Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Disruption in special needs adoptions : a British Columbia review Berland, Jeremy Quentin 1990

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

Item Metadata

Download

Media
831-UBC_1990_A5 B47.pdf [ 3.77MB ]
Metadata
JSON: 831-1.0107111.json
JSON-LD: 831-1.0107111-ld.json
RDF/XML (Pretty): 831-1.0107111-rdf.xml
RDF/JSON: 831-1.0107111-rdf.json
Turtle: 831-1.0107111-turtle.txt
N-Triples: 831-1.0107111-rdf-ntriples.txt
Original Record: 831-1.0107111-source.json
Full Text
831-1.0107111-fulltext.txt
Citation
831-1.0107111.ris

Full Text

DISRUPTION IN SPECIAL NEEDS ADOPTIONS: A BRITISH COLUMBIA REVIEW By Jeremy Quentin Berland B.S.W., The University of British Columbia, 1981 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SOCIAL WORK in THE FACULTY OF GRADUATE STUDIES (School of Social Work) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA AUGUST 1990 (c^Jeremy Quentin Berland, 1990 National Library of Canada Bibliotheque nationale du Canada Canadian Theses Service Service des theses canadiennes Ottawa. Canada K1A 0N4 The author has granted an irrevocable non-exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of his/her thesis by any means and in any form or format, making this thesis available to interested persons. The author retains ownership of the copyright in his/her thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without his/her per-mission. L'auteur a accorde une licence irrevocable et non exclusive * permettant a la Bibliotheque nationale du Canada de reproduce, preter, distribuer ou vendre des copies de sa these de quelque maniere et sous quelque forme que ce soit pour mettre des exemplaires de cette these a la disposition des personnes interessees. L'auteur conserve la propriete du droit d'auteur qui protege sa these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. ISBN 0-315-64073-1 Canada 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. (Signature) Department of f ^ A - p S - t ^ o /CS The University of British Columbia Vancouver, Canada Date f o DE-6 (2/88) 1 1 ABSTRACT This paper describes a study of the adoptions of 82 "special needs" children placed for adoption in British Columbia between 1985 and 1989. The adoptions of 41 of these children were not completed. This study examines variables related to the children placed for adoption to determine those factors which appear to be related to disrupted adoptive placements. The paper traces the development of adoption through history and reviews the findings of major research studies in special needs adoption disruption. Using the findings from the review and adding hypotheses that appeared to be missing from other studies, a file review schedule was developed. With permission from the B.C. Superintendent of Family and Child Service, the schedule was used to review the children's adoption files. Analysis of the data obtained indicated that the age of the child at the time of adoption placement was a significant factor in adoption disruption. In addition, the age when the child was legally free for adoption was found to be significant. In both cases, the younger the child, the lower the risk of disruption. An additional important finding of this study is that children who are members of a sibling group are more likely to have their adoptions completed than children without siblings and children placed alone. i i i The pre-care experiences of children were thought to be an important factor in adoption disruption. The study shows that some experiences have a significant effect, notably those in which the extent of the biological parents' disability is clear to the child prior to the adoption placement. The presence of multiple special needs was not significantly associated with disruption except in the case of boys identified as having emotional/behavioural problems. The paper links the findings to those of other researchers, identifying implications for policy and practice. The resilience of the children studied and their ability to withstand serious trauma in their early years is an unanticipated finding of the study. Recommendations for addressing the findings suggest greater emphasis on maintaining sibling attachment for children in care, increased emphasis on assisting children to understand their family and personal history, and broad public education to eliminate myths about special needs adoption. I V TABLE OF CONTENTS ABSTRACT LIST OF TABLES ACKNOWLEDGEMENTS INTRODUCTION Hypotheses ADOPTION IN HISTORY < Infanticide Adoption in Early Civilizations Religion Western Adoption Tradition . . Development of Adoption Laws Adoption Law in Canada . . . . Adoption in British Columbia REVIEW OF ADOPTION DISRUPTION LITERATURE Definition of Terms Adoption Special Needs Adoption . . . . Adoption Disruption Review of Literature on Disruption . METHOD Subjects Measures Advantages of document study . Disadvantages of Document Study File Review Instrument xi viii x 1 6 10 11 12 14 16 17 22 24 32 34 34 35 35 37 63 64 66 67 69 72 V Validity and Reliability 7 3 Data Analysis 7 5 . . 77 Summary 7 R RESULTS Difference in outcome between boys and girls. . . 80 Children placed with siblings 8 1 Contact with siblings 8 2 Children who are older 8 3 Age when legally free 8 3 Age at placement 8 7 Family of origin 9 1 Returns to Biological Family 9 5 Children placed with relatives or with long-term foster parents 9 7 Number of placements and total length of time spent in foster homes or group homes . . 98 Number of placements 9 8 Time in foster homes and group homes 103 Reason for entering care 107 (A) Abandoned 10 7 (B) Physically abused 109 (C) Sexually abused 112 (D) Emotionally abused 113 (E) Parental alcohoi abuse 114 (F) Parental drug abuse 116 (G) Relinquished by biological parent . 116 (H) Parent deceased (I) Chronic neglect (J) Parent mentally ill (K) Parent mentally handicapped . . . (L) Other reason for entering care . . Multiple Special Needs Post-placement issues DISCUSSION Difference in outcome between boys and girls . . Children placed with siblings Implications for policy and practice . . . Contact with their siblings Children who are older Implications for policy and practice . . . Family of origin Implications for policy and practice. . . . Children placed with relatives Implications for policy and practice . . • Number of placements and total length of time spent in foster homes Implications for policy and practice . . . Pre-care experiences (A) Abandoned (B) Physically abused . (C) Sexually abused (D) Emotional abuse vi 117 118 119 .119 .119 .120 .124 128 129 129 131 132 133 135 137 139 140 140 142 144 145 146 147 148 149 (E) Parental alcohol abuse (F) R e l i n q u i s h e d (G) Chronic neglect (H) Other reasons for entering care . . . Implications for policy and practice Children with multiple special needs Other observations Areas for further research Summary FOOTNOTES REFERENCES APPENDIX A - File Review Schedule vxi 150 151 151 152 155 157 160 161 162 165 167 .172 T.I ST OF TABLES Table 1 Characteristics of disrupted and completed groups Table 2 Age When Legally Free (years) by Gender Table 3 Age when Legally Free (years) by Membership in a Sibling Group Table 4 Age at Placement (years) by Gender Table 5 Age at Placement (years) by Membership in a Sibling Group Table 6 Years Spent Living With Biological Family by Gender Table 7 Years Spent Living with Biological Family by Membership in a Sibling Group Table 8 Number of Foster and Group Home Placements by Gender I X Table 1 0 2 Number of Foster and Group Home Placements by Membership in a Sibling Group Table 10 . . . . 1 0 5 Months in Foster and Group Home Placements by Gender Table 11 1 0 7 Months in Foster and Group Home Placements by Membership in a Sibling Group Table 12 1 1 0 - 1 1 1 Reason for entering care by Disrupted and Completed Groups, Controlling for Gender and Membership in a Sibling Group. Table 13 1 2 2 - 1 2 3 Child's Special Needs by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group Table 14 126 - 127 Post-placement Factors by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group ACKNOWLEDGEMENTS This paper is dedicated with love and affection to my family for their enduring support and encouragement over the past five years. The assistance and guidance of my committee members, Madeline Lovell and Glenn Drover, has been invaluable. Their suggestions and corrections throughout have helped to make this a much more readable and cogent work. I am also deeply indebted to the Superintendent of Family and Child Service, Leslie Arnold. Without Ms. Arnold's approval and support, this project could not have been undertaken. The assistance of the staff of the Adoption Section of the Ministry of Social Services and Housing is gratefully acknowledged. 1 INTRODUCTION "Felton has iearned that one of the convicted killers had been taken from his mother at age seven months and cycled through foster homes until he was returned to his natural mother at sixteen. ' I think the community is partly at fault, I am partly at fault, for not doing enough,' he says 'This child had no lasting relationship since infancy. Nobody did anything for him. Maybe we should do more to help people..."(Kroll, 1990, p.13) As our society becomes increasingly complex, the importance of striving to make sense of the issues we are daily confronted with takes on added significance. The subtle inter-play between social forces is often at the heart of social problems. This paper examines the British Columbia perspective on one emerging issue - special needs adoption. More particularly, the paper reviews a smaller aspect: that of adoption disruption. Social workers in the child welfare arena feel experientally that adoptions frequently break down. Evidence for this is unclear, but the belief appears to be strongly held by many practitioners. The purpose of this study is to examine the adoption disruption phenomenon and to develop a picture of the types of situations where disruption is most likely to occur. Due to the nature of the problem, an interview with either adoptive parents or adopted children was not 2 considered appropriate for this level of enquiry. Better understanding of the problem is needed before one can begin to interview the participants in disrupted adoptions. The pain of the adoption breakdown and the impossibility of obtaining a complete sample led to a decision to use a file sampling technique to gather information about adoptions which disrupted. Permission was obtained from both the Superintendent of Family and Child Service and the Deputy Minister of Social Services and Housing to examine sealed adoption records. This permission was probably unique in Canada and certainly had not been granted before in British Columbia. Armed with such formidable access, there was initially great temptation to attempt to discover as much as possible about the dynamics of each adoption. However, such an approach tends to dilute the usefulness of research due to the very large number of possible variables. Accordingly, the factors chosen for analysis were those which relate directly to the child or children and the placement process. By conducting a thorough file review for each child, it was hoped that a profile of factors tending to lead to disruption might appear. Since the closed adoption records theoretically contain all of the known information about a child, including both primary and secondary sources, one can reasonably assume that a complete profile of the child could be developed. The initial hypothesis of the research was that the pre-care experiences of the child coupled with placement history while in care, were critical factors in adoption outcome. Further reading in the disruption literature pointed to factors relating to the age of the child and to the presence or absence of siblings. While some studies have indicated that the child's sex is an important variable, there is no unanimity on this issue in the research. As the study was focussing retrospectively on the cases, it was possible to exclude the forecasting element that often forms a basis for adoption studies. Such studies must develop models for prediction and methods of testing the prediction that require observation over lengthy time periods. At the time of data collection for this study, the outcomes were already known. Thus, it was possible to identify any number of demographic variables and prepare summaries for each (eg: time in care, number of placements, nature of special need, number and location of siblings). This union of data collection need and file availability helped to ensure that the primary hypotheses could be tested. The conclusions of this study will contribute to the body of adoption knowledge generally, and be helpful to adoption workers in B.C. as they place special needs children for adoption. The study focusses exclusively on special needs children as this is the area which seems to be fraught with the greatest risk for both the child and the 4 adoptive parents. Disruption among healthy infants placed for adoption is very rare. As Kay Donley (1989) says, anyone can place a healthy infant for adoption, but it takes - special skill and expertise to place successfully a special needs child. This study will build on previous studies of its type and will seek to confirm the findings of other researchers. Since there are subtle differences between jurisdictions and among researchers with respect to definitions and practice, this study will be uniquely British Columbian in focus. For the purposes of this study, a special needs child is defined as any child with a physical or mental handicap and any child over the age of two years. Studies such as this may help to change the image of the special needs adoptive child by focussing on those special needs which present special challenges for placement. The theoretical propositions tested were all derived from an extensive literature review and discussions with Ministry of Social Services and Housing Adoption Section staff. The thesis is a simple one - that there are some characteristics of children whose placements have disrupted which are common to the group, and which may highlight those placements most at risk of disruption. The particular focus of this study is the identification of child related variables which affect the outcome of special needs adoption placements.As indicated in 5 the chapters to follow, adoption is a social practice which has existed and evolved for centuries. Until the supply of healthy infants was exhausted in the late 1960's, adoption was primarily a vehicle to satisfy the needs of parties other than the children involved. In ancient times, adoption was a means to guarantee title and ownership of property. Later, it became a means to enter wealthy families and improve one's rank in society. Still later, adoption fell out of favour when the large church organizations took in abandoned children to nurture fidelity to the church itself. Childless couples met their desire for family by adopting the relinquished child. The placement of children from urban areas, while motivated by good intentions, led to exploitation of some of those children in the farms and households where they were placed. The history of adoption, as described below, is not always an attractive account. The growth of the permanency planning movement finally began to focus attention on the needs of the children as a primary consideration beyond those of any other party to adoption. Rather than seeking to provide children for families, the goal became finding families for children. This shift in perspective was necessitated by the needs of the children available for placement, as well as the recognition by social workers of the obligations they faced as guardians of these children. Previous studies of adoption outcome examined the 'success or failure' of the adoption as 6 measured through the eyes of the adoptive parents. There was little in the literature to suggest that the children themselves might have something to say about their situation. The permanency planning movement changed this perception. In a similar vein, researchers began to look at the factors that contribute to successful adoption of special needs children. This shift from adoptive parents' needs or agency needs to children's needs is critical to the study of adoption disruption. Adoption of special needs children is but one element in a continuum of child welfare services provided in most post-industrial countries. Adoption fits into that continuum near the end of a long line of services to families and children. The services include: family planning information, pre and post-natal care, community drug and alcohol awareness programs, child abuse prevention programs in schools, parenting training programs, family counseling programs, crisis intervention programs, transition houses, subsidized housing programs, child protective services, child guardianship, and finally adoption services. This list is not meant to be exhaustive, but serves to indicate that adoption does not exist in isolation from other programs in the community. A number of these services have probably been provided to special needs children and their families of origin. Since special needs children are, by definition, older, we 7 can assume that most have lived with their family of origin for some period of time. Whether it is possible to correlate broad descriptions of care received prior to entering care by the state is one of the questions posed by this research. Ultimately, the research question becomes one of prediction and assessment of cases. Hypotheses A set of hypotheses was developed, based on a careful review of previous adoption disruption studies and related literature. In addition, the gaps in previous studies were identified and hypotheses developed to answer some of those questions, e.g. the effect of pre-care experiences on adoption, the effect of contact with siblings.The following hypotheses are posed in this research project. Some additional issues are considered, such as post-placement factors, but the following items form the core of the discussion: 1) There will be no significant difference in outcome between boys and girls. 2) Children placed with siblings will be likely to experience fewer disruptions. 3) Children who are able to maintain contact with siblings are more likely to experience fewer disruptions. 4) The disrupted group will be marked by children who are older at the age when legally free and older at the time of placement. 8 5) The length of time that a child spends with his/her family of origin will inversely affect outcome. 6) Children placed with relatives or with long-term foster parents will not experience disruption at the same rate as those children placed with 'strangers'. 7) The number of placements and total length of time spent in foster homes or group homes prior to placement will inversely affect outcome. 8) The pre-care experiences of children will affect adoption outcome. Those children with histories of physical and sexual abuse will fare poorly in adoption. 9) Children with multiple special needs will face more disruptions. By placing adoption in a broad historical perspective and linking that view to modern developments, the study will demonstrate the continually evolving nature of this familial event. At this stage of awareness of the needs of children in state care, it is vital that we refine our knowledge of those components that will lead to permanent families. We must develop a clear consensus of the fundamental core issues and develop a shared language for describing the adoption experience. 9 The following chapter reviews the history of adoption. By placing it in a larger historical perspective, adoption can be seen as an evolutionary process which has experienced a great many changes in focus over the years. The third chapter discusses the literature on special needs adoption in general and focusses particularly on those studies concerned with special needs adoption disruption. Chapter Four describes the methods used in this study and the theoretical underpinnings for those methods. The fifth chapter presents the findings of this study and links them to the hypotheses presented in Chapter Three. The final chapter addresses the results and examines them in light of previous studies and offers explanations for the findings. This study will add to the existing body of knowledge about special needs adoption by identifying those factors or combinations of factors which may be indicative of children at risk of adoption disruption. The study will also present a profile of the special needs children placed for adoption in British Columbia. 10 ADOPTION IN HISTORY Sargon, the mighty king, King of Akkad, am I. My mother was a vestal, my father I knew not ... In mv city, Azupirani, which is situated on the banK of the Euphrates, my mother, the vestal, bore me. In a hidden place she brought me forth. She laid me in a vessel made of reeds, closed my door with pitch and dropped me down into the river, which did not drown me. The river carried me to Akki, the water carrier. Akki the water carrier raised me as his own son. Akki the water carrier made of me his gardener. In my work as gardener I was beloved by Istar, I became the king, and for forty-five years I held kingly sway. (Kadushin & Martin, 1988, p.534) It has become commonplace that the face of adoption has recently changed - that healthy white infants are no longer available for childless couples. However, placed in its larger, historical perspective, this development can be seen as simply another stage in the adoption cycle. Adoption itself is a practice recorded in the earliest annals of our history. Benet (1976), in carefully tracing the development of adoption, makes the point that the practice of adoption can be traced to our earliest history. She refers to the legendary founders of the City of Rome, Romulus and Remus, who were suckled by a wolf, and to Moses saved from the river and raised as her own by the daughter of a Pharaoh. Kadushin and Martin (1988) refer to Sargon as quoted above. The Bible is replete with references to adoption and the 11 care of children. For the ancients, at least, adoption seem to have been a way to set apart the leaders and to erect elaborate veils of mystery around them. "Their founders must be seen as belonging unequivocally to the new order rather than to the old, and one way to achieve this discontinuity is to make a mystery of their parentage and the circumstances of their birth." (Benet, 1976, p. 22) Infanticide The practice of infanticide was certainly the more common manner of dealing with children born with some defect, or those whose parents had died, or children the family could not afford. Adoption seems largely to have been the practice of the ruling classes in ancient times. Infanticide at birth or infant exposure was common among all classes and was practiced on a very wide scale in all parts of the world. "The child who was worth rearing was 'born in due time' was 'perfect in all its parts' and had a vigorous cry." (Kadushin & Martin,1988 p. 37) The practice of child abandonment, while often resulting in death, also resulted in children being sold or taken into slavery and prostitution. The practice appears to have been motivated primarily by economic concerns for the family of origin. Our folk and fairy tale literature provides ample reference to the practice.1 The advent of Christianity with its respect for the sanctity of life began to slow the rate of infanticide, but the practice 12 continued until the 1800's in Europe. In some parts of the world the practice continues to the present day. Adoption in Early Civilizations Adoption as an alternative to infanticide was not an option for the same economic reasons that led to the practice of exposing children. However adoption is known to have been practised since at least the time of Sargon in Babylon, 2800 B.C. Economic more than social reasons were also the basis for adoption when it was practised. The ownership of property and succession rights were generally the concerns underlying adoptions in ancient times. Usually these were adoptions of desirable older children or young adults. Although family was seen as the preferred arena for adoption in these circumstances, the adoption process was viewed as an arrangement between two sets of parents to facilitate arrangements for property holdings. Distinctions were generally drawn between 'physical paternity' and 'social paternity'. These distinctions reflected the male role in property ownership. In many cultures, the survival of the family group, complete with its property and values, was all important. It was imperative to ensure continuity by either procreation or incorporation into the family unit. The economic or inheritance concerns of the parent (usually the father) and the selection of a male heir were the primary factors in a decision to adopt. If an heir was 13 not available due to death or infertility, adoption was considered. The cradle of Western Civilization, the Tigris/Euphrates Valley, was the home of the oldest written law which mentions adoption. The Code of Hammurabi was an ! attempt to identify some of the potential pitfalls of adoption. "It deals with some of the risks inherent in every adoption: that the adoptive parents will treat the child differently from a natural child; that the child will suffer | from a change of caretaker; and that the adopted child and ij I family will be unsuited to each other." (Benet, 1976, p. 23) I j The Hammurabic Code also contains provisions for wet-nursing and apprenticeship. Valid social reasons for adoption | existed in Ancient Babylon just as they did in other parts | of the world then (and now). It is significant that | Hammurabi foresaw the inter-personal difficulties adoption I can pose even though adoption existed for more explicit economic reasons such as owning of property, passing on a title, and establishing claim to an improved position in society. The nomadic desert tribes had little property to pass on other than livestock. Adoption was not practised by the Israelites, although other forms of "heir assurance" were undertaken such as legitimation, fathering children by concubines and the "law of levirate". The latter practice is defined as "the custom among the Jews and some other nations by which the brother or next of kin to a deceased man was 14 bound under certain circumstances to marry the widow" (O.E.D., 1971) - and by extension to care for his children. The Roman and ancient Chinese traditions provided for the adoption of both children and adults. This was chiefly to maintain the ancestor cults but also developed as a means to consolidate a power base or, in the case of adults, to develop a more aristocratic lineage. Goody (1969) reports that "from Julius Caesar and Augustus onwards, a considerable number of emperors, failing to beget sons, adopted them instead" (p.60). As with many other early cultures, adoption was almost solely concerned with the preservation of the male line. A notable exception was the Chinese practice of adoption of a future daughter-in-law, pledged early in life to marry. By placing her into her future family, her own family saved often scarce resources. In all of the early societies described above, adoption was primarily the purview of the wealthy. The poor, with no property or inheritance rights to speak of, were more likely to turn to infanticide or abandonment as a solution to the economic problems they faced with children. Religion Among the religions which have developed from the nomadic tribes, Judaism did not embrace the concept of adoption for inheritance purposes. There was, however, a tradition of caring for orphaned and needy children within the community. Islam remains strongly opposed to adoption. 15 This conviction rests on several bases: Allah knows all truth and therefore cannot be misled by the fiction of adoption; there is a patriarchal tradition wherein children can be legitimated by recognition; divorce is uncomplicated; polygamy is common; and women are in a subservient position. The Christian institutions, most notably the Catholic church, have presented conflicting positions on adoption over the centuries. One of the foremost concerns of most organized religions is the bedrock need to ensure the preservation and continuance of the institution itself. Adoption to the church therefore, was, favoured as a recruitment technique over adoption by individual families. As a result of the Christian churches' disapproval of infanticide, children were placed in institutions, monasteries, convents, hospitals and orphanages across Europe. Once placed, they were encouraged to stay within the organization (other possibilities being severely limited) and were inculcated to its traditions and needs. While Kadushin and Martin (1988) take the view that the establishment of institutions for the care of children was based on a societal obligation to provide an alternative to infanticide, Benet (1976) strongly suggests that there was a powerful element of self-interest in the churches' willingness to do so. It may be that both views are historically accurate, the one complementing the other. Benet's (1976) portrait of church as parent, caring for its 16 young and leaving them a legacy of work to do certainly fits with the inheritance basis of adoption of the early civilizations. Western Adoption Tradition One cannot minimize the importance of inheritance and property rights as a major factor in the development of formalized adoption laws. With the fall of Rome and the predominance of the northern tribes - a more nomadic group than the inhabitants of the Mediterranean basin whom they vanquished - much of the established social order of Southern Europe fell into disorder. The Christian church remained relatively intact and began a period of growth which led to the establishment of many large institutions. Apart from the organized activities of the larger churches, there was no legal adoption tradition in Europe until the late 19th Century. Feudal ties in Europe established that property rights (including care and control of children) were vested in the local feudal chief. The twin developments of a feudal state and a large Christian church established rules by which the custody of children was to be governed until the 19th Century. Under the feudal system, inheritance was no longer a family matter but one which the feudal chief conferred. Property was held only at his pleasure for the lifetime of its owner. On his death, property reverted to the chief. Kinship ties became subject to the needs and demands of the feudal state. (Trappings of 17 this system are extant today in Great Britain. Entailed estates still exist where property, by law, must be passed on through the blood line. If no direct descendant of the previous title holder is living, the search for a blood relative must go back a generation to the descendants of the previous title holder. The perception that adoption presented a threat to blood or family ties acted to delay the enactment of any formal adoption statute in England until 1926.) Goody (1969) notes that adoption was rarely considered to be a service concerned with the well-being of children. He traces the development of adoption through several cultures (as does Benet [1976]) and provides overwhelming evidence of adoption in the forms and for the purposes described above. Goody refers to three distinct functions of adoption as identified through his comparative survey: (i) to provide homes for orphans, bastards, foundlings and the children of impaired families; (ii) to provide childless couples with social progeny; (iii) to provide an individual or couple with an heir to their property. (Goody, 1969, p. 57) Development of Adoption Laws In both its common law and legislative history, adoption holds a place within a broader framework of laws 5452 relating to property (as described above) and guardianship. While this latter factor was dealt with extensively in some early law, it became largely a matter for the common law in England and other European countries up until the early years of the twentieth century. Developments through the Middle Ages as the feudal, ecclesiastical and common law developed led to ten types of guardianship: 1) In chivalry ... the right of a feudal lord to services of his tenant as soldier or knight ... thus the lord was entitled to recompense in the form of service by the heir of his tenant in the event of the tenant's death. 2) In socage ("the holding of lands in consideration of ... services of husbandry" ) -similar to above but land-based. The child heir of a deceased tenant could be made the ward of his "nearest relative to whom the estate could not descend" and was allowed advice and guidance. The obligation was maintained to the feudal lord for service but the inheritance of property on reaching age fourteen was maintained. 3) By nature and for nurture - "by nature applied to the eldest son who — was sole heir — by nurture extended to the other children." 4) Ecclesiastical - orders could be made for guardianship of both the young person and the estate. 5) By special custom - guardianship arrangements peculiar to specific manors 6) By election of the infant. 7) By prerogative - applied to members of royal families only. 8) Ami prochein and ad litem - orders made for the duration of legal action in which a young person was involved - had no effect on custody or estate management. 9) Testamentary or statutory, and 10) Chancery - Feudal tenures were abolished in England in 1660, thus eliminating certain forms of guardianship. By this act,the father was given "absolute authority to dispose of... custody after his death." (Clarke, 1957, p. 260-262) In describing previous theories about the reasons for the development of codified adoption law, Goody places such developments in a broader perspective, utilizing the three functions described above. In particular, Goody contrasts his findings with those of Maine (1931). Maine postulated that adoption was a manifestation of "the major long-term change in man's (sic) political development as the evolution of society from one based on kinship to one based on 20 contiguity. Adoption is seen as a mechanism in the process of transition, a legal fiction that permitted the revolution to take place imperceptibly." (Goody, 1969, p. 68-69) However, this mechanism provided few benefits for the child. Benefits to be gained from adoption, if there were any, accrued to the adoptive parent. Benet (1976), Kadushin and Martin (1988) and Tizard (1977) suggest that, historically adoption was primarily concerned with the wealthy and privileged class. Further, its central focus was on the needs of the adopting male parent. Emotional or social considerations respecting the children, their mother, or the adopters' wives were rarely taken into account. Adoption does not enjoy a body of common law history. Formal adoption laws were not written in European countries until the early decades of this century. North American adoption law is generally considered to have its roots on legislation first passed in either Texas or Massachusetts in 1851.2 Kadushin and Martin (1988) suggest that the development of such laws (prompted perhaps by socio-economic forces springing from the industrial revolution ) were "probably delayed by the development of the relation between master and apprentice so that orphans and children of indigent parents could be bound out to obtain care in this way" (p.535). The former colonies of England enacted specific adoption legislation before the 'mother of Parliaments' did. 21 To some degree, this was required because social pressures due to massive immigration and the lack of an established social structure (such as church organizations, poor houses, large estates, extended family) led to few resources for children. Here again, child welfare legislation, in its broadest sense, can be seen as an attempt to cope with social pressures in the community. The American states were far ahead of the British House of Commons in identifying the need for adoption codes. While the early statutes lacked the legal complexity of later laws, they apparently shared some common characteristics: "a requirement that both biological parents consent; a joint petition by both adoptive parents; judicial review and decision; and complete severance of the relationship between biological parents and the child" (Kadushin & Martin, 1988, p.535) . The United States adopted specific legislation relating to children prior to 1900 while Britain, Canada and Australia lagged far behind. This is not to say, however, that similar institutions were not developed in these countries despite the legislative deficiency. In addition, and sadly, the poor house flourished as a resource for children until the abolition of Poor Law principles in the late 1800's. The investigative review of prospective adoptive parents was not fully institutionalized in adoption 22 legislation in the United States until the mid-1930's. Even then, adoption remained primarily the purview of the adopters - the children awaiting adoption were displayed for selection. Due to the excesses of this practice attention was devoted to standards for adoption and the requirement to provide a service to adoptive parents and children. The Child Welfare League of America published its standard for adoption practice in 1938. Adoption Law in Canada Adoption legislation in Canada dates from an 1873 New Brunswick statute. The British Columbia Adoption Act dates from 1920 and has been amended several times, most recently in 1990. Across Canada each province developed its own legislation concerning the protection and adoption of children. By the mid 1920's, all provinces with the exception of Newfoundland had both child protection legislation and adoption legislation enacted. However, the services mandated by these acts were originally provided by private societies whose activities were largely confined to the urban areas. The development of new legislation required provincial governments to become more involved in the delivery of child welfare services generally. This required that provinces ensure that both rural and urban services were developed. In British Columbia, as in many other provinces, this resulted in a provincial government delivery system in areas 23 outside Vancouver and Victoria. In the large cities, private societies delivered child welfare services. However, a provincial bureaucracy was required to be established in order to regulate and control the adoption service. The drafters of the early Canadian adoption legislation were considering the acts to be for the adoption of healthy infants and not the mentally and physically handicapped, older and special needs children. Hepworth (1980) reports that adoption legislation in the first half of the twentieth century had two primary objectives: 1) "The promotion of the welfare of adopted children in part through removal of the stigma attaching to such terms as bastardy and illegitimacy", and 2) "the clarification of the legal relationship of the adopted child to both ... natural and legal adoptive parents, thus making the rights of succession to property more certain" (Hepworth, 1980, p. 131). While Hepworth is undoubtedly correct, Hoggett and Pearl (1983) quoting Tizard reflect on other possible drives for the promotion of adoption - "the primary purpose of the adoption is seen to be the satisfaction of the desire of a married couple to rear a child; at the same time, a home is provided for a child whose natural parents are unable to rear it" (Hoggett & Pearl, 1983, p. 487). Adoption is both a legal and a social process. Ignoring one at the expense of the other is a common theme in both the literature and practice, one suspects. It is clear, 24 however, that the legal and social processes associated with adoption have changed over the past eighty years. Adoption in British Columbia Since this project is concerned exclusively with the British Columbia adoption picture, the following brief history will trace adoption legislation in B.C. with reference to the larger Canadian picture as required. MacDonald (1984) has summarized the following legal requirements for adoption which are common in Canada: 1) The adoptive parent ... must come within one of the categories of persons eligible to adopt... 2) There must be a valid written consent to adoption signed by the natural parents ... the natural mother only, if unmarried; and the Director of Child Welfare only, if the child is a permanent ward. A consent is also required from the adopted child if she has attained a minimum age, which varies from seven years in Ontario to twelve years in most other provinces. No adoption consent is valid unless the child is a minimum number of days old at its signing. This varies in the respective statutes from four days in the Northwest Territories to fourteen days in Prince Edward Island. 3) The child to be adopted must normally have been in the adoption home for a period of at least six months... 4) There must be a report prepared for the court, under the authority of the provincial Director of Child Welfare, assessing the merits of the_ proposed adoption and making a recommendation to the court. 5) The prospective adoptive parents must petition a judge for adoption of the child, declaring their readiness to assume on a permanent basis full parental responsibilities. 6) The judge, after considering the contents of the adoption petition and the report of the 25 Director of Child Welfare, must be satisfied prior to making an adoption order that the adoptive parents possess the ability and capacity to guide, maintain, and educate the child properly. He must also be convinced that the proposed adoption will be in the best interests of the adoptee. (MacDonald, 1984, p. 43-44) British Columbia's adoption legislation dates from 1920 and was titled "An Act respecting the Adoption of Children". This early Act "contained no provision for secrecy of adoption records. No report to the court was required on the fitness of the adoptive parents, although notice of application ... was required to be served on the Superintendent of Neglected Children." (MacDonald, 1984, p.45) The Adoption Act was amended in 1935 and provided for the "secrecy of documents". The new Act required the Superintendent of Neglected Children to "cause investigation to be made as to: -(a) The circumstances and character of the petitioner: (b) The fitness of the petitioner to assume parentage by adoption: (c) The fitness of the unmarried minor for adoption: (d) The mental and physical fitness of the unmarried minor's natural parents. (R.S.B.C. 1936, Chapter 6) In the early stages of development of social programs in B.C., child welfare services were provided in the two large urban areas by private societies and in the rest of the province by a centralized provincial government department. However, as in other provinces and jurisdictions, early child welfare policy was not as intricately interwoven with adoption policy as is now the case. There are few accounts of provincial adoption policy between enactment of the original legislation in 1920 and its wholesale revision in 1957. However, amendments between those years clarified and regulated some areas. For example, the 1936 amendments stipulated that the child had to have resided with the adoptive parents for at least one year prior to the application to the Court. The inheritance rights of children were also clarified in the 1936 Act. In 1957, a new adoption code was brought in by the provincial legislature. The 1957 Act maintained the requirement that the child reside with the adoptive parents for at least one year prior to the hearing date. An amendment in 1964 reduced this requirement to six months. This section has not subsequently been amended. MacDonald (1984) points to this new Act as setting a new standard for clarifying the relationship of adoptee to adoptive parents. In light of earlier confusion and concern about inheritance and property rights, this may be seen as both a positive and yet confounding development. "It was specifically declared that upon adoption a child 'for all purposes' became the child of his adoptive parents and 'for 27 all purposes' ceased to be the child of his natural parents."(MacDonald, 1984, p.47) This legal necessity was later to become a confounding development due to societal changes, which led to increased demands for 'right to know' legislation by adoptees, and the changing demographic picture of adoption. Later conflict about the place of the adopted child in the family and in society was reflected in the Royal Commission on Family and Children's Law (1975). The Commission recommended a better blending of legislation primarily concerned with the protection and welfare of children. In so doing, it tackled the inter-face between adoption, state guardianship and child protection. Among the recommendations made to address the confounding issues relating to the question of cessation of guardianship, adoptees' search for family of origins and subsidy payments to parents of special needs children were the following: 1) Where parental rights have been voluntarily or involuntarily terminated and the Superintendent has permanent guardianship of the child . . . such termination of parental rights should include termination of the requirement to obtain the consents of the parents to the adoption of the child . . . . 23) The best possible data base should be established for each child to include the obstetrical history and all significant factors relating to the mother, her pregnancy and her offspring . . . . 27) New legislation should require that a written summary of non-identifying background information be given to all adoptive parents at the time of placement and . . . to those who have adopted in the past . . . . 28) The applications by adopted persons for disclosure of identifying information about natural parents or other relatives should be made only to the Supreme Court . . . . 35) Where the best interests of a child will be served by being adopted by a particular adoptor and the adoption would not be possible without subsidization, the Superintendent should be able to enter into an agreement with the prospective adoptors to provide financial assistance as needed to the adoptive family. (British Columbia, 1975, summary) The Royal Commission was very far reaching, making 53 recommendations in the area of adoption alone. Societal changes have already rendered some of those recommendations out of date. For example, transfer of guardianship to the Superintendent on signing of adoption consents is current practice. It is also now Ministry policy that the consent of the natural father must be obtained. Where this is not possible, the issue of the lack of the father's consent must be addressed by the Superintendent in her report to the court. Others are only now seeing fruition - a tribute to the vision of the Commissioners. Regarding some of the other recommendations, few today would argue that an adoption registry is unnecessary or frivolous. Yet, the commission in 1975 could not recommend even a passive registry based on the evidence before it at the time. Recent changes to the Adoption Act and the Vital Statistics Act resulted in the creation of a passive 29 adoption registry. While this does not yet meet the demands of adoptees for full disclosure of information, it is a step forward. Similarly, the provincial government recently announced amendments to the Adoption Act to permit "assisted adoptions". While the program is too new to measure any impact, this is an encouraging development. The Family and Child Service Act (1981) enshrined other suggestions of the Royal Commission and modernized the Protection of Children Act. Further overhauling and modernizing of the Adoption Act has been planned for some time but has not yet materialized. Recently announced changes will amend the legislation to control private adoptions. The legislative arena is not the only focus of change in the adoption field. Rather, legislation and policy has had to change to maintain pace with developments in society at large. Hepworth (1980) details some of these changes in his analysis of the declining birthrate in Canada. In 1959, a record number of children were born in Canada:479,000; of these infants 20,000 were born out of wedlock and 7,000 of these were born to mothers under 20 years of age. In 1959-60 there were 12,800 adoptions. By 1969 the total number of births fell to 370,000, but the number of illegitimate births rose to 34,000 and in 1979 reached the highest recorded point of 35,600. In 1969-70 and 1970-71 adoptions also reached a peak, 20,300 and 20,500 respectively . . • births reached a low of 343,000 in 1973 and then rose to 359,000 in 1975 and 365,000 in 1976. Adoptions followed a similar pattern but reached a low of 14,600 in 1975-76 before rising again in 1976-77 to 16,200. (p. 132) 30 The declining birth rate and subsequent decline in the availability of healthy infants for adoption are among the critical factors that led to the drive for placement of children with special needs. The reasons for the decline in numbers of infants have been well-documented: changing social values about single parenthood, easier access to birth control, more access to abortion, expanded social programs for teenage parents, as well as the decline in the number of women of child-bearing age in the population at large. However, it is not merely the quest for children that has spurred the development of programs for adoption of children with special needs. The permanency planning movement has been growing across the Western democracies since at least the early 1960's. It was the convergence of public opinion (perhaps due to the dearth of healthy infants) with social policy research which pushed permanency planning into the forefront. During the 1960's a significant shift occurred in the focus of adoption services. As a result of the concern for the growing number of children remaining in long-term foster care, as well as the concern for children who appeared to be 'lost' in the foster care system, pressure emerged to find more permanent living arrangements for them. The concept of 'special needs' embodies a commitment to increase home-based and adoption services for black, older and handicapped children, many of whom have been in long-term care . . . Recognition that poor children, too have a right to a home and family led to greater utilization of adoption services for them. This effort, in turn, produced more widespread provision of adoptive homes for 31 black and other minority group children." (Kadushin & Martin, 1988, p.537) In British Columbia, a special needs child is considered to be any child over the age of two years. At one time, minority children of any age were considered to be special needs, but that designation has changed as the supply of healthy infants for adoption has diminished. This is a similar process to that experienced by many American adoption agencies which have found that the boundaries of special needs become ever more elastic as they discover that virtually all children are adoptable. This chapter has examined adoption in its most general sense and more specifically, how adoption has evolved in British Columbia. Adoption has moved from a process concerned principally with property and inheritance (and the power associated with property) to a social program primarily concerned with the well-being of children. While further developments are necessary to ensure that the child's best interest is always served, the changes of the last thirty years have addressed some of the worst excesses of the previous millennia. 32 REVIEW OF ADOPTION DISRUPTION LITERATURE It was hard seeing children that young know that much. They had the minds and thoughts of adults They knew about drugs and sex and dice. They'd pretend they was shooting up. And they'd steal too... Steve used to sleep sitting up out of fear of being beaten. His mama and grandma used to beat him for wetting the bed. And they would get in the bed and try to 'be with' one another. And I'd say, 'Who'd you see doing like that ?' They'd say 'Mom and her boyfriend'. They'd say their mama'd pull their clothes off and be laying there with the men. I called the psychiatrist, and she said Steve has seen too much of adult life to be a child. (Nelson, 1985, p.64) The previous chapter traced the development of adoption as a social institution over time. Such a review must paint the broad picture and cannot focus on the myriad small changes which have occurred over time in a complex area such as adoption policy. The lack of infants available for adoption, coupled with the perception that older children were drifting in foster care, has led to increased emphasis on the adoption of older, special needs children. The burgeoning field of permanent planning for children in care can provide some useful pointers for adoption studies.3 In particular, we can learn from post-care interviews with former children in care about their experiences so as to improve the quality of service for children currently in care. One of the major findings of the permanency planning movement has been the lack of planning for children. This 33 has led to foster care 'drift' where children are left in foster homes with little planning for permanent placement being carried out by the agency responsible. Concern about the numbers of children left adrift in foster care prompted legislation in the United States to ensure that children are either returned to their own families promptly or are placed in suitable adoptive homes. Permanency planning is the systematic process of carrying out, within the brief time-limited period, a set of goal-directed activities designed to help children live in families that offer continuity of relationships with nurturing parents or caretakers and the opportunity to establish lifetime relationships. (Fein, Maluccio et al, 1983, p.486) Partly due to this drive for permanent homes for children, greater attention has been paid to defining the factors that lead to successful adoption. Children are being placed for adoption today who were relegated to back wards of institutions for the handicapped only twenty years ago. Agencies have recognized that they must examine their own practice to determine if placement plans are appropriate. If adoption is to be a real option for special needs children, we need to develop a clear data base about the risks and rates of disruption. Armed with this information, agencies can make informed judgements about placement plans and decisions. This chapter will focus on one aspect of the placement of 'special needs children' - placement disruption. The 34 chapter will provide some working definitions, and review the relevant literature from the past two decades. Definition of Terms Definition of terms is important in this area as a great deal of research in the adoption field lacks clarity. For example, using some definitions, children might be counted as having a disrupted adoption even if they remained living with the same family under changed legal status (eg from adoption home to foster home). Similarly, significant differences may exist between children removed due to difficulties prior to completion of the adoption and children returned to agency care many years after completion. In the latter case, one cannot control for any number of events which may have occurred in the home and are unrelated to the adoption. Finally, lack of clarity in definition may lead to special needs children placed for adoption being considered equally with healthy infants for research purposes. Adoption In its stark simplicity, Barth and Berry's (1988) definition of adoption has great utility. "Adoption creates or expands a family through the legal severance of biological ties of a child to his birth parents and the establishment of new ties to an adoptive family" (p. 7). Kadushin and Martin offer a similar view - "adoption provides permanent substitute care for children whose birth 35 parents are unable or unwilling to provide the necessary support these children need" (p. 533). Special Needs Adoption While there are many interpretations of the term 'special needs', it will be defined according to current policy of the B.C. Ministry of Social Services and Housing. For the purposes of this policy, a special needs child is defined as any child over the age of two years, or a child of any age with specific disabilities. (This definition would be seen as too broad by a specialized special needs adoption agency like Spaulding for Children which identifies children over the age of six years as fitting the special needs category. Other agencies specializing in the adoption field in the United States consider a child over eight years to be special needs.) Adoption Disruption Adoption disruption has proven to be a more elusive subject for definition. There is wide discrepancy among the various definitions and not much consistency in their application in research. Disruption definitions generally acknowledge that disruption involves the child or children ceasing to be under the care and control of the adoptive parents. However, there does not appear to be an accurate method of collecting data given the wide range of definitions used by agencies, researchers and governments. 36 For example, Barth and Berry (1988) discuss the following sub-sets of the larger definition. Unofficial disruptions - in which the child's departure from the home is not reported to the adoption agency - may be more common than formal disruptions, but they are not reflected in the statistics Fost-adopt disruptions occur when foster parents care for a foster child with an understanding that adoption will follow the relinquishment of the child, and then decide not to adopt the child (before or after the child's relinquishment), (p. 21) Occasionally, the definition of a disrupted adoption is described in a manner required to facilitate the gathering of information. Due to the secrecy and confidentiality associated with adoption, record searches are difficult to achieve and interview subjects difficult to locate. Partridge et al. (1986) in their disruption study at the University of Southern Maine, defined disruption as "any adoptive placement which had been initiated and terminated within a 2 1/2 year period" (p. 7). Zwimpfer (1978) remarks that her review of "previous studies of adoption breakdown . . . are a heterogeneous selection, having widely varying definitions and methodologies which make direct comparisons with each other . . . quite impossible"(p. 9). It is also important to distinguish adoption disruption from dissolution. Adoption dissolution is generally considered to refer to those cases where a child is returned to agency care after the completion of the adoption. Mixing 37 the terms disruption and dissolution can lead to confusion about the nature of the research conducted and the validity and reliability therein. For the purposes of this study, adoption disruption is defined as any adoption placement which ended prior to an adoption order being granted, where the child or children were removed from the home. This definition excludes those situations where the child remains in the home as a foster child or where a child's status changes (eg. from permanent ward to change of guardianship under the Family Relations Act). Similarly, an adoption was not considered to have disrupted if the child died prior to granting an adoption order, except where the death was due to abuse or neglect. This definition is generally consistent with others. Kadushin and Martin (1988) write that "adoption disruptions are defined as permanent removal of the child from the adoptive home at any time before legal finalization of adoption" (p. 588). Barth and Berry (1988) state that "the designation of disruption indicates that the agency received custody of the child or was informed at the end of the placement when a child informally emancipated" (p. 22). Review of Literature on Disruption The review below of adoption disruption literature will be largely confined to those studies which clearly discuss adoption disruption. While this summary does not include adoption outcome studies, it is important to note that this body of literature has made a valuable contribution to the field of adoption disruption. Adoption disruption is a phenomenon associated largely with special needs children. Disruption was not generally considered to be a factor in the adoption of the healthy infants who comprised the largest part of the adopted population prior to 1970.4 As a result of the emerging nature of special needs adoption, there is not an abundance of literature on disruption. This gap is rapidly filling as the whole area of special needs children is receiving a great deal of attention. Kadushin (1967) conducted a study of 91 children placed for adoption when they were between five and twelve years of age. The sample group comprised forty-nine boys and forty-two girls. The researchers used a combination of interviews with adoptive parents and detailed file review for each child. To his surprise, Kadushin found that the adoptive parents reported a 'successful adoptive experience' in 82-85% of the cases. While this was not a disruption study in that the children continued in their placements, it is indicative of a growing awareness about the pre-adoption experiences of the children. The effect these experiences might have on the children was a prime concern. Given the conditions under which these children lived during their most impressionable years -in poverty, inadequately housed, with alcoholic, promiscuous parents who frequently neglected them, sometimes abused them, and only rarely offered them the loving care that is the prerequisite for 39 wholesome emotional development - how can one explain the generally favourable outcome of these placements ? (Kadushin, 1967, p. 25) Kadushin cites several sources as answers to his question and concludes that three factors are predominantly responsible for the success achieved by the children in the study - "the biological factor of constitutional resiliency . . . the sociological factor of upward displacement, which reinforces self acceptance [and] . . . the more important factor of making a therapeutic milieu available to such children" (p. 31). Unfortunately, Kadushin does not offer any explanation for the lack of success in the 15-18% of adoptions judged to be unsuccessful. In her widely cited article on older child adoption, Bass (1975) reports on her own unpublished survey of failed adoptions in the Bay Area of California. The study reviewed the cases of 53 children returned from adoption from 1972-74. The study appears to use Kadushin and Seidl's definition of adoption failure: "the removal of an adoptive child any time between placement and legal adoption" (Bass, 1975, p.503) . Bass found that age was the single most significant factor in the failure of an adoption. The older the child at the time 6f placement, the greater the hazards in that placement. Forty-two of the children (79%) were over the age of 2, and of these 19 (35% of the total) were in the 9-to-12 year age range. The most frequently stated reasons for the failures were behaviour problems of the children; disruption of the family functioning; 40 inability of the child to meet ... expectations; and internal marital or family problems, (p. 79) Bass does not indicate the research methodology, nor is the sampling technique and control group selection described. The author suggests more involvement by prospective adoptive parents and more emphasis on full sharing of information. Tizard's (1977) study reviewed the adoption of 30 children who were placed for adoption after the age of two. The study was conducted in England under the auspices of several British child care agencies. All of the children were in institutional care prior to placement. The children and families were interviewed 2 1/2 years after placement and again at 6 years after placement. Tizard echoes Kadushin in her finding that there must be a therapeutic element at work in adoption. She found that all of the placements were stable, although three couples "expressed reservations so serious as to amount to dissatisfaction . . . . The only behaviour problem . . . was attention-seeking behaviour . . . . In contrast the children who remained in institutions or who had been restored to their natural families had more frequent and more severe problems" (Tizard, 1979 p. 537). Tizard advances the position that early childhood experiences of abuse and neglect are not crucial. Early experiences have immediate effects which, if not reinforced, will fade in time. They will not 41 per se have long term influences, other than as a link in the developmental chain. The fact that they often appear to have a decisive long-term influence is because early experiences are usually reinforced It is just as likely to be the child's subsequent upbringing which affects his personality development as his early experience or loss. (p. 14) Zwimpfer's comparison of completed and disrupted adoptions in New Zealand established that adoption research had to move beyond the question of the success or failure of the placement. The study was not concerned with [the] broad band of unhappy adoptions, but with those which have resulted in the extreme solution of the removal of the child from its adoptive home. These . . . situations may not be potentially any worse or more unhappy than those which . . . find their later expression in attendance at psychiatric clinics or legal courts, and it may be that factors that predispose some adoption situations to total breakdown rather than other expressions of unhappiness may lie in demographic factors. (Zwimpfer, 1983, p. 170) Zwimpfer used a combination of file reviews and interviews with social workers to conduct her study. She did not interview either the adopted children or their parents. Her study comprised a comparison of eighty disrupted adoptions and eighty adoptions which were finalized. Due to variations in policy and practice over the years and areas included in the study, Zwimpfer chose to include adoption breakdowns (dissolutions) as well as disruptions. She then matched a control file for each disrupted adoption. The study was limited to children aged under seven at the time 42 the department was first aware of problems. "Once a child began to attend school and thus be more independent of the small family, other variables could be operating to an uncertain degree [and we] did not want to get involved with breakdowns involving adolescents as this introduces yet a further set of psychological and social considerations." (Zwimpfer, 1978, p. 19) Zwimpfer developed a checklist for use in the file review and compiled a list of 60 variables. These related to items about the adoption process, the child, the adoptive parents and the probationary period after placement. Due to difficulties in accessing file information which was not routinely forwarded to a central records department, Zwimpfer was unable to calculate a disruption rate. Her results showed many similarities to previous studies. She noted that length of marriage of adoptive parents seemed to be an indicator of 'vulnerability to breakdown' and correlated that with the age of the adopting parents. She further underscored the finding that the age of the child at the time of placement was very significant. This study clearly confirmed the significance of the age of the child ... and the differences were apparent from as early as one-month old at the time of placement. Most (70 in total) successful adoptions involved children placed before one-month, whereas in the breakdown group, only 32 were placed by one month... the significance of this distribution was confirmed statistically. (Zwimpfer, 1983 p.171) Zwimpfer found the gender of the child did not play a significant role in breakdown. However, she noted that children having ethnic backgrounds different from the adoptive parents were over-represented in the breakdown group. One of the most interesting findings of Zwimpfer's study was her analysis of the concept she named 'matching for marginality*. Briefly stated, the hypothesis was that difficult children were often placed with marginally acceptable adoptive parents. Zwimpfer cites Kadushin and Maas as authorities to support her view that social workers tend to develop images of acceptable families for adoption and of children for whom adoption is a viable plan. Maas concluded that these people were more flexible in their adoptive parent role expectations ...but agencies ... were perhaps imposing their own image onto parents of the type of child who was adoptable by t h e m — Kadushin — hypothesized that through a process of accommodation not unlike that of the economic marketplace, applicants of marginal eligibility would be more flexible in their acceptance of 'different' children in order to be considered more favourably by the agency." (Zwimpfer, 1983, p. 174) Zwimpfer's findings are significant in that they reveal that the success or failure of an adoption placement might lie in factors resting outside the purview of the adoptive family. The study also highlights the combination of factors likely contribute to adoption disruption rather than some intrinsic defect in either party. Hall (1981) studied all case records of children placed from 1975 to 1979 by a placement agency in Illinois. She focussed on children over 5 years of age. The total sample was 101 cases. The study was somewhat flawed due to the combination of qualitative and quantitative data collection and the possible subjectivity of the approach. Using a structured instrument for data collection, Hall confined her range of possible responses in such a way as to preclude alternative explanations for behaviours. She studied factors relating to adopting parents, adopted children and natural parents. Briefly stated, Hall found that there were no significant variables associated with the natural parents of the child which influenced disruption. However, Hall did find that there were "significant differences between the adoptive child in a successful adoption and the ... child in a disrupted adoption with regard to the presence of emotional disability, the child's adjustment to the adoptive parents and other children in the adoptive home, and the child's capacity to establish inter-personal relationships." (Hall, 1981, p. iii) Hall repeats Kadushin and Tizard's view that children are remarkably resilient in spite of early parental 1 deprivation. She found no difference between disrupted and : completed adoptions in terms of number of prior placements in foster care or returns to natural parents. In addition, > \ ! ji 45 the study revealed no effect of age or gender of the child on completion. However, Hall indicates that disruption was highly associated with emotional disability. She also posits that the child's ability to establish relationships is a significant factor in the disruption or completion of adoption. Hall cautions that this finding may be due to the child's worker interpreting conditions already in existence and thus may be an effect rather than a cause. When Hall examined the variables relating to the adoptive parents, the age of the adoptive father was found to be significant. Hall reported that the occupational level of the father was also a significant factor. While Hall raises some questions about this finding with respect to the particular nature of the clients served by the agency, she offers no explanation. The study is useful in that it provides some guideposts for further study and identifies those variables generally considered to be important. While Hall presents an array of tables and chi square scores, she provides no analysis of the findings and does not link them to previous studies. This lack of analysis of findings in the study renders it less than helpful in resolving the questions it raises about disruption. Cohen's 1981 study of adoption breakdown in Ontario is the only Canadian study to date. Cohen studied 320 adoption breakdowns which had occurred between 1975 and 1978. She used a file review technique and selective interviews with 46 agency personnel. Cohen found that the mean age on entering care was five and a half and the mean age for first adoption placement was seven and a half. Mean age at time of disruption was eight and a half. On average, the children had had two foster placements with an average stay in each of fifteen months. The children in this study were reported to be in good health without handicaps. Cohen did not make any comparison with completed adoptions nor did she offer much descriptive analysis of her study. The article describes a limited study of eight adoptive families, four of whom experienced an adoption disruption. This was a qualitative study and concludes that "a complex interactional process must occur for an older child to become truly a member of a family by adoption. Bonding, autonomy, initiative, and industry all enter into the process and provide stress. Both family and child have to be able to deal with this in an ongoing way." (Cohen, 1981, p. 130) These conclusions must be viewed with caution given the very small sample in the study (n = 8). The sampling technique is not identified in the monograph, raising questions of validity and reliability. A more substantive research project was undertaken by Festinger in 1984 in New York. This project was undertaken primarily to determine a disruption rate and uniform calculation of that rate. In addition, the author attempted to unravel some of the conflicting data about adoption, 47 particularly of special needs children. Festinger describes at great length the difficulties in arriving at a uniform calculation for adoption disruption. This is in part due to the lack of a uniform 'start-date' to begin counting adoptions and the difficulties one encounters in attempting to count those children already in placements. The study followed the "first 12 months of adoptive placement of 482 children who were placed alone, without siblings, and 415 children who were in sibling groups of two to five children I placed together ... in placements supervised by 36 voluntary agencies as well as the public services" (Festinger, 1986, p. 10-11). The study examined the files for each case and surveyed social workers by means of a structured interview and questionnaire. Festinger calculated an overall disruption rate of 8.2%. However, on closer analysis, some sub-groups within the larger sample showed quite different rates. Like other studies, Festinger found that the age of the child at the time the adoption agreement was signed was a significant factor. The placements of 13.1% of children age 11 or older did not hold, in contrast to 4.5% for children ages six to ten. These differences existed regardless of sex, race or religious affiliation. There were also considerable age differences between those whose placement disrupted (mean =11.2 years) and those who were adopted (mean = 9.8 years) (p. 15) Festinger's study also showed some variation from previously held beliefs about sibling placement. She found that children placed alone had a disruption rate higher (10.7%) than children placed with siblings (5.6%)5. While there were other findings of lower significance relating to gender, ethnicity, and religion, Festinger found that disruption was primarily related to the age of the child and whether the child was placed alone or with siblings. The study examined the placement history of each child as well as the placement experience. Like Hall, Festinger examined the number of returns to natural parents and the total number of foster placements. Her conclusion was that children whose placements disrupted were more likely to have had a previous adoption disruption. Further, they were more likely to have experienced more placements and to have had more types of placements than the children whose adoptions did not disrupt. Unlike Hall, Festinger found few adoptive parent characteristics to have a significant effect on disruption. She examined occupational status, age, race and education. None of these factors was found to influence outcome. The study also looked at the previously assessed problems of the child sample. Festinger found that children placed alone tended to have more problems and were more prone to disruption than children placed with siblings. She also notes that these children were more likely to have had multiple placements - possibly due to the problems they were experiencing. Siblings placed together have a natural support group at hand that can moderate the appearance of problems. It is also possible ... that some were separated from their siblings ... because of their problems, or perhaps some developed problems in response to such a separation. (Festinger, 1986, p. 33) Coyne and Brown (1985) reported on their 1980 study of adoption of developmentally disabled children. They reviewed the cases of 693 children placed for adoption across the' United States and Canada. While their primary focus was on the 'adoptability' of such children, a secondary finding of the study was a review of those placements which disrupted. They found an overall disruption rate of 8.7% (n = 60). As with studies described above, the age of the child was seen as being significant6. They found that for children aged below 7 years, the rate of disruption was low at 3.3%. However, the rate for children over 8 was considerably elevated at 17.7%. The study also examined foster parent adoptions as opposed to adoption by a new set of adoptive parents. Foster parent adoptions had a lower disruption rate (4.4%) than adoptions by families new to the child (10.4%). Disruption did not appear to be influenced by the type of developmental disability the child presented. However, there were missing data elements which prevented analysis of the degree of disability. Based on the limited responses received, the authors reported an inclination toward a lower rate of disruption for children with serious impairment. The study showed no significant difference on the variables of gender or race. Acting under the auspices of the Child Welfare League of America, Nelson (1985) undertook a study of special needs adoptive families. She studied 177 families who had adopted 257 children. The families were identified by both private / and public agencies in diverse areas of the United States. To be eligible for this study, the families must have "adopted a sibling group of three or more children; or a child who was at least eight years old at the time of placement ... or a child with an impairment ... that was likely to impose at least a moderate limitation on functioning." (Nelson, 1985, p. 8) It is a sign of the evolving nature of special needs adoption that Nelson was able to consider children under the age of eight years 'easy to place'. As with other studies, the project used a combination of file review and interview techniques to gather the information needed. Since the focus of this project was the adoptive families, Nelson devotes much attention to documenting the demographic data concerning the adoptive parents. She also describes the adoption placement process and supports available after legalization. As the study families had been selected on the basis of a completed adoption, there were no disruptions in the group. However, there were dissolutions in five families who had a total of seven children placed. Nelson studied adoptive parents' satisfaction with the adoption. Inasmuch as one can link this issue to outcome, the findings are an interesting contrast to most disruption studies. Nelson found that the satisfaction of the adoptive parents is not affected by either the number of previous placements or the length of time in care. Her research also indicated that satisfaction is not affected by the reason the natural parents' rights were terminated. Factors found to have a negative impact on adoptive parents' satisfaction included: an isolated child who was both difficult to reach and not given to reaching out, children placed where there was some degree of legal risk, children who had experienced previous adoption disruption and, situations where the parents' expectations about the child were disappointed. Nelson links this latter point to the need for full disclosure of all pertinent information about the child and adequate preparation of families by the placing agency. Of the seven children whose adoption ended, six were male. Five of the seven were placed with a sibling and six out of the seven were between five and seven years old at the time of their first placement. Nelson makes a number of recommendations about the placement and preparation of children and about the recruitment of families for special needs children. She emphasizes the need for adoptive parents to feel in control of the adoption process. Kagan and Reid (1986) reported on their study of another sub-set of the adoption picture - the adoption of emotionally disturbed youths. This study looked at the adoption placements of 78 children placed by a New York State agency. All of the children were described as being hard to place due to severe emotional and learning disabilities. The majority of the youths were boys and almost all had been neglected by their biological families. Over 50% were physically abused, and 91% had been placed in institutional treatment centres. Most had lived in a series of foster placements that began at an early age. Mean ages at the time of first adoptive placement and time of ... follow-up were 11.0 and 16.4 respectively, (p. 65) The researchers interviewed social workers and child care workers and reviewed files where possible. Kagan and Reid did not generate a disruption rate for their sample but state that of the total sample, 71% were legally adopted. Just over one-half of the sample had at least one adoptive placement disrupt. Given the particular focus of this study,and the sample selected, emphasis was placed on determining whether the behavioural and learning problems already identified would have an effect on adoption outcome. 53 The prospect of reattachment to another set of parents may be perceived ... as intolerable because of the pain remaining from previous bonds to biological and/or foster parents... such youths were assumed to be much less likely to succeed in adoptive families ... success with older youths was hypothesized to be directly correlated with assessments of the ability of the adoptive families to experience and manage the intense depression and rage of these youngsters, (p. 66) In contrast to other studies noted above, the researchers found that age at first placement, age at surrender and total time in care prior to being freed for adoption did not correlate with disruption. Certain behaviours such as property damage, aggressiveness and psychotic episodes were also found not to influence disruption. Kagan and Reid do however echo other studies in their finding that the gender of the child did not influence outcome. Factors considered to be significantly associated with disruption were: those children found to have been physically abused, children who were physically aggressive when first placed, total length of time the children spent in care prior to placement and the number of placements making up the total time. The authors also found that the assessment of the adoptive family was a key variable in the prediction of the family's coping skills and mechanisms-i In addition, the authors suggest that adoption of adolescent boys by single mothers may present special difficulties due to the 54 emergence of sexual issues and the concomitant problems associated with not having a strong bond with the child. Where the children have some form of permission from birth or foster parents to establish ties with a new family, Kagan and Reid indicate that the prospect for successful adoption is more likely. As with other studies, the authors strongly suggest that the provision of post adoption services is crucial to the success of adoptive placements. Partridge et al.(1986) at the University of Southern Maine Center for Research and Advanced Study, looked at data from 235 placements covering six agencies, private and public - urban and rural. The researchers used a combination of data collection techniques including interviews with agency staff, questionnaires and case record questionnaires. They also interviewed adoptive families and their workers. Although 60% of the placements were male, analysis of the disrupted adoptions revealed no significant differences on the basis of gender. Similarly there was no correlation between race and disruption. Data from this study are summarized below in table form. Characteristic Average Average age at separation 3 years Available for adoption 5.8 years Prior placements 3 % or N Physically abused 66% Sexually abused 52% Emotionally abused 73% More than 5 placements 31% Physical,mental handicap 75 (p. 34-35) This project did calculate a disruption rate and arrived at a figure of 8.6% after making allowance for uneven distribution of disrupted cases in the sample, This figure is very similar to Festinger's as reported above. Predictors of disruption, or factors which seem to predispose the children to disruption were: age (11 years vs 6.6), history of serious trauma, number of previous placements, time available for adoption. "There is every indication that children in the disrupted group are more experienced with loss, instability, and maltreatment and appear to be more emotionally disturbed and behaviourally dysfunctional than their counterparts in continuing placements."(p. 44-47) In 84% of the cases, the social worker described the child's 'behaviour, constitution or personality' as reasons 56 for the disruption of the placement. A further finding was that when children had already experienced an adoption disruption, half experienced another disruption. Finally, the authors comment that the child related factors are more associated with disruption than the family or agency factors. Barth and Berry's (1988) recent study is perhaps the most exhaustive look at the phenomenon of adoption disruption to date. The study followed 927 children placed for adoption after three years of age. The study sample was drawn from agencies in 13 Northern California counties. One of the study goals was to describe those characteristics which distinguish disrupted adoptions from completed adoptions. The following extract from their recently published book, indicates the hypotheses explored in the research. It is reprinted here as the model undertaken in this study is similar. They hypothesize that: 1) Children adopted by their foster parents have an added resource and are thus less likely to disrupt. 2) Children with more historical problems of abuse and neglect will display more behaviour problems and will be more likely to disrupt. 3) Children who have developed close relationships to prior caretakers will benefit by retaining contact with them. Thus, open adoptions, for children attached to prior caretakers, will be more stable than closed ones. 4) Children who have been in foster care longer will have endured more prolonged uncertainty and stress, and will be more likely to disrupt. 5) Families with continuous sources of support (no changes in social workers) will adjust to the placement with less likelihood of disruption. 57 6) Children with unresolved separations from prior caretakers will be more likely to disrupt. 7) Adoptive parents with prior experience as foster or adoptive parents will use those resources to create a more stable placement. 8) Parents with inadequate information and/or preparation prior to placement will have misleading expectations and will be more likely to disrupt. 9) Sibling placements into homes with no other children will proceed more smoothly than sibling placements into homes with other children. 10) Subsidies will be associated with smoothness of placements, with a lesser association to disruption. 11) Adoptions will proceed more smoothly when families have the social support of friends, ( relatives, and informal networks (church, school,, etc.). 12) Families in which the marital relationship is strained by the adoptive placements will be more likely to disrupt. 13) Parent support groups will provide resources that help families to maintain adoptions. 14) Adoptions in which the agency is available throughout preparation, the trial period, and in post-legalization difficulties will be more likely to remain intact. 15) Placements in which the child is reported to increase reciprocity toward his parents will be more likely to proceed smoothly and remain intact." (p. 63-64) The authors also set out to attempt to predict which adoptions were more likely to disrupt. A secondary task was to examine the disruption rate and attempt to clarify the calculation of such rates in order to provide more uniform measures. The authors established a disruption rate of 10% over the four years the study followed children. They note that average time from adoption placement to disruption was eighteen months and some placements may yet have disrupted after the study was completed. For this reason, they calculated that the final rate might be closer to 11%. Like Festinger, Barth and Berry realized that there may be variations within the overall rate which provide more helpful information. They point out that the disruption rate for children not adopted by foster parents was much higher than for those children adopted by their foster parents (17% vs 6%). Yet, even within the sub-set of foster parent adoptions, there were other differences which need elaboration. i The difference in the stability of foster parent placements holds, however, only for white children, who had a stability rate for foster parent adoptive placements of 94 percent, compared with 81 percent for new adoptions. Minority children had 90-percent stability rates for both fost-adoptions and new adoptions. Foster parent adoptions are less stable for boys than girls. (Barth & Berry, 1988, p. 229) As with practically all of the studies mentioned above, Barth and Berry found that the age of the child at the time of placement was a highly significant variable. Children in the disrupted group were found to have a mean age of 9 at placement compared to 7 years for the completed group. As Kagan and Reid (1986) found, boys tended to have a higher disruption rate (54% vs 38%). This correlation was not considered to be strongly related to disruption. The child variables considered to be highly associated with disruption were: age of the child, history of prior disruptions, and previously noted behaviour problems. If the child was 59 opposed to the adoption, the risk of disruption was considered to be elevated. In summary, with few exceptions, the studies cited above have found that the age of the child at the time of adoptive placement is a critical factor in adoption outcome. The range of ages considered to fall within the high risk category varied widely from one month (Zwimpfer, 1978) to 11 years (Partridge et al., 1986). This range is somewhat misleading as it reflects only the average age of children in the disrupted group. This does not indicate that these are the threshold ages for stopping adoption placement. The literature also displayed some ambivalence about the utility of sibling placements. While Barth and Berry (1988) comment that the placement of siblings in a home with no other children is contra-indicative of disruption, others have written that sibling placements are more prone to disruption e.g. Kadushin (1971). Many authors have commented on the pre-care experience of the children. While early writers like Kadushin (1967) and Tizard (1977) commented on the resilience of children, some of the more recent studies have focussed on the abuse and neglect children have suffered and the effect this early deprivation has had on subsequent adoption. Foster parent adoptions are a new enough feature of the child welfare system that they receive little attention in the literature. Nelson (1986) was looking primarily at adoptive family functioning and not at disruption, she noted that satisfaction was not as great in fost-adopt situations. Proch (1982) also found less satisfaction in foster parent adoptions. Yet, Barth and Berry (1988) and Festinger (1986) found that foster parent adoptions disrupted less than new adoptions. The rate at which adoptions disrupt has been a theme throughout the studies described above. Partridge et al (1986) estimate a range from 8.6% to 20% across the studies they reviewed. Within these rates however, significant ' differences exist when specific sub-sets are more closely examined. The whole question of a disruption rate as a single uniform figure appears to be open to debate. Adoption practice, policy and legislation vary widely across jurisdictions. Absolute comparison of any study is difficult without a careful analysis of the differences between areas. The presence or absence of adoption subsidies is thought to be a factor in adoption decision-making but service provision is universally considered to be a factor of prime importance. This chapter has identified some of the highlights of adoption disruption research over the past two decades. The primary focus of the review has been on those factors which reflect on the needs and characteristics of the children placed for adoption since this is the area of research described in the chapters to follow. Early on in the decision making process for this research project, it was 61 decided to focus exclusively on the child related variables. While it is clear that these variables do not exist in a sociological vacuum, it was considered important to identify as many features of the children whose placements disrupted as possible. Maintaining the child as the central focus of study is a first step toward debunking some of the myths in special needs adoption which hold that the issues which lead to adoption disruption are largely child-based. While this may be true in some cases, policy and practice should be* based on research not opinion. The identification of those children considered most at risk of adoption disruption will enable policy makers to frame policies which will address the needs of this very special group of children in state care. This chapter and the one previous have indicated the evolving nature of adoption over the millennia. It appears that until very recently adoption was primarily concerned with the needs of all of the parties to it, excepting the children. Drawing the focus back onto the children themselves is, in part, an attempt to place them at the centre of adoption studies rather than on the periphery. This study will contribute to the body of knowledge about special needs adoption in several important ways. It will look at the adoption population in British Columbia and examine similarities with studies as described above. It will also provide an extensive picture of the children currently placed for adoption in this Province. This new knowledge base will provide a platform for the development of policy as well as a jumping-off point for further research. 63 METHOD As in many jurisdictions, British Columbia's Adoption Act requires the sealing of adoption records upon completion of the adoption order. The legislation was drafted in this way in order to provide a degree of finality to the adoption. The 'as if born' terminology was taken to mean that all official records of the child's life prior to the adoption were to be filed away, and referred to only under very special circumstances. This, and similar legislation across Canada and the United States, has presented researchers with problems in obtaining an adequate data base for research into adoption. Concerns about the confidentiality of adoptive parents and adopted children have always loomed large in the decision to allow researchers access to closed files. However, the increased emphasis on the adoption of special needs children has meant that administrators of adoption programs have recognized the importance of research based practice in this area. The B.C. Superintendent of Family and Child Service has championed this research project as a first attempt to identify some of the important issues surrounding special needs adoption disruption. SUBJECTS Permission was granted in the fall of 1987 to review the files of children whose adoption had disrupted prior to granting of the adoption order. In addition, access was permitted to the files of a similar number of children whose adoptions were completed. A list was generated, by the Ministry of Social Services and Housing, of all of the disrupted adoptions since 1985. As all adoption records are held centrally in Victoria, B.C., all files were available through the Adoption Section. By virtue of the method used to generate the list of disrupted adoptions, (i.e. files where the adoption order was not made), some of the cases were not appropriate for study. Examples were: cases where the child died prior to the completion (n = 2); cases where the child remained living with the adoptive parent on a foster basis (n = 2); and cases where the child's status changed due to a successful application by the caregiver under the Family Relations Act (n = 1). In addition, ten files were not reviewed as they were missing or contained insufficient information. These revisions to the original list created a revised list of 41 children. Using the Ministry's existing working definition of special needs children - any child over the age of two years, sibling groups of any age, children with special health needs - all of the children in the disrupted group are well described as 1 special needs'. A control group of files was selected from the shelves of the adoption section. The files were selected on the basis of choosing every 15th file, starting at a random point on the shelves. If the file selected was not that of a special needs child, the next file was chosen. This purely random method of selection was necessary as there was no file database of special needs adoptions completed during the 1985-89 period. Since all adoption files are held centrally, the method of selection can be said to reliably represent a cross-section of the' completed adoptions of special needs children in British Columbia. In the period, 1985 -1989 there were 1,452 adoption placements in British Columbia. Children with special needs accounted for 881 of these placements. The study can be said to accurately represent a picture of the general population of special needs children placed for adoption. However, generalizability of the sub-groups within the larger sample to the larger population of children in care or children available for adoption is limited. This study represents an exploratory first step toward identifying the characteristics of sub-groups within the larger sample. A statistical data base describing all of the information available about children placed for adoption is not in place. Accordingly, this study reliably represents those children whose adoptions were reviewed, but cannot be representative of all of the children potentially available for adoption. 66 Measures A secondary analysis methodology was chosen for this project. Consideration of availability of information, access to case records and ethical issues relating to direct discussion with adoptive parents were factors in this decision. The personal pain surrounding the adoption disruption for both parent and child makes interviewing either party difficult. Further, the accuracy of the information provided is always subject to the interpretation placed on it by the subject. A structured interview was not possible in these circumstances given the limited time and resources available. An ethical researcher would have to ensure that adequate counselling and support services were available at the time of the research intervention in order to avoid further trauma for participants. Ministry of Social Service policy over the last decade has been to request of adoptive parents their consent to engage in follow-up research. The implication is that if adoptive parents do not consent to being contacted for research purposes, they should not be approached. This indicates that a complete sample of both completed and disrupted adoptive placements would be very difficult to achieve. Secondary analysis or document study, while recognized as a valid and useful method of research (Bailey,1978, Grinnell, 1985) has some disadvantages and advantages which must be addressed. The following application of Bailey's 67 model to discuss document study addresses the issues he raises. Advantages of document study 1) Inaccessible subjects - The adoptive parents and children are not ideal subjects for interviews or structured questionnaires. Conducting such a procedure on either the parents or the children in isolation would provide some valuable information. However, the emotional cost would potentially be high and perhaps not worth the potential , suffering. In addition, there is a problem of access to subjects. Gochros (1985) stresses the need for "evidence of the proposed respondents' willingness to be interviewed. If the sample refuses, the result is nothing but a list of non-respondents" (p. 329). 2) Nonreactivity - The data used are based on file material collected for recording and documentary purposes. The file material was not written by the subjects. The 'data collection method'(a structured instrument as described below) did not change the data collected nor did its collection influence the behaviour of the subjects. 3) Longitudinal analysis - As stated above, the study describes the disruption of adoptions from 1985 to 1989. Secondary analysis permits the examination of the data where primary source methods present difficulties in timing, cost and accessibility. This is an empirical as opposed to experimental form of research. 4) Sample size - Although in this case, the sample is relatively small due to a limited population base, the consideration of sample size is important. As indicated above, contact with adoptive parents is only possible in those cases where prior consent to contact for research purposes has been granted. Generating a sample of large enough proportion of both completed and disrupted placements would be a potential problem. 5) Spontaneity - It is assumed that the file material fairly represents the observations of the recorder at the time the events transpired. Documents are filed as required by statute. Assessments completed by third parties represent the state of knowledge available about the case at that point in time. 6) High quality - Given a standard reporting format and auditing procedures to maintain quality recording for many mandated services, one can reasonably expect that certain key documents will be available on files. Where files were available, there was no case where all needed demographic information was missing. A change in Provincial file storage policy in 1988 meant that some file material was no longer kept centrally for disrupted adoptions. While all material related to the child had previously been kept on file in the records section, when adoptions disrupted some material was returned to the district office serving the child. The principal impact of this change on the data collection method was that corollary information was not always present. This complicated data collection in some cases. However, sufficient numbers of files continued to have the necessary information for purposes of this research. Disadvantages of Document Study 1) Bias - Although the file material was not intended to be used for research purposes, the study was designed to probe for information that is known to be routinely recorded. Bias on the part of the recorder may enter assessments of families or children. An additional complication is the inherent bias in adoption home studies where applicants tend to paint the rosiest possible picture of themselves. This research project focussed on that information which was available at the time of placement, and therefore questions of bias relate principally to the interpretation of events by the recorder. In many ways, this is a central feature of the research investigation. Social workers' effectiveness is evaluated in part in their clients' case files. As with most human beings, few social workers are sufficiently honest . . . to record their failures and errors of judgement in agency records. Client files may be biased so as to present the social work practitioner in the best possible light. Certain data which result in the evaluation of the effectiveness of social workers are especially suspect. (Weinbach, 1985, p.75) 2) Selective survival/incompleteness - This was a problem encountered. Some file material was deleted due to changes in file retention policy. Other material was filed on 70 related files, such as those of the child's non-adopted sibling or parents, and not retained on the subject file. Where material was not available in sufficient amounts the case was deleted from the sample. Careful reading of all related files occasionally provided the missing information and a case was not deleted from the sample until this was done. The research instrument was constructed with some foreknowledge of the material likely to be found commonly in ministry files. Thus, the problem of data incompleteness' was contained as much as possible. 3) Lack of availability - Owing to the degree of cooperation received from the records section of the Ministry, and the permission granted to view files, the only question of availability concerned the physical whereabouts of a handful of files. These files could not be located and may have been misfiled, sent back to the originating office in error or sent out for microfilming. 4) Sampling bias - The children's names were drawn from a list of all adoptions which were not completed from 1985-1989. This list was culled by removing those cases which did not fit the disruption definition earlier established. Files were completely read prior to removal from the sample. 5) Limited to verbal behaviour - This jls a deficit in the design as no amount of file recording or assessment can adequately describe the pain and anguish that must accompany each adoption disruption. Nor can file recording adequately 71 capture the full range of verbal expression and behaviour. However, the purpose of the research was to identify those factors associated with disruption. Given that the initial hypotheses were that there were certain factors thought to associate with disruption such as length of time in care, pre-care experiences, history of abuse, examination of those issues could be accomplished through file review. 6) Lack of a standard format/coding - File recording policy and format were unchanged through the review period. The' requirements of the Adoption Act mandate that certain forms and materials be present. However, files dating back over several years occasionally held information about cases where the information presented seemed to indicate that sexual abuse was a possibility in the lives of the children. Wide scale awareness of this form of abuse did not really take hold until 1980 in B.C. and the file records may be limited in this area. Basing their methodologies on the hypothesis that the agency files held valuable data about the children, Zwimpfer (1978), Festinger (1986), Partridge et al (1986) and Barth and Berry (1989) all used secondary analysis to conduct their studies of adoption disruption. Zwimpfer identified the emotional toll personal interviews would take in her decision to apply a secondary analysis technique. Choosing the analysis of adoption files as my sole method was a simple and obvious choice for me: because I knew that the experience of an adoption breakdown is a very stressful and, partly by 72 virtue of its rarity, a guiltladen one as well, I at no time contemplated any personal interviews with the families concerned, as rich in information as it would be; the only other information obtainable therefore was in the adoption files of the Department of Social Welfare and therefore I decided to limit my research to the examination of those files". (Zwimpfer, 1978, p.19) File Review Instrument The file review instrument can be found in Appendix A. It was designed to address the central hypotheses, examining demographic variables such as age at placement, age when legally free, gender, and time spent in a number of different categories of alternate care. The hypotheses, as described above, queried the role of pre-care experiences in adoption disruption. Accordingly, the instrument reflected a range of pre-care experiences and included categories for 'other' types of experience not previously identified. The instrument was constructed to capture information known to be in case files. The discoveries of other researchers in this area were used to build the instrument and to tailor it to the working hypotheses. As mentioned above, foreknowledge of the material likely to be found in files aided in the development of the instrument. A simple coding formula was used which was designed to shorten the time required for recording the findings from each case. All ages were converted to months for simplicity of calculation and because many of the children experienced their first foster placement before their second birthday. Validity and Reliability Considerations of the validity of the data must always loom large in research. The data collection instrument was designed to ensure that the data collected were both accurate and accurately represented the child's situation as described in the file. Bostwick and Kyte (1985) define validity as having two parts: "the instrument actually 1 measures the concept in question, and the concept is measured accurately" (P.161). The concepts chosen for the research were measurable by the instrument. Replication of the data collection would yield identical results. Internal checks were incorporated into the instrument to ensure accuracy. For example, ages at certain key points were measured. These were derived from source documents such as birth registration, placement notices, court reports and file recording. In order to maintain validity, an additional check was performed to account for missing months in the child's file history. Where the information sought concerned nominal data, cases were given a positive count on the item only when file material confirmed the presence of the factor. Questions of circumstance were carefully reviewed for accuracy. If a child was reported in the file or court material to have been sexually or physically abused, this information was 74 recorded. When the file material led to a suspicion that the child may have been abused, based on behaviour and circumstances, but there was no confirmation, the abuse would not be recorded as confirmed. In this instance, a note was made querying abuse as an issue for the child. If the suspected abuse was not subsequently confirmed, the case was not included in the count of children who experienced abuse. The information gathered from the files can be reasonably considered to be reliable, given that much of1 the information was based on official, factual documents. Reliability testing was necessary in some categories. This was particularly the case with items relating to siblings. A number of different questions were posed about siblings to ensure that all possible combinations of possibilities were examined. Instrument categories were designed to avoid ambiguity and to gather the information in a manner which led to ease of interpretation. An additional check was performed to examine whether any different special needs were identified after initial adoption placement. This check examined for reports that the child disclosed previous physical abuse, sexual abuse or neglect. This was to ensure that there were no false negative reports in the earlier assessment of the reason the child entered care. In addition, the files were checked to see if the adoptive parents reported that the child was 75 exhibiting different special needs than had been identified to them at the time of placement. Data Analysis The data were analysed using the statistical package, SPSSPC+. A number of strategies were employed to draw the maximum benefit possible from the data. Complete descriptive information was generated for all variables. This included, where applicable, mean, median and modal values for interval data; frequency tables, standard deviation and variance,' count and frequency tables for all nominal data. Missing information was not included in any calculation of statistics. In addition, a review of the raw data revealed some areas of anomaly, where the information was not missing, but could not be reliably used. In these cases, the variable was recoded as missing. For interval data, a t-test analysis was used. For all nominal categories, a chi-square analysis was used. In this computation, cases with cells with an expected frequency of less than five were not considered. The chi-square statistic was taken after the Yates correction was calculated to ensure a higher level of reliability of the measure (Norusis, 1988). All variables were measured against the dependent variable of adoption completion. An additional test was performed, controlling for children who were members of sibling groups and controlling for gender. Where the 76 variables were measured at the interval level, an analysis of variance method was used to test for significance about the differences between the means. For variables measured at the nominal level, the chi square analysis was used. Where the chi-square statistic was not valid for any of these combinations, the phi statistic was examined to consider the strength of the relationship. In addition, the lambda score was reviewed for an indication of the degree of association. A test for co-variability was not performed. This limits the study to description of the principal variables. A discriminant analysis to test for the effect of each of the nominal variables on the others would be a very useful follow-up study to this research. There were a number of variables about which information was gathered but will not be reported in the following chapter. In general, these variables were not considered to represent reliably information which was relevant to the study. For example, all of the children's files contained a 'disability' rating. This rating is highly subjective on the part of the social worker and cannot be considered a valid measure of the child's emotional, behavioural or physical disabilities. In any event, this information was captured in a more consistent manner through other variables. Other variables are not reported due to the very small number of cases in the sample. For example, children's 77 placements in institutions or residential treatment centres. The numbers in these cases were very small and statistically not significant. Among the post-placement variables, there are a number of areas not reported. Among these are the number of months between home study and placement and child placed as requested. In the former case, the use of this variable would have been inconsistent with the position taken above that the focus of the study was on the children. In the latter case, the information was interesting to > gather but would have required a great deal more development to become a truly useful piece of information. Summary The data collection instrument and the data collection process were carefully matched to capture as much of the relevant material from the files as possible. This chapter has identified the process, the subjects and the apparatus and has considered questions related to this areas. Subsequent chapters will illustrate the findings of the data collection and a discussion of their significance. 78 RESULTS This chapter presents the results of the data analysis. It is presented in the same format and order as the hypotheses. To begin with, however, it is important to present a picture of the sample so as to provide an overview of the children. The total sample size was 82. As indicated in Table 1, there were equal numbers of boys and girls in the sample. The sample was composed of 60% non-caucasian children. Most of these children were Native Indian. However, there was no correlation between race and completion. INSERT TABLE 1 HERE There were 44 children (54%) who were members of a sibling group. As indicated below, there were several dimensions to this variable that were shown to be significant. Numbers of siblings ranged from a low of 1 to a high of 4. Many of the children had siblings who were not placed for adoption with them. These were adult siblings, children still living with their biological parents, children in other adoption homes or children in other care arrangements. There were 52 children (63.4%) in this category. 79 Table 3 Characteristics of disrupted (n = 41) and completed (n = 41) Characteristic Gender of child Caucasian Siblings Age when free for adoption " Age when placed " Months wittj biological family " Child's wait (months)b Completed Male (44%) No (66%) Yes (61%) 4.68 6.36 38.44 25.75 Disrupted Male (56%) No (53%) Yes (42%) 6.84 8.15 57.9 20.11 X2 or t .7805 .3677 3.972* -2.84* -2.11* I - 2 . 2 6 * 1.06 " Mean values provided " Established by subtracting age when legally free from age at placement. * 2 <-05 80 The children's ages at various points in the adoption process are described below. Following Barth and Berry's (1988) example, a calculation of the time waiting for placement following termination of parental rights was undertaken. The result was not significant but indicated that, on average, the children in the completed group waited 25 months for placement. The children in the disrupted group waited an average of 20 months. Hypothesis # 1: There will be no significant difference in outcome between boys and girls. There were 82 cases in the sample. Of these, 41 were boys and 41 girls. Three children were counted twice. That is, two siblings experienced two disrupted adoption placements. A third child experienced an adoption disruption, but was later placed in another adoption home where the adoption was completed. Of the total population, 18 of the boys experienced an adoption disruption and 23 of the girls' adoptions disrupted. While there were more boys than girls in the disrupted group (n = 23, 56%), knowledge of the gender alone of the child reduces by only 12% the prediction of eventual completion or disruption of the adoption (lambda = .12195 dfl, n = 82). 81 The chi square analysis of adoption outcome by gender of the child alone, revealed no significance, chi-square (1, n = 82) = .780 p >.05.) All of the results described below are differentiated on the basis of gender so as to provide additional clarification on this point. Hypothesis # 2: Children placed with siblings will be likely to experience fewer disruptions. There are a number of variables to consider with respect to the presence and/or absence of siblings for the children in the sample. Just over one half of the children in the sample were members of a sibling group at the time they were available for adoption placement (n = 44). The numbers of siblings available for adoption at the same time as the subject child ranged from a minimum of 1 to a maximum of 4. Of the children with siblings available for placement, 17 (20.7%) experienced an adoption disruption. Of the remaining 38 children without siblings at the time of adoption placement, experienced 24 disrupted placements. There was a significant correlation between the availability of siblings at the time of placement and completion of the adoption, chi-square (l,n = 82) = 3.972, p <.05). Not all of the children in the sample with siblings were with all their siblings at the time of placement. For 82 those children who had siblings elsewhere, and were members of a sibling group placed for adoption, there was a significant positive correlation with adoption outcome, chi-square (1, n = 44) = 7.724, £ =-0054. The correlation was also strong for children who were members of a sibling group and placed in the same home as their siblings. For these children there was a correlation between completion of the adoption and the presence of siblings, chi-square (1, n = 44) = 5.714, £ =.0168. Other i significant findings with respect to siblings will be addressed below as they are relevant to each section. Hypothesis # 3: Children who are able to maintain contact with siblings are more likely to experience fewer disruptions. There was insufficient information available on the files to extract adequate data to test this hypothesis. 83 Hypothesis # 4: The disrupted group will be marked by children who are older at the age when legally free and older at the time of placement. Age when legally free The mean age when legally free for adoption for the entire sample was 5.76 years (69.183 months). Mean age for all boys was 6.16 years (73.926 months), with a range from 5 months to 146 months. The mean age for all girls was 5.37 years (64.439 months, with a range from 7 months to 143 months. INSERT TABLE 2 HERE For children whose adoptions were not completed, the mean age overall was 6.84 years (82.098 months). The mean age for boys (n = 23) was 6.63 years (79.562 months), while the mean age for girls (n = 18) was 7.11 years (85.333 months). For the completed adoption group, the mean age overall was 4.68 years (56.268 months). The mean age when legally free for boys (n = 18) was 5.56 years (66.722 months), while the mean age for girls (n = 23) was 4 years (48.087 months). Table 2 Age When Legally Free (years) by Gender Overall Disrupted Completed Mean age S.P. Mean Age S.P. Mean Age S.P Male 6.16 3.67 6.63 4.12 5.56 3.02 n 41 23 18 % 50 28 22 Female 5.37 3.46 7.11 3.35 4.0 2.95 n 41 18 23 % 50 22 28 Overall 5.76 3.57 6.84 3.76* 4.68 3.05* n 82 41 41 % 100 50 50 * 2 = -0057 85 The age at which the child became legally free was significant. A t-test analysis indicated a significant correlation between age when legally free and adoption completion.(f (80) = 1 . 5 2 , p < .05). However, the analysis of variance indicated that there was no significant correlation between the gender of the child, age when legally free and adoption completion. For the children in sibling groups (n = 44), the mean age when they became legally free for adoption was 5.34 years. For the group of children who had siblings available for adoption at the time of placement, 17 (21%) experienced a disruption. The remaining 27 children (33%) had a mean age of 4.53 years. INSERT TABLE 3 HERE The children who did not have siblings available for adoption at the time of placement (n = 38) had a mean age of 6.25 years when they were legally free for adoption. Of this group of children (characterized as 'alone' in Table 3), 24 (29%) experienced a disruption. The disrupted group of children without siblings available at the time of placement had a mean age of 7 years when they were legally free for adoption. The remaining 14 children (17%) whose adoptions were completed had a mean age of 4.98 years when they were legally free for adoption. 86 Table 5 Age when Legally Free (years) by Membership in a Sibling Group Overall Disrupted Completed Mean age S.D. Mean Age S.D. Mean Age S.D Sibling 5.34 2.96 6.61 3.15 4.53 2.59 n 44 17 27 % 54 21 33 Alone 6.25 4.16 7.00 4.20 4.98 3.89 n 38 24 14 % 46 29 17 Overall 5.76 3.57 6.84 3.76 4.68 3.05 n 82 41 41 % 100 50 50 87 An analysis of variance procedure revealed that there was no significant correlation between the age at which the child became legally free for adoption, adoption completion and membership in a sibling group. Age at placement For the entire sample (n = 82) the mean age at placement in the adoption home was 7.26 years (87.134 months). For all boys, the mean age at placement was 7.81 years (93.707 months), with a range of 18 months to 149 months. The mean / age for all girls was 6.71 years with a range of 7 months to 182 months (80.561 months). INSERT TABLE 4 HERE For the disrupted group (n =41) the mean age overall at time of placement was 8.15 years (97.853 months). The mean age for boys (n = 23) was 7.74 years (92.913 months), while the mean age for girls (n = 18) was 8.68 years (104.166 months). In the completed group (n = 41) the mean age overall was 6.37 years (76.415 months). The mean age for boys was 7.89 years (94.722 months), while the mean age for girls was 5.17 years (62.087 months). 88 Table 5 Age at Placement (years) by Gender Overall Disrupted Completed Mean age S.P. Mean Age S.P. Mean Age S.P Male 7.8 3.90 7.74 4.03 7.89 3.83 n 41 23 18 % 50 28 22 Female 6.71 3.89 8.68 3.28 5.17 3.68 n 41 18 23 % 50 22 28 Overall 7.26 3.90 8.15 3.70* 6.36 3.95* n 82 41 41 % 100 50 50 * 2 <•05 89 While the mean age at placement was not considerably higher for boys in the completed group, the girls' mean age was different by over three years. A t-test analysis of the total sample indicated that the age at placement was a significant factor in adoption completion, f (80, n = 82) = 1.13, p < .05. However, an analysis of variance procedure indicated no significant correlation between age at placement, child's gender and adoption completion. For children who were members of a sibling group at the time they were available for adoption placement, the mean age at placement was 7.23 years (n = 44). Of these children, 17 (21%) experienced an adoption disruption. This group had a mean age of 8.16 years. INSERT TABLE 5 HERE The children who were not members of a sibling group, identified as 'alone' in Table 6, (n = 38) had a mean age at the time of placement of 7.31 years. Of these children, 24 (29%) experienced an adoption disruption. The disrupted 'alone' group had a mean age of 8.15 years. The remaining 14 (17%) children in the 'alone' group whose adoptions were completed, had a mean age of 5.87 years. An analysis of variance indicated that there was no significant correlation between the age at placement, membership in a sibling group and adoption completion. 90 Table 5 Age at Placement (years) by Membership in a Sibling Group Overall Disrupted Completed Mean age S.D. Mean Age S.D. Mean Age S.D Sibling 7.23 3.45 8.16 2.91 6.63 3.68 n 44 17 27 % 54 21 33 Alone 7.31 4.43 8.15 4.24 5.87 4.53 n 38 24 14 % 46 29 17 Overall 7.26 3.91 8.15 3.70 6.37 3.95 n 82 41 41 % 100 50 50 91 Hypothesis # 5: The length of time that a child spends with his/her family of origin will inversely affect outcome. The mean time with biological family for the entire sample (n = 78) was 4 years (48.423 months). The mean time for all boys (n = 39) was 4.72 years (56.743 months) with a range of 0 to 138 months. The mean time for all girls (n = 39) was 3.34 years (40.102 months) with a range of 1 month to 137 months. INSERT TABLE 6 HERE For the disrupted group (n = 40) the mean time overall with biological family was 4.83 years (57.9 months). The mean time with biological family for boys (n = 22) was 5.13 years (69.591 months) and the mean time for girls was 4.44 years (53.389 months). For the completed group (n = 38) the mean time overall with biological family was 3.20 years (38.447 months). The mean time with biological family for boys (n =17) was 4.20 years (50.470 months), and the mean time for girls (n = 21) was 2.39 years (28.714 months). Table 6 Years Spent Living With Biological Family by Gender Overall Disrupted Completed Mean S.D. Mean S.D. Mean S.D Male 4.73 3.32* 5.13 3.75 4.21 2.68 n 39 22 17 50 28 22 g. Female 3.34 3.08* 4.45 3.18 2.39 2.72 n 39 18 21 % 50 23 27 Overall 4.03 3.26 4.83 3.48** 3.20 2.82** n 78 40 38 % 100 51 49 * £ =.0595 ** £ <.05 93 There was a significant correlation between adoption completion and months lived with biological family, f (74.18, n = 80) = 1.52, £ <.05. The analysis of variance indicated that there was a moderate correlation between gender, months lived with biological family and adoption completion, f (1, n = 82) = 3.659 2 = -0595. For children who were members of a sibling group at the time they were available for adoption placement, the mean years spent living with their biological families was 3.56 i (n = 43). For the sibling group members whose adoptions disrupted (n = 17), the mean years spent with biological family was 4.53. The remaining 27 children who were members of a sibling group, whose adoptions were completed, lived with their biological families for 2.91 years. INSERT TABLE 7 HERE For those children who were not members of a sibling group at the time they were available for adoption placement (n = 35), the mean number of years spent living with biological family was 4.63. Children who were not members of a sibling group, whose adoption disrupted (n = 24), lived with their biological families 5.04 years. The remaining 14 children in the 'alone' group, whose adoptions were completed, lived with their biological families for 3.85 years. Table 6 Years Spent Living with Biological Family by Membership in a Sibling Group Overall Disrupted Completed Mean S.D. Mean S.D. Mean S.D Sibling 3.56 2.68 4.53 2.91 2.91 2.36* n 43 17 27 % 55 21 33 Alone 4.63 3.80 5.04 3.90* 3.85 3.65 n 35 24 14 % 45 29 17 Overall 4.03 3.26 4.83 3.48* 3.20 2.82* n 78 40 38 % 100 51 49 * p <.05 95 An analysis of variance procedure indicated a significant correlation between membership in a sibling group, years spent living with biological family and adoption completion, f (35.37, n = 49), = 2.71, 2 <.05. Returns to Biological Family Many of the children lived with their biological families for several, separate extended periods of time before they eventually entered permanent care (n = 59). I These children were removed from their parents' care on different occasions and returned to the parents prior to the apprehension that led to the making of a permanent order of guardianship. While the mean number of returns to live with biological family overall, was 1.8 times, the maximum number of returns was 5 (n = 59). For the disrupted group, the mean number of returns to the biological family was 1.59. For the completed group, the mean number of returns was 2.07. For all the boys in the sample (n = 30), the average number of returns to the biological family was 1.9 with a range from 1 to 4 returns. For the boys in the disrupted group, the average number of returns was 1.7 (n = 17). For boys in the completed group (n = 13), the average number of returns to biological family was 2.1. 96 For all the girls in the sample (n = 29), the average number of returns to the biological family was 1.8 with a range from 1 to 5 returns. For the girls in the disrupted group (n = 15), the average number of returns was 1.5. For girls in the completed group (n = 14), the average number of returns was 2.1. An analysis of variance procedure indicated no correlation between number of returns to biological family, child's gender and adoption completion, f (l,n = 59),p >.05. For children who were members of a sibling group at the time they were available for adoption (n = 39) the mean number of returns was 1.97. For the 'alone' group, the mean number of returns was 1.5. There was no significant association between sibling group membership, number of returns to the biological family and adoption completion. There was no significant finding with respect to the number of returns to biological parents, adoption completion and any of the pre-care experiences the children may have weathered. 97 Hypothesis # 6; Children placed with relatives or with long-term foster parents will not experience disruption at the same rate as those placed with strangers. Of the 82 children studied, 19 were adopted by their existing foster parents or by relatives. Of these 19 placements, 3 concluded with a disruption. Of the 19 placements with known adoptive parents, 2 of the 3 1 placements which disrupted were with relatives, while 3 of the completed adoptions were with relatives. A calculation of the mean time spent living with the adoptive families prior to adoption was not possible due to limitations imposed by file material. There is a high degree of correlation between placement with known adoptive parents and adoption completion, t (80, n = 82) = 3.54 £ <.001. 98 Hypothesis # 7: The number of placements and total length of time spent in foster homes or group homes prior to placement will inversely affect outcome. Number of placements The mean number of foster home or group home placements experienced by the sample (n = 78) was 4.66. The minimum number was 1 and the maximum, 11. The children in the disrupted group had a mean number of placements of 4.7 with a range from 1 to 11 placements. For the completed group, the mean number of placements was 4.5 with a range from 1 to 11 placements.The mean number for boys (n = 39) was 4.38 with a range from 1 to 9 placements. The mean number of placements for girls (n = 39) was 4.94 with a range from 1 to 10 placements. INSERT TABLE 8 HERE For the boys in the disrupted group (n =21), the mean number of placements in a foster home or group home was 4.19. For the girls in the disrupted group (n = 16), the mean number of placements was 5.56. 99 Table 8 Number of Foster and Group Home Placements by Gender Overall Disrupted Completed Mean S.D. Mean S.D. Mean S.D Male 4.38 2.46 4.19 2.31 4.61 2.68 n 39 21 18 % 50 27 23 Female 4.95 2.78 5.56 2.36 4.52 3.01 n 39 16 23 % 50 20 30 i Overall 4.67 2.62 4.78 2.40 4.56 2.83 n 78 37 41 % 100 47 53 * £ <.05 100 For the boys in the completed group (n = 18), the mean number of placements was 4.6, while for the girls (n = 23) the mean was 4.52 placements. An analysis of variance procedure indicated that there was no significant difference in outcome between boys and girls based on numbers of placements in foster homes or group homes. For children who were members of a sibling group at the time they were available for adoption placement, the mean number of placements was 5.6 (n = 42). For sibling group members whose adoptions were completed, the mean number of foster or group home placements was 5.5 (n = 27). For sibling group members whose adoptions disrupted, the mean number of foster or group home placements was 5.73 (n = 15). INSERT TABLE 9 HERE For children who were not members of a sibling group at the time they were available for adoption placement ('alone') the mean number of placements in foster or group homes was 3.6 (n = 36). For members of the 'alone' group whose adoptions were completed, the mean number of foster or 101 group home placements was 2.6 (n = 14). For members of the 'alone' group whose adoptions disrupted, the mean number of foster or group home placements was 4.13 (n = 22). An analysis of variance procedure indicated that there was no significant difference in outcome between the children who were members of a sibling group and those who were 'alone' based on the number of foster or group home placements, f (1, n = 78) = .1384, £ >.05. 102 Table 9 Number of Foster and Group Home Placements by Membership in a Sibling Group Overall Disrupted Completed Mean S.D. Mean S.D. Mean S.D Sibling 5.62 2.69 5.73 2.74 5.56 2.71 n 42 15 27 % 54 19 35 Alone 3.56 2.09 4.14 1.96 2.64 2.02 n 36 22 14 % 46 28 18 Overall 4.67 2.63 4.78 2.41 4.56 2.84 n 78 37 41 % 100 47 53 103 However, an analysis examining the relationship between gender, number of foster or group home placements and adoption outcome, controlling for sibling membership, revealed a significant correlation for gender, f (1, n = 78) = 13.978, p <.001. In this case, the mean number of placements was similar for both boys and girls in the group of siblings available for adoption at the same time - a mean of 5.61 placements and little variation across gender or completion. For the 'alone'group, the mean was 3.56 placements. The boys in this group whose adoptions were completed had a mean of 3.4 placements, while the girls whose adoptions were completed had a mean number of placements of 1.86. The 'alone' boys whose adoptions were not completed had a mean of 3.6 placements, while the 'alone' girls whose adoptions were not completed had a mean of 5 placements. Time in foster homes and group homes. The mean number of months spent living in either a foster home or a group home by all of the children in the sample (n = 82) was 44.76 months. However the range shows that the minimum number of months was 1 while the maximum was 140 months. For the children whose adoptions disrupted, the mean amount number of months spent living in foster or 104 group homes was 39.70. The range was from 1 month to 119 months. For the children whose adoptions were completed, the mean number of months was 49.82 with a range from 4 months to 141 months. INSERT TABLE 10 HERE For boys overall, the mean number of months spent in either a foster home or a group home was 43.24, with a minimum of 1 month and a maximum of 140 months. For girls the figure was 46.29, with a minimum of 4 months and a maximum of 135 months. For the boys in the disrupted group, the mean time spent in foster or group home placements was 33.74 months, while for the girls the figure was 47.33 months. For the boys in the completed group, the mean time spent in foster or group home placements was 55.39 months, while for the girls the figure was 45.48 months. The analysis of variance procedure indicated that the number of months spent in foster or group home placements did not affect the outcome of the adoption when the sample was split by gender. 105 Table 10 Months In Foster and Group Home Placements by Gender Overall Disrupted Completed Mean S.P. Mean S.P. Mean S.P Male 43.24 32.2 33.74 26.7 55.38 35.3 n 41 23 18 % 50 28 22 Female 46.30 31.8 47.33 31.7 45.48 32.6 n 41 18 23 % 50 22 28 Overall 44.76 31.9 39.71 29.4 49.83 33.7 n 82 41 41 % 100 50 50 106 For children who were members of a sibling group at the time they were available for adoption, the mean number of months spent in foster or group home placements was 48.7. The minimum number of months was 1 with a maximum number of 141. Sibling group members whose adoptions disrupted spent an average of 35.2 months in foster or group homes. Those whose adoptions were completed spent an average of 57.3 months in foster or group homes. INSERT TABLE 11 HERE The children in the 'alone'group spent an average of 40.2 months in such placements with a range from 10 months to 119 months. Children in the alone group whose adoptions disrupted spent an average of 42.9 months in foster or group home placements. Those whose adoptions were completed spent an average of 20.2 months in foster or group home placements. 107 Table 11 Months In Foster and Group Home Placements by Membership in a Sibling Group Overall Disrupted Completed Mean S.D. Mean S.D. Mean £LD Sibling 48.7 34.8 35.2 26.3 57.3 37.1 n 44 17 27 % 54 21 33 Alone 40.2 27.9 42.9 31.6 35.5 20.2 n 38 24 14 % 46 29 17 Overall 44.7 31.9 39.7 29.4 49.8 33 7 n 82 41 41 % 100 50 50 * £ <.05 108 Hypothesis # 8: The pre-care experiences of children will affect adoption outcome. Those children with histories of physical and sexual abuse will fare poorly in adoption. Reason for entering care (A) Abandoned (n = 81) Fifteen of the sample children entered care as a result of abandonment. The adoptions of 6 of these children were disrupted, chi-square (1, n = 81) = .269, p >.05. There was no correlation between abandonment and adoption disruption where the child was not a member of a sibling group. INSERT TABLE 12 HERE Children who were members of a sibling group available for adoption at the time of placement (n = 44) who entered care as the result of abandonment (n = 8) experienced no disruptions. Children who were in the 'alone' group (n = 37) and who were entered care due to abandonment (n = 7) experienced 6 disrupted adoptions. 109 The chi-square analysis of abandonment and adoption outcome, controlling for sibling membership, indicated a significant correlation, chi-square (1, n = 44) = 4.325, 2 <.05\ (B) Physically abused (n = 80) Thirty of the sample children entered care as a result of having been physically abused. The adoptions of 10 of these children disrupted. A chi-square analysis indicated an association between physical abuse and adoption completion, chi-square (l,n = 80) = 3.632, £ = .056. Of the group of children who entered care as a result of having been physically abused, 13 were boys and 17 were girls. Disruptions occurred in 3 of the boys' placements and in 7 of the girls' placements. A chi-square analysis of physical abuse as a reason for entering care with adoption outcome, controlling for gender revealed no significance for the girls, chi-square (1, n = 40) = .0000, 2 >-05. However, a significant correlation between physical abuse as a reason for entering care and adoption completion existed for boys, chi-square (1, n = 40) = 6.134, £ = .0133. 110 Table 12 Reason for entering care by Disrupted and Completed Groups, Controlling for Gender and Membership in a Sibling Group. Reason for Disrupted % Completed % X 2 entering care Abandoned 7.4 11.1 .269 Boys 9.8 14.6 .6613 Girls 5.0 7.5 .0000 Sibling 0 18.2 4.3257* Alone 16.2 2.7 .9883 Physically 12.3 CM 3.63* Abused Boys 7.5 25.0 6.134* Girls 17.5 25.0 .0000 Sibling 4.7 37.2 7.2069* Alone 21,6 10.8 .0008 Sexually 11.1 13.6 .0166 Abused Boys 5.0 17.5 3.4770" Girls 17.5 10.0 1.7089 Sibling 11.6 16.3 .0006 Alone 10.8 10.8 .1516 Emotionally 7.5 11.3 .2168 Abused Boys 5.0 17.5 3.4770" Girls 10.0 5.0 .7241 Sibling 2.3 14.0 .8911 Alone 13.5 8.1 .0000 Parental Alcohol 23.8 40.0 6.2257* Abuse Boys 25.0 30.0 1.0447 Girls 22.5 50.0 4.0949* Sibling 20.9 53.5 3.0291 Alone 27.0 24.3 .7903 * 2 <.05 ** p <.005 *** p <.001 * 2 = .0622 b £ = .0073 Table 12 (continued) Reason for Disrupted % Completed % X 2 entering care Parental Drug 13.8 7.5 1.4635 Abuse Boys 15.0 7.5 .1752 Girls 12.5 7.5 .7736 Sibling 14.0 7.0 2.7832 Alone 13.5 8.1 .0000 Relinquished 6.3 0 3.6327 Boys 12.5 0 2.8282 Girls 0 0 -Sibling 4.7 0 1.2821 Alone 8.1 0 .6221 Parents Deceased 3.8 0 1.4920 Boys 2.5 0 .0000 Girls 2.0 0 .9099 Sibling 4.7 0 1.2821 Alone 2.7 0 .0000 Chronic Neglect 25.6 41.0 4.0169 Boys 23.1 33.3 2.3096 Girls 28.2 48.7 .3894 Sibling 22.0 53.7 .6928 Alone 29.7 27.0 1.6396 Parent Mentally 111 7.4 8.6 .0000 Boys 7.3 9.8 .1274 Girls 7.5 7.5 .2222 Sibling 9.1 4.5 1.1368 Alone 5.4 13.5 2.5674 Other Reason 2.5 20.3 10.9301 Boys 5.1 12.8 1.1285 Girls 0 27.5 8.9438 Sibling 2.4 19.0 1.8101 Alone 2.7 21.6 10.4652 * £ <.05 ** 2 <.005 *** 2 <-001 " 2 = -0567 112 Children who were members of a sibling group (n = 44), who entered care due to physical abuse (n = 18) experienced 2 disruptions. The chi-square analysis indicated that this was a significant finding, chi-square (1, n = 43) = 7.206, £ = .0073. Children in the 'alone' group who entered care due to physical abuse (n = 12) experienced 8 disruptions. There was no significant correlation between physical abuse and adoption disruption where the child was not a member of a sibling group. (C) Sexually abused (n = 80) Twenty of the sample children were sexually abused. Of these, 9 experienced an adoption disruption. There was no correlation between sexual abuse and disruption, chi-square (1, n = 80) = .0166, p > .05. Of the group of children who entered care as a result of having been sexually abused, 9 were boys and 11 were girls. Disruptions occurred in 2 of the boys' placements and in 7 of the girls' placements. The chi-square analysis of adoption outcome for children who entered care as a result of sexual abuse, controlling for gender, indicated no significance for girls, chi-square (1, n = 40) = 1.708, £ >.05. However for boys, the chi-square analysis indicated a tendency for sexual 113 abuse as a reason for entering care to influence adoption outcome, chi-square (1, n = 40) = 3.477, p = .0622. Of the group of children who entered care as a result of being sexually abused, 12 were members of a sibling group. The adoptions of 5 of these children disrupted. There were 8 children in the 'alone' group who entered care as a result of having been sexually abused. The adoptions of 4 of them disrupted. The chi-square analysis of adoption outcome with children who entered care as a result of being sexually abused, controlling for gender indicated that there was no significant correlation (See Table 12). (D) Emotionally abused (n = 80) Fifteen of the children entered care due to emotional abuse. The placements of 6 of these children disrupted. The chi-square analysis indicated that there was no correlation between emotional abuse and adoption disruption, chi-square (1, n = 80) = .2168, p >.05. Of the group of children who entered care due to emotional abuse, 9 were boys and 6 were girls. Disrupted adoptions were experienced by 2 of the boys and 4 of the girls. The chi-square analysis of adoption completion with emotional abuse, controlling for gender, indicated a 114 tendency toward correlation with disruption for the boys, chi-square (1, n = 40) = 3.477, £ = .0622". The chi-square analysis indicated no correlation for the girls (See Table 12). Of the children who entered care as a result of emotional abuse, 7 were members of a sibling group at the time they were available for adoption placement. Of these children, 1 experienced an adoption disruption. There were 8 children in the 'alone' group who entered care as a result of emotional abuse. The adoptions of 5 children disrupted. The chi-square analysis of adoption outcome with emotional abuse, controlling for sibling membership, indicated that there was no correlation (See Table 12). (E) Parental alcohol abuse (n =80) The parents of 51 of the children were reported to abuse alcohol. The adoptions of 19 of those children whose parents were alcoholics disrupted. The chi-square analysis indicated that there was a correlation between alcohol abuse and adoption disruption, in that the adoptions of 62% of the children in the sample whose parents abused alcohol were completed compared to 31% of the children whose parents did not abuse alcohol, chi-square (1, n =.80) = 6.225, p <.05. 115 Of the children who entered care due to parental alcohol abuse, 22 were boys and 29 girls. The adoptions of 10 of the boys and 9 of the girls whose parents were alcoholics, disrupted. The chi-square analysis of adoption completion with parental alcohol abuse, controlling for gender, indicated no correlation for the boys (See Table 12). However, for the girls, the chi-square analysis of adoption completion with parental alcohol abuse, controlling for gender, indicated a significant correlation, chi-square (1, n = 40), = 4.094, p <.05.9 Children who were members of a sibling group at the time they were available for adoption whose parents abused alcohol (n = 32) experienced 9 disrupted adoptions. The chi-square analysis of adoption completion with parental alcohol abuse, controlling for membership in a sibling group, indicated no correlation (See Table 12). There were 19 children who were in the 'alone' group who entered care due to parental alcohol abuse. The adoptions of 10 of them disrupted. The chi-square analysis of adoption completion with parental alcohol abuse, controlling for membership in a sibling group indicated no correlation (See Table 12). 116 (F) Parental drug abuse (n = 80) The parents of 17 of the children in the sample were reported to abuse drugs. Of these, the adoptions of 11 children disrupted. There was no correlation between parental drug abuse and adoption disruption (See Table 12). The parents of 9 of the boys and 8 of the girls were reported to abuse drugs. The adoptions of 6 of the boys and 5 of the girls disrupted. The chi-square analysis of adoption completion with parental drug abuse, controlling for gender, indicated no correlation (See Table 12). The parents of 9 of the children who were members of a sibling group at the time they were available for adoption, were reported to abuse drugs. The chi-square analysis of adoption completion with parental drug abuse, controlling for membership in a sibling group indicated no correlation (See Table 12). (G) Relinquished by biological parent (n = 80) There were 5 children in the sample who were relinquished by their biological parents. All of these adoptions disrupted. There was a moderate relationship between disruption and relinquishment by biological parents, chi-square (l,n = 80) = 3.632, p = .0S6. 117 All 5 of the children relinquished by their parents were boys. Although this finding is interesting, the chi-square analysis of adoption completion with relinquishment, controlling for gender indicated no significance. There were two children (brothers) who were members of a sibling group at the time they were available for adoption, relinquished by biological parents, whose adoptions disrupted. The chi-square analysis of adoption completion by relinquishment, controlling for membership in a sibling group, indicated no correlation (See Table 12). (H) Parent deceased (n = 80) There were 3 children who entered care as their parents were deceased. None of these adoptions was completed. The chi-square analysis of adoption completion with parents deceased indicated no correlation (See Table 12). There were 1 boy and 2 girls who entered care due to the death of their parents. The chi-square analysis of adoption completion by parents deceased indicated no correlation (See Table 12). There were 2 children who were members of a sibling group at the time they were available for adoption placement who entered care due to the death of their parents. The chi-square analysis of adoption completion with parents 118 deceased, controlling for membership in a sibling group indicated no correlation (See Table 12). (I) Chronic neglect (n = 78) There were 52 children who entered care due to chronic neglect. Of these, the adoptions of 20 were disrupted. The chi-square analysis of adoption completion with chronic neglect indicates a significant relationship, chi-square (1, n = 78) = 4.016, p <.05. There were 22 boys who entered care due to chronic neglect, the adoptions of 9 of these children disrupted. There were 30 girls who entered care due to chronic neglect. The adoptions of 11 of these children disrupted. The chi-square analysis of adoption completion with chronic neglect, controlling for gender, indicted no correlation (See Table 12) . There were 31 children who were members of a sibling group at the time they were available for adoption placement who entered care due to chronic neglect. The adoptions of 9 of these children disrupted. The chi-square analysis of adoption completion with chronic neglect, controlling for membership in a sibling group, indicated no correlation (See Table 12). 119 (J) Parent mentally ill (n = 81) There were 13 children who entered care due to their parents' mental illness. Of these, 6 experienced an adoption disruption. There was no significant finding for the group as a whole nor for the sub-groups by gender and sibling group membership (See Table 12). (K) Parent mentally handicapped (n = 80) There were 3 children who entered care due to their parents' mental handicap. Of these children, 1 experienced an adoption disruption. There was no significant finding with respect to this variable for the group as a whole nor for the sub-groups by gender and sibling group membership. (L) Other reason for entering care (n = 79) There were a number of other reasons for children entering care. Among these were: "parents' crisis oriented lifestyle", mentally handicapped parents' history of drinking and violence, death of mother and effect of this on father, mother refusing to protect child from abusive boyfriend or common-law husband, adoptive parents' inability to provide care (where adoption was previously completed), history of family conflict, lack of proper supervision/care, parents' transience, murder by parent of another child. 120 There were 18 children who entered care for 'other reasons'. Of these, 16 adoptions were completed. The chi-square analysis of adoption completion with other reason for entering care indicates that there is a very significant correlation between this variable and adoption completion, chi-square (1, n = 79) = 10.93, £ <.001. There were 7 boys who entered care due to 'other reasons'. The adoptions of 2 of them disrupted. A chi-square analysis of adoption completion by 'other reason', controlling for gender, indicated no correlation (See Table 12). There were 11 girls who entered care for 'other reasons'. None of these adoptions disrupted. The chi-square analysis of adoption completion with 'other reason', controlling for sibling membership, indicated a significant relationship, chi-square (l,n = 40) = 8.943, p <.005. Hypothesis # 9: Children with multiple special needs will experience a greater number of disruptions. All of the children in the sample were, by definition, special needs children. All were over two years of age so this variable was not considered critical. The group was differentiated by a variety of difficulties. The category, 121 other special need, identified children with the following types of special needs: behavioural difficulties ( n = 13), history of multiple abuse (n =10), special need of sibling (n = 1), family history of mental illness (n = 2), sexual acting out (n = 3), transiency of parents (n = 5) and speech and language delays (n = 2). INSERT TABLE 13 HERE 122 Table 14 (continued) Post-placement Factors by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group Special Need Non-caucasian Boys Girls Sibling Alone Uncertain prognosis Boys Girls Sibling Alone Neonatal Abstinence Syndrome Boys Girls Sibling Alone Fetal Alcohol Syndrome Boys Girls Sibling Alone Born Drug Dependent Boys Girls Sibling Alone Disrupted 26.8 24.4 29.3 18.2 36.8 1.2 2.4 2.6 1.2 2.4 2.4 4.9 5.3 Completed 32.9 31.7 34.1 45.5 18.4 4.9 4.9 4.9 10.5 1.2 2.4 2.6 1.2 2.4 2.6 1.2 2.4 2.6 >C' .8113 2.3209 .0031 2.2261 .0256 .8519 .0488 .3050 2.7204 .0000 .0000 .0000 .0764 .0000 .0000 .0764 .0000 .3050 .0000 .0000 123 Table 14 (continued) \ Post-placement Factors by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group Special Need Disrupted Specific Medical 7.4 Problem Boys 12.2 Girls 2.5 Sibling 4.5 Alone 10.8 Multiply Handicapped 1.2 Boys 2.4 Girls Sibling Alone 2.6 Other Special Need 14.6 Boys 9.8 Girls 19.5 Sibling 15.9 Alone 13.2 Completed % 4.9 2.4 7.5 10.8 1.2 2.4 2.6 23.2 29.3 17.1 29.5 15.8 2C" .1440 1.0196 .0454 1.1685 .1516 .0000 .0000 .0000 1.8671 8.3367** .3570 .0199 1.1519 ** p <.005 124 There was no single defined special need which stood out as significant in terms of effect or number (with the exception of the sibling group category which is discussed in some detail above). Some children presented with more than one special need at the time of adoption placement. The classification was checked for each 'special need' the child exhibited. When the cross-tabulation controls for whether the child is a member of a sibling group, no significance was noted for any of the special needs listed. There was no significance attached to the gender of the child with respect to any of the categories of special need identified with the exception of boys who were identified as having 'other special needs'. These were most commonly identified as emotional/behavioural problems. Post-placement issues The analysis revealed that there were few areas of significance among the variables examined. Table 14 indicates the results of this secondary check of the file material. As can be seen, the examination of whether the child revealed abuse or neglect prior to placement, after adoption was not a significant factor. However, none of the cases where the child revealed emotional abuse, neglect or 125 sexual abuse were completed. As the numbers of these cases were very small, statistical calculation was not performed. The 'differences' the child exhibited compared to the picture presented to the adoptive parents was significant in only one area, that of emotional behaviour. Although the chi-square analysis is not absolute in this case, it is noteworthy that all five of the cases where the file material reported that the child was emotionally different were cases which disrupted prior to completion. INSERT TABLE 14 HERE 126 Table 14 (continued) Post-placement Factors by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group Factor Disrupted Child placed as requested 32.0 Boys 34.3 Girls 30.0 Sibling 22.5 Alone 42.9 Child requires treatment 23.7 Boys 23.7 Girls 23.7 Sibling 22.5 Alone 25.0 Local treatment available 42.9 Boys 41.2 Girls 44.4 Sibling 44.4 Alone 41.2 Child medically different 2.5 Boys 5.1 Girls Sibling 2.4 Alone 2.6 Child emotionally different 6.3 Boys 5.1 Girls 7.5 Sibling 4.8 Alone 8.1 Completed 46.7 37.1 55.0 55.0 37.1 21.1 18.4 23.7 17.5 25.0 42.9 41.2 44.4 38.9 47.2 2.5 2.6 2.4 2.4 2.6 (Fisher) 1.6181 .0000 3.0510 .2160 1.2974 .7269 .0027 .8594 5.1709* 1.0519 2.5715 (.1764) (.2914) (.2022) (.2647) .0000 .0000 .0000 .0000 .0000 3.7536*" .3795 2.2128 1.4116 .6221 * £ <.05 " 2/4 cells had expected values < 5 127 Table 14 (continued) Post-placement Factors by Disrupted and Completed Adoptions Controlling for Gender and Membership in a Sibling Group Factor Disrupted % Completed 2C Child behaviourally different 7.7 Boys 7.9 Girls 7.5 Sibling 7.1 Alone 8.3 Child was sexually abused 4.2 Boys 5.7 Girls 2.7 Sibling 2.6 Alone 5.9 Child was physically abused Boys Girls Sibling Alone 1.3 2.6 2.4 (Fisher) 2.9898 .1746 2.2128 1.3815 .6800 2.0713 .5931 .0646 .2213 .2295 Child was neglected 2.7 Boys Girls 5.3 Sibling Alone 5.9 Probation extended 15.9 Boys 16.2 Girls 18.9 Sibling 15.9 Alone 20.0 13.4 18.9 10.8 20.5 6.7 .7691 1.1152 .0419 10.1666* .0146 3.0061 .0419 1.1047 * 2 < .05 128 DISCUSSION Charlie was born in September 1978 and is a handsome, healthy, active youngster who has been in foster care almost half his life. He is active in the Boy's Club and loves to play most sports. Charlie is in a regular classroom where he likes to read, does average work and has a desire to learn. He needs assistance in talking about feelings and reacting appropriately - it is easier for Charlie to act out his feelings than to talk about them, and it can be hard to remember rules. Charlie needs a family who is patient, firm and able to set consistent guidelines. He is, however, a charming little boy who could be a loving family member. (Adoptalk, Summer 1990. p.19) The discussion below describes a lot of 'Charlies'. Despite handicaps and disadvantages, it is vital to keep uppermost the idea that the subjects of this research paper are children. It is all too easy to dive into the numbers and statistics and forget the very human face of adoption today. It is the face of a lonely child waiting to become a "loving family member". Through understanding the nature of special needs adoptions and of the disruptions that will inevitably ensue, it is hoped that greater knowledge of the issues will lead to improved practice and policy. The discussion below follows the same format as the previous chapter, centered on the core hypotheses. 129 Hypothesis # 1: There will be no significant difference in outcome between boys and girls. The analysis indicated that there was no significant difference in outcome attached to gender. While there were more boys than girls in the disrupted group, and more girls than boys in the completed group, this difference was not significant. There were, however, some significant differences noted when the analysis of other factors controlled for gender. Knowledge of the gender of a child without consideration of any other factors will not assist in identifying children with an elevated risk of adoption disruption. This finding is consistent with previous disruption studies. Barth and Berry also noted the prevalence of boys in the disrupted group, but indicated that there were also more boys in their total sample. The influence of gender is discussed in each section below. Hypothesis # 2: Children placed with siblings will be more likely to experience fewer disruptions. The findings discussed in the previous chapter point soundly to the confirmation of the hypothesis. Children placed with their siblings who were available for adoption at the same time experienced far fewer disruptions than children placed alone. This issue is met with some diversity 130 of opinion in the literature. Kadushin and Martin (1988) report that "the research generally suggests that placing siblings together increases the risk of disruption" (p.590). However, Barth and Berry's (1988) research supports the hypothesis as does Festinger's (1986). There are no reasonable alternative explanations as to why there are more siblings in the completed adoption group other than the question of whether this group of siblings is truly representative of the groups of siblings in care who are awaiting adoption placement. However, the same caution holds for the children in the 'alone' group. In addition, the random selection of completed adoption files tends to narrow the likelihood that this particular sample is markedly different from other adopted children. Were this the case, more pronounced differences between the disrupted and completed groups would likely have been observed. Common sense tells us that siblings ought to do better when placed together - the children have natural allies, a we11-developed sense of their own history, and security in the knowledge of having endured a shared history of abuse and/or neglect. Kagan and Reid (1986) examined this point in their review of youths placed for adoption. A youth's ability to maintain himself or herself in an adoptive family is related'to the intensity of the anxiety and anger youths carry with them from their biological families. The six children . . who were placed with their siblings were less' emotionally disturbed, had more supports and were most likely to perceive themselves in a normal context than the other placed youths. 131 Additionally, such children were less likely to have been the scapegoats in their biological families . . . Children placed apart from siblings would conceivably be more likely to experience (and accept) the burden of pathology in their family, (p.70) Implications for policy and practice While common sense tells us that there is a logical basis for the success of sibling adoption placements, prevailing wisdom assigns a greater degree of risk to them. The direct implication of this finding is that greater emphasis must be placed on disseminating findings of research to adoptive applicants (and the public generally). In addition, efforts must be made to encourage siblings placed apart to maintain contact. Policy must be clear on this point and require social workers to locate siblings and explore their circumstances prior to placing a child apart from his/her brothers and sisters. The policy should require that if placement together is not an option, arrangements should be made for contact throughout childhood. In those rare circumstances where children must be kept apart due to concerns about inter-sibling abuse, the children's wishes about contact must be respected. Each of the sections below reviews sibling presence in addition to the main finding. 132 Hypothesis # 3: Children who are able to maintain contact with their siblings are more likely to experience fewer disruptions. As noted in the results chapter, there was insufficient information available to test this hypothesis. While instinctively it appears that it must be true, based on the finding with respect to sibling placement, there was not a means of deriving accurate information from the files. Visiting may have taken place and not been reported, or may never have taken place due to a number of factors including lack of knowledge that the child had siblings! This is an important area in the development of the child's sense of self and should be examined further. One hears stories and reads accounts of children separated when very young and placed in different homes. When later re-united, they discover to their sadness that they are missing a common history to help them interpret the events in their lives. Tragically, in the fairly small communities of British Columbia, these children may have been within a short distance of each other and yet never know the comfort of contact. Further investigation of the role that sibling.visiting plays in adoption completion is called for. It is a shortcoming of this study that this issue could not be canvassed. While there are a number of variables which would 133 affect the quality and quantity of sibling visiting (such as common parentage, age differences, common life experience, effect of inter-sibling abuse), the question appears to bear greatly on the quality of life of adoptees. Hypothesis # 4: The disrupted group will be marked by children who are older at the age when legally free and older at the time of placement. This hypothesis was supported by the data analysis. As with all other adoption disruption studies, the age at the time of placement was a critical factor in the eventual outcome of the adoption. However, it as important to examine the age at which the child became legally free. This point has not always been amplified in previous studies. While there was no significant finding with respect to the length of time that the children waited from termination of parental rights to adoption placement, an average wait of up to 25 months is clearly far too long for children who have lived for only 87 months. To frame the wait in months is important, for children only have 228 months of childhood in British Columbia. When, on average, it has taken up to 69 of those months before the child is legally free, during which s/he may have been subjected to very low levels of care, the urgency of early placement becomes apparent. 134 The age at which the child becomes legally free assumes additional importance when one considers the range of experiences that the children may have had during that period. Taking both developmental and social factors into consideration may account, in part, for the difficulty the older child has in forming new attachments. The evidence suggests that three somewhat different experiences can each produce the affectionless and delinquent character: a) lack of any opportunity for forming an attachment to a mother-figure during the first three years, b) deprivation for a limited period - at least three months and probably more than six - during the first three or four years, c) changes from one mother figure to another during the same period. (Bowlby, 1951, p.51) Given the very wide range of studies confirming the importance of the age of the child to adoption outcome, there are no reasonable alternative explanations which account for the significant finding of this study: age is an important and significant factor. While the age of the child is an important detail to know, it is really only a symbol for the issues that age represents. With increasing age comes increasing memory. Unlike the child placed as a newborn infant, the special needs child placed at 56 months comes complete with a history which might include a horrendous chronicle of abuse and neglect. The child may also come with an intact recollection of a family, which although not ideal, may represent a fantasy hard to live up 135 to for the new adoptive parents. The older child may also have extended family who have been and may remain important anchors in his/her life. Implications for policy and practice Writers and practitioners such as Cohen (1987) and Archer (1988) have pointed to the necessity of capturing the child's history in a manner which renders it accessible to him or her and which puts the history in a context which serves to explain subsequent life events. Working with the special needs child to prepare for adoption placement requires that the social worker take the time to understand the sequence of events which led to the granting of a permanent order. The child may have no active memory of the precipitating events and may only dimly understand the reason s/he is now to be placed with a new family. The child at 56 months may also have a developing sense of being different in a society which treasures the intact nuclear family as the ideal. A child without parents, cared for by an anonymous state could well feel isolated and alone. S/he might be unaware of other children in this position. Fanshel and Shinn (1978) comment on this sense by former children in care that they were alone in the world and somehow very different than other children. These former wards, now young adults, shared a feeling of distance from their peers and maintained a belief that they alone were responsible for their affairs in life. 136 Age also brings understanding. For the older special needs child, this may be the understanding that placement in an adoptive family means abandoning a known, if uncomfortable, status for a new and possibly frightening position. The adopted child who is told that he was chosen by his adoptive parents has to deal with the underlying reality that he is not a child of choice, that he was abandoned rather than chosen by his natural parents, and that his adoptive parents may have had no other choice of raising a child. (Blum, 1983, p.144) Wherever and whenever possible, the child should be involved in the placement planning process. The empowerment of the child will assist him/her to become an active participant in the adoption rather than a passive recipient of other people's good will. As practised by the Special Needs Adoption Demonstration Project (SNADP), this empowerment of the child takes the form of encouragement to learn his/her history, to re-visit former foster parents, to learn about early life experiences, and to address the issues related to a sense of abandonment by the birth parents. 137 Hypothesis # 5: The greater length of time that a child spends with his/her family of origin will inversely affect outcome. This issue is clearly related to the previous hypothesis regarding age. The analysis indicated a moderate relationship between time lived with family of origin and adoption completion. Those children whose adoptions were not completed lived, on average 19 months longer with their families of origin. When membership in a sibling group is controlled, the level of significance rises. This is consistent with the finding above about the significant effect of sibling placement. There is very little discussion in the literature regarding this point. Although the age of the child receives a great deal of attention, the time spent with biological family does not. One must then speculate about the joint and confounding effects of exposure to a wide range of abuse and neglect, and attachment to a biological parent. Partridge et al. (1988) associate disruption with increased levels of trauma and need. "There is every indication that children in the disrupted group are more experienced with loss, instability and maltreatment and appear to be more emotionally disturbed and behaviourally dysfunctional than their counterparts in continuing placements." (p. 47) However, Partridge et al. did not find 138 that returns to the birth parents were significantly associated with disruption. The previous chapter indicated that the number of returns to the birth parents was not significant. This finding is consistent with that of other studies. There were no significant differences between boys and girls on the variable of returns to biological family. This finding, while consistent with the findings of other researchers, is surprising. It would seem that children who have experienced multiple returns to their family of origin, would be more prone to disrupted placements due to the twin factors of parental attachment and unsettled early years. It maybe that those factors work in favour of the child's ability to join a new family (or at least work against pre-disposing the child to subsequent disruption). The child who has been returned to the family of origin more than once following an apprehension has been able to witness first-hand the steady decline of the parents' ability to provide adequate care and may have experienced repeated physical and sexual assaults. For these children "family romance fantasies" as described by Wieder (1977) may simply not exist. Nickman (1985) writes of the losses of adoption: "Adoptees have a need to mourn several kinds of loss: the overt or object losses, status losses, and covert losses such as the assault on self-esteem which arises from the knowledge of not having remained with one's original parents." 139 Implications for policy and practice. Adoption "mythology" holds that the child returned frequently to parents will be difficult to place for adoption. The evidence from this study does not support such a view. Previous studies on adoption and in the field of attachment indicate that the child who has been able to form an attachment early with a significant mother-figure may be able to repeat that process with another parent. Placing the child who has been returned to parents several times is probably less complicated by the number of returns, but more likely complicated by the age at which the adoptive placement finally happens. The child who has been returned to parents on several occasions is usually older and thus, has more active memory of past events. Such children will also have experienced a range of foster home placements and alternate caregivers. Keeping in mind a range of factors including the age at which the child is free for placement, the adoption social worker ought not to feel constrained by the issue of returns to biological parents as a predictor of disruption. The adoption plan in such cases should be discussed openly with the biological parents who may feel that the age of the child and the frequent returns might mean that the child will remain in foster care and not find a permanent family. 140 Hypothesis # 6: Children placed with relatives or with long-term foster parents will not experience disruption at the same rate as those placed with strangers. The analysis indicated a very significant correlation between placement with known adoptive parents and adoption completion. This is not at all surprising and has been supported by previous studies. Alternative explanations are not supportable in this case. The children placed with their existing foster parents or relatives experienced a much lower incidence of adoption disruption. Barth and Berry's (1988) extensive review of disruption pointed strongly to the success of adoption by foster parents. "Foster parent placements may have lower disruption rates because the child and family learn if the placement will work during the foster care stage." (Barth and Berry, 1988, p. 91). An additional important factor is that the child and parents will have begun to understand each other's behaviour prior to an adoption commitment having been made. Once the plan changes to adoption, each party will need to be able to analyse the behaviour and demeanour of the other in order to accurately predict sources of tension and conflict. Implications for policy and practice Foster parent adoptions were actively discouraged until quite recently. They were seen as a 'back-door' route to adoption and were disfavoured. Due to the different 141 requirements for home-study, selection criteria, matching, and payment, it was felt that foster parents should not adopt the children in their care. While this attitude has changed during the last five years, policy and program changes need to be made to catch up with this and other research findings. Foster parents interested in adopting the children placed with them should be encouraged to do so. The home study guidelines should be reviewed for efficacy in such situations. Supports and services provided during the foster care period should be withdrawn gradually rather than at the point of adoption completion. The new adoption subsidy program may assist with some aspects of this latter point. Policy should be re-written to reflect a requirement that existing foster parents be approached to see if they are interested in adoption before referral to the central adoption registry is made. This policy needs to be circulated in the child welfare community, including members of the legal profession, so that a clear statement can be made about the desirability of placement stability for children removed from their parents' care. Wherever possible, foster parent adoption needs to be canvassed prior to the permanent order hearing. Birth parents who are aware that their child will be adopted by the foster parents may appreciate the stability offered and be willing to accept 142 their child's best interest as the test rather than undergo the excruciating pain of a protracted hearing. Hypothesis ft 7: The number of placements and total length of time spent in foster homes or group homes prior to placement will inversely affect outcome. The findings presented in the previous chapter with respect to this hypothesis are consistent with those of other researchers with respect to the total length of time spent in foster or group homes. The total length of time the child spends in foster homes or group homes does not negatively affect adoption outcome. However, the finding of no relationship between total number of placements and disruption is not completely consistent with that of other research. Hall (1986) and Barth and Berry (1988) found that the number of placements was not a significant contributing factor to adoption disruption. However, Partridge et al. (1986) and Festinger (1986) found that the number of placements was significant. Differences in study outcomes may be related to sample size, differing agency practices, different interpretation of file data, or a combination of these factors. The mean number of placements was found to be higher in this than in other studies. One could speculate that this study, as the most recent, identified that foster home 143 placements are no longer as stable as they may once have been. Child welfare agencies across Canada and the United States are experiencing great difficulty in recruiting and retaining foster homes. The increased numbers of foster home placements for the children studied in this research may reflect this new reality in foster care. There are significant differences between the sibling and 'alone' groups when gender is examined. This finding, that the children in the 'alone' group experienced a much wider range of placements (when split by gender) than the children in the sibling group is not commented on elsewhere. Only the 'alone' girls whose adoptive placements disrupted experienced elevated numbers of foster or group home placements (n = 8). That the sibling group experienced higher numbers of placements is not surprising given the difficulty social workers report in finding foster placements for sibling groups. Further review of the ages of the 8 'alone' girls whose adoptions disrupted indicates that they were, on average, the oldest when legally free (88.2 months) and the oldest when placed for adoption (109.5 months) of the entire group. Taking these ages into consideration, the number of placements becomes explicable. These children experienced higher numbers of placements because they were in care for longer periods of time. The previous discussion about the 144 significance of age as a predictor of disruption holds for the analysis of this subgroup of the total sample. The hypothesis is not proven in this case. The number of placements and the total length of time in foster or group homes does not affect adoption outcome. Implications for policy and practice While the hypothesis was not proven, the finding about the number of placements should be cause for some concern. The mean number of placements experienced by children in British Columbia is higher than those of children in other studies. In the Hornby et al (1986) study the disrupted group had 3.9 placements and the completed group, 2.7 placements. Festinger (1986) reports a mean of 2.7 for the disrupted group and 1.3 placements for the completed group. Barth and Berry (1988) report a finding of 2.9 foster home placements overall. While the recruitment and retention problems are very real and pose serious problems for child welfare administrators, the evidence is clear that the numbers of placements are higher than in other jurisdictions. Although it is encouraging that there is not a correlation between the numbers of these placements and adoption completion, there is a need to continue to highlight resource shortages and the impact these have on children in care. A review of the reasons for the numbers of moves children experience is called for. A case by case analysis of each move should 145 highlight areas where similarities exist in reasons for changes in placements. Where Ministry policy was identified as the problem, documentation would then be available to justify changes to reduce the number of moves. Where other factors are identified, these should be analysed in consultation with foster parent representatives and efforts made to ameliorate the condition. Hypothesis # 8: The pre-care experiences of children will affect adoption outcome. Those children with histories of physical and sexual abuse will fare poorly in adoption. The children in the sample had a wide range of pre-care experiences, often being subjected to multiple forms of abuse before entering permanent care. Barth and Berry (1988) do not address the pre-care experience question directly. Partridge et al (1986) consider "prior incidence of abuse and/or neglect [to be] consistently a strong predictor " of adoption disruption (p.44). Festinger (1986) does not directly address the issue but states that "the reasons for placement were quite similar for children whose adoptive placements eventually disrupted and for those who were adopted" (p.21-22). Nelson (1985) reports that "parents' satisfaction is unaffected by ... the reason for severing the biological parents' rights" (p.72). Hall (198) found that "there is no 146 significant difference between successful and disrupted adoptive children with regard to the reason for their coming into care" (p.75). None of these reports indicate whether the variable controlled for either sibling status or gender. The results in the previous chapter indicate that in some instances there is a clear relationship between the reason the child entered care and subsequent adoption disruption. There are certainly many factors to consider in this regard, but certain groups of issues appear to stand out in the analysis. (A) Abandoned There were no disruptions among the members of sibling groups of children who had been abandoned. Kadushin and Martin (1988) refer to abandonment as "the ultimate neglect — feelings of frustration; indifference, inattentiveness, and lack of concern and awareness of the child's condition and basic needs are associated with neglect" (p.230). When one examines this description while keeping in mind the earlier discussion about the effect of siblings on adoption outcome, the deductive conclusion must be that siblings are able to meet each other's needs in ways which their biological parents cannot. Once placed in the adoptive home, this pattern of collective support continues and provides the children with the foundation they require to successfully adapt to a new family. 147 (B) Physically abused Children who entered care as a result of being physically abused generally fared better in adoption than children who were not physically abused. This held particularly for boys and members of sibling groups. Kadushin and Martin (1988) define physical abuse as "excessive and inappropriate and hence unacceptable violence toward children — some definitions of abuse stress the fact that the behaviour is deliberate with intent to harm the child. Other definitions give [more] consideration to the danger to the child, whether it is intentional or not." (p.229) In contrast to Partridge et al. (1986), this study found in certain cases, prior physical abuse was not predictive of adoption disruption. This may be due to a combination of factors including the element which Kadushin and Tizard call "resilience". A sense of mutual support and strength in the face of obviously poor care may explain the lower disruption rate for siblings who have been abused. More generally, children who have been physically abused to the point that they are apprehended may have had ample opportunity to accept that their biological parents cannot care for them. These children have physical scars to remind them of their parents' inability to protect them from harm. They may be inclined to enter a new family with optimism that they will be well-cared for. 148 The boys were not over-represented in the physical abuse category. However, they fared better in adoption than girls who had been physically abused. There is not enough information available to determine the extent of the injuries suffered or the time period over which they were endured. These factors may have influenced the finding. Further research is needed to clarify this point and to examine the effect of gender on physical abuse. (C) Sexually abused There was no tendency toward significance on this variable overall, or for membership in a sibling group. The latter may be understood by recalling that sexual abuse is most often a private matter which all members of the family may not be privy to. However, for boys there was a tendency toward a significant correlation between adoption completion and sexual abuse. This may be due to more "passive acceptance by the daughter [which] reflects the effects of prior developmental history ... the profile of the child who does become involved in incest suggests a child who is more compliant, more passive, and less confident in her capacity to reject the parent." (Kadushin & Martin, 1988, p.301) One would expect that more girls than boys were sexual abuse victims and this was the case in this study. It should be noted however, that some of the sample children entered care prior to 1982 when sexual abuse awareness was not as discerning as it is today. As a result, there may have been 149 a much higher incidence of sexual abuse among the children which was not identified at the time they entered care. (D) Emotional abuse The National Clearing House on Child Neglect and Abuse defines emotional neglect ... as failure to provide the child the emotional nurturing or emotional support necessary for the development of a sound personality, as for example subjecting the child to rejection or to a home climate charged with tension, hostility and anxiety producing occurrences which result in perceivable problems in children. (Kadushin & Martin, 1988, p.235) Emotional abuse as a reason for entering care was noted for only 15 of the sample children. Of these, 9 were boys and 6 were girls. Although the numbers were small, there appears to be a tendency for boys who were emotionally abused to experience fewer disruptions than girls. While boys may be more frequent targets of abuse in the early years, the question of the effect of emotional abuse on adoption completion may have more to do with early role stereotyping than gender alone. The child who is both emotionally abused and expected to be a 'brave little soldier' may develop a self-protective shell which encourages reaching out to form new bonds when the old ones are irreparably harmed by emotional cruelty. Partridge et al. (1986) identify emotional abuse as predictive of disruption. They do not list their selection criteria, and it may be that this study would have replicated those findings with a similar definition. For the 150 purposes of this study, emotional abuse was noted only if the file made specific reference to it. (E) Parental alcohol abuse There was a significant finding overall with regard to parental alcohol abuse and adoption completion. This was also the case for girls but was not repeated for siblings, children placed alone, or boys. Once again, inadequate file detail does not permit an analysis of the range of behaviours which parental alcohol abuse covers. Only field observation would provide the wealth of data necessary to confirm this finding with conviction. Knowing the boundary confusion which often arises in neglectful, alcohol abusing families, one immediately thinks that sexual abuse may be a factor in this finding. However, this contrasts with the earlier observation that sexual abuse is not significantly associated with either disruption or completion for girls. Role modelling may again play a part in this issue. Many children who grow up exposed to parental alcohol abuse commit themselves to an abstemious lifestyle, due to the abhorrence they hold for alcohol. That same drive, coupled with a societal requirement that girls be well-behaved, compliant, and uncomplaining may foster a desire for permanence with a new family. This area requires a great deal more research in order to confirm this finding and/or provide alternative explanations. 151 (F) Relinquished There was a tendency overall for children who were relinquished to fare more poorly in adoption than children who were not relinquished. However, the numbers were very small (n = 5) and conclusions cannot be drawn on the basis of the chi-square analysis alone. It is interesting to note that all 5 of the children were boys and all 5 of their adoptions disrupted. Unresolved feelings of loss may account for this phenomenon. It is also likely associated with an array of behaviours developed to mask the pain of loss. (G) Chronic neglect There was a significant finding overall that chronic neglect was associated positively with adoption completion. The finding was not repeated when controlling for membership in a sibling group or by gender. Partridge et al. (1986) made a similar finding. Deprivation of necessities and inadequate supervision are the two most frequent forms of neglect. When deprived of necessities, the child is not adequately fed or provided with adequate and appropriate clothing. The child's living conditions are dilapidated with inadequate heating, insufficient protection from possibility of injury and lack of proper bed or bedding ... the neglectful parent fails to adequately meet her own needs even as she fails to meet the basic needs of her children ... the neglectful parent may be rejecting the parental role. (Kadushin & Martin, 1988, p.230-231) The child reared under neglectful conditions is able to see first hand the inadequacy of his/her parent. When 152 removed from such circumstances and provided with adequate food, shelter and clothing, the child may be ready to make new commitments which signal an end to neglect. This situation is somewhat akin to the circumstance where the child is relinquished. In that case, the parent makes a conscious decision to cease caring for the child. In the case of neglect, the parent demonstrates to the child his/her inability to parent and abdicates the role. For the child, the effect may be the same as relinquishment. This is the very point that Kadushin (1967) sought to make in his study of children adopted when older. These children . . . made two important shifts in moving from their own home to the adoptive home. . . . . from a home that offered little in the way of meeting their needs in terms of affection, acceptance, support, understanding, and/or encouragement to the adoptive home, which offered some measure of these essential psychic supplies. They also made a change from a lower-class, multi-problem, generally disreputable family living in a slum-ridden area of the community to a middle-class, reputable home in an area of the community that had some status. (Kadushin, 1967, p.29) (H) Other reasons for entering care As noted in the previous chapter, there were a number of reasons that fell into this category. Predominantly they are characterized by parental lack of interest in continuing to care for the children. As with chronic neglect, the children can be expected to have understood their parents' inability to provide care as a form of abandonment and sought security in a new, permanent family. Of all the 153 reasons for children entering care and the associated analyses, this factor produced the highest significance levels. This was true for the sample overall, for girls and for children in the 'alone' group. There was no significant finding for boys or for children who were members of a sibling group. This set of reasons for entering care reflects some very unique circumstances. The children appear to have held secondary positions in the family, subservient to the unmet needs of their parents. Kadushin's picture of the neglected child above describes even more closely the children in this group. The highly significant finding for children in the 'alone' group tends to reinforce these children as searching for the security and stability their own families were unable to provide. Although they experienced the events of their childhood on their own, there is a quality of personal rejection among this set of reasons which is so final that the children must believe that they are free to create new ties to an adoptive family. Children will be more likely to succeed in adoptive families if they have permission from their biological parents to experience their losses, express feelings of fear and anger and develop the capacity to form new bonds with an adoptive family. Conversely, children who cannot live with their biological families but remain bound to the pressures and conflicts they experienced . . . are likely to go through a series of placements and will be less likely to succeed in an adoptive family. (Kagan, 1986, p.71) 154 The hypothesis cannot be stated to be proven or disproven. There are a number of pre-care experiences which affect adoption completion. The inter-play between and among these factors requires closer examination. The study is limited in its ability to accurately assess individual cases on the basis of file information. Additional information about social and behavioural functioning of the sample would augment the data and might assist in identifying personal attributes of the sample children which have assisted them to overcome their pre-care experiences. As this is a limited study in the sense that the dependent variable - adoption completion - by definition narrows the scope to a short time period, it is possible that some of the completed adoptions will still end in dissolution. The long-term effects of some of the pre-care experiences of the children are difficult to measure as there is little long-term follow-up of adopted children. For example, fetal alcohol syndrome remains a poorly understood disability, the effects of which are lifelong. In many cases, behavioural and learning disabilities do not become apparent until adolescence and are very difficult to accurately diagnose. The detection of sexual abuse of children now represents 22% of the provincial child protection caseload (British Columbia, 1990). However, the long-term effectiveness of the treatment of sexually abused children is still unknown. As these children approach 155 latency, emotional and behavioural problems may arise which will threaten the security of the adoption placement. An additional consideration about this finding is that there may be additional information about specific behaviours and emotional disabilities which was not collected due to the difficulties identified in describing the full range of behavioural difficulties in file recording. Implications for policy and practice The testing of this hypothesis has resulted in an analysis that tends to break down another of the great adoption myths - that the pre-care experiences of some children mark them as being unadoptable. As indicated above, there are no experiences that can conclusively result in adoption disruption. To the contrary, it could be said that a sibling group with a background of abandonment may forge new bonds with a family more easily than children who are sexually abused. One factor is not more important than any other, it is the combination of pre-care experiences along with the child's age, sibling status and unique qualities which determine adoption completion or disruption. As noted above, this research has focussed exclusively on child related variables. Further research is necessary to examine the process by which adoptive parents and waiting children are matched. When parents and child are adequately prepared for the adoptive placement do adoptions disrupt less frequently ? 156 Adoption mythology about the adoptability of children rests, assuredly, on experience in recruiting families for children. Fear and lack of knowledge on the part of prospective adoptive parents about the difficulties of adopting sibling groups or children with histories of abuse and neglect are certainly areas which need to be addressed. Demonstrating to practitioners and adoptive parents that older child adoption is not inherently perilous, and providing post-placement support to families are two tangible services that can be provided to strengthen the viability of special needs adoption. Dissemination of this and other related research findings will help to reduce dependence on adoption mythology as a basis for permanency planning. There is also a need to educate therapists about adoption so that they will be able to incorporate theories about attachment, adoption and pre-care experience into their therapy sessions with children. The information base about children in the care of the Superintendent needs to be expanded so as to afford the opportunity to track individual cases from entry to care through to adoption placement. Armed with this information, it would be possible to identify those combinations of factors thought to contribute to placement instability. 157 Hypothesis # 9: Children with multiple special needs will experience more disruptions. As indicated in Table 14, this hypothesis is not proven. The nature and extent of the children's special needs did not influence adoption completion except in one case - that of boys with 'other special needs'. These special needs are most frequently emotional/behavioural difficulties. These problems were identified prior to placement. Knowing in advance that the child to be placed had behavioural difficulties, adoptive parents of boys appear to be able to manage the behaviour and proceed to finalize the adoption. This finding confirms that of Kagan and Reid (1986). This survey demonstrates an encouraging rate of success in helping severely emotionally disturbed youths to find and maintain relationships with adoptive families. These were children with long histories of multiple placements, institutionalization for treatment of learning and emotional problems, neglect, physical abuse, and behaviours that were dangerous to themselves and others. With this background a legal adoption rate of 70% is very high. (p.69) Neither this study nor that of Kagan and Reid (1986) are able to comment on the matching process as described above. It may be that there are elements of adoptive parent selection which are able to explain this particular phenomenon. In the case of the Kagan and Reid study, it would be useful to know the extent of abuse suffered by the 158 children in the sample. Recent dramatic increases in the numbers of children born drug dependent or suffering from neonatal abstinence syndrome will dramatically change the picture of the very young child placed for adoption. This study did not capture information about this group of children as the time period studied pre-dated the emergence of this disturbing trend. The recognition of children with fetal alcohol syndrome or fetal alcohol effect is sometimes not made until the child is well into the adolescent years. Dorris (1988) has written movingly of his decade long struggle to identify the nature of his adopted child's disability. His story is not uncommon and it must be assumed that larger numbers of children will be diagnosed as having fetal alcohol syndrome in the next few years. This study has not been able to accurately capture information regarding this group of children. Research into the effect on adoption placement of fetal alcohol syndrome and parental drug abuse during pregnancy are urgently needed. Placement options for these children will be severely hampered by our lack of knowledge about the long-term effects on children and families of these preventable conditions. None of the cases where the child was discovered, after placement, to have been emotionally abused, neglected or sexually abused prior to the adoptive placement, were 159 completed. This finding requires further follow-up. Although the numbers were very small, the sample size may not have permitted an adequate testing of this issue. It remains unclear whether it was the behaviours that sprang from this prior, undisclosed abuse, which prompted the disruption, or whether other factors were at play. Examination of subsequent adoptive placements of these children is needed to examine the long-term effect of the disclosures. Kadushin (1967) indicates that there is a therapeutic component to adoption. "Adoption is not psychotherapy but its psychotherapeutic potential is like a good marriage, a true friendship, a new and satisfying job, an enjoyable vacation. It can help to repair old hurts." (p.31) There may well be therapeutic value in adoption, but it should never be used for that purpose. Kay Donley warns about the 'middle-class child improver* - a group of adoptive parents who fare poorly in her estimation with behaviourally dysfunctional children. While these parents, frequently professionals with university training, are motivated by the best of intentions - to help a needy child, their inability to accept the very slow pace of change in the children's lives confuses and frustrates them. The interplay between adoptive parent variables and children with special needs requires further research in order to determine whether Donley's theorem about the types of parents and children who do well together is generalizable. 160 Other observations As noted in Chapter 5, there are disadvantages to secondary analysis as a research methodology. One is forced to rely on the observations and findings of others in order to conduct the research. In this case, the journey of discovery through the files provided revealing glimpses of adoption and protection practice across time and place. A purely subjective observation from reading over 85 files is that in those cases where the social worker had clearly made adoption planning an important casework goal, there were fewer difficulties with placements. In addition, in those cases, the file information was readily available and showed consistency. This was not the case for every file. One child's worker indicated that the child was gregarious, energetic and involved in a number of community activities. The worker recommended seeking a home in an urban setting where the child would be able to continue her activities. By the time the child was placed for adoption, she had had a change of social worker. She was placed for adoption with a family who lived in a very remote part of the Province. Her adoption was not completed. Cases originating from specialized adoption units reflected the greater concentration and expertise available in such units. These cases were characterized by frequent references to case planning activities and goal-setting. It will remain for a future researcher to determine if these 161 cases disrupt less frequently than others. There appeared to be no significant difference between cases originating in rural or urban areas. Areas for further research There are a number of areas for further research which logically follow from this project. Foremost among these would be a project to track this same sample of children. Did the children whose adoptions were completed remain with their adoptive families after placement ? Were the children in the disrupted group subsequently placed for adoption, and were those adoptions completed ? The significance of the presence or absence of siblings was demonstrated for the children studied, however the circumstances of the absent siblings were not. A follow-up review of children separated from their siblings is called for. Adoptive parent variables including a qualitative review of the home study material would yield a wealth of information about special needs adoption. Allowing that the face of adoption has changed, research needs to examine whether the change is only skin deep for adoptive parents. Do they agree to adopt special needs children due to the lack of healthy infants or is their motivation based on a desire to offer a child a home ? Beyond the simple demographic characteristics of the adoptive parents, are there features which mitigate for or against successful special needs adoption ? The role of the adoption social worker requires review. As the author of the assessment which will ultimately convince another social worker to place a child, the adopting parents' social worker holds great power. Is this authority always exercised in the most careful manner, avoiding bias ? The role of planning for children has received a great deal of attention in the literature in the last two decades. Is this information incorporated into the daily practice of workers ? How do planning activities influence adoption ? Summary This research has demonstrated that there are very few children for whom adoption is not a viable option. There are no surprises contained in this study. The research confirms, on reflection, what we should implicitly understand about special needs adoption. This and other studies point to the adoptability of children-in-care. Greater emphasis must be placed on planning for the adoption of older children and preparing them for the experience. The significant finding of this study is not that age inversely affects adoption completion, but rather that age represents a history that can be neither ignored nor forgotten. The adoption files reviewed contained some very poignant and tragic histories. There were none so lamentable as those concerning siblings who were separated. There is no question that this practice is occasionally a valid means of 163 dealing with particular issues in a family. However, the Ministry needs to focus on this area in order to educate social workers about the value of sibling contact and placement. Adoption files represent a chronicle of each child's life in care. They must be complete and adequately rich in detail to answer the inevitable questions about the past which haunt us all. The development of special needs adoption demonstration projects to test some of these ideas in B.C. has been an excellent first step. It is now important to expand these concepts throughout the province. Training for social workers must include sessions on planning for children and on adoption related issues. This must occur in the universities as well as within the Ministry of Social Services and Housing. While there are risks in adoption, particularly for older children, there are not any situations which appear to dictate that children will fare poorly in adoption. This research has concentrated on only one aspect of adoption -that of the children. This will assist practitioners to recognize that the children alone cannot be held responsible for the outcome of their adoptive placements. Adoption must focus on the needs of the child awaiting placement, establishing as a first order priority a thorough review of the circumstances of the child's entry to care. 164 Children of adoption bring with them to their new families a variety of issues. Some are known and visible but many are not. The discovery of the range of these issues and their effect on the child and her/his belief system is the key function of the social worker planning for adoption. Having painted a broad picture of the children placed for adoption, it is left to future researchers to examine the role of the other key participants in adoption, adoptive parents and their families, and social workers. All of these people bring with them to the adoption a particular set of values, beliefs and practices which may influence the adoption. Knowing those child factors which are thought to be critical to adoption outcome and reviewing them in light of findings about the other key participants will enlarge and brighten our understanding of the special needs adoption process. 165 FOOTNOTES 1) See for example: Hansel and Gretel; Snow White. In more recent times one could look to such examples as Kipling's The jungle book and even Burroughs' Tarzan. 2) See Kadushin (1988), Tizard (1977), Whiteford (1988). 3) See for example: National Commission on children in Need of Parents (1979) Who knows ? Who cares ? Forgotten children in foster care.; Stone, Norman, Stone, Susan (1983) The prediction of successful foster placement Social Casework Vol 64 #1 January 1983; Festinger, Trudy (1983) No one ever asked us: A postscript to foster care. New York: Columbia University Press 4) See for example: Lipman, Margaret (1984) Adoption in Canada: Two decades in review in Sachdev, P. (1984) Adoption: Current issues and trends. Toronto: Butterworth and Co. Hepworth (1980) 5) Festinger does not provide a level of significance for this finding. Her study report is confined to percentage levels of disruption. She does indicate where findings are statistically significant but additional information is not provided. 6) Levels of significance were not reported. Percentage figures were provided in this study of the adoption of developmentally disabled children. 166 7) This analysis must be viewed with some caution as the n was small and two of the four cells had expected frequencies of less than five. 8) This finding must be viewed with some caution as two of the cells had expected frequencies of less than 5. 9) This finding must be viewed with some caution as there was one cell with an expected value less than 5. REFERENCES 167 Adoption Act, Revised Statutes of British Columbia, Chapter 4 (1979). An Act respecting the Adoption of Children, Revised Statutes of British Columbia, Chapter 6 (1924). An Act respecting the Adoption of Children, Revised Statutes of British Columbia, Chapter 6 (1924). Archer, J. (1989, May). A description of the special needs adoption demonstration project. Paper presented at Ministry of Social Services and Housing Adoption Conference. Vancouver, B.C. Barth, R., Berry M. (1988). Adoption and disruption: Rates, risks, and responses. Hawthorne, New York: Aldine de Gruyter. Bailey, K. D. (1978). Methods of social research. New York: The Free Press. Bass, C. (1975). Matchmaker - matchmaker: Older child adoption failures. Child Welfare LIV:71 502-512 Benet,K. (1976). The politics of adoption. New York: The Free Press. Blum, H. P. (1983). Adoptive parents: Generative conflict and generational continuity. The Psychoanalytic Study of the Child. Vol. 38 141 - 163. Bostwick, G. J. Jr., Kyte, N. S. (1985). Validity and reliability. In R. M. Grinnell (ed.). Social work research and evaluation, (pp. 161 - 184). Itasca, Illinois: F. E. Peacock Publishers. Bowlby, J. (1953). Child care and the growth of love. Harmondsworth, England: Penguin Books Ltd. British Columbia. (1975). Fifth report of the royal commission on family and children's law. Part VII Adoption. Victoria: Queen's Printer. British Columbia. (1990) Ministry of Social Services and Housing. Services for people. Annual report 1988-89. Victoria: Queen's Printer. 168 Churchill, S., Carlson, B., Nybell, L. (eds) (1979). No child is unadoptable. Beverly Hills: Sage Human Services Guides. Clarke, H. (1957). Social legislation (2nd ed.). New York: Appleton, Century, Crofts Inc. Cohen J. (1981). Adoption breakdown with older children. Toronto: University of Toronto, Faculty of Social Work. Cohen J., Adoption breakdown with older children, in Sachdev, Paul (1983) Adoption: Current issues and trends. Toronto: Butterworth and Co. Cohen J., Westhues A. (1987). How to reduce the risk: Healthy functioning families for adoptive and foster children. Toronto: University of Toronto Press. Coyne, A., Brown, M. (1985). Developmentally disabled children can be adopted. Child Welfare LXIV, 6 November December 1985 607-615 Csapo, M. (1989). Children in distress: A Canadian perspective. Vancouver:Centre for Human Development and Research. Donley, K. (1989, May). Preparing children for adoption. Paper presented at Ministry of Social Services and Housing Adoption Conference, Vancouver, B.C. Dorris, M. (1989). The broken cord. New York: Harper and Row. Fanshel, D., Shinn, E. B. (1978). Children in foster care: A longitudinal investigation. New York: Columbia University Press. Festinger, T. (1983). No one ever asked us: A postscript to foster care. New York: Columbia University Press. Festinger, T. (1986). Necessary risk: A study of adoptions and disrupted adoptive placements. Washington. D.C.: Child Welfare League of America. Gochros, H. L. (1985). Research interviewing. In R. M. Grinnell (ed.). Social work research and evaluation. (pp. 306 - 342}. Itasca, Illinois: F.E.Peacock Publishers. Goody, J. (1969). Adoption in cross cultural perspective. Comparative Studies in Society and History. Vol.11 55 - 78. 169 Grinnell, R. M. Jr. (1985). Social work research and evaluation. Itasca, Illinois: F.E.Peacock Publishers. Grotevant, H., McRoy, R., Jenkins, V. Emotionally disturbed. adopted adolescents: Early patterns of family adaptation. Family Process, Vol 27 Dec.1988 439-457^ Hall, J. (1981). A comparative analysis of successful and disrupted adoptions in a private child welfare agency. (Doctoral dissertation. St. Louis University 1981). Dissertation Abstracts International, 43, 5489. (University Microfilms No. 8207404). Hepworth, H. P. (1980). Foster care and adoption in Canada. Ottawa, Ontario: Canadian Council on Social Development. Hoggett, B. M., Pearl, D. S. (1983). The family, law and society, cases and materials. Toronto: Butterworth and Co. Hornby, H. (1986). Why adoptions disrupt . . . and what agencies can do to prevent it. Children Today, July-August Kadushin, A. (1967). Reversibility of trauma: A follow-up study of children adopted when older. Social Work Vol. 12 No. 4 22 - 33. Kadushin A., (1970). Adopting older children. New York: Columbia University Press. Kadushin, A., Martin, J.(1988). Child welfare services. (4th ed.). New York: MacMillan Publishing Company Kagan R. M., Reid, W. J. (1986). Critical factors in the adoption of emotionally disturbed youths. Child Welfare XV 1, January February 1986 63 - 73. Katz,L. (1986). Parental stress and factors for success in older child adoption. Child Welfare Vol LXV #6 Nov - nPr 1986 569 - 578. Kennedy G.D. (1955). The legal effects of adoption. The Canadian Bar Review, Aug. - Sept. 1955, Vol. XXXIII No. 7. Kroll, Michael (1990, Feb./Mar.). Outfront: Pro-life parents. Mother Jones Magazine, pp. 13 - 16. Lipman M. (1984). Adoption in Canada: Two decades in review. In Sachdev, P. (ed.). Adoption: Current issues and trends. (pp. 31 - 42). Toronto: Butterworth and Co. MacDonald, J. A. (1984). Canadian adoption legislation: An 170 overview. In Sachdev, P. (ed.). Adoption: Current issues and trends. (pp. 43 - 61). Toronto: Butterworth and Co. Nelson, K. (1985). On the frontier of adoption. Washington. D.C.: Child Welfare League of America. Nickman, S. L. (1985). Losses in adoption: The need for dialogue. The Psychoanalytic Study of the Child. Vol. 40. 365 - 398. Norusis, M. J. (1988). SPSS/PC+ V2.0 base manual. Chicago: SPSS Inc. Partridge, S., Hornby, H., McDonald, T. (1986). Legacies of loss: Visions of gain - An inside look at adoption Disruption. Portland: University of Southern Maine Center for Research and Advanced Study Proch, K. (1982). Differences between foster care and adoption: Perceptions of adopted foster children and adoptive foster parents. Child Welfare Vol. LXI No. 5 259 - 268. Sachdev, P. (ed.) (1984). Adoption: Current issues and trends. Toronto: Butterworth and Co. Tizard, B. (1977). Adoption: A second chance. London. England: Open Books Publishing Ltd. Tizard, B. (1979). Adopting older children from institutions. Child Abuse and Neglect, Vol. 3 No. 2 535 - 538 Waiting Children. (Summer 1990). Adoptalk. p.19 Weinbach, R. W. (1985). The agency and professional contexts of Research. In R. M. Grinnell (ed.). Social work research and evaluation, (pp 66 - 82). Itasca, Illinois: F.E. Peacock Publishers. Whiteford, H. (1988). Special needs adoption: Perspectives on policy and practice. Unpublished master's thesis, University of British Columbia. 171 W i e£?F' H- (1977). The family romance fantasies of adopted children. The Psychoanalytic Quarterly. Vol 46. 185 - 199. Zwimpfer, D. (1978). Early indicators of adoption breakdown. Unpublished master's thesis Victoria University of Wellington, Wellington, New Zealand. Zwimpfer, D. (1983). Indicators of adoption breakdown. Social Casework 64,~3 169 - 177 172 APPENDIX A File Review Schedule Case Identification [ J Age when legally free Age at adoption placement If more than one placement, complete one review schedule each, using same case ID [ ] Months [ ] Months 3) Age at adoption completion N/A 4) Age at adoption disruption N/A 5) Gender of child 1 - Male 2 - Female Placement History 6) Biological Family N/A 7) Returns to biological family N/A 8) Foster home/group home N/A 9) Number of foster/group home placements N/A 10) Residential care/treatment N/A [ ] Months [999] [ ] Months [999] [ ] [ ] Months [999] # [99] [ ] Months [999] [_J # [99] [ ] Months [999] 173 11) Number of residential care/treatment [ ] placements # N/A [999] 12) Previous adoptive home [ ] Months N/A [999] 13) Number of previous adoptive placements [ ] # N/A [99] 14) Placement with relatives [ ] Months N/A [999] 15) Number of relative placements [ ] # N/A [99] 16) Hospital or institution [ ] Months N/A [999] 17) Number of hospital/institution [ ] placements # N/A [99] 18) Other placement [ ] Months N/A [999] 19) Number of other placements [ ] # N/A [99] 20) Specify other placement 21) Months missing birth to adoption [ ] placement Months N/A [999] 174 Nature of Special Need As described at time of 1st placement Yes No Unknown N/A Missing 1 2 3 4 9 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Older than 2 years [ Mentally handicapped [ Member of sibling group [ If yes, total number of siblings [_ available for adoption at same time N/A [99 Non-caucasian [ If yes, specify ethnic origin (both parents if applicable) Uncertain prognosis [ Neonatal addiction syndrome [ Fetal alcohol syndrome [ Drug dependent at birth [_ Specific medical problem If yes, specify medical problem. Multiply handicapped [_] Other special need [_] If yes, specify nature of other special need. 175 Reason for entering care 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Yes No Unknown N/A Missing 1 2 3 4 9 Abandoned Physically abused Sexually abused Emotionally abused (If specifically mentioned in file] Parental alcohol abuse Parental drug abuse Relinquished by biological parent Parent deceased Chronic neglect Parent mentally ill Parent mentally handicapped Other reason for entering care Specify other reason Number of times in care prior to permanent committal (or adoption consent) Number of siblings 52) Age/gender of siblings N/A Gender Age [_J # [99] [ _ ] # [99] 176 Child's functioning as listed 1 - Average 2 - Mild to moderate disability 3 - Severely handicapped 53) Physical functioning [ ] 54) Intellectual functioning [ ] 55) Behavioural functioning [_] 56) Child continued to have contact with [_] biological family after permanent committal/adoption consent. 1 - Never 2 - Occasionally (birthdays/ holidays) 3 - At regular intervals less than twice/month 4 - Frequently more than 3 visits/month 5 - N/A 9 - Missing Post-placement variables 1) Length of time between study and [ ] placement Months N/A [999] 2) Type of child placed was as requested [_] 1 - Yes 2 - No 3 - Unknown 4 - N/A 9 - Missing 3) Comments regarding difference from file 4) Child requires treatment services (known [_] prior to placement) Code as above 5) Treatment services are available locally [_] Code as above 6) Child is of same ethnic background as parents Code as above 7) If different, specify child's background and adoptive parents'. 8) Child is medically different than described Code as above for all remaining questions 9) Child is emotionally different than described 10) Child is behaviourally different than described 11) Information was on file but not given to adoptive parents 12) New information surfaces after placement of child 13) Child was sexually abused 14) Child was physically abused 15) Child was neglected 16) Child has siblings elsewhere 17) Siblings placed together in this home 18) Siblings placed together elsewhere 19) Siblings placed apart 20) Regular contact with siblings maintained (if not placed in this home) 21) Adoption completed 22) Adoption probation extended 23) If Adoption probation extended, what were reasons ? 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0107111/manifest

Comment

Related Items