LIBRARY
Michigan State
L University

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

DATE DUE DATE DUE DATE DUE

iaélfl'ms, fl . ll

l fl_

 

 

 

 

 

 

 

 

 

 

 

 

JD

 

 

 

 

 

 

 

ll
D1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

if—

MSU I. An Affirmative Action/Equal Opportunity institution
cMMpma-M

 

 

 

 

 

 

 

 

 

THE RELATIONSHIP BETWEEN PARENTAL PERCEPTION
OF CHILD BEHAVIOR AND
PARENTAL CHILDHOOD MEMORIES:
A COMPARISON OF MOTHERS OF CONDUCT-DISTURBED
AND NON-CONDUCT-DISTURBED BOYS
By

Jerry Stuart Adams

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1992

ABSTRACT
THE RELATIONSHIP BETWEEN PARENTAL PERCEPTION
OF CHILD BEHAVIOR AND
PARENTAL CHILDHOOD MEMORIES:
A COMPARISON OF MOTHERS OF CONDUCT-DISTURBED
AND NON-CONDUCT-DISTURBED BOYS
By

Jerry Stuart Adams

The purpose of this study was to establish a connection between a mother’s childhood
experiences and her current parenting. The study proposed that the mother’s
perceptual style is the result of an affective-cognitive model regarding parent-child
interactions and this mental model is based largely on the mother’s internal
representation of her own childhood relationships. The participants were 71 mothers
of conduct-disturbed boys (MCDB) and 62 mothers of normal boys (MNB). The
mothers completed a) the Child Behavior Scenes, an instrument designed to assess
their deviancy-classification schema (i.e., the tracking and labelling of positive,
negative and neutral child behaviors), b) the Child Behavior Rating Scales, designed
to obtain the mothers’ judgments regarding the degree of aversiveness and the degree
of desirability of specific child behaviors, and c) the Adult Attachment Questionnaire

(AAQ), a questionnaire format of the Berkeley Adult Attachment Interview (George,

Kaplan & Main (1985). The results indicated that the MCDB are overly inclusive in

the classification of deviancy and are more "sensitized" and "reactive" to child

misbehavior. The were no differences in terms of the classification of prosocial

behaviors. While the two groups of mothers did not differ in the content of their
childhood memories, there were significant differences in the organizational qualities
of these recollections. The MCDB childhood memories were significantly more
fragmented and disorganized, and lacking in specific detail. The MCDB were
significantly more "unresolved" and less individuated from their childhood
relationships. Overall, the psychopathology of the child was related to the mother’s
affective-cognitive processing, synthesis, and integration of her own childhood
experiences but not the actual experiences in—and-of-themselves. While the
organization of the childhood memories did not account for the MCDB vigilance
towards deviancy, there were some significant relationships between the organization
of the mother’s recollections and her current perceptual style. The study offers
support for the construct of an internal mental model bridging past experiences with

current functioning. Also, the AAQ provided a good alternative to the full-length

interview.

ACKNOWLEDGEMENTS

I would like to thank Dr. Robert Caldwell, chairperson, Dr. Hiram Fitzgerald,
Dr. Albert Aniskiewicz and Dr. Ellen Strommen for serving as members of my
dissertation committee.

I would like to express my special appreciation to Dr. Caldwell for his
conscientious guidance, support, and valuable suggestions. I am very thankful to Dr.
Caldwell for his patient and continued encouragement of my own autonomy and
competency. As a result of his conduct and persistent respect for my integrity, the
completion of this research resulted in fostering my own growth and development
intellectually, professionally and emotionally.

I would like to acknowledge the cooperation and commitment of the
participating mothers and the cooperation of the participating school districts. I would
like to thank the raters for their input and effective scoring of the AAQ, and the
clinicians that helped in the development of the Child Behavior Scenes. Relatedly,
an important thank you to Dr. Mary Main for granting permission to use the Adult
Classification System. I would like to express my gratitude to Suzy Pavick for her
expert assistance in getting the final draft of this manuscript printed and to the

graduate school. A special thank you to Jim McComb for his support and assistance

with the statistics.

iv

 

t.p

Finally, I am very appreciative of the endless support and encouragement
offered by my friends, especially Chet, Dori and Bella. A special thank you to Ann
Wagner for her fi'iendship, kindness, and the many hours of phone consultations, and
to Ellen Ives, a cherished friend, a continual source of encouragement and motivation,

and a very thorough reader of the final draft. A very special heartfelt thank you to
Dr. Sam Plyler.

LE F NTENT

PAGE
CHAPTER 1--INTRODUCTION .................... 1
Conduct Disorders ........................ 7
Classification ........................ 8
Coercion Theory ..................... 12
Irnpressionistic Overview ................ 22
Discipline and a Review of the Empirical
Literature on Family Interactions .......... 24
Parental Cognitive Variables .................. 39
Parental Perception and Labelling of
Child Behavior ..................... 43
Parental Categorization of Deviant,
Neutral and Prosocial Behaviors ........... 47
The Perspective of the Present Study ............. 53
Attachment Theory .................... 60
Parental Recollections .................. 67
Summary .......................... 73
Hypotheses ............................. 75

Group I - Maternal Perception of Child Behavior . 76
Group II - Maternal Childhood Recollections . . . . 79
Group III - Relationship Between Maternal

Perception and Childhood Recollections ...... 81
CHAPTER 2-—METHODS ........................ 83
Participants ............................. 83
Mothers of Conduct Disturbed Boys ......... 84
Mothers of Normal Boys ................ 85
Demographic Variables ................. 88
Participant Selection ................... 90
Instruments ............................. 91
Child Behavior Checklist ................ 92
Family Information Sheet ................ 93
Family Togethemess Scale ............... 93

Child Behavior Scenes .................. 94

Child Behavior Rating Scales ............. 98
Preliminary Questions ................. 100

Adult Attachment Questionnaire ........... 101

vi

Training ......................... 111
Determination of Inter-Rater Reliability ...... 112
Procedure ............................. 114
CHAPTER 3--RESULTS ........................ 118
Group I - Maternal Perception of Child Behavior . . . . 118
Group II - Maternal Childhood Recollections ....... 128
Group III - Relationship Between Maternal
Perception and Childhood Recollections ..... 133
Sub-Sample Analysis ...................... 135
CHAPTER 4--DISCUSSION ..................... 151
Maternal Perception ....................... 153
Maternal Recollections ..................... 171
Maternal Recollections and Perceptual Style ........ 185
Summary and Implications .................. 200
APPENDD( A - Letter Distributed to Mothers ........... 205
APPENDIX B - Phone Interview and Fact Sheet .......... 206
APPENDDI C - Consent Letter .................... 208
APPENDIX D - Feedback to Participants .............. 211
APPENDDI E - Family Information Sheet ............. 212
APPENDIX F - Family Togethemess Questionnaire ....... 213

APPENDDI G - Instructions for the Child Behavior Scenes . . . 214

APPENDIX H - The Experts’ Final Determination of the Positive
(P), Negative (N), or Neutral (Blank) Value for Each

Behavior in the Child Behavior Scenes ........... 216
APPENDIX I - Child Behavior Rating Scales I/II ......... 228
APPENDIX J - Adult Attachment Questionnaire .......... 240
References ................................. 257

vii

9.

TABLES

. Principal Characteristics of Unsocialized

Aggressive and Socialized Aggressive Conduct
Disorder as Derived from Multivariate Statistical

Studies .............................

Child Behavior Checklist Alphas,

Group Means, and Group Comparisons ........

. The List of Behaviors That Are Represented
on the Extemalizing Scales of the CBCL ........

. Demographic Variables, Group Means
and Group Comparisons ..................

. Percentage of Agreement Between Raters
and for Scales During Rater Training ..........

The Percentage of Inter-Rater Agreement-Within-One

for Raters and Scales ....................

Hypothesis 1: Are MCDB Overly Inclusive in the
Categorizationof Deviancy and Overly Exclusive

in Their Categorization of Prosocial Behavior? . . . .

Hypothesis 2: Are MNB Overly Inclusive in the
Categorizationof Prosocial Behaviors and Exclusive

in Their Categorization of Deviancy? ..........

Intercorrelations Between the Perceptual Variables . . . .

10. Relationship Between Enmeshment and the

Mother’s Deviancy - Classification Schema .......

11. Comparisons Between the MCDB and the MNB

on the Content of Their Childhood Recollections . . .

12. Comparisons Between the MCDB and the MNB

on the Affective-Cognitive Organizational Properties

of Their Childhood Recollections .............

viii

PAGE

...86

...87

...89

.. 125

.. 126

.. 130

13.

14.

15.

16.

17.

18.

19.

20.

21

22.

Intercorrelations Between the Scales of the AAQ .......

Relationship Between the Affective-Cognitive
Organization of Childhood Memories and the Mothers’

Deviancy Categorization Processes ..............

Comparison of the Deviancy Categorization

Process Between MCDB and MNB in the Sub-Sample . .

Intercorrelations Between the Perceptual Variables

for the Sub-Sample ........................

Comparisons for the Sub-Sample Between
the MCDB and the MNB on the Content
of Their Childhood Recollections

Comparisons Between the Sub-Sample of MCDB and
the MNB on the Affective-Cognitive Organizational

Properties of Their Childhood Recollections ........

Relationship Between the Affective-Cognitive Organization
of Childhood Memories and the Mother’s Deviancy

Categorization Processes in the Sub-Sample ........

Partial Correlation Coefficients Between the
Perceptual Variables of the CBS With the Effects of

Responsivity Held Constant ..................

. Intercorrelations Between the Childhood

Experiences with the Mother and the Affective-

Cognitive Organizational Scales of the AAQ ........

Intercorrelations Between the Childhood
Experiences with the Father and the Affective-

132

134

140

144

147

179

Cognitive Organizational Scales of the AAQ ........ 180

ix

CHAPTER I

INTRODUCTION

It is a generally held belief in our culture that parents’ childrearing practices
will have a direct effect on their offsprings’ development. New parents frequently
refer to books like Dr. Spock’s for advice and suggestions regarding the correct
childrearing strategy, as if this will help to insure a "positive" outcome. This
generally held belief also has a tradition within the social sciences and is exemplified
by Baumrind’s (1971) now classic study which demonstrates that specific parental
attitudes and behaviors can have differential effects, either salutary or detrimental, on
the child’s emotional and social development. Currently, however, there is general
agreement that the parental influences are not unidirectional, but rather reciprocal
with the child also having an influence on the parent (Bell, 1979). Within this more
SOphisticated conceptual schema there is a whole array of potentially Significant family
interactions, including not only parent to child and child to parent, but also marital
interactions and those between siblings. Clearly, most contemporary theories on
child/adolescent development now emphasize these family interactions as important
factors contributing to the development of personality and psychopathology.

This emphasis on family interactions is exemplified in the research literature

on childrens’ conduct disorders. The association of specific family variables (e.g.,

1

"an5
ti... .,|

.-,-.
11 ‘

43., .:

'hv- ..

if
.
fi“.

y

Par-s.
‘

....

 

 

. \
W.
.
do "'
.‘_.}:
\.
P's
‘\ T;
"K
\
p.‘
e 1.;
I'\ P
r.
.. ‘
~."|1
5“
c
“I
0v
3‘

2

family problem solving, discipline and/or parental monitoring) with the occurrence
of a conduct disorder is now a well established fact consistently supported by recent
comprehensive reviews of the literature (Hetherington & Martin, 1986; Patterson,
1982; Rutter & Garmezy, 1983; Rutter & Giller, 1983; Snyder & Patterson, 1987).

One of the key factors that distinguishes families of children with a conduct
disorder from "normal" families appears to be differences in parental styles or
characteristics (i.e. , the differing ways in which a parent engages in or performs their
parental duties/ responsibilities). ”There is abundant evidence to support a growing
consensus that antisocial children have parents who lack family management skills"
(Patterson, 1986, p.432). Thus, there are certain apparent deficiencies in the
parenting style (e. g. , inept discipline practices) that are associated with the occurrence
of a conduct disorder.

It is essential to point out, however, that the behavioral family therapy
literature has been criticized as over-emphasizing the position that the basic processes
contributing to the child’s conduct disturbance is a parenting skills deficit (Griest &
Wells, 1983). Griest and Wells (1983) stated that only recently has the ”treatment
vehicle” (i.e. , the parent) been identified as potentially requiring intervention in areas
other than family management skills. "In this expanded model it is hypothesized that
the parent’s cognitive, psychological, marital and/or social adjustment is directly or

indirectly related to conduct disorders in children" (p.38).
Whether the focus is on inept family management and/or parental cognitive

factors, the fact that there are a variety of family variables and specifically parental

.‘up A

n‘. t."

.
”t; a

in...

Liz-.2.

“’30"

.L. g

32‘
.I .
c’i-

3
characteristics that differentiate parents with normal children from the parents of

children with a conduct disorder raises two interesting and related issues. First is the
issue that Rutter and Giller (1983) referred to as "cause correlate or consequence? "
This question raises interpretation issues regarding causality and the meaning of the
associations between parental/family variables and disturbances of conduct. Is the
parents’ behavior a direct cause for the conduct disturbance? Are the associations
caused by some other third factor, for example, genetic, neighborhood, or social
class? Or, as suggested by Patterson (1986b), is the parental behavior a response or
consequence of the child’s deviancy (a notion supported by recent research, e. g.,
Anderson, Lytton & Romney, 1986)? These are difficult questions to resolve and, as
Rutter and Giller (1983) pointed out, it is not always possible to determine causality
or its direction. If, however, a relevant parental factor can be shown to have
anmdated the child’s disturbance, then this supports an interpretation regarding
causality. Overall, Rutter and Giller (1983) concluded that ”it seems likely that the
causal influences are bi-directional but that, at least with some of the family variables,
the preponderant effect is likely to be from parent to child" (p. 190).

Relatedly, the second issue is raised within a broader context by Belsky
(1984):

By tradition, students of socialization have directed their primary energies

toward understanding processes whereby parents’ childrearing strategies and

behaviors shape and influence their offsprings’ development. It is of interest

to learn that, while great effort has been expended studying the characteristics

.I as
5 Jul
-.....,,

‘tée.

Finn.

~..._

 

 

.\

\

 

 

\
,
K
\l'l-
\‘U
‘

4

and consequences of parenting, much less attention has been devoted to

studying why parents parent the way they do--beyond, of course, social-class

and cross-cultural comparisons and investigations examining the effect of the

child on parenting behavior. (p.83)

In other words, why do parents of children with a conduct disorder have
characteristics that differentiate them from " normal " parents? What are the
determinants of the parental functioning or, more specifically the parental
dysfunction?

The present study is designed to shed some light on these important questions.
This study will examine the relationship between parental cognitive variables and the
child’s conduct disturbance. As the literature indicates, the parents of conduct
disordered children do exhibit a variety of deficiencies in family management,
especially discipline. These deficiencies are consistently associated with the child’s
disturbed behavior. The present Study will expand this notion of inept family
management, as suggested by Griest and Wells ( 1983), and examine whether there
are underlying parental cognitive factors that account, in part, for the child’s
mismanagement.

This study is interested in the parental cognitive factors that may determine the
parent’s perception of their child’s behavior. It is only common sense that the ways
in which the parent perceives the child’s behavior will color or bias a wide variety
0f parent-child interactions. This is certainly true in regards to disciplinary

interactions. The ways in which a parent defines the child’s behavior will largely

I11.

S-J

LI

tl'

5

determine the parental response. The parent’s classification of a behavior as deviant,
as opposed to neutral or prosocial, is a critical initial step in the disciplinary process.

It is the position of the present study that parents may contribute directly to the
development of antisocial behavior by failing to adequately and consistently label,
track, and ”consequate" deviant behaviors, as well as by not responding positively to
prosocial behaviors. The present study will examine the parents’ deviancy-
categorization processes (Patterson, 1986b). Hence, this study is designed to examine
whether the parents of conduct disturbed children exhibit a different perceptual style,
or rather, a different cognitive schema for categorizing behaviors as deviant, neutral,
and/or prosocial, as compared to the parents of normal children.

Additionally, this study will examine why it is that these parents behave as
they do. In other words, what are the possible determinants for the parents’
perceptual style? Belsky (1984) provides a model for examining the determinants of
parenting and provides fOr three general sources of influence on parental functioning:
1) the parent’s own developmental history and psychological resources; 2) the child’s
characteristics and 3) the contextual sources of stress and support. The present study
is concerned with the first factor, and will assess whether the quality of the parents’
recollections regarding their own childhood is related to both their current perceptual
style and the psychopathology of their children.

The position of the present study is that the parents’ perceptual style is the
result of an affective-cognitive model regarding parent-child interactions. This mental

(or representational) model is based largely on the parent’s own childhood

6

relationships, and therefore, the ways in which the parents recollect their own
childhood experiences should exhibit the structural/organizational properties of this
mental model. It is proposed that parents of conduct disturbed children have a mental
model that is defensive in its orientation (i.e. , it is organized to distort and exclude
certain classes of information). The present study proposes that this defensively
oriented affective-cognitive model will be related to the parents’ perceptual style, as
evidenced by their deviancy-categorization schema.

The review of the literature is organized into three main sections. The first
section on Conduct Disorders is designed to provide an understanding of the family
characteristics and parental variables that differentiate families with a conduct
disturbed child from normal families. The empirical literature is reviewed with a
focus on those component behaviors that are likely to be involved in disciplinary
interactions. The second section on Parental Cognitive Factors will suggest that these
observed differences in family interactions can, in part, be accounted for by
differences in parental cognitive processes. The research on both the parental
perception of child behavior and the parents’ deviancy-categorization schema will be
reviewed. The final section, Perspective of the Present Study, summarizes the
previous two sections and develops the conceptual foundation/position for this study.
Specifically, this section offers a detailed accounting to the following question: How

is it that the quality of the parents’ recollections regarding their own childhood can

be related to current caregiving behaviors?

 

1.5:" .
, A
...uu. n;

e .
W I
‘ .1

y
fi“:~e..
"We“ ‘

 

1:11“
I
i

 

-Q:€T

I

 

‘.
\.

\I ‘h.

O ‘-1.

. .
.\.
\ .

a up

WIS

This section will review some of the current thinking and research on Conduct
Disorders. The focus will be on the variety of parental characteristics and family
interaction patterns that characterize families with conduct disordered children. There
will be an emphasis on those parent-child interactions that are involved in the
effective implementation of discipline.

This section will be divided into four sub-sections:

1. Initially, it will be helpful to briefly review the literature on the
classification of Conduct Disorders and provide some sense of the scope of the
behavioral disturbance that defines this diagnostic category.

2. It will also be useful to provide a theoretical framework for understanding
the family processes involved in the development and maintenance of a conduct
disturbance. This conceptual framework will be provided by reviewing the extensive
work on Coercion Theory (Patterson, 1982).

3. An impressionistic overview of the research findings on the relationship
between family factors and conduct disorders will be presented.

4. This final sub-section commences by presenting a conceptual understanding
of discipline, including those component behaviors involved in the parents’
disciplinary practices. With an emphasis on these component behaviors, a review of
the empirical literature is undertaken, examining the behavioral differences between
family interactions in families with a conduct disturbed child as compared to "normal"

families. This sub-section will conclude with a review of the research specifically

8
focused on the relationship between parental " family management skills" , especially

discipline, and conduct disturbance.

Classification. Disturbances of conduct represent a major dimension of
child/adolescent psychopathology. Rutter and Garrnezy (1983) reported that general
population studies and studies of clinic samples "have been consistent in showing that
disturbances of conduct (as indicated by fighting, temper tantrums, defiance,
destructiveness, and generally uncooperative or inconsiderate behavior) constitute one
of the most frequent causes of concern regarding children’s behavior" (p. 814).
Factor analytic studies (as reviewed in Achenbach & Edelbrock, 1978; Quay, 1979,
1986a) have consistently shown that disturbances of conduct constitute one of the
main dimensions of behavior during childhood and adolescence. In fact, Quay
(1986b) reported that conduct disorders are "the most prevalent" form of childhood
and adolescent psychopathology. "No other disorder of childhood and adolescence
is so widespread and disruptive of the lives of those who suffer it and of the lives of
others, and thus deserves continued research investigation" (p.64).

There does appear to be general agreement that disturbances of conduct are
heterogeneous. While clinical classification systems have a variety of sub-divisions,
for example DSM-III has four subtypes of conduct disorders, factor analytic studies
(e.g., Quay 1979, 1986a) have suggested that the evidence is very strong for the
existence of two forms of conduct disorders; namely, an Unsocialized Aggressive and
a Socialized Aggressive pattern. The results of Quay’s (1986a) multivariate statistical

analysis of 61 studies in which descriptors of behavior have been analyzed and factor

 

a
i-“

 

Table 1

PM u. -11.. .' 'I f n 0 tr A" ‘ iv -. (13 i ' ed Are Sive
r riv fr m M l'v ' 'tisi ie "

Wm ' ' A 8 iv

Fighting, hitting,assaultive (42)" Has "bad" companions (10)

Disobedient, defiant (40) Truant from school (8)

Temper tantrums (39) Truant from home (6)

Destructiveness (35) Steals in company with others (6)

Irnpertinent, "smart", impudent (31) Belongs to a gang (6)

Uncooperative, resistant, Is loyal to delinquent friends (5)

inconsiderate, stubborn (28) Stays out late at night (5)

Attention seeking, "show-off" (26) Lies, Cheats (2)

Dominates, bullies,threatens (26) Steals at home (2)

Disruptive, interrupts, disturbs others (26)
Boisterous, noisy (25)

Irritability, "blows up" easily (23)
Negative, refuses directions (22)

Restless (21)

Untrustworthy, dishonest, lies (20)
Hyperactivity (18)

 

'The analysis is based on 61 studies spanning almost 40 years in which descriptors of
behavior have been analyzed and factor loadings reported. (Quay, 1986a, p.12)

l"Total number of studies in which characteristic was found associated with this
dimension is in parentheses.

10
loadings reported are presented in Table l. The dimension labeled Unsocialized

Aggressive Conduct Disorder is most frequently associated with behaviors that are
generally considered aggressive, disruptive, and noncompliant. In the Socialized
Aggressive Conduct Disorder, involvement with peers in illegal or at least norm-
violating behavior is central.

Further support for this position is provided by Loeber and Schmaling (1985b).
They attempted to identify patterns of antisocial child behavior, reviewing 28 factor
and cluster analyses of child behavior assessment instruments covering over 10,000
children. They concluded that antisocial behavior can be conceptualized as
unidimensional and bipolar. One end of the dimension was characterized by overt or
confrontive behaviors, such as fighting and other forms of aggressiveness. The other
pole was characterized by covert or clandestine antisocial behaviors, such as alcohol
and/or drug use, truancy, and theft.

While there is statistical support for distinguishing between these two patterns
of conduct disorders, Rutter and Gannezy (1983) suggested that the major justification
for the differentiation between unsocialized/overt and socialized/covert patterns of
deviant behavior comes from evidence showing that they differ in terms of family
background and of outcome. Patterson’s (1982) distinction between "social
aggressors" and "stealers" is conceptually consistent, as he provided preliminary
evidence which suggests that there are different developmental and family processes
that underlie the emergence of overt/confrontive antisocial behavior compared to

covert/concealed antisocial behavior. Relatedly, Loeber and Schmaling’s (1985a)

J.
b\§

1;:

5
pl

'6

“H

l..‘

ll

findings offer further support for the utility of differentiating between an overt and
a covert classification which is based on a typology of antisocial behavior, as the data
does reveal some significant differences in terms of family processes, specifically
parental characteristics.

It is essential that research on conduct disorders take into account these
potentially distinct patterns of antisocial behavior and clearly specify the relevant
dimensions of the population under study. Not only will this aid in progressing the
state of the art in terms of reliability, generalization, and replicability of results, but
also, if covert and overt antisocial behavior do represent developmentally distinct
patterns, then greater knowledge of these patterns will facilitate prediction,
prevention, and treatment (Loeber & Schmaling, 1985b).

The present study is focused on the family processes involved in the
Unsocialized Aggressive Conduct Disorder, referred to by Patterson (1982) as "overt
antisocial" or "Social Aggressors". Families of "overt antisocial" children have been
characterized as "enmeshed" , with the parents exhibiting an overinclusiveness in their
definition of problematic behaviors (Snyder & Patterson, 1987). Thus, the
interactions that characterize these families easily lend themselves to specific
hypotheses regarding the cognitive variables involved in the parent’s deviancy-
categorization processes. With this specific sub-category as a focus, this literature
review will make every attempt to distinguish which type of Conduct Disorder is

being investigated. As Quay (1986b) pointed out, however, this will be rarely

possible:

12

Much of the research on conduct disorder has failed to consider the differences

between the two disorders even though there is now considerable evidence for

basic differences in their causes, correlates and consequences. The frequent

failure of researchers to distinguish between the two disorders often makes it

difficult to know to which group, if either, the findings pertain. (p.35)

W A conceptual framework that aids in the understanding of
conduct disturbances can be best provided by reviewing the extensive work of
Patterson (Patterson, 1976, 1982, 1984, 1986a, 1986b, 1986c; Patterson, Dishion, &
Bank, 1984), at the Oregon Social Learning Center. His clinical work and
observational studies on the development of "antisocial" children, spanning over the
past 20 years, represents perhaps the most extensive and comprehensive research on
conduct disorders. The observational research allows for one to enter the family
domain and explore the "nuts and bolts" of family interactions. His research
"provides one of the rare hard data looks at the family system and the operating or
functional levels within such a system" (Wahler, 1982, p.82).

Patterson is a proponent of a "social interactional" perspective where the
stimulus-response (S-R) associations are no longer the fundamental units of analysis.
Rather, it is the patterning or relationships among these S-R events that are the basic
units of study. Thus, the focus is on the specific patterns of family interactions, or
rather the extended patterns of action and reaction (i.e., a succession of S-R events).
From this perspective, each family member’s behavior serves as both a response to

an antecedent event, as well as a stimulus or antecedent event for subsequent

13

behavior. The underlying assumption of Patterson’s (1982) "coercion theory", shared
by other behavioristic theories, is that the determinants of behavior (e.g., antisocial
behavior) can be identified through the monitoring of the target behaviors’ objectively
observable antecedents and consequences (Robinson & Jacobson, 1987).

Patterson’s " coercion theory" has its root in his clinical work (e.g. , Patterson,
1975; Patterson & Reid, 1973) with the families of antisocial boys. The focus of the
clinical approach was on parent training and this emphasis on parental
behaviors/characteristics and responsibilities remains central to his conceptualizations.
From this clinical work and subsequent research has emerged the "conviction" that
the parents’ child-rearing practices (i.e., their "family management skills") play a
major role in determining both prosocial and antisocial child behaviors. Patterson’s
(1982, 1986a, 1986c) emphasis on the critical role of "inept family management skills"
in the development of antisocial behavior demonstrates this conviction. "The parents’
failure to teach reasonable levels of compliance sets in motion a process of coercive
exchange with family members. The effect of these exchanges later generalize to the
school setting to produce academic failure and poor peer relations" (Patterson, 1986a,
p.433).

At the heart of the conceptual foundation is Patterson’s position that it is the
family that becomes a training ground for the child’s antisocial behavior. There is
the underlying assumption that antisocial behavior is learned and that this education
takes place within the myriad of family interaction sequences. It is Patterson’s

position that there are patterned exchanges of social interactions (i.e., "microsocial

:18".

r
I... A,

1.4“

l4
exchanges") which provide the vehicle for training. The families of conduct-disturbed

boys are characterized by their "coercive" interactions (i.e., they attempt to employ
"control-through-pain" techniques). In other words, these families are not
characterized by cooperation and other prosocial behaviors, but rather it is as if
family members attempt to "bully" one another as they frequently engage in aversive
behaviors. It is assumed that the parents’ failure to control and limit these negative
"microsocial exchanges" provides for the basic training ground for antisocial
behavior. According to Patterson (1986a):
The assumption central to the general model is that failure by parents to
effectively punish garden variety, coercive behaviors sets into motion
interaction sequences that are the basis for training in aggression. The process
set into motion involves family members in patterned exchanges of aversive
behaviors; the exchanges are such that both members train each other to
become increasingly aversive. This process is labeled coercion. (p.436)
The model is truly interactive, as each family member simultaneously shapes
the behavior of the other and at the same time is being shaped by the family
interaction. For example, as the child becomes more coercive, he/she becomes more
difficult to discipline, thus enhancing the effects of the inept disciplinary practices.
Relatedly, Patterson (1986) pointed out that the identified deviant child can be seen
as both residing in and contributing to a behaviorally disordered family system. In

Patterson’s (1976) words the "child is both victim and architect of a coercive system."

15

Again, the focus is on the patterned irritable exchanges between the problem child,
his mother, and siblings:

The effect of these negative microsocial exchanges is to Shape the behavior of

aggressor and victim simultaneously. The training in coercion or pain-control

techniques is in turn thought to be a necessary prelude to training in physical
fighting. It is assumed that the first or basic training phase is the result of
persistent failure of the parents to use effective discipline in controlling
coercive exchanges among family members. The patterned irritable exchanges
increase in frequency and amplitude, which further disrupts parental efforts to

discipline. (Patterson, Dishion, & Bank, 1984, p. 254)

A few evaluative comments regarding Patterson’s conceptualization will aid
in elaborating his position. First, the above quotation clearly exemplifies the
hypothetical tie between aversive behaviors and actual aggression or assault.
Patterson (1986) does suggest that "these children Show a regular progression from
learning to be noncompliant to learning to be physically assaultive" (p.436). As the
rates of coercion increase, the length of the coercive interactions increase and thus
"...presumably, increases in the amplitude of aggression are also likely to be
reinforced" (p. 436). This is the essence of Patterson’s (1982) "escalation hypothesis,
which describes the process by which a dyad moves from the exchange of rather
innocuous aversive events to higher amplitude aggressive behaviors which

characterize child abuse and spouse abuse" (p. 13).

16

In order to evaluate Patterson’s research and conceptualizations it is essential
to grasp the issues involved in the operationalizing of aggression. A general
definition for the term aggression would involve one person inflicting pain upon
another (Patterson, 1982). Yet, is every aversive event to be classified as aggression?
Patterson’s stated goal is to establish an empirical foundation for a small subset of
aggressive phenomena, namely those aversive events that occur in family interactions.
"Within the family those events will be labeled as aggressive which can be shown to
be both aversive and contingent" (Patterson, 1982, p.12) (i.e., the likelihood of an
aversive event is predicated on a specific antecedent event). Much of Patterson’s
(1982) observational research is based on the Family Interaction Coding System
(FICS) which has 14 categories of aversive behaviors. This Total Aversive Behavior
(TAB) score serves as a general measure of the level of coerciveness for the problem
child, as well as each family member. The behaviors included are negativism,
noncompliance, teasing, yelling, whining, issuing a negative command, crying,
disapproval, dependency, ignoring, engaging in repetitive behavior to the extent that
it becomes annoying, humiliating, destruction of property, and physical aggression.

Obviously, these fourteen events do not refer to commonly defined high-
intensity aggression like shooting or stabbing. Patterson’s ( 1982) position, however,
is that these fourteen aversive events "...are not an unwanted concern with trivia..."
(p.13) because it is hypothesized that it is in fact these ubiquitous family events that
can escalate to high intensity aggression. Clearly, these fourteen behaviors reflect a

wide spectrum of aversive events employed in a contingent manner by family

17
members. Patterson (1982) believes that while these behaviors differ in topography,

(e. g, cry versus physical assault) they all reflect a common process and are members
of a common set:

Members of this set share several characteristics: 1) they are aversive; 2) they

are used contingently; 3) they produce a reliable impact upon the victim; and

4) the reaction of the victim has both a short-term and a long-term effect upon

the aggressor. (Patterson, 1982, p.13)

Robinson and Jacobson (1987) are skeptical of the actual support for this
"escalation hypothesis". They state that physical aggression is a relatively rare event
and that even verbal aggression is not a high-base-rate behavior during a l-hour home
observation. The scarcity of these behaviors creates problems for a functional
analysis since it may be difficult to find enough behavior to evaluate. Robinson and
Jacobson (1987) state that "coercion theory circumvents this problem by studying a
class of behaviors termed aggressive that are not, at face value, obviously aggressive.
These behaviors are called coercive, but perhaps a more neutral term would be
aversive" (p.132). In stating this critique it is as if they have ignored Patterson’s
stated goal to investigate a specific subset of aggressive behavior, those involved in
family interactions. While Robinson and Jacobson do give credit that Patterson’s
research establishes an operational basis for the study of punishing events in family
interactions, nevertheless, they question whether this constitutes a basis from which
to study aggression. Their final conclusion regarding the escalation hypothesis does

seem warranted: While children who are reported to be aggressive engage in

Ii
..-

‘Ar

.,}‘

\

..ru

18
significantly higher levels of annoying behavior (Loeber & Schmaling, 1985a;

Patterson, 1982), "...the chain of events that leads to full-scale aggression has not yet
been delineated and remains hypothetical" (Robinson & Jacobson, 1987,p. 133).

Understanding the escalation hypothesis involves grasping the critical role of
negative reinforcement. In fact, it is the principal mechanism in coercion theory.
Negative reinforcement is an arrangement in which an aversive event is followed by
a response which, in turn, has a neutral or positive outcome. The child’s behavior,
as well as his/her parent’s, is said to be negatively reinforced when that behavior
results in the termination of some kind of noxious/aversive event. When the parent
" gives in" to some kind of aversive child behavior (e.g. , nagging) then the parent is
negatively reinforced for acquiescence because the aversive child behavior has
stopped. Similarly, when a parental request/command is met with an aversive
noncompliance and then not followed through, the child is negatively reinforced since
the noxious stimuli (i.e., the parental command) has been removed.

While Patterson acknowledges that positive reinforcement may play a role in
the development of antisocial behavior, coercion is really based on this negative
reinforcement arrangement. "In distressed families, the three-step, escape-avoidant
arrangements (attack-counterattack-positive outcome) may occur hundreds of times
each day" (Patterson, 1986a,p. 436). It is these interactions that form the basis of the
coercion hypothesis. "Negative reinforcement is not only responsible for the
maintenance of many coercive events, it is also the means by which the aggressor

trains a victim to comply or submit" (Patterson, 1982, p. 141). In terms of the

5‘9!

.-‘ _.‘

905511

t-“‘

l9
escalation hypothesis, Patterson (1982) speculates that there is a relationship between

negative reinforcement and the escalation in the intensity of aggression: "Given a
successful outcome for an increase in amplitude, negative reinforcement increases the
likelihood of a high-density response with future presentations of that same stimulus"
(p. 159).

Patterson (1982) does present some interesting descriptive data to support his
notion that in both normal and clinical samples, " . . .the stage is frequently set for the
possibility of a negative reinforcement arrangement" (p. 148). In other words, the
child frequently finds him/herself in the position of "counterattack" (i.e. , reacting to
aversive antecedents). A computer was used to identify the antecedent events
supplied by each family member to the target child’s coercive behaviors. The results
indicated that given a coercive response in the child, there was "a slightly higher"
likelihood of an aversive antecedent for the sample of "out of control boys" (.32), as
compared to the matched non-problem sample (.25). Results from another sample of
"social aggressors" again demonstrated that about one-third of the conduct disturbed
child’s aversive behaviors were reactions to aversive events and not initiations.
"Clearly, then, much of children’s aggression is not attack but counterattack
behavior" (Patterson, p. 148).

Additionally, what is the likelihood that an aversive antecedent will be
presented to the child? Given that the child is engaged in a prosocial behavior, what
is the likelihood that a family member will "crossover" (i.e., launch an unprovoked

attack)? In Patterson’s study, the clinical sample was made up of two groups, stealers

\n“

N \

20

and social aggressors. The results indicated that each family member was
significantly more likely than their normal counterparts to "crossover" and initiate an
attack. Mothers were more likely than fathers or siblings to engage in crossover
attacks. In the social aggressor sample, mothers responded to the child’s prosocial
behavior with an aversive response 5 % of the time, as compared to only 2% for the
normal mothers. Finally, while the problem children were significantly more likely
to crossover towards their mothers than normal children, the problem child was less
likely than their own parents to initiate a conflict.

Patterson interpreted this data as supporting the position that a child in the
clinical sample must learn to cope with aversive intrusions by other family members
and if the coping is successful in terminating these intrusions, then "this would mean
that negative reinforcement mechanisms play an important role in maintaining
coercive child behavior" (Patterson, 1982, p. 150). In fact, Patterson reported that
there was a significantly greater proportion of coercive behavior maintained by
negative reinforcement schedules in the clinical sample (around 21%) compared to the
normal sample (15%).

Some skepticism about the critical role of negative reinforcement has been
raised (e.g., Knutson, 1982; Robinson & Jacobson, 1987; Wahler & Dumas, 1986).
First of all, while the proportion of coercive behavior maintained by negative
reinforcement is significantly different between groups, Patterson does not take into
account discrepant base rates of aversive behavior among aggressive and control

families. Given that there are higher rates of aversive behavior among all family

21

members within the problem families, the aggressive child’s deviant behavior is more
likely to be associated with the termination of aversive behavior by chance alone
(Knutson, 1982). Robinson and Jacobson ( 1987) suggested that while the differences
in the proportion of coercive behavior maintained by negative reinforcement are
significant, they represent small absolute differences. Furthermore, on a strictly
conceptual basis, Knutson (1982) pointed out that all aversive/conflictual interchanges
must eventually end, and hence negative reinforcement of something will inevitably
take place. Inasmuch as aversive events are a ubiquitous occurrence, then so is
negative reinforcement. Robinson and Jacobson (1987, p. 139) concluded that
"negative reinforcement may not be causally related to deviant behavior but may be
an artifact of its definition."

The fact that negative reinforcement plays such a central role in Patterson’s
coercion theory might suggest that the overall conceptualization needs to be viewed
with some caution. While Wahler & Dumas (1986) also acknowledged some
conceptual difficulties with negative reinforcement (e.g., it is unable to account for
the long-term maintenance of some coercive exchanges) they nevertheless point out
that the literature broadly supports the coercion hypothesis. This conclusion, in
general support of Coercion Theory, will be further substantiated by reviewing the
empirical literature on family interactions. The review of the empirical literature will
demonstrate that there are differences in the quality of family interactions that

characterize those families with a conduct disturbed child, as compared to normal

22
families. These differences are frequently in the direction that would be predicted by

Coercion Theory.

W. The overall impression is quite clear; namely, there
are a variety of parental characteristics and family interaction patterns that
characterize families with conduct disordered children (Hetherington & Martin, 1986;
Olweus, 1980; Patterson, 1982; Rutter & Garmezy, 1983; Rutter & Giller, 1983;
Snyder & Patterson, 1987;). The support for these conclusions comes from an
established body of research that includes observational studies, both in the laboratory
and home setting, that have attempted to delineate and specify the actual maladaptive
interactive processes that correlate with the global questionnaire and interview data
(e.g., Patterson, 1982). It is important to point out that the most popular
methodology is correlational and hence definitive statements regarding cause and
effects are still lacking. Recently, however, path analytic models have been employed
(e. g. , Olweus, 1982; Patterson, 1986a) providing for a step towards causal statements.

Rutter and Giller (1983) suggested that this accumulation of data can be
organized around six significant dimensions/factors that are involved in the mediation
of family influences on the development of conduct disorders: (a) characteristics of
the parents; (b) intra-farnilial discord; (c) weak relationships with parents; ((1)
ineffective supervision and regulation of the child’s behavior and activities; (e)
socioeconomic conditions; and (f) family size.

While the present study is focused primarily on factors (a) and (d),

obviously all of the six factors are interrelated. For example, it seems reasonable to

23
assume that the overall family atmosphere and quality of relationships (i.e. , factors

(b) and (c) ) are reciprocally interdependent with the ways in which the parents
perform their parental responsibilities. Thus, given the reciprocal nature of the
influences that characterize the family (i.e., the family-systems’ effect) any
categorization of the family influences involved in the development of a conduct
disorder is solely for the sake of convenience.

Snyder and Patterson’s (1987) survey of the literature is specifically relevant
for the present study because they organized the family interaction data with an
emphasis on parental behaviors/responsibilities. They organized the relevant family
interaction patterns and characteristics associated with a conduct disorder around five
themes: (a) discipline, (b) positive parenting, (c) monitoring, ((1) conflict and problem
solving, and (e) sociodemographic characteristics. Within each of these themes a
variety of cross sectional and longitudinal studies were reviewed, noting how much
of the variance in the antisocial behavior had been accounted for by the variety of
measures used to assess each theme. Overall, Snyder and Patterson ( 1987)
concluded:

The data consistently suggest that parenting practices and family interaction are

associated with the development of antisocial and delinquent behavior. These

findings are consistent in studies using cross sectional and longitudinal designs,
and are consistently replicated using different modes of measuring the various
parenting and family interactional constructs. The same variables are

implicated whether delinquency [or antisocial behavior] is defined using self-

24

report measures, [observational studies] or official arrests and convictions. (p.

233)

I' .10.__.r' -. . .‘. viw f - . urrrr..- Li 1 - - On FIJ' ; I n
This study is concerned with the effects of parental cognitive factors as they relate to
the development and maintenance of the child’s conduct disorder. The literature to
be reviewed in this section will clearly demonstrate that parental discipline, and the
component behaviors involved in its effective implementation, is consistently
associated with the child’s conduct disturbance (Snyder & Patterson, 1987). It is very
likely that parental cognitive factors play a crucial role in the parent’s disciplinary
behaviors. In order to develop and advance this position, it will be essential to
identify those parental/ family characteristics that differentiate families with a conduct
disturbed child from normal families, emphasizing those component behaviors
involved in disciplinary practices. In the next section on Parental Cognitive
Variables, the issue as to whether these parental/family differences can be accounted
for, in part, through an understanding of the parent’s cognitive processes will be
addressed.

For now, it will be helpful to provide a conceptual schema for discipline.
What is discipline? And, what are the parental and family characteristics that might
be relevant factors in terms of the effectiveness of the disciplinary practices? Snyder
and Patterson (1987) pointed out that "discipline is a complex construct that refers to

methods used by family members, [primarily parents], to discourage behavioral

25
excesses or antisocial behavior in children" (p.220). They provided three aspects

involved in effective methods of discipline:

(1) the accurate definition and labelling of certain behaviors as excessive or

antisocial; (2) the consistent tracking of those behaviors over time and across

settings; and (3) the consistent and contingent use of effective but not harsh

methods to inhibit those behaviors. (p. 220)

Similarly, Berkowitz (1973) suggested that the timing, consistency, and intensity of
the parental response are key factors in the effectiveness of discipline.

There is another class of significant variables that are likely involved in the
efficacy of discipline, namely, the overall family atmosphere or emotional tone,
including the quality of relationships. The interpersonal context of the discipline is
a critical factor. Parental punishment is more effective in suppressing undesirable
behavior when it is administered by a warm, nurturing, and loving parent, as
compared to a cold, harsh, or negative parent (Berkowitz, 1973; Wahler, 1969).
Snyder and Patterson (1987) referred to this factor as "positive parenting " and
suggested that it may be directly involved in enhancing the "reinforcement value" of
the parent.

Relatedly, the frequency with which the parent responds to and values
prosocial behavior may be an important factor in terms of discipline. Berkowitz
(1973) pointed out that there is a likelihood of inhibiting a deviant behavior by
strengthening responses that are incompatible with that behavior. Kazdin (1982)

referred to the phenomenon of "response covariation" (i.e., two or more behavioral

26

responses are related). There are covariations among the responses that comprise a
child’s behavioral repertoire. When responses are related, changes in one response
affects the other related responses. The suggestion is that prosocial and antisocial
behaviors may be part of a response covariation. Wahler and Dumas ( 1987) pointed
out that "it is evident from the observational studies that the organization of a child’s
prosocial behaviors with respect to his or her problem behaviors is sometimes
predictable" (p. 593). Russo, Cataldo, and Cushing (1983) did report that problem
children that exhibited the intended goal of an increase in compliance also exhibited
an incidental reduction in their problematic behaviors.

The research that differentiates the parents/families of conduct disordered
children from normal families will now be reviewed. The review will highlight the
following component behaviors/interactions that are suggested to be involved in
effective discipline: family atmosphere, parental attitudes and emotional tone, and the
contingent and non-contingent reinforcement of both prosocial and deviant behavior.
(The literature on parental labeling and tracking of child behavior will be reviewed
separately in the subsequent section on parental cognitive variables.)

The overall family atmosphere and the parents’ attitudes towards their children
do seem to be directly related to the presence of child behavior problems. While both
normal and deviant families generally exhibit more positive than negative behavior
towards their children, regardless of the child’s behavior (Snyder, 1977; Wahl,
Johnson, Johansson & Martin, 1974), the families of antisocial children do exhibit a

significantly greater amount of aversive or displeasing behaviors (Patterson, 1982;

27
Snyder, 1977). Snyder (1977) observed the behavioral sequences and interactions

within both normal and conduct disordered families. The data revealed that 4.7 % of
the behaviors emitted in problem families (e.g. , between father, mother, and son)
were displeasing, and this was two times the rate observed in normal families. Each
family member was significantly higher than the normal counterparts in exhibiting
aversive behaviors (Patterson, 1982; Snyder, 1977).

These results are characteristic of an overall negative and aversive family
atmosphere. This characterization supports Patterson’s (1982) concept of a bilateral
trait which implies that the child’s performance of coercive behavior is the outcome
of two dispositions. Each member brings his or her own disposition to act irritably.
Therefore, the bilateral hypothesis requires that one examine not only the child’s
reactions to others, but also the reactions of family members to the child. In fact,
Patterson ( 1982) did report that measures of the mother’s irritable reactions to the
child account for over 40% of the variability in the child’s TAB score. Patterson
suggested that "these findings support the idea that the problem child’s performance
level is related in some fashion to the manner in which other persons react to him"
(p.256). He concluded that aggressive children do live in families with aggressive
people. This conclusion is firmly supported in the literature and is generally
congruent with the position of Coercion Theory.

Family interactions involving behaviorally disturbed children, especially when
the parents are involved, are consistently and repeatedly characterized by their

emotional negativity (Delfini, Bemal, & Rosen, 1976; Johnson, Wahl, Martin, &

28
Johansson, 1973; Lobitz & Johnson, 1975b; Olweus, 1980), by their criticalness and

rejecting qualities (Forehand, King, Peed, & Yoder, 1975; Olweus, 1980; Patterson,
1986b), and by higher levels of irritability (Patterson, 1982). Many studies have
emphasized the frequency/rates and quality of parental commands (Delfini et al.,
1976; Forehand et al., 1975; Forehand, Well & Sturgis, 1978; Johnson & Lobitz,
1974; Lobitz & Johnson, 1975b) as indicative of the overall negative and controlling
family atmosphere.

Olweus (1980) employed a path analysis on two different representative
samples of Swedish boys, one group age 13 and the other 16 years old, in order to
examine the familial and temperamental determinants of aggression in adolescent
boys. Retrospective interview data were scored for the parental variables (i.e. ,
negativism, pernrissiveness for aggression, and the use of power-assertive discipline
methods) as well as the determination of the boy’s temperament. Aggression scores
were independently obtained via peer ratings. The mother’s negativism was a
composite score which characterized the principal caretaker’s basic emotional attitude
towards the boy in the first five years of life. Mother’s permissiveness (an accounting
of discipline style) and mother’s negativity were the most important causal factors,
as the entire causal model accounted for between 21% and 33 % of the variance in the
boy’s aggression. "If the mother is negative, rejecting and indifferent to the boy when
he is young (and later), this is likely to result in the adolescent boy becoming

relatively more aggressive and hostile toward his environment" (Olweus, 1980,

p.651). There was also a substantial indirect effect of mother’s negativism and the

29

use of power assertive disciplinary techniques such as strong threats and corporal
punishment on the development of higher adolescent aggression levels. Clearly, the
study demonstrated that the basic emotional attitude of the mother and the lack of
clear limits, or rather an inept discipline style, are directly involved in the
development of a child’s conduct disturbance. Patterson (1982) pointed out that the
difficult child, in conjunction with an overly permissive, highly irritable mother is a
familiar theme evidenced in the literature on conduct disorders.

Lobitz and Johnson (1974) have demonstrated a causal relationship between
an increase in both parental commands and negative responding with an increase in
child deviancy. Parents of normal children were asked to attempt to make their
children look bad on certain days of home observation. The results showed that
parents can raise deviancy levels in normal children, even when the effect of child
compliance was removed from the deviancy score, by increasing the rate of negative
responding and the rate of commands. Lobitz and Johnson (1975a) reported similar
results from a follow-up study which included a sample of deviant children. A
reduction in the problem child’s deviant behaviors, from a base rate, was associated
not only with a decrease in parental negativeness but also with an increase in parental
positiveness.

Lobitz and Johnson (1975b) concluded that it appears as if deviant children
"...were being subjected to a greater proportion of negative feedback and parental
control than their non-deviant counterparts, regardless of whether their behavior was

deviant or non-deviant" (p. 368). Lobitz and Johnson (1975b) analyzed the parent’s

30

response to child deviant behavior and child nondeviant behavior separately. The
parents of problem children were significantly more negative to both deviant and
nondeviant behavior than were the parents of the normal children. Thus, the parental
negativity was associated not solely with deviant behavior per se, but rather with the
child’s deviancy status/label. These parents also issued significantly more commands
to their children. Relatedly, Delfini et a1. ( 1976) found that the parents of deviant
boys produced a significantly greater number of negative commands (i.e., their
commands were stated in an angry, threatening, humiliating, or nagging manner).
The rate of overall commands was correlated with disruptive behavior in the deviant
group but not in the normal group. It is the quality of the parental commands that
appears to be a relevant variable. Forehand, Wells, and Sturgis (1978) found that
fewer vague parental commands and more parental rewards issued in a free play
laboratory task, together accounted for 55% of the total variance in the child’s
observed compliance in the home setting. These studies demonstrate that the child’s
deviant behavior and/or deviancy status is consistently related to both parental
negativity/criticalness and parental control or commands.

It has already been mentioned above that the mother’s irritability accounted for
the largest proportion of variance in the child’s performance of antisocial behavior.
It is interesting to note Patterson’s (1982) four measures of irritability: crossover,
counterattack, punishment acceleration, and continuance. Crossover, mentioned
above, refers to the start-up of a conflict, given that the other person is reacting in

a neutral or positive fashion. Counterattack, also mentioned above, refers to the

31

second step in the process (i.e., once there is a start-up) the person has a choice to
ignore or attack back. Given an initiation by a family member and a counterattack,
punishment acceleration refers to the likelihood that the aversive exchange will
continue. The final measure, continuance, refers to the disposition to persevere in
one’s irritable reactions regardless of the response of the other family members.
Family members of antisocial children were significantly higher on the measures of
crossover, counterattack and continuance than their matched normal counterparts.
For punishment acceleration there were nonsignificant trends for mothers, fathers, and
siblings. The antisocial child, however, did Show significantly greater punishment
acceleration than normal children. (This issue of responsivity to punishment will be
addressed below.) In terms of the continuance measure, comparisons across three
samples (e.g. , normals, stealers, and aggressors) were made for six dyadic units (e.g. ,
mother interacting with child, child with mother, child with father, father with child,
child with siblings, and siblings with child). With the exception of the sibling
interactions involving the target child, all other comparisons showed a significantly
greater tendency in the clinical groups to continue negative/aversive interactions
regardless of how the other members of the dyad reacted.

The data just reviewed highlights the negative/aversive ambience which
characterizes the families with an antisocial child. Conduct disordered children
appear to come from generally dysfunctional families. The data on parental
negativeness, criticalness, and family irritability does seem to offer further support

for Patterson’s ( 1982) bilateral hypothesis, mentioned above, which stresses the

32

importance of the family members’ reactions to the target child. Furthermore, it is
reasonable to suggest that the data can be seen as supportive of Coercion Theory,
which would predict that as parents become more negative so do their children and
vice versa, in a reciprocal escalation of the use of aversive events to control each
other. Clearly, the correlational evidence is supportive of this position. In fact, this
relationship between parental negativity (taken as a class of aversive responses) and
problematic child behavior has also been demonstrated in normal families (Johnson
et al., 1973; Lobitz & Johnson, 1975b) where the best overall predictor of the rate
of child deviance was the parent’s overall negativism.

There is research to suggest that the most effective parental response to a
child’s aversive behavior is no response at all (Patterson, 1982; Sallow, 1972).
Patterson (1982) reported that for both the clinical and normal samples, the child
would perform a single coercive act and then stop 66% of the time, given no parental
punishment. Johnson et a1. (1973) reported, however, that even with the normal
sample, the more deviant child gets more attention, either positive or negative, no
matter whether he is being deviant or nondeviant. Patterson (1982) reported that
parents of antisocial children tend to discipline more. However, "in effect, their
attempts to punish coercive child behavior made things worse" (p.130). Patterson
(1982) frequently observed a pattern where the parent nags and threatens but seldom
follows through so in the long run they do not "train" the child in appropriate
behavior. These parents are aversive without the inhibitory effect. "Nattering" is

Patterson’s term for this ineffective parental discipline.

33

Hence, while deviancy does get attention, this parental attention does not
generally have the effect of suppressing the ongoing deviant behavior. It is therefore
important to look at what are the differences in the child’s responsivity to parental
punishment/consequences. Snyder (1977) reported that in normal families there is an
overall decelerative effect, (i.e, aversive consequences to displeasing behaviors tend
to decrease the rate of the behavior under baseline). For the problem families,
however, this was not true. For the family unit overall, aversive consequences to
displeasing behavior increased the rate of displeasing behavior over baseline. Snyder
(1977) concluded:

. ..that problem and nonproblem families can be differentiated on the basis of

their responsivity to consequents. In nonproblem families, punishment was

associated with a suppression of displeasing behavior. In problem families,
punishment was associated with an acceleration rather than a suppression of

displeasing behavior. (p. 534)

In terms of the target child, however, parental discipline had an acceleration
effect for both normal and problem boys, (i.e,, aversive consequences to deviant
behavior tends to increase the rate of the deviant behavior (Snyder, 1977). Patterson
(1982), however, reported that in normal families the impact of aversive consequences
varied as a function of the class of aversive responses. The normal sons’ overall
deviant behavior, as indicated by the TAB score, was slightly accelerated following
parental punishment. Parental discipline was effective towards suppressing both

social aggression (e. g., physical negative and teasing) and hostility (e. g., disapproval,

34

whine, and negativism). On the other hand, the aversive reactions by the parents of
antisocial children did not effectively suppress ongoing aversive behavior in any of
the specified response classes. In a direct comparison between the samples, conduct
disordered boys were significantly more likely to persist in aversive behavior, given
a negative consequence, than were the normal boys (i.e., the deviant children were
significantly more likely to accelerate when punished (Patterson, 1982; Snyder,
1977)).

Snyder ( 1977) suggested that the lack of responsivity is the result of a history
of noncontingent reinforcement. It is reasonable to assume that the relative
ineffectiveness of the parental discipline may be related to its random, haphazard, and
noncontingent employment. Furthermore, Wahler (1969) suggested that a lack of
contingent reinforcement also has an effect on family members responsiveness to
social reinforcers (i.e. , family members’ prosocial behaviors). Wahler (1969)
demonstrated, in a two case clinical study, that increasing the parental reinforcement
value (i.e., the parents’ effectiveness) was accomplished by training the parents to
be more contingent with their positive and negative reinforcement. This increase in
contingent reinforcement coincided incidentally with an increase in the positive
interactions between child and mother.

The literature regarding the parents’ contingent and noncontingent responses
consistently demonstrates that parents of deviant children are generally more
noncontingent in their responses. Studies by Wahl et a1. (1974) and Snyder (1977)

offered empirical support that nonproblem families do generally discriminate in their

35

responses to deviant or prosocial behavior. While there was some tendency for
positive reinforcement of deviant behavior, deviant behavior received significantly
more negative and neutral consequences and significantly fewer positive consequences
than did nondeviant behavior. Clearly, nonproblem families appear to respond
contingently. Even though Johnson et a1. (1973) provided further support that normal
parents do discriminate between deviant and nondeviant behavior, they nevertheless
provided interesting evidence to support the notion that family members can be
characterized by a response class, or a generalized tendency to respond in a given
manner regardless of the antecedent. The best support for this notion was in terms
of a negative response tendency. "A tendency toward negative responding may be
shown irrespective of prior child behavior" (Johnson et a1. 1973, p. 51).

Families with problem children, on the other hand, do not appear to respond
contingently. Parents of deviant children are significantly more likely to provide
positive reinforcement for negative behavior (Patterson, 1982; Sallows, 1972/1973;
Shaw, 1971/1972; Snyder, 1977) and are more likely to punish prosocial behavior
(Patterson, 1982; Shaw, 1971/1972). Patterson’s measure of "crossover" is an index
of punishing prosocial behavior. On the basis of sequential analysis of 10 normal
families and 10 families of behavior problem children, Snyder (1977) concluded:

In the problem families there were no contingencies. The probability of

receiving a positive, neutral or aversive consequent was independent of the

behavior displayed. Members from nonproblem families were more

discriminating. Their responses fostered the development of prosocial

36

behavior and discouraged the display of deviant behavior. For a member from

a nonproblem family, it paid to act in a way that was pleasing to other families

members. In problem families, the payoff was nearly the same for both

pleasing and displeasing behavior. (p. 533)

Overall, this review of family interactions and the examination of the
component behaviors involved in the parent’s disciplinary practices (i.e. , the overall
aversive family atmosphere, the parents’ negative, critical, and controlling attitudes,
and the greater tendency towards the noncontingent reinforcement for both prosocial
and deviant behavior) would strongly suggest that the parental disciplinary practices
of behaviorally disturbed children are relatively ineffective when compared to the
parents of normal children. Therefore, it will come as no surprise that the
effectiveness of the parent’s discipline appears to be related to the development of a
conduct disturbance. In their review of the literature, Snyder and Patterson, (1987)
reported that "the various measures of discipline in these studies account for between
1 to 40 percent of the variance in child antisocial behavior. Discipline is modestly
but significantly related to official delinquency, aggression, stealing, and drug use"
(p. 221).

Relatedly, recent studies examining family-management skills, in which
discipline plays a major role, also suggest that inept family management is critically
involved in the development of conduct disorders (Patterson, 1986a,1986b,l986c;
Patterson & Dishion, 1985; Patterson, Dishion, & Bank, 1984; Patterson &

Stouthamer-Loeber, 1984). The family management skills that have been studied

37

include: (a) discipline, (b) monitoring (i.e., knowing where, when, and with whom
the child is with), (c) positive reinforcement/parenting (i.e., the overall positive tone
and contingent reinforcement of prosocial behavior), and ((1) problem solving. The
intercorrelations of these variables were examined (Patterson, 1986c; Patterson &
Stouthamer-Loeber, 1984) and the significantly positive correlations suggested that
they are defining a common trait. Disruptions in one kind of parental practice
appears to be accompanied by disruptions in other practices as well. The highest
correlation was between discipline and positive parenting. The parents’
ineffectiveness in their implementation of discipline appears to be related to a lack of
prosocial interactions with their children. This would be consistent with Wahler’s
(1969) clinical study reviewed above.

While Patterson and Stouthamer-Loeber (1984) reported that parental
monitoring accounted for the most variance in the path analysis model, the discipline
variable was also shown to make a significant contribution in the two measures of
delinquent activity. The parental monitoring variable was also found to be
significantly related not only to delinquency, but also to relationships with deviant
peers and deficient social skills (Patterson & Dishion, 1985). In another path
analysis, Patterson, Dishion, and Bank (1984) reported that "inept parental discipline"
was a significant variable. The path from inept discipline to "negative microsocial

exchanges" (i.e., aversive sibling interactions) was statistically significant in a
bidirectional process. The effect of inept parental discipline on the antisocial criterion

of physical fighting was mediated by the negative microsocial exchanges. This was

38

taken as supportive of the position that the family becomes a training ground for
aggression. The weakness of the study, however, was that the construct of Discipline
was defined by only a single indicator.

In a recent study, Patterson’s (1986a) stated goal to use multiple methods and
multiple agents to define each construct was accomplished in regards to the construct
of "inept discipline". Once again, the study supported the importance of discipline in
terms of the child’s conduct disturbance. The strongest path was from inept discipline
to antisocial behavior, as this accounted for approximately 40% of the variance.
Also, the parent’s inept discipline was bi-directionally related to the target child’s
coercive exchanges with other family members. Patterson (1986a) interpreted the
results that "something as ordinary as the level of the parents’ family-management
skills" sets into motion an elaborate and complex process involving the child’s family
relationship, and the child’s academic performance, and peer contacts. The path
models clearly suggest that family management, and specifically parental discipline,
originates these processes involved in the development and maintenance of a conduct
disorder.

In summary, parental characteristics and family interaction patterns consistently
differentiate families with a conduct disturbed child from normal families. Clearly,
the literature strongly suggests that these families are dysfunctional. The family
interactions that are involved in the implementation of discipline appear to be
significantly deficient in these problem families when compared to normals.

Furthermore, the literature does suggest that disciplinary practices do play a crucial

39

role in terms of the development of a conduct disturbance. While the results are
largely correlational, the path models do promote a more causal explanation. It is
reasonable to propose that the parent’s inability to create an emotionally positive
family atmosphere, as well as their difficulties in providing clear limits and the
ineffective inhibition of the child’s deviant behavior (i.e., their inept family
management skills) lie at the very heart of the child’s conduct disturbance.
P ni 'v V ' les

Research literature that emphasizes the importance of the parent’s cognitive
functioning in terms of the child’s behavioral disturbance will be reviewed in this
section. Data that suggest that parents of conduct disordered children do not
accurately perceive the child’s behavior will be presented. Two related perceptual
deficits will be discussed: (a) Parents of behavioral disturbed children do not
accurately label their children as there are apparent discrepancies between the parent’s
perceptions of the child’s behavior and the child’s actual behavior. The parental
perceptions of the child’s adjustment do not appear to be based solely on the child’s
actual behavior; and (b) parents of conduct disordered children exhibit some deficits
in the categorization of prosocial, neutral, and deviant behavior.

Bugental and Shennum (1984) stated that the child socialization literature has
overlooked the intervening role of the parents’ cognitive factors in the interaction
process. Recently, other investigators have also made note of the need to understand
the role of parental cognitive variables in the development of the child’s conduct

disturbance (Griest & Wells, 1983; Lobitz & Johnson, 1975b; Patterson, 1986b;

40
Robinson, 1985). For example, Griest and Wells (1983) stated: "Although this tOpic

has received little attention in the past, it appears to us that future research is needed
in the area of cognitive parent variables and their relation to parent’s perceptions of
child behavior" (p. 41).

Obviously, the ways by which a person views and understands the world
around them affects the ways in which that person acts and interacts. Thus, it is only
common sense that cognitive factors play a major role in determining human
interactions. The ways that parents perceive their children, therefore, undoubtedly
affects how the parent interacts with the child and, consequently, how the child
behaves.

Adhering to this line of logic, it is the position of the present study that
parental cognitive factors play a critical role in the development and maintenance of
a conduct disorder. The present study proposes that differences in the parents’
cognitive/perceptual behavior may be related, in part, to the differences in the
component behaviors involved in the parents’ disciplinary actions. The parent’s
negativity, criticalness, lack of contingent reinforcement, and overall inept discipline
of the child’s behavior (reviewed in the previous section) may, in part, be the result
of the ways in which the parent tracks, defines/labels, and processes the child’s
behavior. Clearly, "parents who cannot accurately perceive different types of
behavior will probably not respond to them in an appropriate manner" (I-Iolleran,
Littrnan, Freund, & Schmaling, 1982, p. 548). Furthermore, it is reasonable

to speculate that an inaccurate parental perceptual style can lead to an inaccurate

41
overall assessment of the child and this can result in a feedback loop; the child’s

deviant behavior can be viewed as an attempt to be congruent with the parent’s
negative label or impressions. It is reasonable to suggest that there is a reciprocal
interactive process between the parental labeling of the child and the child’s actual
behavior. Rickard, Graziano and Forehand (1984) suggest that "parents, by virtue
of their unique relationship to the child, may serve not only to initiate but to
perpetuate child deviant behavior, and negative child self-evaluations" (p. 184), as they
caution against parental self-fulfilling prophecies.

Support for the feasibility of this general position is provided by a creative and
well executed study by Bugental and Shennum (1984). (This study has been referred
to by Shaver (1984) as "having all the markings of a classic".) Bugental and
Shennum (1984) have demonstrated the effects of parental attitudes and beliefs on
ongoing parent-child interactions. They clearly demonstrated that certain parental
beliefs (e. g., in this case beliefs regarding the causes of positive outcomes in parent-
child interactions) acted as "selective filters or sensitizers" to specific child behaviors.
Thus, the parental beliefs determined the nature and amount of the adult’s reaction
to different child behaviors. Furthermore, there was evidence to support the
prediction that adult attributions acted in a self-fulfilling fashion (i.e. , the
communication patterns that followed from parental beliefs served to elicit the
behavior patterns that maintained those beliefs). The parental attributions did not
directly cause child behavior, but rather acted as central moderators in the

transactional process. "In our conception, attributions act as cognitive filters--

42
determining what stimuli will be responded to and in what fashion" (Bugental &

Shennum, 1984, p.78).

Clearly, "parental cognitive factors" can refer to a wide variety of behaviors
including: perceptual styles, attitudes, beliefs, expectations, specific
knowledge/information, and the accumulation of prior experiences. A few studies
have examined the relationship between parent cognitive variables, perceptions of the
child, and deviant child behavior. For example, Cahill (1978) reported that parents
labeled their normal children as "difficult" when the child’s behavior did not meet the
parents expectations. Interestingly, many parental expectations exhibited a lack of
knowledge regarding normative child behavior. Addin'onally, Rickard, Graziano, and
Forehand (1984) found that a scale measuring parental beliefs, attitudes, and
expectations did differentiate parents’ of clinic-referred children from parents of
normal children and there was some relationship between these cognitive measures
and actual behavior in the home for the clinic sample. Overall, the parents’ beliefs
about disciplinary practices and their knowledge of child development norms were the
two most important cognitive variables that differentiated the groups. The results
were taken as supportive of the possibility that some children are inaccurately labeled
as deviant, when in fact the principle locus of the "problem " is not the child’s
behavior, but the interpretation of the child’s behavior by the parents. In other
words, both a lack of knowledge about developmental norms and/or unrealistic

expectations may result in inaccurate parental interpretations of the child’s behavior.

43
P P r ' lin f ' B h vi The focus ofthe present

study on parental cognitive factors emphasizes the parent’s overall perception of the
child and the child’s actual behavior. There is now mounting evidence to suggest that
parents do not always perceive their child’s behavior accurately (Griest & Wells,
1983; Rickard et al., 1984). In the studies reviewed in the previous section on
conduct disorders, families with behaviorally disturbed children were compared with
normal families. Studies did indicate that the deviant child’s observed behavior was
significantly different from that of normal children (e.g., Delfini et al., 1976;
Forehand, King, Peed, & Yoder, 1975; Griest, Forehand, Wells & McMahon, 1980;
Lobitz & Johnson, 1975b; Patterson, 1982; Snyder, 1977). It would appear
reasonable to conclude that these studies Show that parental perceptions are accurate
since these children were either significantly more deviant and/or more noncompliant
than the matched controls. Thompson and Bemal’s (1982) investigation into the
factors associated with parental labeling of children referred for conduct disturbances
also concluded that parental perceptions were accurate.

Even though statistically significant differences did exist between the deviant
and the normal children on behavioral variables, there was, nevertheless, considerable
overlap in the rates of deviant behavior ( e.g., Delfini et al., 1976; Griest et al.,
1980; Lobitz & Johnson, 1975b). In other words, there are a variety of studies that
suggest that there is not a straightforward relationship between the parent’s perception
of child deviancy and the child’s actual behavior. Delfini et a1. ( 1976) reported that

" . . .approximately half of the deviant boys appeared no different from the normal ones

44
in the behaviors measured" (p.246). This rate of overlap is consistent with that found

by Lobitz and Johnson (1975b) and suggests that parental perception of the child is
not totally explained by the child’s actual behavior. Some children are clearly
inaccurately labeled as deviant.

In fact, Rickard, Forehand, Wells, Griest, and McMahon (1981) directly
addressed the issue of differences between normal, clinic-referred behaviorally
deviant, and clinic-referred behaviorally non-deviant children, as they confirmed the
existence of a group of children who were labeled as deviant but exhibited behaviors
comparable to the normal group. While the Clinic Deviant group was significantly
more deviant in behavioral measures collected in the home, the Clinic Non-Deviant
group was not different from the Non-clinic group on the behavioral measures. In
terms of parental perceptions of maladjustment, however, the parents in both clinic
groups perceived their children as significantly more maladjusted than the non-clinic
parents. Thus, the parents’ perception of the clinic-nondeviant children was an
inaccurate assessment of the childrens’ actual behavior.

Parental perceptions appear to be potentially inaccurate as they are not always
based on actual child behavior. There is now a body of research to suggest that the
objective nature of the child’s behavior is not the sole determinant, or even the most
important factor involved in the parent’s perception of the child. On the basis of their
research with families of pre-teenage behavior problem children, Christensen,

Phillips, Glasgow, and Johnson (1983) reported that "no significant relationship exists

between child negative behavior and parental perception of child behavior problems,

45

even when the score for child negative behavior includes only behavior directed
toward the parent and is weighted by the parent’s reaction to those behaviors"
(p.162). (The data indicated that parental perceptions of child behavior were
associated with marital discord and parental negative behavior towards the child, but
not the behavior of the target child!) Lobtiz and Johnson (1975b) pointed out that
while their data exhibited a great overlap on the child behavioral variables, there was
not a lot of overlap on the parent perception variables. In fact, results from a
stepwise discriminant analysis correctly classified 90% of both the clinic-referred and
non-clinic-referred children based solely on the parent attitude measures. Parental
perception of child adjustment appears to be a better discriminator of children labeled
as behavioral disordered than the actual child behavior (Griest et al. , 1980).
Overall, the conclusion is that there must be factors involved in the parental
labeling process other than the target child’s actual behavior. Research reviewed in
the previous section on discipline strongly suggest that parental negativity is
associated with child/adolescent conduct disturbance. Yet, the correlations between
the child’s actual behavior and the parental perception scores were negligible (Lobitz
& Johnson, 1975b). This lack of a correlation, combined with frequently found
overlap between "deviant" and normal groups on behavioral variables strongly suggest
that ". ..factors in addition to child deviant behavior must be contributing to negative
parent attitudes..." (Lobitz & Johnson, 1975b, p. 369). It is the goal of the present

study to shed light on what might be contributing to the parental negativity.

46

There are two general factors that have attracted a lot of attention in the
literature, namely, marital satisfaction (Christensen et al., 1983; Forehand, Brody,
& Smith, 1986; O’Leary & Emery, 1984;) and parental psychopathology and/or
personal adjustment (Griest et al., 1980; Rickard et al., 1981; Wahler, 1980; Wahler
& Afton, 1980). The literature suggests that these factors may be more important in
determining the parental perception of the child’s adjustment than the actual child
behavior. It does not appear, however, that there is a direct link between these
factors and the parental perception of the child’s behavior, but rather the relationship
is interactive. For example, Griest et al. (1980) found that maternal perception of
child behavior was the best discriminator between the clinic and nonclinic groups.
These investigators then examined whether child behavior or maternal adjustment best
predicted parental perceptions of their children. For the nonclinic group, maternal
perceptions of child adjustment was best predicted by the child’s actual behavior. For
the clinic group, on the other hand, maternal perceptions of child adjustment were
best predicted by an interaction of child behavior and maternal adjustment. Similarly,
Forehand et al. (1986) concluded that it is the combination of marital satisfaction and
actual child behavior that are important determinants of parental perception of child
adjustment. "When a child is disruptive (noncompliant, deviant or conflictual) and
his/her mother is dissatisfied with her marriage, the child is perceived as more
maladjusted than when the only problem is child disruption or an unsatisfactory

marriage exist" (p.48).

47

Another possible factor that may explain the parent’s negativity and inaccurate
assessment of the child’s behavior is the parent’s cognitive schema for the
classification of deviancy and prosocial behavior (Patterson, 1982, 1986b).

W The
literature on the parental categorization of deviant, neutral, and prosocial behavior
comes primarily from the Oregon Social Learning Center (Holleran, Littrnan, Freund,
Schmaling, & Heeren, 1982a; Holleran, Littrnan, Freund & Schmaling, 1982b;
Lorber, 1981; Patterson, 1982, 1986b; Schmaling & Patterson, 1984). Even though
this research, in terms of methodology and breadth of results, is still in its formative
stages, it is nevertheless of critical importance for the present study. A major
component of the present study is based largely on the ideas and hypotheses
(Patterson, 1982; 1986b) as well as the methodology (especially Holleran et al..
1982b) contained in this emerging body of research. Most of the research to be
reviewed in this section, however, is from unpublished reports (with the exception of
Holleran et al., (1982b and Patterson, 1982 & 1986b) and the conclusions reached
should be interpreted more cautiously and tentatively and taken as suggestive.

Patterson (1986b) speculated that a parent’s disposition to react in a positive
or negative fashion contingent on the behavior of the child is related, in part, to the
parent’s schema for categorizing deviant and prosocial behavior. "It is assumed that
parents differ in what it is that they classify as deviant (e. g. , some parents believe that
temper tantrums are normal but that stealing is deviant). The deviancy-categorization

process, in turn, determines which child behaviors will be monitored and which will

48
be punish " (p. 247). On the basis of years of clinical experience, Patterson (1982)

speculated that there are specific qualities in the parental deviancy-categorization
process of those parents with children characterized by an Undersocialized Aggressive
Conduct Disorder (i.e., "Social Aggressors"):

Rather than being distant and unattached, the parent of the Social Aggressor

tends to be overly enmeshed and very irritable. It is hypothesized that when

differentiating between coercive and normal acts, these parents will tend to be
overly inclusive, i.e., they tend to classify many behaviors as deviant that
parents of Normals would classify as borderline events. The mother of the

Social Aggressors becomes an impassioned participant, scolding and

threatening at the slightest provocation. As noted earlier, the irritable

reactions make the situation worse. (Patterson, 1982, p. 249)

The proposition is that these parents tend to view trivial instances of child
misbehavior as worthy of high intensity scolding and these threats are not effective
discipline techniques Since the parents frequently fail to follow through with their
punishments (Patterson, 1982).

Not only do the parents of conduct disordered children possibly have overly
inclusive categories for classifying deviant behaviors, but it is hypothesized that these
parents may have overly exclusive categories for classifying prosocial behaviors.
Patterson does ask if the two classification schemata are independent. It is reasonable
to suggest that they are not independent. As a logical parallel to Kazdin’s (1982)

"response covariation", it is reasonable to propose that there is a "behavior

49

classification covariation" where a change in one of the classification schemata would
affect the other related classification schemata. In other words, if the two
categorization schemata are related, then an emphasis on deviancy would inherently
result in a de-emphasis on prosocial behavior, or vice versa. Furthermore, if this
deviancy-categorization hypothesis is valid, then parents with this overinclusive
schema would be experiencing and responding to a greater amount of child deviant
behavior, while experiencing less positive behavior. Clearly, this might account, in
part, for the parent’s negativity, criticalness, and tendency to start a conflict when the
child is behaving in a neutral or prosocial fashion (i.e. , Patterson’s "crossover").
Finally, an overly inclusive category for labeling behavior as deviant in conjunction
with an overly exclusive category for defining praiseworthy behaviors, might be a
critical factor in accounting for the inaccurate parental perceptions of the child’s
overall behavior that was discussed above, most striking in Rickard’s et al. (1981)
Clinic non-deviant group.

Even though the data are limited, there is some support for the deviancy-
categorization hypothesis. Lorber (1981) had the mothers of conduct disordered
children and the mothers of normal children view videotapes of parent-child
interactions. The mothers were asked to push different buttons to identify events as
prosocial or antisocial. While viewing the identical segments of videotape
interactions, mothers of aggressive children indicated that the stimulus child they were
viewing emitted a significantly greater number of antisocial behaviors than did

mothers of the normal children. This result would be in keeping with the hypothesis

50

that these mothers are overly inclusive in the categorization of deviant behavior and
thus see more deviancy. Patterson (1986b) reported that these mothers of antisocial
children classified many behaviors as deviant that were not classified as deviant by
trained observers. The mothers from the two samples, however, did not differ in
terms of the accuracy (determined a priori by trained behavioral observers) of
detecting aversive or prosocial child behavior.

Schmaling and Patterson (1984) employed a similar tracking paradigm but did
not find differences in labeling processes between mothers of problem and
nonproblem children. It is important to note, however, that these mothers were
exposed to an initial videotape reflecting either positive or negative parent-chfld
interactions. These investigators suggested that since the mothers were selectively
reactive to child behavior dependent upon the type of scene they viewed first, this
may account for the lack of differences in the labeling processes. Due to the initial
exposure the mothers were predisposed to respond to the ambiguous child behaviors
in accordance to the prevailing valence of the initial scene.

Rather than using a videotape apparatus, Holleran et al. (1982b) employed a
paper—and-pencil measure of parental labeling and tracking. They presented the
parents with a script containing 10 short vignettes made up of a determined amount
of prosocial, deviant, and neutral child behavior. The parents were asked to
differentially underline events they categorized as prosocial or antisocial child
behaviors. This did prove to be a useful paper-and-pencil method for evaluating the

parents’ tracking behavior. These investigators, using a "signal detection format",

51

found that the parents of normal children were significantly more sensitive to
prosocial behavior than were the parents of problem children. (Sensitivity was really
agreement with the classification of behavior segments made by the experimenter.)
This would support the contention that parents of antisocial children are overly
exclusive in their categorization of prosocial behaviors. A second significant finding
was found on the variable for identification of negative behaviors as positive, as
parents of normal children were more prone to this error. Finally, in a pre-
publication draft of the research (Holleran et al., 1982a) the investigators reported
that they asked the parents of the two samples to rate statements about deviant child
behavior on their aversiveness. The parents of the antisocial children rated the
various behaviors as significantly more aversive than did the normal parents.
Finally, the studies just reviewed did attempt to investigate whether the
parents’ tracking and categorization behaviors measured in the laboratory were related
to the parents’ observed behaviors in the home (Holleran et al. , 1982b; Lorber, 1981;
Schmaling & Patterson, 1984). There were a variety of significant correlations
between the various scores from the laboratory tasks and observational data of the
mothers’ interactions with their children at home. For example, Schmaling and
Patterson (1984) reported tendencies to label neutral behavior as positive on the
analog videotape task were significantly correlated with the proportion of observed
maternal behavior that was positive. Also, Holleran et al. (1982b) reported that lower
sensitivity to negative behaviors on the tracking task was associated with higher rates

of positive behaviors in the home. They suggested that the normal parents’

52

" insensitivity" to negative behavior may, in some way, be responsible for the positive
social atmosphere in the home. (It was suggested previously that sometimes the most
appropriate response to negative child behavior is no response!)

Overall, the results do hold promise. For present purposes, the paper—and-
pencil tracking instrument appears sensitive to detect differences in parental perceptual
styles and the various correlational findings with the home observation data does offer
some support for the external validity and generalizability of this laboratory
perception assessment paradigm. The fact that the parents of conduct disordered
children rate deviant behavior as significantly more aversive is also an interesting
finding. It does appear that the deviancy-categorization hypothesis is supported and
seems worthy of further investigation. Additional data, especially from another
research setting, would be valuable. Also, the specific sub-type of conduct disorder
that makes up the sample in Holleran et al. (1982b) is not clearly specified. A
worthwhile test of the overinclusivelunderinclusive categorization hypothesis would
be to specifically examine the hypothesized processes in the parents of children with
an Undersocialized Aggressive Conduct Disorder, focusing on the number of neutral
behavioral segments that are judged to be positive or negative, as well as the deviant
and prosocial behaviors that are inaccurately classified. Finally, it would be
interesting to examine whether parents differ not only in terms of their ratings of

aversiveness, but also in their ratings of the desirability of specified child behaviors.

53
W

The present investigation proposes that the parents’ present parenting style,
specifically the parental cognitive factors involved in their deviancy-categorization
schemata/processes, will be associated with the parents’ recollections regarding their
own childhood. The rationale for this position will be developed in this final section.

It will be helpful to review what has been established in the previous sections.
Clearly, there are a variety of parental qualities that are associated with the child’s
conduct disturbance. Coercion theory helps to provide a conceptual basis for grasping
how the parent-child interactions foster the development of antisocial behavior. The
parent’s frequent and inept attempts at family management, including discipline, in
fact "trains" the child along the developmental path of disturbed conduct. The child’s
increasing behavioral disturbance only exacerbates the parent’s ineffectiveness.
Paradoxically, the parental attempts to eliminate certain behaviors only serves to
fortify them.

These parents have been characterized as "nattering" or nagging, negativistic,
critical, and controlling. It is suggested that these parents do not always view the
child and his/her behavior accurately and other factors beyond the child’s actual
performance are involved in the perceptual deficits. Clearly, the parental
characterization and labeling of the child must strongly influence the child and the
overall quality of the parent-cth interactions. The present study will attempt to
provide further support for the deviancy categorization hypothesis, namely, that the

parents of children with an Undersocialized Conduct disturbance are overly inclusive

54

in their classification schema of child deviancy and relatedly, underinclusive, (or
rather overly exclusive), in their classification schema of desirable/prosocial
behaviors.

On the assumption that this deviancy-categorization process is in fact taking
place, it is reasonable to suggest that the inaccurate classification of the child’s
behavior sets into motion an elaborate process. Patterson (1986b) reported that
parent’s overinclusiveness for deviant behaviors correlated with the parent’s
disposition to scold and nag during home observations. The parents perceive the
child negatively and there is consistent evidence supporting an association between
parental negativity and child deviancy, even within normal samples. Constant
inaccurate tracking and labeling of the child’s behavior can easily lead to an
inappropriate, chaotic, and ineffective management of the child’s behavior. It should
come as no surprise that the child appears non-responsive to parental control, and in
fact parental attempts to discipline can accelerate the frequency of the child’s deviant
behaviors. As these parent-child interactions increase in frequency and
aversiveness/intensity, the parental perception and overall negativity towards the child
appears justified and it seems likely that the child’s behavior and self-concept will
inevitably be altered/modified (Patterson, 1986b). The fact that these children do
exhibit a greater level of deviancy is possibly the outcome of a positive feedback
100p, or the child’s attempt to live up to the perceived parental "expectations" (i.e.,

to act in accordance with how their parents perceive them).

55

It should be clear that parental cognitive factors play a critical role in
determining the parent-child interactions and these parent-child interactions appear to
be consistently associated with the child’s conduct disturbance. Also, these parental
cognitive variables play an important role in the parents’ deficient family management
skills. In this final section, the hypotheses regarding the determinants of the parental
behaviors will be developed. Why is it that these parents lack the necessary skills
involved in effective parenting? Specifically: Why is it that these parents are
overinclusive in their definition of deviancy and underinclusive in their definition of
prosocial child behaviors?

Even though Patterson suggested that inept family management is the result of
faulty training, he is of the opinion that the parents’ negative cognitive style and their
deviancy-categorization schema are very likely the result of living with a deviant child
and being constantly exposed to deviant behaviors (Patterson, 1986b). The position
of the present study, however, is that the parent arrives at the parenting situation with
a predisposition to respond in a negative manner. The parents’ cognitive processes
can be characterized by an anticipatory response set geared towards responding in a
critical or negative fashion and relatedly, there is a reluctance or hesitancy to respond
in a praiseworthy or complimentary mode. Therefore, while Patterson would suggest
that this response set originates through the actual interaction history with the child,
the present study is suggesting that the propensity to respond in a generalized negative

manner is acquired prior to the initiation of parent-cth interactions.

56

It is the position of the present study that the parents’ response proclivity is the
product of their representational model regarding parent-cth interactions and further,
that this mental model is based largely on the parents’ own childhood experiences.
Belsky (1984) has recommended that the parent’s ontogenic origins and psychological
resources may be an important influence in determining the quality of the parental
functioning. He reviewed literature that supports the contention:

.. .that developmental history shapes personality and psychological well-being,

which in turn influences parental functioning. Indeed, a hypothesis that I

advance is that, in general, supportive developmental experiences give rise to

a mature healthy personality, that is then capable of providing sensitive

parental care which fosters optimal child development. (Belsky, 1984, p. 86)

The present study proposes that parents of antisocial children had childhood
experiences that were lacking, to varying degrees, in these "supportive developmental
experiences", and these experiences are reflected in the functional organization of
their mental model regarding parent-child interactions. Thus, it is hypothesized that
the parents of antisocial children have an affective-cognitive model regarding parent-
child relationships that is organized in a defensive or negative orientation, (i.e., it is
oriented towards limiting and/or distorting the access or availability of specific classes
of information).

Underlying Belsky’s (1984) position is a widely accepted assumption,
regardless of theoretical orientation, that one’s childhood experiences can affect later

adult functioning, including parental functioning. For example, it is a generally held

puma:
lacing
mg
1936:.
SUCK“:
general;

P133213:

313mm

migrat-

57

position that parents are ineffective in family management skills because they are
lacking in effective role models and simply were not exposed to beneficial child-
rearing practices (Patterson, l986a, 1986b; Robinson, 1985; Wahler & Dumas,
1986). It is now clearly established that there is an inter-generational continuity to
coercive, antisocial behaviors as these "behavior patterns are passed on from one
generation to the next at a rate well beyond chance" (Wahler & Dumas, 1986, p. 50).
Patterson (1986b) interpreted these data as supporting the idea that inept family
management is the result of poor training in family skills as ”...faulty parenting
practices are handed down from one generation to another" (p. 440).

"While the modeling principles have intuitive appeal, the processes through
which a child comes to imitate a particular model remain poorly understood" (W ahler
& Dumas, 1986, p. 62). Adults, and to a limited extent children, have the cognitive
potential and capacity to remember, to reflect upon past experiences, to compare and
integrate past and current events and ultimately, to learn from these experiences. The
fact that a parent was subjected to coercive parenting as a child does not inherently
lead to the repetition of this parenting style with his/her own children. Rather,
exposure to a coercive model could very easily result in the parent’s determination
to acquire additional information and parenting skills. While childhood experiences
can be determinants of adult functioning, there can be a broad spectrum of effects
ranging from the blind repetition of relationship patterns to the conscious conviction

not to model or repeat the parents’ behavior.

1m

58

The present study is concerned with the ways in which parents have mentally
represented and organized their own childhood experiences and the resultant affective-
cognitive model regarding parent-child relationships. Bugental and Shennum (1984)
pointed out that adults come to the parenting situation with a history of social
experiences with children and, at the very least, they have been exposed to the model
of their own parents. They suggested that these experiences are encoded as causal
beliefs about the forces operating in such relationships. The point is that the adult has
a pre-established organized and integrated conceptual system of these experiences
(i.e. , of parent-child interactions).

A few comments regarding the properties of this parent-child mental model are
now in order. First, childhood experiences exist only to the extent that they are given
meaning. In other words, the experiences of a child are likely to be important only
in relation to the way in which the child conceptualizes his/her experiences (Crowell
& Feldman, 1988; Hinde, 1980). The child’s experiences are internally represented
within a coherent system, referred to here as a mental (or affective-cognitive) model
of parent-child relationships. It is within this organization or conceptual schema that
the experiences attain their meaning. Secondly, within this representational system,
the parent is not encoded solely as a role-model-to-be-imitated, but rather it is the
experiences or the relationship between parent and child that are mentally represented
(Sroufe & Fleeson, 1986); and this would include both affective and cognitive
components. Finally, the mental model is dynamic and not static. The child’s

experiences are organized schematically not categorically (i.e. , they are not encoded

59

as merely static "pictures" but rather experienced events of actions and reactions).
It is in this way that the model is not perceived as a static entity. Rather, the parent-
child mental model is most helpfully conceived as structured processes serving to
obtain or limit access to relevant information (Main, Kaplan, & Cassidy, 1985).

It is the goal of this study to explore how this mental or representational model
is organized. Is it a model that exhibits a degree of integration and understanding
regarding the parents’ own childhood experiences? Or, rather, is it a mental model
that is defensive in its structure? In other words, is the model prone to distort,
restrict, and/or inhibit certain classes of information? Furthermore, another goal of
this study is to examine the relationship between the organization of this mental model
and the parents’ current behavior, specifically their perceptual style. The present
study hypothesizes that there will be a relationship between the degree of
defensiveness of the parents’ recollections regarding their own childhood and their
current deviancy-categorization processes. It is proposed that parents with a defensive
orientation towards their own childhood will be more prone to exhibit a negative
perceptual style towards their own children (i.e., they will exhibit an overly inclusive
definition of deviancy and an overly exclusive definition of prosocial behavior).

It is incumbent upon this presentation to address one final issue, namely: How
is it that the parents’ defensive orientation towards their own childhood would be
related to their current deviancy-categorization behavior? What is the link between
the parents’ defensiveness regarding their own childhood experiences and the apparent

negative perceptual style directed towards their own child?

Brena

“Ma-9
lI’LEl

‘&
V»

g... .
‘3‘ h

60
W. In order to approach this critical issue it will be essential

to direct attention to the current status of attachment theory (Bretherton & Waters,
1985) as recent advances, both conceptually and methodologically, are extremely
apropos for the present study. For example, recent reconceptualizations have focused
on the representational level of attachment relationships (Main, Kaplan, & Cassidy,
1985). Thus, there is an emphasis on the role of cognitive/mental structures in
parent-child interactions. Additionally, there is an emerging body of research to
suggest that retrospective accounts of the mother’s childhood are associated with the
quality of current parental behavior (Crowell & Feldman, 1988; Grossman, Fremmer-
Bombik, Rudolph, & Grossman, 1988; Main et a1. 1985 ; Main & Goldwyn, 1984;
Morris, 1981; Ricks, 1985). It is the goal of this section to clearly articulate the
conceptual logic underlying these results. This will be accomplished mostly by
detailing Mary Main’s research (Main & Goldwyn, 1984; Main, et al., 1985).
While the present study in not focused on attachment theory per se, recent
drinking in this area does provide a conceptual foundation for the present
investigation, as well as corresponding methodological support. Thus, it will be
helpful to briefly review attachment theory. Attachment theory, as presented by John
Bowlby (1973, 1980, 1982) is concerned with a goal-corrected motivational-
behavioral control system involved in the infants’ proximity and security-seeking
interactions with the mother or attachment figure. According the Bretherton (1985),

the term "attachment system":

61

...refers to a psychological organization hypothesized to exist within a person.

This system is so constituted that feelings of security and actual conditions of

safety are highly correlated, although the correlation is by no means perfect.

Seen from an outside observer’s viewpoint the system’s set-goal is to regulate

behaviors designed to maintain or obtain proximity to and contact with a

discriminated person or persons referred to as the attachment figure(s). From

the psychological vantage point of the attached person, however, the system’s

set-goal is felt security. (p. 6)

The assessment of the infant’s attachment system is conducted in the Strange
Situation (Ainsworth, 1972) a laboratory paradigm of separation and reunion episodes
between mother and infant. On the basis Of the observed infant behavior towards the
mother during reunion, the infant’s attachment behavior can be classified as: (a)
Securely attached: these infants approach the mother and seek physical contact with
her if they had been distressed by the separation, or if they had not become
distressed, they still greeted the mother and seek interaction and/or contact; (b) Non-
attached or avoidant: these infants avoid the mother on her retum, or (c) Insecure
attached: these infants exhibit angry resistent behavior mixed with seeking proximity
and contact. The assessment of attachment usually takes place when the infant is
between 12-18 months old.

Bretherton (1985) pointed out that the validity of the Strange Situation as an
assessment of the infant’s quality of attachment is derived primarily " . . .from the

systematic and extensive correlations of infant behavior in this situation with maternal

62
and infant behavior in the home throughout the first year of life" (p.15). In fact,

even though the Strange Situation does not directly assess the parent’s behavior,
Morris (1981) pointed out that the correlations with maternal behavior are so
consistent that researchers frequently equate infant classification with specific
characteristics of the parent’s caretaking behaviors, implying that there are parallel
individual differences in the parents. It appears that "the main variable that
discriminates mothers of securely attached babies from mothers of anxiously attached
babies is sensitive responsiveness to the infant’s signals" (Morris, 1981 , p.312).
Mothers of securely attached infants have been independently judged to be high on
maternal warmth, affectionate behaviors, and maternal sensitivity (Bretherton, 1985;
Sroufe & Fleeson, 1986). On the other hand, insecurely attached and/or avoidant
infants have mothers that were independently judged as rejecting and/or inconsistent
to the infant’s bids for physical contact (Bretherton, 1985; Main & Goldwyn, 1984).

This association between the infant’s avoidance and the mother’s rejection is
of critical importance in terms of the present study. While reviewing the empirical
literature on family/parental characteristics of families with a conduct disturbed child,
it was consistently shown that the parent’s rejecting and negative attitude towards the
child was associated with the child’s deviant behavior. It is beyond the scope of this
paper to explore the implication that perhaps antisocial children were, in fact, rejected
infants in regards to their attachment behavior, suggesting a developmental continuum
with similar underlying processes. It is interesting to note, however, that infants

avoiding their mothers at 12-18 months demonstrated a similar pattern of behavior at

any

i.

m

3:;

63
six years-old (Main et al., 1985). Furthermore, in a follow-up study, Sroufe (1983)

reported a number of aversive behavior patterns in 4-6 years—old children associated
with early avoidance of the mother. Maternal rejection appears to be related to the
development of unpredictable bouts of hostility, avoidance, and a lack of empathy for
the distress of others (Main & Goldwyn, 1984). It is as if rejection (by the mother)
fosters rejection (by the infant/toddler). Is this the same as aversiveness fostering
aversiveness as proposed within coercion theory?

While Patterson might account for the infant’s avoidance and the mother’s
negativity, both examples of aversive behaviors, via a negative reinforcement
paradigm, attachment theory offers a disparate point of view. Within attachment
theory it is suggested that the infant’s avoidance of the mother is a self-protective and
therefore adaptive response. It is presumed that the infant initiates physical
contact/proximity (i.e., attachment behavior) when under stress and/or perceived
threat, and this is alleviated by the attachment figures responsiveness. When the
mother is not responsive, however, the infant’s attachment system is continually
activated without termination. Bowlby proposed that continual exposure to a rejecting
mother may lead, for self-protective or defensive purposes, to a deactivation of the
attachment behavioral system. According to Morris (1981), the repeated rebuffs by
the rejecting mothers lead to automatic withdrawal and avoidance behavior in
situations that usually arouse attachment behavior in secure infants. Morris explained

the automatic nature of the avoidance as "...skipping over the phase of intensified

64

proximity and contact-seeking that the infant " knows " would lead to rejection" (p.
313).

Main et al. (1985) have proposed a reconceptualization of attachment
processes, suggesting a shift from the behavioral level, as exemplified in the Strange
Situation, to an emphasis on the representational level. It is this new focus on the
representational processes and the accompanying research that is of direct relevance
for the present study. This reconceptualization promotes research strategies regarding
adult attachment with a focus on representation and language. Main et al. (1985)
suggested that differences in attachment relationships relate to individual differences
in mental representation (i.e. , the individual’s " internal working model" of
attachment). (This parallels the parent’s mental model regarding parent-child
relationships.) They defined the internal working model of attachment as :

...a set of conscious and/or unconscious rules for the organization of

information relevant to attachment and for obtaining or limiting access to that

information, that is, to information regarding attachment-related experiences,
feelings and ideations.

...[these are] models that direct not only feelings and behavior but also

attention, memory, and cognition, insofar as these relate directly or indirectly

to attachment. Individual differences in these internal working models will
therefore be related not only to individual differences in patterns of nonverbal

behavior but also to patterns of language and structures of mind. (p. 67)

65
The infant’s avoidance of the mother takes on added meaning from this

perspective or reconceptualization. The infant’s avoidance is seen as an organized
shift of attention away from all information relevant to attachment. According to
Main and Goldwyn (1984):

In looking away from and ignoring the attachment figure in stress situations,

the rejected infant may be seen as acting to exclude "attachment-relevant-

information" from firrther processing. Presumably the infant is avoiding the
pain which has been associated with repeated experiences of the rejection of

her/his attachment. (p. 210)

It is reasonable to speculate that the infant’s approach-avoidance conflict (i.e. ,
seeking proximity and avoiding nonresponsiveness or rejection) can be quite
disorganizing and unsettling. Hence, the shift of attention away from the attachment
figure, including potentially disturbing affects, may help in the maintenance of a self-
organization or cohesiveness (Main & Weston, 1982).

It is interesting to note that this emphasis on the affective-cognitive
organization of attachment behavior provides for an accounting of intergenerational
continuity in attachment. It is hypothesized that the internal working model of the
rejected child will be organized in such a manner as to shift attention away from
attachment experiences and the feelings involved in those experiences. When the
rejected child is a parent, however, this organization of their internal working model
would result in a similar nonresponsiveness towards their own offspring, since

attachment relevant information will be excluded. It is in this way that there can be

66

intergenerational continuity. The parents have developed information-processin g or
representational patterns which result in the exclusion and distortion of information,
and the subsequent nonresponsiveness to the child’s attachment seeking behavior.
The child then institutes Similar avoidant and self-protective affective-cognitive
representational processes and the pattern of rejection continues to the next
generation. The research to be reviewed below is generally supportive of this
speculation regarding the intergenerational continuity in attachment.

There is some empirical support for this emphasis on representational
processes. Main et al. (1985) have examined the attachment behavior and the
"internal working models" of attachment in a group of six year-olds and their parents.
This group had been previously assessed in the Strange Situation when the same
children were infants. The reunion behaviors of the six year-olds (i.e. , the security
of their attachment) was significantly correlated with their security of attachment
assessed during infancy.

Furthermore, a variety of measures designed to assess the six-year olds’
attachment on a representational level were significantly associated with the child’s
attachment classification made during infancy. For example, the child’s secure versus
insecure responses to a family photograph were compared to an independently derived
classification of the attachment behavior during infancy. Main et al. (1985) pointed
out that "since the photograph could not respond to or control the child’s response,
the child’s response to the photograph could not simply be an example of an

interaction pattern being maintained over several years between child and parent"

67

(p.78). Thus, it is not actual behavioral interactions as in the Strange Situation, but
rather some mental representation or mental model that must account for differences
in the child’s responses to pictures of their families. Overall, the results were
consistently supportive of the reconceptualization emphasizing the representational
processes and the value of the " internal working model" construct.
W5. Research is now emerging to support the position that
a mother’s retrospective account of her own childhood is significantly related to her
current parental behaviors (Crowell & Feldman, 1988; Grossman et al., 1988; Main
et al., 1985; Main&Goldwyn, 1984; Morris, 1981; Ricks, 1985). The recollections
of childhood attachments, as assessed by either interviews (Crowell & Feldman,
1988; Grossman et al., 1988; Main & Goldwyn, 1984; Morris 1981) or a
questionnaire (Ricks, 1985) appear to have a significant relationship to the mother’s
ability to serve as a secure base for her child. "The evidence reviewed compellingly
indicates that the quality of a mother’s caregiving behavior is strongly related to her
memories of childhood relationships" (Ricks, 1985, p.226). It is not only the specific
content of the childhood memories but also the organizational properties of these
memories that are related to current caregiving behaviors (Crowell & Feldman, 1988;
Grossman et al., 1988; Main et al., 1985; Main & Goldwyn, 1984; Ricks, 1985).
The present study will examine the relationship between parental recollections
of their childhood and their current perceptual style of child behavior (i.e., their
deviancy-categorization processes). The recent studies mentioned above begin to

offer support for the position that parental childhood memories are related to the

68

parents’ current behaviors. The specific link between these memories and current
behavior is provided conceptually via the construct of "internal working model". The
parental recollections are taken as representative of their mental model regarding
attachment behavior. It is hypothesized that adults who were subjected to parental
rejection during childhood will have developed representational processes organized
in a manner which shifts attention away from receiving information that would
activate these emotionally painful experiences. This limiting of the access/availability
of specific classes of information would then account for the parents’ apparent lack
of responsiveness to their own children. The parental nonresponsiveness on a
behavioral level corresponds with the limiting of access to specific information on the
representational level.

Specific support for this position was initially provided by Mary Main’s
research on adult recollections of childhood experiences (Main et al., 1985; Main &
Goldwyn, 1984) and has subsequently been further supported in the literature
(Crowell & Feldman, 1988; Grossman et al., 1988). The design of the present study
is largely based, both conceptually and methodologically, on Main’s research.
Therefore, it will be essential to review this research in detail.

In order to elicit the parents’ childhood recollections and assess their internal
working model of attachment, the Berkeley Adult Attachment Interview protocol was
developed (George, Kaplan, & Main, 1984). This structured interview was designed
"to probe alternately for descriptions of relationships, specific supportive memories,

contradictory memories, assessments of relationships in childhood and current

w
11

Oil,

31131"

111116

it)

. .‘Ir
\1
‘ I
‘3: r7.
M4

a: l‘ u

“it‘d

69

assessments of the same experiences and relationships" (Main & Goldwyn, 1984, p.
211). (A questionnaire format of this interview is used in the present study.) The
interview protocols were initially scored for the "apparent rejection" of the mother
in her own childhood. "The rating was based not only upon the parent’s direct
response to questions regarding rejection, but also upon the way in which these
responses were integrated into the rest of what the parent said or remembered about
her experiences (Main & Goldwyn, 1984, p. 212).

Additionally, the protocols were also scored for quality of the recollections
and/or the degree of distortions or defensiveness. These measures were: (a)
Idealization of mother: the extent to which the subject seemed to cover up or idealize
an apparently neglecting, interfering, or rejecting mother; (b) Anger with mother
now: this rating was for conscious, direct feelings or expressions of anger towards the
mother; (c) Inability to recall childhood: this was simply the frequency with which
the subject insisted that she could not remember her childhood; and (d) Coherency of
the transcript: this rated the mother’s ability to talk about her past and evaluate its
influence coherently. Incoherent protocols exhibited unrecognized contradictions
and/or the subject frequently lost track of the topic, with interruptions in the flow of
the interview conversation.

The results indicated that "apparent rejection" in childhood was strongly
associated with a defensive orientation toward childhood memories. Apparent
rejection was positively correlated with an inability to recall, an idealization of the

parent (similar to Ricks, 1985) and negatively correlated with the coherence of the

prom;

R1321}

70
protocol (i.e. , the less coherent the protocol the greater the likelihood of apparent

rejection). The salient cognitive properties of the childhood recollections of mothers
that were apparently rejected can be characterized by the distortion of information
(i.e. , idealization) the disorganizing and splintering of information (i.e. , incoherent
transcript) and the exclusion of information (i.e. , inability to recall). As predicted,
the mothers who were apparently rejected during infancy/childhood did in fact have
a mental model that was defensively organized to block access or shift attention away
from attachment related information.

Furthermore, the mother’s "apparent rejection" by her mother was strongly
associated with her own infant’s avoidance, as assessed five years earlier (Main &
Goldwyn, 1984). On the basis of the consistently strong correlations reported in the
literature between the infant’s avoidance of the mother in the Strange Situation and
the mother’s independently observed rejection of the infant (e.g., r > .80), Main and
Goldwyn inferred that there is a strong tendency for mothers who have been rejected
by their mothers to reject their own children. Additionally, Main et al. (1985)
reported that the interviews were rated for the parents’ security with respect to
experiences, ideas, and feelings surrounding attachment. There was a strong
correlation between the infant’s security of attachment, assessed five years prior, and
the implicit security in the mother’s internal working model of attachment. According

to Main and Goldwyn (1984), the overall results indicated:

If

m

.9 .-A
77!"-
..
_ V

b,"

71

...a systematic relationship between a woman’s (often indeliberate or
unconscious) representation of her own mother as rejecting during the
Berkeley Adult Attachment Interview and her evident rejection of her own
infant. We find that a mother’s rejection of her own infant is related to
cognitive processes which seem to distort and attempt to negate her own early
experiences of maternal rejection" (p. 205).

In other words, it is reasonable to interpret the data as suggesting that the
mother’s apparent failure to integrate her own childhood experiences, as demonstrated
in her defensively organized internal working model, is a critical determinant of
parental rejection and the corresponding infant’s avoidance. For example, the
mother’s inability to recall her childhood and her idealization of her parents was
significantly related to her own infant’s likelihood to avoid her. On the other hand,
there was a significant negative relationship between the coherence of the mother’s
recollections and the infant’s observed avoidance. Further, the mother’s ability to
directly experience anger and resentment, rather than a defensive idealization, was
also significantly associated with a more securely attached infant. No matter what the
mother’s childhood experiences, if she could speak freely and coherently of her
feelings and experiences regarding her own childhood, then her infant was more
likely to be securely attached and not avoidant. Grossman et al. (1988) reported
strikingly similar findings.

It is therefore reasonable to propose that the mother’s ability to integrate and

differentiate or separate herself as an adult from her experiences as a child is a

33th
flpene

Mia

72

critically important characteristic in determining the quality of current parenting
behaviors. To the extent that this integration has taken place, and the parents have
"come to terms" with their childhood experiences, then there would be less of a self-
protective need to distort and block potentially pain-evoking or unsettling attachment
relevant information. In fact, Main and Goldwyn (1985) did describe one mother
whose recollections of her unhappy childhood were "exceptionally coherent" . This
mother had not idealized her rejecting parents, nor had she blocked from her
awareness childhood experiences. Rather, she had come to terms with the
experiences and had genuinely understood and forgiven her parents. This mother had
the least avoidant infant in the sample.

Clearly, the data support the position that it is not the overall childhood
experience but rather the degree to which it has been understood and integrated that
is related to the parent’s caregiving behavior. This is of pivotal importance for the
present study. It is not solely the quality of the childhood relationships per se, but
rather the ways in which these relationships are cognitively represented and organized
that are related to current parental behaviors.

To summarize, where the parent’s own experiences and feelings are not

integrated, restrictions of varying types are placed on attention and the flow

of information with respect to attachment. These restrictions appear in speech
in the form of incoherences and in behavior as insensitivities. (Main et al.,

1985, p. 100)

1':th

rug.

3,9" 3111

Milan:
Italic,-
&13110
E31113

k .
-fiflqh
{‘“sd‘l

73

Thus, there is support for the position that the defensive orientation of the
parent’s internal working model regarding attachment does limit access to attachment
relevant information, thereby affecting parental responsiveness to the infant and the
overall quality of the parent-child interactions. Crowell and Feldman (1988) offered
support for this postion in an older sample of toddlers, ages 2-5 years-old. They
reported an association between the mother’s internal models of relationships and the
quality of the mother-child interactions observed during a semistructured laboratory
play and problem-solving session.

Sum. In summation, the current conceptualizations and research in
attachment theory offer support for the basic tenets of the present study. Parental
recollections regarding childhood are related to the parents’ current caregiving
behaviors. The conceptual bridge between these parental memories and current
parental behaviors is the construct of an internal mental model or structured/organized
affective-cognitive processes that guide and determine, to a large extent, the quality
of the parent-child interactions. The organizational properties of this parent-child
mental structure appear related to the degree to which the childhood experiences were
conflictual and emotionally problematic. Thus, this mental structure can be
organized, for apparently self-protective reasons, in such a manner as to actively shift
attention away from certain classes of information. This can result in the limiting of
the flow and access of information or rather in a disruption in the organization of the
information being processed. On a behavioral level, this can be evidenced as

rejection and/or nonresponsiveness.

rarer
2th
1m
rarer:
rat

mater

ltd :n

74

The present study is proposing that the negativity that characterizes the parents
of antisocial children is a parallel of the rejection and nonresponsiveness that
characterizes the mothers of avoidant infants. The present study is examining the
parents’ negative perceptual style (i.e. , as reflected in their hypothesized deviancy-
categorization schema) as one aspect of this overall negativity. It is hypothesized that
the parental negativity is the result of a defensively organized mental model regarding
parent-child relationships. It is assumed that the affective-cognitive processes are
organized to shift attention away from information regarding parent-child
relationships, in order to avoid potentially pain-evoking information. The negativity
and criticalness of parents with a conduct disturbed child, as consistently observed in
the empirical literature, is assumed to coincide on an affective-cognitive level with the
exclusion and distortion in the processing of information relevant to parent-child
relationships.

Specifically, it is hypothesized that the parent’s deviancy-categorization schema
is based on an attentional selectivity. An overly exclusive classification schema of
prosocial behaviors is interpreted as a selective inattentiveness, or a Shifting of
attention away from this class of information. This results in the parent’s
nonresponsiveness, rejection, and even irritability or negativity (e.g., Patterson’s
"crossover") to the child’s prosocial behaviors. In other words, the lack of
responsiveness of the mother to the infant’s proximity and security seeking is assumed
to be a parallel to the parent’s reluctance to define and thus respond positively to the

child’s prosocial behaviors. Prosocial behaviors are affiliative in nature and

75

frequently promote praise, affection, and emotional closeness. It is in this way that
prosocial behaviors can be seen as relationship oriented, or as "attachment relevant
information". And, the parents’ overly exclusive schema for the classification of
prosocial behaviors can be interpreted as a Shifting of attention away from this
information. It is hypothesized that the parent’s defensively organized affective-
cognitive model regarding parent-cth relationships will be related to the parent’s
reluctance or hesitancy to define child behavior as praiseworthy or prosocial.

On the assumption that the parental classification schemata regarding deviant
and prosocial child behaviors are not independent, then the parents’ overly exclusive
categories for classifying prosocial behaviors will be related to an overly inclusive
category for defining deviant behaviors. The parent’s hypothesized shift of attention
away from attachment related information and the subsequent overly exclusive
classification schema for prosocial behaviors, results in an emphasis on classifying the
child’s behavior as deviant. Hence, the selective inattention or nonresponsiveness to
prosocial behaviors, coexist with a selective attentiveness to deviant behaviors. It is
hypothesized that the parents’ defensively organized mental model regarding parent-
child relationships will be related to their willingness/eagemess to define child
behavior as deviant.

Macs

There are three groups of hypotheses. The first group, hypotheses 1-7, is

about the mothers’ perception of child behavior. The second group, hypotheses 8-10,

is focussed on the mothers’ recollections of their own childhood experiences. The

76

third group, consisting of hypothesis 11, will examine if there is a relationship

between the mothers’ perception of child behavior and the affective-cognitive

organizational characteristics of their childhood memories.

Group I, The first group of hypotheses is about the mothers’ perception of

child behavior. The first two hypotheses are specifically focussed on the deviancy

categorization processes and will explore if the mothers’ of conduct-disturbed boys

(MCDB) differ from the mothers’ of normal boys (MNB) in their definition and

classification of deviant, neutral, and prosocial behaviors. The hypotheses will

examine if the MCDB both label and track deviant and prosocial behaviors differently

than the MNB. (The data are from the CBS, CBRSI, and CBRSII.)

Hmthesis 1.

The MCDB will be overly inclusive in their
categorization of deviancy and overly exclusive (or
underinclusive) in their categorization of prosocial
behavior. The MCDB will have a greater tendency to
perceive behaviors as deviant and will be less prone to
define and identify behaviors as prosocial. Thus, as
compared to the MNB, the MCDB will: (a) correctly
identify a significantly greater number of deviant
behaviors, (b) will incorrectly identify a significantly
greater number of positive behaviors as negative, (c) will
identify a significantly greater number of neutral

behaviors as negative, (d) will have a "negative response

77

tendency", (i.e., out of all the behaviors identified, for
each individual mother, a significantly greater proportion
will be negative), and (e) the MCDB overemphasis on
deviancy will also be reflected in their ratings of
behaviors as significantly more aversive.

The MNB will be more likely to perceive behaviors as
prosocial and less likely to define and identify behaviors
as deviant. These mothers will be overly inclusive in
their categorization of prosocial behaviors and
underinclusive (or exclusive) in their categorization of
deviant behaviors. Thus, the MNB will (a) correctly
identify a greater number of prosocial behaviors, (b) will
incorrectly identify a greater number of negative
behaviors as positive, (c) will identify a greater number
of neutral behaviors as positive, (d) will have a "positive
response tendency" (i.e., out of all the behaviors
identified, for each individual mother, a significantly
greater proportion will be positive), and (e) the MNB
emphasis on prosocial behaviors will also be reflected in

their ratings of behaviors as significantly more desirable.

78
The third hypothesis deals with the issues of whether there is a "behavior

classification covariation" where an emphasis on deviancy would inherently result in
a de-emphasis on prosocial behaviors and vice versa.

WILL A mother’s emphasis in classifying behaviors as deviant
will be related to a de-emphasis in classifying behaviors
as prosocial and vice-versa. Hence, an over-
inclusiveness in the categorization of deviancy will be
related to an over-exclusiveness in the categorization of
prosocial behaviors.

The final hypotheses within this group will explore the association between the
family’s level of enmeshment and the mother’s perceptual style, as well as examine
related issues concerning the mother’s degree of engagement or involvement.
Patterson (1982, 1986b) clearly characterized the MCDB as being overly enmeshed
and very irritable. He described these mothers as being easily provoked and thus
becoming quickly involved and engaged with the child. This high level of
engagement or reactivity combined with the mother’s tendency to be overly inclusive
in the classification of deviancy (thus frequently perceiving the child’s behavior as
worthy of reprimand) results in recurrent aversive and negative mother-chfld
interactions. (The data are from the CBS, CBRSI, CBRSII, and the FTQ.)

Hymmesis 4. The MCDB will describe their current families as
significantly more enmeshed than the description of the

MNB current families.

79
W. The MCDB will identify a significantly greater number

of behaviors, (i.e., both negative and positive).

H i 6. The level of enmeshment in the family will be
significantly related to the total number of behaviors
identified.

W. The level of enmeshment will be related to the mother’s
deviancy-classification schema.

519.111.11- The second group of hypotheses, hypotheses 8 through 10, will
examine the mother’s mental model regarding parent-child relationships, focussing on
the ways in which she has mentally represented and organized her own childhood
experiences. These predictions will explore not only the specific content of these
childhood experiences but also the affective-cognitive organizational properties of
these memories. (The hypotheses are based on the scores from the Adult Attachment
Questionnaire (AAQ), and this includes the Preliminary Questions.)

Hypothesis 8 will examine differences in the quality of the mother’s childhood
relationships/experiences.

H e is . The childhood recollections of the MCDB will exhibit

(both directly expressed and inferred) a significantly
greater amount of problematic and/or conflictual
emotional experiences in childhood than the recollections
of the MNB. The childhood recollections of the MCDB

will exhibit a greater amount of Unloving, Rejection,

80

Neglect, Role-reversal, Pressure to Achieve, and Abuse

than the recollections of the MNB.

Hypothesis 9 involves establishing a relationship between the defensiveness of

the mother’s recollections of her own childhood and the psychopathology of her child.

Do the MCDB have a mental model regarding parent-child relationships that is more

defensive in it’s orientation than the MNB?

W2.

The childhood recollections of the MCDB will be more
defensively organized (i.e., Incoherent, ldealized, Lack
of Detail (or specific recall), and Unresolved (or lack of
resolution and/or separation-autonomy) than the

recollections of the MNB.

Hypothesis 10 will examine if the quality of the childhood experiences are

related to the organizational properties of the maternal recollections. In other words:

Is there a relationship between the "apparent" childhood experiences and the mothers’

current state of mind (i.e. , the affective-cognitive organizational properties) regarding

these experiences?

Hypgmesis 0.

Problematic childhood experiences will be associated
with childhood memories that are currently defensively
organized. A defensive organization refers to the
tendency to shift attention away from specific
information regarding childhood relationships. Hence,

problematic childhood experiences will be related to the

81

mothers’ tendency to distort, (i.e., Idealize), restrict
(i.e., Lack of Detail) and splinter or disorganize, (i.e.,
Incoherent, and Unresolved) the information regarding

these childhood experiences.

QM. Hypothesis 11 represents the major thrust of the present study. It

is an attempt to establish a link between the mother’s current perceptual style and the

manner in which she has encoded or internally organized/represented her own

childhood experiences.

This hypothesis will explore if a mother’s deviancy

categorization processes, (i.e., her current perceptual style) are related to the

affecitve-cognitive organization of her childhood memories, (i.e., her mental model

of parent-chfld relationships).

HmthcsiLli.

Mothers who are defensive in their organization of
childhood memories (i.e., those that display a tendency
to distort, restrict, and/or splinter or disorganize the
information) will exhibit a corresponding need to shift
attention away from positive, prosocial behaviors as these
behaviors are assumed to be relationship oriented. In
other words, this defensiveness or shift of attention away
from relationship oriented information will result in a
selective inattentiveness to prosocial behaviors and a
subsequent over-emphasis (or selective attentiveness) to

deviant behaviors. Thus, a mother’s defensiveness

82
regarding her own childhood will be related to her

deviancy categorization processes (i.e., the greater the
defensiveness, the more likely the mother will exhibit a
categorization schema that is overly inclusive of deviant

behaviors and overly exclusive of prosocial behaviors).

CHAPTERZ

METHOD

Barium

Only mothers were selected as participants in this study. It is reasonable to
assume that mothers’ and fathers’ influences on a child’s development are not
identical. This assumption is clearly supported in regards to the development of a
child’s conduct disturbance (Hetherington & Martin, 1986; Patterson, 1982). For
example, there are differences between mothers and fathers in terms of their
perception of child deviance (Schaughency & Iahey, 1985) as well as differences in
their responses to deviant behaviors (Patterson, 1982). Since there is support for the
position that gender is an important characteristic involved in parenting (Frank,
Jacobson, Hole, Justkowski, & Huyck, 1986; Perry, Wells, & Doran, 1983) only
mothers were chosen.

This choice is consistent with previous research reviewed in the introductory
chapter. In the literature on conduct disorders (e.g., Anderson et al. , 1986; Forehand
et al., 1975; Griest et al., 1980; Rickard et al.,1981) and the literature on parental
perception (Furey & Forehand, 1986; Forehand et al., 1986; Rickard et al., 1984)
mothers have been the primary participants. In fact, some studies referred to

"parental" characteristics when actually they were examining only the mothers’

83

84
characteristics (e. g., Rickard et al., 1984). Additionally, the preliminary research

using the Adult Attachment Interview has been based primarily on mothers. While
some of the previous research on conduct disorders (e. g. , Olweus, 1980; Patterson,
1982; Snyder, 1977) and parental perception has utilized both parents as participants
(e.g., Christensen et al., 1983; Thompson & Bernal, 1982) in studies where only one
parent participated, it was always the mother. Perhaps the best reason for having
chosen mothers over fathers as participants in the present study is that the conceptual
foundation underlying the hypotheses emphasizes the mothers’ influence on conduct
disturbance and the coercive quality of the mother-son interactions (Hetherington &
Martin, 1986; Patterson, 1982).

The participants were 133 mothers of boys ranging in age between 6-11 years-
old. On the basis of the mother’s responses during the initial phone interview, the
mothers were divided into two groups:

u i r e B DB

The determination of a conduct disturbance was based primarily on each
mother’s description of her son. As a group, this determination was supported by the
sons’ scores on the Child Behavior Checklist (CBCL) (Achenbach & Edelbrock,
1983). There were 71 mothers that described their sons during the initial phone
interview as having behavior problems at home, either " fairly serious behavior

problems" or "mild behavior problems". Correspondingly, the sons’ scores on the

CBCL were, as a group, significantly higher than the comparable scores of the sons

85

in the "nonnal"/control group. (The CBCL scale scores and groups comparisons are
presented in Table 2).

Furthermore, the differences between the two groups on the Extemalizing
Scales of the CBCL, and specifically Scale VIII, Aggressive, are particularly relevant
in regards to the determination of a conduct disturbance. Scale VIII is the most
important scale given that the items on this scale have many similarities to Patterson’s
(1982) "Social Aggressors", and this type of conduct disturbance has also been
referred to an " overt antisocial" pattern (Loeber & Schmaling, 1985 ; Patterson, 1982)
since the child’s behavior is characterized by overtly confrontive antisocial acts which
primarily involve an abrasive interaction with another person. (Table 3 lists the
behaviors that make up these scales.) This pattern of conduct disturbance is
highlighted by behaviors that are basically noncompliant, argumentative, aversive, and
aggressive. For the MCDB, the sons’ mean score on the Aggressive Scale was 16.86
(SD = 7.70). Not only is this significantly higher than the control group, this mean

score is also in the 95m percentile on the normative standardized scale (Achenbach

& Edelbrock, 1983).
M rm B NB

The determination of the son as "normal" (i.e., non-conduct-disturbed) was
based primarily on the mother’s description of her son’s behavior. During the initial
phone contact, there were 62 mothers that described their son’s behavior to be no
more disruptive than the average boy. The mothers described their sons’ behavior

as either "what you’d expect for his age" or as "very well behaved." The

86

 

 

 

Table2
LJ '- ir 1' Alo.hs r . Me s .. . r . oomari it
Me; Alpha M DB MNBb 1. (49
M 5.12 Mu S.

Schizoid .57 2.30 2.11 1.60 1.51 2.21' (126)
Depressed .82 7.15 4.89 3.95 3.16 4.54" (121)
Uncom: .77 4.00 3.00 2.63 2.27 2.99” (128)
mumcatrve
Obsessive- .77 5.42 4.36 3.31 3.47 3.06" (131)
Compulsive
Somatic .63 1.48 1.80 .94 1.37 1.98’ (131)
Complaints
Social .70 3.39 2.63 1.89 1.86 3.86“ (125)
Withdrawal
Hyperactive .79 6.30 4.05 3.19 2.76 5.21” (124)
Aggressive .90 16.86 7.70 8.03 5.54 7.66" (126)
Delinquent .77 3.00 2.60 .82 1.18 6.34” (100)
Internal 20.35 12.96 12.42 8.82 4.17" (123)
External 26.15 12.44 12.04 7.83 7.92" (119)
Total 46.61 22.42 25.53 14.31 6.54" (120)
Problems

 

‘I_r=7l. b11:62.

’p < .05, two-tailed.

”p < .01 , two-tailed.

87

 

 

Table 3

031-1 ot-w-g‘c eExte 'in .- fth
CBCL
Erratum: A s iv Mm
Acts too young Argues Destroys own
Can’t concentrate Brags things

Hyperactive
Confused
Daydreams
Destroys own
things

Impulsive

Poor schoolwork
Clumsy

Prefers young kids
Speech problem

Cnrel to others
Demands attention
Disobeys at home
Disobeys at school
Poor peer
relations

Jealous

Fights

Lies, cheats
Unliked

Attacks people
Screams

Shows off
Stubborn

Moody

Sulks

Swearing

Excess talk
Teases

Temper
Threatens

Loud

Destroys others
things
Disobeys at school
Bad friends
Lies, cheats
Runs away
Sets fires
Steals at home
Steals outside
home
Swearing
Truant
Vandalism

88

mothers’ determination of their sons’ behaviors as within normal limits is
corroborated by the scores on the CBCL. As a group, each CBCL scale score (with
the exception of "depressed" which is in the high normal range) is within the lower
realms of the standardized normal range. Additionally, as mentioned above, Table
2 illustrates that every scale score for the "normal" boys is significantly lower than
those of the CBCL scores describing the conduct disturbed boys.
E l . j! . l l

The means and standard deviations of the demographic variables for both
groups are shown in Table 4. Descriptive data were obtained for the following
variables: the age of the sons, the age of the mothers and fathers, years of education
for both the mothers and the fathers, family income, and the number of children in
the family including the birth order of the son. As shown in Table 4 there are no
statistical differences between the two groups on any of these demographic variables.

Information was also obtained regarding the mother’s relationships to her son.
In over 95% of the cases, the participant was the "natural" mother of the son. There
were a few cases involving step-mothers and a few involving adoption, but there was
no systematic differences between the groups, X2 (2, N = 133) = 1.76, p > .05.
Step-mothers were included in the study as long as they had been living within the
reconstructed family unit for a minimum of 3 years. This criterion was established
in order to allow a sufficient amount of time for the step-mother to be involved as a

parent and have an influence on the boy’s development.

ID

 

 

 

Table 4
D hi V l r M us It Gr om risons
Variable: MCDB MNB L (at)

M L Mm _
Son’s Age 8.64 1.56 8.32 1.85 1.05 (120)
Mother’s Age 36.68 4.51 37.53 4.45 1.07 (124)
Father’s Age 38.48 5.27 39.19 5.45 .75 (122)
Mother’s 15.02 2.23 15.44 2.11 1.08 (130)
Education‘
Father’s 15.99 2.73 16.18 2.85 .39 (126)
Education‘
Family 58.12 32.31 51.38 27.36 1.30 (130)
Income"
Children in 2.64 1.03 2.62 1.70 .08 (97)
Family
F n V ' Chi-89m
% of Firstboms 50.70 59.68 n.s.
% of Middle child 26.76 19.35 n.s.
# of Sons in Therapy 9 1 4,44‘
# of Sons with 24 8 7.06“
History of Therapy
# of Mothers in 8 5

Therapy

 

‘In years. bIn thousands.

’p (.05, two-tailed. ”p <.01, two-tailed.

90

Finally, information was obtained regarding the family’s involvement in
seeking treatment for psychological problems. Table 4 reveals that there were some
significant differences in terms of seeking psychological intervention for the sons.
There were a significantly greater number of conduct-disturbed boys currently
involved in therapy (this included contacts with a school counselor), and a
significantly greater number that had past involvement in treatment. These
differences are understandable since the mothers of these boys perceive their child to
have behavior problems and thus are more likely to seek professional intervention.
P ' i 1 ' n

The participants in this study were recruited from five school districts located
within a 20 mile radius of Lansing, Michigan, the state capital. Lansing and the
surrounding area has a population of approximately 200,000.

Each of the respective school districts agreed to distribute a letter from the
researcher to all of the mothers of all male students in their elementary schools (K
through 6"I grade). The letter was sent home with the boys, usually as part of a
weekly packet of school-related announcements. Four out of the five school districts
wrote a brief endorsement and approval of the study. The school district’s letter
appeared on the reverse side of the recruitment letter. Once the letter was distributed,
the school districts had no other involvement with the research. The recruitment
letter specifically asked those mothers that were interested in the study to contact the

researcher directly in order to obtain more information.

91
A copy of the letter distributed to the mothers through the schools is included

in the Appendix. The letter offered the mother free attendance to a parenting skills
workshop in exchange for her participation in a study exploring important factors in
parent-child interactions, especially those between mother and son. The letter
attempted to locate "both mothers who consider(ed) their son to be a behavior
problem in the home, as well as mothers who regard(ed) their son’s behavior to be
no more disruptive than the average boy ".

Those mothers not interested in the skills workshop were given the option to
choose a $15.00 remuneration. Of the 133 participants, only 12 mothers chose the
financial remuneration. There was no significant difference between the groups in
terms of the mothers selecting the workshop or the financial remuneration, X’ (1, N
= 133) = 3.16, 9 >05.

The half-day workshop was described as being designed to help increase child
management skills and to aid and offer some specific suggestions in dealing with their
sons’ annoying and disturbing behaviors at home. Approximately 50 mothers
eventually attended the parenting skills workshops which were held a few months
after the completion of data collection.

Linnaeus

A variety of paper-and-pencil instruments were used in this research. Some
are standardized instruments used in previous research and some were created and/or
modified for the present study. The instruments used were: the Achenbach Child

Behavior Checklist (Achenbach & Edelbrock, 1983) the Family Information Sheet,

92
the Enmeshment Scale from Bloom’s (1985) Family Activities and Relationship

Questionnaire, the Child Behavior Scenes, the Child Behavior Rating Scales, the
Preliminary Questions, and the Adult Attachment Questionnaire.
W
"The Child Behavior Checklist is designed to obtain parents’ descriptions of
their children’s behavior in a standardized format" (Achenbach & Edelbrock, 1983 ,
p.10). The parent is asked to rate 118 different behavior problems (e.g., "demands
a lot of attention") on a three point scale. The parent rates the extent to which that
behavior describes their own child (e.g., 0 = not at all, 1 = sometimes, and 2 =
very often). Various syndromes of behavior problems were identified by principal
component factor analyses of the CBCL items rated by parents of children referred
for outpatient treatment. Relevant for the present study were the results for the 6-11
year—old boys. There were nine factors that emerged from their analyses. While the
scales were developed on a clinical population in order to detect clinically significant
syndromes, normative data have also been collected. Achenbach and Edelbrock
(1983) report test-retest reliability of mothers’ rating for boys 6-11 years-old as
follows: the Aggressive Scale was .95, the Delinquency scale was .95, and the
Hyperactive scale was .92. The median one week test-retest correlation reliability for
the mothers’ ratings for all scales at this age group was .76. All the above
correlations are reported to be significant at p = .05 or less.
In order to determine the suitability of the CBCL scales for the present sample,

a reliability analysis was performed. Using the current sample, Table 2 shows the

93
Alphas calculated for each scale. The Alphas range between .57, for the Schizoid

Scale, to .90, for the Aggressive Scale.
Eamilxhrfsznuaticnfim

This is a straightforward information sheet designed for the present study in
order to obtain the participant’s demographic data. There are questions about family
size, ages of family members, educational history of mother and father, mother’s
relationship to son, and approximate family income. A copy of the Family
Information Sheet is included in the Appendix.

This is the Enmeshment Scale (i.e., Scale 15) from Bloom’s (1985) Family
Activities and Relationship Questionnaire. The questionnaire is made up of 75-items
that have been factor analyzed in four separate data collections/ studies. The analyses
have consistently yielded 15 factors, or dimensions, of family functioning. The
Enmeshment Scale, one of the 15 factors, is intended to assess "the extent to which
family members are seen as insisting on interdependence to the exclusion of individual
action" (p.237).

It is a five item scale (e.g., "family members find it hard to get away from
each other") and the mothers were asked to rate the extent to which each statement
describes her current family (e. g., 1 = very untrue for our family, 2 = fairly untrue
for our family, 3 = fairly true for our family, and 4 = very true for our family).

Thus, the FT Q produced a score that ranged between 5-20. Bloom (1985) reported

94
norms for this scale as having a Mean = 9.5 and a SD. = 2.9. (A copy of the

Family Togethemess Questionnaire is included in the Appendix.)

Bloom (1985) concluded that there is satisfactory reliability for the
Enmeshment Scale, reporting for each separate study the Cronbach Alpha at .72, .67
and .78, respectively. The average inter-item correlations on this Scale are reported
to be .34, .30 and .42 for the three studies, respectively. On the other hand, an
initial attempt at validation was unsuccessful as the Enmeshment Scale did not
significantly differentiate intact families from families that subsequently were
disrupted by divorce and separation (Bloom, 1985). This validation study, however,
suffered from its reliance on retrospective data (i.e. , it was based on college students
retrospective accounts of family life).

A reliability analysis was performed on the FTQ for the present sample. The
mean score was 9.15 (S1; = 3.09). The Alpha was .80, with the mean inter-item
correlation = .46.

hi1 vi r n

The Child Behavior Scenes was developed specifically for the present study
in order to examine parental perception of child behaviors; specifically, the parental
tracking and labelling of positive and negative child behaviors. This instrument was
designed to help test the hypotheses regarding differences in the mothers’ deviancy-
categorization processes. The Child Behavior Scenes is comparable to both the
methods (i.e., the stimuli and the procedures) and the goals of research reported by

Holleran et al., (1982a, 1982b).

95

The scenes are four short scenarios depicting a variety of interactions primarily
between boys and other family members; there are some peer interactions. The titles
of the four scenes describe the context in which the child behaviors take place: After
School, Getting Ready For Dinner, Dinner, and Around The House. Some of the
behaviors for the fourth scene, Around The House, were suggested by Schmaling
(1983). There are a total of 180 different child behaviors that are embedded in the
four scenes. A behavior is defined by the use of a verb.

The mothers were instructed to read through each of the four short stories and
to underline with a red pen all those child behaviors that they considered negative,
undesirable, aversive, hurtful, and/or inappropriate, and to underline with a green pen
all those child behaviors that they considered positive, desirable, helpful, appropriate,
or prosocial. The mothers were asked to imagine that it was their own son that was
engaged in the behaviors that were depicted.

When the Child Behavior Scenes was administered, the mothers were given
a detailed instruction sheet which was read aloud by the researcher. The instruction
sheet explained that the stories were designed to help study the different ways that
parents follow or keep track of their children’s behaviors, including how parents vary
in the way they judge different child behaviors. The instructions directed the mothers
to read the stories carefully and to pay attention to what the boys do, what they think,
how the boys react and respond, what they say, and how they feel. Specific examples

were provided on the instruction sheet in order to help the mothers identify the

96
separate segments of child behaviors. (A copy of the instructions for the Child

Behavior Scenes is included in the Appendix.)

Each of the 180 behaviors were pre-experimentally classified as either a
positive, negative, or neutral behavior in accordance with the following procedures:
The 180 behaviors were listed in the same order in which they appeared in the
stories, thereby providing the same contextual cues that the mothers bad. This list
was then rated by 5 Clinical Psychologists, all psychotherapists. The therapists all
have a Ph.D and had been in Clinical practice between 5-25 years; the mean number
of years in practice was 14.4. The approximate percentage of their Clinical practice
that involved working with children and families ranged between 30%-100% , (M =
58%). The therapists were asked to individually rate each of the 180 different child
behaviors as either a positive, negative, or neutral behavior. The specific instructions
to the therapists were as follows: For the negative behaviors:

"Those behaviors that you judge to be clearly negative, inappropriate, hurtful,

aversive, and/or undesirable please mark with an N, for Negative. "
The instructions for the positive behaviors were:

"Those behaviors that you judge to be clearly prosocial, appropriate, helpful,

and/or desirable please mark with a I_’ for Positive."

Finally, the instructions for judging a behavior as neutral were:
"This will leave a variety of behaviors unmarked. The unmarked behaviors
will be those that you cannot consider to be either clearly _N_egative or Positive.

These behaviors may be neutral (i.e., they don’t seem either desirable or

97

undesirable) or they may be behaviors that you cannot clearly judge without

more information regarding a specific context (i.e., they are ambiguous)"

The final classification of these 180 child behaviors was based on the following
criteria: A behavior was classified as positive if 4 out of the 5 of the Psychologists
(i.e., 80%) judged the behavior to be positive. A behavior was classified as negative
if 4 out of the 5 Psychologists (i.e., 80%) judged the behavior to be negative. The
remaining behaviors were classified as "neutral" . In other words, those behaviors
classified as neutral were the behaviors on the list that were either left unmarked by
the therapists, and/or those behaviors that did not obtain the 80% criteria.

There were 61 behaviors that reached the criteria to be classified as "positive".
There were 42 behaviors that reached the criteria to be classified as "negative". The
remaining 77 behaviors were classified as Neutral. (The final classification of each
of the 180 behaviors that comprise the Child Behavior Scenes is presented in the
Appendix.)

The Child Behavior Scenes yielded a variety of scores for each participant: the
number of correctly identified positive behaviors, the number of correctly identified
negative behaviors, the number of positive behaviors identified as negative, the
number of negative behaviors identified as positive, the number of neutral behaviors
identified as positive, the number of neutral behaviors identified as negative, the total
number of behaviors identified as negative, the total number of behaviors identified

as positive and, the total number of behaviors identified. There is also a proportion

5C0

a :1
5 e1

13?

WCI‘E

and I

focus:
tgat‘

Iran

31cm

98

score (i.e., the proportion of behaviors identified as negative), which controls for
each mother’s specific response rate.
hi1 B vi r R ' e

The Child Behavior Rating Scales were developed for the present study. They
were designed to obtain the mothers judgments regarding the degree of aversiveness
and the degree of desirability of specific child behaviors. Holleran et al., (1982a)
reported using a similar procedure in examining the parental ratings of child behaviors
in terms of the degree of aversiveness.

The Child Behavior Rating Scales were divided into two sections. Section I
focussed on the mothers’ ratings regarding the degree of misconduct of certain
negative behaviors. Section II focussed on the mothers’ ratings of the desirability of
certain positive behaviors.

Section I of the Child Behavior Rating Scales is comprised of 63 statements
or examples of child misconduct. The mothers were instructed to rate each statement
on a six-point-scale. The mother’s ratings reflected her own judgment as to the
seriousness of each misbehavior (e.g., 5 = an extremely serious misbehavior, 4 =
serious misbehavior, both 2 and 3 = moderate, l = slight misbehavior and 0: not
at all a misbehavior). Thus, the mothers’ scores on Section I, based on only 47 of
these statements, could potentially range between 0-235 .

The statements represent a variety of categories of misconduct: aggression,
Stcalling, destruction of property, drug use, disobedience, deception, and

allnoying/aversive social behaviors (e.g., insults, nagging, forgetfulness, and

99

inconsiderate of others). The variety of misbehavior ranges from the very extreme
(e.g., "He took out a knife and threatened his teacher.") to the moderate (e. g., "He
fed the dog an hour later than he was supposed to.") and includes some statements
that might not be considered a misbehavior at all (e.g., "Stew reminded his parents
that the movie started in 30 minutes"). (A copy of the Child Behavior Rating Scales
I appears in the Appendix.)

Section II of the Child Behavior Rating Scales is comprised of 63 statements
or examples of prosocial child behaviors. The mothers were instructed to rate each
statement on a six-point-scale. The mother’s ratings reflected her own judgment as
to the desirability of each prosocial behavior (e. g., 5 = an extremely desirable
behavior, 4 = highly desirable, both 2 and 3 = moderate, l = slightly desirable, and
0 = not a clearly desirable behavior). Thus, the mothers’ scores on Section II, based
on only 53 of these statements, could potentially range between 0-265.

The statements in Section 11 represent a variety of categories of prosocial
behaviors: volunteerism, charity, cooperation, participation in groups and social
activities, positive peer involvement, interests and hobbies, academic/intellectual
pursuits, responsibility with money (e.g., earning and saving money) restraint of
aggression, appropriate emotional expression (e.g. , appreciation and care) and
honesty. The variety of prosocial behaviors ranges from extremely desirable (e. g.,
"Drake donated all of this weeks allowance to the fund for needy children") to
moderately desirable (e.g., " He washed his hands for dinner."), and include some

Statements that might not even be considered a prosocial behavior (e. g., "Ted put

SONIC

Ben

Prior
4.

linen
itch I
56:90
on m,
. (“pl

Ilfms3

h!

- n

"all:
\

100
some more sugar in his chocolate drink."). (A copy of the Child Behavior Rating

Scales II appears in the Appendix.)

When the Child Behavior Rating Scales were administered, the mothers were
given a detailed instruction Sheet. These instructions were read aloud by the
researcher. The instruction sheet explained that the Child Behavior Scales were
designed to study how parents judge certain behaviors. For both sections of the Child
Behavior Rating Scales it was made clear that each behavior on the list was performed
by a boy. The mothers were asked to imagine that the boy performing the behavior
was the same age as her own son. Examples regarding the use of the scales and how
to complete the ratings were provided.

A principal Component Factor Analysis with a Varimax Rotation was
performed on Section I and II, respectively, in order to determine the underlying
dimensions measured by the instrument and to create more homogeneous scales. In
each respective analysis one main factor emerged. For the Child Behavior Checklist
Section I there were 47 items, from the 63 original items, that loaded .40 or higher
on the one underlying dimension. The internal consistency of these 47 items that
comprised Section I was measured by Alpha = .94. For Section II, there were 53
items, from the 63 original items, that loaded .40 or higher on the main factor.
These 53 items comprising the scale for Section II had a Alpha = .95.

This is a set of 10 questions that served as a summary of the mother’s

relationships with her own parents. Five questions are focussed on the overall quality

3110

1011*!

'1
I
\_.
a: T
.4le
‘
A

101

of experience with the mother (e. g., loving, neglecting, rejecting, pushed to achieve,
and role-reversing) and the same five questions are repeated in regards to the
relationship with the father. For each question the respondent could choose one of
five options that best reflected her own experiences (e.g., in describing how loving
the relationships were with her parents, the mother could choose between very lacking
in love, lacking in love, neither loving nor actively unloving, loving, and very
loving).

The Preliminary Questions were designed specifically for the present study as
an instrument to be used in conjunction with the Adult Attachment Questionnaire (see
below). The Preliminary Questions were administered up to four weeks prior to the
administration of the Adult Attachment Questionnaire. These quickly obtained
summary statements were subsequently either supported or contradicted by the
specific content of the experiences recalled in the detailed responses to the full length
questionnaire. Thus, the Preliminary Questions do not yield a specific score, but
rather they functioned as an aid in scoring the organizational properties of the
mother’s presentation of attachment-related information, particularly in terms of its
overall coherency and consistency. (A copy of the Preliminary Questions can be
found in the Appendix.)

A l A n ' ' AA

The Adult Attachment Questionnaire is an open-ended questionnaire that asked

the mothers to describe their childhood relationships with their parents and to give

memories supportive of these descriptions. Specific questions were asked about

 

1110

11.11:

1:98

mil

ital

quest

Semi-

102

feelings of rejection, experiences of being upset, hurt, or ill, and separations. The
mothers were also asked about their own understanding of why their parents acted the
way they did, the effects of these early relationships on their adult personality, the
influence of other significant relationships with adults, and the current status of the
mother’s relationships with her parents, including changes over the years.

The AAQ is based on the Adult Attachment Interview (George, Kaplan, &
Main, 1985). (An original version of the AAQ was designed by Hazan & Shaver
(1988) but has not yet been used in research. The current version is a slight
modification, discussed below, of their efforts.) The AAQ is an attempt to obtain
attachment-relevant information (i.e. , regarding parent-chfld relationships) in a
questionnaire format that is comparable to the information obtained from the 1-2 hour
semi-structured interview. In fact, the specific questions from the interview are used
as the items on the questionnaire. Fifteen of the 18 questions that comprise the
interview do appear on the AAQ, with an added question regarding the elicitation of
parental attention. In essence, the same basic questions are asked, but the ability to
probe and ask for clarification is obviously not part of the questionnaire format.
Other minor changes from the Hazan and Shaver (1988) version of the AAQ have
been made in terms of wording and formatting. These changes include the addition
of the inquiry on the elicitation of parental attention, as well as the elimination of the
i"quiry into significant losses during childhood; (issues of unresolved grief were not

explored in the current research.) (A copy of the AAQ is included in the Appendix.)

'sm:

prom

quest:

m,

P

C“

'015

Mm;

103
The Adult Attachment Rating and Classification System (Main & Goldwyn,

1989) provided the basis for the scoring system that was used to score the AAQ.
This scoring system was originally designed for use with the Adult Attachment
Interview. It is a scoring system for assessing and ultimately classifying an adult’s
"state of mind with respect to attachment" (Main & Goldwyn, 1989, p. 2). Each
protocol is given several nine—point ratings on various scales. Then, based on the
patterning of these scale scores, each protocol is assigned a single overall
classification representing the individual’s internal working model (or present state of
mind) with respect to attachment.

A pilot study using the AAQ was undertaken in order to determine the
questionnaire’s feasibility and usefulness as a data collection procedure and to assess
the applicability of the Adult Attachment Rating and Classification System. Thus, the
pilot study provided the data necessary in the development/modification of the scoring
manual for the specific use with the AAQ (i.e., verifying which of the previously
used scales were also scoreable given the format of the AAQ) as well as providing
the protocols necessary in the training of raters for the present study.

The participants in this pilot study were 50 undergraduates from a large mid-
westem university who were enrolled in an introductory Psychology class.
Participation in research was part of the course requirements. The students were told
that this was a pilot study aimed at developing a questionnaire that would explore
Childhood memories. The Preliminary Questions and the AAQ were both administered

at the same time and in a group Situation.

mum

.ilQ
score:
0kg
Sales
Mei

Pitta

l
.Iane‘
imam

"3 30ml]

104
On the basis of the data/results from this pilot study using the AAQ, the

differences between the AAQ and the Attachment Interview, and the differences in
the conceptual focus between the present study and the original work of Main and
Goldwyn (1989), slight changes and modifications in both the content of the AAQ and
the original scoring system were implemented. In modifying the scoring manual for
the AAQ every effort was made to maintain the conceptual integrity, spirit, and
nuances of the Adult Attachment Rating and Classification System. The actual
changes included: the development and/or elimination of specific questions on the
AAQ, the development of new variables/scales, the elimination of certain scale
scores, the establishment and/or modification of new criteria used for the assignment
of specific scale scores, the modification of the original nine-point scales to five-point
scales, and the use of only certain rating scales in the present study (i.e., the
participants’ overall attachment—classification was not determined since it was not a
particularly relevant issue for the present study).

The AAQ yielded 12 scores; 10 scale scores and two composite scores. The
ratings for each scale/variable were based on the information obtained primarily from
the entire AAQ protocol and, to a limited extent, from the information obtained from
the Preliminary Questions. An essential feature of the scoring system is that the
entire protocol is taken into account rather than any single answer in isolation.
Hence, the assignment of scale scores were not based on single conclusive statements
in the protocols; rather, the raters were directed to look at the total picture, looking

for contradictions and unsupported statements. For example, the choice of " not at all

ti,
3312‘
e

‘h l
slljdk“

.L1 to
Mint

[:0

5F

105
rejecting", exhibited either on the Preliminary Questions or initially presented in the

questionnaire, may still be part of a protocol scored for rejection on the basis of either
a lack of supportive evidence and/or on the basis of contradictory content.

Six of the ten scales are experiential, representing the mother’s "inferred"
childhood experiences. This is in accordance with Main and Goldwyn’s (1989)
caution that these scales represent "the probable experiences of each parental figure
during childhood" (p. 2). The six scale scores regarding the mother’s probable
childhood experiences are: Loving/Unloving, Rejection, Neglect, Role-Reversing/
Involving, Pressure to Achieve, and Abuse. For each of these six scales, separate
ratings were given to the subject’s experiences with mother and father during
childhood. Brief descriptions of the six scales assessing the mother’s probable
childhood experiences are as follows:

Lgvinglflnlgving. This scale assesses the extent to which the participant seems

to have experienced her mother or father as loving or unloving during

childhood. The scale is designed to assess whether the subject had a firm
sense of the emotional support and availability of the parent, especially in
times of trouble.

Rejecfion. This scale assesses the extent to which the parent rejected and/or

avoided the child’s moves towards dependence and attachment. The scale is

specifically focussed on that rejection which aimed at shifting the child’s
attention away from feelings of attachment. "The essential quality being

assessed as rejection is the quality of turning back or (turning) away of the

106

child’s dependence, affection, attention, need and attachment." (Main &

Goldwyn, 1989, p.18)

Neglect. This scale assesses the extent to which the parent was physically and

psychologically unavailable, unresponsive, or inaccessible. It is a rating of a

parent that is uninvolved, preoccupied, otherwise occupied, inattentive, or

absent.

Role-Reversing/Invglving. This scale assesses the extent to which the parent
made him/herself an object of the child’s attention, involving the child
in his/her physical or psychological care, and/or needing the child’s
involvement for the maintenance of his/her security, everyday
functioning, and well-being.

W. This scale assesses the extent to which the child was

pressured to achieve some particular Status or position. It goes beyond

encouragement to do well, but rather the rating is based on the extent to which
the child’s attachment and needs were responded to only in connection with
specific achievements or accomplishments.

Abuse. This scale is intended to rate the degree of physical, sexual and/or

emotional abuse that the participant experienced during childhood. This scale

assesses the extent to which the parental response or behavior is inappropriate,
extreme, excessive, and exhibits an overall disregard for the child’s well being
and individuality; at the extreme, the basic parental attitude is one of contempt

towards the child.

 

 

 

107
This Abuse scale was not part of the Main and Goldwyn (1989) scoring system

but was added because it seemed like a potentially important aspect of one’s
childhood experiences.

Wm. This variable is a composite score

of the six inferred-experiences scales (i.e., Overall Problematic Emotional

Relations = Unloving + Rejection + Role-Reversal + Neglect + Pressure

to Achieve + Abuse.) The composite score is taken as an overall index of the

mother’s emotional difficulties encountered as a child in regards to attachment-
related experiences.

The AAQ yielded another five scores or variables that were intended to assess
the mother’s present state of mind with respect to attachment (Main & Goldwyn,
1989). These variables refer to the mental representation of attachment-related
experiences and to the organization of the information relevant to these experiences.
Hence, these scales, (there are four scales and one composite score), assess the ease
with which information is accessed, the apparent distortion which certain information
has undergone, the overall organization of attachment-related information, and to
specific feelings. As Main and Goldwyn ( 1989) point out, each of these scales "deal
with the current representation of the self in relationship to attachment figures,
attachment-related experiences and attachment-related phenomena" (p.40). The
following "present state of mind" scales were used:

Incoherent. This scale reflects the extent to which the mother is able to present

and assess her past and its influences in a fluid, cohesive, and consistent

108

manner; a coherent protocol is easily understandable and permits or suggests

few or no alternative interpretations to those already offered by the participant

herself.

Main and Goldwyn (1989) do point out that this variable stands as the over-
riding assessment of an individual’s state of mind with respect to attachment. In
previous research, Coherency has accounted for the largest amount of variance in the
dependent variables (e. g, Main & Goldwyn, 1984).

Idealizafion. This scale assesses the extent to which the mother retains some

image of a "perfect" parent, separated from actual experience and held

separate from actual experience. The scale assesses the discrepancy between
the overall viewpoint of the parent (e. g., on the Preliminary Questions) and
the specific content of the attachment-related experiences.

W. This scale assesses the actual amount and quality of memory

(i.e. , specific content) that is demonstrated within the protocol. This scale

assesses the extent to which the subject provides actual memories and recalls

specific events and circumstances of their own childhood. The scale assesses
the degree of specificity of the recollections, rating the richness/depth of the
historic content that is contained in the written protocol.

This scale, developed for the current research, is intended to be a memory
scale assessing information availability or information accessibility. It is loosely
based on the Main and Goldwyn (1989) scale named "Frequency and Strength of

Insistence Upon Inability to Recall Childhood". Main and Goldwyn (1989) point out

1'11

qua

78$;

aft:

Tiem

109
that it is not possible to assess the subject’s actual inability to recall childhood.

Therefore, Main and Goldwyn (1989) rate the frequency and strength of the mother’s
insistence upon her inability to recall childhood. The Inability to Recall scale is an
assessment based on the amount of times and the degree of insistence that the
interviewee states she can’t remember.

The primary reason that this Inability to Recall scale was determined
inappropriate with the AAQ involves the differences in format between the
questionnaire versus the actual interview. The pilot data revealed that references to
one’s own memory, and specifically the disclaimer of "I can’t remember", is not a
response used with sufficient frequency in written protocols; it apparently does have
a frequent usage within the verbal discourse of the interview. Thus the present scale,
rather than being based on the subject’s reflection and self-appraisal regarding their
own ability to remember childhood, is based on the actual amount and quality of
memory (i.e., the specific content) that is demonstrated within the protocol.

Unresolved; Seoeeao'on god Autonomy Seale, This scale assesses the extent

or degree to which childhood relationships, especially difficult emotional

experiences, have been understood, resolved, and integrated. The scale
assesses not only a cognitive awareness and/or understanding of the influences

(both positive and negative) of one’s own childhood experiences on adult

relationships and functioning, but also the extent to which the participant had

achieved an emotional distance, separateness, and psychological independence

from her childhood relationships.

110

This scale was developed for the current research. Data from the pilot study
suggested that a participant’s " state of mind" regarding attachment-related experiences
is significantly influenced by the degree to which childhood conflicts have been
worked-through and resolved. Main and Kaplan (1984) present evidence supporting
the importance of a Resolution of Childhood variable, suggesting that it is not the
quality of the mother’s childhood attachments per se, but rather it is the extent to
which she has come to terms and resolved her past that is of critical importance.

Relatedly, Main and Goldwyn (1989) make a distinction regarding the quality
of the affect represented and exhibited by the participants. In their Anger scale they
differentiate between an "involved/pre-occupi " anger versus an "autonomous" form.
In the pro-occupied expression the subject expresses anger towards the parent which
indicates current involvement in feelings of anger and an emotionally charged
derogation of the parent (i.e. , a lack of separation and resolve). Similarly, the
current Unresolved scale assesses not only the relevant cognitive representations (i.e.,
the self appraisals) but also takes into account the quality of the affects that are
represented.

Totfl Defensiveness. This is a composite score for the "state of mind"

variables; (Total Defensiveness = Incoherence + Idealization + Lack of

Detail + Unresolved). This composite score is taken as an overall index of

the defensiveness of the mother’s cognitive model regarding her own childhood

experiences.

3

 

I» r‘

111
Raters

The AAQ was scored and rated by five raters; four were graduate students in
Clinical Psychology and one was a Clinical Social Worker. Half-way through the
actual scoring of the data one rater withdrew because of time constraints. While the
raters volunteered to participate primarily to familiarize themselves with Main and
Goldwyn’s Adult Attachment Rating and Classification System, they were also
financially compensated $10.00 per protocol. The raters were not aware of the
specific hypotheses being tested and they were blind as to group membership of the
participants whose protocols they were scoring.

liainiog. The raters all met together with the researcher for approximately
20 hours of training. The training was conducted primarily on the pilot data, with the
very end of training involving actual participant protocols.

Training emphasized familiarization and learning of the scoring manual. There
were 12 protocols that each rater scored independently over the course of the training.
The raters would meet as a group with the researcher, compare scores and discuss the
protocols. These discussions, which accounted for the bulk of the training, were
focussed on points of disagreement. Disagreements between the raters appeared to
be the result of carelessness, ambiguities in the scoring system, the complexity of the
data and the scoring system, and personal biases. Due to the very personal and
subjective nature of the protocols, clinical judgments were susceptible to biases,
distortions, and "transference-type" reactions, similar to those experienced in a

therapeutic setting. These issues were addressed as needed. Finally, the training

112
allowed the raters to acquire a familiarity with the potential range of data (e.g.,

between a very coherent protocol versus a strikingly incoherent protocol) as
prototypical examples were provided.

Determination of Inger—Raier Reliability. The statistic used to determine inter-
rater reliability was the percentage of agreement-within-one. Thus, for example, a
protocol assigned a score of "3" for Incoherent was considered to be in agreement if
another rater scored the same protocol and scale a "2." The rationale for using this
as an index of reliability is derived from the very nature of the data. The data are
full of complexity, subtlety, and nuance. It is a questionable undertaking to assume
you can assign numbers/scores to describe an event as intricate as one’s childhood
relationships. Inherent in the very nature of this task is a heavy reliance on subjective
clinical judgement. While there may be clear differences between a score of "1" and
a score of "3", the difference on some protocols between a score of "3" and a score
of "2" can be so elusive as to be meaningless. Clearly, this is not an exact science
and a significant amount of measurement error, no matter how determined the effort,
is to be expected. Thus, for present purposes "agreement-within—one" will be
considered essential agreement.

As mentioned above, each of the five raters, respectively, independently scored
all twelve of the protocols used in the training. This subsequently allowed for many
pairwise comparisons in order to determine inter-rater reliability as each individual
judgement could be compared with that of the four other raters. A minimum criterion

level of 80% agreement-within—one was considered acceptable. Table 5 presents the

1 13
Table 5

 

 

 

Pr mnB eenRter for eDurin R rT inin
_R_a_te_r_s

I II III IV V
% of Exact
Agreement‘ 50 52 57 50 57
% of Agreement
Within-one‘ 88 90 92 91 93
Scale; m % Agreement wimio-one
Unloveb 52 91
Rejectionb 49 88
Neglectb 59 95
Achieveb 63 95
Role-Reversalb 50 85
Abuseb 69 94
Incoherentc 36 83
Idealizationc 52 86
Recallc 51 90
Unresolvec 40 90

 

‘Based on 782 pairwise comparisons. bBased on 270 pairwise comparisons.
cBased on 135 pairwise comparisons.

nut.

fie

as.

I?»

114

inter-rater agreement across the twelve training protocols. At the end of training the
inter-rater agreement ranged between 83 % agreement within-one (for the Incoherence
scale) and 95% agreement within-one (for both the Pressure to Achieve and Neglect
scales).

In order to determine if this acceptable level of inter-rater agreement within-
one was being maintained during the actual scoring of the data, two different
reliability checks were undertaken. The initial reliability check was shortly after
training was completed and the first 20 protocols had been scored. The second
reliability check was performed about half-way through the scoring, after 60 protocols
had been completed. For both of these reliability checks ten protocols were used,
each protocol was scored by two raters; hence, each rater was compared once with
every other rater. Table 6 presents the combined data from these two reliability
checks. All scales, except for Idealization, maintained an adequate level of inter-rater
agreement. The Idealization scale was at 90% agreement- within-one at the first
check but this level of agreement was not maintained.

Proeedure

The experimental procedure was very straightforward. It can be divided into

three segments: (a) the initial phone interview; (b) questionnaires completed at home,

and (c) the three instruments administered during the data collection session.

115

 

 

 

Table 6
r f In r-R r A r m n -Within-On for R rs d Sc
Ratere

Scale I II III IV V Scaiclmal
Unlove 100 88 94 94 88 93
Rejection 100 88 100 100 88 95
Neglect 81 94 75 88 100 88
Role- 100 88 94 88 94 93
Reversal
Achieve— 88 88 88 88 88 88
ment
Abuse 100 94 100 94 100 98
Incoherent 100 75 100 88 88 90
Idealiz-
ation 88 63 63 88 75 75
Recall

100 88 88 100 75 9O
Unresolve

100 100 100 88 88 95
Rater’s 95 87 91 92 91
Total

 

Note. The scales totals for the first six scales are based on 40 pairwise comparisons;
the remaining four "state of mind" variables are based on 20 pairwise comparisons.
The rater’s totals are based on 128 pairwise comparisons.

 

116

As requested in the letter sent home through the schools, the mother’s first
contact with the researcher was over the telephone. This initial phone contact
consisted of a ten minute interview. (The outline of the phone interview and the
Initial Phone Contact sheet completed during the conversation are in the Appendix.)
The goals of the phone interview were: (a) to establish contact/rapport; (b) to
determine the conduct-disturbed versus non-conduct—disturbed status of the mother’s
son; (c) to obtain some demographic information; (d) to determine the mother’s
choice of either the parenting skills workshop or the financial remuneration, and (e)
to begin tentative scheduling for the data collection session. At the completion of the
phone interview, it was explained that a packet would be mailed to them within a
week.

This packet contained a letter of introduction with an explanation of their
participation in the study, a consent form, the CBCL, the Family Togethemess
Questionnaire, and the Preliminary Questions. The total packet took the mothers
approximately 30-40 minutes to complete. The mothers were asked to return the
completed packet within 7 to 10 days in a postage paid envelope.

The participants were then contacted by phone and scheduled for the 3-4 hour
research session. A week prior to the data collection a letter confirming time and
place was sent to the mother. During the 48 hours preceding the session a follow-up
phone call was made for final confirmation.

At the research sessions all the instruments were administered in a group

setting. Both MCDB and MNB participated at the same time. The average size

117

group was about fifteen mothers. Upon arrival and check-in with the researcher, the
mother was asked to complete the Family Information Sheet. After everyone had
arrived, the mothers were formally welcomed to the data collection session; a brief
overview of the research as well as the agenda for the 3-4 hour session was provided.
The Child Behavior Scenes was then administered. After the entire group had
completed this instrument both sections of the Child Behavior Rating Scales were
administered.

This was followed by a 10-15 minute break, at which time refreshments were
provided. Following the break, the AAQ was administered. This left approximately
two and a half hours to complete the AAQ. After 75 minutes the researcher
announced the time and he suggested how much time was left. Practically all the
mothers completed the AAQ between 1.5-2.5 hours. As the individual participants
completed the AAQ and turned in their completed data packet, they were thanked for
their time, given a debriefing sheet, and allowed to depart. Those mothers who

choose the financial remuneration were paid at this time.

CHAPTER3

RESULTS

This chapter will be organized into four sections. The first three sections
correspond to the three groups of hypotheses; namely, ( 1) those hypotheses that
examine the differences in the mothers’ deviancy categorization processes; (2) those
hypotheses that explore differences in the mother’s recollections of her childhood
experiences and (3) the hypothesis that examines the relationship between the
mother’s perception of child behavior and the affective-cognitive organizational
characteristics of her childhood memories.

‘ The final section of this chapter will present data from an analysis performed
on a sub-sample of the original sample. The rationale for the sub-sample and the
results of this analysis will be presented in the fourth section of this chapter.

Green I; Meterng Pergotion of Chile Behavior

This group of hypotheses explores differences in the mothers’ perception of
child behaviors, specifically their deviancy categorization processes.

Hyoomesi 1. As can be seen in Table 7, there is general support for this

hypothesis with 4 out of the 6 comparisons yielding statistically significant results.

118

119
Table 7

"A9L ii . ‘ ' M DB .Vl'l 1“] IV in ' 1:". . -|.Ol'l Of DCVi-

ve x 1 ive in Th ir ti n of Proso ial Beh vior?

 

 

Comparison: MCDB m
1. The MCDB correctly identified a M 36.79 35.24
significantly greater number of negative S_D_ 4.51 5.71
behavioral segments than did the MNB. t (131) = 1.74’
2. There was not a significant difference M .36 .38
between the MCDB and the MNB in terms SD .63 .81
of the number of prosocial behavioral I(114) = .16 n.s.
segments judged incorrectly as negative.
3. The MCDB judged a significantly greater M 9.75 8.26
number of neutral behavioral segments to be E 4.52 4.62
deviant than did the MNB. 1(131) = 1.88'
4. Overall, the MCDB identified a greater M 47.61 44.61
number of behaviors as negative than did the S_ 7.99 9.44

MNB. §(131) = 1.98"

120

Table 7 continued

 

Comparisons

 

5. Within each mother’s specific response
rate, the MCDB identified a significantly
greater proportion of behaviors as negative

than did the MNB.‘

6. The MCDB did not rate behaviors as
significantly more aversive than the ratings

of the MNB.

M DB MNB
M 40.78 38.13
5;) 7.28 6.15

i(131) = 2.25"

M 132.67 126.67
52 23.26 25.10

3(131) = 1.43 n.s.

 

Note. Comparisons 1-5 are based on data from the CBS. Data for comparison 6

comes from the CBRSI.

‘This is a percentage score calculated, for each mother, as follows: (Total number of

behaviors identified as negative / Total number of behaviors identified, (i.e. , as either

positive or negative» x 100. The score describes the mother’s response tendency.

'o < .05, one-tailed. "o < .05,two-tai1ed.

121
As compared to the MNB, the MCDB: (a) correctly identified a significantly greater

number of negative behaviors; (b) judged a significantly greater number of neutral
behaviors to be negative; and (c) out of the 180 total behaviors (i.e. , positive,
negative, and neutral), the MCDB identified a significantly greater number of the
behaviors to be negative. Finally, for each mother, her " negative response tendency"
was calculated, i.e., (the mother’s total # of behaviors identified as negative / the
mother’s total number of all behaviors identified) x 100. Thus, for each mother this
was a percentage of her specific responses that were negative. This response
tendency was significantly more negative for the MCDB.

While the data support the position that the MCDB are more likely to perceive
behaviors as negative, it does not, however, support the prediction that MCDB are
conversely less prone to define and identify behaviors as prosocial. There was no
difference between the groups of mothers in the number of positive behaviors
incorrectly identified as negative, (a judgement that was very rare in this sarnple) and,
even though the MCDB did rate behaviors as more aversive than the MNB, the
differences in their ratings was not statistically significant.

W. As can be seen in Table 8, there is little support for the
predictions that the MNB are more likely to perceive behaviors as prosocial and less
likely to identify behaviors as deviant. The MNB were not different from the MCDB
in terms of (a) the number of correctly identified prosocial behaviors; (b) the number

of negative behaviors perceived as positive; (c) the number of neutral behaviors

 

122

 

Table 8
i 2' Ar M v r1 Inl iv in Ca riza'on f Prosoi
v x 1v Th H n vi ‘7
Catamarans. M DB MNB

 

1. The number of prosocial behaviors correctly
identified by the MNB was not significantly

different from those identified by the MCDB.

2. The was no difference between the MCDB
and MNB in their tendency to incorrectly
identify negative behavioral segments as

prosocial.

3. The MNB did not identify a significantly
greater number of neutral behavioral segments

to be prosocial behaviors than did the MCDB.

4. Overall, the number of behaviors identified
as positive was not significantly different for the

MCDB as compared to the MNB.

M 49.45 51.02
S 9.30 9.95

1(131) = .94 n.s.

IE

.97 1.34
E 1.91 1.76

;(l3l) = 1.15 n.s.

IZ

21.73 22.53
SD 13.15 12.76

t(131) = .35 n.s.

IZ

72.18 74.89

21.65 21.67

If!)
U

i(13l) = .72 n.s.

1 23
Table 8 continued

 

 

5. Within each mother’s specific response rate,
the MNB identified a significantly greater
proportion of behaviors as positive than did the

MCDB.‘

6. The MNB did not rate behaviors as
significantly more desirable than the ratings of

the MCDB.

M DB MNB
M 59.22 61.86
SD 7.28 6.15

1(131) = 2.25"

M 188.17189.13
S 30.41 34.81

1(131) = .17 n.s.

 

Mote. Comparisons 1-5 are based on data from the CBS. Data for comparison 6

comes from the CBRSI.

‘This is a percentage score calculated, for each mother, as follows: (Total number of

behaviors identified as positive / Total number of behaviors identified, (i.e. , as either

positive or negative» x 100. The score describes the mother’s response tendency.

”o < .05 ,two-tailed.

124
judged to be positive; and ((1) out of the 180 total behaviors, there was no difference

in the number of behaviors judged to be positive. Also, the MNB did not rate
behaviors as any more desirable than the same behaviors rated by the MCDB. The
only support for this hypothesis is that the " response tendency" of the MNB was
significantly more positive than the MCDB. This percentage score, however, is the
inverse of the " negative response tendency" score derived in Hypothesis 1.

W1. Table 9 shows the intercorrelations between all of the
perceptual variables derived from the CBS, the CBRSI and CBRSII, and the FT Q.
There is no support for the position that there is a "behavior classification
covariation" where an emphasis on deviancy would inherently result in a de-emphasis
on prosocial behaviors and vice versa.

In fact, while the a priori predictions were in terms of inverse relationships,
the results indicate positive relationships, with 3 out of the 4 relevant correlations
revealing significant positive relationships. As can be seen in Table 9, there is a
significant relationship between the number of neutral segments judged to be positive
and the number judged to be negative, I. (133) = .37, o <.001. There is a
significant relationship between the number of correctly identified positive behaviors
and the number of correctly identified negative behaviors, I (133) = .53, o < .001.
And, there is a significant relationship between the mother’s ratings of aversiveness
and her ratings of desirability, ; (133) = .40, o < .001. There is not a relationship
between the number of positive behaviors incorrectly judged as negative and the

number of negative behaviors incorrectly judged as positive, ; (133) = .10, o > .05.

l 25
Table 9

 

22mm

mole l 2 3 4 5 6 7 8 9 10

 

.NegNeg‘

.NegPos" .18“

.NegNeut‘ .51" .21*

. PosPos‘ .53** .05 .34“

.PosNeg‘ -.23* .10 -.02 .12

.PosNeut‘ .35" .12 .37“ .74“ -.33**

.NegPro‘ .06 .05 .15* -.74** -.39** -.72**

.CBRSI'I .15“ .07 .29“ -.13 .00 -.01 .27"

.CBRSIIi -.13 .03 .04 -.02 .19* .15* -.09 .40“
10.l'~'l’Qi -.01 -.04 .02 .11 .08 .16“ -.l6* -.08 -.01
”Response" .63" .18* .57“ .89“ .23“ .91" -.60** .02 .05 .13

\OOOQGUI-bWNr—

 

Note. All the correlations are based on u = 133

‘# of Negative statements correctly identified. "# of Positive statements judged to be
negative. °# of Neutral statements judged to be negative. dtil of Positive statements
correctly identified. °# of Negative statements judged to be Positive. ’# of Neutral
Statements judged to be positive. gProportion of negative responses. 1‘The mothers’
ratings of the degree of negativity of negative behaviors. ’The mothers’ ratings of the
degree of desirability of prosocial behaviors. JT he Family Togethemess Questionnaire
is an index of the family’s enmeshment. kResponse is the total number of behaviors

identified on the CBS and is taken as an index of the mother’s responsivity.

’o < .05, one-tailed. "o < .01, one-tailed.

126

 

 

Table 10
R ' nshi B e n Enme hment Mother’s Devianc -Classifi a 'on Sch m
Correlation wim

M11}. Enmeshmen
PHVBI. l. E 1i 1'
Total # of Behaviors Identified .126
# of Neutral Behaviors Judged to be Negative .020
# of Negative Behaviors Correctly Identified .012
# of Prosocial Behaviors Judged to be Negative -.036
Mother’s Ratings of Behaviors as Aversive -.082
Mother’s Response Tendency--Proportion of Total

# of Behaviors Identified Judged to be Negative -. 163’
Inv R ' 11 hi Pr icted°
# of Neutral Behaviors Judged to be Prosocial .159’
# of Prosocial Behaviors Correctly Identified .109
# of Negative Behaviors Judged to be Prosocial .079
Mother’s Ratings of Behaviors as Desirable -.012

 

‘fl = 133

’o < .05

in

her

:‘tSp
a 512
511m:

A.

127
W. This set of hypotheses explores the association

between the family’s level of enmeshment and the mother’s perceptual style. (1' able
9 is relevant for Hypothesis 6 and 7.) While the MCDB did describe their families
(on the FTQ) as significantly more enmeshed M = 9.69, SD = 3.41) than the
descriptions of the MNB M = 8.53, SD = 2.57), t (128) = 2.23, o < .05, two-
tailed, they did not identify (i.e., level of responsiveness) a greater number of total
behaviors, (both positive and negative combined), M = 119.79, SD = 25.71) than
the total number of behaviors identified by the MNB M = 119.50, SD = 28.37),
1(131) = .06, [2 >05.

As can be seen in Table 10, Hypotheses 6 and 7 are not supported, as there
is virtually no relationship between the mother’s deviancy-classification schema (i.e. ,
her perceptual style) and her description of her family as enmeshed. As seen in Table
10, there is not a significant relationship between enmeshment and the "mother’s
responsivity" (i.e., the total number of behaviors identified). Even though there is
a significant inverse relationship between the mother’s negative response tendency and
enmeshment, and a significant positive relationship between the number of neutral
behaviors identified as positive and enmeshment, both of these relationships, however,
are not in the direction predicted a priori. Finally, Table 9 indicates that no other
significant relationships exist between enmeshment and any of the remaining
perceptual variables.

The results for this group of seven hypotheses can be summarized as follows:

There is general support for Hypothesis 1 as the MCDB do have a greater tendency

128

to perceive behaviors as deviant; but contrary to the predictions, they are not "less
prone to define and identify behaviors as prosocial". Hypothesis 2 is not supported
as there is virtually no difference between the groups in terms of their perception of
prosocial behaviors (i.e. , the MNB were not more likely to perceive behaviors as
prosocial). Furthermore, Hypothesis 3 is not supported as an emphasis in classifying
behaviors as deviant is m related to a de-emphasis in classifying behaviors as
prosocial.

In support of Hypothesis 4, the MCDB did describe their current families as
more enmeshed. This variable, however, is not related to the other perceptual
variables. For example, the MCDB did not identify a greater number of behaviors,
as predicted in Hypothesis 5, and the mothers’ level of responsivity (i.e., the number
of behaviors identified) was not related to level of enmeshment as predicted in
Hypothesis 6. Finally, Hypothesis 7 is not supported as there is virtually no
relationship between the level of enmeshment and the mothers’ deviancy-classification
schema.

' M rn hildh Recollections

Hypotheses 8-10 are focussed on the mother’s recollections of her childhood
relationships. The analysis explores not only the specific content of these childhood
experiences, but also the organizational properties of these memories (i.e., the
mother’s mental model). The data are primarily from the AAQ and only minimally

from the Preliminary Questions.

129

 

 

Table 11
u or or 3' "r - DB . d : MNB n - ontent ofTheir Childhod

Recollections

mm M DB MNB 1‘. L411

Unloving M 7.14 7.18 .13 (131)
SD 2 04 1.97

Rejection M 6.34 6.59 .73 (131)
SD 2.02 1.97

Neglect M 4.51 4.37 .40 (131)
SD 2.17 1.91

Role—Reversal M 4.77 5.18 1.19 (131)
52 2.01 2.01

Pressure to Achieve M 3.95 4.53 1.59 (131)
SQ 1.89 2.33

Abuse M 4.93 4.82 .31 (131)
so 2.16 2.00

Total Problematic

Relations M 31.66 32.69 .74 (131)
Si; 8.35 7.72

 

Note. All of the scales were scored in the same direction with higher scores indicating

greater problematic experiences. ‘For all of the 115 reported, o > .05.

130

 

 

 

Table 12
91101:. on. B een ‘ M DB .. . ' MNB on e Affec'v- o'ni-v
Pr ' f Th ir hildh R c lle 'on
Scales. MCDB MNB _I: am
Incoherent M 3.58 3.29 1.73' (131)
S_12 .91 .97
Unresolve M 3.57 3.27 1.88’ (131)
112 .85 .99
Idealized M 2.58 2.54 .24 (131)
SQ 1.12 1.06
Lack of Recall M 3.10 2.81 1.67’ (131)
SD .96 1.05
Total
Defensiveness M 12.84 11.91 1.75' (131)
Si; 2.81 3.28

 

Note. All of the scales were scored in the same direction with higher scores

indicating greater affective-cognitive disorganization or defensiveness.
‘M = 133.
’o < .05, one-tailed.

131
As seen in Table 11, the MCDB did not produce memories that exhibited a

significantly greater amount of problematic emotional experiences. For each specific
scale, as well as for the Total Problematic Relations score, the childhood recollections
of the MCDB were not Significantly different in content than the memories recalled
by the MNB.

On the other hand, as can be seen in Table 12, Hypothesis 9 is generally
supported as the affective-cognitive organizational properties of the mothers’
childhood recollections were significantly different. The childhood recollections of
the MCDB were significantly more defensive (i.e., Total Defensiveness) than the
recollections of the MNB. As compared to the MNB, the childhood recollections of
the MCDB (a) were significantly more incoherent; (b) were significantly lacking in
specific details (i.e., greater inability to recall); and (c) exhibited a significantly
greater amount of unresolve regarding their own childhood experiences. The two
groups did not differ in terms of the amount of idealization exhibited in their
childhood memories.

Table 13 shows the intercorrelations between all of the scales of the AAQ.
Most of the Experiential Scales are significantly correlated with each other, as are all
of the affecitve-cognitive organizational scales.

However, as can be seen in Table 13, there is virtually no support for
Hypothesis 10, since there is practically no relationship between problematic

childhood experiences and a defensive organization of childhood memories. For

132

 

 

Table 13

In r ' n Between Scales f th AA

5.9% l 2 3 4 5 6 7 8 9 10 ll
1. Unlove

2. Reject .82*

3. Neglect .63" .50"

4. Rol-Rev .33“ .25“ .12

5. Achieve .25" .24" .06 .11

6. Abuse .51" .54" .24" .14* .17*
7. Total‘ .85" .81" .62" .50" .50" .66"
8. Incober .11 .06 .01 -.03 -.04 .05 .03

9. Unresolve .16"I .09 .08 -.01 -.15"' .15“ .05 .71“
10.NORecall .15* .07 .13 -.05 -.l7* .00 .00 .54“ .51“
11.1dealize -.05 -.04 -.09 .03 -.03 -.14* -.08 .48“ .33" .25""ll

l2.TotDefen" -.05 .06 .04 -.02 -.12 .01 .00 .87" .81" .74“ .69*

 

fl = 133. ‘Total Problematic Relations. "Total Defensiveness.

’o < .05, one-tailed. "o < .01, one-tailed.

133

example, there is literally no relationship between the two composite variables (i.e. ,
Total Problematic Relations and Total Defensiveness).

The three relevant significant relationships are between Unloving and
Unresolved, Unloving and Lack of Recall, and Abuse and Unresolved. The other
significant relationships (between childhood experience and affective-cognitive
organization) found in Table 13 are not supportive of the hypothesis since the a priori
predictions were in terms of positive relationships and the findings are inverse.

In summation of the results for the three hypotheses in Group II, the mothers
did not differ in terms of the quality of their childhood experiences. Contrary to the
prediction in Hypothesis 8, the MCDB did not produce childhood memories that were
more emotionally problematic than the MNB. On the other hand, Hypothesis 9 is
supported as the MCDB did produce childhood recollections that were significantly
more defensive and disorganized. Finally, there is little support for Hypothesis 10
which predicted that problematic childhood experiences would be associated with a
defensive affective-cognitive organization.

r 111' Rel 'onshi BtweenMam Prc 'n hilh Re 11 'n

The results for Hypothesis 11 are mixed. While there is some support to
establish a link between the mother’s current perceptual Style and the affective-
cognitive organization of her own childhood experiences, there are only a few
Significant relationships and they are not always in the direction predicted a priori.

As can be seen in Table 14, many of the variables have no relationships at all.

134

 

 

Table 14

’ '.-.III 1:0 5' ‘1' - i ' I'v- I- '-'v Or-..I'on If hilthIiI M‘Cfl',
WW

mm mm MHZ—611 _NoReeail ME IDIDLEE
ME

NegNeg" -.14* -.06 -.12 -.17* -.16*
NegPos° .02 -. 11 -.06 .06 -.03
NegNeutd .09 . 13 -.04 .00 .06
PosPos‘ —.19* -.05 -.10 -.16* -.16*
PosNegf .11 .04 .01 .12 .09
PosNeut‘ -.02 .04 -.01 .06 .02
NegProII .08 .05 -.02 .00 .03
CBRSI .18“ .08 .10 .21** .18*
CBRSII . 12 .02 -.05 .09 .05

 

Note. All the correlations are based on M = 133. 'Total Defensiveness. "# of Negative
statements correctly identified. ‘# of Positive statements judged to be negative. ‘# of
Neutral statements judged to be negative. at of Positive statements correctly identified.
’# of Negative Statements judged to be Positive. g# of Neutral statements judged to be

positive. blProportion of negative responses. '9 < .05, one-tailed. "o < .01, one-tailed.

135
Total Defensiveness is significantly inversely related both to the ability to

correctly identify Negative statements and to the ability to correctly identify Positive
statements. While the hypothesis predicted this inverse relationship between
Defensiveness and the number of positive behaviors judged to be positive, it also
predicted a positive relationship between Defensiveness and the number of negative
statements judged to be negative. Additionally, consistent with the a priori
predictions, Defensiveness is significantly related to the mothers’ ratings of behaviors
as aversive.

It is important to note that both Incoherence and Unresolve have the exact
same pattern of relationships to the same three perceptual variables as described above
for Total Defensiveness, both in terms of what was originally predicted and the actual
results.

- l i

Upon completion of the analysis reported above, a second statistical analysis
was undertaken on a sub-sample selected from the original 133 participants. The
rationale for this analysis is twofold; first, this second analysis provided an
opportunity to make the group of MCDB more homogeneous in terms of the type of
conduct disturbance. Secondly, the sub-sample analysis allowed for a testing of the
hypotheses between more "extreme" and/or distinct groups by selectively enhancing
the magnitude of the deviancy, as well as the level of "normalcy".

The literature reviewed in the introductory chapter clearly states that

disturbances of conduct are heterogeneous. There are at least two distinct types of

136

conduct disorders, an Unsocialized Aggressive and a Socialized Aggressive pattern.
Researchers have suggested that there are different developmental and family
processes which underlie the emergence of overt/confrontive antisocial behavior as
compared to covert/concealed antisocial behavior.

While the conceptual foundation for the present research is clearly based on
assumptions pertaining to what Patterson (1982) referred to as an " overt antisocial
pattern" or "Social Aggressors" , the determination of a conduct disturbance for the
original 71 participants included in the group of MCDB was based solely on the
mothers’ judgement of their sons as having a behavior problem. Hence, the type of
conduct disturbance was not determined and/or specified.

Even though the literature strongly suggested differentiating between the two
types of conduct disturbances, this was not, however, readily possible with the present
sample. The data from the CBCL indicated a strong relationship between the
Aggression (i.e., overt antisocial) and the Delinquency (i.e. , covert antisocial) scales,
I (133) = .69, o < .001. While this significant relationship is based somewhat on
overlapping items, it nevertheless suggests that making a theoretical distinction
between the types of conduct disturbances may be an easier task than the practical
differentiation, as the current data support the notion of a " mixed " type of conduct
disturbance (Loeber & Schmaling, 1985a).

The formation of a sub-sample did provide an Opportunity to "purify" or refine
the original sample in terms of the type of conduct disturbance under investigation.

The criteria used for inclusion in this refined conduct-disturbed group was based

 

del

CE

5165

relz

“it

Opp

.1-
I")

r
(I!)
H

ind ,

137
primarily on criteria established by Patterson (1982). He differentiated between

"Social Aggressors" and "Stealers"; the designation of a "Stealer" was based on the
criterion of was "stealing at high rates" as defined by "three or more events" (p.318).
In fact, infrequent incidences of stealing were reported for both the "A g gressors" and
the "Normals".

Thus, for the present sub-sample analysis "stealers" were excluded. The
determination of a " stealer" was based on the mother’s response to two items on the
CBCL; one item referring to stealing at home and the other item referred to stealing
outside the home. On the CBCL, if a mother indicated "frequently" to either of the
stealing-related items, or if she checked-off "occasionally" to both of the stealing-
related items, then the participant was excluded in the sub-sample analysis. A mother
was not excluded on the basis of checking-off "occasionally" to only one of the
stealing-related items.

The second rationale for the formation of a sub-sample was to provide an
opportunity to test the hypotheses under an "extreme case" situation, leaving no
doubt about the conduct disturbed status of the target child. Even though the data
from the CBCL, especially for the Aggression scale, indicated that the MCDB were
significantly different from the MNB, there was, however, considerable overlap in
the initial sample between the two groups in their distribution of Aggression scores.
For example, there were some MNB who scored in the upper percentiles on
Aggression, as well as those MCDB who said that their son was a behavioral problem

and yet checked-off practically no items on the Aggression scale. Hence, this sub-

II

138

sample analysis provided an opportunity to "clean up " the groups, creating more
homogeneity within each group, and thus a greater differentiation or distinction
between the groups in terms of their aggression versus non-aggression status (i.e.,
their conduct-disturbed versus non-conduct disturbed status).

The new groups were formed as follows: From the original sample of 133
participants, those mothers with the 30 highest scores on the Aggression scale of the
CBCL were included in the group of MCDB. (There were actually 36 mothers since
some of the Aggression scores had a frequency of more than one participant.) A few
of these mothers were excluded because the target son was determined to be a
"stealer" , or because during the initial phone interview the mother had not considered
her son to be a behavior problem. There were 28 mothers eventually included in this
group of MCDB for the sub-sample analysis.

The mothers that making up the group of MNB had the lowest 30 Aggression
scores from the original 133 participants. (Once again some of the scores were
assigned to more than one participant; thus, the 30 lowest scores included 35
mothers.) Five of these mothers were excluded because during the initial phone
interview they had detemiined that their son was a behavior problem at home. There
were 30 mothers that were included in this group of MNB for the sub-sample
analysis.

As expected, there was an extremely large difference in the Aggression scores,
as the MCDB M = 24.07, SD = 4.5) were significantly higher than the MNB M

= 3.63, SD = 3.1), i (47) = 20.09, p < .001, two-tailed. Additionally, there was

139

a similarly large difference between the groups in terms of the total amount of
problems described on the CBCL, as the MCDB M = 63.96, SD = 16.2) described
a significantly higher number of problems than the MNB M = 15.37, S_D = 10.8),
t(46) = 13.37, p <.001, two-tailed.

All of the original hypotheses were re-tested except for Hypothesis 10 which
focussed on issues of internal consistency for the AAQ. Table 15 shows the results
of Hypothesis 1 and Hypothesis 2. A few of the results are different from the initial
analysis. All of the originally significant results were significant for the sub—sample
except that the two groups did not differ in terms of their " response tendency". Also,
different from the intial analysis, the MCDB rated behaviors as significantly more
aversive than did the MNB, and the MCDB also rated behaviors as significantly more
desirable than did the MNB.

Once again there is no support for the notion of a "behavior classification
covariation". Table 16, however, does reveal some interesting relationships between
measures from different instruments. For example, the mother’s ratings of
aversiveness (i.e., CBRSI) are significantly related: (a) to a "negative response
tendency"; (b) to her proclivity to see neutral behavior as negative; and (c) is
significantly inversely related to the number of correctly identified positive behaviors.
Furthermore, the mother’s ratings of desirability (i.e., CBRSII) are significantly
related to her proclivity to judge neutral behavior as positive, and the number of
negative behaviors incorrectly judged as positive. These two rating scales (i.e., the

CBRSI and CBRSII) are also significantly correlated.

140
Table 15

Comoeo'son of the Deviancy Categorization Process Between MCDB and MNB in the
SultSampIIL

 

mm: M DB _N_BM

 

l. The MCDB correctly identified a

significantly greater number of negative M 37.89 34.80

behavioral segments than did the MNB. S_D 3.76 6.24
;(48) = 2.30‘

2. There was not a significant difference

between the MCDB and the MNB in terms of M .46 .36

the number of prosocial behavioral segments SD .63 .67

judged incorrectly as negative. I (56) = .57 n.s.

3. The MCDB judged a significantly greater M 10.89 7.53

number of neutral behavioral segments to be M) 4.89 4.55

deviant than did the MNB. t (56) = 2.71”

4. Overall, the MCDB identified a greater M 50.04 43.40

number of behaviors as negative than did the SD 7.11 9.80

MNB. 1(56) = 2.93"

14 1
Table 15 continued

 

Comparisons

 

5. Within each mother’s specific response rate,
the MCDB did not identify a significantly
greater proportion of behaviors as negative than

did the MNB.

6. The MCDB did rate behaviors as
significantly more aversive than the ratings of

the MNB.

7. The number of prosocial behaviors correctly
identified by the MNB was not significantly
different from those correctly identified by the

MCDB.

8. The was no difference between the MCDB
and MNB in their tendency to incorrectly
identify negative behavioral segments as

prosocial.

M 39.55 37.63
SQ 7.18 6.20

1(56) = 1.09 n.s.

M 141.75 123.26
SQ 21.88 24.29
t (56) = 3.04”

M 52.50 51.13
SQ 7.00 11.06

i(49) = .57 n.s.

l3

1.11 1.50

Ian
D

2.31 2.04

i(56)= .69 n.s.

142
Table 15 continued

 

 

Comparisons M DB MEI;

9. The MNB did not judge a significantly M 25.25 22.13
greater number of neutral behavioral segments SD 13.17 12.56
to be prosocial. t(56) = .92 n.s.

10. Overall, the MNB did not identify a M 78.89 74.77
significantly greater number of behaviors as SQ 19.85 22.38
positive than did the MCDB. ;(56) = 1.09 n.s.

11. Within each mother’s specific response M 60.45 62.37
rate, the MNB did not identify a significantly SQ 7.18 6.20

greater proportion of behaviors as positive than

did the MCDB.

1(56) = 1.09 n.s.

143

Table 15 continued

 

 

Comparisons M DB m
12. The MNB did not rate behaviors as M 198.50 181.07
significantly more desirable than the ratings of _SQ 25.19 36.27
the MCDB. t(56) = 2.11"

 

While this difference is statistically significant, it is not in the direction predicted.

'o < .05, two-tailed. "o < .01,

144
Table 16

:r I ' III_ B in 'Pr‘rIIrt V 'bls fr 'SUI-.

 

manual
2mm 1 2 3 4 5 6 7

 

1. NegNeg‘

2. NegPos" .19

3. NegNeutc .44" .28“

4. PosPos‘ .58“ .12 .26*

5. PosNeg‘ -.32** -.04 -.07 .16

6. PosNeutf .23“ .15 .33“ .67** .44**

7. NegPro‘ .12 .06 .25* -.66** -.53** -.71**

8. CBRSI“ .09 -.03 .28“ -.3l** -.12 -.15 .45“

9. CBRSIIi -.17 -.06 .04 -.05 .24* .25“ -.16 .39“
10.13'TQi -.04 .07 .04 .03 .11 .18 -.15 -.02

11.Responsek .59** .23* .54** .87“ .28“ .88“ -.56** -.12

.11

.10 .11

 

Note. All the correlations are based on M = 58. ‘# of Negative statements correctly
identified. b# of Positive statements judged to be negative. °# of Neutral statements
judged to be negative. “I! of Positive statements correctly identified. “19’ of Negative
statements judged to be Positive. f# of Neutral statements judged to be positive.
‘Proportion of negative responses. hThe mothers’ ratings of the degree of negativity
of negative behaviors. iThe mothers’ ratings of the degree of desirability of prosocial
behaviors. JThe Family Togethemess Questionnaire is an index of the family’s
enmeshment. kResponse is the total number of behaviors identified on the CBS and

is taken as an index of the mothers’ responsivity.
‘o < .05, one-tailed. "o < .01, one-tailed.

145
The MCDB M = 10.5, SQ = 4.0) did describe their families as significantly

more enmeshed than the MNB M = 8.3, SQ = 2.5), t (45) = 2.48, o < .05, two-
tailed. While the MCDB M = 129, SQ = 21.5) did respond to more behaviors than
did the MNB M = 118, SQ = 29.4), this difference was not statistically different,
1 (56) = 1.58, o > .05. Also, as can be seen in Table 16, there is no relationship
between responsivity and enmeshment. In fact, the level of enmeshment was not
significantly related to any of the perceptual variables.

As can be seen in Table 17, the childhood recollections of the MCDB did not
exhibit a significantly different amount of problematic emotional experiences in
childhood than did the recollections of the MNB. The only exception was that the
MNB did experience a significantly greater amount of role-reversal.

While the content of these memories did not differ, similar to the initial
analysis, the MCDB did produce memories that were significantly more defensive and
disorganized than the recollections of the MNB. As compared to the MNB, the
MCDB childhood recollections were significantly more incoherent, were more
unresolved regarding problematic emotional childhood experiences, and they exhibited
a significant lack of specific details. Table 18 has the results of these comparisons.

Although frequently not in the direction of the a priori predictions, Table 19
shows that there are some relationships between the affective-cognitive organization
of childhood memories and the perceptual variables. For example, the composite
score of Total Defensiveness is significantly inversely related to the "accuracy"

scores, i.e, correctly identified positive behaviors and correctly identified negative

146

 

 

Table 17
II I.«._r-Ir fIr ‘ I- -. ole B 111 ‘ M DB I I ‘ MNB In ‘ Inten

0 Th' ' h R llec'on

Contendeales: M DB m t— (510

Unloving M 7.34 7.27 . 12 (56)
_SQ 2.00 2.11

Rejection M 6.57 6.64 . 13 (56)
SQ 2.08 2.07

Neglect M 4.67 4.56 .18 (56)
E 2.21 2.85

Role-Reversal M 4.41 5.5 2.34“ (56)
SQ 1.61 1.95

Pressure to

Achieve M 4.09 4.52 .71 (56)
SQ 2.11 2.52

Abuse M 5 .55 4.75 1.35 (56)
SQ 2.27 2.22

Total Problematic

Relations M 31.66 32.69 .74 (131)
S_D 8.35 7.72

 

Note. All of the scales were scored in the same direction with higher scores

indicating greater problematic experiences.

'9 < .05

147

 

 

Table 18
II It; I B . I-S Ile ofM DB I. I ‘ MNB on th Aff tiv-
' 'v ' fTh ' hi1 h c 11

55913.5; M DB L“ I: (SID

Incoherent M 3 .78 3. 10 2.54” (56)
S_Q .91 .97

Unresolve M 3.64 3.10 1.93'" (56)
S_Q .96 1.16

Idealized M 2.67 2.49 .56 (56)
SQ 1.31 1.15

Lack of Recall M 3.18 2.72 1.67. (56)
SQ .96 1.05

Total

Defensiveness M 13.28 11.41 2.04" (56)
SQ 3.18 3.74

 

Mote. All of the scales were scored in the same direction with higher scores

indicating greater affective-cognitive disorganization or defensiveness.

‘fl = 58.

’o < .05, one-tailed. ”o < .05, two-tailed.

148

 

 

Table 19

’t‘ !°l LI : Aff 'Iv- I'mIV‘ Ort. 'zaII'n If hilIIhIlI Mmoris
..-II J' Def, D‘v. ..;°I.Q..IInP,I; ; '1 - .I- ., .1-
Bergem Ineohereni idifl'eeo No Reg! unreeolveo Ioneieri:
NegNegb -.17 -.14 -.16 -.27* -.23*
NegPos° .07 .00 -.06 .02 .00
NegNeutd .18 .20 .01 .07 . 15
PosPos‘ -.29* -.19 -.35** -.22* -.32**
PosNegf .12 .12 -.07 .17 .11
PosNeut‘ . 15 .07 -.02 . 19 . 12
NegProII .09 . 12 .14 -.04 .09
CBRSI .43" .19 .38" .31“ .40"
CBRSII .16 .11 .15 .03 .14

 

Note. All the correlations are based on M = 58.

‘Total Defensiveness. b# of Negative statements correctly identified. °# of Positive
statements judged to be negative. “I! of Neutral statements judged to be negative. °#
of Positive statements correctly identified. 9:, of Negative statements judged to be
Positive. 8# of Neutral statements judged to be positive. "Proportion of negative

responses. ’2 < .05, one-tailed. "p < .01, one-tailed.

149

behaviors. Total defensiveness is also significantly correlated with the mothers’
aversiveness ratings. The dimensions of incoherence, inability to recall, and
unresolve, all exhibit a similar, but not identical, pattern of results.

In summation, the results of the sub-sample analysis are as follows: The
MCDB are overly inclusive in their categorization of deviancy, but not exclusive in
their categorization of prosocial behaviors. There are no differences between the two
groups of mothers in terms of their classification of prosocial behaviors. And, there
is no indication that an emphasis in classifying child behaviors as deviant has an
inverse relationship with an emphasis on classifying behaviors as prosocial. Different
than the initial analysis, the MCDB did, as predicted, rate behaviors as significantly
more aversive; and, contrary to the original predictions, the MCDB also rated
behaviors as significantly more desirable. Finally, the MCDB describe their families
as more enmeshed, but enmeshment has no significant relationship with any of the
other perceptual variables.

The results of the sub-sample analysis indicate that the two groups of mothers
do not exhibit any significant differences in terms of the content of their childhood
recollections. The MCDB do, however, produce childhood recollections that are
significantly more defensive and disorganized than those produced by the MNB.
Finally, although basically not in the direction originally predicted, there are some
significant relationships between the mother’s affective-cognitive organization of

childhood memories and her perception of child behaviors. The Lack of Detail scale

150

exhibits this significant relationship to the perceptual variables only in the sub-sample
analysis.

It is important to point out that these results from the sub-sample analysis are
basically the same pattern of findings as the results from the initial analysis. The
differences between the two analysis are slight and subtle. The main effect of
"purifying" the sample was a slight general increase in the effect size (i.e., the
magnitude of significant correlations was increased and/or the significant differences
between the group means were enhanced). The main exception to this generalization

is that in the sub-sample analysis the MCDB gave higher ratings to both deviancy and

prosocial behaviors.

CHAPTER4

DISCUSSION

Throughout this study, the ideas presented and explored have been organized
around three main groupings: maternal perception, childhood recollections, and the
"linkage" or relationship between perception and the childhood recollections. Within
each of these three respective domains, the statistical results are equivocal with a
mixture of both supportive and non-supportive findings. For example, while there
is not total support for the deviancy categorization hypotheses, there is, however,
evidence to support the notion that MCDB are more inclusive in their classification
of deviant behaviors and they are more "sensitized" to deviant behaviors. And, while
the MCDB do not exhibit differences from the MNB in the content of their childhood
memories, the affective-cognitive organization of these recollections does appear to
be significantly more defensive/disorganized. Finally, even though the hypothesis
predicting a relationship between the mothers’ deviancy classification schema and the
affective-cognitive organization of their childhood memories is not supported, there
are, however, significant findings that do offer support, albeit suggestively/tentatively,
for the underlying notion of linkage, and thus for the construct of an internal mental

model.

151

152
Overall, the conceptual framework proposed in the introductory chapter cannot

be accepted in its entirety since a substantial number of the hypotheses are not
supported. On the other hand, there are a variety of significant findings that do
support/confirm the a priori predictions, as well as a number of results that are both
interesting and germane to the issues raised in the Introduction.

This chapter will reconcile these statistical results with the conceptual model
being investigated. Within each group of hypotheses, it will be useful to: (a) examine
the pattern of significant and non-significant findings, (b) present the pos hoc analyses
undertaken to help clarify some of the results, (c) explore various methodological
issues, and (d) raise relevant questions\issues and propose potential directions for
further research. Finally, this chapter will conclude with a brief summary and
discussion regarding the implications of this research.

Throughout this chapter the interpretation of the data will be approached with
a fair degree of caution and the findings will generally (but not completely) be
regarded as preliminary, suggestive, and encouraging of further research. The
rationale for this conservative perspective in the interpretation of the data is as
follows: The sheer number of statistical tests performed during the analysis of the
data inflates the probability of a Type I error (Myers, 1979). Hence, even though
both the actual number of significant findings and the pattern of these significant
results (i.e., many of the significant results fit together suggesting a conceptual
cohesiveness/coherency) argue against the likelihood of these being spurious findings,

a degree of caution or tempered enthusiasm, nevertheless, still needs to be exercised

153
in the interpretation of this data. Additionally, while many of the hypotheses were

directional and therefore the use of one-tail tests was appropriate, some of the
hypotheses, especially in the large sample analysis, would not have been supported
or accepted if a two-tail criterion had been applied. This suggests that the "effect
size", or the actual differences between the means is, in some cases, actually quite
small. Therefore, patterns of significant results will be emphasized rather than
focusing on any one specific significant finding. Finally, much of the methodology
employed in this research, while based on previous research, was created specifically
for this study (i.e., the instruments were being used for the first time in this study).
This lack of a proven track record for many of the instruments also favors a
conservative approach towards the interpretation of the data.
Mejem Perception

Many of the results examined in this section are from hypotheses which were
derived directly from Patterson’s (1982,1986b) observations and conceptualizations;
specifically, his position that mothers of antisocial children are overly enmeshed and
very irritable, and have a perceptual style (i.e. , deviancy-classification schema) which
promotes negative interactions (i.e., these parents tend to classify many more child
behaviors as deviant and also tend to view many of the child’s acts as worthy of high
intensity scolding). As reviewed in the introduction, Patterson hypothesized that this
inclusiveness in the classification of deviancy is combined with an exclusiveness in
the classification of prosocial behavior; thus further contributing to the

coercive/negative and irritable ambience that characterizes these families.

154

It will be helpful, in the following examination of the results, to make a
distinction between the mother’s deviancy classification processes, and her ability to
track or identify specific child behaviors. These can be conceptualized as two
distinct, yet overlapping, cognitive processes. Obviously, it is not always possible
to determine the specific influence between tracking skills and the deviancy-
classification schema since these cognitive processes can be so highly interrelated.
Nevertheless, it seems reasonable to propose that the ability to correctly identify a
behavior, either as deviant or prosocial, while reflecting both cognitive processes, be
interpreted as primarily an indication of tracking. On the other hand, classifying a
neutral behavior as either positive or negative is more likely to reflect an underlying
deviancy-classification schema. Additionally, "errors" in labelling (e. g., classifying
a deviant behavior as positive) are also very likely to be an indication of one’s
classification schema.

There is a pattern of findings which clearly indicates that the MCDB have both
an overly inclusive cognitive schema for categorizing child behaviors as deviant and
are more sensitized and attentive in the tracking of deviant behaviors. (This is one
of the areas of significant findings that do not need to be interpreted with excessive
caution.) Consistent with previous findings (Lorber, 1981), the MCDB did rate a
significantly greater number of behaviors to be deviant. In the initial analysis, these
mothers identified a significantly greater proportion of behaviors to be deviant, as
compared to the MNB; thus, exhibiting a "negative response tendency". (This

proportion score is the percentage of all behaviors identified as deviant calculated for

155
each mother’s specific response rate.) Patterson (1986b) reported that in the Lorber

study, the differences were not in terms of accuracy, but rather the mothers of
antisocial boys tended to judge behaviors as deviant that were not classified as deviant
by trained observers. In the present study, however, the MCDB do demonstrate a
significantly greater accuracy in their tracking of deviant behaviors, as well as a
significantly greater tendency to also classify neutral child behaviors as deviant.
Thus, the greater overall number of behaviors rated as deviant is the result of both
tracking and the deviancy-classification schema.

The superior ability of the MCDB to identify/recognize deviancy suggest that
these mothers are primed and highly attuned to respond to any, and practically all,
displays of child misbehavior. For example, in the sub-sample analysis, the MCDB
had an average "hit" rate for deviant behaviors of 90% (i.e., correct identification of
deviant behaviors) compared to an 82% hit rate for the normal mothers. It’s as if no
deviant behaviors escape these mothers’ attention.

The interpretation of these findings can be turned around to suggest that
normal mothers exhibit an insensitivity to deviancy. The results suggest that normal
mothers do not track deviant behaviors with the same vigilance as do the MCDB. As
noted in the introductory chapter, Holleran et al. (1982b) reported that lower
sensitivity to negative behaviors on the tracking task was associated with higher rates
of positive behaviors in the home. They suggested that the normal parents’
"insensitivity" to negative behavior may be in some way responsible for the positive

social atmosphere in the home. It was also suggested previously that sometimes the

156

most appropriate response to negative child behavior is no response (Patterson, 1982;
Sallow, 1972).

The MCDB also exhibit a significantly greater inclusiveness in their
classification of deviancy. In both the initial and sub-sample analysis, these mothers
classified a significantly greater number of neutral behaviors to be deviant, as
compared to the normal mothers. In the sub-sample, the MCDB responded to 14%
of the neutral behaviors as if they were deviant, compared to only 8 % for the normal
mothers. This finding strongly supports the position that the MCDB have an overly
inclusive deviancy-categorization schema.

Relatedly, as mentioned in the Introduction, one of Patterson’s (1982)
measures of the mother’s irritability is a "crossover". This refers to the mother
initiating an aversive interaction (i.e., when she reacts negatively to the child’s
prosocial or neutral behavior). The findings regarding the tendency of the MCDB to
view neutral behaviors as deviant promotes the perspective that the mother’s
irritability, as evidenced by her "crossover", may very likely be, in large part, the
result of an underlying labeflmg or categorization process.

The MCDB not only perceive a wider range of child behaviors to be deviant
and are inclined to be more sensitive/vigilant towards these deviant behaviors, but
furthermore, these mothers also appear to be more highly "reactive" to displays of
child misbehavior. This final contention is supported by the findings in the sub-
sample analysis that the MCDB rated child misbehavior to be significantly more

aversive as compared to the ratings of normal mothers. (In the large sample analysis

157

there were no differences between the mothers’ ratings.) This would support
Patterson’s observation that these mothers react strongly at the slightest provocation
and tend to view trivial instances of child misbehavior as worthy of high intensity
scolding.

It is reasonable to conceptualize the existence of a behavior-discrimination
schema and to speculate that this cognitive system has a direct effect on the mothers’
responsiveness. In other words, there are discriminatory functions involved in the
cognitive processes of making a series of differentiations and evaluations that
ultimately lead to the mothers’ ratings regarding the severity of the child’s
misbehavior. Furthermore, the outcome of these cognitive processes (i.e. , the
ratings) is likely to play a critical role in governing a mother’s level of
responsivity/reactivity. For example, the affective response to a misbehavior judged
to be "extremely serious" is bound to be quite different than the reaction to a behavior
judged to be "a slight misbehavior". Thus, the cognitive processes are directly linked
with the modulation of the mothers’ affective response. This line of speculation
implies the importance of a cognitive-affective dimension that is likely to be involved
in determining the overall quality and intensity of a mother’s responsiveness.

Clearly, a meaningful pattern of results emerges when examining the five
significant findings: The results for the MCDB exhibit (a) a greater number of
behaviors identified as deviant, (b) a greater proportion of all identified behaviors
identified as deviant behaviors, (c) a greater number of deviant behaviors correctly

identified, ((1) a greater number of neutral behaviors classified as deviant, and (e)

158

higher aversiveness ratings. Certainly, this pattern of results is supportive of
Patterson’s observations and speculations. In other words, compared to normal
mothers, the MCDB perceive a greater number of behaviors to be deviant, and there
is a greater likelihood that the MCDB will perceive a neutral and/or relatively benign
child behavior as a misbehavior, judge the "misbehavior" to be relatively more
serious, and hence initiate an aversive disciplinary interaction with the child. It is in
this way that the mothers’ deviancy classification schema (and their behavioral-
discrimination schema) combine to increase the probability of an aversive mother-
child interaction and thus contribute directly to the aversive and highly irritable
atmosphere that characterizes the families of conduct disturbed children. It is
reasonable to conclude that the MCDB perceptual style and, relatedly, their response
style are very likely implicated in the maintenance (and possibly the development) of
the child’s conduct disturbance. (It remains an open question whether the mother’s
perceptual style is the cause, consequence, or cause and consequence of the child’s
behavioral disturbance.)

Even though a very consistent picture has unfolded, it is nevertheless
imperative to highlight some aspects of the results that are not necessarily supportive
and compatible with this overall portrayal. The most striking feature of the proposed
model that has virtually no support in the data is the lack of differences between the
two groups of mothers in terms of their perception and categorization of prosocial
behaviors. Inherent in the deviancy-categorization hypothesis is the contention that

the MCDB inclusiveness in the classification of deviancy is combined with an

159

exclusiveness in the classification of prosocial behaviors. Underlying this perspective
is the assumption of a "behavioral classification covariation".

While the MCDB exhibit an inclusiveness in classifying deviancy, this
"sensitivity" is not at the expense of an insensitivity (or exclusion) towards the
classification of prosocial behaviors. There are no differences between the two
groups of mothers in terms of the overall number of prosocial behaviors identified,
the number of prosocial behaviors accurately identified, or the number of neutral
behaviors classified as prosocial. The only significant difference (found only in the
sub-sample analysis) was that MCDB rated prosocial behaviors to be significantly
more desirable. This result, to be discussed below, is contrary to the proposed
model. Clearly, important elements of the deviancy-classification hypothesis are not
supported as the MCDB do not exhibit an exclusiveness (or insensitivity) in the
classification of prosocial behaviors.

Relatedly, the underlying assumption of a behavioral classification covariation
is generally not supported. There is virtually no support for the position that an
emphasis on deviancy will covary with a de-emphasis in the classification of prosocial
behaviors and vice versa. The results reported in the previous chapter (and only
minimally modified by the post hoc analysis reported below) exhibited three
Significant positive relationships rather than inverse as predicted.

Three out of the four correlations performed in the analysis of the behavioral
classification covariation hypothesis used variables derived from the CBS. As can be

seen in Table 9, all of the perceptual variables derived from the CBS are significantly

160
related to the level of responsivity. The more the mother responded and

underlined/identified behaviors on the CBS, the higher her scores were on all of the
variables computed. Hence, the correlation table is so conspicuously influenced by
response rate that it is not possible to determine if the variables covary due to some
underlying dimension other than response rate. In order to make this determination,
partial correlations were computed, holding the effects of responsivity constant. The
results of this post hoc analysis are presented in Table 20.

Even though the results from this post hoc analysis do not affect the outcome
of this hypothesis, the partial correlations do, however, help to clarify some of the
data. For example, the partial correlations demonstrate that in fact there is no
relationship between correctly identifying positive behaviors and correctly identifying
negative behaviors. Nor is there a relationship between incorrectly classifying
negative behavior as positive and incorrectly classifying positive behaviors as
negative. The partial correlations did, however, reveal that the mother’s tendency to
classify neutral behaviors as negative has a significant inverse relationship to her
tendency to classify neutral behaviors as positive. This being the only significant
result of the post hoc analysis, the decision remains not to accept this hypothesis.
Finally, if the classification of deviancy does not covary with the classification of
prosocial behaviors, then this raises an interesting issue as to whether there are
independent classification schemata that may have different functional and/or

structural properties?

1 61
Table 20

Peggi’ Qogeieo'on Coefficients Between the PerceoMal Variables of the CBS Wim the

fR nivi H1 n

 

Esteem
Madame 1 2 3 4 5 6

 

l. NegNeg‘

2. NegPos" .07

3. NegNeut‘ .24* .13

4. PosPos‘ -.O9 -.24* -.45**

5. PosNeg‘ -.49** .06 -.19* -.19*

6. PosNeut‘ -.69"'* -.11 -.44** -.36** .30"

7. NegPro‘ .70“ .20"' -.75** -.56** -.32** -.52*"'

 

Note. All the correlations are based on _N_ = 133

‘# of Negative statements correctly identified. "# of Positive statements judged to be
negative. °# of Neutral statements judged to be negative. ‘# of Positive statements
correctly identified. °# of Negative statements judged to be Positive. ’# of Neutral
statements judged to be positive. gGiven each mother’s individual total response rate,
this is the proportion of her negative responses.

’12 < .05, one-tailed. ”o < .01, one-tailed.

162

In the analysis of the behavioral classification covariation hypothesis one of the
unexpected findings was a significant relationship between the mother’s ratings of
aversiveness and her ratings of desirability. The explanation of this curious finding is
open to speculation. The data suggest that a proclivity to be excessive in the
interpretation or evaluation of deviancy is related to a proclivity to be excessive in the
interpretation or evaluation of prosocial behaviors. Perhaps the positive correlation
between these two rating scales indicates the underlying cognitive-affective dimension
of a behavioral-discrimination schema (proposed above) involved in formulating
evaluations and interpretations.

There are other variables in the data that can be thought of as reflecting this
cognitive dimension involved in evaluative and interpretive processes. Of particular
interest are "over-interpretations" which are defined as the tendency to attribute
meaning, significance, clarity, and/or certainty to the perceptual field. For example, the
classification of neutral behaviors as either positive or negative can be comprehended as
an instance of an over-interpretation since the mother is attributing meaning to a situation
which is, at best, ambiguous, and perhaps even meaningless.

Even though completely speculative, there are some significant correlations,
nonetheless, across different instruments, which lend support, albeit modestly, to this
notion of an underlying cognitive dimension involved in evaluative and interpretive
processes. For example, Table 9 and Table 16 show that a statistically significant
relationship does exist between the tendency to over-interpret neutral behaviors as

deviant (i.e., the CBS) with the excessive evaluation of deviancy (i.e., CBRSI). A

163

significant relationship also exists between the over-interpretation of neutral behaviors
as positive and the excessive evaluation of desirability (e.g. , CBRSII). It is important
to note that the partial correlations Show that interpreting neutral behavior to be
deviant is significantly inversely related to interpreting neutral behaviors to be
positive.

While clearly conjecturable, the data can be understood to suggest that there
is both a general tendency to over-interpret, as well as some evidence which points
towards systematic interpretive biases, either towards deviancy or prosocial,
respectively. There are some trends in the data, including findings already reviewed,
to suggest that MCDB have a greater tendency to over-interpret and that there is a
systematic bias towards attributing negative meaning to the child’s behavior. For
example, in the sub-sample analysis, the MCDB M = 36.14, S_D_ = 14.95)
responded to more neutral behaviors (both positively and negatively) than the normal
mothers M = 29.67, SQ = 15.04) and this difference approached significance, t
(56) = 1.64, o < .053, one-tailed. Furthermore, as mentioned above, the MCDB
differed from the MNB in terms of a greater classification of neutral behaviors as
deviant and in the sub-sample analysis, these mothers gave significantly higher ratings
to both deviant and prosocial behaviors suggesting that their evaluations of child
behaviors are more extreme, regardless of the negative or positive value of the
behavior.

Undoubtedly, this is a Speculative line of reasoning, but one that nonetheless

raises some interesting questions for further research. The most obvious question is

164

why do some mothers over and/or misinterpret? It would make for a fascinating
study to explore the nature of these attributions in order to establish if there is some
systematic bias that is involved in the over-interpretations. For example, is the
attribution of specific meaning to a neutral behavior based on the mother’s affective
state, therefore indicating that the distortions are situationally determined, or, are they
based on her own developmental and/or personality issues? Hence, are some of the
over-interpretations "projective" in nature, or are they reactive? Furthermore, is the
mother’s responsiveness to an apparently benign child behavior frequently
personalized? In other words, are the mother’s negative attributions regarding the
meaning of the child’s behavior frequently combined with the attribution of intent
(e.g., "he just does that to annoy me")? These issues go beyond a simple construct
of a behavioral-classification and/or discrimination schema in an attempt to elaborate
upon the conceptualization and expand our understanding of these hypothetical
cognitive-affective processes. In order to grasp the complexity of the processes
involve, it will be necessary for future research to address issues related to the
specific nature, quality, and possible " motivations" of this cognitive-perceptual and
interpretive system.

The final step in this line of speculation is to propose that these over-
interpretations lead to a higher level of reactivity and that the MCDB higher reactivity
level as observed by Patterson (1982) (and suggested by the current data) may play
a role in determining the mothers’ "inept family management skills" (Patterson, 1982,

1986a,1986c), specifically in terms of discipline. As proposed above, there is a

165

cognitive-affective dimension which largely determines the mother’s reactivity level.
The evaluation and interpretation of the child’s behavior is likely to govern and
modulate the intensity of the mother’s affective response. Consistent negative
attributions regarding the value and severity of the child’s behaviors is likely to lead
to consistently higher levels of intensity and reactivity. Over time, the mother’s
higher levels of reactivity are likely to lose their effectiveness. Analogously, in
listening to loud music there is a sensory adaptation that takes place so that the music
no longer sounds loud. If a mother responds with intensity to a relatively mild
misbehavior, how does she increase her impact for a serious misbehavior? Clearly,
mothers who respond with a generally higher level of intensity to all behaviors, both
positive and/or negative, must lose their effectiveness in terms of both discipline and
their "parental reinforcement value" (Snyder & Patterson, 1987; Wahler, 1969). In
summation, the hypothetical process begins with the mother’s tendency to over-
interpret her son’s behavior which then arouses her to a generally higher level of
responsivity with an intensified reaction. Over a period of time, the mother’s
relatively unmodulated (and/or undiscriminating) responses subsequently lose their
effectiveness and hence, this results (perhaps interacting with a history of
noncontingent reinforcement (Snyder, 1977; Wahler, 1969)) in the child’s observed
nonresponsiveness (Patterson, 1982; Snyder, 1977).

Previous research has consistently demonstrated that parents of behaviorally
disturbed children are generally more noncontingent in their responses. The current

results offer mixed support for this characterization and some of the nonsupportive

166

results are very likely due to a methodological issue involving the CBS. It is the
position of this study that the mother’s classification of the child’s behavior
determines, in large part, the quality and intensity of her response and thus, the
"misclassification" of a behavior might easily result in a noncontingent response. For
example, the MCDB tendency to be excessive in their evaluations of the child’s
behavior and their greater proclivity to perceive relatively benign or neutral behaviors
as deviant are both likely to be cognitive-perceptual factors which play a critical role
in determining the mothers’ noncontingent responses; an example of which would be
Patterson’s (1982) "crossover ".

On the other hand, there are other perceptual "errors" not found in the data
which would also logically lead to noncontingent responses. For example, there is
no support for the position that MCDB are more likely to provide positive
reinforcement for negative behavior (Patterson, 1982; Sallows, 1972/1973; Shaw,
197 1/ 1972; Snyder, 1977) Since there are no differences between the groups of
mothers in terms of their tendency to incorrectly identify a deviant behavior as
prosocial. Similarly, there are no data congruent with the position that MCDB are
more likely to punish prosocial behavior (Patterson, 1982; Shaw, 1971/1972) since
the groups did not differ in terms of the incorrect identification of prosocial behavior
as deviant. It is very likely that these two measures did not differentiate the two
groups of mothers because the CBS did not elicit, as intended, these specific types of
perceptual "errors". For example, in the entire sample (i.e., bl = 133) there were

virtually no examples of incorrectly identifying negative behaviors as positive (i.e.,

167
M = .38; SQ = .72). Improving the instrument so that it has a greater ability to

elicit these types of "errors" would be very helpful for future research. Having data
for these specific types of perceptual errors would help to explore issues of perceptual
bias and distortion and might help to further elucidate the relationships between
parental perceptual style and parental response, specifically how noncontingent
parental responses may be based on the "erroneous" perception of the child’s
behavior.

Even though specific recommendations regarding further instrument
development are required, overall, the CBS was a valuable research tool, especially
in terms of illuminating the MCDB’ negative response tendency. A review of Table
20 does show a variety of expectable relationships among those scales of the CBS and
this offers further support for the validity and internal consistency of this instrument.
Many of the variables which involve the classification of behaviors as negative are
significantly related with each other and are significantly inversely related to those
variables measuring the classification of behaviors as positive. For example, the
correct identification of a deviant behavior is significantly related to the classification
of a neutral behavior as deviant and the overall proportion of behaviors identified as
deviant, respectively; while exhibiting significant inverse relationships to both the
number of neutral and negative statements identified as prosocial, respectively.

This last correlation supports the dimension of a perceptual accuracy for the
classification of deviancy and demonstrates that as accuracy (i.e. , correctly identifying

deviant behaviors) increases, inaccuracy (i.e., identifying deviant behaviors as

168

positive) decreases. (This logical relationship is not an inherent outcome of the
measurement system as it was possible that "accuracy" could demonstrate a different
association with "inaccuracy" .) A similar inverse relationship between accuracy and
inaccuracy exists in regards to the classification of prosocial behaviors, but there are
no statistical relationships which suggest an overall dimension of a behavioral
classification accuracy (i.e., the correct identification of deviancy is not related to the
correct identification of prosocial behaviors.) Again, this suggest independent
classification schematta.

Furthermore, although modest, there are, however, a few significant
relationships among variables across different instruments, mentioned above, and this
provides support for the construct validity of both the CBS and CBRSI and CBRSII.
(See Table 9 and Table 16.) For example, the higher the mother’s ratings of
deviancy on the CBRSI, the more neutral items she identified as deviant, the greater
proportion of behaviors identified as deviant, and in the sub-sample analysis, the less
accurate she was in identifying prosocial behaviors. This indicates an overall
negative response tendency or proclivity. And, the higher the mothers ratings of
desirability on the CBRSII, the more neutral items she identified as prosocial and the
more deviant behaviors she incorrectly identified as prosocial. This suggests an
overall positive response tendency.

While there is much support for the CBS, not all of the intercorrelations,
however, are readily explainable. This is particularly true in the pattern of

correlations between those variables involving the classification of behaviors as

169

positive. For example, it is not readily apparent why the accurate identification of
positive behavior is inversely related to classifying neutral behaviors as positive or
identifying deviant behaviors as positive, respectively. It is unclear whether this
reflects a measurement issue or some actual underlying dimension of the cognitive
schema involved in the classification of prosocial behaviors.

Finally, the remaining hypotheses within this group deal with aspects of the
model that predicted that the MCDB would described their families as more enmeshed
and that this enmeshment is related to various dimensions of the mothers’ perceptual
style. The concept of enmeshment denotes a lack of clear individual "boundaries"
and less individual privacy/space, with a greater emphasis on interdependence; all
together this implies higher levels of interaction and reactivity. This is consistent
with Patterson’s (1982) observations that families of conduct disturbed boys are more
enmeshed and highly irritable or reactive.

While it is true that MCDB do describe their families as significantly closer
(i.e., more enmeshed) than normals, this characterization of family life, however, has
no apparent relationship to the mothers’ perceptual style. For example, the data
provide no support for a relationship between family closeness and the mother’s
responsivity level (i.e., response rate on the CBS). Additionally, there is no support
for the characterization that enmeshment is related to a negative response style (i.e.,
irritability). In fact, there are a few extremely modest significant associations
suggesting that the closer the family, the more positive the response style. For

example, the closer the description of the family: (a) the more behaviors that were

170
classified as prosocial (l: (133) = .15, o < .05,one-tailed; (b) the greater the mother’s

tendency to perceive neutral behaviors as positive, and (c) the greater her positive
response tendency (r = .16, o < .05, one-tailed).

A closer examination of the actual scores on the FT Q indicates that while the
MCDB did describe their families as significantly closer than the descriptions of the
normal mothers, the actual scores do not justify the characterization of these families
as enmeshed. In the sub-sample analysis, where the MCDB exhibited the highest
FTQ scores, the group mean score was 10.5. This is a five-item scale and thus the
average response to each item was a 2.1. This rating represents the choice of "fairly
urine for our family." In other words, the MCDB clearly do not describe their
families as " insisting upon family interdependence at the expense of individual
autonomy" (Bloom, 1985, p.237) and therefore, it is inaccurate to refer to these
families as enmeshed. (If the overall scores had represented the choices of "fairly
true" or "very true of our family", then this characterization might have been
appropriate.) Furthermore, the actual differences between the two groups on the FT Q
is in terms of the MCDB giving scores of "f_aMy untrue" versus the MNB’ scores of
"my untrue". It is not clear what this difference means and it seems unlikely that
it actually represents a qualitative difference in terms of family functioning.

One of the structural properties which characterizes families is that of family
boundaries which are rules that define who participates in interactions and the extent
of that involvement. This assessment is made along a continuum between

"disengaged" at one extreme (indicating inappropriate and rigid boundaries with a

171

lack of interdependence and interactions), through a "normal" range of clear
boundaries, and ending at "enmeshment", the other extreme, indicating diffuse
boundaries with a heightened sense of belonging at the sacrifice of individual
autonomy (Minuchin, 1974). The FT Q does not assess the family’s functioning
throughout the entire continuum but rather examines only the extreme end of
enmeshment.

It is reasonable to conclude that this restricted range of the FTQ consequently
did not allow for an adequate test of those hypotheses exploring the potential
relationship between family functioning and the mother’s perceptual style. It is
incumbent upon future research to assess the quality of the family’s boundaries across
the entire continuum. This will allow for a more thorough examination of the
following issues: Given the full range of the continuum, are families of conduct
disturbed boys more towards the emneshed end of the continuum? Is there a
relationship between the family functioning/structure in terms of boundaries and the
mother’s perceptual style, especially in regards to her level of responsivity and
irritability?
M 011 c 'ons

Throughout this study there has been an emphasis on the critical role played
by parental affective-cognitive factors in determining parent-child interactions. It is
assumed that the mother’s perceptual and response style is the product of a mental
model regarding parent-child interactions. Essential to this perspective is the position

that this parent-child mental model is based largely on the mother’s own childhood

172

experiences. The hypotheses in this section examined the quality and parameters of
this mental model by exploring both the content and the organizational properties of
the mothers’ own childhood recollections.

The results indicate that while the content of the childhood recollections does
not differ between the two groups of mothers, the organizational properties, however,
do appear to be related to the psychopathology of the child. The MCDB produced
childhood recollections that were significantly more disorganized and defensive than
those memories of the normal mothers. Furthermore, even though the initial results
do not offer support for Main and Goldwyn’s (1985) contention that problematic
childhood experiences, specifically rejection, are related to a defensive organization
of this information, a post hoc breakdown and re-analysis of the same data does,
however, offer supportive evidence consistent with Main’s findings. Finally, it will
be useful in this section to include some comments on methodological considerations
regarding the use of the AAQ and the accompanying scoring system.

In certain aspects, the findings that the content of the childhood memories did
not differ between the two groups of mothers are contrary to current common
wisdom. The widely held belief that there is a direct relationship between childhood
experiences and adult functioning, specifically parental functioning, is not promoted
by these results. There is no support for the position that MCDB, as compared to
normals, have had more problematic childhood relationships as these mothers do got;

appear to be lacking in certain " supportive developmental experiences" (Belsky,

1984).

mc
res,
exp

0111

hit:
Ol’cf;

8131:

info”
15.1011
ECO”:

tents ‘

173
In both the initial and sub-sample analysis, the only significant finding was that

the MNB experienced a significantly greater amount of role-reversal. In the initial
analyses each of the content/scale scores (e. g., Unloving, Rejection, Neglect, etc.)
were actually combined total scores as the participant’s experiences with her mother
was combined with the participant’s experiences with her father. In a post hoc
analysis these scores were analyzed separately. Once again the results indicated that
there were virtually no differences between the two groups in the quality of the
mothers’ childhood relationships with their own mother and with their own father,
respectively. The only significant result was that the group of normal mothers
experienced a significantly greater amount of "being pushed to achieve" from their
own mothers, as compared to the MCDB, 1 (110) = 1.88, o < .05, one-tailed.

On the other hand, the data clearly indicate that the MCDB produced
childhood recollections that were significantly more defensive and disorganized. Even
though some of the specific significant findings exhibited only small mean differences,
taken as a whole, however, the pattern of significant results advances a consistent
overall picture. This portrayal shows that the flow and accessibility of relationship-
related information was significantly more restricted for the MCDB.

The childhood recollections of the MCDB demonstrated a relative lack of
information accessibility (i.e., Lack of Specific Detail) and appeared to be more
disjointed, splintered and contradictory (i..e., Incoherent). Relatedly, their
recollections demonstrated a relatively greater lack of resolution, integration, and

hence, a lack of separation and individuation regarding childhood experiences and

174

relationships (i.e., Unresolved). It is in this way that the parent-cth mental model
appears to be more defensive. In other words, the MCDB parent-child mental model
can be characterized by the exclusion and/or restriction of information, by a lack of
integration and synthesis, and by the disorganization, splintering, and fragmentation
of relationship-relevant information.

These results from the AAQ, in terms of both the lack of differences in content
combined with the significant differences in organizational properties, strongly suggest
that it is not the quality of the childhood experiences per se that determines adult
functioning, specifically parental functioning. Rather, it is very likely the affective-
cognitive representation of these experiences that is of critical importance. The
psychopathology of the child was related to the mother’s affective-cognitive
processing, synthesis, integration, and overall representation of her own childhood
experiences and not the actual experiences in-and-of-themselves. Apparently, it is in
the encoding, processing and resolution/integration, or lack thereof, of the mother’s
childhood relationships that is likely to be of critical importance in terms of the
psychopathology of the child.

This interpretation is consistent with previous research on parental
recollections. (It is important to note that previous studies had reported some
differences also in the quality/content of childhood experiences.) As reviewed in the
Introduction, it is not only the specific content of the childhood memories but also the
organizational properties of these memories that are related to current caregiving

behaviors (Crowell & Feldman, 1988; Grossman et al., 1988; Main et al., 1985;

112111
r120
corr
111151
Inth
for 3

p5}: i1

513?”
11985)
filigree
’0 Whit

13301181

311 then
hmonm
“hen 1
C1hmel
Whom.

tale '
“’«4'1111n5jJ

175
Main & Goldwyn, 1984; Ricks, 1985). What seems to be a critical factor in the

relationship between childhood events and adult functioning is the mother’s ability to
"come to terms" with emotionally difficult childhood experiences. It is the degree to
which the emotional impact of childhood experiences has been openly understood and
integrated into a cohesive, coherent, fluid/flexible and non-contradictory awareness
(or a mental model of integrated understanding) that appears to be related to the
psychopathology of the child, and assumptively to the parent’s caregiving behaviors.

A few of the protocols stood out as remarkable examples of having resolved
extremely difficult childhood experiences. Similar to reports by Main and Goldwyn
(1985) these protocols were notable for the striking contrast between the extreme
degree of family dysfunction and unhappiness in childhood, compared to the extent
to which these extremely problematic childhood experiences had been integrated and
resolved. In these protocols the emotional turmoil and negative impact was not
denied or blocked access, but rather had been openly processed and integrated into
an evenly balanced representation of events and their repercussions. These protocols
demonstrated a high degree of resolution, understanding, and forgiveness as the
written presentations were noticeably clear, detailed and yet concise, and subsequently
extremely poignant. These few mothers demonstrated a courage to openly struggle,
confront, and to eventually learn from their particularly problematic early childhood
relationships. Ultimately, the struggle of these women is a tribute to the potential of

the human spirit.

176

A critical question remains unanswered: When childhood experiences are
basically similar, what is it that determines why one mother develops a defensively
organized mental model, while another mother struggles towards achieving a
meaningful resolution? It is assumed that cognitive/mental defensiveness is related
to painful childhood experiences and its function is to block access or to prevent these
painful affects from being evoked. How is it that some mothers come to terms with
very difficult childhood experiences and become exemplary in terms of parenting
(Main & Goldwyn, 1985) while other mothers have a defensive mental model forever
blocking access to these painful events?

Obviously, there is a potentially wide variety of intervening factor/variables,
beyond the scope of this study, that could possibly effect this "outcome" into
adulthood. Clearly, life experiences and specifically adult relationships have the
potential to have a profound impact on one’s relationship-related mental models.
Information obtained in this study suggests that involvement in psychotherapy is not
necessarily the determining factor as the groups did not differ in their involvement
with psychotherapy. It is possible that the timing of difficult childhood events may
be a critical factor in determining the emotional impact and subsequently the degree
of defensiveness. For example, the effects of a divorce would be different on a 4
year-old than a 12-year-old. Furthermore, having some " healthy " and supportive
relationships or meaningful contact with an adult may play a critical role in
ameliorating the negative impact of dysfunctional family relationships. Undoubtedly,

one can speculate on a wide range of factors and it is likely that there are many

177

relevant aspects which interact with each other. This is clearly an arena that raises
many interesting, valuable, and widely relevant issues and questions for future
research.

It is important to point out that Main and Goldwyn (1985) reported a
relationship between apparent rejection in childhood and a defensive orientation
towards these childhood memories. Inherent in the conceptual foundation underlying
the AAQ and the accompanying scoring system is the establishment of this
relationship. It is assumed that the defensive orientation (i.e., the exclusion and/or
the fragmenting of information) is functional and is established in order to block
certain types of painful and/or anxiety provoking memories from awareness. The
establishment of this statistical relationship between problematic experiences and
information inaccessibility (i.e., defensiveness) is an integral step in the construct
development and validation of the entire conceptual schema, and specifically in
regards to instrumentation.

The results reported in the previous chapter do not offer support for this
critical/pivotal conceptual relationship between problematic childhood experiences and
a cognitive/mental defensiveness towards these memories. In a post hoc re-evaluation
of this hypothesis, however, specific support was established. As noted above, in the
initial analysis the experiential scales were combined scores; the participant’s
experiences with her mother were combined with the experiences with her father.
For this re-examination it was reasoned that defensive representational processes

could potentially be developed to block painful memories/infonnation relevant to

178

experiences with only one specific parent; combining the experiential scores for both
mother and father might possibly cancel out the effects or negative impact from one
parent and thus obscure valuable information. Hence, in the post hoc analysis of the
data relevant to Hypothesis 10, correlations were performed separately between the
affective-cognitive organizational scales and the experiences with mother, and the
experiences with father, respectively.

AS can be seen in Table 21, four out of the five affective-cognitive
organizational scales are significantly correlated with experiencing the mother as
Unloving, and three out of the five scales are significantly correlated with the
mother’s apparent Rejection. (Table 22 reveals that these relationships do not exist
in terms of the experiences with the father.) While these significant relationships are
not large statistical associations, nonetheless, the results that emerge from this re-
evaluation of the data are strikingly consistent with those results reported by Main and
Goldwyn (1985) on the basis of their in—depth interview. The lack of a similar
relationship between the experiential scales and the Idealization scale does not
conform to previous results. It is very likely that the marginal inter-rater reliability
for this scale weakened its usefulness (i.e., the scales’ power to differentiate groups.)

The results indicate that the defensiveness in the mother’s recollections, or
mental model, is related, to some extent, to problematic experiences with her own
mother and not necessarily in terms of the relationship with her father. Given that
35 years ago the typical family structure was characterized by the mother as the

primary caregiver, it is reasonable to suggest that mothers were generally more

[2'

’To

It

179

 

 

Table 21

II; ILA-I113 '; ‘ hilIIhOd Exr-rince with the MI er - I
Aff 'v- "v ' 'n es ftheAA

Seeiee l 2 3 4 5 6 7 8 9 10 11
1.Unlove

2. Reject .82**

3. Neglect .55** .46**

4. Roi-Rev .37** .19* .12

5. Achieve .24” .20“ .00 .23“

6. Abuse .47" .48** .22** .10 .18“

7. Total‘ .88" .80“ .59** .54“ .47** .63“

8.1ncoher .14* .12 -.03 -.05 -.13 .05 .02

9. Unresolve .22** 20*“ .05 -.03 -.17* .17* .12 .71**

10.NoRecall .23** .22** .12 -.05 -.16* .07 .ll .54** .51**

ll.Idealize -.07 .00 -.13 -.03 -.10 —.16* —.13 .48“ .33“ .25**

l2.TotDefen" .16* .17* .oo -.05 -.l8* .04 .03 .87" .81** .74**

.69**

 

ii = 133. ‘Total Problematic Relations Between the Participant and her Mother.

l"Total Defensiveness for the entire AAQ Protocol.

’o < .05, one-tailed. "p < .01, one-tailed.

4. Roi
5. Ach

6. Abu

180

 

 

Table 22

In rr ' B n hildhood Ex rience with the F r n
'v- "v r ' 'n sof eAA

Scales 1 2 3 4 5 6 7 8 9 10 ll

1. Unlove

2. Reject .80“

3. Neglect .62** .44**

4. Rol-Rev .11 .14* -.02

5. Achieve .35“ .37** .13 -.01

6. Abuse .54“ .59“ .14* .08 .29"

7. Total‘ .88“ .86** .59” .30” .55** .71**

8.1ncoher .06 -.01 .05 .02 .05 .04 .05

9. Unresolve .06 -.06 .09 .02 -.10 .06 .02 .71**

10.NoRecall .04 -.09 .11 -.03 -.16* -.06 -.05 .54“ .51"
ll.Idealize -.03 -.06 -.04 .08 .04 -.08 -.02 .48** .33** .25**

12.TotDcfen" .04 -.08 .07 .02 -.05 -.01 .00 .87** .81** .74** .69**

 

N = 133. ‘Total Problematic Relations Between the Participant and her Father. t’Total
Defensiveness for the entire AAQ Protocol.

'o < .05, one-tailed. ”p < .01, one-tailed.

influ

mod.

child

101111

achi

1101181

andt

cohes

repre.

181
influential and thus had a greater impact on the structure of the participant’s mental

model. Furthermore, it was the mother’s unavailability and her turning-away of the
child’s attention, affection, and needs that appear to be the critical experiences related
to the development of a defensively oriented mental model. It is understandable that
a child’s repeated experiencing of her mother as generally unavailable, emotionally
nonsupportive, and rejecting could cause a sufficiently intense level of emotional pain
and distress so as to initiate (as a means of self-protection in order to maintain
cohesiveness) the development of a defensive and distorting affective-cognitive
representational style of these events.

Overall, it is clear that the AAQ and the accompanying scoring system has
been an effective research instrument. This study is an initial attempt to utilize a
questionnaire format of the Adult Attachment Interview, with subsequent
modifications to the scoring system. The fact that there is support for some of the
a priori predictions in terms of group differences and that some meaningful significant
relationships, albeit modest, were obtained between the scales is a promising
endorsement regarding the potential usefulness of the methodology. In other words,
the data offer some important support for the validation of both the instrument (in
terms of internal conceptual consistency) and the overall conceptual schema. Even
though the magnitude of the associations id not as large as those reported in Main and
Goldwyn’s (1985) preliminary report based on the in-depth interview, the similarities

in the pattern of results suggest that the questionnaire format may be an appropriate

alternative methodology.

using

exper

few 1

isimr

inlhe

Weigh:

disadv.

501116

Ittolle.
311 ovel
Emmi
Partitip
10 63131]

the data

31.08am

182

As compared to the in-depth interview, there are some definite advantages to
using the questionnaire format. For example, there is significantly less time and
expense involved in the data collection process. The data can be collected within a
few hours in a group setting using a standardized format (i.e., there is no need to
train and hence establish inter-interviewer reliability). Once the data are collected it
is immediately ready to be scored thereby eliminating the time and expense involved
in the transcription of the verbal interview.

Obviously, the specific needs of the research require that these advantages be
weighed against the disadvantages of the written protocols. The prominent
disadvantage is that the questionnaire format does not allow the researcher to "probe" ,
and establish needed clarification of the participants recollections. For example, in
some of the written protocols there were many gaps and disparities in the
recollections and it was solely up to the individual rater to attempt to piece together
an overall picture. Finally, the research on the Adult Attachment Interview has not
emphasized issues of scoring at the scale level since the ultimate goal is to assign the
participants an attachment-classification. Clearly, future research is needed in order
to establish whether this attachment classification can be determined on the basis of
the data from the AAQ.

The scoring system, which was adapted from Main and Goldwyn ( 1989)
appears to have some applicability and usefulness. It is important to point out,
however, that neither the researcher nor the raters had any direct training in the

original scoring system and furthermore, modifications were made in order to

183

accommodate the written format. Therefore, it is reasonable to assume that there may
be many substantive differences between the Main and Goldwyn scoring system and
the system that was ultimately used for this study. Additionally, the three scales
created for this study, namely, Abuse, Lack of Specificity/Recall, and
Unresolved/Autonomy all had good inter-rater reliability. In fact, the Abuse scale
had the highest reliability among all ten scales, and the two "new" affective-cognitive
scales proved to be useful research tools.

It is interesting to note that all of the Experiential scales are significantly
intercorrelated. One possible interpretation of this inter—relatedness among the scales
is that the scoring system did not sufficiently discriminate between different
dimensions or aspects of problematic experiences and that in fact all the scales are
measuring basically the same underlying experience. On the other hand, an
alternative explanation is that there is a wide range of problematic childhood
experiences and that these experiences do in fact overlap and co-exist. For example,
it makes sense that Unloving will be highly correlated with all the remaining
Experiential scales since the other scales are intended to assess some of the various
ways in which a parent might be unloving. It is also quite possible that a parent can
be unloving, rejecting, abusive, and role-reversing. Relatedly, it is consistent that
Neglect (taken as an index of the lack of emotional/nurturing interactions) is not
related to either Role-Reversal or Pressure to Achieve since both of these scales are

indices of over-involvement. Hence, it is reasonable to interpret these

184

intercorrelations as suggesting that when the caregiving is lacking and inadequate, this
creates a wide variety of concomitant problematic emotional experiences for the child.

Finally, even though the inter-rater reliabilities established for the present study
were adequate, it might be more effective to score these protocols within a group
format. This suggestion originates from the group experiences in the raters training
where the final decisions for scoring the protocols were derived through a group
process. Each individual rater would come to the training session with two protocols
scored independently. When scores were divergent the group process would
eventually arrive at a general consensus.

The rationale for proposing this type of scoring procedure is as follows: It
cannot be overly emphasized that the data from the AAQ is extremely sensitive,
intimate, poignant, and emotionally powerful. The participating mothers took the task
seriously and became very involved in the process of recollecting. Most of the
mothers wrote for over two hours, some cried periodically, and practically all of them
were deeply absorbed in their own private experience and introspection. Many
mothers reported that the overall experience was extremely worthwhile and they were
able to gain valuable insights through the process of completing the questionnaire.
Many mothers reported that they spontaneously saw meaningful connections between
their own childhoods and their current adult functioning and relationships, especially
gaining relevant understandings regarding their current parenting style. Ultimately,
these mothers shared very personal and emotionally vulnerable aspects of their

childhood and of themselves. (It would be interesting to compare the level of self-

185

disclosure between the "privacy" and anonynrity of the questionnaire format with the
level of openness exhibited in the interview data which involves disclosure in the
presence of another person.)

Inevitably, these private and heart-rending accounts have a strong and
noticeable impact on the reader/rater. The recollections are stirring and deeply
moving and consequently the reader is highly susceptible to their own personal bias
and interpretive distortions. Even though all the raters had previous clinical training,
nevertheless, they all unavoidably exhibited some " transference " type distortion/bias
in their interpretation of the protocols. The training procedure inadvertently
demonstrated the potential moderating effect of a group process towards the individual
biases evoked in the scoring of this data. Personal sources of bias were counter-
balanced by the group and within this context they became useful sensitivities
providing valuable sources of input. Consequently, having group-derived scores for
the protocols is a way to minimize and constructively utilize the sources of individual
bias and distortion that are inherent in the interpretation of this material, thereby
assigning consistently more "objective" and accurate ratings to the data.

M R 11 'nandPrce a1 1

The major thrust of this study is to demonstrate a link between a mother’s
childhood experiences and her current adult functioning. Specifically, the purpose is
to establish a connection between the manner in which a mother has internally
organized/ represented her own childhood experiences and her current perceptual style.

Conceptrrally, this connection is maintained through the construct of an internal

.1. 'y 1
My),

 

186
mental model regarding parent-cth interactions. It is hypothesized that this parent-

child mental model, based largely on the mother’s own childhood experiences, plays
a crucial role in determining the mother’s behavioral classification schema. Hence,
Hypothesis 11 is an attempt to establish support for the overall notion of linkage and
is specifically designed to explore the relationship between a mother’s deviancy
categorization processes (i.e. , her current perceptual style) and the affective-cognitive
organization of her childhood recollections (i.e., her mental model of parent-child
relationships).

Even though the Specific predictions involved in Hypothesis 11 are not
supported, there are some significant findings that do promote, albeit tentatively, the
notion of linkage. The a priori predictions were based on a very specific line of
reasoning. These specific pathways that were predicted in order to demonstrate the
connection between the mother’s past and her current perceptual style are not found
in the data. Nevertheless, a pattern of significant associations does emerge and these
results are not inconsistent with the overall conceptualization suggesting that the
connection between the mother’s own childhood experiences and her current parenting
behavior, specifically her perceptual style, is maintained, in part, via the construct of
an internal parent-child mental model.

For example, the specific proposition that the mothers’ affective-cognitive
defensiveness would be associated with a negative perceptual style was not supported.
In other words, it is not t__e that the more disorganized and defensive the mother’s

childhood recollections, the greater her tendency to exhibit a categorization schema

 

187

that is overly inclusive of deviant behaviors and overly exclusive of prosocial
behaviors. Interestingly, there is, as predicted, a significant inverse relationship
between the mother’s affective-cognitive defensiveness and her correct identification
of prosocial behaviors. Clearly, this could be interpreted as part of a negative
response style suggesting that the mothers’ tendency to distort, restrict, splinter,
and/or disorganize attachment-related information is in fact (as predicted) related to
a corresponding need to shift attention away from prosocial behaviors. But this
interpretation, which proposes a " selective inattentiveness" towards prosocial
behaviors, is likely to be erroneous since the exact same pattern of results is exhibited
in regards to the identification of deviant behaviors. Therefore, the results do not
favor an interpretation of attentional "selectivity" (or nonresponsiveness to prosocial
behaviors) but rather, the data suggest that the defensiveness in the mother’s parent-
child mental model appears to be associated with a non-specific overall cognitive-
perceptual ineffectiveness or inefficiency regarding the ability to correctly identify
child behaviors, both deviant and prosocial.

It is incumbent upon this discussion to offer a plausible explanation as to why
the predicted associations were not found in the data, and at the same time, to provide
a conceptual framework in order to understand the significant associations which are
evidenced. Clearly, a wide variety of assumptions were involved in the derivation
and ultimate rationale for the a priori predictions. It appears that some of these
assumptions are suspect. A re-examination of these underlying postulates will help

to provide an explanation regarding the lack of support for the a priori perspective

188
and relatedly, help to establish a slightly modified or alternative

understanding/perspective in order to account for the pattern of significant
associations that are present.

In order to examine some of these questionable assumptions, it will be helpful
to briefly review the rationale for the specific predictions involved in Hypothesis 11:
The literature emphasized how the MCDB are characterized by their irritability,
criticalness, and negativity. This negativity was believed to be exhibited, in part, via
a negative perceptual style (i.e., the mothers’ deviancy-classification schema) and
hence, it was predicted that the MCDB would be overly inclusive in their
classification of deviancy and underinclusive in their classification of prosocial
behaviors. It was also reasoned that this hypothesized inclusiveness towards
classifying deviancy was the result of an insensitivity, and specifically an
inattentiveness, towards prosocial behaviors. This reasoning is based on the pivotal
assumption that the negativity which characterizes the MCDB is a parallel process to
the rejection and nonresponsiveness which characterizes the mothers of avoidant
infants. In other words, this hypothesized nonresponsiveness to prosocial behaviors
was seen as reflecting the mothers’ need to shift attention away from attachment-
related information (i.e., their defensively organized parent-chfld mental model) as
prosocial behaviors were assumed to be more affiliative in nature and hence more
relationship oriented. Thus, rejection and nonresponsiveness at a behavioral level was
assumed to correspond to a blocking of information and subsequent inattentiveness at

the representational level. It was reasoned that turning away from attachment-related

189

information (i.e. , the mental/cognitive defensiveness) resulted in a nonresponsiveness
and inattentiveness towards prosocial behaviors and subsequently an over-emphasis
on deviancy.

Clearly, one of the most obvious underlying assumptions that proved to be
erroneous is the "behavioral classification covariation hypothesis". Inherent in the
formulation of the specific predictions is the assumption of a classification
covariation. But the data (reviewed previously in this chapter) clearly do not support
this underlying assumption. Therefore, it was not possible to obtain significant
correlations between the defensiveness in the mothers’ mental model and an inclusive-
exclusive classification schema, since this pattern of covariation in the classification
of child behaviors does not exist in the data.

Perhaps the most influential assumption involved in the formulation of the a
priori predictions is that "the negativity that characterizes the MCDB is a parallel
process to the rejection and nonresponsiveness that characterizes the mothers of
avoidant infants". A corollary of this pivotal assumption is that the MCDB adult
attachment-classification is likely to be that of "Dismissing of Attachment" (Main &
Goldwyn, 1989) and hence, by implication their sons would correspondingly be
characterized by an "insecure-avoidant attachment".

This set of related assumptions, which upon re-examination are questionable,
had a far-reaching impact in the formulation and specific direction of the a priori
predictions. Concomitant with these assumptions is the perspective that the MCDB

affective-cognitive defensiveness has a Specific purpose or function. In other words,

 

 

190

the specific attachment-classification of Disnrissing implies a distinct strategy in the
organization and defensiveness of the mental model. Hence, while it is true that
adults classified as "Pre-occupied by/Entangled in Past Attachments" (Main &
Goldwyn, 1989) also have a defensively organized parent-child mental model that is
designed to "block access to attachment-relevant information", the overall function
or defensive strategy of this particular affective-cognitive organization is different
from that of Dismissing adults.

The main point is that the specific organization (or predominant orientation)
of the particular parent-cth mental model leads to an explicit set of predictions
regarding the mothers’ behavioral classification schema. Predicting a lack of
responsiveness to affiliative behaviors, with a subsequent over-emphasis on deviancy,
derives directly from the assumption that the mental model of these mothers has a
Dismissing orientation. It is this assumption, however, that may have been in error.
Clearly, this would account for the lack of confirming results. Furthermore, the
actual findings are in fact more congruent with what might be expected if the MCDB
attachment-classification was that of Preoccupied.

In order to advance this line of reasoning it will be helpful to clarify the
distinctions that differentially characterize the affective-cognitive organizations related
to these two types of insecure attachments. Main and Goldwyn (1989) do point out
that while there are some similarities in the attitudes of insecure adults, nonetheless,
"the strategy the adult uses in her thinking about attachments and about relationships

is usually predominantly either Disnrissing or Preoccupied" (p.103).

191

"The Dismissing adult generally attempts to ignore, avoid, or devalue
attachment relationships and experiences" (Main & Goldwyn, 1989, p.102). The
strategy is to de-activate and detach from the emotional impact of relationships which,
for the child, were frequently characterized by being unloved and rejected. The
defensive deactivation of these events is maintained via their mental model with a
primary emphasis on the lack of recall (i.e. , inaccessibility to the potentially painful
events) and/or high idealization (i.e. , self-deception regarding the impact of the
events). The distinguishing features of their current parenting behaviors are likely to
emphasize a rejection, a nonresponsiveness, and an emotional distancing;
correspondingly, their own children are likely to exhibit a distancing, rejection and
avoidance of close contact.

The Pre-occupied adult, on the other hand, "is unable to free herself from
either her active concern with her relationships, her traumatic past, or (exhibits an)
inability to move beyond her early relationships" (Main & Goldwyn, l989,p.103).
While these mothers appear to be "objective", analytical, and very aware of the
impact of childhood experiences, their pre-occupation, in fact, reveals a weakened
sense of self, separateness, and resolve. This restricted sense of self is probably a
way to minimize awareness of the negative emotional impact. The emotionality,
when present, seems to lack integration as if it is "unmetabolized" (Main & Goldwyn,
1989). These adults frequently over-intellectualize in an ineffective manner, and with
their weak sense of separateness they remain chronically engaged/preoccupied, never

able to work-through and establish a boundary or finality to these negative

 

"311.3,?
.

 

192
experiences. The Preoccupied adult characteristically experienced a lack of support

as a child, not in terms of rejection, but rather they frequently came from families
that were extremely role-reversing and involving/enmeshing (Main & Goldwyn,
1989). The defensive strategy of their mental model is highlighted by a marked
incoherence (i.e., the fragmenting and splintering of attachment-relevant information
in order to minimize the emotional awareness/impact) and a preoccupation, lack of
resolution, and lack of sense of self. All of these qualities aid in denying the
irreversibility and finality of these past events. The hallmark of these mothers’
current parenting behavior is their inconsistency, ambivalence, and high reactivity.
Correspondingly, their own children are likely to exhibit angry resistant behavior that
is mixed with seeking contact and closeness (i.e., reflecting a fundamental
ambivalence regarding attachment.)

The pivotal (and very likely dubious) assumption that the MCDB are likely to
be classified as Dismissing was based on the characterization of these families as
rejecting and negative. However, the conduct-disturbed families are also regularly
described as inconsistent, enmeshed, and highly irritable and reactive. These
additional descriptors do not fit the characterization of an emotionally distant and
detached family atmosphere as one would expect with that of Dismissing mothers, but
rather they are very congruent with the characterization of Preoccupied mothers. In
other words, the inconsistency, high reactivity, enmeshment, "emotionality" and

intensity described in the literature on conduct-disturbed families, and generally

 

 

193
supported with the present findings, have a striking similarity with the

characterization of the Preoccupied adult.

Furthermore, it is reasonable to propose that the child’s conduct disturbance
appears to exhibit behaviors of an insecure resistant/ambivalent attachment, rather
than an insecure/avoidant attachment. This would correspond with the mother’s
classification of Preoccupied rather than Dismissing. Greenberg and Spletz (1988,
p.206) point out that conduct problems "can be viewed as strategies for gaining the
attention or proximity of caregivers who are unresponsive to the child’s other
signals." But these inappropriate behaviors also seem to reflect an underlying
ambivalence; while the behaviors may be an attempt to seek proximity, they also
seem to express an angry resistance and emotional distancing. The enactment of the
annoying (i.e., distancing) behaviors exhibited by Social Aggressors, the specific
conduct disturbance under investigation (e.g. , teasing, excessive talking, showing off,
stubbom, argues, brags, demands attention, etc) clearly necessitates close and
frequent interactions with another person, specifically the caregiver. This is
contrasted with Delinquency which is much more likely to reflect an insecure-avoidant
attachment, as this type of behavioral disturbance involves a noticeable detachment
or lack of interaction with the caregiver; for example, Delinquency is characterized
in the literature by the lack of parental supervision.

On the basis of this line of reasoning, it is reasonable to conclude that the
MCDB do not, as originally assumed in the formulation of the a priori predictions,

have a parent-child mental model that is primarily Dismissing of attachment. Rather,

 

194

it is likely that these mothers would be primarily classified as Preoccupied.
Obviously, only further research can ultimately resolve this and other related issues.

Apparently, one of the shortcomings of this study is its sole reliance on scale
scores to describe the mother’s internal parent-child mental model, rather than
determining the mother’s overall adult attachment-classification. Main and Goldwyn’s
(1989) higher-order variables (i.e., the attachment-classification) do seem to provide
an added dimension and depth of understanding to the data not possible from only the
individual scale scores. The attachment-classification provides a meaningful way of
organizing the different scale scores into a coherent whole. This typology provides
a useful conceptual framework and allows for differential predictions/hypotheses in
terms of further research. For example, Is the parent’s adult attachment-classification
related to the child’s behavioral disturbance? Do the different insecure attachment-
classifications relate to different types of "pathology" or behavioral disturbances in
the children? Are Social Aggressors predominantly from families with parents
classified as Pre-occupied, while the StealerS/Delinquent have parents that are
primarily classified as Dismissing? Is their an interaction between adult attachment-
classification (including the classification of Secure/Free-Autonomous) and the
parents’ current behavioral classification schema?

The conjecture that the MCDB may in fact have a Preoccupied attachment-
classification would (as stated above) undoubtedly lead to a different set of
expectations regarding the pattern or pathway of linkage between the mother’s parent-

child mental model and her current perceptual style. The affective-cognitive

 

 

195

organization of a primarily Preoccupied parent-child mental model would be reflected
in an inconsistent and highly reactive perceptual style. In other words, rather than
expecting a nonresponsiveness to prosocial behaviors, as originally predicted, it is
likely that the defensive strategy of Preoccupied mothers would result in an erratic,
confused, and ineffective perceptual style.

It is possible to interpret the significant relationships exhibited in Table 14, and
Table 19 for the sub-sample analysis, as consistent with this perspective. For
example, overall, affective-cognitive defensiveness (as measured through the four
different affective-cognitive scales and the composite score) is significantly inversely
related to both the number of correctly identified deviant behaviors and the number
of correctly identified prosocial behaviors; as the defensiveness increases, the
mother’s ability to accurately perceive both deviant and prosocial behaviors slightly
decreases. Also, defensiveness is significantly related to higher aversiveness ratings
of child misbehavior. Overall, this pattern of significant findings does indicate that
the more the mother’s childhood recollections were disorganized,
splintered\fragmented, lacking in detail, and unresolved in terms of childhood
relationships, the less the mother’s accuracy in correctly identifying the child’s
behaviors, and the more likely she is to be highly reactive (i.e., higher ratings) to
deviant behaviors.

It is important to point out that the magnitude of the Significant correlations
only ranges between very modest to moderate and the observed pattern of significant

associations does not exist for all of the defensiveness scales. For example,

 

196

Idealization does not exhibit any associations with the perceptual variables and the
Lack of Detail/Recall scale only exhibits significant relationships in the sub-sample
analysis. (Interestingly, the Unresolve scale is conceptually synonymous with the
status of Preoccupied and this scale does exhibit the strongest or most impressive
associations with the three perceptual variables; this implies that the greater the
preoccupation the less accurate the mother’s perceptual style and the more intense her
reaction to child misbehavior.)

Overall, assuming that the MCDB are in fact Preoccupied adults, the general
pattern of associations are consistent with expectations. The defensive affecitve-
cognitive organization appears to have resulted in a slight decrease in cognitive-
perceptual efficiency and effectiveness and is likely to have also interfered with affect-
modulation. A cornerstone of this study is the logical supposition that the mother’s
accurate identification, evaluation, and labelling of the child’s behaviors are
prerequisites for an appropriate parental response. It is reasonable to propose that an
overall reduction in the capacity to correctly identify ’child behaviors combined with
a less modulated affective response may be factors involved in the MCDB’
inconsistent, more highly emotional, and noncontingent response style. As discussed
below, there are obviously other competing intervening factors which also affect the
outcome of the mother’s perceptual style and subsequently, the ultimate quality of her
responses. Nevertheless, it is reasonable to conclude that the data do, albeit

tentatively, suggest a linkage or connection, maintained via the parent-child mental

 

5A

197
model, between the mother’s childhood experiences and the quality and

appropriateness of her responses to her son’s behavior.

The significant positive correlations, especially in the sub-sample analysis,
between the mothers’ ratings of child misbehavior (i.e., CBRSI) and the various
defensiveness scales offer the most consistent pattern of results promoting the overall
conceptualization. There are four conditions that are necessary in order to confidently
support the conceptual model. Namely, there needs to be: (a) a significant difference
between the groups in terms of the level of cognitive/mental defensiveness; (b) a
significant difference between the groups on specific perceptual variables; (c)
significant correlations between defensiveness and the perceptual variable, and (d) the
direction of the correlations needs to be logically congruent with the scores on the
perceptual variable.

The relationship between the mothers’ affective-cognitive defensiveness and
their level of reactivity to deviancy satisfies all four of the stated criteria. The
MCDB are both significantly more defensive and more reactive to deviancy.
Therefore, the moderate positive statistical associations between the various scales of
defensiveness and the mother’s reactivity to deviancy strongly implicate the MCDB
defensively organized parent-child mental model. This pattern of results suggests that
the affective-cognitive defensiveness may account, at least in part, for the MCDB
higher levels of reactivity and irritability. It was previously proposed that higher
reactivity may reflect a behavioral-discrimination schema that has a direct effect on

the quality of the mother’s responsiveness (i.e., a cognitive-affective dimension).

 

198

Therefore, the positive relationship between reactivity to deviancy and
cognitive/mental defensiveness suggest that the mother’s parent-child mental model
may play a role in the regulation and modulation of the mother’s affective
response/reaction. It was already suggested that habitually higher levels of parental
reactivity may contribute to an ineffective discipline style. Hence, the data can be
interpreted as consistent with the overall conceptualization that the manner in which
a mother has mentally organized her own childhood experiences may ultimately affect
certain aspects of her current parenting behaviors.

On the other hand, a more qualified interpretation is evoked in order to explain
the relationships between cognitive/mental defensiveness and the decreased capacity
to identify the child’s behavior. Only three out of the four criteria required to firmly
support the overall conceptualization are fulfilled. The lack of support comes
primarily from the inverse association between defensiveness and the mothers’ ability
to correctly identify deviant behavior. One of the goals of this study was to
demonstrate that the organizational properties of the mothers’ parent-child mental
model would "o ex a crucial role" in detemrining their behavioral-classification
schema. This objective was not realized. Even though there is a connection between
the organizational properties of the mother’s memories and specific perceptual
variables, the pattern of the data definitely indicates that there are other significant
factors which may be more substantially involved in determining the mother’s overall

classification schema, especially in terms of the MCDB sensitivity to deviancy.

 

 

199
For example, it has been clearly demonstrated that the MCDB exhibit a

heightened sensitivity to deviant behaviors and they also exhibit childhood memories
that are significantly more defensive. Contrary to the goals of this study, however,
this heightened awareness towards identifying deviancy is not accounted for by the
organizational properties of the mothers’ childhood memories since the data indicate
that as defensiveness increases the ability to perceive deviancy decreases. This is the
fourth criteria listed above and the inverse relationship between defensiveness and the
perceptual variable clearly does not fulfill this condition (i.e., the inverse direction
of the correlation is not logically congruent with the group differences in identifying
deviant behaviors.)

Obviously, there must be some unexplained factor(s) and/or interaction which
accounts for the MCDB sensitivity to deviancy. In other words, even though the data
suggest a link between the way in which a mother organizes her childhood
recollections and specific aspects of her current perceptual style, the results do not
ultimately help to explain the MCDB vigilance or heightened awareness to deviant
behaviors. This raises issues as to the complexity of interacting forces that are
involved in deriving the final outcome/product of a behavioral-classification schema.
A better understanding of the nature of these other factors (e.g., perhaps the history
of the child’s chronic misbehavior results in a response set and/or the mother’s
emotional state due to her marital relationship or the quality of social support systems)
might help to explain the MCDB vigilance towards deviancy. This is definitely an

area that necessitates further research.

200

In summation, the very nature of the connection between the Specific defensive
affective-cognitive properties of the mother’s childhood recollections and her current
functioning remain largely, but not completely, elusive and unexplained. Clearly, the
specific and detailed pathways of this connection and the actual amount of the
variance that can be accounted for, are critical issues in need of replicated findings
and further research. It is clear that other (and as of yet unexplained) factors, besides
the organization of childhood memories, also have a strong impact on the mother’s
behavioral-classification schema and ultimately on her response style.

On the other hand, the statistical associations, albeit modest, between the
affective-cognitive organization of the mother’s recollections and specific aspects of
her perceptual style are evidenced. The way in which a mother identifies, evaluates,
and classifies child behaviors can be explained or accounted for, to a modest degree,
by an understanding of the way in which she has organized her own childhood
experiences. Undoubtedly, there is a link between the way in which a mother
organizes her own childhood memories and her current functioning, specifically her
cognitive-affective functioning and certain aspects of her perceptual style. The results
of this study do encourage continued research designed to further explore and
elucidate the construct of an internal parent-child mental model which serves as a
mediator between past childhood events and current adult functioning.
W

The goal of this study was to demonstrate a link or connection between a

mother’s own childhood experiences and her current parenting style, specifically her

 

 

 

 

201

perceptual style. This connection between past events and current functioning is
maintained via the construct of a parent-child mental model which is based on the
mother’s own childhood experiences. In other words, the internal representation (and
specifically the affective-cognitive organization) of the mother’s own childhood
experiences is believed to guide and govern the quality of her current responses to her
children.

While a child’s conduct-disturbance is a legitimate research topic in its own
right, the family and relationship dynamics hypothesized by Patterson (1982), and
believed to be at the core of the child’s behavioral disturbance, do provide a useful
paradigm for examining the relationship between the mother’s own childhood
experiences and her current parenting behaviors. This is primarily due to the
importance ascribed to the mother’s cognitive processes (i.e. , her behavioral
classification schema) in the development and maintenance of the child’s conduct
disturbance. The ways in which a mother perceives her son’s behavior undoubtedly
affects how she interacts with him, and consequently, how the child behaves. This
classification of the child’s behavior involves hypothetical cognitive processes and it
is here that the connection (via the mother’s internal parent-child mental model)
between past and present is established. The organizational properties of the mother’s
parent—child mental model reflect how relevant information is processed, accessed,
and integrated; a defensive organization refers to the tendency to distort, restrict

and/or inhibit certain classes of information. It is this representational model that is

202

believed to play a critical role in determining the mother’s current cognitive-
perceptual behaviors.

In order to support this overall conceptualization it was necessary to (a)
establish differences between the two groups of mothers in regards to their perceptual
style (i.e., behavioral classification schema); (b) to establish differences between the
groups in regards to the content and organizational properties of the mother’s
childhood recollections (i.e., the parent-cth mental model); and (c) to establish a
meaningful connection between the mother’s perceptual style and her mental model.
Even though there were supportive findings partially satisfying each of these criterion,
overall, however, the results were equivocal. For example, the data indicated that
the MCDB are more sensitive and inclusive in their classification of deviancy and
they also exhibited childhood recollections that were more disjointed, fragmented,
lacking in specific detail, and were generally more unresolved regarding childhood
relationships. But the overall conceptual model (in regards to this specific perceptual
variable) was not supported because the mother’s vigilance towards identifying
deviancy was not explained by her affective-cognitive defensiveness.

On the other hand, there was a pattern of Significant results that did conform
to the three requirements listed above. The MCDB were more reactive to child
misbehavior and the results indicated that the affective-cognitive defensiveness did
have a moderate association with higher levels of reactivity to deviancy. The
organizational qualities of the maternal recollections appeared related to the mother’s

"deviancy-discrimination schema” involved in the evaluation of the child’s behaviors.

 

203

In other words, the mother’s childhood recollections, via the construct of a mental
model, appeared to play a role in her affect-modulation. It had been proposed that
chronically higher levels of reactivity may contribute to an irritable family atmosphere
and an ineffective disciplinary style, ultimately impacting on the child’s behavioral
status. Hence, these findings do support the overall conceptual model suggesting that
the mother’s own childhood experiences have an influence, via her parent-child
mental model, on her current parenting style.

Clearly, support for the overall conceptualization was mixed. The mother’s
parent-child mental model was associated with her affect-modulation and a decreased
capacity to correctly identify child behavior. Furthermore, a defensive organization
of the mother’s childhood recollections was related to the psychopathology of the
child. However, an understanding of the specific nature of this connection between
the mother’s mental model and the child’s conduct disturbance does remain elusive
and in need of further research. The MCDB vigilance towards identifying deviancy,
an important characteristic of their perceptual style, is not explained by the affective-
cognitive organization of their childhood recollections. Obviously, the conceptual
model needs to be expanded in order to accommodate other significant
factors/variables and/or potential interaction effects.

Given the equivocal nature of these results, the impact and implications of this
study need to be qualified. The greatest impact of this research is in terms of its
methodological and conceptual contributions. This study has introduced several useful

research instruments. While these instruments, (e.. g, CBS, CBRSI, CBRSII, and the

 

204

AAQ and the accompanying scoring system) are in need of continued development
and further refinement, they nevertheless proved to be useful research tools. Of
particular significance is the use of the AAQ, which has some definite advantages
over the in-depth interview. Within the domain of attachment-theory, there is
currently much interest in exploring the significance of maternal recollections in
regards to current parenting. Establishing the effectiveness and usefulness of the
AAQ is a potentially important methodological contribution to an already active area
of research.

Furthermore, this study has made a conceptual contribution which may have
applicability to a wide range of research interests. Even though the specific dynamics
of the connection remain elusive, nevertheless, this study offers supportive evidence
for the construct of an internal mental model which serves as mediator between past
events and current functioning. There is a diversity of research perspectives
interested in exploring the coherence of behavior patterns across generations, (e.g.,
Belsky & Pensky, 1988; Simon, Whitbeck, Conger and Chyi-In, 1991; Wahler &
Dumas, 1986). Many researchers are investigating behavioral continuity across
generations, with particular interest in destructive and/or abusive behavior patterns
(e.g., Kaufman & Zigler, 1987). It is very likely that an understanding of the way
in which past events are encoded and mentally represented (i.e., organized,
processed, and integrated) may be an important explanatory factor in regards to the

intergenerational transmission of various behaviors.

 

 

APPENDICES

APPENDIX A

LETTER DISTRIBUTED TO MOTHERS

205

Dear Mothers:

lam offering FREE attendance to a PARENTING SKILLS WORKSHOP in exchange for
your participation in a study involving mothers of boys ages 6-11 years-old.

I am a Child and Family Therapist in private practice and completing my graduate studies
at Michigan State University. I will be conducting this study, which is part of the
requirements for my Ph.D., under the close supervision of Dr. Robert Caldwell, Professor
of Psychology at MSU.

The present study will explore some important factors involved in parent—child interactions,
especially those between mother and son. As most of you are already aware, the mother-
child relationship is of critical importance to a child’s growth and development. Parent-child
interactions, however, are a two-way street; the quality of these interactions can have a
clearly noticeable effect not only on the child, but also on the parents, siblings, and the
overall family atmosphere. It is the goal of this study to expand our current knowledge and
understanding regarding the possible effects of mother-child relationships.

I am writing this letter to personally ask you to participate in this study. I am attempting to

locate moMerg who goneioer Meir son go be e Lieh evior problem In Me home, as well as

mI r .1 Ira' =._I a ir I’n IIeh- viI ”I II III In N 1!'Vh_= -ver _Ix,‘

Participation in the study in quite simple. Participation will involve meeting for me session
lasting approximately three to four hours. At this session you will participate with a group
of mothers asked to fill out some rating scales and a questionnaire. All the information
obtained will remain strictly confidential. If you are interested and want further details

mun-1m.

Past experiences suggest that participating in the study may prove to be an interesting,
thought provoking, and personally valuable experience. Plus, as a token of my appreciation
for your participation, you will have your choice of either a S1599 remuneration, or FREE
ADMI'I'TANCE to a PARENTINQ SKILLS WORKSHQP. The half-day workshop will
emphasize increasing child management skills and is designed to aide and offer some specific
suggestions in dealing with your son’s annoying and disturbing behaviors at home.

PLEASE CALL 351-7810 if you are interested in participating in this study and attending
the Parenting Skills Workshop. Feel free to call even if you just want to obtain further
information.

Your participation in this study will be greatly appreciated.

I look forward to speaking to you soon.

Sincerely ,

Jerry S. Adams, M.A. ***0ver***

 

 

 

 

APPENDIX B

PHONE INTERVIEW AND FACT SHEET

206
INITIAL PH NE INTERVIEW

Thank you for calling about the study. What school does your son go to?

How old is your son? What is his date of birth?

"The next question, which has four choices, is designed to help in the determination of
whether or not you regard your son to be a behavior problem in the home. Please listen
carefully to the complete question and then pick the me choice that best describes how you
currently regard yours son’s behavior.

Choice 1. I consider my son to have fairly serious behavior problems at home.

Choice 2. I consider my son to have mild behavior problems at home.

Choice 3. I consider my son to basically M have any behavior problems at home; what
you’d expect for his age.

Choice 4. I consider my son to definitely LL91 have any behavior problems at home; he is
very well behaved. "

If 1 or 2 ask for some examples.
Is you son currently seeing a therapist or counselor? Has he attended therapy in the past?

Is anyone in the family seeing a counselor? Has anyone in the family seen a counselor in
the past? Who? When?

Explain their overall involvement in the study; what their participation will entail. Include
information about the feedback sheet immediately after data collection and the written report
once the data is analyzed.

Explain the three questionnaires, and the consent form, that will be sent in the mail.

Ask about the Workshop or financial remuneration?

Best times to schedule research session?

Any questions they may have.

Name(spelling) and address.

Thank them for their interest and participation.

 

 

207
INITIAL PHONE CONTACT

Preliminary Code

 

 

Mother’s Name:

 

Phone Number:

 

Address:

 

 

Best times to contact:

 

 

 

Where did she hear about study:

 

 

 

Son’s DOB: Age:

Description of

Son’s Behavior: Conduct—disturbed Normal
Examples:

 

 

 

 

Family Members:

 

 

 

 

Is anyone currently in TX? Hx:

Hx of TX:

 

Her Questions:

 

 

Best times to schedule session:

 

 

Tentative plan: Workshop Money

Other notes:

 

APPENDIX C

CONSENT LETTER

208
Dear Mother:

I would like to personally thank you very much for deciding to participate in the
Mother-Child Behavior Project. I think you will find that it will be a worthwhile
experience.

I am writing this letter to welcome your participation in the study, to briefly review,
as we discussed on the phone, the purpose of the study, and to have you fill out some
preliminary screening information, including the Consent to Participate Form. (For
those of you who already received a letter sent home through your son’s school, the
first page of this letter will repeat much of the information already presented.)

Let me briefly review the purpose of the study and the nature of your participation.
The present study will explore some important factors involved in parent-child
interactions, especially those between mother and son. As most of you are already
aware, the mother-child relationship is of critical importance to a child’s growth and
development. The recent emphasis on the teaching of parenting skills highlights the
importance of the parent’s role in term of parent-child interactions. Parent-child
interactions, however, are a two-way street; the quality of these interactions can have
a clearly noticeable effect not only on the child, but also on the parents, siblings, and
the overall family atmosphere. It is the goal of this study to expand our current
knowledge and understanding regarding the possible effects of mother-child
relationships.

At the end of the research session I will be giving you a feedback sheet describing the
research in greater detail. As soon as the study is completed, the general results and
implications will be shared with all participating mothers. Finally, for those of you
participating in the Parenting Skills Workshop, a more in-depth description of the
ideas and concepts that form the basis for this study will be presented.

The design of the study is quite simple and straightforward. Participation will involve
meeting for we session lasting approximately three to four hours. At this session you
will be part of a group of mothers that will be asked to fill out three different
questionnaires. The first two involve the labelling and rating of a wide variety of
child behaviors, and the third is a questionnaire regarding your own background.

As can be expected, all information obtained will be held strictly confidential. Your
name will not appear on any of the research material. The questionnaires will be
coded to protect the identity of each participant.

As a token of my appreciation for your participation in the study, I am offering you
either a $15.00 remuneration, or attendance to the half-day workshop on Parenting
Skills. The workshop will emphasize training in child management skills and is
designed to aide and offer specific suggestions in dealing with your son’s annoying

209

and disturbing behaviors at home. Overall, the Workshop will attempt to meet the
specific needs and interests of the participating mothers.

Attached to this cover letter you will find the Cengnt M Pagflcinate Form. Also
enclosed, you will find ( 1) the Child Behavier Cheeklist, (2) the Emily

W, (3) the Preliminaey Questions, and (4) a pre-addressed retum
envelope.

PLEASE SIGN THE CONSENT FORM AND FILL OUT THE THREE
QUESTIONNAIRES. After you have completed the three questionnaires, please
mail them, along with the signed Consent Form, back to me in the envelope provided.

As a reminder, please (19 net put your name on any of the questionnaires. As you
will notice, the forms are coded so that all the information will remain confidential.
Your personal CODE NUMBER is

 

After I receive the completed forms, I will be calling you to set up the appointment

for the group session. If you have any questions in the meantime, please feel free to
call me at 351-7810.

Thanks again for your interest and participation in the study. I’ll be speaking to you
soon.

Sincerely,

 

Jerry S. Adams, M.A.

Signed I

Date;

210
N T P IPA RM

I have freely consented to participate in a scientific study being conducted by
Jerry S. Adams, under the supervision of Dr. Robert Caldwell, Professor of
Psychology at Michigan State University.

The nature of the study has been explained to me and I understand the
explanation and what my participation will involve.

I understand that by participating in this study I am entitled to either a $15.00
compensation or participation in a workshop directed toward increasing
Parenting Skills.

I understand that I am free not to participate at all, not to answer certain
questions and/or to discontinue my participation in the study at any time
without penalty.

I understand that the results of the study will be held in strict confidence and
I will remain anonymous in any report of the research findings. Within these
restrictions, results of the study will be made available to me at my request.

I understand that if there are any questions or concerns that arise as a result

of participation in this study, that I can contact Jerry Adams, at (517) 351-
7810.

Signed by:

 

Date:

 

APPENDIX D

FEEDBACK TO PARTICIPANTS

211

FEEDBACK TO PARTICIPANIS

Researcher: Jerry Adams, M.A.
Supervisor: Robert Caldwell, Ph.D.

The aim of this research project is to examine the relationship between a mother’s
current parenting style and her own childhood memories and experiences.

It is general knowledge that we all learn from our own past experiences. For better or
for worse, our past experiences have an influence on our current behavior. This is very
much the case in terms of the ways in which we engage in our parental responsibilities.
A mother’s current parenting style is significantly influenced by the ways her own
parents behaved and performed their parental responsibilities.

For example, most of us have entered adulthood and the role of parenting determined _th
to behave or engage in certain behaviors that we did not like in our parents. We have
all heard parents say things like, "my mother was too strict and I would never be that
way with my own child." On the other hand, there are many of our own childhood
experiences that we would like to provide for our own children. For example, ”I try to
be as patient with my child as my mother was with me. " It is in this way that a mother’s
current parenting style can be influenced by her own childhood experiences.

The current research is intended to further explore this relationship between a mother’s
childhood memories and her current parenting style. One very important aspect of
parenting style has to do with how a parent defines and evaluates child behavior. The
ways in which a mother defines a child’s behavior is obviously an important factor
influencing the nature and quality of the parent-child interaction. Hence, the current
study is an attempt to understand the relationship between a mother’s childhood memories
and which behaviors a mother thinks are negative, inappropriate and worthy of
punishment and which behaviors are positive, appropriate and thus worthy of praise.

I would like to thank you for your time, interest, and participation in this study. After
the data is analyzed and the study is completed, I will be mailing out a summary
regarding the general results and implications. For those of you who have signed up for
the Parenting Skills Workshop, I look forward to our continuing the discussion of these
ideas, as well as other concepts and specific suggestions related to child management.

Finally, it is possible that you may feel unsettled about the memories and information
that you recalled today. If these feeling do not quickly subside and continue to be
upsetting, or if you have general comments and questions regarding the study, please feel
free to call me, Jerry Adams, at 351-7810.

 

APPENDIX E

FAMILY INFORMATION SHEET

Ple‘

(_‘a
34"
p—n.

2 12
FAMILY INF RMATION EI‘

Please answer all questions as completely as possible.
1. Mother’s Date of Birth:

Father’s Date of Birth:

Son’s Date of Birth:

2. How many children are there in you current family?
Please list the children’s ages:

Boys:
Girls:

 

 

3. Mother’s relationship to son? (Please check one)
Natural Parent
Step-Parent
Other (Please Specify)

How many years have mother and son been living together?

4. What is your highest level of education completed?
(Please be specific, e. g., one year of junior college)

 

 

5. State your type of employment. (Please be specific, e. g., English teacher at Junior
High School)
6. What is your husbands highest level of education completed?

(Please be specific, e. g., one year of junior college)

 

 

7. State your husband’s type of employment. (Please be specific, e. g., English teacher
at Junior High School)
8. Please state the approximate annual family income. This should include both parents’

combined income.

Thank You.

APPENDIX F

FAMILY TOGETHERNESS QUESTIONNAIRE

- ..\\

21 3
FAMILY TOGETHERNESS QUESTIONNAIRE

Instructions: This brief questionnaire is designed for parents with one or more children in order
to give you an opportunity to describe your family. There are no right or wrong answers. Each
question has four possible answers, and you simply mark the square under what you believe to
be the best answer for each question. The four possible options are (1) very untrue for our
family; (2) fairly untrue for our family; (3) Fairly true for our family; and (4) very true for our
family.

You may feel that some statements are very true for some family members and less true or untrue
for others. Judge each statement in terms of most family members.

Remember, we would like to know what your family seems like to you. Do not try to figure out
what other people might think of your family, but give us yen; impressions of your family. Be
as careful and honest as you can, and be sure to answer all 5 of the questions.

 

l 2 3 4
Statement Very Fairly Fairly Very
untrue untrue true true
for our for our for our for our
family family family family

 

1. Family members find it
hard to get away from each
other.

 

2. It is difficult for family
members to take time away from
the family.

 

 

 

 
 
  

3. Family members feel pressure
to spend most free time together.

 

4. Family members feel guilty
if they want to spend time alone.

   

 

5. It seems like there is never
any place to be alone in our
house.

 
   

 

 

 

 

 

 

APPENDIX G

INSTRUCTIONS FOR THE CHILD BEHAVIOR SCENES

 

or 1
diff
the

YOL

thel
posi
you
is e:

film
the b
oldi-

For e

214
INSTRUCTIONS FQR THE CHILD BEHAVIQR SCEEES

The following stories are designed to help study the different ways that parents follow
or keep track of their child’s behaviors, and how parents vary in the way they judge
different child behaviors. Clearly, the way in which a parent both identifies and defines
the appropriateness of their child’s behavior is an important part of parenting.

You will be asked to read several short stories involving boys that are interacting in a
variety of situations. The children will be interacting with their peers, their teachers and
their families. You will be asked to identify those behaviors you think are desirable and
positive and to identify those behaviors that you think are negative and undesirable. As
you think about the behaviors as negative or positive, try to imagine that your own son
is engaged in the behavior that you are reading about.

W As you read through the stories, please underline with the w Pen
all those child behaviors that you consider negative, undesirable, aversive, hurtful,
and/or inappropriate.

INSIIIBCCTIQNS; As you read through the stories, please underline with the CM
pen all those child behaviors that you consider positive, desirable, helpful, appropriate
or prosocial.

Part of the task involves being able to recognize different segments of child behavior.
The boys in the stories will be engaged in a wide variety of behaviors. It will be
important to read carefully and pay attention to what the boys do, what they think, how
the boys react and respond, what they say, and how they feel. All these are examples
of different child behaviors.

For example: Bob walks home from school. Tom, one of his friends, calls out: "see
you tomorrow." On Bob’s way home, he crosses the street. He is
looking for something on the ground. Bob picks up a rock and throws the
rock into the woods.

This example has the following behavioral segments:

1 Bob walked home from school.

2 Tom called out, "see you tomorrow."

3 Bob crossed the street.

4. Bob was looking for something on the ground.
5 Bob picked up the rock.

6 Bob threw the rock into the woods.

215

It is very important that you underline the entire behavioral segment that you have
identified. The stories are full of many different examples of child behaviors. You may
decide that a sentence has more than one behavior and therefore it is important that each
behavior that you think is either negative or positive be completely underlined.

Past experience has shown that mothers go about this task in many different ways. It
may be helpful to let you know that there are no right or wrong answers. You are not
being tested. It is most important that you just be yourself and identify only those
behaviors (as positive or negative) that make sense to you. The study is interested in
what you, as a mother, think about different child behaviors. So ...... relax, take your
time, and please read carefully.

Firm reminger: underline with RED all those child behaviors you think are negative and
underline in GREEN all those child behaviors that you think are positive. (If you need
to review the definition of positive and negative behaviors, please refer to the above

paragraphs marked as INSTRUCTIQNS.

 

If you have any further questions, please feel free to ask the researcher.

If you have no further questions, you are now ready to begin reading the stories.

PLEASE TURN TO THE NEXT PAGE.

APPENDIX H

THE EXPERTS’ FINAL DETERMINATION OF THE POSITIVE (P),
NEGATIVE (N), OR NEUTRAL (BLANK) VALUE FOR EACH
BEHAVIOR IN THE CHILD BEHAVIOR SCENES

10.

ll.

12.

13.

14.

15.

16.

17.

18.

19.

220.

21.

216

AFTER S HOOL

Bobby says goodbye to the teacher ........................

He picks up his book bag ................................

and then he starts to walk home ..........................

On his way home, he sees a younger boy across the street.
and Bobby notices that the boy looks upset ...............

He walks over to the younger boy .........................

He realizes the boy is his neighbor Tom ..................
Tom is crying ............................................

Brad, an older student, is teasing Tom for crying ........
He is pushing and shoving Torn ............................
Bobby asks Brad to stop teasing Tom ......................
This doesn’t work, so Bobby threatens Brad ...............
Bobby throws Brad’s books into the street ................
Brad picks up his books .................................

and he walks away ........................................

Tom thanks Bobby for helping him ............................
Tom then asks Bobby to come to his house for a snack .....
Bobby says, "thanks, but I can’t. I need to get home"...
Bobby is lying about nwding to go home ..................
Bobby walks across the street ............................

and enters the little store ..............................

 

 

 

zH‘FH

 

Z

 

 

 

 

 

 

(scene. 1)

Si

If.)

217

The story owner greets Bobby

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

and Bobby responds by saying, "hello Mr. Niles.
How are you?" .....................

Bobby stands and looks at all the different candy bars...
He looks to see where the store owner is standing ........

Then, he quickly puts a piece of candy into his
jacket pocket .....

and he walks out the door ................................

He then eats the stolen candy bar ........................
When he gets home, his mother is happy to see him ........
He gives her a hug .......................................

Then he walks across the room into the kitchen ...........
He is starting to feel guilty about stealing the candy...

He decides to take out the trash .........................

Mother seems surprised at his behavior.

33.

He then tells his mother about stealing the candy ........

 

 

 

 

 

 

 

 

 

._2__

She informs him that he will have to miss his favorite activity for a week.

34.
35.

36.

Bobby says: "that makes me really mad." ..................
He starts to whine : "that just isn’t fair!" .............

He kicks the chair in anger ..............................

His mother tells him to go to his room.

37.

38.

39.

40.

As he walks away, he swears at her .......................
Then, he goes to his room as asked .......................
After some time, he apologizes for swearing ..............

He promises to never swear again .........................

 

 

 

 

 

 

218

ETTIN READY R DINNER

1. John asks his mother, "what time will dinner be ready?"..

She says in 20 minutes.

2. John asks what they are having for dinner ................

His mother answers: "We will be having salad, peas and meatloaf."

3. Paul, John’s brother, walks into the room ................
4. Paul interrupts the conversation .........................

5. He starts complaining that he hates peas .................
6. Paul then starts whining: "I want to eat spaghetti!" .....
His mother impatiently yells at him to stop whining.

7. Paul starts crying .......................................
Mother says: "quit being a baby and stop crying!"

8. Paul stops crying ........................................

9. He then says to his mother: "It hurts my feelings
when you yell at me." .............

She replies, "don’t be so sensitive". As mother walks away,

10. Paul says, "you know, sometimes it’s really hard
to talk to you." ..................

Dad comes home from work. He calls out, "hello, I’m home."
11. Paul greets dad at the door ..............................
12. He gives his father a hug and a kiss .....................

13. Paul tells his father that mother has hurt his feelings.

 

Z

 

2

 

Z

 

 

 

 

 

M

II
.6

Mot

Mother calls out from the kitchen and asks for volunteers to set the table for dinner.

14.

15.
16.

17.

18.
19.

20.

21.

22.

219

John says, "it’s not my turn to set the table,
I did it last night." .......................

Paul says: "Alright, I’ll do it." ........................
He begins to set the table for dinner ....................

While setting the table, Paul drops a glass that
breaks all over the floor ........................

John laughs at Paul for breaking the glass ...............

John says, "I’m going to tell mother you broke the glass."

John runs to the kitchen to tell his mother that
Paul has broken a glass .....................

Paul explains that it was an accident ....................

He apologizes for breaking the glass .....................

Mother accepts his apology.

23.

24.
25 .
26.
27.
28.

29.

Paul tells his brother, "it makes me mad when you
try to get me in trouble. " .............

John responds by grinning .................................
Paul says to his brother John: " you are an idiot! " .......
and then Paul hits his brother on the arm ................
Paul gets a broom to sweep up the broken glass ...........
He cleans up the broken glass ............................

He finishes setting the table ............................

Mother announces that dinner is ready.

 

 

 

 

 

 

 

 

 

 

2

 

Z

 

Z

 

 

 

 

 

 

1.

220

DINNER

Todd is watching TV ......................................

Mother informs Todd that dinner is ready.

2.

Todd says, "I’ll go upstairs and tell everyone
that dinner is ready." .......

Mother says, " thank you Todd."

3.

Todd goes upstairs .......................................

The door to his father’s bedroom is closed.

4.

Todd knocks on the bedroom door ..........................

His father calls out, "come in."

5.

Todd enters his father’s room and says,
"dad, dinner is ready." ......

Dad responds, "thank you son, I’ll be right down".

6. Todd then goes to Steve’s room, his older brother ........
The door is closed.
7. Todd opens the door without knocking .....................
8. Steve yells out, "get out of here you creep!" ............
9. Steve is secretly looking at a magazine of naked women...
10. He offers Todd a dollar if he doesn’t tell

about the magazine ......
11. Todd takes the dollar ...................................
12. and he promises not to tell his parents

about Steve’s magazine .......
13. Todd remembers to wash his hands before dinner ..........

 

 

 

 

 

 

 

 

(scene.3)

221

14. He reminds his brother to wash his hands too ............. P

 

Dad is also washing. They all go downstairs for dinner.

15. Steve volunteers to help serve dinner .................... P

 

16. Todd sits down at the table ..............................

17. Steve asks Todd what he wants to drink ................... P

 

18. Todd answers that he wants pop ...........................
Dad says, "not again, please drink something else tonight."

19. Todd says, "that makes me mad when I can’t
have what I want to drink." .......

 

Father reminds him that there is a limit on how much pop they can drink.
20. Todd complains "that isn’t fair!" ........................ N

Dad responds: "life isn’t always fair son." Mother asks Todd to pick something else to
dfink.

21. Todd begins to whine: "I’m not going to drink anything!
And I’m not going to eat!" ............. N

 

Dad ignores Todd and states, "I’m very hungry. Steve, what did mother make for
dinner?"

22. Steve answers his father, "we are having salad,
chicken and mixed vegetables."....

 

23. Steve brings the food to the table ....................... P

 

24. Todd is sitting at the table pouting ..................... N
Mother asks father, "how was your day dear?" Father starts to talk about his day.

25 . Todd interrupts the conversation ......................... N

 

Mother and father continue to talk and finish their conversation.

26. Steve asks, "dad, will you please pass the chicken." ..... P

 

222

27. Steve takes a piece of chicken ...........................

28. and he offers, "does anyone else want
some more chicken?" .......

29. Todd says "no thanks." ...................................
30. Todd has stopped pouting .................................
31. Todd eats all of the food on his plate ...................

32. Todd tells his mother: "I really liked the chicken.
It’s my favorite!" .....................

33. Steve asks if he can be excused from the table .....

34. He says that he is going up to his room to
start his homework...

Father tells him that he may leave. On his way upstairs,
35. Steve teases the dog with a piece of chicken ...........
Father then asks Todd to help clear the table.

36. Todd brings all the dishes from the table
into the kitchen .............

37. He then begins to load the dishwasher ..................
When this is finished,

38. Todd tells his father,
"I better begin my homework now.".

39. He asks if it will be alright to do his homework
at the dinning room table...

40. Todd asks his father for help with his
math homework...

Mother finishes straightening up in the kitchen.

 

 

 

 

 

 

 

 

 

 

 

223

(scene.4)

AR THEHUE

Mother and father are sitting and relaxing in the den. There is a fire burning in the
fireplace. Their four boys are doing different things around the house.

 

 

 

 

 

 

1. Peter is reading a book .................................. P
2. Roger is working on his model airplane ................... P
3. Mike is outside playing basketball ...................... P
4. Jake is wandering around the house ....................... __
5. Jake walks into the kitchen .............................. __
6. He pockets the $1.00 that his father left
on the kitchen counter ...... N
7. He starts to look through his parents’ mail .............. N
8. He takes a cookie out of the cookie jar .................. __
9. and he goes into the living room to eat it ...............
10. Jake yells towards the den: "there’s nothing to do!" ..... N

 

11. Jake is bored ............................................

12. and he proceeds to pick leaves off some plants
in the living room ...... N

 

There is a book of matches on the table.

13. Jake lights a match ...................................... N

 

14. He throws the burning match into the ashtray .............

15. He looks through some magazines on the coffee table ......

16. He reads an articles on ocean cruises ....................

 

17. He then stacks the magazines neatly ...................... P

 

18. Jake looks out the front window ..........................

224

19. He watches his brother and friends playing
basketball on the driveway ........

 

Mother walks past the living room and sees Jake staring out the window. She says
impatiently: "why don’t you find something to do?" She goes into the kitchen and calls
out, "Jake, the lid is off the cookie jar. Did you have one?"

20. Jake replies, "nope, it wasn’t me." ...................... N

 

21. Jake walks back into the kitchen .........................

 

Mother is pouring a glass of beer to bring to dad in the den. The phone rings and
mother goes to answer it. The call is apparently for her and she starts talking.

22. Jake stands in the kitchen for a minute ..................

23. and then reaches over and takes a sip of dad’s beer ...... N

 

24. He makes a face at the taste .............................

The family cat comes into the kitchen.

 

25 . Jake pulls the cat’s tail ................................ N
26. Then Jake sits on the floor .............................. _
27. He pets the cat for awhile ............................... P

 

28. He picks up the cat ......................................
29. and then he holds the cat in his lap .....................
Mother gets off the phone and returns to the kitchen.

30. Jake says, "I heard a good joke the other day." .......... P

 

31. He proceeds to tell his mother the joke .................. P

 

Mother laughs.

32. Jake feels good because his mother liked the joke ........ P

 

Mother musses his hair in an affectionate way.

33. He pushes her hand away .................................

225

34. and he says, "stop it, mom." .............................

As mother leaves the kitchen she says, "why don’t you fwd the cat."

35 . Jake says "sure, no problem. I’ll do it right now." ......
36. He asks, "should I give the cat the canned stuff. " .......
Mother replies, "alright, but only half of it."

37. Jake feeds the cat ........................................

38. He follows his mother’s instructions .....................
39. He lays on the kitchen floor .............................
40. Jake watches the cat eat .................................

41. He then decides to give the cat some milk ................

The cat drinks the milk and then strolls off into the other room.

42. Jake leaves the milk carton on the counter ...............
43. Jake goes into his bedroom ...............................
44. He begins to read a comic book ...........................

45. He then decides to go see if he can play
some basketball.

46. He puts on his tennis shoes ..............................

47. He asks his brother Mike if he can join the game .........
They need an extra player to make the teams even.

48. Jake is happy because he will be able to play basketball.
49. He is a good basketball player ...........................

50. Jake scores a lot of points during the game ..............

Nonetheless, his team looses.

P

 

 

 

 

 

 

 

 

 

226
51. Jake is a good sport about loosing ....................... P

 

52. He tells the other team, "good game guys!" ............... P

 

The other boys agree it was a good game. After deciding to play again tomorrow, they
all say good-bye and go home.

 

 

 

53. Mike, however, is complaining about loosing .............. N
54. Mike accuses the other team of cheating on the score ..... N
55. Jake attempts to console his brother ..................... P

56. He says to Mike: "chill out, it’s only a game.
We’ll beat ’em next time. We’re going to
play again tomorrow." ................................. P

 

57. Mike starts to cry about how much he was

 

 

fouled during the game ....... N
58. Jake is becoming annoyed with Mike’s behavior ............
59. He shouts at Mike, "shut up!! Quit being such
a poor sport!" ..... N
60. Mike throws a small tantrum by yelling
and pounding his fists on the garage door... N

 

61. Jake goes back into the house ............................

62. He is feeling tired ......................................

63. He takes some root beer from the refrigerator ............
64. He sneaks another cookie from the cookie jar .............
65. He takes the root beer and the cookie to his bedroom .....
66. He eats the snack on his bed .............................

67. He decides to take a shower ..............................

ll|||21 ll

68. After the shower, Jake falls asleep on his bed ...........

Mother wakes Jake and tells him that they are all going out for pizza.

227

69. At first he is grumpy for being awaken ...................
70. Then, Jake is excited ....................................

71. He really likes pizza ....................................

APPENDIX I

CHILD BEHAVIOR RATING SCALES I/II

228
HILD BEHAVI R RATIN AL

The Child Behavior Rating Scale was devised to examine how parents judge certain child
behaviors. The purpose of the scale is to obtain ygu; judgement, as a parent, of both
negative and positive child behaviors. Section I emphasizes child misbehavior. Section
II emphasizes positive or desirable child behaviors.

53119111

Section I of the Child Behavior Rating Scale is designed to help us understand how
parents rate the degree of misconduct of certain negative behaviors.

In this section of the Child Behavior Rating Scale you will be asked about your own
judgement regarding the seriousness of misbehavior. You will be asked to read through
a list of child behaviors and to give your own opinion as to how important or serious you
think each negative behavior really is.

For example, most parents would agree that lying to a friend is a negative and
undesirable child behavior. On the other hand, parents will disagree when judging the
degree of seriousness regarding this misconduct. While some parents would consider
a child lying to a friend to be only a slightly serious misbehavior, other parents would
consider this a serious or perhaps even an extremely serious misbehavior. It is
interesting to point out that even professionals regarded as "experts" in the area of child
behavior would disagree about the level of misbehavior. Obviously, there are no right

or wrong answers to these evaluations, but rather only differences of judgement and
opinion.

Specifically, you will be presented with a list containing a variety of primarily negative,
undesirable, aversive, hurtful, and/or inappropriate child behaviors. You will be
asked to rate each misbehavior in terms of it’s seriousness. Each behavior on the list

is performed by a boy. You are asked to imagine that the boy is the same age as your
own son.

INSTRUCTIONS CONTINUE ON THE NEXT PAGE
PLEASE TURN TO THE NEXT PAGE

229

Each behavior will be rated on the following six-point scale:

 

Slight Serious
Misbehavior Misbehavior
O 1 2 3 4 5
\ \ \ \ \ \
Not at all a Moderate Extremely
Misbehavior Serious Misbehavior
W Please rate each behavior on the list according to this scale. Please

rate each behavior separately. Please show your rating of each behavior by circling the
appropriate number that corresponds with your judgement. For example, the actual list
of child behaviors will look like this:

Please Circle That Number Which Best Reflects Your Judgement of
Each Behavior

0 1 2 3 4 5 The boy lied to his teacher.
0 l 2 3 4 5 Johnny started spitting at the girl.
0 1 2 3 4 5 Bobby was talking too loudly at the dinner table.

Lets look at the first behavior as an example. If you decide that the boy’s lying to his
teacher is not really a misbehavior at all, then you would circle the number "0". If you
think that his lying is only slightly negative or a slightly serious misbehavior then you
would rate the behavior a "2", and you would circle the number "2". On the other hand,
if you thought that lying was a serious or extremely serious misbehavior then you
would circle either "4" or "5" depending on your judgement. The ratings of "2" or "3"
are in the moderate range, with "3" obviously leaning towards the more serious
misbehavior. Once again, you would circle the number that best reflects your actual
judgement.

Remember, there are no right or wrong answers and you are not being tested. The
purpose of the Child Behavior Rating Scale is to help in obtaining your opinion. Please
read each behavior carefully and circle the number that best reflects your judgement of
that behavior.

If you have any further questions now is the time to ask the researcher. If not, you are
ready to begin.

Please wait until the researcher announces:

"Please turn to the next page and begin".

230

 

HILD BEHAVI R RA ALE
SECTION!
Slight Serious
Misbehavior Misbehavior
O l 2 3 4 5
\ \ \ \ \ \
Not at all a Moderate Extremely
Misbehavior Serious Misbehavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345

012345
012345
012345

The boy lied to his teacher.

Johnny started spitting at the girl.

Steve reminded his parents that the movie started in 30 minutes.

Bobby was talking too loudly at the dinner table.

He never told his mother about the phone message.

The boy was interrupting his parents’ conversation.

Pete forgot to lock the door of the house as asked.

The boys were trying to smoke cigarettes in their secret club house.

The boy said to his younger sister: "nobody loves you cuz you’re a little brat."
He came home 20 minutes later than he was told to.

He was screaming as loudly as he could at his mother.

He turned in his book report two weeks late.

Todd threatened the other boy that he would kill his cat.

Chuck was listening in on his mother’s phone conversation.

Fred was blasting his radio.

Al decided to break the mirror in the school bathroom.

The boy lied to his mother about having no homework because he wanted to watch
TV.

Bill told his father that he disagreed with many of his opinions.

Sam refused to do his chores around the house.

The boy was whining because he couldn’t have any ice cream.

231

Section I page, 2

 

Slight Serious
Misbehavior Misbehavior
O l 2 3 4 5
\ \ \ \ \ \
Not at all a Moderate Extremely
Misbehavior Serious Misbehavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345

012345

012345

012345
012345

Bob forged his mother’s signature on his report card.

Rick was in the backyard lighting matches.

Ted was secretly looking at a magazine of naked women.

The boy interrupted the teacher without raising his hand.

He stole some money from his father’s wallet.

John was urinating on the living room floor.

He threw a rock and broke the neighbor’s window because he was mad at them.
He was very angry when he said to his mother: "you’re a mean women!"

Tom’s room was messy because his clothes and toys were all over the floor.

The boys were smoking a marijuana cigarette.

He told his mother that the chicken was overcooked.

He was calling people on the phone and hanging up on them.

He quickly opened the car door and stole a few music tapes.

Gary opened his parents’ bedroom door without knocking.

Sam was bored and no one was home. He decided to see what beer tasted like and
he drank a can of beer from the refrigerator.

Even though his lights were off in the bedroom, Phil was using his ear phones to
secretly listen to his radio.

He fed the dog an hour later than he was supposed to.

Randy was snooping in his parent’s room looking for his upcoming birthday gift.

He was shooting the BB gun at the birds in the tree.

ll:

Pl

232

Section I page, 3

 

Slight Serious
Misbehavior Misbehavior
0 1 2 3 4 5
\ \ \ \ \ \
Not at all a Moderate Extremely
Misbehavior Serious Misbehavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345

The boy told his father that certain rules around the house made him angry.
When the teacher turned around, he shot a rubber-band at the teacher’s back.
This year Dave never wished his mother a happy birthday.

The boy didn’t wash his hands before dinner.

The boy was kissing a girl on the way home after school.

Paul copied his friends homework because he didn’t do it himself.

After the teacher told him to sit down, Ted made an obscene gesture to the teacher.
He said to his father: "sometimes you can really be a jerk!”

He took out a knife and threatened his teacher.

Torn pretended that the little plastic bag was full of marijuana.

After being pushed and shoved by Nick, Chris started swearing at him.
Steve was "playing Doctor" with the girl next door.

He didn’t put the butter back into the refrigerator.

Jack was playing with his father’s rifle.

He said to his father: "I’ll be out playing for the rest of the afternoon."

Jack told his mother: "I hate you!"

He continued to watch TV and ignored his mother’s pleas for him to go to bed.

Setti

Not 2
list

Pleas

233

Section I page, 4

 

Slight Serious
Misbehavior Misbehavior
O 1 2 3 4 5
\ \ \ \ \ \
Not at all a Moderate Extremely
Misbehavior Serious Misbehavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345

012345

012345
012345

012345
012345
012345

Ned’s feelings were hurt and he told his parents, "I’m never going to speak to
either one of you ever again!”

On his way home from school, the boy took a rock and scratched the side of a
parked car.

He persistently kept nagging his mother to take him shopping for some new jeans.
Whenever Sam’s mother asked him about taking out the trash he would say, "Gee,
I forgot."

The boy asked if he could stay up late to finish watching his favorite TV show.
He complained to the teacher that she was giving too much homework.

Paul was sitting at the dinner table playing with his food rather than eating it.

YOU HAVE NOW COMPLETED sscnoN r OF THE CHILD BEHAVIOR RATING SCALE

PLEASE TURN THE PAGE AND READ THE INSTRUCTIONS FOR SECTION H

234
HILD BEHAVI R RATIN AL

Instructions for Section II

In Section I of the Child Behavior Rating Scale you were asked to give your judgement
regarding negative child behaviors. Section II will now emphasize positive or prosocial
child behaviors.

You should begin the Instructions for Section 11 only after you have completely rated
all the behaviors listed in Section I. The Instructions for Sections H assume that you
are now already very familiar with the procedure and format of the Child Behavior
Rating Scale. Section H is basically the same as the first section, except now you will
be rating different behaviors in terms of their W.

W

In Section II of the Child Behavior Rating Scale you will be asked about your own
judgement regarding the desirability of certain positive child behaviors. Specifically, you
will be presented with a list containing a variety of primarily positive, desirable,
helpful, appropriate and/or prosocial child behaviors. You will be asked to rate each
positive behavior in terms of it’s desirability. Each behavior on the list is performed by
a boy. You are asked to imagine that the boy is the same age as your own son.

Each behavior will be rated on the following six-point scale:

 

 

I. _ ' .0 NO I I3! DIFFEREN ls '!___ \L RA I a I-"
Slightly Highly
Desirable Desirable
0 l 2 3 4 5
\ \ \ \ I \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Remember, there are no right or wrong answers and you are not being tested. The
purpose of Child Behavior Rating Scale is only to obtain your opinion. Please read each

behavior carefully and circle the number that best reflects your judgement of that
behavior.

If you have any further questions please raise your hand and ask the researcher. If you
have no questions, then you are ready to begin Section II.

Please turn to the next page and begin Section II

235

BEHAVI R RAT G AL

E TI N 11
INSTRUCTIONS Please rate each behavior on the list according to this scale. Please

rate each behavior separately. Please show your rating of each behavior by circling the
appropriate number that corresponds with your judgement.

 

Slightly Highly
Desirable Desirable
0 1 2 3 4 5
\ \ \ \ \ \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345
012345

The boy woke up early in the morning in order to do his paper route.
Pete let his friend know how much he cared about him.

Ted put some more sugar in his chocolate drink.

Bob helped his blind neighbor walk to school.

He wrote down the phone message on the pad where messages were kept.
He washed his hands for dinner.

George volunteered to help set the table for dinner.

Tim asked the teacher if he could turn his report in a few days late.
Charles was very excited about a new stamp he added to his collection.
Terry decided to watch TV by himself in his room.

Carlos was elected president of his church youth group.

The boy told his father that he was mad at him for missing his hockey game.
During the conversation at dinner, Pete told a joke and everyone laughed.
Jesse practiced his musical instrument three days a week.

Dan told his parents that his brother was smoking cigarettes.

236

Section H page, 2

 

Slightly Highly
Desirable Desirable
0 1 2 3 4 5
\ \ \ \ \ \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345

012345
012345
012345
012345

012345
012345
012345

012345
012345
012345
012345
012345
012345
012345

He called his parents to tell them that he would be a half-hour late.

Ted got five dollars for mowing the neighbors lawn.

Tom stayed up a half-hour past his bedtime in order to finish his homework.

Ben said: "I really don’t like what we are having for dinner. Can I make a
sandwich?"

He made his mother a nice card for Mothers Day.

Stan tried to stop the fight at school.

At the dinner table, Ned passed the potatoes to his father.

When his mother changed her mind, Paul said "I’m really bummed. I hate it when
you do that!"

Rick saved enough of his money to buy his favorite album.

Clearance spent Sunday afternoon writing letters to his friends from camp.

He took the long way home from school because it was such a nice spring day
outside.

Tyronne gave his dog a bath.

Fritz loaned three dollars to his friend from school.

Alan tried to get his parents to st0p fighting.

Don wouldn’t let his friend copy off of his test paper.

The boy asked his parents if they could go to the school play.

On Halloween, Stuart got over three pounds of candy in his bag.

He helped the younger boys build their tree house.

237

Section 11 page, 3

 

Slightly Highly
Desirable Desirable
0 1 2 3 4 5
\ \ \ \ \ \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345

012345
012345
012345
012345
012345
012345

012345
012345
012345

012345
012345
012345
012345

012345

Billy’s discussion with his father about nuclear weapons turned into a heated
debate.

Buck turned on his night-light before falling asleep.

Jeff listened with interest to the stories about his friend’s trip to Mexico.

Jerry made toast and juice for his parents in order to surprise them with breakfast.

Gil turned the TV volume very low since everyone in the house was still sleeping.
Nick announced that he was wearing socks with his shoes.

The boy politely asked if they could possibly go somewhere else for dinner since
he really didn’t like the food at the restaurant where they were going.

Philip told his dad that he was scared about going away to summer camp.

Max showered every evening before going to bed.

He knew his friend was allergic to chocolate so he bought some different types of
candy to share.

Justin read a story to his younger sister.

Dick was bored watching the cartoons so he went outside.

Jim said to his mother, "I wish you’d pay more attention to me."

His younger brother teased him and attempted to provoked him, but Mario was
able to go outside and walk off his frustrations.

Frank’s room was very neat and clean.

238

Section H page, 4

 

Slightly Highly
Desirable Desirable
O 1 2 3 4 5
\ \ \ \ \ \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345

012345

012345

012345
012345

012345
012345

012345
012345
012345

Drake donated all of this weeks allowance to the firnd for needy children.

Mark brushed his teeth every morning before school.

He knew his mother was sick, so he cleaned up the entire kitchen.

He told his friend that they couldn’t bring their sandwiches to his bedroom because
there was no food allowed outside of the kitchen.

Even though Sam was really mad and he felt Nick deserved it, Sam stopped
himself from swearing at his friend.

Tommy told the coach that he needed more encouragement in order to stay on the
team.

The boy told his mother that he was going to bed early.

During the dinner conversation Buck said, "I was really embarrassed today when
the teacher asked me a question and I didn’t know the answer."

The boy asked if he could stay up late to finish watching his favorite show.

He made an effort to keep away from Mike because Mike was always getting into
trouble.

Dick asked his dad for help with his vocabulary list.

Jeff spent the entire afternoon by himself taking an enjoyable walk.

Rich signed up at school to have a pen-pal.

239

Section H page, 5

 

Slightly Highly
Desirable Desirable
0 l 2 3 4 5
\ \ \ \ \ \
Not a Clearly Moderate Extremely
Desirable Desirable
Behavior Behavior

Please Circle That Number Which Best Reflects Your Judgement of Each Behavior

012345
012345
012345
012345

Carl gave up his bus seat to an older women.
Shawn wanted to learn more about his cultural background.
Larry decided to stay home and finish reading his book.

Brad asked his mother if he could sleep over at his friends house.

APPENDIX J

ADULT ATTACHMENT QUESTIONNAIRE

240
W

This questionnaire is designed to improved our understanding of how parents feel that
their own childhood experiences have had an effect on them as parents and as people.
The following questions concern your childhood relationships with your parents (or
parental figures). Most of the questions will ask about your early relationships with your
family and what you think about the ways these relationships might have affected you.
The questions are mostly about your childhood, but a few of the questions are about your
later years and what is going on right now.

To get yourself in the right frame of mind, take a few minutes to think about yourself
as a child. Try to imagine the house(s) you lived in and the school(s) you attended. Can
you get an image of the physical surroundings, or a mental picture of the neighborhood?

Try to remember what your parents and siblings looked like and how they behaved. Try
to imagine yourself as far back in time as you can.

_ ° V a II III 1.7.! I: I . 5. I. '." .I'I ‘91 II ..I 2. II .I “1.1., 1. man _I J. In
fact, it may be easiest for you to give MW of situations or circumstances
when filling out the questionnaire. Please be as specific as you can when recalling an
event, (e. g., referring to what was done, what was said, how people reacted etc.).

Remember, there are no right or wrong answers. The questions are designed to gather
information regarding your early relationships with your family. Let yourself be as
relaxed as possible so that you may feel free to write down whatever comes to mind.
Please do not be concerned with spelling, punctuation, and/or grammar, but rather please
answer the questions in as much detail/content as possible. (You may use the backs of
pages or additional pages if you wish. But please label these clearly.)

The questions are designed to be answered in order. Please do not look ahead in the
Questionnaire. Rather, start with Question 1 and when you have W

Wk proceed to Question 2. Once a question has been completed, please
do not go back and change the answer. Finally, please use a pen and not a pencil.

(Please note the time you started the questionnaire, as I will want to know how much
time it took to complete.)

What time is it now?

PLEASE TURN TO THE NEXT PAGE AND BEGIN THE QUESTIONNAIRE

241

1. Starting from as far back as you can remember, please describe your relationships
with your parents. (Try to provide specific examples.)

242

2. If you had to choose four adjectives that reflect your ehildhoog eeIaMLrgIIjn with
m mether which four would you choose?

 

 

3. Now, for e_aM adjective say why you think that particular adjective is a good
descriptor of your childhood relationship with your mother. Why did you choose that
adjective? (It may be easiest to give a smific example of a time when she did or said
a certain thing or acted in a particular way.) Remember that detafl is important.

First adjective:

Second adjective:

Third adjective:

243

Fourth adjective:

4. If you had to pick four adjectives to describe the most important emotional reactions
you had to your mother (the most important ways she made you feel or affected my:
fleeing), which four would you choose?

 

PI nin ih ’ nhn

244

5. If you had to choose four adjectives that reflect m c_hi_lg_h_qed mm with
me father which four would you choose?

 

6. Now, for _e_a_el_r_ adjective say why you think that particular adjective is a good
descriptor of your childhood relationship with your father. Why did you choose that
adjective? (It may be easiest to give a smifie exaMQIe of a time when he did or said
a certain thing or acted in a particular way.) Remember that detail is important.

First adjective:

Second adjective:

Third adjective:

245

Fourth adjective:

7. If you had to pick four adjectives to describe the most important emotional reactions
you had to your father (the most important ways he made you feel or affected 19!:
LLIings), which four would you choose?

 

8. During your childhood it is very likely that you felt closer to either your mother or
your father. To which parent did you feel the closest and why? Why wasn’t there this
feeling with the other parent?

246

9. When you were upset as a child what would you typically do?

These are a series of follow- -up questions concerning how you got comforted as a child.
Searching back in your memory as far as you can, W what you would
typically do if you were in need of comforting. For each question, please say whether
you would tell anyone, and if you did, whom did you tell, and what would that person
typically do (e. g. hold you, talk to you, tell you a joke, play a game, ignore you, tease
you, etc).

9a. When you were emotionally upset as a child what would you do? Were there
"typical" things or events that made you emotionally upset? (e. g., hurt feelings,
disappointments, bad dreams, being scared etc.) Please describe W or
exaMQIes that you remember. If you did turn to your parents when you were upset, how
did they comfort or soothe you?

247

9b. What would happen when you were physically hurt or injured? Please describe some
W of being hurt or injured. Did you tell anyone? Whom did you tell?
How did they comfort or soothe you?

9c. Finally, what would happen when you were ill as a child. Again, please provide
W of when you were ill, what you would do, and how others, specifically
your parents, would respond.

248

10. How did you go about getting attention as a child? How did you get your parents
to respond and/or attend to you and your needs? (Again, We: will be
helpful.)

10B. Did your parents take an interest in your activities and accomplishments? How
much importance was placed on your activities and your achievements? (New nrevide
smifle examples).

249

10C. Did your parents seem to take more of an interest in your activities and
accomplishments than in your emotional concerns and needs? Or, was there a balance

between the two?

Were there some things about you and/or your activities that interested your parents more
than other needs and/or activities?

11. During your childhood, were there any adults, other than your parents, to whom you
felt especially close? If so, please describe the relationship and the circumstances, for
example, say what their relationship was to you, how much time you spent with them,

how old you were at the time, etc.

250

12. During your childhood, were you ever separated from either of your parents for
what felt like a long time? If so, for how long? How old were you? What was the
reason for the separation? How did you feel about it at the time?

13. When you were a child, were your parents ever threatening to you in any way,---
maybe because they were angry, or for discipline, or maybe just jokingly? For example,
did your parents ever threaten to hurt you, to leave you, or send you away? Did they
ever call you names and/or say that you were a bad child? Please give Mtge
W. Did this happen frequently? How do you think this experience(s) affects you

now as an adult?

251

14. Did you ever feel rejected as a young child? Of course, looking back on it now,
you may realize it was not really rejection, but this specific question is concerned with
whether you ever fielt rejected as a child? How often did you feel rejected? What were
the circumstances? How old were you and what did you do? (Again, it may be easiest

to give Wales.)

14A. Why do you think your parents did those things? Do you think he/she realized
he/she was rejecting you?

252

15. At this point in the questionnaire you have probably shared a lot of childhood
memories regarding your relationship with your parents, remembering a variety of
specific interactions. In your opinion, why did your parents behave the way they did

toward you when you were a child?

16. What effect, if any, did your early experiences and relationships with your parents
have on your adult personality? On your adult relationships with friends and lovers?
Are there any experiences that you feel were a set-back in your development?

253

17. Have there been many changes in your relationship with your parents (or remaining
parent) since childhood? In what way(s) have these relationships changed over the years?
In what way(s) have they remained the same?

17A. Describe your current relationship with each of your parents. What is the
relationship like for your now as an adult?

254

18. For most of us, there were some aspects of our childhood relationships with our
parents that were, to varying degrees, troublesome, conflictual, and/or problematic. Do
these issues remain unresolved and still active? Or, have these troublesome aspects of
your childhood relationships with your parents been resolved and/or put to rest?

Please dflfibe in deMiI the process of how you have dealt with these troublesome or
problematic aspects of your childhood relationships.

19. Is there any particular thing which you feel you learned above all from your own
childhood experiences? What would you hope your child might have learned (or will
learn) from his/her experiences of being parented?

255

Now that you have almost completed the questionnaire, the next two questions focus on
your own subjective feelings regarding your ability or inability to remember your own
childhood. These questions are concerned with how LIE—Ed about your ability or
inability to recall your own childhood experiences.

1. How well do you feel you could remember your own childhood experiences? (Please
circle the number that best reflects your own judgment. Do not pick a rating between two
numbers.)

 

Vague Memories Clear Memories
0 1 2 3 4
\ \ \ \ \
Basically Unable to Average Extremely Vivid
Remember Childhood Ability to Recall Memories with
Strong Ability
to Recall

2. Now that you have rated your own ability to remember your childhood, how
confident are you about this rating? (For example, say that you rated yourself as having
only vague memories (#1), how confident are you that you only have vague memories
of childhood.) (Please circle the number that best reflects your level of confidence. Do
not pick a rating between two numbers.)

 

Slightly Strongly
1 2 3 4 5
\ \ \ \ \
Not at all Fairly Totally
Confident Confident

3. Were some aspects of your childhood easier to remember than others? What was
easy to remember? And what was difficult to remember?

Thank you very much for taking the time to fill out this questionnaire!
Please note what time it is now

 

PLEASE CONTINUE TO THE NEXT PAGE.

256

Since this questionnaire is in the continual process of development and modification, it
is very important to get your reactions to answering these questions. Please take a few
minutes to let me know what it was like for you to answer these questions. Anything
you want to share regarding your reactions to the questions will be helpful. (Also, if you
have not already done so, please note the amount of time it took to complete the
questionnaire.)

Finally, if for any reason you are feeling upset or unsettled and would like to talk about

your reactions to the various questions, please feel free to call me directly at 351-7319.
Leave your name and number and I will get back to you as soon as possible.

Thank you very much for your time and cooperation.

Jerry Adams

LIST OF REFERENCES

REFERENCES

Achenbach, T.M. , & Edelbrock, C. S. (1978). The classification of child
psychopathology: A review and analysis of empirical efforts. Psyehelegiefl Bellep'e,
85, 1275-1301.

Anderson, K.E., Lytton, H., & Romney, D. (1986). Mothers’ interactions with
normal and conduct-disordered boys: Who affects whom? WM,
22, 604-609.

Bauer, W.D., & Twentyman, C. T.(1985). Abusing, neglectful, and comparison
mothers’ responses to child-related and one child-related stressors. JeemaLef

CenseltM g Me Clinieal Peyehelegy, i3, 335-343.

Baumrind, D. (1971). Current patterns of parental authority. Qeyelepmeptal
Bulldog» fl, 1-103.

Bell, R.Q. (1979). Parent, child, and reciprocal influences. WW,
MOO), 821-826.

Belsky, J. (1984). The determinants of parenting: A process model. Chm
WI 5;, 83-96'

Belsky, J., & Pensky, E. (1988). Developmental history, personality, and family
relationships: toward an emergent family system. In R. Hinde & J. Stevenson-Hinde

(EdS) WWW (pp 193 217) New York

Oxford University Press.

Berkowitz, L. (1973). Control of aggression. In B. Caldwell & H. Ricciuti (Eds. ),
RviWIfhflwvelmenrth 'IIleIm 1!,0 -._,III
(pp. 95- 140). Chicago: University of Chicago Press.

Bloom, BL. (1985). A factor analysis of self-report measures of family functioning.
Fmily Preeess, 23(2), 225-239.

Bretherton, I. (1985). Attachment Theory: Retrospect and prospect. In I. Bretherton
& E. Waters (Eds. ), Growing points in attachment theory and research, hieeegeapbe
ef Me Seei Iety fer Regatta; M C hrle Qevelepment, 511 (1-2 Serial No. 209), 3- 35.

Bugental, D.B., & Shennum, W.A. (1984). "Difficult" children as elicitors and
targets of adult communication patterns: An atuibutional-behavioral transactional

analysis. Monographs of the Society for Research in Child Development, _4_9_ (l , Serial
No. 205), 1-70.

257

258

Cahill, M.F. (1978). A search for the meaning of the "difficult" child. moon
WILL 12, 18,154.

Christensen, A., Phillips, 5., Glasgow, R., & Johnson, S. (1983). Parental
characteristics and interactional dysfunction in families with child behavior problems:
A preliminary investigation. Journal of Abnormal Child Psychology, HO), 153-166.

Crowell, J. A., & Feldman, S. S. (1988). Mothers’ internal models of relationships and
children’s behavioral and developmental status: A study of mother-child interaction.

mm L9 1273- 1235

Delfini, L.F., Bernal, M.E., & Rosen, P.M. (1976) Comparison of deviant and
normal boys in home settings. In E]. Mash, L.A. Hamerlynck, & L.C. Handy
(Eds.), W New Yorszmnner/Mazel.

Forehand, R., Brody, G., & Smith, K. (1986). Contributions of child behavior and
marital dissatisfaction to maternal perceptions of child maladjustment. mm
W. 25.. 4348.

Forehand, R. ,King, H, Peed, 8., & Yoder, P. (1975). Mother-child interactions. .
Comparison of a non-compliant clinic group and a non-clinic group. Behayio;

W21, 13,
79-84.

Forehand R., Wells, K.C., & Sturgis, E.T. ( 1978). Predictors of child noncompliant
behavior in the home. Journal of Consulm 2 god glinioal Psyohology, fl, 179-180.

Frank, S., Jacobson, 8., Hole, C., Justkowski, R., & Huyck, M. (1986).
Psychological predictors of parents’ sense of confidence and control and self- versus
child-focused gratifications. Developmengl Psychology, _2__2_, 348-355.

George, C., Kaplan. N.. & Main. M. (1985)- WWW
Unpublished manuscript, University of California, Berkeley.

Greenberg, M., & Speltz, M. (1988). Attachment and the ontogeny of conduct

problems. In J. Belsky & T Nezworski (Eds.), Clinioo; Imoligo'ons of Attoohmont
(pp.l77-218). I-Iillsdale, NJ: Lawrence Erlbaum Associates.

Griest, D.L., & Wells, K.C. (1983). Behavior family therapy with conduct disorders
in children. Behavior Thoraoy, 1_4, 37- 53.

259

Grossmann, K., Fremmer—Bombik, E., Rudolph, J ., & Grossmann, KB. (1988).
Maternal attachment representations as related to patterns of infant-mother attachment
and maternal care during the first year. In R. Hinde & J. Stevenson-Hinde (Eds. ),

WW (pp 241-262) New York Oxford

University Press.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment
process. J fP r ' d i P h l , 5;, 511-524.

Hetherington, E. M., & Martin, B. (1986). Family factors and psychopathology in

children In HC Quay & JHS Worry (EdS) WW
ohildhotxl (3rd ed. ). New York. Wiley.

Hinde, R. A. (1980). Family influences. In M. Rutter (Ed.), Wm
WM (pp. 47-63). London: Heinemann Medical.

Holleran, P. A. Littrnan, D. C., Freund, R. D., Schmaling, K. B., & Heeren, J.
(19823) ,_ Igle II .III.C I I I; A’Oltn MulI,_II If "._I ‘
-._ I III iIV II‘ - in I Drain- If I ._ ._ I Ii I‘ .‘II LI. ' . Unpublished
manuscript, Oregon Social Learning Center, Eugene, Oregon.

Holleran, P. A., Littman, D. C, Freund, R. D., & Schmaling, K. B. (l982b). A
signal detection approach to social perception: Identification of negative and positive

behaviors by parents of normal and problem children. MW
Psyohology, _9, 547-557.

Johnson, S.M., & Lobitz, GK. (1974). Parental manipulation of child behavior in
home observations. Joomfl of Aooligz! Bohavior Aoalysig, 2, 23-31.

Johnson, S.M., Wahl, G., Martin, S., & Johansson, S. (1973). How deviant is the
normal child? A behavioral analysis of the preschool child and his family. In R.

Rubin, J. Brady, & J. Henerson (Eds.), Advaogo in bohovior thorooy; Vol, 4 (pp.
37-54). New York: Academic Press.

Kaufman, J ., & Zigler, E. (1987). Do abused children become abusive parents?
Amorim Journal of Qflhoosychiam, 2, 186-192.

Kazdin, A. (1982). Symptom substitution, generalization, and response covariation:
Implications for psychotherapy outcome. Psychologigol Bolleoo' , fl, 349-365.

Knutson, J .F. (1982). Perspective on chapter 7. In G. Patterson, goeroivo famfl y
process (pp.l65-l66). Eugene, Oregon: Castilia Press.

260

Lobitz, G.K., & Johnson, S.M. (1975a). Parental manipulation of the behavior of
normal and deviant children. Child Dovoloomont, go, 719-726.

Lobitz, G. K. & Johnson, S. M. (1975b). Normal versus deviant

children: A multi-method comparison WW 3
353- 374.

Loeber, R., & Schmaling, K. (1985a). The utility of differentiating between mixed

and pure forms of antisocial child behavior. f A ' P l ,
13, 315-336.

Loeber, R, & Schmaling, K. B. (1985b) Empirical evidence for overt and covert

patterns of antisocial conduct problems: A meta-analysis. MW
Embology ll 337- 252

Lorber, R. (1981). Parental tracking of childhood behavior as a function of family
stress (Doctoral dissertation, University of Oregon, 1981). Wm
Lutemational, 51.2.(5), 2066-13.

Main, M. & Goldwyn, R. (1984). Predicting rejection of her infant from mother’s
representation of her own experience: Implications for the abused-abusing

intergenerational cycle. Child Aoooeg Q, Nogloot, 8, 203-217.

Main, M. & Goldwyn, R. (1985). MW
Unpublished manuscript, University of California, Berkeley.

Main, M. & Goldwyn, R. (1989). Adolt agachmon; gm 2 god clgooifioao'on §y§gm.
Unpublished manuscript, University of California, Berkeley.

Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood and
adulthood: A move to the level of representation. In I. Bretherton & E. Waters
(Eds.), Growing points in attachment theory and research, Monoggoho of the Sooiety
for Regroh in Child gvoloomont, S_Q (1-2 Serial No. 209), 66-104.

Main, M., & Weston, D. (1982). Avoidance of the attachment figure in infancy:
Descriptions and interpretations. In C.M. Parkes & J. Stevenson-Hinde (Eds.), Ibo

olaoo of agaoflont in human behavior. New York: Basic Books.

Minuchin, S. (1974). Familios and Family Thomy. Cambridge, MA: Harvard
University Press.

Myers, J. (1979). Fundamengls of Exmrimental Design (3rd ed.). Boston: Allyn and

Bacon.

261

Morris, D. (1981). Attachment and intimacy. In G. Stricker (Ed.), My (pp.305-
323). New York: Plenum.

O’Leary, K.D. , & Emery, RE. (1984). Marital discord and child behavior problems.

In M. Levine & P. Satz (Eds.), WW (pp-
345-364). Baltimore: University Park Press.

Olweus, D. (1980). Familial and temperamental determinants of aggressive behavior
in adolescents: A causal analysis. v 1 m n P h l , 16(6), 644-660.

Patterson, G. R. (1975). Multiple evaluations of a parent-training program. In T.

Thompson & W Dookens (EdS) W (pp 299-
322). New York: Academic Press.

Patterson, G. R. (1976). The aggressive child: Victim and architect of a coercive
system. In E]. Mash, L..A Hamerlynck, & L..C Handy (Eds. ), Bgohaflor
MW. New York. Brunner/Mazel.

Patterson, G. R. (1982), W. Eugene, Oregon: Castilia Press.

Patterson, G.R. (1986a). Performance models for antisocial boys. Amofloao
Poyohologist, 4_1, 432-444.

Patterson, G. R. (1986b). Maternal rejection: Determinant or product for deviant

child behavior? In W. Hartup & Z. Rubin (Eds.), WWW (pp.
73-94). Hillsdale, NJ: Lawrence Erlbaum Associates.

Patterson, G.R. (1986c). The contribution of siblings to training for fighting: A
microsocial analysis. In D. Olweus, J. Block, & M. Radke-Yarrow (Eds.),
D'VlmtI‘II‘, If ‘nIIII. .. I ’II II 1 Be ViIr' R .'.._ I Thri , I I ‘
(pp. 235-261). Orlando, Florida: Academic Press.

Patterson, G.R., & Dishion, T.J. (1985). Contributions of families and peers to
delinquency. Criminology, 23, 63-79.

Patterson, G. R., Dishion, T. J., & Bank, L. (1984). Family interaction: A process
model of deviancy training. W, _Q, 253-267.

Patterson, G.R., & Reid, LB. (1973). Intervention for families of aggressive boys:
A replication study. Behovior Research and Therapy, Ll, 383-394.

Patterson, G.,R., & Stouthamer-Loeber, M. (1984). The correlation of family
management practices and delinquency. Child Devolopmont, 55, 1299-1307.

262

Quay, H. (1979) Classification. In H.C. Quay & J..S Werry (Eds.),
W (2nd ed”) New York: Wiley

Quay, H. (1986a). Classification. In H..C Quay & J..S Werry (Eds.),
W (3rd ed”) New York: Wiley

Quay, H. (1986b). Conduct disorders. In H. C. Quay & J. S. Werry (Eds. ),
P l i i r f hil h (3rd ed..) New York: Wiley.

Rickard, K.M., Forehand, R., Wells, K.C., Griest, D.L., & McMahon, R.J. (1981).
Factors in the referral of children for behavioral treatment: A comparison of mothers
of clinic-referred deviant, clinic-referred non-deviant and non-clinic children.

MW, 12. 201-205-

Rickard, K.M., Graziano, W., & Forehand, R. (1984). Parental expectations and
childhood deviance in clinic-referred and non-clinic children. M
mam. 11(2). 179-186.

Ricks, M. H. (1985). The social transmission of parental behavior: Attachment across
generations. In I. Bretherton & E. Waters (Eds. ), Growing points in attachment

theory and research W
50 (1-2 Serial No. 209), 211-227.

Robinson, EA. (1985). Coercion theory revisited: Toward a new theoretical

perspective on the etiology of conduct disorders. W, 5, 597-
625.

Robinson, E. A., & Jacobson, N. S. (1987). Social learning theory and family
psychopathology: A Kantian model in behaviorism? In T. Jacob (Ed.,) Family
mtegacg'gn nags! psyg hgpathglggy; Thggrias, mgmgig, Q51 findings (pp. 117-162). New

York: Plenum Press.

Russo, D. ,Ca,taldo M. ,& Cushing, P. (1981). Compliance training and behavioral
covariation in the treatment of multiple behavior problems. Wed

Behavigr Agflygi 15, 1_4(3), 209- 222.

Rutter, M, & Garmezy, N. (1983). Developmental psychopathology. In P Mussen
(Ed. ), I. . ”Ck 9f hil. 3 h. 10‘ V01 4 9; «.-_.':!IflI11'I 1- K 1 Q .A -1.
Wm (pp. 776-911). New York: Wiley and Sons.

Rutter, M., & Giller, H (1983). Juvenile delingggngy; Trend§ aad pefipegg'veg
England: Penguin.

263

Sallows, GO. (1973). Responsiveness of deviant and normal children to naturally
occurring consequences (Doctoral dissertation, University of Oregon, 1972).
W 3302), 6092-3.

Schmaling, K. B. (1983). T hni r n n r ° n f hil v' .
Unpublished manuscript, Oregon Social Learning Center, Eugene, Oregon.

Schmaling, K. B., & Patterson, GR. (1984, November). W
I‘ .. «. I I I .1,__ hild I‘thrgII Ir'a {In I ‘ hilI in e Im. Paper
presented at American Association for the Advancement of Behavior Therapy,
Philadelphia.

Shaver, P. (1984). Attributions and child-caretaker interactions: Commentary by
PhilipShaver. h f i er ° ' v mn,52(l,
Serial No. 205), 71-76.

Shaw, DA. (1972). Family maintenance schedules for deviant behavior (Doctoral

dissertation, University of Oregon. 1971). WWW. 32(9).
5458-B.

Simon, R.L., Whitbeck, L.B., Conger, R.D., & Chyi-In, W. (1991).

Intergenerational transmission of harsh parenting. v 1 n P h l , 22, 59-
69.

Snyder, J .J . (1977). A reinforcement analysis of interaction in problem and non-

problem children. Jddmal of Abndrmfl Esydhdldgy, _8_6, 528-535.

Snyder, J.J . & Patterson, GR. (1987). Family interaction and delinquent behavior.

In H. Quay (Ed.), fiadbdok of jdvgnile delingdangy (pp. 216-243). New York:
Wiley & Sons.

Sroufe, L.A. , & Fleeson, J. (1986). Attachment and the construction of relationships.

In W. Hartup & Z. Rubin (Eds.), Rl ' 11 hi v 1 n (pp. 51-71).
Hillsdale, NJ: Lawrence Erlbaum Associates.

Thompson. R.J., & Bernal, ME. (1982). Factors associated with parent labeling of

children referred for conduct problems. 152qu df Abndrmal Child Psyghdlogy, IQ,
191-202.

Wahl, G., Johnson, 8., Johansson, 8., & Martin, S. (1974). An operant analysis of
child-family interaction. Behavior Thgmy, 5, 64-78.

Wahler, R.G. (1969). Oppositional children: A quest for parental reinforcement
control. Joumg df Appligd Behavidr Analysis, 2, 159-170.

264

Wahler, R.G. (1980). The insular mother: Her problems in parent-child treatment.
WWW, L3. 207-219.

Wahler, R.G. (1982) Perspective on chapter 4. In G. Patterson, mm
was (pp. 82-83). Eugene, Oregon: Castilia Press.

Wahler, R.G., & Afton, AD. (1980). Attentional processes in insular and noninsular
mothers: Some differences in their summary reports about child problem behaviors.

End d Bahavidr lihggapy, 2, 25-41.

Wahler, W.G., & Dumas, J.E. (1986). "A chip off the old block": Some
interpersonal characteristics of coercive children across generations. In P. Strain, M.

Guralnick, & H. Walker (Eds.), hil r n’s ' B h v' r v1 11
Wu (pp. 49-91). Orlando: Academic Press.

Wahler, W. G. & Dumas, J. E. (1987). Family factors in childhood psychology:

Toward a coercion-neglect model. In T. Jacob (Ed.),Emily_in_tg_1'a9_d9__aad
MW: (pp 581-627) New York: Plenum

Press.