UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A decision making model of child abuse reporting Beck, Kirk A. 2000

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

Item Metadata

Download

Media
831-ubc_2000-486044.pdf [ 7.36MB ]
Metadata
JSON: 831-1.0089769.json
JSON-LD: 831-1.0089769-ld.json
RDF/XML (Pretty): 831-1.0089769-rdf.xml
RDF/JSON: 831-1.0089769-rdf.json
Turtle: 831-1.0089769-turtle.txt
N-Triples: 831-1.0089769-rdf-ntriples.txt
Original Record: 831-1.0089769-source.json
Full Text
831-1.0089769-fulltext.txt
Citation
831-1.0089769.ris

Full Text

A DECISION MAKING MODEL OF CHILD ABUSE REPORTING by KIRK A. BECK BA. , Simon Fraser University, 1990 M.A., Simon Fraser University, 1993 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies DEPARTMENT OF EDUCATIONAL AND COUNSELLING PSYCHOLOGY, AND SPECIAL EDUCATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March 2000 © Kirk A. Beck, 2000 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. D e p a r t m e n t o f ^ f r / W W / f r W ^ fsujj^ , ^  UmJhi The University of British Columbia Vancouver, Canada Date A/^uX n , MCd DE-6 (2/88) Abstract This study applied Ethnographic Decision Tree Modeling (Gladwin, 1989) to the field of child abuse reporting to investigate the factors that influence decisions to report possible child abuse. Participants were licensed psychologists in the lower mainland of British Columbia. Using ethnographic interviews, participants were asked to discuss a recent case in which they reported possible child abuse and the factors that were influential in their decision making. Based on the data from 34 cases, six factors were identified: (1) Were there any signs of or risk factors for child abuse or neglect? (2) Did the signs or risk factors meet your threshold to report as you understand the law? (3) Was there some other value to report other than a legal one? (4) Were you concerned that reporting would cause harm? (5) Were you able to minimize the harm that would result from reporting? and (6) Did the reasons to report outweigh the reasons to not report? These six factors were presented in a decision tree to illustrate the relationship between factors and decision outcome. This preliminary group model was then tested using the case experiences of a separate yet similar group of registered psychologists in British Columbia. Results found that the preliminary model accurately predicted the reporting outcome of 93% (33 of 36) of the cases in the new sample. Errors in the model were identified, and suggestions were made to improve its predictive ability. The results are evaluated in light of the decision tree produced. Implications for theory building, naturalistic inquiry, clinical practice, policy, and future research are discussed. Table of Contents Abstract ii List of Tables v Acknowledgements vi CHAPTER I Introduction 1 CHAPTER II Review of the Literature 4 Child Abuse Reporting Laws 4 Duty to Report in British Columbia 6 Reporting Laws Across Canada 7 Case Law 11 Professional Reactions 13 Ethics 16 Models of Ethical Decision Making 20 Decision Making 24 Mandatory Reporting of Child Abuse 36 CHAPTER III Methodology 47 Phase 1: Model Development 47 Phase 2: Model Validation 58 CHAPTER IV Results 63 Phase 1: Model Development 63 Phase 2: Model Validation 97 CHAPTER V Discussion 121 References 142 Appendix A The Child, Family and Community Service Act of B.C 152 Appendix B Child Abuse Legislation in Canada 155 Appendix C Child Protection in Canada 157 Appendix D The BC Psychologist Newsletter 159 iii Appendix E Initial Contact Letter: Model Development Stage 161 Appendix F Follow-up Letter: Model Development Stage 163 Appendix G Informed Consent Form 165 Appendix H Initial Contact Letter: Model Validation Stage 167 Appendix I Questionnaire 169 iv List of Tables Tables 1. Demographic Characteristics of Participants: Model Development Stage . 48 2. Demographic Characteristics of Participants: Model Validation Stage 59 3. Model Development Data 67 4. Types of Maltreatment Reported Model Development Stage 68 5. The Nature of Child Abuse Reported: Model Development Stage 69 6. Emotional and Behavioural Indicators: Model Development Stage 73 7. Source of Verbal Disclosure: Model Development Stage 74 8. Risk factors for Potential Abuse: Model Development Stage 80 9. Types of Maltreatment Reported: Model Validation Stage 98 10. The Nature of Child Abuse Reported: Model Validation Stage 99 11. Source of Verbal Disclosure: Model Validation Stage 102 12. Risk factors for Potential Abuse: Model validation Stage 105 13. Model Validation Data 117 V Acknowledgements First, I would like to express my appreciation to my research committee. To Dr. Haverkamp, I am grateful for her ongoing support and encouragement of the research project and the time that she has offered to this project. To Dr. Cochran, I thank for his contribution to the early developments of this project and for his guidance with respect to the methodology used in this study. My thanks also go to Dr. Ogloff for his wisdom and his encouragement to do research in the area of psychology and law. Second, I am grateful to the psychologists who participated in this study. Without their interest to participate, this research would not have been possible. Third, I am also very grateful to my family for their unfailing support of me as a person in whatever goal I have sought out. I am thankful for my friends who have kept me humble and have taught me a lot about striking balance. Finally, I wish to thank my wife and best friend for her support of my academic pursuits and for me as a person. She was always there when I needed her. She has my warmest gratitude and appreciation. This dissertation was supported in part by a scholarship from the Social Sciences and Humanities Research Council of Canada (SSHRCC: 752-96-1922) awarded to the author. vi Chapter I Introduction During the last two decades, the number of children reported abused or neglected has risen steadily (Avis, 1992). In British Columbia, the Ministry for Children and Families received over 33,000 reports of alleged abuse in the 1998/1999 fiscal year (British Columbia's Ministry for Children and Families Annual Report, 1999). This statistic represents an increase of approximately 6,000 cases when compared to 1993/1994. Given that many cases of maltreatment are never detected, including situations of marginal physical abuse and psychological maltreatment, these figures may represent the "tip of the iceberg." In an effort to protect children from maltreatment, mandatory reporting laws have been enacted across Canada and the United States. These laws require professionals and, in many cases, the general public to report suspected abuse to child protection authorities. Within the helping professions, psychologists play an important role in the identification, reporting, and treatment of abused children, yet studies suggest that psychologists selectively report cases of child abuse (Barksdale, 1988; Beck & Ogloff, 1995; Finkelhor, Gomes-Schwartz, & Horowitz, 1984; Haas, Malouf, & Mayerson, 1988; Kalichman & Craig, 1991; Kalichman, Craig, & Follingstad, 1989, 1990; Swoboda, Elwork, Sales, & Levine, 1978; Williams, Osbourne, & Rappaport, 1987). Researchers have identified a number of factors that influence clinicians' decision making process. Some of these factors include the severity of abuse (Zellman, 1992), the type of abuse (Beck & Ogloff, 1995; Williams et al., 1987), years of professional experience (Barksdale, 1988), clinicians' expectations of reporting consequences on the individual or family (Kalichman, Craig, & Follingstad, 1989), age of the child (Kalichman & Craig, 1991), knowledge of the law (Beck & Ogloff, 1995; Swoboda et al., 1978), and clarity of legal requirements (Besharov, 1991). Given the large numbers of variables identified, it 1 remains unclear how any particular variable combination may affect reporting decisions in specific circumstances. The purpose of the present study was to identify the factors that psychologists use in their decisions to report cases of possible child abuse. Furthermore, a focus of this study was to develop a model that represents the relationship between decision criteria and outcome. The findings generated from this study have both practical and theoretical significance. Wi th regard to practice, the accuracy o f reporting can be improved once decision criteria are made explicit (Kalichman, 1993). With regard to theoretical development, the findings of this study can serve as a basis to expand existing conceptual models, and to develop new frameworks that attempt to describe the factors that influence child abuse reports. In order to help ensure that children receive the care and nurturing they need to develop, it is critical to continue to pursue knowledge of the factors involved in reporting decisions. A s this study w i l l be the first of its kind in Canada, its contributions to the area of child abuse reporting w i l l be original. The present study was designed to answer the following question, "What factors contribute to psychologists' decisions to report cases of possible child abuse?" To answer this question, a qualitative study was conducted. This specific method that was used is called Ethnographic Decision Tree Modeling ( E D T M , Gladwin, 1989). The focus of E D T M is on the development and verification of a decision tree model that can be used to explain real life decisions. A decision tree model is a graphic way o f showing the relationship between decision criteria and outcome. In the present study, data were collected from a group of registered psychologists in the Lower Mainland of British Columbia. Through semi-structured interviews, psychologists were asked to specify the criteria that they used in reporting actual cases of possible child abuse. The data from 34 cases were combined to develop a preliminary group model that described the key factors involved in reporting decisions. After the 2 group model was constructed, a questionnaire was developed to verify the model. During this stage, a separate yet similar group of registered psychologists was sampled to test the preliminary group model. The results, as will be described later, found that the preliminary group model accurately predicted the reporting outcome of 93% (33 of 36) of the cases in the new sample. This dissertation is divided into a number of key sections. In Chapter 2, the literature on child abuse reporting will be provided. It will include a review of child abuse reporting laws, professional reactions to reporting laws, codes of ethics and related principles, ethical decision making models, research on decision making, and studies in the field of child abuse reporting. Chapter 3 describes the EDTM methodology used in this study. The results of this study are presented in Chapter 4. Chapter 5 will discuss the decision tree model and its implications for theory building, naturalistic research, clinical practice, policy, and future research. 3 Chapter II Review of the Literature This chapter reviews various aspects of mandated reporting of child abuse. First, the purpose and history of reporting legislation are reviewed, along with details of British Columbia's reporting statute, case law, and professional reactions to mandatory reporting. Second, professional ethics are discussed, including the relevant codes for psychologists that relate to mandatory reporting. Third, several prominent models for ethical decision making are reviewed. Fourth, the literature on decision making theory is summarized. This section focuses on various biases and inconsistencies in the way people use information to make decisions about risk. Fifth, research on failure to report child abuse is presented. This chapter concludes with an analysis of the limitations of previous studies and advances a rationale for the present study. Child Abuse Reporting Laws Purpose Child protection legislation is designed to protect and enhance the health and welfare of children. When parents fail in their duty to protect children, the state may intervene in the child's best interest. In many cases, child protection social workers offer supportive services to the family, such as counselling, in-home support, and respite care. If no other measures are available or adequate to ensure the child's safety, the child may be removed from the home. A provincial court judge makes this decision. In addition to providing safeguards for children, provincial statutes are aimed at protecting the integrity of the family unit. Reporting provisions recognize that the family is the preferred environment for the care and upbringing of children and that the state should adopt the least intrusive means of intervention to promote the child's best interest 4 (Canadian Family Law Guide, 1988). However, the state must intervene when parents are unable or unwilling to provide protection and safety for their children. History Although children have been the victims of violence and neglect for many centuries, mandatory child abuse reporting laws have been in existence only since the early 1960's. The catalyst for the development of reporting legislation came from medical descriptions of non-accidental physical injuries to children. In particular, the introduction of the term, "battered child syndrome" (Kempe, Silverman, Steele, Droegemueller, & Silver, 1962) attracted unprecedented professional and public attention to the prevalence of child abuse. In their paper, Kempe and his colleagues described children who suffered multiple bone fractures, soft tissue injuries, malnutrition and trauma uncharacteristic of accidental injuries. In addition, these authors discussed the reluctance of physicians to believe that parental abuse may be the cause of a child's injury, or to report suspicions of abuse to the proper authorities. Following the publication of Kempe's research, initiatives were made to develop model mandatory reporting legislation (Kalichman, 1993). The purpose of the legislation was to bring forward cases of the battered child syndrome (Paulsen, 1967). Thus, initial reporting laws applied only to medical professionals. Over the years, the persons mandated to report have expanded. Many reporting statutes today stipulate that mental health professionals, as well as all members of the public, are required to report when they believe that a child is in need of protection. Many other features of mandatory reporting have undergone revision over the past 20 years. British Columbia's reporting law is an example of how reporting laws have evolved. Given that this study examines child abuse reporting in British Columbia, the reporting law for this province is discussed in the next section. 5 Duty to Report in British Columbia Given that the nature of this study is on the reporting of child maltreatment, it is necessary to review B.C.'s child abuse reporting law. This law can have considerable influence on the behaviour of psychologists. The legal basis for mandatory reporting in British Columbia is outlined in section 14 of the Child, Family and Community Service Act (1996). This Act is presented in Appendix A. According to the Child, Family and Community Service Act (referred to hereafter as the C.F.C.S.A.), any person is required to report child abuse in accordance with its provisions. This duty applies to members of the general public as well as to professionals. As stated in subsection (2)(a)(b), the obligation of professionals to report supersedes a claim of confidentiality, with the exception that attorneys are exempt when there is a claim of privilege. The circumstances that require a report are broadly defined in the C.F.C.S.A. A child needs protection when he or she has been, or is likely to be, physically harmed, sexually abused or exploited, or physically harmed because of neglect by the child's parent, or where the child has been, or is likely to be, physically harmed, sexually abused or exploited by another person and the child's parent is unwilling or unable to protect the child. Significantly, emotional harm is recognized in this act, along with the criteria that define it. The degree of certainty required for a report is "reason to believe" that a child needs protection. This obligation must be fulfilled promptly. Mandated reporters are shielded from civil or criminal liabilities that otherwise might result from reports, unless a person knowingly reports false information that a child needs protection. Failure to comply with the provisions of the statute constitutes an offense, with a maximum penalty of a $10,000.00 fine and/or six months' imprisonment. Section 14(7) of the C.F.C.S.A. specifies 6 that the six-month time period for prosecution under the Offense Act (1979) is inapplicable for non-compliance with the duty to report. A Comparison of Reporting laws Across Canada In this section, the relevant provisions of provincial and territorial statutes dealing with reporting child abuse will be summarized. The purpose of this review is to show the subtle but important ways in which reporting laws differ (see Appendix B). The results of the present study may reflect sensitivity to some of these differences. For more specific information about reporting statutes in Canada, the reader is encouraged to review the actual provincial statutes that are referenced in Appendix C. Age of a Child Among the jurisdictions in Canada, there is variation in the definition of a child by age. Most commonly, a child is defined as being under the age of 18 in Alberta, Manitoba, Quebec, Prince Edward Island, the Yukon, and Northwest Territories. Saskatchewan, Ontario, Nova Scotia and Newfoundland define a child as being under 16 years of age. In British Columbia and New Brunswick a child is any person under 19 years old. Nature of Duty All provinces and territories in Canada, except the Yukon Territory, require that child abuse be reported. The Yukon statute makes it permissive to report (i.e., a person "may report" a child in need of protection), but it is not mandatory to do so. The term "shall" report is most commonly used to impose the reporting obligation. The statutes in British Columbia, Quebec, and Prince Edward Island state that a person "must" report within the conditions of the law. Definition of Child in Need of Protection With the exception of New Brunswick and Quebec, each provincial statute defines a "child in need of protection." New Brunswick details "when [the] security or 7 development of child [is] in danger," and Quebec describes when a child is "considered endangered." As summarized in the Canadian Family Law Guide (1988, pp.2745-2746), the circumstances upon which a child may be found to be in need of protection are as follows: (1) the child has been physically or sexually abused (Alberta, British Columbia, Manitoba, New Brunswick, Newfoundland, Nova Scotia, Ontario, Prince Edward Island, Quebec, Saskatchewan, and Yukon); (2) the child has been deprived of necessary care, supervision, and control (Alberta, British Columbia, Manitoba, New Brunswick, Newfoundland, the Northwest Territories, Nova Scotia, Ontario, Prince Edward Island, Quebec, Saskatchewan, and Yukon); (3) the child's life, health or emotional welfare have been endangered (Alberta, British Columbia, Manitoba, Newfoundland, the Northwest Territories, Nova Scotia, Ontario, Prince Edward Island, Quebec, Saskatchewan, and Yukon); (4) the child has been abandoned (Alberta, British Columbia, the Northwest Territories, Nova Scotia, Ontario, and Yukon); (5) the child is living in unfit or improper circumstances (Prince Edward Island, Manitoba, New Brunswick, Newfoundland, the Northwest Territories, Nova Scotia, Quebec, and Saskatchewan); (6) the child is living in a situation of severe domestic violence (Manitoba, New Brunswick, Newfoundland, Nova Scotia, Prince Edward Island, and Saskatchewan); (7) the child has been deprived of necessary medical attention (Alberta, British Columbia, Manitoba, New Brunswick, Newfoundland, the Northwest Territories, Nova Scotia, Ontario, Prince Edward Island, Quebec, Saskatchewan, and Yukon); (8) the child is beyond the control of the person caring for him or her (Alberta, Manitoba, New Brunswick, Newfoundland, the Northwest Territories, and Prince Edward Island); 8 (9) the child is not receiving proper education or is failing to attend school (New Brunswick, Newfoundland, Nova Scotia, Prince Edward Island, and Quebec); and (10) the child, being under the age of 12, has committed an act which, were the child over the age of 12, would amount to a criminal offense (New Brunswick, Newfoundland; the Northwest Territories, Nova Scotia, Ontario, Quebec, and Saskatchewan). Mandated Reporters of Abuse Initially, only physicians were required to report because they were presumed to have the most contact with abused children and they were trained to detect symptoms of maltreatment (Paulsen, 1966). Today, provincial statutes have expanded to impose a duty to report on all persons. The language used to communicate this duty does vary somewhat, however. British Columbia, Manitoba, Newfoundland, Ontario, and the Yukon Territory specify "a person." Alberta, New Brunswick, and Quebec require "any person." The Northwest Territories, Nova Scotia, Prince Edward Island, and Saskatchewan state "every person." Degree of Certainty Required The degree of certainty a reporter must reach before making a report varies from province to province and, in some instances, is based on the status of the reporter (i.e., professional vs. member of the general public). Across all provincial statutes, the threshold effectuating the duty to report for a professional refers to "reasonable grounds" or language that is substantially and legally similar such as "reasonable and probable grounds," "reason to believe," "reasonable belief," "reasonable suspicion," or "reasonable and probable cause." In four jurisdictions, the standard to report for non-professionals is simply for the reporter to have a general level of "information." These provinces include Newfoundland, 9 Nova Scotia, New Brunswick, and the Northwest Territories. Ontario and Quebec require members of the general public to report when they have "reasonable grounds to believe." When Reporting Is Required Mandated reporters are required to report as soon as the obligation arises. The laws in Canada specify that reports be made "forthwith," "promptly," "immediately," or "without delay." The reporting laws of Saskatchewan and the Yukon Territory are silent on this issue and do not specify a time line. Exemptions to the Duty to Report Protection for confidential information does not apply with respect to mandatory reporting in all provinces in Canada. Prince Edward Island and the Yukon, however, do not address the issue of confidential information. Lawyer-client privilege is preserved in all provinces except Newfoundland and Nova Scotia. The Yukon territory makes no mention of confidential relationships altogether, possibly due to the fact that the Yukon invites but does not mandate the reporting of abuse (Martz, 1995). Sanctions Imposed For Failure to Report Under all reporting laws, except those of the Yukon, failure to report is an offense for which the penalties range from $500.00 and/or six months in jail in the Northwest Territories to $10,000.00 and/or six months in imprisonment in British Columbia and Newfoundland. While failing to report is not an offense in the Yukon, making a false report carries a penalty of up to $5,000.00 and/or six months imprisonment. Similarly, British Columbia and Nova Scotia make false and malicious reports an offense. In both Ontario and New Brunswick, the failure to report constitutes an offense for professionals only. 10 Immunities Every province in Canada provides civil immunity to reporters. The intent of this provision is to encourage reporting by removing the threat of legal action. Only in cases where the report is made "maliciously," "in bad faith," "without reasonable grounds," or is intentionally "false" is that person liable. Summary This section has reviewed the purpose and history of child abuse reporting laws, along with detail of British Columbia's statute. Additionally, the central features of reporting statutes in Canada were compared. In this review of the statutes relating to the duty to report, it is important to note that this thesis is not intended as a definitive reference, given that these laws do change fairly often. The reader is encouraged to review his or her provincial statute when making a reporting decision. Case Law In the preceding section, mandatory reporting laws were examined. The purpose of this section is to illustrate how reporting statutes have been interpreted and applied by the courts. Court rulings can play an important role in guiding the professional conduct of psychologists. This section reviews case law that relates to the duty to report in Canada. Although mandatory reporting laws have been in place for some time, to my knowledge, there have been no cases in Canada in which a psychologist has been prosecuted for failure to report. Nonetheless, there are several important court rulings involving health professionals and service providers that may have application to psychologists. Some of these cases are reviewed. In an Ontario case, R. v. Stachula (1984), a fourteen-year old girl and her mother went to their family doctor, Dr. Stachula, to report that she might be pregnant, and that she had been having sexual relations with her older brother over the past several years. The 11 doctor referred the young girl to a health centre, where it was determined that she was in fact pregnant. A therapeutic abortion was arranged. While Dr. Stachula did ultimately report the incident of sexual abuse to child protection services, he did not do so until three months following his initial meeting with the mother and daughter. Therefore, the doctor was charged with failing to report. At the end of the trial, the doctor was acquitted based on two issues. First, the Crown did not prove that the abuser, a sibling, had "charge of the child." Ontario law requires that the abuser has or has had charge of the child in order for the duty to report to apply. Second, the Crown failed to present evidence establishing a standard of care expected of Dr. Stachula. Counsel introduced evidence of a doctor who specializes in paediatrics (with a subspecialty in child abuse), which is not the class of persons represented by Dr. Stachula, who is a family practitioner. In R. v. Kates (1986), two operators of a day nursery were charged with failure to report several incidents of suspected child abuse that were committed by their employee. The judge found that the Crown's position on abuse was a misunderstanding of the Child and Family Services Act. The judge ruled that the Act only applies to parents or the persons exercising the parental right, and not someone who has temporary control of a child, such as a daycare worker, bus driver, or teacher. Therefore, it was found that the Crown had no case against Mr. and Mrs. Kates. In a more recent case, R. v. Rahalkar (1995), a doctor was charged with failing to report suspected sexual abuse of a child. The evidence at trial was that the young girl had told her mother and Dr. Rahalkar about an incident of sexual touching by her uncle about three weeks after it occurred. Similar to the Stachula and Kates decisions, the doctor was acquitted because the judge could not conclude that the doctor had reasonable grounds to suspect that a person in charge of the child had molested her. The judge said that the best conclusion that could be drawn from the evidence was that the uncle was alone with the child at the time of the incident, but that that does not equate with the uncle having charge of the child. 12 In another Ontario case, R. v. Cook (1985), a medical doctor, Dr. Cook, was charged with failing to report child abuse after she learned that her patient's daughter was sexually abused by her stepfather over the past two years. Dr. Cook advised her patient to confront her husband with the issue. The mother did so and then told Dr. Cook that the abuse had stopped. Dr. Cook did not make a report of child abuse and was later charged. At trial, Dr. Cook was acquitted, but the decision was appealed. Since there was no evidence that the abuse continued, and the only evidence before the doctor was that in the past the child has been abused, the charge was not established and the original acquittal was upheld. Summary On the basis of the Stachula, Kates, and Rahalkar cases, judges have established that, in Ontario, a person must "have charge of the child" in order for the reporting law to apply. In other words, Ontario's Child Welfare Act only applies to an adult exercizing authority over a child. In addition to this finding, the R. v. Cook case addresses the issue of whether a professional should report abuse to the authorities even when it could be detrimental to the family. The courts appear to recognize the dilemma that doctors and mental health professionals face in deciding to report possible child abuse. On the one hand, it is clear that a person must report when there is reasonable grounds to believe that a child needs protection. On the other hand, the R. v. Cook case recognizes that insistence on reporting can cause substantial harm to children and families as well. This issue and others are discussed in the next section. Professional Reactions Since the enactment of legislation, various authors have expressed strong opinions about child abuse reporting laws. This section will briefly highlight the major concerns related to mandatory reporting. 13 One criticism of mandatory reporting is that the definitions of child abuse are vague and unclear. As a result, there is widespread confusion about what is and what is not child abuse. According to Besharov (1991), this confusion leads to high rates of reports that are not substantiated by child protection services. In fact, it is estimated that approximately 60% of all reports are dismissed by child protection services (Besharov, 1991; Eckenrode, Powers, Doris, Munsch, & Bolger, 1988). As a result of many claims being dismissed, mandated reporters have become discouraged, and families are not receiving the assistance they require (Zellman & Antler, 1990). Other researchers have stated that reports of child abuse, while beneficial to some children, result in the child being further victimized when that child is removed from the home (Bailey, 1982; Newberger, 1983). Similarly, removing the abusive parent from the home can be harmful as well. Bailey (1982) argues that children can experience separation loss when a parent is removed from the home. A child who discloses abuse may feel guilt associated with the disclosure, particularly when the disclosures leads to prosecution or removal of a parent from the home. Newberger (1983) recommends the use of more effective policies designed around prevention and treatment instead of criminalizing family problems. A major criticism cited in the literature is that the benefits of mandatory reporting are not sufficiently weighed against the breach of confidentiality that may be involved (Miller & Weinstock, 1987). According to Harper and Irvin (1985), perceived harm to the therapeutic relationship is the most common reason for failure to report by psychologists. This issue will be more fully elaborated in the section on ethics. Still another concern that is widely expressed among clinicians is that mandatory reporting places limitations on clinical judgement (Ansell & Ross , 1990). Under current law, mandatory reporters are left with choosing between reporting every possible indication of abuse or risking the possibility of criminal charges, civil liability, and professional sanctions. As stated by Ansell and Ross (1990), failure to report possible 14 abuse often occurs among experienced, ethical psychologists who choose "their client's welfare over mindless obedience to reporting laws" (p.399). Reluctance to become involved in legal proceedings has contributed to failure to report among many psychologists and other mental health professionals (Swoboda, Elwork, Sales, & Levine, 1978). Fear of being sued has been cited as another reasons for noncompliance. Bailey (1982), for example, reported that some physician's fear that they will be prosecuted for making an unfounded report. These fears are, of course, erroneous in that mandated reporters are immune from liability (e.g., for defamation) if they report in "good faith." According to Zellman and Antler (1990), psychologist's perception of the ineffectiveness of child protection services is an important predictor of failure to report. Many of the clinical psychologists that were sampled indicated that they decided not to report because previous cases had been mishandled or ignored by child protection services. Consequently, these psychologists tended to seek a higher degree of certainty that abuse had in fact occurred before they reported. The data from this study reflects the anger and frustration of mandated reporters in that their reports have been ignored or overruled by inexperienced and poorly trained child protection workers. In response to these concerns, some authors have proposed that mental health professionals be exempt from reporting less serious cases of child abuse when a family is engaged in treatment (Emery & Laumann-Billings, 1998). Finkelhor and Zellman (1991) have extended this idea and specified a system of flexible options for reporting child abuse. In this way, they argue that compliance with the law may be improved and burdens on the child protection system may be reduced. The key elements of their proposal include: (1) certify reporters; (2) establish reporting options for registered reporters (e.g., delay reporting); (3) the new reporting options would still require that a report be filed but that no identifying information about the family would be provided; (4) in the case of a delayed report, provide identifying information if the case becomes more serious; (5) child 15 protection services would periodically review records submitted by registered reporters to monitor the system; and (6) the guidelines would specify circumstances under which deferred or confidential reporting are not permitted. Summary Despite the above concerns, psychologists have, for the most part, widely accepted the legitimacy of their obligation to report child abuse and they have incorporated these laws into their practices (Cram & Dobson, 1993). The next section introduces ethical codes and principles that relate to the duty to report. Ethics In addition to mandatory reporting laws, codes of ethics can aid psychologists in decision making. This section discusses the purpose of codes of ethics, and reviews the dilemma between maintaining client confidentiality and upholding the child abuse reporting law. Ethical principles relevant to child abuse reporting are presented and guidelines for practice are reviewed. Codes of Ethics Codes of ethics are normative statements that serve to guide ethical and professional practice. Codes of ethics not only signify professional identity but they mark the maturity of a profession (Mabe & Rollin, 1986). According to Corey, Corey, and Callahan (1993), ethical codes are partially designed to protect both the client and the psychologist from harm. Psychologists that practice in accordance with accepted standards have a defense in a malpractice lawsuit. Ethical behaviour involves more than following a code, however. In many instances, psychologists must exercise professional judgement in order to clarify, interpret, and apply an ethical principle or standard (Corey et al., 1993; Smith, McGuire, Abbott, & Blau, 1991). For these reasons, codes of ethics have been the subject of much criticism. A 16 number of authors have expressed concerns about the inherent ambiguity of codes of ethics (e.g., such as when ethical principles conflict with laws), and that these codes may be inadequate in providing sufficient direction to practitioners (Tymchuk, 1986). The next section focuses on the provision of confidentiality and its relation to child abuse reporting. Ethical Standards and Mandatory Reporting of Child Abuse The enactment of mandatory child abuse reporting laws has created concern for many practicing psychologists. This concern relates to the conflict between laws and ethical standards. Mandatory reporting of child abuse legislation poses significant challenges to psychologists, who must strike a balance among their client's best interest, confidentiality, and reporting obligations. Psychologists' ambivalence regarding this particular ethical-legal dilemma is reflected in the literature on their reporting behaviour. Initial studies indicating the failure of psychologists to report child abuse prompted researchers to further investigate the factors involved in reporting decisions (Muehleman & Kimmonsl981; Swoboda et al., 1978). Over the past 20 years, a concern about breaching confidentiality has been a major reason cited for psychologists' failure to report. Kalichman and his colleagues (1989), for example, found that 42% of licensed psychologists surveyed believed that reporting suspected child abuse has negative consequences with respect to the progress of family therapy. Research investigating the actual effects of reporting on the therapeutic relationship, however, has been incongruent with psychologists' perceptions. Harper and Irvin (1985), for example, conducted a qualitative study and found that, when reporting was done in the context of ongoing therapy, clients were unlikely to discontinue with treatment. In fact, with many of these families, reporting was found to have a positive effect on the therapeutic alliance. In another study, Watson and Levine (1989) evaluated 65 cases of reported child abuse and found that the majority of cases involving a report did not 17 have any negative effects on the therapeutic relationship. The authors found that 76% of the families did not terminate following a report of suspected child abuse and that, in one-third of cases, reporting was positive when it involved a report of a third party. In conclusion, Watson and Levine suggest that reporting child abuse may not necessarily be harmful to the therapeutic relationship; rather, the researchers state that it is trust, not absolute confidentiality, that is essential for the therapeutic relationship. In addition to studying the impact of the legal mandate to report child abuse on the therapeutic relationship, researchers have assessed the extent to which psychologists are aware of and communicate the limits of confidentiality to their clients. In one study, Baird and Rupert (1987) found that only about one-half of the psychologists that they surveyed informed clients of the limits to confidentiality in the first session. A more recent study of psychologists across three states in the United States found that almost 20% of their participants indicated that they sometimes, rarely, or never provide the limits of confidentiality to their clients. Over 5% of psychologists misleadingly tell their clients that everything disclosed in therapy is confidential. More encouraging results are reported by Otto, Ogloff, and Small (1991) who report that over 90% of psychologists in their sample addressed the issue of confidentiality with their clients. Codes of Ethics Relating to Duty to Report Although mandatory reporting laws have historically posed an ethical-legal dilemma for many psychologists, codes of ethics have now addressed the duty to report. Two codes of ethics are particularly relevant to psychologists in British Columbia: the Canadian Code of Ethics for Psychologists (Canadian Psychological Association, 1991), and the Ethical Standards of Psychologists in British Columbia (College of Psychologists of British Columbia, 1985). The relevant principles relating to the duty to report will be highlighted below. 18 Standard 1.40 of the Code of Ethics for Psychologists (Canadian Psychological Association, 1991) states that psychologists: Share confidential information with others only with the informed consent of those involved, or in a manner that the individuals involved cannot be identified, except as required or justified by law, or in circumstances of actual or possible serious physical harm or death (p. 13). Standard 11.36 (Canadian Psychological Association, 1991) is also relevant. It states that psychologists: Do everything possible to stop or offset the consequences of actions by others when these actions are likely to cause serious physical harm or death. This may include reporting to appropriate authorities (e.g., the police) or an intended victim, and would be done even when a confidential relationship is involved (p. 18). According to the Code of Ethics for Psychologists (Canadian Psychological Association, 1991), ethical practice also requires that psychologists inform their clients of the legal duty to report child abuse (Standard 1.17). In British Columbia, similar standards apply with respect to the duty to report. Principle 5 of the Ethical Standards of Psychologists in British Columbia (College of Psychologists of British Columbia, 1985) states that psychologists reveal confidential information to others "only with the consent of the person or the person's legal representative, except in circumstances in which not to do so would result in clear danger to the person or to others" (p.6). In summary, current codes of ethics have made it very clear that there are limits to confidentiality and that child abuse must be reported. In this way, ethical codes (and legislation) have removed psychologists' ability to use discretion about what to do once child abuse is suspected. 19 Guidelines for Psychologists In response to legal-ethical difficulties in reporting suspected child abuse, a number of researchers have developed guidelines for psychologists (Kalichman, 1993; Walters, 1993). These guidelines are summarized in point form: (1) know your provincial child abuse reporting law; (2) review the limits of confidentiality with clients; (3) consult with others (colleagues, professional associations, child protection services); (4) report whenever you suspect a child is in need of protection; (5) inform clients of your intention to report; (6) limit disclosure to what is essential; (7) keep accurate case notes; (8) minimize negative consequences of disclosure; (9) follow up on reports with child protection services; and (10) seek ongoing training. Summary This section has reviewed various aspects of professional ethics in relation to mandatory reporting, including codes of ethics, ethical dilemmas, relevant principles, and guidelines for psychologists. In the next section, models of ethical decision making are presented. Models of Ethical Decision Making In a further effort to aid professionals in their ethical decision making, several authors have developed models regarding process components essential to the resolution of ethical issues (Canadian Psychological Association, 1991; Corey, Corey, & Callahan, 1984; Kitchener, 1984; Rest, 1984; Tymchuk, 1986). Unfortunately, these models are entirely conceptual in nature and are not grounded in empirical research. Nonetheless, from a normative perspective, these models merit review. In consideration of ethical issues for psychologists, Tymchuk (1986) developed a model that is intended for decision making, as well as the training of students. Tymchuk's model consists of seven levels: (a) determining who should participate in the decision; (b) 20 determining available alternatives; (c) determining who should decide which alternatives to implement; (d) determining which alternatives to implement by considering the consequences, risks, and benefits for each one; (e) reviewing procedures; (f) implementing the alternative selected; and (g) evaluating the decision. Kitchener (1984) suggests two levels of reasoning when faced with an ethical situation, the intuitive level and the critical-evaluative level. The intuitive level is one's immediate response or "ordinary moral judgement" to an ethical situation that is based on previous beliefs, knowledge, assumptions, and experience. Exclusive reliance upon an intuitive level of moral reasoning, however, is inherently problematic. For example, intuitive thinking is permeated with cultural values that are detrimental to women and other historically undervalued groups (Hill, Glaser, & Harden, 1995). Also, intuitive responses are susceptible to principles of moral relativism, such as when psychologists claim client-beneficence as a justification for their sexual improprieties with patients (Holroyd & Brodsky, 1977). Kitchener (1984) states that the inadequacies of the intuitive level make critical-evaluative reasoning "necessary to guide, refine, and evaluate our ordinary moral judgement" (p.44). There are three aspects of the critical-evaluative level of moral reasoning: (a) rules which include professional codes of ethics and laws; (b) the principles of autonomy, nonmaleficence, beneficence, justice, and fidelity; and (c) ethical theory. Corey, Corey, and Callahan (1984) emphasize that ethical decision making is a highly personalized process and that self-exploration and self-development strategies are critical in exercising sound judgement. Responsible judgement implies that professionals consult with colleagues, maintain current standards through readings and educational activities, and engage in an ongoing process of self-examination. More recently, Corey, Corey, and Callahan (1993) reviewed various models on ethical decision making and proposed seven steps. They include: (a) identify the problem; (b) identify the potential issues involved; (c) review the relevant ethical guidelines; (d) obtain consultation; (e) 21 consider possible and probable courses of action; (f) consider the consequences of various decisions; and (g) choose the best course of action. Adopted from Tymchuk's (1986) model, the new Canadian Code of Ethics for Psychologists (Canadian Psychological Association, 1991) highlights seven basic steps in ethical decision making. The decision making steps are to (a) identify the ethically relevant issues and practices; (b) develop alternative courses of action; (c) analyze the risks and benefits (short-term, ongoing, and long-term) of each course of action on all persons involved or likely to be affected; (d) choose a course of action that is based on existing principles, values, and standards; (e) act; (f) evaluate the results of the act; and (g) assume responsibility for the consequences of the action, including correcting negative consequences, if any, or re-engaging in the decision making process if the ethical issue is not resolved. The CPA model adds that members are expected to seek necessary consultation when such resources can enhance (i.e., add knowledge or objectivity to) the decision making process. In the critical analysis of specific ethical situations, Curtin (1978) states that certain elements have direct impact on ethical choices. For instance, factors such as levels of immediacy, the emotional impact of situations, culture and society are important influences in choice of action. Curtin proposes that a model of decision making ought to incorporate background information, identification of ethical components, ethical agents (i.e., persons involved in the decision making), identification of options, application of principles, and resolution. In proposing this model, Curtin cautions professionals against making decisions for clients (insofar as possible), and encourages them to facilitate, not manipulate, the decision making process. In recent years, three models have been developed that are directly applicable to the reporting of child abuse. In one of these models, Brosig and Kalichman (1992), propose a three tiered approach to decision making in child abuse reports. They suggest that a combination of legal factors, clinician characteristics, and situational factors interact to 22 influence whether a clinician chooses to report. Common legal factors affecting child abuse reporting include the clinicians' knowledge of the reporting law, wording of the law itself, and the procedures specified within the law. Clinician characteristics that affect the probability of reporting include years of experience, training, attitudes and experience in making child abuse reports. Situational factors that influence clinicians' decision making include victim attributes, type of abuse, severity of abuse, and availability of evidence. A more recent publication by Kalichman (1993) adds organizational factors to the original model. Organizational factors include ethical guidelines, support for reporting, and institutional policy. While Brosig's and Kalichman's model outlines several factors influencing reporting decisions, it has provided little information regarding the decision making process. Furthermore, it has yet to be tested by any empirical study. In a second model of child abuse reporting, Kalichman (1993) goes beyond his earlier work in identifying specific factors involved in reporting decisions to describe procedures involved in making a reporting decision. According to Kalichman's model, there are four key steps involved in making a reporting decision: (1) assessing the extent to which the indicators meet the clinicians' reporting threshold, (2) evaluating the cause for the symptoms, (3) consulting with colleagues, and (4) making preliminary contact with child protection services. In review of this model, it appears that the critical element in reporting decisions relates to the notion of a reporting threshold. Kalichman's model predicts that psychologists will become increasingly inclined to report as they surpass a reporting threshold. One of the strengths of Kalichman's (1993) model is that it is presented in the form of a decision tree. His decision tree model shows how various procedures can be used to formulate an overall judgement of whether or not to report. While Kalichman is to be credited for taking a first step to describe key procedures involved in decision making, this model was adapted from a model that was developed in duty-to-warn cases. Therefore, the 23 credibility of this model in child abuse reporting decisions is yet to be determined. Future research is needed to establish the validity of this model. Stadler's (1989) model makes another contribution to the literature in that it outlines a strategy for reporting child maltreatment. Presented in a decision tree format, Stadler's model describes four options for reporting child abuse and neglect. The first option involves offering the client the opportunity to self-report. If the client chooses this option, a report is made during the session. Second, if the client does not wish to self-report, the counsellor reports while the client listens. The third option involves the counsellor offering to report from another room while the client waits. Fourth, the counsellor reports in the absence of the client. Stadler's model is hierarchical in nature and it is derived from principles of client autonomy and beneficence. As with the other models, Stadler's work is limited in that it too has not been tested in any formal way to determine its usefulness and credibility. Summary This section has reviewed several decision making models in ethics. While these models are useful in guiding professional behaviour, they may not reflect the process of actual decision making behaviour. In other words, there is no established link between these models and the decision choices of professionals in the real world. Ethical choice models are therefore needed that are derived from empirical data. Research is also needed that takes into account various factors such as the particular dilemma and its context, as well as characteristics of the decision-maker, and the decision environment. In the next section, research on decision making is presented and related to child abuse reporting. Decision Making Reporting child abuse is a process involving human judgement and decision making. Examining the decision making literature on how people organize, interpret, and 24 use information has important implications for psychologists who report child abuse. This section will present a selected framework in which the decision making literature can be understood. Given that child abuse reporting decisions are often made without predictability of consequences, this review focuses on decisions that are made under risk and uncertainty. Until quite recently, the field of decision research was rigorously experimental, relying upon precise mathematical (or normative) models. Consequently, the studies that will be reviewed in this section are embedded in this tradition, and may not directly relate to child abuse reporting. Nonetheless, elements of decision making research (e.g., heuristics, time pressure, etc.) have implications for the present study. The limitations of statistical models to predict real-world events offers much support for the present qualitative study. Historical Sketch It has only been within the last 40 years that decision making has received substantial attention in the psychological literature (for in-depth reviews, see Ajzen, 1996; Payne, Bettman, & Johnson, 1992; Slovic, Lichtenstein, & Fischhoff, 1988). The publication of The Theory of Decision Making by Ward Edwards (1954) is generally recognized as the beginning point for psychologists interested in decision making. In his review article, Edwards introduced psychologists to the literature on risky and riskless choice, utility, and game theory. This paper spawned a proliferation of theoretical and experimental studies on decision making. Two streams followed from this literature. One stream is the theory of riskless choice. The theory of riskless choice has its origins in the theory of utility maximization. The basic principles of utility theory are that decision makers are fully informed about the possible courses of action and their consequences; are infinitely sensitive to differences in alternatives; and are rational so as to maximize expected utility. The second stream that 25 use information has important implications for psychologists who report child abuse. This section will present a selected framework in which the decision making literature can be understood. Given that child abuse reporting decisions are often made without predictability of consequences, this review focuses on decisions that are made under risk and uncertainty. Until quite recently, the field of decision research was rigorously experimental, relying upon precise mathematical (or normative) models. Consequently, the studies that will be reviewed in this section are embedded in this tradition, and may not directly relate to child abuse reporting. Nonetheless, elements of decision making research (e.g., heuristics, time pressure, etc.) have implications for the present study. The limitations of statistical models to predict real-world events offers much support for the present qualitative study. Historical Sketch It has only been within the last 40 years that decision making has received substantial attention in the psychological literature (for in-depth reviews, see Ajzen, 1996; Payne, Bettman, & Johnson, 1992; Slovic, Lichtenstein, & Fischhoff, 1988). The publication of The Theory of Decision Making by Ward Edwards (1954) is generally recognized as the beginning point for psychologists interested in decision making. In his review article, Edwards introduced psychologists to the literature on risky and riskless choice, utility, and game theory. This paper spawned a proliferation of theoretical and experimental studies on decision making. Two streams followed from this literature. One stream is the theory of riskless choice. The theory of riskless choice has its origins in the theory of utility maximization. The basic principles of utility theory are that decision makers are fully informed about the possible courses of action and their consequences; are infinitely sensitive to differences in alternatives; and are rational so as to maximize expected utility. The second stream that 25 emerged from Edwards' (1954) review is decision making under risk, which has its origins in simple gambles of chance. The theory of risky choice is defined as a decision in which one must choose between outcomes where the probabilities of this outcome are not completely known. Maximization plays a role in these theories; however, the quantity to be maximized becomes expected utility, due to the uncertainty involved in decision making. A few years after Edwards' (1954) review of decision making appeared in the psychological literature, Simon (1956) proposed an alternative view to utility maximization. Simon coined the term bounded rationality, which asserts that cognitive limitations force decision-makers to construct and rely on simplified solution rules. Accordingly, predicting behaviour involves understanding how decision-makers construct these simplified solution rules. The key to simplification, according to Simon, was to replace utility maximization with the satis/icing principle: outcomes are first classified as satisfactory or unsatisfactory with respect to each of the relevant attributes; the first alternative that satisfies the level of aspiration for every attribute is selected. In comparison to utility maximization, satisficing does not call for detailed exploration of all the available alternatives, and requires limited cognitive computational capacities. More recently, the field of decision making has focused on understanding the cognitive processes used during decision making. This emphasis differs from previous inquiry in that it describes the knowledge structures that are used during decision making. Instead of looking at decision making only at the point of decisions between alternatives, the whole process is examined. Current decision research also examines the nature of the knowledge used and the cognitive operations involved in making a choice. This tradition has been highly influenced by the field of artificial intelligence, which aims to develop machines capable of reproducing some human cognitive functions. 26 Normative and Descriptive Theory There are two dominant theories in decision making, normative and descriptive theory. Normative theory seeks to prescribe or recommend how decisions ought to be made; it assumes that people act in a rational manner (Mullen & Roth, 1991). As such, behaviour is regarded as the consequence of logically consistent decision procedures. In other words, people act to maximize performance based on utility (value) or probability. Within this perspective, lack of agreement between a person's choice and a normative model is taken as evidence of irrationality. Normative decision theory makes use of explicit numerical models for the evaluation of human decisions. One of the earliest models is the utility model. According to this model, a person chooses the action with the largest utility and objective probability of payoff (Loke, 1996). Another subgroup of normative theory is the dynamic model. This model attempts to account for risky decision making behaviour, as in gambling situations or where information is limited and the outcome is ambiguous. Here, the decision-maker chooses the action which will give the largest utility and subjective probability of payoff (Loke, 1996). These models have been criticized, however, because all relevant information is not always available to the decision maker, and they provide only limited information about the processes that underlie decisions (Ajzen, 1996). Descriptive theory, on the other hand, is concerned with people's beliefs and preferences as they actually are, not as they should be and is more directly relevant to the current study. Analysis focuses on what individuals actually do in a decision making situation. An example of a descriptive model is the decision tree. A decision tree is a diagram that shows the relationship between decision criteria and responses over time. It describes the factors that decision-makers use in a particular situation. The focus on descriptive models has produced research that casts some doubt about the adequacy with which people use information to make judgements and draw inferences (Kahneman & Tversky, 1973). Some of this research will be presented next, emphasizing 27 three aspects of the decision environment: problem structuring, heuristics, and restructuring decision problems. Decision Environment Prior to making a decision, the decision-maker is presented with a host of challenges. Decision structuring is perhaps one of the most important of these challenges. The purpose of decision structuring is to put a decision problem in a form in which it can be solved (Mullen & Roth, 1991). It involves problem recognition and value analysis, generating alternative choices of action, and considering the possible outcomes. After structuring the decision problem, the decision maker must estimate the probabilities of outcomes associated with the different alternatives, evaluate the subjective values of the outcomes, integrate these judgements to choose a preferred course of action, and then carry-out the decision. Finally, depending on the degree of satisfaction with the results, the decision-maker can restructure the problem, reassess the subjective probabilities and values of possible outcomes, and choose another alternative. In this section, research that relates to each of these stages will be reviewed. Decision Structuring There is a growing interest in understanding how problems are structured or framed, although the amount of research in this area is limited. In most of these studies, participants are presented with a decision problem in which the alternatives and attribute values are specified. Then, participants are asked to solve the particular problem. This strategy allows researchers to maximize experimental control. When faced with unstructured decision problems, research shows that people often fail to generate alternative actions. In one study, Gettys, Manning, Mehle, and Fisher (1980) asked participants to list all possible hypotheses about various tasks, such as why a car might not start. Participants then estimated the probability that the actual cause of the 28 problem was included in the list of generated hypotheses. The authors found that participants were not able to generate important hypotheses about, for example, why a car would not start; furthermore, they tended to believe that the hypotheses that they proposed were more complete than they actually were. These findings suggest that psychologists may be limited in their ability to fully consider a range of hypotheses about clinical issues, and that they may be overconfident about the alternatives that they generate. Estimating Probabilities After structuring a decision problem, the decision-maker must consider the probabilities surrounding the problem and the possible outcomes of the decision. In this subsection, various limitations in the way that people think about risk and uncertainty are discussed. These limitations stem largely from the use of various heuristics—mental strategies people use to reduce difficult tasks to simpler judgements (Tversky & Kahneman, 1974). These heuristics are valid in some circumstances, but in others, they may lead to serious errors in decision making. Examples include availability, representativeness, anchoring and adjustment, and the base-rate fallacy. The Availability Heuristic. The availability heuristic is a strategy that is used to estimate the probability of an event when the relative frequencies are not presented directly. It refers to the ease with which associations come to mind (Tversky & Kahneman, 1973). In most cases, this heuristic is valid because frequent events are typically easier to recall than instances of less frequent events, and likely occurrences are easier to imagine than unlikely ones. However, since availability is affected by subtle factors unrelated to likelihood, such as familiarity, recency, and emotional salience, reliance on it may lead to an overestimation of probabilities. For example, Lichtenstein, Slovic, Fischhoff, Layman, and Combs (1978) found that the frequencies of dramatic, well-publicized causes of death such as accidents, 29 natural disasters, fires, and homicide were overestimated while less dramatic causes of death such as stroke, diabetes, emphysema, and asthma were underestimated. With respect to child abuse reporting, the availability heuristic is likely to influence the preconceptions of psychologists. Recent publicity about cases of child abuse may lead them to overestimate the occurrence of abuse. Consequently, subtle signs of abuse may lead to heightened suspicions of child maltreatment resulting in unfounded reports of abuse. Representativeness. Another strategy for probability reasoning is the representative heuristic. This heuristic involves an assessment of the probability of an uncertain event by judging the degree to which that event is similar to an appropriate mental model or another event (Kahneman, Slovic, & Tversky, 1982). In other words, it is the degree of similarity in essential properties between events or in the process by which events are generated. Decision-makers using the representative heuristic can be led astray either by attending to characteristics that are normatively irrelevant or by disregarding characteristics that are normatively important. The availability heuristic has important implications for the present research. That is, a psychologist who attends to normatively irrelevant factors (e.g., race, gender, socioeconomic status) or disregards important information (e.g., a child who alleges then retracts child abuse) may overreport or underreport child abuse, respectively. Anchoring and Adjustment. Individuals typically begin the prediction process by starting off from an initial value that is usually adjusted in order to accommodate additional information. When prediction is based on incomplete information, the subsequent adjustment is smaller than it should be. Anchoring will occur regardless of whether the decision-maker is told of the initial value or whether the initial value is elicited from the person. Anchoring and 30 adjustment effects have been noted in the literature (Tversky & Kahneman, 1974). This tendency has important implications for psychologists who make reporting decisions. For example, if, as a starting point, a psychologist suspects abuse, it is more likely that he or she will make a report despite learning information contrary to the established belief. Similarly, a psychologist who initially judges that no abuse has occurred to a child may fail to make a report despite learning new information suggesting child maltreatment. The Base-rate Fallacy. The base-rate fallacy is the tendency to ignore base-rate frequencies or to overutilize diagnostic information. Base-rate information is the rate at which an incident occurs, whereas diagnostic information is any information relevant to problem identification. A number of studies have shown that people tend to give less weight to base-rate statistics, and greater weight to diagnostic information (for a review see Borgida & Brekke, 1981). For example, Meehl and Rosen (1955) noted that clinical psychologists often disregard base-rate information when making predictions of rare events, such as suicide. In another study, Kahneman and Tversky (1973) found base-rate neglect in an experiment that manipulated probabilities. Participants were shown brief personality descriptions and asked to assess the likelihood that an individual described was either a lawyer or an engineer. The individual was allegedly sampled from a group consisting of 70 engineers and 30 lawyers, or 70 lawyers and 30 engineers. The odds that any particular description belongs to an engineer rather than to a lawyer is higher in the first condition, and lower in the second condition. The inverse relationship exists for lawyers. Results found that the two conditions produced essentially the same probability judgements. Kahneman and Tversky attributed this neglect to a reliance on representativeness, the similarity of the descriptions to one's stereotype or mental image of an engineer or lawyer. Essentially, prior probabilities were ignored when a description was introduced, even when this description was uninformative. These findings suggest that psychologists, when 31 judging the likelihood of child abuse, may give less weight than warranted to base-rate statistics and, correspondingly, more weight than warranted to diagnostic information. Verbal Phrases of Probability. People often use verbal phrases, such as the words likely or probable, to express their degree of uncertainty. The difficulty that this causes for communication, even among experienced professionals, is that there is little consensus about the level of probability to which these terms refer. For example, attempts to discover the numerical values equivalent to these labels have shown great variability among physicians (Bryant & Norman, 1980). When asked to give a numerical probability equivalent to the term moderate risk, for example, doctors' responses ranged from 0.20 to 0.80; for probable, answers ranged from 0.30 to 0.95. These findings have important implications for psychologists who report child abuse to the Ministry. If terms such as likely abused have different meanings for different professionals, it is possible that reports of child abuse may convey an inaccurate impression of the degree of risk or severity. Probability as Confidence. People tend to be overconfident when judging the probability of events. Slovic, Fischhoff, and Lichtenstein (1980) have illustrated this error in judgement. Participants were given pairs of lethal events (e.g., drowning and suicide) and asked to judge which occurred more frequently, and then estimate the chances that they are wrong. Results showed that most participants estimated that their chances of being wrong were about 1:100 or less; in actuality, their judgements were wrong about 1:8. This tendency has implications for reporting child abuse. That is, professionals who report or choose not to report abuse may be overconfident about the accuracy of their judgements when, in fact, they are mistaken. This may have dire consequences for a child. 32 Within the domain of cognitive psychology, there are three possible explanations for this phenomenon. First, Fischhoff, Slovic, and Lichtenstein (1977) noted the tendency for people to believe that their memories are an exact representation of their experiences, whereas evidence suggests that memory is a reconstructive process in which errors are sometimes incorporated as facts. Second, Pitz (1974) suggested that, in a series of inferences, the uncertainty in the earlier stages may not be carried over into the later stages. Third, Koriat, Lichtenstein, and Fischhoff (1980) emphasized the degree to which people search their memory only for confirming, not disconfirming, evidence concerning an initially favoured answer. Restructuring Decision Problems According to Coupey (1990), restructuring is the application of operations to a set of information to produce a new problem representation. Restructuring might involve transforming information (e.g., rounding off, standardizing, or performing calculations), rearranging information (e.g., the order of attributes), or simplifying by eliminating information (Payne, Bettman, & Johnson, 1992). The purpose of restructuring is to reduce the amount of perceived conflict in the decision or the degree of difficulty (Payne, Bettman, & Johnson, 1992). One study that showed that individuals restructure information was conducted by Coupey (1990). Coupey gathered information about restructuring by allowing half of the participants to take notes while solving decision problems and then coding the restructuring operations that they used. When participants were given poorly structured information, it was found that they used a variety of restructuring methods, and tended to utilize alternative based strategies to solve the problems. Coupey suggests that individuals restructure decision problems so that they can later perform a more accurate heuristic using a reasonable amount of effort. 33 Problem Characteristics The findings discussed in the previous section suggest that various heuristics can influence decision making under uncertainty. In this section, three characteristics of the problem will be discussed that have been shown to influence decision making. These characteristics are time pressure, availability of information, and framing effects. Time Pressure. Time pressure is an important factor in decision making and is relevant to child abuse reporting decisions. Research suggests that decision-makers cope with time pressure in several ways. One way to respond to time pressure is to increase or accelerate the processing of information (Ben Zur & Breznitz, 1981; Miller, 1960; Payne, Bettman, & Johnson, 1988). Alternatively, decision-makers under time pressure may focus on the most important information or on negative information (Ben Zur & Breznitz, 1981; Payne, Bettman, & Johnson, 1988; Wright, 1974). This strategy is also called filtration (Miller, 1960), meaning that only a subset of the most important information is processed. Finally, people may change their decision strategies altogether under time constraints. This may involve avoidance (Ben Zur & Breznitz, 1981; Miller, 1960) or the use of a noncompensatory decision rule (Zakay, 1985). In noncompensatory strategies there are no trade-offs; that is, a good value on one attribute can not make up for a bad value on other attributes. For example, a psychologist may still file a report of child abuse although the abuser expresses interest to continue with therapy. Availability of Information. In addition to time pressure, decisions may vary depending on the amount of information available. If there is a lack of relevant information, then poor decisions can be made as a result of ignorance. Similarly, too much information may result in only irrelevant information being considered. One method of making a decision with partial 34 information is to infer a missing value. The inferred value may be, for example, the average value (Slovic & MacPhillamy, 1974). An alternative method is to assign less favorable values (Yates, Jagacinski, & Faber, 1978). Availability of information is a particular concern for psychologists who make reporting decisions. Typically, decisions to report abuse are made on the basis of incomplete information and, in some cases, on evidence that is contradictory. In such cases, clinicians may rely on other sources of information such as consulting with colleagues, referring to ethical codes or legal statutes, for example, or, on the heuristic (e.g., availability). Framing Effects. The way in which information is presented or framed can influence judgements and decisions. In fact, people's perceptions of risk and their subsequent choices are sometimes dramatically altered by changes in the presentation of risks. For example, McNeil, Pauker, Sox, and Tversky (1982) found that choice between alternative treatments for lung cancer was influenced by whether the outcome was framed in terms of mortality rates or in terms of survival rates. Similarly, choices between two intervention strategies for dealing with a crisis situation differed depending on whether the strategies were described in terms of likelihood of lived saved or lives lost, even though the objective information, the number of lives lost, was the same in each case (Tversky & Kahneman, 1981). According to Tversky and Kahneman (1981) there is a tendency to respond differently to risk when facing a prospective gain than when facing a prospective loss. When faced with a prospective gain, people tend to be risk averse. When faced with a prospective loss, people tend to be risk seeking. This pattern of preference has been demonstrated in other studies (Eraker & Sox, 1981; Slovic, Fischhoff, & Lichtenstein, 1982). Framing effects have important implications for psychologists who make reporting decisions. For example, psychologists who frame reporting decisions as positive (e.g., a 35 report will benefit the child) may be more likely to report. In contrast, when decisions to report are framed negatively (e.g., a report will disrupt therapy), psychologists may be less inclined to file a report. Summary As stated in the introductory paragraph, normative models have dominated research on decision making. The goal of these models is to produce data that are internally valid and reliable. Their use involves the collection of laboratory data that can be easily scorable and used mostly to investigate input/output connections. These models are limited, however, in that they do not predict real-life events or permit a direct examination of the decision making process. Additionally, normative models assume that the decision-maker has all of the relevant information available to make a choice. In actuality, decision making, as in the case of child abuse reporting, often involves making a decision on the basis of incomplete or ambiguous information. Decision making related to child abuse reporting is, furthermore, a complex and difficult process because of time stress and the potential harm to children. Clearly, there is a need for descriptive models in decision research that will predict real-life decisions such as reporting child abuse. Mandatory Reporting of Child Abuse To this point, various aspects of mandatory reporting of child abuse have been discussed (e.g., B.C.'s reporting law, case law in Canada, professional reactions, ethics, etc.) This section reviews the empirical research on reporting child abuse and the factors that influence decision making: legal issues, clinician characteristics, and situational factors (Brosig & Kalichman, 1992a; Kalichman, 1993). This review is limited by focusing only on the research contributions of mental health professionals, and excludes research on teachers, physicians, and nurses. 36 Legal Factors A variety of legal factors have been studied in an effort to account for inconsistent compliance with reporting laws. At the most basic level, Swoboda et al.(978) surveyed mental health professionals about their knowledge of and compliance with the child protection law. With a return rate of 37%, Swoboda and his colleagues found that 17% of mental health providers were unfamiliar with Nebraska's child abuse reporting statute and, in turn, were often noncompliant with it. In a more specific assessment of professional familiarity with reporting laws, Williams, Osborne, and Rappaport (1987) surveyed a range of professionals, including psychologists, about their knowledge of the elements found in Louisiana's reporting statute. Results showed that respondents were aware of mandatory reporting laws and were inclined to comply with them. Current research supports the view that almost all psychologists are familiar with their respective reporting statutes (Beck & Ogloff, 1995; Brosig & Kalichman, 1992a). Taken together, these studies suggest that psychologists have become increasingly familiar with their reporting obligations; however, knowledge of the law does not necessarily correspond with reporting behaviour (Beck & Ogloff, 1995; Kalichman & Craig, 1990; Kalichman et al., 1989; Swoboda et al, 1978). To further investigate the failure of mental health professionals to comply with the duty to report, several studies have asked practitioners to assess the relative importance of the law in making their reporting decisions. Muehleman and Kimmons (1981) surveyed 39 psychologists and found that compliance with the law was most often rated the least important factor affecting their decision to report abuse, with the child's life and the protection of confidentiality being their paramount considerations. Likewise, Kalichman and Craig (1991) found that psychologists most frequently indicated protecting the child as the most important factor while adherence to the law was less influential. Still other research suggests that adherence to the law differs between professionals and is based on their personal history of reporting. Kalichman and Brosig (1993), for 37 example, found that those professionals who consistently reported child abuse were guided by an interest to uphold the law and protect children, whereas psychologists who have inconsistently reported were more influenced by situational characteristics, such as confidence that abuse has occurred and concerns about the effects of reporting on the family. This result is similar to Wilson and Gettinger's (1989) finding that school psychologists who were willing to report abuse were guided, more than non-reporters, by legal issues. Another legal factor that has gained attention in the literature is the effect of specific statutory wording on decisions to report child abuse. In one study, Kalichman and Brosig (1992) found that the tendency to report child abuse decreased when the law was narrow in its reporting requirements, whereas tendencies to report abuse increased consistently when the law was more broadly defined. Kalichman and Brosig concluded that specific reporting requirements affect reporting, and that broadly defined statutes may result in over-reporting. These results have been replicated by Brosig and Kalichman (1992b) who found that psychologists were less inclined to report an abusive adult when the law was limited in scope. Clinician Characteristics Along with legal factors, clinician characteristics have also been found to influence child abuse reporting decisions. One characteristic of clinicians that has been associated with mixed results in reporting decisions is training in child abuse. A survey by Nightingale and Walker (1986) found that professionals who have previous training in child abuse identification were more likely to suspect and report cases of child abuse. In contrast, Kalichman and Brosig (1993) surveyed 226 licensed psychologists from two states and found that psychologists who had received child abuse training in workshops and continuing education were more likely to have not reported child abuse than those without such training. Similarly, a survey by Pope and Bajt (1988) suggest that breaking a 38 law or formal ethical principle occurs even among the most well trained psychologists, with 7 (21%) of the 34 identified breaches involving a refusal to report child abuse. Wilson and Gettinger (1989) also found that degree of training was not significantly related to reporting child abuse. Collectively, these findings suggest that the decision to report child abuse is not straightforward, and that even well-intentioned and trained professionals struggle, in some cases, over whether to report every suspicion of child maltreatment. In addition to training in child abuse detection, the amount of professional experience seems to be a factor relating to compliance with the reporting law. Barksdale's (1989) qualitative study of 10 psychotherapists found that those professionals with less work experience tended to experience more conflict in reporting, and were therefore less willing to report suspected abuse. Therapists with more experience had more confidence that reporting would not ruin the therapeutic relationship. In contrast, a study by Haas, Malouf, and Mayerson (1988) found that psychologists who had fewer years of experience chose to report sexual abuse. The authors suggest that "more experienced therapists [may] develop a somewhat greater level of cynicism about their ability to actively intervene and change circumstances" (p.41). Many clinicians are concerned that reporting will harm the therapeutic relationship (Ansell & Ross, 1990; Kalichman et al., 1989). A survey by Finkelhor, Gomes-Schwartz, and Horowitz (1984) found that 52 percent of mental health professionals admitted that they failed to comply with the mandatory reporting law by not reporting their most recent case of sexual abuse. The authors speculate that mental health workers weigh the promise of confidentiality more heavily than their duty to report. Another clinical characteristic that likely affects reporting behaviour is clinicians' attitudes and beliefs about the reporting legislation and child protection system. For example, Swoboda et al. (1978) found that 87%> of psychologists would not report a vignette describing a father's continued mental and physical abuse of his children. The authors concluded that a lack of familiarity with the law was probably not the most 39 important factor in failing to comply because Nebraska's reporting statute was provided in the first part of the survey. Instead, negative attitudes toward the law may be more of a barrier to reporting child abuse than is ignorance. A recent study by Beck and Ogloff (1995) found that 12% of British Columbia psychologists sampled admitted that they failed to report a case of suspected child abuse in the past year; lack of confidence in child protection services was a key factor in their reasoning not to report. The reviewers cited above have come to different conclusions regarding the influence of clinician characteristics on decision making. It is likely that some of these differences are due to weak and discrepant methodologies. Clearly, much more research is needed in this area—particularly research that employs methods other than the commonly employed written analogue design. Situational Factors Another key factor to understanding the clinical reasoning of mental health professionals is the situation or set of conditions that are associated with the abuse. These factors include characteristics of the victim, type of abuse, and amount of evidence available (Brosig & Kalichman, 1992a). In a review of the literature, specific characteristics of the victim appear to influence reporting. Although Kalichman et al. (1989) found that the child's sex had no effect on clinicians' tendencies to report, other studies have found that the child's age, race, and social class can influence reporting decisions. Kalichman and Craig (1991) investigated 328 psychologists from Minnesota and Oklahoma and found that clinicians were more likely to report a younger victim being physically abused than an older victim experiencing the same maltreatment. Results of a Hampton and Newberger (1985) study show that child abuse is under-reported among white and affluent families. Still other research suggests that the child's reaction and verbal disclosures of abuse are important. For example, Kalichman, Craig, and Follingstad (1988) found that a child responding to 40 questions about her family by crying and acting fearful resulted in many clinicians tending not to report. In contrast, studies have shown that professionals will report a child who makes a verbal statement of abuse (F inlay son & Koocher, 1991; Kalichman & Craig, 1991; Kalichman et al., 1988, 1989), unless the child withdraws the statement, in which case clinicians' tendencies to report significantly lessen (Attias & Goodwin, 1985; Zellman, 1992; Zellman & Antler, 1990). A second variable found to influence reporting is the type of abuse suspected. Wilson and Gettinger (1989) found that school psychologists were more likely to report physical and sexual abuse than neglect or emotional abuse. Beck and Ogloff s (1995) survey of psychologists replicated this finding. Similarly, Williams et al. (1987) found that physical abuse was more likely to be reported than psychological abuse, and Nightingale and Walker found that sexual abuse was more likely to be reported than neglect. These findings suggest that there are differential reporting tendencies among professionals that are based on the type of abuse, with sexual and physical abuse being more suspected and reported than emotional abuse. It is likely that sexual and physical abuse are more often reported because they are easily identified by mental health professionals, and can more easily be substantiated in court. Emotional abuse, by contrast, may be more difficult to define, identify, and substantiate (Melton & Davidson, 1987). A growing body of literature suggests that the clinicians' level of certainty that abuse has actually occurred influences reporting decisions. Kalichman et al. (1988) found that 89% of mental health professionals who failed to report suspected child abuse indicated that they lacked confidence that child abuse was occurring. Similarly, Finlayson and Koocher (1991) found that reporting decisions increased as a function of level of suspicion. A recent study by Beck and Ogloff (1995) found that the degree of certainty that abuse was occurring accounted for almost 70% of the variance in decisions to report, using clinical vignettes. Of the psychologists who indicated that they did not report child abuse in the past year, the belief that there was not enough evidence to do so was the most 41 common reason for not reporting. Taken together, these findings suggest that degree of certainty to report may be related to the amount of evidence available to the clinician. Support for this concept comes from Finlayson and Koocher (1991), who showed that specific signs of sexual abuse were more likely to be reported than diffuse and non-specific signs. Finlayson and Koocher's findings also suggest that reporting tendencies occur along a continuum. Support for this interpretation comes from Muehleman and Kimmons' (1981), who found that professionals were more inclined to report as evidence of abuse increased. Still other research shows that observable signs of abuse (e.g., a bruise) increase tendencies to report (Kalichman & Brosig, 1992). As previously indicated, several studies have shown that a child's verbal disclosure of abuse influences reporting tendencies, as well (Finlayson & Koocher, 1991; Kalichman & Craig, 1991; Kalichman et al., 1988, 1989). Other research has found that a child who is currently being abused is more likely to be reported than a child who has suffered abuse in the past (Wilson & Gettinger, 1989). In summary, these results show that a variety of situational factors influence reporting decisions. It is not known, however, to what extent differences in findings are due to methodological differences, and to what extent these findings demonstrate the complex interactions between variables. Methodological Limitations of Previous Research On the basis of this review, several methodological limitations of existing research warrant mention. To date, research on mandatory reporting of child abuse has relied primarily on analogue methods, (i.e., paper-and-pencil self-report surveys). Barksdale's (1989) study that interviewed a group of psychotherapists is an exception in the literature. Because little naturalistic research has been conducted on reporting child abuse, analogue researchers have tended to hypothesize about and examine factors believed to relate to child abuse reporting without having those hypotheses deeply informed by data from field 42 studies. As a result, knowledge of reporting behaviour is limited because professionals can only respond to the factors that are presented to them, often in analogue form. Second, there are four major threats to external validity with the analogue research reviewed here. First, the extent to which participants' responses to the vignettes match their actual clinical behaviour is questionable. It is possible that responses reflect preferred or ideal choices rather than actual behaviour. Second, the amount of information provided in most of the clinical vignettes is limited, and respondents have no opportunity to obtain additional information before rendering a decision. In actual practice, psychologists typically evaluate their suspicions with follow-up questions and cannot do so when responding to written clinical vignettes. A third limitation with the vignette studies is that the data obtained may not generalize beyond the particular descriptions of the vignettes presented. Finally, it is conceivable that the vignettes focus on dilemmas that may be unfamiliar or irrelevant to some respondents. A final limitation with research in the literature is that no study has attempted to develop a model of the criteria that psychologists use in decisions to report possible child abuse. At present, the key factors that are used in actual reporting decisions are unclear. Therefore, it seems appropriate to use a qualitative approach to explore how reporting decisions evolve, what factors influence reporting, and the constraints to reporting. This knowledge would have important implications for both research and practice. The purpose of the present study was to avoid the methodological limitations of past research, and produce an empirically based model of reporting criteria. Thus, I used Ethnographic Decision Tree Modeling (Gladwin, 1989) to answer the following research question: What factors contribute to psychologists' decisions to report cases of possible child abuse? An overview of EDTM is presented next. 43 Overview of EDTM There are two key stages in ethnographic decision tree modeling (Gladwin, 1989). The first stage involves building a decision tree model, which identifies key decision criteria and outcomes for a group of people. The second stage involves testing the preliminary model to determine its predictive ability. These two stages will be discussed, along with a brief description of the steps involved. There are five steps involved in building a decision tree model. Once the decision to be studied is clearly defined, the first step is to specify the decision options (i.e., report or not report child abuse). The second step in constructing a decision tree is to conduct preliminary, or pilot, interviews. The purpose of the pilot interviews is to gain experience in following the procedures of the method. Furthermore, pilot interviews serve to determine the usefulness and application of a decision tree method for a particular research question. The third step involves selecting a sample of decision-makers and interviewing them. The fourth step in developing a decision tree model is to elicit the criteria that decision makers use in their decision process. Decision criteria are the particular factors, or situations, which cause decision-makers to select one decision option over another. Discovering decision criteria involves two key steps (Gladwin, 1989): (a) the interviewer looks for contrasts (or exceptions to the rule) over decision makers, space, or time; and (b) once a contrast in decision behaviour is found, the interviewer elicits the decision criterion upon which it is based. Once criteria have been identified, the next step is to draw a decision tree for each participant interviewed. Each individual decision tree is a sequence or series of explicit statements (or hypotheses) about the conditions leading to the selection of each decision option or outcome. Each decision criteria is passed or satisfied along a path to the particular choice. To ensure that each decision tree is descriptively valid, the researcher tries to use the language and categories that decision-makers use themselves. Individual trees are later combined to form a group decision model. Decision criteria are arranged in a hierarchical order so that they are logical and predictive of the choices of the 44 sample. The selection and ordering of decision criteria is preliminary at this stage. The validation stage will show incorrectly specified and ordered criteria. As with the development stage of a decision model, the verification stage involves a series of steps. The first step in this process is to design a questionnaire to test the model. Once the questionnaire is designed, the second step is to select an independent sample of decision-makers whose data will test the model. The questionnaire is composed of one factual question regarding the decision outcome and a series of yes/no questions that embody the decision criteria, or choice points, of the model. The factual question serves to elicit the decision outcome from participants before asking them about the reasoning behind their decision. For example, the entry condition in the present study was, "Did you report child abuse or neglect?" Participants are then asked to respond to specific criteria leading up to their choice to report (e.g., "Were there any signs of or risk factors for child abuse or neglect? "). The third step in verifying the model is to calculate the error rate. Errors constitute the number of times the model is not able to explain a decision made by a participant. For example, the model is in error, based on the answers to the test questions, when it predicts that a person will not report when in fact the person did report. By counting the number of errors the model makes, the researcher can determine the accuracy of the model. Validation of the model is achieved when 85% or more of decisions are predicted (Gladwin, 1989). If the model predicts less than 85% of decisions, the model may require modification. Modification of the model may involve adding new criteria or reordering the path of the tree. Alternatively, decision criteria can be generalized. That is, two or more decision criteria are combined into one criterion (i.e., a specific category is subsumed under a more broader or general category). The purpose of generalizing decision criteria is to (a) eliminate redundancy, and (b) cluster the decision criteria logically. Finally, if new decision criteria are added, then the new model is tested with a new, independent sample of participants. 45 Recently, ethnographic decision tree modeling has been applied to medical and psychological choices. For example, studies have investigated the decision making processes of drug users who share needles and are at risk for HIV (Johnson & Williams, 1993), health care choices of cancer patients (Montbriand, 1995), infant feeding behaviour (Bauer & Wright, 1996), and treatment choices of childhood diarrhea (Ryan & Martinez, 1996). Child abuse reporting would seem to be a fruitful area to apply EDTM for several reasons. First, given the prevalence of child abuse in North America, there is a need to identify the factors that influence reporting decisions. Second, much of what we know about decision criteria is limited by the analogue methods used. Typically, participants are asked to respond to controlled vignettes in which the researcher manipulates certain factors. Instead, EDTM, as a qualitative method, allows participants to describe the full range of criteria that they use in their decision making in actual cases. Third, given that there is much communication involved in the reporting of child abuse (e.g., with families, colleagues, child protection services), the process through which these decisions are made is likely to be accessible through verbal elicitation, such as ethnographic interviews. Fourth, because of the potential adverse effects of failing to report child abuse (balanced with an interest to make reports that can be substantiated by child protection services), it is believed that psychologists rely upon specific standards for making decisions to report possible child abuse. In what follows, I attempt to develop a model of the factors involved in the reporting of child abuse in a specific cultural context. First, however, I will describe the procedures used in the design of the study. 46 Chapter III Methodology This chapter presents the EDTM methodology used in studying the factors psychologist's use when reporting cases of possible child abuse. This chapter is divided into two sections: model development and model validation. The procedures for selecting and contacting participants, data collection, and data analysis are presented for each stage of the research. Phase 1: Model Development Participants Interviewees. Twenty registered psychologists in British Columbia participated in the model development stage. Of the twenty psychologists, 13 were female and 7 were male. All but one of the participants had received a doctoral degree in counselling or clinical psychology; the other participant received a master's degree in a psychology-related discipline. Participants mostly worked in medical hospitals or in private practice. Table 1 shows the gender, professional degree, and place of employment of participants. Interviewer. Because the characteristics and viewpoints of the interviewer are an important part of the qualitative research process, a description of the interviewer is provided. I served as the primary researcher and interviewer for all data collection and analysis. I am a white, 3 3-year-old male, counselling psychology student with 7 years experience working as a counsellor. At the time of the study, I worked half-time in a university-based counselling centre providing personal, academic, and career counselling. I have published two studies in the area of child abuse reporting, and I am familiar with research in the field. 47 Table 1 Demographic Characteristics of Participants n % Gender Female 13 65 Male 7 35 Professional Degree Obtained Ph.D. 16 80 Ed.D. 3 15 M.A. 1 5 Place of Employment Medical Hospital 6 30 Private Practice 4 20 University/College 2 10 Rehabilitation 2 10 Mental Health Centre 2 10 Community Agency 1 5 Psychiatric facility 1 5 School system 1 5 Corrections 1 5 Note: to become a registered psychologist in British Columbia, a candidate must have completed a doctoral degree in psychology. However, at the time the legislation was enacted in 1977, there were many psychologists practicing at the master's level. Therefore, these individuals became registered psychologists under the grandfather clause. Committee Members. Three university professors served as committee members. Two professors were from counselling psychology, one of whom was female and the other male, and the other 48 member was male and from clinical psychology. The primary supervisor was female and has extensive knowledge of professional ethics and has written journal articles in the area. Another committee member holds a law degree and a doctoral degree in clinical psychology. He has expertise in legal issues and professional ethics, and has written extensively in these areas. The final committee member has expertise in qualitative methodology as well as counselling practice. Procedure Pilot Study. Four interviews were conducted prior to the beginning of the research project. Of the four participants that were interviewed, two held Ph.D. degrees in counselling psychology and were working as counsellors. The other two participants had obtained master's degrees in social work and were also working as clinicians. Data were transcribed and reviewed by the researcher and senior supervisor. Analysis revealed a number of criteria that were used in participants' reporting decisions, making it possible to develop preliminary models. Therefore, it was decided that the project could formally proceed. The transcriptions of the pilot interviews were also used to help train the interviewer in the procedures of EDTM and data analytic strategies. The pilot data was not included in this study because participants did not meet the criteria for participation (see below). Criteria for Selection. Registered psychologists from British Columbia were selected for participation in this study. Psychologists were chosen for participation for three key reasons. First, psychologists, by the nature of their work, are in a position to encounter cases of potential child abuse. Second, in light of some research suggesting that psychologists do not always report suspected child abuse (Barksdale, 1989; Beck & Ogloff, 1995; Finkelhor et al., 1984; Haas et al., 1988; Kalichman & Craig, 1991; Kalichman, Craig, & Follingstad, 1989, 49 1990; Swoboda et al., 1978; Williams et al., 1987), I felt that it was important to better understand the factors that guide their decisions to report possible of child abuse. Third, I sought to extend my previous research in this area. There were several criteria upon which psychologists were selected. First, in an effort to maximize the homogeneity of the group, registered psychologists from the College of Psychologists of British Columbia (CPBC) were selected. In British Columbia, to become a registered psychologist, a candidate is required to demonstrate competency in core areas (e.g., knowledge of psychological theory, research, and ethical codes/laws pertaining to psychologists). Second, psychologists who practice in the Lower Mainland of British Columbia were specifically sampled because they represented the largest data pool for this homogeneous group of psychologists. Third, it was required that participants be able to recall a specific time in which they made a report of child abuse, and articulate the factors that were important in their decision(s). Due to potential errors in memory, psychologists were asked to recall a case within the past two years. In some cases, participants were able to describe case specific details that dated back three years and, therefore, were included in the study. Fourth, participants were asked to talk about a case that occurred in British Columbia. In this way, variations in reporting laws across jurisdictions were held constant. Finally, it was required that the actual case be resolved so that the participant could speak from beginning to end about their decision making process of abuse reporting. It is important to note that I made no attempt to actively recruit participants who did not make a report of possible child abuse or neglect, and their respective decision criteria. There were several important reasons for this decision. First, I was concerned about learning information about a child who had been harmed, abused, or was otherwise in need of protection, and whether I would be required by law to report. In this way, the promise of confidentiality to participants could be compromised. Second, given that blatant non-reporting is a criminal and ethical offense, I questioned the extent that psychologists 50 would volunteer to participate in the study in which there was no protection of anonymity. Psychologists were asked to participate in a face to face interview. Finally, because of potential ethical and/or legal implications for psychologists in discussing a failure to report case, I questioned the validity of data provided by such participants. Contacting Participants. Prior to the start of the study, an advertisement was placed in the BC Psychologist newsletter in the Fall of 1997 (see Appendix D) to make prospective participants aware of the upcoming research. One hundred and twenty-five (of a pool of approximately 600) registered psychologists were randomly selected from the 1996 Directory of the College of Psychologists of British Columbia. Psychologists were mailed letters describing the study with an invitation to participate (Appendix E). To increase response rates, a second letter was sent out three weeks later (Appendix F). As part of the letter, psychologists were told that they would be called by telephone unless they indicated that they did not wish to participate in the study. Once telephone contact was made, I introduced myself and briefly explained the nature of the research. I then asked prospective participants if they had reported a case of possible child abuse or neglect and if they were able to recall the factors that were important in making that decision, and if they would be willing to be interviewed in person. If a participant agreed to be interviewed, I then set up a date, time, and place to meet with that person. In virtually all cases, I visited the place of work of the participant. During the initial phone contact, I requested that the interviews be audiotaped, and that participants complete the informed consent form (Appendix G) prior to the beginning of the interview. Prior to the interview, participants were encouraged to think about the factors that guided their decision(s) to report. The goal here was to ensure that participants reflected carefully before the interview. In fact, it was my clear impression that all 51 participants had thought a great deal about their case(s) before the interview, with many of them consulting case notes and sometimes even quoting from them. Of the 125 letters sent out, 21 (or 17% of) psychologists agreed to be interviewed in person. Nine letters were returned by the post office as undeliverable, 33 psychologists were unresponsive to initial mailings and follow-up phone calls, and 3 psychologists indicated that they would be out of the country during the time of the research. Forty-six psychologists declined to participate because they did not have any reporting experiences within the last 2-3 years, and 13 declined for other reasons (e.g., did not work as a clinician, too busy). Interview Protocol. Interviews lasted between 40 and 90 minutes, and began with a warm-up exercise about current job activities (i.e., "Please tell me a little bit about what you do in your work?" "How long you have worked here?" "What kinds of clients do you see?"). The purpose of this exercise was to develop rapport so that participant's could more easily talk about their experiences, as well as to assist me in my understanding of the context surrounding the report. I then reviewed the informed consent form, reminded psychologists of the taping, and fielded any questions that emerged. Participants received their own copy of the consent form. Interviews began with the orientation statement. The orientation statement was intended to help participants to describe their experiences in reporting. The orientation statement read as follows: Deciding to report child abuse is often complex and involves consideration of many factors. In order to understand more about the factors that are important in reporting child abuse, I would like to learn about your professional experiences where abuse was suspected. This includes times 52 when you have been quite certain that abuse has occurred, as well as cases that have not been as clear-cut. During the time we have together, I would like to hear your story, from beginning to end, of a time when you reported a case of child abuse. So, with as much detail as you can recall, I would like you to tell me how you decided to report child abuse, and what factors were important in making your decision. If you have a number of cases in mind, I want you to focus on one case until you are finished. Then, you may begin to tell me another experience, and so on. I'll have some questions to ask of you later. Please do not use actual names in describing the case. Do you have any questions before we begin? If not, you may begin now. Once participants began to talk about their reporting experience, I did not interrupt until they were finished telling their story. The purpose of this procedure was to capture participants' natural story without influence. I did, however, probe further when participants' descriptions were unclear or incomplete. However, most participants provided case descriptions that were quite detailed and comprehensive. Throughout the interview process, participants were asked to link their decision to specific material from their cases, rather than responding in terms of hypothetical abstractions. During the interview, I recorded the criteria that participants identified as important in their decision, and the order in which the criteria were considered. After participants were finished describing the factors that guided their reports, I asked open-ended questions to clarify and amplify decision points. Follow-up questions also provided an opportunity to explore what, if any, other factors played a role in reporting the case (as opposed to not reporting the case). Participants often contrasted the reasons why they did report with reasons why they would not have reported. When participants had multiple reporting experiences, I asked them if there were any differences in criteria between cases. At the end of each case, I re-narrated the decision criteria, and order of criteria, back to 53 participants so that they could confirm if I had correctly identified their criteria. Although I did not keep a record of the number of times participants corrected my summaries, it was my sense that this occurred in very few cases. Eliciting Decision Criteria. To elicit decision criteria, Gladwin's (1989) method was followed. Given that only one sub-group of participants were sampled (i.e., only those psychologists who reported child abuse), I elicited contrasting data in the following ways. First, child abuse reporting was conceptualized as a process that has a number of stages; it begins with "don't report" and over time leads to "report." Often decisions to report evolve over days, weeks, and sometimes months, and they are dependent upon the information that the psychologist has available to him/her. In the present study, psychologists who ultimately reported child abuse often experienced times, within that particular case, when they did not report. Second, during the ethnographic interviews, participants were asked why they decided to report (at the point that they did) as opposed to not report. This procedure is referred to as contrastive questioning. Third, participants were invited to compare and contrast decision criteria with other cases that they have experienced. In this way, participants served as their own contrast (i.e., within person differences). Frequently, participants had multiple occasions when they reported child abuse; it was my impression that psychologists had a well-developed sense of the criteria that they use to both report and not report. Fourth, similar to the last point, contrast data were also collected between participants. That is, while all psychologists in the study did ultimately report their case, they did so for different reasons. Hypotheses about criteria were confirmed across participants when I checked the criteria with participants. 54 Transcripts. All interviews were transcribed verbatim for each participant by a professional typist. Each participant was assigned a code number to maintain confidentiality. I listened to each tape and corrected the transcripts. Data Analysis. All transcripts were read in their entirety by the researcher. Of the 21 interviews conducted, one interview was eliminated because the data lacked adequate detail as the participant had to leave abruptly. Since participants had from one to three cases when they reported child abuse, the unit of analysis was each individual case for which a report was made, not the actual number of participants. Although this procedure gave more weight to the criteria of individuals who reported more cases of child abuse, analysis showed that the additional cases served to validate previously established criteria rather than to contribute new criteria. Thus, 37 cases were generated from the 20 participants. Three of these cases were eliminated; one was a hypothetical story, one was too general to use, and the other one involved someone else making the actual report of child abuse and not the participant. According to Gladwin (1989), 20-25 is an acceptable number of cases to develop the preliminary model. On the basis of the transcripts, I identified criteria that participants themselves described as influencing their decision making process. In identifying criteria, great effort was made to use the psychologists' own words, and to avoid making inferences. For the first several transcripts, the primary researcher and senior supervisor met and openly examined all possible criteria and the wording of criteria. In this way, we generated a shared view of the core criteria for the first several cases. I sent a copy of selected interview transcripts to the other committee members for their review and feedback. Their goal was to examine the selection and wording of criteria. The feedback provided by committee members was examined with the aim of arriving at the most accurate and 55 effective identification of decision criteria. When new criteria were identified, I reanalyzed previous cases (for which those criteria had not been identified). For the remaining cases, I met regularly with the senior supervisor to discuss whether criteria had been appropriately identified, and whether the wording was clear and concise. With respect to the decision outcome of individual models (i.e., "report" "don't report"), I considered (in advance of the study) the possibility of a third option: "consult with the Ministry for Children and Families" (hereafter referred to as "the Ministry"). I treated such cases as the psychologists' attempt to gather more information to aid in the decision making process. After consulting with the Ministry, it was believed that psychologists would still be in a position to either report or not report the incident. Ordering and Charting Criteria. Once identified, decision criteria were arranged and charted in sequence. There were several checks in place to determine the order of criteria. First, because participants were asked to describe their case and respective decision criteria from beginning to end, the sequence of criteria appeared natural in participants' stories. Second, as mentioned previously, I provided summaries of the criteria and order of criteria for each case. Participants were invited to correct or confirm my identification and order of criteria. Third, there was a logical manner to the ordering of criteria. For example, detecting signs of possible child abuse naturally proceeds consideration of whether these signs meet a legal threshold to report. As a final check to the ordering of criteria, committee members continually provided feedback in not only the identification of criteria but the sequencing as well. As with the identification of criteria stage, the comments of committee members were incorporated, and revisions were made accordingly. Again, I went back to previous transcripts and reanalyzed the order of data. This process was repeated until the senior supervisor and I felt that no further revisions were needed. 56 Saturation. Data collection stopped when it was determined that saturation was reached. Saturation occurs when no new criteria are added in the development of individual models. I sensed that the data were stable after 24 individual models had been developed and that additional cases were unlikely to change the results in any significant way. I then tested saturation by collecting additional data. No new criteria were added as a result of 10 additional cases. Therefore, I decided that I would proceed to develop the group model. Development of the Group Model. After individual models were developed, the purpose of the next step was to construct a group decision model. To develop the group model, I examined each criterion that was represented in every individual model. By proceeding sequentially through each of the 34 individual models, I created group decision points. Group decision points reflected one content area or idea. Often group decision points were generalized to include a broader range of idiosyncratic factors; however, I did try to preserve the essence of psychologist's meaning when developing these broader categories. For each decision criterion, I counted the number of cases that applied to that criterion. To begin the group model, I started with one that applied to every case. Similarly, a second identified criterion applied to all but one case and was a natural extension of the first criterion. I continued with this process for all of the core criteria, using a combination of frequency counts and logic. After each criterion was identified, I then reviewed all of the individual models to see if any criteria were omitted. Additionally, I did a final reading of each transcript and notes to ensure that all criteria were represented and that they appeared in sequence. Next, I met with the senior supervisor to examine all possibilities regarding the core criteria and wording. When new decision points were created, I reanalyzed all previous cases. Changes in the group model were discussed and agreed on with the senior supervisor before being 57 finally accepted. After several revisions, six criteria were identified in the group model. These six criteria will be fully described and explained in the results section. Phase 2: Model Validation Participants Interviewees. Twenty-one registered psychologists participated in the model validation stage of the study. Participants were 10 females and 11 males. Seventeen participants held doctoral degrees. Participants mostly worked in private practice, rehabilitation, and school settings. Table 2 shows the gender, professional degree, and place of employment of participants. Procedure Pilot Study. To review the procedures for the validation stage, one pilot interview was conducted with a registered psychologist. After the informed consent form was reviewed and signed, the participant was provided with a copy of the questionnaire (described below) and asked to complete it. Next, the participant was invited to talk about the case so that additional data could be collected. Upon review of this procedure, I decided to switch the order of tasks so that participants first talked about their case and then were asked to answer the research questions. Having participants describe their cases at the beginning of the interview served to stimulate participant's recall of events and criteria prior to answering questions about the case. I facilitated the completion of the questionnaires by asking participant's the questions myself. This process made it clear to determine when the model was in error. The question period ended once a decision outcome was reached. 58 Table 2 Demographic Characteristics of Participants % Gender Female 10 48 Male 11 52 Professional Degree Obtained Ph.D. 16 76 Ed.D. 1 5 M.A./M.Ed 3 14 B.A. 1 5 Place of Employment Private Practice 7 33 University/College 2 10 Rehabilitation 3 14 Mental Health Centre 1 5 Community Agency 1 5 Psychiatric facility 3 14 School system 4 19 Note: psychologists with a degree other than a doctoral degree were grandfathered as registered psychologists. 59 Criteria for Selection. As with the model development phase, the same procedure for sample selection was used in the model validation phase with the exception of two changes. First, psychologists were randomly selected from the 1998 Directory of the College of Psychologists in British Columbia and not the 1996 directory. The new directory was published during the time of the ongoing research project. The new directory provided more accurate information on place of work, current address, and the inclusion of new members. Second, psychologists that were randomly selected to participate in the first phase of the study were eliminated from the pool of prospective participants in the second phase of the research. In this way, a separate (yet similar) group of registered psychologists, not part of the model development stage, were sampled to test the preliminary model. Contacting Participants. After removing earlier names from the list, one hundred (out of approximately 475) registered psychologists were randomly selected from the 1998 Directory of the College of Psychologists of British Columbia and were mailed a letter requesting their participation (Appendix H). Psychologists were informed of the nature of the research project and were invited to participate. As part of the letter, I indicated that I would make a telephone follow-up if I did not hear back from prospective participants. This procedure was used to increase response rates. Prior to the validation stage, I established that approximately 30 reported cases would be sufficient to test the group model. This number is consistent with Gladwin's (1989) recommendations. Of the one hundred psychologists who were randomly contacted to participate in the model validation stage, 23 agreed to participate and were interviewed. This represents a response rate of 23%. Fifty-six psychologists indicated that they had not had an occasion to 60 consider making a report or that they did not work as a clinician. Eighteen psychologists did not respond to the initial mail out and follow-up phone calls, two letters were undeliverable, and one psychologist had recently died. Interview Protocol. Interviews lasted between 30 and 60 minutes. Interviews began with the same warm-up exercise and procedures (i.e., reviewed informed consent form, talked about taping, fielded questions, read the orientation statement) as with the model development stage. Next, participants were asked to describe an occasion in which they reported possible child abuse in their professional work. This procedure was intended to collect additional information for each criterion, and to stimulate participant's memory so that they could more easily answer the research questions. Once participants described details of their case, they were asked to answer a series of questions about that incident. Questionnaire Development. The questionnaire included one factual question regarding the decision outcome, and a series of yes/no questions that embodied the decision criteria identified in the model development stage. The factual question was the starting point; it was intended to elicit the decision outcome (i.e., reported abuse) from participants before asking them about the reasoning behind their decision. Next, participants were asked to answer questions on the specific criteria leading up to their decision to report (e.g., "Were there any signs of or risk factors for child abuse or neglect?"). Questions were asked in the order diagrammed in the model (Appendix I). Each time a response diverged from that predicted by the model, I collected additional information to help identify the error in the model. Data Analysis. Of the 23 participants that were interviewed, two were eliminated because participants did not satisfy the requirements of the study. One participant said that while 61 she was involved in the reporting decision, she did not formally make the report herself. The other case was reported in another province in Canada. Again, the unit of analysis at this stage was each individual occasion when a report was made, not the actual number of participants. Thus, 36 cases were generated from 21 participants. Gladwin (1989) reports that 25-30 cases are adequate with a homogeneous group. To calculate the accuracy of the decision model, I counted the number of times when the model did not explain the decision of a participant. The accuracy of the model was determined by dividing the number of errors by the total number of cases, yielding a percentage. 62 Chapter IV Results The results of this study are divided into two parts. The first part presents the findings from the model development stage, which identifies the factors involved in psychologists reporting decisions. The second part of this chapter presents the questionnaire data that sought to validate the preliminary model. Qualitative and quantitative data from both stages will be provided to illustrate each decision criterion represented in the model. This section concludes with some suggestions to improve the present model. Phase 1: Model Development Using the methodology of EDTM, a model was developed that identifies the criteria that psychologists used when reporting cases of possible child abuse. Based on their reporting experiences, a series of six criteria were identified. These six criteria are: (1) Were there any signs of or risk factors for child abuse or neglect? (2) Did the signs or risk factors meet your threshold to report as you understand the law? (3) Was there some other value to report other than a legal one? (4) Were you concerned that reporting would cause harm? (5) Were you able to minimize the harm that would result from reporting? and (6) Did the reasons to report outweigh the reasons to not report? The group model is presented in Figure 1. The boxes in Figure 1 represent each decision criterion, and the diamonds are the decision outcomes (i.e., either to report or not report). The criteria follow a hierarchical order that matches the chronology of participants' conversations. Along the arrows, "Yes" and "No" responses outline the sequence that criteria are considered until the decision outcome is reached. The dotted line portrays a tentative link in the model, whereas the 63 solid lines represent a stronger link. It is important to note, in advance, that no causal sequence is established from the data, although the model does suggest a plausible order. The following description outlines how a given individual would proceed through the model. The model begins with the psychologist becoming alerted to possible child abuse. At this point, the psychologist identifies some sort of sign or risk factor (the word "indicator" is used synonymously for signs and risk factors) for abuse. Without any sort of indicator, there would be no basis for a report. Once an indicator is noted, the psychologist then assesses the extent to which the indicator meets his or her legal threshold to report. In other words, do the indicators fulfill the psychologist's legal definition of a reportable circumstance? If the indicator does not meet the psychologist's legal threshold to report, the psychologist may still report if he or she determines that there is some other value to report (e.g., therapeutic benefit, moral obligation). If the psychologist does not perceive that there is any other value in making a report, then the model predicts that a report is not made. If the legal threshold to report is met, the psychologist then considers the potential harm in reporting the case. If the psychologist perceives that no harm is associated with reporting, then the psychologist reports at this point. If, however, the psychologist believes that reporting will cause some sort of harm, then the next criteria is considered which asks if the psychologist can minimize the harm. If the psychologist is able to minimize the harm, a report is then made. If the psychologist is unable to minimize the perceived harm, the psychologist advances to the next decision point which involves balancing the reasons to report against the reasons to not report. If the reasons to report outweigh the reason to not report, the model predicts that the psychologist will report. If not, the model suggests that the psychologist may not report. 64 No 1 START 1 Gather In HERE 1 formation! Were there any signs of or risk factors for child abuse or neglect? Yes 1 Report unless Yes Were you able to minimize the harm that would result from reporting? No Did the reasons to report outweigh the| reasons to not report? -No-Did the signs 6r risk factors meet your threshold to report as you understand the law? 1 Yes -No-No Was there some other value to report other than a legal one? Yes 1 Were you concerned that reporting would cause harm? -Yes-Yes Figure 1. The preliminary group model for child abuse reporting. 65 In this section, each criterion will be described in the order in which it appears in the model (from top to bottom). Selected quotes are offered to illustrate and represent the existence of each criterion. In some cases, verbatim responses support more than one criterion; there is some overlap in participants' descriptions. In reviewing these quotations, the reader should note that the quotations have been taken out of context, and that other factors may also be contributing to the overall decision to report. Some participants considered all of the decision criteria represented in the model and some participants addressed only a few criteria. Table 3 shows which participants addressed which criteria in the model. It is also important to note that participants moved through the decision making process at different speeds, depending on the circumstances of the case. Criterion 1: Were there any signs of or risk factors for child abuse or neglect? The model begins with participants becoming aware of possible occurrence of child maltreatment. In all 34 cases, participants described signs and risk factors that alerted them to the existence of possible child abuse. Based on re-occurring themes that emerged from the data, I developed four categories for signs of abuse and two categories for risk factors. Signs of abuse included: (1) physical signs, (2) changes in the child's affect and/or behaviour, (3) disclosures, and (4) direct observation and other non-verbal communications (e.g., drawings, poetry, etc.). Risk factors for abuse were related to: (1) the potential perpetrator, and (2) the child. The signs and risk factors that participants discussed are presented below; however, in order to help contextualize this data, some general information about the types of abuse reported are presented first. 66 Table 3 Model Development Data Case Signs of or risk factors for child abuse? Signs/risk factors meet your threshold to report as you underst-and the law? Some other value to report other than a legal one? Concern-ed that reporting would cause harm? Able to minimize the harm that would result from reporting 9 Reasons to report outweigh the reasons to not report? Did you report suspected child abuse or neglect? A l Yes Yes Yes Yes Yes Bl Yes Yes No Yes B2 Yes Yes No Yes CI Yes Yes No Yes Dl Yes Yes Yes Yes Yes D2 Yes Yes Yes Yes Yes D3 Yes Yes No Yes El Yes Yes No Yes E2 Yes No Yes Yes Fl Yes Yes Yes No Yes Yes F2 Yes Yes Yes No Yes Yes GI Yes Yes No Yes HI Yes Yes No Yes H2 Yes Yes Yes No Yes Yes 11 Yes Yes Yes No Yes Yes K l Yes Yes Yes Yes Yes K2 Yes Yes Yes No Yes Yes LI Yes Yes Yes No Yes Yes M l Yes Yes Yes Yes Yes Nl Yes Yes Yes No Yes Yes N2 Yes Yes Yes No Yes Yes 01 Yes Yes No Yes 02 Yes Yes Yes Yes Yes 03 Yes Yes No Yes PI Yes Yes No Yes Ql Yes Yes Yes No Yes Yes Rl Yes Yes No Yes R2 Yes Yes No Yes R3 Yes Yes No Yes SI Yes Yes No Yes S2 Yes Yes No Yes S3 Yes Yes Yes No Yes Yes Tl Yes Yes No Yes T2 Yes Yes No Yes 67 Table 4 shows the types of maltreatment reported by participants. Of the 34 cases that were reported, 18 involved sexual abuse and 16 were physical abuse. Within the 34 cases, 5 involved neglect, and 2 involved emotional abuse. On several occasions, multiple types of child abuse were reported; therefore, the types of abuse identified do not equal the actual number of reported cases (n=34). Table 4 Types of Maltreatment Reported to the Ministry for Children and Families Reports % Sexual abuse 18 44 Physical abuse 16 39 Neglect 5 12 Emotional abuse 2 5 Total 41 100 Information was also gathered on the nature of abuse and neglect. Table 5 shows the frequency of the specific types of maltreatment for sexual and physical abuse. For sexual abuse, 8 participants said that the child was fondled, 3 said that the abuse involved sexual intercourse, and two mentioned anal intercourse. For physical abuse, 10 participants described occasions when the child had been hit. Less frequently, participants said that the child was kicked, pushed, and shaken. With respect to the 5 neglect cases, participants described the inability of the parent(s) to care for the needs of a child (e.g., supervision, food, over-medicating a child). One participant talked about a failure to thrive case in which a child was not growing according to developmental expectations of weight, height, and intellectual development. The two emotional abuse cases involved one parent making 68 verbal statements to the child that he was worthless, and the other involved a parent providing differential treatment to her children (i.e., favouring one child over another). Table 5 The Nature of Child Maltreatment Reported to the Ministry for Children and Families Frequency % Sexual abuse 1 Fondling 8 44 Sexual intercourse 3 17 Anal intercourse 2 11 Unspecified 5 28 Total 18 100 Physical abuse Hit with hand 10 67 Kicked 1 7 Pushed 1 7 Shaking 1 7 Unspecified 2 13 Total 15 101 Note: percentages were rounded off to the nearest whole number and, therefore, do not equal 100%. Within the various types of abuse that were reported, participant's detected a number of signs and risk factors. It is important to note that some of the indicators of abuse occurred in isolation while other signs and risk factors occurred in combination with other indicators. 69 Signs. 1. Physical Signs In 6 of the 34 cases that were reported, participants indicated that they either observed a physical sign of abuse or received a verbal disclosure that a child had been physically injured. In all of these cases, participants conducted a thorough assessment of the nature, cause, severity of the injury, and general history of abuse; they talked to children, siblings, and parents as part of their assessment. Overall, participants indicated that physical signs of abuse played an important factor in their decisions to report abuse. Participants linked physical signs to severity of abuse. For example, one participant said: .. .to me, [causing a bruise] is very clearly abuse because it has left an injury. Had it been a sharp slap but really didn't leave a mark, then that's a little more unclear because it's hard to gauge one person's description of the amount of force used and so on. This participant also added that her decision to report was guided by the father (the perpetrator) stating that he has a right to hit his children. The psychologist said, "Yes, and I think it was that [meaning the bruising], plus the father's statement.. .that he's entitled to do this." In a different case, a 7-year-old girl told her school psychologist that she had been kicked in the back and slapped in the face by her mother earlier that morning. As part of her assessment of the child's injury, the psychologist asked the child, "do you have any marks?" In further assessing the severity of the mother's actions and the child's injury, the psychologists further said: .. .as I was thinking through this, okay kicking not hitting, [kicking] seems more violent than hitting, and it wasn't a kick on the rear end... [I'm] thinking, woo kicking, on the back, mm, then I also asked her.. .is that all that happened, and [she said] no, my mother slapped me across the face.... 70 In another case, a medical psychologist observed that a young child had very low body weight for her age. In talking with the child's parents, the psychologist was told that the young child was not able to swallow. However, after observing the nurse feed the child the food that the parents said that the child does not like, the psychologist became suspicious. In following-up on these suspicions, the participant said that she further talked with the parents and the other children in the family. The psychologists said that she concluded that the child's lack of development was not attributed to a medical condition (i.e., organic lesion in the brain) but rather it was caused by the parent's interest "to maintain the child i l l . " A report of child abuse was then filed. In two separate cases, participants said that they reported child abuse based on physical evidence that was described in a doctor's medical records. In one case, the nature of the physical injury involved trauma to the muscular structure of the child's sphincter. In the other case, the participant talked more generally that the medical records had confirmed sexual abuse of a little girl . In yet another case, a client, who was seeking a divorce from her husband, had presented the psychologist with photographs of bruises on her daughter's legs, alleging that her husband was abusing the child on weekend visits. In talking to the child, the psychologist said that she was unable to verify the allegation of abuse and therefore a report was not made on that basis. In the end, the case was ultimately reported because the psychologist felt that the mother had developed a paranoid disorder and that the child was at risk for emotional abuse. 2. Changes in the Chi ld 's Affect and/or Behaviour Participants identified emotional and/or behavioural indicators of abuse more frequently (n-24) than physical injuries. These indicators tended to be salient for different types of abuse. Table 6 presents some of the emotional and behavioural indicators of sexual and physical abuse that participants described. With regard to sexual abuse, two 71 participants described emotional changes in the child. One participant said that the child was afraid of his father, and the other participant described the child as being anxious and uncertain as to how to behave in front of his father. Behavioural indicators of sexual abused were varied and included inappropriate sexual behaviour (n=l), encopresis (n=l), enuresis (n=l), aggressive behaviour (n=2), being overly affectionate (n=l), withdrawal (n=l), crying (n=2), and dissociation (n=2), In addition, one participant talked about a child having poor personal boundaries and poor attachment. Emotional signs of physical abuse that were noted were the same as for sexual abuse. Two participants talked about children being afraid of their dad, and two participants talked about the child being very angry. Behavioural signs of physical abuse included hyperactivity (n=2), aggression toward other child and siblings (n=2), and crying (n=l). 3. Disclosures Participants relied heavily on verbal disclosures in their decisions to report possible child maltreatment. In fact, 26 of the 34 reported cases involved a verbal disclosure. Table 7 shows the source of the disclosure and the frequencies. I distinguished between the disclosures in three ways: (1) general description of the abuse, (2) case specific details of the abuse, and (3) emotional reaction of the child at the time of the disclosure. 72 Table 6 Emotional and Behavioural Indicators of Child Maltreatment Type of Emotional Behavioural Abuse Indicators Indicators Sexual Abuse Unduly anxious Age-inappropriate sexual behaviour Fear of perpetrator Aggressive behaviour Poor personal boundaries Poor attachment with others Encopresis/enuresis Overly affectionate with father Crying Dissociation Withdrawn Physical abuse Afraid of perpetrator Hyperactive Angry Aggressive toward other children Anxious Crying 73 Table 7 Source of Verbal Disclosure Frequency % Child 11 38 Adult perpetrator 7 24 Non-offending parent 3 10.5 Family member 3 10 Victim who is now an adult 1 3.5 Child perpetrator 1 3.5 Another professional 1 3.5 Friend of child 1 3.5 Teacher 1 3.5 Total 29 100 Note: in several cases there was more than one person who disclosed abuse. A. Description of Abuse In most cases, participants received explicit descriptions of abuse from the child. Descriptions of abuse typically involved information about the type of abuse, context in which it occurred, information about the perpetrator, and when it happened. For example, one case involved a 13 year-old girl and her friend meeting with their school psychologist to disclose sexual abuse that was occurring in the home. The psychologist recalled the child's story as follows: .. .she and a friend came to me and the friend sort of did the talking in the beginning and she picked up.. .and she basically described a situation where she and two younger sisters that were living with their father, in a rural area, 74 had all experienced sexual abuse, and that it had been going on for a while... Another incident that was reported involved a psychologist receiving a direct disclosure of physical abuse from the perpetrator and his wife. In this case, the psychologist was providing support to the father to assist him in coping with the effects of a recent head injury (i.e., severe headaches, intolerant of noise). The participant said: .. .some of the crisis that occur with any client in this population, and did occur with this client as well, is the increase in frustration without being able to return to work, and the sort of wearing down with the chronic pain...in talking about how stressed out everyone was, what came out from the wife and husband together was that the dad had hit one of the children, and hard enough to toss the child part-way across the room... B. Details of Abuse in the Child's Story Several participants said that they sought supportive information (i.e., idiosyncratic details) about the occurrence of abuse before they reported. These participants stated that idiosyncratic details enhanced the credibility of the child's statements and increased their confidence that abuse had occurred. For example, one participant said: One of the things the child said.. .was that 'he pushed me and I fell down. He pushed me and I fell down the stairs and I fell into the slugs.' Now no child is going to tell you something [like that]. You look for those kind of idiosyncratic details in the child's statements, not just that he pushed her. When a child says fell into the slugs, that's what she remembered. So the slugs were, they are in B.C. They're in the garden. She got pushed off the stairs. So those kinds of statements, they make the child's statements much more credible. 75 Another participant, who worked in a youth detention centre and said that he is used to hearing "tales," said that he relies on details and the internal consistency and logic of a disclosure: .. .he [a 16-year-old young man] told me of events that had happened to him at the hands of a teacher. Now in my thinking, I am making some subjective assessments of his honesty by his completion, as I have to decide, is this a real story or is he just a good teller of tale. And as he went on details unfolded, and I asked him some questions to extract some details. I came to the conclusion that that whether or not these events happened as he was describing, they were described in sufficient fullness as to fit my obligation as I understood it to report. C. Child's Affect During Disclosure Another variable that increased participant's likelihood of reporting was the emotional response of the child at the time of disclosure. Reactions from children ranged from anxiety to fear and sadness. For example, in one case a little girl broke down and started to cry in describing how her mother hit her. The psychologist said: .. .the child came to me with her little friend.. .and she sat there and she was sniffling, a little uncomfortable.. .and so she started crying again and I said so what happened this morning.. .and she said 'my mom kicked me.' One participant connected the child's anxiety when talking about his father to the occurrence of sexual abuse. She said: He had to go to the toilet in mid-session, and I have found that over the years that when children have to pee, when they have been working on something and then all of a sudden they have to go, and that anxiety that is created within them. So he brought up his father.. .and.. .out of the blue I asked him who had touched his private parts .. .and he said my father did... 76 Similarly, another psychologist talked about the child's affect being congruent with his disclosure of sexual abuse. She said: .. .he pooped his pants when he was talking about the sexual abuse that involved anal penetration, he was crying when he talked about it. So not just the verbal part, but all the stuff that went along [with it]. 4. Observation and Other Non-verbal Communications Participants also made judgements to report child abuse and neglect based on direct observation. For example, one participant expressed concerns about a father's mental state and ability to care for his four children. This man was involuntarily admitted to the hospital after a suicide attempt. The psychologist said: This guy at the morning group presented as, 'I'm okay, my goal is to learn again how to play guitar, and I feel that I'm ready for discharge and I don't need anybody's help'.. .when we met he looked depressed, he told me he was not eating, he was not sleeping very well, and all that he was able to concentrate on was how he will revamp his business... [when I tried] to gather information about his family it always ended up, 'I'll take care of my family.' How will you do it? 'By starting my business again.' So how will that help the children? How will you cook, how will you do the everyday tasks? That was not important.. .1 learned that he wasn't functioning for a couple of years minimum.. .anyway, so I was thinking well social services should get involved here. Another participant made two separate reports of abuse after meeting with a mother and her two children to do a court assessment. This case involved the mother using cocaine and being involved in a relationship in which there was violence. With respect to the first report, the participant said: 77 .. .the mother came in with the boys, extremely controlling, obviously under the influence of substances, children that screamed and were hollering, a boy who at 5-6 years of age was completely out of control, emotionally. I've never seen a child so out of control. I was very, very concerned about the mother's capacity to parent. He was so emotionally distraught. Very, very concerned about her uncontrollable and violent nature, she was on cocaine, as you know cocaine contributes to violence in people and physical abuse, very concerned about that. I phoned the Ministry of Human Resources and reported my concerns after having done this assessment.. .the older boy was extremely aggressive physically with the younger boy. I was very much afraid that he was going to hurt the little boy, very aggressive. So I was concerned about physical abuse on the part of the mother and particularly on the part of the boyfriend of the mother who was a drug abuser and extremely aggressive in my office. Another participant talked about the parent of a special needs child who gave the child large doses of medication to calm him. The participant said: There was a paediatrician working at the centre [who had] put the child on Ritalin. So then we started noticing that the child would some days come to the day-care and [he] seemed more subdued than other days. And we were kind of wondering what that was all about and asking mom about her use of medication, and she basically [said] 'well the doctor said I could give it to him whenever I wanted to.' And we were very concerned about that because she was very punitive and blaming of the child. One participant suspected sexual abuse on the basis of various artwork and writing. She said, "...[in] vague ways [she] started to talk about abuse, pictures, drawings, poetry, really vague." Another participant suspected child abuse during play therapy. She said: 78 .. .he was extremely aggressive and would use all the fighting toys, got the swords out and the hatchets out, but he was fair, I would get toys also and then he would be the victim. He would fall on the floor, and he would die.. .it seemed intentional that he was going to touch my breasts.. .he wanted to be held and he loved it.. .but a lot of kids don't do that, and what is he 7 or 8 years old, and socially he's closer to a 5 or 6 year old. Risk Factors Along with signs of child abuse, recognition of risk factors for abuse was another starting point in the decision making process for participants. Some of the risk factors occurred with signs of abuse, while other risk factors occurred on their own. Participants identified various factors that place the child at risk for the occurrence of potential child abuse. Table 8 presents 14 risk factors that participants identified. I divided the risk factors into two categories: perpetrator variables and child variables. Among the perpetrator variables, participants identified cases that involved: domestic violence; substance abuse; history of mental illness; parent with a head injury; parent sexually abused as a child; suicidal parent; cultural beliefs/practices about discipline; parent rigid in thinking; previous involvement with the Ministry; and the inability of the non-offending parent to protect the child. Risk factors involving domestic violence, substance abuse, mental illness, and suicidal parents, on their own, resulted in a formal report being made to the Ministry. The other risk factors did not lead to a direct report, but when combined with other concerns, such as signs for abuse, a report was made. 79 Table 8 Risk Factors for Potential Child Abuse Risk factor Frequency Parent Variables Domestic violence 5 Substance abuse 5 History of mental illness 3 Head injury 3 Parent sexually abused as a child 3 Suicidal parent 3 Parent rigid in thinking 2 Other parent unable to protect the child 2 Cultural beliefs about discipline 1 Child Variables Age of child (e.g., child under two) 4 Child unable to communicate 1 History of child abuse 1 Fetal alcohol child 1 Special needs child 1 Participants mentioned several factors that they believed placed the child at risk for maltreatment. The most frequently mentioned child risk factor was the child's age. Specifically, participants indicated that younger children were more at risk. Other child factors that were related to risk included a history of child abuse, and children with special needs. With child risk factors, unlike parental risk factors, participants tended to not report on the basis of these factors; however, they stated that their concerns for abuse increased, and they subsequently became more watchful for potential abuse. 80 Criterion 2: Did the signs or risk factors meet your threshold to report, as you understand the law? The second decision point in the model involves a determination of whether or not the signs or risk factors meet the psychologists' legal threshold to report. The term "legal threshold" was used to capture the broad range of circumstances that may warrant a legal duty to report. In 33 of the 34 cases, participants said that the indicator(s) met their legal threshold to report as they understand the law. There was variability, however, in participants' degree of confidence that abuse had occurred. Some participants stated that they were quite confident that abuse had occurred while others were less certain. For example, one participant said that he was not entirely sure if abuse had occurred but wanted someone to follow-up on the case. Was this incident true, in fact, I do not know, but I didn't need to know that. I needed to know that it was believable, and that this person wanted to own that in the face of knowing that resources would be required. Similarly, another participant stated, "It was clear to me that there was a possibility [of sexual abuse] and the possibility was non-trivial." In a different case, the participant felt a higher level of confidence that abuse had occurred. This case involved a verbal disclosure from the father who had hit his child. The psychologist said, ".. .it was clearly physical abuse.. .the incident itself was clearly reportable." Another psychologist expressed her degree of confidence that abuse had occurred, "I felt that it was really clear that it needed to be reported because those children in the home were being abused, clearly abused." Similarly, another psychologist said, ".. .it was very clear to me that it's blatant sexual abuse. There is no ambiguity about it." While there was much variability in confidence levels of abuse, participants seemed to be clear about their role with regard to reporting. Six participants, in particular, said that they did not see it as their role to determine whether or not abuse had occurred. Rather, 81 they stated that their job was to forward the information to the Ministry for Children and Families so that they could make the ultimate determination of whether or not abuse occurred. For example, one participant said: I came to the conclusion that whether or not these events happened, as he was describing, they were described in sufficient fullness as to fit my obligation as I understood it to report. That is to say, it wasn't for me to make the final decision of the truthfulness or falsehood or exaggeration, if there were any of these things. It was a report of repeated sexual abuse and it was clear that if they had happened it fell well within the obligation to report, and so report I did. The above quotation also reflects the participants' awareness of his legal duty to report possible child abuse. In addition to the legal requirement, several participants mentioned that they had a moral and ethical responsibility to report possible child abuse. In talking about one case, one participant discussed the moral aspect of reporting child abuse. He said: I think I have to uphold the law in that sort of scenario, so that's one thing. The other thing is that I have worked in my training with kids who were abused and it's terrible, it's terrible to be exposed to that. Think of a kid that goes home and is frightened every night and is afraid of his own dad. It's terrible and those kids are so defenseless. I would, independent of the law, it would just irk me if there were something I could do to try and intervene and stop that because that's totally unacceptable.. .1 have a touch more emphasis perhaps on the second issue, not the legal issue, that's more a sense of look this kid is at risk and there's no one there. If nobody's stopping this behaviour it's going to go on and on, and often escalates is my 82 understanding, so that's the primary reason, but I feel pretty strongly about the legal issues too, I really do. Other participant's noted their ethical duty to report. For example, one participant said, "... [reporting] is a legal responsibility and it's also a professional, ethical issue." In addition to citing issues of law and ethics about reporting, participants most frequently stated that protecting the child was their primary concern. A typical response was: "...the best interest of the child is always my mandate." The wording, "as you understand the law," reflected individual differences in interpreting legal responsibilities. For example, several participants mentioned that when the victim of historical child abuse is no longer a child, they viewed their reporting responsibilities as extending to other children (siblings, students) with whom the perpetrator (father, teacher) may have ongoing contact. One participant explained: I'm dealing with an adult, so I'm wondering.. .is the perpetrator, the alleged perpetrator, is this person currently in a position of responsibility and authority over children, in that kind of situation of putting other children at risk. If they are still in that position, and in further exploration she thought that this person was still teaching, that for me was a red flag. I was pretty clear with her about my responsibility in coming up with information and that I would need to be alerting the authorities with regard to what I knew about this situation. With respect to the issue of confidentiality, participants were clear that they were not able to protect information relating to child abuse. Seven participants said that as a matter of practice, they detail the limits of confidentiality to their clients during therapy. One participant said: The mother called me one day and said that her little boy who was six had bruising and the father had given him a "licking" that had left bruises.. .1 had told them the limits of confidentiality and that I have an obligation to 83 report. So I referred her back to that and let her know that I was going to have to report. Criterion 3: Was there some other value to report other than a legal one? Criterion 3 is included in the model to capture those instances where the legal threshold may not be met, yet the psychologist believes that there is value (e.g., therapeutic value) in reporting the case. Although this criterion was addressed by only one participant in the model development stage, I decided to keep this criterion because it extends the reasons for which a report can be made (i.e., other than the legal duty to report). Given that there is little support for this criterion, it is distinguished from the other criteria within the model by a dashed-box. In the one case, the psychologist said that she reported a case of historical child abuse because her client felt that it would be beneficial for her therapy. In this case, the psychologist stated that the case did not meet her legal duty to report because her client was currently an adult, and that the perpetrator presented no risk to other children (e.g., he was old and ill). Nonetheless, the psychologist did report as requested by her client. This decision was made mutually by the client and psychologist. Criterion 4: Were you concerned that reporting would cause harm? While all 34 cases were ultimately reported, 16 participants believed that reporting would cause some sort of harm. Participants expressed concerns for the therapeutic relationship, the child, themselves, the family, and the perpetrator. Concerns about how the Ministry would handle the case and whether the reported cases would be substantiated were widely held. On many occasions, participants identified multiple concerns about the effects of reporting. Therefore, the number of concerns (n=26) exceed the number of cases in which these concerns arose. Each concern is presented below. 84 Harm to Therapeutic Relationship. Most frequently, participants believed that reporting possible child abuse would cause harm to the therapeutic relationship. In fact, 9 participants had expressed this concern. Some selected quotations are provided to illustrate this concern. For example, one participant said that she was concerned that reporting would interfere with her client's willingness to openly disclose information in therapy. The participant said it this way: I have an ongoing working relationship with her [meaning the client]...and she's really come a long way.... In the early time with her she was skittish, there were a lot of things she didn't disclose.. .so the fact that she disclosed this information to me was important and I thought it would be very unfortunate if I had to react in such a way that she would use in the future— not to disclose certain information. This participant went on to say that she was afraid that her client would terminate therapy, "I mean I was afraid that she would quit treatment." As a result of reporting, the participant added that, in her case, this perception was true and reporting did impact on the therapeutic relationship. She said: .. .in fact, we did have some problems with that [meaning trust] later on because sometimes I think her psychiatric state was such that she couldn't take care of the kids properly but she didn't want to tell me how bad it was. She didn't want me getting the Ministry involved. Similarly, another participant commented on the negative impact of reporting on the therapeutic relationship after making a report of physical abuse. He said: .. .it's very, very clear that I had to report it, which I did, but I think that it had a negative effect, at least short term, it had a negative effect on the therapeutic relationship. As I recall, in the end, maybe, we recovered that a little bit, but it was certainly a set back." 85 Two other examples of participants' statements were: ".. .if I phone up and make a report, that could be the end of the therapy..." and "... when [you're] starting to build rapport and then in some way you have to call social services, it's kind of oh dear, this really blows things out of the water." Harm to Child. Of the 16 cases where concerns of harm were expressed, harm to the welfare of the child was suggested in 5 cases. On two separate occasions, one participant expressed her concern about the psychological effect when children are apprehended. The participant said, "we don't act lightly when we take the child away from the people that they're bonded to. It's very serious to do that because you know you're doing psychological damage to a child when you remove them from the people that they're bonded to, even if it's pathological bonding, as it was in this case. That's going to be difficult for that child so you don't do this lightly at all." Another participant talked about the child's feelings of fear and guilt after disclosing to the psychologist that his dad was physically abusing him. The participant said, "this boy didn't want to get his dad into trouble. That was very, very clear [to me]." In another case, the psychologist learned that reporting physical abuse could jeopardize the physical safety of the 17 year-old teenager whom the father was abusing. The teenager told the psychologist that if a report was made, her safety would be in danger or she would kill herself. The participant described the case in this way: [she told me that] if it [meaning the abuse] was reported, she would kill herself. Or if that wouldn't happen her father was likely to put a hit out on her and she had some evidence from friends and so on that in the part of the community that they were in, that this was a possible scenario. That to challenge her father's authority in that way would lead to pretty dire consequences for her. 86 The teenaged girl in this case also threatened that if a report was made then she would deny her allegation. In another case, one that involved a custody and access assessment, the psychologist reported child abuse after she had received a verbal disclosure from a 7-year-old girl that her father was sexually abusing her. Prior to the release of the assessment report in a court proceeding, the psychologist was concerned that the child would be in danger for disclosing the abuse. The psychologist described the situation in this way, "I had severe concerns about the safety of this child after this report comes out, and the father reads what the child has said." The Ministry's Response. Participants' perception about the effectiveness of Ministry workers was mentioned as a concern in reporting abuse in four cases. Their concerns ranged from the lack of response from the Ministry to fears about apprehension. For example, one participant believed that if she reported possible child abuse then nothing would happen, ".. .they [meaning the Ministry for Children and Families] do nothing." Similarly another participant said, "I have found frequently [that] when you report nothing happens. That has been my experience..." In contrast, one participant feared that her report would lead to the child being apprehended by the Ministry social worker. She stated that: .. .one has fear that over zealous social workers will come and kidnap the children, kind of thing, and they won't listen or they will over-react, and in these delicate situations where my therapeutic alliance with the client is on the line, I want to make sure it is done in a way that doesn't interfere with my ability to work with them, that they don't feel betrayed, and at the same time balance that with the fact that if there are children in need of protection then they need protection. 87 As another example, one participant suggested that the Ministry is inconsistent in taking action. She said, "different social workers within offices deal with things differently.. .some of that is philosophy and some of that is caseload." Harm to the Psychologist. Four participants described negative effects to themselves in reporting child abuse. Concerns ranged from experiencing emotional harm in making a report to fear of legal repercussions. With regard to the emotional impact of reporting, one participant said: ... [reporting is] stressful as hell, it really is, it's such a drastic thing to do.. .you're completely reshaping their identity, probably for life.. .so I found it [meaning reporting] very, very serious, and awesome and overwhelming... Two participants, both of whom work in a hospital, said that they were concerned about legal liability. One of these two participants said: I'm also concerned to do it [meaning report] properly so that I'm not going to get blamed for something going wrong...you cover your butt, you cover your butt, much of the work here has involved suicide and things going very wrong. Still another participant talked about his negative experience in reporting in that he did not have the support to report from his immediate supervisor with whom he worked. In describing the case, he said: [I went to] my supervisor.. .and she discouraged any sort of report. However, I went ahead and made contact with the Ministry in that case, because I've learned over time never to trust my supervisor about filling their legal obligations, and I've also learned over time that my responsibilities are my responsibilities and that I have a duty to fulfil them regardless of what, you know, other people perceive as their responsibilities. 88 Harm to Perpetrator. In two cases that were reported, participants talked about the potential harm to the perpetrators after they self-disclosed abuse. In one of these cases, the psychologist was afraid that her client would harm herself. The psychologist said: I was concerned that she [meaning the neglectful mother about whom a report was being made] would be overwhelmed and hurt herself. She's quite capable of self-destructive acts, she cuts herself and she makes herself tense... [she was] afraid that the children would be taken from her. That the father would also use the fact that she had been hospitalized a couple of times, and was now in trouble enough to have social services intervene, that he might try to go for custody.... In the other case, the psychologist expressed his concern about the implication of reporting for the perpetrator. In this case, a 19-year old female disclosed (in group therapy to treat eating disorders) that she was having fantasies of sexually abusing one or more of the young boys at her mother's day-care. In learning this information, the participant said that he was compelled to report but that, in doing so, was worried about labelling the young woman. The participant said: I felt awful that this girl could be committed or called all kinds of names like child molester and so on and that we were making a huge impact on this person's life and of course on her family's life... Harm to Family. In only one case did a participant indicate that reporting would cause harm to family members. In this case, a 13-year-old girl disclosed to the psychologist that her uncle had sexually abused her many years ago. At the time of disclosure, the girl also pleaded with the psychologist that he not report the case because of the harm it would cause to the 89 family. The participant said, "she told me [that] if I reported it, it would destroy the family... [and in disclosing] it caused this girl a great deal of pain and anguish...". Unsubstantiated Report. One participant expressed concern about making a report that later would be unsubstantiated by the Ministry. He said: .. .at the time it was very hard to determine whose story was true, and even the damage that can occur through incorrect reporting, and having seen some of that, it made me a little gun-shy, a little reluctant to just jump in and report as soon as I think it.... Criteria 5: Were you able to minimize the harm that reporting would cause? Participants that discussed the potential harm of reporting also described the extent to which they were able to minimize the harm. Of the 16 cases in which participants expressed some concern about reporting, 6 of them said that they were able to reduce the harm. In doing so, participants then reported the child abuse. These cases will be discussed in this section. The 10 cases in which participants felt that they could not minimize the perceived harm will be discussed under Criterion 6 (next section). Four of the six cases in which participants said that were able to minimize harm related to concerns about the therapeutic relationship. To minimize the harm resulting from reporting, participants spoke of engaging their clients in the reporting process and being open with them. In each of the four cases, the perpetrators disclosed the abuse to the psychologist and took responsibility for their behaviour. One participant said that she invited the perpetrator to self-report; the participant said that, in this way, the father (who had hit his son) could actively seek his own assistance rather than being forced to receive service. The participant described the case as follows: 90 .. .in the meeting with the two parents there I recommended that they contact the Ministry of Social Services to engage a worker to come to the house to help them with the kids, that this was an isolated incident at this time but was something that was something of concern.. .and what I did is I gave them control with the provision that they had a time line in which to do this [meaning self report].. .so I tried to do it in a positive framework that they are there to help the family and that is really how I do see it.. .they're not there to police the family.. .1 also made it clear that if they didn't follow through, that I would be following through, and my number one concern was the well being of the children.. .in my experience having worked with Ministry of Social Services in [participant named another province], they had the most success with parents who self report and engage the service and then they're investigated, that was the point...that they see it as a life line to them, so not that Dr. X got on their case and reported something that was really a secret. In a different case, the psychologist received an allegation of physical and sexual abuse from a family member of her client. In talking about the allegation in a therapy session, the mother denied that her daughter had been abused, and further stated that she did not wish to involve social services. In an effort to maintain a trusting and open relationship with the client, the psychologist indicated that she would have to report the allegation. The psychologist said that she elicited her client's involvement in the reporting process. In the third case, a teenage boy disclosed to his parents and his therapist that he had sexually abused his sister's girlfriend. During the family counselling session, the psychologist told the family that this situation needed to be reported. Again, as with the other participants, this psychologist invited the parents to report their son's case. The 91 participant reasoned that he wanted to empower the family to take responsibility and maintain the therapeutic relationship. In his words, he said: .. .the issue for me.. .is the therapeutic relationship that I have with them. And me reporting would be a last resort, you know, it empowers them if they report. But I still think.. .the bottom line is that if they didn't report, I would. In the fourth case relating to concerns about the therapeutic relationship, a 19-year-old university student told her psychologist that her father while growing-up had physically abused her. The young woman also stated that she no longer was living in the home but that she had two younger sisters, who, she felt, were subjected to similar kinds of physical abuse. The psychologist described the situation this way: I felt that it was really clear that it needed to be reported because those children in the home were being abused, clearly abused.... I felt I needed to tell her that we need to do something about it.. .the most important thing to do was to report abuse. And what I typically do with that is I approach it from the perspective of the client reporting because I think it is a therapeutic issue. I think it is about possibly being able to do something to change what is going on and I don't want to take the power away from the client. Also, it's a therapeutic issue in the sense that if I phone up and make a report, that could be the end of therapy and I didn't want to jeopardize that. The fifth participant who felt that she was able to reduce the harm was concerned about being sued if she reported. The psychologist said that she protects herself from being liable by actively consulting with other professionals prior to reporting. She said: .. .1 presented it to the doctor on call, we agreed this was a complicated case with significant suicide risk, he recommended that I contact car 86 and inquire whether they would do a home assessment.. .it's a combination of 92 sharing the medical/legal responsibility by consulting and pulling in others, so it's not me making the decision by myself. The final case where harm to the child was minimized involved the 17 year-old girl who threatened suicide and stated that her father would kill her if he found out that she reported. In this case, the psychologist consulted with her colleagues, and ultimately consulted with a Ministry social worker. The outcome was as follows: ... [the Ministry social worker] offered to come and meet with the client if she were willing, without hearing her name but to talk with her about her options and alternatives, [and in terms of] what the Ministry might be able to provide for her if she wanted it pursued, [and] what she could do for her own protection, etc. In the end, the participant said that the social worker went to the child's home and had a family counselling session. Criterion 6: Did the reasons to report outweigh the reasons to not report? Participants who were unable to minimize the harm that they perceived would result from reporting considered this final decision point. This factor involved analysis of the benefits and detriments of reporting. This factor reflected the tension between the legal obligation and ethics. On the one hand, psychologists recognized their legal responsibility to report possible child abuse; however, in some cases, reporting abuse was perceived to be detrimental, particularly for the child. Nonetheless, in all 10 cases where the psychologist was not able to minimize the harm, a report was still made. These participants felt that reporting outweighed not reporting, and in every case they cited that their ethical/legal obligation to report outweighed any potential harm to report. For example, in discussing the risk of making a false report, one participant said: 93 .. .if you make a false report you can do a great deal of damage to the system. And if it truly is false, you've been entirely counter-productive in what you were trying to do. But that, the balance of that is that you weigh it up against a stronger desire to report and error on the side of the child. So if anything, I might have a tendency to report too soon but I think I kind of want to have that balance. Another participant described a situation where the abused child wanted to protect his father from the law. The participant said: What was difficult for me in that situation was that this boy didn't want to get his dad into trouble, and that was very, very clear. You know, and for me, it's also very, very clear that I had to report in which I did.... In a different case, the participant expressed concerns that if she reported, then the Ministry would not take any action. The participant described her situation like this: I have found that frequently when you report nothing happens. That has been my experience and so it doesn't necessarily help, which is something that I always think about but I still feel obligated to do something and then to support the client to deal with whatever the outcome is. Another participant talked about her decision making process with respect to the tension she felt between the law and her ethical responsibilities as a psychologist. This case involved the child victim threatening that she would commit suicide if the psychologist reported. The participant said: We were weighing our ethical responsibility to the young woman in terms of assessing [the] risk [that she might commit suicide if we reported].. .as well as our obligation to report and, you know, it's one of these [cases] I think where there is no clear cut solution. You know, your legal obligation 94 and you also have an ethical one if you feel that you really are compromising your client's safety in the short term. Factors Considered but Not Used This section will briefly highlight some of the factors that were consider as potential decision criteria, and the reasons why these factors were not ultimately used in the model. This section will address the following issues: the length of time in which a report was made; subjective confidence rating in the occurrence of possible child abuse; and participants being supported by their colleagues in making a report. This section will conclude with a comment on studying the phenomenon of child abuse reporting. As interviews unfolded, the notion of a time dimension emerged as a factor that I considered integrating into the model. What I noticed was that participants varied in the length of time before which they filed a formal report of possible abuse. Some cases were reported immediately and some cases were reported several weeks after the suspicion arose. Generally speaking, reports that were made immediately tended to relate to the immediate danger to the child. Similarly, but not consistently, reports that were delayed related to a lower risk of harm to the child or other children. In light of this data, I tried to incorporate "level of risk" or "degree of harm" or "severity of abuse" into the early stages of the model. However, the data suggested that these factors inconsistently predicted reporting behaviour. That is, different participants were equally inclined to report immediately when the "level or risk," "degree of harm," or "severity of abuse" was perceived as high, low, or moderate. Therefore, the current model was unable to represent these important variables. A second factor that I considered was participant's level of confidence in the occurrence of child abuse. Similar to the notion of "severity of abuse," subjective confidence ratings seemed to occur along a continuum. While participants invariably reported when they were confident that abuse had in fact occurred, a number of participants 95 reported cases when their suspicions were low (i.e., they had a hunch abuse had occurred). Therefore, confidence ratings were not included in the overall model because this factor did not consistently predict reporting behaviour. Third, participants often consulted with other professionals (e.g., colleagues, supervisors, etc.) before making a decision to report. This factor was mentioned as an important step in the process of reporting. However, on its own, consulting was not identified as a key factor in reporting. Rather, it served to inform participants of factors that are represented in the model. It is worthwhile to note that it would have been impossible to include this factor in the model because participants consulted with others throughout the reporting process and not at any one particular point in time. As a final comment to this section, I recognize that there are many important individual factors involved in the reporting process, some of which are represented in the current model. However, given that the focus of this method is on the development and verification of a group model, it was not possible to represent all of the factors involved in decision making, such as potential factors related to the psychologist, legal issues, situational factors, and the like. The next section presents data that served to validate the preliminary group model. 96 Phase 2: Model Validation After the preliminary group model was developed, it was then formulated into a questionnaire format for the second phase of the study (see Appendix I). Each decision criterion from the group model was transformed into a single question requiring a "yes" or "no" answer. The main purpose of this phase was to verify and establish the predictability of the decision tree model. Based on participants' answers to the questions representing each decision point, the goal of the model was to predict the decision outcome. Throughout this stage, Gladwin's (1989) method was strictly followed. A similar yet separate group of registered psychologists that reported child abuse served to validate the model. Their responses, based on 36 cases of reported child abuse, are presented below. First, however, relevant data are presented that add to the criteria that were developed in the model development phase. The data here are presented using the same general format that was used during the presentation of data during the model development stage. This section will begin with a description of the types of abuse reported and it will conclude with the results of the validation of the model. Criterion 1: Were there any signs of or risk factors for child abuse or neglect? Table 9 shows the types of maltreatment that were reported in the validation phase. The types of abuse presented here are proportionally similar to the data in the first phase. Of the 36 cases, sexual abuse was reported in 15 cases and physical abuse was reported in 13 cases. On two occasions, participants reported multiple types of child abuse. 97 Table 9 Types of Maltreatment Reported to the Ministry for Children and Families Reports % Sexual abuse 15 42 Physical abuse 13 36 Neglect 7 19 Emotional abuse 1 3 Total 36 100 The specific types of abuse reported are shown in Table 10. For sexual abuse, children were most often fondled. In many of the sexual abuse cases the participant did not specify the type of abuse. In the physical abuse cases, 7 involved the child being hit with either the perpetrator's hand or an object (e.g., stick, strap, or belt). The neglect cases often involved multiple indicators including lack of supervision, not attending school, inadequate sleep and nourishment. The only case of emotional abuse involved name-calling. 98 Table 10 The Nature of Child Maltreatment Reported to the Ministry for Children and Families Frequency % Sexual abuse Fondling 4 29 Oral sex 1 7 Unspecified 9 64 Total 14 100 Physical abuse Hit with hand 4 33 Hit with object 3 25 Kicked 2 17 Pushed 2 17 Unspecified 1 8 Total 12 100 Neglect Lack of supervision 3 25 Not attending school 2 17 Consistent hunger 2 17 Inadequate sleep 2 17 Filthy living conditions 1 8 Poor hygiene 1 8 Not given medication properly 1 8 Total 12 100 Note: the child neglect cases often involved multiple concerns and, therefore, exceeded the number of reported cases for that type of abuse (n=7). 99 Within the various types of abuse that were reported, participants detected a number of signs and risk factors. These signs and risk factors are presented below. Again, it is important to note that some of the indicators of abuse occurred in isolation while other signs and risk factors occurred in combination with other indicators. Signs. 1. Physical Signs As with the data in the development phase, physical injuries were not commonly mentioned by participants. In only three of the 36 cases did a participant mention that s/he had either heard of or observed a physical sign of possible abuse. In one of the three cases, a school psychologist said that he noticed that a young girl looked tired and believed that she was not getting adequate sleep at night. His specific words were, "her eyes were darkened and she seemed very sleepy in school." In addition to this information, the psychologist learned that the girl was behaving in a sexual manner (i.e., using sexual language, acting "provocatively") in the playground, and that some of her artwork suggested sexual abuse. On the basis of this information, the psychologist then made a report of possible child abuse. In the second case, the psychologist received a verbal disclosure from a 12-year-old girl that her father was physically abusing her, and that recently he had hit her on her back. The psychologist assessed the nature of the child's injury and, although there was no bruising, he felt that the incident was reportable, noting that the hit had caused physical pain to the girl. In the third case, the participant reported a case of physical abuse stating that bruising had caused a serious injury, enough to warrant a report. The participant said: .. .there was a lot of history of physical, pretty severe physical abuse up until about 6 months ago. The last episode.. .involved quite a bit of bruising, 100 and shoving and throwing objects, which I think, that, that's quite violent. In my, in my template of what violence is, that's quite violent. 2. Changes in the Child's Affect or Behaviour In addition to the emotional and behaviour signs of child maltreatment identified earlier, participant talked about several other signs. For example, in a case involving sexual abuse, one participant identified a suicide attempt as a sign of abuse. With regard to physical abuse, another participant described a case in which a young boy experienced dissociation. Several participants mentioned sleep disturbances when talking about cases of sexual and physical abuse, and neglect. Most participants talked about children presenting as anxious, distressed, angry, and depressed in describing cases of child neglect. One of these participants said: ...there were a number of behavioural indicators, he [meaning the boy] appeared to be anxious and distressed and hypervigilant a lot of the time, and then we were also concerned that he wasn't arriving with a proper lunch, [he] didn't seem to have breakfast in the morning... 3. Disclosures In 31 of 36 cases, the means by which participants learned of possible child abuse was by verbal disclosure. Table 11 shows that the abuse was disclosed by the child in most of these cases (n=20). Adult perpetrators, non-offending parents, and other family members disclosed on three occasions each. 101 Table 11 Source of Verbal Disclosure Frequency % Child 19 61 Adult perpetrator 2 6.5 Non-offending parent 3 10 Family member 3 10 Victim who is now an adult 2 6.5 Child perpetrator 1 3 Friend of child 1 3 Total 31 100 4. Observation and Other Non-verbal Communications Five of the 36 cases were reported on the basis of observation alone. Three of these five cases involved child neglect; the other two involved possible sexual abuse. One of the reports of child neglect involved the child infrequently attending school, or attending school without being adequately fed in the morning. The psychologist said, "[the child] arrived at school hungry, and was kept home from school without explanations... [and] whose mother was very unreliable about following through with appointments and that sort of thing...". In another case, the participant reported child neglect after observing the child's living situation during a home visit. In this case, the psychologist reported to the Ministry a mentally handicapped child who was wearing soiled clothing, and where there was decaying food in her home environment. In a similar case involving a home visit, the 102 participant reported child neglect after he noticed that there was no parental figure in the home, there was no food available to the children, and that the two children looked emaciated. The psychologist later learned that the father was in jail for driving a vehicle while impaired, and that the children were living alone. The other two cases involved reports of possible sexual abuse. In both cases, the school psychologist observed a child acting in sexually inappropriate ways. For example, one psychologist said: .. .when [the boy] came to my office there were a number of signs and symptoms as far as I was concerned, just in [terms of] sexualized behaviour, that this boy was being abused and you know my red flags went up.. .he was doing things like rubbing his genital area up and down on the chair, on the arm of the chair that I was sitting in, coming and rubbing me up and down, kissing my hand and my face and you know, just being, you know way too sexual.... The other psychologist said: .. .there was sufficient basis I felt to suspect that there was some inappropriate sexual kind of at least exposure or stimulation or something happening...[the girl] was really behaving inappropriately in the classroom, was quite pro vocative... was drawing quite explicit, all showing some indication of sexual stimulation at least. Risk Factors. As with the data from the model development phase, recognition of risk factors for abuse was another starting point in the decision making process for participants in the validation stage. Typically, risk factors occurred with other signs of abuse; however, some of the risk factors occurred in isolation and did result in a report of child abuse. 103 In describing cases of child abuse, participants often identified risk factors for the occurrence of child maltreatment (n=20). Table 12 lists the specific types of risk factors under the headings: parent variables and child variables. In addition to these two categories, a third category called "circumstances" was created. Factors such as stress (e.g., financial), and opportunity and access to harm a child were included in this category (n=3). Among the parent factors, participants most frequently talked about substance abuse and history of mental illness (n=6). These factors invariably led to a direct report of child abuse. With regard to child risk factors, special needs children were perceived as high risk for child abuse (n=3). Participants also expressed concerns about children who had a history of being abused (n=3). Participants said that they did not report child abuse based on these factors, but they did indicate that their concern for potential abuse did increase as a result of these factors. 104 Table 12 Risk Factors for Potential Child Abuse Risk factor Frequency Parent Variables Domestic violence 1 Substance abuse 3 History of mental illness 3 Other parent unable to protect the child 1 Belowaverage in intelligence 1 Child Variables Non-verbal child 1 History of child abuse 3 Fetal alcohol child 1 Special needs child 3 Circumstances Opportunity and access 2 Stress in family 1 Criterion 2. Did the signs or risk factors meet your threshold to report, as you understand the law? In 35 of 36 cases, participants indicated that the signs of abuse met their legal threshold to report the case to the Ministry. The one case that did not meet the participants' legal threshold is discussed in the next section. Participants talked about a number of legal issues in the cases they chose to describe. For example, many (but not all) participants said that they explained to their clients, in the first session, the limits of confidentiality. One participant said, "I always indicate the limits to confidentiality before starting any 105 interview, so that the person can decide up front whether or not there are certain pieces of information that they are going to disclose. Similarly, another participant stated: .. .1 have a description that I go through with folks.. .about confidentiality and the limits of confidentiality. And there are three different kinds of circumstances in which I am obligated to break that confidentiality and one of those is if there is reasonable belief that there is child abuse currently taking place. And if that's the case, then I'm required, I'm required to report, report to the Provincial Child Welfare Authorities. So everyone gets that before we talk about anything. As with the model development data, safety of the child was a common concern for most participants. A typical response from participants was, "Well, for me, the most important thing was, are these children really at risk.. .when it became clear enough to me that the children were at risk, then I made the decision to report. The level of certainty required before making a report tended to vary across participants. While many participants reported cases when they were fairly certain that child abuse had occurred, some participants reported when they were less confident. For example, one participant said: .. .1 don't make decisions concerning the credibility of the disclosure. I just take what I get. If it is a concrete disclosure, then I report it. If I have suspicions that something has occurred.. .1 will report my suspicions and call them suspicions. Another issue that emerged for some participants involved unreported incidents of historical abuse. In two cases, participants described a case in which their clients talked about abuse that occurred more than 20 years ago. While both participants stated that they did not have a legal duty to report the historical abuse, a report was made to protect other children from possible abuse. One participant described the case this way: 106 .. .she [meaning the client] had been sexually assaulted by her father when she was a little girl and I am seeing her as an adult in her mid-30's.. .Now this would have, again this would have been you know, 20, 25, years ago that the assault took place, so under normal circumstances, I believe I have no obligation to make any kind of report here. However, what she was concerned about, is that one of her sisters was a mother and her father was now in a position to be spending time alone with her child. And there was nothing that she, that my client said to me that suggested that the father had ever gotten any kind of treatment, had ever acknowledged that he had a problem, and she was worried, right, that he was assaulting the grandchild. The other participant said: .. .this person that [she] met had engaged in inappropriate sexual behaviour toward her when she was a minor.. .she didn't [want to report], she didn't want to dredge that up, and then the issue became [her] realizing that he was, he was at that time in a house where there were minor children also present and so the issue that came forward was... that.. .a report [should] be made [because] there was knowledge of past abuse in this case. Before reporting, many participants constructed an "abuse" label in order to report. For example, one participant said: I couldn't get a reading of how hard the kick was but it didn't matter at that stage to me, this was, you don't do that, and that was an abusive situation.. .1 know the man, a big strong guy, and we are dealing with a little grade three [child].. .at that stage.. .that moved it from an accident and bad parenting to an abusive situation, and I made the call that moment. With respect to the time dimension of reporting, participants varied in terms of the length of time before they decided to report the case of child abuse. Typically, in emergency situations, a report was made immediately, whereas historical abuse cases 107 tended to be reported somewhat later. One example of a participant's timeline is as follows: .. .in terms of the speed at which I would report, it has to do with the safety of the child and whether they are in the situation at the moment. If they are in the situation at the moment I would report immediately and have somebody actually come to my office at the time and not have the child leave. Other situations, if it is an historical abuse situation, meaning that they no longer have any further contact with somebody, then I would report within the next 24 hours, and people would follow-up. But, it, to me the key issue is the safety and terror of the child, and the parent's ability to follow-up and assist the child in what they are going through after they disclose. Because you know, some kids fall apart and they can't cope, and making sure that a report is made prior to any fall-out, where the offender isn't going to become aware that a disclosure has occurred, because I don't want that to happen before police start to investigate it. So, I'm really careful. Timelines are really careful, important to me, and protection... In addition to the legal responsibility to report suspected child abuse, several participants talked about having an ethical obligation to report. For example, one participant said, "I reminded her that there was an ethical as well as a legal obligation to report..." Another participant talked about the impact of not reporting on her sense of ethics, morality, and personal conscience. She stated, "...if I...never even consulted with the Ministry, I would be in big trouble. I mean morally, ethically, and you know, conscience-wise, I would have difficulty living with myself. 108 Criterion 3: Was there some other value to report other than a legal one? Only one participant indicated that he did not feel that the case met his legal threshold to report child abuse. The participant did report, however, stating that there would be some therapeutic value for the child if he did report. The participant stated: .. .this case hadn't met the threshold to report, my own threshold, but I did also differentiate the fact that it doesn't, it didn't seem to be a clear legal requirement because this had been a case of historic abuse and the reason I pursued it this way was that I felt it would be of therapeutic value to this adolescent who I felt was still dealing with unresolved issues around it to have the issue brought up, brought out and reported and dealt with and confronted. Criterion 4: Were you concerned that reporting would cause harm? In 21 of 35 cases, participants perceived that reporting would have a negative impact. In this section, four concerns around harm are presented: harm to the child, the family, therapeutic relationship, and concerns about the Ministry's response. Concern regarding harm to the therapist was not mentioned in the validation stage. Harm to the Child. Harm to the child was noted in 8 of the 21 cases. Typically, participants expressed concerns that a child would be subjected to further abuse for having made the initial disclosure of abuse. Participants talked about the potential for "backlash" or "retaliation." Some examples are provided: Well, my concern about harm to the child is that the father would retaliate even further toward the child for disclosing these things... There might have been backlash, yes. Social Services calling, and then the parents getting all upset, who did you talk to, that kind of thing. 109 I was concerned about the child, having been punished for telling. I was concerned about the child. If the mother found out. .. .my largest concern would be the harm to the child. I was most concerned about that. I felt too that if it wasn't managed or handled carefully the child could be at risk for some retaliation for having given information and so on that led to the report. Another participant talked about potential concerns around the child committing suicide. In this case, the child asked the psychologist not to report the abuse and threatened that a report was made, then she would harm herself. The participant described the situation like this: So I told the client that I would need to consult [with the Ministry] and that led to the client pleading in every way that she could to stop me from consulting, including threatening her own life. In one other case, the participant expressed concern that if a report was filed against her husband (i.e., for abusing his children) and if she decided to remove the children from the home, then she felt that he might harm one of their children. The participant said, ".. .one of the reasons why I didn't report right away... [was that] the husband had indicated that if the wife had attempted to leave, that he would kill one of the children...". Harm to Family. Three participants said that they believed that it would be harmful to the family if a report was made. All of these cases related to fears about the break-up of the family. For example, one participant said: .. .1 was very concerned about the break-up of the family related to this [meaning a report], because even though this person, who was the parent, 110 had a bad temper, he was just a decent guy and I was afraid people wouldn't see that... Similarly, another participant stated, "I thought this family would be broken-up, these parents would be further devastated and they were really a good family...". Harm to the Therapeutic Relationship. Six of the participants believed that the therapeutic alliance would be harmed if a report was filed, but all recognized that a period of anger would have to be worked through. In one case, a participant said that he reported historical sexual abuse of a 17 ye. old girl who no longer had contact with her perpetrator. He said: I thought of all the issues about reporting and about not reporting, and about her risk, and I finally made the call, and I made the call against her wishes. I explained to her very carefully why I felt it was important for me to make the call. It did damage our therapeutic relationship. [The report] did not result in a conviction. It did result in some services being brought in a specialized nature, which she did take advantage of. And it was probably five years before she came back and talked to me, and at that stage said 'yes I understand', and then continued to see me as a client for another two years. But that was a pretty lonely five years in the sense of, did I make the right decision? In another case, the participant also expressed concerns about the therapeutic relationship ending. He said, "I was afraid that these people would be gone and that I wouldn't be able to follow-up on this boy anymore. I l l The Ministry's Response. Six of the participants expressed reservations about involving the Ministry when abuse was suspected. Concerns mainly related to perceptions around apprehension. Several quotations illustrate this point: .. .my concern about harm [was].. .that those people who are the authorities, as in the Ministry, would not see the whole thing in context and wouldn't see the value of the relationship between the parent and the child and nurture that along rather than cut it off based on an incident. I was probably a little concerned about what the Ministry might do, because you always hear about those weird things that happen.. .there's cases where the abuse seems obvious and nothing is done. I was more concerned that nothing would be done in a situation where something absolutely needed to [be done]. I'm not terribly impressed with government services. I don't know whether it's due to cutbacks or whether it's due to generally aspects of ineptitude, but I believe that both kinds of errors, incorrect rejections and incorrect identifications, are made by Social Services all the time... I had concerns that they, I immediately had concerns that they were going to take the two girls, the three girls and I was right. They were going to, they almost did. One participant said that he was concerned that the Ministry would not act on his report, "[that] they would not do anything at all." Criteria 5: Were you able to minimize the harm that reporting would cause? Eleven of the 21 participants said that they were able to minimize the harm that would result from reporting. The 10 participants who stated that they were not able to offset the adverse effects of reporting are discussed in the next section. In all but two of the 11 cases, participants said that they minimized harm (i.e., to the therapeutic relationship, 112 family, Ministry) by engaging their clients in the reporting process and being open with them. For example, one participant who expressed concerns that the father might retaliate against the child for disclosing physical abuse, and the uncertainty of the Ministry's response to the incident, stated that she minimized harm in this way: To just be straight forward, honest, no bullshit, and to be directive toward the Ministry and to provide people who are involved in the situation, that are out of their usual context, with some direction. Understanding their mandate, but also not knowing they don't have the opportunity to be experienced in everything, so they need the support too. I see it as support. If you support the police, if you support the Ministry, you support the parents, you support everybody. But your primary safety is kids and that is the priority regardless of everything. Other examples of how harm was reduced (across the various concerns) involved statements like: .. .by keeping them [meaning the family] informed, you know, on a moment to moment basis. .. .by involving them in the process. Giving them choices [about whether they would like to be involved in filing the report]. .. .getting the mother involved.. .1 continually had her involved as much as I could in the scenario, even though she was denying it. .. .it helped a little by the relationship I had with the parents. In another case, the participant said that her client threatened to harm herself if she reported the suspected child abuse. In this case, the participant said that she was able to minimize the harm to her client by making a verbal contract with her client. The participant said, "well, she promised, she promised me [that] she would not commit suicide at the end of the session, so harm to self, that one, that risk was minimized." 113 In a case that involved a divorce between two parents, a child reported that her father had hit her when she was visiting with him on the weekend. In disclosing this information to the psychologist, the mother expressed concern that if a report was made then the children might be harmed in some way for telling. Additionally, the mother stated that her children were planning on spending the following weekend with their father. To minimize the potential harm in reporting this case, the psychologist said: .. .and I told the children, I said, you have the option, you do not have to go to your father's [home] if you don't want to. And then they said that they didn't and so they weren't going to, they weren't seeing their father... Criterion 6: Did the reasons to report outweigh the reasons to not report? While 10 participants indicated that they had concerns about reporting the suspected child abuse, all of them went on to report the abuse in the cases they chose to describe. Participants stated that ultimately they were guided to report because of their legal duty and interest to protect children. In one case, for example, the psychologist was concerned that her report could potentially result in the children being apprehended. In the end, the psychologist reasoned that her decision to report was based on her legal obligation. She stated: .. .my concern was that if, if the Ministry for Children and Families were involved.. .let's say they were taken away from the mother and put into foster care, um, I have certainly heard stories about foster care and so I was concerned about that.. .regardless of any of those factors, I would still have reported it anyway because of the obligations that I had and the fact that the situation was clearly not a tenable one and those uncertainties, those potential negative outcomes were vague enough and they certainly didn't outweigh the specific dangers that I felt the children were in. 114 In another case, the psychologist learned from the child that he was "slammed-up" against a wall by his parent. The psychologist said that the dilemma in reporting this case was that the child said that if a report was made then he would deny it. The following quotation illustrates how the psychologist viewed the situation: I was seeing the child for individual therapy at the time. He was not prepared to tell me anything about what happened. He had let the cat out of the bag and he was, you know, he screwed up, like he was under instruction from home not to and he had [disclosed].... if the Ministry did come in to investigate.. .if he wasn't going to tell me he certainly wasn't going to tell a stranger, and we would have lost the therapeutic relationship with the family, and we also would have an investigation that where they come in and got no results, can put a child at risk.... we had some real concerns what the repercussions would be if the investigation was initiated at this point... [I reported] because under the law I have a clear duty to report.. .and professionally I have a responsibility to act in the best interest of the child.... I think you are on shaky grounds if you don't report when you have got a disclosure. One participant mentioned that he was concerned about the effect of a report on the therapeutic relationship with his 17-year old client. He said that opted to report in order to comply with the law: .. .the law was the law. I felt I had reasonable grounds at that stage to believe that this had been happening [meaning sexual abuse] and this child was at risk. They outweighed it and I reported. I didn't have very high hopes for anything really happening at that stage and that is where my ambivalence came about. But I think you get to the point where it does outweigh, you know, I have got to report for those reasons. So ya, they outweighed it. 115 Model Verification Data from 36 cases were used to verify the group model. To determine the accuracy of the model, in responding to particular decision criteria, each participant's actual decision outcome was compared to the decision outcome predicted by the model. When the predicted outcome of the model did not match participants' actual decision outcome, I considered the discrepancy to be an error in the predictive capability of the model. For example, the model was considered to make an error when it predicted that a participant would not report when in fact they did report. When an error was identified, I elicited information from the participant to better understand why the model failed to account for his or her decision outcome. To calculate the accuracy of the model, the number of errors were divided by the total number of cases, yielding a percentage of accuracy. It was expected that discrepancies between the models predicted outcome and participants decision criteria leading to a report to be minimal. Gladwin (1989) indicates that a model is adequate if it predicts more than 85% of decisions. Overall, the preliminary model accurately predicted the reporting behaviour of 93% of the cases that participants described. Of the 36 cases that were discussed by participants, there were 3 errors identified in the questionnaire. Each error will be discussed below. Table 13 shows which participants addressed which criteria in the validation of the model. 116 Table 13 Model Validation Data Case Signs of Signs/risk Some Concern- Able to Reasons Did you or risk factors other ed that minimize to report report factors meet your value to reporting the harm outweigh suspected for child threshold report would that the child abuse? to report other than cause would reasons to abuse or as you a legal harm? result not neglect? underst- one? from report? and the law? reporting ? Pilot Bl Yes Yes N/A Yes No Yes Yes CI Yes Yes N/A Yes Yes Yes C2 Yes Yes N/A No Yes C3 Yes Yes N/A Yes No Yes Yes Dl Yes Yes N/A No Yes D2 Yes Yes N/A Yes Yes Yes D3 Yes Yes N/A Yes Yes Yes E l Yes Yes N/A No Yes E2 Yes Yes N/A No Yes Fl Yes Yes N/A No Yes GI Yes Yes N/A Yes Yes Yes G2 Yes Yes N/A No Yes G3 Yes Yes N/A Yes Yes Yes HI Yes Yes N/A Yes Yes Yes H2 Yes Yes N/A Yes No Yes Yes 11 Yes Yes N/A Yes No Yes Yes Jl Yes Yes N/A Yes No Yes Yes K l Yes Yes N/A Yes Yes Yes K2 Yes Yes N/A Yes No Yes Yes K3 Yes Yes N/A Yes No Yes Yes LI Yes Yes N/A Yes Yes Yes M l Yes Yes N/A No *No* Ol Yes Yes N/A Yes No Yes Yes 02 Yes Yes N/A Yes No Yes Yes Qi Yes Yes N/A No Yes Rl Yes Yes N/A No Yes SI Yes Yes N/A Yes Yes *No* Tl Yes Yes N/A No Yes T2 Yes Yes N/A No Yes T3 Yes Yes N/A No Yes U l Yes Yes N/A Yes Yes Yes U2 Yes No Yes Yes U3 Yes Yes N/A No Yes VI Yes Yes N/A Yes Yes Yes Wl Yes Yes N/A No Yes W2 Yes Yes N/A Yes •No/Yes* Yes Yes A n asterisk (*) denotes errors 117 Of the three errors that were identified, two of the errors occurred because the model failed to predict participants' decision outcomes. In both of these cases, participants answered "no" when the model predicted "yes" to the question "Did you report child abuse or neglect." When asked about why they did not report the possible child abuse (despite the fact that that they both indicated that there were signs of abuse, and that the case met their legal threshold to report), each participant indicated that their client made the report themselves. Therefore, fidelity to EDTM methodology suggests that these two cases produce an error in the model. When these cases were further discussed, each psychologist said that he/she would have reported if the client had not done so. Additionally, both psychologists said that they made a follow-up call to the Ministry for Children and Families to confirm that their clients did in fact report the incident. It is important to note, however, that there was no way to confirm whether or not a follow-up call was made. It appears that this practice, as it was described, is risky to both the child and the practicing psychologists. With regard to the other error, a participant answered both "yes" and "no" to the question "were you able to minimize the harm that would result from reporting?" The rational that the participant provided was that she had a number of concerns about harm that would result from a report, some that she felt that she could minimize and one that she felt that she could not minimize. Given that the model does not permit participants to consider two or more criteria at a time, and that it failed to predict her decision outcome in -responding to specific decision criteria, the model is in error. Improving the Model. In addition to identifying errors in the model, I made notes about how the model might be strengthened. Three areas were identified: the wording of the criteria, the inclusion of criteria not relevant, and potential criteria. It is important to note that the 118 suggestions to follow did not produce any additional errors in the model; however, they help to clarify and build on the present model. Each w i l l be discussed below. With respect to the wording of the model, two recommendations are offered. First, I obtained feedback from participants that the word "possible" child abuse should be added to the phrase, "Were there any signs of or risk factors for child abuse?" The phrase would then read, "Were there any signs of or risk factors for possible child abuse?" The inclusion of the word "possible" is key because it suggests a lower level of confidence in reporting the case to the Ministry. Second, the words "report" or "reporting" should be more fully defined to mean "a formal report made to a child protection office." On two occasions I was asked to define the word "report" as participants said that they were initially thinking that it meant report to "a supervisor, program manager, principal, or other person," other than a Ministry social worker. In retrospect, the criterion "Was there some other value to report other than a legal one?" could have been dropped from the model. Review o f the data from both the model development and verification stage shows that this criterion was infrequently used as a factor in reporting decisions (n=2). The literature on reporting behaviour does offer some support for this criterion, however (Kalichman, 1993). A s it stands now, it is suggested that this criterion be more appropriately worded, "Was there a non-legal reason to report?" I believe that this change in wording makes the category clearer. During the validation stage, I identified data that suggested a new category that could have been incorporated into the model. Specifically, several participants talked about activating resources to reduce the risk for potential abuse. For example, one participant talked about a mother having access to respite care, and how this program served to reduce her stress level in raising her child who had attention deficit disorder. In another case, a psychologist talked about involving parents in parenting groups. Sti l l another participant stated that she was able to offset the risk of harm by enlisting other family members and relatives to monitor the home situation. In light of this data, I went back to the original data 119 and found further evidence for this category. Thus, it may be the case that if psychologists are able to intervene and offset risk factors for abuse, then their impetus to report may be lessened. This premise, of course, needs to be validated by research. Summary This section reviewed the results of the present study that involved the development and verification of a group model in reporting child abuse. The current model shows the ordering of six decision criteria that were important in psychologists' reporting decisions. When tested, the model predicted the decision outcome of 93% of the cases. In the next section, the results of this study will be more fully discussed. 120 Chapter V Discussion This final chapter begins with a general overview of the decision tree model that was developed to answer the research question, "What factors contribute to psychologists' decisions to report cases of possible child abuse?" Following, the factors within the model are highlighted. The contributions of the model are discussed with regard to theory building and naturalistic research. This chapter then reviews some methodological issues related to the overall model. Recommendations are offered for psychologists and policy, and suggestions are advanced for future studies. This section concludes with a discussion of the limitations of the present study. The Decision Tree Model Overview This study applied EDTM (Gladwin, 1989) to the field of child abuse reporting. Open-ended ethnographic interviews made it possible to elicit psychologists' perceptions of the criteria that they used in reporting possible child abuse. This approach contrasts with more structured methods, in which the criteria are generated by the researcher. Based on the data from the ethnographic interviews, criteria were identified and individual decision tree models were developed. These individual models were later combined to form a group model. The focus of the group model was to identify common criteria that psychologists used in their decision making, and to show the relationship between decision criteria and outcome. Given that the focus of this study was to develop a group model, broad criteria were developed to encompass case-specific criteria advanced by the participants. Consequently, the criteria within the model are general, containing both emic and etic elements. Thus, the current model best reflects broad categories of criteria that psychologists used in their 121 decision making, rather than case-specific criteria. For example, the second criterion in the model, "legal threshold to report," was developed by the researcher so that it would capture the wide range of individual differences in reporting behaviour. That is, some psychologists tended to make a report based on subtle signs of abuse and others required more explicit indicators. The term "reporting threshold" is widely recognized in the child abuse reporting literature (Finlayson & Koocher, 1991; Kalichman, 1993), and reflects the fact that there are individual differences in psychologist's subjective appraisals of what constitutes "suspected abuse." Despite diversity in participants' backgrounds (e.g., age, gender, work setting), the model shows that there were common criteria that psychologists used in their decisions to report possible child abuse. This finding suggests that there is regularity and consistency of criteria in the decision making process. Therefore, while this study does not capture all individual factors, validation of the decision tree model implies that the process is not wholly idiosyncratic. The decision tree model represents criteria in a certain sequence. The relationship between decision criteria and outcome are diagrammed. The arrows within the model show the order in which psychologists discussed decision criteria. The flow of the model is based on logic, and it is consistent with the sequence in which psychologists described their decision criteria during the interviews (i.e., from beginning to end). The model shows that some psychologists were guided by as few as three key factors in their decision making, whereas others considered up to five criteria before making a reporting decision. There were individual differences in the amount of time in which psychologists cycled through the model. Some participants reported immediately, and others did so over several days or weeks. Based on psychologists' comments, there appears to be a negative relationship between psychologists' perceptions that a child is in danger and their response time. That is, as the danger to a child increased, the psychologist's reported response time 122 decreased (i.e., reports were made more quickly). Similarly, psychologists' comments also suggest that they considered fewer criteria as the danger to a child increased. While danger to a child or severity of abuse were important factors in reporting for many psychologists, they are not represented in the present model. This finding was unexpected given that severity of abuse is widely recognized in the reporting literature as being a salient factor in reporting decisions (Finlayson & Koocher, 1991; Kalichman & Craig, 1991). It appears, however, that the absence of this factor relates to the fact that psychologists reported along a continuum of abuse severity. That is, some psychologists reported circumstances in which severity ratings were high and others reported when abuse severity was low. This finding is assumed to be very positive in that psychologists seem to be reporting the full range of circumstances surrounding possible abuse, and not just the most obvious cases of severe maltreatment. Research findings revealed that psychologists identified fewer physical signs of abuse, as compared to emotional and behavioural signs of abuse. Given this finding, psychologists may have had to draw more inferences about what was going on in the family and, therefore, there could have been more opportunity for the decision making process to be influenced by heuristics such as representativeness, time pressure, and availability of information. These heuristics, in turn, may have led to more expedient or delayed reports, and possible errors in decision making. It is important to note, however, that the present model does not address the issue of whether or not a report was valid or warranted. The types of abuse that psychologists reported most frequently related to sexual and physical abuse. This finding is consistent with studies cited in the literature (Kalichman, 1993) and may reflect the actual proportion of maltreatment that occurs. Alternatively, it may be that these types of abuse are reported because they are perceived as being more serious than other types of abuse or that they involve an intention to harm a child, whereas child neglect, for example, often involves the failure of caregivers to provide for the child's 123 basic needs (e.g., food, clothing, shelter, supervision). In light of this result, it is important to note that the present model may reflect more of the criteria surrounding sexual and physical abuse and less of criteria relating to emotional abuse and child neglect. It is unknown how the model would have changed, if at all, had there been a greater frequency of cases reported that involved emotional abuse and child neglect. It is conceivable that with more child neglect cases, for example, criteria relating to the ability of the psychologist to intervene—before reporting—may have emerged. Examples of ways that a psychologist may intervene prior to reporting could be assisting the client to access other resources in the community, such as food banks, parenting skills courses, and respite care. It is important to note that, while there was some support for this criterion, it needs to be more fully studied in future research. Decision Criteria As reflected in Criterion 1 (Were there any signs of or risk factors for child abuse or neglect?), psychologists in this study identified a wide-range of signs and risk factors. While a few psychologists reported possible child abuse based on subtle indicators, such as unexplained behaviour change and emotional distress, most reports were made following a verbal disclosure from a child or perpetrator. This finding suggests that verbal statements of abuse may increase psychologists confidence that abuse has occurred, thereby increasing the likelihood that they will report. Other research has found that psychologists are more likely to report, than not report, after receiving a verbal disclosure from a child (Finlayson & Koocher, 1991; Kalichman & Craig, 1991). A key finding represented in the current model is that reporting decisions are made in relationship to individual reporting thresholds (Criterion 2: Did the signs or risk factors meet your threshold to report as you understand the law?). The model shows that psychologists ultimately reported possible abuse once they felt the indicators met their legal threshold to report. However, when indicators did not meet their reporting threshold, 124 the model shows that psychologists did not report unless there was some other value in reporting (e.g., therapeutic value). This finding is consistent with earlier research linking reporting thresholds with reporting behaviour (Brosig & Kalichman, 1992a; Finlayson & Koocher, 1991; Kalichman et al., 1989). The identification of individual differences in reporting thresholds is an important finding in that it may explain much of the variability in whether or not a given case will be reported by psychologists. Furthermore, individual variation in reporting thresholds may influence the extent to which the Ministry substantiates a case. For example, psychologists who require only a minimal level of suspicion may detect subtle signs of abuse but they may also identify cases in which child maltreatment has not actually occurred. Consequently, cases that are reported based on low reporting thresholds may not be substantiated by child protective services. Similarly, psychologists with high reporting thresholds may correctly identify child abuse but they may also fail to detect more subtle signs of maltreatment. Therefore, high reporting thresholds may account for many of the unreported cases of child maltreatment. In moving down the decision tree, the model shows that some psychologists continued in a decision making process, despite believing that their thresholds to report had been met. Criterion 4 (Were you concerned that reporting would cause harm?) shows that a number of psychologists considered the potential consequences of reporting as part of their decision making. Most commonly, these psychologists said that they were worried about disrupting the therapeutic relationship, a parent retaliating against a child for disclosing abuse, and uncertainty about how child protective services would handle the case (i.e., that they would do nothing or would overreact and apprehend the child). This finding is important in that it identifies various constraints to reporting, which may play a key role in reporting behaviour (i.e., if, and when, a report is made). This finding also sheds light on how non-legal factors influence reporting decisions. While some clinicians invariably 125 follow the law, others consider a combination of legal and non-legal factors in their reporting decisions. The present model also reveals that psychologists attempt to minimize the perceived harm before they report possible child abuse (Criterion 5: Were you able to minimize the harm?). Efforts to minimize harm invariably occurred when psychologists perceived that the severity of abuse was low (e.g., cases of historical abuse when there was no current risk of harm to the child or other children), and when they had an ongoing therapeutic relationship with their clients. Typically, psychologists said that they tried to minimize harm by being open and "up-front" with their clients about the need to report. They invited families to become engaged in the reporting process, and many offered their clients the option to self-report within a specified time frame. Other psychologists said that they consulted with a Ministry social worker to help ensure optimal services for their clients. Taken together, these findings suggest two important points. First, while psychologists recognize their legal duty, they also appear to be acting ethically in carrying-out their reporting obligations. The practice of delaying a report, as described by psychologists, seems consistent with Standard II.2 (Responsible Caring) in the Code of Ethics for Psychologists (Canadian Psychological Association, 1991), which states that psychologists should "avoid doing harm to clients." Second, it appears that some psychologists have informally adopted a flexible reporting procedure that resembles the one proposed by Finkelhor and Zellman (1991), in which reports of child abuse are delayed in less serious cases. A critical criterion in decision making is assessing the overall benefits and detriments to reporting (Criterion 6: Did the reasons to report outweigh the reasons to not report?). As shown in the model, psychologists who believed that reporting would be beneficial (e.g., to the child, family) decided to ultimately report the case. When psychologists were mixed about reporting, they seemed to place greater emphasis on the potential to protect children from abuse, and less weight on the risks associated with 126 reporting. Thus, it appears that the legal duty to report was central in the minds of this sample of psychologists who ultimately decided to report possible child maltreatment. The present model also implies that psychologists who perceive reporting as being detrimental in some way may not report maltreatment, although the circumstance may meet their legal duty to report. This point, however, was not fully explored in the present study. It is suggested that future research further investigate the factors associated with failure to report cases. One of the most common dilemmas addressed in the child abuse literature relates to breach of confidentiality imposed by the legal mandate to report child abuse. Yet, psychologists in the present study were clear that confidentiality cannot be maintained in cases of possible child abuse and that reporting is mandatory. In fact, many psychologists in this study stated that they reviewed the limits of confidentiality with their clients at the outset of counselling. Thus, it appears that psychologists who reported child abuse did not perceive a dilemma between the legal mandate to report and breaching confidentiality. This finding suggests that psychologists have accepted and integrated changes to the ethical standard of confidentiality (i.e., that there are limits to confidentiality) with the emergence of B.C.'s child abuse reporting law. Prior to the inception of B.C.'s child abuse reporting law in 1981 (which has since been recodified), it was an ethical violation to breach client confidentiality. When compared to previous studies over the past 20 years (Swoboda et al., 1978; Muehleman & Kimmons, 1980; Kalichman et al., 1989), this research finding suggests that while breaching confidentiality was once a concern for psychologists, it no longer seems to be a factor in reporting decisions. This finding is not surprising given that one of the requirements to become a registered psychologist in British Columbia involves being informed of ethical codes and relevant laws pertaining to psychologists. It would be worthwhile to examine, however, if psychologists who fail to report place a higher value on confidentiality, or other ethical principles, than they do on the legal mandate to report. 127 Theory Building The present model is consistent with aspects of previous models and extends our understanding of the factors involved in child abuse reporting decisions. In particular, the findings of this study are congruent with Kalichman's (1993) model, which emphasizes the centrality of a reporting threshold in decision making. The present results suggest that as indicators surpass a given individual's threshold, that person will become increasingly inclined to report. The inverse is true when indicators are below one's reporting threshold. An important aspect of the present model is that various constraints to reporting were identified. This finding extends previous work in the field in that it highlights that, despite the legal requirement to report, psychologists consider non-legal factors in their decision making. The present study shows that psychologists were concerned about the impact of a report on children and their families, particularly when their clients expressed motivation in therapy and when "the possible abuse" was considered marginal. In contrast, severe abuse was reported immediately and without consideration to any constraints. The structure of the present model is similar to Kalichman's (1993) integrative model, which diagrams categories in a decision tree format. A limitation of Kalichman's (1993) model, however, is that it is not grounded in research. Rather, this model was adapted from a model in duty-to-warn cases. Therefore, Kalichman's (1993) model warrants empirical testing to determine its validity. In contrast, the present model has a number of strengths that enhance the quality of the findings. In particular, this study interviewed psychologists who had actually reported a case of possible child abuse. Thus, the present model reflects the criteria that psychologists said that they used in their decision making. Furthermore, the preliminary model in this study was validated by a second and independent group of psychologists. There were three factors that Kalichman (1993) described which were not identified in the present study. They were evaluating the cause for symptoms, consulting with colleagues, and making preliminary contact with child protection services. In review of 128 these categories, it appears that Kalichman's (1993) model focuses more on the procedures of reporting, rather than on the actual criteria involved in reporting decisions. Therefore, differences between the present model and Kalichman's (1993) work appear to be a product of the different objectives of the work, procedures versus criteria. Nonetheless, it is worthwhile to note that the data in the present study are consistent with the procedures described within Kalichman's (1993) model. For example, interview data show that psychologists are continually involved in evaluating the cause for symptoms. When the indicators are subtle, for example, some psychologists indicated that they probe and assess as to the nature of the indicators. When psychologists were uncertain if the indicators met their legal duty to report (and when the consequences for reporting were perceived as negative), their comments indicated that they consulted with Ministry workers and colleagues. The present model also validates parts of Brosig and Kalichman's (1992) model, which contends that a combination of legal factors, clinician characteristics, and situational and organizational factors interact to influence reporting. In particular, the present findings demonstrate the importance of legal factors in reporting decisions (e.g., knowledge of the law). Given the wide range of individual differences in reporting thresholds, however, it appears that psychologists have varying levels of clarity about or agreement with their statutory requirement, as reflected in the presence of non-legal factors in their decision making. The transcript data also show that there are differences between psychologists with respect to the influence of situational factors on reporting decisions. For example, the data show that availability of evidence and severity of abuse inconsistently played an important role for psychologists in identifying and reporting possible child abuse. That is, some psychologists reported when there were few signs and when severity was minimal and others reported when there were multiple signs of severe abuse. The absence of these factors in the present model may be explained by the research methodology that was used, in that it was impossible to represent individual differences in the group model. It is 129 recognized, however, that these factors and other factors (e.g., clinician and organizational factors) may play a critical role in decision making. These factors deserve further consideration in future research. A key point that Brosig and Kalichman (1992) make in describing their model is that any or a combination of their criteria may facilitate or inhibit reporting. This point is important in that it addresses constraints to reporting, which were identified in the present study. Although the present findings support particular categories of their work (e.g., legal factors), the most significant difference between models is that Brosig and Kalichman (1992) did not represent the relationship between their criteria and decision outcome, whereas I did attempt to develop an interactive and sequential model. Indeed, all of the factors of Brosig and Kalichman's (1992) model impact upon reporting decisions. In fact, their model is based on summarizing key findings in the literature. However, their model is limited in that it does not show how any variable combination affects reporting decisions across various circumstances, whereas the present model does. A third model to which the present study can be compared is Stadler's (1989) decision tree, which describes various reporting procedures that can be employed once it is determined that an abuse report is necessary. As discussed previously, Stadler's model describes four courses of action by which a psychologist may effectuate a report. Stadler's model is based on the ethical principles of beneficence (i.e., preventing harm) and autonomy (i.e., respecting an individual's right to make decisions). Relative to Stadler's model, there are strong indications that psychologists in this study used what Stadler refers to as an Option 1 response. Essentially, an Option 1 response is where clients are invited to contact the Ministry themselves. Psychologists in this study stated that they invited their clients to self-report because they wanted to maintain a positive and healthy therapeutic relationship with their clients. When the risk of harm to a child was low, a number of psychologists made arrangements for their clients to report on their own. As part of the agreement, these psychologists stated that they later called the Ministry office to verify if 130 the client had made a report. If a client did not file a report within the agreed upon date, an agreement was made that the psychologists would then report. The present study found little support for the other Option-responses that Stadler proposes. The contribution of the present study, relative to Stadler's work, is that it is descriptive in nature (i.e., identifying what psychologists actually do), whereas Stadler's model is purely prescriptive (i.e., identifying what psychologists should do, could do) and lacks empirical validation. The present findings also support aspects of Tymchuk's (1986) model of ethical decision making. As with Tymchuk's (1986) model, the present model shows that psychologists consider the consequences, risks, and benefits in reporting child abuse. The contribution of this finding is that it shows that psychologists consider their ethical responsibilities in effectuating their legal duty to report. While this practice is inconsistent with the law, it sheds light on the discrepancy between the law and actual practice. It is worthwhile to note that consideration of the consequences of reporting child abuse is also congruent with Kitchener's (1984) model of ethical decision making, from the critical-evaluative level. That is, principles of autonomy, nonmaleficence, and beneficence are considered along side of legal standards. Summary A major shortcoming in the literature is the notable lack of models and theory to guide empirically testable hypotheses concerning the reporting of child abuse. In my review of the literature, the present study represents the first empirical test of a decision making model of child abuse reporting. While the present study offers a model of reasonable fit, researchers are encouraged to test and expand this model. Naturalistic Research The use of EDTM (Gladwin, 1989), as a naturalistic method, adds to the literature on the reporting of child abuse in that it represents a shift in focus away from quantitative 131 research. While this study introduced a new method (i.e., Gladwin, 1989) to the field of child abuse reporting, the findings of the present study are strikingly similar to other studies in the literature, which have used quantitative designs. For example, the criterion representing a reporting threshold and the various constraints to reporting have been cited in other studies (Barksdale, 1988; Beck & Ogloff, 1995; Finlayson & Koocher, 1991). This finding suggest that there is consistency in decision criteria across research methods, and that the present findings are not a product of the research design. Through EDTM (Gladwin, 1989), it was possible to elicit the criteria that psychologists used in their decision making without having their responses limited by the researcher's hypotheses. Typically, in child abuse reporting studies, participants are only permitted to respond to a narrow range of factors that the researcher has identified. The present research design also made it possible to learn about decision making that occurred in the real world. Through ethnographic interviews it was possible to investigate participants' decision making from beginning to end, rather than having them respond to specific aspects of their judgement. The use of EDTM (Gladwin, 1989) as a naturalistic research method helped to capture the complexity of reporting decisions. For example, the data indicate that there are multiple purposes in child abuse reporting decisions. Psychologists in this study not only felt a duty to report possible maltreatment, but they were concerned about the outcome of their report. Furthermore, follow-up questions during the interviews clarified key decision criteria and the relationship between factors. Up to this point, no study has sought to investigate how any particular variable combination may affect reporting decisions. In my view, criticism that naturalistic decision making approaches lack adequate verification procedures appears to be unwarranted. With respect to the present study, for example, once the preliminary model was developed, it was then tested on a separate group of psychologists where its predictive ability was determined. Rather than simply indicating 132 that the model was reflective of psychologist's behaviour, errors permitted refinement and suggestions were made to improve the model. Summary EDTM (Gladwin, 1989) is a relatively new research method used in qualitative studies. Nonetheless, this method appears to be a useful tool in generating a decision making model, which offers a glimpse into key factors involved in reporting decisions. In the next section, I will discuss three methodological issues influencing the decision making model that was developed. Methodological Issues There are three methodological issues that impact the overall decision making model. First, an important aspect of the decision tree model that was developed is the influence of framing effects. Framing effects refer to the relationship between how the problem is presented and how the problem is solved. With respect to the present study, the decision making model is necessarily influenced by the manner in which the research question was framed. Specifically, psychologists were asked to discuss a time when they decided to report a case of child abuse, and the factors contributing to that decision. Given that participants were not asked to discuss times when a report was not made, the present model is likely restricted in its complexity due to the absence of failure to report cases and criteria. Therefore, the model should be evaluated with sensitivity to the decision frame that was used. Second, to say that the present model achieved validation in 93% of the cases refers to the percentage that participants' decision outcomes (i.e., of reported cases) were successfully predicted by the model in responding to particular decision criteria. However, it is important to note that, during the validation phase, participants had no opportunity to contribute new decision criteria, with the exception to clarify errors in the model. Instead, 133 participants were asked to respond in a "yes/no" format to the researcher's questions about their use of decision criteria. Therefore, the validation stage served to endorse the use of previously established criteria, rather than to generate new criteria. Third, the flow of the model is based on logic, and it is consistent with the sequence in which psychologists described their decision criteria during the interviews (i.e., from beginning to end). While ethnographic interviews made it possible to identify decision criteria, it remains unknown, however, to what extent the order of criteria represents what actually happened. For example, participants may have reconstructed the sequence of criteria so that it would be logical and coherent to themselves and the researcher. The methodological procedure of inviting participants to think about their decision making before the interview may have facilitated a reconstructive process. Therefore, it is important to emphasize that the model is based on retrospection and may not reflect the true sequence of criteria. Rather, it is important to think of this model as having a plausible sequence or flow, but that the true sequence of decision criteria is unknown. Recommendations for Psychologists On the basis of the research findings, a number of recommendations for psychologists can be offered. First, given that much individual variation in reporting thresholds leads to a decision to report, it is recommended that psychologists become more familiar with the threshold criteria that are specified in B.C.'s child abuse reporting law. It is believed that increasing their ability to recognize the threshold standard will lead to more accurate and expedient reports. To accomplish this goal, it is suggested that: (1) psychologists be given a copy of B.C.'s child abuse reporting statute, and (2) that child protective services and legal personnel be available, on an ongoing basis, to meet with psychologists, and other mental health professionals, to review reporting criteria and discuss interpretations of B.C.'s current statute. It would also be worthwhile to review the procedures for reporting, including the time-line in which a report is required to be made. 134 Many psychologists in the present study reported some time after they were alerted to possible child maltreatment. A number of psychologists in this study believed that reporting possible child abuse would have a negative impact on the therapeutic relationship. While this concern is distinct from concerns about breaching confidentiality, it suggests the importance of reviewing limits of confidentiality with families at the beginning of the intake session. In this way, psychologists can begin to build an open and honest relationship with clients and families. Psychologists may also want to periodically review limits of confidentiality with their clients to minimize the potential for clients to misunderstand these limits. It is important to note that a number of psychologists in this study volunteered that they use this kind of procedure as a standard part of their practice when they first meet with an individual or family. However, given that the researcher did not formally ask each participant about their policies regarding the duty to report, this recommendation is based on the assumption that not all psychologists employed this kind of procedure in their practice. Third, given that reporting decisions are often an emotionally laden, multifaceted, and difficult process, it is suggested that psychologists and other mental health professionals regularly consult with colleagues to discuss cases in which a report may be required. In this way, professionals can receive some degree of objective support in the decision making process. Furthermore, consulting with others and documenting the decision making process may protect psychologists from possible liability that may ensue following a report to the Ministry. Fourth, it is recommended that psychologists engage in an ongoing process of monitoring their attitudes and the extent to which these beliefs might impact upon their decision making. Results of this study found that some psychologists were reluctant to immediately report due to negative beliefs about the reporting system. The literature in the field of child abuse is rapidly increasing. Therefore, it is suggested that psychologists stay current with literature on the identification of child 135 maltreatment. Psychologists can expand their knowledge base in recognizing early signs of abuse and neglect through such avenues as professional development workshops and peer-reviewed journal articles. Finally, from a training perspective, graduate programs in psychology and related fields can use this model to strengthen students' knowledge and skills in making reporting decisions. Emphasis can be placed on the distinction between legal and ethical aspects of the model. Also, students can be presented with clinical vignettes and invited to engage in a decision making process themselves. In this way, students can develop a greater sense of self-efficacy and confidence in their ability to deal with reporting decisions. Policy Recommendations Psychologists in this study felt strongly about the need to prevent child maltreatment. However, they appeared less confident about how best to protect children once maltreatment is identified. As the present model shows, psychologists considered a range of non-legal factors in their decision making (e.g., perceived harm to children and families), although their legal duty to report was met. This finding suggests that there is a gap between the prescriptive intentions of the law (i.e., that all cases of child maltreatment are to be reported immediately) and the clinical reality of reporting decisions. On the basis of this finding, two policy recommendations can be advanced so that the law can be more effectively applied. First, given individual differences in reporting thresholds, policy makers should ensure that statutory language is clear so that a reasonable person will know with some certainty the boundary between abuse and nonabuse. Specifically, it would be important to distinguish child abuse from parental discipline, and to examine the role of cultural factors. Admittedly, this is an area of ambiguity. Second, there is a need to clarify what constitutes "reason to believe" a child needs protection. Along this line, it would be important to distinguish the term "reason to believe" from clinical hunches, intuition, and professional 136 impressions. Providing some case examples might be one way to clarify the term "reason to believe." Given negative perceptions toward the Ministry, there is a need to maintain a co-operative relationship between psychologists and Ministry workers. Psychologists must continue to recognize their duty to report, and co-operatively work with Ministry workers. Chi ld protection workers must provide adequate supportive services and their departments must be adequately funded, staffed, and supervised. Psychologists may want to speak with a district supervisor i f they believe that their concerns are not being adequately addressed. Future Research The findings of this study suggest a number of possible avenues for future research. First, studies are needed to build on the findings of the present model. Replications in other provinces and the use of larger samples with diverse case characteristics w i l l be necessary to assess the generalizability of the model. Similarly, studies examining the factors involved in cases not reported are needed to determine i f different criteria are used when reports are not made. Reasons behind failure to report child abuse may make an even larger contribution to the reporting literature, and expand the current model. Given the ethical and legal implications of conducting this kind of research, it is recommended that researchers develop unique ways to balance the collection of data with participant anonymity. For many psychologists in this study, the ethical issue that they grappled with was how to best protect children from abuse, not their duty to report. Therefore, it would be fruitful to conduct research on reporting procedures so that recommendations can be made for clinical practice. Areas that would be particularly promising include identifying the procedures that psychologists use in reporting, the relationship between Ministry personnel and psychologists, specific information contained in reports, and follow-up procedures. A n important issue not addressed by this study is the extent to which the criteria identified in this study led to accurate reporting of child maltreatment. In fact, few 137 psychologists spoke about the outcome of their reports. Additionally, the researcher did not prompt psychologists to speak about the outcome of their reports, as this question went beyond the scope of the present study. Therefore, it would be fruitful for researchers to better understand the criteria that relate to both accurate and inaccurate reports of child abuse. Data in the present study reveal the complexity of indicators of child maltreatment. For example, many psychologists reported maltreatment based on various risk factors, such as when there was domestic violence or when a parent was mentally ill. In British Columbia, the requirement to protect children from future harm (i.e., if a child is "likely to be" harmed) adds immeasurably to the subjectivity of reporting decisions. Therefore, it would be worthwhile to conduct research toward the development of a psychological profile of parents who may abuse or neglect their children. Such a profile could be used to identify high-risk families for the purpose of education and prevention of child maltreatment. The findings of this study support the need for more ecologically valid research that examines factors involved in decision making. Future studies may want to validate and build on the present model with a particular focus on a time dimension. Incorporating time into this model would add significantly to its complexity, as well as to allow researchers to better examine how long, and under what circumstances psychologists may delay before filing a report of possible child abuse. One aspect of reporting child abuse that is not represented in the current model is the potential influence of stress on decision making. Review of the transcript data found some evidence that psychologists experienced stress in reporting child maltreatment. However, it is not known to what extent the possible effects of stress impaired or enhanced their decision making. Therefore, it would be valuable for future research to more fully investigate the potential influence of stress on decision making, and the factors that might mediate stress. 138 Finally, the data in this study show that the various indicators of possible child abuse are an important part of clinical decision making. Clearly, there is a need to further develop descriptions of abuse indicators, as well as a more detailed analysis of the relationship between indicators and levels of suspicion. Limitations of the Present Study The results of this study must be considered in light of several methodological limitations, and these shortcomings should be addressed in future research. First, the results of this study are based on psychologists' retrospective accounts about reporting that occurred weeks and, in many cases, months prior to the study. Therefore, the findings are limited with respect to memory failures, and the possibility that participants were not completely aware of or reconstructed their decision criteria. Furthermore, it is possible that participants described the flow of their decision making in a logical order that may not have matched the actual sequence in which they considered criteria. Consequently, no true causal sequence or ordering can be established from these data. The findings are best viewed in the context of psychologists' recollections and constructions of their reporting experience. Second, the robustness of the model is likely limited due to the decision frame that was used in this study. It is conceivable that the overall model would have been more complex and expansive had a broader frame been used (i.e., cases that also included no reports). With respect to the validation stage, it is important to note that participants did not have the opportunity to contribute new decision criteria. Rather, their role was to endorse/not endorse the use of previously established categories, and not to generate new categories. This may have limited the comprehensiveness of the model as well. A third limitation of this study is the small sample size and number of cases, which limits the generalizability of the model. However, it must be emphasized that the number of cases in this study is typical of other studies using this methodology. Validation of the 139 overall model was achieved, however, despite the relatively small sample size. In addition, the generalizability of the model is limited by the homogeneity of the sample, as all participants were affiliated with the same professional organization. Sharing a common professional organization may have created a homogeneity in psychologists' decision criteria. In the future, it would be important to include participants from other professional organizations to examine the generalizability of the model. A fourth limitation of this study was the fact that the sample was a self-selected group of psychologists who were interested in sharing their personal experiences in reporting child abuse. It may be that the sample of psychologists who volunteered may be an idealized group of clinicians, who perceive themselves as most knowledgeable with regard to reporting requirements. It is important to also note that the data may have been influenced by demand characteristics, as is common in with research on legal and ethical issues. Fifth, given that the focus of this study was on the development of a group model, it was not possible to represent all idiosyncratic factors involved in reporting decisions for each case. Consequently, criteria within the present model are limited to general categories. Such factors as characteristics of the psychologist, situation, and victim attributes, for example, are not represented. Finally, although I tried to identify the criteria that participants discussed, the categories that were developed may be limited by my meaning structures. Conclusion The purpose of this study was to develop a model that identifies the key criteria that psychologists used in reporting cases of possible child abuse. Additionally, this study sought to validate the criteria within the model with a separate group of psychologists. The present model depicts decision criteria in a sequence that matches the order in which psychologists described their decision(s) to report. There are some concerns, however, 140 about the validity of the sequence of decision criteria. For example, psychologists may have reconstructed the order of criteria to make it more logical. Notwithstanding this concern, the present model shows the interrelationship between indicators of possible maltreatment and the notion of a reporting threshold. In addition, this study shows that psychologists engaged in a decision making process, even after their duty to report has been met, and that they considered the potential outcome of a report on families. The model further points out that, in acting ethically, psychologists actively attempt to minimize potential harm to children and families before they report. When they are unable to minimize potential harm, the data support the contention that psychologists are guided by their legal and ethical duty to report. The model suggests the importance for psychologists to discuss the limitations of confidentiality with clients, to become more aware of B.C.'s reporting law, and to consult with colleagues and supervisors in difficult cases. The model also suggests that policy makers may want to clarify the boundaries between abuse and parental discipline, and to specify what constitutes "reason to believe" a child needs protection. By examining and understanding the factors that psychologists use in their decision making process, it is hoped that a number of changes can be made in counselling practice and policy so that more abused children can be identified and protected. 141 References Ajzen, I. (1996). The social psychology of decision making. In E.T. Higgins & A.W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp.297-325). New York: The Guilford Press. Ansell, C., & Ross, H. (1990). Reply to Pope and Bajt. American Psychologist, 45, 399. Attias, R., & Goodwin, J. (1985). Knowledge and management strategies in incest cases: A survey of physicians, psychologists and family counsellors. Child Abuse and Neglect, 9, 527-533. Avis, J.M. (1992). Where are all the family therapists? Abuse and violence within families and family therapy's response. Journal of Marital and Family Therapy, 18, 225-232. Baird, K.A., & Rupert, PA. (1987). Clinical management of confidentiality: A survey of psychologists in seven states. Professional Psychology: Research and Practice, 18, 347-352. Barksdale, C. (1989). Child abuse reporting: A clinical dilemma? Smith College Studies in Social Work, 59, 170-182. Bauer, M.C., & Wright, A. L. (1996). Integrating qualitative and quantitative methods to model infant feeding behavior among Navajo mothers. Human Organization, 55, 183-192. Beck, K.A., & Ogloff, J.R.P. (1995). Child abuse reporting in British Columbia: Psychologists' knowledge of and compliance with the reporting law. Professional Psychology: Research and Practice, 26, 245-251. Ben Zur, H., & Breznitz, S.J. (1981). The effects of time pressure on risky choice behavior. Acta Psychologica, 47, 89-104. 142 Besharov, D.J. (1991). Child abuse and neglect reporting and investigation: Policy guidelines for decision making. Policy Issues, 15, 35-50. Borgida, E., & Brekke, N. (1981). The base-rate fallacy in attribution and prediction. In J. Harvey, W. Ickes, & R. Kidd (Eds.), New directions in attribution research (Vol.3, pp. 63-95). Hillsdale NJ: Erlbaum. British Columbia's Ministry for Children and Families Annual Report. (1999). Data Analysis Branch. Victoria, British Columbia. Brosig, C.L., & Kalichman, S.C. (1992a). Clinicians' reporting of suspected child abuse: A review of the empirical literature. Clinical Psychology Review, 12, 155-168. Brosig, C.L., & Kalichman, SC. (1992b). Child abuse reporting decisions: Effects of statutory wording of reporting requirements. Professional Psychology: Research and Practice, 23, 486-492. Bryant, G.D., & Norman, G.R. (1980). Expressions of probability: Words and numbers [Letter to the Editor]. New England, Journal of Medicine, 302, 411. Canadian Family Law Guide. (1988). Don Mills, Ontario: CCH Canadian Ltd. Canadian Psychological Association. (1991). A Canadian Code of Ethics for Psychologists. Ottawa, Ontario: Author. Child, Family and Community Service Act. British Columbia. S.B.C. 1996, c.46. College of Psychologists of British Columbia. (1985). Ethical standards of psychologists. British Columbia, Canada. Author. Corey, G., Corey, M.S., & Callahan, P. (1993). Issues and ethics in the helping professions (Fourth Edition). Pacific Grove, California: Brooks/Cole Publishing Company. Corey, G., Corey, M., & Callahan, P. (1984). Issues and ethics in the helping professions. Pacific Grove, California: Brooks/Cole Publishing Company. Coupey, E. (1990). Decision restructuring in consumer choice. Unpublished doctoral dissertation, Duke University. 143 Cram, S.J., & Dobson, K.S. (1993). Confidentiality: Ethical and legal aspects for Canadian psychologists. Canadian Psychology, 34, 347-363. Curtin, L.L. (1978). A proposed model for critical ethical analysis. Nursing forum, 17, 12-18. Eckenrode, J., Powers, J., Doris, J., Munsch, J., and Bolger, N. (1988). Substantiation of child abuse and neglect reports. Journal of Consulting and Clinical Psychology, 56, 9-16. Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 51, 380-417. Emery, R.E., & Laumann-Billings, L. (1998). An overview of the nature, causes, and consequences of abusive family relationships. American Psychologist, 53, 121-135. Eraker, S.A., & Sox, H.C. (1981). How decisions are reached: Physician and patient. Annuals of Internal Medicine, 97, 262-268. Family and Child Service Act. British Columbia. S.B.C. 1980, cl 1. Finkelhor, D., Gomes-Schwartz, B., & Horowitz, J. (1984). Child sexual abuse: New theory and research. New York: Free Press. Finkelhor, D., & Zellman, G. (1991). Flexible reporting options for skilled child abuse professionals. Child Abuse and Neglect, 15, 335-341. Finlayson, L.M., & Koocher, G.P. (1991). Professional judgement and child abuse reporting in sexual abuse cases. Professional Psychology: Research and Practice, 22, 464-472. Fischhoff, B., Slovic, P., & Lichtenstein, S. (1977). Knowing with certainty: The appropriateness of extreme confidence. Journal of Experimental Psychology: Human Perception and Performance, 3, 552-564. Gettys, C.F., Manning, C , Mehle, T., & Fisher, S. (1980). Hypothesis generation: A final report of three years of research. Decision Processes Laboratory, University of Oklahoma, Norman, OK. 144 Gladwin, C.H. (1989). Ethnographic decision tree modeling. California: Sage Publications. Haas, L.J., Malouf, J.L., & Mayerson, N.H. (1988). Personal and professional characteristics as factors in psychologists' ethical decision making. Professional Psychology: Research and Practice, 19, 35-42. Hampton, R.L., & Newberger, E.H. (1985). Child abuse incidence and reporting by hospitals: Significance of severity, class, and race. American Journal of Public Health, 75, 56-60. Harper, G., & Irvin, E. (1985). Alliance formation with parents: Limit setting and the effect of mandated reporting. American Journal of Orthopsychiatry, 55, 550-560. Hill, M., Glaser, K., & Harden, J. (1995). A feminist model for ethical decision making. In E.J. Rave & C C . Larsen (Eds.), Ethical decision making in therapy: Feminist Perspectives (pp. 18-37). New York: The Guilford Press. Holroyd, J.C., & Brodsky, A.M. (1977). Psychologists' attitudes and practices regarding erotic and nonerotic physical contact with patients. American Psychologist, 32, 843-849. Johnson, J., & Williams, M L . (1993). A preliminary ethnographic decision tree model of injection drug users' (IDU) needle sharing. The International Journal of the Addictions, 28, 997-1014. Kahneman, D., Slovic, P., & Tversky, A. (Eds.) (1982). Judgement under uncertainty: Heuristics and biases. New York: Cambridge University Press. Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80, 237-251. Kalichman, S.C. (1993). Mandated reporting of suspected child abuse: Ethics, law, and policy. Washington, DC: American Psychological Association. 145 Kalichman, S.C., & Brosig, C.L. (1992). The effects of child abuse reporting laws on psychologists' reporting behavior: A comparison of two state statutes. American Journal of Orthopsychiatry, 62, 284-296. Kalichman, S.C., & Brosig, C.L. (1993). Practicing psychologists' interpretations of and compliance with child abuse reporting laws. Law and Human Behavior, 17, 83-93. Kalichman, S.C., & Craig, M.E. (1991). Professionals psychologists' decisions to report suspected child abuse: Clinician and situational influences. Professional Psychology: Research and Practice, 22, 84-89. Kalichman, S.C., Craig, M.E., & Follingstad, D.R. (1990). Professionals' adherence to mandatory child abuse reporting laws: Effects of responsibility attribution, confidence ratings, and situational factors. Child Abuse & Neglect, 14, 69-77. Kalichman, S.C., Craig, M.E. (1990). Victims of incestuous abuse: Mental health professionals' attitudes and tendency to report. In The Victimology Handbook: Research Findings, Treatment, and Public Policy. New York: Garland Publishing. Kalichman, S.C., Craig, M.E., & Follingstad, D.R. (1989). Factors influencing the reporting of father-child sexual abuse: Study of licensed practicing psychologists. Professional Psychology: Research and Practice, 20, 84-89. Kalichman, S.C., Craig, M.E., & Follingstad, D.R. (1988). Mental health professionals and suspected cases of child abuse: An investigation of factors influencing reporting. Community Mental Health Journal, 24, 43-51. Kempe, C.H., Silverman, F.N., Steele, B.F., Droegemueller, W., & Silver, H.K. (1962). The battered-child syndrome. Journal of the American Medical Association, 181, 17-24. Kitchener, K.S. (1984). Intuition, critical evaluation and ethical principles: The foundation for ethical decisions in counseling psychology. The Counseling Psychologist, 12, 43-55. 146 Koriat, A., Lichtenstein, W., & Fishhoff, B. (1980). Reasons for confidence. Journal of Experimental Psychology: Human Learning and Memory, 6, 107-118. Lichtenstein, S., Slovic, P., Fischhoff, B., Layman, M., & Combs, B. (1978). Judged frequency of lethal events. Journal of Experimental Psychology: Human Learning and Memory, 4, 551-578. Loke, W.H. (1996). Models of judgement and decision making: An overview. In W.H. Loke (Ed.), Perspectives on judgement and decision making. Maryland, U.S.A.: Scarecrow Press, Inc. McNeil, B.J., Pauker, S.G., Sox, H.C., & Tversky, A. (1982). On the elicitation of preferences for alternative therapies. New England Journal of Medicine, 306, 1259-1262. Mabe, A.R., & Rollin, S.A. (1986). The role of a code of ethical standards in counseling. Journal of Counseling and Development, 64, 294-297. Meehl, P.E., & Rosen, A. (1955). Antecedents probability and the efficacy of psychometric signs, patterns of cutting scores. Psychological Bulletin, 52, 194-216. Melton, G.B., & Davidson, HA. (1987). Child protection and society. American Psychologist, 42,112-115. Miller, J.G. (1960). Information input overload and psychopathology. American Journal of Psychiatry, 116, 695-704. Miller, R.D., & Weinstock, R. (1987). Conflict of interest between therapist-patient confidentiality and the duty to report sexual abuse of children. Behavioral Sciences and the Law, 5, 161-174. Montbriand, M.J. (1995). Decision tree model describing alternative health care choices made by oncology patients. Cancer Nursing, 18, 104-117. Muehleman, T., & Kimmons, C. (1981). Psychologists' view on child abuse reporting, confidentiality, life and the law: An exploratory study. Professional Psychology, 12, 631-638. 147 Mullen, J.D., & Roth, B.M. (1991). Decision making: Its logic and practice. Maryland, U.S.A.: Rowman & Littlefield Publishers, Inc. Newberger, E.H. (1983). The helping hand strikes again: Unintended consequences of child abuse reporting. Journal of Clinical Child Psychology, 12, 307-311. Nightingale, N.N., & Walker, E.F. (1986). Identification and reporting of child maltreatment by Head Start Personnel: Attitudes and experiences. Child Abuse and Neglect, 10, 191-199. Offence Act. British Columbia. R.S.B.C. 1979. c.305, s.4. Otto, R.K., Ogloff, J.R., & Small, M.A. (1991). Confidentiality and informed consent in psychotherapy: Clinicians' knowledge and practices in Florida and Nebraska. ForensicReports, 4, 379-389. Paulsen, M.G. (1966). Legal protections against child abuse. Children, 13, 32-48. Paulsen, M.G. (1967). Child abuse reporting laws: The shape of the legislation. Columbia Law Review, 67, 73-92. Payne, J.W., Bettman, J.R., & Johnson, E.J. (1992). Behavioral decision research: A constructive perspective. Annual Review of Psychology, 43, 87-131. Payne, J.W., Bettman, J.R., & Johnson, E.J. (1988). Adaptive strategy selection in decision making. Journal of Experimental Psychology: Learning, Memory, and Cognition, 74,534-552. Pierce, R.L., & Pierce, L.H. (1985). Analysis of sexual abuse hotline reports. Child Abuse and Neglect, 9, 37-45. Pitz, G.F. (1974). Subjective probability distributions for imperfectly known quantities. In L.W. Gregg (Ed.), Knowledge and cognition. Potomac MD: Erlbaum. Pope, K.S., & Bajt, T.R. (1988). When laws and values conflict: A dilemma for psychologists. American Psychologist, 43, 828-829. 148 Pope, K.S., Levenson, H., Schover, L.R. (1979). Sexual intimacy in psychology training: Results and implications of a national survey. American Psychologist, 34, 682-689. R. v. Cook, (1985), 46 R.F.L. (2d) 174, Ontario Court of Appeal. R. v. Kates, (1986), unreported, Ontario Provincial Court (Family Division). R. v. Rahalkar, (1995), unreported, Ontario Court of Justice (General Division). R. v. Stachula, (1984), 40 R.F.L (2d) 184, Ontario Provincial Court (Family Division). Rest, J.R. (1984). Research on moral development: Implications for training counseling psychologists. The Counseling Psychologist, 12, 19-29. Ryan, G.W., & Martinez, H. (1996). Can we predict what mothers do? Modeling childhood diarrhoea in rural Mexico. Human Organization, 55, 47-57. Simon, H.A. (1956). Rational choice and the structure of the environment. Psychological Review, 63, 129-138. Slovic, P., Fischhoff, B., & Lichtenstein, S. (1980). Informing people about risk. In L. Morris, M. Mazis, & B. Barofsky (Eds.), Product labeling and health risks: Banbury Report 6 (pp. 165-181). Cold Spring Harbor NY: Cold Spring Harbor Laboratory. Slovic, P., Fischhoff, B., & Lichtenstein, S. (1982). Response mode, framing and information processing effects in risk assessment. In R.M. Hogarth (Ed.), New directions for methodology of social and behavioral science: Question framing and response consistency, (pp.21-36). San Francisco: Jossey-Bass. Slovic, P., Lichtenstein, S., & Fischhoff, B. (1988). Decision making. In R.C. Atkinson, R.J. Hemstein, G. Lindzey, & R. Duncan Luce (Eds.), Steven's handbook of experimental psychology (2nd ed., Vol. 2, pp.673-738). New York: Wiley. Slovic, P., & MacPhillamy, D. (1974). Dimensional commensurability and cue utilization in comparative judgement. Organizational Behavior and Human Performance, 11, 174-194. 149 Smith, T.S., McGuire, J., Abbott, D., Blau, B. (1991). Clinical ethical decision making: An investigation of the rationales used to justify doing less than one believes one should. Professional Psychology: Research and Practice, 22, 235-239. Stadler, H.A. (1989). Balancing ethical responsibilities: Reporting child abuse and neglect. The Counseling Psychologist, 17, 102-110. Swoboda, J.S., Elwork, A., Sales, B.D., & Levine, D. (1978). Knowledge of and compliance with privileged communication and child-abuse-reporting laws. Professional Psychology, 9, 448-457. Tversky, A., & Kahneman, D. (1973). Availability: A heuristic forjudging frequency and probability. Cognitive Psychology, 5, 207-232. Tversky, A., & Kahneman, D. (1974). Judgement under uncertainty: Heuristics and biases. Science, 185, 1124-1131. Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211, 453-458. Tymchuk, A.J. (1986). Guidelines for ethical decision making. Canadian Psychology, 27, 36-43. Walters, D. (1993). Mandatory reporting of child abuse: Legal, ethical, and clinical implications within a Canadian context. Canadian Psychology, 36, 163-181. Watson, H., & Levine, M. (1989). Psychotherapy and mandated reporting of child abuse. American Journal Orthopsychiatry, 59, 246-256. Williams, H.S., Osborne, Y.H., & Rappaport, N.B. (1987). Child abuse reporting law: Professionals' knowledge and compliance. The Southern Psychologist, 3, 20-24. Wilson, C.A., & Gettinger, M. (1989). Determinants of child-abuse reporting among Wisconsin school psychologists. Professional School Psychology, 4, 91-102. Wright, PL. (1974). The harassed decision maker: Time pressures, distractions, and the use of evidence. Journal of Applied Psychology, 59, 555-561. 150 Yates, J.F., Jagacinski, CM. , & Faber, M.D. (1978). Evaluation of partially described multiattribute options. Organizational Behavior and Human Performance, 21, 240-251. Zakay, D. (1985). Post-decisional confidence and conflict experienced in a choice process. Acta Psychologica, 58, 75-80. Zellman, G. (1992). The impact of case characteristics on child abuse reporting decisions. Child Abuse and Neglect, 16, 51-1 A. Zellman, G.L., & Antler, S. (1990). Mandated reporters and CPS: A study in frustration. Public Welfare, 48, 30-37. 151 Appendix A The Child, Family and Community Service Act of British Columbia (1996) 152 The Child, Family and Community Service Act of British Columbia (1996) Section. 14. Duty to report need for protection. (1) A person who has reason to believe that a child (a) has been, or is likely to be, physically harmed, sexually abused or sexually exploited by a parent or other person, or (b) needs protection under section 13(l)(e) to (k) must promptly report the matter to a director or a person designated by a director. (2) Subsection (1) applies even if the information on which the belief is based (a) is privileged, except as a result of a solicitor-client relationship, or (b) is confidential and its disclosure is prohibited under another Act. (3) A person who contravenes subsection (1) commits an offense. (4) A person who knowingly reports to a director, or a person designated by a director, false information that a child needs protection commits an offense. (5) No action for damages may be brought against a person for reporting information under this section unless the person knowingly reported false information. (6) A person who commits an offense under this section is liable to a fine of up to $10,000 or to imprisonment for up to 6 months, or to both. (7) The limitation period governing the commencement of a proceeding under the Offense Act does not apply to a proceeding relating to an offense under this section. Section 13. When protection is needed. (1) A child needs protection in the following circumstances: (a) if the child has been, or is likely to be, physically harmed by the child's parent; (b) if the child has been, or is likely to be, sexually abused or exploited by the child's parent; 153 (c) if the child has been, or is likely to be, physically harmed, sexually abused or sexually exploited by another person and if the child's parent is unwilling or unable to protect the child; (d) if the child has been, or is likely to be, physically harmed because of neglect by the child's parent; (e) if the child is emotionally harmed by the parent's conduct; (f) if the child is deprived of necessary health care; (g) if the child's development is likely to be seriously impaired by a treatable condition and the child's parent refuses to provide or consent to treatment; (h) if the child's parent is unable or unwilling to care for the child and has not made adequate provision for the child's care; (i) if the child is or has been absent from home in circumstances that endanger the child's safety or well-being; (j) if the child's parent is dead and adequate provision has not been made for the child's care; (k) if the child has been abandoned and adequate provision has not been made for the child's care; (1) if the child is in the care of a director or another person by agreement and the child's parent is unwilling or unable to resume care when the agreement is no longer in force. (2) For the purpose of subsection (l)(e), a child is emotionally harmed if the child demonstrates severe (a) anxiety, (b) depression, (c) withdrawal, or (d) self-destructive or aggressive behaviour. 154 Appendix B Child Abuse Legislation in Canada 155 8-1 Vi 3 O c o. s-\ Vi T3 Exceptions to immunity 1 Immunity Failure to report an offence Privileged communicat ions Confidentia lity Time line Degree of certainty Mandated reporters Nature of duty Age of child malicious reporting or without reasonable & probable grounds for the belief Yes Yes: max. $2,000.00 or 6 months applies does not apply forthwith reasonable and probable grounds any person mandatory Under 18 Child Welfare Act Alberta if the person knowingly reports false information Yes 1 Yes: max. $10,000.00 or 6 months, or both applies does not apply promptly reason to believe a person mandatory Under 19 Child, Family and Community Service Act British Columbia bad faith or failure to comply with the statute Yes 1 Yes: summary conviction applies does not apply forthwith information that leads the person reasonably to believe a person mandatory Under 18 Child and Family Service Act Manitoba bad faith or failure to comply with the statute Yes | Yes: for professional s only applies does not apply without delay information causing suspicion Prof: information that reasonably ought to cause suspicion 1 Public: any person mandatory Under 19 Family Services Act New Brunswick malicious reporting or without reasonable cause Yes | Yes: max. $10,000.00 or 6 months, or both does not apply does not apply immediate-ly report all information that a child has been or may be abused or neglected Prof: reasonable grounds to suspect 1 Public: info. a person mandatory Under 16 Child Welfare Act Newfound-land bad faith or failure to comply with the statute Yes | Yes: max. $500.00 or 6 months, or both applies does not apply without delay information of Prof: reasonable grounds to suspect 1 Public: every person mandatory Under 18 Child Welfare Act Northwest Territories false and malicious reporting Yes | Yes: max. $5,000.00 or one year, or both does not apply does not apply forthwith information indicating need of protective services Prof: reasonable grounds to suspect abuse Public: every person mandatory Under 16 Children and Family Services Act Nova Scotia malicious reporting or without reasonable grounds for trip hplipf*nr suspicion Yes Yes for profs only max. $1,000.00 applies does not apply forthwith reasonable grounds to believe Prof: reasonable grounds to suspect 1 Public: a person mandatory Under 16 Child and Family Services Act Ontario bad faith or failure to comply with the statute Yes Yes: max. $1,000.00 applies not stated forthwith knowledge or reasonable and probable suspect every person mandatory Under 18 Family and Child Services Act P.E.I Bad faith Yes Not specified applies does not apply without delay reasonable grounds to believe Prof: has reasonable grounds to believe 1 Public: any person mandatory Under 18 Quebec malicious reporting or without reasonable grounds Yes Yes: max. $5,000.00 or 6 months, or both applies to crown too does not apply none reasonable grounds to believe every person mandatory Under 16 Child and Family Services Act Saskatchew an Malicious and false reporting Yes No not stated not stated none reasonable grounds a person permissive Under 18 Children's Act Yukon o > a* OO CD r CD in \t I—* • O 13 B' o s p a . 156 Appendix C Child Protection Legislation in Canada 157 Child Protection Legislation in Canada Alberta: Child Welfare Act, S.A. 1984, c. C-8.1, ss. 3, 4, 91(4). British Columbia: Child, Family, and Community Service Act, S.B.C. 1996, c. 46, s. 14. Manitoba: Child and Family Services Act, S.M. 1985-1986, c. 8 (C.C.S.M. C80), ss. 18, 18.1-18.4. New Brunswick: Family Services Act, S.N.B. 1980, c. F-2.2, s.30. Newfoundland: Child Welfare Act, R.S.N. 1990, c. C-12, s. 38(1),(2),(3). Northwest Territories: Child Welfare Act, R.S.N.W.T 1974, c. C-3, s. 30.1(1),(2),(3). Nova Scotia: Children and Family Services Act, S.N.S. 1990, c. 5, ss. 22(2)(a),(c),(e),(f),(h),(i),0), 23, 24, 25. Ontario: Child and Family Services Act, S.O. 1984, c. 55, s. 68. Prince Edward island: Family and Child Services Act, R.S.P.E.I. 1988, c. F-2, ss. l(l)(a), 14(1),(4),(5),49. Quebec: Youth Protection Act, R.S.Q. c.P-34.1. Saskatchewan: Child and Family Services Act, S.S. 1989-1990, c. C-7.2, ss. 12, 13(1), 74(4). Yukon Territory: Children's Act, R.S.Y.T. 1986, c. 22, s. 115. 158 Appendix D The BC Psychologist Newsletter 159 The BC Psychologist Newsletter Kirk Beck is a doctoral candidate in Counselling Psychology at U.B.C. He is conducting research on the reporting of child abuse. Psychologists will be contacted at random. Your participation is greatly appreciated. 160 Appendix E Initial Contact Letter: Model Development Stage 161 Appendix F Follow-up Letter: Model Development Stage 163 Appendix G Informed Consent Form 165 Appendix H Initial Contact Letter: Validation Stag 167 Appendix I A Decision Making Model of Child Abuse Reporting Questionnaire 169 A Decision Making Model of Child Abuse Reporting Questionnaire Instructions: On the basis of the case that you just described, please answer the following questions in a "yes/no" format. 1. Did you report suspected child abuse or neglect? 2. Were there any signs of or risk factors for child abuse or neglect? 3. Did the signs or risk factors meet your threshold to report as you understand the law? 4. Was there some other value to report other than a legal one? 5. Were you concerned that reporting would cause harm? If "Yes", please specify the type of harm: a) to the child; b) to the therapeutic relationship; c) to the family; d) to you as the psychologist; e) the response from the MCF; f) Other. 6. Were you able to minimize the harm that would result from reporting? 7. Did the reasons to report outweigh the reasons to not report? Yes No Yes No Yes No Yes No Yes No Yes No Yes No 170 

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.831.1-0089769/manifest

Comment

Related Items