Guardian ad Litem and judicial decision-making in cases of child abuse and neglect

MISSING IMAGE

Material Information

Title:
Guardian ad Litem and judicial decision-making in cases of child abuse and neglect
Physical Description:
vii, 74 leaves : ; 29 cm.
Language:
English
Creator:
McDanal, Cynthia Sutton
Publication Date:

Subjects

Subjects / Keywords:
Legal Guardians   ( mesh )
Decision Making   ( mesh )
Child Abuse   ( mesh )
Jurisprudence   ( mesh )
Department of Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Health Related Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 51-53).
Statement of Responsibility:
by Cynthia Sutton McDanal.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002390141
notis - ALZ5024
oclc - 51314415
System ID:
AA00009050:00001

Full Text
















GUARDIAN AD LITEM AND JUDICIAL DECISION-MAKING
IN CASES OF CHILD ABUSE AND NEGLECT







By

CYNTHIA SUTTON MCDANAL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1994













ACKNOWLEDGEMENTS

The author extends appreciation to her parents, Mr. and Mrs. M. R. Sutton, and

her grandparents, Mr. and Mrs. D. A. Ball, for their continued patience and

encouragement. Particular gratitude is extended to the author's husband, Mr. Charles

McDanal, Jr., for his unconditional guidance and support.














TABLE OF CONTENTS

ACKNOWLEDGEMENTS ................................. ii

LIST OF TABLES ...................................... v

ABSTRACT .......................................... vi

INTRODUCTION ....................................... 1

O verview ........ ... ....... .. ... .... .. ... ... ..... 1
Definition of Abuse ................................. 3
Incidence and Prevalence ................................ 4
Role of Child Protection Agencies ......................... 6
The Guardian ad Litem Program .......................... 8

REVIEW OF LITERATURE ................................ 11

Theoretical Models of Decision-Making .................... 11
Empirical Studies of Decision-Making ..................... 15

RESEARCH OBJECTIVES AND HYPOTHESES ................... 20

M ETHOD ............................................ 22

Subjects .... .. .......... ... ...... .. ..... ..... .. 22
M measure ........................................ 22
Procedures ....................................... 25

RESULTS ............................................ 26

Comparison of Judges' and Guardians' Decisions .............. 26
Rate of Removal by Abuse Type ......................... 28
Logistic Regression Equations .......................... 30
Comparison of Cases of Physical Abuse Only and Neglect Only ...... 37

DISCUSSION ......................................... 41

REFERENCES ........................................ 51

iii














APPENDIX

APPENDIX


APPENDIX

APPENDIX


APPENDIX E


BIOGRAPHICAL


GAL QUESTIONNAIRE FOR DEPENDENCY CASES .

CORRELATIONS AMONG INDEPENDENT
VARIABLES ............................

RATE OF REMOVAL BY TYPE OF ABUSE ........

DESCRIPTIVE STATISTICS FOR VARIABLES IN
CASES OF PHYSICAL ABUSE ALONE AND NEGLECT
ALONE ...............................

PEARSON THREE-WEEK TEST-RETEST
CORRELATIONS .........................

SKETCH .................................













LIST OF TABLES
Table page

1 Comparison of Guardian Recommendations and Judicial
D decisions ..................................... 27

2 Results of Logistic Regression Analysis Using Guardian
Recommendations to Predict Judicial Decisions .............. 27

3 Variable Values In Cases Where Judges and Guardians Disagreed
Regarding Case Outcome ............................ 29

4 Results of Logistic Regression Analysis Using Demographics to
Predict Judicial Decisions . . . 32

5 Results of Logistic Regression Analysis Using Demographic
Variables to Predict Guardian Recommendations ............ 32

6 Results of Logistic Regression Analysis Using Legal and Case
Variables to Predict Judicial Decisions . . 34

7 Results of Logistic Regression Analysis Using Legal and Case
Variables to Predict Guardian Recommendations . 34

8 Results of Logistic Regression Analysis Using Family
Psychological Variables to Predict Judicial Decisions . 36

9 Results of Logistic Regression Analysis Using Family
Psychological Variables to Predict Guardian Recommendations 36

10 Results of Logistic Regression Analysis Using All Significant
Predictor Variables to Predict Judicial Decisions . 38

11 Results of Logistic Regression Analysis Using All Significant
Predictor Variables to Predict Guardian Recommendations ... 38

12 Distribution of Interval Variables . . 42

13 Frequencies of Dichotomous Variables . . 42













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

GUARDIAN AD LITEM AND JUDICIAL DECISION-MAKING
IN CASES OF CHILD ABUSE AND NEGLECT

By

Cynthia Sutton McDanal

April, 1994

Chairman: Jacquelin Goldman
Major Department: Clinical and Health Psychology

It has long been recognized that cases involving child abuse and neglect confront

the legal system with the difficult moral and practical dilemma of whether to remove

children from their parents' care. While broad guidelines are provided under federal and

state laws, there remains a great deal of ambiguity regarding appropriate factors to

consider when making such decisions. Several researchers have put forth rational models

to guide decision-making and a few have examined samples in attempts to determine

which factors are most relevant to court personnel. However, these studies generally

have not sought to generate predictive models for such decision-making, and they tend

to rely on examination of case records rather than direct report from those involved in

the decision-making process. The current study sought to extend our current knowledge

of the determinants of case disposition in cases of child abuse and neglect by considering

the influence of potential factors as reported by the child's court-appointed representative,

vi








or Guardian ad Litem. Guardian recommendations for placement were considered as

well as case adjudication to determine whether significant differences exist in the

decision-making processes of the two groups. Judicial decisions were also compared to

Guardian recommendations to determine the extent to which judges and Guardians

concur.

Results indicated that while judges are highly likely to concur with

recommendations of the Guardians, they tend to place a somewhat different emphasis on

the variables they use to determine whether to remove a child from his home. Judges

were found to rely most heavily on the presence of allegations of sexual abuse and on

the degree of parental warmth and support for the child. Guardians, however, were

found to rely most heavily on the presence of previous reports of abuse against the child.

Neither judges nor Guardians appeared to place heavy weight on much of the information

regarding family psychological functioning. Implications for this finding are discussed.


vii













INTRODUCTION

Overview

Recent estimates of the prevalence of child sexual abuse, physical abuse, and

neglect indicate that almost two million children per year become victims of some form

of maltreatment (National Center on Child Abuse and Neglect, 1988). Numerous studies

have documented the host of psychological and behavioral problems that often result from

such abuse (see Lamphear, 1985; Brown & Finkelhor, 1986; Houck & King, 1988, for

reviews). Such research clearly indicates that child abuse is a serious problem of wide

scope in our society.

The considerable task of assessing and intervening in cases of child abuse and

neglect has fallen to child protection agencies and the judicial system. To facilitate this

process, specific guidelines have been set, and Guardian ad Litem programs have been

established. Working within the guidelines, the Guardian ad Litem serves as a court-

appointed legal representative of the child. The Guardian gathers information concerning

the family and makes a recommendation to the court regarding appropriate placement of

the child and treatment for the family. It is then the responsibility of the judge to make

a final decision in the best interests of the child, the family, and society.

Such a process requires a great deal of subjective judgment, especially on the part

of the Guardian. Several researchers have put forth theoretical models to guide decision-

making in such situations (e.g., Wasserman & Rosenfeld, 1988; Faller, 1988), and other








2

researchers have sought to describe the decision-making process by considering social,

family, and child characteristics which are associated with case disposition (e.g., Katz,

Hampton, Newberger, Bowles, & Snyder, 1986; Finkelhor, 1983).

Although the approaches applied in these latter studies are more direct in that they

analyze actual placement decisions, the studies themselves are largely descriptive and

have limitations in methodology or scope. The present study seeks to extend our current

knowledge of the decision-making process involved in cases of child abuse and neglect

by identifying factors which influence the Guardian's recommendation and the final

adjudication of the case. A large number of potential determinants will be considered.

Guardians and judges will be evaluated separately in order to determine whether

differences exist in the decision-making processes of the two groups. The results of such

analyses can then be compared to theoretical models of decision-making in order to

evaluate the appropriateness and effectiveness of the process.

Historical Background

The problem of child abuse and neglect is not a new phenomenon. On the

contrary, children have been mistreated by adults since the beginning of civilization

(Radbill, 1968). Practices such as harsh discipline, child labor, incest, ritual sacrifice,

and infanticide attest to the extent to which child abuse is deeply rooted in cultural and

religious history (Eskin, 1980). The historical view of children as "property" of their

parents and the generally accepted concept that family matters were private contributed

to such a situation.








3

It was not until the late nineteenth century that society began to shift its view of

children as property to include the concept of children's rights. The landmark case of

"Mary Ellen" in 1870 marked the first time that child abuse was recognized by a U.S.

court. The court intervened by removing Mary Ellen from her home and by prosecuting

her custodian. This case demonstrated the need for a child protection agency, and

shortly thereafter, the New York Society for the Prevention of Cruelty to Children was

chartered by the New York legislature (Landau, 1982). This agency served as the

prototype for many that followed.

Definition of Abuse

While the above case brought about a recognition that parental freedom is not

limitless and that children have a legal right to be protected from cruelty and neglect,

there has been no clear consensus of exactly what constitutes child abuse. Wasserman

and Rosenfeld (1986) note that early definitions of child abuse generally included only

the most severe cases, while recent definitions have been expanded to include any form

of corporal punishment. Emotional abuse has recently been recognized in the literature,

but it has also proved difficult to define precisely. State statutes and mandatory reporting

laws defining child abuse typically refer to harm or threatened harm to a child's physical

or mental well-being, but they do not provide specific guidelines.

Sloan (1983) has put forth a definition of abuse which is operationally useful. He

divides child abuse and neglect into four categories: physical abuse, neglect, sexual

abuse, and emotional maltreatment. Physical abuse of children refers to any non-

accidental injury caused by the child's caretaker. Neglect involves inattention to the








4
basic needs of the child, including food, shelter, clothing, medical care, and supervision.

Sexual abuse includes any interaction between a child and an adult in which the child is

being used for the sexual stimulation of the adult. Finally, emotional maltreatment

includes such behaviors as blaming, belittling, or rejecting a child; unequal treatment of

siblings; or lack of concern for the child's welfare. Such definitions provide more

specific criteria for determining the occurrence of abuse or neglect, and are frequently

used by child protection workers.

Incidence and Prevalence

The exact scope of child maltreatment in this country is subject to wide debate.

One problem in determining the incidence of child abuse is the lack of a standard

definition regarding exactly what constitutes abuse and neglect. A second problem is that

different research designs yield different estimates of child abuse. Daro (1988) notes that

interviews with a random sample of households or individuals consistently project higher

incidence and prevalence rates than indicated by formal reporting data.

The National Center on Child Abuse and Neglect (1988) has responded to such

difficulties in estimation of the incidence of child abuse by recognizing five "levels" of

abused and neglected children. Level 1 includes all abused and neglected children known

to state and local Child Protective Services (CPS) agencies. At Level 2 are those

children who are not known to CPS, but who are known to other investigative agencies,

such as police, courts, or health departments. Level 3 includes children who are not

known to any Level 1 or 2 agency, but who are known to professionals in other

community institutions, such as hospitals, schools, daycare centers, and mental health








5

clinics. The abused and neglected children at Level 4 are known only to individuals

outside of Levels 1, 2, and 3, such as neighbors or other family members. Finally,

Level 5 includes those children who have not been recognized by anyone as abused or

neglected. Although this description of all five levels is valuable in that it formally

recognizes that unreported cases of abuse do exist, the Center's incidence study includes

cases only at Levels 1 through 3.

The results of this study indicated that 10.7 children per 1,000 were physically

abused in the United States during 1988, or a total of 675,000 children. Rates of neglect

were even higher 15.9 children per 1,000, or a total of over one million children

nationwide. The rate of all forms of child sexual abuse was estimated to be 2.5 per

1,000, or a total of 159,000 children.

As mentioned above, random sample interview studies generally yield higher

estimates of the prevalence of child abuse, probably because they tap unreported abuse

as well as abuse known to public agencies and community professionals. For example,

Strauss and Gelles (1986) conducted a national study on the incidence and prevalence of

family violence by telephone interviewing a probability sample of 6,002 households.

Child abuse data were based on the 1,428 households in which there was a child between

the ages of 3 and 17 years, and in which two caretakers were present. The results of this

study indicated that 19 children per 1,000 were subject to at least one incident of "very

severe violence" during 1985.

Random samples of the prevalence of sexual abuse also yield significantly higher

estimates than the Center's study. Russell (1983) surveyed a probability sample of 930








6

women residents of San Francisco about any experience of sexual abuse they might have

had. Her findings indicated that 38% of the women had experienced at least one incident

of sexual abuse before the age of 18 years. Finkelhor (1984) reported that 6% of all

males and 14% of all females in his random sample had experienced sexual abuse before

the age of 16 by a person at least five years older.

Regardless of the specific definition or research methodology utilized, virtually

all estimates of the incidence and prevalence of child abuse and neglect indicate that it

is an increasingly prevalent problem in our society.

Role of Child Protection Agencies

Since the case of Mary Ellen, it has been the responsibility of the court and child

protection agencies to assess and intervene in cases of child abuse and neglect.

Considering the amount of abuse and neglect that has been documented, the lack of a

standard definition of abuse, and the precarious balance between individual rights

protected under the fourteenth amendment and the need to protect children, this is an

imposing responsibility.

In order to carry out the tasks of assessment and intervention in a fair and

effective manner, specific guidelines for child protective services have been put forth.

Although the specific details of such guidelines differ by state, the general principles of

the process remain the same. Landau (1982) describes six major aspects of the child

protective service process:

The referral process. All 50 states now have mandatory reporting laws designed

to encourage the reporting of suspected abuse and neglect. Most reporting systems also








7
include provisions which encourage investigation, intervention, and treatment services.

Many states have also set up central abuse registries.

Definitions. As mentioned above, each state has a slightly different definition of

what constitutes child abuse and neglect, and all state definitions are rather broad.

However, differences are minimal, and the use of a broad definition provides the

practitioner with leeway in decision-making. Most states distinguish between abuse and

neglect, and therefore define them separately.

Critical elements of a child protection system. Landau notes that there are a

number of criteria for a child protective agency to be an effective system. Practitioners

must have accurate knowledge of the incidence of child abuse and neglect. The system

must have a strong and well-publicized mandatory reporting law and an effective central

abuse registry. A specially trained Child Protection Unit must be available at all times.

The court system must be capable of dealing fairly and effectively with families, and

treatment and rehabilitative programs should be available for those families. Finally,

interdisciplinary communication and a system of coordination should be used to

implement the network of services for families.

Assessment of cases. A specially trained child protection worker must

systematically gather information regarding the child, the family, and the circumstances

of their lives. The worker must then use this information to determine whether abuse

has occurred; if so, to determine why it occurred; and to identify appropriate

interventions.








8

Initiation of the legal process. It is the responsibility of the child protection

worker to decide whether to initiate court proceedings. Guidelines for the consideration

of court action include situations in which (a) the child is determined to be in imminent

danger, (b) attempts at treatment have failed, and/or (c) the family refuses to cooperate

with the investigation. Caseworkers must also consider issues regarding the nature of

evidence and the possible negative consequences of bringing a case into court.

Emergency removal. All states have statutes which allow for the emergency

removal of abused or neglected children from the home, even prior to the filing of a

petition. A court order may or may not be necessary. All states have specific criteria

regarding the conditions necessary to justify emergency removal, and most have specific

requirements describing what must be done following the removal.

Caseworker-attorney interchange. Landau (1982) also stresses the importance of

the establishment of rapport between the child protection worker and the agency attorney

as they work to prepare cases. Clear guidelines regarding decision-making are

beneficial.

The Guardian ad Litem Program

Almost all states have implemented Guardian ad Litem or similar programs to

carry out the above functions, and those that have not require that legal counsel be

appointed to represent the child in abuse and neglect hearings (Sloan, 1983). A Guardian

ad Litem is a court-appointed legal representative who participates on behalf of children,

who are legally unable to initiate or function in litigation without adult assistance. The

function of Guardians is to identify the child's best interests in the litigation, and they








9
are legally obligated to do everything within their power to represent those interests

(Whitcomb, 1988).

Wiig (1982) notes that although appointment of a Guardian ad Litem is a

condition of the federal Child Abuse Prevention and Treatment Act (PL93-247), the Act

does not define the functions of a Guardian, distinguish the Guardian's role from that of

the child's social worker and the child's attorney, or clearly define the relationship of the

Guardian to the court. She conducted a limited study of six Guardians in an attempt to

clarify these roles.

Wiig's findings led her to conclude that the following ten activities might be

defined as the functions of the Guardian because they were distinguished from the actions

of the child's lawyer and social worker in the cases she examined: (1) to function as a

mediator in an attempt to achieve consensus among parties prior to court proceedings;

(2) to coordinate between juvenile court and other court departments; (3) to provide

continuity across the entire legal process; (4) to persuade the social worker to consider

alternative placements, (5) to minimize trauma to the child and to provide continuity of

relationship with the child: (6) to insure court attention to cases by placing them on the

calendar; (7) to make recommendations in the best interests of the child; (8) to identify

special needs of the child and to insure that the child receives attention to those needs;

(9) to prevent continuances not in the best interests of the child or to request an attorney

to do so; and (10) to request independent counsel for the child (pp. 93-94). Wiig further

notes that while the attorney or social worker might in some cases fulfill some of these








10
roles, the functions numbered 3, 5, and 8 are distinct and may be considered strictly

functions of the Guardian ad Litem.

While it is generally accepted that the Guardian fulfills these functions and

ultimately makes a recommendation to the court regarding disposition of the case, the

process by which the Guardian arrives at the recommendation is less clear. Several

researchers have noted the problems inherent in such a process, and have put forth

theoretical models of decision-making to guide practitioners. These models are discussed

in detail below.













REVIEW OF LITERATURE

Theoretical Models of Decision-Making

Wasserman and Rosenfeld (1986) have noted several difficulties facing judges who

must decide when to intervene in cases of possible abuse or neglect, and these difficulties

also apply to Guardians as they struggle to make the most appropriate recommendation

based on the information they have. Such problems include the aforementioned lack of

clarity in the definition of abuse, difficulties in proving abuse, limited alternatives

available other than foster care, problems in the foster care system, and the large number

of children in abusive and neglectful situations.

Wasserman and Rosenfeld (1986) also note the positive and negative aspects of

separation that must be considered when deciding whether to remove a child from his

home. Separation provides for the safety of the child, sets limits on parental behavior,

and lets the child know that society will intervene to protect him if his parents do not.

It may also provide an opportunity for the child to live in a safer, warmer environment.

Separation also frees the parents from the burdens of childrearing, thereby giving them

the opportunity to make constructive changes in their situation.

Potential negative consequences of separation must also be carefully considered.

Even though they may be abused, most children have an affectional bond with their

parents and see that relationship as a source of identity and well-being. Thus, even when

faced with abuse, many children express a desire to remain with their families. Also,

11








12

feelings of inadequacy and failure in both parents and children may follow separation.

The authors state that children may even perceive themselves as responsible for the

separation.

In order to make appropriate decisions given such a difficult situation, Wasserman

and Rosenfeld (1986) recommend that practitioners consider the nature and quality of the

parent-child bond. They state, "A judge who is provided with careful evaluation of the

parent-child relationship is in a better position to decide wisely" (p. 524). In addition

to interviewing all individuals involved in the case, the authors recommend that the

caseworker visit the family's home. The caseworker should note issues such as how well

the parent knows the child, how able and willing the parent is to care for the child, how

reasonable the parent's beliefs about discipline are, and how much nurturance the parent

provides the child. Once the caseworker has obtained this information, he or she can

make a reasonable evaluation of the relationship between parent and child. According

to Wasserman and Rosenfeld, this evaluation can then be used to make the most

appropriate decision regarding disposition of the case.

While the characteristics of the family relationships likely have legal and clinical

utility for decision-making, an approach considering only this factor seems overly

simplistic. Based on clinical experience with over 300 cases of child sexual abuse, Faller

(1988) has proposed a model of decision-making in cases of intrafamilial child sexual

abuse which is more comprehensive than the approach described above. Like

Wasserman and Rosenfeld (1986), she believes that the parent-child relationship is an

important consideration, especially the relationship between mother and child. However,








13

Faller also believes that other parental characteristics are crucial to decisions about

intervention.

These characteristics differ between mothers and fathers. According to Faller's

theory, important maternal factors are (1) the level of the mother's dependency,

especially on the father-perpetrator; (2) the degree of love and nurturance she conveys

to the victim; and (3) the degree to which she responded protectively when she

discovered the sexual abuse. Important paternal characteristics are (1) his overall level

of functioning; (2) his superego functioning; and (3) the severity of the sexual abuse.

In order to use such information to choose intervention strategies, Faller has

identified four types of cases based on different combinations of the above characteristics.

Type I cases are ones in which both parents have many strengths. In such cases Faller

recommends family therapy, without court involvement. In Type II cases, the mother

has many positive characteristics, but the father exhibits many areas of problematic

behavior. For these cases, Faller suggests maintaining the mother and child intact and

referring them for treatment but advises that the father-perpetrator be excluded from the

family. Legal intervention may be necessary to remove the father from the home.

In Type III cases, mothers are dependent, hostile toward their children, and

unprotective. Although the father has sexually abused the child, he functions much better

than the mother and he often has a closer relationship with the child than does the

mother. According to Faller, this is the most difficult type of case for which to

determine appropriate intervention. She recommends therapy for all family members for

three to six months, at which time careful re-evaluation should be conducted. During








14

this time period, it may be optimal to place the child with a relative, which allows for

less disruption in the child's life while still protecting him or her from further sexual

abuse.

In Type IV cases, the mother is similar to the Type III mother, and the father

exhibits poor overall functioning and shows no guilt over the (usually) extensive abuse.

Faller recommends that parental rights be terminated in such cases, and that the children

receive treatment for at least one year. Criminal prosecution may also be warranted for

the perpetrator.

Although Faller's approach to decision-making is comprehensive and specific, it

does have several limitations. First, her approach assumes the availability of alternative

placements for the child (i.e., placement with a relative or for adoption), which may not

always be readily available. Second, some possible combinations of parental

characteristics do not match any of the four case types described. Decision-making for

these families is therefore more ambiguous. Third, this theoretical model has been

proposed for cases of father-child sexual abuse only. It is not clear whether such a

model might be adapted to cases of physical abuse or neglect, or to cases of sexual abuse

by a non-father perpetrator. Finally, though her model is based on clinical experience,

it has yet to be empirically tested.

Several researchers have conducted analogue studies with child protection workers

or have analyzed actual case disposition data in an attempt to empirically describe aspects

of the decision making-process in cases of child physical abuse or neglect. Such cases

are reviewed below.








15

Empirical Studies of Decision-Making

Meddin (1984) conducted an analogue study of criteria used by child protection

workers to determine placement decisions in cases of abuse and/or neglect. She surveyed

81 child welfare workers using a questionnaire which described six situations containing

two instances each of physical abuse, sexual abuse, and neglect. Respondents were

provided with information on eight variables: age, gender, and race of the child; the

child's experience of previous placements; the severity of the current incident; risk to the

child of further abuse and/or neglect; and the cooperation and functioning of the primary

caretaker. For each situation, respondents indicated whether they would place the child

in foster care; they then described what variables influenced their decisions.

Results of Meddin's study indicated that the primary variable used to determine

placement of the child was risk of further abuse and/or neglect. Such a finding seems

of dubious value, given that caseworkers are rarely, if ever, provided with that

information directly. More often, assessment of risk to the child is a part of the

caseworker's decision-making process.

Meddin did, however, note that four other variables clustered around risk to the

child, and she suggested that these variables might serve as operational definitions or

indicators of risk. The four variables were severity of the current incident, functioning

and cooperation of the primary caretaker, and age of the child.

While Meddin's study is useful in that it indicates several variables which seem

to be important in caseworkers' placement decisions, it does have several limitations.

In actual case situations, child protection workers typically have a great deal more








16

information than the eight variables Meddin provided. It is impossible to evaluate the

importance of this additional information using the analogue research design employed

here. Additionally, there is the possibility that the caseworkers' actual decisions differ

from what they report in a hypothetical situation.

Several empirical studies of actual case dispositions have been conducted in an

attempt to overcome such limitations. Runyan, Gould, Trost, and Loda (1981) examined

the records of the North Carolina Registry of Child Abuse and Neglect to determine

which social, family, and child characteristics were most influential in the decision to

place a child in foster care. According to Runyan et al., these records contained all

theoretically relevant factors as well as demographic information.

Results of the study indicated that several child and family characteristics put

children at increased risk for placement. These variables included parents who perceive

severe punishment as acceptable, parents with substance abuse problems, severe child

injuries, child abandonment, and several other specific types of abuse. Age of the child

or parent, the presence of several maltreated children in one family, and several other

specific types of abuse were found to be not significant. Race, income, and education

were also nonsignificant, suggesting that there is no systematic bias in placement

decisions. Two non-family variables were also found to be associated with a higher

chance of placement-cases referred by courts or police and geographic area.

Runyan et al. also employed a logistic regression model to predict placement in

foster care given all the above variables. They then compared this to actual placement.

Results of this procedure did not strongly associate the above characteristics with removal








17

from the home. In fact, the variance explained by the model was only 0.168. Runyan

et al. interpreted this result to mean that assignment to foster care approximates a random

process across a large population.

An alternative explanation for such a finding may be that child protection workers

use different criteria for making placement decisions for different types of abuse cases

(i.e., sexual abuse, physical abuse, or neglect). Since these data were collapsed across

type of abuse, such differences may have canceled each other out.

Katz, Hampton, Newberger, Bowles, and Snyder (1986) examined factors that

influence the decision to remove children from their parents' care in cases of abuse

(physical or sexual) versus neglect (nonphysical injury). These researchers examined the

records of 185 suspected abused or neglected children at Children's Hospital, Boston, in

an attempt to reveal how demographic characteristics, family history, family stress, the

nature of the injury, and aspects of the medical encounter influence the outcome of the

case.

Interestingly, the results of the study indicated that children with nonphysical

injuries were more likely to be removed from their parents' care than were those with

physical injuries. The authors suggested several possible explanations for this finding,

including the possibility that neglect may be perceived by clinicians as evidence of a

chronic family problem rather than a single mishap.

Katz et al. also conducted Chi-square analyses to test the influence of other

independent variables on placement, although these analyses were collapsed across cases

of abuse and neglect. Their findings indicated that the following factors predicted








18

removal from the home: Medicaid-eligible families, previous reports of suspected

maltreatment, and suspicion of the mother being involved in the maltreatment. Minority

families, families in which the child was under age six, and families that were

experiencing more than one stressor were more likely to have their children returned.

Surprisingly, severity of condition was not significantly associated with outcome.

The results of this study yielded several interesting and even counterintuitive

findings. Again, however, because different categories of abuse were not considered

separately, the possibility of overgeneralization exists.

Finkelhor (1983) analyzed data on 6,096 cases of sexual abuse which were

officially reported in the U.S. in 1978 in an attempt to evaluate how the cases were

handled. His specific aim was to determine whether children were being unnecessarily

removed from their homes. Such an approach seems useful in that his consideration of

sexual abuse cases alone minimizes the risk of overgeneralization or of having significant

effects washed out by differences among groups.

Finkelhor compared the overall placement rate of sexual abuse cases to the

placement rate of physical abuse cases in the National Clearinghouse sample. His

findings indicated that foster care placement occurs significantly more often in cases of

sexual abuse, which led him to hypothesize that sexual abuse results in more outrage than

does physical abuse, regardless of the degree of danger to the child.

Finkelhor's main finding of the study was that foster care does not seem to be

used indiscriminantly in cases of sexual abuse. For example, the most important

predictor of foster care placement was whether or not the victim initially reported the








19

abuse. A preference to be placed outside the home was more likely to be met than a

preference to stay. Also, foster care is more likely to be used in the case of an older

child than a younger child. This seems reasonable given that older children are typically

better able to adjust to living without their families.

Other factors predisposing to placement in foster care included abuse committed

by parents (especially if both parents were involved), very large families, the presence

of other types of abuse or neglect, and the presence of several other family stressors,

such as spouse abuse, substance abuse, family discord, and mental health problems.

Such findings led Finkelhor to conclude that "the characteristics of the children who are

placed in foster care do suggest some rational relationship between the disposition and

the facts of the case" (p. 199).

Finkelhor's data also suggest that foster care is not used indiscriminantly in

"vulnerable" families. For example, he found that foster care placement is not more

likely to occur with poor families. Further, black children were not more likely to be

removed from their homes than were white children Thus, according to Finkelhor's

findings, there does not seem to be any systematic bias in the decision-making process.

The strengths of Finkelhor's study include its large N, its consideration of a large

number of potentially relevant factors, and its statistical analysis of only one type of

abuse. This methodology does not, however, indicate whether any of these factors

cluster together to form a meaningful aggregate of predictive characteristics. Also,

similar information regarding cases of child physical abuse and of neglect is needed.













RESEARCH OBJECTIVES AND HYPOTHESES

The current study seeks to extend our knowledge of the determinants of case

disposition by considering the influence of potential factors in cases of physical abuse and

in cases of neglect. Guardian ad Litem recommendations for placement will be

considered as well as final judicial decisions, in order to determine whether significant

differences exist in the decision-making processes of the two groups. Judicial decisions

will also be compared to Guardian recommendations to determine the extent to which

judges rely on those recommendations.

Based on review of the available literature, it is hypothesized that several groups

of variables will predict Guardians' and judges' decisions to place the child outside the

home in cases of child physical abuse and neglect. Among demographic variables, it

is hypothesized that increased age of the child will predict placement outside of the

home, but that family race and income will not be related to case disposition.

A second category of variables hypothesized to be related to case disposition

includes those which are descriptive of the abuse situation. Specifically, it is proposed

that abuse perpetrated by parents, very severe abuse, the presence of other types of

abuse, and previous reports of abuse will predict placement of the child outside the

home.

A third category of variables includes those that describe the psychological state

of the family. It is specifically hypothesized that poorer parent psychological








21
functioning, high degree of parental noncompliance with the investigation, the presence

of spouse abuse and substance abuse, high degree of family discord, and low degree of

parental warmth and support for the child will all predict placement outside the home.













METHOD

Subjects

Subjects for this study included 94 families involved in child physical abuse or

neglect cases in the state of Florida. All families were selected from adjudicated cases

in 12 judicial circuits. Only one abused or neglected child from each family was

considered in this study in order to avoid misleading duplication of family and case

variables. Child age ranged from 1 year to 17 years (mean =4 years, 9 months). Family

income ranged from $0 to $65,000 per year (mean=$8,661). Fifty-nine percent of the

families were white, 34% were black, 4% were Hispanic, and 3% were other races.

Measure

Data for this study were collected using a questionnaire designed by the team of

investigators to sample demographic, legal, and psychological factors relevant to the

family (see Appendix A). Demographic and legal information was of public record and

included the age, sex, and race of each family member; who the alleged perpetrator was;

family income; occupation and education level of each parent; whether or not the abuse

was substantiated; past history of abuse; whether criminal charges were brought against

the perpetrator; whether mental health evaluation was recommended and/or completed

for each family member; which placement options were recommended by the Guardian;

and which placement options were adjudicated. Placement options included the

following: (1) perpetrator removed from the home, (2) parents retain custody with

22








23
supervision, (3) temporary custody awarded to a relative, (4) temporary foster placement,

(5) child placement in a treatment facility, (6) placement for adoption, (7) child placed

in a detention facility, and (8) not dependent and returned home. A category designated

"other" was also provided in case final adjudication included a placement not described

above. For the purposes of this study, these options were collapsed to "returned home"

(options 1, 2, or 8) or "removed from the home" (options 3 through 7).

Psychological information for each case was provided by the Guardian, who made

judgments based on his or her interviews with the family. Judgments were made on

visual analog scales on the following issues: (1) who the non-abusing caretaker supports;

(2) severity of the abuse/neglect; (3) child's distress over the current family situation; (4)

conflict between the parents; (5) parents' willingness to report family problems in

general; (6) amount the family favors family activities; (7) amount the family members

clarify and/or speak for other family members; (8) degree to which the physical needs

of the child were met (and who they were met by); and (9) degree to which the emotional

needs of the child were met (and whom they were met by).

Guardians also rated the following issues for each parent: (1) conflict between

parent and child; (2) warm and positive feelings for child; (3) involvement in child's

daily life; (4) reliance on the child; (5) psychological issues (loneliness, depression,

anxiety, irritability); (6) satisfaction with social support; and (7) response to the Guardian

in general. Guardians also indicated on visual analog scales their confidence in each of

their judgments.








24
Additional information was requested regarding whether the parents were abused

or neglected as children, whether the parents are accused of substance abuse and/or

spouse abuse, whether the parents have lost jobs or moved within the past year, whether

the parent or child suffers a mental or physical handicap, and whether the child exhibits

any behavioral problems.

Reliability. Guardian's judgments on the psychological issues were compared to

those of licensed clinical psychologists and law students (Goldman et al., in preparation).

A videotape of a Guardian interviewing expert Guardians who were acting as family

members involved in a child abuse scenario was viewed by 22 Guardians, 17 clinical

psychologists, and 15 law students. The script for the mock interview was written by

the investigators in conjunction with two Guardian supervisors so that it would be

representative of an authentic family situation. After viewing the videotape, all subjects

completed the portion of the research questionnaire which involved psychological issues.

No significant differences were found between the Guardians', psychologists', and

lawyers' judgments on any of the psychological issues, indicating that no specific bias

could be attributed to the subjects' training or their area of expertise.

Test-retest reliability was also calculated for each of the 29 items rated. The 22

Guardians viewed the mock interview twice, at three week intervals. Of the variables

used in this study, Pearson correlations were significant for three variables at the p< .05

level (level of conflict, maternal and paternal warmth for child), for one variable at the

p<.01 level (maternal noncompliance), and for one variable at the p=.001 level

(severity of abuse/neglect). Pearson correlations were found to be non-significant for








25
paternal noncompliance and parental psychological functioning (depression, anxiety,

irritability, and loneliness). It should be noted, however, that the videotape viewed by

the Guardians contained relatively little information regarding the variables found to have

non-significant correlations, and it is therefore hypothesized that the low correlations

were due more to Guardians' lack of confidence in their ratings than to instability in their

judgments over time.

Procedures

Data for each family were collected by the Guardian ad Litem (GAL) assigned

to that case. Each judicial circuit and each Guardian participated in this study on a

volunteer basis. Guardians are assigned cases on an individual basis by the GAL director

of each circuit. Each Guardian is appointed to a family when the case first becomes

active, and that Guardian represents the child until the case is adjudicated. For the

purpose of this study, Guardians collected data on each family when the case reached the

stage of adjudication. Thus, the cases used in this study represent those that were not

settled out of court and therefore may represent the most severe cases.













RESULTS

Comparison of Judges' and Guardians' Decisions

The cross-tabular display provided in Table 1 shows the frequency with which

judges agreed with the recommendation of the Guardians ad Litem. Results indicate that

the child was removed from the home by the judge in 78 cases and returned home in 16

cases. Guardians recommended removal from the home in 69 cases and recommended

returning the child to the home in 25 cases. Table 1 also shows that the judge agreed

with the Guardian's recommendation (to remove the child from the home or to return the

child home) in 79 of the 94 cases. There were only 12 cases in which the judge-but

not the Guardian-felt that the child should be removed, and only 3 cases in which the

Guardian-but not the judge-felt that the child should be removed. The Kappa statistic

(K) was calculated to determine the degree of concurrence observed between judges and

Guardians, and the obtained value of .53 indicates good overall agreement (beyond

chance) between the two groups. It can therefore be concluded that while Guardians

appear to be slightly more conservative in their recommendations to remove children

from their homes, there was considerable agreement between the two groups regarding

placement of the child.

To further assess the relationship between Guardian recommendations and judicial

decisions, a logistic regression analysis was conducted using Guardian recommendation

as a predictor variable for judicial decision (See Table 2). Results indicated that the

26










Table 1. Comparison of Guardian Recommendations and Judicial
Decisions
Guardian Recommendation
Judicial Decision
Removed Not Removed Total
Removed 66 cases 12 cases 78 cases
(70%) (13%) (83%)
Not Removed 3 cases 13 cases 16 cases
(3%) (14%) (17%)
Total 69 cases 25 cases 94 cases
(73%) (27%) (100%)
K=.53


Table 2. Results of Logistic Regression Analysis Using Guardian
Recommendations to Predict Judicial Decisions

Predictor Predicting Judicial Decisions
Variable DF Parameter Standard Chi- Probability of >
Estimate Error square Chi-square
Intercept 1 0.08 0.40 0.04 0.84
Guardian 1 -3.17 0.71 19.77 0.00'
Recommendation


p=.0001


Hosmer-Lemeshow goodness-of-fit statistic=6.24 with 8 DF, p=.62








28
Guardian's recommendation for removal is significantly associated (p=.0001) with a

higher probability of judicial decision to remove the child from the home. Predicted

probabilities based on the parameter estimate obtained for the Guardian recommendation

variable indicate that the probability of removal is 96% in cases where the Guardian

recommends it and 48% in cases where the Guardian recommends against it. It should

be noted again, however, that these data are based on only 25 cases in which the

guardian recommended returning the child to the home. The Hosmer and Lemeshow

goodness-of-fit statistic for the model (6.24 with 8 DF, p=.62) indicates that the

observed responses do not significantly differ from responses which would be predicted

based on this model (Hosmer & Lemeshow, 1989).

The cases in which the judge disagreed with the Guardian's recommendation for

placement of the child were examined further to determine whether any trends among the

predictor variables could be identified. Examination of the values of the predictor

variables in those cases, however, did not reveal any differences which might account

for the difference of opinion between the judges and the Guardians (see Table 3).

Rate of Removal by Abuse Type

It was also of interest to compare the rate of removal depending on the type of

abuse or neglect allegations present. The overall distribution of abuse types in this

sample (whether removed or not removed) indicated that 80% of the cases involved

neglect, 52% involved physical abuse, and 28% involved sexual abuse. Among the cases

in which removal occurred, 83% involved allegations of neglect, 50% involved

allegations of physical abuse, and 24% involved allegations of sexual abuse (see










Table 3. Variable Values In Cases Where Judges and Guardians Disagreed
Regarding Case Outcome
J G C F R P S M P S P W C N S S
U U H A A A E T R E S A 0 0 U P
D A 1 M C R V Y E X Y R N N B A
G R A I E P E P V A F M F C A B
E D G N E R E R B U L O B U
I E C R I S E U N M U S
A P T P S C P S E
N Y E T E
R N W Y 79 N N N 37 5 42 N N
R N 2 0 W Y 93 Y N Y 51 27 44 54 Y N
R N 12 W Y 84 Y N Y 65 85 20 37 Y N
R N 3 0 N Y 97 N N N 69 68 2 27 N N
R N 3 0 N Y 91 N Y N 96 13 65 96 Y N
R N 5 0 N Y 64 N Y N 25 47 16 5 Y N
R N 1 8 W Y 81 Y N N 59 14 54 64 N N
N R 6 8 W Y 89 Y Y Y 46 50 2 47 N N
R N 2 0 N Y 86 N N N 68 46 48 Y N
N R 1 N Y 82 N N N 99 0 0 N N
R N 2 W Y 69 Y N N 63 47 45 21 N Y
R N 3 N Y 94 N N N 88 21 18 Y N
R N 17 W Y 85 Y N N 52 57 94 15 N N
R N 16 W N 95 N N Y 19 90 14 N N
N R 4 20 N Y 60 N N N 54 81 78 33 Y N
"." represents missing data; "Y" = YES; "N" = NO
JUDGE= Adjudication (R=Child removed from home; N=Not removed)
GUARDIAN= Recommendation (R=Recommended removal; N=Did not)
CH1AGE= Child age (in years)
FAMINC= Annual family income (in thousands of dollars)
RACE= Family race (W=White; N=Non-White)
PARPERP= Was the alleged perpetrator a parent?
SEVERITY= Severity of abuse (99= Highest possible severity level)
MTYPES = Were multiple types of abuse alleged?
PREVREP= Were there previous reports of abuse?
SEXABUSE= Was sexual abuse alleged?
PSYFUNCT= Parental psychological functioning (99=Poorest possible level)
WARM= Parental warmth/support for child (99=Highest possible level)
CONFL= Family conflict (99= Highest possible level)
NONCOMP= Parental noncompliance (99= Highest possible level)
SUBABUSE= Was parental substance abuse alleged?
SPABUSE= Was spouse abuse alleged?








30
Appendix C). Results also indicated that children were removed from their homes in

87% of the cases involving allegations of neglect, in 80% of the cases involving

allegations of physical abuse, and in 73 % of the cases involving allegations of sexual

abuse. While these results are complicated by the fact that multiple types of abuse

occurred in 45% of the cases, they nonetheless indicate removal was most likely when

allegations of neglect were present and least likely when allegations of sexual abuse were

present.

Logistic Regression Equations

The primary purpose of this study was to determine which demographic variables,

abusive situation characteristics, and family psychological characteristics influence

Guardians' and judges' decisions to remove the child from the home in cases of child

abuse and/or neglect. Because the outcome variables (the judge's placement of the child

or the Guardian's recommendation for placement) and several predictor variables are

dichotomous, logistic regression was used. In order to test for the possibility of

misleading multicollinearity among the independent variables, correlation matrices were

obtained for each model. The correlations among the independent variables ranged from

.01 to .50, indicating that the results of the analyses were not compromised by significant

multicollinearity.

Two series (one for judges' decisions and one for Guardians' recommendations)

of three logistic regression models were run-one model for each category of variables

(demographic variables, abusive situation characteristics, and family psychological

characteristics). The results of each of these models will be discussed separately.








31
Finally, two logistic regression equations including only the significant predictor

variables from the above models were tested-one using judges' decisions as a dependent

variable and including all independent variables found to significantly predict those

decisions and a second using Guardians' decisions as a dependent variable and including

all independent variables found to significantly predict their decisions.

Demographic Variables

Because information regarding family income was provided in only 54 of the 94

cases, it was not possible to test the impact of this variable. Results of the logistic

regression equation including child age and family race as predictors indicated that

neither variable significantly impacted the decision to place the child outside the home

(see Table 4). The analysis using the Guardian's recommendation as the dependent

variable yielded similar results, as neither child age nor family race emerged as a

significant predictor of a Guardian's decision to recommend placement outside of the

home (see Table 5).

Abusive Situation Characteristics

Among variables related to aspects of the abusive or neglectful situation, it was

hypothesized that abuse perpetrated by parents, very severe abuse, the presence of

multiple types of abuse, and previous reports of abuse would predict placement outside

the home. It was not possible to test the influence of the "parent perpetrator" variable,

however, as 96% of the cases in this sample involved parent perpetrators. When the

remaining variables were included in the logistic regression equation using the judge's










Table 4. Results of Logistic Regression Analysis Using Demographics to Predict
Judicial Decisions

Predicting Judicial Decisions
Demographic Variables DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -1.88 0.55 11.90 0.00
Child Age 1 0.00 0.06 0.00 0.96
Family Race 1 0.38 0.59 0.41 0.52





Table 5. Results of Logistic Regression Analysis Using Demographic Variables to
Predict Guardian Recommendations

Predicting Guardian Recommendations
Demographic Variables DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -1.50 0.47 10.36 0.00
Child Age 1 0.03 0.06 0.36 0.55
Family Race 1 0.34 0.51 0.44 0.51








33
decision as a dependent variable, none emerged as significantly related to removal of the

child from the parents' home (see Table 6). In the analysis using the Guardian's

recommendation as a dependent variable, only the presence of previous reports of abuse

(p < .05) emerged as significantly related to recommendation for placement outside the

home (see Table 7). In this model, the negative parameter estimate indicates that the

absence of previous reports of abuse or neglect (which was treated first in the model) is

associated with a smaller probability of removing the child from the home. Guardians

were not more likely to recommend removal in cases in which the abuse was very severe

or in which multiple types of abuse occurred.

Family Psychological Characteristics

Several variables describing the psychological state of the family were

hypothesized to affect judges' decision-making in cases of abuse or neglect. Specifically,

it was proposed that poorer parent psychological functioning, high degree of parental

noncompliance with the investigation, the presence of spouse abuse and substance abuse,

high degree of family discord, and low degree of parental warmth and support for the

child would predict placement of the child outside of the home. However, only the

degree of the parents' warmth and support for the child (p < .01) emerged as significantly

related to judges' decision-making when the variables were simultaneously entered into

the prediction equation (see Table 8). The positive parameter estimate indicates that

lower levels of parental warmth and support (which were treated first in the model) were

associated with a larger probability of having the child removed from the home.










Table 6. Results of Logistic Regression Analysis Using Legal and Case Variables to
Predict Judicial Decisions

Predicting Judicial Decisions
Legal and Case Variables F Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -0.47 0.89 0.27 0.60
Severity of Abuse 1 -0.15 0.01 1.46 0.23
Multiple Abuse Types 1 0.19 0.59 0.10 0.75
Previous Reports 1 -0.25 0.57 0.19 0.66


Table 7. Results of Logistic Regression Analysis Using Legal and Case Variables to
Predict Guardian Recommendations

Predicting Guardian Recommendations
Legal and Case Variables F Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -0.88 0.89 0.97 0.32
Severity of Abuse 1 -0.00 0.01 0.09 0.76
Multiple Abuse Types 1 0.15 0.51 0.09 0.77
Previous Reports 1 -1.08 0.51 4.46 0.03*


Sp < .05








35
The results of the analysis using the Guardian's recommendation as an

independent variable yielded similar results (see Table 9). Like judges, Guardians were

more likely to recommend removal of children from homes in which parents showed

little warmth and support for them (p< .05). Interestingly, however, none of the other

variables regarding family psychological functioning significantly impacted the

Guardians' decision-making. It therefore appears that neither judges nor Guardians rely

heavily on information regarding the psychological functioning of the family when

making decisions regarding the placement of the abused child.

Models Including All Significant Predictors

In order to determine the best overall predictive model of judicial decision-

making, a logistic regression equation was run which included only the predictor variable

found to significantly impact judicial decisions from the above models-level of parental

warmth and support. Again, the positive parameter estimate obtained for the predictor

variable in this model indicates that lower degrees of warmth and support were more

associated with removal from the home. Calculation of predicted probabilities based on

the obtained parameter estimates indicated that the probability of removal was 98% in

cases where parental warmth and support was rated 0 (the lowest possible level) and 50%

in cases where parental warmth and support was rated 99 (the highest possible level).

The Hosmer-Lemeshow goodness-of-fit statistic (7.90 with 8 DF, p=.44) indicates that

the observed responses do not differ significantly from those which would occur based

on the predicted probabilities from the model (see Table 10).










Table 8. Results of Logistic Regression Analysis Using Family Psychological
Variables to Predict Judicial Decisions

Family Psychological Predicting Judicial Decisions
Family Psychological ----------------
Variables DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -4.79 1.76 7.41 0.01
Psychological Functioning 1 0.01 0.02 0.11 0.74
Parental Warmth and Support 1 0.04 0.02 8.14 0.00
for Child
Family Conflict 1 0.00 0.01 0.00 0.96
Parental Noncompliance 1 0.02 0.01 2.32 0.13
Substance Abuse 1 -0.17 0.71 0.06 0.81
Spouse Abuse 1 -0.78 0.87 0.81 0.37

*p < .01



Table 9. Results of Logistic Regression Analysis Using Family Psychological
Variables to Predict Guardian Recommendations

Family Psychological Predicting Guardian Recommendations
Family Psychological ----------------
Variables DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -3.18 1.39 5.21 0.02
Psychological Functioning 1 0.02 0.02 1.18 0.28
Parental Warmth and Support 1 0.03 0.01 493 0.03*
for Child
Family Conflict 1 -0.00 0.01 0.06 0.80
Parental Noncompliance 1 0.01 0.01 0.99 0.32
Substance Abuse 1 -0.42 0.60 0.49 0.49
Spouse Abuse 1 -0.80 0.72 1.23 0.27


p < .05








37
A similar analysis was conducted using the independent variables found to

significantly impact the decisions of the Guardians. When the variables of previous

reports of abuse and parental warmth and support for the child were entered into the

equation, only the presence of previous reports of abuse (p=.05) emerged as strongly

related to Guardian's decisions. Since inclusion of the variable quantifying parental

warmth and support for the child did not significantly add to the predictive power of the

equation, the analysis was conducted again using only the presence of previous reports

as a predictor variable. The negative parameter estimate indicates that the presence of

previous reports of abuse was associated with a higher probability of making a

recommendation for removal. The Hosmer-Lemeshow goodness-of-fit statistic (9.00 with

8 DF, p=.34) indicates that the observed responses do not significantly differ from those

which would occur given the predicted probabilities based on this model (see Table 11).

Comparison of Cases of Physical Abuse Only and Neglect Only

An original aim of this study was to determine whether judicial and Guardian

decision-making processes differed between cases of physical abuse alone and neglect

alone. Because the sample included only 11 cases solely involving physical abuse and

36 cases solely involving neglect, it was not possible to analyze the data using logistic

regression equations-which require a larger sample size to be interpretively useful.

Analysis using the Shapiro-Wilk statistic (W) indicated that all interval independent

variables were normally distributed in the samples. The means of these variables were

compared using t-tests in an attempt to determine whether potentially important

differences exist among demographic, legal, or family psychological characteristics










Table 10. Results of Logistic Regression Analysis Using All Significant Predictor
Variables to Predict Judicial Decisions

Predicting Judicial Decisions
All Significant Predictors
DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -3.73 0.86 18.87 0.00
Parental Warmth and
SuParenta l Warmth and 1 0.04 0.01 9.33 0.00'
Support for Child

*p < .01

Hosmer-Lemeshow goodness-of-fit statistic=7.9 with 8 DF, p=.44






Table 11. Results of Logistic Regression Analysis Using All Significant Predictor
Variables to Predict Guardian Recommendations

Predicting Guardian Recommendations
All Significant Predictors
DF Parameter Standard Chi- Probability of >
Estimate Error Square Chi-Square
Intercept 1 -0.57 0.30 3.50 0.06
Previous Reports of Abuse 1 -1.02 0.49 4.25 0.04*


* p < .05


Hosmer-Lemeshow goodness-of-fit statistic=9.00 with 8 DF, p=.34








39
between the two groups (see Appendix D). In addition, the frequencies of the

dichotomous independent and dependent variables in each group were reviewed, although

the number of cases was insufficient to support Chi-square analysis.

Results of comparisons between cases of physical abuse alone and neglect alone

using t-tests indicated that the level of severity of neglect tended to be more extreme than

the level of severity of abuse (t=-2.55, p=.02) and that there tended to be less conflict

between parents of neglected children than between parents of abused children (t=2.09,

p=.04). No significant differences between the two groups were found in child age,

family income, level of parental psychological functioning, degree of parental warmth

and support for the child, or parental noncompliance with the investigation.

Visual review of the frequencies of dichotomous variables in the two groups

revealed the following: (1) Judges were somewhat more likely to remove neglected

children from their homes (89% of neglected children as opposed to 73% of abused

children), (2) Guardians were less likely to recommend removal of neglected children

than abused children (75% of neglected children as opposed to 82% of abused children),

(3) Among neglect cases, there tended to be fewer cases involving previous reports of

abuse or neglect against the child (47% as opposed to 57% in cases of physical abuse),

(4) In cases of neglect alone, there were fewer cases involving spouse abuse (14% as

opposed to 27% in families of abused children), and (5) Reports of substance abuse were

more frequent among parents of neglected children (72% as opposed to 55% in cases of

physical abuse). Family race and the number of non-parent alleged perpetrators did not

differ significantly between the two groups.








40
Because these findings are based on very small samples, caution must be used in

their interpretation. The number and magnitude of the differences found, however,

indicate the possibility of substantive differences between the characteristics of cases of

physical abuse only and neglect only. It thus logically follows that important differences

may exist in the process of determining appropriate adjudication for the two different

groups.













DISCUSSION

The results of this study provide descriptive information about cases involving

allegations of child abuse and/or neglect cases which are served by the GAL program in

the state of Florida. Such descriptive information about the cases included in the current

study is provided to establish a framework for interpreting and understanding the results

which were obtained. The distribution of demographic, abusive situation, and family

psychological characteristics of the sample are presented in Tables 12 and 13. As noted

above, 59% of the families in the current study were white and 41% were African

American, Hispanic, and other races (34%, 4%, and 3%, respectively). Of the 56 cases

in which information on family income was provided, the mean income was $8,6660 per

year, suggesting that the majority of the families live in poverty.

All interval variables related to the abusive situation and to the family's

psychological functioning in the study were rated on visual analog scales with values

ranging from 1 to 99. The average level of severity of abuse was 76 (with 99 being the

most severe), indicating that these cases generally represent the most severe incidents of

abuse and neglect. At least one of the parents was the alleged perpetrator in 96% of the

cases, and there existed previous reports of abuse in 50% of the cases. Multiple types

of abuse were reported in 45% of the cases, and sexual abuse allegations were present

(either alone or in combination with abuse and/or neglect) in 28% of the cases.










Table
Variable

Child Age
Family Income
Severity of Abuse
Parental Psychological
Functioning
Parental Warmth and
Support for Child
Family Conflict
Parental Noncompliance


12. Distribution
N

90
56
94


of Interval Variables
Mean S.D.

4.77 4.46
8.66 12.61
76.05 21.85


56.04

47.33

46.53
37.12


20.98

29.42

31.15
26.20


Variable

Family Race

Parent Perpet


Multiple Typ

Previous Repi

Sexual Abuse

Spouse Abuse

Substance Ab


Table 13. Frequencies of Dichotomous Variables
Frequency
White 55
Non-White 39
Alleged 90
Not Alleged 4

es of Abuse Alleged 42
Not Alleged 52
Yes 47
orts of Abuse N 47
No 47
Alleged 26
Not Alleged 68
Alleged 20
Not Alleged 74

use Alleged 61
Not Alleged 33


Median

3.00
5.50
84.00


54.25

48.50

47.00
33.25


Percent
58
42
96
4
45
55
50
50
28
72
21
79
65
35








43
There were allegations of spouse abuse (in addition to child abuse) in 21% of the

cases, and allegations of substance abuse were made in 65% of the cases. The average

rating for overall psychological functioning of the family was 56 (with 99 being the

poorest possible level of psychological functioning), and the average rating for parental

warmth and support for the child was 47 (with 99 being the highest possible level of

warmth and support). The average conflict level reported for the families was 47 (with

99 being the highest possible level of conflict), and the average degree of parental

noncompliance with the investigation was 37 (with 99 being the highest possible level of

noncompliance).

One of the major aims of this study was to determine the frequency with which

judges agree with Guardians ad Litem regarding case outcome. Cross-tabular display of

Guardian recommendation and actual case outcome indicated that the two groups agreed

in 79 of the 94 cases, and the positive Kappa statistic indicated that this level of

agreement is beyond what would be expected by chance alone. In addition, results of

a logistic regression analysis indicated that Guardian recommendation for removal is

significantly associated with a higher probability of judicial decision to remove the child

from the home. The results indicate, in fact, that the probability of a judicial decision

for removal is 96% when the Guardian has recommended that the child be removed.

The main objective of the current study was to determine which demographic,

legal, and family psychological variables are most predictive of decisions to remove (or

recommend removal of) children from their homes in cases involving child abuse and

neglect. It is important at this point to note again that all cases included in this study had








44
reached the point of adjudication and therefore likely represent the most severe cases to

enter into the judicial system. Consideration of this population is felt to be important,

however, as it is these cases in which judges are most likely to render decisions.

Interestingly, very few of the variables assessed were found to be significantly associated

with judicial and Guardian decision-making.

In addition, while judges were quite likely to agree with the recommendations of

Guardians, their decisions were significantly associated with somewhat different

information from the Guardians. Review of the cases in which judges and Guardians

disagreed did not reveal any trends which would account for such differences. It is

therefore possible that this discrepancy is due either to chance or to the influence of

another factor (or factors) which was not accounted for in this study.

Overall, information regarding the degree of parental warmth and support for the

child was most predictive of final case outcome. The results of the separate analyses

using Guardians' recommendations as the dependent variable indicated that the presence

of previous reports of abuse and on the degree of parental warmth and support for the

child are significantly associated with those recommendations. The results of a logistic

regression equation using only these predictor variables, however, indicated that only the

presence of previous reports of abuse or neglect against the child is significantly related

to Guardians' recommendations for removal. Thus, at least in this sample, recurrent

abuse against a child is the single most predictive factor of Guardians' recommendation

that he or she be removed from the home.








45

The results of previous studies have indicated that judges are more likely to

remove older children from the home than they are to remove younger children (e.g.,

Katz et al., 1986). In this sample, however, child age had no discernible effect on either

judges' or Guardians' decisions. Previous studies have also found that the race of the

family is not significantly associated with decisions of judges. The results of the current

study lend support to that finding, suggesting that racial discrimination on the part of

court personnel is not present in the process of adjudicating cases of child abuse and

neglect.

Several legal variables hypothesized to be related to judicial and Guardian ad

Litem decision-making were not found to significantly predict the outcome of the cases

included in this study. Neither judges' nor Guardians' decisions were significantly

associated with the fact that the abuse was very severe or on the fact that multiple types

of abuse were alleged. Interestingly, while the presence of previous reports of abuse was

found to be the most significant predictor of Guardian's decisions to recommend removal

of children from their homes, this information was not significantly related to judicial

decision-making. Among the variables relating to the psychological functioning of the

family, only the level of the parents' warmth and support for the child was found to be

significantly associated with decisions regarding case outcome. Information regarding

the psychological functioning of the parents, the parents' compliance with the

investigation, the degree of family discord, or the presence of allegations of spouse or

substance abuse among the parents was not related to judges' or Guardians' decisions.








46

One source of concern which arises from the results of the current study is the

lack of association between judges' and Guardians' decisions and the severity level of the

alleged abuse. Florida Statute section 39.464 stipulates that grounds for termination of

parental rights include "egregious abuse," which is defined as "conduct that endangers

the life, health, or safety of the child." Yet visual analysis of the data indicated that a

number of children were returned home in cases where the severity of the abuse was

judged to be very high. If one assumes that children who are very severely abused are

at risk for greater injury in the future (if not at greater risk for more frequent future

abuse), this finding reflects a potentially serious oversight on the part of court personnel

involved in dependency case decision-making. In addition, there was little association

between judges' decisions and the presence or absence of previous reports of abuse or

neglect against the child---despite the fact that Florida Statute section 39.464 stipulates

that "continuing abuse or neglect" is grounds for termination of parental rights.

Another area of concern is the lack of association between judicial and Guardian

decisions and information regarding the psychological and interpersonal functioning of

the family. The results obtained here differ from those obtained by Meddin (1984) in

her study of child protective direct service workers and supervisors. Subjects in that

study were found to treat the functioning and cooperation of the parents as important

indicators of future risk to the child. Despite considerable evidence that psychological

disturbance of parents (particularly depression), aberrant parenting, and marital discord

frequently lead to abuse and neglect of children and subsequently to poor adjustment of








47

those children (e.g., Lynch and Roberts, 1982; Martin, 1979; Green, 1976), factors such

as these were not found to predict decisions to remove children from their homes.

Review of Florida Statute section 39.467 provides a likely explanation for the

above finding. Judges are required to consider "the love, affection, and other emotional

ties existing between the child and the child's present parent or siblings"-which they

appear to do, given the significant association between the degree of parental warmth and

support for the child and final adjudication. However, while judges are instructed to

consider "the capacity of the parent or parents to care for the child to the extent that the

child's health and well-being will not be endangered upon the child's return home," there

is no specific requirement to consider the psychological functioning of the parents.

Again, given our current knowledge about the detrimental effects of poor parent

psychological functioning on the well-being of children, failure to consider these factors

when determining whether to remove children from their homes would represent a gross

oversight.

Results of comparison of the variables among groups of children who were

physically abused (but not neglected or sexually abused) and those who were neglected

(but not physically or sexually abused) also indicated several potentially important

differences in both case type and outcome-most notably that while Guardians are

somewhat less likely to recommend removal of children in cases of neglect alone, judges

are somewhat more likely to adjudicate removal. Such a finding raises questions

regarding whether the decision-making processes of judges and Guardians differ when

allegations include different types or combinations of abuse or neglect. As a result, it








48

is recommended that future research consider groups of physically abused, sexually

abuse, and neglected children separately rather than risk "washing out" of potential

differences among them.

The current study has identified a number of variables important to judges and

Guardians in determining custody of abused and/or neglected children, as well as a

number of issues which may warrant more attention on the parts of these two groups.

There are, however, several methodological limitations present in this study which

deserve consideration. First, there are limitations which concern the generalizability of

the results obtained here. Despite considerable effort, the researchers were unable to

collect data on every case referred to the GAL program during the period of study.

Some of the judicial circuits in the state of Florida were more willing to cooperate than

were others, and even the most cooperative were able to provide data only from

Guardians willing to complete our forms on a volunteer basis. Our sample is therefore

not random and may not accurately represent the population of GAL referrals in the state

of Florida. Additionally, the current results are based on a relatively small sample, and

some of our findings may therefore be sample specific. Caution must thus be used when

generalizing the results obtained in this study to other populations. In addition, the

sample used in this study did not include a wide distribution of family income levels or

child ages. Because the sample included mostly very low income families with young

children, generalization of these findings to populations with different demographic

characteristics is not advised, as judges and Guardians may well weigh information

differently in cases involving wealthy families or those with older children.








49

While federal laws regarding cases involving allegations of child abuse or neglect

provide some degree of uniformity in determining case disposition, it may be that some

states have legal guidelines and political climates which differ from the state of Florida.

It is therefore recommended that the results obtained here not be generalized to other

geographic regions prior to further study in those areas. Another aspect of the current

sample which deserves consideration is that the cases involved both initial dependency

hearings and termination of parental rights hearings. Thus, some of the children were

removed from their homes on a temporary basis, while others were permanently

removed. Because court personnel may use somewhat different criteria when deciding

whether parental rights should be severed, it is recommended that additional research be

conducted on these two groups independently of each other.

Future research is clearly needed to assess the generalizability of the current

findings to other populations. It is also recommended that a similar analysis be

conducted using a larger sample of a comparable population in order to replicate these

findings and to determine whether the process of Guardian ad Litem and judicial

decision-making changes over time.

Further research is also needed to clarify the results obtained in the current study.

One rather puzzling finding of this study was that while judges tend to agree with the

recommendations of Guardians, they appear to place a somewhat different emphasis on

the variables they use to determine the final adjudication on the cases they consider.

This finding suggests that while judges rely on case information provided to them by the

Guardians, they may not rely heavily on the Guardians' recommendations. Future








50

research using Guardian recommendation as an independent variable to predict judicial

decisions is thus warranted to further clarify this situation. In addition, it is

recommended that future studies be conducted which consider abused children and

neglected children as separate groups in order to determine whether important differences

exist in the decision-making of judges and Guardians between the two groups.

Another area requiring further research is the lack of consideration given to

psychological aspects of the family in making decisions regarding removing children

from their homes. Potential explanations for this finding include the possibility that

judges and Guardians are unaware of the importance of such factors on the short- and

long-term well-being of the children involved, that they are not adequately educated

about how to assess the information, and/or that they do not feel that the consideration

of such information is warranted under the legal guidelines for determining case

disposition. Future research aimed at elucidating these or other reasons is therefore

suggested.













REFERENCES


Alter, C.F. (1985). Decision-making factors in cases of child neglect. Child Welfare,
64, 99-111.

Brown, A., & Finkelhor, D. (1986). Impact of child sexual abuse: A review of the
research. Psychological Bulletin, 99, 66-77.

Daro, D. (1988). Confronting child abuse: Research for effective program design.
New York: The Free Press.

Eckenrode, J., Powers, J., Doris, J., Munsch, J., & Bolger, N. (1988). Substantiation
of child abuse and neglect reports. Journal of Consulting and Clinical
Psychology, 56, 9-16.

Eskin, M. (1980). Child abuse and neglect: A literature review and selected biblio-
graphy. Washington, DC: U.S. Department of Justice.

Faller, K.C. (1988). Decision-making in cases of intrafamilial child sexual abuse.
American Journal of Orthopsychiatry, 58, 121-128.

Finkelhor, D. (1983). Removing the child-prosecuting the offender in cases of sex-
ual abuse: Evidence from the National Reporting System for Child Abuse and
Neglect. Child Abuse and Neglect, 7, 195-205.

Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York:
Free Press.

Goldman, J., Albanese, I., Graves, L., McDanal, C.S., Sorenson, E., & Ward, M. (in
preparation). Reliability of Guardian ad Litem ratings on psychological issues in
cases of child abuse and neglect.

Green, A. (1976). A psychodynamic approach to the study and treatment of child-
abusing parents. Journal of Child Psychiatry, 15, 414-429.

Houck, G.M., & King, M.C. (1989). Child maltreatment: Family characteristics and
developmental consequences. Issues in Mental Health Nursing, 10, 193-208.









Hosmer, D.W, & Lemeshow, S. (1989). Applied Logistic Regression. New York:
John Wiley & Sons.

Katz, M.H., Hampton, R.L., Newberger, E.H., Bowles, R.T., & Snyder, J.C. (1986).
Returning children home: Clinical decision-making in cases of child abuse and
neglect. American Journal of Orthopsychiatry, 56, 253-262.

Knudsen, D.D. (1988). Child protective services: Discretion, decisions, dilemmas.
Springfield, IL: Charles C. Thomas.

Lamphear, V.S. (1985). The impact of maltreatment on children's psychosocial
adjustment: A review of the research. Child Abuse and Neglect, 9, 251-263.

Landau, H.R. (1982). The abused and neglected child. Practising Law Institute.

Lynch, M.A., & Roberts, J. (1982). Consequences of Child Abuse. London, England:
Academic Press.

Martin, H.P. (1979). Child abuse and child development. Child Abuse and Neglect,
3, 415-421.

Meddin, B.J. (1984). Criteria for placement decisions in protective services. Child
Welfare, 63, 367-373.

National Center on Child Abuse and Neglect (1988). Study findings: Study of na-
tional incidence and prevalence of child abuse and neglect. Washington, DC:
U.S. Department of Health and Human Services.

Radbill, S.X. (1968). A history of child abuse and infanticide. In B.E. Helfer & C.H.
Kempe (eds.), The battered child. Chicago: University of Chicago Press.

Runyan, D.K., Gould, C.L., Trost, D.C., & Loda, F.A. (1981). Determinants of foster
care placement for the maltreated child. American Journal of Public Health, 71,
706-711.

Russell, D.E.H. (1983). The incidence and prevalence of intrafamilial and extra-
familial sexual abuse of female children. Child Abuse and Neglect, 7, 133-146.

Sloan, I.J. (1983). Child abuse: Governing law and legislation. New York: Oceana
Publications, Inc.

Strauss, M.A., & Gelles, R.J. (1986). Societal change and change in family violence
from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and
the Family, 48, 465-479.








53

Wasserman, S., & Rosenfeld, A. (1986). Decision-making in child abuse and neglect.
Child Welfare, 65, 515-529.

Whitcomb, D. (1988). Guardians Ad Litem in criminal courts. Washington: U.S.
Department of Justice.

Wiig, J.K. (1982). Functions of the Guardian ad Litem in child abuse and neglect
proceedings-Los Angeles Juvenile Court. In H. Landau (ed.), The abused and
neglected child. Practising Law Institute.













APPENDIX A


GAL QUESTIONNAIRE FOR DEPENDENCY CASES

Thank you for your time and support of our efforts to collect the following
information on dependency cases. Our goal is to provide information which will aid
the GAL program in its role as the advocate for children in dependency cases.

The questionnaire is divided into three broad sections. This first section deals with
basic demographic and legal information of your case. This information is similar to
that found in your case records.

The second sections requests information on individual and family psychological
issues. These questions may be easily answered after your interviews with the child's
family. This section requires that you mark on a line to indicate your rating of the
intrafamilial relationships. Some of these questions require a single mark on the line,
such as for conflict between the two parents:

CONFLICT BETWEEN THE TWO PARENTS:

I -----------------------------------/ -------------------
Very Little Very Much

While other questions require you to make two ratings on each line, one for the
mother-child relationship and one for the father-child relationship:

CONFLICT BETWEEN EACH PARENT AND THE CHILD(ren) who was
reportedly abused or neglected, with an M, F, SM, and an SF.
M F
I -----------------/ ------- /---------------------
Very Little Very Much

Finally, the third section surveys information from the disposition hearing on the case
in terms of your recommendations to the court and the court's final determination.

Please return this form when completed to your GAL program's secretary.

Thank you again for your help.










Dependency Case

DEMOGRAPHIC AND LEGAL INFORMATION


Who is filling out this form?

Date of this hearing?_

Circuit:_


Guardian Coordinator/Director

Judge Code:


County:


Circle any immediate family members who have not been interviewed by the
GAL: (Circle all that apply)


Mother Father Child

Stepmother Stepfather


Are the parents represented by counsel:


Other (specify)

Grandparent


Yes


Report for all family members living with the child victim (and the alleged
perpetrator if he/she does not live with the child) the following information:
Race: C=caucasian, H= hispanic, B=black, O=other
V/P: V=victim of abuse/neglect, P=alleged perpetrator
Relation: (to the victims) P=parent, SP=step parent, GP= grandparent, S=sibling,
BF= boy/girlfriend of the parent, FP= foster parent, AP=adoptive parent,
ST=stranger, O =other


AGE SEX RACE


V/P RELATION


For each parent report the following information:
Yearly Years of
Income Education


MOTHER
FATHER
STEPMOTHER
STEPFATHER


Age


Other


Occupation








56
Has one or more of the parents lost their jobs within the last year? Yes No

Has the family moved within the last year? Yes No

Circle the one category into which the reported child victim falls:

Born Prematurely Physically Handicapped Mentally Retarded

Neither Unknown

Which parents) may have difficulty caring for their children) due to a physical
handicap, serious illness, or injury?

Mother Stepmother Father Stepfather Neither

For the following questions, place a check mark in the box corresponding to any
alleged or substantiated problem for any parentss:
ALLEGATIONS SUBSTANTIATED
M S M F S F NMS M F SF N
OT O A TAE T O A T A E
T ET T E T I T E T T E T I
H PH H PH T H PH H PH T
E E E E H E E E E H
R R R R E R R R R E
R R
Alcohol Abuse
Drug Abuse
Hitting the Child
Child Physical Abuse
Child Sexual Abuse
Child Emotional Abuse
Child Neglect
Spousal Emotional
Abuse
Spousal Physical Abuse
Previous Reports of
Child Abuse or Neglect










INTERPERSONAL AND FAMILY ISSUES: Please evaluate items 1-9 from your
interviews with the family members. Place a mark on the line to indicate your rating
of the family. Also, please circle the number of each question if you have included
information on that issue in your report to the court.

1. Who does the non-abusing caretaker SUPPORT:


Child Victim
Child Victim


-------Reported Abuser
Reported Abuser


CONFIDENCE in your rating.


Total


None


2. SEVERITY of the abuse/neglect:


Very Slight
Very Slight


---Very Severe
Very Severe


CONFIDENCE in your rating.


None


Total


3. CHILD'S DISTRESS over the current family situation:

Very Little-
Very Little


Very Much


CONFIDENCE in your rating.


None


Total


4. CONFLICT BETWEEN THE TWO PARENTS:

Very Little-
Very Little


Very Much


CONFIDENCE in your rating.

None Total
None Total


----------------------------------------------------------------------------


-----------------------------------------------------------------


! - - - - - - - - - -


----------------------------------------------------------------------------


_!


.I


I








58

5. PARENTS' WILLINGNESS TO REPORT FAMILY PROBLEMS in
general:
-------------------------------------------------------------------------I
None Completely

CONFIDENCE in your rating.

I--------------------------------------------------------------------------I
None Total

6. Amount the FAMILY FAVORS FAMILY ACTIVITIES:

- -- ---- ---- --.. -. --..-.------------------------
Too Little About Right Too Much

CONFIDENCE in your rating.

i-------------------------------------------------------------------------I
None Total

7. Amount that the family members CLARIFY AND/OR SPEAK FOR other
family members.
i-------------------------------------------------------------------------I
Very Little Very Much

CONFIDENCE in your rating.

i -------------------------.. --.. -------------.....................I
None Total

8. DEGREE TO WHICH THE PHYSICAL NEEDS OF THE CHILD(ren)
were met, prior to the recent report of abuse/neglect.

i ----..---------. ----------------------------............. I
None Total

CONFIDENCE in your rating.

i ------------------------------ ----------........ ............................
None Total

These needs were primarily met by (circle one):

Mother Father Both Other (specify)










9. DEGREE TO WHICH THE EMOTIONAL NEEDS OF THE CHILD(ren)
were met, prior to the recent report of abuse/neglect.


None


Total


CONFIDENCE in your rating.


Total


None


These needs were primarily met by (circle one):


Mother Father Both


Other (specify)


For items 10-16, use the following letters that apply to this family to represent your
evaluation of the parents on these issues. Place a mark on the line to represent your
rating of each parent and label each mark with one of the following letters. Please
make sure that you make separate ratings on each of them, including your ratings of
the CONFIDENCE you have in your rating on each of them.

M=Mother, SM= Stepmother, F=Father, SF=Stepfather

10. CONFLICT BETWEEN EACH PARENT AND THE CHILD(ren) who
was reportedly abused or neglected, with an M, F, SM, and SF.

-----------------------------------------------------------I
Very Little Very Much

For each parent, is there as much conflict with other children?


Yes No


Father:


Stepfather:


Yes No


Mother:


Stepmother:


CONFIDENCE in each of your ratings with an M, F, SM, and SF.

None ------Total------------ ---------------i
None Total


----------------------------------------------------------------------------


---------------------------------------------------------------------------


!
!


_1








60

11. WARM AND POSITIVE FEELINGS FOR THE CHILD(ren) who was
reportedly abused or neglected held by each parent with an M, F, SM,
and SF.


Very Little


Very Much
Very Much


CONFIDENCE in each of your ratings with an M, F, SM, and SF.


None


Total


12. INVOLVEMENT of each parent with the child(ren)'s daily lives (e.g., peer
relationships, child's feelings, minor illnesses) with an M, F, SM, and SF.


No Involvement


Excessive Involvement


CONFIDENCE in each of your ratings with an M, F, SM, and SF.

None Total
None Total


13. EACH PARENT'S RELIANCE ON THE CHILDREN FOR
COMPANIONSHIP, COMFORT, AND PROBLEM-SOLVING
F, SM, and SF.


with an M,


I---------------------------------------------------------------------
Very Little Very Much

CONFIDENCE in each of your ratings with an M, F, SM, and SF.

-------------------------------------------------------------I
None Total


14. PSYCHOLOGICAL ISSUES for each parent with an M, F, SM, and SF.

Lonely
Slightly Highly
Slightly Highly


----------------------------------------------------------------


! ----------------------------------------------------------------------------


i ----------------------------------------------------------------------------


I










Depressed
.. .. .... .. .. .. .... .... ... ... .. .. ... .. .. .. .. .. .. .... .... .... .. ... .. ...-


Slightly


Anxious


Slightly


Irritable


Highly



Highly


-------Highly
Highly


Slightly


CONFIDENCE in each of your ratings with an M, F, SM, and SF.

I --------------------------------- -.....--- .-.--- ..-............. ..........- !


None


Total


15. SATISFACTION WITH SOCIAL SUPPORT that each parent receives for
any problems from relatives, friends, coworkers, and community agencies
with an M, F, SM, and SF.


Inadequate


Excellent


CONFIDENCE in each of your ratings with an M, F, SM, and SF.

i--------.----.--------------------------------------------------------..........................................................-
None Total

16. EACH PARENT'S RESPONSE TO YOU in general with an M, F, SM,
and SF.

i-------------------------------------------------------------------------
Very Friendly Hostile/Rejecting

CONFIDENCE in each of your ratings with an M, F, SM, and SF.

i -----------------------------------------------------------I
None Total

17. If either parent was abused/neglected as a child, report the types of that
abuse (P=physical, S=sexual, E=emotional, N=Neglect, and N/A=none)


Mother Father_


i --- --- -- -- -- ------ ---- ------ -- -- -- -- -- --- --- -- -- -- -- --- -- -- -- -- -- ----- -


I


Stepmother


Stepfather










18. Check the following problems if any child was reported to exhibit the
problem:

Child verbally fights with parents

Child physically fights with parents

Child physically fights with siblings

Child verbally or physically fights with non-family members

Child is more promiscuous than would be expected for
his/her age

Child has been absent from school for more than 2 weeks
in the last year, and it has not been due to any
medical illness

Child has run away from home frequently in the past

Child steals from family members or others

Child is withdrawn and does not participate in normal daily
activities, such as schoolwork or family activities

Child refuses to do what his parents tell him to do,
such as going to school or doing household chores

PLEASE CHECK TO MAKE SURE THAT YOU HAVE CIRCLED THE
NUMBER OF THE QUESTIONS WHICH CONTAIN INFORMATION YOU
REPORTED TO THE COURT

Report the total number and dates of any past report of abuse or neglect and for
the present report. (If no abuse write in N/A; if the neglect was chronic, use C in
place of the number of occurrences.)

DATES TOTAL NUMBER

PAST


PRESENT








63
Report whether criminal charges are being brought against the alleged
perpetrator for abuse/neglect in this case and previously.

This case: YES NO Previously (how many):

DISPOSITION

Does the alleged perpetrator accept responsibility to the abuse/neglect?

YES NO

Indicate who requested a mental health evaluation for any family members, and
who completed it:
REQUESTED BY:(l =judge 2=Guardian 3=Attorney for Parent 4=HRS 5=N/A)
COMPLETED BY:(1 =Clinical Psychologist 2= Psychiatrist 3= Social Worker
4= Other Psychologist 5= Other 6= Not Completed)
Requested By Completed By
For the Mother
For the Father
For the Stepmother
For the Stepfather
For the Child(ren)

Circle any family members) for whom the GAL has recommended therapy:

Mother Stepmother Father Stepfather Children None

Do the parents admit dependency? YES NO

Please circle type of case:

Initial Dependency Termination of Parental Rights








64


Check any of the following options it they were being considered, then check
which option was recommended by the Guardian, and which option was
adjudicated.


Perpetrator Removed From the
Home
Parents Retain Custody With
Supervision
Temporary Custody Awarded to
a Relative
Temporary Foster Placement
Child Placed in a Treatment
Facility
Placement for Adoption
Child Placed in a Detention
Facility
Other (specify)_
Not Dependent and Returned
Home


Considered


Guardian


Adjudicated













APPENDIX B


CORRELATIONS AMONG INDEPENDENT VARIABLES


Table B-1. Estimated Correlations Among Demographic Variables Entered
in Logistic Regression Equation Predicting Judicial Decisions
Child Age Family Race
Child Age 1.00 0.16
Family Race 0.16 1.00


Table B-2. Estimated Correlations Among Demographic Variables Entered
in Logistic Regression Equation Predicting Guardian Recommendations
Child Age Family Race
Child Age 1.00 0.18
Family Race 0.18 1.00










Table B-3. Estimated Correlations Among Legal and Case
Variables Entered in Logistic Regression Equation Predicting
Judicial Decisions
Severity of Multiple Types Prior Reports
Abuse of Abuse of Abuse
Severity of Abuse 1.00 -0.25 -0.06
Multiple Types of -0.25 1.00 -0.21
Abuse
Prior Reports of -0.06 -0.21 1.00
Abuse


Table B-4. Estimated Correlations Among Legal and Case
Variables Entered in Logistic Regression Equation Predicting
Guardian Recommendations
Severity of Multiple Types Prior Reports
Abuse of Abuse of Abuse
Severity of Abuse 1.00 -0.26 -0.12
Multiple Types of -0.26 1.00 -0.18
Abuse
Prior Reports of -0.12 -0.18 1.00
Abuse











Table B-5. Estimated Correlations Among Family Psychological Variable Entered in
Logistic Regression Equation Predicting Judicial Decisions


Psychological
Functioning


Psychological
Functioning
Parental
Warmth
Family
Conflict
Parental Non-
compliance
Substance
Abuse
Spouse Abuse


1.00

0.09

0.09

-0.23


Parental
Warmth


0.09

1.00

0.33

0.43


Family
Conflict


0.09

0.33

1.00

-0.01


-0.20 -0.16 -0.24


-0.02


0.07


-0.07


Parental
Non-
compliance
-0.23

0.43

-0.01

1.00

0.04


-0.01


Table B-6. Estimated Correlations Among Family Psychological Variable Entered in
Logistic Regression Equation Predicting Guardian Recommendations

Psychological Parental Family Parental Substance Spouse
Functioning Warmth Conflict Non- Abuse Abuse
compliance
Psychological 1.00 0.17 0.03 -0.20 -0.23 -0.01
Functioning
Parental 0.17 1.00 0.29 0.39 -0.09 -0.01
Warmth
Family 0.03 0.29 1.00 -0.06 -0.18 -0.11
Conflict
Parental Non- -0.20 0.39 -0.06 1.00 0.14 -0.01
compliance
Substance -0.23 -0.09 -0.18 0.14 1.00 -0.17
Abuse
Spouse Abuse -0.01 -0.01 -0.11 -0.01 -0.17 1.00


Substance
Abuse


-0.20

-0.16

-0.24

0.04

1.00

-0.20


Spouse
Abuse


-0.22

0.07

-0.07

-0.01

-0.20

1.00













Table B-7. Estimated Correlations Among Significant Predictor Variables
Entered in Logistic Regression Equation Predicting Guardian
Recommendations
Prior Reports Parental Warmth
of Abuse and Support

Prior Reports 1.00 0.05
of Abuse
Parental Warmth 0.05 1.00
and Support













APPENDIX C


RATE OF REMOVAL BY TYPE OF ABUSE


Table C-1. Cross-Tabular Analysis of Removal Rate by Presence of Neglect
Allegation Judicial Decision
Frequency Not Removed Removed Total
Total Percent
Row Percent
Column Percent
Not Neglected 6 13 19
6.38 13.83 20.21
31.58 68.42
37.50 16.67
Neglected 10 65 75
10.64 69.15 79.79
13.33 86.67
62.50 83.33
Total 16 78 94
17.02 82.98 100.00











Table C-2. Cross-Tabular Analysis of Removal Rate by Presence of Physical
Abuse
Allegation Judicial Decision
Frequency Not Removed Removed Total
Total Percent
Row Percent
Column Percent
Not Physically 6 39 45
Abused 6.38 41.49 47.87
13.33 86.67
37.50 50.00
Physically Abused 10 39 49
10.64 41.49 52.13
20.41 79.59
62.50 50.00
Total 16 78 94
17.02 82.98 100.00


Table C-3. Cross-Tabular Analysis of Removal Rate by Presence of Sexual
Abuse
Allegation Judicial Decision
Frequency Not Removed Removed Total
Total Percent
Row Percent
Column Percent
Not Sexually 9 59 68
Abused 9.57 62.77 72.34
13.24 86.76
56.25 75.64
Sexually Abused 7 19 26
7.45 20.21 27.66
26.92 73.08
43.75 24.36
Total 16 78 94
17.02 82.98 100.00













APPENDIX D


DESCRIPTIVE STATISTICS FOR VARIABLES IN CASES OF
PHYSICAL ABUSE ALONE AND NEGLECT ALONE


Table D-1. Distribution of Interval Variables in Cases of Physical Abuse Only
and Neglect Only
Physical Abuse Only Neglect Only T-Test
Variable N Mean S.D. N Mean S.D. Value

Child Age 11 4.36 3.59 35 3.60 3.34 0.63
Family Income 6 13.17 9.00 21 4.95 7.30 2.05
Severity of Abuse 11 53.45 24.27 36 74.14 21.04 -2.55*
Parental Psychological 10 48.66 18.65 30 58.99 23.34 -1.42
Functioning
Parental Warmth and
Parental Warmthand 11 57.45 18.83 36 54.94 31.06 0.33
Support for Child
Family Conflict 11 54.18 17.06 29 37.54 32.64 2.09*
Parental Noncompliance 11 37.36 23.63 36 31.89 28.78 0.64


* p < .05












Table D-2. Frequencies of Dichotomous Variables in Cases of Physical Abuse
Only and Neglect Only
Physical Abuse Only Neglect Only
Variable Frequency Percent Frequency Percent


Adjudication

Guardian
Recommendation

Family Race

Parent Perpetrator

Previous Reports of
Abuse

Spouse Abuse

Substance Abuse


Removed
Not Removed
Remove
Do Not Remove
White
Non-White
Alleged
Not Alleged
Yes
No
Alleged
Not Alleged
Alleged
Not Alleged








73

APPENDIX E

PEARSON THREE-WEEK TEST-RETEST CORRELATIONS


Table E-1. Pearson Three-Week Test-Retest Correlations of Psychological
Variables
Variable r2
Parental Conflict .52*
Maternal Warmth .48*
Paternal Warmth .49*
Maternal Depression .36
Paternal Depression .18
Maternal Anxiety .68***
Paternal Anxiety .21
Maternal Irritability .60**
Paternal Irritability .09

p<.05
** p<.005
*** p<.001













BIOGRAPHICAL SKETCH

Cynthia McDanal is a psychology associate in the state of Maryland. She

currently works under supervision in a private practice conducting psychotherapy for

children and performing neuropsychological evaluations to rule out learning

disabilities and attention disorders in children, adolescents, and adults. Upon

completion of the requirements for the degree of Doctor of Philosophy, Cynthia will

pursue certification as a licensed clinical psychologist in Maryland and the District of

Columbia. She graduated magna cum laude from Stetson University in 1987 and

earned a Master of Science degree in Clinical and Health Psychology from the

University of Florida in 1990.







I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.



JacqueUn Goldman, Chair
Professor of Clinical and Health Psychology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.



J e ohnson
P fessor of Clinical and Health Psychology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.



Michael Robinson
Associate Professor of Clinical and Health
Psychology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.



Stephen ggs
Associate Professor of Clinical and Health
Psychology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for t de ee of Doctor of Philosophy.




Professor & Law








This dissertation was submitted to the Graduate Faculty of the College of
Health Related Professions and to the Graduate School was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.


Dean, College of Health Related Professions


Dean, Graduate School


April, 1994







































UNIVERSITY OF FLORIDA
3II 11 1111I I I IIIII lill I 1 ll 0I I
3 1262 08554 5324