Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of experience on counsellor schema of sexually abused clients Sheehan, Heidemarie 1995

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

Item Metadata

Download

Media
831-ubc_1995-0412.pdf [ 7MB ]
Metadata
JSON: 831-1.0054137.json
JSON-LD: 831-1.0054137-ld.json
RDF/XML (Pretty): 831-1.0054137-rdf.xml
RDF/JSON: 831-1.0054137-rdf.json
Turtle: 831-1.0054137-turtle.txt
N-Triples: 831-1.0054137-rdf-ntriples.txt
Original Record: 831-1.0054137-source.json
Full Text
831-1.0054137-fulltext.txt
Citation
831-1.0054137.ris

Full Text

THE EFFECT OF EXPERIENCE ON COUNSELLOR SCHEMA OF SEXUALLY ABUSED CLIENTS by HEIDEMARIE SHEEHAN B.A., Carleton University, Ottawa, Ontario, 1983 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTERS 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 July, 1995 © Heidemarie Sheehan, 1995 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 of Cf lUhSef lM^ H y c f a / o ^ ftcvf^y o{ &UC^{v^\ The University of British Columbia Vancouver, Canada Date ^ f l - O f l . f T -DE-6 (2/88) Abstract The sexual abuse schemas of 97 counsellors were compared on the bases of experience (general counselling, specific sexual abuse counselling, and combined counselling), current client population (working with sexually abused clients or not), and whether they had their own history of childhood sexual abuse. Three inferential measures on a questionnaire were used to evaluate the dimensions of schematic complexity: organization, breadth, and extremity. They were, in order: two versions (one explicit and one disguised) of an "intake" vignette describing an adult woman with a history of childhood sexual abuse; a 100 list of adjectives to describe a sexually abused client; and 20 statements concerning sexual abuse effects. The primary finding was a strong relationship on the dimension of organization: counsellors were significantly more likely to identify sexual abuse as a treatment issue in response to an explicit sexual abuse vignette version than to a disguised version. There was no experience effect but post hoc analyses found significant results on likelihood of identifying sexual abuse in response to the disguised version when counsellors' client focus and personal childhood history of sexual abuse were considered. Current sexual abuse counsellors were more likely to identify it than non sexual abuse counsellors, and counsellors who had themselves been sexually abused as children were more likely to identify sexual abuse than were counsellors who had not. No differences were found on measures of breadth or extremity. Post hoc analyses suggested a tendency for counsellors of high and low experience to respond differently to a list of adjectives and a list of sexual abuse effects but these results did not reach significance. I l l Table of Contents Page Abstract 1 1 Table of Contents '» List of Tables v i Acknowledgment 1 X Chapter One - Introduction 1 Chapter Two - Literature Review 12 Issues in Sexual Abuse Counselling 12 Knowledge ' 3 Attitudes 1 5 Counsellor Preparedness 17 Education and Training 18 Schema 20 Schematic Properties 21 Organization 21 Breadth 2 2 Extremity 2 2 Prototypes 2 3 Person Schemas 2 3 Expertise 2 ^ Schema Activation 26 Effect of Counselling Process 26 Hypothesis Testing 2& Cognitive Complexity 30 Schema Measurement 31 iv Experience . 32 Relationship to Clinical Judgement 33 Effect of Amount of Information 34 Summary 37 Research Questions 38 Hypotheses 40 Chapter Three - Method 41 Design 41 Variables 41 Sample 42 Questionnaire 47 Pilot 52 Procedure 53 Analysis 55 Chapter Four - Results 57 Homogeneity of Sample 57 Preliminary Analyses 57 Experience Variable 60 Hypothesis One 63 Hypothesis Two 66 Hypothesis Three 69 Hypothesis Four 72 Exploratory Analyses 75 Post Hoc Analyses 85 Chapter Five - Discussion 107 Overview 108 Identification of Sexual Abuse in Response to Vignettes I l l Breadth of Schema: Number of Adjectives 116 Breadth of Schema: Negative and Positive Adjectives 119 Extremity of Schema 120 Duration of Therapy 124 Current Sexual Abuse Counsellors 126 Counsellor's Personal Abuse History 128 Training Activities 130 Limitations 130 Future Research 133 Implications for Counsellors 134 Summary and Conclusions 136 References 140 Appendices 145 Appendix A. Questionnaire Part A: Vignette 145 Appendix B. Questionnaire Part B: Characteristics.. 148 Appendix C. Questionnaire Part C: Sexual Abuse Effects 151 Appendix D. Questionnaire Part D: Demographics 154 Appendix E. Agency Letter Regarding Pilot Study 157 Appendix F. Agency Letter Regarding Research Study 159 Appendix G. Questionnaire Package 161 Appendix H. Participant Follow Up Letter 166 Appendix I. Ethical Approval 168 vi List of Tables Page Table 1. Demographic Characteristics of Sample 44 Table 2. Counsellors' History of Childhood Abuse and Adult Assault 46 Table 3. Association of Vignette Version and Identification of Sexual Abuse (Total Sample) 59 Table 4. Association of Gender and Identification of Sexual Abuse 60 Table 5. Association of Age and Amount of Experience 62 Table 6. Hierarchical Log Linear Analysis of Vignette Version by Experience by Identification of Sexual Abuse 64 Table 7. Association of Type of Counselling Experience and Identification of Sexual Abuse 65 Table 8. T-test Results Comparing the Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Levels of General, Sexual Abuse, and Combined Experience (Vignettes SA and NSA) 67 Table 9. T-test Results Comparing the Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Levels of General, Sexual Abuse, and Combined Experience (Vignette NSA) 68 Table 10. T-test Results of Mean Scores of Polarized Ratings of Adjectives by Experience Level (Vignettes SA and NSA) 70 Table 11. T-test Results of Mean Scores of Polarized Ratings of Adjectives by Experience Level (Vignette NSA) 71 Table 12 Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Combined, General and Sexual Abuse Experience Levels (Vignettes SA and NSA) 73 Table 13. Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Combined, General and Sexual Abuse Experience Levels (Vignettes SA and NSA) 74 Table 14. Association of Predictor Variables and Anticipated Duration of Treatment (Vignettes SA and NSA) 77 Table 15. Association of Predictor Variables and Anticipated Duration of Treatment (Vignette NSA) 78 Table 16. T-test Results for Scores of Negative and Positive Adjectives Selected by Counsellors (Vignettes SA and NSA) 80 Table 17. T-test Results for Scores of Negative and Positive Adjectives Selected by Counsellors (Vignette NSA) 82 Table 18. Individual Adjectives Receiving Significantly Different Ratings from Low and High Experienced Counsellors (Vignettes SA and NSA, and Vignettes NSA only) . 84 Table 19. T-test Results of Responses to Sexual Abuse Effects (Part C) by Low and High Combined Experience Groups (Vignettes SA and NSA) . . . . 86 Table 20. T-test Results of Responses to Sexual Abuse Effects (Part C) by Low and High Combined Experience Groups (Vignette NSA) 88 Table 21. Comparison of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors on Amount of Counselling Experience 90 Table 22. Comparison of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors on Work Setting 91 Table 23. Comparison of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors on Self-Reported Level of Expertise of Counselling Issues 92 Table 24. Association of Current Sexual Abuse Counselling and Identification of Sexual Abuse (Vignettes SA and NSA) 93 Table 25. Association of Current Sexual Abuse Counselling and Identification of Sexual Abuse (Vignette NSA) ' viii Table 26. T-test Results Comparing Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors 95 Table 27. T-test Results of Mean Scores for Polarized Ratings of Adjectives Between Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors (Vignettes SA and NSA, and Vignette NSA only) 96 Table 28. Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors (Vignettes SA and NSA, and Vignette NSA only), . 97 Table 29. Association of History of Childhood Abuse and Adult Assault and Identification of Sexual Abuse (Vignettes SA and NSA) 99 Table 30. Association of History of Childhood Sexual Abuse and Identification of Sexual Abuse (Vignette NSA) 100 Table 31. Variables Associated with Identification of Sexual Abuse Responses to Ambiguous Cues (Vignette NSA) 102 Table 32. T-test Results Comparing Mean Number of Training Activities Undertaken by Counsellors 105 ix Acknowledgment I wish to acknowledge my supervisor, Beth Haverkamp, for her expertise, wisdom, and support throughout this research. I feel privileged to have been under the guidance of such an invaluable source of knowledge and strength. I would also like to thank the other members of my committee, Richard Young and Angela Henderson, for their perspectives on the study and their belief in the importance of this research. In addition I would like to thank my family for their love, encouragement, and faith, and to acknowledge the practical help provided by Erika and Gary. I wish to recognize all my friends for their on-going support throughout this endeavour. In particular, I want to acknowledge Kim for her professional insight, sense of humour, and perspective; Susan for her unwavering friendship and unequivocal belief in me; Sharon for her loving spirit and optimism; and William for his presence. Finally, I am indebted to the team of counsellors at VISAC (Vancouver Incest and Sexual Abuse Centre) for their perspectives, assistance and guidance throughout this project. This work is dedicated to my grandmother, Helen Shaver, (1908-1994) who never failed to be in my corner, and whose unending love is a precious gift. Counsellor Schema 1 Chapter One Introduction Childhood sexual abuse has become one of the therapeutic community's largest focus areas due to the increasingly high numbers of clients requesting counselling specific to that issue. Recent empirical research indicates that the incidence of sexual abuse among clinical populations approaches 50% (Alpert, 1990); implicit in these statistics is the likelihood that service providers in many venues will encounter clients with sexual abuse treatment issues. This raises questions of training for the clinicians who will be providing the counselling. Alpert and Paulson's (1990) study found that the curricula of most university professional degree programs in psychology did not incorporate specific training to prepare therapists to work with this population. In addition to training, counsellor reactions to clients are influenced by schemas, which are integral to the counselling process. Schemas are the cognitive structures that guide a perceiver's perception, memory, and inference about people and situations (Fiske & Taylor, 1984). Schemas are fundamental to the therapist's way of conceptualizing the client and presenting problem, and the development of a treatment plan. Questions arise, therefore, about how counselling experience affects a therapist's schema, and subsequent treatment, of sexually abused clients. These questions are of great importance in the counselling profession because every counsellor is faced with the considerable task of helping clients make sense of the complexity of their problems and life stories, which are richly imbued with meaning. The counsellor's role is to facilitate, through the process of clinical judgement, client growth and understanding. The process of change is usually through the intangible process of social influence (Heppner & Frazier, 1992) and can be affected by "perceived expertness, attractiveness, and trustworthiness" (p. 144). This process of clinical judgement is fraught with ambiguities and potential sources of bias (Spengler & Strohmer, 1993). Among these influences are stereotypes regarding client variables and therapeutic issues, and the therapists' beliefs and perspectives based on their particular theoretical orientations (Snyder & Thomsen, 1988) with their attendant preconceptions (Turk & Salovey, 1985). Therapeutic effectiveness relies on factors such as the counsellor's general knowledge and about specific therapeutic issues, their Counsellor Schema 2 beliefs, and attributions about the clients themselves (Heppner & Frazier, 1992) and the presenting issues. For example, a clinicians' attributions of relative blame towards a client who is the victim of a sexual assault and to her offender may affect whether the therapist focusses on the woman's responsibility in the event, or her pain. Counsellors' education, personal experiences, backgrounds, and beliefs all have an impact on their clinical judgement or their way of viewing their work and clients. The ability to integrate and utilize client information is fundamental to the clinical decision making processes of diagnoses, interpretation and choice of interventions (Holloway & Wolleat, 1980). Schemas are "preexisting knowledge structures... [that] are used to filter the wealth of information that continually confronts the clinician as he or she tries to make sense of incoming stimuli, to answer complex questions, and to make difficult judgements" (Turk & Salovey, 1985, p. 20). A counsellor's schema for a particular presenting problem is in a different stage of development depending on previous exposure to the clients, the type and amount of training, and the degree of experience with and knowledge about the phenomenon. Hypothesis testing about client problems is guided by the counsellor's schema (Snyder & Thomsen, 1988). These schemas also have implications for decisions made about the course of therapy. Clinical judgement and treatment are concomitantly affected by the schemas counsellors have developed through their accumulated expertise. Therapists with different amounts of clinical experience possess varying degrees of knowledge, levels of technical skill, and relationship-building abilities. Counsellor information gathering and cognitive strategies may benefit from experience (Mallinckrodt & Nelson, 1991). However, research on the effect of counsellor experience on clinical decision-making has produced contentious and equivocal results. Further, such research has not examined how therapists "think or process client information" (Spengler & Strohmer, 1993, p. 4). Recently there has been a recognition of child sexual abuse as a problem area. Counsellors working with sexually abused clients encounter several specific challenges. There has been rapid growth in the knowledge base of the etiology and treatment of sexual abuse, accompanied by changes in the understanding of the emotional, psychological and behavioural sequelae. It is a highly complex Counsellor Schema 3 phenomenon with a multitude of dynamics, effects, and presentations (Briere & Runtz, 1993). There is a lack of adequate training for professionals in the field (Alpert & Paulson, 1990), which is complicated by a surfeit of ethical ambiguities regarding intervention not only with decision making regarding reporting and confidentiality, but of professional competence and preferred treatment modality (Daniluk & Haverkamp, 1993). The process of clinical judgement is particularly complex. There appear to be inherent biases at all stages of the therapeutic process, such as filtering information, ignoring counter evidence, using confirmatory hypothesis testing strategies, and relying on heuristics (Rabinowitz, 1993). The combined intricacies of clinical judgement, ethical considerations, and the complicated and manifold issues related to childhood sexual abuse suggest the need to learn more about the way in which these interact. The definition of childhood sexual abuse varies somewhat when considered from therapeutic or legal perspectives. For the purposes of this study, the description offered by Sgroi, Blick, and Porter (1982) was used as a conceptual framework: "Child sexual abuse is a sexual act imposed on a child who lacks emotional, maturational, and cognitive development. The ability to lure a child into a sexual relationship is based upon the all-powerful and dominant position of the adult or older adolescent perpetrator, which is in sharp contrast to the child's age, dependency, and subordinate position. Authority and power enable the perpetrator, implicitly or directly, to coerce the child into sexual compliance" (p. 9). The utility of this definition lies in its avoidance of age parameters, unlike legal definitions which necessarily differ from country to country and jurisdictional boundaries within individual countries (as the age of "majority" varies); and by focusing on the relationship between abuser and abused as opposed to outlining specific behaviours or events. Further, it precludes a requirement of jurisprudence: an individual may have experienced sexual abuse as a child regardless of whether a criminal charge had been laid or conviction pronounced. Child sexual abuse, then, is any sexual act committed by an older person towards a child where more authority or power is ascribed to the older person by virtue of age, size, or relationship. In the introduction to her book, Handbook of Clinical Intervention in Child Sexual Abuse. Sgroi (1982) offered the following observation about sexual abuse therapy: "the field is too new and the body of accumulated knowledge and skills is too small and inadequately tested for anyone to claim that he or she Counsellor Schema 4 has the answers" (pp. 5-6). She commented that "every clinician . . . who deals with cases of child sexual abuse today is, in a sense, a pioneer exploring largely uncharted territory" (p. 6). Little more than a decade has passed since the publication of her innovative work, providing minimal time for empirical longitudinal scrutiny of her treatment principles. As testimony to Sgroi's prophecy that skills and knowledge would grow and change, the intervening decade has witnessed the proliferation of information through the media and the popular and academic presses. Although the knowledge has grown considerably, there remain few definitive answers about treatment in this evolving field. The relatively recent clinical focus on the phenomenon has its roots in the widespread reporting of childhood sexual abuse during the past 15 years. Although the use of children for the gratification of adult sexual needs is universal and historic, having been traced back through the centuries in many cultures (Rush, 1984), public acknowledgment of its existence is relatively recent. The prevalence of sexual abuse in the general population is, by most accounts, high. Demographic trends indicate that family constellations are changing to more blended and single parent families, which are associated with higher rates of intrafamilial childhood sexual abuse. However, the apparent increase in abuse that could have been attributed to the changing face of families primarily reflects a rise in disclosures by victims (Crewdson, 1988). Although discrepancies exist about the actual prevalence due to factors such as methodological problems, the sensitivity of the subject, and ethical limitations when doing such research, the consensus clearly substantiates the existence of sexual abuse among a high proportion of women and men. Survey results from the United States in the early 1980s indicate that between 22% to 50% of women and 16% to 33% of men have been victims of some type of intrusive or "non contact" sexual abuse before the age of eighteen (Crewdson, 1988). Added to this is Finkelhor's (1984) observation that"... almost universally the problem is conceded to be far greater than statistics on reported cases would indicate" (p. 1). Current research confirms earlier reports (Finkelor, Hotaling, Lewis, & Smith, 1990), indicating that the statistics were not anomalous and the problem has not abated during the decade since the findings were originally publicized. Counsellor Schema 5 Response to the phenomenon has been extensive. The personal aid movement has produced countless "self-help" books (e.g., Bass & Davis, 1988; Blume, 1990); the professional realm has published entire clinical texts devoted to specialized sexual abuse treatment modalities (e.g., Courtois, 1988; Friedrich, 1990); and the academic press has witnessed the inclusion of sexual abuse literature in the fields of psychology, psychiatry, social work, women's studies, and counselling, occasionally with entire journal volumes devoted to the topic (e.g., Professional Psychology: Research and Practice. 1990; Journal of Consulting and Clinical Psychology. 1992). Information is available on the effects, identification, assessment, treatment, jurisprudence, and ethics of sexual abuse and its intervention. Absent from the literature is research on the professionals who are providing therapy for the clients, their readiness to do so, and their responses to the individuals seeking their help (Parisien & Long, 1994). The magnitude of the evolving knowledge base in the field presents a practical difficulty for clinicians attempting to select relevant and appropriate resources. Despite the wealth of the literature, much of it is drawn from personal experience and clinical observations, and empirical research remains in infancy. Additionally, few professionals have time to acquaint themselves with the range of research and writing, and may select a limited number of sources to increase their knowledge. Depending on what is selected, this may result in misinformation, skewed or biased views, or a sense of being overwhelmed by the nature of the work. The selection of resources by clinicians and the manner in which they integrate the new knowledge in their practice may be affected by their schemas of sexually abused clients. In his review of the historical context, Finkelhor noted that known cases of childhood sexual abuse were "relatively rare" prior to 1975 (p. 200). When the reporting tripled during the subsequent three years, professionals " . . . had a hard time developing expertise, experience and procedures for handling a so quickly emerging problem.... Communities have found... that as they develop people and agencies with specialized capabilities in the area of sexual abuse, these resources are quickly swamped" (p. 200). Finkelhor's (1984) observation that "the management and treatment of cases of child sexual abuse have posed a serious challenge to communities across the country" (p. 200) continues to ring true as social service agencies, mental health facilities, and criminal justice systems struggle to keep up with the tide of Counsellor Schema 6 disclosures and the requests for help. This type of demand for services increases the pressure on therapists who lack either the expertise or knowledge to counsel survivors. Coupled with survey results indicating the high prevalence of sexual abuse among clinical populations and the possibility of clients concealing their histories at the intake level (Courtois, 1988), inexperienced therapists may well find themselves working with a sexual abuse survivor. This may occur knowingly in communities with limited resources, or unknowingly in cases where clients initially disguise their reasons for requesting treatment either consciously, or due to their own lack of awareness of underlying abuse histories (Briere & Conte, 1993). The complexity of therapy is exemplified by the emotional and behavioural aftereffects of childhood sexual abuse. There can be considerable variation in individual symptomatology: the manifestations may include post traumatic stress reactions, cognitive distortions, psychological disturbances, emotional upset, impaired life functioning, relationship problems, and behavioural difficulties (Briere & Runtz, 1993). Compilations of the symptoms can be found elsewhere in many sources (e.g., Briere & Runtz, 1993; Kluft, 1990). Courtois's (1988) synopsis of the range of effects is telling: "In adulthood, victims as a group show impairment of some sort when compared with non victims; 20% of these victims show serious psychopathology. In all likelihood [research] findings underestimate aftereffects. . . . " (p. 89). The effects are multidimensional, as untreated sequelae produce "secondary elaborations" in the form of depression, eating disorders, substance abuse and somatization disorders (p. 90), and it is these symptoms for which survivors often initiate treatment. Presentation at the onset of treatment may be disguised by other issues by those survivors who may "deny, dissociate, and repress the abuse and their reactions to it" (Courtois, 1988, p. 94). To prevent the possibility of abuse being overlooked in treatment, clinicians must therefore be familiar with the psychological and behavioural manifestations of sexual abuse, and, in addition, be alert to the possibility of masked symptoms. The multitude and permutations of sequelae and the potential level of disturbance is a strong justification for specialized therapist training and the need for a particular level of expertise. Not to be ignored are the ethical implications of working with survivors. The principles of beneficence (doing good) and nonmaleficence (do no harm) are some of the core considerations for all Counsellor Schema 7 therapists, and are magnified for those lacking the requisite knowledge and training to provide adequate counselling. The question of who is competent to provide therapy must be examined within the context of ethical principles. "The counselor who fails to be of assistance to the survivor, based on a lack of skill, knowledge or expertise in the treatment of sexual abuse, is clearly acting in an unethical fashion...." (Daniluk & Haverkamp, 1993, p. 17). Ethical questions permeate other aspects of the work: "given the rapidly changing state of our knowledge in this area, counselors working with adult survivors of abuse must be committed to continuing education and supervision (beneficence and nonmaleficence)" (p. 17), to ensure therapists remain abreast of new developments, upgrade their knowledge bases, and refine their skills in conjunction with emerging information. A further argument in support of specialized training is underscored by the principle of fidelity: "sexual abuse must not be the target of the untrained professional's interventions. . . . The counselor must not consider the provision of well-intentioned but unskilled service as being better than no service at all" (p. 18). The continuum of potential perils for an inexperienced therapist, and thereby for the client, range from being ineffective and incapable of "doing good" as a result of lack of knowledge, to unwittingly entering a realm requiring advanced skills, and intervening in such a manner that clients deteriorate. Rarely do therapists maliciously or deliberately employ inappropriate methods; however, the risks of unintentionally practicing poor counselling increase in new fields, and can be exacerbated with emotionally charged issues like sexual abuse. Given the statistics that indicate 25% to 50% of the population have experienced some type of childhood sexual abuse, it is reasonable to assume that a significant proportion of clinicians will have abuse in their own histories. The evocative nature of the material that may manifest in clinical sessions raises questions about the repercussions for sexually abused therapists who have or have not reached resolution regarding their own experiences (Pope & Feldman-Summers, 1992). Questions arise about the relative sensitivity to abused clients by clinicians with their own histories of abuse; whether they are more likely to recognize the indicators of sexual abuse in their clients or have "blinders" in this area. How might one's own experience become central to the counsellor's understanding and stereotypes? In a national survey of Counsellor Schema 8 psychologists (Feldman-Summers & Pope, 1994), 23.9% reported sexual or physical abuse in childhood, and 40% of that group reported an amnesiac period about some or all of the abuse (p. 638). Briere and Conte (1993) found a higher percentage of amnesia (59.3%) among their clinical sample of adult survivors of sexual abuse. Speculation arises, therefore, about the proportion of psychologists who may be continuing to forget their own abuse, and how vulnerable they may be to re-traumatization or vicarious traumatization when working in this field. Nuttall and Jackson (1994) conducted a survey of the prevalence and effects of childhood abuse among clinicians. Citing studies that discuss the influences on clinician selection, perception, and interpretation of information, the authors describe the "compelling need" (p. 456) to investigate variables including "belief systems connected to personal history and professional training," both powerful moderators of clinician behaviour. All of these cognitive processes regarding the integration of information about the self and others mediated by schemas, are therefore additional important counsellor variables to examine. Based on their finding of prevalence rates comparable to those reported in other (non clinical) populations, they suggest that clinicians' histories of childhood abuse may have "significant effects on clinical data processing and decision-making" (p.469). The influence of one's personal experience with childhood sexual abuse on clinical practice, although as yet undetermined, is relevant and potentially significant, and may occur via the counsellor's schema. Providing adequate therapy is further complicated by the recent controversy over "false memories," and concerns that the combination of suggestible clients and inadequately trained therapists has, on occasion, resulted in unfounded incidents of childhood sexual abuse being "remembered" by clients (Wylie, 1993). Highly regarded professionals in psychology have adopted positions on both sides of the debate about the existence of "false memories." That cases have occurred cannot be disputed; the controversy, however, has been exacerbated by the emergence of the False Memory Syndrome Foundation in 1992 (Enns, McNeilly, Corkery, & Gilbert, 1995). This advocacy group has received wide publicity and assisted in the filing of malpractice lawsuits against implicated counsellors. Given the potential for litigation and discreditation, therapists should necessarily exercise caution during particular phases of Counsellor Schema 9 treatment of sexually abused clients. They should also, however, maintain an awareness that there have been some cases of false recollections elicited by the inappropriate interventions of misinformed practitioners; implicit in this fact is that counsellors should be prudent in their therapeutic process. Competence (the ethical principle of beneficence) is undoubtedly at issue, and underscores the need for the existence of comprehensive education. The state of research on counsellors who work in the sexual abuse field is in its infancy. There has been only minimal exploration in many of the areas, such as the differential effectiveness of various therapeutic approaches, and the comprehensiveness and relevance of counsellor education. Underlying many of the therapeutic tasks is the cognitive structure, schema, which guides the essential information processing and which has also received little attention in counselling work (Heppner & Frazier, 1992). The role of counsellor experience on schema remains unclear. There appears to be a positive relationship between experience with a phenomenon and schema and that "experienced counsellors may. . . have more complex cognitive schemata for organizing client information into case conceptualizations" (Mallinckrodt & Nelson, 1991, p. 136). Mature schemas are more complex, can integrate discrepant cues, and are more resistant to change; developing schemas facilitate the assimilation of new information (Fiske & Taylor, 1984). Counsellors' sexual abuse schemas have not been directly investigated. In this evolving field, with its changing knowledge base, there may be advantages to both mature and emergent schemas. When undertaking the study, this researcher was unable to find existing instruments appropriate for the investigation. Since commencing the research, two other studies have been located and are currently in progress. One is sponsored by a local (Lower Mainland of British Columbia) organization, the Service Providers for Adults Network (SPAN), and was distributed during the fall of 1994 through a mailing and at conferences in the Greater Vancouver Region. Responses are currently being collated. The other survey was contained in a Washington State based journal entitled Treating Abuse Today. 4(6), published in the November/December 1994 edition with a return date of January 1995. Results have not yet been reported. The emergence of these studies is encouraging as it indicates that counsellors in the field of sexual abuse are now being studied. However, the content of both surveys was limited to areas of focus such as Counsellor Schema 10 counsellor's theoretical orientation; size of caseload; time spent working with survivors and specific treatment issues; use of clinical consultation and supervision; experience with professional liability; and legal and ethical issues as they relate to clinical practice. Still missing from the literature, then, are surveys that more directly address schema or the application of training to practice and clinical decision-making. The purpose of this initial inquiry was to investigate the relationship between experience and counsellors' schemas of sexually abused clients. In particular, it was expected that the sexual abuse schemas of counsellors with greater amounts of general counselling experience and sexual abuse counselling experience would be broader and more complex than those of less experienced counsellors. Greater complexity was expected to be reflected by more highly organized schemas that promoted the recognition of subtle informational cues. The schemas of experienced counsellors were also expected to contain more elements (greater breadth) and be more moderate (schematic dimension of extremity) than those of less experienced counsellors. On a more personal note, I want to disuss my assumptions and biases that inform this research. I myself am a counsellor in a sexual abuse agency and I have worked in the field for approximately seven years. My understanding of the work developed through clinical experience and from attendance in pertinent training courses and conferences. The subject of this research evolved from my recognition of the difficulties counsellors face in order to become sufficiently knowledgeable about the field. This results both from the evolving nature of the work, and because of the minimal attention given to the subject in the formal academic arenas. I hold some assumptions about clients and about counsellors, which will be delineated below, and which have influenced the research design regarding what participant tasks I chose to include, and how I interpreted the results. First, I believe the repercussions of this childhood trauma manifest in clients in widely varying manners, with a broad range of adaptive strategies adopted by individuals who have been sexually abused. The amount of therapy necessary for effective resolution may therefore vary considerably from one client to another. The program in which I work provides services for clients with minimal resources and whose life situations are often extremely troubled and chaotic, compounding the difficulties the clients may experience Counsellor Schema 11 in establishing a trusting relationship with a counsellor. I work from the assumption that clients rarely enter counselling strictly with sexual abuse issues, but with the other life stressors which may need to be addressed before the trauma can be resolved. I also have assumptions about the therapists doing the work. First and most importantly, I know the field is extremely complex and the level of trauma experienced by the clients may be profound. I therefore believe that therapists must be responsible for acquiring a particular skill level and that they be well informed about the issues and interventions, and seek training and consultation to ameliorate their knowledge levels. Although I work with clients identified as having been sexually abused, I also maintain that therapists must be able to recognize indicators of a history of sexual abuse to ensure clients have the opportunity to address the issues. Because clients may be reluctant to initiate a discussion about sexual abuse, I believe it is the responsibility of the therapist to ask questions of a client that will provide opportunities for disclosure to ensure the issue is not overlooked in therapy. Although my bias is clearly that a counsellor's ability to recognize a sexual abuse history is a positive skill, it is not my intention to imply that the failure to do so is bad counselling practice. I am curious about the differences among counsellors and the variables associated with this skill, and I recognize that there may be valid reasons why clinicians do not identify sexual abuse; that this may reflect a stylistic difference. My work, assumptions and biases about the field and counsellors, may therefore influence the study, and discussion and interpretation of the results. Counsellor Schema 12 Chapter Two Literature Review There is a paucity of literature addressing the subject of sexual abuse counsellors; no studies have been located that examine their schemas. Related research in the area of practitioner knowledge of, and attitudes towards, childhood sexual abuse will be reviewed to provide a context for this study. Literature on the effects of training for therapists, and the ability of clinicians to integrate research data in practice, is considered. Counsellor readiness for working with sexually abused clients will be surveyed briefly. The relevance of a counsellor's own history of childhood sexual abuse is introduced. The theme of adequate education in the field of sexual abuse, which permeates the literature and provides support for the need for additional research in the area, will be reviewed. Schema will be discussed in view of its effects for the counsellor and subsequently on the therapeutic process, and the relevance of schema in the area of sexual abuse treatment is addressed. The research on counsellor experience and its relationship to clinical judgement is reviewed. Issues in Sexual Abuse Counselling Despite the proliferation of literature in the field of childhood sexual abuse, scant attention has been paid to one important factor in the equation: the counsellor. What information does exist is limited to a few select areas. There is a small and growing body of research on the attribution of responsibility in incest and sexual abuse cases that examines the diverse amounts of blame accorded victims, offenders, and situational factors by samples of different professionals (Collings & Payne, 1991; Doughty & Schneider, 1987; Jackson & Ferguson, 1983; Reidy & Hochstadt, 1993). Several studies have been located which address the emotional reactions or countertransference issues of counsellors working with sexually abused clients (Emerson, 1988; Grossman, Levine-Jordano, & Shearer, 1990; Reynolds-Mejia & Levitan, 1990). No empirical research has, however, been identified that explicitly examines counsellors' schemas with this client population. The assessment of therapist knowledge and training is, however, becoming an important area of investigation. Counsellor Schema 13 Knowledge. Finkelhor (1984) first raised questions about sexual abuse knowledge and its application by professionals in 1984: "How knowledgeable are they? Are they using the resources that are available? How deep are the differences in philosophy among various agencies? Are some agencies and professionals more conspicuously successful than others?" (p. 201). His were pertinent and practical questions. Although the thrust of research in the sexual abuse field has been to ascertain the prevalence, understand the sequelae, and assess treatment modalities, little attention has been paid to how well the resulting information explosion has been assimilated by practitioners. Dye and Roth (1990) allude to this concern with their discovery that "several studies have demonstrated that most practicing clinical psychologists do not read or base their work on psychotherapy research [and]. . . it is likely that most assault victims who currently seek psychotherapy are seen by therapists who are not familiar with [ i t ] . . . ." (p. 193). It is not clear whether time constraints, inadequate access to resources, lack of interest, or misplaced confidence are the underlying reasons for this alarming finding. Considering the breadth of the potential deleterious effects of sexual abuse, more investigation is necessary to ensure counsellors are sufficiently prepared to provide adequate and appropriate therapy based on updated information. Failure to do so raises serious ethical questions, specifically related to the principles of beneficence and nonmaleficence (Daniluk & Haverkamp, 1993). The earliest empirical research on clinician knowledge emerged soon after Finklehor's questions were raised. Attias and Goodwin (1985) investigated the knowledge and management of incest cases by pediatricians, psychologists, psychiatrists, and family counsellors (N = 108), as a means of ascertaining what gaps remained in professional education in the area of sexual abuse. The respondents reported relatively high degrees of relevant knowledge and clinical experience. However, the results of their survey indicated that some professionals retained misperceptions about sexual abuse such as believing that a disclosure represented a child's fantasy. There were also errors by respondents on questions regarding the correlation between physical and sexual abuse, and the likelihood of siblings being at risk. A split existed along gender lines. Male respondents made more errors on all the questions: "gender rather than discipline or personal clinical experience was an important predictor of response on key items... [and] the impact of Counsellor Schema 14 such gender differences on case management decisions may be magnified because males are overrepresented (80%) among the medical professionals who are most influential...." (p. 532). The study, although raising unsettling questions about the practitioners, was based on a limited sample, and selected from a telephone book which may not have included specialized agencies or services. The report was published in 1985, which is relatively old research in this rapidly evolving field. Five years later, Hibbard and Zollinger (1990) examined trends in knowledge in a larger sample of a variety of professionals (N = 902) who worked in some capacity with victims of childhood sexual abuse: medical personnel (39 physicians, 202 nurses), 236 child protection workers, 20 lawyers and judges, 118 law enforcement personnel, 57 psychologists, and 223 assorted service providers. The purpose of the survey was to identify pertinent issues that continued to be sources of confusion and error in knowledge and perceptions of childhood sexual abuse. Results of the true/false questionnaire indicated that, overall, the respondents had a good knowledge base; child protection workers were the most knowledgeable of the professionals, followed by psychologists and doctors. However, misunderstanding remained in some areas such as not recognizing the possibility that children may have positive feelings about the experience, and that abuse does not necessarily involve intercourse or physical force. Formal training was found to have a mild positive effect on knowledge, with statistically significant effects on half of the items. The results demonstrated positive growth in practitioners' knowledge bases since the Attias and Goodwin study. However, the sample was comprised primarily of professionals with experience and some relevant training; there was no comparison with novice practitioners. Furthermore, participants were self-selected and probably motivated in the area, as they were professionals who had accepted invitations to an educational program on childhood sexual abuse. An added weakness was that the questionnaire primarily addressed issues of disclosure and post disclosure procedures, and included only two questions about possible indicators. The majority of correct responses were for procedural issues. Hibbard and Zollinger concluded their study with a recommendation for continued education for all professionals that "needs to expand beyond the normal professional training received.... Further research . . . is necessary to address important issues regarding the most effective educational format to train professionals...." (pp. 353-4). Counsellor Schema 15 Their recommendation should be extrapolated to target practitioners who do not necessarily indicate an interest in learning about the subject or do not regularly work in the field, as their knowledge base may consequently be relatively impoverished. In a study unrelated to sexual abuse counselling, Rabinowitz (1993) reviewed research on decision-making biases among mental health practitioners. He noted several areas of clinician "oversight," one of which was sexual abuse. Citing data from several studies Rabinowitz noted that although over 50% of patients admitted to hospital had been sexually abused as children, and many of these individuals met DSM-III criteria for post-traumatic stress disorder arising from the abuse, staff failed to identify the sexual abuse histories in more than half of the cases. "Staff reported that sexual abuse was rare among patients. However, patients reported that staff had never asked them about abuse" (p. 302). It should be noted that the studies included in Rabinowitz's review date back only to 1988. The relative recency of his findings offer important information regarding the "invisibility" of childhood sexual abuse and suggest the need for increased education in the area of childhood sexual abuse. Attitudes. Knowledge on its own is insufficient for effective intervention with sexually abused clients. Also salient are counsellor variables including therapists' attitudes towards sexual abuse victims, an area that has received minimal attention in the literature. Adams and Betz (1993) noted the oversight. "This omission is particularly striking given that incest is one of the most emotionally provocative issues that counselors may address with a large number of clients" (p. 210). Further, attitudes are among "the most important issues regarding the treatment of sexual assault survivors" (p. 210). A number of dangers related to attitudes are inherent, including the potential for stereotyping, bias, and countertransference, all of which affect the therapeutic process. The ensuing clinical errors may by exacerbated for counsellors without appropriate preparation and education. Adams and Betz recommend further research in the area of attitudes, which "is important in guiding training interventions" (p. 210). Adams and Betz (1993) studied the effect of gender on counsellors' attitudes. They used the "Jackson Incest Blame Scale" (JIBS) (Jackson & Ferguson, 1983) to assess respondents' attributions of responsibility regarding childhood sexual abuse, and the "Incest Attitudes Survey" (IAS) (Ciccone, 1982) Counsellor Schema 16 to ascertain counsellors' attitudes about the phenomenon. Both scales were normed on samples of undergraduate college students. Overall, in comparison with established norms, the sample of counsellors in the Adams and Betz study were much less likely to engage in victim-blaming attitudes. Relevant and significant results of their study included the following: female counsellors had broader definitions of incest and were less likely to perceive incest reports as fantasies, and female counsellors held stronger beliefs that the clients would overcome the effects of abuse. The authors acknowledged the limited generalizability of their study due to the use of a small sample (N = 111; 67 female and 44 male counsellors), confined to one type of setting, university counselling centres. Other methodological problems were related to the low reliabilities of subscales of the JIBS and IAS, and the limitations of the available norms for both instruments. They recommended further research to "examine the conceptualization and measurement of incest related attitudes" (p. 215). Other components of attitudes are the stereotyping and myths associated with the victims of sexual crimes. Although the client population for Dye and Roth's (1990) research was sexual assault victims as opposed to women with histories of childhood sexual abuse, their concerns regarding the knowledge and attitudes of psychologists, social workers and psychiatrists (N = 257) are relevant. They administered an author-constructed questionnaire in which the "Rape Myth Acceptance Scale" (Burt, 1980) was embedded. Five additional subscales were designed to measure knowledge of symptoms and themes, treatment choices, and attribution of responsibility; Cronbach's alphas for the subscales ranged from .436 for "Discounting Sexual Assault" to .756 for "Knowing Themes." One positive finding was that the therapists did not generally endorse rape myths. There was an inverse correlation between age and knowledge of treatment themes: knowledge increased as therapist age decreased. Professional degree was significantly related to knowledge, with social workers scoring highest, followed by psychologists, then psychiatrists. There was virtually no consensus on common treatments. However, the correlation found between attitudes and treatment decisions suggested that clinical judgement may " . . . reflect negatively prejudiced assumptions held by the therapist" (p. 209) as opposed to being Counsellor Schema 17 based on the individual needs of the client. For example, therapists who endorsed "rape myths" were found to reflect this attitude in their treatment approach, by focusing on the client's role in the abuse. There were some methodological difficulties with the study: the use of a dated instrument; an author-generated scale with serious psychometric deficiencies (low reliability, no validity information); and a possible sampling bias that resulted in overrepresentation by female social workers and psychologists, and underrepresentation of male psychiatrists. However, the authors raised the crucial issue that "the extent to which the current research emphasis on sexual assault-specific treatment has been integrated into clinical practice remains unclear" (p. 193). Due to the potential implications for the counselling process, there is clearly a need to further investigate the relationship between training and counsellor perceptions of, and attitudes towards, their clients. Overall, the review of literature about professionals' knowledge and attitudes indicates the area is under-researched. Although there seems to be an improvement in the level of general knowledge among experienced clinicians, general deficiencies remain in both knowledge base and the application of that knowledge to practice principles. According to Enns et. al. (1995), "theoretical and treatment recommendations about working with adult survivors of sexual abuse have outstripped research regarding the efficacy of various treatment practices. Psychologists need to build a body of knowledge to fill in gaps in our understanding" (pp. 203-4). In addition, the area of attitudes warrants further investigation particularly given the emotionally evocative nature of childhood sexual abuse. The research on counsellor knowledge and attitudes about childhood sexual abuse has often employed psychometrically weak measures; thus the conclusions are preliminary and cannot be generalized to the full spectrum of counsellors. Counsellor Preparedness. The related area of the application of information in practice warrants consideration. Duncan (1987a, 1987b) was one of the first authors to address the issue of counsellor preparedness for treating sexual abuse. Based on the reported prevalence, she determined that specialized treatment was justified (1987a). She raised the ethical question of counsellors ".. . providing treatment in an area where specialized training is necessary, but may not have been completed for a variety of Counsellor Schema 18 reasons.. . . However, it is the professional's responsibility to seek whatever information or training is necessary because the 'fly by the seat of the proverbial pants' approach is clearly not enough. There are real risks involved here, especially for the victim" (p. 62). The purpose of her articles was to provide counsellors with information about dealing with childhood sexual abuse, to offer guidelines for the work, and to raise pertinent legal and ethical issues. She also asserted the importance of remaining abreast of research developments as the field continues to evolve. She cautioned readers to be aware of bias in the literature regarding treatment modalities, and of one's personal biases that may be exacerbated by societal stereotyping. A limitation of her article is the reliance on older research, particularly regarding gender stereotyping, which may have undergone changes since that time. However, the ethical concerns raised and the need for training cannot be understated. Counsellor preparedness was also a theme in Frazier and Cohen's (1992) investigation of research on the sexual victimization of women and ensuing implications for counsellor training. In describing the rationale for their study, they cited the dearth of research in the area, the lack of data regarding counsellor preparation, and the potential for unintentional secondary victimization of clients by inadequately trained counsellors. In a sample of female university counselling centre clients (N = 82), they found a high correlation between sexual victimization and serious concerns including eating disorders, suicidal ideation, and further assault (all significant at the .001 level); self-mutilation, depression, family abuse, and weight issues (significant at the .01 level); and substance abuse and anger management, significant at .05 (p. 150). Frazier and Cohen's concern was that"... some mental health professionals may be unhelpful to victims.. . simply because they lack knowledge about the issues" (p. 151). They therefore advocated an increased emphasis in training programs in the areas of therapeutic process, attitudes, beliefs, myths, and the potential affective responses of counsellors towards the clients. The methodologically-sound research was based on one counselling centre, so results must be generalized with caution. Although significant new empirical data were not realized in this investigation, its importance is twofold. It highlighted the need for attention to training and education for counsellors in the area of sexual victimization, and introduced attitudes and beliefs as significant issues warranting investigation. Counsellor Schema 19 Education and Training. The deficiencies in education vis-a-vis sexual abuse were also noted by Pope and Feldman-Summers (1992) in their national survey of clinical and counselling psychologists (N = 290). They raised the issue of training specific to the area of sexual abuse and observed,"... the professional literature suggests that graduate training programs have largely ignored abuse as a specific content area" (p. 353). The purpose of their study was to ascertain respondents' perceptions of the adequacy of abuse-specific graduate training they received and their perceived competence in working with abused clients, and to determine the correlation of these variables with gender and level of education. The adequacy of graduate training programs regarding several types of physical and sexual abuse was rated by respondents on a five-point scale that ranged from one indicating "very poor; e.g., little or no attention devoted to the topic" (p. 356), to five which denoted "very good." The modal response, for all but one of the categories as rated by both male and female participants, was 1.0. In other words, participants frequently judged their training to be "very poor." A regression analysis incorporating other variables yielded the statistically significant result that the likelihood of a higher rating increased with recency of graduation, possibly indicating that training is improving. Pope and Feldman-Summers (1992) noted some difficulties in their research design such as sampling, content and wording that may have limited their findings. They were therefore cautious about the generalizability of the results to the larger population of clinical and counselling psychologists. Due to the strikingly low ratings accorded to graduate training programs, the authors expressed concern about the lack of attention given to the areas of physical and sexual abuse. And, although it appeared that there was a trend towards improvement, "significant changes in graduate programs... are needed if they are to be regarded as being any better than 'very poor' in training in the abuse areas in the eyes of practitioners" (p. 358). Their findings highlight the impoverished state of appropriate counsellor education regarding sexual abuse. Alpert and Paulson (1990) also argued in favour of furthering graduate level education in the field of childhood sexual abuse and presented descriptions of two graduate level courses offered at their university, one in research and theory, and the other a practicum course. By way of introducing and justifying the Counsellor Schema 20 courses, they described professional training as "in a pioneer state" (p. 367), which echoed Sgroi's comments in 1982. Their comprehensive literature search revealed " . . . no reports or studies concerned with training around child sexual abuse at the graduate school level. For the most part, psychologists work clinically with child sexual abuse victims or adult survivors of child sexual abuse for the first time in their externships or internships" (pp. 366-7). Alpert and Paulson alerted the profession to the possibility that counselling trainees and psychology interns are not receiving adequate preparation for this work. There is consensus among the few researchers who have investigated the state of graduate level education in the area of childhood sexual abuse that existing programs fall far short of what might be considered adequate training for practitioners. Given the nature of this significant social problem, appropriate and relevant counsellor education is critical. Training specific to this field may also have a role in ameliorating counsellors' schemas. Schema General information about schemas and specific features will be presented, followed by their role and influence on the therapeutic process. Discussions on the phenomenon of hypothesis testing, and the concept of cognitive complexity are included because of their relevance to schema. This section will be concluded with some comments on the difficulties and practicalities of schema measurement. Turk and Salovey (1985) provide a clear definition of schemas and their role for counsellors. Schemas are "preexisting knowledge structures [that] are used to filter the wealth of information that continually confronts the clinician as he or she tries to make sense of incoming stimuli, to answer complex questions, and to make difficult judgments" (p. 20). The prior knowledge individuals have in their schemas about people, events or phenomena permits perceivers to shape their perceptions and make sense of new instances. Schemas contain bench-marks for inclusion in a category, and criteria for how many characteristics and attributes must match between a new instance and the category in order for the new example to be accepted as a member of that category. The criteria vary as a function of schema maturity, as do the number of characteristics contained in a particular schema (see discussion on "Schema Properties" below). Schemas have an impact on memory, expectations, generalizations, and inference. Counsellor Schema 21 The strength of a schema's effect on inference increases with the degree of similarity between the phenomenon and a prototype of that phenomenon. Research in counselling has demonstrated that the more highly symptomatic the client in comparison to established norms and criteria regarding particular diagnoses, the greater the reliability and accuracy of a clinician's diagnosis of a client's disorder (Fiske & Taylor, 1984). Implicit in this is the obverse: clients who do not manifest standard indicators of a particular problem may be misdiagnosed. In addition, counsellors may generalize from a few very visible characteristics which could conceivably be representative of more than one complaint, and may also result in erroneous interpretations. It can be speculated that reliance on immature schemas due to the lack of a sound knowledge base may therefore contribute to counsellors drawing conclusions based on limited data. In the case of clients who knowingly or unwittingly disguise their abuse histories with other complaints, inexperienced counsellors may fail to recognize that alternative explanations (i.e., sexual abuse) account for the presenting symptomatology. Schematic Properties. Research has shown that as a schema develops, it becomes "more abstract, more complex, more organized, more moderate, and more conservative" (Fiske & Taylor, 1984, p. 173). As the schema matures due to frequency of experiences with similar categories or events, the more sophisticated it becomes. A broader range of cues from the environment can be processed and assimilated, rapidly and easily. There are also disadvantages to mature schemas. They are more conservative, or resistant to change, as they are composed of more elements. The perseverance effect, which causes schemas to endure in spite of contradictory evidence, fuels this intransigence. The perceiver may tenaciously and creatively seek to assimilate exceptional cases rather than surrender his or her schema (Fiske & Taylor, 1984). Finally, new learning of related concepts is facilitated by developing, rather than mature, schemas. A traditional mechanic, for example, who now has to use computerized scopes to identify problems may have difficulty understanding what the print-out means in practical terms, whereas the schema of an apprentice who lacks that history may be better able to integrate the new technology. Organization. A schema's organization depends on the number and structure of the links among the elements. A schema becomes more intricately organized and more sophisticated with experience. Counsellor Schema 22 Greater experience promotes the ability to assimilate ambiguous information, and to extrapolate from seemingly unrelated cues or events. Experience "creates more organized schemata and allows greater capacity for managing inconsistencies" (Fiske & Taylor, 1985, p. 174). A person with a more developed schema is better able to attend to, rather than dismiss, schema-discrepant information. The information contained in a mature schema becomes more highly structured and can be accessed readily and quickly as the greater degree of organization frees "processing capacity" (p. 174). A less developed schema restricts the perceiver's ability to recognize anything other than clearly schema-consistent information and cues. Breadth. With maturation, the schema becomes more complex, with a greater number of characteristics. Essentially, the range and type of elements increases with experience and exposure to phenomena. The schema matures to include not only attributes about a person or event, but information about function, goals, and so on. The schema therefore becomes broader and more comprehensive. An automobile mechanic's schema of a car engine, for example, complete with knowledge of the parts, how they work together, and the ability to recognize what is causing problems, is more mature than that of a car owner who does not even know how to check the oil level. The mature schema contains more dimensions and a broader range of characteristics and attributes which provides a more complete picture from which generalizations can be made: the mechanic can transfer general knowledge of an engine to many models of cars. Extremity. Schema maturation, as the perceiver encounters more examples of a phenomenon, minimizes the extremes of judgement. Schemas become tempered with the addition of information, moderating any existing extremes attributed to the person or category. Generalizations based on limited information and reliance on stereotypes is more common in developing schemas. The experienced perceiver of a phenomenon will be less likely to categorize it as either one way or the other. A more mature schema will permit the perceiver to recognize the range of behaviours or attributes between opposing poles (Fiske & Taylor, 1984). For example, a person whose experience with individuals who have developmental delays is limited to an isolated contact, may believe their personalities to be docile because that was the demeanour of the individual who was encountered. Someone who has a family member with a similar Counsellor Schema 23 developmental delay or who works in the field will have a more complex schema that enables the individual to be cognizant of the wide range of personality styles. Schema extremity has implications for counselling because of the inherent risks of skewed perceptions about clients and issues. Prototypes. Schemas are different from prototypes. Prototypes "can be thought of as a particularly salient exemplar of a given category" (Turk & Salovey, 1985, p. 21) and develop from recurring exposure to, and experience with, members of a group or category (Smith & Zarate, 1990). A prototype more closely resembles an instance with specific characteristics which may or may not be relevant for the category, and which evokes particular attributions and expectations of that phenomenon. A classic example of a prototype is a "red sports car." The colour red is not a necessary condition for a vehicle to be categorized as a sports car, and most people would agree that sports cars can be any colour. But one's prototype of a sports car frequently includes a colour, and that colour is often red. Prototypes contain all known, relevant or irrelevant, attributes. Schemas, in contrast, have more latitude and "permit some features to be unspecified" (Fiske & Taylor, 1984, p. 148). Clinicians may develop idiosyncratic prototypes that "reflect [their] tendency... to encode, process, and retrieve information. . . based on previous experience (e.g., experiences with certain types of clients)" (Turk & Salovey, 1985, p. 21). In other words, counsellors may encapsulate their views of clients and their issues based on prototypes containing attributes that are not necessarily characteristics pertinent to that issue. For example, a counsellor may develop a prototype of a client with a substance abuse problem that includes a belief about socioeconomic status because the counsellor's experience with addictions has regularly been with clients who received financial assistance. Even though an individual's financial circumstances are not a necessary precondition for drug addiction, this attribute may be contained in the counsellor's prototype. Such cognitive strategies can lead to inferential errors about the clients and treatment planning, and can be compounded by the tendency of clinicians to "chunk. . . information into a congruent whole to avoid overload of stimuli" (Rabinowitz, 1993, p. 305). Counsellors may therefore ignore other cues, limiting their perceptions. Counsellor Schema 24 Person Schemas. There are four primary categories of schema (Fiske & Taylor, 1984). Self schemas are generalizations that inform the individual about his or her own psychology, personality and behaviours, and are based on past experiences. Role schemas include information regarding broad social categories that place an individual in society, such as age, race, and occupation. Stereotypes are a particular class of role schema. Event schemas aid in the understanding of what typically occurs in certain social situations or on particular occasions. Person schemas, a subtype of prototypes, contain knowledge about the traits and goals of categories of people or individuals, and are of particular importance for this study, namely the schemas counsellors have about sexually abused clients. Person schemas are comprised of an individual's traits that assist the perceiver to categorize that individual based on the relevance of these characteristics to the perceiver's schema. Schemas determine what information is consistent for a given person or category of person, and what is most readily remembered. Categorization of an individual depends on the degree of similarity of that individual to the range of characteristics contained in the perceiver's schema. When a sufficient number of features approximate the typical summary representation, or schema, the person is accepted as an instance of the category (Smith & Zarate, 1990). Assumptions about characteristics, personality, and behaviour are then made about the person based on the beliefs associated with the category of person. Schemas may be about any of a number of groups including race, appearance, occupation, psychological diagnosis, areas of residence, and so on. A prototypic member of a category may be integrated into the schema and used as a referent for future cases, which can skew the schema by introducing non-relevant attributes. Person schemas have particular significance in the counsellor-client interaction because they guide the counsellor's perceptions, beliefs, and assumptions about the client. A client may be categorized on the basis of a number of variables including appearance, socioeconomic status, family history information, and personal issues as described by the client. Accepting the client as an instance of a particular category too quickly or based on a schema irrelevant to the context (e.g., employing a schema about socioeconomic status rather than one pertinent to the counselling situation) can lead to misdiagnosis or clinical oversights. Counsellor Schema 25 The implications for therapy will be discussed in more detail (see discussion on "Effect on Counselling Process" below). Expertise. As is discussed elsewhere (see "Experience"), expertise differs from experience. Experience is a necessary but not sufficient condition for expertise (Hillerbrand & Claiborn, 1990). Additional essential elements are distinguished skills in the field and a high level of specialized knowledge which is well organized and easily accessible, similar to the description of a complex schema. In their study comparing novice with expert clinicians, the criteria for expert psychologists included longevity of practice and nomination by a body of peers for "above-average diagnostic ability" (Hillerbrand & Claiborn, 1990, p. 685). Spengler and Strohmer (1993) note that a protracted duration of clinical experience is not synonymous with expert status, as it has not been demonstrated that experts are relatively free of bias or error. Identifying true experts is a considerable task, and little is known about the cognitive processes of an expert counsellor. Research on experts versus non experts indicates that the likelihood of noticing and utilizing discrepant information increases with the amount of one's expertise and the subsequent accessibility of a particular schema (Higgins & Bargh, 1987). Expert schemas integrate information more quickly and with less confusion to produce "larger perceptual units" (Fiske & Taylor, 1984, p. 158). Experts "excel at retrieving information and making judgements that are routine. A novice... may know the same answer but reach it more slowly and have less information to back it up" (p. 158). Experts are known to possess the ability to reflexively apply their knowledge and cognitive conceptualizations to familiar situations and can extrapolate their expertise to novel situations. When presented with reasoning problems, experts, in comparison to novices, make "qualitatively different inferences in their reasoning, focus on different problem features,... [access] different knowledge in memory, and [focus] on different case information" (Hillerbrand & Claiborn, 1990, p. 684). They will therefore follow their reasoning to different conclusions. Novices are more likely to overlook key elements and attend to superficial characteristics. Experts have a more sophisticated method of integrating Counsellor Schema 26 information and matching it to schemas. Therefore, the cognitive processing abilities of an expert are distinct from a novice, and will therefore differentially affect the clinical judgement processes of counsellors with varying levels of expertise. Schema Activation. Whether a schema will be activated when the perceiver is presented with particular information depends on the extent of its prior use. The more frequently or recently a schema is employed, the more likely it will be primed, which has been described as somewhat like being on standby (Fiske & Taylor, 1984). There are variable implications of this feature for counsellors. The schema of a counsellor who is working regularly with sexually abused clients is more likely to be primed due to its probable maturity and to the regularity with which it is operationalized. A counsellor whose sexual abuse schema is less comprehensive and not as accessible, cannot draw on it as readily, may miss some indicators, and take longer to classify the client. This may be positive, however, because the unavailability of prototypes and heuristics in a novice's emerging schema may lead the less experienced counsellor to conduct a more thorough assessment of the client, and to be tentative when generating hypotheses. In other words, schemas could affect a counsellor's approach in various ways, and there may potential benefits for both mature and less developed schema. Effect on Counselling Process. The counselling process is one that is exceptionally complex partly because of the multitude of factors practitioners bring to the interaction including their experiences, attitudes, and expectancies, all of which are encompassed in the concept of schema. Schemas are "theories or concepts that guide how people take in, remember, and make inferences about raw data" (Fiske & Taylor, 1984, p. 140), which, for counsellors, would therefore shape their conceptualizations and clinical judgements about clients, and influence the direction of the therapy. "The most fundamental principle suggested by schema research is that people simplify reality.... The schema contains knowledge about what information would be congruent.. . what would be incongruent. . . and what is irrelevant" (p. 141). "Cognitive organization of previous experience... commonly referred to as schemata... have been hypothesized to guide the selection and interpretation process" (Higgins & Bargh, 1987, p. 378). This premise is fraught with implications for the counselling process because schemas act as filters for Counsellor Schema 27 information: the type of information attended to, how it is processed, and the hypotheses generated. This, in turn, will affect the interaction, and the manner in which the counsellor guides the course of therapy (Haverkamp, 1993). The schemas employed by therapists in their work develop as a function of personal and professional variables, the latter including experience, training, and theoretical orientation. Little research has been conducted on the role of schemas for the counsellor and the therapeutic process, and counsellor schema has been recommended as a vital area of investigation (Heppner & Frazier, 1992) Schemas are integral to clinical judgement, one of the decision making processes in therapy (Turk & Salovey, 1985). Schemas influence the counselling process at various stages, and can result in bias. The initial concerns or manner of presentation by a client upon commencement of therapy may effect categorization by the counsellor of that client based on the client's degree of similarity to a client or problem prototype. As a result of person schemas, individuals' distinct characteristics may be minimized or overlooked by counsellors seeking cues that permit them to fit the client into a prototypic classification. Furthermore, "once a perceiver places a person into a particular category, the perceiver is likely to misremember the presence of category-consistent but never-seen attributes" (Fiske & Taylor, 1984, p. 152). Non-existent traits or behaviours may be erroneously attributed to clients merely because they fit the model of a particular client type. This is an example of a representativeness heuristic: how closely the individual case resembles features intrinsic to the counsellor's schema can result in overlooking disconfirming data (Turk & Salovey, 1985). At the intake level, there is a potential for bias from a clinician's theoretical orientation as a result of the inherent preconceptions that necessarily underly the kinds of questions asked and the type of information sought. For example, a counsellor who works only with individuals may assume the issues are intrapsychic and neglect to ask about family issues. Continually relying on a particular perceptual lens and a familiar set of questions reinforces the schema, may colour expectations, and may preclude the investigation of alternative hypotheses about a client's presenting problem. Rabinowitz (1993) noted that during the input phase, clinicians form first impressions which anchor judgements and provide the foundation for their hypotheses. These first impressions have consistently been demonstrated to be formed Counsellor Schema 28 quite rapidly and then endure, often regardless of the subsequent presentation of more information which may or may not substantiate the original hypothesis. The assessment and decision-making processes of counselling are influenced by the utilization of heuristics, which are "shorthand procedures for reducing information" (Turk & Salovey, 1985, p. 20). Data are "organized around a central theme, typically the clinicians' initial hypothesis, and are synthesized to support that hypothesis by using confirmatory strategies...." (Rabinowitz, 1993, p. 305). A preference has been found for anecdotal information over systematic or statistical evidence which can lead to inaccuracies if clinicians' prototypes about clients or issues include faulty or exceptional information. A therapist unfamiliar with the traumatic effects of recent immigration, for example, may not consider this factor when working with a new Canadian family experiencing parent-teen conflict, choosing instead to reduce the problem focus to the current conflict. As the course of therapy continues, there may be a tendency to misconstrue the client's emerging changes. A counsellor who observes a reduction in her client's expressed anger and determines that the direction of therapy should be maintained because progress is being made, may not notice that alcohol intake has increased concomitantly because a substance abuse problem was not defined at the outset. It may therefore not be addressed in therapy. The complex cognitive processing that persists throughout the stages of therapy is thus affected by the counsellor's schema. Hypothesis Testing. Although investigating the role of hypothesis testing is beyond the scope of this study, it receives mention because of its connection to schema and the subsequent implications for counselling. A perceiver's purpose for observing an event or person will affect the activation of a schema: goals such as impression formation, behaviour anticipation, or empathizing, all have an effect. Hypotheses, which are filtered through schemas, have a critical role in the selective attention to information. The phenomenon of hypothesis testing (Snyder & Thomsen, 1988) is the process in which "people attempt to use their interactions as opportunities to test the accuracy of their beliefs and intuitions about other people" (p. 129), often through strategies of selective questioning to gather information in support of the hypothesis. Counsellor Schema 29 Throughout therapy,". . . as therapists proceed to test their hypotheses, they may preferentially search for, preferentially solicit, and preferentially elicit information that is relevant to and consistent with their beliefs and expectations" (Snyder & Thomsen, 1988, p. 131). Consistent or expected information is "preferentially encoded" (Higgins & Bargh, 1987, p. 379), recalled more readily, integrated more definitively, and more resistant to change. This phenomenon will thus also affect the counsellor's clinical judgement. The literature suggests a number of repercussions as the belief system becomes more entrenched: the person is more likely to seek out confirmatory evidence which increases the likelihood that initial impressions will solidify and persevere. As a result of confirmatory bias, the seeking of information supportive of one's hypothesis to the neglect of contradictory cues, there is the " . . . inherent risk... that counselors may draw premature conclusions about client problems and make inaccurate assumptions and decisions about diagnosis and treatment" (Haverkamp, 1993, p. 303). All counsellors rely on the process of hypothesis testing. If the schema lacks the breadth of information which thereby limits the available data base from which to draw, the counsellor with an undeveloped schema may necessarily rely more on confirmatory or biased hypothesis testing. Higgins and Bargh (1987) reviewed a series of studies undertaken by Snyder and Swann during the later 1970s and concluded that participants "had confirmatory strategies for testing hypotheses about people, where 'confirmatory' questions are nondiagnostic in that they do not allow answers that disconfirm the hypothesis" (Higgins & Bargh, 1987, p. 398). The tendency for employing hypothesis confirming questions was determined to be a function of the unavailability of an alternative hypothesis. Although there was a tendency to use diagnostic questions, people "will display a preference for hypothesis-matching questions when the hypothesis under consideration is the only accessible alternative and is believed to be true" (pp. 401-2). Counsellors lacking a comprehensive knowledge base in the field of sexual abuse who encounter a non-disclosing client with ambiguous presentation may not have that hypothesis at their disposal, and may elicit information that substantiates some other diagnosis. Appropriate treatment may therefore not be provided. Counsellor Schema 30 Strohmer and Spengler (1993) were involved in several studies to investigate the phenomenon of hypothesis testing and initially found that counsellors with varying levels of experience employed open-ended questioning strategies that were unbiased. Further research by the same authors (Strohmer & Spengler, 1993) yielded somewhat different results, with all participants demonstrating a high degree of confirmatory bias, a finding that held for experienced and inexperienced therapists. More clarity was achieved on this issue once the variable of positive versus negative client information was factored in: their literature review "suggested that clinicians tend to favor the collection of negative client information over positive client information" (p. 5), an "equally troubling clinician bias" (p. 5). The implication of a strong confirmatory bias with respect to negative hypotheses and information, compared with a bias for positive data is that clinicians " . . . will test a negative hypothesis in a confirmatory way and a positive hypothesis in a disconfirmatory way" (p. 6). The later studies did not address the variable of experience, therefore the relationship between experience and negative bias is unclear. Research on confirmatory bias has begun to " . . . challenge... the contention that counsellors are typically neutral in their approach to client problem identification..." (Haverkamp, 1993, p. 314). The inattention to all aspects of a client's issues and experience may attenuate clinical judgement and treatment decision-making. Cognitive Complexity. Another stream of research addressing how therapists process information and make decisions about clients is in the area of cognitive complexity, a cognitive or mental ability on which individual differences emerge, which may have an impact on counsellor schema. Spengler and Strohmer (1993) have been involved in several studies addressing this area. Their work on cognitive complexity derived in part from the unsatisfactory research results on other counsellor attributes such as age, experience, confidence, and theoretical orientation, which do not further the understanding of how counsellors process information. Individuals with higher cognitive complexity have at their disposal multiple dimensions for construing a person and that person's behaviour. Research to date has found that counsellors who have greater cognitive complexity are " . . . more adept at utilizing incongruent or contradictory client information,... search for more bits of information,... and integrate greater amounts and types of information.... Counsellor Schema 31 Cognitively complex individuals have also been found to be less prone to invoke cognitive simplification strategies... thought to underlie errors in clinical judgment" (p. 8). Schemas, on the other hand, have been variously described as cognitive structures, themes, or maps, based on preexisting knowledge. Schematic complexity refers to the intricacy and sophistication of the structure, rather than to an ability as with cognitive complexity. In their study of clinical bias, Spengler and Strohmer found support for their hypothesis that higher complexity clinicians surpassed their lower complexity counterparts in their ability to assimilate discrepant information and to refrain from applying stereotypes to clients. This finding has implications for the process of counsellor decision making, as cognitive complexity may. mediate the therapist's schema by enhancing its organization and breadth and thus the therapist's consequent ability to consolidate a wider range of cues and data. Furthermore, Spengler and Strohmer posit that if cognitive complexity is a state as opposed to trait variable, it may be amenable to change, which has implications for counsellor education and training. Schema Measurement. Cognitive processes, by their nature, cannot be readily subjected to empirical measurement. Evaluation is generally achieved through inferential means, as the processes cannot be directly observed. Measuring "traces," or records, is a common strategy for assessing processes such as attention (Fiske & Taylor, 1984). For example, one method to determine the amount of attention paid by a participant to different people in a meeting is to provide separate videotapes of each attendee, and to monitor the amount of time the participant spends watching each tape. Similarly, allowing participants to have control over the slide changing mechanism while viewing particular slides provides a means to assess the amount of time spent attending to the different stimuli being presented. Although the actual attention is not measured per se, a record of its manifestation is. What cannot be controlled or accessed are factors such as whether the participant's attention is truly on the stimulus or whether the individual is thinking of other things (e.g., daydreaming). Similarly, schema research utilizes inferential means of evaluation. There are various methods of assessment, some of which include the following examples. To demonstrate the impact of schemas on Counsellor Schema 32 understanding, psychologists monitor the accuracy, speed, and focus of memories: how accurately participants recalled previously presented information; how quickly they remembered pertinent data; and what different things were reported (Fiske & Taylor, 1984). To study the effect of prototypes on memory, participants can be given a list of adjectives that are said to describe a particular type of person such as an extrovert. At a later time they are asked to select which adjectives on a second list were originally presented on the first. The results are significant if the participants identify words prototypic of an extrovert that were not on the first list (Fiske & Taylor, 1984). The thought listing procedure has been used recently in research on counsellors' responses to clients presenting with sexual abuse, physical abuse, and role conflict issues (Parisien & Long, 1994). Cognitions of the counsellors were obtained by this self report method which had participants list their self talk. Research on clinical bias (Strohmer & Spengler, 1993) used client narratives with varying amounts of confirmatory or disconfirmatory units of information. Participants were asked to recall the information at a later date to ascertain the way client information was remembered by the clinicians in the study. Thus various strategies can be employed to tap cognitive processes. Investigating sexual abuse schema would utilize similar methods. In summary, the emotionally evocative nature of sexual abuse treatment and its evolving knowledge base makes the area particularly susceptible to biased cognitive processing, which has consequences for the counselling field. One of the fundamental structures that underlies cognitive processing, and ultimately clinical judgement, is schema. As a cornerstone of many of the processes in counselling, the importance of schemas cannot be understated. Further research is necessary to determine the nature of their role in the counselling process. Experience The literature on counsellor experience describes empirical research dating back several decades which examined the effect of experience on various aspects of the process, including clinical judgement and decision-making. Research on the variable of experience in counselling is one of several areas that Spengler and Strohmer (1993) identified as having been "criticized for inconsistent and contradictory Counsellor Schema 33 results, as well as a lack of relevance to tasks meaningfully associated with clinical decision-making" (p. 4). Inconclusive findings about the nature of the relationship between counsellor experience and various aspects of the therapeutic process pervade the literature. Counsellor experience may differentially influence various aspects of the therapeutic process. Mallinckrodt and Nelson (1991), in their study on the relationship of counsellor experience and therapeutic alliance, noted the diverse effects of experience. They suggested that experience serves to amplify a clinician's knowledge base and helps them develop "more efficient strategies for collecting and processing new information" (p. 133), leading to greater sophistication of case conceptualization abilities. Experience effects do not, however, generally emerge on the variables of listening and empathic reflecting due to the emphasis in counsellor training on the acquisition of those skills. Experience working in a particular field or with a specific phenomenon increases knowledge and ameliorates the cognitive apparatus and skills that serve as the basis for information processing (Mallinckrodt & Nelson, 1991). Counsellors, knowingly or not, rely heavily on their schemas, the "cognitive structure that represents organized knowledge about a concept and that guides how people select, remember, and make inferences about information in their environments" (Heppner & Frazier, 1992, p. 158). Schemas aid the perceiver to filter, understand and integrate information, and permit clinicians to develop hypotheses about their clients (Turk & Salovey, 1985). The use of schemas permeates the therapeutic process, from the initial information processing stage through client conceptualization, clinical judgement, and later inferences. The schemas operating for any particular counsellor develop as a result of training, theoretical orientation, and experience (Turk & Salovey, 1985). Relationship to Clinical Judgement. Many of the studies that provide the foundation for current knowledge in the area of experience and clinical judgement were undertaken in the 1960s. The overall conclusion of a survey of that body of research (Wiggins, 1973) was that training and experience were not positively related to clinical judgement. A more recent review of the literature was conducted by Garb (1989), who followed up on the 1982 recommendation of the American Psychological Association task force charged with evaluating the merits of training and education in the field. The task force had noted Counsellor Schema 34 that "there is no evidence that professional training and experience are related to professional competence" (Garb, 1989, p. 387) and recommended that research be undertaken addressing the value of specific training and educational requirements. Garb's intention was to update the literature by including previously omitted studies and more recent research. He considered 55 studies that investigated the existence of differences between graduate students with differing amounts of experience, and clinicians of varying amounts of experience. Following is a synopsis of Garb's findings that are relevant to this research project. In many of the studies, clinical judgement is equated with making assessments based on psychological test data. Contrary to what one might intuitively expect, amount of experience did not discriminate among mental health practitioners. There was little to distinguish experienced clinicians from their less experienced counterparts or from graduate students (Garb, 1989, p. 388) on measures of clinical predictive validity using standardized personality assessment instruments. Participants of diverse levels of experience, who were provided with test results of clients or patients, were not differentially able to accurately predict treatment issues. Similarly, when assigned tasks to evaluate the results of projective tests, accuracy of client classification was not significantly related to training and experience (p. 390). However, training did appear to have a positive effect on validity when using particular and specific assessment instruments. Garb offered several possible explanations for the findings. He questioned whether the results may have been a function of the outdated training of clinicians who had been working in the field longer. He also suggested that clinicians may have difficulty learning from experience because of misleading or biased feedback, or the unavailability of feedback for particular tasks. Finally, Garb considered the cognitive processes of inadequate hypothesis testing strategies, hindsight bias, and the fallibility of memory as factors contributing to the failure of experienced clinicians to exhibit greater accuracy in clinical judgement tasks. It can be surmised that the cognitively complex process of clinical judgement may not be amenable to measurement in this manner and studies of this type may thus be obscuring existing differences. Effect of Amount of Information. Another variable in the process of clinical judgement, the amount of information provided, was considered in some research. Weiss (1963) cited several studies that examined Counsellor Schema 35 coi as the effect of the amount of information on clinical predictive validity resulting in equivocal findings. Studies supporting the contention that accuracy can be augmented by increasing the available information mcluded that this holds only if the data are provided in case history form or as part of intake information as opposed to test data (p. 257). Furthermore, only a few types of information, primarily class membership information (e.g., age, sex, occupation) were found to account for accuracy among participants (p. 258). Therefore, the moderating variable may be the type, as opposed to amount, of information provided. Brenner and Howard (1976), in their investigation of clinical judgement, favoured the description of clinical judgement used by Calvin and Cutin in 1962: "a synthesis of impressions ~ a process that involves reaching a decision by weighing and combining bits of information" (p. 721), which is a schema based process. Having noted that a common finding in the literature was that "experienced clinicians were no more accurate than less experienced clinicians" (p. 722), and citing the common assumption that an increase in information would concomitantly increase the accuracy of a judgement, Brenner and Howard, like Weiss, introduced into their investigation the variable of amount of information provided. The instrument used in their study was the "Therapy Session Report" which contained several assessment areas, including some that allowed broader scope for speculation and interpretation. Their large sample was comprised of six groups of 18 judges each (three experience groups plus three control groups). Their primary hypothesis was that, as increasing amounts of information were provided, experienced therapists would strive to integrate all of it, resulting in confusion, less attention to the client, and a concomitant reduction in accurate empathy (p.722). Novice clinicians, however, were predicted to increase in their accuracy with the provision of more information as they would selectively attend to information in an attempt to avoid being overwhelmed. Brenner and Howard found support for their hypothesis at a statistically significant level on the majority of the subsections of the Therapy Session Report. The authors provided several possible explanations for their findings (p. 726). They suggested that the entire amount of information emerging in a therapy session from verbal and non-verbal cues may exceed that which can be realistically integrated regardless of experience level. It may also be that Counsellor Schema 36 experienced therapists were able to intuit and understand the client's process at a deeper level than can the client, and in reflecting it, left the unaware client feeling misunderstood: "a clinician's hypothesis with regard to unconscious material or one that is based on a combination of much disparate information may be valid, yet, paradoxically, may distance him from the patient" (p. 726). The intermediate-level therapists were most accurate overall, suggesting that they used "an optimal amount of information" (p. 727). Clavelle and Turner (1980) compared masters-level social workers (N = 11), Ph.D. level psychologists (N = 13) and paraprofessionals who had completed an eight-week training course for psychology/social work assistants (N = 32) on their information-gathering styles and on their decision-making about critical issues derived from emergency intake interview information. In response to all the questions posed about the "patient", no significant differences were found among the three groups of participants on their decisions. The questions pertained to patient disposition (whether to make a psychiatric admission or administer psychiatric medication) and the risk of suicide. There were no significant differences between the groups on the levels of within-group consensus. There were some individual differences by profession on the amount of confidence about their decisions but no overall statistically significant between-groups differences. For example, the psychologists were significantly more sure of their decisions on four of the six questions than were the paraprofessionals, and the social workers over the paraprofessionals on one question. There was, however, a significant experience effect within the paraprofessional group: the greater number of years of experience, the more confident they were about their decisions. Clavelle and Turner suggested several explanations (p. 837) to account for the lack of significant differences between various experience and professional groups. The cases used may have been too complex to allow the greater knowledge of psychologists and social workers to become apparent. The study's assignment may have been too great a departure from actual clinical tasks to permit the professionals' expertise to become evident. A further possibility was that the mean age of the professionals (social workers, M = 33.4 years; psychologists, M = 31.2 years; paraprofessionals, M = 28.2 years) was relatively young and an older, more experienced group might have performed differently. Finally, the study Counsellor Schema 37 involved "emergency decisions" which may have been less prone to between-groups differences than would other types of clinical decisions. Such conjectures about the failure of research to yield empirical evidence for experience differences on clinical judgement were reiterated by Brenner and Howard (1976) who raised the issue of confounds in previous studies: "because of the need to obtain quantifiable data, many of the . . . studies have used tasks of questionable relevance to the daily activities of clinical psychologists" (p 722). An alternative explanation may be that counsellors' schemas are so entrenched that the amount or type of information provided does not affect their perceptions or inferences about clients. This is speculative as it has not been researched, but may help elucidate the equivocal findings. In summary, the empirical evidence supporting a positive relationship between experience and accurate clinical judgement is scant. Because this general finding does not conform to expectations, researchers have sought to understand why no association has emerged. As noted above, there are several confounds that surface regularly which may have obscured the existence of experience effects. Further ambiguities may be due to the delineation between expertise and experience which may be misunderstood or used interchangeably in error. Research in this area may need to use different types of experimental tasks to measure clinical judgement. None of the studies surveyed directly addressed counsellor schema, which permeates the clinical judgement process, and which may be more likely to yield evidence of an experience difference. Summary The sexual abuse literature has consistently recommended that more research be completed on counsellor variables including knowledge, attitudes, and effectiveness of training and education. The importance of this research has been suggested by the ethical and practical concerns about who should be working with this potentially highly traumatized population; and whether the educational system should better prepare counsellors for providing quality service. Further research on counsellor schema is also necessary. The examination of the relationship between counsellor experience and schema is warranted given the conflicting conclusions from the research on Counsellor Schema 38 experience and schema: the former suggests experience is not positively correlated with clinical judgement, whereas the outcomes from studies on schema suggest experience enhances complexity. To date, no studies have compared the sexual abuse schemas of counsellors with diverse amounts of experience. The current study was not designed to supplement the existing body of evidence about counsellors' knowledge in the area of sexual abuse, but to investigate differences of schema complexity. It was beyond the scope of this study to replicate research in the areas of schema and confirmatory bias, hypothesis testing, and cognitive complexity, although the results may have general implications for the field of counsellor schema. It was expected that results would lend themselves to descriptions of counsellors' sexual abuse schemas, which may provide directions for future research and for the nature of professional education. The generalizability of the results were limited to counsellors in the area of sexual abuse. Accordingly, the purpose of this initial investigation was to determine the relationship between counsellors' professional experience with sexual abuse and their schemas of sexually abused clients. The primary research question is, how does experience affect a counsellor's schema of sexually abused clients? A positive relationship between experience and the complexity of a counsellor's schema is anticipated: that there will be a concomitant maturation of a counsellor's sexual abuse schema as a function of counselling experience and exposure to sexually abused clients. Research Questions. Several questions regarding the relationship between counsellor experience and sexual abuse schema were under investigation in this study, and are outlined as follows. The fundamental question, "is there a difference between the sexual abuse schemas of experienced and inexperienced counsellors?" was operationalized through a questionnaire designed to evaluate some components of schema complexity, specifically organization, breadth, and extremity. The question, "do the sexual abuse schemas of experienced and inexperienced counsellors differ on schematic complexity, with experienced counsellors demonstrating greater complexity (organization)?" was operationalized by recording the frequency of sexual abuse answers generated in response to ambiguous cues. Counsellor Schema 39 The question, "do experienced and inexperienced counsellors differ on schematic complexity, with experienced counsellors demonstrating greater complexity (breadth) of schema?" was operationalized by the number of adjectives chosen to describe sexually abused clients. The question ,"do experienced and inexperienced counsellors differ on schematic complexity, with inexperienced counsellors demonstrating less complexity (extremity)?" was explored by assessing whether inexperienced counsellors have a greater tendency to polarize their responses to descriptors of, and effects for, sexually abused clients. Additional exploratory questions were addressed in the research project. "Do differences exist between experienced and novice counsellors on anticipated treatment length?" was evaluated by a question specifically asking for a recommended duration of treatment regarding a hypothetical client. Finally, "do experienced and inexperienced counsellors differ on the amount of negative schematic inference, with inexperienced counsellors attributing more negative inference to sexually abused clients?" was evaluated by the frequencies of negative descriptors given for a sexually abused client. Counsellor Schema 40 Hypotheses Hypothesis One: Experienced counsellors will exhibit greater organization of schema: they will identify sexual abuse as a treatment issue more frequently than inexperienced counsellors in response to ambiguous cues in written client vignette (i.e., in response to the disguised version of the vignette). Hypothesis Two: Experienced counsellors will demonstrate greater breadth of schema. When presented with a list of adjectives, they will select a greater number of adjectives to describe a sexually abused client than will inexperienced counsellors. Experienced counsellors will rate more adjectives as "sometimes likely" or "frequently likely" than will inexperienced counsellors. Hypothesis Three: Inexperienced counsellors will exhibit greater extremity of schema than experienced counsellors. When presented with a list of adjectives, inexperienced counsellors will be more likely to make polarized ratings ("never" or "frequently") to describe a sexually abused client. Hypothesis Four: Inexperienced counsellors will demonstrate greater extremity of schema than experienced counsellors. Inexperienced counsellors will select polarized ratings for items describing potential effects of sexual abuse (scoring on a line graph between zero and 100 percent), while experienced counsellors will tend to score the items towards the mean. Exploratory questions. The exploratory question of whether differences exist between experience groups on anticipated treatment length will be addressed by participant responses to a question which directly asks respondents to recommend the duration of therapy for a hypothetical client. Inexperienced counsellors are expected to predict a shorter duration of treatment as a reflection of less understanding of the complexity of the area. Inexperienced counsellors, due to less complex sexual abuse schemas, are expected to more frequently select negative adjectives (negative schematic inference) than experienced counsellors when asked to describe a sexually abused woman. Counsellor Schema 41 Chapter Three Method Design The study design was a quasi-experimental field study (Gelso, 1978), combined with a descriptive survey design using a self-report questionnaire. The results will lend themselves to describing the characteristics of the different groups of counsellors and their schemas. The core research question examined the relationship between experience and schema, the two variables of interest in this study, which were operationalized by a questionnaire developed for this purpose (see "Questionnaire"). Variables-Independent Variables. The inclusion in the questionnaire of two versions of a case vignette provided the experimental manipulation in this study. Both vignettes contained the same sexual abuse issues and cues; one had the addition of explicit information about a sexual abuse history. The experimental independent variable was the sexual abuse cues; whether the participant received the explicit or ambiguous version of the vignette. The descriptive independent variable was counsellor experience and was derived from the counsellor-provided responses to the personal information section. Specifically, respondents were asked to indicate both the amount of general counselling experience and the amount of sexual abuse counselling experience they had. Analyses were performed on the basis of high and low general experience; high and low sexual abuse experience; and the combined experience of those counsellors with high amounts of both general and sexual abuse counselling experience, and low amounts of both general and sexual abuse counselling experience. Dependent Variable. Schema complexity is the dependent variable, specifically the dimensions of organization, breadth, and extremity. The construct of schema was operationalized in this study through the use of the questionnaire. The dimension of organization was operationalized by the identification of sexual abuse as a treatment issue in response to ambiguous cues; breadth was operationalized by the number of descriptors chosen for sexually abused clients; and extremity was operationalized by the Counsellor Schema 42 frequency of polarized responses to adjective descriptors ("Part B" of the Questionnaire), and sexual abuse effects of sexually abused clients ("Part C" of the Questionnaire). Sample In total, 180 questionnaires, equally divided between one of the independent variables, the explicit and ambiguous vignettes, were distributed to counselling centres in the Lower Mainland of British Columbia and to classes of graduate students in the Department of Counselling Psychology at the University of British Columbia. Ninety-seven were returned (53% response rate), and were fairly evenly represented between vignette versions: fifty ambiguous, non-sexual abuse identified vignettes (51 %) and 47 explicit sexual abuse vignettes (49%) were received. Considerably more female than male therapists responded: 84 women (86.8%) and 13 men (13.4%). Participants ranged in age from 24 to 64 years (X = 40.8, SD = 8.68). (See Table 1). Education. Table 1 provides the distribution of highest degree held by respondents. The most frequently reported academic degrees held by respondents were M.A. (N = 36, 37.1% of total population), and M.S.W. (N = 22, 22.7% of total population). In addition there were eight respondents with M.Ed, degrees (8.2%), primarily in counselling psychology. Sixteen respondents (16.5%) reported holding "other degrees", all but two of these were masters level degrees or were masters level graduate students. Eleven other respondents who were initially included in the B.A. group (N = 26) were, in fact, graduate students in a masters program in counselling psychology or social work. Some respondents have been represented in more than one category because of dual degrees. Eighty-seven (89.7%) of the counsellors were masters-level clinicians or in a masters-level graduate program. Work Location. Participants were asked to report on the types of settings in which they have worked or currently work. It was not possible to ascertain from their responses which, if any, was their present place of employment; some counsellors reported multiple locations depending on their work histories. Four choices were provided: agency, institution, private practice, and other. The latter category was seldom used. As Table 1 indicates, fully 81 (83.5%) of counsellors have had experience in a counselling agency setting; 19 (19.6%) in an institution; and 28 (28.9%) in private practice. Counsellor Schema 43 Experience. As Table 1 indicates, there was a considerable range of general counselling experience reported by participants, from no experience (presumably from the pool of graduate students) to 30 years (X = 8.4; SD = 7). The overall amount of reported sexual abuse counselling experience was no experience to 15 years (X = 3.5; SD = 3.9). The range of experience had a bimodal distribution with highly experienced counsellors in one group and very inexperienced counsellors (probably due to the inclusion of graduate students) comprising the other. Current Client Population. A question was posed to ascertain whether respondents were currently working with adult clients who had histories of sexual abuse. Seventy-seven of the counsellors (79.4%) indicated that yes, their present caseload included adult survivors; 20 (20.6%) did not. Theoretical Orientation. Nine different theoretical perspectives were presented and respondents were asked to indicate which orientation(s) constituted their particular counselling framework. They were not limited in the number they could choose. Some trends were noted (see Table 1): the most popular approach was client-centered, endorsed by 70.1 % of the respondents. Other frequently-noted theories were systemic (61.9%), feminist (57.7%), and cognitive (49.5%). All the remaining perspectives (adlerian, existential, experiential, gestalt, and solution-focussed) were utilized by forty percent of the sample or less. Counsellor Schema 44 Table 1 Demographic Characteristics of Sample Characteristic N Percent M SD Range Age Counselling Experience (years) General Sexual Abuse Gender Male Female Education B . A . a MA. M.Ed. M.S.W. Ph.D. Otherb Work Setting Agency Institution 97 97 40.8 8.68 24 - 64 13 84 26 36 8 22 3 16 19 13.4 86.8 26.8 37.1 8.2 23.7 3.1 16.5 83.5 19.6 8.4 3.5 7.0 3.9 0-30 0-15 Private Practice 28 28.9 Counsellor Schema 45 Characteristic N Percent M SD Range Currently Working with Sexual Abuse No 20 20.6 Yes 77 79.4 Theoretical Orientation Adlerian 3 3.1 Client-Centered 68 70.1 Cognitive 48 49.5 Existential 23 23.7 Experiential 39 40.2 Feminist 56 57.7 Gestalt 26 26.8 Solution-focussed 35 36.1 Systemic 60 61.9 aThis group includes eleven respondents who were graduate students in counselling or social work. DThis group includes 14 respondents who held masters degrees or were masters level graduate students. Personal Abuse History. Participants were asked to indicate whether they had themselves experienced childhood abuse or adult assault. AH but two counsellors (N = 95) responded to this question. Twenty-eight (29.5%) of the counsellors indicated they had not experienced any abuse; the remaining 67 (71.5%) indicated one or. more forms of childhood abuse or neglect, or were victims of an assault (physical or sexual) as an adult (see Table 2). The reported incidence of childhood sexual abuse (24.7%) was Counsellor Schema 46 consistent with respondents to a national survey of psychologists (Pope & Feldman-Summers, 1992), 26.9% of whom indicated they had been sexually abused in childhood; and with a more recent survey of clinicians (Nuttall & Jackson, 1994). The incidence among participants in the latter survey ranged from 10.9% for pediatricians to 21.3% for social workers. The middle groups were comprised of psychiatrists (15.6%) and psychologists (18.1%). The overall prevalence rate for the entire sample was 17%. However, when gender was controlled, 13% of males and 20% of females reported sexual abuse. The disparities among the studies may be due to limited empirical data available to date which precludes a broad statistical base; sampling (e.g., national versus local surveys; various professions); gender differences; and definitional ambiguities. Table 2 Counsellors' History of Childhood Abuse and Adult Assault Abuse History Frequency Childhood N % Emotional Abuse 38 39.2 Neglect 20 20.6 Physical Abuse 16 16.5 Sexual Abuse 24 24.7 Adult Physical Assault 10 10.3 Sexual Assault 7 7.2 None of the above 28 29.5 Note: Participants may have indicated more than one type of abuse experience. Counsellor Schema 47 Questionnaire As this study was relatively novel, no existing instruments had been located to evaluate counsellor schema with respect to sexual abuse. The "Attribution of Rape Blame Scale" (Ward, 1980) was considered but is outdated and specific to sexual assault, the dynamics of which are notably different than those of childhood sexual abuse. In addition, it does not address the issue of schema which is pertinent to this study. Similarly, the "Jackson Incest Blame Scale" (Jackson & Ferguson, 1983) was eliminated because, as an adaptation of Ward's scale, it merely substituted the word "incest" for "rape" in each of the twenty items. There were no additional revisions to reflect the distinct dynamics of child sexual abuse. Weaknesses in the scale were compounded by sampling biases (age, education and geographic variables) in the university undergraduate pool of respondents used as the normative sample. Psychometric data for the instrument are sparse (Adams & Betz, 1993) and it does not measure counsellor schema. For numerous reasons, therefore, the "Jackson Incest Blame Scale" is not an appropriate instrument for this study. A comprehensive literature search failed to unearth other potential means of evaluation. As discussed earlier, schema is not readily amenable to measurement by objective means because it is an unobservable cognitive process. Consequently, inferential strategies must be employed to operationalize this construct. To facilitate the collection of the required data, a questionnaire was developed. Considerations in the development of the questionnaire included: the necessity that it be research-based with respect to the issues of sexual abuse in order to accurately reflect the current knowledge base in the field; an adequate operationalization of schema; and provision of sufficient information about the counsellors' level of experience to enable classification into different categories. The reliance on empirical research in the field of sexual abuse provided content validity for the questionnaire. Construct validity for the schema variable was provided by employing inferential means of evaluation. Because of the introductory use of this qeustionnaire, there is no evidence of concurrent validity or predictive validity. A pilot test of the original questionnaire was completed for further validation and to increase its reliability. The researcher reviewed the schema literature to learn what assessment procedures have been used, some examples of which were cited earlier. The crucial consideration was to develop a means of obtaining Counsellor Schema 48 data from which counsellor schema could be inferred. As part of this process, the questionnaire was administered in a pilot study (see below), and the original version was revised in view of the feedback received. The questionnaire was modelled after ways other researchers have attempted to evaluate schema (e.g., Haverkamp, 1993). However, the author recognizes that even if expected results emerge, replication of the study and additional research will be required to fully assess the construct validity of the questionnaire and support this method as a successful evluation of schemas. Schematic complexity was the primary variable of interest and the questionnaire was constructed to provide several means to evaluate counsellor schema. The questionnaire is comprised of four sections which will be referred to as "Vignette", "Characteristics", "Sexual Abuse Effects" and "Demographics." Three dimensions of complexity were under investigation. Organization was evaluated by the Vignette, and operationalized by whether the respondents identified sexual abuse as a treatment issue from ambiguous clues. Breadth of schema was evaluated by the "Characteristics" section, operationalized by the amount of characteristics, in this case adjectives, comprising the schema. Extremity was evaluated by the "Characteristics" and "Sexual Abuse Effects" sections, and was operationalized by the manner in which participants scored the items: in a polarized or moderate manner. The data from the questionnaire provided means by which to infer the complexity of counsellors' schemas. Part A: Vignette. The purpose of the vignette was to investigate the complexity of counsellor schema by ascertaining whether respondents recognized subtle cues and subsequently identified sexual abuse as a salient treatment issue. There were two versions of the vignette (see Appendix A),one providing an explicit indication of sexual abuse, the other offering sexual abuse symptoms but no label. They were distinguished by the addition of a final paragraph in one version that was omitted in the other. This manipulation comprised the experimental condition in the study. The vignette was written in consultation with a sexual abuse therapist with twenty years experience in the field, to ensure that it accurately depicted an adult woman with a history of childhood sexual abuse. The case history described an adult woman with several presenting concerns (e.g., anger, problems with intimacy, anxiety), all of which were possible sexual abuse related issues. In addition, her present family Counsellor Schema 49 constellation and the provided family of origin information included dynamics often noted among sexually abused clients, and contained clues drawn from the literature that suggest a sexual abuse history. The versions were distinguished by a final paragraph which indicated the client had also disclosed memories of childhood sexual abuse; accordingly, one version was explicit, one was ambiguous. The ambiguous vignette was in some ways similar to what Courtois (1988) referred to as "disguised presentation": a client who does not reveal, at the intake level, her history of childhood sexual abuse. The vignettes will be referred to as Vignette NSA (the ambiguous, non-sexual abuse version) and Vignette SA (the explicit sexual abuse version). The instructions to participants indicated that the vignette was "intake" information for a hypothetical client about which the respondent would be asked questions. Three questions followed the information. The first two questions, which asked the counsellor to identify primary treatment issues and list the clues that led the counsellor to those hypotheses, were designed to ascertain whether experience was related to the likelihood of identifying sexual abuse from the clues, and whether the respondents attended to subtle cues. Responses to the first question were used to evaluate Hypothesis One. Responses of "childhood sexual abuse," "sexual abuse," or "incest" were deemed to be sexual abuse responses. Participant responses permitted the evaluation of the schematic dimension of organization. The third question asked the counsellor for an estimation of treatment duration. This was an exploratory question, intended to provide information about differences in clinical approach between counsellor experience groups. PartB: Characteristics. The purpose of this section (see Appendix B) was to assess schematic breadth. Respondents were asked to consider a "sexually abused female client," and to indicate how likely each adjective was to be characteristic of such a client. Because some adjectives are stereotypically ascribed to one gender over the other, this type of potential gender confound was reduced by stating "female client". The response choices of "never," "rarely," "sometimes," and "frequently" were provided. This section assessed Hypothesis Two, the schematic dimension of breadth, through the number of adjectives selected. It was also used to evaluate Hypothesis Three regarding extremity of schema and one of the exploratory questions regarding negative inference. Counsellor Schema 50 The "Characteristics" section consisted of a list of one hundred adjectives culled from the "Adjective Check List" (Gough & Heilbrun, 1983). The original pool of 300 words was considerably reduced, primarily to contain the size of the questionnaire. The ACL was selected because its range of descriptors is broad and generic; consequently the list would not be comprised of a disproportionate number of sexual abuse descriptors or psychological traits, a potential confound if the researcher had developed the item pool. Of the 100 items, 50 were positive attributes (e.g., confident, imaginative, kind), and 50 had negative connotations (e.g., moody, scared, unstable). In selecting the adjectives for inclusion, items were eliminated from the ACL for several reasons: those that were unlikely to be used to characterize anyone (e.g., zany, whiny); those that might be misunderstood (e.g., mannerly, hasty); and those which could be interpreted as positive or negative traits depending on the subjective value attributed to it (e.g., methodical: some people may view this as a strength, others as an undesirable quality). Embedded in the final list were 15 additional adjectives generated by a group of five experienced sexual abuse clinicians as characteristic of female adult survivors of childhood sexual abuse (e.g., hypervigilant, creative). Part C: Sexual Abuse Effects. This section was designed to investigate Hypothesis Three, the extremity variable of schema complexity, as opposed to the content, or knowledge of sexual abuse issues. It was comprised of 20 statements of potential effects of sexual abuse culled from the literature (Courtois, 1988; Briere & Runtz, 1993) and from an earlier study (Hibbard & Zollinger, 1990) that investigated professionals' knowledge of child sexual abuse. This reliance on the literature supported content validity for this section of the questionnaire. The items (see Appendix C) were related to the experience of sexual abuse (e.g., "the sexual abuse involved physical force"); the emotional effects (e.g., "cannot tolerate being alone"); the behavioural manifestations (e.g., "abuse alcohol or drugs"); and the resulting interpersonal dynamics (e.g., "attempt to anticipate the needs of others"). Respondents were asked to indicate, on a line graph of 0-100, the percentage of adult survivors of childhood sexual abuse who would manifest the effects listed. The zero pole represented "none;" 100% Counsellor Schema 51 indicated "all." The graph was marked in increments often percentage points which provided space to circle a number or make a new mark for a point in between. Part D: Demographics. The purpose of this section (see Appendix D) was to gather demographic data on the participants for comparison purposes and for assignment to experience level. Respondents were asked to provide information on age, gender, work setting (e.g., agency, institution, private practice), theoretical orientation, and highest level of education attained. They were asked how many years of experience they had in general counselling and sexual abuse counselling. Other items were related to such questions as type and amount of supervision received, specialized training programs, and additional degrees. A rating scale was provided with a list of counselling issues: counsellors were asked to rate their degree of knowledge and skills on a continuum from zero (novice) to ten (expert). The issues were pertinent to therapy for sexual abuse: adult counselling; gender and power issues in families; human sexuality; legal and ethical issues; normal child development; sexual abuse; substance abuse; and trauma. The intent of this section was to corroborate the other experience factors, and to ascertain the respondent's level of self-rated expertise. The question "do you currently work with adult clients who have a history of sexual abuse" was posed, followed by a question to ascertain the number and types of means by which the counsellor learned to work with clients who have been sexually abused. A list of fifteen possible sources (e.g., consultation, in-house training programs, reading relevant journals and books, workshops, etc.) was provided for counsellors to select, as well as a generic "other" category for any sources not included. The final question asked whether the respondent had a personal history of childhood neglect, sexual, physical, or emotional abuse, or adult physical or sexual assault. An explanatory note preceded the question to elucidate the rationale for its inclusion. The note read as follows: "This final question has been included because the therapist's realm of personal experience is a potent source of knowledge and insight and is relevant to the work you do," to assure respondents that the purpose of the question was relevant to the research. Counsellor Schema 52 Pilot Due to the inaugural nature of the study and the use of a researcher-designed questionnaire, it was decided to conduct a pilot study of the questionnaire prior to final data collection as a means of enhancing its utility and validity. To prevent contamination and depletion of the sample pool in the Lower Mainland geographical area, two Ontario-based counselling centres specializing in sexual abuse were identified by the director of Vancouver's centre, and contacted by a letter (see Appendix E) that requested the assistance of agency staff in the project. In addition, a class of graduate counselling psychology students at the University of British Columbia was solicited, as a comparison "inexperienced" group. The therapists who chose to assist fulfilled the dual roles of participant and consultant. They were asked to complete the questionnaire, with full disclosure of its purpose, goals and intent, as they would if participating in the study. They were also asked to provide comments and feedback on items as to how understandable they were, how representative of sexual abuse issues, and any concerns or questions that arose while completing the questionnaire. Eleven therapists returned their questionnaires. Five had received Vignette SA; six returned Vignette NSA. Five had two years experience with sexually abused clients or less; six were experienced therapists with five years or more of direct sexual abuse experience. Because of the full disclosure of the purpose of the questionnaire, the responses to identified treatment issues on the vignette section were not analyzed in the pilot. There was consensus that the questionnaire was too long and might deter counsellors from participating, particularly as time constraints are common in the counselling field. The therapists noted the elapsed time from start to finish was between one hour and one and one-half hours. The first part, the vignette was the section that demanded the most time. The decision was made, therefore, to shorten the amount of information provided in the vignette and to the reduce the number of questions in Part A. The choice of questions to eliminate was the result of feedback during the pilot, and a decision to keep questions directly related to schema complexity rather than other aspects of schema. Counsellor Schema 53 Part B, the adjective list, was retained with some adjustments to the instructions primarily as an acknowledgment of the difficulty pilot participants reported in attempting to generalize about a population of clients. The wording of the final version contained the proviso, "recognizing that generalizing is difficult...." Considerable revisions were undertaken with Part C, which was originally comprised of fifty items culled from the literature. The first version contained a 100 point Likert scale on which respondents were to rate the likelihood that the statement typified an adult woman in counselling for sexual abuse issues. Due to the feedback received, the directions were altered and the method of responding changed. Directions were revised to ask participants to indicate the percentage of the population of adult woman attending sexual abuse counselling who would present with the stated effects. In addition, the original fifty items were pared down to twenty. Items were eliminated if they were identified by pilot participants as too easily misinterpreted, or were too vague or confusing. Items that most noticeably discriminated among the pilot counsellors with differing amounts of experience were retained for the final version. Procedure Potential participants were canvassed from numerous counselling centres throughout the Lower Mainland of British Columbia. The Directory of Services for the Lower Mainland: The Red Book, the directory of community resources in Greater Vancouver, was the source from which agencies were identified. Centres that provide sexual abuse counselling, family counselling, and alcohol and drug treatment were selected because of the likelihood of reaching counsellors with both specialized and generic experience. Several graduate classes in counselling psychology at the University of British Columbia were canvassed, with the expectation of reaching less experienced counsellors.' The first contact with the agencies was made by letter to the counselling supervisor (see Appendix F), that explained the nature of the research and the request for participation from agency staff. A telephone call was placed several weeks later to each supervisor to ascertain whether the agency or program had agreed to take part. In some cases there was no agency support for the study, and consequently no further contact was made. If there was interest, the required number of packages containing questionnaires equally Counsellor Schema 54 divided between Vignette SA and Vignette NSA were forwarded to the program supervisor for distribution to the team. Included in each counsellor's package was an introductory letter; a notice of informed consent complete with a return form if a summary of results was desired; information about the incentive and a ballot form; the questionnaire and return envelope; and a sealed debriefing letter (see Appendix G). The letter described the focus of the research as "client conceptualization" with the purpose of answering questions about the effect of experience and training. Counsellors were reassured that the "questionnaire is not a test of . . . knowledge, nor will it be assessing... skills," as there was the possibility that such an interpretation might be made, thereby affecting the counsellors' willingness to respond and the manner in which the questionnaire would be completed. The anticipated time commitment of thirty minutes was provided, and the incentive (a draw prize for a gift certificate at a Vancouver restaurant) was explained. Anonymity and confidentiality were assured. Directions were listed, and telephone contact numbers for the student researcher and faculty advisor provided. The informed consent form reiterated the guarantee of anonymity and confidentiality, and outlined the method by which prize draw ballots could be returned without revealing the identity of the respondent. Ballots were to be sealed in a second envelope and separated from the questionnaire immediately upon receipt by the researcher, without opening the sealed ballot. Telephone contact numbers were again provided. The lower portion of the consent form could be returned by the respondent (under separate cover) if a summary of findings was desired. The final document in the package was a debriefing letter, primarily intended for recipients of the disguised version of the vignette. The letter identified the sexual abuse focus of the research and provided an explanation for the initial disguise. It was stapled closed, with the words "read last" clearly written on the front to protect the integrity of the experimental condition. Approximately two week after the initial mailing, follow-up letters (see Appendix H) were sent to each program or agency, again via the supervisor. Sufficient letters were sent to reach all the counsellors who received the original package. Counsellors were thanked for their participation and asked to disregard the Counsellor Schema 55 reminder letter if they had already completed their questionnaire. If they had not yet done so, they were reminded about the research project and that voluntary participation was again being requested. To access a potentially less experienced pool of counsellors, three graduate counselling psychology classes at the University of British Columbia were attended personally by the student researcher after consulting with the faculty members responsible for each course. An announcement was made about the nature of the research disguising the sexual abuse focus, and the time commitment and incentive were explained. A sufficient number of questionnaires for all class members were distributed, for the purpose of ensuring that non-participating students could not be distinguished from those who chose to respond. In place of a follow up letter, a reminder announcement was made in each class several weeks following the initial presentation. Analysis All responses were coded to facilitate the statistical analyses. Omitted responses were coded as "missing data". More complete descriptions of analyses conducted for each hypothesis are reported in relevant sections of the "Results". A hierarchical log linear analysis was employed initially to ascertain whether the experimental independent variable (the vignette condition) had an effect on the independent variable, experience, and its association with the dependent variable schema. Hierarchical log linear analysis is a means to assess the relationship among "nested" variables or characteristics (Arnold, 1992). Chi square tests were used for between-group comparisons of frequencies of responses to the "Vignette" section (Hypothesis One: organization). T-tests were used to compare the groups on the mean number of adjectives selected in the "Characteristics" section (Hypothesis Two: breadth). Hypothesis Three (extremity) was evaluated with t-tests to compare counsellor groups on the mean scores of polarized ratings on the adjectives list. T-tests were also employed to evaluate the exploratory questions: to assess between-group differences in mean length of recommended treatment, and to compare counsellor experience groups on mean scores of negative and positive adjectives. Counsellor Schema 56 To determine which group of counsellors tended to score the items from the "Sexual Abuse Effects" section towards the poles of the line graph (Hypothesis Four: extremity), the data were converted to absolute values, ranked, and compared using a Mann-Whitney test, a nonparametric procedure that does not require a normal distribution of scores. The purpose of this initial inquiry was to determine whether differences in schematic properties regarding sexually abused clients existed among counsellors of varying experience levels. Results of the analyses for all hypotheses and exploratory questions are presented in Chapter Four. Counsellor Schema 57 Chapter Four Results Homogeneity of Sample As reported in the Method section, the sample was evaluated for educational background, work setting, and theoretical orientation (see Table 1, p. 43). An analysis of the reported academic degrees indicated that 87 (89.7%) of the participating counsellors were masters-level clinicians, primarily with counselling psychology, education, or social work degrees, or in a graduate program in counselling psychology or social work, indicating homogeneity for this sample on the variable of graduate education. The sample appears to be fairly homogeneous with regard to work setting, with fully 81 (83.5%) of the counsellors having previous or current experience in a counselling agency. The two other categories were institution (N = 19, 19.6%) and private practice (N = 28, 28.9%). Finally, respondents were asked about their adherence to a particular theoretical perspective(s). A client-centered approach was the most popular orientation, endorsed by 70.1% of the respondents. A systemic framework was the next most frequently selected theory (61.9%), followed by feminist practice (57.7%). These results provided some degree of confidence that the responding counsellors, as a group, had a comparable level of academic training, shared similar experiences in work location, and approached their clients from similar theoretical orientations. Preliminary Analyses Vignette Condition. The first hypothesis of the research was based on the possibility of an existing difference in responses to the experimental condition of the two vignette versions: Vignette SA, the sexual abuse identified vignette, and Vignette NSA, the non-sexual abuse vignette. The two versions were identical with the exception that Vignette SA included a final statement noting that the client disclosed a history of child sexual abuse by her stepfather. Equal numbers of both versions were distributed and were fairly evenly represented in the returns: 50 Vignettes NSA (51 % of the sample) and 47 Vignettes SA (49% of the sample) were returned. Counsellor Schema 58 The first level of analysis was conducted to ascertain whether there was a vignette effect and whether the experimental condition would be maintained. Counsellors were deemed to have identified sexual abuse as a treatment issue for the vignette if they included "sexual abuse," "childhood sexual abuse," or "incest" in their list of possible treatment issues. A range of answers were provided such as "sexual problems," "marital difficulties," "intimacy," "post-partum depression," and "trauma" but only responses that clearly specified childhood sexual abuse in one of the three ways cited above were considered to meet the criteria of a "sexual abuse" response. As can be seen in Table 3, a chi square test to calculate the frequency of "sexual abuse" responses to the question "what have you identified as the client's primary treatment issue" yielded a statistically significant difference (JC 2 = 24.05, df = 1, < .001) between Vignette SA and Vignette NSA. Only 6% of the counsellors receiving the sexual abuse vignette did not identify sexual abuse, compared with 52% of the counsellors receiving the disguised version. This strong difference necessitated that the experimental condition be maintained in further analyses. Results indicated that counsellors presented with written "intake information" about a client were more likely to identify sexual abuse as a treatment issue when it was explicitly stated that she had a history of abuse. Counsellor Schema 59 Table 3 Association of Vignette Version and Identification of Sexual Abuse (Total Sample) Sexual Abuse Identified Row Vignette No Yes Total Percentage Total Sample* Vignette NSA 26 24 50 51.5 Vignette SA 3 44 47 48.5 Column Total 29 68 97 100 Percentage 29.9 70.1 100 100 Female Respondents Only** Vignette NSA 20 21 41 48.8 Vignette SA 3 40 43 51.2 Column Total 23 61 84 100 Percentage 27.4 72.6 100 100 * J C 2 = 24.05, df= 1,rr<.001 **x2= 18.45, df = .001 Given the strong vignette effect, and the equal distribution of participants who did (N = 26, 52%) or did not (N = 24, 48%) identify sexual abuse in response to Vignette NSA, the decision was made to consider this group (N = 50) alone for some of the planned analyses. The distribution of counsellors who did (N = 44, 94%) and did not (N = 3, 6%) identify sexual abuse as a treatment issue for Vignette SA, however, was so skewed, that subsequent combined group analyses would likely be confounded. Counsellor Schema 60 Gender. Of the 97 respondents, 84 (86.8%) were women, and 13 (13.4%) men. To determine whether a gender difference existed, a chi square analysis of gender by identification of sexual abuse was conducted. It did not yield a significant difference (see Table 4). As can be seen from Table 3, male respondents were relatively evenly distributed among the cells. There was a slight tendency for males to fail to identify sexual abuse in Vignette NSA but the effect was small enough that it would not result in a gender confound. As there was no substantive gender effect, the male respondents were retained in the sample for subsequent analyses. Table 4 Association of Gender and Identification of Sexual Abuse Sexual Abuse Identified Row Gender No Yes Total Percentage Male (N= 13) 6 7 13 13.4 Female (N = 84) 23 61 84 86.6 Column Total 29 68 97 100 Percentage 29.9 70.1 100 100 xl = 1.89, df = 1, not statistically significant. Experience Variable The decisions made with regard to categorizing counsellors on the variable of experience are described below. Respondents indicated their years of experience for both "general" and "sexual abuse" counselling. There was a considerable range of general counselling experience reported by participants: from no experience (presumably from the pool of graduate students) to 30 years (X = 8.4; SD = 7). Groups were Counsellor Schema 61 formed by using cut points after four years of experience and before nine years, eliminating the middle group to ensure adequate sample size and an ample range to make the groups distinct, an acceptable procedure for establishing comparison groups (Holloway & Wolleat, 1980). Thus, "low general counselling experience" included all counsellors reporting zero to four years of general counselling experience (N = 33); "high general counselling experience" included all counsellors reporting ten years experience or more (N = 36). The overall amount of reported sexual abuse counselling experience was none to 15 years (X = 3.5; SD = 3.9). In this specialized field, it was decided that four years of experience would adequately demarcate an "experienced" counsellor of sexually abused clients from those who are inexperienced. Consequently, "low sexual abuse counselling experience" included therapists with one year or less reported sexual abuse counselling experience (N = 37); the groups subsumed by the title "high sexual abuse counselling experience" included therapists with four years or more of sexual abuse counselling experience (N = 40). The combined experience groups were comprised of the general and sexual abuse counselling experience groups. Those counsellors included in the low (N = 26) group had low general counselling experience combined with low sexual abuse counselling experience; the high (N = 23) group consisted of high general counselling experience and high sexual abuse counselling experience.. The decision was made to separate the respondents into two experience groups, high and low, eliminating the middle group from the analyses, with the expectation that greater variation in counselling experience would increase the likelihood of finding of an experience effect. The "combined" group initially had four levels (high general, high sexual abuse; high general, low sexual abuse; low general, high sexual abuse; and low general, low sexual abuse). However, as the cell sizes of the two middle groups was extremely small (N=1; N=9), only the high and low groups were retained for analysis. Initial analyses were conducted with a combination of general and sexual abuse counselling experience, and were followed with analyses by general experience only, and sexual abuse experience only, to determine whether emerging differences could be associated with a particular type of counselling experience. Counsellor Schema 62 Age and Experience. The age range for the sample was 24-64 years old (X = 40.8, SD = 8.68). T-tests for experience and age were significant: counsellors who reported more years of experience were significantly older than their less experienced counterparts. This difference held for general experience (t = -4.14, df = 68, p < .001), and combined general and sexual abuse experience (t = -3.87, df = 47, < .001) (see Table 5). This type of finding makes intuitive sense; it would be expected that counsellors who have been in the field longer would be concomitantly older. The lack of a statistically significant interaction of age and sexual abuse experience may be accounted for by the relatively new specialization in the counselling field which has not necessarily provided sufficient time for counsellors to accrue many years of experience. Table 5 Association of Age and Amount of Experience Age Type of Counselling Experience X SD t-value General Low (N = 33) 37.39 8.65 High (N = 37) 45.43 7.60 -4.14* Sexual Abuse Low(N = 40) 39.13 9.08 High (N = 36) 42.25 8.44 -1.55 Combined Low(N = 26) 36.5 8.17 High (N = 23) 45.39 7.86 -3.87* *{L < .001 Counsellor Schema 63 Hypothesis One The first hypothesis under investigation was that inexperienced counsellors would generate fewer sexual abuse responses to Question One of the vignette: "what have you identified as the client's primary treatment issues (i.e., what will be the focus of therapy)?" Respondents were deemed to have provided a sexual abuse answer if "childhood sexual abuse" or "incest" were noted by the participant. Analysis. Table 6 provides the results of a hierarchical log linear analysis that was used to ascertain the existence of an experience effect: whether the "high combined experience" group differed significantly from the "low combined experience group" on frequency of sexual abuse responses. A vignette by experience by frequencies of sexual abuse responses was conducted. There was a significant main effect (x2 = 24.49, df = 7, p_, < .001); however, a backward elimination of variables indicated that the strong difference could be accounted for by the vignette variable only; experience was not associated with the likelihood of generating sexual abuse responses. Counsellor Schema 64 Table 6 Hierarchical Log Linear Analysis of Vignette Version by Combined Experience by Identification of Sexual Abuse Factor Observed Count Expected Count Chi Square (df = 4) Vignette NSA (N = 28) S/A Identified - No Low Experience 9 8 High Experience 7 8 S/A Identified - Yes Low Experience 6 6 High Experience 6 6 Vignette SA (N = 21) S/A Identified - No Low Experience 0 .5 High Experience 1 .5 S/A Identified - Yes Low Experience 11 10 High Experience 9 10 1.84 Counsellor Schema 65 Seemingly, the content difference between the vignettes was a strong predictor: regardless of experience level, counsellors were significantly more likely to identify sexual abuse as a treatment issue when the sexual abuse was explicitly articulated in the vignette. Chi square tests on the frequencies of sexual abuse responses were calculated for the three categories of experience groups (general, sexual abuse, and combined) after factoring out the vignette condition. No significant results were found (see Table 7). Thus, the first hypothesis, that there would be a positive relationship between experience level and the identification of sexual abuse, was not borne out in the results. Table 7 Association of Type of Counselling Experience and Identification of Sexual Abuse Sexual Abuse Identified Type of Experience No Yes Total Percentage Chi Square (df=l) General Low(N = 26) 9 17 26 53.1 High (N = 23) 8 15 23 46.9 .000 Sexual Abuse Low(N = 40) 13 27 40 52.6 High (N = 36) 10 26 36 47.4 .200 Combined Low(N= 15) 9 6 15 53.6 High (N= 13) 7 6 13 46.4 .108 Counsellor Schema 66 Hypothesis Two The second hypothesis was based on the "Characteristics" section, which asked participants to "consider a sexually abused female client and circle the letter which represents whether, in your opinion, the characteristic is never likely, rarely likely, sometimes likely, or frequently likely to be shown by a sexually abused client." It was expected that, as an indicator of greater schematic complexity (breadth), experienced counsellors would select an overall greater number of traits. It was hypothesized that the experienced group would have higher means of adjectives rated as "sometimes" or "frequently," both of which indicated the adjective was selected as representative of sexually abused women. Given the vignette difference, analyses were conducted both for combined Vignettes SA and NSA, and Vignette NSA alone. Analysis: Combined Vignettes SA and NSA. The number of adjectives given a rating of "sometimes likely" or "frequently likely" were tabulated for each counsellor. T-tests were calculated to compare the mean number of selected adjectives between the low and high combined experience groups. The t-tests failed to yield significant differences (see Table 8), a finding that held when further comparisons were conducted using the general counselling and sexual abuse counselling experience categories. Counsellor Schema 67 Table 8 T-test Results Comparing the Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Levels of General. Sexual Abuse, and Combined Experience (Vignettes SA and NSA) Number of Adjectives Experience X SD t df General Low(N = 33) 82.49 8.18 High (N = 37) 82.27 9.17 .10 68 Sexual Abuse Low(N = 40) 81.93 9.39 High (N = 36) 85.03 6.18 -1.68 74 Combined Low (N = 26) 83.92 6.92 High (N = 23) 84.52 6.5 -.31 47 Analysis: Vignette NSA. Because a strong effect had been found in the preliminary analyses with the vignette by identification of sexual abuse, a decision was made to examine these means within the sample of counsellors who received the non-sexual abuse vignette only (Vignette NSA). A t-test was calculated on the differences between the mean number of adjectives checked by the low and high combined experience groups of counsellors. However, no significant results were found. The results of additional t-tests using the general and sexual abuse experience categories were also non-significant (see Table 9). Counsellor Schema 68 Table 9 T-test Results Comparing the Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Levels of General. Sexual Abuse, and Combined Experience (Vignette NSA) Adjectives Experience X SD t df General Low (N = 20) 82.60 9.28 High(N_= 19) 81.47 10.29 .36 37 Sexual Abuse Low(N = 21) 82.67 10.67 High (N= 18) 84.11 6.07 -.51 37 Combined Low(N=15) 85.27 6.83 High (N = 13) 83.46 6.48 .71 26 Summary. As evaluated by the number of traits attributed to sexually abused clients, an association between experience and breadth of counsellor schema was not found. This result did not support the hypothesis that experienced counsellors have a greater breadth of sexual abuse schema than do less experienced counsellors. Counsellor Schema 69 Hypothesis Three Hypothesis three addressed the schematic property of complexity by evaluating the extremity of counsellors' responses to the adjectives that potentially described a sexually abused client. It was hypothesized that the mean number of polarized responses ("never" and "frequently") would be greater for inexperienced counsellors than for experienced counsellors. Analyses were planned for the sample of combined vignettes (SA and NSA) and for Vignette NSA alone. Analysis: Combined Vignettes SA and NSA. Adjective ratings were recoded prior to conducting the analysis. Adjectives rated as "never" and "frequently" (the poles) were assigned a value of two; ratings of "rarely" and "sometimes" were assigned a one. Individual participants' means were calculated and grand means for each experience group were computed. The grand means were then compared by use of a t-test. A higher mean (i.e., closer to two) would indicate a greater tendency to rate the adjectives with polarized scores. Table 10 indicates that no significant differences were found on the basis of experience using the general, sexual abuse, or combined experience groups: the tests failed to discriminate between low and high experience for the mean scores on polarized adjectives. Counsellor Schema 70 Table 10 T-test Results of Mean Scores for Polarized Ratings of Adjectives by Experience Level (Vignettes SA and NSA) Polarity of Responses Experience Level X SD t df p.. Combined Experience Low(N = 23) 1.72 .156 High (N=18) 1.67 .127 1.11 39 .27 General Experience Low (N = 29) 1.69 .171 High (N = 31) 1.68 .140 .10 58 .92 Sexual Abuse Experience Low (N = 36) 1.72 .156 High (N = 29) 1.69 .127 .53 63 .59 Vignette NSA. The same method of analysis was used to investigate the hypothesized relationship between experience and extremity of responses with the group of counsellors who received Vignette NSA only. Again, no significant differences for the mean scores for polarized adjectives was found for any of the counselling experience categories (see Table 11). Counsellor Schema 71 Table 11 T-test Results of Mean Scores for Polarized Ratings of Adjectives by Experience Level (Vignette NSA) Polarity of Responses Experience Level X SD t df p^  Combined Experience Low(N=14) 1.70 .173 High (N=12) 1.66 .126 .77 24 .45 General Experience Low (N = 18) 1.68 .184 High (N= 18) 1.66 .157 .26 34 .79 Sexual Abuse Experience Low(N = 20) 1.69 .185 High (N= 16) 1.68 .124 .10 34 .92 Summary. The hypothesis that counsellors of varying experience levels would differ on schematic complexity (extremity) was not supported by these findings: there were no differences between experienced and inexperienced counsellors regarding the extremity of their ratings of adjectives potentially describing a sexually abused client. Counsellor Schema 72 Hypothesis Four The final formal hypothesis also assessed the schematic property of complexity, by comparing the groups on the extremity of their responses to the twenty items in Part C of the questionnaire, which were statements of potential effects of sexual abuse (psychological, behavioural, and interpersonal). Participants were instructed to indicate, on a percentage line graph marked in increments of ten points ranging from zero to 100, the percentage of adult survivors they believed would exhibit the listed effects. Similar to Hypothesis Three, it was hypothesized that inexperienced counsellors would be more likely to respond with extreme answers, while the experienced group would tend to rate their responses towards the mean. Analysis: Combined Vignettes SA and NSA. The statistical test chosen for this analysis was the Mann-Whitney test. The absolute values of each item were calculated, resulting in a range of scores from zero to 50. The scores were then tabulated to produce a single score for each participant, with counsellors who tended to rate on the extreme ends of the scale having higher scores, whereas counsellors who tended to rate more moderately having in lower scores. A Mann-Whitney test was then conducted on the combined experience groups (low experience, N = 22; high experience, N = 21) but failed to produce a significant difference between the experience groups (see Table 12). Analyses of general counselling experience and sexual abuse counselling experience also failed to yield significant between group differences (see Table 12). Counsellor Schema 73 Table 12 Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Combined, General, and Sexual Abuse Counselling Experience Level (Vignettes SA and NSA) Part C Ratings Experience Mean Rank Z p_. Combined Experience Low(N = 22) 22.11 High(N_=21) 21.88 -.06 .95 General Experience Low (N = 28) 32.30 High (N = 35) 31.76 -.118 .91 Sexual Abuse Experience Low (N = 35) 32.96 High (N = 34) 37.10 -.859 .39 The results indicated that experience was not associated with the tendency for counsellors to polarize their responses. On the basis of this particular means of evaluation it could not be concluded that experience was associated with extremity of responses. Therefore, the hypothesis that inexperienced counsellors would rate the sexual abuse effects items with more polarized scores in comparison with experienced counsellors was not supported. Analysis: Vignette NSA. The same analysis, a Mann-Whitney test, was conducted for recipients of Vignette NSA only. Counsellors were compared on combined, general, and sexual abuse counselling experience. Table 13 presents the results, which were non-significant. Counsellor Schema 74 Table 13 Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Combined. General, and Sexual Abuse Counselling Experience (Vignette NSA) Experience Part C Ratings Mean Rank Z IL Combined Experience Low(N= 11) High(N= 13) General Experience Low(N= 15) High(N= 19) Sexual Abuse Experience Low(N= 17) High (N= 18) 14.00 11.23 19.4 16.0 19.6 16.5 .957 .989 .34 .32 .908 .36 Counsellor Schema 75 Exploratory Analyses Two exploratory analyses were planned in addition to those directly related to the hypotheses under investigation. The first examined the relationship between counsellor experience level and the counsellors' recommended duration of therapy for a hypothetical client based on information contained in the vignettes. It was expected that results might provide indirect support for the hypothesis that inexperienced counsellors had less complex sexual abuse schemas. The second avenue of exploration assessed for differences between counsellor groups on the valence of characteristics attributed to a "typical female adult client with a history of sexual abuse." It was anticipated that inexperienced counsellors would choose more negative adjectives than experienced counsellors, ascribing to them a skewed set of attributes, an indication of a less complex schema. Duration The vignette section of the questionnaire included the following question: "what is your estimate of the anticipated duration of therapy?" This was a planned exploratory analysis to determine whether experience was associated with expectations regarding treatment length. Consistent with the hypotheses that the schemas of experienced counsellors would be more complex, it was anticipated that the greater complexity would be reflected in their predictions for length of therapy. It was expected, then, that experienced counsellors would predict a longer course of treatment as a reflection of their understanding of the intricacies and complexity of sexual abuse treatment. Comparison by Vignette. The first analysis assessed whether differences existed between the mean number of months of therapy on the basis of vignette. The mean of counsellors who received Vignette NSA (N = 46; X = 12.3; SD = 9.46) was lower than the mean for the group of counsellors who received Vignette SA(N = 41; X = 16.2; SD = 9.52). Table 14 shows that, although there was a greater tendency for counsellors reading the sexual abuse vignette to anticipate a need for longer treatment, the result approached but did not reach statistical significance (t = -1.88, df - 85, p_, < .06). Comparison by Combined Experience Level. As can be seen from Tables 14 and 15, t-test analyses conducted for the combined experience groups on mean length of treatment did not yield significant results for either the combined vignettes sample or for only the counsellors who received Vignette NSA. Counsellor Schema 76 Comparison of Sexual Abuse and Non Sexual Abuse Counsellors. A subsequent analysis was conducted for the counsellor groups distinguished by whether or not they currently worked with sexually abused clients. T-tests were conducted for the entire sample (combined Vignettes SA and NSA) and again for the group who received Vignette NSA only. No significant differences emerged on mean duration of recommended treatment for any of these comparisons (See Tables 14 and 15). Comparison by Personal Childhood Abuse History. T-tests were used to compare the anticipated length of therapy recommended by those counsellors who had their own history of childhood sexual abuse and those who did not. As can be seen in Table 14, when the entire sample (Vignettes SA and NSA) was used, a statistically significant difference was evident (t = -2.32, df = 83. rx < 05), indicating that the group of counsellors who had themselves experienced childhood sexual abuse anticipated that a significantly longer duration of therapy was required for the client. Comparisons with Vignette NSA. A t-test on the Vignette NSA sample did not, however, bear the same results: there was no statistically significant difference for anticipated duration of treatment between counsellors with and without their own histories of sexual abuse. The analyses comparing current sexual abuse counsellors with non sexual abuse counsellors approached statistical significance, with current sexual abuse counsellors predicting considerably longer durations of treatment. Counsellor Schema 77 Table 14 Association of Predictor Variables and Anticipated Duration of Treatment (Vignettes SA and NSA) Duration (months) Variable X SD t Vignette Version Non-Sexual Abuse (N = 46) Sexual Abuse (N = 41) Combined Counselling Experience Low (N = 26) High(N= 18) Current Sexual Abuse Counsellors No(N= 19) Yes (N = 63) History of Childhood Sexual Abuse No(N = 61) Yes (N = 24) 12.35 9.46 16.17 9.52 -1.88" 12.54 9.72 17.50 9.99 -1.65 11.21 9.39 14.97 9.59 •1.52 12.69 8.99 17.92 10.27 -2.32 ** *approaching statistical significance, p_. = .06 **p_, < .05 Counsellor Schema 78 Table 15 Association of Predictor Variables and Anticipated Duration of Treatment (Vignette NSA) Duration (months) Variable X SD t Combined Counselling Experience Low (N= 15) 11.87 10.39 High(N= 10) 14.70 8.54 -.71 Current Sexual Abuse Counsellors No(N= 13) 8.54 7.57 Yes (N = 33) 13.85 9.80 -1.75* History of Childhood Sexual Abuse No (N = 32) 10.97 8.89 Yes(N= 12) 15.83 10.21 -1.55 * approaching statistical significance, p_. = .08. Summary of Findings. An analysis was conducted to ascertain whether the vignette version was related to recommended duration of treatment for the hypothetical client. Further comparisons were conducted on the basis of several counsellor variables (experience, current client population, and personal history of childhood sexual abuse). There was a tendency, although non-significant, for experience to be associated with recommended duration of treatment, as indicated by the considerable differences in means (see Table 15). Counsellors with their own history of child sexual abuse recommended a significantly longer duration of treatment than counsellors who had not experienced sexual abuse. The results of two other comparisons, between vignette versions, and between current sexual abuse counsellors and non sexual abuse counsellors who received Vignette NSA, approached statistical significance. Counsellor Schema 79 Characteristics: Additional Analyses Statistical tests for the adjective list (Part B of the questionnaire) did not produce significant findings with regard to the number of characteristics selected by respondents to describe a typical female adult survivor of childhood sexual abuse. Although there were no overall between-group differences when the ratings for all adjectives were collapsed, further discrete comparisons were conducted. Negative and Positive Adjectives An additional exploratory question was whether counsellors of different experience levels varied in the valence of characteristics they attributed to clients with sexual abuse histories. In other words, inexperienced counsellors, consistent with the hypothesis of less complex schemas, were expected to attribute more negative adjectives to the clients than clinicians with more experience. Experienced clinicians were expected to demonstrate a more balanced (negative and positive) schema. Combined Vignettes SA and NSA. Equal numbers of positively and negatively valued adjectives had been included in Part B of the questionnaire. The mean scores for all negative adjectives were calculated for each participant, and again for the category of positive adjectives. Grand means were then calculated for the low and high combined experience groups, and compared with a t-test. Table 16 provides the results of the combined vignettes (SA and NSA) sample. There was a difference on the mean scores for negative adjectives between experience groups that approached statistical significance (p_, = .07). The high experience counsellors tended to score more highly on the negative adjectives, suggesting that they chose more negative characteristics to describe a sexually abused client than did less experienced counsellors. This result was not in the expected direction. The same calculations were completed on the basis of current sexual abuse counsellors and those not working with this population. The comparison between counsellors currently working with sexually abused clients and those who are not produced a statistically significant differences (p_. < .02) for negative adjectives. The current sexual abuse counsellors scored higher than their counterparts on the negative adjectives, suggesting that current sexual abuse counsellors chose more negative characteristics to describe a sexually abused client. Counsellor Schema 80 No differences emerged when low and high counsellor groups, and current sexual abuse counsellors and non sexual abuse counsellors, were compared on positive adjectives. Table 16 T-test Results of Scores for Negative and Positive Adjectives Selected by Counsellors (Vignettes SA and NSA) Counsellor Groups X Adjective Scores SD t df p_. Negative Adjectives Combined Experience Low (N = 26) High (N = 21) Current Sexual Abuse Counsellors No (N = 20) Yes (N = 72) Positive Adjectives Combined Experience Low (N = 25) High (N = 20) Current Sexual Abuse Counsellors No(N= 18) Yes (N = 72) 3.25 3.36 3.22 3.35 3.11 3.08 3.03 3.06 .229 .159 -1.88 45 .07* .206 .211 -2.40 90 .02** .258 .188 .45 43 .66 .207 .287 -.35 88 .73 * approaching statistical significance **p_, < .05 Counsellor Schema 81 Vignette NSA The same tests were calculated for the sample of Vignette NSA only, for the counsellor comparison groups of combined experience and current client focus. Table 17 indicates that the current sexual abuse counsellors rated negative adjectives significantly higher than did counsellors not working with this population. This result was not in the expected direction, indicating that the counsellors currently working with sexually abused clients attributed more negative characteristics to sexually abused clients than did the counsellors not working with this population. There were no significant differences between combined experience groups on the negative adjectives. No between-group differences emerged for the low and high combined experience counsellors, or for the current sexual abuse counsellors and those not working with sexually abused clients, on the mean scores for positive adjectives. Counsellor Schema 82 Table 17 T-test Results of Scores for Negative and Positive Adjectives Selected by Counsellors (Vignette NSA) Adjective Scores Counsellor Groups X SD t df p_. Negative Adjectives Combined Experience Low(N= 15) High(N= 13) Current Sexual Abuse Counsellors No(N= 14) Yes (N = 35) Positive Adjectives Combined Experience Low (N = 12) High (N = 35) Current Sexual Abuse Counsellors No(N= 14) Yes(N= 13) * £ < .05 Adjectives: Item-by-Item Analysis. No significant between-group differences were found from the comparisons of adjectives rated as "sometimes likely" or "frequently likely" when the data were collapsed over the 100 adjectives. To investigate more discrete differences, chi square tests were conducted between the low and high combined experience counsellors on each individual adjective. Eleven of the 100 3.29 .189 3.36 .166 -1.04 26 .31 3.22 .168 3.37 .236 -2.11 47 .04* 3.04 .184 3.03 .347 .03 45 .97 3.08 .222 3.07 .207 .14 25 .89 Counsellor Schema 83 adjectives were rated differently by low and high combined experience groups either to a statistically significant degree (N = 9) or to a level that approached significance (N = 2). Table 18 presents the list of adjectives for which differences were found. Three of the adjectives were positive attributes (calm, fairminded, giving); the remainder were negative: anxious, ashamed, bitter, fearful, forgetful, hypervigilant, sad, and suspicious. No patterns were evident from the results, although eight of the nine adjectives exhibiting statistically significant differences were negative characteristics. Two of the three positive adjectives did not reach statistical significance; the only positive attribute that did show a clear difference was "fairminded", with the low experience group rating it higher (as more likely to be characteristic of a sexually abused client) than did the high experience group. The other finding was that the exhibited difference was primarily a greater tendency for the high experience group in comparison with the low experience counsellors to rate the adjectives as "frequently likely" to be characteristic of a client who had been sexually abused. In other words, the more experienced counsellors selected more negative adjectives, and rated them in a more polarized way. These findings were contrary to the hypotheses that more experienced counsellors would be less likely to choose negative adjectives, and to be more moderate, rather than extreme, in their ratings of the adjectives. Counsellor Schema 84 Table 18 Individual Adjectives Receiving Significantly Different Ratings From Low and High Experienced Counsellors (Vignettes SA and NSA) Counsellors Low Experience High Experience Adjective Rating3 Adjective Rating Adjective 2 3 4 2 3 4 Chi Square Significance Anxious 0 6 20 0 1 22 3.49 .06 Ashamed 0 6 20 0 0 23 6.05 .04 Bitter 0 12 14 0 18 5 5.29 .02 Calm 7 18 1 13 19 0 4.92 .08 Fairminded 2 16 7 3 20 0 8.49 .04 Fearful 0 10 16 0 2 21 5.82 .02 Forgetful 2 21 3 0 14 9 6.24 .04 Giving 0 22 4 0 14 9 3.5 .06 Hypervigilant 0 10 16 0 3 20 4.05 .04 Sad 0 10 16 0 3 20 4.05 .04 Suspicious 0 19 7 0 9 14 5.74 .02 a Numbers correspond to ratings as follows: 2 = rarely likely; 3 = sometimes likely; 4 = frequently likely Counsellor Schema 85 Post Hoc Analyses Additional analyses were conducted to investigate the following: (a) the association of current sexual abuse counselling with schema; (b) the relationship of counsellors' own history of childhood sexual abuse and sexual abuse schema; (c) the relative amounts of sexual abuse-related training opportunities undertaken by different groups of counsellors. In addition, discrete analyses (item-by-item) were conducted on the sexual abuse effects (Part C) section of the questionnaire. Results are reported here. Sexual Abuse Effects (Part C): Additional Analyses. Although Hypothesis Four, that there was a greater likelihood for inexperienced counsellors to provide extreme responses to the items than experienced counsellors, was not substantiated by the planned statistical analyses, further evaluation of the responses was undertaken. Item-by-Item Comparison (Vignettes SA and NSA). A comparison was conducted on each of the 20 sexual abuse effect items to identify any individual item differences between combined experience groups. T-tests were calculated to ascertain whether the mean scores differed between the high and low combined experience groups (Vignettes SA and NSA). Eight of the 20 items produced significant differences; a ninth approached significance (see Table 19). The remaining 11 items did not yield significant differences. The results were not all in the expected direction: the high experience group was closer than the low experience group to the mean score of 50 on only three of the eight items, in contrast to the hypothesis that counsellors with less experience would be more likely to polarize their responses. Counsellor Schema 86 Table 19 T-test Results of Responses to Sexual Abuse Effects (Part O by Low and High Combined Experience Groups (Vignettes SA and NSA) Combined Experience3 Low(N = 26) High(N = 23) Effect X SD X SD t Trust others indiscriminately 26.92 15.94 32.61 22.81 -1.02 .31 Avoid physical touch 51.73 18.16 55.65 15.32 - .81 .42 Wear provocative clothing 32.69 15.38 42.61 15.14 -2.27 .03 Obsessive about personal hygiene 45.00 19.24 48.70 20.96 - .64 .52 High achieving in work or school 53.08 16.68 52.17 17.04 .19 .85 Deny negative feelings 50.20 18.96 61.30 18.66 -2.04 .05 Abuse alcohol or drugs 53.27 19.64 63.26 16.76 -1.90 .06 Have dissociative episodes 50.92 27.82 68.70 25.46 -2.32 .02 Had a positive body experience 48.33 23.88 50.65 14.79 - .40 .69 Anticipate others needs 69.42 13.44 79.13 13.79 -2.49 .02 Diagnosis of borderline personality 28.80 22.04 39.13 23.14 -1.58 .12 Readily trust their counsellors 25.42 10.99 34.09 17.02 -2.10 .04 Misinterpret neutral stimuli 51.20 20.27 64.55 29.21 -2.31 .03 Choose sexually abusive partners 39.81 19.42 49.55 20.64 -1.68 .09 Believe everyone can tell 46.15 19.87 65.46 19.93 -3.35 .00 Are sexually promiscuous 41.54 17.59 48.18 19.67 -1.23 .22 Counsellor Schema 87 Combined Experience3 Low (N = 26) High (N = 23) Effect X SD X SD t p. Fail to remember until adulthood 43.20 17.25 50.23 19.79 -1.30 SA involved physical force 50.00 26.08 40.0 22.89 1.40 Cannot tolerate being alone 42.40 14.80 48.64 23.16 -1.11 Attempted suicide at least once 39.40 21.03 54.77 21.18 -2.49 a Low = Low SA Experience (one year or less) and Low General Experience (four years or less) High = High SA Experience (four years or more) and High General Experience (ten years or more) Item-by-Item Comparison (Vignette NSA). T-tests were calculated to compare the means of low and high combined experienced counsellor groups using the sample comprised solely of Vignette NSA. Few items yielded significant differences (see Table 20): two were statistically significant and an additional two approached significance. All four of the items were also represented in the combined vignette versions sample. Counsellor Schema 88 Table 20 T-test Results of Responses to Sexual Abuse Effects (Part C) by Low and High Combined Experience Groups (Vignette NSA) Combined Experience3 Low (N = 15) High (N = 13) Effect X SD X SD t Trust others indiscriminately 26.67 16.33 34.62 21.45 -1.11 .28 Avoid physical touch 52.67 21.20 53.08 14.37 - .06 .95 Wear provocative clothing 32.00 13.73 39.23 13.21 -1.41 .17 Obsessive about personal hygiene 44.68 19.59 42.31 20.88 .31 .76 High achieving in work or school 54.00 18.82 52.31 17.39 .25 .80 Deny negative feelings 49.29 20.18 57.69 21.66 -1.04 .30 Abuse alcohol or drugs 51.33 19.59 57.31 17.87 - .84 .41 Have dissociative episodes 44.60 29.04 63.85 28.15 -1.77 .09 Had a positive body experience 46.92 27.50 48.08 11.46 - .14 .89 Anticipate others needs 70.00 13.63 76.15 15.57 -1.12 .28 Diagnosis of borderline personality 27.14 23.02 37.69 19.91 -1.35 .19 Readily trust their counsellors 22.86 9.95 35.39 14.36 -2.65 .01 Misinterpret neutral stimuli 52.86 23.67 56.15 17.09 - .41 .68 Choose sexually abusive partners 36.67 19.15 46.15 21.52 -1.23 .22 Believe everyone can tell 43.00 19.25 61.54 21.93 -2.38 .03 Are sexually promiscuous 42.00 18.97 41.54 18.64 .06 .95 Counsellor Schema 89 Combined Experience3 Effect Low (N X = 15) SD High (N X = 13) SD t E* Fail to remember until adulthood 44.29 15.55 48.85 20.43 - .66 .52 SA involved physical force 47.33 28.90 41.54 22.67 .58 .56 Cannot tolerate being alone 40.00 17.97 41.54 17.25 - .23 .82 Attempted suicide at least once 36.43 15.98 50.39 21.45 -1.93 .07 3 Low = Low SA Experience (one year or less) and Low General Experience (four years or less) High = High SA Experience (four years or more) and High General Experience (ten years or more) These results do seem to suggest a difference in how the experience groups tended to score the items although not always in the expected direction. In other words, it can be inferred that there may be a difference in the sexual abuse schemas of counsellors with varying levels of experience; however, further exploration would be necessary to understand the nature of the differences. Current Sexual Abuse Counsellors. A question was included in the demographic section to learn about the current client population of participants: "do you currently work with adults who have a history of sexual abuse?" Twenty respondents (20.6%) indicated "no;" the remaining 77 (79.4%) indicated that they did. It should be noted here, however, that some of the agencies included in the sample pool were programs for children, so child therapists working in the field of sexual abuse may have responded with a "no" to this question. The "no" group may also have included graduate students who had not yet entered the field or were presently studying and therefore not counselling any clients. Alternatively, they may have been counsellors who had not identified sexual abuse as an issue among their clients. It was decided to conduct comparisons between the groups who reported that they were and were not presently working with adult survivors. Counsellor Schema 90 Experience. The groups were first compared on some demographic information to determine whether differences existed. When compared on experience level, the group who currently were working with a sexually abused client population had significantly more general experience and sexual abuse experience than the counsellors who were not currently working with sexual abuse issues (see Table 21). Table 21 Comparison Of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors on Amount of Counselling Experience Type of Experience Amount of Experience (months) X SD t df General Current SA Counsellors Non SA Counsellors Sexual Abuse Current SA Counsellors Non SA Counsellors 113.69 52.70 50.78 7.95 85.60 61.86 48.14 16.64 -2.98* 95 -3.91* 95 * < .005 Work Setting. Participants were asked to indicate where (agency, institution, private practice) and for how long they had been working as counsellors. Chi square tests yielded a significant difference on only Counsellor Schema 91 one of the three locations, agency (x 2 = 10.11, df = 1, JL < .002), indicating that counsellors who were presently working with this client population, when compared with the group not working with sexually abused clients, were more likely to work in an agency setting (see Table 22). A possible confound of this result, however, may have been a sampling bias, as only agencies and graduate students were sampled. Table 22 Comparison of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors by Work Setting Work Setting Agency Institution Private Practice Counsellor Group N N N Current SA Counsellor 69 14 25 Non SA Counsellor 8 5 3 Chi Square 10.11* .432 2.45 *p_, < .002 Self-Reported Expertise. Using a 10 point Likert scale (0 = novice; 10 = expert), participants were asked to rate their knowledge and skills on eight counselling areas, all of which were pertinent to the field of sexual abuse counselling. T-tests were used to compare group means on the self-reported levels of expertise. Table 23 provides the results and indicates that the group of counsellors currently working with adult survivors rated themselves significantly higher than non sexual abuse counsellors on all the issues, with the exception of "child development." These findings are discussed further in the Discussion section. Counsellor Schema 92 Table 23 Comparison of Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors on Self-Reported Level of Expertise of Counselling Issues Current Sexual Abuse Counsellors (N = 97) No (N = 20) Yes (N = 77) Issue3 X SD X SD t Adult Counselling 4.9 2.6 7.5 1.7 -5.38** Child Development 5.4 2.2 6.0 2.2 -1.22 Legal and Ethical Issues 4.2 2.2 3.3 2.3 -2.17* Gender and Power 4.1 2.6 6.7 2.3 -4.49** Sexual Abuse 3.8 2.3 6.5 2.3 -4.71** Sexuality 4.5 2.3, 5.6 2.0 -2.14* Substance Abuse 3.9 2.7 5.8 2.8 -2.69* Trauma 5.0 2.8 6.7 2.2 -2.84* 3 Respondents rated themselves on a 10 point scale where 0 = novice, 10 = expert. *p_, < .05 **rx< .001 Vignette Received. A comparison was made for these groups on which vignette they received. Although not statistically significant, the group currently working with this client population were more likely to receive the sexual abuse vignette rather than the non-sexual abuse vignette (x^ = 3.43, df = 1, rx < .06). Counsellor Schema 93 Identification of Sexual Abuse (Vignettes SA and NSA). A chi square test was used to determine whether the identification of sexual abuse in the vignette varied as a function of the counsellors' present client load. Results indicated that the current sexual abuse counsellors were overwhelmingly more likely to identify sexual abuse as a treatment issue than were those not presently working in the field (x 2 = 7.57, df = 1, p_. < .006). This finding suggested that working with sexually abused clients was associated with the likelihood of identifying sexual abuse from information presented in written form. (See Table 24): those counsellors who indicated they currently worked with sexually abused clients were more likely than non sexual abuse counsellors to identify sexual abuse as a treatment issue in response to the written vignette. Table 24 Association of Current Sexual Abuse Counselling and Identification of Sexual Abuse (Vignettes SA and NSA) Identification of Sexual Abuse Row Current S/A Counselling No Yes Total Percentage No(N = 20) 11 9 20 20.6 Yes(N = 77) 18 59 77 79.4 Column Total 29 68 97 100 Percentage 29.9 70.1 100 100 x 2 = 7.58df = l , p u < .006 Counsellor Schema 94 Vignette NSA. A chi square test was used to compare the groups who received Vignette NSA only. Results were statistically significant (see Table 25), indicating that the group of current sexual abuse counsellors were more likely than non sexual abuse counsellors to identify sexual abuse as a treatment issue in response to the subtle cues of Vignette NSA. Table 25 Association of Current Sexual Abuse Counselling and Identification of Sexual Abuse (Vignette NSA) Identification of Sexual Abuse Row Current S/A Counselling No Yes Total Percentage No(N= 14) 11 3 14 28.0 Yes (N = 36) 15 21 36 72.0 Column Total 26 24 50 100 Percentage 52.0 48.0 100 100 x 2 = 5.50,df = 1, p x .02 Characteristics: Number of Adjectives Checked. Hypothesis Two, which stated that inexperienced counsellors would demonstrate a less complex schema by rating fewer adjectives as "sometimes likely" or "frequently likely," was extrapolated for the purposes of comparison by the groups of counsellors who were or were not currently working with sexually abused clients. It was expected that clinicians who were currently working with this population would check more adjectives, as had been hypothesized for experienced counsellors. The first analysis was conducted with a combined sample of Vignette SA and Vignette NSA. A t-test was conducted to compare the mean number of adjectives rated as "sometimes" or "frequently" by the Counsellor Schema 95 current sexual abuse counsellors and those not presently working with sexually abused clients. Results were not significant (see Table 26). A further comparison was calculated using Vignette NSA only; differences between the groups approached but did not reach statistical significance (p_. < .09). Table 26 T-test Results Comparing Mean Number of Adjectives Rated as "Sometimes" or "Frequently" by Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors Adjectives Counsellor X SD t df Vignettes SA and NSA Current SA Counsellors (N = 77) Non SA Counsellors (N = 20) Vignette NSA Current SA Counsellors (N = 36) Non SA Counsellors (N = 14) 82.32 8.07 84.70 8.41 80.75 9.24 85.71 8.18 1.16 95 1.76* 48 *p_. < .09, approaching significance As with the comparisons of counsellors on the basis of experience level, results did not indicate the existence of a relationship between current client load and schematic complexity as evaluated by the number of adjectives selected by respondents to describe a sexually abused client. Polarity of Characteristics. These two groups of counsellors were also compared on their tendency to polarize their ratings of adjectives in Part B of the questionnaire. Adjective ratings were recoded prior to conducting this analysis. Adjectives rated as "never" and "frequently" (the poles) were assigned a value of Counsellor Schema 96 two; ratings of "rarely" and "sometimes" were assigned a one. Individual participant means were calculated. Grand means for the two counsellor groups were computed and compared by use of a t-test. A higher mean (closer to two) would indicate a greater tendency to rate the adjective in a polarized way. T-tests were conducted for the entire sample (Vignettes SA and NSA) and for the Vignette NSA recipients only. As presented in Table 27, there were statistically significant differences for comparisons of the entire sample and for the Vignette NSA group only. In both cases, the counsellors not currently working with sexually abused clients had higher means than current sexual abuse counsellors, indicating a greater tendency to rate the adjectives in the expected direction: with polarized scores. Table 27 T-test Results of Mean Scores for Polarized Ratings of Adjectives Between Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors (Vignettes SA and NSA. and Vignette NSA only) Polarity of Responses Counsellor Group X SD t df Combined Vignettes SA and NSA Current Sexual Abuse Counsellors No (N = 18) 1.77 Yes (N = 66) 1.67 Vignette NSA Current Sexual Abuse Counsellors No(N=12) 1.78 Yes (N = 33) 1.63 130 .143 2.80 82 .006 .129 .155 3.01 43 .004 Counsellor Schema 97 Sexual Abuse Effects (Part C). A method of analysis similar to that used to address Hypothesis Four, which assessed the schematic property of complexity by comparing the groups on extremity of responses to the twenty sexual abuse effects items in Part C, was used to compare the groups of counsellors based on their current client population. The absolute values of each item score were calculated, resulting in a range of scores from zero to 50. The values were then tabulated to produce a single score. Counsellors who tended to rate on the extreme ends of the scale received higher scores, whereas counsellors who tended to rate more moderately received lower scores. A Mann-Whitney test was used to compare the groups. The analyses were conducted on two samples: the combined Vignettes SA and NSA, and Vignette NSA only. Neither test produced a significant result (see Table 28), indicating there was no relationship between counsellor group and the tendency to rate the items in a polarized fashion. Table 28 Mann-Whitney Test Results Comparing Responses to Sexual Abuse Effects (Part C) by Current Sexual Abuse Counsellors and Non Sexual Abuse Counsellors (Vignettes SA and NSA. and Vignette NSA alone) Part C Ratings Counsellor Group Mean Rank Z Vignettes SA and NSA Current Sexual Abuse Counsellor No(N=17) 41.03 Yes(N_=71) 45.33 -.624 .53 Vignette NSA Current Sexual Abuse Counsellor N o ( N = l l ) 20.59 Yes(N = 33) 23.14 -.570 .57 Counsellor Schema 98 Summary of Current Sexual Abuse Counsellors. All of the same tests used in the primary analyses to compare counsellors on the basis of experience were used to compare counsellors who were currently working with sexually abused clients with those who were not. The current sexual abuse counsellors reported significantly more experience than those not working in the field. There were no significant effects from the analyses comparing number of adjectives selected, polarity of adjectives, and polarized responses to sexual abuse effects items. One finding, identification of sexual abuse as a treatment issue, was significantly different between the two groups. Current sexual abuse counsellors were significantly more likely to identify sexual abuse for the two vignette versions combined (SA and NSA) and the disguised version (Vignette NSA) alone. As it was a post hoc analysis, and the two groups were of unequal distribution, the results must be treated with caution. They may, however, prove to be an interesting avenue for further research. Counsellors' Own History of Sexual Abuse Respondents were asked about their own histories of childhood neglect, emotional, physical, and sexual abuse, and adult physical or sexual assault. Post hoc analyses were conducted to determine whether this counsellor variable was associated with any of the schematic properties being evaluated in this study. Twenty-four of the respondents (24.7%) indicated they had been sexually abused as children, compared with 71 (73.2%) who did not. Two respondents did not answer this question. Identification of Sexual Abuse (Vignettes SA and NSA). The first analysis was to determine whether there was an association between a history of any of the types of childhood abuse or adult assault and the identification of sexual abuse in the vignette section of the questionnaire. The entire sample (Vignettes SA and NSA) was used. As Table 29 indicates, the only variable that yielded a significant result was the history of childhood sexual abuse (x^ = 4.45, df_= 1, fx < .03). This strong result suggested that a counsellor's personal experience with childhood sexual abuse, as opposed to other types of abuse, was associated with whether they would detect sexual abuse from the information provided in a written "intake" vignette. Counsellor Schema 99 Table 29 Association of History of Childhood Abuse and Adult Assault, and Identification of Sexual Abuse (Vignettes SA and NSA) History Sexual Abuse Identified No Yes Chi Square (df= 1) Childhood Physical Abuse No Yes Childhood Emotional Abuse No Yes Childhood Sexual Abuse No Yes Childhood Neglect No Yes Adult Physical Assault No Yes Adult Sexual Assault No Yes 26 2 19 9 25 3 25 3 24 4 53 14 38 29 46 21 50 17 61 6 26 62 2 5 2.67 1.02 4.45* 2.55 .596 .003 * p_, < .03 Counsellor Schema 100 Identification of Sexual Abuse (Vignette NSA). A chi square test was conducted with Vignette NSA only, to assess whether counsellors with a personal history of childhood sexual abuse were more likely to identify sexual abuse as a treatment issue in response to ambiguous cues. The other abuse histories were not included in this analysis because of the lack of significant results for these variables from the combined vignettes sample. Results (see Table 30) indicated that a counsellor's personal history of childhood sexual abuse was strongly associated with the likelihood of identifying sexual abuse as a treatment issue in response to ambiguous cues. Counsellors who were themselves sexually abused as children were more likely to identify sexual abuse in response to subtle cues than counsellors who were not sexually abused. Table 30 Association of History of Childhood Sexual Abuse and Identification of Sexual Abuse (Vignette NSA1 Sexual Abuse Identified Childhood Sexual Abuse No Yes Total Percentage No (N = 36) 23 13 36 75.0 Yes (N= 12) 2 10 12 25.0 Column Total 25 23 48 100 Percentage 52.1 47.9 100 100 x 2 = 8.04,df= 1,EL<.005 Counsellor Schema 101 Additional Analyses. All analyses conducted between other counsellor groups were replicated to compare counsellors who did and did not report personal histories of sexual abuse. They were compared on the number of adjectives selected to describe a sexually abused client; the polarities of their ratings of the adjectives and of sexual abuse effects, and number of positive and negative adjectives. No significant results emerged for any of the subsequent analyses. It appeared that the only variable in this study associated with the counsellor's own history of childhood sexual abuse was the identification of sexual abuse as a treatment issue in response to the ambiguous cues of Vignette NSA, suggesting that counsellors with their own histories of childhood sexual abuse were more likely than counsellors without a sexual abuse history to recognize sexual abuse in response to subtle cues. In addition, there was the finding, as reported earlier, of a significant difference on anticipated duration of therapy, with counsellors who had themselves been sexually abused recommending a longer duration of treatment than counsellors who had not been abused. Identification of Sexual Abuse in Response to Ambiguous Cues (Vignette NSA). Fifty non sexual abuse vignettes were returned: 24 (48%) of these respondents identified sexual abuse from the provided cues; 26 (52%) did not. The question arose of who comprised the group of counsellors able to identify the sexual abuse. Multiple variables were evaluated to determine if they were associated with the identification of sexual abuse (see summary presented in Table 31). Some of these results are reported elsewhere and have been replicated here for the purposes of summarizing the findings. Counsellor Schema 102 Table 31 Variables Associated with Identification of Sexual Abuse Responses to Ambiguous Cues ("Vignette NSA) Variable Sexual Abuse Identified No Yes Chi Square General Experience Low (N = 20) High (N=19) Combined Experience Low (N= 16) High (N= 13) Sexual Abuse Experience Low (N = 21) High (N=18) Current Sexual Abuse Counsellors No (N= 14) Yes (N = 36) Own History of Childhood Sexual Abuse No (N = 36) Yes (N= 12) 12 10 9 7 12 11 15 23 2 8 9 7 6 9 10 3 21 13 10 .215 .108 .626 5.50* 8.04*=" Counsellor Schema 103 Sexual Abuse Identified Variable No Yes Chi Square Agency Work No (N = 8) 4 4 Yes (N = 42) 22 20 .015 Institution No(N = 40) 19 21 Yes (N = 9) 6 3 1.08 Private Practice No(N = 35) 19 16 Yes (N = 24) 17 7 .168 *rx < .05 **rx < .005 This table indicates that the only variables significantly associated with the likelihood that counsellors would identify sexual abuse as a treatment issue in response to the subtle cues of Vignette NSA was whether or not they were currently working with this population, and whether or not they had their own history of childhood sexual abuse. Non sexual abuse counsellors were less likely to identify sexual abuse than those currently working in the field (p_. < .05). Counsellors with their own experiences of sexual abuse were more likely than those without a sexual abuse history to identify sexual abuse (p_. < .005). There was no relationship between general experience, sexual abuse experience, combined experience, or work location, and the identification of sexual abuse. A t-test comparing counsellors' mean age was similarly non-significant: those who identified sexual abuse were older (X = 41.96, SD = 6.81) than those who did not (X = 39.65, SD = 8.18), but the difference was not significant (t = -1.08, df = 48, tx = .286). Counsellor Schema 104 There was, therefore, no support for the hypothesis that experience was related to organization of schema, as measured in the current study. Learning As a way of evaluating the range of training activities undertaken, participants were asked to indicate how many and which means they had used to learn about sexual abuse treatment (total possible listed was 15). This question was not considered as a function of the vignette received; therefore the total sample was included in the analyses. T-tests were used to compare group means on the number of learning methods. Significant differences were found (see Table 32), with all the high counselling experience groups noting a greater number of training activities than low counselling experience groups. Results for the combined experience category (t = -5.29, df = 47, < .005), for the general experience category (t = -2.12, df = 60, rx < 04) and for the sexual abuse experience category (t = -6.12, df = 74, p_. < .001) were all statistically significant. The group of counsellors currently working with sexually abused clients indicated a higher number of learning activities than did those who did not have report working with sexually abused clients (t = -5.07, df = 95, < .001). Similarly, counsellors with their own histories of sexual abuse had significantly higher rates of learning activities than counsellors who did not personally experience sexual abuse (t = -5.16, df=93,rr<.001). Counsellor Schema 105 Table 32 T-test Results Comparing Mean Number of Training Activities Undertaken by Counsellors Number of Training Activities Variable X SD t df Combined Experience Low (N = 26) High (N = 23) General Experience Low (N = 33) High (N = 29) Sexual Abuse Experience Low (N = 40) High (N = 36) Current Sexual Abuse Counsellors No (N = 20) Yes (N = 77) Own History of Childhood Sexual Abuse No(N = 71) Yes (N = 24) 3.385 2.24 7.91 3.12 -5.29** 47 5.09 3.27 6.83 3.16 -2.12* 60 4.25 2.55 8.08 2.91 -6.12** 74 3.65 2.89 7.23 2.79 -5.07** 95 5.58 2.81 9.00 2.81 -5.16** 93 * p_, < .05 **p, < .005 The results indicated that the more experience a counsellor had, either as a function of years of general counselling experience or in the specific field of sexual abuse, the greater number of learning activities he Counsellor Schema 106 or she was likely to have undertaken. Two explanations can be offered to account for this significant difference: length of time in the field may have provided a concomitant amount of time to participate in training activities. Alternatively, as experience increased, so did the recognition of the need for training and hence more opportunities were sought. Counsellors with their own sexual abuse histories were also more likely to participate in training activities; whether this result was an anomaly related to amount of counselling experience, or a function of personal interest, cannot be determined from the results. Counsellor Schema 107 Chapter Five Discussion This initial investigation was designed to evaluate, by inferential means, the relationship of counsellors' experience to their sexual abuse schemas. The goal of the study was to test the hypothesis that the sexual abuse schemas of inexperienced counsellors would be less complex than those of counsellors who have more experience, as indicated by less organization, less breadth, and greater extremity. It was anticipated that higher amounts of general counselling experience and specific sexual abuse counselling experience would be differentially associated with the participants' schemas. Specifically, it was hypothesized that, as an indication of greater organization of schemas, experienced counsellors would be more likely to identify sexual abuse as a treatment issue in response to the disguised version of the case vignette than to the explicit sexual abuse version, which were identical except for the explicit statement of sexual abuse in the latter version. Experienced counsellors' greater breadth of schema, as hypothesized in this study, was expected to be manifested by a greater number of adjectives selected to . describe a sexually abused client. The anticipated extremity of inexperienced counsellors' schemas in comparison with those of counsellors having more experience was evaluated by the polarity of responses to the adjective list, and to the list of 20 items describing various sexual abuse effects. Inexperienced counsellors were expected to polarize their responses in these two sections more frequently than were experienced counsellors as a manifestation of greater extremity, and therefore less complexity, of schemas (Fiske & Taylor, 1984). Additional exploratory questions were examined. The first assessed participants' recommended duration of treatment for the hypothetical client, to ascertain whether counsellors with more experience would predict a longer course of therapy for the sexually abused client in the vignette. Secondly, experienced and inexperienced groups were evaluated on their tendency to attribute negatively valued adjectives to sexually abused clients with the expectation that less experienced counsellors would exhibit a greater tendency. Post hoc analyses were undertaken using different counsellor groups as bases of comparison to evaluate whether other variables were associated with complexity of schema. In particular, Counsellor Schema 108 counsellors currently working with sexually abused clients were compared with those who were not, and counsellors with their own personal histories of childhood sexual abuse were compared with those who did not experience sexual abuse as children. It was expected that counsellors having current contact with the client group, and those with their own histories of sexual abuse would have more complex schemas as a result of their schemas being more developed through experience with the phenomenon (Fiske & Taylor, 1984). Because of greater schematic organization that permits recognition of subtle as well as explicit cues, they would be more likely to identify sexual abuse as a treatment issue in response to the disguised vignette; select a greater number of adjectives to describe sexually abused clients (schematic breadth); and be less polarized in their adjective ratings and responses to sexual abuse effects items (schematic extremity). It should be noted that the questionnaire focussed on female sexually abused clients. One of the reasons for this decision was to reflect the higer prevalence of sexual abuse for women than men. Gender was also identified to minimize a possible bias or confound, as there are some effects of sexual abuse for males that are different than those for females. The discussion, conclusions and speculations should therefore be considered with respect to female clients who have been sexually abused. Results and any conclusions can only be generalized bo female clients. Overview Contrary to predictions, there was little empirical evidence for the hypothesis that the dependent variable, schema, would vary with the independent variable, experience. This was true for the dimensions of breadth and extremity. Indirect support for an experience effect was found when current sexual abuse counsellors (who were significantly more experienced) were compared with non sexual abuse counsellors; the sexual abuse counsellors identified sexual abuse as a treatment issue in response to the disguised vignette more frequently than the non sexual abuse counsellors. The literature on counsellor experience and clinical judgement supports the general finding of no experience effect, as research has consistently demonstrated that greater experience does not increase accuracy with respect to clinical judgement (Grebstein, 1963; Wiggins, 1973; Clavelle & Turner, 1980; Counsellor Schema 109 Garb, 1989). However, the process of clinical judgement involves the cognitive processes of information gathering, integration, generalization, and inference, all of which are functions of schemas (Turk & Salovey, 1985). Schemas are fundamental to the way in which the therapeutic process is guided by the counsellor (Haverkamp, 1993), although the exact nature of the interaction of counsellor schemas and the therapeutic process remains unclear (Heppner & Frazier, 1992). Clinical judgement is therefore related to schemas, but it cannot be assumed that research on clinical judgement provides evidence about schemas. The emerging research into cognitive complexity may further illuminate the findings. Cognitive complexity, rather than experience, seems to be associated with accuracy in clinical judgement (Spengler & Strohmer, 1993). Counsellors of greater cognitive complexity are more abstract thinkers and therefore ask more questions, generate more hypotheses, and are more accurate than clinicians who are cognitively simple, or more concrete in their thinking (Holloway & Wolleat, 1980). Spengler and Strohmer (1993) have found that clinicians with greater cognitive complexity surpassed lower complexity counsellors in their ability to avoid stereotyping, simplification, and generalization, all of which are processed through schemas. The variable of cognitive complexity may therefore affect the organization and breadth of counsellor schema. This factor may mediate the cognitive processes and abrogate any experience effect. A counsellor with less experience may be able to apply superior cognitive complexity to a problem or phenomenon and utilize this personal resource to understand a situation, thereby being less constrained by a lack of experience; and a counsellor who is less cognitively complex may not always benefit from more experience. It can be speculated that the results of the current study reflect differences, not of counsellors' experience, but of their relative cognitive complexity. Cognitive complexity, like schema, has an influence on the ability to integrate discrepant information (Spengler & Strohmer, 1993). The current study found a significant difference on participants' ability to identify sexual abuse as a treatment issue in response to the ambiguous vignette version. Whether this was a function of greater schematic organization or of cognitive complexity cannot be categorically concluded. It may be that cognitive complexity rather than schema complexity contributed to the participants' ability to recognize the schema-discrepant information. If Counsellor Schema 110 cognitive complexity was a mediating factor, then, it may help explain the lack of an experience effect. Further investigation of the effect of cognitive complexity on schema might provide an interesting avenue of research, to continue the exploration of the relationship between the two factors and determine whether they have, for example, an additive or exponential effect. The possibility that schema content, rather than complexity, was being evaluated in this study must be given serious consideration. Was the questionnaire accessing the sexual abuse knowledge of participants rather than the schematic properties under investigation in this study? While greater knowledge can be associated with, and help create, more complex schemas, there is no necessary connection. A person may have a large base of information (schema content) but not have many links between individual facts (schema organization). The findings in this study would then be consistent with some research on sexual abuse knowledge. For example, Hibbard and Zollinger (1990) found that formal training had a mild positive effect on knowledge. The finding that more adequate abuse-specific training was reported by more recent graduates (Pope & Feldman-Summers, 1992) may suggest that these factors serve to overcome the handicap of less experience. Recent graduates and current students with little experience may have received more relevant training based on current and emergent research. Their knowledge base may therefore be greater and updated in comparison with experienced clinicians who have been unable to learn of the latest empirical studies. Although the analysis of learning activities undertaken by clinicians in this study yielded significantly higher numbers of methods by more experienced counsellors, there was no information on the quality or extent of the learning. .If the present survey has to some extent accessed knowledge, counsellors with less experience may not be entirely hindered, therefore differences may not emerge between the experience groups. An additional possible explanation for the lack of significant findings has its source in Dye and Roth's (1990) study of clinician stereotypes. They found an inverse correlation between age and knowledge of treatment themes: as therapist age decreased, knowledge increased. The present study found a significant age effect on the experience variable: counsellors with more experience were significantly older than lesser Counsellor Schema 111 experienced therapists. The results may reflect that clinicians who have been in the field significantly longer, and therefore out of school for a longer period of time, have outdated knowledge (Garb, 1989). It may again be a case of the benefits of experience being cancelled out by another factor, in this case age. It can be speculated that other factors have influenced these results. Fiske and Taylor (1984) observed that the frequency or recency of a schema's use can maintain it in a state of readiness and easy accessibility, a phenomenon known as priming. Snyder and Thomsen (1988) noted that expectations based on prior contact with individuals will affect one's hypotheses. Clinicians are known to make diagnoses extremely quickly and then proceed to seek supporting evidence that is corroborated by their experience and knowledge of a particular presenting problem (Snyder & Thomsen, 1988). Given that 79.4% of the respondents indicated that they currently worked with adult survivors of sexual abuse, a high proportion of the counsellors, regardless of experience level, could conceivably have been completing the questionnaire with a primed sexual abuse schema. Additionally, the pervasive attention in the media to the phenomenon of childhood sexual abuse, and its prevalence among the clinical population, may be priming the sexual abuse schemas of counsellors as a group. Cases of child sexual abuse are now regularly reported in the media to the public, so even counsellors who do not work in the field are likely to have the phenomenon in their awareness. Given the high incidence in clinical populations, there is an increased likelihood that non sexual abuse counsellors will hear from colleagues and in case consultations about its existence and effects. The volume of professional information about sexual abuse that circulates cannot be entirely disregarded. In sum, counsellors as a profession have knowledge, to at least a limited degree, of the phenomenon of child sexual abuse. Consequently, their schemas may be frequently primed and therefore readily activated regardless of their experience or clinical focus. Identification of Sexual Abuse in Response to Vignettes A chi square analysis indicated a very strong association between the vignette version received and the identification of sexual abuse as a treatment issue (x^ = 24.05, df = 1, p_. < .001). Participants receiving Vignette SA were more likely than those receiving Vignette NSA to have included the hypothesis of sexual Counsellor Schema 112 abuse in their list of treatment issues identified. This relationship was not, however, mediated by counsellor experience level. Clinicians of low and high levels of experience (general counselling experience, specific sexual abuse counselling experience, or a combination of the two) were similarly able to identify sexual abuse as a treatment issue in both versions of the vignette. The hypothesis that experience would be related to counsellors' likelihood of recognizing subtle or discrepant cues (Fiske & Taylor, 1984) was therefore not supported. The findings from the present study are inconsistent with the schema literature which suggests that the more experience one has with particular concepts, phenomena, or categories, the more easily the perceiver can absorb discrepant information and extrapolate from seemingly unrelated cues (Fiske & Taylor, 1984); and that training and experience function to develop counsellors' schemas (Turk & Salovey, 1985). Perhaps, then, the lack of significant experience differences may be attributed to experimental confounds or methods of evaluation (see "Limitations"). The vignette was an abridged "case history" that included information about the hypothetical client's current family structure, a brief developmental history, and her current presenting concerns. The disguised version was further restricted in amount of content by the omission of the sexual abuse disclosure. Given the limited information provided, counsellors' responses about identified treatment issues may have reflected a caution about implying any preexisting conditions such as those of early childhood traumas like sexual abuse. There may have been some reluctance to develop hypotheses based on minimal information which is not necessarily poor therapeutic practice. In fact, recent researchers have urged counsellors to refrain from precipitous hypothesis formation (e.g., Turk & Salovey, 1985; Heppner & Frazier, 1992). Written comments provided by some participants stating they anticipated their identified treatment issues may change as the counselling process evolved provided support for the possibility that some respondents may have been exercising appropriate caution by suspending judgement. Results may therefore not necessarily reflect a knowledge or schema difference but a stylistic one. Furthermore, the questions used to elicit counsellors' responses were not completely open-ended and therefore may not have provided ample latitude for counsellors to speculate about the nature of the client's history. Counsellor Schema 113 The questionnaire was designed to provide various means of operationalizing different dimensions of schema complexity. As discussed previously, as a cognitive process, schema measurement is notably difficult due to its invisibility. Inferential means are required to evaluate schema. The vignette condition produced a strong effect with counsellors identifying sexual abuse more frequently in response to the explicit rather than disguised version. However, the hypothesis that experience would be associated with identification of sexual abuse was not borne out. One possibility is that this method of evaluating schema was not capable of achieving that goal; perhaps a difference existed but was not detected by this method. That the disguised vignette version produced great variation in responses (52% did not identify sexual abuse; 48% did) suggests the counsellors' sexual abuse schemas were therefore differentially activated for this condition. When schemas ". . . are actively involved in the comprehension of related information, they facilitate the understanding of material that might otherwise be unintelligible" (Higgins & Bargh, 1987, p. 378). However, the remainder of the questionnaire, in which the sexual abuse focus of the study was clear, did not reveal differences between the vignette groups. Schemas would by then have been activated, and no differences emerged between counsellor groups when, for example, the questionnaire task asked for a description of a sexually abused client, and there were no differences on the number of adjectives between the counsellors who did and did not identify sexual abuse in the disguised vignettes. Those who did not identify sexual abuse in the disguised version appeared to have a schema to draw on, but it was not necessarily activated in the ambiguous condition. The results are inconclusive as there was no strong relationship between counselling experience and identification of sexual abuse. However, the findings of an equal distribution of counsellors who did (N = 26, 52%) and did not (N = 24, 48%) identify sexual abuse responses on Vignette NSA warranted further investigation about what counsellor variables contributed to these results. Significant differences did emerge for two groups of counsellors: those who currently work with sexually abused clients listed sexual abuse as a treatment issue more frequently than the non sexual abuse counsellors, and counsellors with their own histories of sexual abuse identified sexual abuse more frequently than counsellors who had Counsellor Schema 114 not themselves been abused. The current sexual abuse counsellor group was significantly more experienced, lending indirect support for an experience effect. It can be surmised that both groups of counsellors (current sexual abuse counsellors and those with their own histories) have sexual abuse schemas that are regularly primed because of their professional and personal experience with, and exposure to, the issues. It would be important to learn what other factors distinguished the counsellors who did identify sexual abuse in response to subtle cues from those who did not. If the reasons were related to variables that are amenable to change, as with knowledge, interventions such as training can be taken to remediate the situation. This finding may have implications for counsellor education. If it is more a function of counsellor variables such as motivation, arousal and attention, variables that affect involvement and complexity of counsellors' responses (Spengler & Strohmer, 1993), more focus would need to be given in education and in supervision to ameliorate these influences. One possible explanation for the limited support for an experience effect derives from a study of „ experience, amount of information and clinical judgements (Brenner & Howard, 1976). These authors found that the more information received by experienced (but not inexperienced) therapists, the less accurate their empathic inferential judgements became. Variance in responses increased (with therefore more errors) when a judge increased the amount of information to which he attended (p. 726). Brenner and Howard suggested that these therapists may have "attempted to integrate as much of the information as they possibly could.. . [whereas] the least experienced therapists. . . probably were overwhelmed by the complexity of the therapy session itself and the experimental task and, as a consequence, became more accurate the more they heard" (p. 726). In the current study, the experienced group's attempt to integrate all the information in the vignette may have interfered with their ability to be accurate. Less experienced counsellors, however, may have been better able to understand the situation in the fully disclosed vignette because of the provision of additional information. An experience effect may therefore not have been manifested. It should be noted, however, that the activities undertaken by participants in the Brenner and Howard study, which involved Counsellor Schema 115 listening to an audio tape of a session and completing the "Therapy Session Report," were considerably different from the questionnaire completion task of this study. Although providing interesting speculation to explain the lack of an experience difference, Brenner and Howard's conclusions cannot therefore be generalized with confidence to explain the results of the current study. A further caveat is that Brenner and Howard's findings contradict those of other researchers such as Weiss (1963), who found that increasing the amount of available information had a positive effect on behavioural predictions, provided the information was in the form of case history data (as opposed to, for example, psychological profiles). It must be remembered, too, that empirical evidence supporting a relationship between experience and clinical judgement is limited, as indicated in literature reviews by Wiggins (1973) and Garb (1989). As noted earlier, the application of this body of research to a study of schema must be done with discretion. The lack of a significant gender difference was consistent with previous research, although the current sample contained few men. Kendall-Tackett and Watson (1991) investigated the variables associated with the likelihood that professionals would conclude, without benefit of a disclosure, that a child had been sexually abused. They noted a gender effect in response to vignettes containing ambiguous information: female clinicians were more likely than male to identify sexual abuse (p. 387). In addition, Attias and Goodwin (1985) discovered in their survey of clinician knowledge about sexual abuse that male professionals made uniformly more errors than did female participants. In the present study, male and female counsellors alike identified sexual abuse in response to Vignette SA. Twice as many of the male respondents did not identify sexual abuse in response to Vignette NSA as those who did. However, the small number of male participants and the skewed distribution of male (13.4%) versus female (86.8%) respondents in the present study do not allow for generalization of the findings. An additional element must be considered. Although every effort was employed to disguise the sexual abuse focus during data collection to protect the integrity of the vignette section, there remains the possibility that the fully informed supervisors and directors who were charged with distributing the packages may have unwittingly exposed the study's focus. Furthermore, this researcher, as a counsellor in Counsellor Schema 116 a Vancouver sexual abuse center, is known to some degree in the community. Participants may have made assumptions about the nature of the study given the researcher's employment. A final conjecture arises from the current controversy in the media about repressed memories and the highly publicized civil actions directed at therapists who have allegedly elicited "false memories" of childhood sexual abuse from their clients. Rabinowitz (1993) noted that the process of clinical decision-making in the mental health field is "complicated by beliefs that diagnoses can be destructive labels" (p. 308). This particular deterrent may not be as compelling in counselling settings as in psychiatric facilities, but the field of sexual abuse may have as its counterpart the "false memory" debate. Awareness of the debate may have instilled caution among therapists when "diagnosing" sexual abuse as the client's issue. Furthermore, in light of the increasing attention to repressed memories in the media and the therapeutic community, some participants may have surmised the study was about "false memories" and consequently been reluctant to identify sexual abuse if it was not explicitly identified. This would account for the sharp difference in responses to the explicit versus subtle vignettes. Breadth of Schema: Number of Adjectives The frequency with which respondents rated the adjectives as "sometimes likely" and "frequently likely" to be characteristic of an adult woman with a history of childhood sexual abuse were tabulated and compared with t-tests. The analyses were conducted twice: on the sample comprised of both Vignettes SA and NSA; and for Vignette NSA only. None of the results were significant, indicating that experience level was not associated with breadth of schema as inferred by this means of evaluation: counsellors with more experience did not select a significantly greater number of adjectives to describe a woman who had been sexually abused. The lack of significant results held when the vignette condition was controlled, suggesting that the identification of sexual abuse in one of the vignettes did not have a priming effect on how counsellors characterized sexually abused individuals. This finding is inconsistent with schema research which indicates that, with maturity, a schema develops and contains more elements (Fiske & Taylor, 1984). It had been anticipated that experienced counsellors, with broader schemas, would be more likely to view individuals with a history of sexual abuse Counsellor Schema 117 as they would people in the general population (with a similar range of characteristics) and choose more adjectives than would inexperienced counsellors with their hypothesized narrower schemas. The results may, however, reflect the research that indicates therapists do not remain abreast of current research nor apply it in practice (Dye & Roth, 1990). A counterbalancing of schemas may have occurred between clinicians who have been in the field for years but have fallen behind with the emerging understanding, and more novice clinicians whose current learning has allowed them to reach a knowledge level concomitant with that of their experienced counterparts. Further, the current study did not permit comparison with a list of adjectives that do or do not definitively describe a sexually abused client. The respondents as a group rated the list overwhelmingly with "sometimes," suggesting a tendency to score in the middle. Although the directions included a caveat acknowledging the difficulty in generalizing, there may have been an unwillingness among counsellors to categorize, reminiscent of the effect suggested by Rabinowitz (1993) regarding the reluctance on the part of clinicians to diagnose their clients. The lack of significant differences may also be due to the method; perhaps rating a list of adjectives does not access schematic breadth. There may have been more effective means of evaluating this schematic property that would yield differences, and may be worthy of further investigation. For example, asking an open-ended question for participants to "describe a sexually abused client" would provide considerably more latitude for counsellors to generate what is contained in their schemas. Responses could be evaluated for schematic content (information and knowledge) and breadth (number of components). Alternatively, the fact that differences did not emerge may reflect a sensitivity on the part of counsellors to the multidimensionality of client populations. In this case, the failure to discover a significant finding is therefore a positive indication, suggesting that this sample of counsellors is refraining from stereotyping or limiting their versions of sexually abused clients. Item-by-Item Analysis. The low and high combined experience groups of counsellors were compared for their ratings of the individual adjectives. Eleven of the 100 adjectives were rated differently to a statistically significant degree (N = 9) or approached significance (N = 2). Three of the adjectives were positive attributes and the remainder were negative. No patterns were evident from the particular Counsellor Schema 118 adjectives that yielded significant scoring differences between low and high experienced counsellors although negative adjectives did seem to be overrepresented in the results. This is similar to the findings from the comparison of negative and positive adjectives (see below). These results suggest that the inexperienced and experienced groups of counsellors exhibited some differences in their ratings on individual adjectives. However, the number of adjectives receiving differential ratings from the two experience groups is not sufficient, nor is the trend in the differences significant enough, to make assumptions or generalizations from the results. Of interest, however, is the tendency for the high experienced counsellors to rate these particular adjectives as "frequently likely" more often than the low experienced group, in contrast to expectations that inexperienced counsellors would polarize more adjectives ("never likely" or "frequently likely") than experienced counsellors. This tendency may be attributed to schema content: perhaps these characteristics (ashamed, anxious, fearful, forgetful, hypervigilant, sad, and suspicious) are noted on a more regular basis by counsellors who have experience with sexually abused clients. Four of these were among the embedded sexual abuse descriptors (anxious, ashamed, hypervigilant, and sad) as opposed to Adjective Check List items. Consistent with this explanation may be a suggestion that low experience counsellors hold expectations that sexually abused clients will present in a more positive way, which may not in fact reflect the reality of the serious emotional and psychological sequelae (e.g., Courtois, 1988; Kluft, 1990). Given this speculation, it may be appropriate to conduct more investigations into counsellor perceptions and stereotypes about this population. Certainly, if particular groups of counsellors are more likely than others to perceive these clients with such characteristics, there may be differences in treatment approaches as a result. It would be useful to know the extent of the differences and whether these perceptions are reality-based, or a by-product of myths and stereotypes. Such variations may have implications for clinical practice, training, and the ethical treatment of sexually abused clients: are the principles of beneficence or nonmaleficence in danger of being compromised (Daniluk & Haverkamp, 1993) as a result of the variation in counsellor views? Counsellor Schema 119 The results of these analyses are of limited validity and generalizability because of the small number of adjectives yielding significant differences. Furthermore, there was a lack of consistency among the results. For example, "fearful" was rated significantly differently between the counsellor experience groups, whereas its synonym "scared" did not reach significance; "hypervigilant" was significant to the .05 level but "wary," a similar descriptor, was not. The results may be more reflective of differences in schema content than complexity. The existence of some, albeit small, trends suggests the need for further research, perhaps in the area of schema content (i.e., knowledge). Breadth: Positive and Negative Adjectives. An additional analysis was conducted using the data from the characteristics section: the mean scores of positive and negative adjectives were tabulated and compared for differences between experienced and inexperienced counsellor groups. Some differences that approached statistical significance (p_. = .07) emerged on the mean scores of negative adjectives. The experienced group tended to give higher ratings for the negative adjectives in comparison with the inexperienced counsellor group. This result was not in the expected direction, as it had been hypothesized that counsellors with less experience would be more likely to make negative inferences about the clients. The tendency for experienced counsellors to rate negative adjectives higher was supported in the additional analysis that compared counsellors who currently work with sexually abused clients and those who do not. Current sexual abuse counsellors rated negative adjectives with higher scores significantly more often than the non sexual abuse counsellors (p_. < .02). As noted previously, the current sexual abuse counsellors had significantly more general and sexual abuse counselling experience lending indirect support for the experience effect. The hypothesis was that counsellors with less complex schemas would consequently demonstrate a tendency to categorize sexually abused clients in a negative manner. The results, while significant, were in the opposite direction. The clinical experience of counsellors often has a greater influence on the clinician than empirical research (Snyder & Thomsen, 1988). It can be speculated, therefore, that the results reflect the clinical experience of counsellors who encounter sexually abused clients: perhaps sexually abused clients do present with more negative attributes. Indeed, the literature is replete with studies that list the Counsellor Schema 120 serious effects associated with childhood sexual abuse (e.g., Courtois, 1988; Briere & Runtz, 1993). Sexually abused clients do present "with higher levels of psychological symptoms than nonabused clients" (Busby, Glenn, Steggell, & Adamson, 1993, p. 384), and this may result in a less positive presenting disposition. Finally, some of the "negative" characteristics may actually be adaptive for a child experiencing sexual abuse. For example, a child who is "wary," "guarded," "mistrustful" and "rebellious" may have developed these ways of being as coping or survival methods in the face of danger. Therefore it cannot be necessarily concluded that counsellors who rate clients with more negative characteristics were viewing these traits as maladaptive and negative. While not supporting the hypothesis of schema breadth, perhaps the finding suggests a naivete on the part of uninformed therapists. This finding alludes to the need for increased awareness by clinicians that sexually abused clients may not have the range of positive and negative attributes as other client groups. The results must be considered within the context of subjective interpretations of the adjectives by the researcher and participants. While efforts were undertaken to minimize such bias, there may still exist schematic differences for the attributions of these characteristics. For example, "rebellious" was included in the negative adjective list, yet some people may, based on their own experience or values, consider this a valuable trait. A sophisticated selection procedure was not used to assign adjectives to positive or negative categories; the choices were made subjectively by the researcher and consultants. This method was a weakness in this section. A more thorough analysis of the valence of the adjectives would be necessary before these results could, with confidence, be accepted. Extremity of Schema In accordance with the previous schema literature, it had been anticipated that the schemas of less experienced counsellors would be more extreme. This hypothesis was evaluated with analyses of data from the adjectives (Part B) and the 20 items describing sexual abuse effects (Part C). It was expected that counsellors with less experience would tend to polarize their responses to the items. In other words, less experienced counsellors were expected to more frequently give adjectives the ratings of "never likely" and "frequently likely" in comparison with the experienced group. Less experienced counsellors were expected Counsellor Schema 121 to select percentages closer to the extremes on the zero to 100 point line graph in response to the list of sexual abuse effects, in comparison with experienced counsellors who would score the items closer to the mean. Analyses were conducted using the total sample of Vignettes SA and NSA, and the sample of Vignette NSA alone, but failed to yield significant differences between the counsellor experience groups. These results are inconsistent with the schema literature that suggests schema maturation evolves from experience, which minimizes the extremes in the schema (Fiske & Taylor, 1984). For the current sample, the results indicated that less experienced counsellors did not show a polarization effect, and counsellors with greater counselling experience, in contrast, did not demonstrate moderation of schema extremity. However, perhaps it cannot be assumed that experience is always associated with maturity, especially in this rapidly changing area (sexual abuse). As the list of treatment effects in Part C was culled from the literature and may represent knowledge, this section, in retrospect, may be more suitable as a means of evaluating schema content. It may, therefore, be more appropriate to consider the findings in light of the clinical judgement literature, which has focussed on knowledge and accuracy of judgement. The results failed to discriminate significantly between the two experience groups and are consistent with the research that concluded amount of experience was not positively associated with accuracy of clinical judgement (Wiggins, 1973; Clavelle & Turner, 1980; Garb, 1989). As cited previously, the sexual abuse literature has thus far produced equivocal findings regarding an experience effect on knowledge. For example, Hibbard and Zollinger (1990) found a positive growth in practitioners' knowledge bases since an earlier study (Attias & Goodwin, 1985). In contrast, Rabinowitz (1993) noted sexual abuse as a prominent oversight of clinicians. Because this section of the questionnaire was not designed as a test of knowledge, it is beyond the scope of this study to assess the degree of accuracy among responses. It does, however, highlight the need for further research in this area, as little has been investigated with respect to what clinicians actually know about sexual abuse. It is rare to find a study that measures extent and accuracy of knowledge (Finkelhor, 1984; Dye & Roth, 1990; Rabinowitz, 1993). Because the Counsellor Schema 122 information base is changing and there are a multitude of factors to understand (e.g., Briere & Runtz, 1993), counsellors may have widely varying degrees of knowledge. Several explanations to account for the lack of significant findings in the analyses for schema extremity (e.g., tendency to moderate responses on the adjective list) have been offered elsewhere. Other methodological difficulties may have affected responses to the sexual abuse effects items of Part C. For example, the statistical test chosen for the analyses of the sexual abuse effects may not have been the best way to assess the responses to the items; as differences in the scoring of individual items may have been obscured by including all 20 in the analyses. In other words, a small tendency to score in a particular manner by a group of counsellors may not have emerged when the scores for all 20 items were combined. Specific items may have differentiated among counsellor groups while others did not. A between-groups difference might therefore have been masked. It was for this reason that further analysis was conducted (see below). Individual Item Differences. An item-by-item analysis did reveal a tendency for the two counsellor groups to respond differently to the list of sexual abuse effects, although not always in the expected direction or reflecting schema extremity. Collapsing the data over the 20 items, then, may have obscured existing differences, and the variation in responses may not have been related to polarity as had been hypothesized. For example, the item "trust others indiscriminately" received a low overall rating by the entire sample (X = 33.3, SD = 24.01), by the high experience group (X = 32.61, SD = 22.81), and by the low experience group (X = 26.92, SD = 15.94), suggesting a generally held belief by counsellors that clients who were sexually abused as children are unlikely to exhibit this behaviour. Although one of the effects of sexual abuse can be the tendency to trust indiscriminately (Courtois, 1988), the converse (difficulty trusting) may be the effect observed more regularly by this sample of counsellors. The results may reflect the favouring of counsellors' clinical experience over empirical evidence (Snyder & Thomsen, 1988). Similarly, the item "may be diagnosed as borderline personality disorder" was included because of the recognition that survivors have, or are misdiagnosed with, this personality disorder (Briere & Runtz, 1993). Counsellor Schema 123 The means of the two experience groups (high: X = 39.13, SD = 23.14; low: X = 28.8, SD = 22.05) were noticeably different but did not reach statistical significance. It can be speculated that the low experience group might be unfamiliar with the diagnosis and select lower percentages. Written comments provided by several respondents revealed that the high experience group, out of respect for their clients, were reluctant to impose a psychiatric label, consistent with Rabinowitz's (1993) conclusions. While neither group scored this item in the expected manner, there was a difference that approached statistical significance which is potentially meaningful in terms of suggesting between group differences. Another confound may result from the population sampling bias. The majority of respondents worked in non-profit agency settings which are mandated to provide accessible service. As a result, the clients are generally more marginalized (have fewer resources and less income) than individuals who can afford to pay for private therapy. The literature does indicate that there is a differential response by adults who were sexually abused as children, for a number of reasons. Some are more positively adaptive, such as exhibiting greater competence in overcoming difficulties, or becoming high achievers (e.g., Courtois, 1988; Howe, Herzberger & Tennen, 1988). This latter group may be underrepresented among the client populations of this sample of counsellors and therefore not demonstrate the range of each effect. These results may therefore be partially attributed to sampling biases. This researcher believes that much could be learned by an in-depth item-by-item analysis of this section. While not supportive of the hypothesis related to extremity of schema, the discrete comparisons may allude to differences in schema content as reflected by differences in knowledge and attitudes about the issues and treatment philosophies. Counsellor attitudes, as noted by Adams and Betz (1993) are highly influential in sexual abuse treatment. Differing definitions of sexual abuse, for example, may affect how serious the counsellors perceive the effects to be. A counsellor who believes sexual abuse has significant injurious effects may perceive a client to be fragile and severely impaired. This counsellor might then tend to score all the items that cite negative effects, in a manner that strongly suggests agreement with the effect. A counsellor whose attitudes are reflective of idiosyncratic responses to the experience of childhood sexual abuse, however, may be more moderate in the responses. Although a large proportion of the counsellors in Counsellor Schema 124 this sample subscribe to a client centered theoretical orientation, attitudes are also influenced by colleagues, personal experiences, choice of reading material and so on. As attitudes were not controlled for in this study, they may have had an effect on the results. Although beyond the scope of the present study, such an investigation would be worthy of further exploration. As suggested previously, the finding that the experience groups were not extremely diverse with their responses may indicate that the knowledge base of counsellors is consistent across experience levels. This would be encouraging if it could be concluded that current education and training are addressing the issues; alternatively it would be discouraging if the implication is that experienced counsellors are not benefiting from their experience. Duration of Therapy Participants were asked to predict the duration of therapy for the "client" portrayed in the vignette. A t-test revealed a vignette difference that approached statistical significance: a longer course of treatment was suggested in response to Vignette SA than to Vignette NSA, indicating that counsellors predicted therapy would need to be longer for a client who self-identified as sexually abused than for a client who did not. This finding suggests that counsellors generally believe sexual abuse warrants longer treatment than other kinds of concerns. The obverse of this finding, however, suggests that a client with the same presenting issues who does not identify sexual abuse in her history will not be deemed as needing the same amount of therapy as one who does. This could suggest that counsellors who may draw quick conclusions and use biased hypothesis testing methods (Snyder & Thomsen, 1988), or who lack the schematic organization that permits them to recognize discrepant cues, could overlook an appropriate course of action. The client may therefore not be afforded the type and duration of treatment she requires, and raises questions about treatment planning. This may also be indirectly supported by what Strohmer and Spengler (1993) noted as a disturbing clinician bias: the tendency to elicit negative client information. The course of any treatment plan is contingent on a multitude of individual differences; if clinicians suggest longer treatment for the client in Vignette SA, whose only difference is the disclosure of sexual abuse, it may be evidence of a negative bias Counsellor Schema 125 in the belief that a sexually abused client needs significantly longer therapy. Sexual abuse treatment research to date has not identified definitive answers regarding length of treatment. The reasons for the differences in this study may be due to experience with this client population; alternatively, it may suggest a clinician bias. Further research would help elucidate this finding. There was no experience effect on duration for the combined vignettes sample, which is consistent with the literature on experience and clinical judgement (Wiggins, 1973; Garb, 1989). The duration noted by the counsellors currently working with sexually abused clients was also not significantly different from their non-sexual abuse counsellor counterparts. The mean recommended duration of treatment was considerably higher for the experienced group (X = 17.50 months) than the inexperienced group (X = 12.54 months). The current sexual abuse counsellors also recommended longer treatment (X = 14.97 months) than the non sexual abuse counsellors (X = 11.21 months). Neither result reached statistical significance, but the results suggest possible differences, alluding to the possibility that counsellors with experience with this client population tend to believe longer treatment is warranted. The only comparison group of counsellors distinguished in the duration analysis was the cohort comprised of those who have personal histories of sexual abuse versus those who do not. The group who had experienced abuse themselves was more likely to recommend a longer duration of therapy (p. < .05), a result which is substantiated in the child abuse literature. At least one study has indicated clinicians with their own histories of child abuse (sexual, emotional or physical) viewed the abuse of others as more serious than did those without their own history (Howe, Herzberger, & Tennen, 1988). It can be speculated that a perception of abuse as serious would lead to a recommendation that longer treatment would be necessary to remedy the effects of the abuse. This may have had an effect on the perceptions of counsellors with their own history of child sexual abuse regarding treatment length. When the analysis was conducted for Vignette NSA alone, there again was no experience effect. The group of counsellors with a personal history of childhood sexual abuse did not differ significantly in this analysis from counsellors without sexual abuse in their backgrounds. However, the group of counsellors Counsellor Schema 126 currently working with sexually abused clients predicted a longer duration of therapy, a result which approached statistical significance. As schemas have an effect on counsellor expectations for therapy, these findings suggest that the current sexual abuse counsellors have different schemas than those not working with sexually abused clients, lending indirect support for a difference in sexual abuse schemas between these counsellor groups. As noted elsewhere (see the discussion below on current sexual abuse counsellors), this may also provide indirect support for an experience effect as this cohort is significantly more experienced than the non-sexual abuse counsellors. Current Sexual Abuse Counsellors Post hoc analyses were conducted using, as the basis for comparison, the group of counsellors who identified themselves as currently working with sexually abused adult clients versus those who indicated they were not. Given that there is a possibility of skewed results due to the unequal cell sizes (considerably more counsellors in this sample are working in the field than not), and that these are post hoc analyses, it remains interesting to note the differences between the groups. Current sexual abuse counsellors were significantly more likely than the non-sexual abuse counsellors to identify sexual abuse as a treatment issue in response to the subtle cues of Vignette NSA. In response to the same vignette, this group predicted a longer duration of therapy, a result that approached statistical significance (see previous discussion). These findings are consistent with schema research which discusses the priming effect (Fiske & Taylor, 1984). Schemas are primed and more readily activated when used recently or frequently. Research indicates that schemas mature with experience with a phenomenon (Turk & Salovey, 1985; Higgins & Bargh, 1987). Having a more sophisticated organization of schema complexity that developed from experience increases the likelihood of recognizing discrepant or subtle information. The schemas of the group of counsellors currently working with sexually abused clients, then, are more likely to be in a state of readiness, for ease of access. The results support, and are consistent with, schema research. The findings are also consistent with research regarding variables associated with professionals' detection of sexual abuse. Kendall-Tackett and Watson (1991) found that the strongest effect on counsellors' recognition of a client's history of sexual abuse from behavioural indicators was the counsellors' expectation that sexual Counsellor Schema 127 abuse would be a factor. Consequently, counsellors currently working with this population may be more likely to expect the issue to surface, particularly if they are familiar with statistics about the prevalence of sexual abuse among clinical populations, or work in clinics set up to serve this clientele. These results suggest that further research is warranted to compare counsellors who do and do not regularly work with this population. As noted previously, there was a significant difference on mean ratings of negative adjectives between these two groups of counsellors: current sexual abuse counsellors rated negative adjectives more highly than did non sexual abuse counsellors. This result indicates a difference in counsellor perceptions of sexually abused clients: those working with the population have a greater negative inference about them. The results have been discussed elsewhere. It is worthy of note here because it provides additional evidence that these two counsellor groups have difference schemas regarding sexually abused clients. The group of current sexual abuse counsellors, in comparison with those not working in the field, had significantly more years of general counselling experience (X = 9.4 years for current sexual abuse counsellors compared with X = 4.4 years for non sexual abuse counsellors; p_. < .005) and sexual abuse counselling experience (X = 4.2 years for current sexual abuse counsellors compared with X = 7 months for non sexual abuse counsellors p_. < .001). Perhaps, then, there is indirect support for the hypothesis about an experience effect on schema. Although experience on its own did not produce significant differences, experience combined with current contact with this client population seems to be related to the ability to recognize sexual abuse in response to subtle information. This finding suggests the possibility of an interaction effect between experience and contact with the client population. The results are supported in the schema literature on the effect of expectations: "clinicians' impressions of clients may be guided... by what they expect to observe. Moreover, what they expect to observe is likely to be what they will observe" (Turk & Salovey, 1985, p. 24). Current sexual abuse counsellors are more likely to expect a client to have a history of sexual abuse due to their familiarity with the phenomenon and the frequency with which they see it in practice. As Snyder and Thomsen (1988) explained, "if the client is not a clear-cut case... therapists may categorize the client in terms of the Counsellor Schema 128 potentially applicable categories most cognitively accessible to them" (p. 134). A counsellor whose caseload includes sexually abused clients will readily have this treatment issue available for hypotheses about new clients. In addition, the likelihood of recognizing discrepant cues increases with the amount of perceiver expertise (Higgins & Bargh, 1987). The comparisons between these two groups on the self-reported levels of expertise on relevant counselling issues also yielded significant differences. The current sexual abuse counsellors rated themselves as having more expertise on all items with the exception of "child development." It is heartening to know that counsellors in the field perceive themselves as having a higher level of expertise, which increases the likelihood they will fulfill their ethical obligations (Duncan, 1987b). The results from the comparisons of counsellors currently working in the field versus those who do not must be treated with caution given that they were post hoc and cell sizes were considerably dissimilar. They do, however, provide useful information about the association of regular contact or experience with a client group, and schemas. Current sexual abuse counsellors were more likely to recognize sexual abuse from subtle cues (greater organization of schema) than counsellors not working in the field. The primary implication for counsellors is that those who do not work in the area of sexual abuse may need to gain additional knowledge or understanding of this client population, and the effects of sexual abuse, prior to working with them. Counsellor's Personal Abuse History Clinicians' own history of childhood abuse has begun to receive attention in the literature (Howe, Herzberger, & Tennen, 1988; Pope & Feldman-Summers, 1992; Nuttall & Jackson, 1994). A question about participants' histories was included in this study. Twenty-four (24.7%) counsellors indicated they had been sexually abused as children, which is consistent with surveys on the general population (e.g., Finkelhor, Hotaling, Lewis, & Smith, 1990) and two recent surveys of clinicians (Pope & Feldman-Summers, 1992; Nuttall & Jackson, 1994). Analyses were conducted using these groups of counsellors for comparison purposes. The group who had themselves been sexually abused were, in comparison with counsellors who had not, significantly more Counsellor Schema 129 likely to identify sexual abuse as a treatment issue in response to the subtle cues in Vignette NSA. This result is consistent with Nuttall and Jackson's (1994) finding that clinicians who had their own histories of abuse were significantly more likely to believe sexual abuse allegations. Depending on the effect the experience of abuse has had on the counsellor (Nuttall & Jackson, 1994), and subsequently on the sexual abuse schema, this result may be accounted for by the representativeness heuristic (Turk & Salovey, 1985): if the indicators in the vignette resembled the counsellor's own experience, he or she might be better able to recognize sexual abuse in spite of the disguised presentation. Again, there is support in the schema literature for this interpretation because an individual's experience with a phenomenon or category positively influences the perceiver's ability to recognize it from seemingly unrelated cues (Fiske & Taylor, 1984; Higgins & Bargh, 1987). A counsellor's personal experience with sexual abuse would necessarily contribute to his or her sexual abuse schema. Self schemas may also have relevance for this finding. Individuals have self schemas containing information, derived from past experience, about their self concept (e.g., personality attributes, strengths and weaknesses). These schemas guide one's information processing about themselves as well as aiding memory retrieval (Fiske & Taylor, 1984). In addition to guiding information about the self, self schemas have been found to affect one's perceptions of others (Heppner & Frazier, 1992): people readily recognize traits in others that resemble traits in themselves. Information on self schemas can be applied to the case of counsellors with their own sexual abuse histories being better able than counsellors without that history to identify sexual abuse in someone else. It may be that the person schemas, about sexually abused clients, and the self schemas are activated simultaneously when particular cues are presented, amplifying the likelihood of recognition of sexual abuse. This is speculative and would need to be investigated further to determine how much of the self schema can be generalized to others. In other words, are counsellors who have been sexually abused only able to recognize sexual abuse that directly resembles their own experience or do they generalize beyond their own experience? Counsellor Schema 130 A chi square test did not yield a significant association between history of childhood sexual abuse and experience. The group of clinicians with their own history were therefore not significantly more experienced than counsellors who did not have a history of childhood sexual abuse. The finding that counsellors with a personal sexual abuse history were able to recognize sexual abuse from subtle cues regardless of counselling experience level is noteworthy. This variable could be important for counsellors working in the field and is worthy of further exploration: to quote Nuttall and Jackson (1994), "there may be significant effects on clinical data processing and decision-making" (p. 469) as a result of clinicians' histories of childhood abuse. These findings must be generalized with caution given that these were post hoc analyses. Training Activities Participants had been asked to indicate how many methods they had utilized to learn about sexual abuse treatment. Comparisons of counsellors on the bases of experience, current sexual abuse counselling, and personal histories of childhood sexual abuse were all statistically significant. Counsellors with high experience and those presently working in the field indicated participation in greater numbers of training methods. This result may be a function of duration in the work force (it should be recalled that the current sexual abuse counsellors had significantly more experience). It can also be speculated, however, that counsellors who have more experience in the area recognize the need for specialized training. Conversely, it may be that they work in the field because they have the training. It is not clear why those counsellors with personal histories of sexual abuse were pursuing training in this area. They may be doing so from personal interest, or the result may be associated with variables such as experience or current work focus. Regardless of the explanation, however, it is encouraging to note the amount of attention being paid to training in this field and it can be speculated that there is motivation among counsellors to pursue training in this area. The choices of learning methods did not include university courses other than a category for "practicum," so the majority of activities were outside the realm of formal academics. Pope and Feldman-Summers (1992) had found a significant improvement in the academic training of recent graduates, but the Counsellor Schema 131 current study did not assess whether the focus of formal education includes sexual abuse treatment; conclusions cannot therefore be drawn about the state of academic or formal training for sexual abuse. Limitations Confounds in this study can be grouped into two main categories: limitations to the questionnaire, and sampling biases. As has been stated in the schema literature, the cognitive structure is notably difficult to "measure" because its invisibility is not amenable to direct observation. Any method chosen to evaluate schema, then, may be subject to confounding that is not readily apparent. It is possible, therefore, that the questionnaire did not provide an adequate means of evaluating counsellor schema, and that more sophisticated or alternative methods would be more likely to yield significant results. Construct validity may therefore have been limited. The researcher had not intended to employ this questionnaire as a formal psychological measure but as a series of stimuli through which counsellor schema could be evaluated. The questionnaire, although not validated through rigorous psychometric assessment, was developed in collaboration with counsellors who had a substantial amount of sexual abuse expertise, to enhance construct validity, and does have the benefit of a pilot study. It is research-based in that items regarding sexual abuse effects and the information contained in the vignette had been drawn directly from the existing body of empirical and clinical literature about sexual abuse sequelae. Content validity was therefore enhanced. However, it must be considered that the tasks included in the questionnaire may not have been reliable means of evlauating sexual abuse schemas. There may have been a sampling bias confounding the results. Participants were self-selected, giving rise to questions about the group of counsellors who did not return their questionnaires. More than half the packages were completed which is a reasonable return rate; however, if the sexual abuse focus of the research was apparent through various means (as speculated earlier), clinicians who did not feel sufficiently knowledgeable in the area may have been reluctant to participate even though the explanatory letter underscored the fact that it was not a test. Given the time expectations for completion (minimum 30 minutes), there may have been a motivational component to who participated. Motivation may also have an effect on the individual's willingness to attend training, undertake additional learning, and remain open to Counsellor Schema 132 new ideas, all of which could have had an effect on the responses, and may have overshadowed or obscured an experience difference if the respondents were significantly more motivated in the sexual abuse field, or counselling in general, than those who did not participate. The demographic information requested distinguished work location on the basis of "agency," "institution," "private practice," or "other," but these categories were not further itemized into such classifications as "sexual abuse agency," "family counselling centre," and so on. Because agencies of several types were canvassed, it is impossible to determine which respondent worked in each particular setting. As Rabinowitz (1993) noted, settings "influence judgement because practitioners expect to see specific types of clients in specific places, [and]. . . can anchor clinical judgement" (p. 300). Participants who worked in specialized sexual abuse centers may therefore have been more likely to identify sexual abuse as a treatment issue, regardless of which vignette they received, because the clientele in their particular work environment may have increased the expectation of "seeing" sexual abuse in the vignette. The priming effect may also have been operating in this instance. Those participants working in sexual abuse agencies or with a high proportion of sexually abused clients on their caseloads may have had the expectation that sexual abuse would be an issue, because their schemas were on "standby" due to the frequency and recency of contact with the phenomenon of sexual abuse. Therefore, their sexual abuse schemas may have been activated whether or not sexual abuse was explicitly noted in the vignette. Sampling may also have had an effect on the assignment of respondents to experimental conditions. Due to the nature of the survey (questionnaire mail-out), random assignment to vignette version was not guaranteed, although equal numbers of both vignette versions were mailed out, and were evenly distributed in the return. Furthermore, the decisions made regarding separating respondents into low and high experience groups resulted in the elimination of numerous participants from the data analysis. Having found a significant association between responses to the first question regarding treatment issues on the basis of which vignette the counsellors received, many comparisons were subsequently conducted on participants who returned the non-sexual abuse vignette only. This decision considerably reduced the cell sizes during comparisons of the combined experience groups (from 49 to 28), limiting the power of the Counsellor Schema 133 analyses. Similarly, the size of the "general experience" pool contracted from 70 to 39, and the "sexual abuse experience" groups from 76 to 39. The diminishing cell sizes, therefore, may have decreased the likelihood of finding significant between-groups differences in this study. Counsellors were grouped by experience level according to their responses to the demographic section of the questionnaire. They were asked to indicate "years of counselling experience (general)" and "years of sexual abuse counselling experience" (see Part D of Questionnaire, Appendix D). A confound may have emerged from the manner in which participants responded. Depending on how the counsellor interpreted the questions, there may have been variations in how they categorized their experience. They may have differentiated the two completely, or subsumed the sexual abuse counselling experience within the "general" category as well as adding it separately. Nor does the question provide information on amount of client contact hours, which could vary significantly and would affect the "amount" of the experience. It is unclear whether a counsellor who reports, for example, 10 years of general counselling experience and three years of sexual abuse counselling experience has a total of only 10 years experience or 13. This ambiguity has potentially broad range ramifications for the experience variable and may have confounded the results of the survey. It should be noted, too, that the range of counselling experience had a bimodal distribution, which can confound the results, and was probably due to the inclusion of students with very little or no experience comprising one group, and fairly well experienced counsellors comprising another. The inclusion of students may also have a confounding effect on the results as students may differ from experienced counsellors in ways other than experience, such as their comfort with research due to exposure to it in the university setting, and a particular level of curiosity arising from their learning activities. The homogeneity of the sample, while beneficial for comparison purposes, limits the generalizability of the results of this survey to counsellors with masters level qualifications, working in an agency setting in the Lower Mainland of British Columbia, and who have experience working with sexually abused adult clients. The results of the survey do not further the body of research on the state of counsellor training and education with respect to sexual abuse. The minimal support for an association between experience and Counsellor Schema 134 schema must be limited to counsellors in the field of sexual abuse, and should not be generalized given the possibility of the confounding of the experience variable noted earlier. Future Research Three general areas are identified for future research. Further investigation is clearly indicated to assess the relationship between counsellor experience and schema. Since both variables have significant influences on the therapeutic process, the equivocal results of the schema and experience/clinical judgement literature with respect to an experience effect, and the lack of conclusive results in this survey, strongly support the need for exploration of the variables of counsellor experience and schema, and the relationship between them. Due to the changing knowledge base in the field of sexual abuse treatment, it would be fruitful to explore what is known and what is not, and by whom. This could aid in the curriculum planning of formal academic institutions and counsellor training programs. In general, more empirical research is needed with respect to the counsellors working in this specialized field, the effect of training, education, and experience. It should be noted that three of the counsellors who returned the explicit sexual abuse questionnaire did not identify sexual abuse as a treatment issue. It may be that these counsellors did miss it, which raises questions about why that happened. Or perhaps these counsellors were operating on the premise that they did not want to identify the issue prematurely in spite of the diosclosure. If a similar result were to emerge in future research, it would be worthy and interesting to further explore the underlying reasons. Finally, as recommended by Nuttall and Jackson, (1994), results from the post hoc analyses of counsellors with their own histories of sexual abuse suggest this is a population that warrants further investigation, to ascertain the effect of personal experience with sexual abuse on the knowledge, beliefs, schemas and, ultimately, the therapy provided by a counsellor. Implications for Counsellors Significant differences on the basis of current client load imply that counsellors working in the field of sexual abuse may have more knowledge and a greater ability to detect the existence of sexual abuse from subtle cues. This suggests that novice counsellors may be wise to seek consultation and supervision while Counsellor Schema 135 they are acquiring the requisite knowledge base, to benefit from others' expertise. A recommendation for supervision, when working in this field, is crucial (Daniluk & Haverkamp, 1993). Counsellors should be encouraged to avail themselves of supplementary training to deepen their knowledge (Hibbard & Zollinger, 1990) and to be as fully apprised about the area as possible before undertaking work in this field, as a means of fulfilling their ethical obligations (Duncan, 1987a). Recent research has suggested that counsellors are not neutral with respect to client problem identification (Haverkamp, 1993); this effect may be reflected in the differences on the ratings of negative adjectives between counsellor experience groups and those currently working with sexually abused clients or not. Strohmer and Spengler's (1993) finding that counsellors were more likely to seek negative information resurfaces with respect to this finding In addition, if the sexual abuse effects items were evaluated for accuracy (i.e., comparing counsellors' estimated percentages with what is documented in empirical research, on the items that have such evidence), more information may emerge about the knowledge and attitudes counsellors have about sexually abused clients. Misconceptions or biases may become apparent. It is incumbent upon clinicians to endeavour to reduce sources of bias in their work by increasing their knowledge base, learning to suspend particular frames of reference, and becoming aware of stereotypes that may influence their perceptions. Therapists must become cognizant of their own biases and examine them honestly to ascertain whether they are disproportionately weighted by negative perceptions. The inherent dangers for hypothesis testing and confirmatory bias (Snyder & Thomsen, 1988; Higgins & Bargh, 1987) have been discussed elsewhere. Clinicians need to learn to gather more information and generate alternative explanations prior to seeking information that confirms initial hunches (Snyder & Thomsen, 1988). They should rely less on short cuts such as representative heuristics which affect a clinician's thinking about a client and his or her problems, and "may lead people to ignore other [relevant] information" (Heppner & Frazier, 1992, p. 162). As the majority of participants held similar academic qualifications, a connection between education and schema cannot be made. It appears that most of the specialized learning about sexual abuse may have Counsellor Schema 136 derived from sources other than formal education. There were statistically significant differences for amount of learning activities for counsellors of all types of experience (general, sexual abuse, combined), current sexual abuse counsellors and those counsellors with their own histories of sexual abuse. This difference may be a function of time (the longer in the field, the more time a counsellor will have had to attend training) but does point to an accumulation of training over the course of time. Given the trend towards graduate level education for therapists, there may be some utility in providing additional information and opportunities within course curricula that directly and more thoroughly address sexual abuse and its treatment (Alpert & Paulson, 1990; Frazier & Cohen, 1992). The results of the survey do not lend themselves to a finding of differences between counsellors of low and high experience on schematic organization, breadth or extremity. Characterizations of sexually abused clients by counsellors with more experience in the sexual abuse field did tend towards negative attributes which may suggest the need for experienced counsellors to assess their biases, or for inexperienced counsellors to recognize the gravity of the effect of childhood sexual abuse on adult women. Arising from this research are indications that schema content (i.e., knowledge) may warrant further investigation. The item-by-item analysis of sexual abuse effects indicated that experienced and inexperienced counsellors differed on their ratings, suggesting that knowledge or beliefs may vary across experience levels. A knowledge deficit may be indicated by the finding that 52% of the counsellors receiving the disguised vignette did not identify sexual abuse as a treatment issue. These results may be interpreted as clinicians not integrating the emerging research in their practice or not becoming familiar with the available information in this field. Counsellors may be well advised to remain abreast of the literature and empirical research that continues to emerge. Summary and Conclusions This survey did not yield many anticipated findings with respect to a definitive relationship between counsellor experience and sexual abuse schema. Amount of self-reported counsellor experience was not found to be associated with the schematic dimensions of organization, breadth, or extremity. It had been expected that both specific sexual abuse counselling experience (due to its direct link to this study) and long Counsellor Schema 137 term general counselling experience (due to an increased likelihood of exposure to, and experience with, sexual abuse issues over time) would promote greater organization and breadth, and a reduction in the extremity, of counsellor schema. The findings did not support these hypotheses. However, there was a significantly greater likelihood of sexual abuse being identified in response to the explicit version of the vignette, rather than the disguised version, by all counsellors. This raises questions about the ability (or willingness, in light of the "false memory" controversy) of counsellors to detect the presence of sexual abuse from subtle cues. Furthermore, the disguised version did discriminate among counsellors who currently work with sexually abused clients from those who do not, indicating that the current sexual abuse counsellors seemed more attuned to the indicators. Greater organization (and therefore complexity) of schema, which permits the recognition of a phenomenon from schema-discrepant information, can be inferred. And, because this group was significantly more experienced, it may be support for an interaction effect between experience and current contact with a particular phenomenon or category on schema complexity. Because the respondents receiving the disguised version of the vignette were evenly distributed between those who recognized sexual abuse and those who did not, speculation arises about those variables that may be associated with the ability to recognize sexual abuse cues. Two such factors are whether the counsellor currently worked in the field of sexual abuse and whether the counsellor had his or her own history of sexual abuse. These two influences, then, have a bearing on the counsellor's likelihood of recognizing the presence of sexual abuse from subtle cues. Differences between groups of counsellors compared by experience, current client population and personal history of childhood sexual abuse did not emerge on the measures of schematic breadth or extremity, indicating the sexual abuse schemas did not vary in terms of the number of elements (with respect to characteristics) or the tendency to view sexually abused clients in polarized ways. There was, however, a difference in the amount of negative adjectives ascribed to a sexually abused client. This result suggested that experienced and inexperienced counsellors, and counsellors working and not working with sexually abused clients, attributed differential amounts of negative inference to sexually abused clients. Counsellor Schema 138 The emergence of significant differences on particular individual adjectives also evokes speculation that differences exist. If the sparse findings are extrapolated, it can be surmised that experienced clinicians, by virtue of more contact with this population, may be better acquainted with the commonly presented demeanour of sexually abused clients and hold more realistic views of them. The individual adjective differences, then, may reflect actual presenting client dispositions. This is certainly an avenue worthy of further exploration, particularly given the implications for ethical standards of practice (Daniluk & Haverkamp, 1993). Post hoc analyses of the list of sexual abuse effects (Part C of the questionnaire) suggested the existence of differences in the response patterns among counsellors of varying amounts of experience. Mean scores of eight of the 20 items were significantly different between experienced and inexperienced counsellors, although not all in the expected direction of experienced counsellors rating closer to the mean. It can be speculated that this represents diverse knowledge levels or beliefs held by counsellors. Further discrete assessment of this type may produce interesting findings. Schema content rather than complexity may have been accessed by the sexual abuse effects items as the items were based on research and therefore reflect counsellor knowledge. Again, because of the existence of a difference between ratings of counsellors of different experience levels, there may be variation in knowledge base. Schema content may be an area to further evaluate, as it appears differences may exist on this basis. In summary, no direct association between experience and schema complexity was found in this study. However, sexual abuse schemas were clearly activated in the vignette condition, evidenced by the strong vignette effect, and the differences in responses to the disguised vignette version. A further finding was that current sexual abuse counsellors and counsellors with their own history of sexual abuse surpassed their counterparts in their recognition of sexual abuse treatment issues. In addition, there was a greater tendency for counsellors working with sexual abuse, than those who were not, to attribute negative characteristics to sexually abused clients. The current client focus of counsellors appeared to be a stronger predictor of recognition of sexual abuse as a treatment issue than did general or sexual abuse counselling experience. Counsellor Schema 139 Overall, the results indicate that a more fruitful investigative approach of this topic would be to apportion the complexity question to the schema research field while considering the content of sexual abuse schemas within the rubric of sexual abuse knowledge research. This would more clearly delineate what is being measured: schema content (knowledge) or schema complexity. General exploration of the latter could then be applied to the specific area of sexual abuse schemas. There may, as well, be a link to the literature on attitudes (Adams & Betz, 1993) and stereotypes (Dye & Roth, 1990). It is hoped by this writer that research in that area would investigate not only what the attitudes and stereotypes are, but the characteristics of the counsellors with different views. In sum, this initial investigation offers numerous avenues of further exploration in the fields of counsellor schema, experience, and sexual abuse. Counsellor Schema 140 References Adams, E.M., & Betz, N.E. (1993). Gender differences in counselors' attitudes toward and attributions about incest. Journal of Counseling Psychology. 40(2). 210-216. Alpert, J.L., & Paulson, A. (1990). Graduate-level education and training in child sexual abuse. Professional Psychology: Research and Practice, 21(5). 366-371. Arnold, C.L. (1992). An introduction to hierarchical linear models. Measurement and Evaluation in Counseling and Development. 25. 58-87. Attias, R., & Goodwin, J. (1985). Knowledge and management strategies in incest cases: A survey of physicians, psychologists and family counselors. Child Abuse and Neglect. 9, 527-533. Bass, E., & Davis, L. (1988). The courage to heal. New York: Harper & Row. Blume, E.S. (1990). Secret survivors: Uncovering incest and its aftereffects in women. New York: John Wiley & Sons. Brenner, D., & Howard, K. (1976). Clinical judgment as a function of experience and information. Journal of Clinical Psychology. 32(3), 721-8. Briere, J., & Conte, J. (1993). Self-reported amnesia for abuse in adults molested as children. Journal of Traumatic Stress, 6(1), 21-31. Briere, J., & Runtz, M . (1993). Childhood sexual abuse: Long-term sequelae and implications for psychological assessment. Journal of Interpersonal Violence, 8(3), 312-330. Burt, M.R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology. 38,217-230. Busby, D.M., Glenn, E., Steggell, G.L., & Adamson, D.W. (1993). Treatment issues for survivors of physical and sexual abuse. Journal of Marital and Family Therapy, 19(4), 377-392. Ciccone, B.S. (1982). The development of an instrument to study incest attitudes (Doctoral dissertation, Temple University, 1981). Dissertation Abstracts International. 42, 4743B. Clavelle, P.R., & Turner, A.D. (1980). Clinical decision-making among professionals and paraprofessionals. Journal of Clinical Psychology. 36(3). Collings, S.J., & Payne, M.F. (1991). Attribution of causal and moral responsibility to victims of father-daughter incest: An exploratory examination of five factors. Child Abuse and Neglect. 15, 513-521. Courtois, C.A. (1988). Healing the incest wound: Adult survivors in therapy. New York: W.W. Norton & Company. Crewdson, J. (1988). By silence betrayed: Sexual abuse of children in America. New York: Harper & Row. Counsellor Schema 141 Daniluk, J.C., & Haverkamp, B.E. (1993). Ethical issues in counseling adult survivors of incest. Journal of Counseling and Development. 72(5). 16-22. Doughty, D.L., & Schneider, H.G. (1987). Attribution of blame in incest among mental health professionals. Psychological Reports. 60. 1159-1165. Duncan, L. (1987a). The preparation of counselors to treat sexual abuse. Texas Association for Counseling and Development Journal. 15. 61-69. Duncan, L. (1987b). The preparation of counselors to treat sexual abuse: Part II. Texas Association for Counseling and Development Journal. 15(2), 151 -160. Dye, E., & Roth, S. (1990). Psychotherapists' knowledge about and attitudes toward sexual assault victim clients. Psychology of Women Quarterly. 14. 191-212. Emerson, S. (1988). Female student counselors and child sexual abuse: Theirs and their clients'. Counselor Education and Supervision. 28, 15-21. Enns, C.Z., McNeilly, C.L., & Gilbert, M.S. (1995). The debate about delayed memories of child sexual abuse: A feminist perspective. The Counseling Psychologist. 23(2), 181-279. Feldman-Summers, S., & Pope, K.S. (1994). The experience of "forgetting" childhood abuse: A national survey of psychologists. Journal of Consulting and Clinical Psychology. 62(3), 636-9. Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York: The Free Press. Finkelhor, D., Hotaling, G., Lewis, I.A., & Smith, C. (1990). Sexual abuse in a national survey of adult men and women: Prevalence, characteristics, and risk factors. Child Abuse and Neglect. 14, 19-28. Fiske, S.T., & Taylor, S.E.( 1984). Social cognition. New York: Random House. Frazier, P.A., & Cohen, B.B. (1992). Research on the sexual victimization of women: Implications for counselor training. The Counseling Psychologist. 20. 141-158. Friedrich, W.M. (1990). Psychotherapy of sexually abused chidren and their families. New York: W.W. Norton & Company. Garb, H.N. (1989). Clinical judgment, clinical training, and professional experience. Psychological Bulletin. 105(3). 387-396. Gelso, C.J. (1978). Research in counseling: Methodological and professional issues. Counseling Psychologist. 8(3), 7-35. Gough, H.G., &Heilbrun, A.B.,Jr. (1980). The adjective check list. Palo Alto, CA: Consulting Psychologists Press. Counsellor Schema 142 Grossman, B., Levine-Jordano, N., & Shearer, P. (1990). Working with students' emotional reactions in the field: An educational framework. The Clinical Supervisor. 8, 23-39. Haverkamp, B.E. (1993). Confirmatory bias in hypothesis testing for client-identified and counselor self-generated hypotheses. Journal of Counseling Psychology. 40(3) 303-315. Heppner, P., & Frazier, P A . (1992). Social psychological processes in psychotherapy: Extrapolating basic research to counselling psychology. In S.D. Brown & R.W. Lend (Eds.) Handbook of Counseling Psychology. 2nd. Ed. (pp. 141-175). N.Y.: John Wiley & Sons. Hibbard, R.A., & Zollinger, T.W. (1990). Patterns of child sexual abuse knowledge among professionals. Child Abuse and Neglect. 14. 347-355. Higgins, E.T., & Bargh, J.A. (1987). Social cognition and social perception. Annual Review of Psychology. 38, 369-425. Hillerbrand, E., & Claiborn, C D . (1990). Examining reasoning skill differences between expert and novice counselors. Journal of Counseling and Development. 68, 684-691. Holloway, E.L., & Wolleat, P.L. (1980). Relationship of counselor conceptual level to clinical hypothesis formation. Journal of Counseling Psychology. 27(6), 539-545. Howe, A.C., Herzberger, & Tennen, H. (1988). The influence of personal history of abuse and gender on clinicians'judgments of child abuse. Journal of Family Violence. 3(2). 105-119. Jackson, T.L., & Ferguson, W.P. (1983). Attribution of blame in incest. American Journal of Community Psychology. 11(3), 313-322. Kandel, M. , & Kandel, E. (1994). Flights of memory. Discover. May 1994, 32-38. Kendall-Tackett, K.A. & Watson, M.W. (1991). Factors that influence professionals' perceptions of behavioral indicators of child sexual abuse. Journal of Interpersonal Violence. 6(3), 385-395. Mallinckrodt, B., & Nelson, M.L. (1991). Counselor training level and the formation of the psychotherapeutic working alliance. Journal of Conseling Psychology, 38(2), 133-138. Nuttall, R., & Jackson, H. (1994). Personal history of childhood abuse among clinicians. Child Abuse and Neglect. 18(5), 455-472. Parisien, L.P., & Long, B.C. (1994). Counselor trainees' self-statement responses to sexually and physically abused clients, and client role conflict. Journal of Counseling and Development. 72(1), 304-309. Pope, K.S., & Feldman-Summers, S. (1992). National survey of psychologists' sexual and physical abuse history and their evaluation of training and competence in these areas. Professional Psychology: Research and Practice, 23(5), 353-361. Counsellor Schema 143 Rabinowitz, J. (1993). Diagnostic reasoning and reliability: A review of the literature and a model of decision-making. The Journal of Mind and Behavior. 14(4). 297-316. Reidy, T.J., & Hochstadt, N.J. (1993). Attribution of blame in incest cases: A comparison of mental health professionals. Child Abuse and Neglect. 17. 371-381. Rush, F. (1980). The best kept secret: Sexual abuse of children. New York: McGraw-Hill. Reynolds-Mejia, P. & Levitan, S. (1990). Countertransference issues in the in-home treatment of child sexual abuse. Child Welfare. 69. 53-61. Sgroi, S.M. (1982). Handbook of clinical intervention in child sexual abuse. Lexington, Mass.: D.C. Heath and Company. Sgroi, S.M., Blick, L.C., & Porter, F.S. (1982). A conceptual framework for child sexual abuse. In S.M. Sgroi (Ed.), Handbook of clinical intervention in child sexual abuse, (pp. 9-27). Lexington, Mass.: D.C. Heath and Company. Smith, E.R., & Zarate, M.A. (1990). Exemplar and prototype use in social categorization. Social Cognition. 8(3), 243-262. Snyder, M. , & Thomsen, C.J. (1988). Interactions between therapists and clients: Hypothesis testing and behavioral confirmation. In D.C. Turk & P. Salovey (Eds.), Reasoning, inference.and judgment in clinical psychology (pp. 124-152). New York: Free Press. Spengler, P.M., & Strohmer, D.C. (1993). Counselor complexity and clinical judgment: Challenging the model of the average judge. Poster presented at the 101st meeting of the American Psychological Association, Toronto, Canada. Stone, M.H. (1989). Individual psychotherapy with victims of incest. Psychiatric Clinics of North America. 12(2), 237-255. Strohmer, D.C, & Spengler, P.M. (1993). Clinical judgment biases and clinical susceptibility: Anew alternative to the old model of the average judge. Paper presented at the 1993 American Psychological Association annual convention, Toronto, Canada. Treating Abuse Today (1994). 1994-1995 Survey of abuse and trauma therapists: Current clinical practice. Treating Abuse Today. 4(6), 22-26. Turk, D.C, & Salovey, P. (1985). Cognitive structures, cognitive processes, and cognitive-behavior modification: II. Judgments and inferences of the clinician. Cognitive Therapy and Research. 9. 19-33. Ward, M.A. (1980). Attribution of blame in rape (Unpublished doctoral dissertation, University of South Dakota, 1980). Dissertation Abstracts International. 41(5). 1934B. Weiss, J.H. (1963). Effect of professional training and amount and accuracy of information on behavioral prediction. Journal of Consulting Psychology. 27(3), 257-262. Counsellor Schema 144 Wylie, M.S. (1993, September/October). The shadow of a doubt. Networker. 18-29. Counsellor Schema 145 Appendix A Questionnaire Part A: Sexual Abuse Vignette Questionnaire Part A: Non Sexual Abuse Vignette i Hb This questionnaire will take approximately one-half hour of your time. While answering the different sections, please do not look ahead to subsequent pages, nor go back to earlier parts to change what you have written. This questionnaire contains a number of questions that ask for your impressions of clients. We recognize that no population can be described in absolute terms as clients have tremendous individual differences. .However, for this research, your best estimate of trends or characteristics is requested. Your first impression is probably the best to record. Thank you. PART A Based on the information provided in this "intake" of a hypothetical client, please answer the questions that follow. The client is a 33-year-old married woman with two children: a five-year-old son and an infant daughter. She has chosen to remain home until her children begin school. Her husband, 36, works full time. Their marriage of seven years is characterized by mutual respect, friendship and genuine caring. The client had left home when she was 16 years old. At that time she used alcohol and marijuana extensively and withdrew from school for about a year. Her lifestyle changed when she was 25 and her stepfather died. She became intensely involved in her education and career. The client has requested counselling for herself because she has recently experienced unexplainable crying spells, sleeps fitfully (although the baby sleeps through the night), has outbursts of rage directed at her husband, has lost interest in their sexual relationship, and describes what sound like panic attacks when she takes her children out for errands, fuelling a reluctance to leave the home. She is becoming concerned about her ability to care for her children. She would like to regain control over her anger, be able to overcome her anxiety about leaving the house, and rejuvenate the intimacy in her relationship with her husband. The client divulges that she has begun remembering childhood episodes of sexual attention by her stepfather from the age of five until her parents separated when the client was 11. She recounts having been touched sexually and being forced to touch her stepfather. She fears she may have more memories that she has not yet recalled. 1) What have you identified as the client's primary treatment issues (i.e., what will be the focus of therapy)? 2) What clues in the information above lead to your choice of issues? 3) What is your estimate of the anticipated duration of therapy? ) L\~± This questionnaire will take approximately one-half hour of your time. While answering the different sections, please do not look ahead to subsequent pages, nor go back to earlier parts to change what you have written. This questionnaire contains a number of questions that ask for your impressions of clients. We recognize that no population can be described in absolute terms as clients have tremendous individual differences. However, for this research, your best estimate of trends or characteristics is requested. Your first impression is probably the best to record. Thank you. PART A Based on the information provided in this "intake" of a hypothetical client, please answer the questions that follow. The client is a 33-year-old married woman with two children: a five-year-old son and an infant daughter. She has chosen to remain home until her children begin school. Her husband, 36, works full time. Their marriage of seven years is characterized by mutual respect, friendship and genuine caring. The client had left home when she was 16 years old. At that time she used alcohol and marijuana extensively and withdrew from school. Her lifestyle changed when she was 25 and her stepfather died. She became intensely involved in her education and career. The client has requested counselling for herself because she has recently experienced unexplainable crying spells, sleeps fitfully (although the baby sleeps through the night), has outbursts of rage directed at her husband, has lost interest in their sexual relationship, and describes what sound like panic attacks when she takes her children out for errands, fuelling a reluctance to leave the home. She is becoming concerned about her ability to care for her children. She would like to regain control over her anger, be able to overcome her anxiety about leaving the house, and rejuvenate the intimacy in her relationship with her husband. 1) What have you identified as the client's primary treatment issues (i.e., what will be the focus of therapy)? 2) What clues in the information above lead to your choice of issues? 3) What is your estimate of the anticipated duration of therapy? Counsellor Schema 148 Appendix B Questionnaire Part B: Characteristics PART B Listed below are words that may be descriptive of people. Recognizing that generalizing is difficult, please consider a sexually abused female client and circle the letter which represents whether, in your opinion, the characteristic is never likely, rarely likely, sometimes likely, or frequently likely to be shown by a sexually abused client. Key: N=Never; R=Rarely; S=Sometimes; F=Frequently absent-minded N R S F fair-minded N R S F adaptable N R S F fearful N R S F affectionate N R s F foresighted N R S F aggressive N R s F forgetful N R S F alert N R s F gentle N R S F angry N R s F giving N R s F anxious N R s F gloomy N R s F artistic N R s F guarded N R s F ashamed N R s F helpful N R s F awkward N R s F high-strung N R s F bitter N R s F hostile N R s F calm N R s F hyper-vigilant N R s F candid N R s F imaginative N R s F careless N R s F impulsive N R s F caring N R s F inhibited N R s F confident N R s F insecure N R s F confused N R s F insightful N R s F conscientious N R s F irresponsible N R s F contented N R s F isolated N R s F courageous N R s F kind N R s F creative N R s F loving N R s F curious N R s F loyal N R s F cynical N R s F mistrustful N R s F deceitful N R s F moody N R s F distractible N R s F optimistic N R s F dominant N R s F original N R s F evasive N R s F outgoing N R s F explosive N R s F perceptive N R s F PART B persevering N R S F pessimistic N R S F proud N R c F rational N R S F realistic N R S F rebellious N R s F reflective N R s F reliable N R s F resentful N R s F reserved N R s F resilient N R s F resourceful N R s F rigid N R s F sad N R s F scared N R s F self-confident N R s F self-destructive N R s F self-reliant N R s F silent N R s F sincere N R s F sociable N R S F spontaneous N R s F stable N R •s F strong N R s F submissive N R s F suggestible N R s F suspicious N R s F thorough N R s F thoughtful N R s F tolerant N R s F understanding N R s F unemotional N R s F unkind N R s F unrealistic N s F unstable N R s F warm N R s F wary N R s F weak N R s F wise N R s F withdrawn N R s F Counsellor Schema 151 Appendix C Questionnaire Part C: Sexual Abuse Effects PART C Circle the number on the line graph which indicates the percentage of adult survivors of childhood sexual abuse you think who would show the following effects (e.g., 0 = none; 100% = all). You may use the space in the column at the right to explain your answer or provide comments on the item itself. 1. Trust others indiscriminately 2. Avoid physical touch 3. Wear provocative clothing 4. Are obsessive about personal hygiene 5. Are high achieving in work or school 6. Deny or are unable to identify negative feelings 7. Abuse alcohol or drugs (prescription or non-prescription) 8. Have dissociative episodes 9. Have had a positive body experience or feelings towards the offender at that time NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% 10. Attempt to anticipate the needs of others NONE ALL 0 10 20 30 40 50 60 70 80 90 100% PART C ' ^ 11. Diagnosis of borderline personality disorder NONE _ — A L L 12. Readily trust their counsellors as potentially dangerous their own children 15. Believe everyone "can tell" they were sexually abused 16. Are sexually promiscuous 17. Fail to remember the abuse until adulthood 18. The sexual abuse involved physical force NONE 19. Cannot tolerate being alone 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100%  ALL 0 10 20 30 40 50 60 70 80 90 100% NONE ALL 0 10 20 30 40 50 60 70 80 90 100% 20. Report having attempted suicide on at least one occasion NONE ALL 0 10 20 30 40 50 60 70 80 90 100% Counsellor Schema 154 Appendix D Questionnaire Part D: Demographics PART D Background Information The following questions are designed to provide background information on the people who respond to this questionnaire. We are trying to make sure we have included the perspectives of counsellors with a wide variety of backgrounds. Please answer as honestly as you can, and be assured that confidentiality will be maintained and that your answers will remain anonymous. Thank you for taking the time to complete this part of the questionnaire. Age: Years of counselling experience (general): Years of sexual abuse counselling experience: Work site (please check any that are appropriate): Agency • Yes • No Institution • Yes • No Private practice • Yes • No Other • Yes • No Gender: • Male • Female Duration: Duration: Duration: Duration: Theoretical orientation (check all that apply): • Adlerian • Client-centred • Cognitive • Existential • Experiential • Feminist • Gestalt • Solution focused/brief therapy • Systemic • Other: Highest degree attained: • High School • College • M.S.W. • Ph.D. (or equivalent) • Diploma • BA. • M.A. • M.Ed. • Other: Training programs completed (e.g., psychiatric nursing, counsellor training, clinical training, social services training programs etc.). Please list: Are there additional degrees, certificates or diplomas have you acquired? • Yes • No If yes, please specify: Do you belong to a professional counselling organization? • Yes • No If yes, please specify: On what percentage of cases do you seek supervision or consultation? What type? (Check all that apply) • individual supervision • agency team supervision • group networking • peer supervision • professional consultation • other: 0/ /o PART D Please circle the appropriate number to indicate the degree of your knowledge and skills on the following counselling issues, where zero (0) indicates no training or experience, and ten (10) indicates expertise in the area: adult counselling gender and power issues in families human sexuality legal and ethical issues normal child development sexual abuse substance abuse trauma novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert novice 0 1 2 3 4 5 6 7 8 9 10 expert Do you currently work with adult clients who have a history of sexual abuse? • Yes • No How did you learn to work with clients who have been sexually abused? • consultation • films, videos, brochures, manuals • formal supervision • in-house training programs • information networks • lay counselling • learning from adult clients • other (please specify) • own experience of abuse • peer supervision • personal therapy • practicum placements • reading relevant journals, books, articles • special course work on own time • workshops/seminars outside of workplace This final question has been included because the therapist's realm of personal experience is a potent source of knowledge and insight and is relevant to the work you do. . Do you yourself have a history of? • childhood physical abuse • childhood emotional abuse • childhood neglect • childhood sexual abuse • adult physical assault • adult sexual assault • none of the above Many thanks for taking the time to respond to this questionnaire. I truly appreciate your involvement. Your participation in my research is making possible a valuable contribution to our knowledge of counsellors and how they do their very important work If you wish a summary of the results of the research, please complete the bottom portion of your consent form and send it to me in a separate envelope. Counsellor Schema 157 Appendix E Agency Letter Regarding Pilot Study Counsellor Schema 159 Appendix F Agency Letter Regarding Research Study Counsellor Schema 161 Appendix G Participant Package: Introductory Letter Informed Consent Incentive Information Debriefing Letter THANK YOU !!!! . . . for your help!!! In appreciation, I am inviting you to enter your name for a draw prize of: Dinner for two at Bridges Restaurant on Granville Island (value $50.00) To enter, please write your name and phone number in the space provided, tear off on the dotted line, and seal it in the attached envelope. Return the envelope with your questionnaire. Be assured that it will be separated from the questionnaire immediately, so I cannot match your name to the questionnaire. Only one envelope will be opened, at the time of the draw, around the end of 1994. Good luck and thanks again. Name Telephone number T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A Department of Counselling Psychology Faculty of Education 5780 Toronto Road Vancouver, B . C . Canada V 6 T 1L2 Tel: (604) 822-5259 Fax: (604) 822-2328 Dear Counsellor Many thanks for taking the time to complete this questionnaire. By now you win have realized that my particular area of focus is sexual abuse. This was not clearly identified at the outset because some of die "case histories" included a final paragraph indicating the "client" disclosed childhood sexual abuse; others did not This was done for the purpose of learning which informational dues are most readily recognizable as sexual abuse indicators. The remainder of the questionnaire provided different means for determining how different counsellors conceptualize clients with sexual abuse histories. I currently work as a counsellor in a sexual abuse counselling centre in Vancouver, the Vancouver Incest and Sexual Abuse Centre (VISAC). I have become aware of how, in this emerging field, we are continually presented with new statistics, knowledge and treatment approaches. I began to wonder how wen we integrate this in our practice and how counsellors with varying amounts of experience and working in different therapeutic venues (yet suH encountering sexual abuse issues) may differ in their ways of viewing clients as the knowledge base changes so rapidly. I believe the results of this study wfll provide valuable information about counseDor conceptualizations in the field of sexual abuse, which may have implications for the type of training that would be valuable. Once again I would like to remind you that a summary of results wfll be made available upon completion of the research, if you would be interested in knowing what I have learned. I greatly appreciate your contribution to my research. Many thanks for your time. Sincerely, Heidi Sheehan Counsellor Schema 166 Appendix H Participant Follow Up Letter Counsellor Schema 168 Appendix I Ethical Approval 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0054137/manifest

Comment

Related Items