Factors affecting psychosocial adjustment in chronically ill children and in their parents

UF00102820 ( .pdf )

FACTORS AFFECTING PSYCHOSOCIAL ADJUSTMENT IN
CHRONICALLY ILL CHILDREN AND IN THEIR PARENTS










BY

BRIAN L. LEWIS


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





UNIVERSITY OF FLORIDA


1981









ACKNOWLEDGMENTS


There are a number of people I wish to acknowledge for

their contributions to this dissertation. I would first like

to thank Kon-Taik Khaw, M.D., Senior Associate in Medicine

in the Cystic Fibrosis Program at Children's Hospital in/

Boston. It was Dr. Khaw who agreed to sponsor my research

at Children's Hospital and through whom access was obtained

to the chronically ill subject populations. Dr. Khaw's

initial receptiveness to the project and his unfailing

support and enthusiasm throughout are testimony to his

concern for the psychological, as well as physical, well-

being of his patients. It was certainly my pleasure to

have made his acquaintance.

Dr. Paul Schauble served as the chairman of my doctoral

committee and was my major advisor throughout my doctoral

studies. I would like to thank Paul for his judicious

guidance, uncanny savvy, and respect for my competencies

and need for independence. More than anyone else, Paul has

been my mentor and role model, and his presence was felt in

numerous and sundry ways throughout the preparation of this

dissertation.

The contributions of each of the faculty members on my

committee has been significant. Dr. Franz Epting, as my co-

chairman, was particularly helpful in the initial, concep-

tualizing stages of the project. I have always appreciated










the unstructured quality of Franz's thought processes.

Dr. Hannelore Wass was a valued resource for her expertise

in the area of death and dying. In my interactions with

Dr. Wass I was continually impressed by her openness,

genuine concern, and kind sensitivity. I would like to

thank Dr. Harry Grater for his charismatic and settling

presence, and Dr. Robert Ziller for my initial introduction

to the disease of cystic fibrosis. Although not a formal

member of my committee, Dr. William Froming was sought out

on numerous occasions for advice on statistical matters,

which he freely gave.

This study would not have been possible without the

support and guidance of a number of individuals at Children's

Hospital in Boston. I would particularly like to acknowledge

the following people for their various contributions: Harvey

Colton, M.D., Ms. Mary Williams, Ms. Jane Wally, Ms. Virginia

Rice, Ms. Julie Henry, and Frank Twarog, M.D., Ph.D. I also

wish to thank the Brookline, Massachusetts,School System for

their sanction of the project, and the concerned Brookline

residents who willingly volunteered to participate as controls

in the study.

Finally, I would like to thank Ms. Stephanie Schmitz

for her editorial acumen, support and patience.


iii


1












TABLE OF CONTENTS


CHAPTER PAGE

ACKNOWLEDGMENTS.................................. ii

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

I. PROBLEM STATMENT................................. 1

II. REVIEW OF THE LITERATURE......................... 4

Introduction................................... 4
Adjustment to Physical Illness............... 6
Adjustment in Children with Chronic
Illness.................................... 8
Family Adjustment to Chronic
Childhood Illness.......................... 17
Psychosocial Aspects of Cystic
Fibrosis (CF)............................... 25
Parental Death Attitude and the
Chronically Ill Child...................... 31
Hypotheses................................... 35

III. METHODOLOGY....................................... 40

Subjects..................................... 40
Instruments.................................. 43
Procedure.................................... 50

IV. RESULTS.......................................... 56

Subject Match.................................. 56
Relationship between Chronic Childhood
Illness and Family Functioning............. 57
Relationship between Chronic Childhood
Illness and Child Adjustment............... 61
Relationship between Chronic Childhood
Illness and Parental Attitudes toward
Death...................................... 64
Relationship between Child Adjustment
and Family Functioning..................... 67
Relationship between Parental Orientation
toward Death and Adjustment in the CF
Child..................................... 68

V. DISCUSSION......................................... 72


I









APPENDICES

A TEMPLER'S DEATH ANXIETY SCALE....................... 86

B THREAT INDEX......................................... 87

C INTRODUCTORY LETTER TO PARENTS OF CF CHILDREN........ 89

D INTRODUCTORY LETTER TO PARENTS OF HEALTHY
CHILDREN..................... ....................... 90

E RESULTS OF ANALYSES BASED ON FATHER'S DATA.......... 91

REFERENCES................................................... 97

BIOGRAPHICAL SKETCH........................................... 105















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



FACTORS AFFECTING PSYCHOSOCIAL ADJUSTMENT IN
CHRONICALLY ILL CHILDREN AND IN THEIR PARENTS


By

BRIAN L. LEWIS


June 1981


Chairman: Paul Schauble, Ph.D.
Cochairman: Franz Epting, Ph.D.
Major Department: Department of Psychology


The present study was an attempt to explore factors

affecting psychosocial adjustment in chronically ill chil-

dren and their parents. Previous research has generally

established that the presence of childhood illness is asso-

ciated with deficits in individual and family functioning;

however, it is unclear whether these deficits are due to

the illness per se, or mediating variables. The purpose

of this study was to increase understanding of dynamics

by examining the impact of the illness, versus the mediating

variables of family functioning and parental orientation

toward death, on adjustment in the chronically ill child.

It was hypothesized that these mediating variables would be

more important to the child's adjustment than would the illness.










In testing this hypothesis, data were obtained from

families of the following three groups of children: children

with cystic fibrosis (N=31), children with chronic asthma

(N=27), and children with no previous history of chronic

illness (N=28). All of the children were Caucasian and

between 7 and 12 years of age. The identified child in

each family completed the Piers-Harris Children's Self Concept

Scale. Parents completed the Family Adaptability and Cohe-

sion Evaluation Scales, the Behavior Problem Checklist, the

Death Anxiety Scale, and the Threat Index. The primary

testing of the hypotheses utilized mother's data.

Results indicated that the presence of childhood illness

was not associated with significant deficits in family func-

tioning or with decreased self-concept in afflicted children.

Illness was associated with an increased frequency in behavior

problems in the children (p < .05 for CF children, p < .01

for asthmatic children). Contrary to expectation, death

threat was significantly lower in mothers of terminally ill

children than in mothers of healthy children (p < .05).

Family functioning was found to be a significant variable

relative to behavior problems of the children, regardless of

illness (p < .001). It was concluded that childhood adjust-

ment was related more to this variable than to the presence

of a chronic illness. The implications of these results

were discussed as were avenues for future research.


vii















CHAPTER I
PROBLEM STATEMENT



All people experience periodic episodes of physical

illness throughout their lives. Recovery from these episodes

usually necessitatesshort-term change in normal habits and

lifestyle. Many people are afflicted with illnesses that

are chronic in nature; i.e., long-term and sometimes

progressive and fatal. For them sickness is a way of life:

recovery is often impossible and exacerbation of symptoms

is a constant concern. Maintenance of a stable condition

may require the adoption of a lifestyle somewhat different

from "normal." The imposition of this illness-related

style of life is often stressful for those affected and

sometimes leads to deficits in psychological and social

adjustment (Lipowski, 1970).

Chronic illness in children is particularly problematic.

The usual process of development is stressful enough for

children without their having to contend with the burden

imposed by a potentially debilitating sickness. More than

adults, children must rely on the adaptive resources avail-

able in the family in order to cope with this stress.

Normal, healthy maturation is dependent on this coping

process.









The topic of this dissertation concerns psychological

and social adjustment in chronically ill children and their

families. Most research that has been done in this area

supports the idea that the presence of chronic childhood

illness leads to deficits in individual and family function-

ing (see reviews by Bakwin & Bakwin, 1972; Magrab, 1978; and

Mattsson, 1972). The problem with this research is that most

studies have utilized very few subjects and inadequate

controls, and have been primarily descriptive in nature,

relying on clinical experience and subjective evaluations

(Gayton & Friedman, 1973).

In a recent review; Tavormina, Kastner, Slater, and

Watt (1976) criticized previous research for the implicit

assumption that chronic illness leads to, or directly causes,

emotional problems in children and families. They suggested

that this assumption is erroneous because it implies that

the chronically ill are necessarily a psychologically

deviant population, when in fact many are emotionally well-

adjusted and lead essentially normal lives. These authors

concluded that a more useful conception of the psychosocial

effects of chronic illness is that of Green (1965), who

suggested that the presence of illness places the child

"at risk" (vulnerable) for emotional problems but is not,

in itself, sufficient to produce psychopathology.

It has been suggested that subsequent research, in

addition to using more rigorous methodology, be concerned

with identifying those specific factors that influence the










psychosocial functioning of the sick child and the affected

family. Gayton and Friedman (1973) proposed that mediating

variables important to consider would be family size, reli-

gion, race and marital stability. Tavormina et al. (1976)

expand this list by including degree of severity of disease,

age of onset and duration, life threatening status and degree

of noticeability of the illness.

The study to be reported in later sections was an attempt

to increase our understanding of the psychological and social

effects of chronic childhood illness by examining some of

the above issues. Two main questions were addressed. The

first was general in nature and concerned the overall effect

of the presence and severity of chronic childhood illness

on individual and family functioning. The second question

was concerned specifically with psychological adjustment

in the afflicted child. This question sought to examine

the relative importance of the illness per se, vs. relevant

mediating variables, in affecting the child's adjustment.

Mediating variables of interest were family functioning

and parental orientation toward death.














CHAPTER II
REVIEW OF THE LITERATURE



Introduction



Mattsson (1972) described chronic illness as "a disorder

with a protracted course which can be progressive and fatal,

or associated with a relatively normal life span despite

impaired physical and mental functioning. Such a disease

frequently shows periods of acute exacerbation requiring

intensive medical attention" (p. 803). There are three

fundamental qualities characteristic of chronic illness

that can be extracted from the above description. First,

the chronic illness is long-term, often beginning at birth

or in early childhood and continuing for years or even life.

Secondly, chronic illness is associated with some functional

disability. This disability can be mental as well as physical.

Thirdly, chronic illness implies the presence of a physical

disease which is active in nature, progressively deteriora-

ting, or fluctuating between good and bad periods and

requiring constant treatment to maintain stability.

One problem in reviewing research in this area is the

lack of consensus as to which diseases are chronic and which

are not. The above definition is not sufficiently specific









to make this distinction in all cases. Mattsson (1972)

reported that when including visual and hearing impairments,

mental retardation, and speech and learning disorders as

chronic disorders, it is estimated that as many as 40% of

all children suffer from one or more chronic illnesses.

This is much higher than the 7% to 10% estimate of illness

of primarily physical origin (e.g., cystic fibrosis, leukemia,

hemophilia, etc.) (Pless, 1968). Travis (1976) stated that

there is a fundamental difference between chronic illness and

physical/mental handicaps (hearing impairments, etc.),

because they pose very different problems of adjustment for

the affected individual. The literature reviewed below is

concerned exclusively with the 7 to 10% population of serious

illness with physical origins. Wherever possible, an attempt

will be made to specify the exact nature of the illnesses

studied.

In the following review a number of studies will be

discussed which have attempted to take a wholistic view of

disease by considering the psychological and social effects

of chronic illness in children. For purposes of clarity

the review will be divided into five main topics. The

first topic is a brief review of various ways in which

adjustment to physical illness has been conceptualized.

Of primary concern in this section is an appropriate defini-

tion "healthy adjustment" relative to the presence of chronic

illness.









The second topic is concerned with adjustment in children

with chronic illness. It is a general overview of the litera-

ture pertaining to the child's own reaction to his/her illness

and illness-related environment. The third topic area is

also general in nature and concerns family adjustment to

chronic childhood illness.

The chronic illness which is the focus of the study

reported in this dissertation is cystic fibrosis (CF).

Literature pertaining to the psychosocial aspects of this

specific illness is reported in detail in section four of

this review. The final topic in this chapter concerns parental

attitudes toward death and chronic childhood illness. Litera-

ture is briefly reviewed which suggests that the parent's

attitudes about death create an atmosphere around the sick

child that can be problematic for his/her adjustment.



Adjustment to Physical Illness



Adjustment to physical illness is quite variable and often

difficult to define. Barker, Wright, Myerson and Gonick (1953)

conceptualized adjustment in terms of the self concept of the

afflicted individual. According to this view, disease affects

the body's function as a tool for action. As this function is

altered, so too is the individual's perception of self. Accord-

ing to Barker et al., a person's post-illness self concept

is affected by his/her pre-illness self concept, the stress

of the illness and its effect on his/her body image, and the

response of family and environment to the illness.










Another way adjustment has been conceptualized is in

terms of "coping processes." Lipowski (1970) defined coping

in the physically ill as "all cognitive and motor activities

which a sick person employs to preserve his body and psychic

integrity, to recover reversible impaired functioning and

compensate to the limit for any irreversible impairment"

(p. 93). He went on to consider various "styles" of coping

which he broke down into cognitive and behavioral aspects.

In the cognitive domain, Lipowski identified two general

modes for dealing with the facts of illness: "minimizing"

and "vigilant focusing." In the behavioral domain, Lipowski

proposes three basic coping styles: "tackling," "capitula-

tion," and "avoidance." Tackling implies the active engage-

ment with tasks imposed by the illness. Capitulation is

seen in people who react passively, often regressing to

levels of helpless dependency. Avoidance is the behavioral

style usually linked with the cognitive "minimization," in

which the illness ceases to become a focus and is in some

cases denied.

Mattsson (1972) proposed an adaptational model similar

to that of Lipowski. He defined "coping" as, "all the

adaptational techniques used by an individual to master a

psychologic threat and its intendant negative feelings in

order to allow him to achieve personal and social goals"

(p. 257). Adjustment is defined as successful coping

which results in effective functioning.









Whether coping processes are considered adaptive or

maladaptive depends on a number of factors. Verwoerdt (1972)

groups these factors into three major areas: "characteristics

of the physical illness" (severity, duration and organ system

involved), "characteristics of the host" (age, sex, body

image and premorbid personality), and "situational factors"

("sick role," family relationships and processes of inter-

action). According to Verwoerdt:

Adaptive coping leads to re-establishment of a
dynamic equilibrium that was disturbed by stress.
Maladaptive coping behavior typically leads to a
vicious circle which, by depleting the patient's
resources, aggravates the very problem it is supposed
to solve. Whether a defense is adaptive or maladapt-
ive depends on its intensity as well as its appropriate-
ness to the particular situation. (p. 133)

In reviewing some of the above theories, Pless and Pinkerton

(1975) proposed what they term an "Integrated Model" of adapta-

tion to physical illness. According to this model, adjust-

ment should be viewed in terms of the impact on self-concept

as well as the breakdown in coping processes. These writers

stressed the fact that healthy adaptation is not always

directly related to the degree of acceptance of the reality

of the condition. They prefer to view adaptation in terms

of "psychological balance" or "freedom from abnormality."



Adjustment in Children with Chronic Illness



Children with chronic illness obviously have a more

difficult time in life than those who are healthy. In

addition to facing the usual problems of growth and develop-









ment, they are forced to adjust to atypical circumstances

such as frequent hospitalizations, retarded growth, special

diets, constant medication treatments and frequent absences

from school. These factors combine to continually remind

these children that they are different from others. The

resultant stress makes them "at risk" or vulnerable to

psychopathology (Green, 1965).

Much evidence currently exists attesting to higher than

normal levels of psychopathology and adaptational difficulty

in children with all types of chronic illness (see reviews

by Bakwin & Bakwin,1972; Diller, 1972; Dinnage, 1970, 1972;

Pilling, 1973; and Rutter & Graham, 1970). In a major review

of three extensive epidemiologic surveys of chronic childhood

illness, Pless and Roghmann (1971) reported results support-

ing this thesis.

The three studies reviewed by the above authors were:

the National Survey of Child Health and Development (Douglas

& Bloomfield, 1958); the Isle of Wight survey (Rutter, Tizard

and Whitmore, 1970); and the Rochester Child Health Survey

(Roghmann & Haggerty, 1970). The National Survey consisted

of data collected from a representative sample of 5000

children born in England, Wales and Scotland during the

first week of March in 1946. Of these 5000 children, 528

were described as having symptoms of chronic illness (Pless

& Douglas, 1971). The Isle of Wight study consisted of a

survey of the total population of 9 to 11 year old children

on the Isle of Wight, England. The 3,271 children were









first screened to determine the incidence of chronic physical

illness. Those determined to have a chronic physical condi-

tion were studied to assess emotional concomitants by

psychiatric interviews, psychological tests and interviews

with parents. In the Rochester study a 1% probability sample

of all children under 18 years of age living in Monroe

County, New York, was selected. Of the 1,756 children

screened, 206 were found to have chronic physical problems.

In this study a matched control group was also selected.

In reviewing the results of these studies, Pless et al.

(1971) reported that social functioning is poorer in chronically

ill children than in the normal controls. Although not signi-

ficant, the results from the National Survey and the Rochester

study revealed that the chronically ill children are more

frequently truant, more often troublesome in school, and

more often socially isolated.

In the Isle of Wight study significantly more chronically

ill children than normal children were found to have psychiat-

ric disturbances (based on examinations by a psychiatrist).

Seventeen percent of the sick children were found to be

psychiatrically disturbed (compared to 7% of the healthy

controls). There was no difference between the two groups

in neurotic or antisocial behavior patterns.

The National Survey looked at childhood adjustment in

terms of behavioral symptom questionnaires which were

completed by parents, teachers and the children themselves.

Twenty-five percent of the sick children were found to have









two or more behavioral symptoms, compared to 17% of the

normal controls. Sick children were also rated by teachers

as more nervous and aggressive (39% vs. 31%).

Based on the results of these three independently

conducted studies Pless et al. concluded that maladjustment

is more frequent in the chronically ill children and that a

high proportion of the social and psychological disturbances

that were found must be attributed to the physical disorder.

Unlike the surveys reported above, most studies in

this area have investigated specific chronic illnesses and

their psychological concomitants. The tendency to rely

primarily on descriptive data obtained from relatively small

samples has caused the vast majority of studies to be of

questionable validity. Still, the cumulative evidence

seems to support the above thesis.

In an early study of adjustment in diabetes, Sterky (1963)

compared data obtained from 145 diabetic children with that

of a matched control group of healthy children. Although

the total frequency of "mentally disturbed" children was

the same in the two groups, Sterky reports that the diabetics

were more likely to show severe symptomotology; i.e.,

emotional liability, aggression and difficulty with companions.

One should be cautious in interpreting these results, however,

because the method used for determining mental stability was

subjective in nature, based primarily on interviews with

mothers.

A more sophisticated study of adjustment in diabetes

was undertaken by Swift, Seidman and Stein (1967). Their









research compared 50 diabetic juveniles with 50 normal

control subjects matched according to age, SES, and race.

Based on psychiatric interviews, information from parents,

and a series of psychological tests, the two groups were

compared on a number of psychological dimensions. The authors

reported that the diabetic youths showed significantly more

pathology on psychiatric classification, self percept,

manifest and latent anxiety, sexual identity, constriction

and hostility. Individually scored test data revealed

greater pathological body image, latent anxiety and depen-

dence. In addition, parental interviews revealed that the

diabetics were more socially maladapted at home and with

peers.

Similar findings have been reported in studies of

children with congenital heart disease (Green & Levitt,

1962; Linde, Rosof & Dunn, 1970; and Thomas, Milman &

Rodriquez-Torres, 1970), chronic renal disease (Korsch

& Barnett, 1961; Korsch, Negrette, Gardner, Weinstock,

Mercer, Grushkin & Fine, 1973), rheumatoid arthritis

(Cleveland, Reitman & Brewer, 1965; Grokoest, Snyder &

Schleger, 1962; and McAnarney, Pless, Satterwhite & Friedman,

1974) and cystic fibrosis. Research pertaining to cystic

fibrosis will be presented in detail in a later section.

Chronic childhood asthma has undoubtedly received

more attention from a psychological perspective than any

other childhood illness. The reason for this is that

asthma's aetiology is uncertain, and emotional factors









have been found to have a significant impact on the course

and stability of the disease (Purcell, 1975; Purcell & Weiss,

1970).

A number of early studies have proposed that the asthma-

tic patient displays a distinctive personality pattern and

that this pattern predisposes him/her to developing asthmatic

symptomotology (French, 1936; Riess & Decillis, 1940; Schatia,

1941). The problem with this early research is that it was

based solely on clinical impression and therefore lacked

objective bases for comparison and replication.

In a systematic exploration of the above hypothesis,

Neuhaus (1958) analyzed-data obtained from 169 subjects,

including children with asthma, children with cardiac disease,

normal siblings and a normal control group. The normal con-

trols were selected to match the sick children according to

age, I.Q., SES, religion and number of siblings. Instruments

used to assess emotional functioning included the Rorschach,

Brown Personality Inventory, and Despent Fables. The results

of this study indicated that the asthmatics were significantly

more maladjusted than the normal controls but not signifi-

cantly different from children with heart disease. These

findings tended to refute the hypothesis of a specific

asthmatic personality profile while supporting the notion

that asthma is similar to other chronic illnesses in its

general adverse effect on psychological adjustment.

A similar conclusion was reached in a more recent study

by Graham, Rutter, Yule & Pless (1967). In reviewing data









obtained in the Isle of Wight survey (previously discussed)

these investigators reported that the incidence of psychiatric

illness was greater in the asthmatic children than in the

normal controls (10.5% vs 6.3%), but was not significantly

different from that for children with other chronic disorders.

Tavormina et al. (1976) criticized many of the studies

reported above for their methodological crudeness and biased

assumptions. They contended that most research had approach-

ed the subject assuming that there was a direct cause/effect

relationship between psychopathology and chronic illness.

Furthermore, the evidence cited to validate this assumption

was usually based on subjective evaluations, clinical impres-

sions and abbreviated projective techniques.

In an attempt to provide a more rigorous test of this

assumption Tavormina et al. gave a battery of psychological

instruments to 144 children with chronic disease (CF, diabetes,

asthma and hearing impairment). The battery of psychological

tests were designed to assess various aspects of emotional

stability. Average scores for the children with each type

of illness were analyzed relative to each other and to stan-

dardized norms. Although there were minor differences in

level of pathology between the different types of illnesses,

the authors reported that overall, the chronically ill

children were not significantly more psychopathological

than normal children. Contrary to prevailing attitudes,

these results suggested that the chronically ill were more

healthy than deviant. The authors did note, however, that









sick children demonstrated a wider range of degree of psycho-

pathology and proposed that future research focus on those

factors that tend to predispose the sick child to psycho-

logical deviance.

Following the lead of Tavormina et al. (1976), Bedell,

Giordani, Amour, Tavormina and Boll (1977) identified "life

stress" as one possible predisposing factor. Citing work

by Holmes and Rahe (1967), and Coddington (1972), which

indicated that psychological functioning was related to

recent experiences requiring life change (life stress),

Bedell et al. hypothesized that psychopathology in the

chronically ill might be more directly related to experienced

life stress than to the illness itself. Using 45 chronically

ill children attending a three-week residential summer camp

as subjects, the authors administered to each two measures

of emotional stability: the Piers Harris Self-Concept Scale

and the State Trait Anxiety Inventory, as well as the child's

version of the Social Readjustment Rating Scale (SRRS).

The last scale was a self-report questionnaire designed to

assess the child's recent experience of events that tend

to cause stress and social readjustment.

After dividing the subjects into either a high or a low

stress group based on scores on the SRRS, the authors compared

the groups on the two measures of emotional stability.

Although there were no differences in state and trait anxiety

between the two groups, there was a significant difference

in self concept. Subjects in the high stress group had


_









significantly lower self-esteem scores than those in the low

stress group. Bedell et al. concluded that "Highly stressed

children perceive themselves as more poorly behaved, less

physically attractive, less able at school work, less popular,

and less satisfied with themselves in general" (p. 240).

When considered in conjunction with the fact that the chron-

ically ill children in general experienced higher levels of

life stress, these findings supported the author's hypothesis

that the higher levels of psychopathology frequently reported

in the chronically ill were more directly related to resultant

life stress factors than to the illness.

Pless, Roghmann and Haggerty (1972) identified "family

functioning" as another variable influencing psychological

adjustment in the chronically ill child. These authors

re-examined data obtained in the Rochester Child Health

Survey (Roghmann et al., 1970) previously reviewed. A

mental health adjustment score was obtained for each of the

206 chronically ill children as well as for 110 healthychil-

dren selected as controls and matched for age, sex, race

and SES. The mental health adjustment score was calculated

using data from three measures: parental descriptions of

child's behavior, child's self report, and teacher's

description of child's behavior. In addition, the child's

family environment was rated according to a family function-

ing index designed to reflect qualities like marital satis-

faction, communication potential and family happiness.

The results of the study indicated that psychological










maladjustment was 10 15% greater among the chronically ill

than among the healthy controls. Furthermore, the sick chil-

dren at highest risk for developing psychopathology were

those from poorly functioning families. In other words, in

terms of emotional stability, family functioning was an im-

portant moderating variable. It was not a sufficient variable

determining psychological functioning, however, for the authors

reported that even when controlling for this variable, the

chronically ill as a group were still at higher risk than

were normals. Apparently, family functioning is one of many

potentially stress-producing variables that interact to affect

the child's stability.



Family Adjustment to Chronic Childhood Illness



Although the physical effects of chronic disease are

confined to the affected child, the psychological/emotional

effects of the illness spread over the entire system within

which the child lives. There is little empirical research

available documenting psychosocial effects of chronic child-

hood illness of parental and family functioning. There

are, however, numerous descriptive accounts of associated

stress factors affecting the process of family adaptation

(see reviews by: Magrab, 1978; and Travis, 1976). Most

agree that the presence of chronic physical illness places

a heavy burden on the family and often results in deficits

in family functioning.










Emotional stress in parents of the chronically ill child

often begins prior to diagnosis. In studying the effects of

cystic fibrosis on family life, McCollum and Gibson (1970)

reported that most parents became frustrated and anxious

some time before diagnosis as they began to realize that

their child was frequently sick and not making the gains

that were expected.

When the diagnosis of serious chronic illness is finally

made, parental anxiety and frustration reach a peak as their

worst fears are actualized. From this point on, parents

begin a long and difficult process of adaptation. The adapta-

tion sequence begins with shock, anger and guilt and eventually

(hopefully) reaches a stage of resignation and acceptance

(Mattsson, 1972).

Based on interview data from 35 hemophiliac boys and

their parents, Mattsson and Gross (1966) emphasized the

significance of the parental feeling of guilt. They stated

that guilt is an inevitable reaction of parents on first

hearing that their child has a chronic illness. Parents

tend to feel that, somehow, they should have been able to

prevent the disease. This guilt reaction is particularly

strong for diseases that are genetic in nature (hemophilia,

CF, etc.), because the parents soon learn that, in effect,

they were the direct cause of the child's suffering.

Mattsson et al. suggested that it is essential for parents,

especially mothers, to resolve their guilt feelings if their

children are to develop normally.










Family adaptation is a difficult concept to operational-

ize. Some researchers have approached the topic, in relation

to chronic childhood illness, by focusing on family structure,

e.g., incidence of parental divorce or separation. Literature

in this area is equivocal. A number of researchers reported

that families of sick children were being torn apart by the

stress and that the incidence of divorce and separation was

particularly high in this population (Sultz, Schlesinger &

Mosher, 1972; Binger, 1969). It is unfortunate that these

authors based their information on very small samples and

didn't carefully document their findings or make adequate

comparisons to control groups.

In a more recent investigation Lansky, Cairns, Hassanein,

Wehr, & Lowman (1978) studied the incidence of marital dis-

cord and divorce in families of children with cancer. They

first collected marital status information from the parents

of 191 children who were being treated for cancer over a

seven year period. They then compared the rate of divorce

in this group to the average rate for the population at

large, for couples with children residing in Kansas and

Missouri (the two states in which all the cancer patients

lived). Contrary to expectation, Lansky found that the

person year divorce rate of 1.19% for the parents of chil-

dren with cancer was actually less than the 2.03% rate for

the general population. In order to explore the incidence

of marital stress, Lansky et al. computed the Arnold Sign

Indicator (ASI), based on the comparison of MMPI profiles










for husband and wife of 38 intact couples whose children

were being treated for cancer. For purposes of comparison,

marital stress was also calculated for a group of 23 couples

who had hemophiliac children and for a normal standardized

group. The author reported significantly more marital

stress in the cancer group than in the hemophiliac group,

and significantly more stress in both the clinical groups

than in the normal control group. It was suggested that the

difference in marital stress between the cancer and hemo-

philiac groups was probably due to the nature of the disease;

cancer is more immediately serious and life-threatening.

Lansky et al. concluded that chronic illness does exacerbate

marital conflict and discord but does not lead to higher

rates of separation and divorce. Although this finding

conflicts with previous reports, this study was methodologi-

cally more rigorous than earlier studies.

Similar findings were reported by Begleiter, Burry and

Harris (1976) in studying divorce in parents of children

with CF. These researchers interviewed 40 parents of CF

children and compared the rate of divorce in these families

to the national average (based on the 1970 United States

Census). They found that the divorce rate of 17.2% in CF

families did not differ significantly from the national

average of 14%. Furthermore, those parents of CF children

who were divorced reported that they did not feel that the

sick child was a contributing factor in the dissolution of










their marriage. Twenty-two of these parents actually felt

that caring for the sick child had brought them closer

together.

Family adaptation to chronic childhood illness has also

been explored via the concept of "family functioning" (Pless

& Satterwhite, 1973). "Functioning" is a dynamic concept

relating to the quality and process of family life rather

than its mere structure. Given the complex nature of family

dynamics, it is little wonder that the majority of studies

reporting family effects have been descriptive in nature

and based primarily on subjective interviews and clinical

experience (Pless et al., 1975). In a recent article,

Stein and Riessman (1980) reviewed previous attempts to

measure the "impact-on-family" of chronic childhood illness.

They concluded that, "To date, no comprehensive measure

exists which can quantify the impact of childhood illness

on a family and which delineates the many facets of this

complex domain. None of the measures meet rigorous scientif-

ic criteria, i.e., demonstrable reliability and validity"

(p. 465).

Although measures have been slow to develop,there is

a growing body of theoretical literature pertaining to

family functioning and chronic childhood illness. It has

been noted that the presence of a chronically sick child

increases the developmental burden on the family (Travis,

1976; Magrab, 1978). There is a natural tendency for parents

to overprotect the sick child, often at the expense of his/her










developing a sense of mastery and autonomy. In a recent

article on primary care setting for the chronically ill,

Harding, Heller and Kesler (1979) emphasized this challenge

to the family and stated:

The developmental tasks of each child are mirrored
by developmental tasks of the family as a whole.
A child does not easily gain autonomy without the
family's encouragement or allowing him to do so.
All families must strike a balance between protection,
nurturance, and dependency on the one hand; and auton-
omy, independence, and self sufficiency on the other.
An inability to strike this balance can lead to
serious problems which affect not only the emotional
well-being and harmony of the family members, but
their health as well. (p. 315)

They went on to say that "The chronically ill child is often

deprived of many opportunities and experiences because of

the decreased expectations ,of well-meaning adults" (p. 314).

The literature on childhood asthma is replete with

references made to maternal overprotection and a tendency

for overly close mother-child bonds (Sandler, 1977; Meijer,

1979). Others have reported dramatic improvement in asthmat-

ic symptomatology following structural family therapy in

which healthy distance was created between mothers and

affected children (Liebman, Minuchin & Baker, 1974).

It is this writer's feeling that a "systems" perspective

(Haley, 1971, 1972) provides a much needed theoretical frame-

work for the study of family adaptation to chronic illness.

Systems theory views the family as a multidimensional com-

plex which has an identity of its own, but which is also

dynamic in nature, constantly being submitted to stimuli

which provoke changes in its structure (Bowen, 1971).










According to systems thinking, healthy family function-

ing requires a degree of flexibility. Flexibility enables

the family to.confront a crisis or stressful event by chang-

ing individual roles as well as transactions between various

subsystems in order to reestablish a state of equilibrium or

balance (Minuchin & Barcai, 1969). If the alliances between

family members are too "rigid," transactions are prevented

and conflicts become insoluble.

One attempt to apply systems theory to the study of

family dynamics in response to chronic disease was made by

Paquary-Weinstock, Appelboom-Fondu and Dopchie (1979).

These researchers studied 25 families of hemophiliac chil-

dren in Belgium. Family dynamics were assessed through

interviews with parents in which information was obtained

concerning type of communication, the existence of alliances,

and type of functioning, both before and after diagnosis of

the disease. The authors reported that families which had

a strong identity and functioned according to an open

communication system prior to the disease had the healthiest

post-diagnosis adaptation. Although initially reacting with

shock and depression, they were eventually able to consolidate

family bonds and restructure the family cell. On the other

extreme, families whose identity and style of communication

were dysfunctional to begin with tended to react to the

presence of the illness by dislocation, and often the

appearance of psychiatric pathology in one of the parents.











There are many flaws in the above study. The authors

based their conclusions on a very small sample and on sub-

jectively interpreted data. Furthermore, they did not

define their terms or describe methods used to analyze the

data. The study does show, however, that there is potential

for meaningful research in this area. What is needed is a

conceptual model for studying family systems which proposes

testable hypotheses and is open to empirical validation.

Olson, Sprenkle and Russell (1979) present such a con-

ceptual system in the "Circumplex Model" of marital and

family systems. After reviewing the literature in the entire

field of family behavior, these writers concluded that two

aspects, "cohesion" and "adaptability," appear to be the

most salient dimensions used to describe family functioning.

They define "cohesion" as: "the emotional bonding members

have with one another and the degree of individual autonomy

a person experiences in the family" (p. 5). "Adaptability"

is defined as t "the ability of a marital or family system

to change its power structure, role relationships, and

relationship rules in response to situational and develop-

mental stress" (p. 12).

In the Circumplex Model, Olson et al. have combined

these two independent dimensions in such a way that families

can be classified according to where they fall on both. By

dividing each dimension into four levels: very low, low to

moderate, moderate to high, and very high, a 4 X 4 matrix

is formed indicating 16 types of family functioning.










According to the Circumplex Model (which is based on

family systems theory), balanced levels on both cohesion

and adaptability make for the healthiest family functioning.

Empirical studies validating this model have been reported

by Olson, Russell and Sprenkle (1980b). This model has also

been found to be useful in monitoring change in family

functioning in response to stress (Olson & McCubbin, 1980a).

It is this writer's feeling that the Circumplex Model

of family systems could provide an excellent conceptual

framework for examining the relationship between family func-

tioning and chronic childhood illness. The research presented

in later sections makes use of this model.



Psychosocial Aspects of Cystic Fibrosis (CF)



Cystic fibrosis is the most common autosomal-recessive

disease found in the white population, and it is thought to

affect one in every 1600 to 2500 live births (Breslow,

McPherson & Epstein, 1981). It is a disease which causes

a chronic pulmonary condition, pancreatic insufficiencies

and elevated levels of sodium and chloride in sweat. It is

the pulmonary aspect of the disease which is the most prob-

lematic and which accounts for almost all the mortality in

CF patients (Doershuk, Wood & Boat, 1976). CF is a life-

threatening illness for which there is presently no cure.

A strict regimen of antibiotic treatment and rigorous physical

therapy has extended the average age of survival to 19 years

(Breslow et al., 1981), but few live into their thirties.










Cystic fibrosis is a particularly stressful chronic ill-

ness both for the family and the afflicted child. As Denning,

Gluckson and Mohr (1976) point out, factors contributing to

the stressful nature of the illness include:

it is a chronic disease for which there is no cure,
it is a genetically transmitted disease for which
there is no carrier test yet available, it is an
expensive disease requiring thousands of dollars
worth of medication each year, and it is a disease
in which no one can accurately assess prognosis
and life span to lend support to the parents,
patient and families who are expected to faithfully
carry out the prescribed treatment and other details
of a complicated medical program day after day
throughout life. (p. 127)

One of the first systematic attempts to investigate the

psychosocial effects of CF on family functioning was a des-

criptive study by Turk (1964). Turk administered a ques-

tionnaire to the parents of 25 children with CF ranging in

age from 3 months to 23 years. He reported that all of the

families studied experienced stress relating to the child's

disease. Of primary concern was financial strain. The CF

child needs continual medical attention as well as special

foods, supplements and home treatment equipment. In most

cases the family was left to bear the financial burden them-

selves, which often meant that the father had to work longer

hours or take a second job. There was also considerable

social stress within these families. Parents reported that

they had less time to devote to their spouses, their leisure

time was drastically reduced, their sexual relations were

strained, and constructive communication between family mem-

bers was inhibited. Furthermore, many parents reported











trouble deciding whether or not to have another child, and

60% said they could not discuss the diagnosis with the

affected child.

In another study with CF patients, Lawler, Nakielny and

Wright (1966) conducted psychiatric interviews with 11 mothers

of affected children. Although again results were presented

only in a descriptive fashion, they did indicate considerable

intrapsychic conflict in most parents. Eight of the 11

mothers were judged to be clinically depressed. Further-

more, many of the mothers reported marriage and family conflicts.

Six of the 11 said that they had recently considered separation

from their husbands and three reported having abstained from

sexual relations with their husbands. Although these results

suggest negative effects of the presence of a chronically ill

child on parental stability and family fucntioning, one must

be cautious'about making such interpretations from descriptive

reports of such a small group of subjects. Lawler et al.

make no reference to any control group; therefore, there is

no basis for comparing families of healthy and those of

sick children.

In a study previously cited (Begleiter et al., 1976), it

was noted that the rate of divorce or separation among CF

parents did not differ from the national average. The study

did find, however, that the rate was higher for CF parents

than for parents of children with most other chronic diseases.

The authors suggested that this finding is probably due to the

specific nature of the disease, especially the fact that it










is genetically transmitted, with a high probability that

subsequent children will either have, or be carriers of,

the disease. Parents are naturally anxious about the pros-

pect of further children and usually decide to limit their

families. Begleiter et al. proposed that it is this realiza-

tion, and not the reality of having to care for the sick

child that causes a higher rate of divorce in parents of

CF children than in parents of children with other diseases.

In a study of 20 children with CF, Tropauer, Franz and

Dilgard (1970) reported that these children had a reasonably

good understanding of their disease and were tolerant of

treatment but they were upset by dietary deprivation,

interruptions of play, and physical limitations. Results

of projective psychological tests indicated that these chil-

dren tended to feel inadequate, insecure and anxious.

Similar results were reported by Cytryn, Moore and

Robinson (1973). In their study of 29 children with CF

between the ages of 8 months and 9 years, 42% were judged

to have significant emotional disturbance.

In a major review of the literature pertaining to

psychosocial aspects of CF, Gayton and Friedman (1973) contended

that, "The studies reviewed are consistent in their conclusion

that the occurence of CF results in psychosocial consequences

for patient, parents and family" (p. 858). The presence of

CF has been found to precipitate financial concerns, communi-

cation difficulties, emotional instability, marital problems

and overall family stress. Gayton et al. are cautious in

their interpretation of these results, however, and they










note a number of problems with this research. Again, most

of these studies have used very few subjects and have been

primarily descriptive in nature. Although this type of

research can give a clinically useful perspective of the

problem, without adequate controls and more sophisticated

instrumentation it is difficult to make meaningful statements

about the specific effects of the chronic illness.

In a more recent study, Gayton, Friedman, Tavormina

and Tucker (1977) attempted to verify results of previous

research while applying a scientifically more rigorous

methodology. As their sample they selected 45 families of

children with CF from a total population of 75 being seen

at the Cystic Fibrosis Clinic at the University of Rochester

Medical Center. The children were between the ages of 5 and

18 with an average age of 11. Parents of the children com-

pleted two questionnaires, the Family-Concept Q-Sort (FCQS)

and the MMPI. The CF children and selected normal siblings

completed three questionnaires; the Piers-Harris Self Concept

Scale, the Missouri Children Picture Series, and the Holtzmann

Inkblot Test.

In exploring psychological adjustment of the CF children,

results of their responses to the above instruments were

compared to those of their healthy siblings as well as appro-

priate normative data (when available). The authors reported

that, except for one of the 22 variables calculated for the

Holtzmann Inkblot Test, no significant differences were

found between the groups. This obviously conflicts with

previous findings.











In exploring parental adjustment, MMPI profiles of the

CF parents were compared to those of parents of noncystic

children (drawn from the population of parents used in a

previous study). The authors reported that fathers of CF

children showed significantly higher scores on 4 MMPI scales

than the comparison group fathers. Mothers of CF children

had significantly higher scores on 2 of the subscales.

This result is consistent with previous research which has

demonstrated negative effects on the emotional stability of

parents caring for the CF child. It is somewhat surprising,

however, that the fathers appeared to be affected more than

the mothers.

In terms of parental perception of family functioning,

responses of parents of CF children were compared to those

of parents of noncystic children (drawn from a population

of parents used in another previous study). The authors

reported that, as expected, having a child with CF tended

to decrease family satisfaction and family adjustment.

The overall results of this study lend only partial

support to the findings of previous research. In discussing

these findings, Gayton and Friedman (1973) stated:

The literature dealing with the psychosocial aspects
of chronic illness often implies that the presence
of a chronically ill child in a family serves as a
psychological stressor, invariably resulting in
damaging psychological consequences for the child
and his family. Although he would not disagree with
the conceptualization of serious chronic illness as
a stressor, this does not preclude the utilization
of coping devices which allows the chronically ill
child and his family to develop healthy adaptations.
(p. 893)










Again, the challenge for future research lies in the identi-

fication of those specific factors that are important in

facilitating healthy coping and adaptation.



Parental Death Attitude and the Chronically Ill Child



Death is obviously a complex and troublesome issue in

the consideration of psychological adjustment in terminally

ill children and their families. There is currently much

disagreement among professional as to the degree to which the

sick child is actually concerned about death, and how directly

the child needs to confront the topic. Whereas some contend

that death is "the central emotional problem of the seriously

ill child" (Vernick & Karon, 1965, p. 396), others do not

agree.

Many experts in the field of treatment for the chronically

ill child contend that death is not an important issue for

the child (Evans & Edin, 1968). Natterson and Knudson (1960)

reported that fatally ill children do not express, or even

experience, anxiety about death until they are at least 10

years old. Relying on staff observations, Morrissey (1963)

reported that less than 30% of all fatally ill children ever

exhibit any anxiety about death. Based on these findings it

has been suggested that discussing death with the sick child

only serves to heighten his/her anxiety by introducing a

subject that the child cannot possibly comprehend. The










implication is that the child's suffering will be minimized

if he/she is shielded from the reality of the disease.

Others have sharply criticized this perspective. Using

data based on interviews with children (between 9 and 20 years

of age) hospitalized for the treatment of acute leukemia,

Vernick and Karon (1965) made the opposite contention. These

researchers reported that children (even very young children)

who were being kept in ignorance about their disease knew much

more about the seriousness of their condition than was im-

mediately obvious. They seemed to be able to know when they

were being lied to and could easily detect anxiety in their

parents. It was also observed that these children really

wanted to talk about their disease, but they felt to do so

would make their parents upset and angry. Vernick et al.

reported that as a result of this frustration, "protected"

children would often withdraw, feel isolated, and tend to

exhibit more behavior problems than children who were aware

of their diagnosis and prognosis. Based on a very success-

ful treatment program initiated by these authors, they

concluded that:

the most helpful way for adults to meet the emotional
needs of the seriously ill child was to provide an
atmosphere in which such children feel completely
free to express their concerns: an atmosphere which
always provided an honest answer to any question.
(p. 397)

Waechter (1971) found empirical support for some of the

contentions of Vernick and Karon. Waechter studied 64 children

between the ages of 6 and 10 in three groups: terminally ill

(CF, leukemia, etc.) children, children with chronic non-









life-threatening disease, and children with just brief ill-

nesses. All of these children were studied while in the

hospital using the following measures: the General Anxiety

Scale, TAT, and parental interviews. Consistent with earlier

reports, fear of death was rarely spontaneously discussed in

any of the children; however, the terminally ill children

had general anxiety scores twice as high as any other group.

Furthermore, in responses to the TAT cards, these children

discussed significantly more themes of loneliness, separation

and death. This finding is particularly significant in light

of the fact that the medical treatment of the terminal group

was not significantly different from that of the chronic

non-terminal group, and that 14 of the 16 terminal children

were supposedly unaware of their diagnosis.

Spinetta and Maloney (1975) present further evidence

supporting the thesis that death is of concern for terminally

ill children. In their study, projective stories told by

27 fatally ill children between the ages of 6 and 10 were

compared with stories told by a matched group of children

with non-terminal chronic illness. Results show that

themes of threat to body function and body integrity were

much more frequent in the stories of the fatally ill chil-

dren. General anxiety also appeared manifest at a higher

level in the stories of the terminally ill.

An interesting finding of the Waechter study (previously

discussed) was a significant negative correlation between

preoccupation with themes of separation and death, and the










degree to which the child was allowed to discuss his/her

disease. The implication of these findings is that even

when shielded from the full nature of their disease, terminally

ill children are aware of the seriousness of their condition,

are concerned about dying, and need to be allowed to express

this concern.

Apparently, the reason many adults feel they must not

openly discuss death with dying children is because of their

own inability to deal with the subject (Vernick, 1973).

It seems reasonable to suspect that parental attitudes toward

death will affect the emotional climate surrounding the

child and ultimately the child's adjustment. The relationship

between psychosocial adjustment in chronically ill children

and parental orientation toward death has not previously

been explored.

In reviewing much of the literature on psychological

and sociological aspects of dying, Simpson (1979) noted that

there has been a paucity of research concerning death atti-

tudes in adults. This seems to be due to the fact that death

is a very sensitive and complex issue. As Simpson pointed

out, attitudes toward death appear to be quite variable,

are difficult to define, and are even more difficult to

reliably measure.

Research examining death attitudes in adults has been

concerned primarily with emotional attitudes--usually fear

or anxiety (Kastenbaum & Costa, 1977). A number of test

instruments have been developed to measure anxiety associated










with the idea of death (Collett & Lester, 1969; Dickstein,

1972; Pandley, 1975; and Templer, 1971). Krieger, Epting

and Leitner (1974) proposed "death threat" as another way

to conceptualize one's attitude toward death. Although

somewhat related to anxiety, "threat" implies a more cogni-

tive orientation. "Death threat" as measured by the Threat

Index (TI) (Krieger et al., 1974) has been found to correlate

only weakly with measures of "death anxiety" (Rigdon, Epting,

Neimeyer & Krieger, 1979) which implies that they are tapping

different dimensions of a more global concept of death orienta-

tion. "Death threat" and "death anxiety" appear to be two

relevant ways to examine the relationship between parental

death orientation and adjustment in chronically ill children.



Hypotheses



The study to be reported in this dissertation is an

attempt to expand our understanding of the psychosocial

concomitants of chronic childhood illness by examining some

of the above issues utilizing a scientifically rigorous

methodology. As has been mentioned, most previous research

in this area has relied primarily on descriptive data obtain-

ed from small samples and has not used adequate controls.

In this study, objective instruments (with established

reliability and validity) are used to obtain data from three

groups of children matched, as closely as possible, by age,

sex, and race. Two of the groups consist of children with










diagnosed chronic illness (CF and asthma). The third group

is a control group composed of "normal" children, i.e.,

children with no previous history of chronic physical illness.

By using these three groups, not only can differences be

studied between chronically ill and normal children, but

also between sick children whose illnesses have very differ-

ent implications in terms of their life-threatening status

(CF is a terminal disease, asthma is not).

The review of the literature indicates that the presence

of chronic childhood illness is often associated with a

deficit in psychological functioning of child and family.

There has been a gradual shift in emphasis over the years

from considering the disease as causative in this process

to viewing the problem as a failure in coping or adaptation.

The reason for this shift is the growing realization that

many children and families do not appear to suffer psychologi-

cal problems in the presence of chronic illness.

There are two main questions addressed in this study.

The first is general in nature and concerns the overall

effect of the presence and severity of chronic childhood

illness on individual and family functioning. Three depen-

dent variables are explored in this regard. They are:

child adjustment (defined in terms of behavior problems

and self-concept); family functioning (defined in terms

of the circumplex model for cohesiveness and adaptability);

and parental orientation toward death (defined in terms of

"death threat" and "death anxiety"). Thus, this study in










some ways replicates previous research. It differs in the

use of a family-functioning model which has not previously

been applied to this population, and in the consideration

of parental death orientation as a relevant dimension that

may be affected.

The second question is more specific in nature and

potentially more important. This question will explore

the effects of family functioning and parental orientation

toward death on the adjustment of the child. In this way,

the relative impact of the illness per se (vs. mediating

variables: family functioning, etc.) can be analyzed in

terms of effects on childhood adjustment.

The following hypotheses are explored:

1. Family functioning will be affected by the

presence and severity of chronic childhood illness in

the following ways:

a. Families of CF children will be more likely

to be functioning at the extremes of the Circumplex

Model than will families of healthy children; i.e.,

they are more likely to be either high or low in

cohesion and adaptability.

b. Families of asthmatic children will be

more likely to be functioning at the extremes of

the Circumplex Model than will families of healthy

children.

c. Families of CF children will be more likely

to be functioning at the extremes of the Circumplex

Model than will families of asthmatic children.










2. Child adjustment will be affected by the presence

and severity of chronic illness in the following ways:

a. Children with CF will have more behavior

problems and lower self concepts than healthy children.

b. Children with asthma will have more behavior

problems and lower self concepts than healthy children.

c. Children with CF will have more behavior

problems and lower self concepts than asthmatic children.

3. Parental orientation toward death (death threat

and death anxiety) will be affected by the presence and

severity of chronic childhood illness in the following

ways:

a. Parents of CF children will be more anxious

about and threatened by death than will parents of

healthy children.

b. Parents of asthmatic children will be more

anxious about and threatened by death than will

parents of healthy children.

c. Parents of CF children will be more anxious

about and threatened by death than will parents of

asthmatic children.

4. Regardless of illness, the tendency for families

to be functioning at the extremes of the Circumplex Model

will be directly related to the number of behavior problems

of the children and inversely related to children's self

concept.

5. In CF families, parental death threat and death

anxiety will be directly related to the number of behavior






39



problems of the children and inversely related to chil-

dren's self concept.















CHAPTER III
METHODOLOGY



Subjects


Eighty-one children and their parents (usually mothers,

preferably both mothers and fathers) participated as subjects

in this study. The children were selected on the basis of

their membership in one of three groups. Group 1 consisted

of children with cystic fibrosis (a chronic, life-threatening

disease). Group 2 consisted of children with chronic asthma

(usually a non-life-threatening disease). Group 3 consisted

of "normal" children (children with no previous history of

chronic illness). Samples were obtained in the following

manner.

Group 1: Thirty-one children with CF were selected

from the population of children being treated on an outpatient

basis in the CF clinic at Children's Hospital Medical Center

in Boston. The following criteria were used in making

this selection:

1. The children had to be between the ages of 7 and

12. Children younger than 7 were excluded because

of the level of verbal ability required to complete

the self concept inventory. Children older than 12

were not included because it was felt that the

quality of adjustment demands changes as one enters









adolescence; consequently, adolescents and children

may differ in terms of healthy coping styles.

2. The children had to have been diagnosed as having

CF for at least two years. The purpose of this

criterion was to insure that the effects to be measured

in individual and family adjustment would reflect more

or less stable and long-term coping styles. Families

of more recently diagnosed children might still be

experiencing short-term effects due to the initial

shock of diagnosis.

3. The children could not have been hospitalized

within the preceding year due to exacerbation of the

illness. The rationale for this criterion is the same

as the above;i.e., to tap enduring adjustment styles

rather than short-term reactions to crisis situations.

4. The children could not have any siblings who also

have the disease. Families in which two or more chil-

dren had been diagnosed as having CF or in which a

child had previously died from the illness were excluded.

It was felt that these families could be at a differ-

ent stage of adjustment than families with just one

sick child.

Of the 31 children in this group, 19 were boys and 12

were girls. Data wereobtained from all 31 mothers and from

20 fathers.

Group 2: Twenty-seven children with chronic asthma were

selected from the population of asthmatic children being

treated on an outpatient basis in the Allergy Clinic at









Children's Hospital Medical Center in Boston. These chil-

dren were selected so as to match the CF children as closely

as possible by age, sex and race (all of the children in the

study were white). In addition, they had to meet the follow-

ing criteria:

1. They had to have been diagnosed as having asthma

for at least 2 years (same rationale as above).

2. They could not have been hospitalized within the

preceding year due to exacerbations of the illness

(same rationale as above).

3. They could not have any siblings who also have the

disease (same rationale as above).

4. They had to be taking daily medication for mainte-

nance of the illness. Asthma is an illness with a

wide range of severity. In some it is hardly notice-

able. The purpose of this criterion was to insure

that all children selected were experiencing an

illness severe enough to require daily attention.

Of the 27 children in this group 16 were boys and 11

were girls. Data wereobtained from 26 of the mothers and

18 of the fathers.

Group 3: Twenty-eight children with no previous history

of chronic illness participated as control subjects in this

study. These children were selected from a single elementary

school in the Boston metropolitan area and matched, as

closely as possible, to the CF children according to age,

sex, and race. An additional criterion used in selecting









this sample was that neither the subject nor any of his/her

siblings could have had a chronic childhood illness. Of

the 28 children in this group 16 were boys and 12 were

girls. Data wereobtained from 27 of the mothers and 18 of

the fathers.



Instruments



Family Adaptability and Cohesion Evaluation Scales (FACES).

This is a self-report instrument developed by Olson, Bell and

Portner (1978) to assess family functioning along the two

primary dimensions of "adaptability" and "cohesion." It

is a clinical and research tool based on the Circumplex

Model (Olson et al., 1979) previously discussed. The Cir-

cumplex Model combines adaptability and cohesion in an

attempt to diagnose marital and family systems and to set

goals for treatment.

FACES is composed of 111 statements concerning various

aspects of family dynamics and functioning. A family member

independently completes the questionnaire by indicating

on a scale from one to four the degree to which each state-

ment is felt to be true of his/her family. A "1" indicates

that the statement is felt to be true of the family "none

of the time," while a "4" means it is true of the family

"all of the time."

From this scale two primary scores are obtained, one

for "cohesion" and one for "adaptability". Scores for









cohesion can range between 162 and 378. Adaptability

scores can range between 126 and 294. In addition, 16

subscales have been proposed, based on factor analysis.

Split-half reliability is very low on these subscales,

however, and the authors recommend that total scores

rather than subscales be used in research.

By dividing each of the two primary dimensions into

four levels (low, low-moderate, moderate-high, and very high)

and then combining them in a single matrix, a 4 X 4 matrix

is formed indicating 16 different types of marital and

family systems. This matrix is shown in Figure 1 with names

the authors use to describe each type of system. According

to the Circumplex Model, the healthiest families are those

located in the moderate ranges of both dimensions (the

four types inside the rectangle) while the unhealthiest

families are those at the extremes on both dimensions (those

underlines in the four corners). In between these two

groups are the eight types of families which are moderate

on one dimension but extreme on the other.

Because of the newness of this instrument, validity

studies are still underway and have not yet been reported.

The Circumplex Model upon which FACES was developed does

appear to have empirical validity in terms of differentiating

families under stress and in setting treatment goals for

family therapy (Olson et al., 1980a;and Olson et al.,

1980b).

The authors reported high internal consistency relia-

bility for the total scores for adaptability and cohesion



















Q bo PC 0 $ v
Q 00 0D

M0 4) X (l 3 S

S x rA COvN s


- >



a I 0 r
C -o
0 r-



41 $8












d0
Of
8 X U
M s G C


Scd v


4O G) O0
r-o 4 O



FzU C0U










-4 0 0)
a (d 4 OS
X td 3 Cl
(D 0. ;.4 "
r-4) *4 Q)
N4C 3 MO


*4


O)









CO
O
w0










hO


4-)


CO )
0



E

d X
a)

r-i


4O


0





cdX
2

4E
0 0



w 0





cI-






.,-4
PC)
COE

*0
r-
(l' Z
a:
Ma


QpH ,-4 O O
0 0 af M b 4N bGM


H a) -
Ao bo bD "




O a







co
E-
00 MQ ,Q








04 0 2 4)









(r = .75 and r = .83 respectively). Correlation with the

Edmond's Conventionality Scale (a measure of social desira-

bility) was very low for adaptability (r = .03) but high

for cohesion (r = .45).

Death Anxiety Scale (DAS). This is an instrument devel-

oped by Templer (1970) to assess anxiety associated with the

thought of death. The DAS consists of a list of 15 statements

concerning death and death-related issues (see Appendix A for

for copy). In completing the scale the subject is to indicate

whether each statement is mostly true or mostly false for him/

her by circling either "true" or "false". The scale has been

administered by itself, as well as embedded within the MMPI,

with no difference in mean scores (Templer and Ruff, 1971).

Items of the DAS were selected so as to insure internal

consistency significant at the .10 level. Templer reported

test-retest reliability (3 weeks apart) of .83. The scale

did not correlate significantly with the Marlowe-Crowne

Social Desirability Scale.

The DAS was found to have predictive validity in a study

by Templer (1970). Twenty-one high death-anxiety psychiatric

patients (identified by their spontaneous verbalizations of

fear of, or preoccupation with, death) scored significantly

higher on it than did a group of psychiatric control patients

who were not overtly preoccupied with death. Concurrent

validity was demonstrated by a significant correlation

(r = .74) with Boyar's Fear of Death Scale (FODS) for a group

of 77 undergraduate psychology students.










Threat Index (TI). The TI is a cognitive measure of

death orientation developed by Krieger et al. (1974) and

based on the theoretical work of George Kelly (1955). Death

threat differs from death anxiety or death fear in that it

is concerned with the degree to which an individual's

construct system is structured to anticipate death, rather

than the affect associated with the topic. Krieger et al.

(1974) defined death threat as:

the reluctance of a person to subsume his present view
of himself, the way he prefers to see himself, and the
concept of death together as elements under the same
poles of a sample of his constructs. It is assumed
that the person who describes both himself and death
with the same pole of a single construct dimension
is organizing his world in such a way as to be able
to see death as a personal reality. The person who
places himself and death on opposite poles of a
construct, however, would have to reorganize his sys-
tem to construe self and death together. (p. 301)

The form of the index used in this study is the TIp40,

which consists of two identical lists of 40 bipolar dimen-

sions. (See Appendix B for copy). On the first list subjects

are to go through each of the 40 pairs of adjectives indicating

the side with which they see themselves more closely associ-

ated (self elements). On the second list they choose which

of the two adjectives in each pair is more closely associated

with the thought of their own death (death elements). The

death threat score is determined by the number of splits,

i.e., the number of times the "self" element is rated toward

one pole of a construct and the "death" element toward the

other pole. Scores for the TIp40 can range between 0 (not

at all threatened) and 40 (very threatened).








In a recent review of the research utilizing the TI,

Rigdon et al. (1979) reported test-retest reliability for

the TIp40 between .87 and .90 for intervals from 4 to 9 weeks.

The instrument was found to be internally consistent with a

split-half coefficient of .96 (Krieger, Epting & Hays,

1979). It did not correlate significantly with the Marlowe-

Crowne Social Desirability Scale.

Since this is the first instrument of its kind, concur-

rent validity could not be assessed by comparing it with other

measures of death threat. As expected however, the TI has

been found to correlate in a consistent manner with other

measures of death orientation (Krieger, Epting & Leitner,

1974; Neimeyer, Dingemans & Epting, 1977; Neimeyer & Dingemans,

1980; and Epting, Rainey & Weiss, 1979). This suggests that

the TI has validity as a measure of death orientation. The

TIp40 correlated moderately (p = .23) with Templer's DAS

(Krieger, 1977). Although this indicates some overlap between

the two scales, it is weak enough to imply that the instru-

ments are measuring two different aspects of death concern.

Behavior Problem Checklist (BPC) (Quay & Peterson, 1979).

The BPC is an instrument which developed from the original

work of Peterson (1961). It consists of a list of 55 problem

behaviors that often occur in childhood and adolescence.

The checklist measures four primary dimensions derived from

factory analysis (i.e., conduct problems, personality prob-

lems, inadequacy-immaturity and social delinquency). It

can be completed by parents, teachers, or anyone else

familiar with the child's behavior.









In completing the BPC, the evaluator is instructed to

go through the list placing a check next to any of the items

which is felt to represent a problem as far as the child is

concerned. A score is obtained for the child by summing

the total number of items checked. Scores can range between

0 and 50; the higher the score, the more behavior problems

exhibited by the child. When appropriate, individual scores

can also be calculated for each of the four factors mentioned

above.

Quay (1977) reviewed the reliability and validity data

for the instrument. The scale was reported to be internally

consistent and to have test-retest reliability around .82.

Interrater reliability tended to be moderately high between

teachers and between parents, but surprisingly low between

teachers and parents. O'Leary and Johnson's (1979) review

suggested that the instrument has high concurrent and con-

struct validity. The scale reliably differentiates groups

of children with known behavior disorders.

Piers-Harris Children's Self Concept Scale (P-H) (Piers

& Harris, 1969). This is an 80 item self-report questionnaire

to assess the child's self-evaluation. It developed from the

original work of Piers and Harris (1964). The items in the

instrument are first person declarative statements of the

type, "I am a happy person", to which the child indicates

whether it is mostly true or mostly false by circling

either "yes" or "no." Scores can range between 0 and 80;

the higher the score the greater the child's self esteem.









The P-H was originally designed for use with children

between the ages of 8 and 17. Because of the reading level

required to complete the questionnaire, standardization

data has not been obtained for children below 8 years of

age. The authors did suggest, however, that when administer-

ed individually the scale might be used successfully below

this level. Piers (1977) reported test-retest reliability

of .76 for a group of 7 year old children over a two month

interval. This compares quite favorably with that obtained

for older children.

Reports of internal consistency for the scale ranged

from .78 to .93 with test-retest reliability estimates

around .72 (Piers and Harris, 1969). A correlation coefficient

of about .65 was found with similar measures of self-esteem,

and of .40 with teacher and peer ratings (Piers, 1977;

Buros, 1972). Some have found the P-H to be a good indicator

of emotional stability (Bedell et al., 1977). Piers and

Harris (1969) reported correlations with measures of anxiety

between -.54 and -.69.



Procedure



Group 1: The procedure for selecting the CF subjects

and administering the questionnaires required a number of

steps. The first step was to review, on a weekly basis,

lists of CF patients with upcoming clinic appointments.

Two weeks prior to their scheduled appointment a letter









was sent to the parents of any and all CF patients who met

the above-stated criteria. (A copy of this letter is included

in Appendix C). In this letter, the purpose of the study

was described and an initial orientation to the procedure

was presented. The letter was thought to be necessary

because of the sensitive nature of the questionnaire involved

(especially those concerned with "orientation toward death").

A week following the mailing of these letters a tele-

phone call was made by the investigator to the parents. The

purpose of this call was to answer any questions the parents

had and to arrange a time to meet during their .upcoming

clinic appointment.

On the day of their appointment, consenting parents

and their children met with the investigator for approximately

20 minutes. During this meeting general information on the

family was obtained from the parents while the children

completed their questionnaire (the Piers Harris Children's

Self Concept Scale). The children filled out these question-

naires in private and were told that no one except the

investigator would have access to their answers. Any help

needed in reading the statements or in defining certain

words was given directly by the investigator.

Parents were then given instructions for completing

their questionnaires (FACES, the Behavior Problem Checklist,

the DAS and the TI). Whenever possible these questionnaires

were completed during the clinic visit. If this was not

possible, the parents were allowed to take them home and









were given a stamped, self-addressed envelope in which to

mail them back. Fathers were seldom present during these

visits. An identical set of questionnaires for the fathers

were sent home with the mothers with the request that they

be completed and returned by mail. It was emphasized that

completion of these questionnaires should be done independently

by each parent.

A total of 45 CF families were initially contacted by

letter and subsequently by phone. Of these, only one family

openly refused to participate. Twelve families who had

initially agreed to participate were unable to be tested

on account of scheduling problems and missing appointments.

One family did not return their questionnaires. Completed

data wereobtained from 31 mothers and their children and

20 fathers. Five of the mothers were single parents. Six

fathers failed to return their questionnaires.

Group 2: For the asthma group introductory letters

were not sent to parents prior to their clinic visits.

The reason for this was that computerized printouts of

weekly clinic appointments were not available for this

group. The procedure used to select these subjects and

administer the questionnaires required the following steps:

In the morning prior to Allergy Clinics, the investigator

reviewed the charts of all the patients with appointments

for that day. From this review a list was made of all the

patients who had a primary diagnosis of "asthma" and who

met the above stated criteria (7 to 12 years old, white,

on daily medications, etc.).









Upon entering the clinic and prior to meeting with

their doctors, parents of these children were approached

by the investigator, informed of the nature of the study

and invited to participate. Consenting parents and children

then followed the same procedure for obtaining information

and completing questionnaires as outlined for the CF group.

Since the Allergy Clinic visits were usually of shorter

duration than the CF visits, very few parents of asthmatic

children completed their questionnaires while in the hospital.

Of the 40 adults who were approached in this manner,

only 2 refused to participate. The reason for these refusals

had to do with the time factor involved. One family was

excluded because the child was a foster child. One other

was excluded because the parents could not speak English.

Thirty-eight asthmatic families agreed to participate in

this study. Although the child's questionnaire was completed

for all 38 children, parental responses were never returned

for 11 of these families (even after follow-up phone calls).

Data wereobtained from a total of 26 mothers and 18 fathers.

Five of these mothers and one of the fathers were single

parents. Of the intact families that participated, 3

of the fathers failed to return the questionnaires.

Group 3: A control group of "normal" subjects outside

the hospital setting was difficult to obtain. A nearby

school system agreed to sponsor the project, but for reasons

of confidentiality, could not provide names of appropriately

matched students. Taking this problem into consideration,









the following procedure was used to obtain and test subjects:

One elementary school was chosen as the source of all con-

trols. A letter was then drafted, addressed simply to

"Parents," describing the study and inviting families to

volunteer (a copy of the letter can be found in Appendix D).

Copies of these letters were distributed by the teachers to

20 randomly selected students in each grade, 2 through 6

(the grades most likely to have children between 7 and 12

years of age). These children were told to bring the letters

home to their parents. The investigator had previously met

with each of these teachers to describe the randomization

process. This process consisted of selecting every other

student from an alphabetical list of all the white students

in the classroom. If more students were needed after going

once through the list, teachers were to start at the begin-

ning and select those not previously chosen.

Enclosed with each letter was a self-addressed post

card which was to be returned to the investigator if parents

agreed to participate. The post cards included a space for

"name", "telephone number," and for indicating the "age"

and "sex" of the targeted child. When the post cards were

received, the investigator called the parents directly and

arranged a time to visit them at their homes in order to

administer the questionnaires. The procedure for administer-

ing the questionnaires was identical to that used with the

two other groups.









From the 100 letters sent home, responseswere received

from 31 families. Because of scheduling difficulties, 2 of

these families could not be seen. The investigator was

eventually able to administer the questionnaires to 29 of

these families. One of these families did not return the

questionnaires by the stated deadline and could therefore

not be included in the analysis. In all, completed data

wereobtained for 27 mothers and 18 fathers. Four of the

mothers and one of the fathers were single parents. Of

the intact families 4 fathers failed to return the question-

naires.










CHAPTER IV
RESULTS



Results will be presented separately for each of the

5 major hypotheses. Although data was collected from both

parents whenever possible, the primary testing of hypotheses

1, 3, 4 and 5 (those relying on parental variables) was

based solely on data obtained from the mothers. This was

done for two reasons. First, it was assumed that it was

usually the mother who was the primary caretaker of the

child. Second, data weremuch easier to obtain from the

mothers than from the fathers, as fathers more frequently

reported not having the time or the inclination to complete

the questionnaires. Although not reported in this section,

separate analyses were done for each of the hypotheses using

data obtained from the fathers. These results are presented

in tabular form in Appendix E.



Subject Match



As previously discussed,an effort was made to match

the children in each group according to race, age and sex.

All of the 86 children studied were white. Table 1 presents

comparison data for the three groups of children on average

age, age range and "percent male." Also included in this

table is the percentage of children in each group living

in single parent families. As is apparent, the three










groups of children selected as subjects for this study were

similar on all of the above criteria.



Table 1. Comparison Data for the Groups of Children Selected
as Subjects



Average Average From Single
Group Age (yrs) Range (yrs) % Male % Parent Families

CF 9.52 7 12 61 19

Asthma 9.18 7 12 56 22

Control 9.46 7 12 57 21




Relationship Between Chronic Childhood
Illness and Family Functioning



In order to test the hypothesis that families caring for

children with chronic illness are more likely to be functioning

at the extremes within the Circumplex Model of adaptability

and cohesion (hypothesis 1), FACES scores were utilized in

the following manner: First, grand means were calculated

for adaptability and for cohesion based on the pooled data

from all 84 mothers. These grand means were 172.77 and

259.08 respectively. Next, a deviation score was calculated

for each of the mothers on each of the two dimensions.

These scores were obtained by taking the absolute value of

the difference between a mother's score on one dimension

and the grand mean for that dimension. In this way, two

deviation scores (one for adaptability and one for cohesion)





58



were calculated for each family based on the mother's

responses to FACES. The deviation score for adaptability

(Dev ADP) can be represented in the following way:

Dev ADP = / ADP ADP /, where ADP is the family's adapta-

bility score and ADP is the grand mean for adaptability.

Similarly the deviation score for cohesion (Dev COH) can

be represented as Dev COH = / COH COH /.

Because families with high scores are those functioning

toward the extremes of the dimensions these deviation scores

make it possible to test hypotheses la, 2b, and Ic. Table 2

shows the means of Dev ADP and Dev COH for the three groups

of families. Differences between the group means were

examined by three MANOVA procedures.



Table 2. Mean Dev ADP and Dev COH Scores for Families
of CF, Asthma, and Control Children
(based on mother's response)




Group N Mean Dev ADP Mean Dev COH


CF 31 13.46 17.60

Asthma 26 9.45 19.16

Control 27 8.70 14.93




Results of MANOVA analyzing the difference between the

deviation scores of the CF and control group are presented











in Table 3. The overall F value of 1.27 was not significant,

and therefore hypothesis la was not supported. Results of

individual ANOVA's for each of the dimensions (Dev ADP and

Dev COH) are also presented in Table 3; these, too, were

not significant.



Table 3. F Scores for MANOVA Examining the Differences between
CF and Control Families on Dev ADP and Dev COH
(based on mother's response)




Source df F Value p


Dev ADP 1 2.32 .134

Dev COH 1 0.67 .415

Dev ADP/Dev COH 2,55 1.27 .289


In Table 4, the result of the MANOVA analysis of the

difference between the asthmatic and control families are

presented. The overall F value of .98 was not significant,

nor were the individual ANOVA's for the two dimensions.

Hypothesis Ib is therefore not supported.









Table 4. F Scores for MANOVA Examining Differences between
Asthma and Control Families on Dev ADP and Dev COH
(based on mother's response)




Source df F Value p


Dev ADP 1 0.13 .723

Dev COH 1 1.44 .235

Dev ADP/Dev COH 2,50 0.98 .384




To test hypothesis Ic, a MANOVA was performed examining

the difference in mean deviation scores between the CF and

the asthma groups. These results are presented in Table 5.

Here, too, the overall F and individual ANOVA's did not yield

significant results.


Table 5. F


Scores for MANOVA Examining Differences between
CF and Asthma Families on Dev ADP and Dev COH
(based on mother's response)


Source df F Value p


Dev ADP 1 1.63 .208

Dev COH 1 0.16 .692

Dev ADP/Dev COH 2,54 1.01 .372










Hypotheses la, Ib, and Ic were all unsupported by the

above results. Caring for a child with a chronic illness

was not related to extreme family functioning within the

Circumplex Model of cohesion and adaptability. Furthermore,

there did not appear to be a difference identified by this

model between families of terminally ill children and those

of children with a non-fatal chronic illness.



Relationship Between Chronic Childhood
Illness and Child Adjustment



Hypotheses 2a, 2b and 2c concern differences between

the three groups of children on two dependent variables,

self concept (P-H score) and number of behavior problems

(BPC scores). Data were analyzed separately for each depen-

dent variable.

Table 6 shows the means and standard deviations for

P-H scores of the three groups of children. Hypotheses 2a

and 2b suggest that the mean P-H scores for the CF and

asthmatic children, respectively, will be less than the

mean for the control children. Hypothesis 2c states that

the mean CF score will be less than the mean asthma score.

Results of the t-tests performed to test these hypotheses

are presented in Table 7. None of these t-tests approached

significance. Based on these results, it appears that the

child's self concept is not significantly affected by the

presence of a chronic illness.










Table 6. Mean P-H Scores and Standard Deviations for CF,
Asthma and Control Children



Standard
Group N Mean Deviation


CF 31 63.87 10.22

Asthma 27 62.82 13.41

Control 28 64.54 7.08

Total 86 63.76 10.41


Table 7. Results of t-tests Examining Differences between
Groups on Mean P-H Scores



Group
Comparisons Value t-Value p


CF vs Control -0.66 -0.24 .809

Asthma vs Control -1.72 -0.61 .545

CF vs Asthma 1.06 0.38 .704


When using behavior problems as the dependent variable,

however, the presence of a chronic illness did appear to

have a significant negative effect on the child's adjustment.

Table 8 presents the means and standard deviations of BPC

scores for the three groups of children. T-tests were used

to examine hypothesized differences between the groups.

These results are shown in Table 9. As expected, CF children










had significantly more reported behavior problems than did

the control children (p <.05). The difference between the

asthma and control children on behavior problems was also

significant (p <.01) and in the predicted direction. Thus,

both hypothesis 2a and 2b were supported when using the BPC

as the dependent variable.


Table 8.


Mean BPC Scores and Standard Deviations for CF,
Asthma and Control Children
(based on mother's response)


Standard
Group N Mean Deviation


CF 31 4.97 5.61

Asthma 26 5.88 5.27

Control 27 2.26 3.22

Total 84 4.38 5.03


Table 9. Results of t-tests Examining Differences between
Groups on Mean BPC Scores
(based on mother's response)




Group
Comparisons Value t Value p


CF vs. Control 2.71 2.12 .037*

Asthma vs Control 3.63 2.72 .004**

CF vs Asthma -0.92 -0.71 .480


*significant at p = .05
**significant at p = .01










Hypothesis 2c was not supported by these data. There

was no significant difference between the mean BPC score

for the CF children and that for the asthma children. The

number of behavior problems was related more to the presence

of a chronic illness than to differences in severity or life-

threatening status.



Relationship Between Chronic Childhood Illness
and Parental Attitudes Toward Death



Hypotheses 3a, 3b and 3c concern differences between

the three groups on two-measures of parental death orienta-

tion, death threat and death anxiety. It was expected that

parents of both groups of chronically ill children would be

experiencing greater death threat and death anxiety than

parents of healthy children. It was also hypothesized that

death threat and death anxiety would be greater in CF parents

than in asthmatic parents, given the life-threatening nature

of the former illness. Again, the two dependent variables

were analyzed separately.

Table 10 presents the means and standard deviations

of DAS scores for the three groups of mothers. Results of

t-tests performed on this data to assess group differences

are presented in Table 11. None of the t-tests were signi-

ficant. Apparently, mothers' death anxiety (as measured

by the DAS) was not affected by the presence or severity

of chronic childhood illness.










Table 10. Mean DAS Scores and Standard Deviations for
Mothers of CF, Asthma, and Control Children



Standard
Group N Mean Deviation


CF 31 7.81 2.94

Asthma 26 6.69 2.75

Control 26 7.96 2.62

Total 83 7.51 2.80




Table 11. Results of t-tests Examining Differences between
Groups on Mean DAS Scores of Mothers



Group
Comparisons Value t Value p


CF vs Control -0.16 -0.21 .835

Asthma vs Control -1.27 -1.64 .104

CF vs Asthma 1.11 1.51 .136




In examining the differences in death threat between the

groups, some very interesting findings emerged. As is apparent

from Table 12, which lists the means and standard deviations

of TI scores for the three groups, the results were in the

opposite direction from what was hypothesized. Mothers of CF

children were least threatened by the thought of death; the

mothers of the healthy children, on the other hand, were the

most threatened of the three groups. T-tests used to make





66



group comparisons (reported in Table 13) revealed that differ-

ences in TI scores between the CF group mothers and the control

group mothers were significant at the .05 level. In other

words, mothers of CF children were significantly less threaten-

ed by the thought of death than were mothers of healthy chil-

dren. T-values for the other comparisons were not significant.

The implication of this finding will be discussed in the

next section.


Table 12.


Mean TI Scores and Standard Deviations for Mothers
of CF, Asthma, and Control Children


Standard
Group N Mean Deviation


CF 31 15.52 9.07

Asthma 23 17.48 7.56

Control 27 20.26 9.52

Total 81 17.66 8.95


Table 13. Results of t-tests Examining Differences between
Groups on Mean TI Scores of Mothers



Group
Comparisons Value t Value p


CF vs Control -4.74 -2.04 .045*

Asthma vs Control -2.78 -1.11 .271

CF vs Asthma -1.96 -0.81 .422


significant at p = .05








Relationship Between Child Adjustment
and Family Functioning



Since there was no relationship found between family

functioning and the presence of childhood illness (hypotheses

1) there was no need to statistically control for effects of

the latter in examining hypothesis 4. To test the hypothesis

that extreme family functioning (defined in terms of Dev ADP

and Dev COH) is negatively related to child adjustment, two

multiple regression analyses were performed with the pooled

data (N=86). Both analyses used Dev ADP and Dev COH as the

independent variables. P-H scores were used as the dependent

variable in the first analysis; BPC scores in the second.

Results of the multiple regression analysis examining

the relationship between Dev ADP/Dev COH and the child's

self-concept (P-H score) are presented in Table 14. An F

value of 2.97 was obtained from this analysis; this value

just missed significance at the .05 level (p = .059).

Although not significant, this finding is in the predicted

direction.

When the reported number of behavior problems of the

child (BPC score) was used as the dependent variable, the

multiple regression analysis (reported in Table 15) yielded

highly significant results (p = .001). In other words, there

was a significant relationship between extreme family function-

ing and the number of behavior problems reported in the chil-

dren. This significant result--along with the tendency reported

above for self-concept--provides support for the Circumplex









theory, which suggests that healthy families function in the

moderate ranges on both cohesion and adaptability.


Table 14. F Scores for Multiple Regression Analysis of the
Relationship between Child's P-H Scores and Family
Dev ADP and Dev COH (based on mother's response)
for the Pooled Data




Source N df F Value p


Dev ADP 86 1 2.26 .137

Dev COH 86 1 7.61 .061

Dev ADP/Dev COH 86 2,81 2.94 .059



Table 15. F Scores for Multiple Regression Analysis of the
Relationship between Child's BPC scores (mother's
rating) and Family Dev ADP and Dev COH (based on
mother's response) for the Pooled Data



Source N df F Value p


Dev ADP


Dev COH

Dev ADP/Dev COH


1

1

2,81


7.09

8.43

7.76


.009**

.005**

.001**


significant at p = .01


Relationship Between Parental Orientation
Toward Death and Adjustment in the CF Child


Within the CF group, Pearson r correlation coefficients

were calculated between each of the measures of child adjust-

ment (P-H and BPC) and the mother's TI and DAS scores. The









correlations between the child's P-H score and the mother's

TI and DAS scores are presented along with p values in

Table 16. Again, death anxiety did not appear to be an

important variable in the analysis. Death threat, on the

other hand, did have a low level although statistically

insignificant relationship to child's self-concept (r = .25).

The relationship, however, was in the opposite direction

than was expected. Mothers who were most threatened by

the thought of death tended to have children with higher

self-concepts.



Table 16. Pearson r Correlation between CF Children's P-H
Scores and Mothers' TI and DAS Scores



Parental
Attitude N r p


Death Threat (TI) 31 .25 .091

Death Anxiety (DAS) 31 -.04 .426


To see if this relationship was unique to the CF group,

Pearson r correlation coefficients were calculated between

P-H scores and the mother's TI scores for each of the

other two groups, and then for the pooled data. Results of

this analysis are presented in Table 17. As is apparent

from this table, the unexpected relationship found in the CF

group between children's self-concept and maternal death

threat also held true for each of the other groups. Further-

more, when the datawere pooled, the correlation became









significant (p = .04). In other words, for all families

regardless of illness, high maternal death threat appeared

to be associated with high self-esteem of the children.

This finding is not easily explained and will be discussed

in the next section.



Table 17. Pearson r Correlations between Children's P-H
Scores and Mothers' TI Scores for CF, Asthma
and Control Groups and for Pooled Data




Group N r p


CF 31 .25 .091

Asthma 23 .20 .187

Control 27 .17 .201

Pooled 81 .20 .040*


*significant at p = .05



Table 18 presents the Pearson r coefficients and p values

for the CF group for correlations between BPC scores and

maternal death threat and death anxiety. Again, death anxiety

was not an important variable in the analysis. The correla-

tion of -.18 between TI scores and BPC scores was neither

significant nor as close to significance as that between

the TI and P-H scores. It is interesting, however, that

this correlation was also in the opposite direction from

what was expected. High maternal death threat was associated

in a low level and statistically insignificant way with few

reported behavior problems in the children.









Table 18. Pearson r Correlations between CF children's
BPC Scores (based on mother's rating) and
Mother's TI and DAS Scores



Parental
Attitude N r P


Death Threat (TI) 31 -.18 .167

Death Anxiety (DAS) 31 .11 .280















CHAPTER V
DISCUSSION



In general, the results of this study support the notion

that chronically ill children and their families adjust fairly

well despite the stresses imposed by the presence of the ill-

ness. Family functioning does not appear to be significantly

different in families of chronically ill children and families

of healthy children. Although there does seem to be a differ-

ence in psychological adjustment between sick and healthy

children, this difference may be more related to mediating

factors (like family functioning) than to the illness per

se. This conclusion is consistent with recent research.

Parental orientation toward death (at least as measured by

the TI) does seem to be affected by the presence of a termi-

nally ill child, but its importance as a mediating variable

affecting the child's adjustment is unclear.

Before discussing the implications of the major find-

ings of this study, a word should be said about limitations

due to the sampling procedures. The procedure used to obtain

subjects from the two clinic populations was somewhat differ-

ent. Prior to their first meeting with the investigator,

parents of CF children received an introductory letter and

then a follow-up phone call from this investigator. In this









way they were well-oriented to the purpose of the study and

had ample opportunity to think about their pariticpation

and to ask questions. Parents of asthmatic children were

not oriented in such a personal and concerned manner. These

parents were approached by the investigator one time only

and immediately given the questionnaires to be completed at

home. This difference in procedure could account for the

lower questionnaire return rate for the asthmatic parents

than for the CF parents, which could be a selection bias.

Selection bias is even more evident in the procedure

used to obtain control subjects. For reasons of confiden-

tiality, families in this group could not be randomly

selected and individually approached. The procedure adopted

required willing families to take the initiative to contact

the investigator rather than vice versa. The obvious draw-

back of this procedure was that only interested and motivated

control families were ultimately selected (only 1/3 of the

families blindly invited). The clinic families did not

have the luxury of anonymously refusing to participate.

Another difference between the control group and the

clinic groups was that the control children were all from

the same town and all attended the same elementary school.

This suggests that they are probably a more homogeneous

group than the groups of chronically ill children. There

is also some reason to suspect a difference in socio-

economic status between the groups. Although SES was not

formally assessed, the control children appeared to be










from predominantly upper-middle class families. The sick

children, on the other hand, appeared to be from families

representing a wider range of socio-economic levels.

These differences in samples and sampling procedures

are limitations of the study which should be taken into

account when interpreting the results. It has been shown

however, that the three groups of children selected as

subjects for this study were similar on a number of

important dimensions, i.e., age, race and sex. In addition

they were all from roughly the same geographic area. As

previously mentioned, few studies in this area have used

healthy control groups for purposes of comparison. It is

felt that the control group used in this study, although

not perfect, does represent a methodological improvement

over the use of standardized norms.

The first major finding in this study was that the

presence of chronic childhood illness was not significantly

related to extreme family functioning (as assessed by

mother's Dev ADP and Dev COH scores on FACES). The

Circumplex Model of family systems proposes that families

under stress have a tendency to move toward the extremes

on adaptability and cohesion. If this model is correct,

and if FACES and the method used to interpret FACES scores

are valid, than there are two possible explanations for

the above result: Either families of chronically ill chil-

dren are not experiencing stress associated with the illness,

or they are successfully coping with the stress that does

exist.











This second explanation seems most plausible. Caring

for seriously ill children has to be inherently stressful

for the affected families. It does not have to lead to

deficits in family functioning, however. Circumplex theory

suggests that healthy families may move toward the extremes

of adaptability and cohesion when initially confronted with

a stressful event but will eventually gravitate back toward

the moderate ranges as they successfully cope with these

events. It is possible that this is exactly what the families

of chronically ill children have done; i.e., they have adapted

to the stress in a normal and healthy manner.

An interesting test of the Circumplex Model would be

to study changes in family functioning relative to chronic

childhood illness over the course of the disease. If the

theory is correct, one would expect healthy families to move

toward the extremes on adaptability and cohesion at the

time of diagnosis and during periods of acute exacerbations

of the illness, but to gravitate back toward the middle as

the illness is stabilized. Dysfunctional families could

therefore be identified by their inability to move back

and forth in this fluid manner.

Another factor that should be considered in interpreting

the above result is that FACES may not be an appropriate

instrument for assessing group differences. To date, FACES

has been used primarily as a clinical tool for identifying

problem families. Although apparently successful in this

regard, it may not be sensitive enough for assessing subtle










differences in group trends. It should be noted that even

though the differences between groups in extreme family

functioning were not significant, they were in the predicted

direction (refer to Table 3). The mean deviation scores

for families of chronically ill children were higher than

those for families of healthy children, and CF families

tended to deviate more than asthmatic families.

The method developed in this study for converting raw

FACES scores on adaptability and cohesion to deviation

scores has not previously been used as a measure of extremity

rating. For this reason it could be concluded that the

method has no established validity and therefore yields

meaningless results. Given the wide range of raw scores

considered to be indicative of healthy functioning, it could

be argued that moderate deviations from the means would not

be particularly important. However, the validity of this

procedure was supported by the relationship found to exist

between deviation scores and child adjustment.

As hypothesized, mother's Dev ADP and Dev COH scores

were significantly related to the number of behavior prob-

lems exhibited by the child. The relationship between

these deviation scores and the child's self-concept was

extremely close to significance. These results are inter-

esting in that they support the use of Dev ADP and Dev COH

scores as measures of extremity ratings while providing

evidence validating the Circumplex Model. Circumplex

theory suggests that dysfunctional families will be located










more toward the extremes on cohesion and adaptability than

will healthy families. If one makes the logical assumption

that there is a higher frequency of poorly adjusted children

in dysfunctional families than in healthy families, then

according to the theory, there will be a higher frequency

of poorly adjusted children in families located at the

extremes of the Circumplex dimensions than in families

located at moderate levels along the dimensions. This is

essentially what was found in this study. This result was

obtained for pooled data regardless of illness. The implica-

tion of this finding in terms of adjustment in chronically

ill children will be discussed below.

Results concerning the relationship between chronic

illness and the psychological adjustment of the afflicted

children were mixed. As expected, chronically ill children

had significantly more behavior problems than healthy chil-

dren. Contrary to expectation, however, they did not differ

in self-concept. These two variables were examined inde-

pendently because they were thought to be measuring very

different aspects of the child's adjustment. Self-concept

relates to an internal feeling or perception of self-worth,

while behavior problems reflect a learned style of outward

interaction with the environment.

In speculating about the reason for the discrepancy

noted above, a possible explanation arises involving the

relationship established between sick child and parent

(especially mother). It has often been noted that there


I









is a tendency for sick children to be overprotected by their

mothers and to be treated as special (Travis, 1976). It

seems quite possible that mothers may compensate for their

own feelings of guilt and helplessness in the face of their

child's misfortune by being more tolerant of misbehavior

in these children and less likely to use consistent discipline.

This could have the effect of making these children more prone

to developing behavior problems without necessarily affecting

the way they feel about themselves. The relationship between

parental discipline and behavior problems in chronically ill

children has not previously been explored and might be a

direction to pursue in future research.

In general, the results of this study concerning adjust-

ment in chronically ill children are consistent with previous

research. There does appear to be some relationship between

childhood illness and the presence of adjustment difficulties

in afflicted children. This relationship is not strong

however, and appears to be more a function of mediating

variables than of the illness per se. In this study,

family functioning did appear to be an important mediating

variable in the interaction between illness and adjustment

variables.

As previously discussed, the relationship found between

extreme family functioning and behavior problems in the

children was highly significant. In fact, it was much more

significant than the relationship found between chronic ill-

ness and behavior problems. This raises the question of











whether the higher incidence of behavior problems reported

in both groups of sick children than in the healthy children

might be due more to differences in family functioning than

to the presence of the illness, even though these differ-

ences were not found to be significant. In any event, it

can be concluded from the results of this study that adjust-

ment difficulties in children (defined in terms of behavior

problems and self-concept) are more related to extreme

family functioning than to the presence of chronic childhood

illness. This conclusion suggests that family functioning

is an important mediating variable to consider in childhood

adjustment to chronic illness.

Parent's orientation toward death was also explored as

a possible mediating variable affecting psychological adjust-

ment in sick children. It was hypothesized that having a

child with a chronic physical illness--especially if this

illness was terminal--would have the affect of making parents

anxious and threatened by the thought of death. Furthermore,

it was felt that with fatally ill children there would be

a relationship between child adjustment and parental orienta-

tion toward death, since parents most threatened and anxious

about death would be least able to provide an open and sup-

portive atmosphere for the child, relative to his/her illness.

In exploring these hypotheses, death anxiety, as

measured by the DAS, was not found to be an important variable.

The DAS is a very short, true-false instrument which may not

be sensitive enough to subtle differences between groups











in orientation toward death. Another possible explanation

for this finding is that anxiety associated with death may

not be a stable entity in parents, but may fluctuate

markedly relative to the situation. This hypothesis is

consistent with results reported by Neimeyer et al. (1977)

that DAS scores and situational anxiety scores increased

significantly in a group of subjects after viewing a 30

minute film depicting war atrocities. If this is the case,

one would expect parental death anxiety to increase during

exacerbations of the illness but to subside when the illness

is stabilized. As previously mentioned, all of the sick

children in this study were in stable condition when tested.

It would be interesting to explore changes in DAS scores

over the course of the illness.

Death threat, as measured by the TI, was found to be

an important variable in the above analysis, but in an

unexpected way. Contrary to prediction, death threat was

significantly higher in parents of healthy children than

in parents of terminally ill children. This results in a

serious question about the validity of the primary assumption

underlying the initial hypothesis. This assumption was that

parents of dying children react to the terminal nature of

the illness in an unhealthy manner. That is, they deny

the reality of the illness, they are fearful of and cannot

accept the prognosis, and (ultimately) they become less

able to conceive of death as a personal reality (death










threat). The fact that the opposite was actually found

to be the case attests to generally healthy rather than

unhealthy parental adjustment to illness. This suggests

that parents of terminally ill children have thought about

the implications of death for their children, and have

consequently restructured their own personal construct

systems in order to make death more compatible with their

perceptions of self.

When the initial assumption is changed concerning a

healthy vs. unhealthy parental response to the implication

of death in terminal childhood illness, the above results

are quite consistent with personal construct theory. This

theory proposes that a person's cognitive orientation toward

his/her world is generally stable (especially for constructs

that are particularly meaningful for the individual), but

that this cognitive orientation does change in response to

significant life experience. Obviously, having a child

with a terminal illness is a significant life experience.

The results of this study show that, as predicted by the

theory, the personal construct systems of mothers are

affected by this experience.

The results obtained concerning the relationship

between maternal death threat and adjustment in the fatally

ill children are difficult to understand. Contrary to

expectation, there was a tendency for greater maternal

death threat to be associated with fewer behavior problems

and higher self-concepts in the CF children. If this is more









than just a spurious result it could call into question

the assumption discussed above, that lower parental death

threat is a healthy response to the presence of terminal

childhood illness.

It is particularly surprising that this relationship

seems to hold true for all three groups of families. One

can only speculate as to the meaning of this finding.

Possibly, what is being measured by the TI is a more general

personality trait than is usually assumed. It could, for

example, be tapping a person's tendency to remain impersonal

and matter-of-fact in the presence of emotionally-related

stimuli. Someone who is able to be objective and emotionally

unresponsive to topics (like death) that usually arouse some

affect in people might also be overly objective and emotionally

unresponsive to people. A parent with this tendency might

have difficulty meeting the emotional needs of his/her

children, which could eventually result in problems of adjust-

ment for these children. This is obviously pure specula-

tion; there is no evidence currently available substantiating

this circuitous connection. However, the finding does pose

some interesting questions which are deserving of further

attention.

In conclusion, the present study supports the contention

that chronic childhood illness has a general stressful effect

which can lead to deficits in functioning, although in most

cases, individuals and families appear to adapt in basically

healthy ways. It also supports the notion that mediating










variables (family functioning, etc.) are more important in

affecting the psychological adjustment of the sick child

than is the illness itself.

More research is needed in this area in order to better

understand the process of adaptation to childhood illness.

Two lines of research would be useful in this regard. One

line would involve continuing the search for mediating

variables that are important in the adjustment process.

Once these variables have been identified and clearly

defined, systematic intervention would be possible with

the goal of insuring a healthy response to the disease.

At present, information concerning these mediating variables

is scattered and often times contradictory.

A second course for future research is in the direction

of longitudinal designs. There is presently little hard

data available concerning changes in individual and family

adjustment during the course of a child's chronic illness.

Cross-sectional studies such as the one reported here seem

to be based on the assumption that illnesses are static

entities which have a uniform effect. Diseases are dynamic,

as are the people who are affected. To really understand

the adaptational process, it is necessary to observe changes

over time. As the adjustment demands change over the course

of the illness, so too will the implications for treatment

and intervention.














Research concerning the psychosocial effects of chronic

childhood illness generally supports the view that the

presence of disease is frequently associated with deficits

in individual and family functioning. There has been a

gradual shift in recent years from considering the disease

as causative in this association to emphasizing the importance

of mediating variables. The present study was an attempt to

increase our understanding of this process by examining the

relative impact of the illness per se versus mediating

variables (family functioning and parental attitudes toward

death) on individual adjustment in the chronically ill child.

Two major questions were addressed in this regard. The

first was general in nature and concerned the overall effect

of the presence and severity of illness on relevant adjust-

ment variables--behavior problems and self-concept of the

afflicted children, extreme family functioning, and parental

death threat and death anxiety. The second question examined

the relationship between adjustment in the children and both

family functioning and parental death orientation, when

controlling for effects of the illness. In this way the

impact of disease was compared to the impact of mediating

variables (family functioning and parental death orientation)

relative to childhood adjustment.

To explore these questions, three groups of families

were studied: families of children with cystic fibrosis,









families of children with chronic asthma, and families of

healthy children. Results of this study indicated that the

presence of chronic childhood illness was not associated

with significant deficits in family functioning or with

decreased self-concepts in the afflicted children. It was

found to be associated with an increased frequency in report-

ed behavior problems for the children. Contrary to expecta-

tion, death threat was significantly lower in mothers of

terminally ill children than in mothers of healthy children.

Maternal death anxiety was not affected by the presence of

childhood illness.

Family functioning was found to be a significant mediating

variable affecting childhood adjustment regardless of illness.

It was concluded that childhood adjustment was related more

to this variable than to the presence of a chronic illness.

The implications of these results were discussed, as were

avenues for future research.










APPENDIX A

TEMPLER'S DEATH ANXIETY SCALE


Below are 15 statements concerning attitudes toward death or death
related issues. For each statement you are to check whether that attitude
is mostly "true" or mostly "false" for yourself. Please be sure to respond
to all 15 statements by checking either "true" or "false".


True False

1. I am very much afraid to die.

2. The thought of death seldom enters my mind.

3. It doesn't make me nervous when people talk about
death.

4. I dread to think about having to have an operation.

5. I am not at all afraid to die.

6. I am not particularly afraid of getting cancer.

7. The thought of death never bothers me.

8. I am often depressed by the way time flies so
very rapidly.

9. I fear dying a painful death.

10. The subject of life after death troubles me
greatly.

11. I am really scared of having a heart attack.

12. I often think about how short life really is.

13. I shudder when I hear people talking about a
World War III.

14. The sight of a dead body is horrifying to me.

15. I feel that the future holds nothing for me to
fear.











APPENDIX B


THREAT INDEX (TIp40)





Below is a list of bipolar dimensions. For each dimensions please
circle the side with which you see yourself more closely associated.
For example, do you associate yourself more with the term "predictable"
or "random"?


predictable

empty

sad

personal

lack of control


random

meaningful

happy

impersonal

control


productive

learning

purposeful

responsible

bad


unproductive

not learning

not purposeful

not responsible

good


satisfied dissatisfied not caring caring


relating to not relating crazy healthy
others to others personality

pleasure pain conforming not conforming


feels bad feels good animate inanimate


objective subjective weak strong


alive dead useful useless

helping others being selfish closed open


specific general peaceful violent


kind cruel freedom restriction


incompetent competent non-existence existence


insecure secure understanding not
understanding

static changing sick healthy


unnatural natural stagnation growth

calm anxious abstract concrete

easy hard hope no hope
















For each dimension circle the side with which you most closely
associate your own death.


predictable

empty

sad

personal

lack of control


random

meaningful

happy

impersonal

control


productive

learning

purposeful

responsible

bad


unproductive

not learning

not purposeful

not responsible

good


satisfied dissatisfied not caring caring


relating to not relating crazy healthy
others to others personality

pleasure pain -conforming not conforming


feels bad feels good animate inanimate


objective subjective weak strong


alive dead useful useless

helping others being selfish closed open


specific general peaceful violent


kind cruel freedom restriction


incompetent competent non-existence existence


insecure secure understanding not
understanding

static changing sick healthy


unnatural natural stagnation growth

calm anxious abstract concrete

easy hard hope no hope














APPENDIX C


INTRODUCTORY LETTER TO PARENTS OF CF CHILDREN





Dear Parents,

We would like to invite you to participate in a psychological study being
conducted at Children's Hospital in Boston. The purpose of the study is to
explore psychosocial affects of chronic childhood illness on families and
afflicted children. We hope to be able to use the information obtained in this
study to better help families, like your own, adjust to the stress of caring
for a child with cystic fibrosis.

If you agree to participate in this study you (the mothers) will be asked
to complete a series of four questionnaires during your next clinic visit.
These questionnaires will require you to think about your present family situa-
tion, the behavior of your child, and certain personal attitudes you have toward
the idea of death. The questionnaires will take roughly one hour to complete.
During this time your child will also be interviewed by the investigator for
about 15 minutes. The purpose of this interview is to assess his/her self
esteem; i.e., how he/she feels about himself/herself. The idea of death will
not be discussed with the child.

At the conclusion of the study all participating families will be given
the opportunity to meet with the investigators to discuss the general findings.
If desired, suggestions might also be made concerning the adjustment needs
of specific families.

A few days prior to your next clinic visit you will be called by one of
the investigators to discuss whether or not you wish to participate. At that
time we will be happy to answer any questions you may have concerning this
study. Participation is, of course, voluntary.

Sincerely,



Kon-Taik Khaw, M.D. Brian L. Lewis, M.ED.
Senior Associate in Medicine Co-Investigator
CyEtic Fibrosis Program Cystic Fibrosis Program
Children's Hospital Medical Center Children's Hospital Medical Center












APPENDIX D


INTRODUCTORY LETTER TO PARENTS OF HEALTHY CHILDREN









Dear Parents,

We, at Children's Hospital, have undertaken a study in which we are examining
psychological effects of chronic childhood illness. We are specifically interested
in looking at how the presence of chronic childhood illness effects family func-
tioning, parental attitudes, and the behavior and self concept of the afflicted
children. We hope to be able to use the information obtained in this study to bet-
ter help families adjust to the stresses of caring for a child with special needs.

Two groups of families are being studied at Children's Hospital; families
of children with cystic fibrosis and families of children with chronic childhood
asthma. In order to determine which effects are due specifically to the presence
of the physical illness it is necessary to compare the data obtained from these
families with that of a control group of "normal" families; i.e., families of
children with no previous history of chronic physical illness. We would like
to invite you to participate as one of these control families.

Participation in this study will only require approximately one hour of your
time. During this hour you will complete four questionnaires. These questionnaires
will require you to think about your present family situation, the behavior of
your child, and certain personal attitudes you have toward the idea of death.
Children in this study will be asked to complete one questionnaire designed to
examine how he/she feels about himself/herself. This will require only about 15
minutes of her/her time. You will have a chance to examine all questionnaires
before you agree to participate.

We have sent these letters to parents of randomly selected Brookline children
who are roughly the same age as those under investigation at Children's Hospital.
If you feel you might be willing to volunteer for this study or if you would
simply like more information about it, complete and return the enclosed, self-
addressed postcard within the next few days. The only information needed on this
post card is your name, telephone number, and age and sex of your child. When
we receive the post card you will be called by the co-investigator, who will
answer any questions you may have. If you then agree to participate, this inves-
tigator will bring the questionnaires directly to your home at a mutually convenient
time. You will then have about a week to complete them and mail them back to
Children's Hospital.

Although this study is primarily concerned with adjustment in families of
children with chronic illness we feel that it will have generalized relevance for
all families. At the conclusion of the study interested participants will be
given feedback concerning the results. Your participation would be greatly
appreciated.

Sincerely,

P.S. We would appreciate your cooperation Brian L. Lewis
in returning the post cards to us Co-investigator
within the next two days. Cystic Fibrosis Program


















APPENDIX E


RESULTS OF ANALYSES BASED ON FATHER'S DATA










Table 20. Mean Dev ADP and Dev COH Scores for Families of
CF, Asthma and Control Children



Mean Mean
Dev Dev
Group N ADP COH


CF 20 12.39 13.88

Asthma 18 8.64 13.94

Control 18 9.02 13.99


Table 21. F Scores for MANOVA Examining the Differences between
CF and Control Families on Dev ADP and Dev COH



Source df F Value p


Dev ADP 1 1.47 .233

Dev COH 1 0.00 .970

Dev ADP/Dev COH 2,35 0.72 .496




Table 22. F Scores for MANOVA Examining the Differences between
Asthma and Control Families on Dev ADP and Dev COH



Source df F Value p


Dev ADP 1 0.03 .874

Dev COH 1 0.00 .989

Dev ADP/Dev COH 2,33 0.01 .988










Table 23. F Scores for MANOVA Examining the Differences between
CF and Asthma Families on Dev ADP and Dev COH



Source dr F Value p


Dev ADP 1 2.14 .152

Dev COH 1 0.00 .985

Dev ADP/Dev COH 2,35 1.06 .359


Table 24. Mean BPC Scores and Standard Deviations for CF, Asthma
and Control Children



Standard
Group N Mean Deviation


CF 20 3.90 5.98

Asthma 17 5.59 7.46

Control 18 1.89 2.14

Total 55 3.76 5.72




Table 25. Results of t-tests Examining Differences between
Groups on Mean BPC Scores



Group
Comparisons Value t Value p


CF vs Control 2.01 1.10 .277

Asthma vs Control 3.70 1.94 .057

CF vs Asthma -1.69 -0.91 .367

PAGE 1

FACTORS AFFECTING PSYCHOSOCIAL ADJUSTMENT IN CHRONICALLY ILL CHILDREN AND IN THEIR PARENTS BY BRIAN L. LEWIS A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1981

PAGE 2

ACKNOWLEDGMENTS There are a number of people I wish to acknowledge for their contributions to this dissertation. I would first like to thank Kon-Taik Khaw, M.D., Senior Associate in Medicine in the Cystic Fibrosis Program at Children's Hospital i Boston. It was Dr. Khaw who agreed to sponsor my research at Children's Hospital and through whom access was obtained to the chronically ill subject populations. Dr. Khaw's initial receptiveness to the project and his unfailing support and enthusiasm throughout are testimony to his concern for the psychological, as well as physical, well being of his patients. It was certainly my pleasure to have made his acquaintance. Dr. Paul Schauble served as the chairman of my doctoral committee and was my major advisor throughout my doctoral studies. I would like to thank Paul for his judicious guidance, uncanny savvy, and respect for my competencies and ned for independence. More than anyone else, Paul has been my mentor and role model, and his presence was felt in numerous and sundry ways throughout the preparation of this dissertation. The contributions of each of the faculty members on my committee has been significant. Dr. Franz Epting, as my co chairman, was particularly helpful in the initial, conceptualizing stages of the project. I have always appreciated ii

PAGE 3

the unstructured quality of Franz's thought processes. Dr. Hannelore Wass was a valued resource for her expertise in the area of death and dying. In my interactions with Dr. Wass I was continually impressed by her openness, genuine concern, and kind sensitivity. I would like to thank Dr. Harry Grater for his charismatic and settling presence, and Dr. Robert Ziller for my initial introduction to the disease of cystic fibrosis Although not a formal member of my committee, Dr. William Froming was sought out on numerous occasions for advice on statistical matters, which he freely gave. This study would not have been possible without the support and guidance of a number of individuals at Children's Hospital in Boston. I would particularly like to acknowledge the following people for their various contributions: Harvey Colton, M.D., Ms. Mary Williams, Ms. Jane Wally, Ms. Virginia Rice, Ms. Julie Henry, and Frank Twarog, M.D., Ph.D. I also wish to thank the Brookline, Massachusetts,School System for their sanction of the project, and the concerned Brookline residents who willingly volunteered to participate as controls in the study. Finally, I would like to thank Ms. Stephanie Schmitz for her editorial acumen, support and patience. iii

PAGE 4

CHAPTER TABLE OF CONTENTS PAGE ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . ii ABSTRACT. . . . . . . . . . . . . . . . . . . . . Vi I. PROBLEM STATMENT ................................. 1 II. REVIEW OF THE LITERATURE ......................... 4 Introduction. . . . . . . . . . . . . . . . . 4 Adjustment to Physical Illness ............... 6 Adjustment in Children with Chronic Illness.................................... 8 Family Adjustment to Chronic Childhood Illness .......................... 17 Psychosocial Aspects of Cystic Fibrosis (CF) .............................. 25 Parental Death Attitude and the Chronically Ill Child ...................... 31 Hypotheses. . . . . . . . . . . . . . . . . . 35 I I I METHODOLOGY. . . . . . . . . . . . . . . . . . . 4 0 Subjects ..................................... 40 Instruments... . . . . . . . . . . . . . . . 43 Procedure. . . . . . . . . . . . . . . . . . 50 IV. RESULTS. . . . . . . . . . . . . . . . . . . . . 56 Subject Match ................................ 56 Relationship between Chronic Childhood Illness and Family Functioning ............. 57 Relationship between Chronic Childhood Illness and Child Adjustment .............. 61 Relationship between Chronic Childhood Illness and Parental Attitudes toward Death ...................................... 64 Relationship between Child Adjustment and Family Functioning ..................... 67 Relationship between Parental Orientation toward Death and Adjustment in the CF Child ...................................... 68 V. DISCUSSION ......................................... 72 iv

PAGE 5

APPENDICES A TEMPLER'S DEATH ANXIETY SCALE ....................... 86 B THREAT INDEX. . . . . . . . . . . . . . . . . . . . 87 C INTRODUCTORY LETTER TO PARENTS OF CF CHILDREN ....... 89 D INTRODUCTORY LETTER TO PARENTS OF HEALTHY CHILDREN. . . . . . . . . . . . . . . . . . . . 90 E RESULTS OF ANALYSES BASED ON FATHER'S DATA .......... 91 REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . 97 BIOGRAPH I CAL SKETCH ........................................ 105 V

PAGE 6

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FACTORS AFFECTING PSYCHOSOCIAL ADJUSTMENT IN CHRONICALLY ILL CHILDREN AND IN THEIR PARENTS By BRIAN L. LEWIS June 1981 Chairman: Paul Schauble, Ph.D. Cochairman: Franz Epting, Ph.D. Major Department: Department of Psychology The present study was an attempt to explore factors affecting psychosocial adjustment in chronically ill chil dren and their parents. Previous research has generally established that the presence of childhood illness is asso ciated with deficits in individual and family functioning; however, it is unclear whether these deficits are due to the illness per se, or mediating variables. The purpose of this study was to increase understanding of dynamics by examining the impact of the illness, versus the mediating variables of family functioning and parental orientation toward death, on adjustment in the chronically ill child. It was hypothesized that these mediating variables would be more important to the child's adjustment than would the illness. vi

PAGE 7

In testing this hypothesis, data were obtained from families of the following three groups of children: children with cystic fibrosis (N=31), children with chronic asthma (N=27), and children with no previous history of chronic illness (N=28). All of the children were Caucasian and between 7 and 12 years of age. The identified child in each family completed the Piers-Harris Children's Self Concept Scale. Parents completed the Family Adaptability and Cohe sion Evaluation Scales, the Behavior Problem Checklist, the Death Anxiety Scale, and the Threat Index. The primary testing of the hypotheses utilized mother's data. Results indicated that the presence of childhood illness was not associated with significant deficits in family func tioning or with decreased self-concept in afflicted children. Illness was associated with an increased frequency in behavior problems in the children (p < .05 for CF children, p < .01 for asthmatic children). Contrary to expectation, death threat was significantly lower in mothers of terminally ill children than in mothers of healthy children (p < .05). Family functioning was found to be a significant variable relative to behavior problems of the children, regardless of illness (p < ,001). It was concluded that childhood adjust ment was related more to this variable than to the presence of a chronic illness. The implications of these results were discussed as were avenues for future research. vii

PAGE 8

CHAPTER I PROBLEM STATEMENT All people experience periodic episodes of physical illness throughout their lives. Recovery from these episodes usually necessitatesshort-term change in normal habits and lifestyle. Many people are afflicted with illnesses that are chronic in nature; i.e., long-term and sometimes progressive and fatal. For them sickness is a way of life: recovery is often impossible and exacerbation of symptoms is a constant concern. Maintenance of a stable condition may require the adoption of a lifestyle somewhat different from "normal." The imposition of this illness-related style of life is often stressful for those affected and sometimes leads to deficits in psychological and social adjustment (Lipowski, 1970). Chronic illness in children is particularly problematic. The usual process of development is stressful enough for children without their having to contend with the burden imposed by a potentially debilitating sickness. More than adults, children must rely on the adaptive resources avail able in the family in order to cope with this stress. Normal, healthy maturation is dependent on this coping process. 1

PAGE 9

The topic of this dissertation concerns psychological and social adjustment in chronically ill children and their families. Most research that has been done in this area supports the idea that the presence of chronic childhood illness leads to deficits in individual and family function ing (see reviews by Bakwin & Bakwin, 1972; Magrab, 1978; and Mattsson, 1972). The problem with this research is that most studies have utilized very few subjects and inadequate controls, and have been primarily descriptive in nature, relying on clinical experience and subjective evaluations (Gayton & Friedman, 1973). In a recent review, Tavormina, Kastner, Slater, and 2 Watt (1976) criticized previous research for the implicit assumption that chronic illness leads to, or directly causes, emotional problems in children and families. They suggested that this assumption is erroneous because it implies that the chronically ill are necessarily a psychologically deviant population, when in fact many are emotionally well adjusted and lead essentially normal lives. These authors concluded that a more useful conception of the psychosocial effects of chronic illness is that of Green (1965), who suggested that the presence of illness places the child "at risk" (vulnerable) for emotional problems but is not, in itself, sufficient to produce psychopathology. It has been suggested that subsequent research, in addition to using more rigorous methodology, be concerned with identifying those specific factors that influence the

PAGE 10

3 psychosocial functioning of the sick child and the affected family. Gayton and Friedman (1973) proposed that mediating variables important to consider would be family size, reli gion, race and marital stability. Tavormina et al. (1976) expand this list by including degree of severity of disease, age of onset and duration, life threatening status and degree of noticeability of the illness. The study to be reported in later sections was an attempt to increase our understandi~g of the psychological and social effects of chronic childhood illness by examining some of the above issues. Two rv ain quest ions were addressed. The first was general in nature and concerned the overall effect of the presence and severity of chronic childhood illness on individual and family functioning. The second question was concerned specifically with psychological adjustment in the afflicted child. This question sought to examine the relative importance of the illness per se, vs. relevant mediating variables, in affecting the child's adjustment. Mediating variables of interest were family functioning and parental orientation toward death.

PAGE 11

CHAPTER II REVIEW OF THE LITERATURE Introduction Mattsson (1972) described chronic illness as "a disorder with a protracted course which can be progressive and fatal, or associated with a relatively normal life span despite impaired physical and mental functioning. Such a dis e ase frequently shows period~ of acute exacerbation requiring intensive medical attention" (p. 803). There are three fundamental qualities characteristic of chronic illness that can be extracted from the above description. First, the chronic illness is long-term, often beginning at birth or in early childhood and continuing for years or even life. Secondly, chronic illness is associated with some functional disability. This disability can be mental as well as physical. Thirdly, chronic illness implies the presence of a physical disease which is active in nature, progressively deteriora ting, or fluctuating between good and bad periods and requiring constant treatment to maintain stability. One problem in reviewing research in this area is the lack of consensus as to which diseases are chronic and which are not. The above definition is not sufficiently specific 4

PAGE 12

to make this distinction in all cases. Mattsson (1972) reported that when including visual and hearing impairments, mental retardation, and speech and learning disorders as chronic disorders, it is estimated that as many as 40% of all children suffer from one or more chronic illnesses. This is much higher than the 7% to 10% estimate of illness 5 of primarily physical origin (e.g., cystic fibrosis, leukemia, hemophilia, etc.) (Pless, 1968). Travis (1976) stated that there is a fundamental difference between chronic illness and physical/mental handicaps (hearing impairments, etc.), because they pose very different problems of adjustment for the affected individual ~ The literature reviewed below is concerned exclusively with the 7 to 10% population of serious illness with physical origins. Wherever possible, an attempt will be made to specify the exact nature of the illnesses 1 studied. In the following review a number of studies will be discussed which have attempted to take a wholistic view of disease by considering the psychological and social effects of chronic illness in children. For purposes of clarity the review will be divided into five main topics. The first topic is a brief review of various ways in which adjustment to phy~ical illness has been conceptualized. Of primary concern in this section is an appropriate defini tion "healthy adjustment" relative to the presence of chronic illness.

PAGE 13

6 The second topic is concerned with adjustment in children with chronic illness. It is a general overview of the litera ture pertaining to the child's own reaction to his/her illness and illness-related environment. The third topic area is also general in nature and concerns family adjustment to chronic childhood illness. The chronic illness which is the focus of the study reported in this dissertation is cystic fibrosis (CF). Literature pertaining to the psychosocial aspects of this specific illness is reported in detail in section four of this review. The final topic in this chapter concerns parental attitudes toward death and chronic childhood illness. Literature is briefly reviewed which suggests that the parent's attitudes about death create an atmosphere around the sick child that can be problematic for his/her adjustment. Adjustment to Physical Illness Adjustment to physical illness is quite variable and often difficult to define. Barker, Wright, Myerson and Gonick (1953) conceptualized adjustment in terms of the self concept of the afflicted individual. According to this view, disease affects the body's function as a tool for action. As this function is altered, so too is the individual's perception of self. Accord ing to Barker et al., a person's post-illness self concept is affected by his/her pre-illness self concept, the stress of the illness and its effect on his/her body image, and the response of family and environment to the illness.

PAGE 14

Another way adjustment has been conceptualized is in terms of ''coping processes." Lipowski ( 1970) defined coping in the physically ill as "all cognitive and motor activities which a sick person employs to preserve his body and psychic integrity, to recover reversible impaired functioning and compensate to the limit for any irreversible impairment" (p. 93). He went on to consider various "styles" of coping which he broke down into cognitive and behavioral aspects. In the cognitive domain, Lipowski identified two general modes for dealing with the facts of illness: "minimizing" and "vigilant focusing.~ In the behavioral domain, Lipowski proposes three basic coping styles: "tackling,' ; "capi tula tion," and "avoidance." Tackling implies the active engage ment with tasks imposed by the illness. Capitulation is seen in people who react passively, often regressing to levels of helpless dependency. Avoidance is the behavioral style usually linked with the cognitive "minimization; in which the illness ceases to become a focus and is in some cases denied. Mattsson (1972) proposed an adaptational model similar to that of Lipowski. He defined "coping" as, "all the adaptational techniques used by an individual to master a psychologic threat and its intendant negative feelings in order to allow him to achieve personal and social goals" (p. 257). Adjustment is defined as successful coping which results in effective functioning. 7

PAGE 15

8 Whether coping processes are considered adaptive or maladaptive depends on a number of factors. Verwoerdt (1972) groups these facotors into three major areas: "characteristics of the physical illness" (severity, duration and organ system involved), "characteristics of the host" (age, sex, body image and premorbid personality), and "situational factors" ("sick role," family relationships and processes of inter action). According to Verwoerdt: Adaptive coping leads to re-establishment of a dynamic equilibrium that was disturbed by stress. Maladaptive coping behavior typically leads to a vicious circle which, by depleting the patient's resources, aggravates the very problem it is supposed to solve .... Whether a defense is adaptive or maladapt'ive depends on its intensity as well as its appropriateness to the particular situation. ( p. 133) In reviewing some of the above theories, Pless and Pinkerton (1975) proposed what they term an "Integrated Model" of adapta tion to physical illness. According to this model, adjustment should be viewed in terms of the impact on self-concept as well as the breakdown in coping processes. These writers stressed the fact that healthy adaptation is not always directly related to the degree of acceptance of the reality of the condition. They prefer to view adaptation in terms of "psychological balance" or "freedom from abnormality." Adjustment in Children with Chronic Illness Children with chronic illness obviously have a more difficult time in life than those who are healthy. In addition to facing the usual problems of growth and develop

PAGE 16

ment, they are forced to adjust to atypical circumstances such as frequent hospitalizations, retarded growth, special diets, constant medication treatments and frequent absences from school. These factors combine to continually remind these children that they are different from others. The resultant stress makes them "at risk or vulnerable to psychopathology (Green, 1965). Much evidence currently exists attesting to higher than normal levels of psychopathology and adaptational difficulty in children with all types of chronic illness (see reviews by Bakwin & Bakwin,1972; Diller, 1972; Dinnage, 1970, 1972; 9 Pilling, 1973; and Rutter & Graham, 1970). In a major review of three extensive epidemiologic surveys of chronic childhood illness, Pless and Roghmann (1971) reported results support ing this thesis. The three studies reviewed by the above authors were: the National Survey of Child Health and Development (Douglas & Bloomfield, 1958); the Isle of Wight survey (Rutter, Tizard and Whitmore, 1970); and the Rochester Child Health Survey (Roghmant & Haggerty, 1970). The National Survey consisted of data collected from a representative sample of 5000 children born in England, Wales and Scotland during the first week of March in 1946. Of these 5000 children, 528 were described as having symptoms of chronic illness (Pless & Douglas, 1971). The Isle of Wight study consisted of a survey of the total population of 9 to 11 year old children on the Isle of Wight, England. The 3,271 children were

PAGE 17

10 first screened to determine the incidence of chronic physical illness. Those determined to have a chronic physical condi tion were studied to assess emotional concomitants by psychiatric interviews, psychological tests and interviews with parents. In the Rochester study a 1% probability sample of all children under 18 years of age living in Monroe County, New York, was selected. Of the 1,756 children screened, 206 were found to have chronic physical problems. In this study a matched control group was also selected. In reviewing the results of these studies, Pless et al. (1971) reported that social functioning is poorer in chronically ill children than in the normal controls. Although not signi ficant, the results from the National Survey and the Rochester study revealed that the chronically ill children are more frequently truant, more often troublesome in school, and more often socially isolated. In the Isle of Wight study significantly more chronically ill children than normal children were found to have psychiat ric disturbances (based on examinations by a psychiatrist). Seventeen percent of the sick children were found to be psychiatrically disturbed (compared to 7% of the healthy controls). There was no difference between the two groups in neurotic or antisocial behavior patterns. The National Survey looked at childhood adjustment in terms of behavioral symptom questionnaires which were completed by parents, teachers and the children themselves. Twenty-five percent of the sick children were found to have

PAGE 18

two or more behavioral symptoms, compared to 17% of the normal controls. Sick children were also rated by teachers as more nervous and aggressive (39% vs. 31%). Based on the results of these three independently conducted studies Pless et al. concluded that maladjustment is more frequent in the chronically ill children and that a high proportion of the social and psychological disturbances that were found must be attributed to the physical disorder. Unlike the surveys reported above, most studies in this area have investigated specific chronic illnesses and their psychological concomitants. The tendency to rely primarily on descriptive data obtained from relatively small samples has caused the vast majority of studies to be of questionable validity. Still, the cummulative evidence seems to support the above thesis. 11 In an early study of adjustment in diabetes, Sterky (1963) compared data obtained from 145 diabetic children with that of a matched control group of healthy children. Althou~h the total frequency of "mentally disturbed" children was the same in the two groups, Sterky reports that the diabetics were more likely to show severe symptomotology; i.e., emotional lability, aggression and difficulty with companions. One should be cautious in interpreting these results, however, because the method used for determining mental stability was subjective in nature, based primarily on interviews with mothers. A more sophisticated study of adjustment in diabetes was undertaken by Swift, Seidman and Stein (1967). Their

PAGE 19

research compared 50 diabetic juveniles with 50 normal control subjects matched according to age, SES, and race. Based on psychiatric interviews, information from parents, 12 and a series of psychological tests, the two groups were compared on a number of psychological dimensions. The authors reported that the diabetic youths showed significantly more pathology on psychiatric classification, self percept, manifest and latent anxiety, sexual identity, constriction and hostility. Individually scored test data revealed greater pathological body image, latent anxiety and depen dence. In addition, parental interviews revealed that the diabetics were more socially maladapted at home and with peers. Similar findings have been reported in studies of children with congenital heart disease (Green & Levitt, 1962; Linde, Rosof & Dunn, 1970; and Thomas, Milman & Rodriquez-Torres, 1970), chronic renal disease (Korsch & Barnett, 1961; Korsch, Negrette, Gardner, Weinstock, Mercer, Grushkin & Fine, 1973), rheumatoid arthritis (Cleveland, Reitman & Brewer, 1965; Grokoest, Snyder & Schleger, 1962; and McAnarney, Pless, Satterwhite & Friedman, 1974) and cystic fibrosis. Research pertaining to cystic fibrosis will be presented in detail in a later section. Chronic childhood asthma has undoubtedly received more attention from a psychological perspective than any other childhood illness. The reason for this is that asthma's aetiology is uncertain, and emotional factors

PAGE 20

13 have been found to have a significant impact on the course and stability of the disease (Purcell, 1975; Purcell & Weiss, 1970). A number of early studies have proposed that the asthma tic patient displays a distinctive personality pattern and that this pattern predisposes him/her to developing asthmatic symptomatology (French, 1936; Riess & Decillis, 1940; Schatia, 1941). The problem with this early research is that it was based solely on clinical impression and therefore lacked objective bases for comparison and replication. In a systematic exploration of the above hypothesis, Neuhaus (1958) analyzed'1:lata obtained from 169 subjects, including children with asthma, children with cardiac disease, normal siblings and a normal control group. The normal con trols were selected to match the sick children according to age, I.Q., SES, religion and number of siblings. Instruments used to assess emotional functioning included the Rorschach, Brown Personality Inventory, and Despent Fables. The results of this study indicated that the asthmatics were significantly more maladjusted than the normal controls but not signifi cantly different from children with heart disease. These findings tended to refute the hypothesis of a specific asthmatic personality profile while supporting the notion that asthma is similar to other chronic illnesses in its general adverse effect on psychological adjustment. A similar conclusion was reached in a more recent study by Graham, Rutter, Yule & Pless (1967). In reviewing data

PAGE 21

~ v ~ ~ --------14 obtained in the Isle of Wight survey (previously discussed) these investigators reported that the incidence of psychiatric illness was greater in the asthmatic children than in the normal controls (10.5% vs 6.3%), but was not significantly different from that for children with other chronic disorders. Tavormina et al. (1976) criticized many of the studies reported above for their methodological crudeness and biased assumptions. They contended that most research had approach ed the subject assuming that there was a direct cause/effect relationship between psychopathology and chronic illness. Furthermore, the evidence cited to validate this assumption was usually based on subjective evaluations, clinical impressions and abbreviated projective techniques. In an attempt to provide a more rigorous test of this assumption Tavormina et al. gave a battery of psychological instruments to 144 children with chronic disease (CF, diabeties, asthma and hearing impairment). The battery of psychological tests were designed to assess various aspects of emotional stability. Average scores for the children with each type of illness were analyzed relative to each other and to stan dardized norms. Although there were minor differences in level of pathology between the different types of illnesses, the authors reported that overall, the chronically ill children were not significantly more psychopathological than normal children. Contrary to prevailing attitudes, these results suggested that the chronically ill were more healthy than deviant. The authors did note, however, that

PAGE 22

15 sick children demonstrated a wider range of degree of psycho pathology and proposed that future research focus on those factors that tend to predispose the sick child to psycho logical deviance. Following the lead of Tavormina et al. (1976), Bedell, Giordani, Amour, Tavormina and Boll (1977) identified "life stress" as one possible predisposing factor. Citing work by Holmes and Rahe (1967), and Coddington (1972), which indicated that psychological functioning was related to recent experiences requiring life change (life stress), Bedell et al. hypothesized that psychopathology in the chronically ill might be more directly related to experienced life stress than to the illness itself. Using 45 chronically ill children attending a three-week residential summer camp as subjcts, the authors administered to each two measures of emotional stability: the Piers Harris Self-Concept Scale and the State Trait Anxiety Inventory, as well as the child's version of the Social Readjustment Rating Scale (SRRS). The last scale was a self-report questionnaire designed to assess the child's recent experience of events that tend to cause stress and social readjustment. After dividing the subjects into either a high or a low stress group based on scores on the SRRS, the authors compared the groups on the two measures of emotional stability. Although there were no differences in state and trait anxiety between the two groups, there was a significant difference in self concept. Subjects in the high stress group had

PAGE 23

16 significantly lower self-esteem scores than those in the low stress group. Bedell et al. concluded that "Highly stressed children perceive themselves as more poorly behaved, less physically attractive, less able at school work, less popular, and less satisfied with themselves in general" (p. 240). When considered in conjunction with the fact that the chron ically ill children in general experienced higher levels of life stress, these findings supported the author's hypothesis that the higher levels of psychopathology frequently reported in the chronically ill were more directly related to resultant life stress factors than to the illness. Pless, Roghmann and Haggerty (1972) identified "family functioning" as another variable influencing psychological adjustment in the chronically ill child. These authors re-examined data obtained in the Rochester Child Health Survey (Roghmann et al., 1970) previously reviewed. A mental health adjustment score was obtained for each of the 206 chronically ill children as well as for 110 healthychil dren selected as controls and matched for age, sex, race and SES. The mental health adjustment score was calculated using data from three measures: parental descriptions of child's behavior, child's self report, and teacher's description of child's behavior. In addition, the child's family environment was rated according to a family function ing index designed to reflect qualities like marital satis faction, communication potential and family happiness. The results of the study indicated that psychological

PAGE 24

17 maladjustment was 10 15% greater among the chronically ill than among the healthy controls. Furthermore the sick chil dren at highest risk for developing psychopathology were those from poorly functioning families. In other words, in terms of emotional stability, family functioning was an im portant moderating variable. It was not a sufficient variable determining psychological functioning, however, for the authors reported that even when controlling for this variable, the chronically ill as a group were still at higher risk than were normals. Apparently, family functioning is one of many potentially stress-produeing variables that interact to affect the child's stability. Family Adjustment to Chronic Childhood Illness Although the physical effects of chronic disease are confined to the affected child, the psychological/emotional effects of the illness spread over the entire system within which the child lives. There is little empirical research available documenting psychosocial effects of chronic child hood illness of parental and family functioning. There are, however, numerous descriptive accounts of associated stress factors affecting the process of family adaptation (see reviews by: Magrab, 1978; and Travis, 1976). Most agree that the presence of chronic physical illness places a heavy burden on the family and often results in deficits in family functioning.

PAGE 25

18 Emotional stress in parents of the chronically ill child often begins prior to diagnosis. In studying the effects of cystic fibrosis on family life, McCo!lum and Gibson (1970) reported that most parents became frustrated and anxious some time before diagnosis as they began to realize that their child was frequehtly sick and not making the gains that were expected. When the diagnosis of serious chronic illness is finally made, parental anxiety and frustration reach a peak as their worst fears are actualized. From this point on, parents begin a long and difficult process of adaptation. The adapta tion sequence begins with shock, anger and guilt and eventually (hopefully) reaches a stage of resignation and acceptance (Mattsson, 1972). Based on interview data from 35 hemophiliac boys and their parents, Mattsson and Gross (1966) emphasized the significance of the parental feeling of guilt. They stated that guilt is an inevitable reaction of parents on first hearing that their child has a chronic illness. Parents tend to feel that, somehow, they should have been able to prevent the disease. This guilt reaction is particularly strong for diseases that are genetic in nature (hemophilia, CF, etc.), because the parents soon learn that, in effect, they were the direct cause of the child's suffering. Mattsson et al. suggested that it is essential for parents, especially mothers, to resolve their guilt feelings if their children are to develop normally.

PAGE 26

19 Family adaptation is a difficult concept to operational ize. Some researchers have approached the topic, in relation to chronic childhood illness, by foctlsing on family structure, e.g., incidence of parental divorce or separation. Literature in this area is equivocal. A number of researchers reported that families of sick children were being torn apart by the stress and that the incidence of divorce and separation was particularly high in this population (Sultz, Schlesinger & Mosher, 1972; Binger, 1969). It is unfortunate that these authors based their information on very small samples and didnot carefully document their findings or make adequate comparisons to control groups. In a more recent investigation Lansky, Cairns, Hassanein, Wehr, & Lowman (1978) studied the incidence of marital dis cord and divorce in families of children with cancer. They first collected marital status information from the parents of 191 children who were being treated for cancer over a seven year period. They then compared the rate of divorce in this group to the average rate for the population at large, for couples with children residing in Kansas and Missouri (the two states in which all the cancer patients lived). Contrary to expectation, Lansky found that the person year divorce rate of 1.19% for the parents of chil dren with cancer was actually less than the 2.03% rate for the general population. In order to explore the incidence of marital stress, Lansky et al. computed the Arnold Sign Indicator (ASI), based on the comparison of MMPI profiles

PAGE 27

for husband and wife of 38 intact couples whose children were being treated for cancer. For purposes of comparison, marital stress was also calculated for a group of 23 couples who had hemophiliac children and for a normal standardized group. The author reported significantly more marital stress in the cancer group than in the hemophiliac group, 20 and significantly more stress in both the clinical groups than in the normal control group. It was suggested that the difference in marital stress between the cancer and hemo philiac groups was probably due to the nature of the disease; cancer is more irnmediat~Jy serious and life-threatening. Lansky et al. concluded that chronic illness does exacerbate marital conflict and discord but does not lead to higher rates of separation and divorce. Although this finding conflicts with previous reports, this study was methodologi cally more rigorous than earlier studies. Similar findings were reported by Begleiter, Burry and Harris (1976) in studying divorce in parents of children with CF. These researchers interviewed 40 parents of CF children and compared the rate of divorce in these families to the national average (based on the 1970 United States Census). They found that the divorce rate of 17.2% in CF families did not differ significantly from the national average of 14%. Furthermore, those parents of CF children who were divorced reported that they did not feel that the sick child was a contributing factor in the dissolution of

PAGE 28

their marriage. Twenty-two of these parents actually felt that caring for the sick child had brought them closer together. Family adaptation to chronic childhood illness has also been explored via the concept of "family functioning" (Pless & Satterwhite, 1973). "Functioning" is a dynamic concept relating to the quality and process of family life rather than its mere structure. Given the complex nature of family dynamics, it is little wonder that the majority of studies reporting family effects have been descriptive in nature and based primarily on subjective interviews and clinical experience (Pless et al., 1975). In a recent article, Stein and Riessman (1980) reviewed previous attempts to measure the "impact-on-family" of chronic childhood illness. They concluded that, "To date, no comprehensive measure exists which can quantify the impact of childhood illness 21 on a family and which delineates the many facets of this complex domain. None of the measures meet rigorous scientif ic criteria, i.e., demonstrable reliability and validity" (p. 465). Although measures have been slow to develop,there is a growing body of theoretical literature pertaining to family functioning and chronic childhood illness. It has been noted that the presence of a chronically sick child increases the developmental burden on the family (Travis, 1976; Magrab, 1978). There is a natural tendency for parents to overprotect the sick child, often at the expense of his/her

PAGE 29

developing a sense of mastery and autonomy. In a recent article on primary care setting for the chronically ill, Harding, Heller and Kesler (1979) emphasized this challenge to the family and stated: The developmental tasks of each child are mirrored by developmental tasks of the family as a whole. A child does not easily gain autonomy without the family's encouragement or allowing him to do so. All families must strike a balance between protection, nurturance, and dependency on the one hand; and auton omy, independence, and self sufficiency on the other. An inability to strike this balance can lead to serious problems which affect not only the emotional well-being and harmony of the family members, but their heal th as well. (p. 315) They went on to say that ~ "The chronically ill child is often deprived of many opportunities and experiences because of the decreased expectations ,of well-meaning adults" (p. 314). The literature on childhood asthma is replete with references made to maternal overprotection and a tendency 22 for overly close mother-child bonds (Sandler, 1977; Meijer, 1979). Others have reported dramatic improvement in asthmat ic symptomatology following structural family therapy in which healthy distance was created between mothers and affected children (Liebman, Minuchin & Baker, 1974). It is this writer's feeling that a "systems" perspective (Haley, 1971, 1972) provides a much needed theoretical frame work for the study of family adaptation to chronic illness. Systems theory views the family as a multidimensional com plex which has an identity of its own, but which is also dynamic in nature, constantly being submitted to stimuli which provoke changes in its structure (Bowen, 1971).

PAGE 30

According to systems thinking, healthy family function ing requires a degree of flexibility. Flexibility enables the family to confront a crisis or stressful event by chang ing individual roles as well as transactions between various subsystems in order to reestablish a state of equilibrium or balance (Minuchin & Barcai, 1969). If the alliances between family members are too "rigid, 11 transactions are prevented and conflicts become insoluble. One attempt to apply systems theory to the study of family dynamics in response to chronic disease was made by Paquary-Weinstock, Appelhoom-Fondu and Dopchie (1979). 23 These researchers studied 25 families of hemophiliac chil dren in Belgium. Family dynamics were assessed through interviews with parents in which information was obtained concerning type of communication, the existence of alliances, and type of functioning, both before and after diagnosis of the disease. The authors reported that families which had a strong identity and functioned according to an open communication system prior to the disease had the healthiest post-diagnosis adaptation. Although initially reacting with shock and depression, they were eventually able to consolidate family bonds and restructure the family cell. On the other extreme, families whose identity and style of communication were dysfunctional to begin with tended to react to the presence of the illness by dislocation, and often the appearance of psychiatric pathology in one of the parents.

PAGE 31

There are many flaws in the above study. The authors based their conclusions on a very small sample and on sub jectively interpreted data. Furthermore, they did not define their terms or describe methods used to analyze the data. The study does show, however, that there is potential for meaningful research in this area. What is needed is a conceptual model for studying family systems which proposes testable hypotheses and is open to empirical validation. Olson, Sprenkle and Russell (1979) present such a con ceptual system in the "Circumplex Model" of marital and 24 family systems. After reviewing the literature in the entire field of family behavior, these writers concluded that two aspects, "cohesion" and "adaptability," appear to be the most salient dimensions used to describe family functioning. They define "cohesion" as: "the emotional bonding members have with one another and the degree of individual autonomy a person experiences in the family" (p. 5). "Adaptability" is defined as "the ability of a marital or family system to change its power structure, role relationships, and relationship rules in response to situational and develop mental stress" (p. 12). In the Circumplex Model, Olson et al. have combined these two independent dimensions in such a way that families can be classified according to where they fall on both. By dividing each dimension into four levels: very low, low to moderate, moderate to high, and very high, a 4 X 4 matrix is formed indicating 16 types of family functioning.

PAGE 32

According to the Circumplex Model (which is based on family systems theory), balanced levels on both cohesion 25 and adaptability make for the healthiest family functioning. Empirical studies validating this model have been reported by Olson, Russell and Sprenkle (1980b). This model has also been found to be useful in monitoring change in family functioning in response to stress (Olson & Mccubbin, 1980a). It is this writer's feeling that the Circumplex Model of family systems could provide an excellent conceptual framework for examining the relationship between family func tioning and chronic childhood illness. The research presented in later sections makes use of this model. Psychosocial Aspects of Cystic Fibrosis (CF) Cystic fibrosis is the most common autosomal-recessive disease found in the white population, and it is thought to affect one in every 1600 to 2500 live births (Breslow, McPherson & Epstein, 1981). It is a disease which causes a chronic pulmonary condition, pancreatic insufficiencies and elevated levels of sodium and chloride in sweat. It is the pulmonary aspect of the disease which is the most prob lematic and which accounts for almost all the mortality in CF patients (Doershuk, Wood & Boat, 1976). CF is a life threatening illness for which there is presently no cure. A strict regimen of antibiotic treatment and rigorous physical therapy has extended the average age of survival to 19 years (Breslow et al., 1981), but few live into their thirties.

PAGE 33

26 Cystic fibrosis is a particularly stressful chronic ill ness both for the family and the afflicted child. As Denning, Gluckson and Mohr (1976) point out, factors contributing to the stressful nature of the illness include: it is a chronic disease for which there is no cure, it is a genetically transmitted disease for which there is no carrier test yet available, it is an expensive disease requiring thousands of dollars worth of medication each year, and it is a disease in which no one can accurately assess prognosis and life span to lend support to the parents, patient and families who are expected to faithfully carry out the prescribed treatment and other details of a complicated medical program day after day throughout life. (p. 127) One of the first sy.stematic attempts to investigate the psychosocial effects of CF on family functioning was a des criptive study by Turk (1964). Turk administered a ques tionnaire to the parents of 25 children with CF ranging in age from 3 months to 23 years. He reported that all of the families studied experienced stress relating to the child's disease. Of primary concern was financial strain. The CF child needs continual medical attention as well as special foods, supplements and home treatment equipment. In most cases the family was left to bear the financial burden them selves, which often meant that the father had to work longer hours or take a second job. There was also considerable social stress within these families. Parents reported that they had less time to devote to their spouses, their leisure time was drastically reduced, their sexual relations were strained, and constructive communication between family mem bers was inhibited. Furthermore, many parents reported

PAGE 34

I trouble deciding whether or not to have another child, and 60% said they could not discuss the diagnosis with the affected child. 27 In another study with CF patients, Lawler, Nakielny and Wright (1966) conducted psychiatric interviews with 11 mothers of affected children. Although again results were presented only in a descriptive fashion, they did indicate considerable intrapsychic conflict in most parents. Eight of the 11 mothers were judged to be clinically depressed. Furthermore, many of the mothers reported marriage and family conflicts. Six of the 11 said that 'they had recently considered separation from their husbands and three reported having abstained from sexual relations with their husbands. Although these results suggest negative effects of the presence of a chronically ill child on parental stability and family fucntioning, one must be cautious'about making such interpretations from descriptive reports of such a small group of subjects. Lawler et al. make no reference to any control group; therefore, there is no basis for comparing families of healthy and those of sick children. In a study previously cited (Begleiter et al., 1976), it was noted that the rate of divorce or separation among CF parents did not differ from the national average. The study did find, however, that the rate was higher for CF parents than for parents of children with most other chronic diseases. The authors suggested that this finding is probably due to the specific nature of the disease, especially the fact that it

PAGE 35

I is genetically transmitted, with a high probability that subsequent children will either have, or be carriers of, 28 the disease. Parents are naturally anxious about the pros pect of further children and usually decide to limit their families. Begleiter et al. proposed that it is this realiza tion, and not the reality of having to care for the sick child that causes a higher rate of divorce in parents of CF children than in parents of children with other diseases. In a study of 20 children with CF, Tropauer, Franz and Dilgard (1970) reported that these children had a reasonably good understanding of their disease and were tolerant of treatment but they were upset by dietary deprivation, interruptions of play, and physical limitations. Results of projective psychological tests indicated that these chil dren tended to feel inadequate, insecure and anxious. Simi+ar results were reported by Cytryn, Moore and Robinson (1973). In their study of 29 children with CF between the ages of 8 months and 9 years, 42% were judged to have significant emotional disturbance. In a major review of the literature pertaining to psychosocial aspects of CF, Gayton and Friedman (1973) contended that, ''The studies reviewed are consistent in their conclusion that the occurence of CF results in psychosocial consequences for patient, parents and family" (p. 858). The presence of CF has been found to precipitate financial concerns, communi cation difficulties, emotional instability, marital problems and overall family stress. Gayton et al. are cautious in their interpretation of these results, however, and they

PAGE 36

29 note a number of problems with this research. Again, most of these studies have used ver y few subjects and have been primarily descriptive in nature. Although this type of research can give a clinically useful perspective of the problem, without adequate controls and more sophisticated instrumentation it is difficult to make meaningful statements about the specific effects of the chronic illness. In a more recent study, Gayton, Friedman, Tavormina and Tucker (1977) attempted to verify results of previous research while applying a scientifically more rigorous methodology. As their sample they selected 45 families of children with CF from a total population of 75 being seen at the C y stic Fibrosis Clinic at the University of Rochester Medical Center. The children were between the ages of 5 and 18 with an average age of 11. Parents of the children com pleted two questionnaires, the Family-Concept Q-Sort (FCQS) and the MMPI. The CF children and selected normal siblings completed three questionnaires; the Piers-Harris Self Concept Scale, the Missouri Children Picture Series, and the Holtzmann Inkblot Test. In exploring psychological adjustment of the CF children, results of their responses to the above instruments were compared to those of their healthy siblings as well as appro priate normative data (when available). The authors reported that, except for one of the 22 variables calculated for the Holtzmann Inkblot Test, no significant differences were found between the groups. This obviously conflicts with previous findings.

PAGE 37

In exploring parental adjustment, hlMPI profiles of the CF parents were compared to those of parents of noncystic children (drawn from the population of parents used in a previous study). The authors reported that fathers of CF children showed significantly higher scores on 4 MMPI scales than the comparison group fathers. Mothers of CF children had significantly higher scores on 2 of the subscales. This result is consistent with previous research which has demonstrated negative effects on the emotional stability of parents caring for the CF child. It is somewhat surprising, however, that the fathers appeared to be affected more than the mothers. In terms of parental perception of family functioning, responses of parents of CF children were compared to those of parents of noncystic children (drawn from a population of parents used in another previous study). The authors reported that, as expected, having a child with CF tended to decrease family satisfaction and family adjustment. The overall results of this study lend only partial support to the findings of previous research. In discussing these findings, Gayton and Friedman (1973) stated: The literature dealing with the psychosocial aspects of chronic illness often implies that the presence of a chronically ill child in a family serves as a psychological stressor, invariably resulting in damaging psychological consequences for the child and his family. Although he would not disagree with the conceptualization of serious chronic illness as a stressor, this does not preclude the utilization of coping devices which allows the chronically ill child and his family to develop healthy adaptations. (p. 893) 30

PAGE 38

Again, the challenge for future research lies in the identi fication of those specific factors that are important in facilitating healthy coping and adaptation. Parental Death Attitude and the Chronically Ill Child Death is obviously a complex and troublesome issue in 31 the consideration of psychological adjustment in terminally ill children and their families. There is currently much disagreement among professional as to the degree to which the sick child is actually concerned about death, and how directly the child needs to confront the topic. Whereas some contend that death is "the central emotional problem of the seriously ill child" (Vernick & Karon, 1965, p. 396), others do not agree. Many experts in the field of treatment for the chronically ill child contend that death is not an important issue for the child (Evans & Edin, 1968). Natterson and Knudson (1960) reported that fatally ill children do not express, or even experience, anxiety about death until they are at least 10 years old. Relying on staff observations, Morrissey (1963) reported that less than 30% of all fatally ill children ever exhibit any anxiety about death. Based on these findings it has been suggested that discussing death with the sick child only serves to heighten his/her anxiety by introducing a subject that the child cannot possibly comprehend. The

PAGE 39

implication is that the child's suffering will be minimized if he/she is shielded from the reality of the disease. 32 Others have sharply criticized this perspective. Using data based on interviews with children (between 9 and 20 years of age) hospitalized for the treatment of acute leukemia, Vernick and Karon (1965) made the opposite contention. These researchers reported that children (even very young children) who were being kept in ignorance about their disease knew much more about the seriousness of their condition than was immediately obvious. They seemed to be able to know when they were being lied to and could easily detect anxiety in their parents. It was also observed that these children really wanted to talk about their disease, but they felt to do so would make their parents upset and angry. Vernick et al. reported that as a result of this frustration, "protected" children would often withdraw, feel isolated, and tend to exhibit more behavior problems than children who were aware of their diagnosis and prognosis. Based on a very success ful treatment program initiated by these authors, they concluded that: the most helpful way for adults to meet the emotional needs of the seriously ill child was to provide an atmosphere in which such children feel completely free to express their concerns: an atmosphere which always provided an honest answer to any question. (p. 397) Waechter (1971) found empirical support for some of the contentions of Vernick and Karon. Waechter studied 64 children between the ages of 6 and 10 in three groups: terminally ill (CF, leukemia, etc.) children, children with chronic non

PAGE 40

33 life-threatening disease, and children with just brief ill nesses. All of these children were studied while in the hospital using the following measures: the General Anxiety Scale, TAT, and parental interviews. Consistent with earlier reports, fear of death was rarely spontaneously discussed in any of the children; however, the terminally ill children had general anxiety scores twice as high as any other group. Furthermore, in responses to the TAT cards, these children discussed significantly more themes of loneliness, s e paration and death. This finding is particularly significant in light of the fact that the medical treatment of the terminal group was not significantly different from that of the chronic non-terminal group, and that 14 of the 16 terminal children were supposedly unaware of their diagnosis. Spinetta and Maloney (1975) pr e sent further evidence supporting the thesis that death is of conc e rn for terminally ill children. In their study, projective stories told by 27 fatally ill children between the ages of 6 and 10 were compared with stories told by a matched group of children with non-terminal chronic illness. Results show that themes of threat to body function and body integrity were much more frequent in the stori e s of the fatally ill chil dren. General anxiety also appeared manifest at a higher level in the stories of the terminally ill. An interesting finding of the Waechter study (previously discussed) was a significant negative correlation between preoccupation with themes of separation and death, and the

PAGE 41

degree to which the child was allowed to discuss his/her disease. The implication of these findings is that even 34 when shielded from the full nature of their disease, terminally ill children are aware of the seriousness of their condition, are concerned about dying, and need to be allowed to express this concern. Apparently, the reason many adults feel they must not openly discuss death with dying children is because of their own inability to deal with the subject (Vernick, 1973). It seems reasonable to suspect that parental attitudes toward death will affect the emotional climate surrounding the child and ultimately the child's adjustment. The relationship between psychosocial adjustment in chronically ill children and parental orientation toward death has not previously been explored. In reviewing much of the literature on psychological and sociological aspects of dying, Simpson (1979) noted that there has been a paucity of research concerning death atti tudes in adults. This seems to be due to the fact that death is a very sensitive and complex issue. As Simpson pointed out, attitudes toward death appear to be quite variable, are difficult to define, and are even more difficult to reliably measure. Research examining death attitudes in adults has been concerned primarily with emotional attitudes--usually fear or anxiety (Kastenbaum & Costa, 1977). A number of test instruments have been developed to measure anxiety associated

PAGE 42

with the idea of death (Collett & Lester, 1969; Dickstein, 1972; Pandley, 1975; and Templer, 1971). Krieger, Epting and Leitner (1974) proposed "death threat" as another way 35 to conceptualize one's attitude toward death. Although somewhat related to anxiety, "threat" implies a more cogni tive orientation. "Death threat" as measured by the Threat Index (TI) (Krieger et al., 1974) has been found to correlate only weakly with measures of "death anxiety" (Rigdon, Epting, Neimeyer & Krieger, 1979) which implies that they are tapping different dimensions of a more global concept of death orienta tion. "Death threat" and "death anxiety" appear to be two relevant ways to examine the relationship between parental death orientation and adjustment in chronically ill children. Hypotheses The study to be reported in this dissertation is an attempt to expand our understanding of the psychosocial concomitants of chronic childhood illness by examining some of the above issues utilizing a scientifically rigorous methodology. As has been mentioned, most previous research in this area has relied primarily on descriptive data obtain ed from small samples and has not used adequate controls. In this study, objective instruments (with established reliability and validity) are used to obtain data from three groups of children matched, as closely as possible, by age, sex, and race. Two of the groups consist of children with

PAGE 43

36 diagnosed chronic illness (CF and asthma). The third group is a control group composed of "normal" children, i.e., children with no previous history of chronic physical illness. By using these three groups, not only can differences be studied between chronically ill and normal children, but also between sick children whose illnesses have very differ ent implications in terms of their life-threatening status (CF is a terminal disease, asthma is not). The review of the literature indicates that the pr e sence of chronic childhood illness is often associated with a deficit in psychological functioning of child and family. There has been a gradual shift in emphasis over the years from considering the disease as causative in this process to viewing the problem as a failure in coping or adaptation. The reason for this shift is the growing realization that many children and families do not appear to suffer psychologi cal problems in the presen c e of chronic illness. There are two main questions addressed in this study. The first is general in nature and concerns the overall effect of the presence and severit y of chronic childhood illness on individual and family functionin g Three depen dent variables are e x pl o red in this regard. They are: child adjustment (defined in terms of behavior problems and self-concept); family functioning (defined in terms of the circumplex model for cohesiveness and adaptability); and parental orientation toward death (defined in terms of "death threat" and "death anxiety"). Thus, this study in

PAGE 44

some ways replicates previous research. It differs in the use of a family-functioning model which has not previously been applied to this population, and in the consideration of parental death orientation as a relevant dimension that may be affected. The second question is more specific in nature and potentially more important. This question will explore the effects of family functioning and parental orientation toward death on the adjustment of the child. In this way, the relative impact of the illness per se (vs. mediating variables: family functioning, etc.) can be analyzed in terms of effects on childhood adjustment. The following hypotheses are explored: 1. Family functioning will be affected by the presence and severity of chronic childhood illness in the following ways: a. Families of CF children will be more likely to be functioning at the extremes of the Circumplex Model than will families of healthy children; i.e., they are more likely to be either high or low in cohesion and adaptability. b. Families of asthmatic children will be more likely to be functioning at the extremes of the Circumplex Model than will families of healthy children. c. Families of CF children will be more likely to be functioning at the extremes of the Circumplex Model than will families of asthmatic children. 37

PAGE 45

2. Child adjustment will be affected by the presence and severity of chronic illness in the following ways: 38 a. Children with CF will have more behavior problems and lower self concepts than healthy children. b. Children with asthma will have more behavior problems and lower self concepts than healthy children. c. Children with CF will have more behavior problems and lower self concepts than asthmatic children. 3. Parental orientation toward death (death threat and death anxiety) will be affected by the presence and severity of chronic childhood illness in the following ways: a. Parents of CF children will be more anxious about and threatened by death than will parents of healthy children. b. Parents of asthmatic children will be more anxious about and threatened by death than will parents of healthy children. c. Parents of CF children will be more anxious about and threatened by death than will parents of asthmatic children. 4. R e gardless of illness, the tendency for families to be fuiictioning at the extremes of the Circumplex Model will be directly related to the number of behavior problems of the children and inversely related to children's self concept. 5. In CF families, parental death threat and death an x iety will be directly related to the number of behavior

PAGE 46

problems of the children and inversely related to chil dren's self concept. 39

PAGE 47

CHAPTER III METHODOLOGY Subjects Eighty-one children and their parents (usually mothers, preferably both mothers and fathers) participated as subjects in this study. The children were selected on the basis of their membership in one of three groups. Group 1 consisted of children with cystic ~ fibrosis (a chronic, life-threatening disease). Group 2 consisted of children with chronic asthma (usually a non-life-threatening disease). Group 3 consisted of "normal" children (children with no previous history of chronic illness). Samples were obtained in the following manner. Group 1: Thirty-one children with CF were selected from the population of children being treated on an outpatient basis in the CF clinic at Children's Hospital Medical Center in Boston. The following criteria were used in making this selection: 1. The children had to be between the ages of 7 and 12. Children younger than 7 were excluded because of the level of verbal ability required to complete the self concept inventory. Children older than 12 were not included because it was felt that the quality of adjustment demands changes as one enters 40

PAGE 48

adolescence; consequently, adolescents and children may differ in terms of healthy coping styles. 2. The children had to have been diagnosed as having CF for at least two years. The purpose of this criterion was to insure that the effects to be measured in individual and family adjustment would reflect more or less stable and long-term coping styles. Families of more recently diagnosed children might still be experiencing short-term effects due to the initial shock of diagnosis. 3. The children could not have been hospitalized within the preceding year due to exacerbation of the illness. The rationale for this criterion is the same as the above;i.e., to tap enduring adjustment styles rather than short-term reactions to crisis situations. 4. The children could not have any siblings who also have the disease. Families in which two or more chil dren had been diagnosed as having CF or in which a 41 child had previously died from the illness were excluded. It was felt that these families could be at a differ ent stage of adjustment than families with just one sick child. Of the 31 children in this group, 19 were boys and 12 were girls. Data wereobtained from all 31 mothers and from 20 fathers. Group 2: Twenty-seven children with chronic asthma were selected from the population of asthmatic children being treated on an outpatient basis in the Allergy Clinic at

PAGE 49

42 Children's Hospital Medical Center in Boston. These chil dren were selected so as to match the CF children as closely as possible by age, sex and race (all of the children in the study were white). In addition, they had to meet the follow ing criteria: 1. They had to have been diagnosed as having asthma for at least 2 years (same rationale as above). 2. They could not have been hospitalized within the preceding year due to exacerbations of the illness (same rationale as above). 3. They could not have any siblings who also have the disease (same rationale as above). 4. They had to be taking daily medication for mainte nance of the illness. Asthma is an illness with a wide range cf severity. In some it is hardl y notice able. The purpose of this criterion was to insure that all children selected were experiencing an illness severe enough to require daily attention. Of the 27 children in this group 16 were boys and 11 were girls. Data wereobtained from 26 of the mothers and 18 of the fathers. Group 3: Twenty-eight children with no previous history of chronic illness participated as control subjects in this study. These children were selected from a single elementary school in the Boston metropolitan area and matched, as closely as possible, to the CF children according to age, sex, and race. An additional criterion used in selecting

PAGE 50

this sample was that neither the subject nor any of his/her siblings could have had a chronic childhood illness. Of the 28 children in this group 16 were boys and 12 were girls. Data wereobtained from 27 of the mothers and 18 of the fathers. Instruments 43 Family Adaptability and Cohesion Evaluation Scales (FACES). This is a self-report instrument developed by Olson, Bell and Portner (1978) to assess family functioning along the two primary dimensions of "adaptability" and "cohesion." It is a clinical and research tool based on the Circumplex Model (Olson et al., 1979) previously discussed. The Cir cumplex Model combines adaptability and cohesion in an attempt to diagnose marital and family systems and to set goals for treatment. FACES is composed of 111 statements concerning various aspects of family dynamics and functioning. A family member independently completes the questionnaire by indicating on a scale from one to four the degree to which each state ment is felt to be true of his/her family. A "1" indicates that the statement is felt to be true of the family "none of the time," while a "4" means it is true of the family "all of the time." From this scale two primary scores are obtained, one for "cohesion" and one for "adaptability". Scores for

PAGE 51

cohesion can range between 162 and 378. Adaptability scores can range between 126 and 294. In addition, 16 subscales have been proposed, based on factor analysis. Split-half reliability is very low on these subscales, however, and the authors recommend that total scores rather than subscales be used in research. By dividing each of the two primary dimensions into 44 four levels (low, low-moderate, moderate-high, and very high) and then combining them in a single matrix, a 4 X 4 matrix is formed indicating 16 different types of marital and family systems. This matrix is shown in Figure 1 with names the authors use to describe each type of system. According to the Circumplex Model, the healthiest families are those located in the moderate ranges of both dimensions (the four types inside the rectangle) while the unhealthiest families are those at the extremes on both dimensions (those underlines in the four corners). In between these two groups are the eight types of families which are moderate on one dimension but extreme on the other. Because of the newness of this instrument, validity studies are still underway and have not yet been reported. The Circurnplex Model upon which FACES was developed does appear to have empirical validity in terms of differentiating families under stress and in setting treatment goals for family therapy (Olson et al., 1980a;and Olson et al., 1980b). The authors reported high internal consistency relia bility for the total scores for adaptability and cohesion

PAGE 52

A D A p T A B I L I T y Fi~ure 1. Sixteen Types of Marital and Family Systems Derived from the Circumplex Model C O H E S I O N DISENGAGED SEPARATED CONNECTED (Very Low) (Mod. Low) (Mod. High) CHAOTIC (Very High) Chaotically Chaotically Chaotically Disengaged Separated Connected FLEXIBLE (Mod. High) Flexibly Flexibly Flexibly Disengaged Separated f Connected STRUCTURED (Mod. Low) Structurally Structurally Structurally Disengaged Separated Connected RIGID (Very Low) Rigidly Rigidly Rigidly Disengaged Separated Connected ENMESHED (Very High) Chaotically Enmeshed Flexibly Enmeshed Structurally Enmeshed Rigidly Enmeshed

PAGE 53

(r = .75 and r = .83 respectively). Correlation with the Edmond's Conventionality Scale (a measure of social desira bility) was very low for adaptability (r = .03) but high for cohesion (r = .45). Death Anxiety Scale (DAS). This is an instrument
PAGE 54

Threat Index (TI). The TI is a cognitive measure of death orientation developed by Krieger et al. (1974) and based on the theoretical work of George Kelly (1955). Death threat differs from death anxiety or death fear in that it is concerned with the degree to which an individual's construct system is structured to anticipate death, rather than the affect associated with the topic. Krieger et al. (1974) defined death threat as: the reluctance of a person to subsume his present view of himself, the way he prefers to see himself, and the concept of death together as elements under the same poles of a sample of his constructs. It is assumed that the person who describes both himself and death with the same pole of a single construct dimension is organizing his world in such a way as to be able to see death as a personal reality. The person who places himself and death on opposite poles of a construct, however, would have to reorganize his S}S tem to construe self and death together. ( p. 301) The form of the index used in this study is the Tip40, which consists of two identical lists of 40 bipolar dimen47 sions. (See Appendix B for copy). On the first list subjects are to go through each of the 40 pairs of adjectives indicating the side with which they see themselves more closely associ ated (self elements). On the second list they choose which of the two adjectives in each pair is more closely associated with the thought of their own death (death elements). The death threat score is determined by the number of splits, i.e., the number of times the "self" element is rated toward one pole of a construct and the "death" element toward the other pole. Scores for the Tlp40 can range between O (not at all threatened) and 40 (very threatened).

PAGE 55

In a recent review of the research utilizing the TI, Rigdon et al. (1979) reported test-retest reliability for 48 the Tip40 between .87 and .90 for intervals from 4 to 9 weeks. The instrument was found to be internally consistent with a split-half coefficient of .96 (Krieger, Epting & Hays, 1979). It did not correlate significantly with the Marlowe Crowne Social Desirability Scale. Since this is the first instrument of its kind, concur rent validity could not be assessed by comparing it with other measures of death threat. As expected however, the TI has been found to correlate in a consistent manner with other measures of death orient~tion (Krieger, Epting & Leitner, 1974; Neimeyer, Dingemans & Epting, 1977; Neimeyer & Dingemans, 1980; and Epting, Rainey & Weiss, 1979). This suggests that the TI has validity as a measure of death orientation. The Tip40 correlated moderately (p = .23) with Templer's DAS (Krieger, 1977). Although this indicates some overlap between the two scales, it is weak enough to imply that the instru ments are measuring two different aspects of death concern. Behavior Problem Checklist (BPC) (Quay & Peterson, 1979). The BPC is an instrument which developed from the original work of Peterson (1961). It consists of a list of 55 problem behaviors that often occur in childhood and adolescence. The checklist measures four primary dimensions derived from factory analysis (i.e., conduct problems, personality prob lems, inadequacy-immaturity and social delinquency). It can be completed by parents, teachers, or anyone else familiar with the child's behavior.

PAGE 56

,-1 In completing the BPC, the evaluator is instructed to go through the list placing a check next to any of the items which is felt to represent a problem as far as the child is concerned. A score is obtained for the child by summing 49 the total number of items checked. Scores can range between 0 and 50; the higher the score, the more behavior problems exhibited by the child. When appropriate, individual scores can also be calculated for each of the four factors mentioned above. Quay (1977) reviewed the reliability and validity data for the instrument. The scale was reported to be internally consistent and to have test-retest reliability around .82. Interrater reliability tended to be moderately high between teachers and between parents, but surprisingly low between teachers and parents. O'Leary and Johnson's (1979) review suggested that the instrument has high concurrent and con struct validity. The scale reliably differentiates groups of children with known behavior disorders. Piers-Harris Children's Self Concept Scale (P-H) (Piers & Harris, 1969). This is an 80 item self-report questionnaire to assess the child's self-evaluation. It developed from the original work of Piers and Harris (1964). The items in the instrument are first person declarative statements of the type, "I am a happy person", to which the child indicates whether it is mostly true or mostly false by circling either "yes" or "no. Scores can range between O and 80; the higher the score the greater the child's self esteem.

PAGE 57

The P-H was originally designed for use with children between the ages of 8 and 17. Because of the reading level required to complete the questionnaire, standardization data has not been obtained for children below 8 years of 50 age. The authors did suggest, however, that when administer ed individually the scale might be used successfully below this level. Piers (1977) reported test-retest reliability of .76 for a group of 7 year old children over a two month interval. This compares quite favorably with that obtained for older children. Reports of internal consistency for the scale ranged from ~78 to .93 with test-retest reliability estimates around .72 (Piers and Harris, 1969). A correlation coefficient of about .65 was found with similar measures of self-esteem, and of .40 with teacher and peer ratings (Piers, 1977; Buros, 1972). Some have found the P-H to be a good indicator of emotional stability (Bedell et al., 1977). Piers and Harris (1969) reported correlations with measures of anxiety between -.54 and -.69. Procedure Group 1: The procedure for selecting the CF subjects and administering the questionnaires required a number of steps. The first step was to review, on a weekly basis, lists of CF patients with upcoming clinic appointments. Two weeks prior to their scheduled appointment a letter

PAGE 58

51 was sent to the parents of any and all CF patients who met the above-stated criteria. (A copy of this letter is included in Appendix C). In this letter, the purpose of the study was described and an intial orientation to the procedure was presented. The letter was thought to be necessary because of the sensitive nature of the questionnaire involved (especially those concerned with "orientation toward death"). A week following the mailing of these letters a tele phone call was made by the investigator to the parents. The purpose of this call was to answer any questions the parents had and to arrange a time to meet during their upcoming clinic appointment. On the day of their appointment, consenting parents and their children met with the investigator for approximately 20 minutes. During this meeting general information on the family was obtained from the parents while the children completed their questionnaire (the Piers Harris Children's Self Concept Scale). The children filled out these question naires in private and were told that no one except the investigator would have access to their answers. Any help needed in reading the statements or in defining certain words was given directly by the investigator. Parents were then given instructions for completing their questionnaires (FACES, the Behavior Problem Checklist, the DAS and the TI). Whenever possible these questionnaires were completed during the clinic visit. If this was not possible, the parents were allowed to take them home and

PAGE 59

were given a stamped, self-addressed envelope in which to mail them back. Fathers were seldom present during these visits. An identical set of questionnaires for the fathers were sent home with the mothers with the request that they be completed and returned by mail. It was emphasized that 52 completion of these questionnaires should be done independently by each parent. A total of 45 CF families were initially contacted by letter and subsequently by phone. Of these, only one family openly refused to participate. Twelve families who had initially agreed to participate were unable to be tested on account of scheduling problems and missing appointments. One family did not return their questionnaires. Completed data wereobtained from 31 mothers and their children and 20 fathers. Five of the mothers were single parents. Six fathers failed to return their questionnaires. Group 2: For the asthma group introductory letters were not sent to parents prior to their clinic visits. The reason for this was that computerized printouts of weekly clinic appointments were not available for this group. The procedure used to select these subjects and administer the questionnaires required the following steps: In the morning prior to Allergy Clinics, the investigator reviewed the charts of all the patients with appointments for that day. From this review a list was made of all the patients who had a primary diagnosis of "asthma" and who met the above stated criteria (7 to 12 years old, white, on daily medications, etc.).

PAGE 60

Upon entering the clinic and prior to meeting with their doctors, parents of these children were approached by the investigator, informed of the nature of the study 53 and invited to participate. Consenting parents and children then followed the same procedure for obtaining information and completing questionnaires as outlined for the CF group. Since the Allergy Clinic visits were usually of shorter duration than the CF visits, very few parents of asthmatic children completed their questionnaires while in the hospital. Of the 40 adults who were approached in this manner, onl y 2 refused to participate. The reason for these refusals had to do with the time factor involved. One family was excluded because the child was a foster child. One other was excluded because the parents could not speak English. Thirty-eight asthmatic families agreed to participate in this study. Although the child's questionnaire was completed for all 38 children, parental responses were never returned for 11 of these families (even after follow-up phone calls). Data wereobtained from a total of 26 mothers and 18 fathers. Five of these mothers and one of the fathers were single parents. Of the intact families that participated, 3 of the fathers failed to return the questionnaires. Group 3: A control group of "normal subjects outside the hospital setting was difficult to obtain. A nearby school system agreed to sponsor the project, but for reasons of confidentiality, could not provide names of appropriately matched students. Taking this problem into consideration,

PAGE 61

54 the following procedure was used to obtain and test subjects: One elementary school was chosen as the source of all con trols. A letter was then drafted, addressed simply to "Parents," describing the study and inviting families to volunteer (a copy of the letter can be found in Appendix D). Copies of these letters were distributed by the teachers to 20 randomly selected students in each grade, 2 through 6 (the grades most likely to have children between 7 and 12 years of age). These children were told to bring the letters home to their parents. The investigator had previously met with each of these teachers to describe the randomization -process. This process consisted of selecting every other student from an alphabetical list of all the white students in the classroom. If more students were needed after going once through the list, teachers were to start at the begin ning and select those not previously chosen. Enclosed with each letter was a self-addressed post card which was to be returned to the investigator if parents agreed to participate. The post cards included a space for "name", telephone number;" and for indicating the "age" and "sex" of the targeted child. When the post cards were received, the investigator called the parents directly and arranged a time to visit them at their homes in order to administer the questionnaires. The procedure for administer ing the questionnaires was identical to that used with the two other groups.

PAGE 62

From the 100 letters sent home, responseswere received from 31 families. Because of scheduling difficulties, 2 of these families could not be seen. The investigator was eventually able to administer the questionnaires to 29 of these families. One of these families did not return the questionnaires by the stated deadline and could therefore not be included in the analysis. In all, completed data w e reobtained for 27 mothers and 18 fathers. Four of the mothers and one of the fathers were single parents. Of the intact families 4 fathers failed to return the question naires. 55

PAGE 63

CHAPTER IV RESULTS Results will be presented separately for each of the 5 major hypotheses. Although data was collected from both parents whenever possible, the primary testing of hypotheses 1, 3, 4 and 5 (those relying on parental variables) was based solely on data obtained from the mothers. This was done for two reasons. First, it was assumed that it was usually the mother who was the primary caretaker of the child. Second, data were much easier to obtain from the mothers than from the fathers, as fathers more frequently reported not having the time or the inclination to complete the questionnaires. Although not reported in this section, separate analyses were done for each of the hypotheses using data obtained from the fathers. These results are presented in tabular form in Appendix E. Subject Match As previously discussed,an effort was made to match the children in each group according to race, age and sex. All of the 86 children studied were white. Table 1 presents comparison data for the three groups of children on average age, age range and "percent male." Also included in this table is the percentage of children in each group living in single parent families. As is apparent, the three 56

PAGE 64

groups of children selected as subjects for this study wer e similar on all of the above criteria. 57 Table 1. Comparison Data for the Groups of Children Selected as Subjects Average Group Age (yrs) CF 9.52 Asthma 9.18 Control 9.46 Average Range (yrs) 7 12 7 12 7 12 % Male 61 56 57 From Single % Parent Families 19 22 21 Relationship Between Chronic Childhood Illness and Family Functioning In order to test the hypothesis that families caring for children with chronic illness are more likely to be functioning at the extremes within the Circumplex Model of adaptability and cohesion (hypothesis 1), FACES scores were utilized in the following manner: First, grand means were calculated for adaptability and for cohesion based on the pooled data from all 84 mothers. These grand means were 172.77 and 259.08 respectively. Next, a deviation score was calculated for each of the mothers on each of the two dimensions. These scores were obtained by taking the absolute value of the difference between a mother's score on one dimension and the grand mean for that dimension. In this way, two deviation scores (one for adaptability and one for cohesion)

PAGE 65

were calculated for each family based on the mother's responses to FACES. The deviation score for adaptability (Dev ADP) can be represented in the following way: Dev ADP=/ ADP ADP/, where ADP is the family's adapta bility score and ADP is the grand mean for adaptability. Similarly the deviation score for cohesion (Dev COH) can be represented as Dev COH = / COH COH /. Because families with high scores are those functioning toward the extremes of the dimensions these deviation scores make it possible to test hypotheses la, 2b, and le. Table 2 shows the means of Dev ADP and Dev COH for the three groups of families. Differences between the group means were examined by three MANOVA procedures. Table 2. Mean Dev ADP and Dev COH Scores for Families of CF, Asthma, and Control Children Group CF Asthma Control N 31 26 27 (based on mother's response) Mean Dev ADP 13.46 9.45 8.70 Mean Dev COH 17.60 19.16 14.93 Results of MANOVA analyzing the difference between the deviation scores of the CF and control group are presented 58

PAGE 66

59 in Table 3. The overall F value of 1.27 was not significant, and therefore hypothesis la was not supported. Results of individual ANOVA's for each of the dimensions (Dev ADP and Dev COH) are also presented in Table 3; these, too, were not significant. Table 3. F Scores for MANOVA Examining the Differences between CF and Control Families on Dev ADP and Dev COH Source Dev ADP Dev COH Dev ADP/Dev COH (based on mother's response) df 1 1 2,55 F Value 2.32 0.67 1.27 p .134 .415 .289 In Table 4, the result of the MANOVA analysis of the difference between the asthmatic and control families are presented. The overall F value of .98 was not significant, nor were the individual ANOVA's for the two dimensions. Hypothesis lb is therefore not supported.

PAGE 67

Table 4. F Scores for MANOVA Examining Differences between Asthma and Control Families on Dev ADP and Dev COH (based on mother's response) Source Dev ADP Dev COH Dev ADP/Dev COH df 1 1 2,50 F Value 0.13 1.44 0.98 p .723 .235 .384 60 To test hypothesis le, a MANOVA was performed examining the difference in mean deviation scores between the CF and the asthma groups. These results are presented in Table 5. Here, too, the overall F and individual ANOVA's did not yield significant results. Table 5. F Scores for MANOVA Examining Differences between CF and Asthma Families on Dev ADP and Dev COH (based on mother's response) Source Dev ADP Dev COH Dev ADP/Dev COH df 1 1 2,54 F Value 1.63 0.16 1.01 p .208 .692 .372

PAGE 68

~----------Hypotheses la, lb, and le were all unsupported by the above results. Caring for a child with a chronic illness was not related to extreme family functioning within the Circumplex Model of cohesion and adaptability. Furthermore, there did not appear to be a difference identified by this model between families of terminally ill children and those of children with a non-fatal chronic illness. Relationship Between Chronic Childhood Illness and Child Adjustment Hypotheses 2a, 2b and 2c concern differences between the three groups of children on two dependent variables, self concept (P-H score) and number of behavior problems (BPC scores). Data were analyzed separately for each depen dent variable. Table 6 shows the means and standard deviations for P-H scores of the three groups of children. Hypotheses 2a and 2b suggest that the mean P-H scores for the CF and asthmatic children, respectively, will be less than the mean for the control children. Hypothesis 2c states that the mean CF score will be less than the mean asthma score. Results of the t-tests performed to test these hypotheses are presented in Table 7. None of these t-tests approached significance. Based on these results, it appears that the child's self concept is not significantly affected by the presence of a chronic illness. 61

PAGE 69

Table 6. Mean P-H Scores and Standard Deviations for CF, Asthma and Control Children Standard Group N Mean Deviation CF 31 63.87 10.22 Asthma 27 62.82 13.41 Control 28 64.54 7.08 Total 86 63.76 10.41 Table 7. Results of t-t-ests Examining Differences between Groups on Mean P-H Scores Group Comparisons CF vs Control Asthma vs Control CF vs Asthma Value -0.66 -1.72 1.06 t-Value -0.24 -0.61 0.38 p .809 .545 .704 62 Wheu using behavior problems as the dependent variable, however, the presence of a chronic illness did appear to have a significant negative effect on the child's adjustment. Table 8 presents the means and standard deviations of BPC scores for the three groups of children. T-tests were used to examine hypothesized differences between the groups. These results are shown in Table 9. As expected, CF children

PAGE 70

had significantly more reported behavior problems than did the control children (p (.05). The difference between the asthma and control children on behavior problems was also significant (p <.01) and in the predicted direction. Thus, both hypothesis 2a and 2b were supported when using the BPC as the dependent variable. Table 8. Mean BPC Scores and Standard Deviations for CF, Asthma and Control Children (based on mother's response) Standard Group N Mean Deviation CF 31 4.97 5.61 Asthma 26 5.88 5.27 Control 27 2.26 3.22 Total 84 4.38 5.03 Table 9. Results oft-tests Examining Differences between Groups on Mean BPC Scores (based on mother's response) Group Comparisons CF vs. Control Asthma vs Control CF vs Asthma *significant at p = .05 **significant at p = .01 Value 2.71 3.63 -0.92 t Value 2.12 2.72 -0.71 p .037* .004** .480 63

PAGE 71

64 Hypothesis 2c was not supported by these data. There was no significant difference between the mean BPC score for the CF children and that for the asthma children. The number of behavior problems was related more to the presence of a chronic illness than to differences in severity or life threatening status. Relationship Between Chronic Childhood Illness and Parental Attitudes Toward Death Hypothes e s 3a, 3b and 3c concern differences between the three groups on twomeasures of parental death orienta tion, death threat and death anxiety. It was expected that parents of both groups of chronically ill children would be experiencing greater death threat and death anxiety than parents of healthy children. It was also hypothesized that death threat and death anxiety would be greater in CF parents than in asthmatic parents, given the life-threatening nature of the former illness. Again, the two dependent variables were analyzed separately. Table 10 presents the means and standard deviations of DAS scores for the three groups of mothers. Results of t-tests performed on this data to assess group differences are presented in Table 11. None of the t-tests were signi ficant. Apparently, mothers' death anxiety (as measured by the DAS) was not affected by the presence or severity of chronic childhood illness.

PAGE 72

Table 10. Mean DAS Scores and Standard Deviations for Mothers of CF, Asthma, and Control Children Standard Group N Mean Deviation CF 31 7.81 2.94 Asthma 26 6.69 2.75 Control 26 7.96 2.62 Total 83 7.51 2.80 Table 11. Results of t~tests Examining Differences between Groups on Mean DAS Scores of Mothers Group Comparisons Value t Value p 65 CF vs Control Asthma vs Control CF vs Asthma -0.16 -1.27 1.11 -0.21 -1.64 1.51 .835 .104 .136 In examining the differences in death threat between the groups, some very interesting findings emerged. As is apparent from Table 12, which lists the means and standard deviations of TI scores for the three groups, the results were in the opposite direction from what was hypothesized. Mothers of CF children were least threatened by the thought of death; the mothers of the healthy children, on the other hand, were the most threatened of the three groups. T-tests used to make

PAGE 73

66 group comparisons (reported in Table 13) revealed that differ ences in TI scores between the CF group mothers and the control group mothers were significant at the .05 level. In other words, mothers of CF children were significantly less threaten ed by the thought of death than were mothers of healthy chil dren. T-values for the other comparisons were not significant. The implication of this finding will be discussed in the next section. Table 12. Mean TI Scores and Standard Deviations for Mothers of CF, Asthma, and Control Children Standard Group N Mean Deviation CF 31 15.52 9.07 Asthma 23 17.48 7.56 Control 27 20.26 9.52 Total 81 17.66 8.95 Table 13. Results oft-tests Examining Differences between Groups on Mean TI Scores of Mothers Group Comparisons Value t Value p CF vs Control -4.74 -2.04 .045* Asthma vs Control -2.78 -1.11 .271 CF vs Asthma -1.96 -0.81 .422 significant at p = .05

PAGE 74

Relationship Between Child Adjustment and Family Functioning Since there was no relationship found between family functioning and the presence of childhood illness (hypotheses 1) there was no need to statistically control for effects of the latter in examining hypothesis 4. To test the hypothesis that extreme family functioning (defined in terms of Dev ADP and Dev COH) is negatively related to child adjustment, two multiple regression analyses were performed with the pooled data (N=86). Both analyses used Dev ADP and Dev COH as the independent variables. P-H scores were used as the dependent 'variable in the first analysis; BPC scores in the second. Results of the multiple regression analysis examining the relationship between Dev ADP/Dev COH and the child's self-concept (P-H score) are presented in Table 14. An F value of 2.97 was obtained from this analysis; this value just missed significance at the .05 level (p = .059). Although not significant, this finding is in the predicted direction. When the reported number of behavior problems of the 67 child (BPC score) was used as the dependent variable, the multiple regression analysis (reported in Table 15) yielded highly significant results (p = .001). In other words, there was a significant relationship between extreme family function ing and the number of behavior problems reported in the chil dren. This significant result--along with the tendency reported above for self-concept--provides support for the Circumplex

PAGE 75

theory, which suggests that healthy families function in the moderate ranges on both cohesion and adaptability. Table 14. F Scores for Multiple Regression Analysis of the Relationship between Child's P-H Scores and Family Dev ADP and Dev COH (based on mother's response) for the Pooled Data Source Dev ADP Dev COH Dev ADP/Dev COH N 86 86 86 df 1 1 2,81 F Value 2.26 7.61 2.94 p .137 .061 .059 Table 15. F Scores for Multiple Regression Analysis of the Relationship between Child's BPC scores (mother's rating) and Family Dev ADP and Dev COH (based on mother's response) for the Pooled Data Source N df F Value Dev ADP 84 1 7.09 Dev COH 84 1 8.43 Dev ADP/Dev COH 84 2,81 7.76 ** significant at p = .01 Relationship Between Parental Orientation Toward Death and Adjustment in the CF Child p .009** .005** .001** Within the CF group, Pearson r correlation coefficients were calculated between each of the measures of child adjust ment (P-H and BPC) and the mother's TI and DAS scores. The 68

PAGE 76

' correlations between the child's P-H score and the mother's TI and DAS scores are presented along with p values in 69 Table 16. Again, death anxiety did not appear to be an important variable in the analysis. Death threat, on th e other hand, did have a low level although statistically insignificant relationship to child's self-conaept (r = .25). The relationship, however, was in the opposite direction than was expected. Mothers who were most threatened by the thought of death tended to have children with higher self-concepts. Table 16. Pearson r Correlation between CF Children's P-H Scores and Mothers' TI and DAS Scores Parental Attitude Death Threat (TI) Death Anxiety (DAS) N 31 31 r .25 -.04 p .091 .426 To see if this relationship was unique to the CF group, Pearson r correlation coefficients were calculated between P-H scores and the mother's TI scores for each of the other two groups, and then for the pooled data. Results of this analysis are presented in Table 17. As is apparent from this table, the unexpected relationship found in the CF group between children's self-concept and maternal death threat also held true for each of the other groups. Further more, when the datawerepooled, the correlation became

PAGE 77

significant (p = .04). In other words, for all families regardless of illness, high maternal death threat appeared to be associated with high self-esteem of the children. This finding is not easily explained and will be discussed in the next section. Table 17. Pearson r Correlations between Children's P-H Scores and Mothers' TI Scores for CF, Asthma and Control Groups and for Pooled Data Group N r p CF 31 '-.25 .091 Asthma 23 .20 .187 Control 27 .17 .201 Pooled 81 .20 .040* *significant at p = .05 70 Table 18 presents the Pearson r coefficients and p values for the CF group for correlations between BPC scores and maternal death threat and death anxiety. Again, death anxiety was not an important variable in the analysis. The correla tion of -.18 between TI scores and BPC scores was neither significant nor as close to significance as that between the TI and P-H scores. It is interesting, however, that this correlation was also in the opposite direction from what was expected. High maternal death threat was associated in a low level and statistically insignificant way with few reported behavior problems in the children.

PAGE 78

Table 18. Pearson r Correlations between CF children's BPC Scores (based on mother's rating) and Mother's TI and DAS Scores Parental Attitude N r p Death Threat (TI) Death Anxiety (DAS) 31 31 -.18 .11 .167 .280 71

PAGE 79

CHAPTER V DISCUSSION In general, the results of this study support the notion that chronically ill children and their families adjust fairly well despite the stresses imposed by the presence of the ill ness. Family functioning does not appear to be significantly different in families of chronically ill children and families of healthy children. Although there does seem to be a differ ence in psychological adjustment between sick and healthy children, this difference may be more related to mediating factors (like family functioning) than to the illness per se. This conclusion is consistent with recent research. Parental orientation toward death (at least as measured by the TI) does seem to be affected by the presence of a termi nally ill child, but its importance as a mediating variable affecting the child's adjustment is unclear. Before discussing the implications of the major find ings of this study, a word should be said about limitations due to the sampling procedures. The procedure used to obtain subjects from the two clinic populations was somewhat differ ent. Prior to their first meeting with the investigator, parents of CF children received an introductory letter and then a follow-up phone call from this investigator. In this 72

PAGE 80

way they were well-oriented to the purpose of the study and had ample opportunity to think about their pariticpation and to ask questions. Parents of asthmatic children were not oriented in such a personal and concerned manner. These parents were approached by the investigator one time only and immediately given the questionnaires to be completed at home. This difference in procedure could account for the lower questionnaire return rate for the asthmatic parents than for the CF parents, which could be a selection bias. 73 Selection bias is even more evident in the procedure used to obtain control subj e cts. For reasons of confiden tialit y families in thfs group could not be randomly selected and individually approached. The procedure adopted required willing families to take the initiative to contact the investigator rather than vice versa. The obvious draw back of this procedure was that only interested and motivated control families were ultimately selected (only 1/3 of the families blindly invited). The clinic families did not have the luxury of anonymously refusing to participate. Another difference between the control group and the clinic groups was that the control children were all from the same town and all attended the same elementary school. This suggests that they are probably a more homogeneous group than the groups of chronically ill children. There is also some reason to suspect a difference in socio economic status between the groups. Although SES was not formally assessed, the control children appeared to be

PAGE 81

I I from predominantly upper-middle class families. The sick children, on the other hand, appeared to be from families representing a wider range of socio-economic levels. These differences in samples and sampling procedures are limitations of the study which should be taken into account when interpreting the results. It has been shown however, that the three groups of children selected as subjects for this study were similar on a number of important dimensions, i.e., age, race and sex. In addition they were all from roughly the same geographic area. As previously mentioned, few studies in this area have used healthy control groups for purposes of comparison. It is felt that the control group used in this study, although not perfect, does represent a methodological improvement over the use of standardized norms. The first major finding in this study was that the presence of chronic childhood illness was not significantly related to extreme family functioning (as assessed by mother's Dev ADP and Dev COH scores on FACES). The Circumplex Model of family systems proposes that families under stress have a tendency to move toward the extremes on adaptability and cohesion. If this model is correct, and if FACES and the method used to interpret FACES scores are valid, than there are two possible explanations for the above result: Either families of chronically ill chil74 \ ctren are not experiencing stress associated with the illness, \ or they are successfully coping with the stress that does exist.

PAGE 82

This second explanation seems most plausible. Caring for seriously ill children has to be inherently stressful for the affected families. It does not have to lead to deficits in family functioning, however. Circumplex theory suggests that healthy families may move toward the extremes of adaptability and cohesion when initially confronted with 75 a stressful event but will eventually gravitate back toward the moderate ranges as they successfully cope with these events. It is possible that this is exactly what the families of chronically ill children have done; i.e., they have adapted to the stress in a normal and healthy manner. An interesting test of the Circumplex Model would be to study changes in family functioning relative to chr0nic childhood illness over the course of the disease. If the theory is correct, one would expect healthy families to move toward the extremes on adaptability and cohesion at the time of diagnosis and during periods of acute exacerbations of the illness, but to gravitate back toward the middle as the illness is stabilized. Dysfunctional families could therefore be identified by their inability to move back and forth in this fluid manner. Another factor that should be considered in interpreting the above result is that FACES may not be an appropriate instrument for assessing group differences. To date, FACES has been used primarily as a clinical tool for identifying problem families. Although apparently successful in this regard, it may not be sensitive enough for assessing subtle

PAGE 83

differences in group trends. It should be noted that even though the differences between groups in extreme family functioning were not significant, they were in the predicted direction (refer to Table 3). The mean deviation scores for families of chronically ill children were higher than those for families of healthy children, and CF families tended to deviate more than asthmatic families. The method developed in this study for converting raw FACES scores on adaptability and cohesion to deviation 76 scores has not previously been used as a measure of extremity rating. For this reason it could be concluded that the method has no established validity and therefore yields meaningless results. Given the wide range of raw scores considered to be indicative of healthy functioning, it could be argued that moderate deviations from the means would not be particularly important. However, the validity of this procedure was supported by the relationship found to exist between deviation scores and child adjustment. As hypothesized, mother's Dev ADP and Dev COH scores were significantly related to the number of behavior prob lems exhibited by the child. The relationship between these deviation scores and the child's self-concept was extremely close to significance. These results are inter esting in that they support the use of Dev ADP and Dev COH scores as measures of extremity ratings while providing evidence validating the Circumplex Model. Circumplex theory suggests that dysfunctional families will be located

PAGE 84

more toward the extremes on cohesion and adaptability than will healthy families. If one makes the logical assumption that there is a higher frequency of poorly adjusted children in dysfunctional families than in healthy families, then according to the theory, there will be a higher frequency 77 of poorly adjusted children in families located at the extremes of the Circumplex dimensions than in families located at moderate levels along the dimensions. This is essentially what was found in this study. This result was obtained for pooled data regardless of illness. The implica tion of this finding in term of adjustment in chronically ill children will be discussed below. Results concerning the relationship between chronic illness and the psychological adjustment of the afflicted children were mixed. As expected, chronically ill children had significantly more behavior problems than healthy chil dren. Contrary to expectation, however, they did not differ 1 in self-concept. These two variables were examined inde pendently because they were thought to be measuring very different aspects of the child's adjustment. Self-concept relates to an internal feeling or perception of self-worth, while behavior problems reflect a learned style of outward interaction with the environment. In speculating about the reason for the discrepancy noted above, a possible explanation arises involving the relationship established between sick child and parent (especially mother). It has often been noted that there

PAGE 85

is a tendency for sick children to be overprotected by their mothers and to be treated as special (Travis, 1976). It seems quite possible that mothers may compensate for their own feelings of guilt and helplessness in the face of their child's misfortune by being more tolerant of misbehavior 78 in these children and less likely to use consistent discipline. This could have the effect of making these children more prone to developing behavior problems without necessarily affecting the way they feel about themselves. The relationship between parental discipline and behavior problems in chronically ill children has not previously been explored and might be a direction to pursue in future research. In general, the results of this study concerning adjust ment in chronically ill children are consistent with previous research. There does appear to be some relationship between childhood illness and the presence of adjustment difficulties in afflicted children. This relationship is not strong however, and appears to be more a function of mediating variables than of the illness per se. In this study, family functioning did appear to be an important mediating variable in the interaction between illness and adjustment variables. As previously discussed, the relationship found between extreme family functioning and behavior problems in the children was highly significant. In fact, it was much more significant than the relationship found between chronic ill ness and behavior problems. This raises the question of

PAGE 86

whether the higher incidence of behavior problems reported in both groups of sick children than in the healthy children might be due more to differences in family functioning than to the presence of the illness, even though these differ ences were not found to be significant. In any event, it can be concluded from the results of this stud y that adjust ment difficulties in children (defined in terms of behavior problems and self-concept) are more related to extreme family functioning than to the presenc e of chronic childhood illness. This conclusion suggests that f amily functioning is an important mediating va~iable to consider in childhood adjustment to chronic illness. Parent's orientation toward death was also explored as 79 a possible mediating variable affecting psychological adjust ment in sick children. It was hypothesized that having a child with a chronic physical illness--especially if this illness was terminal--would have the affect of making parents anxious and threatened b y the thought of death. Furthermore, it was felt that with fatally ill children there would be a relationship between child adjustment and parental orienta tion toward death, since parents most threatened and anxious about death would be least able to provide an open and sup portive atmosphere for the child, relative to his/her illness. In exploring these hypotheses, death anxiety, as measured by the DAS, was not found to be an important variable. The DAS is a very short, true-fals e instrument which may not be sensitive enough to subtle differences between groups

PAGE 87

in orientation toward death. Another possible explanation for this finding is that anxiety associated with death may not be a stable entity in parents, but may fluctuate markedly relative to the situation. This hypothesis is consistent with results reported by Neimeyer et al. (1977) that DAS scores and situational anxiety scores increased significantly in a group of subjects after viewing a 30 minute film depicting war atrocities. If this is the case, one would expect parental death anxiety to increase during exacerbations of the illness but to subside when the illness is stabilized. As previouslmentioned, all of the sick children in this study were in stable condition when tested. It would be interesting to explore changes in DAS scorEs over the course of the illness. Death threat, as measured by the TI, was found to be an important variable in the above analysis, but in an unexpected way. Contrary to prediction, death threat was significantly higher in parents of healthy children than 80 \ in parent3 of terminally ill children. This results in a serious question about the validity of the primary assumption underlying the initial hypothesis. This assumption was that parents of dying children react to the terminal nature of the illness in an unhealthy manner. That is, they deny the reality of the illness, they are fearful of and cannot accept the prognosis, and (ultimately) they become less able to conceive of death as a personal reality (death

PAGE 88

threat). The fact that the opposite was actually found to be the case attests to generally healthy rather than unhealthy parental adjustment to illness. This suggests that parents of terminally ill children have thought about the implications of death for their children, and have consequently restructured their own personal construct systems in order to make death more compatible with their perceptions of self. When the initial assumption is changed concerning a healthy vs. unhealthy parental response to the implication of death in terminal childhood illness, the above results are quite consistent with personal construct theory. This theory proposes that a person's cognitive orientation toward his/her world is generally stable (especially for constructs that are particularly meaningful for the individual), but that this cognitive orientation does change in response to significant life experience. Obviously, having a child with a terminal illness is a significant life experience. The results of this study show that, as predicted by the theory, the personal construct systems of mothers are affected by this experience. The results obtained concerning the relationship between maternal death threat and adjustment in the fatally ill children are difficult to understand. Contrary to expectation, there was a tendency for greater maternal death threat to be associated with fewer behavior problems 81 and higher self-concepts in the CF children. If this is more

PAGE 89

than just a spurious result it could call into question the assumption discussed above, that lower parental death threat is a healthy response to the presence of terminal childhood illness. It is particularly surprising that this relationship seems to hold true for all three groups of families. One can only speculate as to the meaning of this finding. Possibly, what is being measured by the TI is a more general personality trait than is usually assumed. It could, for 82 example, be tapping a person's tendency to remain impersonal and matter-of-fact in the presence of emotionally-related stimuli. Someone who is able to be objective and emotionally unresponsive to topics (like death) that usually arouse some affect in people might also be overly objective and emotionally unresponsive to people. A parent with this tendency might have difficulty meeting the emotional needs of his/her children, which could eventually result in problems of adjust ment for these children. This is obviously pure speculation; there is no evidence currently available substantiating this circuitous connection. However, the finding does pose some interesting questions which are deserving of further attention. In conclusion, the present study supports the contention that chronic childhood illness has a general stressful effect which can lead to deficits in functioning, although in most cases, individuals and families appear to adapt in basically healthy ways. It also supports the notion that mediating

PAGE 90

variables (family functioning, etc.) are more important in affecting the psychological adjustment of the sick child than is the illness itself. More research is needed in this area in order to better understand the process of adaptation to childhood illness. Two lines of research would be useful in this regard. One line would involve continuing the search for mediating variables that are important in the adjustment process. Once these variables have been identified and clearly defined, systematic intervention would be possible with the goal of insuring a healthy response to the disease. At present, information concerning these mediating variables is scattered and often times contradictory. A second course for future research is in the direction of longitudinal designs. There is presently little hard data available concerning changes in individual and family adjustment during the course of a child's chronic illness. Cross-sectional studies such as the one reported here seem to be based on the assumption that illnesses are static entities which have a uniform effect. Diseases are dynamic, as are the people who are affected. To really understand the adaptational process, it is necessary to observe changes over time. As the adjustment demands change over the course of the illness, so too will the implications for treatment and intervention. 83

PAGE 91

Research concerning the psychosocial effects of chronic childhood illness generally supports the view that the presence of disease is frequently associated with deficits in individual and family functioning. There has been a gradual shift in recent years from considering the disease 84 as causative in this association to emphasizing the importance of mediating variables. The present study was an attempt to increase our understanding of this process by examining the relative impact of the illness per se versus mediating variables (family functioning and parental attitudes toward death) on individual adjustment in the chronically ill child. Two major questions were addressed in this regard. The first was general in nature and concerned the overall effect of the presence and severity of illness on relevant adjust ment variables--behavior problems and self-concept of the afflicted children, extreme family functioning, and parental death threat and death anxiety. The second question examined the relationship between adjustment in the children and both family functioning and parental death orientation, when controlling for effects of the illness. In this way the impact of disease was compared to the impact of mediating variables (family functioning and parental death orientation) relative to childhood adjustment. To explore these questions, three groups of families were studied: families of children with cystic fibrosis,

PAGE 92

families of children with chronic asthma, and families of healthy children. Results of this study indicated that the presence of chronic childhood illness was not associated with significant deficits in family functioning or with decreased self-concepts in the afflicted children. It was 85 found to be associated with an increased frequency in report ed behavior problems for the children. Contrary to expecta tion, death threat was significantly lower in mothers of terminally ill children than in mothers of healthy children. Maternal death anxiety was not affected by the presence of childhood illness. Family functioning was -k>und to be a significant mediating variable affecting childhood adjustment regardless of illness. It was concluded that childhood adjustment was related more to this variable than to the presence of a chronic illness. The implications of these results were discussed, as were avenues for future research.

PAGE 93

APPENDIX A TEMPLER'S DEATH ANXIETY SCALE Below are 15 statements concerning attitudes toward death or death related issues. For each statement you are to check whether that attitude is mostly "true" or mostly "false" for yourself. Please be sure to respond to all 15 statements by checking either "true" or "false". 1 I am very much afraid to die. 2. The thought of death seldom enters my mind. 3. It doesn't make me nervous when people talk about death. 4. I dread to think about having to have an operation. 5. I am not at all afraid to die. 6. I am not particularly afraid of getting cancer. 7. The thought of death never bothers me. 8. I am often depressed by the way time flies so very rapidly. 9. I fear dying a painful death. 10. The subject of life after death troubles me greatly. 11. I am really scared of having a heart attack. 12. I often think about how short life really is. 13. I shudder when I hear people talking about a World War III. 14. The sight of a dead body is horrifying to me. 15. I feel that the future holds nothing for me to fear. 86 True False

PAGE 94

APPENDI X B THREAT I N DE X (Tip40) Below is a list of bipolar dimensions. For each dimensions please circle the side with which you see yourself more closely associated. For example, do you associate yourself more with the term "predictable" or "random"? predictable random productive unproductive empty meaningful learning not learning sad happy purposeful not purposeful personal impersonal responsible not responsible lack of control control bad good satisfied dissatisfied not caring caring relating to not relating crazy healthy others to others personality pleasure pain conforming not conforming feels bad feels good animate inanimate objective subjective weak strong alive dead useful useless helping others being selfish closed open specific general peaceful violent kind cruel freedom restriction incompetent competent non-existence existence insecure secure understanding not understanding static changing sick healthy unnatural natural stagnation growth calm anxious abstract concrete easy hard hope no hope 87

PAGE 95

For each dimension circle the side with which you most closel y associate your own death. predictable random productive unproductive empty meaningful learning not learning sad happy purposeful not purposeful personal impersonal responsible not responsible lack of control control bad good satisfied dissatisfied not caring caring relating to not relating crazy healthy others to others personality pleasure pain ---conforming not conforming feels bad feels good animate inanimate objective subjective weak strong alive dead useful useless hel ping others being selfish closed open specific general peaceful violent kind cruel freedom restriction incompetent competent non-existence existence insecure secure understanding not understanding static changing sick healthy unnatural natural stagnation growth calm anxious abstract concrete easy hard hope no hope 88

PAGE 96

APPENDIX C INTRODUCTORY LETTER TO PARENTS OF CF CHILDREN Dear Parents, We would like to invite you to participate in a psychological study being conducted at Children's Hospital in Boston. The purpose of the study is to explore psychosocial affects of chronic childhood illness on families and afflicted children. We hope to be able to use the information obtained in this study to better help families, like your own, adjust to the stress of caring for a child with cystic fibrosis. If you agree to participate in this study you (the mothers) will be asked to complete a series of four questionnaires during your next clinic visit. These questionnaires will require you to think about your present family situa tion, the behavior of your child, and certain personal attitudes you have toward the idea of death. The questionnaires will take roughly one hour to complete. During this time your child will also be interviewed by the investigator for about 15 minutes. The purpose of this interview is to assess his/her self esteem; i.e., how he/she feels about himself/herself. The idea of death will not be discussed with the child. At the conclusion of the study all participating families will be given the opportunity to meet with the investigators to discuss the general findings. If desired, suggestions might also be made concerning the adjustment needs of specific families. A few days prior to your next clinic visit you will be called by the investigators to discuss whether or not you wish to participate. time we will be happy to answer any questions you may have concerning study. Participation is, of course, voluntary. one of At that this Kon-Taik Khaw, M.D. Senior Associate in Medicine CyEtic Fibrosis Program Children's Hospital Medical Center 89 Sincerely, Brian L. Lewis, M.ED. Co-Investigator Cystic Fibrosis Program Children's Hospital Medical Center

PAGE 97

APPENDIX D INTRODUCTORY LETTER TO PARE N TS OF HEALTHY CHILDREN Dear Parents, We, at Children's Hospital, have undertaken a study in which we are examining psychological effects of chronic childhood illness. We are specifically interested in looking at how the presence of chronic childhood illness effects family func tioning, parental attitudes and the behavior and self concept of the afflicted children. We hope to be able to use the information obtained in this study to bet ter help families adjust to the stresses of caring for a child with special needs. Two groups of families are being studied at Children's Hospital; families of children with cystic fibrosis and families of children with chronic childhood asthma. In order to determine which effects are due specifically to the presence of the physical illness it is necessary to compare the data obtained from these families with that of a control group of "normal" families; i.e., families of children with no previous history of chr~nic physical illness. We would like to invite you to participate as one of these control families. Participation in this study will only require approximately one hour of your time. During this hour you will complete four questionnaires. These questionnaires will require you to think about your present family situation, the behavior of your child, and certain personal attitudes y ou have toward the idea of death. Children in this study will be asked to complete one questionnaire designed to e x amine how he/she feels about himself/herself This will require only about 15 minutes of her/her time. You will have a chance to examine all questionnaires before you agree to participate. We have sent these letters to parents of randomly selected Brookline children who are roughly the same age as those under investigation at Children's Hospital If you feel you might be willing to volunteer for this study or if you would simply like more information about it, complete and return the enclosed, self addressed postcard within the next few days. The only information needed on this post card is your name, telephone number, and age and sex of your child. When we receive the post card you will be called by the co-investigator, who will answer any questions you may have. If you then agree to participate, this inves tigator will bring the questionnaires directl y to your home at a mutually convenient time. You will then have about a week to complete them and mail them back to Children's Hospital. Although this study is primarily concerned with adjustment in families of children with chronic illness we feel that it will have generalized relevance for all families. At the conclusion of the study interested participants will be given feedback concerning the results. Your participation would be greatly appreciated. P.S We would appreciate your cooperation in returning the post cards to us within the next two days. 90 Sincerely, Brian L. Lewis Co -i nvesti g ator C y stic Fibrosis Program

PAGE 98

APPENDIX E RESULTS OF ANALYSES BASED ON FATHER'S DATA

PAGE 99

Table 20. Mean Dev ADP and Dev COH Scores for Families of CF, Asthma and Control Children Mean Mean Dev Dev Group N ADP COH CF 20 12.39 13.88 Asthma 18 8.64 13.94 Control 18 9.02 13.99 92 Table 21. F Scores for MANOVA Examining the Differences between CF and Control [amilies on Dev ADP and Dev COH Source Dev ADP Dev COH Dev ADP/Dev COH df 1 1 2,35 F Value 1.47 0.00 0.72 p .233 .970 .496 Table 22. F Scores for MANOVA Examining the Differences between Asthma and Control Families on Dev ADP and Dev COH Source Dev ADP Dev COH Dev ADP/Dev COH df 1 1 2,33 F Value 0.03 0.00 0.01 p .874 .989 .988

PAGE 100

93 Table 23. F Scores for MANOVA Examining the Differences between CF and Asthma Families on Dev ADP and Dev COH Source Dev ADP Dev COH Dev ADP/Dev COH dr 1 1 2,35 F Value 2.14 0.00 1.06 p .152 .985 .359 Table 24. Mean BPC Scores and Standard Deviations for CF, Asthma and Control Children Standard Group N Mean Deviation CF 20 3.90 5.98 Asthma 17 5.59 7.46 Control 18 1.89 2.14 Total 55 3.76 5.72 Table 25. Results oft-tests Examining Differences between Groups on Mean BPC Scores Group Comparisons Value t Value p CF vs Control 2.01 1.10 .277 Asthma vs Control 3.70 1.94 .057 CF vs Asthma -1.69 -0.91 .367

PAGE 101

94 Table 26. Mean DAS Scores and Standard Deviations for Fathers of CF, Asthma, and Control Children Standard Group N Mean Deviation CF 20 5.00 3.01 Asthma 18 6.00 3.07 Control 18 5.33 3.07 Total 56 5.43 3.02 Table 27. R e sults oft-tests E x amining Differences between Groups on M e an DAS Scores of Fathers Group Comparisons CF vs Control Asthma vs Control CF vs Asthma V a lue t -0.33 0.67 -1.00 Value -0.34 0.66 -1.01 p .738 .514 .317 Table 28. Mean TI Scores and Standard Deviations for Fathers of CF, Asthma, and Control Children Standard Group N Mean Deviation CF 17 16.18 9.16 Asthma 18 19.94 11.77 Control 17 24.53 8.87 Total 52 20.21 10.44

PAGE 102

Table 29. Results oft-tests Examining Differences between Groups on Mean TI Scores of Fathers Group Comparisons Value t Value p CF vs Control -8 35 -2.41 .019* Asthma vs Control -4.58 -1.35 .184 CF vs Asthma -3.77 -1.11 .274 *significant at p = .05 Table 30. F Scores for Multiple Regression Analysis of the Relationship betwee n Child's P-H Score and Family Dev ADP and Dev COH (based on father's response) for the Pooled Data Source N df F Value 95 p Dev ADP Dev COH Dev ADP/Dev COH 53 53 57 1 1 2,53 2.45 0.07 1.26 .123 .798 .292 Table 31. F Scores for Multiple Regression Analysis of the Relationship between Child's BPC Score (father's response) and Family Dev ADP and Dev COH (based on father's response) for the Pooled Data Source N df F Value p Dev ADP Dev COH Dev ADP/Dev COH 52 52 52 1 1 2,52 0.03 0.78 0.40 .873 .380 .669

PAGE 103

96 Table 32. Pearson r Correlation between Children's P-H Scores and Father's TI and DAS Scores Parental Attitude Death Threat (TI) Death Anxiety (DAS) N 17 20 r -0.06 -0.12 p .415 .303 Table 33. Pearson r Correlations between CF Children's BPC Scores (father's r~ting) and Father's TI and DAS Scores Parental Attitude Death Threat (TI) Death Anxiety (DAS) N 17 20 r 0.02 0.38 p .465 .051 Table 34. Pearson r Correlations between Children's P-H Scores and Father's TI Scores for CF, Asthma and Control Groups and for Pooled Data Group CF Asthma Control Pooled N 17 18 17 52 r -0.06 -0.14 0.18 -0.08 p .415 .293 .248 .279

PAGE 104

REFERENCES v Bakwin, H., & Bakwin, R. M., Behavior disorders in children (4th ed.). Philadelphia: W. B. Saunders, 1972. Barker, R. G., Wright, B. A., Myerson, L. & Gonick, M. R. Adjustment to physical handicap and illness: A survey of the social psychology of physique and disability. New York: Social Science Research Council, 1953. Bedell, J. R., Giordani, B., Amour, J. L., Tavormina, J., & Boll, T. Life stress and the pyschological and medical adjustment of chronically ill children. Journal of Psychosomatic Research, 1977, 21, 237-242. Begleiter, M. L., Burry, V. F., & Harris, D. J. Prevalence of divorce among parents of children with cystic fibrosis and other chronic diseases. Social Biology, 1976, 23, 260-264. Binger, C. M. Childhood leuKemia; emotional impact on patient and family. New England Journal of Medicine, 1969, 280, 414-418. Bowen, M. The use of family theory in clinical practice. In J. Haley (Ed.), Changing families. New York: Grune & Stratton, 1971. Breslow, J. L., McPherson, J., & Epstein, J. Distinguishing homozygous and heterozygous cystic fibrosis fibroblasts from normal cells by differences in sodium transplant. New England Journal of Medicine, 1981, 304, 1-5. Buros, 0. K. (Ed.). The seventh mental measurements yearbook. Highland Park, New Jersey: The Gryphon, 1972. Clevelanc, S., Reitman, E., and Brewer, E. Psychological factors in juvenile rheumatoid arthritis. Arthritis and Rheumatism, 1965, 8, 1152-1158. Collett, L. J., & Lester, D. The fear of death and the fear of dying. Journal of Psychology, 1969, 72, 179-181. Coddington, B. D. The significance of life events on etiologic factors in the diseases of children. Journal of Psychoso matic Research, 1972, 36, 7-18. Cytryn, L., Moore, V. P., & Robinson, M. E. Psychological adjustment of children with cystic fibrosis. In E. J. Anthony & C. Koupernik (Eds.), The child in his family: The impact of disease and death. New York: John Wiley, 1973. 97

PAGE 105

98 Denning, C.R., Gluckson, M. M., & Mohr, I. Psychological and social aspects of cystic fibrosis. In J. A. Mangos & R. L. Talamo (Eds.), Cystic fibrosis: projections into the future. Miami: Symposia Specialists, 1976. Dickstein, L. S. Death concern: measurement and correlates. Psychological Reports, 1972, 30, 563-571. ~ iller, L. Psychological aspects of physically handicapped children. In B. B. Wolman (Ed.) Manual of child psychology. New York: McGraw-Hill, 1972. Dinnage, R. he handicapped child. ment research review (Vol. 1) Dinnage, R. The handicapped child. ment research review (Vol. 2). Studies in child develop London: Longman, 1970. Studies in child develop London: Longman, 1972. Doershuk, C. F., Wood, R. E. & Boat, T. F. Pulmonary therapy for cystic fibrosis. In J. A. Mangos & R. L. Talamo (Eds.) Cystic fibrosis: projections into the future. Miami: Symposia 4 Specialists, 1976. Douglas, J. W. B., & Bloomfield, J.M. Children under five. London: Allen and Unwin, 1958. Epting, R.R., Rainey, L. C., & Weiss, M. Constructions of death and levels of death fear. Death Education, 1979, 3, 21-30. Evans, A. E., & Edin, S. If a child must die ... The New England Journal of Medicine, 1968, 278 (3), 138-142. French, T. M. Psychogenic factors in asthma. American Journal of Psychiatry, 1936, 96, 87-98. Gayton, W. F., & Friedman, S. B. A review of the psychosocial aspects of cystic fibrosis. American Journal of Diseases in Children, 1973, 126, 856-859. / Gayton, W. F., Friedman, S. B., Tavormina, J. F., & Tucker, F. Children with cystic fibrosis: 1. Psychological test findings of patients, siblings and parents. Pediatrics, 1977, 59, 888-894. Graham, P., Rutter, M., Yule, W., & Pless, I.B. Childhood asthma; A psychosomatic disorder? Some epidemiological considerations. British Journal of Preventive and Social Medicine, 1967, 21, 78-85. Green, M. The management of long-term nonlife-threatening illness. In M. Green & R. J. Haggerty (Eds.), Ambulatory pediatrics. Philadelphia: W. B. Saunders, 1965.

PAGE 106

Green, M., & Levitt, E. E. Constriction of body image in children with congenital heart disease. Pediatrics, 1962, 29, 438-441. 99 Grokoest, A., Snyder, A., & Schleger, R. Juvenile Rheumatoid Arthritis. Boston: Little, Brown, 1962. Haley, J. A family therapy. Changing families. New York: Grune & Stratton, 1971. Haley, J. Strategies of psychotherapy. New York: Grune & Stratton, 1972. Harding, R. K., Heller, J. R., & Kesler, R. W. The chronically ill child in the primary care setting. Primary Care, 1979, 6(2), 311-324. Holmes, T., & Rahe, R. The social readjustment rating scale. Journal of Psychosomatic Research, 1967, 11, 219-225. Kastenbaum, R., & Costa, P. T. Psychological perspectives on death. Annual Review of Psychology, 1977, 28, 225-249. Kelly, G. A. The psychology of personal constructs (2 vols.). New York: Norton, 1955. Korsch, B., & Barnett, H. L. The physician, the family, and the child with nephrosis. Journal of Pediatrics, 1961, 58, 707-715. Korsch, B. M., Negrette, V. F., Gardner, J. E., Weinstock, C. L., Mercer, A. S., Grushkin, C. M., & Fine, R. M. Kidney transplantation in children: Psychosocial follow-up study on child and family. Journal of Pediatrics, 1973, 83, 399-408. Krieger, S. R. Death orientation and the specialty choice and training of physicians (Doctoral dissertation, University of Florida, 1975). Dissertation Abstracts International, 1977, 37, 3616B. Krieger, S. R., Epting, F. R., & Hays, C.H. Validity and reliability of provided constructs in assessing death threat: A self-administered form. Omega, 1979, 10, 87-95. Krieger, S. B., Epting, F. R., & Leitner, L. M. Personal constructs threat, and attitude toward death. Omega, 1974, ~' 299-310. Lansky, S. B., Cairns, N. U., Hassanein, R., Wehr, B. A., & Lowman, J. T. Childhood cancer: Parental discord and divorce. Pediatrics, 1978, 62(2), 184-188.

PAGE 107

Lawler, R.H., Nakielny, W., & Wright, N. A. Psychological implications of cycstic fibrosis. Canadian Medical Association Journal, 1966, 94, 1043-1046. 100 Liebman, R., Minuchin, S., & Baker, L. The use of structural family therapy in the treatment of intractable asthma. American Journal of Psychiatry, 1974, 131(5), 535-540. Linde, L. M., Rosof, B., & Dunn, O. J. Longitudinal studies of intellectual and behavioral development in children with congenital heart disease. Acta Paediatrica Scandinavica, 1970, 59, 169-176. Lipowski, z. J. Physical illness, the individual and the coping process. Psychiatry in Medicine, 1970, l, 91-101. vM agrab, P.R. (Ed.). Psychological management of pediatric problems (Vol. 1). Baltimore: University Park Press, 1978. Mattsson, A. Long-term physical illness in childhood: A challenge to psychosocial adaptation. Pediatrics, 1972, 50, 801-811. Mattsson, A., & Gross, S. Social and behavioral studies on hemophilic children and their families. Journal of Pediatrics, 1966, 68, 952-957. McAnarner, E., Pless, I. B., Satterwhite, B., & Friedman, S. Psychological problems of children with chronic juvenile arthritis. Pediatrics, 1974, 53, 523-528. McCollum, A. T., & Gibson, L. E. Family adaptation to the child with cystic fibrosis. Journal of Pediatrics, 1970, 77, 571-578. Meijer, A. Emotional disorders of asthmatic children. Child Psychiatry and Human Development, 1979, Q(3), 161-169. Minuchin, S., & Barcai, A. Induced family crisis. In J. H. Masserman (Ed.) Science and psychoanalysis (Vol. 14). New York: Grune & Stratton, 1969. Morrissey, J. R. Children's adaptation to fatal illness. Social Work, 1963, 8, 81-87. Natterson, J.M., & Knudson, A.G., Jr., Observations concerning fear of death in fatally ill children and their mothers. Psychosomatic Medicine, 1960, 22, 456-465. Neimeyer, R. A., & Dingemans, P. M. A. J. Death orientation in the suicide prevention worker. Omega, 1980, 11(1), 15-28.

PAGE 108

101 Neimeyer, R. A., Dingemans, P. M., & Epting, F. R. Convergent validity, situational stability and meaningfulness of the Threat Index. Omega, 1977, ~, 251-265. Neuhaus, I. C. children. A personality study of asthmatic and cardiac Psychosomatic Medicine, 1958, 20, 181-186. O'Leary, K. D., & Johnson, S. B. Psychological assessment. In H. Quay & J. Werry (Eds.), Psychopathological disorders of childhood (2nd ed.). New York: John Wiley, 1979. Olson, D. H., Bell, R., & Portner, J. and cohesion evaluation scales. Science, 1978. FACES: Family adaptability St. Paul: Family Social Olson, D. H., & McCubbin, H. I. Circumplex Model of marital and family systems: Application to family stress and crisis intervention. In H. I. Mccubbin (Ed.) Family stress, coping and social support. Minneapolis: Burgess, 1980. (a) Olson, D. H., Russell, C. S., & Sprenkle, D. H. Circumplex Model of marital and family systems II: Empirical studies and clinical intervention. Advances in family intervention, assessment and theory, 1980, !, 129-176. (b) Olson, D. H., Sprenkle, D. H., & Russell, C. S. Circumplex Model of marital and family systems: 1. Cohesion and adaptability dimensions, family types, and clinical application. Family Process, 1979, 18(1), 3-27. Pandley, A. E. Factor analytic study of attitudes toward death among college students. International Journal of Social Social Psychiatry, 1975, 21, 7-11. Paquay-Weinstock, M., Appelboom-Fondu, J., & Dopchie, N. Influence of a chronic disease on the evolution of the family dynamics. Acta Paedopsychiatrica, 1979, 44, 219-228. Peterson~ D.R. Behavior problems of middle childhood. Journal of Consulting Psychology, 1961, 25, 205-209. Piers, E. V. The Piers-Harris children's self-concept scale: Research monograph #1. Nashville: Counselor Recordings and Tests, 1977. Piers, E., & Harris, D. Age and other correlates of self concept on children. Journal of Educational Psychology, 1964, 55, 91-95. Piers, E. V., & Harris, D. B. Manual for the Piers-Harris children's self-concept scale. Nashville: Counselor Recordings and Tests, 1969.

PAGE 109

102 Pilling, D. The child with a chronic medical problem cardiac disorders, diabetes, haemophilia. (National Children's Bureau Report), Slough: NFER Publishing Co., 1973. Pless, I.B. Epidemiology of chronic disease. In M. Green & R. J. Haggerty (Eds.) Ambulatory pediatrics. Philadelphia: W. B. Saunders Co., 1968. Pless, I. B., & Douglas, J. W. B. Chronic illness in childhood: 1. Epidemiological and clinical characteristics. Pediatrics, 1971, 47, 405-414. Pless, I. B., & Pinkerton, P. Ctironic childhood disorder: Promoting patterns of adjustment. London: Kimpton, 1975. Pless, I. B., & Roghmann, K. J. Chronic illness and its consequences: Results from three epidemiologic surveys. Journal of Pediatrics, 1971, 79, 351-359. Pless, I., Roghmann,K.,& Haggerty, B. Chronic illness, family functioning, and psychological adjustment: a model for the allocation of preventive mental health services. International Journal of Epidemiology, 1972, !, 271-277. Pless, I.B. & Satterwhite, B. A measure of family functioning and its application. Social Science and Medicine, 1973, 2, 613-621. Purcell, K. Childhood asthma: The role of the family relation ships, personality, and emotions. In A. David (Ed.) Child personality and psychopathology (Vol. 2). New York: New York: J. Wiley, 1975. Purcell, K., & Weiss, J. H. Asthma. In C. G. Costello (Ed.), Symptoms of psychopathology. New York: J. Wiley, 1970. Quay, G. Measuring dimensions of deviant behavior: The Behavior Problem Checklist. Journal of Abnormal Child Psychology, 1977, 5, 277-289. Quay, H. C., & Peterson, D.R. Manual for the Behavior Problem Checklist. Authors, 1979. Riess, B. F., & DeCillis, O. E. Personality differences in allergic and non-allergic children. Journal of Abnormal Social Psychology, 1940, 35, 104-113. Rigdon, M.A., Epting, F. R., Neimeyer, R. A., & Krieger, S. R. The Threat Index: a research report. Death Education, 1979, 3, 245-270. Roghmann, K. J., & Haggerty, R. J. Rochester child health surveys: 1. Objectives, organization and methods. Medical Care, 1970, ~' 47-59.

PAGE 110

103 Rutter, M., & Graham, P. Psychiatric aspects of intellectual and educational retardation. In R. M. Rutter, J. Tizard & K. Whitmore (Eds.), Education, health and behavior. London: Longman, 1970. Rutter, R. M., Tizard, J., & Whitmore, K. Education, health and behavior. London: Longman, 1970. Sandler, N. Working with families of chronic asthmatics. Journal of Asthma Research, 1977, 15(1), 15-21. Schatia, V. The incidence of neurosis in cases of bronchial asthma as determined by the Rorschach Test with psychiatric examination. Psychosomatic Medicine, 1941, 3, 156-169. Simpson, M.A. Social and psychological aspects of dying. In H. Wass (Ed.), Dying: Facing the facts. New York: Hemisphere, 1979. Spinetta, J. J., & Maloney, L. J. Death anxiety in the out patient leukemic child. Pediatrics, 1975, 56(6), 10341037. Stein, R. E. K., & Riessman, C. K. The development of an impact on-family scale: Preliminary findings. Medical Care, 1980, 18(4), 465-472. Sterky, G. Family background and state of mental health in a group of diabetic schoolchildren. Acta Paediatrica, 1963, 52, 377-390. Sultz, H. A., Schlesinger, E. B., & Mosher, W. E. Long term childhood illness. Pittsburgh: University of Pittsburgh Press, 1972. Swift, C.R., Seidman, F. & Stein, H. Adjustment problems in juvenile diabetics. Psychosomatic Medicine, 1967, 29, 555-571. Tavormina, J.B., Kastner, L. S., Slater, P. M., & Watt, S. L. Chronically ill children: a psychologically and emotionally deviant population? Journal of Abnormal Child Psychology, 1976, i(2), 99-110. Templer, D. I. The construction and validation of a death anxiety scale. Journal of Genetic Psychology. 1970, 82, 80-92. Templer, D. I., & Ruff, C. F. Death anxiety scale means, standard deviations and embedding. Psychological Reports, 1971, 29, 173-174.

PAGE 111

104 Thomas, L.A., Milman, D. H., & Rodriquez-Torres, R. Anxiety in children with rheumatic fever: Relation to route of prophylaxis. Journal of the American Medical Association, 1970, 212, 2080-2085. Travis, G. Chronic illness in children: Its impact on child and family. Stanford: Stanford University Press, 1976. Tropauer, A., Franz, M. N., & Dilgard, V. W. Psychological aspects of the care of children with cystic fibrosis. American Journal of Diseases in Children, 1970, 199, 424-432. Turk, J. Impact of cystic fibrosis on family functioning. Pediatrics, 1964, 34, 67-71. Vernick, J. Meaningful communication with the fatally ill child. In E. J. Anthony & K. C. Koupernik (Eds.), The child in his family: The impact of disease and death. New York: J. Wiley, 1973. Vernick, J., & Karon, M. Who's afraid of death on a leukemia ward? American Journal of Diseases in Children, 1965, 109, 393-397. Verwoerdt, A. Psychopathological responses to the stress of physical illness. Advances in Psychosomatic Medicine, 1972, ~, 119-141. Waechter, E. H. Children's awareness of fatal illness. American Journal of Nursing, 1971, 71, 1168-1172.

PAGE 112

BIOGRAPHICAL SKETCH Brian Lewis was born in Hartford, Connecticut, on January 28, 1952. He lived in rural Connecticut until the age of 14 at which time he moved with his family to Springfield, Vermont. After high school Brian attended Middlebury College in Middlebury, Vermont and graduated with a B.A. in psychology in 1974. In May, 1977 Brian Lewis received a masters degree in Counseling Psychology from the State University of New York at Buffalo. Immediately following, he entered the doctoral program in Counseling Psychology at the University of Florida. Coursework was completed by June, 1979 and during the next year Brian was a psychology intern at the Boston V. A. Medical Center in Jamaica Plain, Massachusetts. Following graduation, Brian plans to move back to Vermont where he has accepted a position as a psychologist in a private medical center. 105

PAGE 113

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. Paul G. Schauble, Chairman Professor of 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 .,,., Fran0p ting, Co-Cij.a1.rman Associate Professor of 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. Psychology

PAGE 114

I certify that I ha v e 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. Hannelore Wass Professor of Foundations of Education 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 Phi lophy. / , /:.-. .........,, ,.,.,.. --:--71;i -Y------/ / / \ ) :: ( J ,,,,, . -/ / r:: :. ,.. . L .,,,,, _ / Robert ZilJ...er-Profes s or of Psychology This dissertation was submitted to the Graduate Faculty of the Department of Psychology in the College of Liberal Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June 1981 Dean for Graduate Studies and Research

PAGE 115

r -------UNIV E RSITY OF FLORIDA I I I IIIIIIIIIIIIII IIIII II I I I I I I I II III II I I I II II II I I 11 11 1 1 1 1 111 111 1 3 1262 08553 5887

PAGE 116

r -------UNIV E RSITY OF FLORIDA I I I IIIIIIIIIIIIII IIIII II I I I I I I I II III II I I I II II II I I 11 11 1 1 1 1 111 111 1 3 1262 08553 5887