play would be about, for example "Today we're playing forest animals". She was also very decisive and direct in assigning me my play roles. She initially sought more permission and suggestions in exploring or choosing toys, frequently doing so indirectly, by commenting on materials she was unsure about and then looking to me. Throughout the course of therapy, she continued to alternate between relying on my suggestions during creative play or contradicting/not seeking my suggestions. 97 Ryan had lots of questions about the identity of toys in the first sessions, followed by direct requests in finding toys and direct and indirect requests for help in how to use them. By the middle of play therapy, he tended to experiment and problem-solve on his own, only asking for help if he still had difficulty. After asking help he would later on frequently contradict the suggestions offered, going back to independent problem-solving. By middle to late sessions, he was very decisive about choices and changes in play, but would increasingly just stop what he was doing and start the next thing, without announcement. By the latter stage of therapy, he was very directive of me in roles and the use of toys. Rather than directing me explicitly, he tended more to use mutually understood code words or gestures to indicate what he wanted me to do, for example "It's time for filling" (referring to using sword covers in the sand). Table 3 Categories of Therapeutic Relationship Annie Ryan Contact with therapist eye contact:brief, indirect at 1 st; later long, direct in emotional play, less in intense play eye contact: brief, indirect at 1 st and later; long, direct in aggr. play; brief, direct in practise play; less in intense play Annie Ryan 98 proximity: steadily increases in all play types; indirect at 1st, direct later physical contact: some in all stages; direct, indirect, intentional Context of personal talk: much, in all relationship stages own toys: frequent, in all stages further relationship: often, in all stages, more at end beyond playroom: after all sessions Assigned roles 1st-protector in all sessions 2nd- participant in 2nd 1/2 of sessions 3rd- helper in 3 sessions 4th- aggressor and reflector in 1 each proximity: steadily increases; seen mostly in aggressive play, later in interactive play physical contact: more in 2nd 1/2; indirect, intentional quite frequent personal talk: moderate, in all stages own toys: once, in middle stage further relationship: none sought beyond playroom: 2nd 1/2 only 1st- cheerleader and participant in all but 2 2nd- aggressor in mid to later sessions 3rd- helper in 5 sessions 4th- reflector in 3 of the last sessions not used- cheerleader not used- protector 99 Annie Ryan Decisions in guidance: permission, suggestions at 1 st, help later (more indirect) guidance: suggestions and help at 1 st (more direct), later help after tries self play independent choice: more later, with announcement; more verbally directive, challenge suggestions later independent choice: more later, esp. problem-solving, challenge suggestions, directive with gestures Pattern Matching Comparing and contrasting the development of the concepts in play between children and across time revealed the following similarities and differences. Categories of play. Annie primarily used dramatic and creative play, while Ryan used mostly aggressive and dramatic play, with a secondary emphasis on practice play. Narrative themes changed over time for both children. For Annie, initial themes were: (a) pervasive personal attack; (b) powerlessness, hopelessness and failure; (c) help sought through strangers; (d) isolation and neglect; and (e) safety as containment. New themes that emerged at different points later in therapy and in differing intensities were: (a) attack in order to meet a need; (b) beginning competence, power and hope; (c) help given through family; (d) community and family; and (e) desire for/fears of freedom. For Ryan, initial themes were: (a) 100 attack on the team base; (b) competition for power/dominance and failure denied; (c) help provided by peers; (d) family/home as safety; and(e) limits as restriction. Later themes that emerged to varying degrees were: (a) some fears of destruction of home; (b) competition for personal mastery, cooperation and failures admitted leading to problem-solving; (c) help sought through family; (d) fears of loss of home and maternal neglect; and (e) limits seen as safety. Annie's boundaries of play included initial variety and exploration in her play followed by a focusing of her play and increased movement. She tended to comply with limits on play activities but challenged limits on the relationship. Ryan on the other hand, changed from initially limited variety and moderate exploration to a great increase in both. While he tended to test the limits on play activities, he complied with limits on the relationship. Finally, Annie initially expressed more positive and later more negative feelings (anger expressed last) and increased fairly quickly in overall expression (particularly through narrative). Ryan initially expressed more negative and later more positive feelings, with a slower overall increase than Annie, finally lagging behind slightly in verbal expression. Categories of therapeutic relationship. Annie's contact with me increased over time in her use of eye contact and proximity, but not in physical contact. She made little eye contact in intense dramatic play but tended to make prolonged, direct eye contact during highly 101 emotional segments of dramatic play. Ryan increased over time in the use of all modes of contact, but also avoided eye contact during intense dramatic play. He made prolonged, direct eye contact during aggressive play. Annie frequently attempted to extend the context of the therapeutic relationship beyond play sessions from early to mid-therapy on. Other than some talk of his personal life and later accompanying me outside the playroom, Ryan on the other hand, made no attempts to extend the relational context. Primary roles assigned to me by the children (in more than half the sessions) were protector and secondarily participant for Annie (never cheerleader); and cheerleader, participant and secondarily helper for Ryan (never protector). Changes in decision-making patterns for Annie were a change from some indirect help-seeking to more independent choice (mostly verbal direction of me in dramatic play and some challenging of me in other play). For Ryan the changes were from frequent initial direct help-seeking to more independent choice as well (expressed differently as nonverbal direction of me through gestures and codes in most types of play, independent problem-solving and lots of challenging me). Changes in Therapeutic Alliance There is agreement in the literature on therapeutic alliance and on play therapy (as has been shown earlier) that the establishment of a working 102 therapeutic relationship is essential to achieving therapeutic goals. The children were seen to vary over time and between each other in the degree to which they moved towards or moved away from relating closely with me as therapist. This dimension could be called self-focus versus other-focus. This dimension of therapeutic alliance applies to four emerging themes in the children's patterns of relating to me in play therapy. These will be described and illustrated, followed by relating them to differences in the children's attachment patterns. Finally, two themes that emerged, pertaining more specifically to attachment behaviours will be examined. Guidance-seeking versus self-direction. At the start of therapy, both children tended to seek my guidance in identifying toys and for suggestions or directions as to how they could be used. They had different strategies in seeking guidance, however. Annie was more indirect, for example making statements rather than direct questions such as "My mother says I have to clean up"; whereas Ryan more frequently asked direct questions such as "How does this work". This trend was not absolute, with some use of the other's strategies. With time, they were both more self-directed in their play and relied less on me for play choices. My impression was that this was more important for Annie to communicate to me, as she usually made a big announcement of her play choice, whereas Ryan tended over time to casually go 103 to his next choice without announcement. This same pattern was seen in the way they directed me. They both challenged my suggestions later on (Ryan more so). In relating these differences to their attachment styles, defended children such as Annie tend to outwardly comply with, but inwardly resist the guidance of others whereas coercive children such as Ryan are overtly resistant to guidance. It could be that overtly challenging my guidance was more of an accomplishment for Annie, leading to her calling attention to her self-direction. Help-seeking versus problem-solvirig. The emergence of this theme was evident in comparing the patterns seen in the decisions in play category. Again, Annie was more indirect and Ryan more direct in strategies for seeking help in play activities. For example, Annie might pause as she struggled with playdough materials and look to me or comment on her difficulty, whereas Ryan might directly ask how to get the materials to do what he wanted. Later on in therapy, Ryan used a lot of independent problem-solving, even after asking for my help when stuck (then backing off from using my help). Annie focused much less on gaining mastery of the materials and engaged in less direct problem-solving. Again, this difference in use of the therapeutic relationship could be related to the different strategies inherent in their differing attachment styles. The difference with this theme is paradoxical at face value, since defended children (like Annie) are typically self-reliant in managing 104 their difficulties, due to an expectation of rejection or withdrawal of help by caregivers. Coercive children on the other hand, are decidedly dependent on others, refusing to solve their problems without others meeting their needs on their own terms. An alternate interpretation could be that the children were using the therapeutic relationship to develop previously unused capacities; for receiving help in Annie and for independent problem-solving in Ryan. Contact-seeking versus exploration. This theme pertains to how the children differed in their focus on the relational bond versus their focus on exploring their environment beyond the relational context. This theme was seen in patterns in the context, contact and boundaries categories. Specifically, Annie attempted over time to expand the boundaries and context of the therapeutic relationship and sought out connection with me during emotional segments of her play (through eye contact). She made sporadic physical contact, however. Defended children tend to maintain the proximity of their caregivers without responsively engaging them; however Annie's engaging strategies towards me may have been with a new expectation of receiving a responsive rather than an interfering or rejecting response. On the other hand, Ryan did not test the boundaries of the relationship nor attempt to increase its context. He sought out equal proximity and eye contact however, and made more physical contact than Annie. The reason for this difference is unclear. 105 Coercive children display increased anxiety with loss of contact with their caregivers, yet Ryan displayed both increased physical contact yet maintained a limited relationship. His play was increasingly exploratory in nature over time, markedly more so than Annie's play. Reflection-seeking versus focus on play. This theme is primarily illustrated in the different ways the children sought out eye contact with me during their play. Towards the middle of therapy and beyond, Annie made prolonged and direct eye contact during emotion-laden segments of her dramatic play, but when the action became particularly intense and fast-paced (as opposed to expressive of strong affect) she made little eye contact, focusing intently on the unfolding drama. A slightly parallel pattern occurred in Ryan's play, whereby he focused intently on intense, action packed segments of dramatic play, making little eye contact; but his use of prolonged direct eye contact was almost exclusively limited to aggressive, interactive play (usually sword play). He also made frequent, brief and direct eye contact during practise play. My impression was that during episodes of little eye contact, the children were preoccupied with the work of play (i.e. symbolically expressing themselves through the unfolding drama), and temporarily less intent on relating to me or on my responses. On the other hand, I interpreted their intense eye contact as a bid for reflection of some important aspect of self that had just been 106 expressed. Following Annie's eye contact with me, I gave reflections of the strong feelings in her play; for example, she would look intently at me after the baby had been crying and calling for its mother (recently killed), followed by me saying "The baby's so sad and scared without its mother!". Following Ryan's long and intense eye contact with me, I would typically respond with a reflection of his competence; "Wow! That was the best sword move yet. You really got me that time!". This could be related to attachment pattern differences in that defended children have limited awareness of their feelings and coercive children have limited self-reliance in coping with difficulties. These differences will be elaborated on further in descriptions of the following two themes that are specifically related to the differing strategies employed by insecurely attached children to meet their needs in the context of relationships; and specifically here, in the context of therapeutic alliance. In summary, the children displayed some differences in their reliance on the therapeutic relationship. Annie seemed to respond more to my use of responsiveness to her needs and reflection of her feelings; whereas Ryan responded more to my mere availability to him as he periodically "touched base" after more independent exploration and problem-solving. The therapist's availability and responsiveness to clients are essential contributors to therapeutic alliance, but they are also characteristic of how attachment figures go about 107 providing a secure base (for relating and exploration) to children that can then develop secure attachment. Self expression. As has been described earlier, insecurely attached children tend to have heightened negative feelings of distress (primarily anxiety or anger) due to having an unavailable and/or unresponsive caregiver. In seeking to manage and reduce these feelings, the typical strategy employed by the defended child is to assume this responsibility oneself; leading to unawareness of their feelings and little communication of affect or need (except for pseudo-communication of positive affect seen as a false brightness). As a result, these children resist eliciting help from others. On the other hand, the typical strategy employed by the coercive child is to put responsibility for management of distress on caregivers; leading to a preoccupation with their negative emotional state, and coercion of others to meet their needs as well as reluctance to responsively negotiate in problem-solving (Crittenden, 1992b). The capacity for awareness of and communication of the child's inner emotional state would likely have a strong impact on his/her ability to develop relational bonds (and hence therapeutic alliance) and to elicit help in meeting needs. The child must be emotionally available to themselves and to others in order to relate to others in a responsive way and in order to communicate their 108 needs. Given that therapeutic alliance includes the development of a mutually responsive bond, as well as mutual goals and tasks related to expressed needs (Bordin, 1979), the child's self expression would likely impact on the development of a therapeutic alliance. The relational patterns of the two children showed some differences in their self expression of feelings to me and in their play. Both children showed most self expression in dramatic play and creative play. Feelings were expressed through the narratives of conflicts dramatized in their play or through the symbolic content of their creative play. While both children increasingly expressed both positive and negative feelings over time; as described earlier, they differed in terms of emphasis. Annie's focus was increasingly on dramatic and creative play and as therapy proceeded there was more intense and greater range of negative expression. Anger surfaced after more vulnerable negative affect, for example with her as a vengeful penguin attacking my character, in role. Ryan however, fairly quickly expressed anger in war, attack and revenge dramas. Later to emerge were vulnerable negative feelings (e.g. fear and sadness) and positive feelings (e.g. pride at completing a ski jump; celebration of cooperative versus competitive friendship). While Annie did more creative play, they both seemed to use it to portray or express their emerging selves (e.g. both doing initially primitive drawings such as boxes that later 109 increased in complexity showing representations or dramas of things important to them). Ryan used dramatic and creative play also, but his play became increasingly varied and less focused on the exclusively expressive types of play. He made less intense eye contact during dramatic play. One interpretation would be that the opportunity for self expression was more salient to Annie and thus surfaced more in play therapy, with her seeking eye contact and reflection of the emotional content of her play. Relating this thematic difference to the children's attachment styles; if Annie was less expressive and aware of her emotions (given her defended attachment style) than Ryan, the opportunity for a therapeutic relationship with an available and responsive therapist might prioritize her use of that relationship for exploration and expression of those emotions. While Ryan was increasingly expressive, this may not have been as much of a priority to him in play therapy (coercive children already being acutely aware and expressive of negative affect). He did use therapy to explore more of his positive feelings, however. Self agency. The differences summarized above in insecurely attached children's strategies for moderating negative feelings and getting their needs met also points to a possible difference in self agency in relating to others as well as in therapeutic alliance in play therapy. Differences in self agency seen in the children's 110 relational patterns refers to their tendency to be self directed in their play as opposed to being more reliant on me for my availability and my response (reflecting and helping/care giving). In terms of patterns and themes that have been described; Annie's play involved less use of the types of play that afforded development of skills and manipulation of her environment (such as exploratory and practise play), but this was seen moderately in her creative play. She did make increasingly independent choices in her play however. Ryan on the other hand made extensive use of practice play and to a lesser extent, exploratory play. He displayed greater variety, exploration and limit testing in his play. He made frequent, brief direct eye contact during practice play, which was followed by my reflection and praise of his growing skill level. His narrative themes in dramatic play also focused on themes of competence, mastery and towards the end, cooperation with rather than coercion of others (which may have indicated the beginning of perspective taking; notably lacking in coercive children and affecting their ability to negotiate effective solutions). He became increasingly casually directive of me, seemingly delighted that I required only code words or gestures to pick up on and respond to what he wanted me to do with him. This would be in marked contrast with his likely experiences with caregivers (as a coercive child) of needing to use disarming strategies in order to coerce his attachment figures to meet his needs on his terms. Given that Annie (as a defended and caretaking I l l child) would already have well developed strategies for meeting her own needs in the face of her attachment figures' withdrawing style; the opportunity to develop an alliance with an available and responsive caregiver might have lead her to relax her vigilant reliance on self care strategies. This is one possible interpretation for her greater reliance on relating to me rather than self-direction in play. This was reflected in her narrative themes, such as fears of isolation, neglect and powerlessness, and a reliance on others for help and surrogate care, and seeking community. Towards the end of her therapy however, there was the beginning emergence of narrative themes of competence, seeking but fearing freedom and personal empowerment. Portraits Annie In comparison to Ryan, Annie's play process demonstrated less self agency, and more self expression and focus on the relational bond with me. These differences became increasingly evident with the progression of therapy. Initially, there was some exploration of the play environment and different play media, later to be followed by an increasing focus on dramatic play and on creative play (to a lesser extent). These play choices offered her media through which she explored and expressed many feelings. These were initially more positive (such 112 as happy and curious) and were soon followed by more vulnerable, negative feelings such as sadness, fear and hopelessness. The last feeling to be expressed was anger (which could be seen as a more empowering negative feeling). Similarly, her narrative themes were initially of isolation, neglect, powerlessness, seeking out community and surrogate care/help from others. Only at the very end of therapy did more empowering themes begin to emerge of competence and ambivalent desires for freedom. Her ways of seeking my help and guidance were more indirect (e.g. looking at me, commenting on the problem or stating what she was unsure about). Despite her lesser tendency for self reliance, she did show greater independent play choices later on, which she usually made a point of enthusiastically announcing to me. In general, there was less problem solving and purposeful manipulation of the play materials than use of the play media for more exclusive self expression. She also focused more on the relational bond with me. Over time, Annie sought increasingly to expand the context and boundaries of the therapeutic relationship, for example requesting for me to visit her at her home. Her ways of making contact with me were less through physical contact and more through the use of eye contact or proximity to me. Her use of eye contact was most prolonged and direct during episodes of emotionally rich dramatic play. These periods were followed by my reflection of the feelings expressed directly or embedded in the 113 narrative dramas. At termination of the play therapy process, Annie expressed much affection to me and a desire for continued contact. She was very emotionally expressive, showing delight and appreciation of the established bond as well as sadness over it ending. Ryan Ryan's play process, in comparison to Annie's, was generally more self directed, less focused on the relational bond, and more focused on exploration and mastery than the expression of feelings. Over time, he used more of the playroom and demonstrated increasing variety in his play, focusing mostly on exploratory play and practise play. These play choices afforded him the opportunity to explore and develop his abilities and gain mastery of his play environment. While he too chose dramatic and creative play (and expressing feelings through them), this was somewhat less so than Annie, making somewhat less eye contact during these types of play as well. The feelings expressed through these play media were initially more negative (with vulnerable negative feelings emerging after anger). More positive feelings emerged later on. Ryan's play choices of exploration and practise of skills to be mastered was also reflected in his narrative themes in dramatic play. These were initially themes of competence, competition and mastery, later including cooperation (rather than an earlier theme of coercion). His ways of seeking my help and guidance were more direct than Annie's, 114 frequently involving direct requests from me. When help was offered, even though both children began to challenge my suggestions later in therapy, Ryan did so more than Annie. He displayed a great deal of independent problem-solving and manipulation of play materials. In his relationship with me, Ryan displayed little testing of the boundaries of the relationship or seeking to expand its context beyond play sessions. He did frequently tested me in my adherence to the set limits of the play context and boundaries, such as safety rules or keeping play (rather than the relationship) within the playroom. He also made contact with me through eye contact and proximity to me, but made more physical contact than Annie. His greatest use of eye contact was during practise play and aggressive play. He also made eye contact during dramatic play, but somewhat less so than Annie. Upon termination, Ryan demonstrated some difficulty expressing and regulating his feelings, initially expressing strong resistance to the end of play and the relationship, followed by a restriction of affect or talking about his feelings in ending. In summary, analysis of Annie's and Ryan's relational patterns would seem to suggest that they differed in their use of the therapeutic relationship to meet varying needs for self awareness, self expression and self direction. Annie seemed to use the relationship as a safe base from which to explore her feelings and needs, and began sharing them with me. Ryan seemed to use me as a safe base from which to explore, manipulate and problem-solve in his play environment, developing his capacity for self agency. These are both areas of relational skills expected to be underdeveloped in each child, relative to their respective attachment styles and their preferred coping strategies. 116 Chapter Six: Discussion Analysis of the differences in development of therapeutic alliance in play therapy with two insecurely attached children revealed different relational patterns between the children and therapist, that are assumed to be related to the children's different attachment styles. Given the exploratory nature of this study, further research would be needed to confirm these relationships in larger groups of children. The two primary patterns that were found related to the children's predominant modes of relating to the therapist and their predominant focus in their process of play. While there were certainly overlaps in the children's patterns, each one had an identifiable preferred style of relating and playing in the context of the therapeutic alliances established. Different patterns emerged in categories of play and of therapeutic relationship. Pattern-matching revealed a few similarities and several contrasts in the children's patterns. Annie's play was characterized by an increasing focus on expressive, dramatic play that powerfully portrayed gradually surfacing conflicts and negative feelings about themes of isolation, powerlessness, neglect and a desire for safety/protection. Following the engagement of therapist as protector and participant in these dramas, and eliciting the therapist's reflection of the themes through the use of selective eye contact, verbal direction, indirect help-117 seeking and attempting to maximize the context/boundaries of the relationship; there was some change in the evolving feelings and conflicts expressed. Towards the end of therapy, there was some evidence in her play of greater expression of negative affect; of hope for community, family and help provided; of empowerment despite ambivalence and fears of freedom; and the very initial emergence of striving for competence. The play modality of self expression was predominant in her use of the play therapy process, rather than exploration and development of mastery in the play environment. She seemed to focus primarily on maintaining the emotional responsiveness of the therapist, both in responding to and reflecting her self expression in therapy and in requests for needs to be responded to outside of the therapy context. Ryan's play was characterized by a preference for aggressive, dramatic and practice play, that increasingly expanded (rather than narrowed in focus) to greater variety and exploration in play. Predominant initial play themes reflected much negative, angry affect in portrayals of domination, competition for power, limiting restriction and a reliance on home base for safety and protection. Ryan's tendency in relating to the therapist was to elicit her help, cheerleading and participation in more varied, active play; and her reflection of his continual exploration and attempts at mastery of the play environment, through the use of selective eye contact, gestural and encoded direction, direct help seeking or 118 independent problem-solving, and frequently testing the limits of play. Towards the end of therapy, there were some evident initial changes in his play and expressed themes including; an increase in expression of positive affect; a self motivated rather than a defensive competition and striving for mastery; provision of safety through protective limits and independent problem-solving; and emerging fears of loss of protection and of home. His predominant play modality was self agency, or mastery and exploration. He used play and the therapeutic relationship to test and develop skills at problem-solving; with or without the input of the therapist. He thus tended to focus on maintaining the availability of the therapist, as a witness and supportive encourager of his developing skills at mastery of self and his play environment. In brief, the overall patterns were for Annie to relate to the therapist as a responsive reflector of her self expression in the therapeutic alliance; and for Ryan to relate to the therapist as an available encourager of his self agency as seen in the therapeutic alliance. Limitations Several elements of this study limit the strength of hypotheses that might be generated from the results and emergent patterns presented. Subject selection presents serious limitations on the generalizability of results, due to the limitation to two subjects and the possibility of subjects' sex being an intervening variable (i.e. due to the limitted subject pool, two subjects of the same sex that met the 119 essential criteria were not available). In other words, the differences found may not be due to different attachment styles but due to the children's differing sex. There is some possibility of contamination of results due to the two children having been friends (unknown to the author at the initiation of play therapy). It is clear from the summary of sessions that the children were aware that each was concurrently receiving play therapy, but it is unclear what impact this might have had on the development of therapeutic alliance with each child, if any. Treatment restrictions place a limitation on implications drawn for practice; due to the short course of play therapy, and termination occurring after the working stage had barely been established. It is expected that different patterns of therapeutic alliance would develop in later stages of the therapeutic process. Needless to say, any hypotheses generated about the relationship between . children's attachment style and the development of therapeutic alliance in play therapy could not necessarily be applied to alliance developing in other styles of play therapy (such as focused play therapy). Several limitations relate to the methods of analysis. While the author had only tentative hunches about expected patterns that might emerge, these may have created a bias in analysis and interpretation of results. However, several of these hunches were not born out by the results found (see final section of the literature review). As indicated in the description of methodology, the author did not 120 remain blind to the children's attachment classifications throughout analysis of results, as was intended. This unfortunate methodological error increases the potential for effects of bias in the open coding of results and pattern-matching for emerging themes. Implications for Theory While theory relating attachment patterns to therapeutic alliance with children is almost non-existent, several clear convergences in therapeutic alliance theory and attachment theory have been presented earlier and related to expected patterns in play therapy with children. The goals and potential outcomes of child-centered play therapy include increased self awareness and self expression; and increased exploration and problem-solving abilities. These might also be seen in varying degrees in children's play. Would attachment patterns be related to differences in children's style or form of play? The emergent patterns in this study would seem to indicate such a relationship. Secondly, the above therapeutic changes are only expected to happen within the context of a therapeutic relationship with an available and responsive therapist (i.e. conducive to secure attachment). It is expected that consistent availability and responsiveness to the child would likely be selectively elicited from the therapist by children, and lead to stronger therapeutic alliance between the two (and possibly improved outcome). Would the potential differences in children's style or form of play 121 affect the development of therapeutic alliance? Again, the implications of the relational patterns found would seem to indicate that yes, they do. Not only did the children's pattern of attachment seem to relate to their style of play, but it also related to their selective way of relating to the therapist. The main implication for play therapy theory that follows (if these results are confirmed in further research) is that different patterns of play are related to attachment patterns; and also call for different ways of relating to the child in therapy, to build and to maintain therapeutic alliance. Specifically, one might expect a defended child (who's preferred strategy for managing feelings of distress is to restrict self and other's access to their feelings and maintain self-sufficiency) to have a greater need for a therapist that responds to their expression of feelings and needs than one who is merely available and does not assist them to move out of their rigid self-isolation and self-reliance. In fact, the defended child in this study did selectively respond to the therapist's responsiveness to her increasing self expression in play therapy. Secondly, one might expect a coercive child (who's preferred strategy is to maintain a heightened awareness of distress, limiting the capacity for clear problem-solving; and to be coercive and dependent on others to meet their needs) to have a greater need for a therapist who is consistently available to meet needs if requested, but allows for and encourages the child's independent exploration and solving of his/her own problems. In fact, the coercive child in the study selectively related to the therapist as an available, encouraging cheerleader, offering help only at the initiative of the child. Implications for Practice From the clinical experience of the author, the relationship between therapist and child can be described as a 'dance of relationship' and is dialectical in nature. There is a constant flow between leading and following, with various invitations in relating being made by the child. The results found indicate that the child with a defended attachment style made more invitations for the therapist to be responsive to her needs as an involved presence and for more reflection of feelings than the coercive child. The child with the coercive attachment pattern on the other hand, made more invitations for the therapist to be simply available if needed and to reflect his exploration and mastery of his environment than the defended child. This 'dance of relationship' is fraught with significance, because it is through the therapeutic relationship that a child's relational traumas are healed. The results of the current study point to the need for clinically useful tools for the assessment of children's attachment patterns and for tracking the development of therapeutic alliance. Instruments that were quick to administer and did not require lengthy microanalysis would assist in identifying which 123 children are likely to respond to or require which components of therapeutic alliance to be emphasized in therapy. Other than an emphasis on initial provision of necessary relational conditions for play therapy, there is very little emphasis in play therapy practise on the evolution of therapeutic alliance, or on how children's attachment styles impact on the therapeutic process or choice of strategies in practice. The results of this study indicate the possibility that children would require different therapeutic approaches and use of self by the therapist, depending on their attachment style. Selective use of therapeutic strategies such as simple attending, reflection of affect, helping or direction to independent choice, relative to differently attached children and different stages in therapy could potentially strengthen the development of therapeutic alliance and the chances for positive therapeutic outcome. Implications for Future Research As has already been highlighted, there is a dearth of process research both in the area of play therapy and in therapeutic alliance. There is a lack as well in attachment research on child attachment patterns beyond the preschool years. Given these broad gaps, even more exploratory studies are needed to examine the nature of relationships between attachment style and therapeutic alliance with child populations. Replications of the current study are needed to verify the results. An extension to studying children with other attachment patterns is also 124 indicated. Investigating these relationships as they unfold in other forms of play therapy would confirm or disconfirm the generalizability of these results across treatment approaches. Research involving more extensive group comparisons would clarify how applicable the hypotheses generated here are to other age groups and to both sexes. Further research potentially generated from these findings would be to investigate further types of alliance formed in play therapy, leading to categorization of therapeutic alliance approaches with children. This could further lead to research into training approaches for play therapists in this 'dance of relationship'. Summary This study examined how the evolution of therapeutic alliance during child-centered play therapy varied between two children with differing patterns of attachment. The two children recruited, both 5 years-old, were a female, classified as insecure-defended (subclassification A3 - caretaking) and a male, classified as insecure-coercive (subclassification C2 - disarming). The children's attachment patterns were determined by two independent raters, using Crittenden's (1992a) classification of quality of attachment of preschool-aged children. Each child received 10 weekly non-directive play therapy sessions, which were videotaped, with the investigator as therapist. Participant-observation was used for data collection, and open coding and pattern-matching was used for 125 data analysis. Pattern differences that emerged in the children's categories of play and categories of the therapeutic relationship were related to changes in therapeutic alliance and in attachment strategies for managing distress. Over time, the A3 child responded more to the therapist's responsive behaviours to her expressed needs and feelings; and engaged in relatively more self expression in her play. Over time, the C2 child responded more to the therapist's mere availability to the child; and tended to engage in relatively more self agency (self direction) in his play. Further play therapy research is needed to confirm the relationship between attachment and therapeutic alliance with children. 126 References Ainsworth, M. D. S. (1989). Attachments beyond infancy. American Psychologist. 44 (4), 709-716. Ainsworth, M. D. S. (1990). Some considerations regarding theory and assessment relevant to attachments beyond infancy. In M. T. Greenberg, D. Cicchetti & E. M. Cummings (Eds.), Attachments in the preschool years (463-488). Chicago: University of Chicago Press. Ainsworth, M. D. S., Blehar, M. C , Waters, E. & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, N. J. : Erlbaum. Axline, V. M. (1969). Plav therapy. New York, N. Y. : Ballantine. Benjamin, L. S. (1974). Structural analysis of social behaviour. Psychological Review. 81 (5). 392-425. Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory Research and Practice. 16 (3), 252-260. Bowlby, J. (1969). Attachment and loss: Vol.1. Attachment. London: The Hogarth Press. Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. New York: Basic Books. Brody, M. (1978). Developmental play: A relationship-focused program for children. Child Welfare. 57 (9), 591-600. Carmichael, K. D. (1993). Preliminary development of an interaction matrix for observation of client-therapist relationships in play therapy. International Journal of Plav Therapy. 2 (2), 19-32. 127 Carter, M. A. (1992, October). The therapeutic alliance with sexually abused children: Counsellor role, psychological development, and resistances. Paper presented at the 2nd International Conference on the Child, Montreal, Que. Cassidy, J. (1988). Child-mother attachment and the self in six-year-olds. Child Development. 59. 121-134. Cicchetti, D. & Barnett, D. (1991). Attachment organization in maltreated preschoolers. Development and Psvchopathology. 3, 397-411. Crittenden, P. M. (1992a). Classification of quality of attachment of preschool-aged children. Coding manual. N. H. : University of New Hampshire. Unpublished manuscript. Crittenden, P. M. (1992b). Quality of attachment in the preschool years. Development and Psvchopathology. 4, 209-241. Crittenden, P. M. (1992c). Treatment of anxious attachment in infancy and early childhood. Development and Psvchopathology. 4 (4), 575-602. Crittenden, P. M., Partridge, M. F. & Claussen, A. H. (1992). Family patterns of relationship in normative and dysfunctional families. Development and Psvchopathology. 3,491-512. Faust, J. & Burns, N. J. (1991). Coding therapist and child interaction: Progress and outcome in play therapy In C. E. Schaefer, K. Gittin & A. Sandgrund (Eds.), Play diagnosis and assessment (663-689). (Eds.) New York: John Wiley and Sons. Freidrich, W. N. (1995). Psychotherapy with sexually abused boys: An integrated approach. Thousand Oaks, CA: Sage Publications. Frieswyck, S. H., Allen, J. G., Colson, D. B., Coyne, L., Gabbard, G. O., Horwitz, L. & Newsom, G. (1986). Therapeutic alliance: Its place as a process and outcome variable in dynamic psychotherapy research. Journal of Consulting and Clinical Psychology. 54 (1), 32-38. 128 Greenberg, M. T., Speltz, M. L., Deklyn, M. & Endriga, M. C. (1992). Attachment security in preschoolers with and without externalizing behaviour problems: A replication. Development and Psychopathology. 3. 413-430. Hartley, D. E. (1985). Research on the therapeutic alliance in psychotherapy. American Psychiatric Association Annual Review. 4, 532-549. Head, T. (1996). Children's and their parents' conceptualizations of attachment. Unpublished doctoral dissertation, University of British Columbia, Vancouver. Hellendoorn, J., van Eijk, D., Vollaard, I. & Harinck, F. (1989). Comparison of imagery interaction play therapy and child centered child therapy. Tijdschrift voor Psychotherapie. 15 (3), 134-147. Horvath, A. O., & Greenberg, L. S. (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology. 36 (2), 223-233. James, B. (1989). Treating traumatized children: New insights and creative interventions. Lexington, MA: Lexington Books. Landreth, G. L. (1991). Play therapy : The art of the relationship. Muncie, IN: Accelerated Development Inc. Mahon, J. M. (1994). A comparison of parent-child attachment to a play therapy program. Dissertation Abstracts International. 55. 464A. (University Microfilms No. AAC 9421880). Main, M. & Cassidy, J. (1988). Categories of response to reunion with the parent at age 6: Predictable from infant attachment classifications and stable over a 1-month period. Developmental Psychology. 24 (3). 415-426. Main, M. & Goldwyn, R. (1990). Adult attachment classification system. Unpublished coding manual, University of California, Berkely. 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 of attachment theory and research. Monographs of the 129 Society for Research in Child Development. 50 (1-2, Serial No. 209). Marziali, E., Marmar, C. & Krupnik, J. (1981). Therapeutic alliance scales: Development and relationship to psychotherapy outcome. American Journal of Psychiatry. 138 (3), 361-364. Moretti, M. M., Holland, R. & Peterson, S. (1994). Long term outcome of an attachment-based program for conduct disorder. Canadian Journal of Psychiatry. 39, 360-370. Moustakas, C. E. (1959). Psychotherapy with children: the living relationship. Greeley, CO: Carron Books. Neufeld, G. A. (1995, April). Impact of divorce on children. Unpublished proceedings of a workshop conducted at Burnaby Psychiatric Services, Burnaby, B. C. Rasmussen, L. A. & Cunningham, C. (1995). Focused play therapy and non-directive play therapy: Can they be integrated? Journal of Child Sexual Abuse. 4 (11. 1-20. Schneider-Rosen, K. (1990). The developmental reorganization of attachment relationships: Guidelines for classification beyond infancy. In M. T. Greenberg, D. Cicchetti & E. M. Cummings (Eds.), Attachment in the preschool years: Theory, research and intervention (pp 185-220L Chicago: University of Chicago Press. Short, R. H. & Boon, B. J. (1990). Counselling as a dialectical process. International Journal for the Advancement of Counselling. 13. 129-143. Speltz, M. L. (1990). The treatment of preschool conduct problems: An integration of behavioural and attachment concepts. In M. T. Greenberg, D. Cicchetti & E. M. Cummings (Eds.), Attachment in the preschool years. Chicago: University of Chicago Press. Strauss, A. & Corbin, J. (1990). Basics of qualitative research. Newbury Park, CA: Sage Publications. 130 Weitzman, J. & Cook, R. E. (1986). Attachment theory and clinical implications for at-risk children. Child Psychiatry and Human Development. 17 (2), 95-103. White, K. E. (1994). Attachment, coping and the therapeutic alliance. Masters Abstracts International. 30, 994. (University Microfilms No. AAC MM92301). Wolcott, H. F. (1990). Writing up qualitative research. Newbury Park, CA: Sage Publications. Yin, R. (1989). Case study research. Newbury Park, CA: Sage Publications. Zetzel, E. R. (1956). Current concepts of transference. International Journal of Psychoanalysis. 37. 369-376. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items