IW

............................... Littk‘aqbq

3 1293 00609 2807
LIBRARY
Michigan State

Quivers!”

 

 

 

#—

#-

This is to certify that the

thesis entitled

HEALTH LOCUS OF CONTROL AND ENGAGEMENT STYLE
IN CHILDREN AND ADOLESCENTS WITH CANCER

presented by '
Dan Allen Mondoux

has been accepted towards fulfillment
of the requirements for

M.A. degree in Psychology

Major pr fessor

 

 

0-7639 MS U is an Afiirman've Action/Equal Opportunity Institution

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before due due.

 

DATE DUE DATE DUE DATE DUE

 

   

   

 

 

[l
, g ,
‘ V 1 '
r . I, ‘ . I 5
.‘ (I -- '3.

r it; ‘ 5;." .1;

24%: v " " I
#6:“—
FEB 0 6 1996 I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

= if!

MSU Is An Affirmative AotIorVEquel Opportunlty Institution

 

 

______

HEALTH LOCUS OF CONTROL AND ENGAGEMENT STYLE
IN CHILDREN AND ADOLESCENTS WITH CANCER
By

Dan Allen Mondoux

A THESIS

Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS
Department of Psychology

1989

ABSTRACT

HEALTH LOCUS OF CONTROL AND ENGAGEMENT STYLE IN
CHILDREN AND ADOLESCENTS WITH CANCER

By

Dan A. Mondoux

This study investigated health locus of control beliefs (HLOC) and engagement
style in children with cancer and matched comparison children. Participants were
twenty-three pairs of children aged eight to 18 (n= 46). Each child with cancer was
matched with a child from his or her own classroom on the basis of sex, date of
birth, and race. The study used McKinney’s Test of Engagement Style, and ParCel
and Meyer’s Multidimensional Children’s Health Locus of Control test. Analyses
revealed that, contrary to predictions, children with cancer were more agent
(perceiving themselves as doing in social situations rather than being done to) than
children without cancer (p < .05). No differences between groups on HLOC were
found. Chance external HLOC was found to be negatively correlated with age for
children without cancer but not for children with cancer. Theoretical implications of
the findings, ramifications for children’s cancer treatment and future research were

discussed.

For Mom and Dad

ACKNOWLEDGMENTS

I would like to express my gratitude to the members of my committee for all
their efforts on my behalf. They each made a special contribution to this project:

Dr. Lauren Hanis, whose counsel and insight have made all the difference.

Dr. Robert Noll, whose suggestions, knowledge, and energy made this project
possible.

Dr. John Paul McKinney, whose enthusiasm has inspired me, whose
perspective has changed my way of thinking about research, and whose friendship I
value.

I want to thank Sarah LeRoy for the road trips, for the neighborhood walks,
and for Timmy.

I thank the Psychology Department for their support during this project.

I would like to thank Cynthia Kaufman for making me remember the reason.

I want to express my thanks and admiration to all the participants in this

project. Your time and effort is appreciated.

iv

TABLE OF CONTENTS

Page
LIST OF TABLES ....................................... vi
LIST OF FIGURES ..................................... vii
INTRODUCTION ........................................ 1
Locus of Control .................................... 7
Demographic Factors ............................. 9
LOC Scales .................................. 9
Engagement Style ................................... 11
Purpose ......................................... .13
Hypotheses and Predictions ............................. 14
METHOD . ......................................... 16
Research Participants ................................. 16
Instruments ....................................... 18
Procedures ....................................... 18
Analysis ......................................... 21
RESULT . . ......................................... 23
gens of Control ................................... 23
est of Engagement Style .............................. 27
DISCUSSION ......................................... 28
CONCLUSIONS ....................................... 32
APPENDICES ........................................ 35
Appendix A: Multidimensional Children’s Health Locus of Control . . . 35
Appendix B: Clinic Family Contact ....................... 37
Appendix C: Comparison Family Contact .................... 38
Appendix D: Consent Form ............................. 39
LIST OF REFERENCES ................................. 40

LIST OF TABLES
TABLE Page

1 Characteristics of the sample: Cancer and Comparison groups ....... 19

2 Pearson Correlations between age and the Locus of Control
Subscales, and Subscale Intercorrelations .................... 25

FIGURE

LIST OF FIGURES

Engagement Style sample card

vii

OOOOOOOOOOOOOOOOOOOOO

INTRODUCTION

The problem of childhood chronic illness and its psychological impact is
becoming increasingly important. Medical advances in recent years have greatly
extended the life expectancies of children with serious chronic illnesses (Hobbs,
Perrin, Ireys, Moynihan, & Shayne, 1984). As a result, growing numbers of
children and families are faced with the burden of long-term chronic illness
(Thompson, 1986). Chronic illnesses can be categorized in a variety of ways
including how serious the disease is (Hobbs et al., 1984) and whether or not it is
primarily of physical origin (Thompson, 1986). Some chronic illnesses, such as
cystic fibrosis and cancer, are highly life-threatening whereas others, such as asthma
and various orthopedic conditions are not nearly so dangerous.

Childhood cancer is a life-threatening chronic illness. Untreated, it is often
fatal. Fifteen years ago, most children who were diagnosed with one of the various
cancers died within five years. The life-expectancy of children with cancer has
improved greatly in recent years. Although the prognosis for some types of cancer
is still very poor, for many other types, patients with proper treatment can live for
years. As cancer research has made surgery, radiation, and chemotherapy more
effective, it has become important to consider the long-term psychological impact
that the experience of pediatric cancer has on individuals (Koocher, O’Malley,

Gogan, & Foster, 1980). Although the prognosis has vastly improved in the last

2
decade, the hardships undergone by children with cancer are Still great. The

specific prognosis is dependent upon the particular type of disease and its stage of
development at diagnosis. The treatments the children need to save their lives are
often painful and the side effects can result in extreme discomfort and nausea. In
addition, children often face more visible side effects such as hair loss from
chemotherapy and the coarsening of facial features from steroid therapy. As
Goldberg (1974) has found, visible side effects of chronic illness can make children
feel out of place among their peers.

If one considers the extent of the difficulties faced by these children, one
might expect a large amount of research to be devoted to understanding these
children and their problems. Until recently, this has not been the case. It has been
suggested that chronically ill children as a class have been comparatively ignored
and that the vast majority of research monies are devoted to basic biomedical
research, while research into psychological issues related to childhood cancer and
other chronic illnesses has been less vigorously supported (Hobbs et al., 1984). A
recent review of the psychosocial literature dealing with childhood cancer has
suggested that this is changing and that scientific inquiry into this area is increasing
(Van Dongen-Melman & Sanders-Woudstra, 1986). One task for the social scientist
and the health care worker involved in such research is to attempt to discover
methods of not only prolonging life but of enhancing the quality of life to the
greatest possible degree.

Koocher (1986) has argued for the importance of improving psychological as

well as medical treatment of pediatric cancer. An important aspect of this

3

treatment, according to Koocher, involves the reduction of anxiety in the child and
the bolstering of the child’s feelings of self-control. This can be accomplished early
in treatment by encouraging the child’s "engagement in the process" through 1
discussions with the child. Such discussions would deal with topics such as the
child’s fears about his or her condition, apprehensions about physical changes
resulting from treatment, and how peers may react to the illness.

Unpleasant life events that cause disruption in living patterns are stressful to
individuals and families. Krantz, Grunberg, and Baum (1985) have described stress
as the mental condition of a person who finds their physical or psychological well-
being threatened. Since poor health by itself is associated with higher levels of
stress (McFarlane, Norman, Streiner, & Roy 1983), it is no surprise that childhood
cancer is also associated with stress (Koocher & Sallan, 1978; Lewis & LaBarbera,
1983) and with psychological problems including anxiety, depression, regression,
emotional withdrawal, and sleep disturbances (Derogatis, et al., 1983; Hobbs et al.,
1984, Van Dongen-Melman & Sanders-Woudstra, 1986). Koocher and Sallan (1978)
assert that the relationship between staff mental health professionals and pediatric
cancer patients and their families is crucial in order to help counter the stresses
produced by the hardships and uncertainties of treatment. It is necessary, therefore,
to improve our understanding of these children so that mental health professionals
will be able to perform this important task adequately.

Some researchers have examined the ways in which families of chronically ill
children cope with stress. In terms of types of coping strategies used, Powers, Dill,

Hauser, Noam, and Jacobson (1985) found that families with seriously ill

4

adolescents tended to respond to stress through the use of cognitive and affective
strategies rather than by attempting to modify the stressful situation. They argue
that this type of solution is probably appropriate given that the causes of the
family’s stress are not likely to respond to active efforts toward change (the specific
illnesses examined were psychiatric disturbance and diabetes). Finally, Powers et al.
found that the two types of illnesses evoked different types of responses from
families. The families of children with psychiatric disorders were more passive in
their response, while the families of diabetic children used more active strategies.

The results of the Powers et a1. study suggest several things. In response to
high levels of stress, the families of adolescents appear to expand their range of
resources in order to deal with their difficulties. Also, it appears that different types
of illnesses generate different types of responses from families. The specific
illnesses examined do not seem to provoke active attempts at situational change
from families. This is probably due to the lack of hope on the part of families that
the illness will ever be "behind" them. In the case of adolescents with cancer, the
life-threatening nature of the disease may result in a different constellation of ceping
strategies. In addition, there is a hope that the illness will be "cured" eventually.
Unfortunately, because it is impossible to truly know that the disease will not recur,
the concept of "cure" is problematic when one considers cancer (Mauer, 1987). In
spite of this difficulty, the existence of hope may result in different strengths and
different weaknesses in families of children with cancer.

As long-term mortality rates have improved, the question of the long-term

psychological impact of pediatric cancer has become increasingly important. At

5
present, there is little information about the ways in which pediatric cancer affects
long-term survivors (Mulhem, Wasserman, Friedman, & Fairclough, 1989), and few
studies have been done (Van Dongen-Melman & Sanders-Woudstra, 1986). The I
research that has been done has reported a variety of adjustment problems including
problems with social skills (Noll et al., 1988; Spirito et al., 1988), depression
(Lansky, List, & Ritter-Sterr, 1986), and school-related problems and somatic
complaints (Mulbern et al., 1989).

In order to enhance both the short and long-term quality of life for the
chronically ill child, physicians, nurses, and mental health professionals must
understand the many different factors involved in the lives of these children. One
of these considerations is that of control. Some degree of uncontrollability is
inevitable for any child with a chronic illness. This uncontrollability may relate to
the onset or course of the illness, the treatments, the clinic visits, and/or the
hospitalizations that are required (Spence, 1987).

Compared to healthy children, children with cancer find themselves much
more dependent upon others for care. They are also regularly forced to undergo
unpleasant and often painful treatment. Treatment and the reasons for a particular
medical procedure or drug can be difficult, if not impossible, for children and even
their parents to understand. This is particularly true when the child did not begin to
feel ill until the treatment started, as is the case with some forms of cancer. It has
been suggested that as a result of their life situation and its associated stress,

pediatric cancer patients may begin to feel that they lack control over events in their

6
lives in both the short term (Nannis, et al., 1982), and the long term (Koocher &
Sallan, 197 8).

A lack of perceived control over illness can, by itself, result in a more
negative evaluation of the disease. Meyerowitz, Williams, and Gessner (1984)
found that among healthy university undergraduates, when cancer was perceived to
be personally uncontrollable, it was seen more unfavorably than when it was
believed that personal control could be exerted. The level of control that physicians
were seen as having over the disease did not influence participants’ attitudes toward
cancer. In other words, the control that some powerful other might exert over the
disease did not seem to affect the negative attitudes engendered in the
undergraduates by a perceived lack of control on the part of the patient. Control
must reside with the patient.

Recognizing the importance of control for children with cancer, Worchel,
Copeland, and Barker (1987) examined the relationship between control-related
coping strategies and emotional adjustment among these children. They examined
four different types of control strategies: Behavioral, Decisional, Informational, and
Cognitive. Behavioral control referred to the number of different behaviors used in
assorted situations associated with medical treatments and non-medical activities.
Decisional control involved the perception of control over treatment (who makes the
decisions). Informational control pertained to question asking and the amount of
information desired. Cognitive control was concerned with frequency of thinking

about and talking about the illness and treatment.

7
Worchel et a1. (1987) found that the best predictor of nurses’ ratings of

positive adjustment was low use of behavioral control. The researchers suggest that
this finding indicates that individuals with a positive adjustment tend to use a few
successful coping strategies rather than trying and rejecting many ineffective
strategies. Worchel et a1. (1987) also found that higher levels of decisional control
were associated with better adjustment and fewer behavioral problems, whereas high
cognitive control predicted nurses’ ratings of passive non-compliance. Informational
control was not related to adjustment ratings.

This study demonstrates that a better understanding of the control beliefs of
chronically ill children and adolescents is needed. The work of Worchel et a1.
(1987) suggests that some types of control can be helpful whereas others appear to
do more harm than good. These different types of control need to be differentiated

and defined more clearly and examined in greater detail.

Locus of Control

Feelings of control are generally measured through the use of some Locus of
Control (LOC) scale. The concept of locus of control originated with Phares (1955)
and James (1957) and was further developed by Julian Rotter (1966). Locus of
control is a construct that refers to an the generalized expectancy that what happens
to oneself is the result of one’s own actions, or conversely is the result of external
factors. People with an internal LOC tend to believe that they are in control of

events in their lives, whereas those with an external LOC tend to believe that either

8

random chance or powerful others control outcomes. It is assumed that this
expectancy is learned through repeated encounters with the social environment. The
social environment is a complex one, and, as a result, issues of control can be
complex as well. Rotter and his colleagues originally attempted to create a complex
scale designed to measure control expectancies for several distinct domains in
addition to generalized LOC (Lefcourt, 1981). This attempt was unsuccessful.
Instead, Rotter developed his widely used one-dimensional I-E (Internal-External)
Scale.

A number of demographic factors have been shown to influence scores on
LOC measures. Using a measure of general locus of control, Stipek (1981) found
that over the course of first grade, children tended to become increasingly internal.
Stipek suggests that this may result from the opportunities that school provides for
children to observe the outcomes of their own behavior. Stipek also argued that
school teaches responsibility directly. This may also contribute to the increased
internality over the course of first grade. Stipek also found that middle
socioeconomic status (SES) children began first grade with a more internal LOC
than children with a lower SES but that this difference disappeared by the end of
first grade. Other researchers, however, have consistently reported differences
between middle SES and lower SES adults (see Lefcourt, 1976). Either this SES
difference is reestablished by adulthood, or as Stipek explains this discrepancy
across ages is due to the type of LOC measure that was used in these studies
(responsibility versus power). In other words, while there may be no difference

between SES groups in the extent to which they attribute positive outcomes to their

9
own actions, there may be a difference in the extent to which these groups feel they

can exercise influence in social situations.

Warm

Another demographic factor that has been found to influence LOC is family
size. Newhouse (1974) examined reinforcement-responsibility (a measure of LOC)
and found that only-born children tended to be less internal than children with
siblings. Newhouse’s findings also suggest that between grades four and six, the
tendency to accept responsibility for success declines. This finding seems to
contradict Stipek’s argument that school teaches responsibility, but it may just
indicate an increasing awareness of the social acceptability of humility or an
increasing belief in luck.

Sex differences have also been reported. Martin and Cowles (1983) found
that boys tend to be more internal than girls (all subjects between six and eight

years). They also found an interaction effect between sex, SES, and educational

program.

W

Since Rotter’s (1966) work, a number of specific DOC scales have been
devised for specific areas of control, in order to handle the complexities arising
from the diversity of human goals. Rather than attempting to create a single all-
encompassing scale, these scales have focused on restricted domains. Strickland

(1989) has remarked on the importance of the relationship between LOC beliefs and

10
health. For the purposes of this study, we are interested in this relationship and in

L.O.C. scales related to health beliefs.

Wallston and Wallston (1981) have devised a number of health-oriented LOC
scales. The first scale was the Health Locus of Control Scale (HLC) (W allston,
Wallston, Kaplan, & Maides, 1976). The HLC scale is a one dimensional scale
with the upper end designated "HEALTH-EXTERNAL", and the' lower end
designated "HEALTH-INTERNAL". With this scale, researchers have found that
chronic illness in general is related to external locus of control on HLC scales (see
Wallston and Wallston, 1981). In response to criticisms by Levenson (1973; 1974;
and 1975), who argued that health locus of control is not one dimensional, Wallston
et al. (197 8) eventually developed the Multidimensional Health Locus of Control
Scales. These scales include three dimensions: INTERNALITY, CHANCE
EXTERNALITY, AND POWERFUL OTHERS EXTERNALITY. This
conceptualization of locus of control is consistent with Levenson’s (1981)
formulation of the construct. The Wallston et al. (197 8) scales are reasonably
reliable (alphas range from .67 to .77), and have fairly low interscale correlations.
However, these scales were designed for use with adults and therefore may not be
useful for research with children.

Parcel and Meyer (1973) developed the Multidimensional Children’s Health
Locus of Control Scale (MCI-ILC). This scale is designed to assess children’s
perceptions of control over illness and health where high scores indicate a relatively
internal sense of control. Katz, Rubinstein, Hubert, and Bleu (1984) used this

measure in their study of psychosocial functioning in newly diagnosed pediatric

11
cancer patients. They found, among other things, that intemality was associated
with depression. The direction of the effect is unclear, however. Either depressed
children felt more responsible for their illness, or: those children who felt more
responsible for their illness were depressed.

Perrin and Shapiro (1985) examined health locus of control beliefs in healthy
children and in children with one of four chronic illnesses. The illnesses selected
were: asthma, diabetes, a seizure disorder, and an orth0pedic condition. The result
was that the control beliefs of asthmatic and diabetic children did not statistically
differ from those of healthy children. Children with a seizure disorder or an
orthopedic condition, however, had consistently lower Internality scores and higher
Powerful Others and Chance External scores as compared to healthy children.

Perrin and Shapiro’s findings suggest that there is something quite different
about the experience of these two sets of illnesses. The nature of this difference
needs further examination, but it may involve any of a variety of factors, including
differences in the biological nature of the illnesses, differences in medical and/or
psychological treatment, and differences in peer reactions to name but three

possibilities.

Engagement Style

A concept that on the surface may appear to be related to locus of control but

that arises from very different origins is that of Engagement Style (McKinney,

1978). This construct is measured through the use of McKinney’s Test of

12
Engagement Style (TES). The TES uses a semi-projective technique involving
several sets of cards depicting scenes of social interaction. Participants are asked
to describe the activity taking place, and their responses are scored as to whether
the "main character" is acting (Agent) or is being acted upon (Patient). Agent
responses are added with a maximum score of 24 (the number of scored cards).
High scores indicate "agency", and low scores indicate "patience". It is assumed
that the respondent identifies with the "main character". Where LOC depends upon
expected outcomes of action, engagement style focuses on whether or not the
individual perceives himself or herself as acting at all. In other words, LOC deals
with cause and effect, whereas engagement style refers to the social actor without
considering the social outcomes for that actor. McKinney (1981) has suggested that
Agency may be necessary but not sufficient for an internal expectancy. However,
no relationship between locus of control and engagement style has been established
empirically.

McKinney (197 8) has found that a number of variables are related to
engagement style. Family size is one such variable. He predicted and found that
children from large families generally have more extreme scores on the TES than
do children from smaller families. In other words, large families tend to produce
children who are either very agent or yery patient. McKinney also has found that
engagement style changes with age. For boys, agency declines between 2nd and 7th
grade, then increases between 7th and 12th grade. For girls, agency declines
throughout the school years (this decline being sharper before 7th grade than after

7th grade). McKinney (1980) has also found that family integration and maternal

l3
employment influence engagement style. Girls whose mothers work outside the

home are more agent than girls whose mothers are not employed outside the home.
Hatch (1979) has used the engagement style construct to explore the way in
which adolescents perceive the transition from living with parents to living on their
own. Hotch examined whether adolescents saw this transition as something they did
themselves, as something that was forced on them, or as a mutual decision
reflecting both agent and patient components. Hotch developed this domain-specific
conceptualization of engagement style and used it as a criterion measure against
which various predictors were compared. Hotch found that family variables
including relatedness and self-sufficiency predicted the manner in which home-
leaving was perceived. Adolescents who experienced a close relationship with their
parents had a more agent style of separation. Hotch argued that these adolescents
would regard the transition to independence as necessary but much less appealing
than would adolescents with a more distant parental relationship. As a result, those
individuals with a high level of relatedness saw themselves as having to be more
agent in the process of leaving home, whereas those with a low level of relatedness

saw the less pleasant home environment as a force pushing them out on their own.

Purpose

In order to treat pediatric cancer patients in the most effective manner, we
must understand how they view the world. The purpose of this research was to

begin to explore the way children with cancer view their social environment. Of

14
particular interest was whether children perceived themselves as acting or being
acted upon, and whether they see their health as being internally or externally

controlled.

Hypotheses and Predictions

The first hypothesis was that as a result of the treatment that they undergo,
pediatric cancer patients would tend to believe that their health depended on factors
beyond their control (see Wallston & Wallston, 1981). Therefore, it was predicted
that children with cancer will have higher extemality scores on the MCI-ILOC scale
than will children who do not have cancer. As noted earlier, this effect has been
found by Perrin and Shapiro (1985) in children with seizure disorders and
orthopedic conditions.

The second hypothesis was that pediatric cancer patients would feel that they
were acted upon by the world more than they were actors in the world because they
are acted upon so frequently during their treatment. Further, this perception of their
social environment would be generalized to other types of social situations.
Therefore, it was predicted that children with cancer will have a lower agency score
on the TES than will children who do not have cancer.

Finally, because of the stresses, cognitive growth, and emotional changes
associated with adolescence (for example puberty, and the greater need for peer
acceptance), it was predicted that the effect of the experience of cancer would be

different for adolescents than for pre-adolescents. Some studies have found that

15
children diagnosed and treated at younger ages are less likely to have adjustment
problems later in life than older children and adolescents (Koocher et al., 1980;
Lansky, List, & Ritter-Sterr, 1986). On the other hand, Allen and Zigler (1986)
report that the harmful effects of cancer on verbal intelligence scores were greater
for younger than for older children. Their argument as to why this is not surprising
involves issues of increasing autonomy and control. They suggest that older A
children are more likely to work on their problems on their own and that they are
more capable of doing so. It follows, that the effect of pediatric cancer might be
different for adolescents than for pre-adolescents. No prediction was made,

however, about the specific nature of this difference.

METHOD

Researchlatticinants

The participants were twenty-three matched pairs of children (n= 46). Fifteen
of the pairs were boys, 8 of the pairs were girls. The difference in the numbers of
each sex was due to random variation. Cancer patient participants in this study
were contacted through the pediatric oncology clinics of two large American
midwestem universities (Michigan State University and the University of Michigan).
Seventeen of the children were recruited from the clinic at Michigan State and 6
from the University of Michigan. Children with cancer involving the central
nervous system were excluded from the study because such children have been
found to have significant psychological problems of physiological origin associated
with the location of the tumor (Aram & Ekelman, 1986; Kun, Mulhem, & Crisco,
1983; Sollee & Kindlon, 1987), the magnitude of the lesion (Danoff, Cowchock,
Marquette, Mulgrew, & Kramer, 1982), and the use of cranial radiation (Duffner,
Cohen, Thomas, & Lansky, 1985; Duffner, Cohen, & Parker, 1988; Ellenberg,
McComb, Siegel, & Stowe, 1987; Kun et al. 1983). Clinic children between the
ages of 8 and 18 years who were attending school regularly and either were
receiving chemotherapy or had stopped treatment within the past 12 months were
contacted and asked to participate in the study. These children were selected
because their disease and treatment had stabilized enough for them to attend school

on a regular basis. This also meant that they were past the family upheaval and

16

l7

trauma associated with diagnosis and the rigorous initial regime of treatment and
were in the maintenance phase of treatment. Of these children, 10 had sustained
relapses, and 12 had been treated with cranial radiation. Of the families contacted,
92% (23 out of 25) agreed to participate.

As part of a larger study examining these children’s peer relationships (Noll,
Bukowski, Rogosch, LeRoy, & Kulkarni, in press), the children and their families
were asked for permission to contact the child’s school. With this permission, the
school principal was approached and the peer study was discussed. With the
principal’s permission, the teachers of each of the children in the study were
contacted and their aid was enlisted in conducting the peer study in their
classrooms. Consent forms were sent home with each of the children (n=575).
These consent forms made no mention of cancer. The children’s parents were asked
for permission to include their children in a study of children’s friendships. Ninety
percent of the parents agreed to permit their children to take part in the study
resulting in the participation of 515 children. This included all of the original clinic
children. The children with cancer were matched from among the other participants
with a child from their own classroom on the basis of sex, closest date of birth, and
race (racial match was based on observation). The parents of potential comparison
subjects were contacted by telephone and asked to participate in a Study examining
the effect of chronic illness on children. In the case of one child with cancer, the
school declined to participate. The control for this child was found by canvassing
in the general neighborth of the participant family to find an appropriate match.

Families were told that it was important to have information from families with

18
healthy children as well as those with chronically ill children in order to understand

how chronic illness affects children and families. The comparison children and their
parents were asked about the health of the child at this point to confirm that they
did not have a chronic illness before they were included in the study. Of

comparison families 82% of those contacted (23 out of 28) agreed to participate.

Iasmlmeuts

It was thought to be important that the measures to be used in the study have
validity across the entire participant age range in order to simplify comparisons
between ages. The LOC instrument that was used was the Multidimensional
Children’s Health Locus of Control (MCHLC)
scale (Parcel & Meyer, 1978) (See Appendix A). Because this scale was designed
for use with children, it was thought to be a better measure for the purposes of the
current research than health locus of conuol scales designed for use solely with
adults.

Since the Test of Engagement Style (McKinney, 1978) can be used over a
wide range of ages, it was well suited to the nwds of the current study. McKinney
(1981) reports that normative means on the TES range between 10.46 for 5th

graders and 11.5 for 12th graders.

Procedure
All data from the participants were collected in their homes by a researcher

from either Michigan State University or Wayne State University. Prior to data

19

Table l

 

 

W (15 boys. 3 girlS)
l6 Leukemia (ALL)' No Severe Chronic
4 Solid tumor Illness

3 Lymphoma

18 On therapy at data collection

5 Off therapy at data collection

AgLQLthsLSamnle
Range: 8-18 years 8-17 years
M: 12.3 years 12.5 years
Race
22 Caucasian 2 Caucasian
1 Black 1 Black
Wis
% married 70% 78%
% divorced/single 30% 22%
Duncan Socimgngmig Index
M: 39.00 42.57
512: 24.89 21.60

 

‘- acute lymphocytic leukemia

20

collection, the child and parents had an opportunity to ask questions about the
Study. The parents signed informed consent forms, which were also explained to
and signed by the child participant. The children were then seated in a quiet place
apart from the parents and were presented with the study instruments. At this time,
data for this study and a related research project were collected. While this was
occurring, the child’s parents were seated elsewhere in the home where they
completed another series of instruments including family demographic questionnaires.
After the data were collected the family was again allowed to ask questions and
were asked their reactions to the instruments. The entire procedure took
approximately one hour.

Children were given the following instructions for the test of Engagement

Style:

This is a set of pictures we’re using to see if everyone can tell what’s
happening in them. They’re all about one girl (boy) named Sally (Billy).
She’s the one with the bow in her hair (He’s the one with the cap on). What
I’d like you to do is tell me what’s happening in each picture. They’re all
about the same girl (boy) named Sally (Billy). Okay? Let’s begin.

(McKinney, 1978, p. 7-8)

Participants were then presented with two practice cards followed by six packets of
test cards. Their responses were recorded verbatim by the tester onto a coding

sheet. These response sentences were later scored for agency. As an example of

21

engagement style scoring, refer to figure 1. One child might describe the figure 1
stimulus as "Sally is giving the girl a horse-back ride." This response would be
scored as agent, because the main character Sally is the active participant. Another
child might respond with: "The girl is riding on Sally." Such a description would
be scored patient because Sally is the object of the other child’s activity. Another
possible response could be: "Sally is giving the girl a ride, while the girl is sitting
on Sally’s back." This response incorporates both agent and patient components and
would be scored 1/2 agent and 1/2 patient.

The MCHLC items were read aloud to each participant, and their responses
were marked by the experimenter. This was done to make the presentation of the
instrument as similar as possible for all participants across the range of ages (8 to
18 years) and the corresponding differences in reading skill. These data were

likewise coded later.

Analysis

The analysis of cross-group data was performed using a repeated measures
analysis of variance. The members of each matched pair were treated as a single
subject from whom data were collected on two occasions. This procedure took full
advantage of the pairwise matching process used to select the comparison group and

was used to reduce the random error in the scores of participants.

22

 

 

RESULTS
Locus of Control

The locus of control data were initially examined using a multidimensional
scoring system that produced three subscale scores: Intemality, Powerful Others
Externality, and Chance Extemality. This method is consistent with Levenson’s
(1981) position regarding the multidimensionality of locus of control and with Parcel
and Meyer’s (1978) discussion of the multiple dimensions that underlie the MCHLC
scale. It was thought that this procedure might reveal meaningful differences that
would otherwise be hidden by the more common unidimensional internal-external
scoring method.

Multidimensional methods of scoring the MCHLC have been used in previous
research (Perrin & Shapiro, 1985; Green & Kolff, 1980). These studies have found
low reliabilities for the subscales, however, and it has been suggested that the use
of LOC subscales with children may be inappropriate (Green & Kolff, 1980).

Given the short length of the MCHLC (20 items) and the correspondingly small size
of the subscales (6, 7, and 8 items), reliability problems are not surprising. In the
current study, the subscale reliabilities were variable. The reliability of the
Intemality subscale was particularly poor (coefficient alphas of .11 for children with
cancer and -.031 for comparison children). The reliability of the Chance Externality
subscale was somewhat better (coefficient alphas of .39 for children with cancer,

and .50 for comparison children). The reliability of the Powerful Others Externality

23

24
subscale, however, was quite good (coefficient alphas of .84 for children with cancer
and .88 for comparison children).

A preliminary MANOVA was carried out with the three LOC subscales to
look for overall group differences. The results of this analysis were marginal
(E(3,42) = 2.228, p < .099). A closer examination of the data seemed appropriate
given the exploratory nature of the study.

An AN OVA revealed no significant sex differences in either the cancer or
comparison groups, and a repeated measures ANOVA disclosed no significant
subscale differences between groups, although differences on the Intemality subscale
approached significance (E(l,22) = 3.43, p < .077). The unreliability of this
subscale makes interpretation of this finding problematic, but it appeared that
children with cancer (M = 6.7, SD = 0.82) might be more internal in locus of
control than children without cancer (M = 6.2, SD = 0.85). The three locus of
control subscale scores were then correlated with the age of the participants. This
correlation was carried out separately for the cancer and comparison groups to
assess the influence of age within each group. It revealed important differences
between the groups

Although scores on both the Intemality and the Powerful-Others External
subscales had similar age correlations for both groups, this was not the case for the
Chance—External subscale. Intemality scores were not significantly correlated with
age for either group of participants (Pearson correlations of -O. 18 for children with
cancer and -0.15 for comparison children). These correlations were not significantly

different from one another (2 score comparison).

Table 2

25

 

 

Children with cancer

AGE

INTERNAL

CHANCE
EXTERNAL

POWERFUL
OTHERS

AGE

-.18

-.23

-.58*

Children without cancer

AGE

INTERNAL

CHANCE
EXTERNAL

POWERFUL
OTHERS

AGE

-.77*

INTERNAL CHANCE POWERFUL
W QIIiERL

.26 .29 ---

W CHANCE POWERFUL
w MRS—

.18 .65* ---

 

* p < .05, two-tailed

26

Powerful-Others scores, on the other hand. were negatively correlated with
age for both groups (Pearson correlations of -.58 for children with cancer and -.77
for comparison children). Older subjects were less external. That is, they were less
likely to attribute health outcomes to powerful others. The correlations for the two
groups were, again, not significantly different from one another
(2 score comparison).

Scores on Chance-Externality were negatively correlated with age for the
comparison group (Pearson correlation of -.66) but were not significantly correlated
with age for children with cancer (Pearson correlation of -.23). For the comparison
group, older children were less likely than younger children to attribute health
outcomes to chance factors. Among children with cancer, however, the likelihood
of crediting health outcomes to chance was unrelated to the individual’s age. The
age correlations for the chance external subscale were significantly different between
the groups, z = 1.77, p < .05.

AS can be seen from Table 2, Chance External and Powerful Others scores
were significantly correlated with one another for the comparison group but not for
the children with cancer. This is consistent with the finding of varying age
conelations for these scores between groups.

Given the variable reliability of subscale scores, the data were further analyzed
using the more common unidimensional method, which produces a single total score
for each participant. Because the external items were reverse scored, high scores
indicated greater internality. This technique was found to be substantially more

reliable (coefficient alphas of 0.64 for children with cancer and 0.75 for comparison

27

children). These scores were then analyzed in the same fashion as the subscale
scores; an ANOVA was canied out to check for cross-group differences, and a
correlation with age was made for each group. This analysis again revealed no
significant differences in overall locus of control between groups. Locus of control
scores for both groups were significantly correlated with age. The Pearson
correlation for children with cancer was 0.50, whereas for the comparison children it
was 0.87. For both groups, older children were more Internal. Interestingly, these
correlations were significantly different from one another, 2 = 2.48, p < .01. This
difference may be due to the differences in age correlations found between groups

for Chance-External scores.

Test of Engagement Style

The Test of Engagement Style proved to be a highly reliable instrument for
the current sample (coefficient alphas of .89 for children with cancer and .76 for
comparison children). Contrary to predictions, however, the analysis of agency
scores revealed that children with cancer M = 11.24, SD = 5.69) showed
significantly higher levels of agency than did their matched controls M = 7.98,
SD = 3.62), E(1, 23) = 4.76, p < .05. Engagement Style scores were not
significantly correlated with age for either group (Pearson correlations of -0.21 for
children with cancer and 0.05 for comparison children). Likewise, ANOVAs

revealed no sex differences in engagement style for either group.

DISCUSSION

The results reveal a number of interesting differences between children with
cancer and children without cancer. Contrary to expectations, children with cancer
had a Want view of social relations than did comparison children. Crucial
questions are raised by this finding: Does Agency influence the individual’s chances
of overcoming illness? If so, how does it affect survival? Based on a review of
go LOC literature, Lefcourt (1982) has suggested that patients with an internal LOC
may tend to avoid behaviors that make their conditions worse. If we can generalize '
this suggestion to individuals with an agent engagement style, this would imply that
higher agency is a positive, adaptive reaction for children with cancer. One
possible reason for the finding lies in the way that pediatric cancer patients are
encouraged by treatment team members to take an active, responsible role in the
management of their disease. The trend in recent decades of actively involving
patients in their own health care (Maclntyre, 1977) may have produced higher levels
of Agency with respect to health as a result. This encouragement has perhaps made
these children more Agent in many aspects of their lives.

One of the two institutions involved in this research has a particular
commitment to encouraging the children to participate actively in their treatment.
From the point of diagnosis on, the family has access to and frequent contact with a

clinical psychologist and a clinical social worker who are integral members of the

28

29
oncology team. The children as well as their families are kept informed about all
relevant aspects of their treatment. This arrangement is highly unusual and places
this institution at the forefront of the movement to encourage patient participation.
Since this is the case, it raises the question: To what extent are these increased
agency findings due to the particular circumstances of the research sample? In
order to address this question, a post hoc analysis was conducted comparing clinic
participants from the two treatment institutions. The analysis revealed no significant
differences between the clinic samples.

The finding (approaching statistical significance) that children with cancer
scored higher on internal health locus of control than comparison children would
appear to be consistent with the encouragement explanation. The findings of N annis
et al. (1982) also support this conclusion. They found that, in cancer patients
between the age of 12 and 21, the cancer patient’s overall sense of control and
feelings of control over medical decisions are positively correlated with the patient’s
knowledge about their disease and treatment.

Unfortunately, the unreliability of the Intemality subscale makes these findings
suspect. The lack of differences in locus of control between groups on the overall
measure is consistent with the findings of Penin and Shapiro (1985). The findings
of the present study suggest two possible explanations. The first is that the
experiences of children with cancer are more similar to those of children with
diabetes or asthma than to those of children with a seizure disorder or an orthopedic
condition. This similarity could involve the children’s perception of the influence

they exert over their illness through treatment, daily routine, and attitude toward the

30

disease. The second possibility is that children with cancer belong to a distinct
third category of illness. Given that cancer is a highly life-tlueatening illness in
contrast to the illnesses studied by Penin and Shapiro, the latter explanation seems
more likely.

The current findings, albeit only approaching significance, are also consistent
with the findings of Steinhausen (1982) that children with certain chronic illnesses
tend to have a greater internal locus of control as assessed by generalized measures.
These illnesses included cystic fibrosis, Crohn’s disease, and ulcerative colitis. This
increased intemality was not found for children with asthma Steinhausen suggests
that a high magnitude of behavioral control is encouraged among individuals with
serious illnesses as the result of severe behavioral restrictions and intensive medical
treatment. If so, the higher Agency scores for children with cancer may also be a
result of this encouragement.

The conelations between age and locus of control are intriguing. It is not
surprising that, among children without cancer, older children are less likely than
younger children to believe that their health is due to the activities of powerful
others or "luck". It is commonly accepted that, in children, locus of control
becomes less external with age. In fact, the Nowicki-Strickland (1973) test is
specifically designed with the developmental progression from External to Internal in
mind.

Among children with cancer, older children are, like the control group, less
likely than younger children to attribute their health outcomes to the activities of

powerful others. What is fascinating is that these older children who have

31
experienced cancer are just as likely as younger children to attribute their health
outcomes to chance, whereas this does not appear to be the case among the healthy
children. This attribution may be a defense against self-blame. It is as though the
sick children are asking themselves, "What did I do to deserve this?", and
answering, "Nothing, because bad luck caused it." This could be very adaptive for
children with cancer and their families and would be consistent with the findings of
Katz et al. (1984) that among children with cancer, a positive relationship existed
between Intemath and depression. This response in conjunction with an agent
stance toward fighting the disease might prove a powerful force in resisting the
psychological stresses associated with the disease.

This finding also provides support for the validity of the chance external LOC
subscale. The scale scores indicate that adolescents with cancer consider the role of
luck in health outcomes to be important. Adolescents without cancer do not seem
to have as strong a belief in the influence of chance. This finding makes sense

and, therefore, strengthens the face validity of the chance scale.

CONCLUSIONS

The current research has generated a number of interesting findings. Children
with cancer were found to be more agent than children without cancer and possibly
to have a greater internal locus of control as well. As they grow older, children
with and children without cancer seem to lose their sense that powerful others
control health outcomes for them. The tendency of children without cancer to
attribute their health outcomes to chance also seems to diminish with age. Children
with cancer, however, do not seem to lose their belief in the influence of chance on -
health.

Issues of measurement have also been raised regarding the use of
multidimensional approaches to the study of children’s health LOC beliefs. Rather
than suggesting, however, that locus of control subscales are inappropriate for the
study of children, the age relationship findings provide support for the use of
multidimensional approaches for the study of children’s health LOC beliefs while
demonstrating the need for longer and more reliable subscales for use with children.
This methodological refinement is essential in order to accurately assess and
understand the components that make up beliefs about locus of control. In addition
to the factors of Intemality, Chance, and Powerful Others, the distinction between

competence and contingency must be addressed.

32

33

The findings discussed above have implications for understanding the coping
strategies used by children with cancer and their families. They indicate that
children and adolescents with cancer perceive themselves as acting on their
environment and possibly exerting control over their health outcomes to a greater
extent than do children and adolescents without cancer. They also appear to
maintain a belief in the influence of random events on their health. This apparent
inconsistency allows them to be active in fighting their illness, but at the same time
it frees them from the guilt and increased depression that might follow if they felt
that they were to blame for their illness.

The findings of this study suggest new directions for future research. It
appears that different categories of chronic illnesses may influence children’s locus
of control in different ways. How are these categories of illnesses different from
one another, and what do illnesses within these categories have in common? There
are a number of possibilities: there may be characteristic family coping styles
related to different kinds of illnesses; the psychological treatment that these groups
receive may differ across categories; or the courses of particular types of illnesses
may have unique psychosocial effects. Other questions need to be answered as
well. Does perceived control affect the types of coping strategies that children use?
How does this affect long-term survival? How do different psychological treatments
affect coping, control beliefs, and survival? Cromwell, Butterfield, Brayfield, and
Curry (1977) discovered a relationship between survival rates and treatment
strategies that accommodated individual LOC beliefs among cardiac patients. The

same may hold for children with cancer. A prospective study of these

34

relationships could provide mental health professionals with information crucial to
their work. One could begin to explore these questions by collecting information
about children’s control beliefs at the time that families are referred to the oncology
clinic and prior to diagnosis. By following up children positively diagnosed for
cancer and examining their control beliefs and coping strategies, the relationships
among these constructs might be better understood. Long-term survival rates of
participants could then be correlated with these data. The effect of different
psychological treatments could be explored either by random assignment (which is
ethically problematic) or by conducting this research at different institutions with
different psychological treatment approaches.

The answers to all of these questions could have an immense impact on the
treatment of chronically ill children. If different psychological treatments in
combination with different coping strategies and sets of control beliefs can be shown
to influence family functioning, child emotional development, and/or survival rates,

an important advance in clinic treatment will have been achieved.

APPENDIX A

35
APPENDIX A

MULTIDIMENSIONAL CHILDREN’S HEALTH LOCU S OF CONTROL
We would like to learn about different ways children look at their health. Here are
some statements about health or illness (sickness). Some of them you will think are
trueandsoyouwillsayYEStothem. Someyouwillthinkarenottrueandso
you will say NO. Even if it is very hard to decide, be sure to say YES or NO for
every statement. Nexer say both YES and NO for one statement. There are no
right or wrong answers. Be sure to answer the way you really feel and not the
way other people might feel.

PRACTICE: Try these statements.
a. Children can get sick. YES NO
b. Children never get sick. YES NO
Try one more statement for practice.
c. When I am not sick, I am healthy. YES NO

NOW DO THE REST OF THE STATEMENTS THE SAME WAY YOU
PRACTICED.

1. Good health comes from being lucky ............................................. YES N O
2. I can do things to keep from getting sick ...................................... YES NO
3. Bad luck makes people get sick. ..................................................... YES NO
4. I can only do what the doctor tells me to do ................................ YES NO
5. If I get sick, it is because getting sick just happens ..................... YES NO
6. People who never get sick are just plain lucky ............................. YES NO
7. My mother must tell me how to keep from getting sick .............. YES NO
8. Only a doctor or a nurse keeps me from getting sick .................. YES NO
9. When I am sick, I can do things to get better ............................... YES NO
10. If I get hurt it is because accidents just happen ............................ YES NO

11. I can do many things to fight illness .............................................. YES NO

12. Only the dentist can take care of my teeth .................................... YES
13. Other people must tell me how to stay healthy ............................. YES
14. I always go to the nurse right away if I get hurt at school ......... YES
15. The teacher must tell me how to keep from having accidents ,

at school ............................................................................................. YES
16. I can make many choices about my health .................................... YES
17. Other people must tell me what to do when I feel sick ............... YES
18. Whenever I feel sick I go to see the school nurse right away.... YES
19. There are things I can do to have healthy teeth ............................ YES
20. I can do many things to prevent accidents ..................................... YES ‘
Subscaleitems

36

Internal scale: 2, 9, ll, 14, l6, 18, 19, 20

Chance external: 1, 3, 5, 6, 10

Powerful Others: 4, 7, 8, 12, 13, 15, 17

N O
NO
NO

NO
NO
NO
NO
NO
NO

APPENDIX B

37
APPENDIX B

Clinic family contact

Hello, is this the home? Good. My name is Dan Mondoux,
and I’m a graduate student at Michigan State working with Bob Noll the
psychologist at the MSU pediatric oncology clinic. Do you remember his asking
you for your permission to release your family’s name and phone number to me for
possible participation in a research project? (good)

What we’re trying to do is to understand some of the effects that cancer has
on children and families. I’m calling now to try to find a time when I could come
out to visit your home. This visit will take about 1 hour of your family’s time.
Since we’re looking at the effect of cancer on the whole family, I would like to
find a time when I could meet with you, your flursbandbzife) and chiltL’sm.
Also, I would like to meet MW brothers and sisters. Does helshe have a
brother/sister over age 10?

NAME?
Alf?-
GIRL OR BOY?

Okay. I do this work at your: convenience in the evening or on the weekends.
I’m wondering when would be a good time for me to visit with your family?

Questions} we’re trying to learn about several things. We’re trying to find out
about what children think about themselves, in school, with friends, in sports and
games. We’re also trying to figure out if all the appointments and treatments and
stuff change children’s feelings of control over their life.

Also, we want to better understand the families of children with cancer and what
they’re like.

APPENDIX C

38
APPENDIX C

Comparison family contact

Hello, is this the house? Good. My name is and I’m a
graduate student at (W SU/MSU) and I’m doing some additional work on the
friendship study that your child participated in at his/her school. Do you remember
that project?

Good. (Explain briefly if not.)

 

I’m contacting you now about your possible participation in further research we’re
doing to learn more about children with serious/chronic illness and their families.
To do this we’re visiting some families with serious illness and some families with
healthy children. We have visited some families with illness in the
area and we hoped that your family could help us learn more about families with
healthy children who are in your area. Your help will take only about an hour of
your family’s time and would involve filling out some questionnaires. What I’d like
to do now is to arrange a time that I could come to your home, meet with you,
your wife/husband, and W to explain the study in more detail and
answer any questions that you have. When would be a good time for me to come
to your home?

mm It will involve filling out some questionnaires.
Wm Ask what it was that led them to say no.

 

APPENDIX D

39
MSU FRIENDSHIP STUDY

RESEARCH PARTICIPATION INFORMED CONSENT FORM

We have freely consented to take part in a research study being conducted jointly
by Michigan State University and Wayne State University.

The research study has been explained to us, and we understand that the purpose of
the research is to learn more about family relationships, and how parents view their
children’s behavior. We understand that the study is also looking at how children
view themselves, their relationships, and how they view health. We understand that
if we agree to participate, we and our child will be asked to fill out several
questionnaires. We have been told that the procedure will take approximately one
hour.

We understand that we are free to discontinue our participation in the Study at any
time without penalty; we are aware that lack of participation in the study will not
result in any penalty for us or our child.

We understand that the results of the study will be treated in strict confidence and
that we and our child will remain anonymous. Within the restrictions described
previously, we understand that general results of the research may appear in
professional journals and may be presented at scientific meetings. General results of
the study will be made available to us at our request.

We understand that our participation and our child’s participation in the study does
not guarantee any beneficial results to any member of our family.

The study has been explained to my child and she/he has assented (verbally agreed)
to participate in the Study.

In the event of any injury resulting from the research, no reimbursement,
compensation, or free mdical care is offered by Wayne State University.

Any questions I have about the project will be answered. (Sarah LeRoy: phone
351-1327 or Dan Mondoux: phone 353-3933)

I understand that I will receive a copy of this consent form.

Mother’s Signature Date
Father’s Signature Date
Child’s Signature Date Project Copy

Witness’s Signature Date Participants Copy

LIST OF REFERENCES

LIST OF REFERENCES

Allen, L, & Zigler, E. (1986). Psychological adjustment of seriously ill children.
- z. in , 25, 708-712.

 

Aram, D. M. & Ekelman, B. L. (1986). Cognitive profiles of children with early
onset of unilateral lesions. Wong, 2, 155-172.

Cromwell, R. L, Butterfield, E. C., Brayfield, F. M., & Curry, J. J. (1977). Acute
._ :_ _ _ 2.1.SaintLouis:TheC..VMosby

  

Danoff, B. F., Cowchock, F. S., Marquette, C., Mulgrew, L., & Kramer, S. (1982).
Assessment of the long-term effects of primary radiation therapy for brain tumors
in children. Cancer. 42. 1580-1586.

Derogatis, L. R., Morrow, G. R., Petting, J, Penman, D, Piasetsky, S, Schmale, A.
M., Henrichs, M., Camicke, C. (1983). The prevalence of psychiatric disorders

among cancer patients WWW £86).
751-783.

Duffner, P. K., Cohen, M. E., & Parker, M. S. (1988). Prospective intellectual
testing in children with brain tumors. MW 23, 575-579.

Duffner, P. K., Cohen, M. E. Thomas, P. R. & Lansky, S. B. (1985). The long-term
effects of cranial irradiation on the central nervous system. Cancer, 56, 1841-1846.

Ellenberg, L., McComb, J. G., Siegel, S. E. & Stowe, S. (1987). Factors affecting
the intellectual outcome in pediatric brain tumor patients. W, 21,
638- 644.

Green, K. E., & Kolff, C. (1980). Two promising measures of health education

program outcomes and asthmatic children. W. 5.0.
332-336.

Goldberg, R. T. (1974). Adjustment of children with visible and invisible handicaps:
congenital heart disease and facial burns. Whole“, 21(5),
428-432.

Hobbs, N., Perrin, J. M., Ireys, H. T., Moynihan, L. C., & Shayne, M. W. (1984).
Chronically ill children in America. Migration—literature, .45(7-8), 206-213.

40

 

41

Hotch, D. F. (1979). ' 2. 2. ' ‘ '
WW. Unpublished dOCtOral drssertatiOn, Michigan State
University, East Lansing.

 

James, W. H. (1957).

W’Unpubushed doctoral dissertation, Ohio State ‘
University, Columbus.

 

Katz, E. R., Rubinstein, C. C. Hubert, N. C., & Bleu, A. (1984, August).
1o 1.3» 11-111 1 1111.111-1 1 1.':11 '11. 10 W143.“

OWL Paper presented at the annual meeting of the American Psychological
Association, Toronto, Canada.

Koocher, G. P. (1986). Psychosocial issues during the acute treatment of pediatric
cancer. Cancer, 58, 468-472.

Koocher, G. P. & Sallan, S. E. (1978). Pediatric oncology. In P. R. Magrat (Ed),
= ‘ __ 1. . 3 (Vol. 1) (pp. 283- 307). Baltimore:

 

Baltimore University Press.

Koocher, G. P., O’ Malley, J. B., Gogan, J. L. & Foster, D. J. (1980). Psychological
adjustment among pediatric cancer survivors. MW

1283191113!!! 21. 163-173

Krantz, D. S. Grunberg, N. E, Baum, A. (1985). Health psychology. Amual
Wireless. 36. 349- 383.

Kun, L. B., Mulhern, R. K., & Crisco, J. J. (1983). Quality of life in children
treated for brain tumors: Intellectual, emotional, and academic function. Journal
QLNeumsurgenl. 58. 1-6.

Lansky, S. B, Vats, T., & Cairns, N. U. (1979). Refusal of treatment a new
dilemma for oncologiSts WNW
Hemamlogyflncolm 1(3). 277-282

Lansky, S. B., List, M. A., & Ritter-Sterr, C. (1986). Psychosocial consequences of
cure. Cancer, 58, 529-533.

Lefcourt, H. M. (1981). Overview. In H. M. Lefcourt (Ed), W411;
locumfmntmlmsmlct (Vol. 1, pp. 1-11). New York: Academic Press.

Lefcourt, H. M. (1982). W (2nd ed.). Hillsdale, New Jersey, Lawrence
Erlbaum Associates Inc.

42

Lewis, S., & LaBarbera, J. D. (1983). Terminating chemotherapy: Another stage in
coping with childhood leukemia. WW
Hemmloaflnmlogx. 5(1), 33- 37.

Lewnson, H. (197 3) Multidimensional locus of control in psychiatric patients.
IoumalmLConmmlLandflim'calEsxcholoax. _1_. 397-404.

Levenson, H. (1974). Activism and powerful others: Distinctions within the concept
of internal-external control W18. 377- 383

Levenson, H. (1975). Multidimensional locus of control in prison inmates. LQumaLQf
Applifimalhxchologx. 5. 342'347

Levenson, H. (1981) Intemality, powerful others, and chance. In H. M. Lefcourt

(Ed). Rmmhmmofmmoumct (Vol 1. PP 15- 63) New
York: Academic Press.

Maclntyre, A. (1977). Patients as agents. In S. F. Spicker & H. T. Englehardt, Jr.
‘ s ._ _ .. - .(p.p 197- 212).

 

Dordrecht, Holland. Raider.

Martin, S., & Cowles, M. (1983). Locus of control among children in various
educational environments. WM, 56, 831-834.

Mauer, A. M. (1987). The concept of cure in pediatric oncology. The American
loumalmtfiediauicflematologyflneology. 9. 58-61.

McFarlane, A. H., Norman, G. R., Streiner, D. L., & Roy, R. G. (1983). The
process of social stress: stable, reciprocal, and mediating relationships. W
HealrhmdjeciaLBehaxior. 24. 160-173

McKinncy. J P (1978) Engagememmmflalecticalmpectixe Paper
presented at the second international conference on mental health, Monterry,
Mexico.

McKinney, J. P. (1978). .. a:
(Manual). Unpublished manuscript.

 

McKinney, J. P. (1980). Engagement style (agent vs. patient) in childhood and
adolescence. WW1, 23, 192-209

McKinney, J. P. (1981) The construct of engagement style: theory and research. In

H M Lefcourt (Ed). W (V01 1.
pp. 359-383), New York: Academic Press.

43

Meyerowitz, B. F... Williams, J. G, & Gessner, J. (1984, August). EemequuLof
. = - . : : - .Poster presented at the annual meeting
of the American PsyChOIOgical Association, Toronto, Canada.

 

Mulhern, R. K., Wasserman, A. L., Friedman, A. G., & Fairclough, D. (1989).
Social competence and behavioral adjustment of children who are long-term
survivors of cancer. Pediatrics, 83(1), 18-25.

Nannis,E. D., Susman,E. J., Strope, B. E., Woodruff, P. J., Hersh, S. R, Levine,
A. S., & Pizzo, P. A. (1982). Correlates of control in pediatric cancer patients and

their families IoumalmLEediauiLflsychologx 1(1). 75- 84

Newhouse, R. C., (1974). Reinforcement-responsibility differences in birth order,

grade level, and sex of children in grades 4, 5, 6. W. 3_4_.
699-705.

Noll, R. B., Bukowski, W. M., Rogosch, F. A. LeRoy, S, & Kulkarni, R. (In
Press). Social interactions between children with cancer and their peers: Teacher

ratings. Wholesal-

Noll, R. B. LeRoy, S. Bukowski, W. M., Mondoux, D. A., Lindauer, H, &

Rogosch, F. A. (1988, August). WW”
friendships. Paper presented at the annual meeting of the American Psychological

Association, Atlanta, Georgia.

Nowicki, S. & Strickland, B. R. (1973). A locus of control scale for children.
MW. 411. 148-154.

Parcel, G. S, & Meyer, M. P. (1978). Development of an instrument to measure
children’s health locus of control. WM 6, 149-151.

Perrin, E. C., & Shapiro, E. (1985). Health locus of control beliefs of healthy
children, children with a chronic physical illness, and their mothers. W
Metrics. .191. 627-633.

Phares,,E...J(l955) .1141" ' ‘ ‘ .=.1. ' ° ' .
Unpublished doctoral dissertation, Ohio State UniverSity, COlumbus.

 

Powers, 8. I., Dill, D., Hauser, S. T., Noam, G. G., & Jacobson, A. A. (1985).

COping strategies of seriously ill adolescents. MW, 5(1),
101-113.

Rotter, J. B. (1966). Generalized expectancies for internal vs. external control of
reinforcement. MW 89. whole.

44

Sollee, N. D. & Kindlon, D. J. (1987). Lateralized brain injury and behavior
Problems in children mm. 1.5. 479-490

Spence, L. (1987). Comprehensive examination. Unpublished manuscript. Michigan
State University, East Lansing.

Spirito, A, Stark, L. J, Williams, C, Hewett, K., Drigan, R., & Andrakites, A.
(1988, April). " ‘ . - . =

 

successfullmatedmhildhoodflnmems Paper presented at the meeting of
the Society of Behavioral Medicine, Boston, Massachusetts.

Steinhausen, H. (1982). Locus of control among psychosomatically and chronically

ill children and adolescents. MW. _Q(4).
609- 616.

Stipek, D. (1981). Social-motivational development in first grade. Contempgm
Educationallmholou. 6, 33-45.

Strickland, B. R. (1989). Internal-external control expectancies: From contingency to

creativity Amencarlliychologist. 44(1). 1- 12

Thompson, RI. (1986). Behavior problems in children with developmental and
learning disabilities. - - .- __ . l. . _. '

Hamster. 3. Whole.

Van Dongen-Melman, J. E. W. M., & Sanders-Woudstra, J. A. R. (1986).
Psychosocial aspects of childhood cancer: A review of the literature. 1611mm

SEW. 21(2). 145- 180

Wallston, K. A., & Wallston, B. S. (1981). Health locus of control scales. In H.

M Lefcourt (134.) W (V01 1.
pp. 189-243). New York: Academic Press.

 

Wallston, B. S., Wallston, K. A. Kaplan, G. D., & Maides, S. A. (1976).
Development and validation of the health locus of control (HLC) scale. .[QumaLof

WW2! 44 530-535

Wallston, K. A., Wallston, B. S., & DeVellis, R. (197 8) Development of the
multidimensional health locus of control (MI-TLC) scales. Healthjducation

Mimoaraphi. 6. 161- 170

Worchel, F. F., Copeland, D. R., & Barker, D. G. (1987). Control related coping

strategies in pediatric oncology patients. WW, _1_2(l),
25- 38.

‘o