37:25“.

. 1.: ; .v. :
«Sim .wmwkmrfl .

. 2...»? r Y.

«3?...
.‘
$i+. .
. I .
.. . . .. .7
.. _ “.31.“. in“? . .
.fizéxév... . . . , 53 La
.2 . ‘ 3:. Fun?
2 .. .. . , .. ‘135 n V
4,. :fiafi. . ‘ , _ ‘ . wwwagm
Ti.
9 .
Emma
. : 5
. . . . . :4
“War. , ‘
:56

d. , x
. . , . . .3! 4.

. .HLJ: ‘ 1 ‘ , . ‘ . , . $5.5...“
d...” ._ . . V V . .. . . . 43..
9.1:“qu

v: ~13:

8......
~ by
. . ..
.3. 79...?
f

.5}? a
. 9...
..

 

 

 

 

 

«I 1.
, .3}. $
1.1.1.2.?

a . .11... .
‘ , . . . . ‘ urine}. «.151
, ‘ . , , . q ..s..\... )3"

 

THESIS

l 32

V?

C7

This is to certify that the
dissertation entitled

Parent Management Training (PMT) for
Children with Disruptive Behavior Disorders:
Motivational Aspects

presented by
Anne Marie Caruso

has been accepted towards fulfillment
of the requirements for

Ph .D . degree in School Psychology

 

Hm F (lemme.

gajor professor

 

Date 12/08/00

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

LEBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE

n21 Q

DEER

DATE DUE

DATE DUE

 

OCT17

5-wa- v

RU

205

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 chIRC/DatoDmpGS—p. 14

 

 

PARENT MANAGEMENT TRAINING (PMT) FOR
CHILDREN WITH DISRUPTIVE BEHAVIOR DISORDERS:
MOTIVATIONAL ASPECTS
BY

Anne Marie Caruso

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

DEPARTMENT OF COUNSELING, EDUCATIONAL PSYCHOLOGY, AND
SPECIAL EDUCATION

2000

ABSTRACT

PARENT MANAGEMENT TRAINING (PMT) FOR
CHILDREN WITH DISRUPTIVE BEHAVIOR DISORDERS:
MOTIVATIONAL ASPECTS

Current well-developed lines of theory and research on parent
management training for children with disruptive behavior disorders focus on
predictors of outcomes, especially as they apply to demographic and
environmental variables. In particular, a great deal is known about the following
traditionally examined predictors: single parenthood, low socio-economic status,
maternal depression, number of psychosocial stressors, and severity of the
child’s behavior problem. Much less is known about potential motivational
predictors in parent management training.

The current study investigated parent’s motivational aspects and
resources that might be included in a central schema in conjunction with
traditionally examined factors which predict outcomes in parent management
training. Featured motivational aspects included parent self efficacy, attributions,
achievement strivings, goal theory, implicit theories, and self schemas. These
motivational resources were examined within the context of
vaIue/interest/appreciation aspects and integrated into a cohesive model of
motivational schema through factor analysis of the Parent Motivational Schema
Scale. Outcomes of parent management training were compared across various

levels of traditional predictors, motivational schemata, and individual motivational

resources in a canonical correlation, discriminant functional analysis, logistic
regression, and series of chi-square analyses.

The results indicate that motivational measures can make significant
contributions and have future implications for parent management training
particularly in decisions regarding customized additional components in

treatment delivery for a particular parent (or subgroups of parents).

Copyright by
ANNE MARIE CARUSO
2000

To my parents (Matilde and Ernesto) and all other parents

ACKNOWLEDGEMENTS

I would like to thank Dr. Harvey Clarizio, dissertation chair, for his guidance,
direction, and support which provided the structure and organization for this
project. The dissertation committee in a collaborative effort, made invaluable
contributions in their areas of expertise. The following were most salient: Dr. Lois
Bader for her guidance in literacy analysis and aspects of the scale; Dr. Walter
Hapkiewicz for his direction in instructional approaches for treatment delivery;
and Dr. Michael Reckcase for his statistical expertise.

I would like to thank all of the other social scientists who have contributed in large
and small fashion. Dr. Jere Brophy was largely instrumental in the development
and especially the content validity analysis of the PMSS. Dr. Jayne Eyberg
graciously sent a copy of her Therapy Attitude Inventory and granted permission
for its use in the current study. The professors, both within Michigan State and in
other institutions have taught me that discovering new knowledge and
disseminating information “is what academia is all about”.

I would like to thank the management team and the board of directors at the
Group Health Centre organization who, collectively, provided the approval and
the practical means with which to carry out this project. Special thanks should go
to Robert Backstrom, Chief Executive Officer, who expressed interest and kept
abreast of the progress of this study. The physicians at the Group Health Centre
(particularly, Dr. Thomas Muhlstein and Dr. David Burrows, Paediatricians), the
psychologists and social workers at Algoma Child and Youth Services, ADHD
Clinic (particularly Dr. Lina Girard, Dr. Keith Lefave, and Leanne Murray), and the
many teachers from the Sault Ste Marie School Boards who referred various
clients and espoused the value of this project. This project truly felt like a
collaborative effort among agencies historically known for their territoriality.

I would like to thank the university undergraduate and graduate students who
helped manage the data. Michael lanni collected and input the data for the pilot
survey. Paul DeMarco input the data for the larger, main study.

I would like to thank my weekly squash partners (Eva Lidstone, Peter Schwartz,
and Scott McGiIIvray) who not only help keep me sane but whose pursuit of
excellence motivates me to do better.

I would like to thank “the girls" (Anna Magli-Smith, Jo Thorman, Nancy Barbon,
Donna Magli and Nancy Vardy) who are lifelong friends spread throughout the
globe but still close.

I would like to thank my family. Matilde Caruso (mother) and Santa DeMarco

(sister) gave me ongoing unconditional support in so many ways. Stan Caruso
(brother), took on the paternal father figure role early and continues to give

vi

immeasurable advice. Patricia Caruso (sister-in-law) acted as consultant and
advisor through many “adverse” situations. Orlando DeMarco (brother-in-law),
was always there to help out. My nephews and nieces (Paul and Patricia
DeMarco and Timothy and Sarah Caruso) who continually tell me how proud
they are of me and that I am their role model. It is their goals, dreams, and
accomplishments of which I am so proud. I thank this family whose love, caring,
entrepreneurial spirit and fortitude, motivate me to be better.

vii

Table of Contents

List of Tables .................................................................................... xi
List of Figures ................................................................................... xv
List of Abbreviations ........................................................................... xvi
Chapter 1 .............................................................................................................. 1
INTRODUCTION ............................................................................................... 1
Statement of the Problem .............................................................................. 1
Definitions ...................................................................................................... 4
Children with Disruptive Behavior Disorders .............................................. 4
Parent Management Training ................................................................... 13
Questions Leading to Hypotheses ............................................................... 16
Organization of the Dissertation ................................................................... 17
Chapter 2 ............................................................................................................ 19
LITERATURE REVIEW ................................................................................... 19
Universality Of Parent Management Training .............................................. 19
PMT as Prevention For Behavior Problems ............................................. 19
PMT As A Cross Categorical Treatment For All Dismptive Behaviors ..... 25
Predictors In Parent Training Interventions .................................................. 41
Traditionally Examined Outcomes And Predictors ................................... 41
Single Parent Families (SPF) ............................................................... 47
Socioeconomic Status (SES) ................................................................ 48
Parent Psychopathology ....................................................................... 49
Psychosocial Stressors ......................................................................... 50
Severity of Child Behavior .................................................................... 51
Other Predictors ................................................................................... 52
Failure to Find Consistent Predictors ....................................................... 52
Parent Cognitions As Predictors .................................................................. 59
Interactions Among Teps And Parent Cognitions ........................................ 63
Potential Motivational Predictors Of Outcomes In Pmt ................................ 65
Self Efficacy Theory ................................................................................. 70
Personal Control Beliefs ........................................................................... 78
Achievement Strivings .............................................................................. 82
Goal Theories ....................................................................................... 84
Implicit Theory .......................................................................................... 90
The Self .................................................................................................... 94
Integrating Inner Motivational Resources ................................................. 96
Pilot Study ..................................................................................................... 106
The Development Of The Parent Motivation Schema Scale ......................... 106
Method ....................................................................................................... 106

viii

Participants ............................................................................................ 1 06

Measure ................................................................................................. 107
Response Scale ..................................................................................... 110
Contributions Towards Validity And Reliability ....................................... 111
Phase 1: Establishing Item Content .................................................... 112
Phase 2: Administering The Scale .................................................... 113

Phase 3: Evaluating The Factor Structure Of The Scale .................... 113
Phase 4: Evaluating Reliability Of The Scale ...................................... 114
Results ....................................................................................................... 115
Discussion ................................................................................................. 127
Hypotheses ................................................................................................ 1 30
Chapter 3 .......................................................................................................... 133
METHODOLOGY .............................................................................................. 133
Participants ................................................................................................ 1 33
Procedures ................................................................................................ 1 34
Setting and Materials ................................................................................. 137
Data Collection .......................................................................................... 139
Traditionally Examined Predictors: ......................................................... 141
Parent Motivational Aspects/ Schema .................................................... 143
Outcomes ............................................................................................... 146
Child Behavior Changes ..................................................................... 147
Parent Satisfaction ............................................................................. 147
Parent Engagement ............................................................................ 148
Design and Analysis .................................................................................. 150
Analysis I: Treatment Effects .................................................................. 150
Analysis ll: Linear Relations Among Continuous Variables .................... 152
Analysis lll: Classifying According to Drop Out ...................................... 153
Analysis lV: Using Motivational Subscales Classify Drop Out ................ 154
Analysis V: Univariate Examination of Discrete Variables ...................... 154
Analysis Vl: Change in Parents’ Motivational Schema ........................... 155
Chapter 4 .......................................................................................................... 156
RESULTS ...................................................................................................... 156
Analysis I: Treatment Effects ..................................................................... 156
Analysis II: Linear Relations Among Continuous Variables ....................... 168
Analysis lll: Classifying According to Drop Out.............. ............................ 178
Analysis IV: Individual Motivational Resources .......................................... 192
Analysis V: Univariate Examination of Discrete Variables ......................... 196
Analysis Vl: Changes in Parent Motivational Schema ............................... 199
Chapter 5 .......................................................................................................... 202
DISCUSSION ................................................................................................ 202
Summary of Results .................................................................................. 202
Hypothesis I: Treatment Effects ............................................................. 202
Hypotheses II, III, and N (All Predictors and Outcomes) ....................... 207

ix

Hypothesis V (Individual Predictors) ...................................................... 210
Hypothesis VI (Changes in Motivational Aspects due to Participation) ..211

Further Comments ..................................................................................... 212
Limitations of the Present Study ................................................................ 215
Instrumentation ...................................................................................... 21 5
Sampling Issues ..................................................................................... 219
Random Assignment .............................................................................. 221
Developmental Considerations .............................................................. 222
Subgroups and/or Comorbidity Considerations ...................................... 223
Treatment Integrity ................................................................................. 223
Reactivity ................................................................................................ 224
Chapter 6 .......................................................................................................... 226
SUMMARY AND IMPLICATIONS ................................................................. 226
Implications for Theory .............................................................................. 226
Implications for Practice ............................................................................. 227
Parent Efficacy and Outcomes in Practice ............................................. 228
Parent Personal Control Beliefs and Practice ........................................ 230
Learned Helplessness in Practice .......................................................... 232

Goal Theory and PMT Practice .............................................................. 232

Self Schemas, Possible Selves, and Practice ........................................ 233

Value Aspects and Practice ................................................................... 234
Summary ................................................................................................... 235
Implications For Future Research .............................................................. 237
Final Thoughts ........................................................................................... 240
APPENDIX A .................................................................................................... 241
Parent Motivational Schema Scale (PMSS) .................................................. 241
APPENDIX B .................................................................................................... 245
Consent Forms .............................................................................................. 245
APPENDIX C .................................................................................................... 252
Decisions on Factor Dimensions ................................................................... 252
APPENDIX D .................................................................................................... 257
Logistic Regression Results Using Factor Scales in Lieu of Subscales ........ 257
APPENDIX E .................................................................................................... 261
Method Of Calculating Socio-Economic Status ............................................. 261
APPENDIX F .................................................................................................... 263
Power Analysis Results for T-Tests ............................................................... 263
BIBLIOGRAPHY ............................................................................................... 264

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

Table 10

Table 11

Table 12

Table 13

Table 14

Table 15

Table 16

Table 17

Table 18

LIST OF TABLES

The Individuals with Disabilities Act (IDEA) Definition Used
to Describe Children with Behavior Disorders ...................................... 5

The Council for Exceptional Children Definition Used

to Describe Children with Behavior Disorders ....................................... 7
Diagnostic Criteria for ADHD ............................................................ 8
Diagnostic Criteria for ODD ............................................................... 9
Diagnostic Criteria for CD ............................................................... 10
Select List of Well Known PMT Programs ............................................ 15

Summaries of DSM Nomenclatures and Prevailing

Theoretical Views of ADHD ............................................................... 28
A Comparison of Disruptive Behavior Disorders .................................... 42
Summary of Studies Examining Predictors in Parent Training ................. 52

Studies Examining Motivational Variables within PMT or
Similar Intervention Contexts ............................................................. 67

Explanatory Style Categorized Within the Three Attributional
Dimensions Regarding Child Misbehavior .......................................... 79

Hypothesized Parent Strategies for Responding to PMT as
Related to their Expectancy x Value Perceptions ................................. 97

Demographic Characteristics of Parent Participants in PMSS
Pilot Study (n = 200) and Treatment and Control Parents (n=93) ............ 108

Demographic Characteristics of Children Whose Parents
Completed the Pilot PMSS (n=293) ................................................... 109

Correlations among the motivation variables, statistical means,
standard deviations, and reliability analysis of motivational aspects
of the PMSS ................................................................................. 116

Eigenvalues, Percentages of Variance, and Cumulative Percentages
for Factors of the 44 - Item Parent Motivation Schema Scale (PMSS) ...... 118

Summary of Items and Factor Loadings of Varimax
Orthogonal Five-Factor Solution ......................................................... 119

Factor Loadings for Varimax Orthoganal F ive-Factor Solution .................. 121

xi

Table 19

Table 20

Table 21

Table 22

Table 23

Table 24

Table 25

Table 26

Table 27

Table 28

Table 29

Table 30

Table 31

Table 32

Table 33

Table 34

LIST OF TABLES

Descriptives for Varimax Orthoganal Five-Factor Solution on all
144 items of the PMSS (n=293) ......................................................... 123

Correlations among Motivation Measures pre- PMT and post-
PMT Program for Treatment (n=64 ) and Control (n=29) ........................ 124

Correlations among Motivation Measures pre- PMT and post-
PMT Program for all Participants (n=93) ............................................. 125

Correlations among Motivation Schema pre- PMT and post-

PMT Programfor Treatment (n=64 ) and Control (n=29) ........................ 126
Test-Retest Reliability Coefficients pre- PMT and post-

PMT Program for Treatment (n=64 ) and Control (n=29) ........................ 128
Demographic Characteristics of Participants in

PMT Program for Treatment (n=64 ) and Control (n=29) ........................ 135
Demographic Characteristics of Children Whose Parents

Participated in the PMT Study (n=93) ................................................ 136
Step by Step Procedures as Detailed in Defiant Children

(Barkley, 1997) .............................................................................. 140
Factor Loadings on the Therapy Attitude Inventory (TAI) ........................ 149
Summary of Specific Hypotheses (abbreviated),

Measures, and Analyses .................................................................. 151
Mean Participant Scores (Standard Deviations) on

Continuous Outcome Variables ......................................................... 157
Mean Change Scores (Standard Deviations) from

CBCL Measures ........................................................................... 164
Analysis of Variance Results for Main Effects and

Interaction Effects of Group, Time Point, and Scale on

Child Behavior Changes (CBCL) ...................................................... 165
Correlations and Standardized Canonical Coefficients

and Their Canonical Variates in Separate Models ................................. 169
Correlations and Standardized Canonical Coefficients

and Their Canonical Variates in a Combined Model .............................. 171
Canonical Analysis of Predictors and Outcome

Variables in PMT ........................................................................... 172

xii

Table 35

Table 36

Table 37

Table 38

Table 39

Table 40

Table 41

Table 42

Table 43

Table 44

Table 45

Table 46

Table 47

Table 48

Table 49

LIST OF TABLES

Correlations and Standardized Canonical Coefficients
Between Traditional and Motivational Predictors and
PMT Outcome Variables and Their Canonical Variates

in a Combined Model without SES ................................................... 174
Correlations, Standardized Canonical Coefficients and

Their Canonical Variates inTraditional and Motivational

Models ....................................................................................... 175
Correlations, Standardized Canonical Coefficients

and Their Canonical Variates in a Combined Model .............................. 176
Canonical Analysis of Predictors and Outcome

Variables in PMT ........................................................................... 177
Means and Standard Deviations of Predictor Variables

as a Function of PMT Drop Out Point ................................................ 179
Predictor Variables in Discriminant Function Analysis

by Traditional, Motivational, and Combined Models .............................. 181
Correlation of Predictor Variables with Discriminant Functions

(Function Structure Matrix) and Standardized Discriminant

Function Coefficients ..................................................................... 182
Correlations Between Discriminating Variables and Discriminant

Functions (Function Structure Matrix) ................................................ 183
Classification Analysis for Drop Out Using Traditional Predictors

Only ............................................................................................. 184
Classification Analysis for Drop Out Using Motivational Predictors

Only ............................................................................................. 188
Classification Analysis for Drop Out Using Both Traditional

and Motivational Predictors in a Combined Model ................................. 191
Logistic Regression Predicting PMT Drop Out .................................... 195
Prevalence (%) of Statistics Canada Income Cutoffs within

Points of Drop Out ........................................................................ 197
Prevalence (%) of Early Developmental Milestones within

Points of Drop Out ........................................................................ 197
Chi-Square Analyses of Participant Characteristics within

Points of Drop Out ........................................................................ 198

xiii

Table 50

Table 51

Table 52

Table 53

Table A1

Table A2

Table 0.1

Table C.2

Table C.3

Table 0.4

Table D1

Table 0.2

Table E.1

Table E.2

Table F.1

LIST OF TABLES

Mean Change Scores and T-test for Pre- and Post- PMSS
Subscales by Treatment (n=37) and Control
Group (n=29) ................................................................................. 200

Mean Change Scores and T-test for Pre- and Post- PMSS
Factors by Treatment (n=37) and Control
Group (n=29) ................................................................................. 201

Summary of Specific Hypotheses (abbreviated), Analyses,
and Decisions with Predictors ............................................................ 203

Salient Motivational Aspects and Value Interventions as

Potential SMEPT modules ............................................................... 236
Agreement Among Independent Raters ............................................. 242
PMSS Items Listed by Motivational Resource ....................................... 245
Descriptives for Varimax Orthoganal Nine-Factor Solution ..................... 256
Eigenvalues, Percentages of Variance, and Cummulative

Percentages for Nine Factor Solution ................................................ 257
Summary of Item and Factor Loadings for Nine-Factor Solution ............. 258
Descriptives for Non-orthogonal Five Factor Solution ........................... 259
Logistic Regression Predicting PMT Drop Out Using Factor Scores ......... 261
Correlation of Individual Items of the PMSS with Drop Out Outcome ......... 262
Hollingshead Two Factor Index ......................................................... 264
Low Income Cutoffs ........................................................................ 265
Power Analysis Results .................................................................. 266

xiv

Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10

Figure 11

LIST OF FIGURES

Scree Plot Yielding Five Factor Solution ............................................... 117
Pre- Post- Mean Scores of CBCL for Treatment and Control Groups ......... 158
Median CBCL Results Across All Scales .............................................. 159
Median CBCL Total Subscale Results ................................................ 160
Median CBCL Internalizing Subscale Results.........................................161
Median CBCL Externalizing Subscale Results ....................................... 162
Box Plot Depicting Interaction of Group * Scale Results ........................... 166
Bar Graph Depicting Interaction of Group * Scale Results ........................ 167
Territorial Map with Traditional Model .................................................. 186
Territorial Map with Motivational Model ................................................ 189
Territorial Map with Combined Model ................................................... 193

XV

LIST OF ABBREVIATIONS

SMePT ....................................................... Schema Mediated Parent Training
PMT .......................................................... Parent Management Training

ADHD ........................................................ Attention Deficit Hyperactivity Disorder
ODD .......................................................... Oppositional Defiant Disorder

CD ............................................................ Conduct Disorder

DSM-IV ...................................................... Diagnostic and Statistical Manual

of Mental Disorders - Fourth Edition

PMSS ........................................................ Parent Motivation Schema Scale
TEP ........................................................... Traditionally Examined Predictors
SE ............................................................ Self Efficacy

OES .......................................................... Optimistic Explanatory Style

PES ........................................................... Pessimistic Explanatory Style

Sch ........................................................... Schematicity

Imp ............................................................ Implicit Theory

PGD .......................................................... Performance Goals-Self Defeating
PGE .......................................................... Performance Goals-Self Enhancing
LG ............................................................ Learning Goals

Val ............................................................ Value

CBCL ......................................................... Child Behavior Checklist (4-18)
HSQ .......................................................... Home Situations Questionnaire
TAI ............................................................ Therapy Attitude Inventory

BDI-ll ......................................................... Beck Depression Inventory- 2nd Ed.
LICO ......................................................... Low Income Cutoffs

SPF ........................................................... Single Parent Family

LV ............................................................. Eactor1 (LG, Val)

IMC ........................................................... Factor 2 (SE, OES. Sch)

PGE .......................................................... Factor 3 (PGE, 1 PGD)

PGD .......................................................... Factor 4 (PGD)

IT .............................................................. Factor 5 (Imp)

xvi

Chapter 1
INTRODUCTION

Statement of the Problem

Parent behavior management training (PMT) is a widely studied
intervention for children with disruptive behavior disorders. Outcome studies
have demonstrated considerable success in reducing symptoms in children. As
well, studies suggest that parent management training (PMT) can also be viewed
as widely applicable. First, parent management training is an effective prevention
method for the development of disruptive behaviors (Adelman & Taylor, 1997;
Hinshaw, Klein, & Abikoff, 1998). Second, put within the context of the shifting
conceptionalizations of Attention Deficit Hyperactivity Disorder (ADHD) and the
development of comorbid disorders including Oppositional Defiant Disorder
(ODD), Conduct Disorder (CD), and possibly other disorders, it can be argued
that selection of specific PMT strategies should not hinge on a specific diagnosis
but on other factors. Furthermore, disruptive behaviors are found among children
with various diagnoses (including developmental disorders) and children who do
not meet criteria for any particular diagnosis (Johnston, 1996). Finally, the
majority of referrals for mental services are due to disruptive behaviors,
regardless of diagnostic status (Kazdin, 1987).

PMT therapies which are concerned with specific childhood outcomes
need not be isolated to specific childhood categorical labels but include all
disruptive behavior problems and address several additional (and possibly)

interacting predictors. Previous studies in parent training have recurrently

1

examined several predictors of successful outcomes in PMT, usually measured
as child behavior changes. Among the predictors examined are the following:
marital status (Holden, 1990; Webster-Stratton, 1985; Webster-Stratton &
Hammond, 1990), socio-economic status (Holden, 1990; Prinz & Miller, 1994;
Webster-Stratton, 1985; Webster-Stratton & Hammond, 1990), parent
psychopathology and/or depression (Frankel & Simmons, 1992; Webster-
Stratton, 1985; Webster-Stration & Hammond, 1990), severity of the child’s
problem (Ruma, 1996), and number of life stressors (Prinz & Miller, 1994;
Webster-Stratton & Hebert, 1993). These predictors are largely demographic and
environmental. Little research has been devoted to examining other predictors
and little research has explicitly addressed potential mediators in parent training
outcomes. This is where the challenge reportedly lies in research (Kazdin &
Weisz, 1998).

For behavioral outcomes, it seems reasonable to examine cognitions
related to the individual-environment interaction. Parent cognitions may be
organized in “schemas” which encompass a wide range of constructs. The notion
that parental beliefs and attitudes should be considered in selecting strategies to
teach effective parenting can be supported by bodies of research literature
specific to social science research in parenting and to educational research in
motivation.

First, research studies have been conducted to demonstrate relationships
between parent beliefs and attitudes and children’s behavior (Goodnow, 1995;

Holden, 1995; Lisi & Sigel, 1995). Surprisingly, parent management training

research which has examined specific predictors and childhood outcomes, has,
for the most part, ignored these potential cognitive mediational predictors, until
recently (Johnston, 1996).

Second, considerable research has examined the relationships among
motivational concepts and Ieaming outcomes (Brophy, 1999; Reeve, 1996).
Since PMT involves educational and learning components, considerable potential
exists for parent motivational aspects acting as mediators in parent training
outcomes. Within this second body of research, less emphasis has been given to
value aspects of motivation which imply that some relevance to personal
schemas must be perceived in order for Ieamers to benefit from the activity
(Brophy, 1998). This value aspect certainly would contribute significantly to
parent expectations regarding PMT and ultimately, outcomes.

The purpose of the present proposal was to examine parent’s motivational
resources, as individual constructs and as integrated motivation schemas as
mediators of predictors of outcomes in parent training. It was hypothesized that
these motivational aspects are mediators of traditionally examined predictors and
have implications for the predictive success of parent training outcomes.
Specifically, the present study investigated the relations between parent’s inner
motivational resources (self efficacy, attribution, goal, implicit, and self schema
theories) and value as related to motivation, and how these predict PMT
outcomes as measured by child behavior changes, parent satisfaction, and
parent involvement. It was expected that these findings would culminate in a

solid rationale and empirical basis for the development and inclusion of Schema

Mediated Parent Training (SMePT) components/modules into traditional PMTs.
Definitions

To better understand the contents of this dissertation, two central and
broad categories of terms must be defined. These include definitions for behavior
disorders as well as definitions for parent management training.

Children with Disruptive _Behavior Disorders

Many approaches have been used to classify children with behavior
disorders. Notably, the term, behavior disorders has been used to describe
children with both external (social, directed at others) and internal (emotional,
directed at self) problems (Hardman, Drew, & Egan, 1999). Children with
externalizing or disruptive behaviors gather the most attention from parents and
teachers and are the most commonly referred (Kazdin & Weisz, 1998). Children
with internalizing disorders (eg. depression, anxiety, and withdrawal) are more
likely to be overlooked (Kazdin & Weisz, 1998); yet, it is this latter group which
qualifies for special education services under the Individuals with Disabilities
Education Act (IDEA) (1997).

The IDEA describes children with behavior disorders under its Seriously
Emotionally Disturbed (SED) definition (Tablet). The current definition states that
SED is “a condition resulting in persistent and consistent maladaptive behavior
which interferes with the student’s Ieaming process” (Hard man, 1999, p. 219).

This federal definition excludes children with social maladjustment (such as

Table 1

The lndivfiuals with Disabilities Act (IDEA) definition y§ed to describe children
with behavior disorders.

 

“Seriously emotionally disturbed” is defined as . .:

1. a condition exhibiting one or more of the following characteristics over a
long period of time and to a marked degree, which adversely affects
educational performance:

a.

b.

c.
d.
e.

an inability to learn which cannot be explained by intellectual, sensory
or health factors;

an inability to build or maintain satisfactory relationships with peers and
teachers;

inappropriate types of behavior or feelings under normal circumstances;
a general pervasive mood of unhappiness or depression; or

a tendency to develop physical symptoms or fears associated with
personal or school problems

2. the term includes children who are schizophrenic or autistic.

3. the term does not include children who are socially maladjusted, unless it is
determined that they are seriously disturbed.

 

(adapted from Hardman, 1999 , p. 219)

Conduct Disorders below) from satisfying criteria for a Serious Emotional
Disturbance (SED) (Forness et al., 1996).

However, children with externalizing problems can also qualify under the
IDEA. For example, children with ADHD can qualify under PoHI and can also
qualify under Section 504 of the Rehabilitation Act of 1973 which can also lead to
special education or accommodations in general education.

The Council for Exceptional Children (CEC) has proposed a more
inclusive definition which has significant advantages over the IDEA definition
(Table 2). These advantages include a) recognizing impairments in adaptive
social, emotional, and behavioral areas; b) using multiple sources, considering
cultural and ethnic factors, and using normative standards; c) examining
prereferral and other interventions before labeling; and, d) potentially including
children labeled as socially maladjusted (Hardman et al., 1999).

The Diagnostic and Statistical Manual of Mental Disorders (4Th ed.; DSM-
IV; American Psychiatric Association, 1994), classifies disruptive behavior
disorders as Attention Deficit Hyperactivity Disorder (Table 3), Oppositional
Defiant Disorder (Table 4), and Conduct Disorders (Table 5). Attention
Deficit/Hyperactivity Disorders are characterized by the presence of six or more
criteria within two broad categories: a) inattention and b) hyperactivity/impuIsivity
(American Psychiatric Association, 1994). Oppositional Defiant Disorders are
characterized by the presence of at least four behaviors consistent with a pattern
of negativistic, defiant, disobedient, and hostile behavior towards authority figures

(APA, 1994). Conduct Disorders (CD) are characterized by the presence of

Table 2

The Council for Exceptional Children definition used to describe children with
behavior disorders.

 

Emotional or behavior disorder (EBD) refers to a condition in which behavioral
or emotional responses of an individual in school are so different from his/her
generally accepted age-appropriate, ethnic, or cultural norms that they
adversely affect educational performance in such areas as self-care, social
relationships, personal adjustment, academic progress, classroom behavior,
or work adjustment.

EBD is more than a transient, expected response to stressors in the child’s or
youth’s environment and would persist even with individualized interventions,
such as feedback to the individual, consultation with parents or families, and/or
modification of the educational environment.

The eligibility decision must be based on multiple sources of data about the
individual’s behvioral or emotional functioning. EBD must be exhibited in at
least two different settings at least one of which must be school related.

EBD can coexist with other handicapping conditions as defined elsewhere in
this law [IDEA].

This category may include children or youth with schizophrenia, affective
disorders, or with other sustained disturbances of conduct, attention, or
adjustment

 

(Adapted from Hardman, 1999, p. 220)

Table 3

Diagnostic criteria for Attention Deficit Hyperactivity Dmrder (ADHD).

 

A. Either (1) or (2):

(1) six (or more) of the following symptoms of inattention have persisted for at least 6
months to a degree that is maladaptive and inconsistent with developmental level.

Inattention

(a) often fails to give close attention to details or makes careless mistakes in
schoolwork, work, or other activities

(b) often has difficulty sustaining attention in tasks or play activities

(c) often does not seem to listen when spoken to directly

(d) often does not follow through on instructions and fails to finish schoolwork,
chores, or duties in the workplace (not due to oppositional behavior or failure to
understand instructions)

(e) often has difficulty organizing tasks and activities

(f) often avoids, dislikes, or is reluctant to engage in tasks that require sustained
mental effort (such as schoolwork or homework)

(9) often loses things necessary for tasks or activities (e.g. toys, school assignments,
pencils, books, or tools)

(h) is often easily distracted by extraneous stimuli

(i) is often forgetful in daily activities

(2) six (or more) of the following symptoms of hyperactivity-impulsivity have persisted
for at least 6 months to a degree that is maladaptive and inconsistent with
developmental level:

Hyperactivity
(a) often fidgets with hands or feet or squirms in seat
(b) often leaves seat in classroom or in other situations in which remaining seated is
expected
(c) often runs about or climbs excessively in situations in which it is inappropriate (in
adolescents or adults, may be limited to subjective feelings of restlessness)
(d) often has difficulty playing or engaging in leisure activities quietly
(e) is often “on the go” or often acts as if “driven by a motor"
(f) often talks excessively
Impulsivity
(9) often blurts out answers before questions have been completed
(h) often has difficulty awaiting turn
(i) often interrupts or intrudes on others (e.g. butts into conversations or games)

Onset before age 7

Impairment present in two or more settings

Clear evidence of impairment in social, academic, or occupational functioning
Symptoms do not occur exclusively during the course of a Pervasive Developmental
Disorder, Schizophrenia, or other Psychotic Disorder and are not better accounted for by
another mental disorder (e.g. Mood Disorder, Anxiety Disorder, Dissociative Disorder, or a
Personality Disorder)

Code based on type:

AttentionDeficit/Hyperactivity Disorder, Combined Type

Attention-Deficit/Hyperactivity Disorder, Predominantly lnattentive Type
Attention-Deficit/Hyperactivity Disorder, Predominantly Hyperactive-Impulsive Type

[119,093

 

(American Psychiatric Association, 1994, pp. 84-85).

Table 4

Djaqnostic criteriafir Oppositional Defi_ant Disorder (ODD).

 

A. A pattern of negativistic, hostile, and defiant behavior lasting at least 6

months, during which four (or more) of the following are present:

(1) often loses temper

(2) often argues with adults

(3) often actively defies or refuses to comply with “adults” requests or rules
(4) often deliberately annoys people

(5) often blames others for his or her mistakes or misbehavior

(6) is often touchy or easily annoyed by others

(7) is often angry or resentful

(8) is often spiteful or vindictive

Note: Consider a criteria met only if the behavior occurs more frequently than
is typically observed in individuals of comparable age and developmental
level.

B.

The disturbance in behavior causes clinically significant impairment in
social, academic, or occupational functioning.

The behaviors do not occur exclusively during the course of a Psychotic or
Mood Disorder.

. Criteria are not met for Conduct Disorder, and, if the individual is age 18

years or older, criteria are not met for Antisocial Personality Disorder.

 

(APA, 1994, pp. 93-94).

Table 5

Criterig for Conduct [lsorder (CDL

 

A. A repetitive and persistent pattern of behavior in which the basic rights others or
major age-appropriate societal norms or rules are violated, as manifested by the
presence of three (or more) of the following criteria in the past 12 months, with at
least one criteria present in the past 6 months:

Aggression to people and animals

(1) often bullies, threatens, or intimidates others

(2) often initiates fights

(3) has used a weapon that can cause serious physical harm to others (e.g. a bat,
brick, broken bottle, knife, gun)

(4) has been physically cruel to people

(5) has been physically cruel to animals

(6) has stolen while confronting a victim (e.g., mugging, purse, snatching, extortion,
armed robbery)

(7) has forced someone into sexual activity

Destruction of property

(8) has deliberately engaged in fire setting with the intention of causing serious
damage

(9) has deliberately destroyed others' property (other than by fire setting)

Deceitfulness or theft

(10) has broken into someone else’s house, building, or car

(11) often lies to obtain goods or favors or to avoid obligations (i.e., “cons” others)

(12) has stolen items of nontrivial value without confronting a victim (e.g.
shoplifting, but without breaking and entering; forgery)

Serious violation of rules

(13) often stays out at night despite parental prohibitions, beginning before age 13
years

(14) has run away from home overnight at least twice while living in parental or
parental surrogate home (or once without returning for a lengthy period)

(15) is often truant from school, beginning before age 13 years

8. The disturbance in behavior causes clinically significant impairment in social,
academic, or occupational functioning.

C. If the individual is age 18 or older, criteria are not met for Antisocial Personality
Disorder.

Specify type based on age at onset:

Childhood-Onset Type: onset of at least criterion characteristic prior to age 10

Adolescent-Onset Type: absence of any criterion characteristic prior to age 10

Specify severity:

Mild: few if any conduct problems in excess of those required for diagnosis and

conduct problems cause only minor harm to others

Moderate: number of conduct problems and effect on others intermediate between

“mild” and “severe”

Severe: many conduct problems in excess of those required to make the diagnosis

or conduct problems cause considerable harm to others

 

(APA, 1994. pp. 90-91)

10

three or more criteria within four broad categories: a) aggression to people and
animals, b) destruction of property, 0) deceitfulness or theft, and d) serious
violations of rules (APA, 1994) (Table 5). A central feature of conduct disorder is
a persistent pattern of behavior in which the rights of others and age-appropriate
social norms are violated (Kazdin, 1997, p. 162).

In the research literature, several terms are sometimes used
interchangeably with conduct disorders, all having to do with varying degrees of
conduct related problems. Juvenile delinquency, (a less inclusive, legal term) and
antisocial behavior are among those terms most closely associated with Conduct
Disorders. Violence is defined as aggressive acts causing serious harm, such as
aggravated assault, rape, robbery, and homicide (Loeber & Stouthamer-Loeber,
1998, p.242). Aggression refers to acts that inflict bodily harm or mental harm on
others and is confined to acts that cause less serious harm or mental harm on
others (Loeber & Stouthamer-Loeber, 1998, p. 242). Thus, aggressive acts are
viewed as less serious than violence which is an extreme form of antisocial or
delinquent behavior (T olan & Guerra, 1994, p.3). In examining the criteria for
conduct disorders, some children with conduct disorders may exhibit behaviors
consistent with antisocial, delinquent, violent, and/or aggressive acts and some
may not, depending upon the constellation of symptoms. For the purposes of this
paper, the distinctions among the disruptive behavior disorders and these other
acts/areas will be retained when required; however, results from studies

examining each of these concepts will be discussed.

11

While recognized in the DSM-IV (APA, 1994), Conduct Disorders (per se)
do not meet the federal guidelines under the IDEA as a disability condition
(Murray & Myers, 1998). In fact, little concordance has been reported between
DSM-IV diagnosis and eligibility for special education services under the
category of SED under the IDEA (Fomess et al., 1996). This may lead to some
compensatory strategies on the part of educational personnel which can be
deemed inappropriate and unethical. For example, school personnel may
misclassify children with CD as SED as the “path of least resistance” and
misplace them in special education self-contained classes for SED (Murray &
Myers, 1998). Children with CD tend not to do well in special education settings
and tend to be disruptive with children who are genuinely emotionally disturbed
(Murray 8. Myers, 1998). Furthermore, putting both disorders within the same
educational setting may exacerbate the problems since educational programs
and treatments for children with conduct disorders and serious emotional
disturbances are not the same (Murray & Myers, 1998).

Clarizio (1996) further differentiates this distinction by suggesting that
placement in special education programs and related interventions for SED is
contraindicated for the children in the late onset CD group. Another concern has
been voiced that children who are socially maladjusted have learned their
behavior patterns and should not be served under the SED definition which has
been interpreted to address only those behavioral disorders that have an
emotional causation (McIntyre & Forness, 1996). The definition of SED continues

to need further revisions (Murray & Myers, 1998) despite various previous

12

proposals and attempts towards potential changes (McIntyre, 1993; McIntyre &
Fomess, 1996; Sweeney, 1995). Since the IDEA definition for behavior disorders
continues to be criticized on these same grounds, i.e., for excluding children
described as socially maladjusted (Hardman et al., 1999), the focus of the current
paper will be on disruptive behavior disorders as defined by the DSM-IV.
Parent Management Trainina

Major treatments for children with behavior disorders address a variety of
general goals within a variety of approaches including the following: a) insight
oriented therapy (for relieving symptoms, treating causes of behavior, developing
therapeutic relationships, understanding unconscious causes of behavior; b) play
therapy (for developing therapeutic relationships, playing out emotional
problems, developing positive peer relationships, teaching social skills,
developing problem solving skills; 0) behavior therapy (for developing positive
peer relationships, teaching language, self-help, academic, social skills and
controlling disordered, unusual, aggressive and general overt behavior); d)
marital and family therapy (for treating causes of behavior, developing
therapeutic relationships, and developing problem solving skills; e) drug therapy
(for relieving symptoms controlling behaviors); and f) group psychotherapy (for
treating causes of behavior, developing therapeutic relationships, developing
positive peer relationships, developing problem solving skills, and understanding
unconscious causes of behavior (Hardman et al., 1999). As noted above, the
selection of the most efficacious approach may be dependent, in part, upon the

classification of behavior disorder.

13

Kazdin (1997) identified four psychosocial treatments which are reported
to have considerable promise in treating conduct disorders. Using carefully
selected criteria for evaluation, Parent Management Training (PMT) was
identified as among the most well developed and researched. PMT primarily
consists of training parents to alter their child’s behavior in the home (and other
settings) through changing coercive, unsupportive patterns of interchanges
between parent and child into more prosocial and reinforcing interchanges
(Kazdin, 1997). Using a more stringent criteria of treatments, Kazdin and Weisz
(1998), identified only three treatments in the efficacious category (i.e.,
established), again including PMT. Efficacious treatments have been shown to
be more effective than no treatment, placebo, and other treatments with findings
replicated by multiple and different researchers (Kazdin & Weisz, 1998). Brestan
and Eyberg (1998) describe PMT as having one of the strongest bases in
controlled outcome research for children with behavior disorders. Webster-
Stratton (1993) describes PMT as having the most effective and promising
results with children with ODD or CD.

A plethora of PMT programs exist and have been shown to be efficacious
in treating the disruptive behavior disorders (ADHD, ODD, and CD) as well as a
wide domain of child problems and ages (Table 6). These PMT programs can be
primarily classified as behavioral management programs or based on operant
learning principles. The behavioral PMT programs have several elements in
common as follows: a) a focus on accurate descriptions of child behavior; b) the

identification of antecedents and consequences; c) an emphasis on teaching

14

Table 6

Select List of Well Known PMT Proqrams

 

 

Author(s) Program Description or Model or
Treatment Manual Name Approach

Barkley (1997) Defiant Children behavioral

Bloomquist (1996) Skills Training for Children with Behavior combination

Dinkmeyer & McKay (1989)

Eyberg (1998)

Forgatch & Patterson (1989)
Forehand & Manon (1981)

Fox & Fox (1990)

Getz & Gunn (1988)
Gordon (1975)

Patterson & Forgatch (1987)

Phelan (1995)

Webster-Stratton (1981 )

Webster-Stratton (1996)

Wells (1993)

Behavior Disorders
Systematic Training for Effective
Parenting (STEP)

Parent-Child Interaction Therapy
(Special lnroads; U. of Florida)

Parents and adolescents living together
Helping the Non—Compliant Child

STAR Parenting Program
(stop, think, ask, respond)

Parent education
Parent Effectiveness Training (PET)

Parents and adolescents living together:
The Basics

1-2-3 Magic

BASIC
ADVANCE

treatment manual and training materials
with video tapes and discussion
questions

Parent Training Manual for Oppositional
Children

behavioral and
cognitive behavioral

Adlerian

behavioral

behavioral
behavioral

cognitive behavior
modification

family systems
Adlerian

behavioral

behavioral

behavioral

behavioral

behavioral

 

15

consistency; d) use of reinforcement techniques; and, e) assigned homework
tasks between sessions (Hutchings, 1996). Overall, PMT refers to a set of
interventions that target parenting skills critical to improving family interactions
and enhancing their children’s socializing skills (Frick, 1998). The term PMT
used throughout this proposal encompasses these behaviorally oriented
approaches except in instances where PMT approaches need to be clearly
differentiated.

Interestingly, it is reported that the parent advocacy/involvement
movement in children’s mental health (including PMT) has its roots in special
education (Bickman, Heflinger, Northrup, Sonnichsen, & Schilling, 1998).
Specifically, the IDEA explicitly includes parent participation in special education
as a due process issue (Bickman et al., 1998; Heflinger, Bickman, Northrup, &
Sonnichsen 1997).

Finally, the term “parents” is used throughout the proposal. It is recognized
that parents may include grandparents, step-parents, and other primary
caregivers or guardians.

Questions Leadinq to Hypotheses

Little or no attention had been focused on demonstrating (a) whether
parent motivational aspects influence outcomes in PMT; (b) whether parent
motivational aspects change as a result of PMT participation; (0) evaluating how
to maximize parent motivational aspects, and ((1) whether maximizing parent
motivational aspects will lead to better outcomes. The present study addressed

the first two questions as follows:

16

'__. we " “‘0'“

1. Do certain motivational resources, either individually or as integrated schema,
increase parent involvement or engagement; and, if so, does the involvement
lead to changes in the way parents benefit from PMT and consequently, lead
to better outcomes for the child and parent?

2. Are the levels of various parent motivational resources stable and consistent
constructs or do they change as a result of participating in parent
management training sessions?

The main hypothesis of the study is that traditionally examined predictors and

motivational aspects are predictive of positive changes in child behavior, parent

satisfaction, and parent engagement.

Orqgnization of the Dissertation

The first chapter introduces the study, defines the problem, and briefly
describes the purpose for the study. The concept of Schema Mediated Parent

Training (SMePT) is introduced and several key terms (disruptive behavior

disorders and parent management training) central to the arguments of the

dissertation, are defined. An argument is presented for the study of motivational
resources as predictors in parent management training outcome research.
Chapter 2 provides a comprehensive literature review and the main body
of the dissertation. Arguments for the wide use of parent behavior management
training are supported by research which 1) discusses parent training as a viable

prevention method for the development of disruptive behavior disorders; 2)

presents a rationale for why treatment may not necessarily need to be tailored to

specific, isolated, identified disorders; and, 3) outlines treatment efficacy studies

supporting parent management training for disruptive behavior disorders. A

detailed description is provided on traditionally examined predictors of PMT

outcomes, including marital status, socioeconomic status, parent

17

psychopathology, number of psychosocial stressors, and severity of the child’s
behavior disorder.

Chapter 2 also presents an argument for considering parent cognitions as
predictors in PMT outcomes. Specifically, research on the influence of student
inner motivational resources on academic achievement is postulated to directly
translate to the influence of parent inner motivational resources on parent training
“achievement”. A pilot study describes the development of an instrument, the
Parent Motivational Schema Scale (PMSS), which was specifically designed to
measure these inner motivational resources.

Chapter 3 describes the methodology, Chapter 4 presents the results, and
Chapter 5 discusses the findings of the study. This latter chapter also discusses
shortcomings and limitations regarding the research study. A summary
containing implications for theory, future research, and practice is presented in
Chapter 6. Based on findings of significant motivational predictors, Chapter 6
presents a detailed rationale towards the development of Schema Mediated
Parent Training (SMePT) and clearly articulates components which may be a

necessary and beneficial addition to PMT approaches.

18

Chapter 2
LITERATURE REVIEW
Universality Of Pgrent Manggement Training
The majority of PMT studies tend to concentrate on primary interventions
for full-blown disorders and isolate specific populations for their research.
Arguments can be made that parent management training may have some
efficacy with all behavior populations. In several replication studies, the Webster-
Stratton PMT approach has been shown to be efficacious with several different
populations as well as superior to wait-list control groups (Webster-Stratton,
1993). First, PMT is a viable prevention intervention for the development of
disruptive behavior disorders. Second, treatment need not necessarily be
directed towards specific homogeneous populations, especially because
treatment results may not necessarily be a result of the nature of children’s
disorders.
PMT as Prevention For Behavior Problems
Cowen and Hightower (1988) define prevention according to two broad
strategies: 1) “primary prevention... seeks to forestall the development of
pathology” and 2) “secondary prevention seeks to identify early signs of
dysfunction and to introduce prompt correctives to short circuit negative
outcomes” (Cowen & Hightower, 1988, p. 777). Within the context of juvenile
delinquency, Davidson and Redner (1988) consider primary prevention as

intervention before behavior is labeled while secondary or tertiary prevention as

19

intervention after formal labeling. It has been reported that schools have done
some work with early secondary prevention and even less with primary
prevention (Cowen & Hightower, 1988).

Many prevention programs seem to target populations at risk or treat
identified behavior disordered populations and sometimes neglect the direct
caregivers such as parents. Even when prevention programs make efficient use
of supportive personnel such as child-aides, they rarely employ the use of
parents as service providers. For example, the Primary Mental Health Project
uses child-aides for children in primary grades at risk for academic difficulties and
uses parent aides who function similar to paraprofessionals in keeping parents
informed (Cowen & Hightower, 1988). There is no mention of using parents as
child-aides. Other examples of prevention programs which do not overtly mention
parent involvement include programs designed for children with severe behavior
problems (Dyer & Larsson, 1997; Wolery & Winterling, 1997), aggression
(Matson & Duncan, 1997) and noncompliant and aggressive behavior in a case
study of a child, JB, who was at risk for a behavioral disorders (Umbreit & Blair,
1996). In fact, in an analysis of experimental studies on intervention for emotional
and behavior disorders from 1980 to 1993, intervention types were characterized
along 6 categories: skills training, self management, antecedent based,
consequence based, peer mediated, and other (Dunlap & Childs, 1996).
Nowhere in the article was parent training explicitly mentioned. These studies
are surprising in light of the fact that Kazdin (1987) has been quoted as writing

“no other intervention for antisocial children has been investigated as thoroughly

20

as parent management training and has shown as favorable results” (Strayhorn
et al., 1993).

In contrast, The Fast Track Multi-Site Demonstration Project directly
addresses the importance of parent involvement (Bierrnan et al., 1996). Their
Conduct Problems Prevention Research Group (CPPRG) advocates the need for
interventions including socializing agents (parents, peers, teachers, etc.) as a
secondary necessary condition in order to affect long term developmental
changes in children at risk for conduct disorder (Bierrnan et al., 1996 ). It should
be noted that the CPPRG view aggression and later conduct disorders as
multifaceted and multiply determined (Bierrnan et al., 1996 ). As a result, they
have designed a particularly extensive and intensive preventive intervention
which will likely be beyond the scope of most independent agencies unless
mechanisms are in place for school linked services or integrated services such
as those described by Adelman and Taylor (1997):

The call is for moving from fragmentation to coordinated/integrated

intervention and from narrowly focused, problem-specific, and

specialist-oriented services to comprehensive general approaches.
(Adelman & Taylor, 1997, p.408)

Similarly, comprehensive services may be made possible through
integrated services among community based mental health and social service
programs in a system of care defined as:

a community based approach to providing comprehensive,

integrated services that are available through multiple agencies and

professionals, in collaboration with families.
(Eber & Nelson, 1997).

21

Several current programs are emphasizing these comprehensive systems
of care wherein agencies share resources (Adelman & Taylor, 1997; Eber &
Nelson, 1997; Fomess et al., 1996). In addition to researching the most
comprehensive of treatments, it might be beneficial for researchers to investigate
the necessary and sufficient conditions, i.e., minimal services required to affect
change in outcomes of at-risk children.

Other studies have addressed parent training as an important intervention
for children at risk for developing conduct disorders. Kaiser and Hester (1997)
listed parent training as first among three components essential for prevention of
conduct disorders. However, they also note that most behavioral interventions
target populations already identified as behavior problems as opposed to children
at risk (Kaiser & Hester, 1997). Ramey et al. (1988) wrote about the Carolina
Early Intervention Program which not only directly involved parents but also
considered care giver characteristics (such as expectations, propensities,
development, supports, stressors, and interactions with the child) as important
factors in their conceptual model for early intervention. Forness et al. (1996)
recommended parent training as an initial intervention in their multistage
approach to prevention and intervention of emotional behavioral disorders. Many
current programs for children with SED have been criticized for emphasizing
comprehensive systems of care without considering prevention in service
delivery systems (McConaughy, Kay, & Fitzgerald, 1998).

Davidson and Redner (1988) developed their prevention training model for

at risk juvenile delinquents around a family focus. Eber and Nelson (1997) also

22

discuss a family focus as important for comprehensive and collaborative planning
and service delivery in their wraparound approach. Wraparound has been
described as “a needs-driven process for creating and providing services for
individual children and their families” (Eber et al., 1997). This latter approach
provides support for direct caregivers (whether teachers, parents, etc.) using an
Interactive team to empower members.

Price et al. (1989) note that not everything called a prevention program is,
in fact, a prevention program. This philosophy is consistent with previous
research suggesting that there have been no known studies in the long term
prevention of conduct disorder (Bierrnan et al., 1996). Long term studies are
necessary to evaluate if prevention interventions with children at risk for conduct
and other behavior disorders are beneficial. Studies advocate up to three to five
year follow ups (Kaiser & Hester, 1997) or even treatment up until children reach
secondary school age (Bierrnan et al., 1996). Some studies indicate that children
at risk for behavior problems have benefited from early intervention prevention
programs (Ramey et al., 1988; Strayhorn & Weidman, 1991 ). Other studies
suggest that, because there is such difficulty in accurately identifying high risk
groups, there is no evidence that prevention affects later juvenile delinquent
behavior (Davidson & Redner, 1988). Strayhorn and Weidman (1991)
demonstrated that a parent-child interaction, preventive mental health
intervention had beneficial long term effects. Specifically, the results of a one
year follow up indicated that improvements in classroom behavior were

correlated with improvements parents had shown in training.

23

Webster-Stratton (1998) evaluated the effectiveness of a parenting theory
based, multi-faceted, prevention intervention which was designed to prevent the
development of ODD and CD in at risk Head Start children. Results indicated
that children in the parenting program showed significantly more Improvements
(fewer CD problems, less ODD, less negative affect, more positive affect) than
children in the control condition who had received the regular Head Start
program.

Tucker, Gross, Fogg, Delaney, and Lapporte (1998) recently examined
the efficacy of a behavioral parent training (BPT) intervention as a prevention
strategy for young children at risk of developing CDs. One of the findings of the
study indicated that the amount of parent participation in the treatment was
correlated with greater gains in parent-child outcomes at one year. Similarly, in a
review of the literature on early intervention and prevention programs, Tucker
and Gross (1997) found that PMT significantly reduced negative child behaviors,
promoted positive parent child interactions, and that these treatment gains were
maintained over time.

The phrase has been coined that “parent training is prevention”.
Furthermore,

prevention activity creates a safe and productive environment and

develops the type of attitudes and capacities that students and their

families need to deal with violence and other threats to safety.
(Adelman & Taylor, 1997, p. 417).

Finally, PMT programs which view parents as equal partners in developing and

implementing interventions rather than helpers while mental health professionals

24

remain the “experts”, lead to more successful outcomes (McConaughy et al.,
1998).
PMT As A Cross Categmn'cal Treatment For All Disrgptive Behaviors

Discussion in this chapter focuses on categorical versus cross-categorical
approaches, and the state of the art with DSM-IV classification systems. Both
issues are used as arguments against the practice of rigidly defining target
populations (based on child diagnoses) for intervention research. Cross-
categorical approaches have several advantages over categorical approaches In
designing intervention studies. Also, DSM-IV classifications are far from reliable
and consistent. In addition, comorbidity becomes a factor.

Presently, a common educational practice of classifying children for
intervention services is through categorical or disorder based approaches
(Hardman et al., 1996). For purposes of intervention research, participants are
divided into discrete groups based on individual characteristics or disorders,
usually based on the Diagnostic and Statistical Manual of Mental Disorders -
Fourth Edition (DSM-IV). The selection of treatment based upon a categorical
approach is based on the assumption that children will respond differentially to
treatments contingent on their diagnosis. For example, it is assumed that children
with Oppositional Defiant Disorder have different characteristics and instructional
needs than children with Attention Deficit Disorders or Conduct Disorders.
According to a categorical approach, grouping each of these children under their
respective labels would be most functional in meeting their needs (Hardman et

al., 1996). Thus, the selection of target populations (usually homogenous groups)

25

for intervention research has the dual intent of maximizing treatment gains as
well as controlling for any extraneous variables.

Several other researchers have argued that categorical approaches are
neither necessary nor beneficial (Hardman et al., 1996). They argue that there is
no need to separate children by categorical labels, traditional categorizing
approaches result in unnecessary separation, and categorizing can be an
expensive label. Considerable debate exists among proponents of cross-
categorical approaches and categorical systems (Hardman et al., 1996). Cross-
categorical approaches do not use the traditional categories in defining and
classifying children with Ieaming and behavior disorders but do identify children
according to the level of severity of their difficulties (Hardman et al., 1996, p. 59).
When serious consideration is given to the parameters or state of the art in the
current classification system, DSM-IV, then a categorical approach may not be
the most functional for intervention research with disruptive behavior disorders.

An example of a disruptive behavior disorders category which may have
lost its utility as a research label is Attention Deficit Hyperactivity Disorder
(ADHD). The literature on Attention Deficit Hyperactivity Disorder is replete with
hard core research, interactive clinical research (actuarial science), popular
psychology, and anecdotal accounts. It should be noted that a historical
overview of attention deficit disorders indicates several diagnostic labels and
descriptions of children with behaviors consistent with ADHD prior to the
publication of the first DSM (Lerner & Lerner, 1991). As cited in Lerner and

Lerner (1991), these categories included Brain Damage Syndrome (Werner &

26

Straus, 1941; 1947) and Minimal Brain Dysfunction (Clements & Peters, 1962)
despite lack of evidence that these children had suffered brain damage. Even
earlier, near the end of the nineteenth century, Heinrich Hoffman, a German
physician and poet described children with ADHD characteristics as “Fidgety
Phils” and George Stillman, an English Physician, described these children as
defective in control of behavior and inhibition (Barkley, 1992). Despite these
descriptions, it was not until the second DSM in which ADHD was recognized
(Table 7). This historical summary supports the observation that the various
Diagnostic and Statistical Manuals of Mental Disorders (DSMs) demonstrate a
lag in research conceptualizations; furthermore, variables and factors other
than rigorous scientific experimentation are instrumental in decisions regarding
the creation and publication of these diagnostic manuals (Barkley, 1993;
Pelham, 1996; Tannock, 1998). As a result, it is posited that specific
intervention strategies should not be based on a generic diagnosis of ADHD.
With the publication of every Diagnostic and Statistical Manual of Mental
Disorders (DSM), changes to the Attention Deficit Hyperactivity Disorder
(ADHD) category have involved specific definitional criteria and nomenclature
(Table 7). In fact, there have occurred major changes regarding what the
criteria are at either a sub-symptom level or addressing the category of ADHD
as a whole. The debate has been about how best to weight the symptoms, how

best to combine them, and so on.

Most recently, diagnosis of ADHD changed drastically from DSM-lll-R to

DSM-IV. A number of changes had been recommended for inclusion in the final

27

Table 7

Summaries of DSM Nomenclatures and TheoreticalfViews of ADHD

 

DSM Nomenclature Proposed but
Rejected Nomenclature

Theoretical Era

 

DSM No recognition NA* None
DSM-ll Hyperkinetic Kyperkinetic Reaction Hyperactivity
(1968) Reaction of Childhood Motor
DSM-III ADD / + H Attention Deficit Disorder Attention
(1980) ADD / - H
DSM-lll- ADHD ID (Impulse Control) Motivational
(1987) U-ADHD HID (Hyperactive Impulse response to
Disorder) rewards
DSM-IV ADHD-IN EFD (Executive Function Self Regulation
(1994) ADHD-HI Disorder) Disinhibition
ADHD-Co Executive
Dysfunction
DSM-V NA SRD + IN Self Regulation
(2001) SRD + HI With lnattention
Self Regulation
with Hyperactivity-
lmpulsivity

 

28

draft of DSM-IV by the Committee on the Disruptive Behavior Disorders
(Barkley, 1993). Empirical field trials and expert consultations provided the
basis for the recommended changes (McBumett et al., 1993). Several studies
provide detailed descriptions of the field trial process for the DSM-IV
(McBumett et al., 1993; Power & DuPaul,1996). Symptom items were field
tested on both normal and clinic referred children at 10 or 11 sites across the
United States, depending on the source (McBumett et al., 1993 and Power &
DuPaul, 1996, respectively). Researchers used predictive power statistics to
determine diagnostic utility (Power & DuPaul, 1996). Experimental items which
did not demonstrate sufficient level of predictive power were excluded from the
diagnostic criteria. Other criteria which remained uncertain, required further

analyses from the field trial data set (Barkley, 1993).

The entire Child and Adolescent Categories Work Group and other APA
governance structures passed approval on the majority of recommendations by
the committee. Other additional diagnostic criteria were recommended by
committee members but did not pass approval and were not incorporated in the

ensuing DSM (Barkley, 1993). These included the following:

adjust DSM-lll-R cutoff for age
symptoms exceed 93 °/o of same age normal children
onset by 6 years of age (arguable)
duration of symptoms of 12 months
symptoms occur across many tasks and situations demanding
persistence
if IQ is below 85, use mental age norms in 1. above
(adapted from Barkley, 1993)

9959539

9’

29

One final recommendation involved re-instating the Attention Deficit Disorder
(ADD) category to describe those children previously labeled ADD without
Hyperactivity (DSM-Ill) or Undifferentiated ADD (DSM-lll-R).

The final published DSM-lV (1994) contains two clusters or lists of
symptoms to focus on: 1) lnattention symptoms and 2) lmpulsivity-Hyperactivity
Symptoms conceptualized as one cluster (Pelham, 1996; Tannock, 1998).
Several lnattention symptoms appear to be taken directly from the Connors
Rating Scale. Six symptoms from this list are part of the criteria for meeting
diagnosis of ADHD - predominantly lnattentive (ADHD-IN). Similarly, six
symptoms from the lmpulsivity-Hyperactivity Symptoms lists are required to
meet diagnosis of ADHD-predominantly Hyperactive-Impulsive (ADHD-HI).
Research has demonstrated that the impulsivity and hyperactivity tend to
cluster together providing an impetus to group the symptoms under one list in
the DSM-IV (Power & DuPaul, 1996). DSM-IV is the first time, Hyperactivity-
lmpulsivity is put in as a separate category. When factor analysis research is
done, symptoms relating to hyperactivity do not separate out from those
indicating impulsivity. Research from the DSM-IV field trials has shown that
impulsivity and hyperactivity load together to form a single behavioral
dimension (Applegate et al., 1997). All but one of the studies reviewed by
McBumett et al. (1993) found a two factor structure: inattention and
hyperactivity/impulsivity. Hinshaw (1994) in his review also supports the notion
of two stable dimensions: inattention/disorganization and

impulsivity/hyperactivity.

30

The DSM-IV conceptualizations of ADHD are based on the belief that
these two clusters of symptoms are distinct in terms of their etiology, in their
clinical course, prognosis, and response to treatment. Researchers argue that
this belief in distinct subtypes, is as of yet still unfounded by empirical data, but
is the very premise upon which the DSM-IV is based (Power & DuPaul, 1996;
Tannock, 1998). Furthermore, the DSM-Ill category of ADHD without
Hyperactivity cannot be conceptualized as the same as the DSM-IV category of
ADHD-Predominantly lnattentive type. The former incorporated symptoms
related to impulsivity while the latter does not. Currently, there is a
conceptualization or belief that three distinct types of children exist which can
all receive this global label of ADHD: ADHD-IN and ADHD-HI as mentioned
above as well as ADHD - Combined type (ADHD-CO). Other researchers argue
that differences in meeting the 7 year old age of onset criteria among subtypes
(ADHD-IN, 43%; ADHD-HI, 100%; & ADHD-CO, 18%) support the validity of
these three distinct types of children (Applegate et al., 1997). These latter
results, however, raise the question about the validity of the age of onset of

impairment criteria rather than an age of onset of symptoms criteria.

As a result of the somewhat arbitrary judgements and beliefs, the DSM-
IV also contains several additional criteria changes which have the potential to
pose considerable problems. First is the criteria that symptoms have to be
present before the age of 7. Tannock (1998) found a low reliability of recall
among parents of ADHD children. Only 33 % of parents were able to recall the

age of onset one year after initial clinic visit. However, Green et al. (1991) found

31

recall of age of onset of ADHD behavior by the mother as stable. While age of
onset is a historically based requirement, there has been little consistent
empirical evidence to support this criterion. First, age of onset is difficult to
establish reliably through retrospective studies and second, age of onset makes
little difference in the reliability of the diagnosis if it is not required (Barkley,
1993; Tannock, 1998). This age cutoff for the onset of ADHD symptoms appear
to be a continuing source of disagreement among researchers. Researchers
have previously argued to keep the diagnostic cutoff at age 7 (Green et al.,
1991), lowering the cutoff to age 6 (Barkley, 1993), and having no cutoff at all
(Power & DuPaul, 1996).

A developmentally appropriate cutoff criterion is proposed because a
marked decline in hyperactive-impulsive behaviors with age, has been found in
the general population (Barkley, 1993). Barkley (1993) argued that a fixed cut
off score is not logical in populations with declining baselines. This fixed
criterion is problematic because clinicians are likely to over diagnose preschool
children and under-diagnose adolescents and adults. From this perspective,
arguments that people outgrow the disorder can be viewed as artifacts of
having a fixed cutoff score in the diagnostic criteria. This issue has become
one of the criticisms of the DSM-IV, i.e., it fails to account for developmental

changes in ADHD presentation (Power & DuPaul, 1996).

Requiring the presence of symptoms in two or more settings (e.g. at
school, work, and/or home) presents tremendous difficulties for clinicians

(Power & DuPaul, 1996; Tannock, 1998). These may include problems with

32

how to operationalize this requirement in practice, how clinicians know how to
select which two settings and informants, and how clinicians decide what
constitutes a setting (e.g., are the school playground and school classroom two

different settings or do they constitute one setting?)

Setting and informant can easily become confounded in any diagnostic
decisions. Tannock (1998) found that there is major discrepancy or
disagreement among informants especially between parents and teachers. It
the teacher states one thing and parents state something else who is the more
credible source, particularly if inattention and hyperactive impulsivity are distinct
problems. Similarly, August and Garfinkel (1993) reported that requiring two or
more settings (home and school) or two corroborating sources may be too
restrictive. Not all researchers have arrived at the same conclusion. Biederman
et al. (1993) concluded that although correlation between parent and teacher
ratings were low to moderate, parent ratings were corroborated by teacher
reports using “broad” teacher diagnoses of ADHD. The DSM-IV stipulates that
problems in impairment must be present in social and/or academic settings in
order to meet criteria (Power & DuPaul, 1996). This leaves room for
considerable interpretation. In fact, many clinicians report paying homage to the
DSMs, following the criteria to some degree, and correcting for errors and doing
an even better job by adding their own idiosyncratic criteria (Applegate et al.,
1997). Barkley (1997) recommends that clinicians locate evidence that the
symptoms existed in several settings but not insisting on agreement between

parents and teachers in order to diagnosis. These elements of clinical

33

judgement create a potentially unreliable diagnostic process (Power & DuPaul,
1996). According to the DSM-IV, the prevalence of ADHD is estimated at 3% to
5% among school age children (APA, 1994). According to its predecessor, the
DSM-lll-R, the prevalence of ADHD was described as common and as
occurring in as many as 3% of school age children (APA, 1987). Surprisingly,
this increase suggests that the prevalence of children diagnosed with ADHD is
higher when requirements stipulate two settings. However, the criteria for
ADHD also stipulated the presence of 8 symptoms as opposed to 6 in the

DSM-lll-R and DSM-IV, respectively.

The field trial researchers also recommended many more symptoms for
the DSM-IV but symptoms were dropped due to nonscientific rationale. An
anecdotal account revealed that one of the higher ups in the APA governing
body decided that there were too many symptoms and wanted some dropped
(Barkley, 1993; Pelham, 1996). For political reasons, the APA governing body
chose two smaller symptom lists, one with hyperactive and impulsive symptoms
congregated together and one with inattention symptoms together and reduced
the list with the claim that clinicians would not be able to remember the larger
lists (Barkley, 1993). The changes appear scientifically questionable since
shortening an item list tends to weaken reliability which in turn weakens the
validity of the diagnostic criteria. The Committee had originally planned that
children who met criteria on the Hyperactive Impulsive would be called ADHD
and children who met criteria on the lnattentive list would be ADD. The DSM-IV

governance bodies changed this recommendation so that all children who meet

34

any criteria are labeled as ADHD but with one of three trailing clauses:
Predominantly lnattentive, Predominantly Hyperactive-Impulsive, or
Predominantly Combined Type. Barkley (1993) viewed this change as
regressive scientifically in that it was a throw back to DSM-Ill where ADD and
ADHD were viewed as sub-types of a common problem, i.e., they were viewed
as the same disorder. Barkley (1993) did not view ADD as a disruptive behavior
disorder and did not even want to include ADD in the Disruptive behavior
disorders section of the DSM. Thus, according to Barkley (1993) DSM-IV
obscured where it could have clarified. Certainly, given an analysis of some of
the field trial studies specifically regarding symptoms (Applegate et al., 1997), it
becomes apparent that research may not be the determining factor in devising

diagnostic manuals.

Other researchers have found evidence consistent with the notion that
ADD and ADHD are separate disorders (August & Garfinkel, 1993). Goodyear
and Hynd (1992) in an extensive review of the literature suggested that
ADHD/H and ADHD/ WO present with distinct behavioral and
neuropsychological differences that distinguish them in clinical presentation,
etiology, and possibly intervention and that the two groups may constitute two
separate disorders. Later in the same article, they qualified this suggestion and
concluded that their review does not permit a categorical answer to the
question of the existence of ADD subtypes. Goodman and Poillion (1992)

synthesized 48 articles and books to examine the agreement regarding

35

characteristics and cause of ADHD. They found 69 characteristics and 38

causes.

Consistent with the notion of separate disorders are several growing
clinical beliefs which are unsubstantiated by conclusive empirical data. One
belief is that the risk factor lies with the Hyperactive-Impulsive Type and that
lnattentive types are less severely impaired (Tannock, 1998). The implications
are severe; if clinicians believe that lnattentive type children have less severe
problems, then it is likely that the lnattentive types will receive less treatment or
intervention (Tannock, 1998). Other related general beliefs are that 1)
lnattentive type are more likely to have more anxiety, depressive, and reading
disorders and 2) The hyperactive-impulsive type are more likely to have more
oppositional defiant disorder, conduct disorder related problems. Tannock
(1998) found that hyperactive impulsive children are more likely to have

conduct disorder but oppositional defiant disorder did not cluster together.

Morgan et al. (1996) found that ADHD/Combined type children are
perceived by their parents to display significantly more externalizing,
delinquent, and aggressive behavior than ADHD/lnattentive type. Findings
from studies suggest that ADHD/Hyperactive-lmpulsive types are at greater risk
for antisocial behavior but no longitudinal studies have been conducted to date
(Power & DuPaul, 1996). One earlier research study identified three different
subtypes of ADHD: ADD without hyperactivity; ADD with hyperactivity; and
ADD with hyperactivity and aggression, ADDHA, (Dykman & Ackerman, 1993).

Other research found no significant difference between ADHD subtypes and

36

rate of affective or anxiety disorder co-diagnosis (Morgan et al., 1996). Other
subtypes have been proposed in the literature including ADHD with and without
aggression, and ADHD with learning disabilities (Power & DuPaul, 1996).
However, ADHD with and without hyperactivity children have similar rates of
Ieaming disabilities when reading and math disabilities are grouped together
(Barkley, 1990; Morgan et al., 1996). The latest ICD also classifies a

hyperkinetic-conduct disorder subtype.

Power and DuPaul (1996) recommend continuing to treat these
subtypes as co-morbid disorders rather than distinct subtypes. For a
nomenclature to be effective, common characteristics must be attributed to the
system to be known and agreed upon by all who use the system. Thus, the
entire diagnostic and nomenclature issue of ADHD is not only complicated by
the heterogeneity of ADHD but also by the prevalence of comorbidity. This has
huge implications for future research directions but especially for current

diagnosis and treatment.

Currently, there are no empirical studies which examine ADHD subtypes
and differential response to various treatments including pharmacological
intervention, contingency management, and modification to academic
instruction (Power & DuPaul, 1996). Yet researchers have made

recommendations based on subtypes:

In terms of intervention, the findings indicate that with combined type
children, it is important to identify and treat aggressive, oppositional, or
delinquent behavior and/or to take a preventative approach (Morgan et
al.1996)

37

Power and DuPaul (1996) suggested that the DSM-IV diagnostic criteria
should be useful for predicting effective lists of treatment for children diagnosed
with ADHD. Recommended treatments for ADHD include parent training, family
therapy, adaptation in school environment, pharmaceutical, behavior
intervention in the classroom and playground, and involvement in parent
support groups (Power & DuPaul, 1996). However, the utility of diagnosis for
treatment is very general and therefore limited in being helpful for designing
behavioral and instructional intervention (Power & DuPaul, 1996). Goodman
and Poillion, (1992) also held the view that diagnosis has limited value for

treatment because of the uncertainty of the nature of ADHD.

Pelham (1996) also presents a strong argument for why clinicians should
place less emphasis on diagnosis for attention deficit disorders and
nomenclature. Because the criteria change all of the time, practicing clinicians
should not be overly concerned or devote inordinate amounts of time to
diagnosis (Pelham, 1996). It is probable that the next DSM (either DSM-lV-R or
DSM-V) will have a different list of symptoms, or different way of grouping
them, and a different way of conceptualizing the disorder. Throughout the
years, they have all been the same children, they all have impairment, and they
have all had problems that required treatment. It would be a rare clinician who
assessed children, found them just short of meeting the number of symptoms,
and sent them on their way. Regardless of whether the child needs 12
symptoms and only meets 10 for criteria, he still has a problem which likely

requires some form of intervention. As a result, Pelham (1996) recommends

38

establishing impairment and designing treatment based on impairment.
However, any decisions regarding impairment also require a cut-off score.
Barkley (1993) argues that how professionals conceptualize the disorder will
affect how clients will be treated. This is problematic considering that ADHD

has been conceptualized from various perspectives for decades.

Given the lack of consensus in the field, treatments cannot be tailored to
ADHD. Thus, an argument can be made that the ensuring recommendations for
multi-modal treatments for children with ADHD can be viewed as a blitz

treatment to cover all contingencies.

The anecdotal accounts from various sources (Barkley, 1994; Pelham,
1997; Tannock, 1998) suggest that specific criteria that are a part of diagnosis
in the DSMs vary tremendously according to many different dimensions or
parameters of which research may be a relatively small part. Basically, a
treatment committee put together the diagnostic criteria and governing bodies
made the final decisions. It was originally predicted that using the DSM-Field
Trial data would improve the reliability, internal consistency, and clarify the
association of ADHD with functional impairment (McBumett et al., 1993).
Regarding ADHD in particular, the DSM continues to be criticized for being far
from fully reliable, specific, or valid (Hinshaw, 1994). “The clinical field has long
grappled with issues of classification, diagnosis, underlying patho-physiology,
associated difficulties, and treatment of youth with this syndrome" (McBumett et
al., 1993). Given the shifting nomenclature, diagnostic criteria, and theoretical

models of ADHD presented above, one can ask the question, how can

39

clinicians decide on effective treatment based on a diagnosis. Inter-clinician
agreement on what the problem is constitutes a prerequisite to linking diagnosis

and treatment.

This categorical — treatment issue with ADHD is further complicated by
the prevalence of comorbidity with other disruptive behavior disorders as well
as other DSM-IV disorders. The research literature is converging to the view
that Attention Deficit Hyperactivity Disorder is likely a neurodevelopmentally
disorder (Barkley, 1993; 1995, 1997). However, the developmental progression
of ADHD into other disruptive behaviors disorders has been viewed within a
behaviorist model. Large percentages of children with ADHD (35 to 60% )
develop Oppositional Defiant Disorder (ODD) by 7 years of age and a large
percentage (30 to 50%) will eventually meet criteria for Conduct Disorders (CD)
(Whalen & Henker, 1998). Predictors of development into ODD behaviors were
found to be parental consistency and predictability. Predictors of development
into CD were related to parental monitoring and tracking. These predictors are
related to parent characteristics. It is also noted that a substantial percentage
(15 to 25%) eventually develop antisocial personality into adulthood (Whalen &
Henker, 1998).

Finally, it has been reported that the nature or type of children’s
disruptive behavior disorders is not predictive of parent training outcomes
(Ruma et al., 1996). Rather the severity of the disorder is a more reliable
predictor. One of the parameters of cross categorical approaches, characterize

children according to the severity of their condition (e.g., mild, moderate, or

40

severe). To objectively evaluate the proposition that PMT is suitable to all
disruptive disorders, Table 8 summarizes similarities on various dimensions
among children with ADHD, CD, and ODD, as well as differences
(Anastopoulos, Barkley, & Sheldon, 1996; Barkley, 1998; Clarizio, 1996; Frick,
1998; Hinshaw & Anderson, 1996; Kazdin, 1996; Marx & Gross, 1998;
McMahon & Wells, 1998; Whalen & Henker, 1998).

Brestan and Eyberg (1998) describe PMT as having one of the strongest
bases in controlled outcome research for children with behavior disorders.
Furthermore, research has documented PMT’s efficacy across a wide age
range of children with various kinds of behavior problems (Brestan & Eyberg,
1998). Thus, these findings provide support for the view that cross-categorical
approaches can be used in PMT intervention research.

PREDICTORS IN PARENT TRAINING INTERVENTIONS

 

Predictors of outcomes in PMT can be grouped according to two broad
categories which may not necessarily be mutually exclusive: 1) traditionally
examined predictors and, more recently, 2) parent cognitions. The first include
primarily demographic types of predictors as well as child characteristics. The
second, parent cognitions, may include attitudes, beliefs, and perspectives as
well as inner motivational resources.

Traditionally Examined Outcomes And Predictors

 

Outcome research has consistently confirmed the efficacy of parent
training as a treatment for changing behavior problems in children (Ruma et al.,

1996). Considerable research has also been conducted throughout the years

41

Table 8

A Comparison of Disruptive Behavior Di_sorders

 

Behavior Disorder

 

 

 

Dimension AD/HD ODD CD
CARDINAL SYPTOMS

Core hyperactive-impulsive defiant aggressive
(disinhibition) irritable antisocial
inattention oppositional infliction of pain
severity fluctuates initiating fights

according to context denial of others' rights
persistent not transient
Number of
Symptoms 6/18 4/8 3/ 1 5
EPIDEMIOLOGICAL

Subtypes Predominantly lnattentive none Childhood-Onset
Predominantly Hyperactive- Adolescent-Onset
Impulsive
Predominantly- Combined

Prevalence 2 % to 6.3 % 5 to 7% 6 to 16% males

2 to 16% 2 to 9 % females

Age of Onset <3years, PHI before 8 years as early as 5 to 6 years
5 to 7 years, Pl usually childhood or adoles.

Sex ratios 3:1 community samples aprox. 1:1 2:1 to 4:1

(male to female)

6:1 to 9:1 clinic samples

 

ASSOCIATED CHARACTERISTICS

Social Problems yes yes yes
Medical Risks yes na na
Conscience Delays no no yes, deficient .
(moral reasoning) normal lack empathy, guilt
Voluntary control no yes yes

over behavior uninhibited, minimal control calculated, planful
Attachment relations typically poor typically poor poor quality
Academic Problems yes yes yes

 

42

Con't Table 8

 

ETIOLOGIES
Heredity Influence 10 to 35% no yes
Neurobehavioral yes negative yes
Patterns strong evidence temperament (strong with early onset)
prefrontal lobes cognitive
characteristics
Dietary Contributions unsubstantiated na na
Environmental Toxinsyes no no
Psychosocial lnfluencyes (on severity) yes (quality of yes (strong with adolescent
Contextual parent-child onset)
interactions)
DEVELOPMENTAL TRAJECTORIES/ PATHWAYS
age change in yes no yes
cardinal symptoms maximal minimal small
influence on pathwaysubtype parent subtype

consistency parent monitoring & tracking

 

 

parent influence on minimal consistency monitoring
symptom progress compensatory strategies predictability tracking
re: symptoms indiscriminant
parenting
characterization heterotypic continuity highly stable heterotypic continuity
INTERVENTIONS
Punishment Need responsive to external need for changeeward dominant learning
in Treatment controls in coercive style
interactions
PMT efficicacious efficacious efficacious
Pharmacotherapy methylphenidate no specific no specific pharma-
antidepressants pharmaco- cotherapy for core
therapy for clonidine for aggression
core lithium for aggression and
symptoms explosiveness
Cognitive-Problem yes na yes
Solving promising efficacious
Choice of target self control compliance covert (stealing, lying,
behaviors self regulation firesetting)

overt (bullying, compliance)

 

43

examining predictors of successful and unsuccessful outcomes in parent
training programs with various disruptive behavior populations. The outcomes
have included a broad range of measures across studies. The predictors have
primarily included parent depression, marital status and adjustment, social
class, number of negative life experiences, maternal insularity, age, and
severity of disorders, (Ruma et al., 1996; Webster-Stratton & Hammond,
1990). There has been considerable interest in how and why some parent
training programs work for some parents and not others. Several years ago,
Webster-Stratton (1985) demonstrated that depending on definition of
treatment outcome criterion, different clusters of predictors of treatment will
have varying success or failures. These treatment outcome criteria can be
grouped into three broad categories: a) changes in child behavior; b) parent
engagement; and, c) parent satisfaction.

Many studies have measured childhood outcomes as perceived by
parents on rating scales. Some studies have examined outcomes of child
behavior through direct observation in home and school. Other studies have
examined parental outcomes and changes in parent’s behavior. One study
reported changes in parent attitude; however, examination of the study revealed
that parental attitude was measured through having the parent rate the child’s
behavior on the Child Behavior Checklist (CBCL) (Webster-Stratton, 1990). The
CBCL appears to have set the standard as an outcome measure of changes in

child behavior.

44

A second category of outcome measures involves parent satisfaction.
Consumer satisfaction has become an essential component of outcome
evaluation of psychological treatments (including parent training), due to the
recent changes in mental health delivery, funding, and emphasis on treatment
effectiveness (Brestan, Jacobs, Rayfield, & Eyberg, 1999) as well as managed
care organization preference (Bilbrey & Bilbrey, 1995, July/August).
Furthermore, (Brestan et al., 1999) view standardized measures of treatment
satisfaction with child treatments, as a priority. Parent satisfaction would be
defined as the extent to which parents liked the PMT process and outcomes
which include the format, techniques, and effects of treatment (Brestan et al.,
1999). Webster-Stratton (1993) reports that Patterson's PMT treatment
approach, Forehand and Mahon’s approach, and Webster-Stratton’s programs
have been extensively evaluated and all have had high reports of parent
satisfaction.

Outcomes have also been assessed through examination of parental
engagement. Several indexes of parental engagement have included premature
drop out and behavioral signs in training which may include sporadic attendance,
missed and late appointments, last minute cancellations, skipped sessions
without cancellation, and failure to complete homework (Prinz & Miller, 1996).
Other signs of inadequate parental engagement include quality of participation in
sessions such as low level of interaction, complaints or arguments, giving up on

strategies after one attempt, and refusing to role play (Prinz & Miller, 1996).

45

Holden et al. (1990) describe three related criteria in assessing parent
engagement outcomes in parent training programs: 1)the proportion of clients
who successfully complete training; 2) the magnitude of client’s behavioral
change; and, 3) the length of time needed to bring about change. Previous
researchers have described outcomes according to two levels: 1) those clients
who terminate treatment prematurely and drop out; and, 2) those clients who stay
in treatment but fail to demonstrate positive change (Webster-Stratton, 1985).

Holden et al. (1990) break these levels down further in their description of
a “continuum of effectiveness”. At one end, certain clients may progress rapidly
and show improvement (Holden et al., 1990). In the middle of the continuum,
clients may experience difficulty in attaining treatment goals. They complete
training and take longer to demonstrate progress but only after training has been
completed (Holden et al., 1990). It is probable that these middle clients would
present as clients at the opposite end of the continuum unless long term
outcomes are evaluated. At this opposite end, clients may drop out from training
and for a variety of reasons are least likely to benefit from training.

Several studies have noted significant changes in parent report measures
but nonsignificant changes in child behavior data (Webster-Stratton, 1985).
Several studies have also reported that immediate post treatment gains were not
maintained on follow up (Webster-Stratton, 1990). Kazdin (1997) summarized the
conclusions of several outcome studies as follows: 1) parent management
training outcomes as measured by parent and teacher reports of deviant

behavior, direct home and school observations, and institutional records (e.g.,

46

truancy, police contact, arrests) indicate significant improvement; 2) the
magnitude of changes place children with conduct disordered within normal
ranges of behavior in comparison to non-referred same sex, age peers; 3) long
term treatment gains from 1 to 3 years have been maintained; 4) studies include
some sparse data on “incidental improvements” such as in marital satisfaction,
maternal psychopathology (depression), family cohesion, and sibling behavior
which were not a direct focus of training.

The most commonly studied predictors of parent training outcomes,
henceforth called traditionally examined predictors (TEPs), include single parent
families (SPFs), socioeconomic status (SES), parent psychopathology
(especially, maternal depression), number of psychosocial stressors, and
severity of the child’s behavior problem.

Single Parent Families (SPF)

Treatment failure has also been related to marital adjustment (Webster-
Stratton & Hammond, 1990). Specifically, single parent families are more likely
to be non-responsive to treatment or drop out of treatment (Webster-Stratton,
1985; Wester-Stratton & Hammond, 1990). In contrast, intact families are more
successful in maintaining treatment effects over time (Webster-Stratton &
Hammond, 1990). No mention was made as to the nature of the marital
relationship and whether or not this would have an impact in parent training.

The Webster-Stratton and Hammond (1990) study examined multiple
predictor measures with a large sample of parents at post treatment and at one

year follow up. They found that marital status and SES were the strongest

47

predictors of maternal behavior with their children post-treatment while marital
status alone was the strongest predictor of paternal behavior (Webster-Stratton &
Hammond, 1990). In contrast, Holden et al. (1990) found that marital status was
not found to be predictive of parent training outcomes. In their case, outcome
was defined as completion of the program.

Marital status appears to be a variable worthy of examination. Single-
parent families are a growing segment of the US. population (25%) and
Canadian population (20%) and usually headed by mothers (Statistics Canada,
1992; US. Bureau of the Census, 1994).

Socioeconomic Status (SE8)

Socioeconomic disadvantage has been defined by low income, low
education, single parent status, poor area of residence, large number of children
in the family, and referral to the clinic by outside agency (Webster-Stratton,
1985; Wester-Stratton & Hammond, 1990). Family SES has been consistently
reported to be related to parent training outcome. Low SES parents were more
likely to drop out of parent training programs (Kazdin & Wassell, 1999; Webster-
Stratton & Hammond, 1990) or do more poorly in parent training in that they were
generally less successful in making positive behavioral changes (Holden et al.,
1990; Kazdin & Wassell, 1999). These results were found at both immediate
post-treatment and long term post treatment. An earlier summary of the research
literature indicated that SES was an important predictor for outcome criteria, i.e.,
the more disadvantaged the family (single, low income, and low education), the

less likely it was to benefit from parent training (Webster-Stratton, 1985).

48

On the other hand, advantaged SES can also be viewed as predictive of
poor outcomes in terms of missed appointments and dropouts. Financially secure
families may be overly committed to individual activities (e.g. volunteering or
church) and recreational activities (sports or clubs) (Prinz & Miller, 1996).
Engagement in these other activities can lead to inflexible or infrequent
scheduling of sessions (Prinz & Miller, 1996). Also, families with extensive other
activities may have disorganized lifestyles which would make it difficult to
practice new strategies learned in parent training (Prinz & Miller, 1996).

Parent Psychopathology

Studies have suggested that pretreatment levels of maternal depression
were significantly related to treatment failure or relapses (Kazdin & Wassell,
1999; Webster-Stratton & Hammond, 1990). Webster-Stratton and Hammond
(1990) also found that paternal depression was just as predictive of reports on
child adjustment.

Not all studies have found maternal depression to be a predictor of parent
training outcomes. One study found that maternal depression as measured by
the Beck Depression Inventory, was not a predictor in parent training outcomes
(Webster-Stratton, 1985). It was hypothesized that this may have been the result
of maternal depression decreasing significantly from pre- to post-treatment
assessment. Similarly, Frankel and Simmons (1992) did not find evidence to
support the notion that maternal depression is a predictor of parent training
outcomes. They found that the only parent personality variable which was related

to parent training outcomes was psychotic distortions. However, it was noted that

49

parents endorsed scale items which reflected helplessness and negativity rather
than psychotic processes (Frankel & Simmons, 1992). Helplessness and
negativity are also consistent with depressive affect. These parents tended to
drop out of treatment (Frankel & Simmons, 1992).
Psychosocial Stressors

Ecological variables have been found to have a significant contribution to
parent training outcomes. For example, stressful life experiences make it difficult
to find time to do assignments, participate in parent training, and follow through
with suggestions (Kazdin & Wassell, 1999; Webster-Stratton & Herbert, 1993).
Also, the degree of life crisis and environmental stresses would be significantly
related to a family's ability to maintain treatment effects (Webster-Stratton, 1985).

Situational demands and constraints have an influence on behavioral
outcomes in treatment. Prinz and Miller (1996) discussed the research in this
area and concluded that a multitude of stressors has an adverse impact on
training outcomes including parental engagement in therapy, reduced treatment
motivation and participation, as well as the parent child relationship. Parents with
a multitude of environmental and personal stressors are more likely to over react
to daily hassles (car repairs, sick family member) (Prinz & Miller, 1996). Parents
may rationalize and attribute drop outs and missed appointments to these types
of situational obstacles (Prinz & Miller, 1996).

Webster-Stratton (1985) hypothesized that an interaction between types of
life experiences and the availability of social supports may be predictive of parent

training outcomes. Wahler’s early seminal research found that single mothers

50

who have contact with other people outside the home have much more treatment
success than mothers who are socially isolated (Dumas & Wahler, 1983).
Maternal insularity increases the probability of parent training program failure
(Wester-Stratton & Hammond, 1990). Other researchers have found that social
insularity may interact with other stressors to affect parenting styles (Dadds 8
McHugh, 1992). High levels of social support was found to be predictive of how
parents responded to treatment, specifically mothers without support failed to
benefit from traditional parent training (Dadds & McHugh, 1992). However, when
access to social support systems was made a part of the treatment, mothers
without their own supports still failed to benefit (Dadds & McHugh, 1992).
Severity of Child Behavior

Studies have discussed the limited effects of parent training for
adolescents (Ruma et al., 1996). Ruma et al. (1966) used archival data to test for
age effects on the outcomes of parent group training. They found that the best
predictor of parent training outcome was the severity of the child’s problems and
concluded that severity of the behavior problem was responsible for the age
effects in parent training (Ruma et al., 1996). This study, however, was limited in
that they had no control or comparison group, follow-up data was not available,
long term effects were not noted, outcome measures were limited to one global
rating of psychopathology among children (the Child Behavior Checklist), and
one informant (mothers) provided the rating (Ruma et al., 1996).

However, other researchers have also found that parents of children with

greater numbers of presenting problems were less likely to complete their parent

51

training program (Holden et al., 1990; Kazdin & Wassell, 1999). Kazdin (1997)
reported that more severe antisocial behavior and comorbidity are predictive of
negative outcomes to parent management training.
Other Predictors

Webster-Stratton (1993) summarized the research on predictors and
concluded fewer treatment gains with the following parent and family
characteristics: marital distress, spousal abuse, lack of a supportive partner,
maternal depression, poor problem solving skills, high life stress, and SES
disadvantages coupled with lack of social support.

Bank et al. (1991) studied time in treatment as a predictor and offense
rates among chronic delinquents as the outcome variables. They found a
significant treatment by time effect with greater reduced rates of serious crimes
with time in therapy (Bank et al., 1991). Holden et al. (1990) found that ethnic
minorities were less likely to complete parent training.

Failure to Find Consistent Predictors

 

Based on the review of studies examining predictors, there is no single
consistent predictor of treatment success or failure (Table 9). For example,
evidence from the Carolina Early Intervention Program indicated that children
demonstrated benefit regardless of parent (mother in this case) intelligence, age,
education, or type of family (Ramey et al., 1988). The apparent failure to find
consistent predictive outcomes based on variables identified in the literature may

be due to several limitations in the research literature.

52

«00:0 000: . 3000500.: «0005090

50 0000 :0 028620

00:29:00 .0 02.2005
00:29:00 :0 026690
00:99:00 :0 02306000 :00

.0958: 0_ 0.0:

3:985 .0800
wmw 30.
00050.83 0.050
0068.000 .mEEmE

mam.» 3:00.00 2050
200005 5600000030
0390 3:00..

900305.00
200000

DO
000

0000=0E00000
w 02000030
00. 00:90:

:8: :95»;

w 0090000 .93. 63000.2 £03

88: 29:05: a 8:20.868;

83: 090800 a .0530... .0020:

 

0858; mmm 30.
000500000 060$ 00
Ewe:_0m_m-000 020000000 .mEcfiE 006000000 .mEQmE 000 38: 000005.568;
093090006 0_E000000_0om
900000 060?.
800030 “000000.: 000. a 3:235 _m_0om 0030000 38: 00.00; a. 8an
000620000 0050096
Esmom £22005 00:053.”. >036

 

 

 

@550; Ewan. 0_ m00«0_000d&0_0:0|0xw 863w :0 EMEEzw

m 298.

53

 

0000000 0: 00:00: >0 00000300.
0000000
50328 200 .o 02.285

00.0090. 00.000

00.0850 006.0000 3:0>00

0.30 090 00 0200600
050.090 U.0 028.0000

00:00.0:000 30.
0.00050 00 00.0.0000 .000:_00_0-000
50 080 00 00:00 .0090 006600

0:00:00 3000 0. 05:3

32050000000 E0000

mum
00000000

5030000 00 3:0>00
0.00000:

5.3 0.0000:0_00 .0000>0_00
00.000000 00:08:00 .0000

0030000 00 3:0>00

0030000 0:00 00 3:0>00

00000000 .0.000000>00
000 :00

00000000 060000000
00.000000: 00.00
000.090.: 0:000>00

000000 .0000 30.

Do

_0_000:00
02000500
.000:_00000

.0800
-:00 0.000

0.00.0000 0030000

0000000

00
5. x0: .0

00

0.00.000 0030000

.0 :00__0E00

.000 0000000

.000: 020002, 0 :6me

that >0_>>o._0 w 000:0... .06me

:00: 060v.

.000: 8859: 0 .9030 .0050

.000: .002 0 2:0

.000: 0090... 0 000000000000;

.000: 00055.0 0 00.00.”.
.000: 00:00.2 0 00000

m 0.00... 3000

54

In a review of parent training research, Rogers Wiese (1992) discussed
the shortcomings in research methodology which had been addressed by
previous reviews as follows: 1) an over-reliance on reporting data about child
behavior changes rather than specific parent behavior changes; 2) inconsistent
use of multiple observers; 3) failure to report inter-rater reliability with the former;
4) failure to use multiple outcome measures; 5) little assessment of
generalization of treatment effects; and 6) limited, if any, follow up data
(Moreland et al., 1982; Rogers Wiese 8. Kramer, 1988).

Subsequent research appears to have addressed many of these
methodology shortcomings. Notably, later parent training research shows an
upward trend in collecting parent change data, increased multiple outcome
measures, and collection of generalization data (Wiese, 1992).

Group studies in parent research also contained several methodological
limitations. These included a failure to gather or include follow up data, failure to
include control groups, minimal collection of treatment integrity data, and fairly
limited reporting of assessment and treatment protocols (Wiese, 1992). These
undermined the internal and external validity of the group results and thus make
it impossible to determine if the intervention is responsible for changes and if
changes generalize over time, settings, and people (Wiese, 1992).

Several researchers have Webster-Stratton and Hammond (1990)
described several confounding issues in many parent training studies as follows:

1. examined only one isolated predictor variable at a time, or combined
predictors in a single index score making it difficult to determine effects of

individual predictors (Webster-Stratton 8. Hammond, 1990)
2. lack of a standard method to assess treatment success or “clinical

55

significance" (Webster-Stratton 8. Hammond, 1990)

3. almost exclusive reliance on single not multiple outcome measures such
as maternal information, neglecting paternal variables (Webster-Stratton 8.
Hammond, 1990) and ignoring parent and child (Tucker et al., 1998)

4. using small homogeneous samples, lack of control groups at long term
follow up, lack of random assignment, and wide variability in training
programs (Tucker et al., 1998).

Two specific examples of methodology shortcomings involve the child’s
age as a predictor and the selection of outcome measures. While age has been
discussed as a predictor of outcomes, studies have also seldom controlled for
different age groups, younger and older children, whose parents are involved in
training (Ruma et al., 1996). Home observations and teacher reports have been
discussed as possibly better outcome measures because they are less biased
and based on performance demands than parent reports of behavior (Webster-
Stratton 8. Hammond, 1990).

Thus, research has yet to demonstrate how several predictors may
function together, how predictors function independently and how to operationally
define treatment success (Wester-Stratton 8. Hammond, 1990). Consequently, it
makes sense to conclude that different predictors would affect different outcomes
(Wester-Stratton & Hammond, 1990).

Using statistical analysis, actual variance among predictors of outcomes in
parent training has been reported to be small (Webster-Stratton 8. Hammond,
1990). The data suggest that the predictor variables which were studied
(including maternal depression, marital situations, SES, life stressors) may have

limited utility in practice. Secondly, the authors suggested that researchers may

need to look for other predictors that determine outcomes in parent training.

56

Kazdin, Holland, and Crowley (1997) label traditionally studied predictors
as “variables of convenience”. Researchers may select variables to study based
on easy access, such as clinic intake forms. Also, the predictors are usually
demographic data which provides little information regarding mechanisms
involved in outcomes nor guidelines on how to intervene in parent training to
prevent negative outcomes and provide better treatments (Kazdin et al., 1997).
This creates the propensity towards exclusionary practices. The solution for
many PMT programs has been to carefully select clients and exclude parents
who have mental health problems (Hutchings, 1996). Another example involves
SES. Knowing that low SES is a predictor of negative outcomes in parent training
does not provide guidelines for treatment.

Kazdin et al. (1997) proposed, studied, and found support for their
hypothesis regarding a conceptual model which would be predictive of parent
training outcomes. The “barriers-to-treatment model” includes predictors which
describe practical obstacles to participation, perceptions of treatment, and
relationship with the therapist (Kazdin et al., 1997). One notable finding was that
the predictor findings related to SES may be due to several mediating variables
such as cost factors for transportation, more severe child-rearing strategies,
more child and/or parent health problems, and discrepant parent - therapist
expectations (Kazdin et al., 1997).

In a meta-analysis of psychotherapy dropout, Wierzbicki and Pekarik
(1993) found that psychotherapy dropout was significantly related to minority

status, racial status, low education, and low SES; however, mean effect sizes

57

were moderate, ranging from .23 to .37. In fact, they concluded that these mean
effect sizes are likely upper estimates of the true value of the effects, due to
limitations in their study (missing data) (Wierzbicki & Pekarik, 1993). Although the
study did not specifically address or examine why SES is related to dropout, the
authors speculated that SES may be related to dropout because of several
underlying factors such as differences among therapists and clients regarding
education, value systems, and expectations (Wierzbicki 8. Pekarik, 1993).

These results suggest that outcome research should focus on more
complex psychological variables rather than demographic characteristics (Kazdin
et al., 1997; Wierzbicki 8. Pekarik, 1993). Furthermore, parent management
training can encompass many variations in treatment within the actual program
and in combination with other treatments (Kazdin, 1997). This would have
implications for outcome studies that attempt to identify predictors of outcome.
One area of study which has not been sufficiently explored is who profits from
single modality treatments such as PMT or from combined treatments (Kazdin,
1997). Also, unexplored is whether certain parents profit from certain types of
PMT treatments as opposed to others. It has been hypothesized that certain
parents and children may be a “good fit” for a particular type of parent training
program (Holden et al., 1990). This proposal is designed in an attempt to build on
previous findings regarding the predictors of PMT, identify those characteristics/
variables in advance to anticipate and provide the most effective training and

prevent parents from dropping out prematurely or progressing slowly.

58

Parent Cognitions As Predictors

 

Attention to parents’ cognitions arose within psychology as a reaction
against the exclusively behavioral accounts of parenting (Goodnow, 1995).
Interest in belief and attitude theories arose partially out of a reaction to simplistic
behavioral explanations for the development of childhood disorders (Goodnow,
1995). Furthermore, interest in the literature on parental beliefs and attitudes is
consistent with Eifert’s (1996) recommendation to use more theory driven
approaches in behavior therapy complemented with individualized and
manualized treatment. Yet, more general parent cognitions, such as childrearing
beliefs and attitudes have been shown to be limited in predicting parent behavior
(Geller & Johnston, 1995). Part of the reason may be the difficulties with defining
constructs. In examining the research literature on beliefs in general, research
studies demonstrate no consensus regarding what beliefs and attitudes really are
and considerable confusion about the constructs. The constructs of beliefs,
attitudes, perceptions, ideas, attributions, values, expectations, knowledge,
cognition, perceptions, judgements, and conceptions have been used
interchangeably throughout the literature (Holden, 1995; McGillicuddy-De Lisi 8
Sigel, 1995). Beliefs have been defined as knowledge or ideas that are accepted
as true; attitudes defined as an evaluative aspect that is an integral component of
the parent’s cognitive orientation as in “positive” or “negative"; values refer to
longstanding goals that a parent holds for a child, rather than truths; both

attitudes and values may include a cognitive component but these are not seen

59

as facts or truth (McGillicuddy-De Lisi 8 Sigel, 1995, p.334). Kohn (1969) in his
seminal study on social class and values defined values as:

the values that parents would most like to see embodied in their

children’s behavior - the characteristics they consider most

desirable to inculcate in their children.

(Kohn, 1969, p.18).

Basically, there is no generally acceptable term for belief nor any agreed
upon definitions for the different cognitive constructs in the literature
(McGillicuddy-De Lisi 8 Sigel, 1995). Consequently, research has shifted to other
areas of parent cognitions.

According to four models, (attribution, information processing,
constructivist, and transactional) parental beliefs are sources of influence on the
developing child, on the parents, and on the parent-child relationship
(McGillicuddy-De Lisi 8 Sigel, 1995). An implicit and explicit assumption is that
parental beliefs guide parental actions with children and consequently will
influence childhood behavior. Some studies attempt to target the nature of
parent’s beliefs and link them to childhood outcomes (McGillicuddy-De Lisi 8
Sigel, 1995). Interestingly, these studies view the age of the child as less
important and do not include the children in their studies. More important is the
issue of measuring these attitudes and beliefs.

Studies have developed child rearing attitude surveys for predominantly
descriptive, theoretical, and methodological purposes (Holden, 1995). Few
studies developed surveys to serve as clinical tools. Frequently studied parental

attitudes toward childrearing include four main categories: 1) attitudes related to

discipline and control (authoritarian, permissive, restrictiveness, physical

60

punishment, democratic versus autocratic); 2) attitudes related to affection
(acceptance, hostility, rejection, approval, warmth); 3) other parental attitudes
(consistency of parenting, encouragement of independence or dependence, toilet
training, role in education/teaching, over-involvement, and worries about
childrearing); and, 4) attitudes related to views about child or child characteristics
(assertiveness of child, conformity/deviance of child, fears of child, judgements
about child behaviors, sex behavior of child, achievement of child, approval of
child, emotional health of child, health of child, sex role expectations of child, and
verbal output of child) (Holden, 1995). There were few psychometrically solid,
formal instruments available for measuring parent cognitions (Johnston, 1996).

Earlier, seminal research utilized parent interviews to measure parent
cognitions. Baumrind (1971) and Baumrind and Black (1967) categorized
parental attitudes into authoritarian, authoritative, and permissive types based in
part, upon interviews.

Attitude surveys sometimes include descriptive beliefs, instrumental
beliefs, behavioral intentions, and values. Parents have descriptive beliefs about
childrearing which are beliefs about how children develop or how parents affect
children (Holden, 1995). For example, the statement, most children are toilet
trained by 15 months is a descriptive belief. Parents also have instrumental
beliefs about childrearing which are defined as the ways in which parents can
achieve particular goals (Holden, 1995). For example, the statement, Ibelieve

physical punishment is the best way to discipline, is an instrumental belief. It is

61

these instrumental beliefs which are of the most interest and may have the most
applicability to outcomes in parent management therapies.

Behavioral intentions are simply the intention to behave in a particular way
(Holden, 1995). Few childrearing surveys explicitly address childrearing practices
which focus on specific behaviors and how often parents behave in that manner.
The surveys that do, include questions such as how often do you require your
child to put away his or her clothes” (Holden, 1995).

Values, perceptions, and self-perceptions are also addressed in some
childrearing surveys. Values are regarded as superordinate categories that
reflect abstract goals or a coherent set of attitudes (Holden, 1995). An example
of the former is to raise a happy child while an example of the latter is respect
one’s parents. Perceptions address parents’ views or reactions to their child
(Holden, 1995) . For example, items assess views about how difficult the child is
(Holden, 1995). Self-perceptions refer to parents’ attitudes about their own
abilities and their relationship with their child (Holden, 1995). For example, an
item here might include the statement, I find the greatest satisfaction in my child.

Research has demonstrated over and over again, that parents’
childrearing attitudes and beliefs explain only a small proportion of the variance
in parenting behavior (Geller 8 Johnston, 1995). This suggests that attitudes and
beliefs are not the sole determinants of parenting behavior and that other
cognitions may play a role (Geller 8 Johnston, 1995). First several investigators

have examined the interaction among parents’ beliefs and attitudes with single

62

parenthood, SES, parent psychopathology, and life stressors. Second,
investigators have begun to examine other cognitions.

Interactions Among Teps And Parent Cognitions

 

Studies have examined the relationship among childrearing attitudes and
several traditional examined predictors (marital status and maternal insularity,
SES, maternal depression, and life stressors) as determinants in parenting
behavior. Marital status and adjustment has also been linked to various
childrearing attitudes. Goldberg (1990) replicated previous work and again found
that mothers in happy marriages reported attitudes of greater warmth and more
encouragement of independence in their children while mothers in unhappy
marriages reported more aggravation with their children. Cox et al. ( 1989) in a
study of 38 couples with infants, reported that mothers in close, confiding
marriages were warmer and more sensitive and fathers in the same marriages
had more positive childrearing attitudes than parents in less happy marriages.

Kohn (1969) hypothesized about the relationship between attitudes and
social class. Kohn found substantial support for his theory which relates social
class to values to parent behavior. It is noteworthy that social class was
evaluated by the Hollingshead Index which is based on education and
occupation. This index leaves out income and subjective class identification. He
proposed that working class parents value conformity to external rules,
emphasize obedience and good manners, neatness, and impose constraints on
their children. Furthermore, these values are the result of life conditions, i.e.,

pa rent job requirements associated with working under supervision, and following

63

instructions. In contrast, parents from higher occupational levels value self
direction, self-control, curiosity, creativity, and are more supportive of
independence in their children in keeping with their job requirements.

Early studies have linked maternal depression to parent attitudes. Scores
on the Parent Attitude Research Instrument (PARI), specifically, hostile
childrearing attitude scores were associated with anxiety and depression in
mothers (Holden, 1995). Brunnquell, Chrichton, and Egeland (1981) in a large
study of 267 high risk mothers described the interaction of childrearing attitudes
and personality variables as the best predictors of quality of care infants received
with implications for exploring intervention in mother-child disturbed relationships.
Neither single attitudes nor single personality variables were discriminant of high
functioning versus lower functioning mothers. Hall et al. (1991) in a study of 225
mothers found low income, single mothers to be at high risk for depressive
symptoms. Maternal depression symptoms predicted parent attitudes which in
turn predicted child behavior (Hall et al., 1991). Research also supports the
findings that depression promotes negative perceptions of child behavior and is
associated with more negative attributions for child behavior (Johnston, 1996).

Maternal insularity and social support has also been linked to various
childrearing attitudes. Cochran and Niegro (1995) found the presence or absence
of social supports impacted on maternal attitudes and adjustment. Crnic and
Acevedo (1995) found that social support and stress predicted maternal attitudes
toward satisfaction in parenting. Crockenberg (1987) in a study of 40 mothers

with two year old children found that mothers who experienced little support and

64

were, themselves rejected by their parents, held negative, punitive, and angry
attitudes towards childrearing. In an extension of the research associating life
stressors with parent perceptions of child deviant behavior, (Johnston, 1996)
reports that major life stressors differentially affect mothers’ perceptions of child
behavior. Thus, links have been drawn among other traditional predictors
(maternal depression and marital adjustment) and parent attitudes.

Potential Motivational Predictors Of Outcomes In Pmt

 

Many of the problems parents face in managing the disruptive and
inappropriate behavior of their children are likely motivational in nature.
Considerable research has been done in the general area of motivation in
parenting behavior and child development outcomes (Goodnow, 1995; Holden,
1995; Lisi 8 Sigel, 1995). One basic premise has been that parent child rearing
“attitudes” and their core philosophies guide their behavior and ultimately, the
child’s development (Gillis-Amold, Crase, Stockdale, 8 Shelley, 1998). The
present proposal contends that many of the problems and negative outcomes in
PMT programs also contain a motivational in nature mediating component.
Furthermore, these motivational mediating components are central to parent’s
inner processes and not only the child’s inner processes (which is an entire other
proposal). A definition of motivation will help illustrate these points:

Motivation is the study of the internal processes that give behavior

its energy and direction......motivation originates from a variety of

sources (needs, cognitions, and emotions), and that these internal

processes energize and direct behavior in multiple ways such as

starting, sustaining, intensifying, focusing, and stopping it.
(Reeve, 1996, p. 2).

65

It has long been established that student motivation has a direct causal relation
to student achievement (Brophy, 1998). Parents involved in PMT programs are
essentially involved in achievement type situations — hence the term, parent
education with some programs. Consequently, it stands to reason that parental
motivational factors influence parent “achievement” in parent management
training programs, and ultimately, translate into goal directed behavior which
facilitates either positive or negative outcomes of PMT.

Reeve (1996) describes eleven inner motivational resources including
psychological needs (self-determination, competence, and relatedness),
cognitive domains (self-efficacy, personal control beliefs, achievement strivings,
goal setting, and the self), and emotional states (curiosity, interest, and
enjoyment/positive effect). Brophy (1998) defines motivation as “a theoretical
construct used to explain the initiation, direction, intensity, and persistence of
behavior, especially goal directed behavior” (Brophy, 1998, p. 3). Organizing
much of the motivational research within an expectancy x value model, he
explains that,

the effort that people are willing to expend on a task is the product

of 1) the degree to which they expect to be able to perform the task

successfully if they apply themselves (and thus the degree to which

they expect to get whatever rewards that successful task

performance will bring and 2) the degree to which they value those

rewards as well as the opportunity to engage in the processes

involved in performing the task itself.

(Brophy, 1998, p.14).

Few of these inner motivational resources and value aspects have been

directly and explicitly examined within the context of parent management training

(PMT) outcome research. Only a handful of studies have alluded to the influence

66

of parent motivation within the context of PMT outcomes and mostly in a
peripheral fashion (Table 10). One study which explicitly examined motivation
within a PMT program, taught parents “motivational” strategies for changing child
behavior and did not address the motivational constructs underlying the
strategies (Koegel, Bimbela, 8 Schreibman, 1996). Another study examined
motivations for foster parenting in a foster parent training program (Gillis-Arnold
et al., 1998). Other studies did not mention the word motivation or related
motivational aspects, although the content suggested elements consistent with
the motivational constructs of goals (First 8 Way, 1995) and value aspects of
motivation (Kazdin, Holland, 8 Crowley, 1997; Kazdin 8 Wassell, 1999).

A few studies directly addressed the motivational aspects of self-efficacy
but not as predictors of PMT outcomes but as a parent outcome of participation
in PMT (Thompson, Ruma, Schuchmann, 8 Burke, 1996; (Tucker et al., 1998);
and, as a pre-post measure of self-efficacy regarding the use of children’s mental
health services (Bickman et al., 1998; Heflinger et al., 1997). Outside the context
of PMT but within a situation responsive nursing program, one study examined
parent’s general self-efficacy and locus of control as related to health promotion
behavior (Ford-Gilboe, 1997).

Those resources pertaining to parent’s cognitive domains (self efficacy
theory, personal control beliefs/attributions, achievement strivings/goal theory,
self theory, and value) are the focus of the present proposal and postulated to be
predictive of outcomes in PMT. Early seminal studies are credited for their

theories of motivation and implications for understanding thought and behavior

67

000000 .00.>0000 0:00 0.
00.0.0. 3009.0 .00.: 000.00:
.0 00:00.0. 00200.00 .00.00

.0500. 00.08.00 :0
0.00:0 00 00.0.0: >000_...0 ._00

0.00 .0.0000: >0 00.02.0E
02.000000: .200 02.0 .0.
0.00:0.000E00 .00.0..00 02.0.0
00.00.00. .0.00:0.:00:.00 0.20
00000. 0. 00.02.00. 0.00. 02.0000

0.00000 00_0>
00.00._0E. .00E.00:
0. 0.0.00: 0.... .0 000

>000...0.._00

00:00.00
.0.00. .0. 05:02.0...
0.02.0000 00

000.20... 002.098 00
00.0.0. .0 .00.:000005. .00.0n. 0...... 00.0..00 .000: .0.000>> w 0.0~0v.

300.90 0005 00”.. 00000 0:00..
.00.0.0 .00:..0.<.00Em .0.00E 000:0> C00 5 00000.0:00
0200.00 ...0.0000> 0...: 00.0..00 .0 .00.0002 0000.05 00000.0...

E0_00...00 0.00.00 .0.00.
00.0.0. .. 003.000.... ”00:00.00 02.000000: 800 3 >205
.0.000 .0. 02.0000... 0.02.0. 0:0 02.0000 .0 0.000.020 .000.0 .0.00.<-0.__.0

 

08:00.0 >0._000 .0002. 0:000 >000....0 ._00 .0.0000 00.0.00 00......0.
0. 00..00.0:.00 2.0.0. .00..0 ._0::00 .0 02000000. 00:02.0 ".210. .00.00 02.. 0:0
>.0>:.000 00:02.9: .0 000.000 0000. 0000: .00.0.0. .0005. 0:00... .0.00E00_0>0n. .00.00 0.00.0 .0005 .s. 08.5-0.0“.
0:.E00. 0>:0:..0.000.. u 000.00.00 00.00000 ._00 .0.000 Cwn. .00.
.00:00:00>0. .00.00_0000.00000 00.00..0:.. 0.00.020 .0.0>00 .0 0.00E0.0 00.00.00: .000 3 >0>> a .0.....
00.00800 0000290. .00.00 300.0... 0005 00“.. 00000 5.000
000 000.00.00 .0.>0000 0:00 :0 .00.0.0 .00E.0z.00.:w .0.00E 000..0> 800: 00....000 .0 000020000
0.00.00 0: 00.0.0: >000...0 ..00 >000_...0.._00 0200.00 ._.0.0000> 0...... 00.0..00 .00.0002 0000.00... 00.00.05
00:02.00. 0. 0....0000.000.00.0000 .00:0:00 0.9.00.0 00:0.0000 >020
05:02.05. 0:65 .0 :20

 

 

 

 

.0.x0.000 :0:00>.0.0. .M..0..m .0 .220 0.5.2. 00.00..0> .0:0:0>:0.2 00.0.E0xw 00.030

or 0.00...

68

 

00.00.00. 00000.0 00..00.00 0.
0000000 02..000 0. 000. .000
0000.000

.00.0.00.. .0. .0000 .000:.00.0

00.00.00
00 5000.0 .00.00 00 .0000 00
00.0.0 .0.00E 000 000 .00.00
02.0000 .000 .00.0.00..

.060. 00.0.0: 00:02.00.

0...... 000..00.0.0. .00.00-0_.00
02.000 0.00. .005 0000.000
.00.0.00.. .0. .00..0 .000:.00.0

0.0.000 .00.0 0000.0. .0050.
.00 0. E0800: .00. 00:08.00
0000.00. .00E.00.. 0. 0.0.000

000.020 0.030.500.0003.
5000.0 ._00 C00. 00.0.0: .00.00 .0.0.>0000

.00000. 008.00.. 000
.000.> 00.0 0.0. 00:00:00.
.00.... 008.... .0000.
5000.0 ..00 0.....0 .00.0000000. .0.>0000
00.0.2.0. 2.0.0.00
000 0.00.0..0
0.02.0. 0.00.
0x00. 0.0 0......
2.00 x.0. .00.000
2.6 00. 0......
”00.0.0 >00.0..0
0.0.0.0000 00

.000 m... 0.02.0...

0. 2.00 00 00.00.000.
00000000. .0820

00.0000. + 032 .N
.0.>0000 .00.0. 00.0000 0. ..
H00:00:.000.

.0.>0000 00.0000.

.020.
00.0.0... .00E00000s. .00.00

.2000 2.6 .5. C000. 0...
-00.. 0:00 9.3.00 0.0.8.0

00:02.00. .0 0.00000
00_0> 00.00..00..

0.208.
.020... 02....

0.00.090
.0.>0000
0...... 00.0....0

0.0..00
5.2. 00.2.6

.0.000..00
w .0>.000.000
._000...00000
0...... 00.0....0

.000..0000000

a $000.00 .0000 .0005 00000.0

.000: 9.50

w .000E00000m .00.00 .0000E00...

.000:
00:52.00 0 0005.0 .0008.

£00.. .2390 .0 0:00: .00.00.

c. 0.00... 0....00

69

related to self efficacy (Bandura, 1997), causal attributions (Weiner, 1979), and
performance and Ieaming goals (Dweck & Leggett, 1988). These and some of
the more recent studies in motivation regarding other domain specific areas
provide the groundwork for application to PMT. Within each motivational section,
a review of relevant studies examining motivational factors which influence
parenting in general are provided as well as a discussion of the sparse, PMT
research which can be related to these motivational aspects. The motivational
factors discussed below (self efficacy, personal control beliefs, achievement
strivings, goal theories, implicit theory, self theory, and value) are those which
are the focus of the present study and predominantly apply to PMT outcomes.

Self Efficacy Theory

 

Self efficacy is concerned with how well or poorly people judge or believe
that they will perform on a task given their skills and a variety of circumstances
(Bandura, 1997, p.37). It is the individuals’ beliefs about their abilities and about
the outcomes of their actions which influence the many ways in which they
behave (Pajares, 1996). For example, individuals avoid tasks in which they feel
incompetent and engage in tasks in which they feel competent (Pajares, 1996).

Self efficacy beliefs originate from four sources of information including
performance accomplishments, vicarious experience (modeling), verbal
persuasion, and physiological state, in decreasing order of potency of influence
(Reeve, 1996). Performance accomplishments are first hand mastery
experiences in which successes increase efficacy while failures decrease

efficacy beliefs (Reeve, 1996). Several caveats to these maxims are given which

70

implicate perceived difficulty of the task, the amount of physical guidance
received, and the timing of the pattern of successes and failures (Reeve, 1996).

The second source of information, vicarious experience, occurs when
people base their efficacy beliefs on a social comparison process with others
(Reeve, 1996). Furthermore, efficacy beliefs are strengthened when observers
watch coping models (who struggle and cope to overcome difficult tasks) as
opposed to mastery models (who expertly perform flawlessly) and when they
watch multiple models (Reeve, 1996).

The third source of information upon which individuals base efficacy
beliefs, verbal persuasion, occurs when others (teacher, parent, coach, peer,
employer, therapist, audience, inspirational poster, song) give a successful pep
talk (Reeve, 1996). The persuader must be credible, trustworthy, and have some
“expertise” in order for verbal persuasion to have an influence on efficacy
expectations (Reeve, 1996).

The least potent source informing efficacy beliefs, physiological state,
comes from experiences prior to, or during task performances (Reeve, 1996).
Autonomic nervous system arousal (increased heart rate, respiratory rate,
trembling hands), somatic arousal (muscle tension, fatigue, pains and aches),
and cortical arousal (rate of thinking, mental confusion) sometimes occur during
some performances (Reeve, 1996). The degree to which they are perceived to
be “under control” heightens efficacy beliefs (Reeve, 1996).

Self efficacy beliefs have been found to have diverse influence on

student’s academic functioning including choice of activities and selection of

71

environments; extent of effort and persistence put forth during task performance;
quality of thinking and decision making during performance; and, emotional
reactions, related to stress and anxiety (Reeve, 1996). Students with low self
efficacy likely “shun and avoid activities and situations they see as likely to
overwhelm their coping capacities” and arrest developmental potentials
(Bandura, 1997; Reeve, 1996). Similarly, parents with low self efficacy may be
likely to drop out of PMT. Students with strong self efficacy beliefs choose active
engagements that contribute to growth of competencies (Reeve, 1996). Students
with low self efficacy slacken their efforts or give up when things get difficult.
Students with high self efficacy persist with coping efforts when things get difficult
(Reeve, 1996). Again, parents with high self efficacy are more likely to follow
through with PMT suggestions and to complete homework assignments.
Students with low efficacy become cognitively erratic when problem solving is
required (Reeve, 1996). Students with high efficacy become highly efficient in
their analytic thinking during complex decision making (Reeve, 1996). This latter
finding is among the most salient in reference to parent efficacy when judging
how well or effectively, they can manage their child’s disruptive behavior. In order
to maximize the efficacy of PMT involvement, parents must use memories of past
events to generate hypotheses about the most effective course of action, analyze
feedback to assess the validity of their plans and strategies, reflect upon
performances and remember which were most effective and which not. Parents
who are low in self efficacy are likely high in self doubt which deteriorates and

impairs the thinking and decision making process, especially in the face of

72

situational stress and problem solving failures (Reeve, 1996). Finally, students
high in self efficacy, visualize success scenarios and react to challenges with
enthusiastic effort, optimism, and interest (Reeve, 1996). Students low in self
efficacy dwell on personal deficiencies, visualize formidable obstacles, and react
to challenges with pessimism, anxiety, and depression (Reeve, 1996). Basically,
events offer potentially aversive experiences and consequences but individuals
offer coping capabilities which can counteract the potentially ovenlvhelming
aspects of the event (Reeve, 1996, p. 89).

The above discussion is central to parent self efficacy beliefs regarding
their own capabilities. Also relevant here, is the teacher efficacy literature which
involves “an instructor’s belief or conviction that he or she can influence how well
students learn, especially students considered to be difficult or amotivated”
(Reeve, 1996, p. 89). Because teacher efficacy influences a teacher’s selection
of activities, effort, and persistence (Reeve, 1996), it is also likely that parent
efficacy influences the same. Parents with high efficacy may also believe that
they can bring about “Ieaming, performance, and achievement” in their children.
Parents with low efficacy avoid offering activities that have the potential to be
unmanageable, give up quickly, and forgo the effort to reteach or try again.
Although measuring teacher efficacy is difficult, three types of items are deemed
to represent the essence of teacher self efficacy (Reeve, 1996):

1. When I really try, I can get through to most difficult students.

2. If a student did not remember information I gave in a previous lesson, I

would know how to increase his or her retention in the next lesson.

3. If a student in my class becomes disruptive and noisy, I feel assured

that I know some techniques to redirect him quickly.
(Reeve, 1996, p. 90)

73

Similar statements from surveys assessing self efficacy and supported through

factor analysis procedures include the following five items:

1.

PWN

If I really try hard, I can get through to even the most difficult or

unmotivated student.

If some students in my class are not doing well in math, I feel that I

should change my approach to the subject.

By trying a different teaching method, I can significantly affect a

student’s achievement.

There is very little I can do to insure that most of my students achieve

at a high level.

I am certain I am making a difference in the lives of my students.
(Midgley, Feldlaufer, & Eccles, 1989)

By altering the domain specificity of these items, these same types of items can

be translated and applied to parent self efficacy:

1.

2.

4.

5.

When I really try, I can get through to my son or daughter when he or
she is being most difficult.

If my son or daughter did not remember the last time we had a “scene”
about a certain house rule, I would know how to increase his or her
memory the next time.

If my child became disruptive and noisy in public places (malls,
church), I feel assured that I know some techniques to redirect him or
her quickly.

By trying a different discipline method, I can significantly affect my
child’s behavior.

There is really very little I can do insure that my child behaves
appropriately.

(With one additional related question),

6.

If I could choose to have children, if I had the decision to make over
again, I would choose parenthood.

Notably, these statements involve judgements which are task and domain

specific. In order to be clear about what is being measured and to differentiate

other expectancy beliefs from self efficacy beliefs, assessment requires context-

specific judgments at a micro-analytic level (Pajares, 1996). From the

74

perspective of more task and situation specific judgements, the following
statements may also apply to parent efficacy:

1. I am confident that I can make the decisions that are made in our home
(parent efficacy).

2. I am confident that I can have my son or daughter listen to me and comply
when I am discussing household chores (self-efficacy for engaging
compliant behavior).

3. I am confident that I can learn the material with the other parents in these
sessions (collective efficacy).

(adapted from Pajares, 1996)

Collective efficacy is defined as “a group’s shared belief in their
capabilities to realize given levels of attainment” (Bandura, 1997). Furthermore,
this collective efficacy has been reported to mediate the influence of students’
socioeconomic status and prior academic achievement (Bandura, 1997).
Coupled with the findings from a number of studies which have demonstrated the
relationship between teacher efficacy and student achievement (Midgley et al.,
1989), parent efficacy beliefs are likely to have similar influences on achievement
in parent training and ultimately child behavior. It is also possible that the group
parent training “collective efficacy” may serve to mediate the effects of
traditionally examined predictors, such as SES, in PMT outcomes.

A series of studies on The Family Empowerment Project is among the
sparse literature found which remotely examines the relationship between parent
self efficacy and PMT (Bickman et al., 1998; Heflinger et al., 1997). Bickman et
al. (1998) utilized a multi-component parent training curriculum to enhance

parent empowerment and increase their involvement in mental service delivery

for their children with special needs. While this training does not meet the

75

definition of PMT nor apply exclusively to children with disruptive behavior
disorders, the program included mental health service’s self efficacy as a key
aspect and novel addition (Heflinger et al., 1997). The four self efficacy sources
of information (performance accomplishments, vicarious experience/modelling,
verbal persuasion, and physiological state) were used as training techniques to
help parents perceive themselves as capable of acquiring and using newly
acquired knowledge and skills about the mental health system to improve their
children’s treatment. The training of mental health services self efficacy
specifically addressed (a) encouraging active participation in decision making, (b)
promoting collaborative relationships between parents and professionals, (c)
modeling participation in a support group (Heflinger et al., 1997).

This series of evaluations indicated that the training intervention increased
parent mental health services efficacy but parent knowledge and self efficacy
was not related to parent involvement, service use, or children’s mental health
(Bickman et al., 1998; Heflinger et al., 1997). This suggests that parents’
confidence in their knowledge of service systems and skills needed to interact
with service systems is not related to better outcomes, in this case as measured
by increased involvement. However, these studies still leave open the question
as to whether parents’ confidence in their knowledge of child management and
skills needed to apply that knowledge, promotes better outcomes in PMT.

Two other studies examined self-efficacy within the context of PMT but
explored self-efficacy as an outcome measure of participation in the PMT

interventions (Thompson et al., 1996; Tucker et al., 1998). Thompson et al.

76

(1996) examined cost effective PMT treatments and evaluated changes in parent
self-efficacy as an outcome of participation. Specifically, the authors found that a
modified, less time consuming and less costly PMT program is an effective
treatment on post test outcome measures including child behavior changes,
parent attitudes and satisfaction with family relationships. As noted previously,
the authors also found that marital status did not predict outcomes in parenting
efficacy or parenting satisfaction (Thompson et al., 1996). However, outside the
context of PMT, Johnston (1996), discusses the reciprocal nature of disruptive
child behaviors and parenting efficacy. This may suggest a link between self-
efficacy and severity of child behavior, another TEP.

Tucker and colleagues, (1998) examined the effectiveness of a behavioral
parent training (BPT) intervention which is another name for PMT. The authors
found that positive changes in parenting self efficacy, parenting stress, and
parent-toddler interactions resulted from participation in a PMT intervention.

A study peripheral to PMT examined general self efficacy and locus of
control, noted below in the attribution section (Ford-Gilboe, 1997). The term
general self efficacy was used to denote a general belief in one’s capabilities to
successfully perform behaviors across situations (Ford-Gilboe, 1997). They
found a moderate correlation between general self-efficacy and health behavior.
The authors point out that the usefulness of general self-efficacy as a predictor of
specific health behaviors has been debated (Ford-Gilboe, 1997). Given the lack
of clarity and domain specificity of efficacy beliefs, it is possible that other

motivational expectancy aspects were measured as well.

77

Personal Control Beliefs

 

Personal control beliefs relate to the level of mastery an individual
feels over life’s outcomes and involve the two psychological constructs of
causal attributions and explanatory style (Reeve, 1996). A causal
attribution is an explanation of why a particular outcome comes to pass
and most often involves the following five attributions: ability, effort, task
difficulty, strategy, and luck (Reeve, 1996). Because any list of causal
attributions is possible, attributions are organized along three dimensions:
1) internal versus external; 2) stable versus unstable; 3) controllable
versus uncontrollable (Reeve, 1996). Internal versus external attributions
involve those that exist within the individual (intelligence) versus those that
exist outside, in the environment (help from others), respectively (Reeve,
1996). Stable versus unstable attributes involve those that are permanent
(personality traits) versus those that are temporary (moods), respectively
(Reeve, 1996). Controllable versus uncontrollable attributions involve
those that the performer can take responsibility (personal effort) versus
those that the performer cannot (luck), respectively (Reeve, 1996).

The three attributional dimensions can be grouped into two
explanatory styles (pessimistic or optimistic) which are cognitive
personality variables that reflect the habitual manner in which people
explain the cause of bad events that befall them (Reeve, 1996, p. 98).
Table 11 illustrates the explanatory styles and related attributional

dimensions within the context of child misbehavior (any behavior or task

78

 

.8? exam sec 8382 .862

 

 

293.0550 o_nm__obcooc3 23:05:08: .m> 223.9280

295mg: 03ma mimic: .m> ofifim

.9:ka .mEQE .manxm. .m> .mEEE

o=m_E_EO o=m_E_mmon_ 229885 .330
23w boamcmaxw

 

 

._o_>mcmnm_s_ EEO GEE—30m
mcoficmEE .mcozantzd. 3.5. m5 c_£_>> nmNtodmumo 23¢, anmcflaxw

 

E. 2an

79

can be substituted here). Hypothetically speaking, parents with optimistic
explanatory styles, attribute their child’s misbehavior to 1) external, 2)
unstable, 3) controllable dimensions and endorse related statements as
follows:

1. When my child is in one of his/her moods, there is usually
something that set him off but can be handled.

2. My child gets oppositional, hostile, or defiant because of situational
circumstances, he can be appeased with appropriate discipline.

3. My child experiences difficulties because of his school or home
environments which can be altered to suit his/her needs.

In contrast, parents with pessimistic explanatory styles attribute their
child’s misbehavior to 1) internal, 2) stable, 3) uncontrollable dimensions
and endorse related statements as follows:

1. My child’s behavior is mostly biological, a big part of his/her
personality, and sometimes, we’re lucky to get through one of
his/her temper tantrums.

2. There is little influential relationship between my behavior and how
my child behaves, especially since s/he has a hostile and defiant
temperament.

3. There is little to be gained from an open, accepting, and supportive
relationship with my child, especially since my child is
unmanageable.

It is this latter, pessimistic explanatory style which predisposes students to
a passive approach to studying and Ieaming and also leads to a helpless
motivational orientation whereas students with an optimistic explanatory
style develop a mastery motivational orientation and persist in the face of
difficulties (Reeve, 1996). Thus, parents with a pessimistic attributional

style may also be predisposed to similar approaches to learning and

consequently, negative outcomes in parent training.

80

Three recent studies discuss the association between parent
attributions and parenting behavior outside the context of PMT. Ford-
Gilboe (1997) examined the impact of motivational aspects (and other
resources) on health promotion among single versus two parent families.
Locus of control, an attributional motivational aspect, defined as the
family’s beliefs about their ability to influence health, was one of the
sources. A second source of motivation was general efficacy defined as a
general belief in one’s capability to successfully perform behaviors across
situations. Both, internal health locus of control and general self efficacy
were positively correlated with participation in health work.

A second study examined mothers’ attributions and general
parenting attitudes and their relation to responses to child noncompliance
(Geller & Johnston, 1995). These authors’ review of the literature
consistently linked parent attributions of child behavior with parenting
responses (Geller 8 Johnston, 1995). Similarly, the study findings also
indicated that mothers’ attributions for both child and parent behavior were
predictive of particular parental responses. In particular, maternal
depressed mood was also associated with more negative attributions for
child behavior, suggesting a link between a well established TEP and a
more pessimistic explanatory style.

Third, in a basic mediational model of parenting children with disruptive
behaviors, Johnston (1996) included the following four parent cognitions: a)

expectations for child behavior; b) perceptions of child behavior; c) attributions for

81

child behavior; and, d) sense of parenting efficacy. Among the four parent
cognitions discussed, the latter two were obviously motivational in nature.
Interestingly, lines of research have examined parent attributions regarding child
intentional (malicious defiance) and unintentional (fidgety, restlessness) child
behaviors, characteristics of CD and ADHD, respectively (Johnston, 1996).
Young adult raters were better able to discriminate between these behaviors
when presented separately than when they co-occurred (Johnston, 1996). Thus,
the raters were found to have difficulty matching their attributions to the
intentionality of the child’s behavior, blaming children for unintentional fidgeting
while, conversely, excusing children for overt defiance (Johnston, 1996). Given
the prevalence for comorbidity among the disruptive behavior disorders, this has
implications for parent attributions and parenting behavior.

If we consider that parent cognitions predict the outcome of parent-child
interactions, these studies support the notion that parent attributions may play a
role in PMT outcomes.

Achievement Stn'vings

 

Students’ adaptive or maladaptive patterns of motivation influence
whether students seek or avoid academic challenges, exert or withhold
effort, persist or give up in the face of failure, and work to develop skills
efficiently (Reeve, 1996). Dweck termed the adaptive motivational pattern
as a mastery orientation and a maladaptive pattern as a learned helpless
orientation (Reeve, 1996). These patterns are not necessarily related to

ability or intelligence but to two types of goals which students adopt in

82

achievement situations (Reeve, 1996). The two types of goals pertaining
to the achievement/competence class of goals (Dweck, 1996) are defined
as follows:

1) Learning goals: students approach academic challenges with the
desire to increase their competence, gain task mastery, improve
knowledge and skill, and understand something new. This
orientation leads students to ask the question, What is the best way
to increase my ability or achieve mastery?

2) Performance goals: students approach academic challenges with
the desire to gain favorable and avoid unfavorable evaluations of
their competence and to measure/evaluate their ability. This
orientation leads students to ask the question, Is my ability
adequate or inadequate?

(Dweck & Leggett, 1988; Reeve, 1996)
Furthermore, perceptions of ability combine with these two goal
orientations to produce different levels of achievement behavior (Dweck,
1996; Reeve, 1996). Specifically, students with high performance goals
and high ability perception sacrifice learning and choose moderate or
moderately difficult tasks to display competence. In response to difficulty,
these students demonstrate a mastery orientation of effective problem
solving. Students with performance goals and low ability perception
sacrifice learning and choose moderately easy tasks to avoid displays of
non-competence. In response to difficulty these students demonstrate a
leamed-helpless response, low quality problem solving strategies,
impaired performance, lowered persistence, and negative affect (Dweck,
1996; Elliot & Dweck, 1988). In contrast, students with Ieaming goals and

either high or low ability choose Ieaming at the risk of displaying mistakes

to increase competence, focus on effort or strategy. maintain positive

83

affect, and demonstrate a mastery orientation of effective problem solving

in response to difficulty or failure (Dweck, 1996; Elliot & Dweck, 1988).
Parents who demonstrate a mastery orientation with Ieaming goals

may endorse statements such as the following:

Making mistakes is a part of Ieaming.

I work hard to learn.

The therapist makes sure I understand the work.

The therapist wants us to try new things.
(adapted from Ames 8 Archer, 1988)

PP’NT“

Altemately, parents who demonstrate a learned helpless orientation
with performance goals may endorse statements such as the following:

Other parents want to know how I am doing on assignments.
I really don’t like to make mistakes.
Only a few parents will get a lot out of this parent training.
I work hard to get the homework assignments done.
Parents feel bad when they do not do as well as others.
(adapted from Ames & Archer, 1988)

WPWN.‘

Parents with exceptionally high scores on performance goals may be
more likely to be obsessed with self-worth protection and public
appearances than with Ieaming goals, i.e., on how to make the situation
better. This is somewhat related to the notion that overall classroom
setting, and possibly school and even community culture may play a role
in students’ goal orientations (Lemos, 1996).
Goal Theories

No published research has been found, to date, which examines
the role that PMT interventions play in parents’ goal orientations.
Dweck (1996) delineates three classes of goals within goal theory

frameworks and patterns. The first class of goals (previously mentioned

84

above) focus on the goals that people adopt in achievement/competence
situations and can be categorized according to Ieaming goals,
performance goals, and work avoidant goals (Brophy, 1998). Learning
goals are adopted when students focus on trying to learn whatever they
need to know in order to develop the skills necessary for the task (Brophy,
1998). Synonymous terms in the literature include learning goals (Dweck
8 Leggett, 1988) and mastery goals (Ames 8 Archer, 1988). Performance
goals are adopted when students focus more on themselves than on the
task, i.e., on their self-perceptions, and on their public reputations (Brophy,
1998). Synonymous terms include performance goals (Ames 8 Archer,
1988; Dweck 8 Leggett, 1988) and ego-involvement or ego-oriented goals
(Midgley, Arunkumar, 8 Urdan, 1996; Nicholls, 1984). Work avoidant
goals are adopted when students refuse to accept the academic challenge
and minimize time and effort to it (Meece 8 Holt, 1998).

The Patterns of Adaptive Learning Survey (PALS) has been factor
analyzed and provides evidence of construct validity and contains items
measuring an ego-oriented goal orientation (performance goals) and a
learning-oriented goal orientation (Midgley et al., 1996). A recent study
(Midgley et al., 1996) examined self-handicapping orientations and
adapted the following five Ieaming oriented items:

1. I feel most successful in school when I learn something I didn’t

know before.
. The main reason I do my work is because I like to learn.

2
3. I like school work the best when it really makes me think.
4. I like school work I’ll learn from even if I make a lot of mistakes.

85

5. Understanding the work in school is more important to me than

the grades I get.

As well as the following four ego-oriented goals:

1.

2.
3.
4.

I would feel successful in school if I did better than the other
students in my class.

I like school work that lets me show how smart I am.

I’d like to show my teachers that I’m smarter than the other kids.
I’d like to show my parents that I’m smarter than the other kids
in my class.

Translating, adapting these items to similar items relating to PMT, the

following items may suggest similar constructs:

Learning Oriented Goals:

 

1.

2.

5.

I will feel most successful in parent training when I learn
something I didn’t know before.

The main reason I am doing my parent training homework is
because I like to learn.

3. I like parent training work best when it makes me think.
4.

I like parent training work I'll Ieam from even if I make a lot of
mistakes.

Understanding the parent training work is more important to me
than getting evaluated (others looking at my behavior and
approving).

Performance Goals:

 

1.
2.
3.

4.

I would feel successful in parent training if I did better than the
other parents in the sessions.

I would like parent training activities that let me show how good I
am at managing behavior.

I would like to show my group leader, therapist, that I am better
at managing behavior than the other parents.

I would like to show my other relatives who are parents (my
parents, siblings, cousins), that I am better at managing
behavior than other parents.

Two possible dimensions of performance (ego-oriented goals) have

been identified (self-enhancing and self-defeating ego orientations) and

analyzed separately (Skaalvik, 1997). The above examples of

performance goal statements (Ames, 1988; Midgely et al., 1996) define

goals in terms of ego-enhancing ego orientations. Students with strong

86

self-enhancing ego orientations adapt performance goals which strive to

demonstrate superior abilities, do better than others, or outperform others

(Skaalvik, 1997). Students with strong self-defeating orientations adapt

performance goals which strive to avoid looking stupid and being

negatively judged by others (Skaalvik, 1997). Factor analysis studies

identifies the following statements as related to self-defeating ego

orientations:

1.

:‘f-SON

N991

When I answer questions in class, I am occupied by how I am
perceived by other students.
When lam working on the blackboard I am concerned about what
my classmates think about me.
At school lam concemed not to make a fool of myself.
When I give a wrong answer in class lam most concemed about
what my classmates think about me.
The worst thing about doing mistakes at school is that other
students may notice.
At school, it is important for me to avoid looking stupid.
At school, I try not to be among the poorest students.

(Skaalvik, 1997)

Applied to PMT and parents, the above statements are reformulated:

1.

2.
3

Whenever I talk in group sessions, lam occupied (I wonder) by
how I am perceived by other parents.

When [participate in the group activities, I am concerned about
what my peers think about me.

. When interacting with my child in public places, I am concerned not

to make a fool of myself.

Whenever I talk about how unsuccessful homework assignments
have been at home, I am most concerned about what other parents
think about me.

The worst thing about making mistakes at home and discussing
them in groups is what other people may notice.

At home, it is important for me to avoid looking stupid.

At group sessions, I try to avoid being among the poorest
students/parents.

87

Students, preoccupied with social comparisons and highly concerned with
not being perceived as stupid, are particularly vulnerable in achievement
situations (Skaalvik, 1997). Thus, parents, similarly preoccupied, may be
predisposed to negative outcomes in PMT programs.

Dweck's (1996) second class of goals, interpersonal or relationship
goals, contain two subclasses of goals which are analogous to learning
and performance goals. Developmental goals involve a focus on building
relationships, seeking intimacy, helping others, and seeking to improve
one’s relationship skills and are more likely to produce mastery-oriented
responses (focus on attributions to controllable factors, maintenance of
positive affect, and increased self-disclosure) in the face of failure or
rejection (Dweck, 1996). Approval goals involve a focus on striving for
validation, seeking approval, acceptance, and are more likely to produce
helpless responses in the face of failure or rejection (Dweck, 1996).

Dweck's (1996) third and fourth classes of goals are viewed as
social dominance and control goals and as hedonic goals (focus on
pleasure or pain or on concrete rewards or punishment), respectively.
When the focus is on dominating, controlling, or seeking power over
others, control goals have been shown to be predictive of hostile
attributions and aggression (Dweck, 1996). All goals are described as
universal in that any individual is likely to pursue all classes at various

times (Dweck, 1996). However, different classes have been shown to

88

create different vulnerabilities when certain goals predominate (Dweck,
1996)

Other theorists have reformulated and discussed other types of
goals, other than strictly achievement oriented have been described in
Ieaming situations. For example, students may adopt up to seven highly
distinct types of goals in order of descending priorities: academic
orientations (working, evaluation, and learning goals), social and
interpersonal orientations (complying, discipline, and interpersonal
relationship goals), and enjoyment orientations (enjoyment goals) (Lemos,
1996). Moreover, these student goals appear to be consistent in that they
bring to the classroom setting a general goal orientation (e.g. working) and
implement specific goals when faced with the real setting (Lemos, 1996).

Performance goals were apparent in the phrases used by parents
describing their experiences and outcomes of participating in parent
education classes (First 8 Way, 1995). Although this phenomenological
study did not discuss parent responses within the context of goal theory,
parent responses interpreted as belonging to the category of happier
outcomes, more confidence, and feeling like “good” parents contained
elements of self enhancing performance goals. For instance, parents
made statements such as “People say. . .here’s somebody we can look up
to. . “I ’ve made a lot of progress..somebody look up to me for what I ’ve
done." “I ’m trying hard to put them [mother qualities] to use and let other

people know that I do have them. . and “my kids will be proud of me”. In

89

this study, transformative Ieaming in which parent training contributed to a
major change in the parent’s lives, was described as an overarching
explanatory theme in parent responses (First 8 Way, 1995).

Implicit Theory

 

Dweck et al. (1991) examined individuals with similar abilities who
set either learning or performance goals and suggested that these
individuals have different implicit theories about the nature of the
intelligence. Individuals who are entity theorists view intelligence (and
other traits, attributes) as a fixed entity which cannot be changed and over
which they have no control (Hong, Chiu, Dweck, 8 Sacks, 1997). Those
individuals who are incremental theorists view intelligence (and other
qualities) as incrementally responsive and malleable to an individual’s
effort (Brophy, 1998; Hong et al., 1997; Levy, Stroessner, 8 Dweck, 1998).
This implicit theory of intelligence has been used to explain why some
individuals favor performance goals over Ieaming goals and vice versa
(Dweck 8 Leggett, 1988).

In addition, the implicit theories have been applied to social
domains and generalized to external attributes. When individuals adapt
entity theories in which other’s attributes are viewed as fixed or
uncontrollable, they are predisposed towards judgment goals which
influence behavior (low initiation and resistance towards change),
cognition (rigid, over-simplified thinking), and affect (contempt) (Dweck 8

Leggett, 1988). Furthermore, individuals who are entity theorists are more

90

likely to give up in the face of difficulties (Dweck 8 Leggett, 1988). Recent
research has implicated the role of entity theory as predictive of biases
towards social stereotyping (Levy et al., 1998), prejudices, evaluative
processing (Hong et al., 1997), and increased emphasis on evaluative
processes rather than processing and contextualizing factors which
mediate behavior (Heyman 8 Dweck, 1998).

When individuals adapt incremental theories in which others’
attributes are viewed as malleable, they are predisposed toward
developmental goals which influence behavior (mastery oriented goal
pursuit), cognition (process analysis), and affect (empathy) (Dweck 8
Leggefl,1988)

Hong et al. (1997) used three domain-general items of implicit
person theory to measure a diverse set of skills as follows:

1. People can do things differently, but the important part of who they
are can’t really be changed.
2. The kind of person someone is, is something very basic about them
and it can’t be changed very much.
3. Everyone is one kind of person and there is not much that can be
done to really change that.
(Hong et al., 1997, p. 303).
In order to depict a unitary theme, avoid monotonous repetition, highly
compelling, and socially desirable responses only three items pertaining to
entity theory were used while explicitly incremental theorist type items
were not depicted. Apparently, this practice was well substantiated by

research (Chiu, Dweck, Tong, 8 Fu, 1997; Hong et al., 1997).

Furthermore, disagreement with the entity theory items can be taken to

91

represent an incremental theory (Chiu et al., 1997). Previous validation
studies have found the scale to have high internal reliability, test-rest
reliability, and construct validity (Hong et al., 1997). Furthermore, the scale
does not correlate with measures of intelligence, academic aptitude, self
esteem, optimism regarding human nature, positivity about the self and
others, and ideological rigidity (Hong et al., 1997), self presentation
concerns, and political attitudes (conservatism and liberalism scales)
(Chiu et al., 1997). These findings suggest that the items are not
confounded with these other constructs.

A more indirect method of assessing implicit theories would be to
ask individuals to explain positive or negative behaviors. For example,
individuals would be told “Alexis stole some bread from the bakery shop”
and then asked to make causal attributions by completing the following
sentence stern: “This probably occurred because...” (Hong et al., 1997). It
has been found that entity theorists generate significantly more personality
trait responses such as “Alexis is a thief’, “Alexis was dishonest” while
incremental theorists tend to generate more process oriented,
psychological state explanations such as “Alexis was hungry’, “Alexis was
desperate” (Hong et al., 1997).

Chiu et al. (1997) studied the association among implicit theories
(entity or incremental) and moral beliefs (duty based or rights-based). The

authors found that entity theory and moral character were associated with

a preference for duty-based moral beliefs which also served to maintain

92

the status quo. In other words, these individuals view other people’s moral
character as fixed, consider duty as the basis to judge morally right
behavior, are less tolerant of deviance, preserve the status quo, and, are
more likely to sanction behavior, use punishment, and not reward
appropriate conduct (Chiu et al., 1997). Incremental theories were
associated with a preference for rights-based moral beliefs which served
to allow, promote, and guide change. Individuals who are incremental
theorists view other people’s moral character as malleable, shapeable,
cultivated, and improvable; consider fundamental principles and rights as
the basis to judge behavior; and, are more likely to use positives,
appreciate appropriate behavior, use more negotiation and education, and
use less punishment. Within the conduct domain, entity theorists were
more likely to use directives while incremental theorists were more likely to
use negotiation and assurance when talking to children with undesirable
behavior.

The influence of implicit theories are likely transferable in parent
training and behavioral outcomes. Using the Chiu et al. (1997) measures
of morality theories as a framework, the following statements may be
viewed as equivalent in the parent training framework:

1. My child’s behavior is something very basic about him/her and it

can’t be changed much.

2. Whether my child follows directions, obeys instructions, and
acts appropriately or not is deeply ingrained in his/her
personality. It cannot be changed much.

3. There is not much that can be done to change my child’s

behavioral traits (e.g. oppositional, aggressive, defiant)
(adapted from Chiu et al., 1997)

93

Just as implicit theory of personality is an individual difference factor that
mediates automatic attitude activation towards others (Hong et al., 1997),
implicit theories may serve as mediators in parent training outcomes. No
studies were found which addressed parent’s implicit theories and PMT.
The Self

Self schemas are defined as domain specific and as “cognitive
generalizations about the self, derived from past experience, that organize
and guide the process of self-related information contained in the
individual’s social experience” (Reeve, 1996, p. 139). These self schemas
serve several valuable and useful functions: a) promote the ability to
encode, evaluate, and retrieve information; b) provide structure for quick
and confident judgments; 0) allow flexible adaptation to information
processing goals; d) enhance sensitivity and attunement to schema
relevant information; e) direct close attention to and favor information
relevant to that domain; and, consequently, f) provide individuals with the
ability to exercise the skills/abilities when needed (Brophy, 1998; Cross 8
Markus, 1994).

Individuals who have developed self schemas are said to be
schematic in a domain. Schematic individuals recognize that they have a
particular ability (in that particular domain) and believe that this ability is of
critical importance which together produce their self schemas (Cross 8
Markus, 1994). Individuals are said to be aschematic when they do not

possess a self schema in a particular domain (Cross 8 Markus, 1994).

94

Aschematic individuals do not recognize that they have a particular ability
(in that particular domain) and do not believe that this ability is of critical
importance, (or at least at lower levels) (Cross 8 Markus, 1994). It has
been postulated that aschematic individuals may be less likely to
anticipate, recognize, and employ the relevant abilities and strategies
needed for completing tasks, especially when tasks are novel and require
considerable structuring (Cross 8 Markus, 1994).

Self schemas have been measured through self ratings on domain
specific direct questions as follows:
How good are you at analytical problem solving?
How good are you at logical problem solving?
How important is analytical problem solving to you?

How important is logical problem solving to you?
(Cross 8 Markus, 1994)

99°F)?"

The function and organization of self schemas have been examined in
several domains such as problem solving and competence (above),
dependence-independence, gender roles, shyness, depression, career
orientation, and Type A personality (Cross 8 Markus, 1994). Students who
were schematic and aschematic performed equally well on a problem
solving task yet demonstrated variable judgements regarding perceptions
of facility and enjoyment of the task (Cross 8 Markus, 1994).

No attention has been given to the consequences of schematicity in
parenting performance with children with disruptive behaviors. Yet, these
same questions can be directly addressed through domain specific, simple

modifications:

95

How good are you at controlling your child’s behavior?
How good are you at managing your child’s behavior?

How important is it to you to control your child’s behavior?
How important is it to you to manage your child’s behavior?

PPN.‘

These questions may also be modified to be even more domain specific or
setting specific (e.g. how good are you at controlling behavior in the
home?)

Self schemas change through discrepant feedback and/or through
future oriented self conceptions which are called possible selves (Cross 8
Markus, 1994). Markus and Nurius (1986) developed the concept of
possible selves which are cognitive representations of oneself in the
future" (Brophy, 1998, p. 216). Possible selves represent those selves the
person could become, would like to become, or is afraid to become (Cross
8 Markus, 1994, p.424). With the creation of more vivid and elaborated
possible selves, individuals tend to perform better toward desired
outcomes (Cross 8 Markus, 1994). Furthermore, being able to view
oneself as “successful" within a particular domain in the future may also
help reduce anxiety or worry during the activities (Cross 8 Markus, 1994).

Integrating Inner Motivational Resources

 

Many of these inner resources can be organized within the
expectancy x value model of motivation. The expectancy x value model
predicts four primary student responses to classroom activities. First,
students react with engagement when they see value to a task and they
are confident they can do it (Brophy, 1998). Second, students react with

dissembling when they recognize the value but do not feel capable of

96

meeting the task demands (Brophy, 1998). Third, students respond
through evasion when task value is low yet they feel capable of successful
performance (Brophy, 1998). Finally, students respond with rejection when
task value perceptions are low coupled with low success expectations
(Brophy, 1998). These same student strategies for responding may be
applicable to parents engaged in parent training programs (Table 12).
Dissembling parents and evading parents may differ only in program
completion; for example, dissembling parents may complete the program
but find excuses such as time constraints while evading parents may
simply drop out.

Obvious parallels can be drawn among the various motivational
aspects of self efficacy theory, attribution theory, personal control beliefs,
goal setting, value, etc. For example, Pintrich and De Groot (1990)
reported that self reports of self-efficacy and value perceptions (interesting
and important course work) were positively related to student performance
on classroom English and Science assignments.

Many causal attribution theorists have focused on individual’s
causal attributions regarding performance on tasks and have neglected
value aspects inherent in many tasks (Brophy, 1998). The interdependent
relationship with self-efficacy and personal control expectations and
parent causal attributions addressing task value aspects of motivation may

be revealed by answers to the following open ended questions:

97

 

.89 {seem 52.. 85m}. .902

 

8.593 35883 .9 beam
.205 among; ”:29QO 9:0...
9 9:58. o~=9ocom 9 29:.
”m9_>=om 96.6 octane m>zom
EoEmmmmcm

GEE 92 :09 $3333

£3, =$525.0: new BEE
”8:99.88 ho mums: 9099a
0.538.0me

So 9:306

L9 xm: :9; ”3:039:02

Eomnm 8 992.8

-E ”8255,: 9: moon
858$ :9:

So @5306

.9 xm: :9; 6956

E cozmgofima .mE_:_E
20.98me 33

 

=9:
m:_m>

mcozmfimaxw
26.. $825

 

 

mcozaoohmn. o2m> x >oc90maxw :9: 9

 

o9m_mm mm ten. 9 ucficoommm .9 $5995 Ewan oo~_mo£oa>I

NV 92m...

98

1. What benefits do I expect to derive from engaging in this task?
2. Why do I seek lavoid opportunities to learn about parent
management?
3. What does parent management training do/fail to do for me?
(adapted from Brophy, 1998; Brophy, 1999).
Subjective task value may have three major components as described by
Eccles and Wigfield (1985). The first, attainment value, involves achieving
success for enhancing self concept or fulfilling achievement, power, or
prestige needs. Thus, parents may sustain interest and appreciate PMT
involvement in order to meet certain personal needs. The second, intrinsic
or interest value, involves engaging in tasks because of the enjoyment of
involvement. Translating this concept of intrinsic value into PMT
involvement, parents may also engage in PMT because of the social
aspects. These parents are interacting with a reference peer group where
they can make social comparisons. The third component, utility value,
involves engaging in the task in order to reach certain goals. In the PMT
case, this may involve the parental perception that PMT will aid in meeting
their goals as explicitly articulated at the start of therapy, though
sometimes, implicitly addressed. These three task value components may
be reflected in the following statements:
1. Involvement in PMT will help me learn how to manage my child’s
behavior and help me feel better about myself.
2. Involvement in PMT will help me Ieam how to manage my child’s
behavior and have better control and power.
3. Involvement in PMT will help me Ieam how to manage my child’s
behavior and other’s won’t look down on me so much, anymore.
4. Involvement in PMT will be interesting and possibly fun, as I will

meet other parents in similar situations.
5. Involvement in PMT will be useful in helping me to reach my goals.

99

F”

Involvement in PMT will be helpful in improving the quality of our
lives.

I think that what I am Ieaming in this group is useful for me to know.
I think that what we are learning in this class is interesting.

It Is important for me to Ieam what Is being taught In these groups.
0. I like what I am learning In these groups.

(Pintrich 8 De Groot, 1990)

doefl

Support for considering the value aspects of motivation within PMT
contexts and antisocial populations comes from the work of (Kazdin et al.,
1997; Prinz 8 Miller, 1994). Kazdin et al. (1997) examined barriers to child
and/or parent treatment and found that parent perceptions of treatment
irrelevance was associated with premature termination from therapy.

Prinz and Miller (1994) found that parents who had been given
opportunities to discuss feelings about being in therapy (and other
personal issues) showed lower drop out rates. Results from both of these
studies suggest that parent perceptions of value may be predictive of PMT
outcomes.

Research has effectively established the interrelationships among
sense of efficacy or expectations for success and the following: causal
attributions for success and failure, learning versus performance goals,
selection of strategies for accomplishing tasks, managing failure, and
handling frustration (Brophy, 1999). Within specific content domains,
research has established relationships among various motivational
variables and perceptions of competence. For example, in comparison to
students who rated their competence as low in mathematics, students who

rated their competence highly also demonstrated learning and mastery

100

goals, more positive emotions, greater enjoyment, and fewer negative
emotions (Stipek, Salmon, Giwin, 8 Kazemi, 1998). Within the parenting
domain, a more pertinent study finding involves the association between
mothers’ feelings of self efficacy and attributions of personal controllability,
i.e., beliefs about their ability to impact child misbehaviors (Geller 8
Johnston, 1995).

Goal theorists have extensively researched motivational factors
(expectations, self-efficacy, attributions) and their influence on individual’s
engagement in tasks and ultimate levels of achievement success (Brophy,
1998). This inter-relationship and mediating influence of goals can be
seen in the functional definition of goals:

Serving as filtering mechanisms for beliefs about the self and

subsequent action and as setting the stage for eliciting motivational

cognitions (such as values, expectations, attributions, and self-
evaluations), particular cognitive operations (mental
representations, monitoring, evaluation of responses and strategic
thinking), and affects about the self, and the tasks and others.
(Lemos, 1996, p.152)

Research has begun to establish the interrelationship among
classes of goals as follows: goals play a critical role in setting up
individual’s attributions; self schemas and possible selves are beginning to
be discussed in relation to people’s classes of goals; the nature of the
relationship between self efficacy, related variables (competency). and
goals is being explored (as noted above) (Dweck, 1996).

Research has yet to establish empirically the connection between

the value, interest, and appreciation aspects of motivation and

101

engagement in domain specific Ieaming activities as well as ideas about
engagement in activities (Brophy, 1999). Theoretically, (Dweck, 1996)
suggests that attitudes (including values and evaluations) translate into
behavior when housed within goal structures; the more value one gives to
something the more it has the power to drive goals. At the very least,
these motivational aspects and value aspects may interact to predict PMT
outcomes. Ideally, all of these aspects comprise the major elements of a
cognitive theory of motivation and may be structurally combined in a
complex motivational schema which impacts on PMT outcomes.

In a model, described as Schema Mediators, the present
dissertation would like to build on the above conceptualizations of inner
motivational resources and value aspects and incorporate the various
components into an integrated model with interrelated parts. Like students
who are high in motivation to learn, parents who are high in motivation to
Ieam may also possess a motivated Ieaming schema which is
triggered/activated whenever they enter a PMT program (Brophy, 1998).
Those parents with most successful outcomes, ideally would possess a
motivation to Ieam schema described as follows:

A network of connected insights, skills, values, and dispositions,

that enable [parents] to understand what it means to engage in

[PMT] activities with the intention of accomplishing their Ieaming

goals and with awareness of the strategies that they use in

attempting to do so.
(Brophy, 1998, p. 14)

Furthermore, these schema are instantly accessible, guide information

processing and problem solving, are accumulated from prior knowledge

102

and experience, and tend to be domain or situation-specific (Brophy,
1999). Those with more elaborated and better developed motivational and
cognitive schemas are more likely to appreciate and benefit form
engagement in more abstract and complex Ieaming activities (Brophy,
1999)

Just as Brophy (1998) points out that the same principles that apply
to student motivation and achievement, apply to teacher motivation and
professional challenges, the same may apply to parent motivation and
PMT outcomes. Furthermore, it has been suggested that “motivated
functioning in [their respective] domain[s] is schema-driven” (Brophy,
1999). It would be worthwhile to examine how these motivational
resources combine to influence parent outcomes in PMT.

Two additional, final points need to be addressed. First, many of
the concepts noted above have been essentially dichotomized and
described as mutually exclusive dimensions such as in task (learning)
versus ego (performance) goal orientations even though evidence
suggests the two goals are interdependent (Ames 8 Archer, 1988; Meece
8 Holt, 1993). Other examples involve studies which examine high
efficacy versus low efficacy, internal versus external attributions. Unlike
the above studies, future studies plan on evaluating the individual
motivational inner resources on a continuum and acknowledge that an
entire spectrum of motivational factors may exist in evaluating various

levels of the concepts.

103

Second, the various parent motivational resources throughout the
PMT interventions may not be static constructs. It has been argued that
“motivation in therapy is not a personal characteristic, stable trait,
something a person does or does not have, or a driving force channeled
by biology and experience” (Kanfer, 1996, p. 2). As such, it is quite
possible that parent motivations in PMT may be subject to a process of
continual “refinement” and adjustment to environmental conditions over
the course of treatment. Research findings suggest that as parents are
involved in the PMT process, their cognitive processes may go through
various phases and changes as they work to cope with their child’s
disruptive behavior (Spitzer, Webster-Stratton, 8 Hollinsworth, 1991).

Johnston (1996) speculates that as parents progress through these
phases, different cognitions (including motivations) may be central at
different times. Furthermore, parent motivations, independent of PMT
participation, may also be subject to the same processes. Consequently,
parent motivational resources and value aspects may change.

Two areas of research have been drawn upon for this dissertation. One
has to do with traditionally examined predictors which are usually consistent with
demographic information about parents and children and their impact on PMT
outcomes. The second has to do with conditions “inside” the parent. The concern
here is predominantly with the parent preferred style of interacting with their child,
nature of parent goals and values, and degree of responsibility for affording

change in their child. This second area can be expanded to include parent

104

motivational resources and values and their influence on outcomes. To consider
only those conditions which are internal or external to the parent and child in
isolation is insufficient. The challenge is to consider several predictors and
eventually find ways of analyzing and predicting influences of various
combinations, and to ultimately develop treatment plans for better outcomes.

The various nine motivational inner resources (self efficacy, optimistic
explanatory style, pessimistic explanatory style, Ieaming goals, self enhancing
performance goals, self defeating performance goals and self schemas) in
conjunction with valuing aspects (interest, appreciation, importance) are
integrated into a smaller number of cohesive motivational schemas in the pilot
study below. The pilot study addresses the first question related to motivational
resources and provides the impetus for the second question which is examined
in the main study:

1. Do the motivational resources and value aspects covary to form central
schemas (4 or 5) among individual parents? And if so,

2. Are these schemas predictive of parent training outcomes?

105

PILOT STUDY
THE DEVELOPMENT OF THE PARENT MOTIVATION SCHEMA SCALE

The Parent Motivation Schema Scale (PMSS) is designed to uncover
particular parent motivational aspects and values which may eventually
contribute to, if not determine, outcomes in parent management training. Nine of
the twelve inner motivational resources are reflected in the PMSS. These
cognitive motivational resources include self efficacy, attributions (divided into
optimistic and pessimistic explanatory styles), learning goals, performance goals
(both self-enhancing and self-defeating), implicit theories, self schemas, and
value. The purpose of study 1 was to establish the reliability, internal validity, and
factor structure of the PMSS.
METHOD
Participants

The pilot survey participants were 200 parents (168 mothers, 3O fathers, 2
unknown) and their children (78 girls, 113 boys, 9 unknown). Families were 91 %
White, 6% Native Canadian, and 1 % Asian Canadian. The children’s mean age
at the time of the survey was 9.25 (SD=4.50).

The treatment and control participants were 93 parents (69 mothers and
24 fathers) and their children (29 girls and 64 boys). Families were 81 "/0 White
and 16 % Native Canadian. The children’s mean age at the point of the treatment

study was 8.23 (so=2.35).

106

Parent gender, education, current annual household income and child
gender as well as child education for all participants (pilot survey, treatment, and
control) are summarized in Tables 13 and 14, respectively.

The group of parents who served as participants live in a small, northern,
Canadian community (population = 80,000). The participant demographic data is
representative of this population.

Measure

In this first study developing the PMSS, items are adapted from previous
factor analysis research. Items are adapted from various instruments with stable
factor structures used to assess student motivation aspects including self efficacy
(Midgley et al., 1989; Pajares, 1996; Reeve, 1996), goal theory (Ames 8 Archer,
1988; Midgley et al., 1996; Skaalvik, 1997), implicit theory (Cross 8 Markus,
1994; Hong et al., 1997), and values (Brophy, 1999; Pintrich 8 De Groot, 1990).
Factor analysis had originally been used to guide the scale construction of each
of these motivational aspects within their respective research studies. Within the
present study, these items are translated and adapted to fit the specific parenting
domain as opposed to the initial content domains of the original studies.

The domain specificity of the PMSS items are based on the principal
investigator’s clinical experiences with PMT. Parents have offered statements in
therapy, written comments in homework assignments, made notes in journals,
and reflections in behavioral diaries, which implicate specific motivational

resources. The domain specificity of these item responses are generated from

107

Table 13

Demcmraphic Characteristics of Parent Participants in PMSS Pilot Study

 

 

 

 

Pilot PMT Groups Total
(n=200) (n=93) (n = 293)

Descriptive fl % Ll °/o _I\_l %
Parent gender

Male 30 15 24 26 54 18

Female 168 84 69 73 230 78
Parent education

Elementary 4 2 3 3 7 2

High School 52 26 32 34 81 28

College (Community) 64 32 38 40 100 34

University (undergraduate) 47 24 13 14 60 20

University (graduate) 28 14 5 5 33 1 1

Other 2 1 0 0 2 1
Ethnicity/ Race

Native Canadian or Inuit 12 6 15 16 27 9

Asian-Canadian 2 1 0 0 2 1

White / Caucasian 182 91 76 81 252 86

Current Annual Household Income

Under $10,000 4 2 10 11 13 4
$10,000 - 14,999 4 2 18 19 21 7
$15,000 — 19,999 3 2 1O 11 11 4
$20,000 - 24,999 11 6 10 11 20 7
$25,000 — 29,999 10 5 1 1 11 4
$30,000 — 34,999 6 3 1 1 7 2
$35,000 - 39,999 12 6 6 6 15 5
$40,000 - 44,999 10 5 3 3 12 4
$50,000 - 54,999 12 6 7 7 19 7
$55,000 - 59,999 11 6 8 9 18 6
$60,000 — 64,999 21 11 4 4 26 9
$65,000 — 69,999 23 12 6 6 3O 10
$70,000 and over 61 31 8 9 67 23

 

Note. Totals do not sum to 100% due to missing data.

108

Table 14

Demographic Characteristics of Children Whose Parents Completed the Pilot PMSS m=293)

 

 

 

fllo_t mtment a_nd Control Totals
Characteristic g % fl % N %
Child Gender
Male 113 57 63 67 172 59
Female 78 39 29 31 105 36
Child Education
Preschool 23 12 2 2 24 8
Junior Kindergarten 6 3 10 11 11 4
Senior Kindergarten 11 6 3 3 14 5
Grade 1 20 10 15 16 34 12
Grade 2 11 6 6 6 16 5
Grade 3 16 8 14 15 28 10
Grade 4 12 6 12 13 25 9
Grade 5 15 8 11 12 24 8
Grade 6 9 5 14 15 22 8
Grade 7 12 6 3 3 15 5
Grade 8 7 4 1 1 7 2
Grade 9 16 8 O 0 17 6
Grade 10 10 5 O O 10 3
Grade 11 6 3 O 0 6 2
Grade 12 9 5 0 0 9 3
Grade 13 / OAC O O O 0 5 2
Other 2 1 1 1 2 1

(special education classes)

 

Note. Totals do not sum to 100% due to missing data.

109

these sources as well as from past informal discussions with professionals
experienced in the field of PMT.

Statements are written to measure parent motivational aspects and values
regarding both a) their children’s behavioral difficulties and b) involvement in
parent management therapies. Some of the values of parents are transformed
into a set of futuristic statements reflecting perceived usefulness, importance,
and relevance of PMT. As such, explicit expectations are unavoidably combined
with value statements.

Response Scale

The PMSS is written in English and can be self-administered in
approximately 15-35 minutes. Parents are required to rate how they feel about a
set of 44 statements, on a scale of 1 to 5, where 1=not true of me at all, 2=just a
little true of me, 3= somewhat true of me, 4 =very much true of me, and 5 =
extremely true of me. The statements are divided into nine subscales based
upon motivational aspects (self efficacy, optimistic explanatory style, pessimistic
explanatory style, Ieaming goals, self-enhancing performance goals, self-
defeating performance goals, implicit theory, self schema/schematicity, and
value). Points earned for responses to each subgroup of statements are
combined yielding a total score for each of the motivational aspects subscales.
High scores on seven of the subscales indicate corresponding high levels of self
efficacy, optimistic explanatory style, pessimistic explanatory style, learning goal
orientation, self enhancing performance goal orientation, self defeating

performance goal orientation, and value regarding PMT. For these seven

110

subscales, low scores indicate lower levels on these seven motivational
resources.

The remaining two subscales (implicit theories and self schemas) are
essentially dichotomized scales. A high implicit theory score indicates parents
may have more of an entity theorist orientation whereas a low implicit theory
score suggests more of an incremental theorist orientation. Similarly, a high
score on self schemas items indicates a parent who is schematic whereas a low
score indicates a parent who is aschematic. These two subtypes are phrased in
a positive fashion based on arguments posed in previous research (Chiu et al.,
1997; Hong et al., 1997).

Items from each of the nine subscales are randomly distributed throughout
the scale to reduce the potential for any further biased responding such as
selecting the same types of responses for clustered similar items. Several
motivational aspects are likely to be present simultaneously with varying degrees
of intensity and /or even struggling for dominance. If parent responses within a
particular subgroup of motivational aspects are inconsistent, the response
pattern is said to be nonconvergent and will not likely load onto the same factor.
It is hypothesized that the practical application of parent motivational aspects lie
in application to PMT practice and approaches and not so much for the purpose
of categorizing or labeling parents.

Contributions towards VaMy and Reliability

Validity data includes content and construct validity, (convergent and

discriminant). Reliability data includes split-half reliability and test-retest reliability.

The development of the PMSS is characterized by four main phases: 1) finding
111

evidence to support the item content and content validity; 2) administering the
scale to a pilot group; 3) conducting a factor analysis of the scale; and, 4)
evaluating the reliability of the scale).

Phase 1: Establishing item content:

First, the survey designer developed a comprehensive list of potential
statements for inclusion in the PMSS. List items were generated by the principal
investigator based upon a review of the professional literature on key
motivational inner resources and values as well as clinical experiences (both
described above). The 55 potential items were randomly ordered in a
questionnaire (see Appendix A for Pilot Study: Pilot Survey).

Second, individual items of the PMSS were checked for agreement by the
principal investigator and a professor who is an active researcher in the area of
motivation. The professor identified motivational aspects with corresponding item
statements in a checklist fashion (see Appendix A for Pilot Study: Agreement
Among Independent Rater Responses). For this portion of the second phase,
items were retained or discarded based on the following simple decision rule: a)
items in which both raters agreed (100%) were retained and, b) items in which
raters did not agree were discussed in order to retain an adequate sample of
motivational resources. By retaining only these items in which the independent
raters agreed are an adequate and representative sample of the respective
motivational aspects, then the scale can be said to have support for content
validity (see Appendix A for Pilot Study: Scale Items Listed by Domain). Based

on these reviewers’ ratings, items were revised and deleted from the scale and

112

the resulting 44 item scale was pilot tested (see Appendix A for Pilot Study: Final
PMSS).
Phase 2: Administering the Scale to the Pilot, Treatment, and Control Groups;
This second draft was administered to the participants. The validity
and reliability of the PMSS was tested on a sample of parents of children
who receive services in an out-patient health clinic and/or attend public or
separate school boards, each mandated by the Ministry of Health and
Ministry of Education, respectively. Parents were asked to read a consent
form indicating their voluntary participation in a research project prior to
completing the survey (see Appendix B for Pilot Study: Consent Form).
These surveys were collected by randomly surveying parents as they
attended a clinic or through a mail out survey to clinic members who
attended the Communication Disorders Department over the past two
years.
The PMSS surveys completed by the treatment and control parents
were collected during the treatment phase of the study. These surveys
were independently analyzed statistically for phases 3 and 4 of the pilot
project. On the basis of the interpretability of the results it was decided to
pool together the surveys from both samples (see Appendix C for
Decisions on Factor Dimensions).
Phase 3: Evaluating the Factor Stmcture of the Scale
Each of the items contributing to the respective motivational aspects were

also summed for a total subscale score. Cronbach alphas were also calculated

113

for these subscale scores. A principal component analysis was conducted on the
subscales to adjust for a smaller sample size, if necessary.

A principal components analysis was then conducted on the individual 144
PMSS items. A varimax rotation was used to determine the number of factors
necessary to describe the data. Pearson correlations were used to assess the
stability of the PMSS. Cronbach’s alphas were calculated for the PMSS total and
factor scores.

Phase 4: Evaluating Reliability of the Scale

The Reliability of the PMSS was tested through internal consistency,
(coefficient alpha reliability). Alpha reliability analyses were conducted with data
from the above parents already participating in therapy in the clinic or school
board. Alpha reliability analyses were also conducted by randomly selecting
patients with appointments throughout the clinic, to complete the questionnaire.

Test-retest reliability analysis of the PMSS was evaluated on a smaller
sample, by using separate and combined responses from the treatment and
control groups of the research study at the pre- and post- intervals. In this
procedure, the same PMSS was administered to the same group of participants
on two different occasions and then the two sets of scores were correlated to
obtain sets of scores. The number of weeks between the first and second
administration of the PMSS ranged from a minimum of 9 weeks to a maximum of
16 weeks. Test re-test reliability coefficients for each of the subscales, total
scores, and factor scores were examined. These reliability coefficients indicate

the degree of stability (consistency) in scores over time and is also referred to as

114

the coefficient of stability. The recommended reliability level is .80 for research
purposes.
RESULTS

The mean PMSS total score was 126.32 (SD =16.33). The PMSS total
scores ranged from 70 to 192. Coefficient alpha for the total items of the PMSS
was .87. The subscale mean scores are presented in Table 15. These results
demonstrate the ability of the PMSS to reflect variability in parent motivation and
provide a heterogeneous sample for principal component analysis of the scales.
Table 15 also contains an evaluation of the correlations among the subscales as
well as a reliability analysis for each of the subscales which comprise the PMSS.
Several subscales are significantly correlated (Table 15). The item reliabilities
range from a high of .91 for the Value Subscale to a low of .32 for the Pessimistic
Explanatory Style Subscale (Table 15). However, the factor structure of the
PMSS did not support the nine subscales reflecting the motivational resource
literature (see Appendix C for Principal Component Analysis Results for Nine
Factor Solution).

The principal component analysis of the individual 144 PMSS items
yielded five factors based on using a scree plot decision rule. The elbow rule in
the scree plot criteria, supports extracting five factors (Figure 1). In addition, all of
these factors have eigenvalues substantially greater than 1 (Table 16). All of the
144 items of the PMSS and their factor loadings are presented in Table 17. Only
the items with the highest factor loadings on the five factors are presented in

Table 18.

115

90.0..
So.vm .

.0629 n ..<> ”Eoszmcow u Iow ”0959: =0__QE_ u as: ”655990 :09 28m 8cmEotoa u Don.
meocmzco :59 053 858.950 u de ”a_moa mEEmm. u 04 ”30:35.9 a_bm b92398 029E686 u wmn.
”3:250:99 a_bw boumcmaxm 039E300 n mmo $0850 :60 n mm .4. 56:890. 5 .090: Sn mE9_ 9:..902

 

 

 

N00 90.0 9.0 9.0 :0 90.0 0N0 N00 N00 000 2%
00.0. 00.9 00.9 9N 00.0 00.0 00.0 0N9 00N 2N 0000300 0.00090
N009 N090 0N0 00.0 00.09 00.0 5.8 00.N 00.9 8.8 000.2 0.888
N9 3. 09 3 0N NN 0m 0 9 0N 00_0> 505082
0N : N 0 0 0 0 N 0 9 23> 52552
3: 3 N 0 0 0 0 N 0 0 0&9. .0 .02
00.9 30. N00- .000 .000 .000 .N0 000 N00- 829

00.9 90.0- .NN.0 00.0- 000 N0. .000 .30 26005050..

00. 9 .0N.0 .0N.0 9.0- . 90.0 .0 90. N00- 8:09: 5:05.

00. 9 .000 .0N.0 .0N.0 .0N.0 L 9.0 mm L 0.80 :08

00. 9 00.0 .N0.0 :m 9.0- 30. om \ 280 Ea

00,— 00.0 .0N.0 .0N.0 0.86 0558..

00.9 90.0- 5.0- 0.5 035588

009 .000 035 0005000

00. 9 68% =8

.909 3<> row as: mméa 00-08 6.. mma 30 mm 0.888

 

 

 

 

 

 

mmsE 9: So 906 mm 5:89:90 .o £0 .95 ___nm__9 uclm .mcozmSoo 98:90 0:09: .8590 .3399; co=m>=oE 65 dcoElm 30:99.00

or 9an

116

Scree Plot
10-.--._m- __ _ __ #_ .__ -0 ._ ,.

.
8‘
6“
4..
2I
(U 7
‘-
.......
..........
.........
- ‘ r-rv . La-
0 ' a. g. L-:‘ —
_. ">o<oo "‘ A
" -_.."vo—.."’ '—--.. 7
o— -9 , t—orofi Ha ..
0 H .-., - s-.,

1 4*? 1'0 ‘13—1‘23"1'g“2§‘23 '2? 37 3'4 3'7 4'0 4'3

Component Number

Figure 1. Scree Plot Yielding Five Factor Solution

117

Table 16

Eigenvalues, Percentages of Variance, and Cumulative Percentages
for Factors of the 44 — Item Parent Motivation Schema Scale JPMSS)

 

 

% of
Factor Eigenvalue Variance Cumulative %
One 8.872 20.164 20.164
Two 4.945 1 1.238 31.401
Three 4.434 10.078 41 .480
Four 1.849 4.201 45.681
Five 1.556 3.537 49.218

 

118

 

3.00 00.0 0N .0 03.0 0N0.0- .0E 0. €0000E. 0. 0.5 0:0 .0.>0:0: 0.0....0 >E 0:...0.€00 .0 0000 E0 . .00

000.0 0N .0 000.0 000.0 00.0- 00:03.0 05 0.0:0: 0:0 .00

.055: .00 .055 50 0.00.. :00 _ .0000E 050.: .0 0:0 :_ 0. 0.20 >E :0:>> .0..

N000- 3 .0 00N.0 mm .0 000.0 .E00.

0. 00... _ 003000: 0. 0:03 0:0.0000 €0.00 >E 00 0.003 _ :0000. 50E 0: .- .00

00.0 :00 000.0 000.0 30.0 .0.0E>:0 .:0:E 00 0:. :0 5500 x00. 0:03 0.050 0:0 .0.>0:00

0.0..:0 >E 0.0:0: 0. 30: E00. 0:. 0.0: _...s 0:0.0000 €0.00 5 €0E0>.0>:. .00

000.0- 300. 000.0 03.0 000.0 .055 0E 000.0E .. :05... .00: 0:0; 0.:0.0000 €0.00 00... ._..s . .5

000.0 0NNO- .00.0- 0000 000.0 0:30.000 000 .00.0.0.. .0. .0 000.00. 0.050

:05 0:. 0. E9505 0.0:. 0. 0003 0.5.0000 €0.00 05 0:.0:0.0.00:3 .00

03.0 03.0 00.0 03.0 30.0 0.0.0: 30:0.

..:0.0 _ 0:.50E00 E00. _ :05... 0:0.0000 €0.00 5 3.0000000 .00E .00. ...>>. ...~

000.0 000.0 .0000 000.0. 000.0 0000.0 €0.00 5 €000. 0: ._..s 0:3 E00. 0. 0E .0. €0000E. 0. ._ .v

.00.0- 0N0.0 000.0 ~000- thd .0.000 >E :000. 050.0: 5 5.00: 0: .._.s 0:0.0000 €0.00 5 €0E0>.o>:_ .00

00.0. 0.0.0 0.0.0- 8.0. 30.0 .00>..

.00 .0 3:030 05 0:.>0.0E. 5 5.0.0: 00 __..s 0:0.0000 €0.00 :. €0E0>.0>:. .0.

000.0- 000.0 000.0- 000.0- 80.0 .._00>E 3000 .0500 .00. 0E 0.0... 0:0 .0.>0:00

0.0..:0 >E 0.0:0: 0. 30: E00. 0E 0.0: __..s 0:0.0000 €0.00 :. €0E0>.0>:_ .0

0.0.0- 000.0- 00 .0 9.0.0 .0.-.0 000.90_E .0 .0. 0 050E . .. :0>0 E0.. E00. _....> . 0:02. 0.5.0000 €0.00 00... . .mm

30.0- 000.0 0 00.0 0.0.0. 2.0.0 0000.0 0005 5 0:.E00. E0 _ .055 0v... =05 . .0.

000.0- 000.0 30.0- 30.0 «3.0 0:000:50 .0_.E.0 :. 9:0.00 .050 .00E

:05. 00 .:0. 20.0000 0:0 05.00.95 0: 0.00? 0000.0 €0.00 5 €0E0>.0>:. .0..

30.0 08.0. 0000 N0 :9 0.0.0 .2000 00 0.00; 000.0 .00.00 0 0. 0.00. 0.00; _ .00; .05 0.0.5. .9

000.0- .000 .000 .000. 000.0 0000.0 .00.00 0. 0.00. 0.08 _ .003 9... 0.00; . .3

000.0 000.0 00.0 0N0.0. 03.0 0500.95 0: 0.002. 000.0 €0.00 0 5 E00. 0.003 _ .055 .05 0.55. <-
02". .30”. 00.: .- 03.0 0:0 0E9.

 

005003 .900”.

 

 

 

.mmmuz. @020. 05.0. :0_S_0m 5.00002“. 0:000:00 x0E..0> .0. 0.050004 .900“. 0:0 E9. .0 EmEEam

n.- 900-.-

119

 

000.0

000.0

000.0

003.0

030.0

000.0

003.0
003.0

000.0

00.20

000.0
00.20-

000.0-

003.0-

000.0-

00..0-

000.0-

000.0

003.0

000.0

00.0.0

000.0

«00.0

00.20
000.0

003.0

000.0

000.0-
003.0

300.0-

00.20-

03.0

000.0-

3.00.0

0.00.0

003.0

000.0

03.0

000.0

003.0

000.0
0 ..md

000.0

33.0

000.0
000.0-

000.0

000.0-

000.0-

003.0-

000.0

000.0-

003.0-

300.0-

000.0

000.0

003.0

00 _..o
000.0-

003.0-

000.0-

000.0
0 3.0

000.0

30.0

000.0

000.0

30.0

000.0-

000.0-

000.0

000.0

000.0

30.0-

0 00.0
000.0-

000.0

000.0

000.0
000.0

03.0

300.0

003.0-

000.0-

000.0-

.:00E 000:0:0

0: ..000 .. 0:0 .0:.E.: .0000 0.00: 00> 0:.50E00 0. .0320: 0.0::0 >5. .00
00500.0. 05:0.» 05...:

.00. .0.>0:0: 0.0::0 0E 00:0:0 0. 0:00 0: :00 .0:. :00E .0: 0. 0.0:: .00
.:00E 000:0:0 0: ..:00 ._ .0. 0: 0:3 .0 ..00 0 0.

.0: .0 000.0 0.00 0:0 .0:0..00..05 000:0 .0:0..00..0 030:0. 0..:0 0E .0:.0:>> .0
..0.>0:0: 0.0_.:0 0E 05.0:0:

.0 E0 _ 0000 30: 30:0 0E .0. .0:. 0052.00 0:0.0000 €0.00 00.: 0.003 _ .0
.0.:0.00 .050 :05 .0.>0:0: 0500:0E .0 .050: E0 _ .05 050000 005.50
.0.:0.00 0E. 0.00.00 0.0 0:? 0050.0. .0:.0 0E 30:0 0. 00... 0.00.5 . .03
.0.:0.00 .0:.0 0:. :05 .0.>0:0:

05.0:0: .0 .000: E0 _ .0:. ..0000. 000.0 €0.00 0E 30:0 0. 00... 0.002, _ .0

. 000:0 0.000.. ..000 _ 9.: 0008. 0.90 o. 0:. .0. 20:005. 0. ._ .050: .< .0.
.0050: 00E 0.0000 .0:.0 .0:; 0. 0000.0

5 E05 0500000.0 0:0 0E0: .0 0050.0.E 050.0E .0000 05:. .0.03 0:... .00
.0E .0000 0.55

0.050 .0:? .0000 00E00:00 E0 _ .00...>..00 000.0 5 0.00.0...00 _ :0:>> .00
.8. 0 9... 00.0.02

.0000 00E00:00 E0 . ..00 05.00 0.050:0 0:0 0:000 5 0..:0 >E 5.3 :0:>> .00
.0.:0.00 .0:.0 00 :03 00 00 .0: 00 >05 00:? 00: .00. 0.00.00 .00

.00.0:0

..05 00 0:0 0E 0. :0.0: .0.:0000 .0 :00 0E 0>0: :00 _ .0:. €05.50 E0 .. 00
.:000.000 .:0.. 0:. 5.3 00500 0: :00

.0: 000.0 .0 :0..00..0 0:. .0 000000: 05:0 0500.0. .0:.>.00 0.00 0..:0 22 .00
0.00.-50

.00E 050: 0. 0:0 .0 0: :0:>> 0..:0 0E 0. :000.5 .00 :00 . .0.. 2.00. _ :0:>> .0
.00 0. .0:; :0 0000. 0E00 0>0: 0:0 00.00.0..0 0E00 30:0. _ .0:.

.00. . .A:0.0:0 .0__0E. 0000.0 0:000 5 0.0: 0:0 0>..00.0.0 0E000: 0..:0 0E ._ .0
00.0:0: 0: :00 .. .0: ..0 E.: .00

.0:. 0:.50E00 2.0000 0. 0.0:. .0000E .055: .0 0:0 5 0. 0..:0 >E :0:>> .3
.0E 0. .:0000E. 0. 0.50:0 .0.>0:0: 0.0::0 0E 0500:0E .0 0000 E0 . .00
0..:00 0. 0.00.:

120

Table 18

Factor Loadings for Varimax Orthoqona_l Five-Factor Solution

 

 

 

Items Load
Factor 1.
7. I think that what I would learn in a parent group would be interesting. 0.843
44. I would like what I could learn in parent groups. 0.839
13. I think that what I would learn in a parent group would be useful for me to know. 0.815
43. Involvement in parent groups would be interesting and possibly fun. 0.812
18. I will like what I am learning in these groups. 0.775
23. I like parent session’s work I will learn from even if I make a lot of mistakes. 0.757
2. Involvement in parent sessions will help me learn how to handle my child’s behavior. 0.728
16. Involvement in parent sessions will be helpful in improving the quality of our lives. 0.721
20. Involvement in parent sessions will be useful in helping me reach my goals. 0.717
4. It is important for me to learn what will be taught in parent groups. 0.699
21. I will feel most successful in parent sessions when I learn something I didn’t know
before. 0.644
38. Understanding the parent session’s work is more important to me than others looking
at my behavior and approving. 0.598
31. I will like parent session's work best when it makes me thing. 0.573
37. Involvement in parent sessions will help me learn how to handle my child’s behavior
and others won't look down on me so much, anymore. 0.457
Factor 2:

40. When my child is in one of his/her moods. I can figure out what set him/her off, and
handle the situation. 0.766
30. I am good at controlling my child’s behavior and this is important to me. 0.747
39. I am good at managing my child’s behavior and this is important to me. 0.740
11. When my child is in one of his/her moods, there is usually something that set him off
but it can be handled. 0.720
8. If my child became disruptive and noisy in public places (malls. church). I feel that I
know some strategies and have some ideas on what to do. 0.697

5. When I really try. I can get through to my child when he or she is being most difficult. 0.641
29. My child gets defying. refusing, angry because of the situation or place but can be

 

 

 

calmed with the right approach. 0.592

27.l am confident that I can have my son or daughter listen and do their chores. 0.413
Factor 3:

36. Parents feel bad when they do not do as well as other parents. 0.695

32. When with my child in public and she/he’s acting out. I am concerned about looking

like a fool. 0.644

35. When I participate in group activities. I am concerned about what others think about

me. 0.567

25. The worst thing about making mistakes at home and discussing them in groups is

what other people may notice. 0.519

19. At home. it is important for me to avoid looking like I can't handle things. 0.477
Factor 4:

9. I would like to show my parent group leader, that I am better at handling behavior than

the other parents. 0.821

17. I would like to show my other relatives who are parents (my parents, siblings.

cousins) that I am better at managing behavior than other parents. 0.777

6. I would like parent sessions activities that let me show how good I am at handling my

child’s behavior. 0.471
Factor 5:

3. Whether my child follows directions, obeys instructions. and acts okay or not is a part

of who he is. It can't be changed much. 0.780

12. My child's behavior is something very basic about him/her and it can’t be changed

much. 0.735

26. There is not much that can be done to change my child's behavior (e.g. hitting.

yelling, refusirE). 0.702

 

121

Together, the five rotated factors accounted for 49.22% of the variance
(Table 16). Factor 1 consisted of 10 value items and 5 Ieaming goal items. This
rotated factor containing a total of 15 items, accounted for 20.16% of the
variance. Factor 2 consisted of 8 items (3 optimistic explanatory style, 3 self
efficacy, and 2 schematicity) which accounted for 11.24% of the variance. Factor
3 consisted of 5 items (4 self defeating performance goals and 1 self enhancing
performance goal) which accounted for 10.08% of the variance. Factor 4
consisted of 3 items (3 self enhancing performance goals) and accounting for
4.20 of the variance. Factor 5 consisted of all 3 implicit theory items accounting
for 3.54% of the variance.

Chronbach’s alphas for the factor structure of the PMSS vary (Table 19).
The Factor 1 coefficient score is excellent (.93). Coefficient alpha for the Factor 2
score is good (.83). Coefficient alpha for the Factor 3 and 4 scores are adequate
(.68 and .71, respectively). Coefficient alpha for the Factor 5 score is also
adequate (.72).

Pre- and post- PMT motivational subscale measures are significant for all
subscales when control and treatment groups are collapsed into total participants
(Table 20). However, correlations of the motivational subscales from pre- to
post— PMT are higher for the control group than for the treatment group (Table
21). Similarly, pre — and post- PMT subscale measures are significant for all
motivational factors when control and treatment groups are collapsed and
significant for all factors, except the implicit theory factor when viewed separately

(Table 22). These results suggest greater consistency in motivational subscales

122

 

 

 

mud id mod 86 mod 8:24. 852%:ma
and no.0 nod and mad mo=___nm__cm 93?. E9.
EN mmN and «fim 3.9 5:350 235%
5+ wmh E. E o FNN mmdm cmos. 28m
9 m? vm mm mm mo:_m> E:E_xm_2
F N o NF 5 mo:_m> E:E_:__2
m m m w m? 9:9. ..o 8952
m2“. So... 925 0?... 95 839.630
208m“.

 

 

swans mmsa 9: co 25: 3; __m 8 8:28 583-82”. .2825 xmetm> .9 823580

9. 2an

123

 

.< x6083. 0. 02m: Em mEog on... .902

 

 

 

#000 000.0- 000.0- 000.0 03.0 .. 50.0 050.0- 0.0.0 500.0- -.<>
N000 500.0- N5P.0- 00.4.0- 00N.0- 0 3.0 500.0 05N.0 100
000.0 300- 03.0 N000- 50N.0 0.0.0 500.0 ms:
000.0 000.0 0000- $0.0 00—..0 . 00..0- w0n.
0N0.0 550.0- N00.0- 05 ..0.- 0Nv.0- n.0n.
No.00 0 50.0- 005.0 500.0- 0-_
0 55.0 NNN.0 000.0 mun.
00¢.0 00¢.0 mwO
..N0.0 mm

6.3—..00
0N0.0 0N0.0 50N.0- 050.0- 00.20 0NP.0 050.0- NN0.0 005.0 -_<>
000 N0 .. .0- 555.0 500.0- 000.0 005.0 «40.0 0 ..0.- 100
N50 0000 000.0 ..F...0- 00.0.0 090 00..0- n..2.
055.0 03.0 N00.0- 00..0- 05_..0 00.0 m0n.
000.0 505.0 005.0- 005.0 00N.0 00m
000.0 003.- 3N0 00N.0 0..
0N0 5VN.0- 00v.0- mun.
000.0 ..0v.0 me
000.0 mm

HEM.

-.<> 100 n..)._ 00a m0n. 0.. mun. mmO mm 00:02.0.)—
:o_.m>_.o.>_ tan. -o.n. him-.00.”.

 

 

AmNHCV E.:—..00 new A hmncv “cm—5500...?

 

co. Emcmoi .:2n. .32-0000 55?. -90 $5305. 00002.22 0095 2200.930

0N 0.00....

124

50.0.... .3. v m .. . mo. v m ..

.00:_0> u -.<> ”3.000;.0500 n ..00 00.805 0.2.08. u as: 00000.00 0.000 0.000 0000.50.00 u 00a.
00.000500 :00. 0.000 000085000 u won. .0.0o0 0:.E00. u 0.. ”0.30 09000.90 0.0.8.0000 u mun.
”0:003:00. 0.30 09000.90 000.8000 n 000 $00050 ..00 u mm .< 50:03.4 0. 0000.. 0.0 0800. 0:... .062

 

 

 

 

005.0 50N0- 0050- 03.0 505.0 034.0 000.0 000.0 000.0 -.<>
00.00 0NNO- 03.0 59.0- 000.0 0¢N0 5¢N0 000.0 100

v0N0 0..0.0 0N00 0050- 520 N050 N0...0- n..>..

005.0 000.0 0500- 000.0- 00...0 3.0.0 m0n.

N000 000.0 03.0- ..0...0 00_..0 00m

5000 000.0 00...0 0..N.0 0..

N500 5.. ..0- ....0.0- mun.

050.0 #300 wwO

0.00 mm

.73 100 as: w0n. D0n. 0.. mun. 0wO mm 0000200.).
00:026.). ._-.>.n.-0._n. ._..>_n.-00n.

 

 

.00.... 0:00.2th :0
.2 E0620 52m. -0000 0:0 ....2n. 0:. 0050005. 0000282 0:080 000000.50

 

..N 0.00...

125

Table 22

Correlations amonlMotivation Schema pre- PMT and post- PMT Program
for Treatment (n=37 ) and Control (n=29)

 

 

 

Post Pre-Motivation

Motivation LV IMC PGD PGE IT
.Fl/LT.

Learning Value 0414*

Se/OES/Sch 0.026 0.594"

Performance Goal/Defeating 0.037 0.263 0.510“

Performance Goal/Enhancing -0.139 0.221 0.531“ 0.698“

lmplicitTheory -0.296 0.074 0.117 0.187 0.267
Control

Learning Value 0.713“

Se/OES/Sch -0.872** 0.714“

Performance Goal/Defeating -0.153 -0.052 0595*

Performance Goal/Enhancing 0.102 0.257 0.604 0.826“

Implicit Theory 075 0.009 0.427 -0.492 0.425
M

Learning Value 0.647“

Se/OES/Sch -0.235 0.582“

Performance Goal/Defeating 0.012 0.176 0.533“
Performance Goal/Enhancing -0.023 0.223 0.559“ 0.732“

Implicit Theory -o.24o 0.073 0.220 -o.002 0.316*

 

*p<.05
**p<.01

126

for the control group (Table 20 and 21). In fact, test-retest reliabilities of the
PMSS indicate highest reliability coefficients for the control group (Table 23).
Test — retest r_‘s ranged from .905 for self enhancing performance goals (control
group) to .421 for implicit theory (treatment group).

DISCUSSION

The results of the present pilot study indicate that the various nine
motivational inner resources can be integrated into a smaller number of
motivational schemas. The results also suggest that the PMSS is a
psychometrically sound instrument for measuring parent motivation regarding
child behavior and participation in PMT programs. Study 1 establishes the factor
structure of the PMSS. This study suggests that the PMSS assesses parent
motivational aspects into five principal components/factors. A reliability analysis
of these factors indicates that the structure is sufficient for the five factors. The
variable item reliability results likely reflect the number of items which comprise
the respective subscales as well as elements of noise in the data. In future, some
items from the scale may be excluded because of a lack of correlation or factor
structure.

The five factors are comprised of specific motivational aspects and can be
named: (1. Learning Value; 2. Internal Motivational Controls; 3. Performance
Goals-Self Defeating; 4. Performance Goals- Self Enhancing; and, 5. Implicit
Theory). Factor 1 consists of scale items from two motivational resources (value
and learning goals) which represent parent’s positive hopes and expectations.

This more existential, transcendental, and idealistic factor can be labeled,

127

Table 23

Test-Retest Reliabilitv Coefficients gre- PMT and post- PMT Program for
Treatment (n=37 )and Control (n=29)

 

 

 

 

Scale or Participants
Factor Treatment Control Total
Subscales
SE 0.771 0.609 0.743
OES 0.753 0.608 0.722
PES 0.456 0.842 0.583
LG 0.478 0.660 0.596
PGD 0.585 0.491 0.834
PGE 0.863 0.720 0.562
IMP 0.421 0.623 0.480
SCH 0.733 0.745 0.652
VAL 0.685 0.819 0.824
Schema
LV 0.586 0.832 0.785
IMC 0.746 0.833 0.736
PGD 0.676 0.746 0.696
PGE 0.822 0.905 0.846
IT 0.421 0.623 0.480
** p < .01
*** p < .001

Note 1. The schema items listed are LV = Learning Value; IMC = Internal
Motivational Controls; PGD = Performance Goal/Self Defeating;
PGE = Performance Goal/Self Enhancing; lT = Implicit Theory.

128

Learning Value (LV). Factor 2 consists of items representing three motivational
resources (optimistic explanatory style, self efficacy, and schematicity) which
appear to represent parent’s internal resources and competency. This more
autonomous, independent, self-related, ownership taking factor can be labelled,
lntemal Motivational Controls (lMCs).

Factor 3 consists only of items pertaining to performance goals including 4
self defeating and 1 self enhancing items. These items appear to represent
resources external to parent’s self resources and influenced by others. This more
relational, social dependent, attributing to others, less controllable factor can
continue to be labelled by its original construct, Performance Goals— Self
Defeating (PGD).

Factor 4 consists of only self enhancing performance goals and is also
labelled by its originally construct, Performance Goals- Self Enhancing (PGE).

Factor 5 consists of all of the Implicit Theory (IT) items which directly
reflect incremental and entity theorist perspectives.

The test-retest reliabilities are generally adequate to excellent and are
most stable for the control group participants. The difference in the pre- and
post- measures across groups may not necessarily reflect unreliability of the
instrument but rather a change in the parent or the parent’s situation. It is
possible that some incidental motivational change may have been occurring in
the treatment group parents and contributed to the weaker correlations from pre-

to post- PMT in comparison to the control group parents.

129

The pilot study addressed the question as to whether the motivational

resources formed identifiable motivational schema. These schema are grouped

together to form five central schemas (LV, IMC, PGD, PGE, IT). Whether these

motivational factors predict outcomes in PMT will be examined in the main study

to address the following questions:

1. Do the traditionally examined predictors (severity of child behavior,
psychosocial stressors, parent depression, socio-economic status, and
single parent families) predict outcomes in PMT(child behavior, parent
satisfaction, parent involvement)?

2. Are the individual motivational inner resources and value aspects
associated with outcomes of parent management training? Specifically,
questions will be asked as to whether these motivational schemas are
predictive of various outcomes in parent training?

3. Do the motivational resources change over time and are they susceptible
to greater change as a result of participation in a predominantly
behaviorally oriented PMT?

Hvootheses:

The following specific hypotheses are offered in the main study:

1.

Parents participating in PMT are likely to demonstrate better outcomes (child

behavior change) in comparison to the parents in the wait list control group.

a) parents in the treatment group will show a larger drop in ratings of total
behavior problems;

b) parents in the treatment group will show a larger drop in ratings of
externalizing behavior problems;

c) parents in both groups will show no change on ratings of internalizing
problems.

. Predictors of PMT will predict outcomes regarding child behavior

change and parent satisfaction as follows:

a) The traditionally examined predictors will continue to be predictive
of outcomes.

b) The motivational factors will emerge among individual parents and
be predictive of outcomes.

c) A combination of traditionally examined predictors and motivational
predictors will be the most predictive of outcomes in parent training.

130

3.

4.

iii)

Predictors of PMT will predict outcomes regarding parent engagement,

i.e. drop out rates:

a) The traditionally examined predictors will continue to be predictive
of parent drop out outcomes.

b) The motivational factors will emerge among individual parents and
be predictive of drop out outcomes.

c) A combination of traditionally examined predictors and motivational
predictors will be the most predictive of drop out outcomes in PMT.

The individual motivation resources will also be predictive of outcomes

as follows:

i) Parent self efficacy for parent behavior management will influence
PMT outcomes:

a) Parents who are high in self-efficacy will likely use more
cognitive strategies, persist longer, and be more consistent
than parents who are not high in self efficacy. Thus, higher
levels of parent self-efficacy will be associated with
sustained and engaged parent involvement.

b) Parents with low self efficacy in the behavior management
domain will be more likely to drop out of parent training.

Parents’ explanatory styles defined as pessimistic or optimistic will
be related to drop out outcomes in PMT.

a) Parents with optimistic explanatory styles will show greater
levels of parent engagement and complete the PMT
program.

b) Parents with pessimistic explanatory styles will show lower
levels of parent engagement and drop out from the PMT
program.

Parents’ goal orientation will influence drop out outcomes in parent

training.

a) Parents who demonstrate a performance goal orientation will
likely focus on the adequacy of their ability, leaving them
vulnerable to the helplessness response, negative affect, setting
up impaired performance and translating into ultimately less
positive outcomes in PMT than parents with a Ieaming gOal
orientation.

b) Parents who demonstrate a learning goal orientation will likely
focus on increasing their ability over time, predisposing them to
effective strategy formation, positive affect, and sustained
performance. Thus, these parents will demonstrate more
positive outcomes in parent involvement.

Parents” implicit theorist orientation is likely to be associated with
drop out outcomes in PMT.

131

vi)

vii)

a) Parents who score higher on the entity theorist scale are more
likely to believe in a fixed social-moral reality, invest in the
status quo, use more punishment and less positives, and thus,
demonstrate little change from pre-, post- testing in terms of
parent outcomes and are more likely to drop out.

Parents who are schematic in parent training are more likely than
aschematic parents to have well-developed possible selves which
serve to energize and motivate strategic use of skills and abilities.
Consequently, higher schematicity scores will be associated with

more positive drop out outcomes.

Parents who demonstrate more valuing aspects of motivation

(interest, appreciation) within the PMT sessions are more likely to

be engaged in PMT group activities, complete homework

assignments, and transfer learning to home situations.

a) Higher value, interest, and appreciation aspects of motivation
will be associated with more positive outcomes on measures of
parent involvement/engagement.

The expectancy x value model of motivation provides the

framework for integrating parent motivation factors relevant to PMT

programs. Specifically, relationships are likely between the

following;

a) the interaction between pessimistic explanatory style and value
will predict parent drop out;

b) the interaction between optimistic explanatory style and value
will predict parent drop out.

5. Parents’ orientations will be associated with drop out as follows:

a) Parents who are schematic are less likely to drop out of PMT;

b) Parents who are aschematic are more likely to drop out of PMT;

0) Parents with an entity theorist orientation will be associated with
parent dropout.

d) Parents with an incremental theorist orientation will be
associated with parent completion of PMT.

6. As a result of participating in PMT, parents motivational schema will
change.

a) Parents in the treatment group are likely to show higher post test
scores on self-efficacy in comparison to parents in the control group.

b) Parents in the treatment group are likely to show more changes on
other inner motivational resources in comparison to parents in the
control group.

c) Parents in the treatment group will show less stability among
motivational aspects/resources.

132

Chapter 3
METHODOLOGY

Participants

Participants are parents of clinic referred children who present as requiring
parent management training. The majority of referrals were made by physicians
(general practitioners and pediatricians) in a health care organization. Analysis of
the history of physician referrals indicates few referrals of children demonstrating
“normal” patterns. In the past 2 years of physician referrals for various learning,
behavior, and language disorders, 98 % were assessed to have symptoms within
clinical ranges. Other participants were referred by psychological and social
work staff from the local ADHD clinic and self initiated contact by phoning for an
orientation session. As a result, it was impossible to determine the number of
parents who had initially been referred. Of the parents who attended orientation
sessions and initial assessment sessions, six of the 99 refused to participate in
the research study.

Ninety-three parents with children who range in age from 4 ‘/2 to 12
years old (at first contact) served as participants. Each parent had children
enrolled in a general education, preschool, junior kindergarten to grade 8
classroom and whom the parent described as demonstrating disruptive
behavior problems. Forty-six children met DSM-IV diagnostic criteria for a
disruptive behavior disorder, 35 were diagnosed with a comorbid

diagnoses, 7 diagnosed with another disorder, and 2 with no diagnosis

133

(Table 24). Detailed demographic information divided according to
treatment and control groups is provided in Tables 24 and 25. Frequency
tables for both parent and child characteristics by group were calculated.
Descriptives were calculated for various other identifying characteristics.
All parents have agreed and signed a consent form (see Appendix B for
Consent Form for Participation in Research Project).

Procedures

Procedures followed three primary phases: 1) orientation interview and
baseline assessment; 2) intervention; and 3) a one-month follow-up.

During Phase 1, participants came into the clinic for an initial
orientation/assessment individual session. At the orientation session, the
intervention study was explained to parents and they were asked if they would
like to be involved. Prior to beginning the study portion of the interview, parents
received information about the study and were asked to sign a consent for
research form (see Appendix B). Contingent upon their written consent, parents
then completed independent additional questionnaires. Although their interview
and/or assessment results were not used in the data analyses, any parents who
did not consent were not excluded from participating in the PMT sessions. The
individual interviews and questionnaire completions took approximately 1 to 1 1/2
hours to complete in the clinic. Some of the parents brought the questionnaires
home and either mailed them back or returned them at the first session. As a

result of this process, 8 questionnaires and survey forms were not returned.

134

Table 24

Demgqraphic Characteristics of Treatment and Control ParticipJants in PMT Study (n = 93)

 

 

 

 

 

 

Participants
Descriptive Treatment Control Total
Q 0/o fl 0/0 fl 0/0
Parent gender
Male 17 26.6 7 24.1 24 25.5
Female 47 73.4 22 75.9 69 73.4
Parent education
Elementary 1 1.6 2 6.9 3 3.2
High School 19 29.7 13 44.8 32 34.0
College (Community) 28 43.8 10 34.5 38 40.4
University (undergraduate) 10 15.6 3 10.3 13 13.8
University (graduate) 5 7.8 1 3.4 5 5.3
Ethnicity/ Race
Native Canadian or Inuit 13 20.3 2 6.9 15 16.0
White / Caucasian 50 78.1 26 89.7 76 80.9
Current Annual Household Income ($)
Under 10,000 8 12.5 2 6.9 10 3.4
10,000 — 14,999 13 20.3 5 17.2 18 6.1
15,000 - 19,999 5 7.8 5 17.2 10 3.4
20,000 - 24,999 8 12.5 2 6.9 10 3.4
25,000 — 29,999 0 0 1 3.4 1 0.3
30,000 — 34,999 0 0 1 3.4 1 0.3
35,000 - 39,999 2 3.1 4 13.8 6 2.0
40,000 — 44,999 2 3.1 1 3.4 3 1.0
45,000 -- 49,999 6 9.4 1 3.4 7 2.4
50,000 — 54,999 5 7.8 3 10.3 8 2.7
55,000 — 59,999 4 6.3 0 0 4 1.4
60,000 — 69,999 4 6.3 2 6.9 6 2.0
70,000 and over 7 10.9 1 3.4 8 2.7
Low Income Cutoffs (Stats Canada)
Low 25 39.1 12 41.4 37 12.6
High 39 60.9 16 55.2 56 19.0
Single Parent Household
No (married, common law, etc.) 39 60.9 19 65.5 58 19.7
I Yes (separated, divorced, widowed) 25 39.1 9 31 34 11.6

 

Note. Totals do not sum to 100% due to missing data.

a n=64 for treatment group
b n=29 for wait list control
c n=93 for total

135

Table 25

Demographic Characteristics of Children Whose PLmnts Participated in the PMT Study (n=93)

 

 

 

 

 

 

 

Participants
Treatment Totals
Characteristic % % %
Child Gender
Male 42 65.6 72.4 63 67
Female 22 34.4 24.1 30 30.9
Child Education
Preschool 1 1.6 1 3.4 2 2.1
Junior Kindergarten 8 12.5 2 6.9 10 10.6
Senior Kindergarten 3 4.7 0 0 3 3.2
Grade 1 13 20.3 2 6.9 15 16
Grade 2 4 6.3 2 6.9 6 6.4
Grade 3 10 15.6 4 13.8 14 14.9
Grade 4 9 14.1 3 10.3 12 12.8
Grade 5 6 9.4 5 17.2 11 11.7
Grade 6 7 10.9 7 24.1 14 14.9
Grade 7 1 1.6 2 6.7 3 3.2
Grade 8 1 1.6 0 0 1 1.1
Special Education 1 1.6 1 1.6 2 2.1
Child Medication
None 34 53.1 22 75.9 56 59.6
Ritalin 26 40.6 4 13.8 30 31.9
Other Psychotropic 4 6.3 2 6.9 6 6.4
Child Diagnosis
None 2 3.1 0 0 2 2.1
ADHD 7 10.9 3 10.3 10 10.6
DBD-NOS 28 43.8 8 27.6 36 38.3
ADHD Comorbid with 23 35.9 2 41.4 35 37.2
Other DBD
Other 3 4.7 4 13.8 7 7.5
Previous Treatments
None 14 21.9 4 13.8 18 19.1
Psychotherapy/Counselling 13 20.3 2 6.8 15 16.8
Leaming/Resource 14 21.9 9 31 23 24.5
Combined 22 34.4 2 41.4 34 37.8

 

Note. Values do not sum to 100% due to missing data.

136

Length of time between orientation interview and the first session varied because
of this process.

Parents were initially randomly assigned to a treatment or wait list control
group. Despite random assignment, the treatment and control groups differed in
terms of specific characteristics (e.g., Ritalin use and previous psychotherapy
treatment).

During Phase 2 (Intervention), participants in the PMT Treatment groups
did not receive a reminder phone call about the intervention sessions.
Intervention closely followed Barkley’s (1997) training procedures in four, two-
hour, weekly sessions and a one month follow up meeting, totaling 8 weeks.

During Phase 3 (Follow-up), PMT Treatment parents participated in a
follow up booster session 1 month after termination of the initial parent
management training. All treatment group participants (PMT and Control
Participants) participated in post-PMT interviews 1 month after initial treatment.
Control Group participants were offered parent management training sessions at
this point.

Settinq and Materials

An out-patient health care organization, The Group Health Centre (GHC),
provided the setting and participants for the proposed study. An initial formal
request was made to the Department Head of Quality Assurance who scheduled
a proposal presentation to the Research and Ethics Review Board of the
organization (see Appendix B for Letter Requesting Permission and Algorithm of

Research and Ethics Review Committee Process).

137

All parent interview and training sessions took place in the same
office/conference room locations throughout the same outpatient health care
organization. Individual orientation interviews and assessments were conducted
in an office in the Communication Disorders Department, located one building
west of the Main Building. This corner office is extremely large and comfortable,
next to the Psychiatry Department. When appointments overlapped and parents
required extra time to complete independent questionnaires, parents were moved
to one of the adjacent offices, depending upon availability. Training sessions
were conducted in the Women's Health Centre Conference Room, which seats
20 people and is located in a third building just south of the Main Building.
Participants had no difficulty locating free parking in the large lots at each of
these clinic locations. Non-clinic members were charged a token fee (to cover the
cost of photocopying and other materials). Othenrvise, the individual
consult/orientation and subsequent training sessions were provided without
charging a fee for service. Sessions were offered in the mornings, afternoons,
and evenings to accommodate parents schedules and babysitting constraints.
Parents were allowed to move among groups in order to facilitate attendance,
especially since several parents were employed in shift work schedules. In this
fashion, several attempts were made to remove some of the barriers to treatment
traditionally discussed in the PMT literature.

The Parent Management training followed the step by step procedures as

detailed in Defiant chiIJdren: A clinicfli’s manual for assessment a_nd parent

138

training (Barkley, 1997). Nine of the ten steps1 were covered in two-hour
sessions as outlined in Table 26. Parents were provided with the detailed manual
handouts which summarized each of the steps covered in training as well as
corresponding homework assignments, developed by the therapist. Procedures
for parent training explicitly followed the parent training manual instructions with
the following exceptions to treatment manual protocol:
1) Steps were combined in order that the parent training sessions were
completed within two months rather than in ten plus weekly sessions;
2) Steps 3, 4, and 5 were supplemented with materials from other parent
training manuals and programs (Alvord, 1973; Greenberg & Horn, 1991;
Wells, 1993);
3) Step 8 was omitted, as noted previously;
4) Coffee breaks and refreshments were provided in between each step;
5) The majority of parents participated in small group instruction (6-8
parents) as suggested in the manual but group size varied as some were
larger (10-12 parents) and some were smaller (3-4 parents) depending
upon attendance dates.
6) Parents were required to practice newly learned skills using standard
homework assignments which were collected in subsequent weeks.
Parents assigned to the Control group did not participate in parent training
until follow up data had been gathered (1 month post-PMT). The particular parent
training approaches/procedures to be utilized varied and were somewhat
dependent upon the present study results.
Data Collection

A battery of self report questionnaires was used to study the constructs

believed to be related to PMT. Data were gathered at Baseline (1-2 weeks prior

 

1 Step 8: Improving School Behavior from Home: The Daily School
Behavior Report Card was omitted as the study data gathering was also
conducted during summer months.

139

 

000005. 00-26.60 £005. 000 000 00.0000 00080 x 0

050.090 0030000 0550 05.0ch x.
0000.0 0.500 0. 000.20 00.00005. H0E0.0o.0 00.00.0000. .5 v

532802 050 9 so ms; 0:053 _>

000505. 000.3005 .050 00< 0:0 0.0.... > 0
000.00 000 00.00 03.00 0.00000 62 0. 00.0.0 :0:>> >_
50.0 0000000000. 000 0000._0E00 00.000000. :. N
03000.0 00 ..
9,280.2 09220 0.2. _ F

 

2E 090 5.82,

 

A500F $00000. 00.0.0.0 0.0.-.00 0. 02.000 00 00500090 0000 >0 d000

0N 0.00.0

140

to training), Pre-PMT (at the first PMT session), Post-PMT (at the last PMT
session), and Post-PMT (1 month after training) as indicated.
Traditional/v Examined Predictors:

Information regarding traditionally examined predictors was collected at
baseline session as part of the evaluation process for the initial intake interview
and assessment. Parents completed a general information sheet to assess
marital status, socio-economic status (SES), and number of psychosocial
stressor predictors. Marital status was coded as 0 for single parent households
and 1 for two parent households.

Hollingshead and Redlich’s two factor index (1958), which utilizes
educational and occupational attainment in its formula, was used to reflect
socioeconomic status level (SES). Socioeconomic status was also designated as
low income or high income based on a formula published by Statistics Canada
(1998). This latter formula is current and utilizes various factors (household
income and number of members supported by this income) to estimate low
income cutoffs (LICO). Thus, SES as a predictor was used as a dichotomous
(LICO) and continuous variable (Hollingshead 2 Factor Index) due to the very
different nature of both measures (see Appendix E for Methods of Calculating
SES)

Parent depression was derived through total scores on the Beck
Depression Inventory -2"d Edition (BDl-ll). The BDI-II is a self-report scale of 21
items assessing level of depression in clinical and normal participants. Each item

contains 4 statements about a particular symptom of depression and are

141

arranged in increasing severity. Parents select the statement that best describes
their affect over the previous two week span. lntemal consistency of the BDI-ll is
high as the coefficient alpha for outpatients and students was .92 and .93,
respectively (Beck, Steer, & Brown, 1996). Evidence of the convergent and
discriminant validity of the BDI-ll indicates a positive correlation (.71) with
depression on the Revised Hamilton Psychiatric Rating Scale for Depression and
robust discriminant validity between depression and anxiety (.47) using the
Revised Hamilton Anxiety Scale (Beck et al., 1996).

Severity of child behavior was derived through parent responses on the
Home Situations Questionnaire (HSQ). The HSQ is a rating scale which provides
a measure of child behavior in 14 contexts, specific to home and public settings.
Behaviors are rated according to whether children have problems in any of those
situations and according to how severe on a scale of 1 to 9 (where 1-miId, 9-
severe). A total number count and mean severity score is calculated. Gender and
age based norms are available for this scale and can be used to interpret the
individual child’s severity of behavior in comparison to peers. The HSQ is
reported to have satisfactory test-retest reliability and to correlate significantly
with other parent completed rating scales (Barkley, 1990).

Information from the interview and questionnaires was used to gather
demographic information and summarized in the Participant Section as well as in
table format (T ables 24 and 25). Data were gathered from each individual
participant (including both parents if involved) for each predictor and

subsequently analyzed.

142

Parent Motivational Aspects/ Schema

No published measures were found in the literature regarding parent
motivation and PMT. Thus, the PMSS was developed specifically for this project.
Parent information on motivational aspects was gathered through the
administration of the PMSS, which is an adaptation of 44 domain specific
questionnaire items taken from the research literature on motivation (see
Appendix A for Parent Motivation Schema Scale). The content validity of the
PMSS was attempted through inter-rater judgements and a factor analysis
procedure on the pilot sample. A content validity index was computed using the
approach by Waltz, Strickland, and Lenz (1991). Based on the raters’
judgements, the original 55 items were maintained, revised, or deleted from the
scale. The resulting 44 items in the scale were pilot tested on 200 parents of
children attending a health care organization, community agency, and/or school
system. Support for the reliability of the PMSS was demonstrated through
internal consistency via Cronbach’s alpha (on both the pilot and study samples)
and test-retest reliability (on the study groups i.e., treatment and control groups).
This scale was administered both pre- and post- PMT.

Parents were asked to rate their motivation in two domain specific
areas: a) regarding participation in the PMT intervention; and, b) '
motivational aspects concerning their child’s disruptive behavior on a 5-
point scale from not true of me at all (1) to extremely true of me (5).
Individual items of the PMSS which reflect nine motivational aspects, were

grouped into nine subscales, respectively. The self efficacy subscale

143

assesses parent concern with how well or poorly people judge or believe
that they will perform on a task give their skills and a variety of
circumstances (Bandura, 1997, p.37). A sample item from this subscale is
“If my child became disruptive and noisy in public places (malls, church), I
feel assured that I know some techniques to redirect him or her quickly. “

The optimistic explanatory style and pessimistic explanatory style
subscales assess attributional aspects of motivation. For example, an
optimistic explanatory style involves attributing child misbehavior to 1)
external, 2) unstable, 3) controllable dimensions while a pessimistic
explanatory style involves attributing child misbehavior to 1) lntemal, 2)
stable, and 3) uncontrollable dimensions. Sample items from these
subscales are, “when my child is in one of his/her moods, there is usually
something that set him off but can be handled" and “ my child’s behavior is
mostly biological, a big part of his/her personality, and sometimes, we ’re
lucky to get through one of his/her temper tantrums”, respectively.

The Ieaming goals subscale assesses parent approach to PMT and
child management challenges as the desire to increase competence and
improve skill. A sample item from this scale includes, ”I like parent training
work I’ll Ieam from even ifl make a lot of mistakes.” The performance
goals-self enhancing subscale assesses the levels parents strive to
demonstrate superior abilities, do better than others, or outperform others
(Skaalvik, 1997). A sample item is, “I would like parent training activities

that let me show how good lam at managing behavior”. The performance

144

goals-self defeating subscale assesses the level parents strive to avoid
looking stupid and being negatively judged by others (Skaalvik, 1997). A
sample here involves, “When interacting with my child in public places, I
am concerned not to make a fool of myself.

A short subscale assessed whether parents were entity or
incremental theorists according to implicit theories. According to implicit
theory definitions (Brophy, 1998; Hong et al., 1997; Levy et al., 1998),
parents who are entity theorists view their child’s behavior as a fixed entity
which cannot be changed and over which they have no control. Those
parents who are incremental theorists view their child’s behavior as
responsive to their management effort . All of the sample items, measure
level of entity theory orientation, with low scores implying parents may
hold an incremental theory. A sample item illustrates this point: “my child’s
behavior is something very basic about him/her and it can‘t be changed
much.” Parents with scores of 7 or higher were deemed entity theorists
while parents with scores less than 7 were deemed incremental theorists.

An even shorter subscale assesses parent schematicity which
involves parent recognition that they have a particular ability in handling
child behavior and belief that this ability is important (Cross & Markus,
1994). The items used to identify schematic and aschematic parents were
summed to arrive at a total schematicity/ self schema score. For further
analysis of parent engagement measured by drop-out or program

completion, parents with scores of 5 or above on these items were

145

classified as schematic. Parents with scores of less than 5 were classified
as aschematic. All other parents who do not clearly fall into schematic or
aschematic categories were not selected at this level of analysis.

Several items comprise the value subscale which is a relatively new
aspect and consideration in motivational research. A sample item of this
scale is, “involvement in PMT will be helpful in improving the quality of our
lives.”

Each of these subscale items is summed to be used in select data
analysis. In addition, the implicit theory and schematicity items are
converted into categorical variables as described above.

A principal component analysis of the PMSS yielded five factors.
Thus, the nine subscales were not supported by the factor analysis. When
the five factor solution was produced, the orthogonal factors were labelled
according to their respective factor loadings. Factor 1 was labelled
Learning Value because it consisted of only Ieaming goals and value
items. Factor 2 was labelled lntemal Motivational Controls because it
consisted of self efficacy, optimistic explanatory style, and schematicity
items. Factors 3, 4, and 5 consisted of the original motivational aspects
and retained the names of Performance Goals-Self Defeating,
Performance Goals-Self Enhancing, and Implicit Theory.

Outcomes
The following measures are organized according to PMT outcome

categories.

146

Child Behavior Changes

Pre—, Post-, and Long-term outcome measures regarding changes in child
behavior were based on information from the Child Behavior Checklist (CBCL/4-
18) (Achenbach, 1991). The CBCL data were obtained from one parent of each
child participant. Parents rate their child’s behavior on 118 problem behavior
items where O = not true as far as you know, 1 = somewhat or sometimes true,
2: very true or often true. The CBCU 4-18 provides raw scores which are
converted to normalized T-scores and percentiles grouped on three scales (Total
Problems, Internalizing, and Externalizing) and 8 syndrome scales. Reliability
and validity data are well established and detailed in the manual (Achenbach,
1991). For instance, reliability information from the manual indicates that test-
retest coefficients for the Behavior Problem scales are high (.95) and also high
for the Social Competence Scale (.99) (Achenbach, 1991). Child behavior
change scores were calculated by subtracting post-CBCL t-scores from pre-
CBCL t-scores for each respective scale (Total, lntemalizing, and Externalizing
Problems). Long term child behavior change scores were calculated by
subtracting long-term CBCL t-scores from pre-CBCL scores.
Parent Satisfaction

Post- outcome measures regarding parent satisfaction were based on
information from the Therapy Attitude Inventory (T Al)(Eyberg, 1993). The TAI is
a brief consumer satisfaction measure of parent training, parent-child treatments,
and family therapy with established psychometric evidence of reliability and

validity (Brestan et al., 1999). Support for internal consistency of the TAI was

147

provided by Cronbach’s alpha coefficients of .91 (excellent) and a stability
coefficient over 4-months of .85 (high). Evidence of external validity (construct
validity) was demonstrated through positive correlations (.36 to .49) between
scores on the TAI and a) the Eyberg Child Behavior Inventory, and b)
observations of behavioral compliance (Brestan et al., 1999). A principal
component analysis on the TAI items yields two factors: 1) a satisfaction with
outcome factor which represents parent satisfaction with child behavior changes
and 2) a satisfaction with process factor which represents parent satisfaction with
the components of treatment, including parenting skills taught and general
approach to treatment (Table 27) (Brestan et al., 1999).

The 10 items of the TAI reflect satisfaction with parenting skills learned,
child behavior changes, and the type of program used, which are all derived to
reflect goals of parent-child treatments (Brestan et al., 1999). Parents rate items
on a scale from 1 (no satisfaction) to 5 (maximum satisfaction) and yield a total
score between 10 and 50.

Parent Engagement

A final outcome measure involved estimates of parent engagement and
parent disengagement. Information for parent engagement characterization was
based upon an exposure measure 1) number of hours attended in all sessions;
2) number of homework assignments completed, and 3) number of participants
who dropped out and did not complete the program. The latter was further
differentiated according to point of drop off as pre-PMT, pari-PMT, post-PMT,

long term-PMT. Given the smaller sample size and lower statistical power,

148

Table 27

Factor loadinfion the Therapy Attitude Inventory (TAI).

 

 

Item Content Factor 1 Factor 2
1. Learned useful techniques for discipling 0.39 0.72
2. Learned useful techniques for teaching 0.11 0.87
3. Status of relationship with child 0.81 0.34
4. Ability to discipline child 0.80 0.32
5. Major behavior problems of child 0.82 0.33
6. Child's noncompliance to commands 0.85 0.29
7. General progress of child 0.83 0.33
8. Problems unrelated to child 0.31 0.69
9. Type program used to improve child behavior 0. 72 0.15
10. General feeling about the program 0.39 0.57

 

Note. Numbers in italic represent strongest factor loadings for each item.
(adapted from Eyberg & Brestan, 1999)

149

engagement was analyzed by dropout while attendance and homework
completion measures were summarized in demographic information.
Desiqn and Analysis

The present study was a longitudinal design consisting of three types of
variables: traditionally examined predictors, motivational predictors, and
outcomes in PMT. The focus of the present longitudinal study was to examine
the relationship among traditional predictors (single parenthood, severity of child
behavior, parent depression, SES, and number of psychosocial stressors) and
potential novel predictors (parent motivation aspects) on three categories of
outcomes in parent training (changes in child behavior, parent satisfaction, and
parent engagement).

Descriptive statistics and displays were calculated before formal
analyses. Initially, several analyses comparing the treatment and control
groups were performed using the individual CBCL pre- and post- t-scores
for each scale as well as behavior change scores, in order to establish any
treatment effects. Several other analyses were performed to test the
several hypotheses of the current study (Table 28).

Analysis I: Treatment Effects

To test the hypothesis that parents in PMT would demonstrate
better outcomes, (greater child behavior change) in comparison to parents
in the wait list control group, the first analysis consisted of a 2 X 2 X 3
factorial design with repeated measures. The 2 groups (treatment and

control) by the two time points (pre- and post-) by the 3 CBCL Scales

150

Table 28

Summa_rv_of Specific Hypotheses (abbreviated). Measuresl and Analyses

 

 

 

Measure
Hypothesis Abbreviated Description Pre- Post- Analysis
I Treatment effects: CBCL CBCL 2'3'3 Factorial
Treatment versus control parents Anova
II Continuous outcome variables CBCL CBCL'
will be predicted by continuous predictors TAl TAI
a. TEP predict child behavior HSQ canonical
change and parent satisfaction Stress Checklist correlation
BDl-ll
Hollingshead Index
b. Motivation schema predict PMSS canonical
child behavior and parent satisfaction correlation
c. combination of both types of same as a. & b. canonical
predictors predict child behavior above correlation
and parent satisfaction
lll Discrete outcome variable (drop out) count of treatment count of treatment
will be predicted by continuous predictors participants participants
a. TEP predict drop out HSQ discriminant
Stress Checklist function
BDl-ll
LICO
SPF
b. Motivation schema predict PMSS discriminant
parent drop out factor totals function
c. combination of both types of same as a. 8. b. discriminant
predictors predict drop out above function
W Individual motivational resources predict PMSS logistic
drop out subscales regression
count of treatment count of treatment
participants participants
V Parents' theoretical orientations are PMSS chi-square
associated with drop out schematicity subscale analysis
implicit theory subscale
count of treatment count of treatment
participants participants
Vl Parents who participate in PMT experience PMSS PMSS pair-wise t-test
a change in motivational schema and/or factor totals factor totals
subscales subscales subscales

 

151

(Total, Internalizing, Externalizing) were compared in a two way analysis
of variance. For the purpose of brevity, all tests for normality and equality
of variance were met unless otherwise specified in the results. The
appropriate post hoc analyses (A Bonferroni post hoc test) was conducted
based on an assumption of equal variances.
Analysis ll: Linear Relations Among Continuous Variables

To test the hypotheses that traditional predictors and motivational
predictors would predict child behavior change and parent satisfaction with
PMT, a canonical correlation was conducted in the second analysis. The
canonical correlation examined the linear relationship among the
continuous predictor variables both traditional (severity of child behavior,
parent depression, number of stressors, SES) and motivational ( the five
factors) with the linear relationship among the continuous outcome
variables (changes in child behavior and parent satisfaction). The
appropriateness of canonical correlation techniques is well established
given the multiple predictors and multiple outcomes. Basic tests of
significance associated with the canonical correlation were considered
and two common tools for interpretations were provided. While
interpretation of loadings or standardized coefficients is the most widely
used , primarily for reasons of multicolinearity among variables within sets
(Levine, 1977; Thompson, 1984), standardized coefficients are described
as the most useful (Rencher, 1995). The standardized coefficients remove

the effects of scaling as well as provide the most information regarding the

152

importance of a given variable in the context of the others (Rencher,
1995).
Analysis Ill: Classifying According to Drop Out

The third analysis was conducted to test the hypothesis that
traditional predictors and motivational predicts would be associated with
parent engagement (i.e., drop out of PMT). This third statistical analysis
employed a discriminant function analysis to evaluate whether the
traditional predictors and/or motivational predictors were useful to screen
or classify participants known to drop out of PMT (parent engagement
outcome). Tests for the equality of variance, independence, and linearity
were met unless othenrvise specified.

The two multivariate techniques above (canonical and discriminant)
were conducted to evaluate the relationships among the predictors and
outcomes in PMT. As a reflection of the mutual influence of the variables
on each other, it was assumed that statistical results would change if
some variables were added or deleted. Interest in the current study was
centered on the joint performance of the traditional and new motivational
predictor variables at hand. Mechanisms were in place to conduct an
evaluation and comparison involving a traditional model and new model of
predictors in parent training. Thus, the two categories of predictor
variables were added to each of the statistical analyses in three stages or

models.

153

First, each multivariate statistical analyses included an analysis
with the traditional predictors in the model. Second, statistical analysis
used the motivational predictors in the new model. This indicated the
maximum possible change on outcomes by manipulating new predictors.
Third, each statistical procedure used both the traditional and motivational
predictors in a combined model which indicated a more realistic appraisal
of possible change on outcomes. These three models (traditional,
motivational, and combined) were compared for each multivariate
statistical procedure to test the third, fourth, and fifth hypotheses (Table
28).

Analysis IV: Using Motivational Subscales to Classify According to Dron Out

The fourth analysis tested the hypothesis that individual
motivational resources would be predictive of PMT outcomes. This fourth
statistical analysis employed a logistic regression running the discrete
category outcome variable (drop out versus completers) with the individual
motivational resources (see Appendix D for Rationale for Using
Subscales). Logistic regression was included because it 1) requires far
fewer assumptions than discriminant analysis; and 2) to specifically
evaluates for any interaction effects among the motivational predictors.
Analysis V: Univariate Examination of Discrete Variables

The fifth analysis tested the hypothesis that parent orientation
(schematic versus aschematic and entity versus incremental) would be

associated with drop out. A series of chi-square analyses were also

154

conducted among the drop out outcome variable of the PMT study and the
parents’ schematicity and implicit theory orientations, respectively, and
among other categorical traditional predictors (single parenthood, low
income cutoff). As well, analyses were also conducted for drop out and
other parent or child descriptors. The Chi-Square Two Way Test for
Association was computed to determine whether or not any of the two
paired variables were independent of each other (Horvath, 1981 ).
Analysis Vl: Change in Parents’ Motivational Schema

Finally, the sixth analysis tested the hypothesis that parents’
motivational schema and individual resources would show greater change
for the treatment group participants than for the control group participants.
A series of t-tests were conducted to test this hypothesis.

Table 28 summarizes the specific hypotheses (abbreviated) with
measures used and analyses employed. An alpha level of .05 was used
for all statistical tests. A Bonferroni adjustment was considered and

employed when multiple statistical tests were conducted.

155

Chapter 4
RESULTS
Analysis I: Treatment Effects

The results of the descriptive information and Factorial Anova test the
hypothesis that treatment group parents would demonstrate better outcomes.
The mean CBCL scores for each group (treatment and control), each time point
(pre-, post-, long term-), and each scale (Total, lntemalizing, and Externalizing
problems) are summarized in Table 29. T-scores greater than 65 are indicative
of clinically significant behavior problems. Treatment group behavior scales
indicate a decline in average behavior problems over pre-, post— , and long term-
PMT for all scales. In fact, behavior for the treatment group participants had
improved from clinically significant ranges at baseline (t-score >= 65) to behavior
within normal limits (t-score <60) in comparison to other children their age at one
month follow up.

Control group scales indicate an increase in average behavior problems
over pre- and post- time periods (Figure 2). Visual inspection of the CBCL data
also suggests a slight decrease in median behavior problems for the treatment
group and a slight increase for the control group, across all scales (Figure 3).
Within scales, the treatment group results indicate a larger decline in total
problems, smaller decline in internalizing problems, and little, if any change in

externalizing problems (Figures 4, 5, and 6). The control group results indicate a

156

.00 .0 .00.0000 000.090. 0:90 .908 .9 0N.0 0.0900
.00" .E0-0oEF 000 .08 0......) 0.0 n 0590 E00009. 0.0 .9 0N.0 0.0E0m .N 0.02
.25.. 0050009 00.00.0000 0009.00: 000 3000. .o.>0...00 0..:0 00

0.500. 000. 000. .0. 0:90 .958 .0.. :03 00 000.98 .00 0.03 0.00 .F 0.02

 

 

 

 

 

 

 

157

 

 

 

I...- ll... 00. F no.9 N_<...

ll.- --- 5.0 00.00 :5.

it: it: 006 00.00 .05-F
0000020050
02.00.0000 .00.00

it: .1..- 0.0.0 F $.00 -_..>_0-0oE F

mm. F F 00.3. 00.0 $.00 ..20-0o0

«v. F F .00.00 00.3 3.00 0.20-0.0
00_N..00.0.0. - 000

E... .1..- 3:0. F 00.00 0.5.0-000. F

00. F F 00.00 00.3 00.3. 0.5.0-000

00.0 F 00.00 00. F F R00 0.20-0.0
9..~._00.0Fxm. - 000

l...- ------ 00. F F 00.00 0.5.0-000. F

00.3 00.00 F00 F000 .:20-00o0

mod F 00.00 00.0 00.00 0.20-0.0
00.0390 .05-F - 000

60. z 60. z
.9500 E00000... 05000.2 0.00050
0000.00.00

 

 

00.00..0> 00.0050 0:50.000 5 000.0300 90000.0. 00.000 0000.00.00 000.).

0N 0.000.

T—score

80<

 

50—
TIME
40-

 

N= H 786’ fii foe " f 34 60
Treatment Control

Figure 2. Pre- Post— Scores of CBCL Results

158

T-score

 

 

. ‘Group
40: ‘
‘ .Treatment
304‘ ,, -_ fl ,, --- -. -,. fir 7 7 - ‘ 2423.35“ Control
N= 98 4e 98 48 98 48

Total lntemalizing Externalizing

Figure 3. CBCL Results Across All Scales

159

T-score

80f ‘

70-

50<

N= ae'HT ’36" if” 26 ' 56'

Treatment Control

Figure 4. CBCL Total Subscale Results

160

 

Total Subscale

l
‘ - Pre-

Post-

 

T-score

   

 

70 —
60 4
50 - 1
‘ Internalizing Scale
" l

40 .

- ‘ T ’ l - Pre-

l 7 l __
30_l,,.__,,-_1- .- _ fig? - - 2:3“ Post-

N = 36 36 20 20

Treatment Control

Figure 5. CBCL Internalizing Subscale Results

161

T-score

 

 

 

50-
l Externalizing Scale
40- r ' , ,
-_ ‘ - Pre-
l -
30.- 77 77 7 77 7 7 -Post-
N = 36 36 20 20

Treatment Control

Figure 6. CBCL Externalizing Subscale Results

162

slight decline in total problems, increase in internalizing problems, and slight
decrease in externalizing problems (Figures 4, 5, and 6).

Means and standard deviations for each of the behavior change scores
are shown in Table 30. From pre— to post- PMT, the treatment group has the
largest average change score for total child behavior while the control group has
the smallest average change score for externalizing behaviors. The majority of
changes from pre— to post— PMT, with the exception of internalizing problems, are
in a positive direction suggesting a decline in child behavior problems. The
control group had a relative large average change score for internalizing
behaviors in a negative direction indicating a perceived increase in these child
behaviors. From pre- to long term- PMT, all of the treatment group average
change scores are noticeably larger across all scales indicating a decrease in
perception of problem behaviors after PMT.

The first formal analysis (Factorial Anova with Repeated Measures)
indicated a significant main effect for CBCL Scale, F (2, 338 ) =22.071, p <. 001,
and for Group * Scale, F (2,338) =7.661, p <.OO1. All other main effects and
interactions were not significant (Table 31). A Bonferroni comparison indicated a
significant difference between the internalizing and total scales, p < .001, and
internalizingand externalizing scales of the CBCL, p < .001. Figures 7 and 8
present graphic representations of the Group * Scale interaction. From visual
inspection, it appears that the mean CBCL t-score relates not only to the
particular scale and to the group of the participants, but also a combination of the

values of those variables. Treatment group participants rate total and

163

Table 30

Mean Change Scores (Standard Deviations) from CBCL Measures

 

 

 

 

 

 

 

Participants
CBCL Scale Treatment (n=37) Control (n=29)
M L_lsd M £011
Pre minus Post
Total 1 .50 5.44 0.45 5.92
Internalizing 0.92 7.16 -1.45 5.90
Externalizing 0.47 4.89 0.25 8.03

Pre minus Long Term

Total 5.58 6.51 ............
Internalizing 4.77 9,13 ............
Externalizing 4.94 6.57 ............

 

Note. Long term CBCLs for control group were not administered.

164

Table 31

Analysis of Variance Results for Main Effects and Interaction Effects
of Group. Time PointLand Scale on Child Behavior Changes (CBCL)

 

 

 

 

 

Source 0 M§ E g9.
Between Subiects
Group 1 95.066 0.251 0.617
Error 1 90 378.247
Within Subiects
Time (pre- , post-) 1 9.803 0.377 0.540
Scale (total, internal, external) 2 574.419 22.071 0000*"
Group * Time 1 28.375 1.090 0.297
Group * Scale 2 199.375 7.661 0.001 ***
Time * Scale 2 4.187 0.161 0.851
Group * Time * Scale 2 9.775 0.376 0.687
Error 2 338 26.026
* g < .05.
** g < .01.
***_g < .001

165

T-score

 

l ‘ 1 - Group

- Treatment

 

.' - - --- - v" . --.

-' .-

N = 98 48 98 4a 7 98 48
Total lntemalizing Externalizing

Figure 7. Boxplot Depicting Group * Scale Interaction

166

T-score

l
Group
- Treatment

7 Control

 

Total N lntemalizing ”7 Externalizing

Figure 8. Bar Graph Depicting Group * Scale Interaction

167

externalizing behaviors with the highest scores. Control group participants rate
the total scale with the highest score and rate both internalizing and externalizing
problems with equivalent scores.

Thus, the results of the first analysis do not support the overall hypothesis
that treatment group participates would demonstrate better outcomes (child
behavior change) in comparison to control group participants. The treatment
effects appear dependent upon CBCL scale and given the long term CBCL
treatment group results, may also depend upon time after treatment.

Analysis ll: Linear Relations Among Continuous Variables

To test the hypothesis that PMT predictors would predict child behavior
change and parent satisfaction, the second analysis (canonical correlation)
determined a linear relationship between the continuous data set of predictors
and outcome variables across the three models.

In the traditional predictor model, only the first canonical variate accounted
for a significant relationship between the two sets of variables ( r = .816, p <
.018). The first variate contained largest coefficients for severity of child behavior
as defined by responses on the HSQ and number of psychosocial stressors in
set 1 (predictors) and internalizing behavior problems (both post and long term),
long term externalizing problems, and parent satisfaction with both process and
outcome in set 2 (outcomes) (Table 32). It is noteworthy that maximum loadings
are evident in severity of child behavior and parent depression for the first set
and all child behavior scores on the second set, highlighting potential differences

between interpreting standardized scores versus loadings or structure

168

Table 32

Correlations and Standardized Canonical Coefficients and Their Canonical Variates
in Separate Models

 

First Variate Second Variate

 

 

Canonical
Correlation coefficient

Canonical
Variable Correlation coefficient

jaditiorLaL Model One

 

 

PREDICTORS:
HSQ-Severity 0.972 0.995 -0.233 -0.098
Number of Stressors 0.179 -0.254 -0.868 -1.034
Parent Depression 0.468 0.154 0.247 0.567
SES 0.096 0.066 -0.396 0.151
OUTCOMES:
Post-
CBC-Extemalizing 0.822 0.044 -0.244 0.929
CBC-Internalizing 0.852 0.984 0.184 0.021
Long-Term-
CBC-Externalizing 0.706 0.620 -0.551 -1.914
CBC-Internalizing 0.637 -0.662 -0.030 0.827
Parent Satisfaction
TAI-1 -0.056 0.457 0.310 0.609
TAI-2 -0.356 -0.380 -0.025 -0.219
Motigtionai Model Two
PREDICTORS:
1.Val, LG 0.852 0.752 -0.325 -0.255
2. SE, OES, Sch 0.575 0.354 -0.206 -0.247
3. PGE, PGD -0.293 -0.047 -0.468 -0.292
4. PGE -0.187 -0.238 -0.100 -0.084
5. lmplicit 0.317 0.306 0.889 0.811
OUTCOMES:
Post-
CBC-Externalizing 0.270 0.41 1 0.254 1.461
CBC-Internalizing -0. 190 -0.477 0.137 -1.077
Long-Term-
CBC-Externalizing 0.407 1.020 0.262 -0.825
CBC-Internalizing -0.104 -0.675 0.364 0.630
Parent Satisfaction
TAI-1 0.290 0.246 0.406 1.057
TAI-2 0.579 0.418 -0.383 0873

 

Model One; = .816, p < .018
Model Two 5 = .795, p < .280

169

coefficients. The second variate (r = .608, p < .376) contained largest
coefficients on number of stressors and parent depression with externalizing
problems (both post and long term), and parent satisfaction (Table 32).

In the motivational predictor model, none of the canonical variates
accounted for a significant relationship (5 = .795, p < .280). Researchers employ
significance tests only as a minimal criteria in deciding which canonical functions
to interpret as a result of the large samples required to invoke the central limit
theorem and to minimize the tendency of canonical correlation to capitalize on
sampling error (Thompson, 1984). Largest coefficients for the first variate
included four predictors (factors 1, 2, 5, 4) with five outcomes (post and long term
externalizing problems, long term internalizing problems, and parent satisfaction
with outcome) (Table 32).

The combined model yielded a significant first variate ([ = .950, p < .016)
with maximal loadings on five predictors (number of stressors, severity of child
behavior, IT, PGD, LV) and maximal loadings on two outcomes (post
internalizing problems and satisfaction with process) (Table 33). The second
variate ( r_= 813, p <.511) yielded maximal loadings on five predictors (severity of
child behavior, stress, parent depression, factor 1, and factor 5) with two
outcomes (externalizing and internalizing problems on the long term outcome). In
comparing the three models only the traditional model and the combined models
had significant canonical variates (Table 34).

As SES failed to load even moderately onto any of the previous models, it

was removed in a subsequent model leaving the remaining variables. While only

170

Table 33

Correlations and Standardized Canonical Coefficients a_nd Their Canonical Variates
i_n_a Combined Model

 

 

 

 

First Variate Second Variate
Canonical Canonical
Variable Correlation coefficient Correlation coefficient
PREDICTORS:

HSQ-Severity -0.606 -0.931 -0.751 -0.641

Number of Stressors 0.036 0.760 -0.524 -0.507

Parent Depression -0.507 —0.448 -0.110 0.312

SES 0.125 -0.208 -0.288 0.280

1.Va|, LG 0.458 0.356 -0.634 -0.369

2. SE, OES, Sch 0.494 0.228 -0.166 -0.092

3. PGE, PGD -0.211 0.438 0.109 0.089

4. PGE 0.000 -0.224 0.187 0.129

5. Implicit 0.037 0.644 -0.235 -0.361

OUTCOMES:
Post-

CBC—Extemalizing -0.394 0.271 -0.822 -0.019

CBC-Internalizing -0.807 -1.127 -0.494 -0.277
Long-Term-

CBC-Externalizing -0.185 0.295 -0.884 -1.358

CBC-lntemalizing -0.500 0.077 -0.446 0.794

Parent Satisfaction
TAI-1 0.110 -0.209 0.146 -0.005
TAI-2 0.591 0.531 -0.013 -0.144

 

[ = .950, p< .016

171

Table 34

Canonical Analysis of Predictors a_nd Outcome Variables in PMT

 

Standardized

Canonical Coefficient

 

 

 

Variable Traditional Model Motivational Model Combined Model
Root 1 Root 2 Root 1 Root 2 Root 1 Root 2
PREDICTORS:
Traditional Predictors
HSQ-Severity 0.995 -0.098 ------------ -0.931 -0.641
Number of Stressors -0.254 -1.034 ------------ 0.760 -0.507
Parent Depression 0.154 0.567 ------------ -0.448 0.312
SES 0.066 0.151 ------------ -0.208 0.280
Motivational Factors
1.Val, LG ------------ 0.752 -0.255 0.356 -0.369
2. SE, OES, Sch ------------ 0.354 -0.247 0.228 -0.092
3. PGE, PGD ------------ -0.047 -0.292 0.438 0.089
4. PGE ------------ -0.238 -0.084 -0.224 0.129
5. Implicit ------------ 0.306 0.811 0.644 -0.361
OUTCOMES:
Child Behaviour Changes
Post-
CBC—Extemalizing 0.044 0.929 0.411 1.461 0.271 -0.019
CBC-lntemalizing 0.984 0.021 -0.477 -1.077 -1.127 -0.277
Long-Term-
CBC-Externalizing 0.620 -1.914 1.020 -0.825 0.295 -1.358
CBC-lntemalizing -0.662 0.827 -0.675 0.630 0.077 0.794
Parent Satisfaction
TAI-1 0.457 0.609 0.246 1.057 -0.209 -0.005
TAI-2 -0.380 -0.219 0.418 -0.873 0.531 —0.144

 

1 Model One: = .816, p < .018
2 Model Two I = .795, p < .280

3 Combined Model [ = .950, p < .016

172

one of the canonical variates accounted for a significant relationship between the
two sets of variables (L: .942, p < .008), the second, not significant variate ([ =
.799, p < .382), is also presented in Table 35. Highest loadings for the predictor
set on the first variate include severity of child behavior, number of stressors,
factor 5 and factor 3. All of the other variables (parent depression, factors 1,2,
and 4) also contribute to the equation with modest loadings. Highest loadings for
the outcome set on the first variate include changes in child behavior on
internalizing behavior problems, externalizing problems, and parent satisfaction
with the program (Table 35). Highest loadings for the predictor set on the second
variate include severity of child behavior, factor 5, number of stressors, factor 1,
parent depression. Highest loadings for the outcome set on the second variate
include long term externalizing and internalizing problems (Table 35).

Canonical variates from one set of variables are subject to change with
any variations in the second set (Levine, 1984). Thus, it was decided to vary the
nature of the second set. The variables in set 2 were changed by using child
behavior change scores in lieu of the post and long term t-scores across scales
(Tables 36 and 37). Neither of the first variates for the traditional model (r = .684,
p < .168) or motivational model (r = .782, p < .081) was significant (Table 36).
However, the first variate for the combined model was significant (r = .908, p <
.028) (Table 38). The standardized coefficients indicate links worth identifying
among the predictors and outcomes. Three of the traditional predictors (stress,
severity of child behavior, and SES) and two of the motivational factors ( 1 and 5)

are highly correlated with the first variate while one traditional (parent

173

Table 35

Correlgtions a_md Standardized Ca7nonical Coefficients Between Traditionai
yd Motivational7Predictors and PMT Outcome Variables a_nd Their Cfianonical Variates

in a Combined Model without SES

 

 

 

 

 

First Variate Second Variate
Canonical Canonical
Variable Correlation coefficient Correlation coefficient
PREDICTORS:
Traditional Predictors
HSQ-Severity -0.664 -0.985 0.695 0.561
Number of Stressors -0.020 0.610 0.477 -0.243
Parent Depression -0.523 -0.361 0.048 0.370
SES ---- ---- ---- ----
Motivational Factors
1.Val, LG 0.412 0.295 0.675 0.352
2. SE, OES, Sch 0.494 0.249 0.236 0.161
3. PGE, PGD -0.223 0.453 -0.180 -0.056
4. PGE 0.012 -0.245 -0.194 -0.199
5. Implicit 0.050 0.583 0.321 0.379
OUTCOMES:
Child Behaviour
Post-
CBC-Externalizing -0.464 0.315 0.777 0.160
CBC-lntemalizing -0.854 -1.184 0.408 0.093
Long-Tenn-
CBC-Extemalizing -0.256 0.178 0.863 1.343
CBC-lntemalizing -0.535 0.133 0.403 07%
Parent Satisfaction
TAI-1 0.164 -0.134 -0.035 0.164
TAI-2 0.585 0.468 0.056 0.078

 

r: .942 , p < .008

174

Table 36

Correlations. Standardized Canonical Coefficients a_nd Their Canonical Variates in

Traditional and Motivational Models

 

First Variate

 

Second Variate

 

 

 

 

Canonical Canonical
Variable Correlation coefficient Correlation coefficient
laditioneLModel One
PREDICTORS:
HSQ-Severity -0.059 -0.410 0.464 0.652
Number of Stressors 0.871 1.172 0.313 0.077
Parent Depression 0.107 -0.077 -0.530 -0.728
SES 0.118 -0.315 0.688 0.418
OUTCOMES:
Child Behaviour Changes
Pre- Post Change Score
CBC-Extemalizing —0.580 -0.498 0.152 0.920
CBC-lntemalizing -0.103 0.584 -0.184 0417
Pre- LongTerm Change Score
CBC-Externalizing -0.781 -0.589 -0.389 -0.550
CBC-lntemalizing -0.330 -0.122 -0.539 -0.880
Parent Satisfaction
TAI-1 -0.437 -0.624 0.195 0.528
TAI-2 -0.006 0.267 0.035 -0.213
Motivational Model Two
PREDICTORS:
1.Val, LG 0724 -0.629 0.515 0.483
2. SE, OES, Sch -0.370 0231 0.378 0.176
3. PGE, PGD 0.123 0.175 -0.357 -0.024
4. PGE 0.399 0.403 0.057 0.011
5. Implicit 0.506 0.547 0.825 0.818
OUTCOMES:
Child Behaviour Changes
Pre- Post Change Score
CBC-Extemalizing 0.561 0.663 0.449 0.484
CBC-lntemalizing 0.016 -0.615 0.349 0.923
Pre- LongTerrn Change Score
CBC-Extemalizing 0.669 0.377 0.093 0.135
CBC-lntemalizing 0.207 0.153 -0.207 -1.417
Parent Satisfaction
TAI-1 -0.073 0.254 0.327 0.398
TAI-2 -0.590 -0.633 0.267 0.093

 

Model One 5 = .684, p < .168
Model Two 5 = .782, p < .081

175

Table 37

Correlations. Standardized Canonical Coefficients and Their Camnical Variates in

a Combined Model

 

 

 

 

First Variate Second Variate
Canonical Canonical
Variable Correlation coefficient Correlation coefficient
PREDICTORS:
HSQ-Severity -0.237 -0.559 0.294 0.197
Number of Stressors 0.338 0.686 0.002 0.680
Parent Depression -0.244 0.301 -0.361 -0.742
SES 0.076 -0.444 0.432 0.250
1.Val, LG 0.768 0.751 0.021 -0.505
2. SE, OES, Sch 0.395 0.099 0.239 0.292
3. PGE, PGD -0.158 0.180 -0.216 0.393
4. PGE -0.201 -0.271 -0.104 -0.343
5. Implicit -0.129 0.361 0.603 0.920
OUTCOMES:
Child Behaviour Changes
Pre— Post Change Score

CBC-Externalizing -0.254 -0.307 0.253 1.181
CBC-Internalizing 0.143 0.977 -0.185 0537

Pre- LongTerm Change Score
CBC-Externalizing -0.274 -O.210 -0.259 -0.559
CBC-lntemalizing -0.476 -0.675 -0.452 -0.867

Parent Satisfaction

TAI-1 0.021 -0.415 0.090 0.532
TAI-2 0.636 0.795 -0.081 -0.217

 

g = .908, p < .028

176

Table 38

Canonical Anglvsis of Predictors fl! Outcome Variables in PMT

 

Standardized

Canonical Coefficient

 

 

 

Variable Traditional Model Motivational Model Combined Model
Root 1 Root 2 Root 1 Root 2 Root 1 Root 2
PREDICTORS:
Traditional Predictors
HSQ-Severity -0.410 0.652 ------------ -0.559 0.197
Number of Stressors 1.172 0.077 ------------ 0.686 0.680
Parent Depression -0.077 -0.728 ------------ 0.301 -0.742
SES -0.315 0.418 ------------ -0.444 0.250
Motivational Factors
1.Val, LG ------------ -0.629 0.483 0.751 -0.505
2. SE, OES, Sch ------------ -0.231 0.176 0.099 0.292
3. PGE, PGD ------------ 0.175 -0.024 0.180 0.393
4. PGE ------------ 0.403 0.011 -0.271 -0.343
5. Implicit ------------ 0.547 0.818 0.361 0.920
OUTCOMES:
Child Behaviour Changes
Pre- minus Post-
CBC-Extemalizing -0.498 0.920 0.663 0.484 -0.307 1.181
CBC-lntemalizing 0.584 -0.417 -0.615 0.923 0.977 0537
Pre- minus Long-Term-
CBC—Extemalizing -0.589 -0.550 0.377 0.135 -0.210 -0.559
CBC-Internalizing -0. 122 -0.880 0.153 -1.417 -0.675 -0.867
Parent Satisfaction
TAI-1 -0.624 0.528 0.254 0.398 -0.415 0.532
TAI-2 0.267 -0.213 -0.633 0.093 0.795 0217

 

1 Model One; = .684, p< .168
2 Model Two 1 = .782, p < .081

3 Combined Model 1 = .908, p < .028

177

depression) and motivational predictor (factor 4) are moderately correlated.
Within the second set, both internalized behavior change scores (post and long
term-) as well as both parent satisfaction factors are highly correlated with the
first variate. Thus, in comparing the three models only the combined model
yielded a significant canonical variate (Table 38).

The results of the second analysis support Hypothesis 2c that a
combination of traditionally examined and motivational predictors would be most
predictive of outcomes in PMT but do not provide support for Hypotheses 2a and
2b that models of traditional predictors and motivational predictors, respectively,
would predict outcomes.

Analysis Ill: Classifying Accordinq to Drop Out

To test the hypothesis that PMT predictors would predict parent
engagement, the third analysis (discriminant function analysis) described a linear
combination of the predictor variables that best separated groups according to
point of drop, i.e., at orientation (pre-PMT), mid-way through the PMT program
(pari-PMT), or after long term follow up (completers).

Table 39 contains group means and standard deviations for traditional
predictors and motivational predictors. Note that the orientation drop out group
had higher mean scores on number of stressors, parent depression, and
Hollingshead SES than the other two groups. Also note that higher Hollingshead
Index scores actually reflect lower SES. The proportion of single parent families
was also higher for the orientation drop out group, next highest for the midway

drop out group, and lowest for the completer group. The proportion of low income

178

Table 39

Means and Standard Deviations of Predictor Variables as a Function of PMT DropOut Point

 

 

 

 

 

 

 

 

 

Point of Drop Outs
Pre-PMT Pari-PMT Completers

Predictor Variable M g M S___D M S79

Editiongl Predictors
HSQ-Severity 4.49 1.58 4.56 2.24 4.55 2.21
Number of Stressors 4.44 2.10 3.32 2.29 2.77 2.28
Parent Depression 16.06 12.15 13.50 10.10 11.34 10.41
SES 51.69 15.23 45.00 13.37 43.49 14.94
Single Parent Family 0.50 0.52 0.38 0.52 0.29 0.46
Low Income Cutoff 0.38 0.50 0.88 0.35 0.74 0.44

Motivational Predictors
1.Val, LG 0.18 1.09 0.52 0.86 0.40 0.70
2. SE. OES, Sch -0.28 1.05 -0.74 0.98 -0.67 0.93
3. PGE, PGD -0.05 1.22 -0.45 1.01 -0.63 1.00
4. PGE 0.09 1.35 0.17 1.10 -0.22 0.89
5. Implicit -0.03 1.18 -0.07 0.67 0.22 1.31

 

Note 1. Single Parent Family is coded 0 for no (married, step-parent, common-law, etc.)
and 1 for yes (single)and Low Income Cutoff is coded 0 for low and 1 for high.

Thus, the mean of these variables is theproportion of cases with a value of 1.

Note 2. Motivational predictors are standardized factor scores, z-scores with a mean of 0 and
standard deviation of 1.

179

was highest for orientation drop outs and next highest for completers. The
midpoint drop out group had the lowest proportion of low income among the
three groups. Among the three groups, the completer group had the lowest
performance goal and the highest implicit theorist mean scores; the pari- drop out
group had the lowest self efficacy/optimistic explanatory style/schematicity,
lowest implicit theorist, highest value and learning, and highest self enhancing
performance goal mean scores; and, the orientation drop out group had the
lowest value/learning goal, highest self efficacy/optimistic explanatory
style/schematicity, and highest self defeating performance goal mean scores.
Traditional Model

The measure of the degree of association between the traditional
discriminant scores and the drop out groups was not significant for either the first
function (p < .125) or the second function (p < .676). Correlations between the
values of the functions and values of the variables indicated that low income
cutoff had the highest correlation with the discriminant function (.769), number of
stressors had the second largest correlation (.544), and single parent family had
the third largest (-.360) in absolute value (Table 40). The latter’s negative value
indicates that small function values are associated with the presence of single
parent families and high values are associated with non-single parent
households. These results are similar to those obtained by the standardized
coefficients (Tables 41 8- 42).

The discriminant function for the traditional model correctly classified

64.4% of the originally grouped cases (Table 43). Of 35 parents who completed

180

Table 40

Wes in Discriminant Fu_nction Analysis by Traditional, Motivational,
and Combined Models

 

Equivalent Significance

 

 

 

 

 

 

Predictor Variables Wilks's Lambda F(df )
IgditionaJ Predictor;
HSQ-Severity 1.000 0.005 0.995
Number of Stressors 0.898 3.166 0.050
Parent Depression 0.964 1.061 0.353
Low Income Cutoff 0.856 4.727 0.013
Single Parent Family 0.962 1.092 0.343
Motivatlmal Predickg
1.Val, LG 0.981 0.489 0.616
2. SE, OES, Sch 0.966 0.900 0.413
3. PGE, PGD 0.980 0.525 0.595
4. PGE 0.975 0.655 0.524
5. Implicit 0.988 0.304 0.739
Combined Model
HSQ-Severity 0.989 0.262 0.771
Number of Stressors 0.924 1.972 0.150
Parent Depression 0.937 1 .162 0.209
Low Income Cutoff 0.914 2.249 0.116
Single Parent Family 0.963 0.913 0.408
1.Val, LG 0.972 0.700 0.502
2. SE, OES, Sch 0.973 0.666 0.518
3. PGE, PGD 0.979 0.520 0.598
4. PGE 0.957 1.083 0.347
5. Implicit 0.993 0.158 0.854

 

Note. Degrees of freedom for Traditional Model, Motivational Model, and Combined Model

are F(2, 56), F (2, 51), and F (2, 48), respectively.

*p<.05

181

Table 41

Correlation of Predictor Variables with Discnmmrit Functions (Function Structgre Matrix)
and Standardized Discriminant Fugion Coefficients

 

 

Correlation with Standardized discriminant
discriminant function function coefficients
Predictor Variable Function 1 Function 2 Function 1 Function 2

mitim Predictors

 

 

 

HSQ-Severity 0.026 0.008 0.518 -0.063
Number of Stressors -0.610 0.544 -0.451 0.617
Parent Depression -0.345 0.364 -0.393 0.296
Low Income Cutoff 0.769 0.470 0.778 0.860
Single Parent Family 0360 0.305 0.062 0.503
Motivational Predictors
1.Val, LG 0436 0.234 -0.493 0.271
2. SE, OES, Sch 0.612 -0.107 0.690 -0.275
3. PGE, PGD 0.464 -0.135 0.514 -0.254
4. PGE 0.344 0.759 0.118 0.857
5. Implicit -0.245 -0.499 -0.340 0446
Combined Predictors
HSQ-Severity -0.055 -0.300 -0.345 -0.320
Number of Stressors 0.486 0.250 0.531 0.570
Parent Depression 0.451 -0.148 0.499 0172
Low Income Cutoff -0.481 0.423 -0.316 0.553
Single Parent Family 0.345 -0.028 0.145 0.133
1.Val, LG 0277 0.205 -0.543 0.230
2. SE, OES, Sch 0.264 -0.223 0.204 -0.352
3. PGE, PGD 0.245 -0.151 0.119 0.005
4. PGE 0.217 0.522 0.267 0.795
5. Implicit -0.082 -0.201 -0.050 0.171

 

182

Table 42

Correlations Between Discriminating Variables and Discriminant Functions
(Function Structure Matrix)

 

 

Variable Function 1 Function 2

 

T_raditional Predictors

 

 

HSQ-Severity 0.026 * 0.008
Number of Stressors -0.610 * 0.544
Parent Depression -0.345 0.364 *
Low Income Cutoff 0.769 * 0.470
Single Parent Family 0360 * 0.305
Motivational Factors
1.VaI, LG 0436 * 0.234
2. SE, OES, Sch 0.612 * -0.107
3. PGE, PGD 0.464 * -0.135
4. PGE 0.344 0.759 *
5. Implicit -0.245 -0.499 *
Combined Model
HSQ-Severity -0.055 -0.300 *
Number of Stressors 0.486 * 0.250
Parent Depression 0.451 * -0.148
SES -0.481 * 0.423
Single Parent Family 0.345 * -0.028
1.Val, LG 0277 * 0.205
2. SE, OES, Sch 0.264 * -0.223
3. PGE, PGD 0.245 * -0.151
4. PGE 0.217 0.522 *
5. Implicit -0.082 -0.201 *

 

183

* largest absolute correlation between each variable and any discriminant function

 

. axe «..00 u 00000 000.000... 2.00000 .0 000.000.00 ..0.0>O .0.0z

 

 

 

 

0.00 .0 o o v. . . v 00 .0.000..00

0.00 .. o 0 0...... . 0 so 020 020-000

0.00 0 o o 0.0.. N 0. .8 00.0 02.0.0000

oxo 0- .x. 0. 0.. m 0 0.00.000.0E 000.0 .0:.0<
.0.0_0E00 ....20-..00 00..0.00..O

 

0.00.00E0s. 000.0 00.0.8.0

 

>.0O 0.0.0.00.fl.000...00.... 00.0: .00 00.0 .0. 0.0>.-00< 0000050006

0? 0.00..-

184

the program, 31 were predicted correctly to be members of the completer group
(88.6 %), while 4 were assigned incorrectly to the orientation drop out group
(11.4 %). Seven (43.8 %) of the parents who dropped out at orientation were
classified correctly while 4 (11.4 %) were misclassified as completing the
program and 1 (12.5 %) was misclassified as dropping out midway through the
program. Even worse, 0 (0%) of parents who dropped out mid-way through PMT
were correctly classified while 1 (12.5 %) was misclassified as dropping out at
orientation, and 7 (87.5 %) were misclassified as completing the program. From
Figure 9, we can see the boundaries which mark off the combination of function
values that result in the classification of the cases into two of the three groups,
i.e., there are no boundaries for the pari-PMT drop out group.

To put these results into decision-making language, the use of the TEP
model to predict PMT drop out outcomes would result in 6.8% false positives,
27.1 % false negatives, and 66.1 % true/correct.

Severity of child behavior and parent depression had the highest
correlation coefficient ( r = .424); and, low income cuttoff and single parent family
have the next highest correlation ( r = -.416). lnterdependencies among the
variables appear to be minimal; thus, an interdependency effect on this
multivariate analysis was minimal.

Motivational MM

The measure of the degree of association between the motivational

discriminant scores and the drop out groups was also not significant for either the

first function (p < .292) or the second function (p < .157). Correlations between

185

Canonical Discriminant

Function 2

 

 

 

-3.0 —2.0 -1.0 .o 1.0 2.0 .o
1. 1 1 1 1 1 1
T l i l T
3.0 i 12
12
12
12
12
12
-~ + + +12 + +
12
12
12
12
12
-- + + +12 + +
12
12
12
12
12
.o 4 + + * 1132 * + +
1133 32
1133* 32
1133 32
133 32
113 32
-1.0 4 + 1133 + 32+ +
1133 32
1133 32
1133 32
1133 32
133 32
-2.0 - 113 + + + 32 +
1133 32
1133 32
1133 32
1133 32
33 32
-3.0 1 32
1 1 1 1 1 1 1
T i l l l 1 i
-3.0 -2.0 -1.0 .o 1.0 2.0 .o

Canonical Discriminant Function 1

Figure 9 — Territorial Map with Traditional Model

186

 

the values of the functions and values of the variables indicated that factor 2 had
the highest correlation with the discriminant function (.612), factor 3 had the
second largest correlation (.464), and factor 1 the third largest (-.436) in absolute
value (Table 41). The latter’s negative value indicates that small function values
are associated with the presence of higher value and learning goals and larger
values are associated with the absence of value and learning goals. These
results are also similar to those obtained by the standardized coefficients (Table
42).

The discriminant function for the motivational model correctly classified
64.8% of the originally grouped cases (Table 44). Of 33 parents who completed
the program, 31 were predicted correctly to be members of the completer group
(93.9%), while 2 were assigned incorrectly to the orientation drop out group (6.1
%). Four (26.7%) of the parents who dropped out at orientation were classified
correctly while 11 (73.3 %) were misclassified. Again, 0 (0%) of the parents who
dropped out mid-way through PMT were correctly classified while 1 (16.7 %) was
misclassified as dropping out at orientation, and 5 (83.3%) were misclassified as
completing the program. The numbered boundaries mark off the combination of
function values that result in the classification of parents into the three groups
(Figure 10).

To put these results into decision-making language, the use of the
motivational model to predict PMT drop out outcomes would result in 3.7% false

positives, 29.6 % false negatives, and 66.7 % true.

187

 

.0.. 0.3 u 00000 0050000 2.00000 .0 000.0800 __0.0>0 .0.02

 

 

 

 

 

0.00 .0 o o ..0 0 00 0.00050
0.00 0 o o 0.0. . 0 5o 090 020-000
0.2 I o 0 Now 0 m. .00 005 02.0.0000
.0.. m .0.. m .00 m 0 200000.000. 0:05 _0:.0<
00.0.0800 0.5.0-000 02.0.0000
2000000002 000.0 00.20000

 

 

>_00 00.20000 .002.0>_.0_2 020: .30 005 00. 0_0>_00< 0000050020

3 030...

188

Canonical Discriminant
Function 2

 

-3.0 -2.0 -1.0 .0 1.0 2.0 .0
l l l l l l l
I I I I I 4T
3.0 w
3
32

3
13333
111133333
1.0 0 111113333 + +
111133333
111113333
111133333
111113333
1111

-o + + + +

1»

-1.0 0 + + +

 

l I

1* 1 1 .
0 -2.0 -1.0 .0

332
3322
12
12
12
12
12
12

12
12
12
12
12
12
12
12
12
12
12
12
12

l
I
1.0

Canonical Discriminant Function 1

2.

Figure 10 — Territorial Map with Motivational Model

189

+

o-+—

.CD—II—

 

Since factor 2 and 4 have the largest correlation coefficient (.169)
interdependencies among the variables appear to be minimal; thus, an
interdependency affect on this multivariate analysis was minimal. However in the
motivational model, Box’s M F test indicated significant results (p < .003),
causing us to reject the null hypothesis of equal population covariance matrices
and conclude that one of the assumptions of discriminant analysis was violated.

Combined Model

 

The measure of the degree of association between the discriminant
scores and the drop out groups was also not significant for the combined model.
The significance level for the first function was .678, indicating that it did not
contribute substantially to group differences. Correlations between the values of
the functions and values of the variables indicated that number of stressors had
the highest correlation with the discriminant function (-.543), low income cutoff
had the second largest correlation (-.481), and parent depression had the third
largest (.451) (Table 42). These results are also similar to those obtained by the
standardized coefficients (Table 41).

The discriminant function for the combined model correctly classified 62.7
% of the originally grouped cases (Table 45). 0f 31 parents who completed the
program, 27 were predicted correctly to be members of the completer group
(77.4 %), while 3 were assigned incorrectly to the orientation drop out group (6.1
%) and 1 to the pan-PMT group (6.5 %). Five (28.6 %) of the parents who
dropped out at orientation were classified correctly while 9 (64.3 %) were

misclassified. Once more, 0 (0%) of parents who dropped out mid-way through

190

. .0.. fine u 00000 0000020 2.00000 00 00000800 __000>0 .0.02

 

 

 

 

E0 00 0.0 . 0.0 0 .0 00.200000

0.00 0 o o 0.0. . 0 so 02.0 020-000

0.00 0 o o 0.00 0 3 .8 02.0 02.0.0000

.00 m 00 m 00 m .10. 2000000000. 00000 _0:.0<
00.200000 020-000 02.0.0000

 

200009005. 0020 00.2005

 

0022 0020.000 0I0_ 000.2005 _002.0>_.0_2 000 00220000 0.0m 0203 .00 005 .0.. 20205.. 0000050005

mv 0B0...

 

191

PMT were correctly classified while 1 ( 28.6 %) was misclassified as dropping
out at orientation, and 5 ( 83.3 %) were misclassified as completing the program.
From Figure 11, we can see the numbered boundaries which mark off the
combination of function values that result in the classification of parents into the
three groups.

To put these results into decision-making language, the use of the
combined model to predict PMT drop out outcomes would result in 7.8 % false
positives, 27.5 % false negatives, and 64.7 % true.

Factor 3 and parent depression had the largest correlation coefficient,
.325, factor 5 and single parent families had the next largest, .304, and then
factor 3 and severity of child behavior, .276. Since these correlations were not
particularly large, interdependencies among the variables appeared to be
minimal once more; thus, an interdependency affect on this multivariate analysis
was also minimal. However, Box’s Test of Equality of Covariance Matrices
indicated significant results (p < .001), causing us to reject the null hypothesis of
equal population covariance matrices and conclude that one of the assumptions
of discriminant analysis was violated.

The results of the third analysis do not support the hypothesis that the
traditionally examined predictors, the motivational predictors, and. a combination
of both would predict parent engagement in PMT.

Analysis IV: Individual Motivational Resources

 

The fourth analysis (logistic regression) tested the hypothesis that the

motivational aspect subscales would be useful in predicting parent drop out or

192

Canonical Discriminant
Function 2

 

 

 

 

-3.0 —2.0 -1.0 o 1.0 2. .o
l l l l l I
I I I I I
3.0-1- 32 «1-
32
32
32
32
32
+ + + +32 «-
32
32
32
32
32
+ + + 32 + --
32
32
* 32
32
33333333333333332
.0 «3333333333333333311111111111*1112 ‘ + v
11111111111111111 12
12
12
12
12
+ + + 12 + ‘-
12
12
12
12
12
-2 o - + + + 12 + 0
12
12
12
12
12
-3.o~l 12 «l
l. l J l l .1
I I I I I I
-3.0 -2.0 -1.0 .0 1.0 2.0 .0

Canonical Discriminant Function 1

Figure 11 — Territorial Map with Combined Model

193

completion of the PMT program (See Appendix D for Rationale for Using
Subscales). Table 46 provides regression coefficients and significance levels for
the variables in the model. Variables with positive root values indicate that as the
variable values increase, the likelihood of parents completing the program
increases. For example, as self efficacy increases (p < .0491 ), the likelihood of
parents completing the program increases and as pessimistic explanatory style
(p < .0010) increases, the likelihood of parents completing the program
decreases. Variables with small root values indicate the variable has a small
partial contribution. For example, the low root values for optimistic explanatory
style, learning goals, both performance goals, and schematicity indicate that
these motivational aspects have a small partial contribution. Interaction effects
indicate that the interaction between pessimistic explanatory style and value is
significant ( p < .0007).

Comparing predictions to the observed outcomes, 33 parents who
completed the program were correctly predicted by the model not to dropout.
Similarly, 19 parents who dropped out were correctly predicted to prematurely
terminate. A total of 11 parents were incorrectly classified, 4 as drop outs and 7
as completers. Of the parents who did not drop out, i.e., completed the program,
89.19% were correctly classified. Of the parents who dropped out, 73.08 % were
correctly classified. Overall, 82.54% of parents were correctly classified by the
model.

The results of the fourth analysis support the hypothesis that a

combination of select individual motivational resources predict parent

194

Table 46

Logistic Regression Predicting PMT Drop Out (n=64)

 

 

 

Varable Beta SE Odds Wald Sig.
Ratio Statistic
Self Efficacy 0.341 0.173 0.148 3.872 0.049 *
Optimistic Explanatory Style 1.385 1.144 0.000 1.465 0.226
(OES)
Pessimistic Explanatory Style -10.784 3.277 -0.322 10.832 0.001 ***
(PES)
Learning Goal -0.144 0.142 0.000 1.037 0.309
Performance Goal 0.129 0.134 0.000 0.928 0.335
-- self defeating
Performance Goal 0.179 0.130 0.000 1.901 0.168
-- self enhancing

Implicit Theory -0.324 0.173 -0.133 3.518 0.061
Schematicity -O.139 0.260 0.000 0.287 0.592
Value -O.633 0.322 -0.148 3.865 0.049 *
OES * Value -0.035 0.025 0.000 1.891 0.169
PES * Value 0.268 0.079 0.332 11.891 0.001 ***
* p_< .05
** g < .01
*** p_ < .001

195

engagement, i.e., drop out. Furthermore, the results also support the hypothesis
that an interaction between pessimistic explanatory style and value predicts
parent drop out from PMT and thus, provides support for the expectanch value
model of motivation.

Analysis V: Univariate Examination of Discrete Variables

 

The fifth analysis tested the hypothesis that two discrete motivational
aspects (implicit theorist orientation and schematicity) as well as select
demographic variables would be associated with drop out. This chi-square
analysis identified isolated associations among predictors and the drop out
outcomes. Approaching significant associations were found between point of
drop out in parent training and 1) low income cutoffs (Chi-square = 8.188, df = 2,
p < .017) and 2) early developmental milestones (Chi-square = 10.077, df = 4, p
< .039) (Tables 47 and 48). Using a Bonferroni adjustment (p. < .004), however,
indicates that these associations are not significant. All other chi-square analyses
also revealed insignificant results (Table 49).

The results of the fifth analysis do not support the hypothesis that parent
implicit theorist orientation predicts parent drop out. The results also do not
support the hypothesis that parents schematicity predicts drop out. However, the
results do support the hypothesis that select demographic variables may be

associated with drop out.

196

Table 47

Prevalence (%) of Statistics Canada Income Cutoffs within Points of Drop Out

 

 

 

Point of Drop Out
Cutoff No Pre- Pari- Chi- p
(_ = 37) (g = 18) (_ = 9) Square
Low 29.7 66.7 22.2 8.188 0.017
High 70.3 33.3 77.8

 

* p < .004 (with Bonferroni adjustment)

Table 48

Prevalence (%) of Ea_rlv Developmental Milestones within Points of Drop Out

 

 

 

Point of Drop Out
Cutoff No Pre- Pari- Chi- Q
U.) = 37) (fl = 18) (11 = 9) Square
within normal limits 89.2 77.8 50.0 10.077 0.039
late 8.1 11.1 50.0
early 2.7 11.1 0.0

 

* p < .004 (with Bonferroni adjustment)

197

Table 49

Chi-Square Analyses of finicipgnt Char_a_cteristics within Points of Drop Out

 

 

Characteristics Chi- df Q
Square

Parent Education 7.088 8 0.527
Parent Ethnicity 4.348 2 0.114
Current Annual Income 29.976 20 0.070
Social Class (Hollingshead) 7.641 8 0.469
Child Medication Usage 7.901 4 0.095
Single Parent Household 1.697 2 0.428
Child Diagnosis 14.768 10 0.141
Previous Treatments 3.245 8 0.918
Schematic/Aschematic 2.839 2 0.242
incremental/Implicit Theorist 0.307 2 0.858

 

* p < .004 (with Bonferroni adjustment)

198

Analysis VI: Changes in Parent Motivational Schema

 

To test the hypothesis that parents in the treatment group would show
change in motivational resources while parents in the control group would not
show change as a result of participating in PMT, a series of t-tests were
conducted (Tables 50 & 51). Parents in the treatment group demonstrated
significant change from pre- to post- PMSS administration on the self efficacy
and approaching significance on the schematicity subscales (Table 50). Parents
in the control group demonstrated no significant differences on the pre- and post-
subscales. Parents in the treatment group demonstrated significant change from
pre— to post- PMSS administration on the factor scales (Table 51). The results
support the hypothesis that treatment group participants would show change on
motivational resources but only on select motivational resources (self efficacy
and possibly, schematicity). The results support the hypothesis that the control

group would show no change.

199

Table 50

Mean Motivational Scores and T-test for Pre- and Post- Subscales by
Treatment (n=37) and Control Group (n=29)

 

Subscale Score

 

 

 

PMSS Pre- Post- Pair-wise
Subscale Mean (Sd) Mean (Sd) T-Test Signif.
Treatment
SE 19.340 (2.45) 20.530 (2.53) -3.282 0.003 *
OES 11.560 (2.37) 11.910 (2.90) -0.809 0.425
PES 3.310 (1.35) 3.340 (1.00) -0.120 0.905
LG 22.270 (3.29) 22.310 (3.62) -0.068 0.947
PGD 10.890 (3.44) 11.030 (6.38) -0.132 0.896
PGE 9.380 (2.95) 8.970 (3.11) 1.132 0.266
IMP 5.250 (2.17) 4.840 (1.92) 0.871 0.390
SCH 5.220 (1.52) 5.970 (1.45) -3.000 0.005 *
VAL 42.280 (5.15) 41.310 (6.55) 0.935 0.357
Control
SE 18.850 (2.34) 19.150 (2.82) -0.433 0.673
OES 11.310 (2.36) 11.460 (2.22) -0.237 0.817
PES 2.690 (0.85) 2.460 (0.66) 1.389 0.190
LG 19.230 (3.94) 19.150 (2.54) 0.078 0.939
PGD 11.230 (3.54) 10.150 (5.24) 0.735 0.476
PGE 8.380 (1.89) 8.310 (2.93) 0.114 p 0.911
IMP 5.620 (2.10) 5.770 (2.59) -0.213 0.835
SCH 6.380 (1.71) 5.310 (1.93) 2.344 0.037
VAL 33.150 (9.51) 32.540 (10.78) 0.273 0.789

 

*significance criteria of .0056 (with Bonferroni adjustment)

200

Table 51

Mean Motivational Scores and T-test for Pre- and Post- Factors by
Treatment (n=37) and Control Group (n=29)

 

Subscale Score

 

 

 

PMSS Pre- Post- Pair-wise

Factor Mean (Sd) Mean (Sd) T-Test Signif.
Treatment

LV 58.28 (4.71) 58.03 (4.47) -4.669 0.869

IMC 23.56 (7.57) 26.78 (8.66) 0.166 0.000 *

PGD 11.31 (3.34) 11.61 (5.98) —0.356 0.724

PGE 5.51 (2.33) 5.05 (2.36) 1.536 0.133

IT 5.41 (2.31) 4.81 (1.88) 1.410 0.167

Control

LV 47.08 (12.22) 46.00 (12.40) 0.399 0.697

IMC 25.07 (4.65) 24.71 (5.85) 0.324 0.751

PGD 11.23 (3.90) 11.46 (6.06) -0.170 0.868

PGE 5.43 (2.24) 4.93 (2.76) 1.202 0.251

IT 5.43 (2.14) 5.57 (2.59) -0.213 0.834

 

significance criteria of .01 (with Bonferroni adjustment)

201

Chapter 5
DISCUSSION

Summary of Results

 

The study findings support several of the overall hypothesis and do not
support several others. Table 52 provides a summary of these findings. A post-
hoc analysis reveals that power is low for the current study. As a result, low
power may be a partial explanation for the nonsignificant results (see Appendix F
for Power Analysis Summary).

Hypothesis I: Treatment Effects

 

The study findings do not formally support the hypothesis that treatment
group participants will report improvements in child behavior in comparison to
control group participants in PMT. However, the results did identify a significant
main effect for the CBCL behavior scale and for the interaction between scale
and group. Thus, it was established that ratings for each of the CBCL scales
depend on the group variable.

Several explanations are consistent with the lack of significant results in
treatment effects. Only about half of the parents who complete PMT programs,
report a significant change in child behavior (Johnston, 1996). While parent
management sessions typically produce changes in child behavior as rated by
parents, changes do not bring children within normal limits and are considerably
weaker than changes produced from medication (Hinshaw et al., 1998). PMT in

combination with social skills training did not find any significant difference

202

>._

own.

.5

8323 S btm>om
9885 Go .mnEsc

0:0...

20326 .0 .mnEac
3323 ho 3:93

8399.8 cacomamzmm 298 new
notoaqsm .828ch .6323 35 8:55 £20665
Co «69¢ 58 Co cozchEoo .0

20:92.8 cozofimzmm E98 new 5328 2:0

62.53% Co: $2550 8605 mEmcom co=m>=22 .n
5339.8 cozoflmzmm E23 was 09.20

notoaaam $2550 .03ch Eco 869a amt .m

2228.5 303528 E 86:65 3 =3
3.3:? oEooSo msosczcoo ._

 

m>oc< 3:98 65:8 899 EoEfioF
a_nmeaam 6c votoaanm 6c .9903 PPN “maootm EoEfiot. _
£9285 co_m_ooo m_m>_m:< 5:98me 3639590. £8509»:
236228536

 

9928i 5:5 mco_m_omo ocm .mow>_mc< $66365an $8.:on 056on Go meEnw

Nm .2an

203

 

6m
mm

038:3: E:
_m> . mum

mun.
mm

0CD:

0:0:

0:05

3:033
25:8

notoaazm E:

notoaanm

:98an E:

notoaaam E:

notoaazm E:

663 8:523

99535

:o_mm9a9
27.52

330:2
EmcEEQmB

:ozoce
Em:_E_._om_c
88:2
EmEEEmE

89:89 .mcozgzoE :_ 3:20 a
mocotoaxo P.2d :_ 99:2th 2.3 9:93

So :9: :Es 3383mm

9m 203953 303985 .mE9mn.

So no.6
869: 89:89 .m:o=m>=oE 3336:.

So :9: 869: 92069:
E womb Son E :ozmcfiEoo .0

So :9: E93
“069: mEocom :o=m>=o_2 .n

So :9: 869: mm» .m

99069: mzo::_Eoo E 69069: on =3)
:30 :93 23:9 oEooSo 9985

3:00 Nm 9an

>_

204

between parents who received training and those who did not (Finch, 1998). It
was also noted that social skills training alone did not produce any significant
differences between treatment and control groups (Finch, 1998). These results
are surprising given the well established efficacy of PMT programs (Kazdin,
1997; Kazdin & Weisz, 1998).

PMT programs which run less than 8 weeks are generally less successful
(Barkley, 1997). The current PMT may have been too short and, thus, provided
limited availability and time for the parents to implement and practice the
strategies and techniques discussed. In fact, the better change scores from pre-
and long term CBCL scales for the treatment group, support this latter
interpretation. It is also possible that parents who are high on self efficacy are
likely to show better child behavior change outcomes regardless of group.
Parents who are more disadvantaged on traditionally examined predictors are
likely to show worse child behavior change outcomes regardless of group.

The significant scale main effect and scale by group interaction are also
consistent with several explanations. The increased internalizing problem scale
results among control group participants is a summary of both pre- and post-
measures. It is possible that the control group experienced increased anxiety
regarding the treatment over the wait time and that this increase in worry,
frustration, and/or anticipation is translated to the parents' perceptions of child
behavior. Parental internal states may have influenced ratings of their children.
Furthermore, the high incidence of comorbidity among anxiety and ADHD and

other disruptive behavior disorders (Stahl & Clarizio, 1999) is consistent with

205

these findings. This result may suggest that parents and children with disruptive
behavior disorders may be at risk for further developing internalizing problems if
left untreated. Comorbidity involving both internalizing and externalizing problems
may partially explain why good outcome measures were not achieved as well as
have implications for future choice of treatment (Stahl & Clarizio, 1999).

Finally, a selection bias may have been in effect in group assignment due
to the small sample size and provides an alternate explanation for this difference.
Parent participants in one particular group spent considerable energy and
resources discussing outside factors, especially school climate, principal
philosophies, and teacher practices inside the classroom and during recess.
These parents firmly believed that these factors were strongly influencing their
children’s behavior and impacting on the efficacy of implementing new parenting
skills. While acknowledging the importance of context, these comments also
revealed perceptions consistent with several motivational aspects including self
efficacy, pessimistic explanatory style, and incremental theories.

The high drop out rate in the current study, discussed below, may also
provide a partial explanation for the differences between treatment and control
groups on scale. It may be that certain participants who dropped out of the
treatment group may have differed from those who stayed on some very
important dimensions/characterstics.

The control group’s increases in behavior problems from pre- to post-
measures are diametrically contrary to the literature which consistently

demonstrates drops in behavior problem rating scores as a direct result of

206

retesting. In school psychology intervention research involving rating forms,
participants have been observed to indicate lower ratings on a second measure
(Barkley, 1997). This increase is also contrary to regression towards to the mean
as a result of retesting.

The present results may have differed from previous research because of
possible differences in wait periods for Canadian and United States clinic
populations. Many of the participants from both the treatment and wait list control
groups had been on an extensively long wait list due to the universal access to
health care in Canada. The control group participants may have had their hopes
raised due to finally being seen but then experienced increased anxieties once
placed on another wait list, as noted above.

Hypotheses II, III, and IV LA” Predictors and Outcomes)

 

Three comparisons of models (traditional, motivational, and
combined predictors) were conducted for several multivariate statistical
procedures. The combined models in the canonical analysis predicted
significance after controlling for the traditional and motivational models,
respectively, suggesting that it may be possible to make a difference in
PMT outcomes by considering and eventually manipulating the more
dynamic motivational predictors.

The results of the canonical correlation supports the hypothesis about the
relationship between the two sets of variables for a sample of parents
participating in a parent management training program. One set of variables

measured traditionally examined predictors and motivational predictors and the

207

other set measured outcomes of PMT. The traditional predictors with the largest
impact included severity of child behavior, parent depression, and number of
psychosocial stressors. The motivational predictors included the five new
motivational factors (LV, IMC, PGE, and IT). The outcome variables included
child behavior change and parent satisfaction measures. While the results
determined that motivational predictors alone were not significantly related to
outcomes, weights assigned to coefficients suggest that certain predictors have
more impact and loadings than other predictors on outcomes of PMT for a
sample of parents at an outpatient health care organization.

Brestan et al. (1999) found that parent expression of satisfaction may be
more closely linked to symptom changes during treatment than to the levels of
child behavior problems after treatment as measured by absolute scales. These
findings are consistent with the stronger relations in the models utilizing behavior
change scores. Furthermore, correlations between parent satisfaction and
behavior problem severity were not significant in previous research studies
(Brestan et al., 1999).

The results of the discriminant analyses suggest that parents who are at
high risk for dropping out of PMT cannot be identified early using the predictors in
the current study. The discriminant function analysis failed to significantly classify
group membership in the traditional predictor model, motivational predictor
model, or combined model. However, classification percentages across
traditional, motivational, and combined models suggested good rates for

predicting completers. (88.6%, 93.9%, and 87.1 %, respectively), and adequate

208

rates for predicting orientation drop outs (56.3%, 73.3%, and 64.3%,
respectively). It is predicting this latter group and identifying most cases of the
smaller group which holds the most value. For example, by identifying everyone
to be completers of a PMT program, the error rate will be very small, since fewer
people actually drop out. However, the results are less valuable, since the goal is
to identify the drop out parents. The intent of the disriminant analysis was to
identify parents to receive special, extended treatment prior to PMT.

Drop out rates are reportedly high in PMT, with estimates ranging from
25 per cent, 40-60 per cent, and 46.86 per cent, respectively (Johnston, 1996;
Kazdin, et al., 1997; Wierzbicki & Pekarik, 1993). Some of the reasons for
attrition and drop out have included subject geographical re-location and
reduction in motivation (Peterson & Bell-Dolan, 1995) and identified social
emotional disorders as at high risk (Fomess & Hoagwood, 1993). Attempts at
preventing drop outs through management of the more static variables included
minimal payments, accommodating parent shift-work, and establishing special
pretreatment techniques to establish a relationship between experimenter and
participants at orientation (Peterson & Bell-Dolan, 1995). Similar attempts to
prevent drop outs in the current study were not particularly successful given the
high drop out rate (42%). It is possible that the high drop out rates demonstrate
the need for additional treatment.

Attrition refers to important differences among participants who drop out of

one group, participants who drop out of other groups, and/or participants who

remain in the study (AATBS, 1995). This attrition is especially problematic in

209

longitudinal research studies in which the final sample is not comparable to the
initial sample because of characteristics of the participants who dropped out
(AATBS, 1995). Furthermore, attrition increases the probability of making a Type
II error (in which no differences are found when differences exist) due to loss of
power (Black & Holden, 1995).

While it was hoped that the discriminant analysis would aid in the
prediction of drop outs and it did not, the results do not preclude the use of
motivational strategies and techniques to prevent premature termination. Mean
scores suggest lower levels of value and Ieaming goals and higher levels of
performance goals among drop out participants. Special treatment and care
continues to need to be administered and can be tailored to target these
motivational aspects with these parents to encourage completion of the program.

The results of the logistic regression analysis support the hypothesis that
original motivational aspects predict drop out among parents involved in PMT.
Self efficacy, value, and pessimistic explanatory style appeared to have the most
contribution to the model. The significant interaction between pessimistic
explanatory style and value also supports the hypothesis of an expectancy times
value effect.

Hypothesis V (Individual Predictors)

 

The results of the univariate analysis (chi-square) support the hypothesis
that one traditional predictor (SES) was associated with drop out of PMT but the
two other motivational predictors measured (schematicity and implicit theories)

were not. The fact that SES failed to be clearly associated with outcomes in

210

many of the other multivariate analyses suggests that SES as a predictor is not
as important within the context of other factors. It is also possible that the effects
of SES were masked due to associations with other variables. Finally, previous
studies have noted that SES is associated with type of PMT program preferred,
i.e., low SES parents prefer more behaviorally orientated programs such as
Barkley’s Defiant Children (Wood & Baker, 1999).

Hypothesis VI (Changes in Motivational Aspects due to Participatiory

Some of the motivational mediators may have been implicitly influenced by the
nature of the parent training program, itself. Similar to behavioral parent training
interventions (Tucker et al., 1998), it is possible that the PMT intervention in the
current study increased parenting self-efficacy through the following sources of
information: a) mastery attainment as parents practiced newly learned skills
using standard homework assignments; b) vicarious learning experiences
through observing other parents and listening to group discussions; c) verbal
reinforcement from other parents and the therapist. These sources of information
appear to be important, inherent components of the PMT intervention and may
have lead to changes in parent motivational resources. For example, the PMT
program was conducted with a group of similar parents. These parents may have
provided coping models in the role plays and other activities, and thus functioned
as effective strategies for increasing self efficacy beliefs. The informative nature
of PMT also implicitly enhances self efficacy. This provides a strong argument for
including strategies which explicitly address motivating mediators. These

motivational enhancing strategies can be easily included in PMT programs.

211

The argument that the nature of PMT itself predisposes parents to a
pessimistic explanatory style can be easily countered. Granted, once their child is
diagnosed / labeled with a disorder, an entire gamut of motivational resources
come into play. Many programs explicitly educate parents that ADHD is
neurodevelopmental and therefore sets the parent up for an internal, stable,
uncontrollable explanatory style which is pessimistic but possibly, realistic.
However, the behavioral manifestations of ADHD relate to similar causal
dimensions relevant to the other disruptive behavior disorders, as follows. Before
beginning PMT, many parents are explicitly counselled that ODD and CD are
largely environmental and can be influenced by factors such as parental
consistency and predictability, and monitoring and tracking of behavior,
respectively (Barkley, 1997). Ultimately, this practice sets the stage for a more
external, unstable, and controllable, therefore, optimistic explanatory style as well
as for a more incremental theorist orientation and increasing value for the
therapy. It is this latter element, value for treatment which has been indirectly
examined as a barrier to treatment (Kazdin & Wassell, 1999). Perceived
relevance of treatment (which was defined as the extent to which treatment was
seen as relevant to the child’s problems and met with parent expectations) was
one of the sources of barriers which was strongly related to child behavior
change (Kazdin & Wassell, 1999).

Further Comments

 

Overall, the traditional predictors appeared to have the largest and most

significant impact on the outcomes measured in this study. Unfortunately, the

212

traditional predictors are primarily subject or demographic variables (Kazdin &
Wassell, 1999). The traditional predictors can also be viewed as static predictors
in that they represent aspects of the participants’ characteristics that are
predictive of outcomes but are not believed to be as readily subject to change.
Several of the motivational predictors in the current study also appeared to have
an impact on outcomes but not to quite as clear an extent. The motivational
predictors would have been more dynamic predictive factors which reflect the
present circumstances and motivation of the participant. Thus, the motivational
predictors may have the potential to be more readily amenable to change.
Dynamic predictors tend to be more unreliable and may change over time.
The lower reliability coefficients among the motivational aspects for the treatment
group than for the control group suggests that the PMT may have been
incidentally addressing some of the motivational aspects throughout the program.
Because the control group did not have exposure to the PMT program at pre-
and post- PMSS measurement points, their reliability estimates may be more
reflective of the stability of the motivational scale. Measurements of the dynamic
predictors involves a great deal of subjectivity. Any unreliability in measurement
leads to underestimation of validity. Therefore, this line of reasoning implies that
dynamic variables must be weaker predictors in comparison to static predictors.
Because by their very nature motivational predictors are necessarily weaker and
because manipulation of the traditional predictors did not enhance outcomes,

arguments can still be made for the inclusion of motivational aspects in PMT.

213

The present study suggests that motivational aspects influence outcomes
in PMT to the largest extent when traditionally examined predictors are
considered. Recent evidence indicates that parent perception of few barriers to
treatment acts as a protective factor for children identified as at risk for showing
little behavior change because of low SES, parent psychopathology, stress, and
severity of behavior (Kazdin & Wassell, 1999). Thus, the findings/ results do
provide some minimal support for the main hypothesis that traditional and
motivational aspects jointly predict outcomes in parent management training.
Furthermore, these results provide support for recent motivational research
arguments that both individual and environmental factors direct motivational
energy (Maehr & Pintrich, 1999). The results highlight the potential impact that
improvements in parent’s motivational aspects can have on children’s behavior
changes, parent’s satisfaction with treatments, and parent engagement. The
motivational aspect results provide tentative evidence for the inclusion of
motivational modules in PMT and support for further theory-testing and
refinement.

Interest in the current study (with the exception of the chi-square
analyses) focused on the joint performance of the set of variables at hand and
reflects the mutual influence of the variables on each other. Considerable
change in results was observed if some variables were added and others
deleted. This change in results likely reflects reality and we want the variables to

behave this way. To ask for the contribution of each variable independent of all

214

other variables is equivalent to requesting a univariate index that ignores the
other variables (Rencher, 1995).

Limitations of the Present Study

 

Several methodological limitations in this intervention outcome study are
acknowledged and at the same time corrective suggestions are presented. The
purpose of the present section is to discuss methodology issues in the parent
training intervention study. Some strategies and techniques for controlling these
issues are included. The following limitations discussed include instrumentation,
sampling issues, specificity, developmental considerations, treatment integrity,
and reactivity.

Instrumentation

 

The present study focused on only a few Ieaming outcomes and some
narrowly conceptualized measures. Consideration of a broader domain of
outcomes than might be typically considered in PMT programs might reveal
outcomes which can be far-reaching in nature and potentially effective. The
present study attempted to measure some of these but simply did not measure or
capture the entire gamut of outcomes. A recent case study of two children with
ADHD/ODD included behavioral observations in the home in addition to the
behavior rating scales of the current study; significant drops in child behavior
problems were noted on CBCL and HSQ, as well as behavioral observations
using the Behavior Management Flow chart in a PMT program (Danforth, 1999).
Another recent case study directly observed outcomes of parent training in a

multiple baseline design and evaluated change according to inappropriate child

215

behavior, compliance, on-task behavior, parent-child interaction, and a parent
praise measure (Greene, Kamps, Wyble, & Ellis, 1999).

Choice of specific instrumentation is related to this limitation in the present
study. While one of the most popular outcome measures utilized in the treatment
outcome research literature is Achenbach’s Child Behavior Checklist (CBCL) and
its related instruments (Drotar et al., 1995), the CBCL has many disadvantages
as well as advantages. The advantages include sound psychometric properties
regarding reliability and validity as an assessment of frequently presenting
symptoms in children from 4-18 years of age; two large demographically diverse
standardization samples (one referred and one non-referred); and ease of
administration and scoring with a relatively inexpensive computer program
(Drotar et al., 1995).

The CBCL, being viewed as a broad-band scale, assesses many facets of
behavior. One disadvantage is that it may be that only a couple of child
behavioral areas might need change. The CBCL does not single out one or two
target behaviors and could mask gains made in particular areas. As for other
disadvantages, the CBCL has been noted to have not only limited sensitivity to
milder, more subtle behavior problems but also to have compressed all raw
scores within the normal range (less than 69th percentile) and assigned them the
same T-score; provides a potentially misleading assessment of Social
Competence which measures selected aspects of social functioning (social
adjustment) and neglects others (social performance and social skills); is biased

in reporting physical symptoms by not distinguishing organic from more

216

emotional and stress related disorders; provides no interpretative information
regarding ethnically, culturally, and racially diverse populations; and, contains
much subjectivity in scoring (Drotar et al., 1995). It is important to note that these
issues are probably not unique to the CBCL and likely applicable to other
measurement instruments and materials used in much intervention research.

First and Way, (1995) discuss outcomes of PMT as facilitating the
development of new ways of thinking, transformative learning, as well as
enhancing specific parenting skills. The immediate goals of PMT are to develop
very specific skills in parents (Kazdin, 1998) and thus, to basically “change”
parenting behavior. No direct formal measure of parent behavior change was
obtained in the current study.

In fact, informal discussion with parents who were extremely satisfied with
the program, discussed gains which were not explicitly measured in the program
and which may be more related to reputation than behaviour change per se.
Several parents commented on the need to separate the behavior from the child
which intuitively implicates a movement from an implicit theory orientation to an
incremental theorist orientation. Several parents commented on how much they
found enjoying being with their children and that it was not a stressful time to
dread and find ways to avoid. The home environment had becomemuch less
stressful and amenable to positive interactions among family members. The
present results suggest that given the potentially wide range of outcomes beyond
mere changes in child behavior and satisfaction, and acquisition of parenting

skills, therapists should consider motivational aspects more explicitly when

217

planning programs to create more meaningful experiences. In considering the
issue of reputation versus behaviour change, it may be that reputation changes
more slowly.

Finally, the current study did not provide for any check of harmful
outcomes. The PMT sessions had a heavy emphasis on positive reinforcement
techniques in the first two weeks and the next two weeks included a punishment
contingency (e.g., time out, response cost, fines). With families who are prone to
negative, violent, and/or abusive interactions, it is possible that application of
punishment contingencies may be difficulty (botched) and even harmful (Barkley,
2000)

Thus, the present study may have measured a specific and narrow
measure of outcome. Methodological rules of thumb are described as follows: to
include multiple measures of treatment outcome in all relevant domains of any
life situations likely to be influenced by treatment effectiveness (Peterson & Bell-
Dolan, 1995); to assess areas of outcomes beyond the child domains (social
functioning, academics) and include parent domains (interactions with child,
stress, and marital discord) as well as broader contexts (family, living conditions,
and neighborhoods) (Kazdin, 1995); to consider the use of multiple informants
even though concordance among raters can be problematic (Tarullo et al., 1995).
Forness and Hoagwood (1993) recommend a multiple diagnostic approach with
several instruments. These recommended practices make instrumentation a
particularly salient issue in school psychology intervention research; logistically, it

would make sense that with more measures there is more opportunity for

218

instrumentation problems coupled with more opportunity to find significant
effects.

Sampling Issues

 

The pre-test results suggest that the treatment and control groups may
have differed prior to treatment on the CBCL scale ratings and several other
characteristics. This lack of group equivalence (Peterson & Bell-Dolan, 1995)
may be a reflection of selection bias or the small sample size. Selection refers to
systematic differences between groups at the beginning of the research study
which are the result of the method used to assign subjects to treatment groups or
control groups (AATBS, 1995).

In addition to systematic differences between groups another problem is
having a non-representative sample of the population because some potential
subjects failed to be included at all (Peterson 8. Bell-Dolan, 1995) and/or because
some potential subjects did not sign the consent forms (Carlson, 1993; Eyberg,
1995). Based on estimates of referrals from the ADHD clinic, well over half of the
parents who were referred for PMT did not initiate contact for an orientation
session. Barkley (2000) reports that estimates of families failing to accept or
attend parent training programs range from 30 % to 50 % or more. Thus, it
continues to remain unknown how these families would fair with PMT.

Based on records from priority/wait lists of referrals from physicians in the
health care clinic, all of the parents who were referred attended orientation
sessions; however, these orientation sessions doubled as initial interviews for

upcoming evaluations and may have biased attendance. Three parents who

219

attended orientation sessions refused to participate in the research component of
the program. Three other parents refused to participate in PMT at all. In any
case, the participants in the study were representative of the population of
interest, i.e., participants who typically participate in treatment (AATBS, 1995;
Peterson 8. Bell-Dolan, 1995). As a result, the participants in the intervention
study are basically a voluntary yet true clinical sample (Peterson & Bell-Dolan,
1995). For example, participants were essentially parents of children who show
some clinical symptoms, children who meet clinical diagnosis, and children who
are clinically referred, and non-referred clinical cases (Peterson & Bell-Dolan,
1995)

Another systematic selection difference is the under-representation of
children from diverse backgrounds and lack of inclusion of different ethnic groups
in intervention research (Foster & Martinez, 1995). Fomess and Hoagwood
(1993) included a comprehensive review which concluded that the literature is
lacking in studying ethnic differences in intervention as well as assessment.
Despite the fact that the US. population has been becoming increasingly
pluralistic, theories of child behavior continue to be based on European American
culture (Foster & Martinez, 1995). The population from which this sample was
drawn is not particularly diverse but representative of small northern Canadian
communities. These results do not generalize to larger, metropolitan, more
diverse communities. Thus, the study has limited external validity which
describes the generalizability of a research study’s results to other people,

settings, and conditions (AATBS, 1995; Porter, 1988).

220

Random Assignment

 

Random assignment was not total and thus, problematic in the clinic. A
few parents who could not attend the upcoming group due to scheduling and
other difficulties, were assigned to control groups. Sometimes, random
assignment was not possible because of other considerations. An apt example of
practical consideration is the situation in which staff from the Children’s Aid
Society referred a parent with disruptive children for parenting sessions under
court order and there were no other current parenting programs being offered in
the community. In this case where random assignment was violated, the parent
was excluded from the research data; however, it became apparent from informal
conversations with CAS staff that other parents had been referred to their
physicians and attended PMT under court order. However, some researchers
have argued that a truly random sample where every individual in the population
has an equal chance of being selected is never the case (Shulman, 1988).

Henggeler, Smith, and Schoenwald (1994) provide two general
persuasive arguments to overcome resistance to random assignment by decision
makers who are concerned about unfairly depriving children and parents of
services. First, empirical evidence indicates that treatments for serious antisocial
youths has minimal effectiveness and reviews have concluded that nothing works
(Henggeler et. al., 1994). Second, random assignment could be presented as the
most ethical and fair method of allocating scarce treatment resources

(Henggeler et. al., 1994). While these two arguments are presented as an

221

“ethical imperative", they can also be viewed as ethically questionable within the
context of effective treatments and adequate resources.

When participants have been involved in previous treatments (and
sometimes concurrent treatments), the results of the study cannot be generalized
to situations in which people will be exposed to the treatment under investigation.
It should be noted that many of these factors (especially reactivity and multiple
treatment interference) can be viewed as direct threats to both internal and
external validity.

Develpomental Considerations

 

The present study chose not to consider the child’s age as a factor given
limited research indicating age as a predictor is also strongly associated with
severity of the disorder. It may have been less limiting to shrink the age range of
the children in the study. Positive results of PMT may be manifested more quickly
or easily when the children are younger (ages 5 to 7) than when they are older
(10-12) (Barkley, 1997). If this is true, age of the child might have become
conflated with the other variables of interest. However, given the small sample
size, it was neither possible to restrict the age of children included in the study
nor ensure that different sample subgroups contained similar age distributions. If
true that severity of child behavior is strongly associated with age, then age as a
factor is a moot point. However, type of child behavior disorder was also not
examined in this cross-categorical approach study. Similar
arguments/justifications to this limitation are applicable. Some research on CD

indicates a hardening around age 8.

222

Subgroups and/or Comorbidity Considerations

 

The present study failed to consider whether there were certain subgroups
(especially subtypes of AD/HD) or comorbid conditions were related to parental
drop out or other outcomes. Responsiveness to PMT may vary according to
diagnosis, especially, subtype, and/or comorbidity (Barkley, 2000). A related
limitation of that the present study seemed to neglect the child’s condition as a
moderator variable (e.g., did parents of comorbid children differ from other
parents?)

Treatment Integrity

 

The present study contained no formal measure providing assurance of
treatment integrity which is the extent that treatment is actually implemented in
the manner it was intended (Black & Holden, 1995; Shapiro, 1987). Specifically,
the intervention did not contain objective observation of treatment delivery and
documentation of the extent to which the intervention was implemented
(Deschler & Schumaker, 1994). However, the study did contain set manualized
procedures which were explicitly followed. These included a) a brief video tape
presentation provided a summary and overview of procedures and strategies to
be discussed prior to each session; b) objectively specified interventions and
handouts; and, c) maintenance of a treatment log documenting subjective
assessment of participant status, participants’ participation, procedures

taught/discussed, and future plans for each session.

223

Reactivity

Reactivity is a phenomenon in which research participants respond to an
independent variable in a particular way simply because they know their behavior
is being observed (AATBS, 1995, p.27). Reactivity can include four types of
phenomena: 1) a Hawthorne Effect; 2) Evaluation Apprehension; 3) Demand
Characteristics; and , 4) Experimenter Expectancy (AATBS, 1995).

The Hawthome effect refers to the tendency of participants to perform better
because of the attention they are receiving as research participants (AATBS,
1995). Evaluation apprehension causes participants to act in ways so as to avoid
negative evaluations (AATBS, 1995). Demand characteristics include cues in the
environment which inform participants of the purpose of the study or suggest to
them what behaviors are expected of them (AATBS, 1995). Experimenter bias
occurs when an experimenter unintentionally (usually) provides participants with
cues (demand characteristics) that inform them of what behavior is expected of
them (AATBS, 1995). Experimenters can act in ways that do not affect the
participants directly but which can bias the results of the study (AATBS, 1995).
The present study included no measure to objectively evaluate the effects of any
reactivity.

Pressley and Harris (1994) view sample size as the least controversial of
methodological issues which should be planned in advance in the study design.
Ultimately, it depends on the research question and a cost-benefit analysis of
what kind of methodological strengths the experimenter prefers and what kinds of

methodological weakness the experimenter will tolerate. The level of power in the

224

current study for detecting differences was low. Thus, it is possible that the study
was not sensitive enough to detect differences which were being studied and/or
the treatment was not strong enough for the statistics which are presented.
Peterson and Bell-Dolan (1995) argue that it would make more sense to
view methodological issues on a continuum rather as typically dichotomous, i.e.,
either the study has methodological flaws or does not. Specifically, we can
conclude that the current study has elements of sound design features which are
attributed to the inclusion of control groups, using multiple outcome measures,
and collecting follow-up data (Dunak et al., 1995). The study used parents of
children from existing clinical groups rather than from the general population. The
study included theory based or developmentally rooted interventions. All of these
features are in diametric opposition to limitations as posed in reviews of other

intervention studies (Durlak et al., 1995).

225

Chapter 6
SUMMARY AND IMPLICATIONS

The results speak to several of the research questions related to
motivational resources. The motivational resources and value aspects formed
identifiable motivational schema and several of these factors influenced
outcomes but, for the most part, they did not. The motivational resources and
value aspects grouped together to form five central schemas among individual
parents (LV, IMC, PGE, and IT). Some of these schemas were predictive of
some parent training outcomes (child behavior change and parent satisfaction
with PMT). Furthermore, some of the individual motivational resources were
predictive of parent engagement. Both the motivational schema and the
individual motivational resources changed over time and were susceptible to
greater change among treatment parents as a result of participation in a
predominantly behaviorally oriented PMT. The latter suggests these resources
are more readily amenable to therapeutic influence.

Implications for Theory

 

Motivation as a construct has largely been applied to educational theory.
The current results may have implications for borrowing constructs from one
setting (classroom) and transferring to another setting (clinical) and suggest

these theoretical constructs are not readily transferable.

226

Implications for Practice

 

The implication for future practice is that although therapists (like
teachers) can do little aboUt an individual’s experience with uncontrollable
outcomes and predictors (socioeconomic status, family constellation,
psychosocial stressors), they can do much regarding how individuals come to
interpret and understand why those outcomes occur (Reeve, 1996).

There are numerous motivational aspects that have developed empirically
validated methods to enhance academic achievement. The present study found
that there are also several aspects of motivation (parent self-efficacy,
schematicity, optimistic explanatory style, learning goals, performance goals, and
value) which combined with traditional predictors can be successful in predicting
various outcomes in parent training programs. These results suggest that
proponents of parent management training programs consider diverting (or
augmenting already considered) effort toward encouraging parents’ inner
motivational resources in order to increase the probability of successful
outcomes. Furthermore, programs with primary motivational goals and strategies
which focus on the motivation to Ieam, the intention of acquiring the knowledge
or skills that activities are intended to develop, are likely to be most successful
(Brophy, 1998).

The results of the present study suggest that useful heuristics/guidelines
can be offered to therapists in providing a comprehensive PMT program targeted
specifically at addressing, compensating for, and/or optimizing these significant

motivational mediators in parent training outcomes. While some PMT programs

227

implicitly provide methods for enhancing motivational factors by their very nature,
motivational mediated parent training would serve to make these more explicit.

Schema Mediated Parent Training may be a useful adjunct to traditional
PMT approaches. While the content of materials will be very similar, participants
in each treatment group will be exposed to different methods of instruction based
upon a set of heuristic guidelines. These heuristic guidelines can be viewed as “a
set of rules of thumb that can be adapted to fit each situation, rather than a rigid
procedure that must be followed in the same way every time” (Newby et al.,
1996, p.120).

Parent Efficacy and Outcomes in Practice

 

For those parents who present with low self efficacy beliefs which
significantly mediates negative outcomes in PMT, explicit guidance and
instruction is warranted. Future practice may consider a comprehensive program
for self efficacy training, consisting of three parts (identifying component skills,
measuring self-efficacy expectations, and supplying positive self-efficacy
information) and using the four sources of information (performance
accomplishments, vicarious experience, verbal persuasion, and physiological
states) (Reeve, 1996). Borrowing the guidelines from the seven-step procedures
recommended for instructors and students, PMT may also incorporate them as
follows:

1. Therapist identifies component skills needed for competent performance.
Instructor measures parents’ self-efficacy expectations on each of those skills
(possibly through using components of the PITS).

2. Therapist models each component skill.

3. Following the modeling, parents perform each component skill through role
playing, and “home” work practice.

228

4. Individually mastered component skills are integrated into competent
functioning during a simulation of the task (using various vignettes). Therapist
graduates the simulations so they consist of a succession of increases in task
difficulties and performance obstacles (vignettes organized in a hierarchy of
problematic behavior).

5. While one parent performs the simulated task, other parents watch on the
sidelines. Parents offer the performer skill-related tips and motivation-related
encouragement.

6. Therapist provides additional instructive modeling, using corrective feedback
as needed.

7. Therapist identifies stressful indicators and warning signs (increased heart
rate, clenched jaw) and models confident demeanor and arousal regulating
techniques (relaxation, imagery) to parents.

(adapted from Reeve, 1996, p. 85)

Teaching efficacy beliefs is both possible and important in that it can build
strong and resilient cognitive beliefs that parents can use to counter the
unpredictable, stressful, and difficult challenges of raising a child with disruptive
behavior tendencies.

Efficacy beliefs vary in how strong and resistant they are to change based
on the amount of information which formed them (Reeve, 1996). For example,
firm efficacy beliefs are based on much, consistent information and require
compelling disconfirmation (Reeve, 1996). lnconclusive efficacy beliefs are
based on little, inconsistent information and easily subject to reanalysis (Reeve,
1996). It is these latter beliefs which seem more characteristic of parents whose
children have disruptive behavior disorders.

Efficacy can be enhanced by the following therapist efforts in PMT:

1) Encouraging parents to set specific, difficult, and attainable goals;

2) Model and cue effective response strategies;

3) Provide feedback that helps students to achieve success;

4) Make attributional statements that help them to appreciate that they are

developing their abilities by accepting challenges and applying consistent

effort.
(Brophy, 1998, p.57)

229

Parent Personal Control Beliefs and Practice

 

Parents with pessimistic explanatory styles can also be explicitly taught to
adapt an optimistic explanatory style in a therapy process consistent with
“attributional retraining”. Attribution retraining was used to teach students to
consider more adaptive attributional analysis to explain school’s aversive events
(Reeve, 1996). The goal of attributional retraining programs is to learn to make
intemal-stabIe-controllable attributions for successes and extemaI-unstable-
uncontrollable attributions for failures (Reeve, 1996). This training is first
accomplished by the therapist extending the range of possible attributions for
failures and suggesting alternative attributions such as ineffective strategy,
insufficient effort, lack of proper training or experience, high task difficulty
(Reeve, 1996). Second, the therapist persuades parents to change their
pessimistic attributions through three procedural techniques:

1) Persuasion: a parent performs a task and a therapist verbalizes the desired
attribution and then attempts to convince the parent of the validity of the
desired attribution;

2) Behavior modification: the therapist encourages parents to make attributions
of ability, effort, task difficulty, and luck for failure but verbally reinforces only
the unstable, uncontrollable attribution (e.g. luck); and,

3) Information: for example, therapists inform parents about the nature of their
children’s disorders.

(adapted from Reeve, 1996, p.111)

This third avenue usually occurs outside of PMT. For example, many therapists

may provide information regarding the nature of childhood disorders prior to

beginning PMT and usually during assessment feedback sessions.

Parents with a Ieamed helplessness orientation require support through

two avenues: 1) changing a nonresponsive environment to a responsive one,

230

and 2) changing attributional style. The latter is easier and accomplished through
attributional training and retraining programs discussed above. The former
requires a systems approach and restructuring of the environment which is not
always possible. Multisystemic treatment approaches may be more
advantageous here (Henggeler, Schoenwald, Borduin, Rowland, & Cunningham,
1998). Furthermore, attributional retraining which explicitly teaches parents to
attribute their successes to their ability as well as reasonable effort, and to
attribute their failures to lack of infomtation or response strategies as well as lack
of effort, (Brophy, 1998) is likely to promote more positive PMT outcomes. In
previous attribution retraining treatments, like with students in achievement
contexts, parents could be taught the following:

1. to concentrate on the task at hand and not worry about “failing”

2. to cope with “failures” by retracing their steps to find their mistake

or by analyzing the problem and finding another approach
3. to attribute their failures to insufficient effort, lack of information or
use of ineffective strategies
(Brophy, 1998, p.86)

Cunningham (1990) incorporates a number of strategies in his PMT
community programs which foster social Ieaming attributions for strategy
choices and child behavior. Among these strategies are the use of
errorless Ieaming, problem solving, and coping models (Johnston, 1996).
Attributional retraining has been incorporated as a component of many
other therapies for example, Beck’s psychotherapy, Baucom and Epstein’s

marital therapy, Alexander and Parson’s family therapy, and even

Cunningham’s community BPT programs (Johnston, 1996).

231

Learned Helplessness in Practice

 

Parents, like students with failure syndrome problems would benefit from
therapy addressing self confidence in their abilities, strategies for coping with
failures, and persistence in problem solving efforts under conditions of difficulty
(Brophy, 1998). Cognitive retraining approaches which include attribution
retraining, efficacy training, and strategy training involve more specific and
elaborate suggestions and use a combination of methods (direct instruction,
modeling with verbalized self instruction, thinking out loud, coaching, and
scaffolded practice) (Brophy, 1998). Students with Ieamed helplessness need to
learn how to cope with frustration and failure rather than a steady diet of
“success” (Brophy, 1998). Ditto for parents and maybe therapists, tool.

Goal Theory and PMT Practice

 

Early, deliberate, and more systematic changes in PMT programs are
required in order to affect any advantageous changes in parent goal. First and
foremost, therapists need to emphasize Ieaming goals since what appears to be
most salient is the individual’s perception of the goal orientation of the
achievement situation (Ames & Archer, 1988). For those parents with a social
orientation and excessive concern with "how their child’s behavior looks. . this
is no easy task. Therapists can promote a Ieaming orientation by emphasizing
what parents can Ieam from PMT activities, including collaborate work where
parents help one another, and treating errors or “behavioral mishaps” as natural
parts of the Ieaming process and rich material for follow up discussion (Brophy,

1998).

232

Dweck’s work suggests several implications for classroom teachers to
encourage their students: 1) to adopt incremental theories rather than entity
theories of abilities; 2) to set learning goals rather than performance goals; and,
3) to adopt a mastery orientation rather than a helpless orientation (Brophy,
1998). Parallel parent management training methods for establishing the first
point are illustrated through the following points:

setting up the therapeutic environment as a Ieaming community,
focusing parent’s attention on self improvement rather than
comparison with other parents,

making mistakes a part of Ieaming,

encouraging mastery through using successive approximations,
portraying activities as opportunities to acquire and not just to

display skills,
. providing appropriate feedback and comments which convey
confidence in reaching goals.
(Brophy, 1998).

Issues related to self worth protection bear special mention as this appears to be
an especially salient predictor of parent training outcomes. Self worth protection
in the classroom is minimized through establishing the classroom as a Ieaming
community and orienting students towards Ieaming goals through models,
socialization, and instruction (Brophy, 1998). Strategies by Covington (1992),
Lehtinen et al. (1995), and Brophy (1998) also suggest helping the students see

how the self worth protection focus limits their performance and development.

Self Schemas, Possible Selves, and Practice

 

Parents who are aschematic can become schematic through discrepant
feedback and elaborating on possible selves. A viable parent study would be to

examine the effect of parents teaching parents on schematicity.

233

Value Aspects and Practice

 

As an initial step in PMT, parents require a rationale for engaging in the
treatment and discussion of its benefits. While this is not new information, it tends
to be relegated to therapist discretion, as well as experience. Few treatment
manuals explicitly address the value aspects of motivation within PMT. The
results of the current study suggest that value enhancement should be explicitly
targeted at the orientation session prior to the start of PMT. Especially needed
are strategies for helping parents come to value what they are Ieaming for its
relevance, empowerment implications, exposure to whole aspects of life, and
potential quality of life application (and not just something which needs to be
done and over with). PMT may focus on building within the parents some
understanding of the human condition as well as tearing down some
misunderstandings and some sense that they are doing the activities engaged in
something worthwhile for themselves, instead of what they should be doing. This
is largely related to the parent need to “buy into” goals (as noted above) and not
simply buying into strategies. (Eyberg, Edwards, Boggs, & Foote, 1998) stresses
the importance of presenting a clear and concise rationale at the beginning of
PMT to avoid possible attrition, loss of cases before outcome and follow-up
evaluations.

Furthermore, even those aspects which are already inherently valued by
parents may not become as valued in PMT unless exposure is 1) scaffolded to
enhance learning as well as appreciation and 2) it connects with their zone of

proximal development (Brophy, 1999)

234

Summary

Given the significance of the findings, all of these suggestions could
culminate in an extensive comprehensive program offering Schema Mediated
Parenting Training (SMEPT) (Table 53). SMEPT needs to consider that to
change efficacy-based helplessness, therapists need to build in parents’
particular skills, competencies, and resiliency perceptions. These target change
in some aspect of the parent. To change personal control-based helplessness,
therapists need to consider some aspects of the environment (some of the more
traditionally examined predictors such as SES, family structure) and some aspect
of the parents’ explanatory styles (especially if pessimistic). It is also noted that
self-efficacy and personal control expectations coincide in an interdependent
relationship (Reeve, 1996). The implication for future practice is that although
therapists (like teachers) can do little about an individual’s experience with
uncontrollable outcomes and predictors (socioeconomic status, family
constellation, psychosocial stressors), they can do much regarding how
individuals come to interpret and understand why those outcomes occur (Reeve,

1996).

235

Table 53

Salient Motivational Aspects and Value Interventions as Potential SMEPT modules.

 

Motivational Aspect

Clinical Intervention

 

Self Efficacy

Attributions

Schematicity

Goals

Value

Verbal Persuasion
Vicarious Modeling
Mastery Exposure
Anticipatory Stress Inoculation

Attribution Re-training

Self confidence
Social Support
Graduated Task Assignments

Environmental structuring
Goal Setting
Goal clarification

Value clarification
Moral Discussion

 

Note. Value clarification sometimes involves moral judgements.

Implications For Future Research

 

Certainly, future studies should examine the efficacy of SMEPT. Despite

the extensive amount of research conducted in parent management training, the

need and potential exists for further studies examining the following issues:

1.

2.

The field is ripe for a meta analysis of predictors and outcomes in PMT;

Further attention to the accurate prediction of the probability of drop out would
assist in determining whether resources (time, etc.) should be invested.
Inaccurate prediction may lead to a waste of resources on very low risk
parents or, on the other hand, failure to adequately treat high risk parents;

Further examination of subgroups and/or comorbidity among disruptive
behavior disorders and the relationship with PMT may be warranted;

The idea of matching levels of treatment to the risk level of the parent for drop
out and to enhance positive outcomes aspect provides a bridge between
assessment, classification, and treatment;

Students with an emphasis on Ieaming goals process information using
deeper level strategies while students with performance goals are more likely
to use surface level memorizing (Brophy, 1998). This has implications for the
generalization of treatment strategies into other areas and is another area for
further research;

Previous research showing increased resistance to therapists may have been
related to attributional implications of therapist teaching in a didactic manner
(implication parents lack knowledge and skill); more Socratic, cooperative,
collaborative, and supportive may allow parents to continue to feel more self-
efficacious while making the behavioral changes;

One area of study which has not been sufficiently explored is who profits from
single modality treatments such as PMT or from combined treatments
(Kazdin, 1997). Also, unexplored is whether certain parents profit from certain
types of PMT treatments as opposed to others. It has been hypothesized that
certain parents and children may be a “good fit” for a particular type of parent
training program (Holden et al., 1990);

Clarizio et al. (1994) posed several questions related to instructional
approaches and values as follows:
Would you favor differ approaches for different students? For
example, which approach would be better for those children who
come from homes that have emphasized order, discipline, and

237

9.

work? How about children from homes that valued openness,
challenge, creativity, and spontaneity? Which of these approaches
would be best if we want students to get the big “Picture”? (Clarizio
etaL,1994,p.172x

Future studies could examine evaluating/measuring parent zone of
proximal development; few have considered parent actual abilities or
intelligence;

10.The interplay between parent cognitions, affect, and behavior is an

11.

area worthy of study (Johnston, 1996). Interventions may need to be
aimed at all of these components; focus on parenting as an integration
of all of these;

Despite links between underachievement, language delays, reading
disabilities, and CD, PMT research has rarely if ever included an academic
skills enhancement program for parents (Webster-Stratton, 1993). Research
is needed in programs which train parents to enhance and help children with
their academics;

12. Programs in which parents who have been through training act as facilitators

and teach parents just entering as parental educators are consistent with
notions of enhancing self efficacy; parents may develop more confidence
when models are more alike and result in better outcomes (Fox & Parroni
Hennick, 1996);

13. Possible selves and latency measures more accurate than simple more like

me, not like me responses, eg. Computer program to measure time to
respond; takes away social desirability component. The latency of the
response may indicate whether the item is readily accessible as part of their
self concept and whether or not the descriptor is not a feature of their working
schema. The speculation that schematic persons, respond faster to positive
items and aschematic persons respond faster to negative items; (Cross &
Markus, 1994) can be evaluated and used as a more precise measure of
schematicity to see if it is in fact predictive of outcomes of parent training;

14.Assess parent satisfaction at multiple points during treatment and follow-up

(future study)(Brestan et al., 1999). Answers to these questions may provide
information to treatment planning and may extend to planning maintenance of
treatment gains (Brestan et al., 1999);

15. Ruma et al. (1996) asked the question “which treatment works best for which

children” and concluded that the question remains largely unanswered. With
regard to parent training, it is hypothesized that a more productive question
would be to change the question to “which mode of treatment works best with
which parents?” Therapists likely adjust their treatment styles, i.e., strategies

238

for instruction, based on an intuitive evaluation of parents. Through
interactions with parents, therapists may come to realize what parents want
and to which mode of instruction they would be most responsive. It would be
highly advantageous to be able to identify those parameters of successful
outcomes and to adjust parent management training to capitalize on those
parameters;

16.The history of Ethnotheory refers to folk perspectives on the nature of children
and of parental roles in child development (Goodnow, 1995). These are
described as less formal and less fully articulated, but characteristic of
members in particular societies or cultures. For example, maternal beliefs
about child development and parenting practices among Quebecois,
Vietnamese, and Haitian mothers living in Canada varied depending upon
their culture of origin (Pomerleau et al., 1991). Quebecois mothers saw
children as unique competent individuals and viewed the mother’s role as
important in stimulating the infant. These mother with origins to France
practiced encouragement of exploratory behavior and presentation of
stimulating toys and activities to their children. Vietnamese mothers, in
contrast emphasized formal Ieaming activities. An important assumption of
Ethnotheories is that beliefs cannot be separated from the cultural milieu
(Goodnow, 1995). Ethnotheories view all members of particular cultures as
possessing an underlying belief structure regarding the children’s growth and
development (Goodnow, 1995). Consequently, it could be hypothesized that
an Ethnotheory perspective would predict that parents from similar cultures
will use similar childrearing methods;

17. Notions of collective efficacy and its influence as well as culture influence on
goal orientation may be another area to study. This is somewhat related to
the notion that overall classroom setting, and possibly school and even
community culture may play a role in students’ goal orientations (Lemos,
1996);

18. Predictors of PMT outcomes regarding disruptive behavior need not be so
domain specific; note, same predictors of negative outcomes for PMT
designed to reduce risks and enhance protection against early substance use
in children (Kosterman, Hawkins, Spoth, Haggerty, & Zhu, 1997);

19. Future research may see if multisystemic approaches decrease learned
helplessness in parents;

20.The development of a teacher motivational schema scale (TMSS) may be of
interest in predicting outcomes with children with behavior problems in the
classroom and ultimately provide an aid in consultation based classroom
management;

239

21. Frequent parent comments among the groups involved the lack of time to
implement some of the strategies, to spend with their children, increasing
demands of work, and their children’s structured extra-curricular activities.
Qualitative studies have dominated the PMT field. The field is also ripe for the
inclusion of some ethnographic studies to further examine some of these
Issues.

Final Thoughts

 

Like some of the more complete packages integrating different theoretical
traditions in mathematics (Bell & Kanevsky, 1996) and reading comprehension
(Shawaker & Dembo, 1996), PMT can combine elements from motivational
treatments into one comprehensive package. In fact, practice in the future
creates entire possibilities for incorporating motivational modules into parent
management training programs for parents of children with disruptive behavior
disorders. Further experimental and new qualitative studies would be the next
step to determining if enhancing parent motivation would foster more optimal

outcomes.

240

APPENDIX A

Parent Motivational Schema Scale (PMSS)
Contents of this Appendix include the following:

1. Inter-rater Agreements on Original PMSS Items (Table A. 1)
2. Scale Items Listed by Domain (Table A. 2)

241

Table A.1

Agreement Among Independent Raters.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ExpertI Student! Decision
STATEMENTS Researcher Literature
Judgment Surveys

1. I am confident that I can learn the material with Efficacy Efficacy Retain
the other parents in education sessions.

2. Involvement in parent training will help me learn Value Value Retain
how to manage my child’s behavior and help me
feel better about myself.

3. Whether my child follows directions, obeys 1. Implicit Retain/
instructions, and acts appropriately or not is Pessimistic theory discuss
deeply ingrained in his/her personality. It cannot 2. Implicit
be changed much.

4. It is important for me to learn what is being Value Value Retain
taught in these groups.

5. Making mistakes is a part of learning. Schemati- Learning Drop

city goal

6. When I really try, I can get through to my child Efficacy Efficacy Retain
when he or she is being most difficult.

7. I would like parent training activities that let me Perform Perform Retain
show how good I am at managing behavior. goal- goal-

Enhancing enhancing

8. I think that what I would Ieam in a parent Value Value Retain
trainm class would be interesting.

9. If my child became disruptive and noisy in public Efficacy Efficacy Retain
places (malls, church), I feel assured that I know
some techniques to redirect him or her quickly.

10.I would like to show my group leader, therapist, Perform Perform Retain
that I am better at managing behavior than the goal- goal-
other parents. enhance enhancing

11. lwork hard to learn. Learning Learning Retain

goal goal

12. By trying a different discipline method, I can Optimistic Efficacy Drop
significantly affect my child’s behavior. style

13. When my child is in one of his/her Optimistic Optimistic Retain
moods, there is usually something that style style
set him off but it can be handled.

14. My child’s behavior is something very basic Pessimistic ImpliCit Discuss
about him/her and it can’t be changed much. style theory

15. I think that what I would Ieam in a parent group Value Value Retain
would be useful for me to know.

16. At group sessions, I always try to avoid being Perform Perform Retain
among the poorest students/parents. goal — goal-

defeatirg defeating

17. The therapist will try to make sure I Optimistic Learning Drop
understand the work. style goal

18. There is really very little I can do to insure that Efficacy Efficacy Retain

my child behaves appropriately.

 

 

 

 

242

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

19. Involvement in PMT will be helpful in improving Value Value Retain
the quality of our lives.
20. I would like to show my other relatives who are Perform Perform Retain
parents (my parents, siblings, cousins), that I am goal- goal-
better at managing behavior than other parents. enhancing enhancing
21. I will like what I am learning in these gLoups. Value Value Retain
22. Other parents want to know how I am doing ? Perform Drop
on assignments. goal-
enhancing
23. The therapist will want us to try new things ? Learning Drop
goal
24. At home, it is important for me to avoid looking Perform Perform Retain
stupid. goal — goal-
defeating defeatig
25. Involvement in PMT will be useful in helping me Value Value Retain
to reach my goals.
26. Iwill feel most successful in parent training Learning Learning Retain
when I learn something I didn’t know before. goal goal
27. If I could choose to have children, if I had the 7 Efficacy Drop
decision to make over again, I would choose
parenthood.
28. I really don’t like to make mistakes. Perform Perform Discuss
goal — goal-
defeating enhancing
29. My child experiences difficulties Optimistic Optimistic Retain
because of his school or home style style
environments which can be altered to
suit his/her needs.
30. I like parent training work I will learn from even if Learning Learning Retain
I make a lot of mistakes. goal goal
31. I am confident that I can make the decisions that Efficacy Efficacy Retain
are made in our home.
32. The worst thing about making mistakes at home Perform Perform Retain
and discussing them in groups is what other goal- goal-
people may notice. defeatig defeating
33. There is not much that can be done to change Pessimistic Implicit Discuss
my child’s behavioral traits (e. g. oppositional, style theory
aggressive, defiant)
34. I am confident that I can have my son or Efficacy Efficacy Retain
daughter listen to me and comply when I am
discussing household chores.
35. The main reason I would do my parent training Learning Learning Retain
homework is because I like to Ieam. goal goal
36. My child gets oppositional, hostile, or Optimistic Optimistic Retain
defiant because of situational style style
circumstances, yet he can be appeased
with appropriate discipline.
37. I am good at controlling my child’s behavior and Efficacy Schematicit Discuss
this is important to me. y
38. I will like parent training work best when it Learning Learning Retain
makes me think. goal goal
39. When interacting with my child in public places, I Perform Perform Retain
am concerned about making a fool of myself. goal- goal-
defeating defeating
40. lwould feel successful in parent training if I did Perform Perform Retain

 

243

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

better than the other parents in the sessions. goal- goal-
enhancing enhancing
41. Involvement in PMT will help me learn how to Value Value Retain
manage my child’s behavior and others won’t
look down on me so much, anymore.
42. When lparticipate in group activities, I am Perform Perform Discuss
concerned about what my peers think about me. goal- goal-
enhancing defeating
43. I work hard to get the homework assignments Learning Perform Discuss
done goal goal-
enhancing
44. Parents feel bad when they do not do as well as Perform Perform Discuss
others. goal- goal-
defeating enhancing
45. Involvement in PMT will help me learn how to Value Value Retain
manage my child’s behavior and have better
control and power.
46. Understanding the parent training work is more Learning Learning Retain
important to me than getting evaluated (others goal goal
looking at my behavior and approving).
47. I am good at managing my child’s behavior and Efficacy Schematicit Discuss
this is important to me. y
48. When my child is in one of his/her moods, I can Efficacy Optimistic Discuss
figure out what set him/her off, and handle the style
sfiuafion.
49. My child’s behavior is mostly biological, and a Implicit Pessimistic Discuss
big part of his/her personality, and sometimes theory style
we ’re lucky to get through one of his/her
tempers.
50. Whenever I talk about how unsuccessful Perform Perform Retain
homework assignments have been at home, I goal- goal—
am most concerned about what other parents defeating defeating
think about me.
51. Only a few parents will get a lot out of .7 Perform Drop/
this parent training. goal- Discuss
enhancing
52. There is little to be gained from an open, Pessimistic Pessimistic Retain
accepting, and supportive relationship style style
with my child, especially since my child
is unmanageable.
53. Involvement in PMT will be interesting and Value Value Retain
possibly fun, as I will meet other parents in
similar situations.
54. Whenever I talk in group sessions, lam Perform Perform Discuss
occupied (I wonder) by howl am perceived by goal- goal-
other parents. enhancing defeating
55. I would like what I could learn in parent training Value Value Value

groups.

 

 

 

 

 

Items in bold were dropped
Items in italics needed to be further discussed to arrive at a consensus to improve content
sampling for each motivational resource
Note: difficulty with performance goal judgments, schematicity, and implicit theories; may be
worthy of discussion in general within scale development section.

244

 

 

.0.:0000:0E::

0. 0..:0 >E 0050 2.0.0800 0..:0 >E 55> :.:0:0:0.9 0200::00 0:0 .0:::0000 .:0:0 :0 E9: 00:.00 0: 0: 0.5. 0. 90: ...

.E0E90:E0:

::0:00 0:0 2:8: 0 00: 050 00:.0 >__0.00:00 .00>0:0: 0..:0 >E 30: 0:0 850:0: >E c0920: 5:80:59 .0::0:.:5 0.::._ 0. 90E.

.0EEE0: :0:E0:

.80.: :0 0:0 :000.5 :00 0: >x00. 9.03 .00E.:0E00 0:0 .>:._0:00._0: 90:05. :0 t0: 0.: 0 ..00.00_o.: >.:00E 0. :0.>0:0: 0.0..:0 >5.
05:332. 200 0205.90. 0:05.08:

 

.0000: 98.0.: :50 0: 090:.0 0: :00 :0.:.s 0E0E:9_>:0 090: :0 60:00 0.: :0 000000: 00:30:50 000:0:0:x0 0..:0 >5.
.05_:_00.0 0:0::9::0 5.3 00000::0 0: :00 0: .000:0:0E:9.0 05:03.0 :0 000000: ::0:00 :0 .0.:00: ..0:0:.00::0 0:00 0..:0 >5.

00.80: 0: :00 :0: ::o E.: :00 :05 0:.50E00 2.000: 0. 90:: .00009 8&5: :0 0:0 5 0. 0..:0 >E :0:>>
0:25:52: 2.00 082280 2:25.30

..>000_.::0 02:00:00: 0:0.0000 000:: 5 0::90: .050 0:: :Es .0:0:0E 0:: E00. :00 _ :0:: E00.::00 E0 .
.co.>0:0: E0..:E00

050008 :0: >000_::0.:.00: 090:0 0.0:0000: 05000050 E0 . :0:>> 2:98 0:0 0E 0: :0:0._ .0::0000 :0 :00 >E 0>0: :00 . :05 E00::00 E0 .
..>000::0 E903 0E0: 50 5 000E 90 :0:: 80.0.000 0:: 00.0E :00 _ :0:: ::00::00 E0 .

.000:E90: 080:0 0.003 _ .5000 :0>0 050E 0: :0.0.000 0:: 00: _ :. .:90..:0 0>0: 0: 080:0 0.000 . :_

.>_0:0::9::0 00>0:0: 0..:0 >E :0:: 900:. 00 :00 _ 0.5. >._0> >__09 0. 90:...

.:0.>0:0: 0.0_.:0 >E :00::0 >_E00:.:0.0 :00 _ .0050E 0:._:.00.0 E9050 0 05>: >m

.>_0.0.:0 :0:

:0 E.: 89.09 0: 00:55.00: 0E00 30:: _ :0:: 090000 .00: . .3980 .0__0E: 0000.: 0..:0: 5 >06: 0:0 0>::20.0 0E000: 0..:0 >E :.
.0E: 08: 0::

boE0E :0: :0 0.: 00090:. 0: 30: 30:0. 0.003 _ .0:: 00:0: 50:80 0 50:0 .8000. 0 00: 03 0E: :00. 0:: :0:EOE9 :0: 0.0 .0::0000 :0 :00 >E :.
.:.:0::_0 :00E 0:.0: 0. 0:0 ..0 0: :0:>> .0::0000 :0 :00 >E 0: :000.5 :00 :00 _ >9: >__09 _ :0:>>

.6850 :00

 

moLDOmmm _mCO=m>=O§ >D DOE: mEm: mmS—n—

N.< 0.00...

245

 

.00.:0: >0E 0.:00: :0::0 :0:>> 0. 0:090 5 E0:: 05000050 0:0 0E0: :0 00x0:0.E 05:.0E :00:0 05:: :90; 0: .r
.0E :00:0 E.::
0::90: :0::0 :0:>> :00:0 00E00:00 :00E E0 . .0E0: :0 :00: 0>0: 0::0E:0.000 0:030E0: .0:0000000:0 30: :00:0 5.0: _ :0>0:0:>>
.:.00>E :0 .00: 0 050E 0: :0: 00E00:00 E0 . .0000_: 0..:0: 5 0..:0 >E :53 05:00:95 :0:>>
.0E :00:0 E.:: 0:00: >E :0:>> :00:0 00E00:00 E0 . .00.:.>.:00 :00:0 0:: 5 9050.00: . :0:>>
.0::90: .050 >: 002090: E0 _ 30: >: C0803 _: 00.:0000 E0 . .8588 :00:0 5 5.0: _ :0>0:0:>>
850000.500:
.9050 00 :03 00 00:0: 00 >0:: :0:>> 00: .00: 0::90:
.0000 0::0E:0.000 :.:030E0: 0:: :00 0: 0:0: 0:03 .
.0559: E90: 0.:: :0 :00 :0. 0 :00 .52, 0::90: 30: 0 >.:O
.00:0:0.E 050E 0: 0:... :.:00 >__09 _
.0::0E:0.000 :0 0500 E0 . 30: 30:0. 0: E02, 0::90: :0::0
.0::90: .050

:0:: 850:0: 0500:0E :0 :0::0: E0 _ :0:: ..050000 00530 .0::90: >E: 0::90: 90 0:3 82:09 .050 >E 30:0 0: 00... 0.003 .
.0::90: .050 0:: :0:: :0.>0:0: 0500:0E :0 :98: E0 _ :0:: .:0.:0._0:: .:0000. :00:0 >E 30:0 0: 00... 0.003 _

.:0.>0:0: 0500:0E :0 E0 _ 0000 30: 30:0 0E :0. :0:: 00.:.>.:00 0550:: E90: 00... 0.003 _

.0:0.0000 0:: 5 0::90: .050 0:: :0:: :98: 0.0 _ :. 0550:: E90: 5 .0:0000000 .00: 0.003 .

:0:.0:m::0.:.00:

0:000 00:02:89:

..05>9::0 0:0 :0.>0:0: >E :0 050.00. 0.050: 090508 05:00 :0:: 0E 0: ::0::0:E. 90E 0. 0:03 0550:: E90: 0:: 0580:9005:

 

000.0:0.E :0 :0. 0 050E . :. :0>0 E9: E00. .2 0:03 0550:: E90: 00... .
.:.5:: 0E 00x0E :. :0:>> :00: 0:03 0550:: E90: 00... _
.E00. 0: 00... _ 000000: 0. 0:030E0: 0550:: E90: >E 0500 E0 _ :0009 50E 0: ._.
90:0: 30:0. :.:0.0 _ 05::0E00 E00. _ :0:>> 0550:: E90: 5 .0:0000000 :00E .00: :5...
005:: 30: >:: 0: 00 0E0; :0.:0:0:: 0: :.
.0003 0:: 0:0:900c0 _ 900 000.0E :0.:0:0:: 0: 5
.E00. 0: 0:0: 0:03 .
.05E00. :0 t0: 0 0. 00x0:0.E 050.05.
3:8...
00:.>...:m E0E0>0.:0<

00:00 N.< 0.:0...

246

 

0:090 000:: 5 0:.E00. E0 _ :0:3 00... _

0:090 000:: 5 E000: 050: 0. :0:3 E00. 0: 0E 5: E0:0:E. 0. :_

055995 0. 000.0 0.:: 5 0:.E00. 90 03 :0:3 :0:: E.:: .

.305. 0: 0E 5: 5:000 0: ___3 :00:0 0.:: 5 0:.E00. ___3 _ :0:3 :0:: E.:: _

.002. :00 :0 5.000 0:: 0:305E. 5 .250: 0: ___3 :.2: 5 E0E0>_0>:.

.0.000 >E :009 0: 0E 05:5: 5 .2000 0: __.3 #5.: 5 E0E0>.0>5

.0:0.:00:.0 5:50 5 0::90: :0::0 :00E __.3 _ 00 .:0: >508: 0:0 05:00:95 0: _._3 ...5.n. 5 E0E0>_0>:.

.90E>:0 .:00E 00 0E :0 :300 0.00. :.:03 05:5 0:0 530:0: 0.0..:0 >E 000:0E 0: 30: E00. 0E :.0: __.3 :2: 5 ::0E0>.0>:_

.530: 0:0 .9E00 :0::0: 0>0: 0:0 530:0: 0.0..:0 >E 000:0E 0: 30: E00. 0E :.0: .__3 .52: 5 E0E0>_0>:.

.:.00>E :00:0 :95: .00: 0E :.0: 0:0 530:0: 0.0..:0 >E 000:0E 0: 30: E00. 05 :.0: ._.3 t2: 5 E0E0>.0>:.
00.0>

.0E 0: E0:5:E. 0. 0.:: 0:0 530:0: 0.0..:0 >E 0500:0E :0 0000 E0 _
.0E 0: E0:5:E. 0. 0.:: 0:0 530:0: 0.0..:0 >E 0:...9E00 :0 0000 E0 .
85200 5005.0005300

3:0:00 .0>.000.000 ..0:0.:.00::0 0.0. 050:: .05.>0:0: 0.0..:0 >E 00:0:0 0: 0:00 0: :00 :0:: :00E :0: 0. 90: .r
.:00E 000:0:0 0: :0::00
:. .>:__0:00:0: 550.: 5 005905 2:80 0. :0: 5 >_0:0.::9::0 0:00 0:0 .0:0.:00::05 0>0:0 .80.:0050 030:0: 0_.:0 >E :0::0:>>
.:00E 000:0:0 0: :.:00 :. 0:0 :0::5: :00:0 0.00: >5> 05::0E00 0. 530:0: 0.0..:0 >5.
8.52: 50.35.
.0::90::0E000:0 :0900: 0:: 0:0E0 050: 0.0>0 0: >:: . .85080 :00:0 :<
.0520 050.00. 0.0>0 0: 0E 5: E0085 0. :. .0E0: :<

0.250 N.< 050:.

247

APPENDIX B

Consent Forms

CONSENT FORM FOR PARENT SURVEY

The Group Health Centre and the College of Education at Michigan State
University support the practice of protection for human subjects participating in
research. A brief description of the present study is as follows:

The goal of this study is to develop a new survey which can be
used when working with parents in the Department of
Communication Disorders as well as parents in other centres. It is
necessary to establish the reliability (each question is stable) and
valid (each question measures what it is supposed to measure. By
completing the enclosed questionnaire, you will help me to develop
a reliable and valid survey.

Your participation is voluntary. Even if you agree to participate you may withdraw
at any time. You have been provided with the opportunity to ask questions about
the study. Should you have more questions about the research, you have been
encouraged to contact Dr. P. Avery, the chair of the Research and Ethics Board,
Group Health Centre (254-4139) and/or Dr. D. Wright, the chair of the University
Committee on Research Involving Human Subjects (1-517-355-2180).

All information will be reported in group format and your name will NOT be
associated with research findings. Any identifying information will not be
included, that is, your name will not be written on any forms.

Principle Investigator:
Anne Marie Caruso, M.A., M.B.A.
School Psychology Doctoral Candidate
Michigan State University

Project Supervisors:
Harvey Clarizio, Ph.D
Department Chair, CEPSE
Michigan State University

You indicate your voluntary agreement to participate by
completing and returning this questionnaire.

248

 

CONSENT FORM FOR PARTICIPATION IN RESEARCH PROJECT

The Group Health Centre and the College of Education at Michigan State
University support the practice of protection for human subjects participating in
research. A brief description of the present study is as follows:

In this study, I am interested in finding out what it is that influences
what works in parent behavior management sessions. I plan on
looking at 1) predictors that have already been studied by other
researchers and 2) parent’s thinking about involvement. The
eventual aim is to use these predictors to develop the best
approach for working with parents and helping them help their
children.

Your participation is voluntary. Even if you agree to participate you may withdraw
at any time. You have been provided with the opportunity to ask questions about
the study. Should you have more questions about the research, you have been
encouraged to contact Dr. P. Avery, the chair of the Research and Ethics Board,
Group Health Centre (254-4139) and/or Dr. D. Wright, the chair of the University
Committee on Research Involving Human Subjects (1-517-355-2180).

All results will be kept strictly confidential and will be stored in your medical
record file upon completion of the study. All information will be reported in group
format and your name will NOT be associated with research findings. Any
identifying information will not be included.

Signature of Participant:
Print name here:

 

 

Signature of Investigator Obtaining Consent:
Print name here:

 

 

Signature of Witness:
Print name here:

 

 

Principle Investigator:
Anne Marie Caruso, M.A., M.B.A.
School Psychology Doctoral Candidate
Michigan State University

Project Supervisors:
Harvey Clarizio, Ph.D
Department Chair, CEPSE
Michigan State University

249

 

Letter Requesting Permission and Algorithm of REB Process

Saturday, April 10, 1999

John Harwood

Chief Executive Officer
Group Health Centre
Sault Ste Marie, Ontario
PBB 1Y5

Dear Mr. Harwood,

Thank you for taking the time to speak with me about a potential summer
research project. I am writing to request permission to conduct a study in the
Group Health Centre.

In this study, I am interested in evaluating predictors of positive outcomes
in parent behavior management training. I plan on looking at two broad
categories of predictors (traditionally examined predictors and parent
motivational resources) within the context of three categories of outcomes
(changes in child behavior, parent satisfaction, and parent engagement). The
eventual aim is to develop a heuristic guideline for the selection of an
instructional intervention based on predictor characteristics of clients, i.e., finding
the “best match” or finding the treatment approach that will maximize the benefit
for parents and children.

I would like to conduct the study in four phases as detailed in an
attachment. These include a 1) a scale development phase, 2) participant
“recruitment” phase, 3) intervention phase, and 4) outcome evaluation phase.

Thus, I am requesting permission to conduct the study in the Group Health
Centre; specifically, for you to supply the setting, materials, and participants with
the understanding that parents may refuse to cooperate in the study. For your
information, I will be mailing a second draft of the proposal which may still
undergo a few minor revisions before meeting all members of my doctoral
committee’s approval.

I appreciate the opportunity and know that this project would be mutually
beneficial. Outcome studies have shown parent behavior management training to
be among the most efficacious treatments when working with children with
behavior problems.

250

 

If it would be helpful, I would be happy to make a brief presentation about
the project to any Group Health Centre staff, you deem necessary. If you would
like any other information, please feel free to contact me by phone ( 517-
8871215) or e—mail (carusan@pilot.msu.edu).

Again, I appreciate your time and consideration of my request and look
forward to seeing you during the Research and Ethics Meeting in May.
Sincerely,

Anne Marie Caruso
Doctoral Candidate

CEPSE
Michigan State University

251

 

 

Phase One (Scale Development and Factor Analysis)

A scale is being developed specifically for the purposes of this study, since no
one has examined motivational resources within the context of parent therapy.
To ensure that each scale item content is consistent with the definition of the
predictors, the content validity of the Parent Motivation Schema Scale (PMSS) is
currently being established by inter-raterjudgements (10 PhD students and the
professor, in a doctoral level motivation course). The scale currently contains 58
items containing 9 aspects of motivational resources which may still further be
reduced.

In order to establish a) whether or not the scale items actually do measure the 9
motivational resources and b) find whether or not they cluster together into
common factors (maybe the 9 motivational resources actually group into two big
Ms, much like intelligence factors group into big and little Gs), the scale needs to
be administered to a large sample of parents and then statistically analyzed.
Based on the factor analysis, we may find other items to eliminate which may not
be necessary for the purposes of the study.

The PMSS will be administered to parents who bring their children to the Group
Health Centre. A summer student will randomly ask parents if they would be
willing to complete a survey and provide direction on the completion of the Parent
Motivation Schema Scale. Estimated completion time is approximately 15-20
minutes. The administration will occur while parents wait in the physician waiting
rooms for their appointments. All survey responses will be unsigned and
unidentified. Similarly, the clinicians in the Communication Disorders Department
will ask parents if they would be willing to complete the survey and provide
direction during a regularly scheduled appointment with their child. Any parent
can decline participation without concern.

Later the reformulated PMSS will be used to identify the predictors of outcomes
in parent training (below).

Phase Two (Identifying Participants / Gathering predictor information)

This phase would involve identifying parents suitable for the study. The majority
of participants will be parents of children already referred to the Communication
Disorders Department for a behavior evaluation. These parents are parents of
children on the waiting list for consults which has always been maintained by the
appointment center.

Parents who have been referred for services to the Communication Disorders
Department will be interviewed. The parent interviews should take approximately
1 hour and all information would be kept confidential. Note that these interviews
are a standard part of the assessment evaluation process whether or not parents
participate in the research project. After the interview, the research project will be

252

 

 

explained to parents and they will be asked if they would like to participate in the
project. They will be assured that they will receive the same services in the
department whether or not they agree to participate. If they agree to participate,
they will be asked to sign a consent form for participation (see enclosed). At
which point, they will be administered the Parent Motivation Schema Scale (final
version).

The Information from the interviews will be used to establish the parent levels on
each of the traditionally examined predictors (family constellation, socio-
economic status, number of psychosocial stressors, severity of child behavior,
and level parent depression, if any). The information from the PMSS will be used
to establish level of parent motivational resources (self efficacy, attributions,
goals, implicit theories, schematicity, and value).

All parents (including those who agreed to participate and those who did not) will
be randomly assigned to an intervention or control group. Information from the
parents who did not sign the consent will not be used in the study but they will
still participate in the intervention phase. Control group parents will be offered
parent management training services at the end of the intervention phase
(approximately two months from the point of assessment). This is not in excess
of past usual wait times for therapy services in the department.

.11.

Phase Three (Intervention/Treatment)

For several summers, the staff from the Communication Disorders Department
have conducted several summer projects, including parent training which is
designed to help parents to work with their children with behavior problems.

The parent management training would begin immediately upon completion of
the interviews and random assignment of parents to intervention or control
groups. This phase would require a larger setting, i.e., larger rooms. Possibly the
Women’s Health Centre’s Conference room can easily be scheduled during non-
use times on a regular weekly basis (as in past summers).

Phase Four (Gathering Outcome Information)

During this final phase, outcome data will be gathered at the final parent therapy
session and at the one month booster session and follow up. During the final
therapy session, outcome data regarding changes in child behavior will be
gathered through having parents complete a post-therapy, Child Behavior
Checklist (pre-therapy measure also used as part of the assessment process).
Outcome data regarding parent satisfaction will be gathered through having
parents complete an evaluation form about the parent training. Finally, outcome
data regarding parent engagement will be gathered through collecting parent
workbooks and evaluating homework completion and attendance (in hours).

 

253

At one month follow-up (the booster session), the CBC and the Parent evaluation
forms will again be re-administered. Parent workbooks will be returned.

Future Phases (Developing our own Parent Training Program)

The findings from this study should have several implications for theory as well
as implications for future practice and therapeutic application. For example, if it is
found that parent self efficacy has a significant effect on outcomes in parent
therapy, then maximum benefit would be derived by designing parent training to
include strategies and techniques on increasing parent efficacy.

A future study (maybe next summer) may consider examining the addition of

motivational modules to treatment packages as something which would further
increase the efficacy of parent training treatments.

254

 

 

APPENDIX C

Decisions on Factor Dimensions

Tables 0.1, C2, and 0.3 present the results of an orthogonal Nine-Factor
Solution using the PMSS. The Nine-Factor solution did not easily support the
factor structure of the nine subscales of the PMSS as items did not readily group
into the theoretical motivational aspects (Table C.3 ).

Similarly, the nonorthogonal five factor solution was not as easy to

interpret and did not contain adequate reliabilities (Table 0.4). The first factor in

the non-orthogonal five solution contained a mix of 17 items (10 value, 5 learning

goal, 1 self defeating performance goal, and 1 self enhancing performance goal

items). The second factor contained 12 items: 3 optimistic explanatory style, 5

self efficacy, 1 schematicity item, 1 implicit theory, 1 pessimistic explanatory

style, and 1 self defeating performance goal. The third factor contained the

following 7 items: 3 self defeating performance goal, 2 self enhancing

performance goal, 1 pessimistic explanatory style, and 1 schematicity. The fourth l

factor contained 2 implicit theory and 1 value item. The fifth factor comprised of

 

one item (self defeating performance goal).
Due to ease of interpretation and better reliabilities for the factor scales, it
was decided to use the rotated five factor solution for the subsequent data

analyses.

255

I Milk

 

 

 

 

0:0- 9 00.0 5.0 00.0 E.: 00.0 00.0 00.0 0052 80.20990

02:. 0: 00.0 00.0 00.0 5.0 5.0 00.0 00.0 00.5050: 052 Em:

0.1 0: 00.: 00.0 :0 00.0 00.0 000 :3 8.0.5: 000:0:0

5.0 00.0 00.0 00.: 5... 00.0 E. 3 9.00 00.00 :82 280

0 : 0 0 0 0 0 0 0: 0&9 :o .8532

052 Emm. :0>0m 5m 0>.n. :00”. 09:... 03... 0:0 825.5000
05:00“.

 

 

8005 00.2: 9: :o 0&2. 3.: =0 8 8:200 8:09.92 0:00.250 x055; 5: 09.55000

.50 0.:0...

256

Table 0.2

Eigenvalues, Percentages of Variance. and Cumulative Percentages
for Varimax Orthogonal 9 Factor Solution

 

 

 

% of
Factor Eigenvalue Variance Cumulative %
One 8.872 20.164 20.164
Two 4.945 11.238 31.401
Three 4.434 10.078 41 .480
Four 1.849 4.201 45.681
Five 1.556 3.537 49.218
Six 1.349 3.067 52.285
Seven 1.243 2.826 55.111
Eight 1.18 2.683 57.793
Nine 1.061 2.411 60.205

 

257

 

 
 

 

we 0.:0: 5 00:0: 0.0 m :000.5 F 0:900... 0:02

 

 

 

 

 

50.0- «mm: .025: 50 c: 000E 90 :0:: 0:90.500 0:: 050E :00 . :0:: E00350 E0 : .00
05.0 .000:
.8539: :00E 0:: :53 E90: 0:: 9.... 05:00: 0.90 0: b: 053.0 0.003 : .8988 :00:0 :< .v:

"0 5:00“.
000.0 .000: .089 8:00.: :50
0: 000:0:0 00 :00 0000:: 000:: 500.500 0E0: 5 50:00 0:: :0 0000000 0E059: 00: 0..:0 >5: .mm

:0 5:00“.
000.0 «mm: .:000.0E 0:: E00: :00 : :0:: E00500 E0 : .:
000.0 6:: .E00: 0: 05: E03: .0:

K 5:00“.
000.0 .000:
000.0000 0:: 5 0:90: 550 0:: :0:: 5:0: 0:0 2: 80.0000 E90: 0.: 55000000 :00: 0:09: : .00
00.0.0 3mm: .0EEEO: :0QEO: 0:: :0 000 :0095 :00 0: >50:
9. 03 00E::0E00 0:0 5.500090: 5.50:: :0 :.:0Q 0.5 0 0:0 50:00:03 >.:00E 0: 530:0: 0.00:0 >3. .3.
000.0 «00.9.: .:00E 00 0E :0 0300 0:00: :.:03 9050
0:0 530:0: 0.05:0 >E 05:0: 0: 30: E00: 0E 90: :.:3 80.0000 00.0090: 0.: E0E0>5>E .00
500.0 «mm: 50:85:50 09.020: 0..:0 >E :0:: 900:: 0: 00 :00 : 05:: b9. 0.: 99:. .0:
000.0 3mm: .0.:000020E00 0.: 0:50 >E 002:0 50.5050
.0....5 >E £3. 9:80:09 02:59:00 0:0 00.30000 .0000 :0 E9: 00500 0: o: 0::... 0.. 90E 0:.

:0 5:00".

05000..
5:00 H. E0:.

 

.000uz.
wm5.n. 0:: 5: :0.:0_0w 5:00:052 5:000:00 x0E:0> 5: 005000.. 5:00“. 0:0 E0:_ :0 >5EEOw

 

0.0 030..

258

 

Table 0.4 P

 

Descriptives for Non-orthogonal Five-Factor Solution

 

 

 

 

Factor
Descriptives One Two Three Four Five
Number of Items 17 12 7 3 1
Scale Mean 59.35 34.63 13.22 5.41 2.17
Standard Deviation 11.44 4.61 3.91 2.01 1.24
Item Alpha Reliabilities 0.91 0.58 0.68 0.42 na
Standardized Alphas 0.92 0.58 0.70 0.44 na

 

 

259

APPENDIX D
Logistic Regression Results Using Factor Scales in Lieu of Subscales

The factor analysis results did not support the subscales. From the
previous analyses, it appears that the factor analysis does not suggest the same
dimensions as the nine motivational aspect subscales. Instead there appears to
be a more collapsed structure of five dimensions. Thus, logistic regression
results using the factor scales instead of the subscales are presented in Table
D1

Humphries, llgren, McGrath, & Montanelli (1969) present information
regarding capitalizing on chance in factor analytic research. It has been argued
that purification of scales sometimes reduces the predictive value of the scale.

A summary of the correlations between individual items of the PMSS and
the child behavior change, parent satisfaction, and parent engagement (drop out)
outcomes is presented in Table D2. It is apparent that only a select number of
items (2 value, 3 self efficacy, 1 learning goal, 2 self defeating performance
goals, 1 optimistic explanatory style, 1 schematicity, and 1 pessimistic
explanatory style) are significantly correlated with the outcomes of the current

study.

260

 

 

 

 

 

 

 

 

28.0 83 08.0 00.0 30.9 085 5:95

a :8 000.0 80.0 80.0 000.0- 808800.80 880850

80.0 08.0 08.0 000.0 83 8838.280 8:005th

82. 82.0 80.0 80.0 000.0 22080 .8898: .0595

:80 08.0 80.0 03.0 83- 80> 02503
8890 800

.90 0.02, 080 m0 900 .98“.

 

 

0900m 5:00“: 0503 :00 005 .520 0002090 0200900”: 00053

_..D 030...

261

 

 

 

 

262

 

 

11 1'10: . -... .
83 83- 33 383- 83- 83- 83- 283 83 m0 .8
:3 83 83 83 83- :3 83 :3 38.0 03 .8
83- 203- 083- 83- 8 3 083 0:3 80.: 80.0 000 .8
83 S3- 83- 33- 83 83- 083- 38.: 83 03 .8
83- 83 83. 0 S3 83- 83- 83- 83 80.0 _m> 30
83 20.83 93 .83 R3 83 83 08.: 80.0 000 .2
83 83 083 0: 3 t 3- 83 83 :.83 80.0 _0> 3:
83 93- 33. 83- 83 N83- 83 3.: 08.: won. .t
20 3 0: 3 £3- 83 83- 9 3- 83- 8 Z 83 _m> .0:
08.9 83- 38.9 083- 3.83 83 .83 83 80.: m0 .8
:83. E 3 83- 83. : : 3- .23. 83- 08.: N2 3 000 .3
83 .383 203 83 80.9 83- 83- 083 83 _a> .2
83. 083- 83 08.? 83 33 083 08.: 08.: as. .8
83. .83 83. 083 83- 83 083. 83 «5.0 000.:
83 083- 02.? 83. 083. .38.? .083. 83 28.0 04 .2
83 93- 80.? R3- 83 83. 83- 383 08.: 00.”. 3
3: 3 83 8 3. 383- 08.9 .283- 83. 38.: 83 00 .0
83 83 083 02 3 0: 09 083 83- 08.: ~83 _0> .:-
083 83 83- 083 83 83 :33 08.: v8.0 000 .0
83 .83 83 .283 80.? 83 83 83 83 m0 .0
08.? 2.083 83 .:3 809 83- 83- 0 83 28.0 :3 .0
:3- R3- 83 83- 83 83 083 28.: 80.: as. .0
83. 083 8 3 83 83 83 83 08. F 80.0 _0> .0
83- 83 83- 083 383- 083 83 083 083 m0 .:
50 no.5 N 5:00“: :- :0:omn_ ESP 055.050wa 953.0595 .90... um :00—)— E0:
.5 8.80 8:80 000 00.20

 

00.0000> 005030

 

00.0000> 025050 £05 025: .0920 029200. :0 0500.950 000 00000300 0.00006 0000.2

N0 030%

 

000005.50 00..-..0000v. 00.00 0E0: 000.20... 5.3 00.0.0.50 0.0 00.00..0> 0:50.00 .00 00.0
0.5.00.0.50 . 000.001 00.0: 00.0.. 000.20... 5.3 00.0.0.50 0.0 00.00..0> 0:50.00 .0 .0 000 00500 000000 000 .052

.00.00-N. .82 8.0 05 .0 00000.00.» 0. 00029.8 ..
60:00-0. _0>0_ 8.0 05 .0 0:80.000 0 00009.8 .

 

000.0- 05 0.0 0 0N0 0 ..N0 .0000. 050.0 00 F0.- 0N00 000.0 .05 .3...
000.0- 00N.0 0 FNO 00N.0 0P 0.0. 0 FN.0 050.0 0000 0 ..00 _0> .0v
0N0.0- 0NNO. .000 ..000. $0080 00N0 N000 905.0 50N. .. mun. .NV
0500 N000- 00N0 00 ...0 v0...0 V000 50N.0 0N v.0 ..00-N QEzmwn. .3
..000. N 00.0 000.0. N000. 00N0- '0 r0. 50N0. 000.0 0 F0-N me .9.
0.00. 00N0 N 0.0 00 F0 00N0. 000.0- 50N.0. 050.0 000.N com .00
000.0- 50 V0. 50N0. 0VNO. 00...0. ..000- 000.0. 50N.F N000 .0.—.00
00 .. .0. 30.0 000.0. 000.0. 000.0- 0N .- .0 0 00.0- 0VN... V000 .0> .50
N000. 0 ..0. 00N0. 00N0. 000.0 0NPO 0.. ..0 00 F... 0000 00.”. .00
00...0 N000. 00N0. 0N0- 00F0. 0:0. 00...? 000... 000.N OOn. .00
000.0. 500.0 000.0 050.0 000.0 05 ..0 50 ..0 50N... V5 fiN .0> 00
00N0 N0 ..0. V000. 00 P0. 00 ..0 0NNO. .000. 505.0 000.0 wmva .00
00 P .0 000.0 N30- N 50 «000.0 000.0. 500.0 ..00... 000.N 000. .N0
030 05 ..0 N0 F0 50 5.0 000.0- 50 v0. 00 F0. 005.0 00.0 0.. ...0
000.0- .5000 «000.0 .0000 5000 N000 500.0 5000 0000 now .00
000.0 00N.0 500.0 000.0 0.00. N..0.0 0.. ..0. 000.0 53.-0 mmO .0N
0.00 03.0 00.0 00.0 0000 000.0 03.0 00N€ 0N0.0 0.. .0N
.... P0 000.0 0000. 000.0.- .0000. 000.0 31.0. NN0.0 v3.0 mm .5N
r000- 00...0- 30.0. 50...0- 0N00 v0N0 v0N0 300 000.0 wwaEE. .0N
0000 N30- 0v00 5 00.0 00 F .0 0N00- 0N00- 00N... 00 _..N Own. .0N

00:00 Nd 0.00..-

263

APPENDIX E

Method Of Calculating Socio-Economic Status

Socio—economic status (SES) was calculated according to two methods.
The first involved the Hollingshead Two Factor Index and the second involved
the most recent low income cutoffs from Statistics Canada.

The Hollingshead Two Factor Index (Hollingshead, 1965) has set the
standard and is the most widely used in most psychological research involving
PMT. The Index has been criticized for being dated and obsolete in its list of
occupations. However, few better alternatives are available and its most recent
update, the Hollingshead Four Factor Index uses marital status in its equation.
Because the Four factor index confounds one of the predictor variables, it was
decided to use the Two Factor Index as a continuous variable measure of SES.

To calculate SES according to the Hollingshead Index, two factors
(occupation and education) are assigned pre-determined scores obtained from
their respective scale positions. The two factors are multiplied by pre-determined
weights. The weights for each factor were originally determined by multiple
correlation techniques. Weights for each factor are as follows:

Table E.1

Factors and Weights used to Calculate Two Factor Index.

 

 

 

Factor Weight
Occupation 7
Education 4

 

The Low Income Cutoffs (LICO) from Statistics Canada (1999) are more

current, relevant to the population in the study, and provide a discrete measure of

SES. Estimates of income are based on a sample of 35,000 households
excluding households in the territories and on Indian reserves. Cut-offs were
selected on the basis that families with incomes below the limits usually spend
more than 54.7% of their income on food, shelter, and clothing, and are therefore
considered to be living in straightened circumstances. LlCOs are updated based
on information from the annual Survey of Household Spending to reflect changes
in the average proportion of family income spent on food, shelter, and clothing.
As a national average, the 1998 low income cut-off is $33,063 (rounded to the
nearest dollar). Seventeen and a half per cent of the Canadian population
sampled is said to be low income.

264

 

-1!

A family with income below the cutoff for its family size and urbanization
classification (size of area of residence) is considered a “low income” family.
Participants were coded 0 for low income or 1 for other by consulting the
following table:

Table E.2

Low Income Cutoffs using Number of Persons in Family Unit and Set Size of
Area of Residence (30,000 to 99,999 population)

 

 

 

Number of Persons Low Income Cutoff

 

14,965
18,706
23,264
28,162
31,481
34,798
7 or more 38,117

mmAwM—s

 

If low income cutoff fell within the survey range results, participants were coded
as low income.

265

 

 

APPENDIX F
Power Analysis Results for T-Tests
The power of a statistical test is the probability that results will be
statistically significant (Cohen, 1988) or in other words, power is the probability of
rejecting the null hypothesis when the null hypothesis is false. In Table E.1,
power analysis results are summarized for a two-tailed, t-test for means using
alpha set at .05 and alpha set at .0056 (bonferroni adjustment), respectively.
Using power equals .80 as a general rule of thumb, power in this study is
obviously low.
Table F.1

Post Hoc Power Analysis Results

 

 

 

 

Factors to Power Analyses
One Two

Effect Size (d) .5 .5
(medium)
Alpha .05 .0056
Sample size n1 64 64
Sample size n2 29 29
Delta 2.4614 2.4614
Critical t (101) 1.9837 2.8310
Power .6837 .361 1

 

266

 

AH _.

 

BIBLIOGRAPHY

Abikoff, H.B. , & Hecktman, L. (1996). Multimodal therapy and stimulants
in the treatment of children with ADHD. In E.D. Hibbs & P.S. Jensen (Eds.),
Psychosocial Treatments for Child and Adolescent Disorders. Empirically Based
Strategies for Clinical Practice (pp. 341-370). Washington, DC: American
Psychological Association.

 

 

Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18
and 1991 profile. Burlington: University of Vermont, Department of Psychiatry.

 

 

Adelman, H. S., & Taylor, L. (1997). Addressing barriers to Ieaming:
Beyond school-linked services and full-service schools. American Journal of
Orthpsychiatry, 67(3), 408-421.

 

 

Alworth, J. (1973). Home token economy: An incentive program for
children and their parents. Champaign, Illinois: Research Press.

 

 

American Psychiatric Association. (1980). Diagnostic and statistical
manual of mental disorders (3rd ed.). Washington, DC: Author.

 

 

American Psychiatric Association. (1987). Diagnostic and statistical
manual of mental disorders (3rd ed., revised). Washington, DC: Author.

 

 

American Psychiatric Association. (1994). Diagnostic and statistical
manual of mental disorders (4th ed.) Washington, DC: Author.

 

 

Ames, C., & Ames, R. (1984). Systems of student and teacher motivation:
Toward a qualitative definition. Journal of Educational Psycholog)’. 76(4), 535-
556.

 

Ames, C., & Archer, J. (1988). Achievement goals in the classroom:
Students' Ieaming strategies and motivation processes. Journal of Educational
Psychology, 80(3), 260-267.

 

 

Ames, R. (1985). Adult motivation in the leadership role: [An attribution
value analysis. Advances in Motivation and Achievement, 4, 333-369.

 

Anastopoulos, A.D., Barkley, R.A., Sheldon, TL. (1996). Family based
treatment: Psychosocial intervention for children and adolescents with Attention
Deficit Hyperactivity Disorder. In E.D. Hibbs & P.S. Jensen (Eds.), Psychosocial
Treatments for Child and Adolescent Disorders. Empirically Based Strategies for
Clinical Practice (pp. 267-284). Washington, DC: American Psychological
Association.

 

 

267

Applegate, B., Lahey, B. 8., Hart, E., Biederman, J., Hynd, G. W., Barkley,
R., Ollendick, T., Frick, P. J., Greenhill, L., McBumett, K., Newcom, J. H.,
Kerdyk, L., Garfinkel, B., Waldman, I., & Shaffer, D. (1997). Validity of the age of
onset criteria for ADHD: A report from the DSM-IV field trials. J. Am. Acad. Child
Adolesc. Psychiatry, 36(9), 1211-1221.

 

 

Association for Advanced Training in the Behavioral Sciences (1995).
Research design and statistics. Westlake Village, California: AATBS.

 

August, G. J., & Garfinkel, B. D. (1993). The nosology of Attention Deficit
Hyperactivity Disorder. J. Am. Acad. Child Adolesc. Psychiatry, 32(1), 155-165.

 

Bandura, A. (1997). Self-efficacy, the exercise of control. New York: W.H.
Freeman & Co.

 

Bank, L., Marlowe, J.H., Reid, J.B., Patterson, G.R., & Weinrott, MR.
(1991). A comparative evaluation of parent-training interventions for families of
chronic delinquents. Journal of Abnormal Child Psychology, 19(1), 15-33.

 

Barkley, RA. (1987). Defiant children: A clinician's manual for parent
training. New York: Guilford Press.

 

Barkley, R. A. (1990). Attention Deficit Hyperactivity Disorder: A handbook
for diagnosis and treatment. New York: Guilford Press.

 

 

Barkley, RA. (1992). ADHD: What do we Know? [Video]. New York:
Guilford Publications, Inc.

 

Barkley, RA. (1993). Attention Deficit Hyperactivity Disorder in children
and adults. Symposium conducted at the meeting of the New England Institute of
Psychiatry, Cape Cod, Massachusetts.

 

 

Barkley, RA. (1995). Taking charge of ADHD : The complete,
authoritative guide for parents. New York: Gulford press.

 

 

Barkley, RA. (1996). Attention-Deficit/Hyperactivity Disorder. In E.J. Mash
& R.A. Barkley (Eds), Child Psychopathology (pp. 63-112). New York: Guilford
Press.

 

Barkley (1997a). ADHD and the nature of self control. New York: Guilford

 

Press.

Barkley, R.A. (1997b). Defiant children: A clinician's manual for
assessment and parent training [2nd Edition]. New York: Guilford Press.

 

268

Barkley, RA (1998). Attention-Deficit/Hyperactivity Disorder. In E.J. Mash
& R.A. Barkley (Eds.), Treatment of Childhood Disorders (2nd Ed., pp. 55-110).
New York: Guilford Press.

 

Barkley, RA. (2000). Commentary: Issues in training parents to manage
children with behavior problems. J. Am. Acad. Child Adolesc. Psychiatry, 39(8),
1004-1006.

 

Beck, A.T., Steer, R.A., & Brown, G. K. (1996). BDI-II Manual. New York:
The Psychological Corporation, Harcourt Brace & Company.

 

Becker, WC. (1977). Parents are teachers. New York: Guilford Press.

 

Bell, 0., & Kanevsky, L. (1996). Promotingpositive achievement
motivation in a regular Grade 2 classroom. Paper presented at the annual
meeting of the American Educational Research Association, New York.

 

 

Bickman, L., Hefiinger, C. A., Northrup, D., Sonnichsen, S., & Schilling, S.
(1998). Long term outcomes to family caregiver empowerment. Journal of Child
and Family Studies, 7(3), 269-282.

 

 

Biederman, J., Farone, S. S., Milberger, S., & Doyle, A. (1993). Diagnosis
of Attention Deficit Hyperactivity Disorder from parent reports predict diagnoses
based on teacher reports. J. Am. Acad. Child Adolesc. Psychiatry, 32(2), 315-
321.

 

Bierman, K. L., Cole, J. 0., Dodge, K. A., Greenburg, M. T., Lochman, J.
E., 8 McMahon, R. J. (1996). Integrating social-skills training interventions with
parent training and family focused support to prevent conduct disorder in high-
risk populations. The fast track multisite demonstration project. Annals New York
Academy of Sciences, 256-264.

 

 

Bilbrey, J., & Bilbrey, P. (1995, July/August). Judging, trusting, and
utilizing outcomes data: A survey of behavioral health care payers. Behavioral
Health Care Tomorrow, 62-65.

 

Black, M.M. & Holden, E.W. (1995). Longitudinal intervention research in
children's health and development. Journal of Clinical Child Psychology, 24(2),
163-172.

 

Bloomquist, M. L. (1996). Skills training for children with behavior
disorders: A parent and therapist guidebook. New York: Guilford Press.

 

 

Brestan, E., & Eyberg, S. M. (1998). Effective psychosocial treatments for
children and adolescents with disruptive behavior disorders: 29 years, 82 studies,
and 5275 kids. Journal of Clinical Child Psychology, 27(2), 179-188.

 

269

Brestan, E. V., Jacobs, J. R., Rayfield, A. D., 8 Eyberg, S. M. (1999). A
consumer satisfaction measure for parent-child treatments and its relation to
measures of child behavior change. Behavior Therapy, 30, 17-30.

 

Brophy, J. (1998). Motivating students to learn. New York: McGraw-Hill
Companies, Inc.

 

Brophy, J. (1999). Toward a model of the value aspects of motivation in
education: Developing appreciation for particular Ieaming domains and activities.
Educational Psychologist, 34, 75-86.

 

Brunnquell, d., Crichton, L., & Egeland, B. (1981). Maternal personality
and attitudes and disturbances of child rearing. Amer. J. Orthopsychiat., 51(4),
680-691.

 

Carlson, C. (1993). The family school link: Methodological issues in
studies of family processes related to children's school competence. School
Psychology Quarterly, 8(4), 264-276.

 

Chiu, C., Dweck, C. S., Tong, J. Y., & Fu, J. H. (1997). Implicit theories
and conceptions of morality. Journal of Personality and Social Psychology, 73(5),
923-940.

 

Clarizio, HP. (1996). Conduct Disorder: Developmental considerations.
Paper presented at the Annual Meeting of the American Psychological
Association, Toronto.

 

Clarizio, H. F., Mehrens, W. A., & Hapkiewicz, W. G. (Eds). (1994).
Contemporary issues in educational psychology (6th ed.). New York: McGraw-
Hill, Inc.

 

Cochran, M., & Niegro, S. (1995). Parenting and social networks. In M. C.
Bomstein (Ed.), Handbook of Parenting: Status and Social Conditions of
Parenting (Vol. 3, pp. 393-418). Mahwah, New Jersey: Lawrence Erlbaum,
Associates.

 

Cohen, J. (1988). Statistical power analysis for the behaVioraI sciences
(2nd ed.). Hillsdale, New Jersey: Lawrence Erlbaum Associates, Publishers.

 

Cowen, J. (1990). Things I have Ieamed so far. American Psychologist,
45(12), 1304-1312.

 

Comings, D. E. (1997). Genetic aspects of childhood behavioral disorders.
Child Psychiatry and Human Development, 27(3), 139-149.

 

270

 

 

 

Covington, M. (1992). Makiriglhe grade: a self-worth perspective on
motivation and school reform. Cambridge, UK: Cambridge University Press.

 

 

Cowan, EL, 8 Hightower, AD. (1988). Alternative approaches in school
based preventive intervention. In R.H. Price, E.L. Cowan, R.P. Lorion, 8 J.
Ramos-McKay (Eds.), Fourteen ounces of prevention: A casebook for
practitioners. Washington, DC: American Psychological Association.

 

 

Cox, M. J., Owen, M. T., Lewis, J. M., 8 Henderson, V. K. (1989).
Marriage, adult adjustment, and early parenting. Child Development, 60(1015-
1024)

 

Crnic, K., 8 Acevedo, M. (1995). Everyday stresses and parenting. In M.
H. Bomstein (Ed.), Applied and Practical Parenting (Vol. 4, ). Mayway, New
Jersey: Lawrence Erlbaum Associates, Publishers.

 

Cross, 8. E., 8 Markus, H. R. (1994). Self-schemas, possible selves, and
competent performance. Journal of Educational Psychology, 86(3), 432-438.

 

Dadds, MR, 8 McHugh, TA. (1992). Social support and treatment
outcome in behavioral family therapy for child conduct problems. Journal of
Consulting and Clinical Psychology, 60(2), 252-259.

 

Danforth, J. S. (1999). The outcome of parent training using the behavior
management flow chart with a mother and her twin boys with oppositional defiant
disorder and attention-deficit hyperactivity disorder. Child 8 Family Behavior
Therapy, 21(4), 59-80.

 

 

Davidson, W.S., 8 Redner, R. (1988). The prevention of juvenile
delinquency: Diversion from the juvenile justice system. In R.H. Price, E.L.
Cowan, R.P. Lorion, 8 J. Ramos-McKay (Eds.), Fourteen ounces of prevention:
A casebook for practitioners. Washington, DC: American Psychological
Association.

 

 

Deschler, 0.0., 8 Schumaker, JB (1994). Grouping intervention research
in the larger context of schooling: A response to Pressley and Harris. Educational

 

Psychology Review, 6(3), 215-230.

 

Dinkmeyer, D. 8 McKay, GD. (1989). STEP: Systematic training for
effective parenting. Circle Pines, MN: American Guidance Service.

 

 

Drotar, D., Stein, R.E.K., 8 Perrin, EC. (1995). Methodological issues in
using the Child Behavior Checklist and its related instruments in clinical child
psychology research. Journal of Clinical Child Psychology, 24(2), 184-192.

 

271

 

I...

 

Dumas, J.E., 8 Wahler, R.G. (1983). Predictors of treatment outcome in
parent training: Mother insularity and socioeconomic disadvantage. Behavioral
Assessment, 5, 301 -31 3.

 

Dunlap, G., 8 Childs, K.E. (1996). Intervention research in emotional and
behavioral disorders: An analysis of studies from 1980-1993. Behavioral
Disorders, 21(2), 125-136.

 

Dunteman, G.H. (1989). Principal components analysis. (Vol. 69). New
York: Sage Publications.

Durlak, J.A., Wells, A.M., Cotton, K., 8 Johnson, S. (1995). Analysis of
selected methodological issues in child psychotherapy research. Journal of
Clinical Child Psychology, 24(2), 141-148.

 

Dweck, C. S. (1996). Capturing the dynamic nature of personality. Journal
of research in personality, 30, 348-362.

 

Dweck, C. S., 8 Leggett, E. L. (1988). A social-cognitive approach to
motivation and personality. Psychological Review, 95(2), 256-273.

 

Dyer, K., 8 Larrson, E. V. (1997). Developing functioning communicating
skills: Alternatives to severe behavior disorders. In N. N. Singh (Ed.), Prevention
and Treatment of Severe Behavior Problems (pp. 121-148). Detroit: Brooks/ Cole
Publishing Company.

 

Dykman, RA, 8 Ackerman, RT. (1993). Behavioral subtypes of Attention
Deficit Disorder. Exceptional Children, 60(20), 132-141.

 

Eber, L., 8 Nelson, C. M. (1997). School-based wraparound for students
with emotional and behavioral challenges. Exceptional Children, 63(4), 539-556.

 

Eifert, G.H. (1996). More theory driven and less diagnosis based behavior
therapy. J. Behav. Ther. 8 Exp. Psychiat, 27(2), 75-86.

 

Elliot, E. S., 8 Dweck, C. S. (1988). Goals: An approach to motivation and
achievement. Journal of Personality and Social Psychology, 54(1), 5-12.

 

Eyberg, S. M. (1993). Consumer satisfaction measures for assessing
parent training programs. In L. VandeCreek, S. Knapp, 8 T. L. Jackson (Eds.),
Innovations in Clinical Practice: A source book (Vol. 12, ). Sarasota, FL:
Professional Research Press.

 

Eyberg, SM. (1995). Jigsaws: Introduction to the special issue on
methodological issues in clinical child psychology research. Journal of Clinical
Child Psychology, 24(2), 214-226.

 

 

272

 

.1_—’—_

Ti"
p. .

 

 

Eyberg, S. M., 8 Boggs, SR. (1998). Parent-child interaction therapy for
oppositional preschools. In C.E. Schaefer 8 J.M. Briesmeister (Eds.), Handbook
of parent training: Parents as co-therapists for children's behavior (2nd ed., pp.
61-97). New York: Wiley.

 

Eyberg, S. M., Edwards, 0., Boggs, S. R., 8 Foote, R. (1998). Maintaining
the treatment effects of parent training: The role of booster sessions and other

maintenance strategies. Clinical Psychology: Science and Practice, 5(4), 544-
554.

 

Faul, F., 8 Erdfelder, E. (1992). A priori, mst-hoc, and compromise power
analysis for MS-DOS [Computer Program]. Bonn, FRG: Bonn University, Dep. of
Psychology.

 

 

Finch, M. K. (1998). Social skills training for preadolescents with problem
behaviors (parent training). Unpublished Dissertation, University of Illinois at
Chicago, Chicago.

 

 

First, J. A., 8 Way, W. L. (1995). Parent education outcomes: Insights into
transformative Ieaming. Family Relations, 44, 104-109.

 

Ford-Gilboe, M. (1997). Family strengths, motivation, and resources as
predictors of health promotion behavior in single-parent and two-parent families.
Research in Nursing 8 Health, 20, 205-217.

 

Forehand, R.L., 8 Mahon, R.J. (1981). Helping the noncompliant child: A
clinician's guide to parent training. New York: Guilford Press.

 

 

Forgatch, M., 8 Patterson, GR. (1990). Parents and adolescents living
together. Eugene, OR: Castalia.

 

Fomess, SR, 8 Hoagwood, K. (1993). Issues in sampling, design, and
implementation of school based mental health services research. School
Psychology Quarterly, 8(4), 291-300.

 

Forness, S. R., Kavale, K. A., MacMillan, D. L., Asarnow, J. R., 8 Duncan,
B. B. (1996). Early detection and prevention of emotional or behavior disorders:
Developmental aspects of systems of care. Behavior Disorders, 21(3), 226-240.

 

Foster, S.L., Martinez, Jr., CR. (1995). Ethnicity: Conceptual and
methodological issues in child clinical research. Journal of Clinical Child
Psychology, 24(2), 214-226.

 

 

Fox, R.A., 8 Fox, TA (1990) STAR parenting: A positive and practical
approach to effective parenting. Bellevue, WA: Star Parenting, Inc.

 

 

273

 

' ““7
r‘r

 

Fox, R. A., 8 Parroni Hennick, E. (1996). Evaluating a training program for
parental educators. Psychological Reports, 79, 1143-1150.

 

Frankel, F., 8 Simmons, III, J.Q. (1992). Parent behavioral training: why
and when some parents drop out. Journal of Clinical Child Psychology, 21 (4),
322-330.

 

Frick, P. J. (1998). Conduct Disorders. In T. H. Ollendick 8 M. Hersen
(Eds.), Handbook of Child Psychopathology (3rd Ed., pp. 213-237). New York:
Plenum Press.

 

Geller, J., 8 Johnston, C. (1995). Predictors of mothers' responses to child
noncompliance: Attributions and attitudes. Journal of Clinical Child Psychology,
24(3), 272-278.

 

Gillis-Amold, R., Crase, S. J., Stockdale, D. F., 8 Shelley, M. C. I. (1998).
Parenting attitudes, foster parenting attitudes, and motivations of adoptive and
nonadoptive foster parent trainees. Children and Youth Services Review, 20(8),
715-732.

 

Goldberg, W. A. (1990). Marital quality, parental personality, and spousal
agreement about perceptions and expectations for children. Merrill-Palmer
Quarterly, 36(4), 531-556.

 

 

Goodman, G., 8 Poillion, M.J. (1992). ADD: Acronym for any dysfunction
or difficulty. The Journal of Special Education, 26(1), 37-56.

 

Goodnow, J. J. (1995). Parents' knowledge and expectations. In M. H.
Bomstein (Ed.), Handbook of Parenting: Status and Social Conditions of
Parenting (Vol. 3, pp. 305-332). Mahway, New Jersey: Lawrence Erlbaum
Associates, Publishers.

 

Goodyear, P., 8 Hynd, G.W. (1992). Attention Deficit Disorder with
(ADD/H) and without (ADDNVO) Hyperactivity: Behavioral and
neuropsychological differentiation. Journal of Clinical Child Psychology, 21(3),
273-305.

 

Gordon, T. (1975). P.E.T., parent effectiveness training: The tested new
way to raise responsible children. New York: P.H. Wyden.

 

 

Greenberg, G.S., 8 Horn, W.F. (1991). Attention Deficit Disorder:
Questions and answers for parents. Champaign, Illinois: Research Press.

 

 

274

 

 

 

Greenhill, LL. (1998). Childhood Attention Deficit Hyperactivity Disorder:
Pharmacological treatments. In P.E. Nathan 8 J.M. Gorman (Eds.), A Guide to
Treatments that Work, (pp. 42-64). New York: Oxford University Press.

 

Greene, L., Kamps, D., Wyble, J., 8 Ellis, C. (1999). Home-based
consultation for parents of young children with behavioral problems. Child 8
Family Behavior Therapy, 21(2), 19-45.

 

Hall, L. A., Gurley, D. N., Sachs, B., 8 Kryscio, R. J. (1991). Psychosocial
predictors of maternal depressive symptoms, parenting attitudes, and child
behavior in single-parent families. NursinLResearch, 40(4), 214-220.

 

Hardman, M. L., Drew, C. L., 8 Egan, M. W. (1997). Human
exceptionality: Society, school, and family. (Fifth Edition ed.). Boston: Allyn and
Bacon.

 

 

Hardman, M. L., Drew, C. L., 8 Egan, M. W. (1999). Human
exceptionalitLSociety, school, and family. (Sixth Edition ed.). Boston: Allyn and
Bacon.

 

 

Heflinger, C. A., Bickman, L., Northrup, D., 8 Sonnichsen, S. (1997). A
theory-driven intervention and evaluation to explore family caregiver
empowerment. Journal of Emotional and Behavioral Disorders, 5(3), 184-191.

 

Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland, M. D., 8
Cunningham, P. B. (1998). Multisystemic Treatment of Antisocial Behavior in
Children and Adolescents. New York: The Guilford Press.

 

 

Henggeler, S.W. Smith, B.H., 8 Schoenwald, SK. (1994). Key theoretical
and methodological issues in conducting treatment research in the juvenile
system. Journal of Clinical Child Psychology, 23(2), 143-150.

 

Heyman, G. D., 8 Dweck, C. S. (1998). Children's thinking about traits:
Implications for judgements of the self and others. Child Development, 64(2),
391-403.

 

Hinshaw, SP. (1994). Attention Deficit Hyperactivity Disorder. Volume 29.
Developmental clinical psychology and psychiatry. Thousand Oaks, California:
Sage Publications, Inc.

 

 

Hinshaw, SP. (1996). Enhancing social competence: Integrating self-
management strategies with behavioral procedures for children with ADHD. In
E.D. Hibbs 8 PS. Jensen (Eds.), Psychosocial Treatments for Child and
Adolescent Disorders. Empirically Based Strategies for Clinical Practice (pp. 285-
310). Washington, DC: American Psychological Association.

 

275

 

‘1

 

 

Hinshaw, SP, 8 Anderson, CA. (1996). Conduct and Oppositional
Defiant Disorders. In E.J. Mash 8 RA. Barkley (Eds.), Child Psychopathology
(pp. 113-152). New York: Guilford Press.

 

Hinshaw, S. R, Klein, R. G., 8 Abikoff, H. (1998). Childhood Attention
Deficit Hyperactivity Disorder: Nonpharrnacological and combination treatments.
In P. E. Nathan 8 J. M. Gorman (Eds.), A Guide To Treatments that Work (pp.
26-41). New York: Oxford University Press.

 

Hoagwood, K. (1993). Introduction: Methodological issues in school-based
mental health services research. School Psychology Quarterly, 8(4), 239-240.

 

Holden, G.W., Lavigne, V.V., 8 Cameron, AM. (1990). Probing the
continuum of effectiveness in parent training: Characteristics of parents and
preschoolers. Journal of Clinical Child Psychology, 19(1), 2-8.

 

Holden, G. W. (1995). Parental attitudes toward childrearing. In M. E.
Bomstein (Ed.), Handbook of Parenting: Status and Social Conditions of
Parenting (Vol. 3, ). Mahway, New Jersey: Lawrence Erlbaum Associates,
Publishers.

 

Hollingshead, AB. (1965). Two factor index of social position. New Haven,
Connecticut: Yale University Station, Department of Sociology.

 

Hollingshead, AB, 8 Redlick, RC. (1958). Social class and mental
illness. New York: John Wiley and Sons.

 

Hong, Y., Chiu, G, Dweck, C. S., 8 Sacks. (1997). Implicit theories and
evaluative processes in person cognition. Journal of Experimental Social
Psychology, 33, 296-323.

 

 

Horvath, T. ( 1981). Basic research statistics. Windsor, Ontario: University
of Windsor Press.

 

Humprhies, L. G., Montanelli, R.G. (1975). An investigation of the parallel
analysis criterion for determining the number of common factors. 193-205

Hutchings, J. (1996). Evaluating a behaviorally based parent training
group: Outcomes for parents, children and health visitors. Behavioral and
Cognitive Psychotherapy, 24, 149-170.

 

 

Johnston, C. (1996). Addressing parent cognitions in interventions with
families of disruptive children. In K. S. Dobson 8 K. D. Craig (Eds.), Advances in
Cognitive Behavioral Therapy (pp. 193-208). Thousand, Oaks: Sage
Publications: Sage Publications.

 

 

276

 

 

 

Kaiser, A. P., 8 Hester, P. P. (1997). Prevention of conduct disorder
through early intervention: A social communicative perspective. Behavioral
Disorders, 2(3), 117-130.

 

Kanfer, F. H. (1996). Motivation and emotion in behavior therapy. In K. S.
Dobson 8 K. D. Craig (Eds.), Advances in Behavioral Therapy (pp. 1-26).
Thousand, Oaks: Sage Publications.

 

Kanigsberg, J.S., 8 Levant, RF. (1988). Parental attitudes and children's
self-concept and behavior following parents' participation in parent training
groups. Journal of Community Psychology, 16, 152-160.

 

Kazdin, A.E. (1987). Treatment of antisocial behavior in children: Current
status and future directions. Psychological Bulletin, 102(2), 187-203.

 

Kazdin, A.E. (1996). Problem solving and parent management in treating
aggressive and antisocial behavior. In E.D. Hibbs 8 PS. Jensen (Eds.),
Psychosocial Treatments for Child and Adolescent Disorders. Empirically Based
Strategies for Clinical Practice (pp. 377-408). Washington, DC: American
Psychological Association.

Kazdin, A. E. (1997a). Parent management training: Evidence, outcomes,
and issues. Journal of Consulting and Clinical Psychology, 65(3), 453-463.

 

Kazdin, A.E. (1997b). Practitioner review: Psychosocial treatments for
Conduct Disorder in children. J. Child Psychol. Psychiat., 38(2), 161-178.

 

Kazdin, A. (1998). Psychosocial treatments for Conduct Disorder in
children. In P. E. Nathan 8 J. M. Gorman (Eds.), A Guide to Treatments That
Work (pp. 65-89). New York: Oxford University Press.

 

Kazdin, A. E., Holland, L., 8 Crowley, M. (1997). Family experience of
barriers to treatment and premature termination from child therapy. Journal of
Consulting and Clinical Psycholfly, 65(3), 453-463.

 

Kazdin, A. E., 8 Wassell, G. (1999). Barriers to treatment participation and
therapeutic change among children referred for Conduct Disorder. Journal of
Clinical Child Psychology, 28(2), 160-172.

 

Kazdin, A. E., 8 Weisz, J. R. (1998). Identifying and developing empirically
supported child and adolescent treatments. Journal of Consulting and Clinical
Psychology. 66(1), 19-36.

 

 

Koegel, R. L., Bimbela, A., 8 Schreibman, L. (1996). Collateral effects of
parent training on family interactions. Journal of Autism and Developmental
Disorders, 26(3), 347-358.

 

 

277

Kim, Jae-On, 8 Mueller, CW. (1978). Introduction to factor analysis. (Vol.
13). New York: Sage Publications.

Kim, Jae-On, 8 Mueller, CW. (1978). Factor analysis: Statistical methods
and practical issues. (Vol. 14). New York: Sage Publications.

Kohn, M. L. (1969). Class and conformity: A study in values. Homewood,
III.: Dorsey Press.

 

Kosterman, R., Hawkins, J. D., Spoth, R., Haggerty, K. P., 8 Zhu, K.
(1997). Effects of a preventive parent-training intervention on observed family

interactions: Proximal outcomes from preparing for the drug free years. Journal
of Community Psychology, 25(4), 337-352.

 

Lehtinen, E., Vauras, M., Salonen, P., Olkinuora, E., 8 Kinnunen, R.
(1995). Long-term development of Ieaming activity: Motivational, cognitive, and
social interaction. Educational Psychologist, 30, 21-35.

 

Lemos, M. S. (1996). Students' and teachers' goals in the classroom.
Learning and Instruction, 6(2), 151-171.

 

Lerner, J. 8 Lerner, S. (1991). Attention Deficit Disorder: Issues and
questions. Focus on Exceptional Children, 24(3), 6.

 

Levine, MS. (1977). Canonical analysis and factor comparison. (Vol. 6).
New York: Sage Publications.

 

Levy, S. R., Stroessner, S. J., 8 Dweck, C. S. (1998). Stereotype
formation and endorsement: The role of implicit theories. Journal of Personality
and Social Psychology, 74(6), 1421 -1436.

 

 

Lisi, M.-De., 8 Sigel, I. E. (1995). Parental Beliefs. In M. H. Bomstein
(Ed.), Handbook of Parenting: Status and Social Conditions of Parenting (Vol. 3,
pp. 333-349). Mahway, New Jersey: Lawrence Erlbaum Associates, Publishers.

 

Loeber, R., 8 Stouthamer-Loeber, M. (1998). Development of juvenile
aggression and violence: Some common misconceptions and controversies.
American Psychologist, 53(2), 242-259.

 

Maehr, M. L., 8 Pintrich, P. R. (Eds.). (1999). Advances in Motivation and
Achievement. (Vol. 11). Stamford, CT: JAI Press Inc.

 

 

Markus, H., 8 Nurius, P. (1986). Possible selves. American Psychologist,
41, 954-969.

 

278

 

 

 

 

Marx, B.P., 8 Gross, AM. (1998). Behavioral treatment. In T.H. Ollendick
8 M. Hersen (Eds.), Handbook of Child Psychopathology (3rd Ed., pp. 581-602).
New York: Plenum Press.

 

Matson, J. L., 8 Duncan, D. (1997). Aggression. In N. N. Singh (Ed.),
Prevention and Treatment of Severe Behavior Problems (pp. 217-236). Detroit:
Brooks/Cole Publishing Company.

 

McBumett, K., Lahey, B. B., 8 Pfiffner, L. J. (1993). Diagnosis of Attention
Deficit Disorders in DSM-IV: Scientific basis and implications for education.
Exceptional Children, 60(2), 108-117.

 

McConaughy, S. H., Kay, P. J., 8 Fitzgerald, M. (1998). Preventing SED
through parent-teacher action research and social skills instruction: First-year
outcomes. Journal of Emotional and Behavioral Disorders, 6(2), 81-93.

 

McMahon, R.J., 8 Wells, KC. (1998). Conduct Problems. In E.J. Mash 8
RA. Barkley (Eds.), Treatment of Childhood Disorders (2nd Ed., pp. 111-210).
New York: Guilford Press.

 

Meece, J. L., 8 Holt, K. (1993). A pattern analysis of students'
achievement goals. Journal of Educational Psychology. 85(4), 582-590.

 

Midgley, C., Arunkumar, R., 8 Urdan, T. C. (1996). "If I don't do well
tomorrow, there's a reason": Predictors of Adolescents' use of academic self-
handicapping strategies. Journal of Educational Psychology, 88(3), 432-434.

 

Midgley, C., Feldlaufer, H., 8 Eccles, J. S. (1989). Change in teacher
efficacy and student self- and task-related beliefs in mathematics during the
transition to junior high school. Journal of Educational Psychology, 81(2), 247-
258.

 

Moreland, J.R., Schwebel, A.I., Beck, 8., 8 Wells, R. (1982). Parents as
therapists: A review of behavior therapy parent training literature - 1975-1981.
Behavior Modification, 6, 250-276.

 

Morgan, A. E., Hynd, G. W., Riccio, C. A., 8 Hall, J. (1996). Validity of the
DSM-IV ADHD predominantly inattentive and combined types: Relationship to
previous DSM diagnoses/ subtype differences. J. Am. Acad. Child Adolesc.
Psychiatry, 35(3), 325-333.

 

 

Murray, BA, 8 Myers, MA. (1998). Conduct Disorders and the Special-
Education Trap. The Education Digest, 48-54.

 

279

 

Nathan, PE, 8 Gorman, J.M. (1998). Summary of treatments that work.
In P.E. Nathan 8 J.M. Gorman (Eds.), A Guide to Treatments that Work, (pp. xix-
xxx). New York: Oxford University Press.

 

Newby, T.J., Stepich, D.A., Lehman, J.D., 8 Russell, JD. (1996).
Instructional technology for teaching and Ieaming. Englewood Cliffs, New Jersey:
Prentice Hall, Inc.

 

Nicholls, J. G. (1984). Achievement motivation; Conceptions of ability,
subjective experience, task choice, and performance. Psychological Review, 91,
328-346.

 

Pajares, F. (1996). Self-efficacy beliefs in academic settings. Review of
Educational Research, 66(4), 543-578.

 

Patterson, GR, 8 Forgatch, MS. (1987). Parents and adolescents living
together--Part I: The basics. Eugene, OR: Castalia.

 

 

Pelham, W. (1996). ADHD in children, adolescents, and adults: Recent
advances. Paper presented at the Meeting of the New England Institute of
Psychiatry, Cape Cod, Massachusetts.

 

Pelham, W., 8 Hoza, B. (1996). Intensive treatment: Summer treatment
program for children with ADHD. In E.D. Hibbs 8 PS. Jensen (Eds.),
Psychosocial Treatments for Child and Adolescent Disorders. Empirically Based
Strategies for Clinical Practice (pp. 311-340). Washington, DC: American
Psychological Association.

 

 

Peterson, L., 8 Bell-Dolan, D. (1995). Treatment outcome research in
child psychology: Realistic coping with the "Ten commandments of
Methodology". Journal of Clinical Child Psychology, 24(2), 149-162.

 

Phelan, T. (1995). 1-2-3 Magic: Effective discipline for children 2-12 (Rev.
2nd ed.). Glen Ellyn, Illinois: Child Management Inc.

 

Pintrich, P. R., 8 De Groot, E. V. (1990). Motivational and self-regulated
learning components of classroom academic performance. Journal of
Educational Psychology, 82(1), 33-40.

 

Pomerleau, A., Malcuit, G., 8 Sabatier, C. (1991). Child-rearing practices
and parental beliefs in three cultural groups of Montreal: Quebecois, Vietnamese,
Haitian. In M.H. Bomstein (Ed.), Cultural approaches to parenting (pp. 45-68).
Hilsdale, WJ: Lawrence Erlbaum Associates.

 

280

Porter, AC (1988). Comparative experiments in educational research. In
R. M. Jaeger (Ed.), Complementary methods for research in education (pp. 391-
414). Washington, DC: American Educational Research Association.

Power, T. J., 8 DuPaul, G. J. (1996). Attention Deficit/Hyperactivity
Disorder: The re-emergence of subtypes. School Psychology Review, 25(3), 284-
296.

 

Pressley, M., 8 Harris, KR. (1994). Increasing the quality of educational
intervention research. Educational Psychology Review, 6(3), 191-208.

 

Price, R., Cowen, E. L., Lorion, R. P., 8 Ramos-McKay, J. (1989). The
search for effective prevention programs: What we Ieamed along the way.
American Journal of Orthopsychiatry, 59(1), 49-58.

Prinz, R. J., 8 Miller, G. E. (1994). Family-based treatment for childhood
antisocial behavior: Experimental influences on dropout and engagement.
Journal of Consultingand Clinical Psychology, 62, 645-650.

 

Ramey, C. T., Bryant, D.M., Campbell, F.A., Sparling, J.J., 8 Wasik, B.H.
(1988). In R.H. Price, E.L. Cowan, R.P. Lorion, 8 J. Ramos-McKay (Eds.),
Fourteen ounces of prevention: A casebook for practitioners. Washington, DC:
American Psychological Association.

Reeve, J. (1996). Motivating others: Nurturinginner motivational
resources. Boston: Allyn and Bacon.

 

Rencher, A. C. (1995). Methods of multivariate analysis. New York: John
Wiley 8 Sons, Inc.

 

Ruma, P.R., Burke, R.V., 8 Thompson, R.W. (1996). Group parent
training: Is it effective for children of all ages? Behavior Therapy, 27, 159-169.

 

Rogers Wiese, MR. (1992). A critical review of parent training research.
Psychology in the Schools, 29, 229-236.

 

Rogers Wiese, MR, 8 Kramer, J.J. (1988). Parent training research: An
analysis of the empirical literature 1975-1985. Psychology in the Schools, 25,
325-330.

 

Shapiro, ES. (1987). Intervention research methodology in school
psychology. School Psychology Review, 16(3), 290-305.

 

Shawaker, T., 8 Dembo, M. (1996). The effects of efficacy buildinj
instruction on the use of Ieaming strategies. Paper presented at the annual
meeting of the American Educational Research Association, New York.

 

 

281

Shulman, LS. (1988). Disciplines of inquiry in education: An overview. In
R. M. Jaeger (Ed.), Complementary methods for research in education (pp. 4-
17). Washington, DC: American Educational Research Association.

 

Skaalvik, E. M. (1997). Self-enhancing and self-defeating ego orientation:
Relations with task and avoidance orientation, achievement, self-perceptions,
and anxiety. Journal of Educational Psychology, 89(1), 71-81.

 

Spitzer, A., Webster-Stratton, C., 8 Hollinsworth, T. (1991). Coping with
conduct problem children: Parents gaining knowledge and control. Journal of
Clinical Child Psychology. 20(4), 413-427.

 

Stahl, N., 8 Clarizio, H. (1999). Conduct disorder and comorbidity.
Psychology in the Schools, 36(1), 41-49.

 

Statistics Canada (1999, December). Low income cut-offs (LICOS).
Available http:/l www.statcan.ca/cgi-bin/sear... TION = Highlight 8 ITEM = 1 8
DAILY = 990414.

 

Stipek, 0., Salmon, J. M., Giwin, K. B., 8 Kazemi, E. (1998). The value
(and convergence) of practices suggested by motivation research and promoted
by mathematics education reformers. Journal for Research in Mathematics
Education, 29, 465-488.

 

 

Strayhorn, J. M., Strain, P. S., 8 Walker, H. M. (1993). The case for
interaction skill training in the context of tutoring as a preventive mental health
intervention in schools. Behavioral Disorders, 19(1), 11-26.

 

Strayhorn, J. M., 8 Weidman, C. S. (1991). Follow up one year after
parent-child interaction training: Effects on behavior of preschool children. J. Am.
Acad. Child Adolesc. Psychiatry, 30(1), 138-143.

 

Tannock, R. (1998). Attention Deficit Hyperactivity Disorder (ADHD):
Understanding and intervention. Paper presented at the 51st Annual OPA
Convention, Toronto, Ontario.

 

 

Tolan, P., 8 Guerra, N. (1995). What works in reducing adolescent
violence: An empirical review of the field. Boulder, Co: Center for the Study and
Prevention of Violence.

 

 

Thompson, B. (1984). Canonical correlation analysis: Uses and
interpretation. (Vol. 47). New York: Sage Publications.

 

 

282

Thompson, R. W., Ruma, P. r., Schuchmann, L. F., 8 Burke, R. v. (1996).
A cost-effectiveness evaluation of parent training. Journal of Child and Family
Studies, 5(4), 415-429.

 

Tucker, 8., 8 Gross, D., (1997). Behavioral parent training: An intervention
strategy for guiding parents of young children. The Journal of Perinatal
Education, 6, 35-44.

 

 

Tucker, 8., Gross, D., Fogg, L., Delaney, K., 8 Lapporte, R. (1998). The
long term efficacy of a behavioral parent training intervention for families with 2-
year-olds. Research in Nursing 8 Health, 21, 199-210.

 

US. Bureau of the Census (1994). Statistical abstracts of the United
States: 1994 (114th ed.). Washington, DC: Author.

 

 

Vuchinich, S., Wood, 8., 8 Angelelli, J. (1996). Coalitions and family
problem solving in the psychosocial treatment of preadolescents. In E.D. Hibbs 8
PS. Jensen (Eds.), Psychosocial Treatments for Child and Adolescent
Disorders. Empirically Based Strategies for Clinical Practice (pp. 497-520).
Washington, DC: American Psychological Association.

 

 

Webster-Stratton, C. (1985). Predictors of treatment outcome in parent
training for conduct disordered children. Behavior Therapy, 16, 223-243.

 

Webster-Stratton, C. (1990). Long-term follow-up of families with young
conduct problem children: From preschool to grade school. Journal of Clinical
Child Psycholog, 19(2), 144-149.

 

 

Webster-Stratton, C. (1991). Annotation: Strategies for helping families
with Conduct Disordered children. J. Child Psychol. Psychiat., 32(7), 1047-1062.

Webster-Stratton, C. (1992). The parents and children videotape series:
Programs 1-10. Seattle, Washington: Seth Enterprises.

 

 

Webster-Stratton, C. (1993). Strategies for helping early school-aged
children with oppositional defiant and conduct disorders: The importance of
home-school partnerships. School Psychology Review, 22(3), 437-457.

 

Webster-Stratton, C. (1996a). Early-onset conduct problems: Does gender
make a difference? Journal of Consulting and Clinical Psychology, 64(3), 540-
551.

 

Webster-Stratton, C. (1996). Early intervention with videotape modeling:
Programs for families of children with Oppositional Defiant Disorder or Conduct
Disorder. In E.D. Hibbs 8 PS. Jensen (Eds.), Psychosocial Treatments for Child

 

283

and Adolescent Disorders. Empirically Based Strategies for Clinical Practice (pp.
435-474). Washington, DC: American Psychological Association.

Webster-Stratton, C. (1998). Preventing conduct problems in Head Start
children: Strengthening parenting competencies. Journal of Consulting and
Clinical Psychology. 66(5), 715-730.

 

 

Webster-Stratton, C., 8 Hammond, M. (1990). Predictors of treatment
outcome in parent training for families with conduct problem children. Behavior
Therapy, 21 , 319-337.

 

Webster-Stratton, C. 8 Hammond, M. (1997). Treating children with early-
onset conduct problems: A comparison of child and parent training interventions.
Journal of Consulting and Clinical Psychology, 65(1), 93-109.

 

Webster-Stratton, C., 8 Hebert, M. (1993). "What really happens in parent
training?". Behavior Modification, 17(4), 407-456.

 

Webster-Stratton, C., Kolpacoff, M., 8 Hollinsworth, T. (1995). Self-
administered videotape therapy for families with conduct-problem childrenl
Journal of Consulting and Clinical Psychology, 57, 550-553.

 

Webster-Stratton, C., 8 Spritzer, A. (1996). Parenting a young child wit
conduct problems. In T.H. Ollendick 8 R. J. Prinz (Eds.), Advances in clinical
child psychology (Vol. 18, pp. 1-62).

 

 

Weiner, B. (1979). A theory of motivation for some classroom
experiences. Journal of Educational Psychology, 71(1), 3-25.

 

Wells, KC. (1993). Parent trainiananual for oppositional children.
Symposium conducted in Toronto, Ontario.

 

Whalen, C.K., 8 Henker, B. (1998). Attention-Deficit/Hyperactivity
Disorders. In T. H. Ollendick 8 M. Hersen (Eds.), Handbook of Child
Psychopathology (3rd Ed., pp. 181-211). New York: Plenum Press.

 

 

Wierzbicki, M., 8 Pekarik, G. (1993). A meta-analysis of psychotherapy
dropout. Professional Psychology: Research and Practice, 24(2), 190-195.

 

Wolery, M., 8 Winterling, V. (1997). Curricular approaches to controlling
severe behavior problems. In N. N. Singh (Ed.), Prevention and Treatment of
Severe Behavior Problems (pp. 217-236). Detroit: Brook/Cole Publishing
Company.

 

 

284

Wood, W. I., 8 Baker, J. A. (1999). Preferences for parent education
programs among low socioeconomic status, culturally diverse parents.
Psychology in the Schools, 36(3), 239-247.

 

285