A
0.1.

.n
.2 3:3.

3'
1!. '
5:5...d

frauzv: 4‘ z .
tomb”; - dll.o..!.

. I
1.2;.
23.53

i :2}...
.53.! I. ,
ii). . ‘ . . [fl ..
I. 21.33“}.
..>.L.5n?u

I:
231. .x
it sun}?
I A». 4
:1 .\ L3.

in!

.fﬂuiau

1.5!...
:1 iii!
E;x.‘3:.l

.549.
.3...» 2

 

 

:7 .
£935. gm weak tn... ”0.11~n.l.| , , . .1: .11. . . , . V ‘ 9.3%. Jun”, 1: .n... .17. l r. V

 

 

THESlS

I131":
\,,0.

lililllilll ;

This is to certify that the

thesis entitled

Maternal and Infant Temperament:
Goodness of Fit

presented by

Mona Ibrahim

has been accepted towards fulﬁllment
of the requirements for

M.A. Psychology

degree in

 

 

 

@WM i

r
'-/ ﬂ Major professor

Date 9’1/'75

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

     

 

LIBRARY
Michigan State
University

 

 

 

PLACE Ii RETURN Boanmwombchockomm ywnocord.
TO AVOID FINES Mum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU isAn Niimdivo WOMEN Oppommhy Intuition
mm:

MATERNAL AND INFANT TEMPERAMENT: GOODNESS OF FIT

By

Mona Ibrahim

A THESIS

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

MASTER OF ARTS

Department of Psychology

1995

ABSTRACT

MATERNAL AND INFANT TENIPERAMENT:
GOODNESS OF FIT

By

Mona Ibrahim

This study examined the mechanisms of the "Goodness-of-Fit" model of temperament by
investigating the relationship between maternal temperament and maternal expectations
and the relationship between the infant's lack of ﬁt with each and maternal ﬁsnctioning. The
study also appraised the differential effects of the direction of the misﬁt on maternal func-
tioning. Participants were 60 ﬁrst-time mothers and their infants. Results indicated that,
overall, after delivery, maternal expectations did not match maternal temperament (for
example, r=‘ . 11 for Quality of Mood), and that a discrepancy between maternal tempera-
ment and infant temperament was more related to maternal functioning than a discrepancy
between maternal expectations and infant temperament (for example, r=.27 vs. r= .04 for a
discrepancy in Quality of Mood and Depression). Furthurmore, taking the direction of the
discrepancy into account when computing correlations between ﬁt scores and maternal

functioning did not improve the chances of finding signiﬁcant relationships.

Dedicated with love to my children
Ashraf Kamel
and
Ayah Kamel
who ﬁll my life with joy

and drive me to become a better person

ACKNOWLEDGEMENTS

I would like to express my love and gratitude for my husband, Ahmed Kamel, for sharing
not only his knowledge of computer software, but also his love and friendship. He bright-
ens my world each day, and I could not have gone this far without his constant support and

encouragement.

I gratefully acknowledge my Chairperson and advisor, Jackie Lerner, for sharing her
vast knowledge on Temperament. Her helpful comments and ideas provided a clear direc-
tion for my research and her encouragement and support were instrumental in the comple-

tion of this study.

I would like to deeply thank me committee member, Ralph Levine, for the countless
hours of meeting, valuable feedback, and a willingness to teach me new statistical soﬁ-
ware. His expert advice contributed signiﬁcantly to the scientiﬁc presentation of this

study’s results.

I also wish to thank my committee member, Ellen Strommen, for her thoughtful and
constructive criticism, which greatly enhanced the quality of this work. Her warm personal

support, even when she was on sabbatical, is much appreciated.

I sincerely thank AJ. Alej ano and Sandra Frassetto, for generously sharing their data set
with me. Although she is not aware of it, AJ. Alejano indirectly taught me a lot about

research and “showed me the rcpes” of thesis writing.

Lastly, I would like to thank and acknowledge my parents for their support in so many

ways and for always instilling in me the belief that I could achieve my goals.

iv

Table of Contents

 

List of Tables ...................................................................................................................... vii
List of Figures ..................................................................................................................... ix
Introduction .......................................................................................................................... 1
Problem Statement ................................................................................................... l
Temperament ............................................................................................................ 2
Deﬁnition of Temperament .................................................................................................. 2

Importance of Temperament ................................................................................................ 3

Structure of Temperament .................................................................................................... 5

Categories of Temperament ................................................................................................. 7

Contextual Demands Regarding Temperament ................................................................... 7

Theoretical Orientation ............................................................................................ 8
Developmental Contextual Perspective ............................................................................... 8

The Goodness-of-Fit Model...... - _ ................................................................ 9

The Goodness-of-Fit Model in Temperament Research ........................................ 10

The Research of Thomas and Chess .................................................................................. 11

The Research of Super and Harkness ................................................................................ 13

Research in the Lerner and Lerner Laboratory .................................................................. 14

Additional Research ........................................................................................................... 23

Focus of the present Study ..................................................................................... 24
Method ............................................................................................................................... 29
Sample ................................................................................................................... 29

Subjects .............................................................................................................................. 29

Recruitment ........................................................................................................................ 29

Procedure ............................................................................................................... 30
Design .................................................................................................................... 30
Measures ................................................................................................................ 32
Mother’s Behavioral Demands (Temperament) ................................................................. 33

Mother's Expectations] Demands ...................................................................................... 34

Infant Temperament ........................................................................................................... 36

Discrepancy Scores ............................................................................................................ 37

Maternal Mental Health Functioning: Depression ............................................................. 38

Maternal Mental Health Functioning: Anxiety .................................................................. 39

Results ................................................................................................................................ 40
Data Assessment .................................................................................................... 4O
Descriptive Statistics .............................................................................................. 41

Variable Intercorrelations ....................................................................................... 43

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Internal Consistency Reliability of Measures ........................................................ 50
Reliability of DOTS-R _ _ «0
Reliability of the DOTS-R:Ethnotheory .................................................... - 52
Reliability of the IBQ - -- - - - 52
Reliability of the Two Outcome Measures - - - '52

Reliability of Discrepancy Between Expectations and Behavior .......................... 52

Stability Analysis ................................................................................................... 5 3
Stability of Mother's Behavioral Demands (DOTS-R) _ 58
Stability of Depression Scores (CBS-D)- - . . - - - 58
Stability of Anxiety Scores (STAI)- - - u. 58

Change Analysis .................................................................................................... 5 8
Analysis of Change in Maternal Behavioral Demands ............. -- 70
Analysis of Change in Maternal Expectational Demands - - -74
Analysis of Change in Maternal Functioning -- 77

Relationship Between Behavioral and Expectational Demands ............................ 80

Absolute Difference Scores and Maternal Functioning ......................................... 82
Correlations Between Discrepancya and Maternal Functioning 86
Correlations Between Discrepaneyb and Matemai Functioning - - - _ 87

Diﬁemnce Scores and Maternal Functioning ........................................................ 89
Correlations Between Discrepancyc and Maternal Functioning 89
Correlations Between Discrepancyd and Maternal Functioning - _ 93

Limitations of the study ......................................................................................... 94

Discussion .......................................................................................................................... 97

Relationship Between Behavioral and Expectational Demands .......................... 103

Absolute Discrepancy Scores and Maternal Functioning .................................... 104

Direction of Discrepancy and Maternal Functioning ........................................... 105

Stability of Maternal Temperament and Maternal Functioning ........................... 106

Change in Maternal Variables After Delivery ...................................................... 106

Future Research Directions....L ............................................................................. 108

Appendix: Instruments Used In The Present Study .......................................................... 110
References ........................................................................................................................ 121

vi

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.
TABLE 19.
TABLE 20.
TABLE 21.

TABLE 22.

TABLE 23.

TABLE 24.

TABLE 25.

TABLE 26.

TABLE 27.

TABLE 28.

List of Tables

Research on Importance of Temperament ................................................... 4
The Goodness of Fit Model in Temperament Research ............................. 19
Questionnaires Used and Point(s) in Time They were Given .................... 32
Retained DOTS-1L DOTS-RzEthnotheory, and IBQ Dimensions ............. 37
Descriptives for Scales ............................................................................... 42
Intercorrelations Among All Variables Used in the Study ........................ 44
Reliability Estimates for Scales ................................................................. 51
Post-Delivery Reliability Estimates for Discrepancy Scores .................... 54
Summary of Stability Analyses ................................................................. 57
Observed Mean Changes in DOTS-R Scores ............................................ 60
Observed Mean Changes in DOTS-R:Ethnotheory Scores ....................... 61
Observed Mean Changes in Matemai Functioning Scores ........................ 62
Analysis of Change in DOTS-R From Time 2 to Time 3 .......................... 67
Analysis of Change in DOTS-RzEthnotheory From Time 2 to Time 3 ..... 68
Analysis of Change in Matemal Functioning From Time 2 to Time 3 ...... 69
Correlations Between Demographics and DOTS-R Change Scores ......... 73
Correlations Between Demographics and DOTS-RzEthnotheory Change
Scores ......................................................................................................... 76
Correlations Between Demographics and CBS-D and STAI Scores ......... 79
Correlations Between DOTS-R And DOTS-R Ethnotheory at Time 2 .....80
Correlations Between DOTS-R And DOTS-R Ethnotheory at Time 3 ..... 81
Correlations of Absolute Discrepancy Between DOTS-R and IBQ with
Depression .................................................................................................. 83
Correlations of Absolute Discrepancy Between DOTS-R and IBQ with
State Anxiety .............................................................................................. 83
Correlations of Absolute Discrepancy Between DOTS-R and IBQ with
Trait Anxiety .............................................................................................. 84
Correlations of Absolute Discrepancy Between DOTS-RzEthnotheory and
IBQ with Depression .................................................................................. 84
Correlations of Absolute Discrepancy Between DOTS-R2Ethnotheory and
IBQ with State Anxiety .............................................................................. 85
Correlations of Absolute Discrepancy Between DOTS-RzEthnotheory and
IBQ with Trait Anxiety .............................................................................. 85
Correlations of Discrepancy Between DOTS-R and IBQ with

Depression .................................................................................................. 90
Correlations of Discrepancy Between DOTS-R and IBQ with State
Anxiety ....................................................................................................... 90

vii

TABLE 29.

TABLE 30.

TABLE 31.

TABLE 32.

TABLE 33.

Correlations of Discrepancy Between DOTS-R and IBQ with Trait

Anxiety ....................................................................................................... 91
Correlations of Discrepancy Between DOTS-R:Ethnotheory and IBQ with
Depression .................................................................................................. 91
Correlations of Discrepancy Between DOTS-RzEthnotheory and IBQ with
State Anxiety .............................................................................................. 92
Correlations of Discrepancy Between DOTS-RzEthnotheory and IBQ with
Trait Anxiety .............................................................................................. 92
Summary of the Findings of the Present Study ......................................... 98

viii

List of Figures

Figure 1: Generic Model Underlying All Stability Analyses .................................... 56

ix

Chapter 1
Introduction

Problem Statement

Children possess characteristics of individuality which—in the context of person-
context bidirectional relations—allows them to be agents in their own development
(Lerner & Lerner, 1983). Results of many studies (Carey & Mcdevitt, 1989; Chess &
Thomas, 1986; Kohnstamm et al., 1989; Strelau, 1983) point to the importance of
temperament as a key characteristic of individuality that contributes essentially to the
efﬁciency and adequacy of human behavior in everyday life. Interindividual differences in
temperament have been found to act as important moderators of children’s success at
coping with the stressors and demands encountered in the key settings of life: the family,

the school, and the peer group.

The functional signiﬁcance of a child’s temperamental attributes in the family setting is
dependent upon the demands and expectations of the parents. The concept of the “goodness
of ﬁt” proposed by Stella Chess and Alexander Thomas examines the relationship between
the individual child's temperament and the demands and expectations of the parents. If a
temperament attribute is congruent with the demands of the parents (i.e. there is a ‘good’
ﬁt), it should produce a positive adjustment. Ifthe same attribute is incongruent with such

demands (i.e. there is a ‘poor’ ﬁt), a negative adjustment would be expected.

This study attempted to take a closer look at the mechanics of the goodness of ﬁt model
by assessing the relationship between the behavioral style of the mother and the
expectations and demands she has with regards to her child’s behavior. In addition to
investigating whether the mother’s expectationai demands ﬁt/misﬁt her own behavioral
style, the study also investigated how children actually ﬁt/misﬁt maternal demands and

expectations. Gaining a better understanding of the dynamics of the inﬂuence of

l

2
temperamental characteristics on the child’s life course would facilitate identiﬁcation of

high-risk patterns of parent-child interaction.

Temperament

Definition of Temperament

The scientiﬁc study of temperamental attributes began with Gesell’s (1937) analysis of
ﬁlm records of children to assess characteristics such as Activity Level and adaptability. He
concluded that “certain fundamental traits of individuality, whatever their origin, exist
early, persist late and assert themselves under varying environmental conditions.”
Nevertheless, one of the most controversial problems regarding temperament studies today
remains the notion of “temperament” itself. Some researchers regard temperament as a
synonym for personality (Eysenck, 1969), while others conﬁne temperament to the

emotional characteristics of behavior (Goldsmith, 1987).

The most widely accepted deﬁnition of temperament is that of Thomas and Chess, who
provided an important stimulus to research on temperament through the New York
Longitudinal Study which began in 1956 and continues into the present day. They propose
that temperament refers to how an individual does things or how he or she responds to
people and to situations, rather than to what the individual does (i.e. the content of
behavior), or to why he or she does it (i.e. motivation), or to the behavioral capacities or
abilities that he or she manifests (Thomas & Chess, 1977). For example, since all children
eat and sleep, focus on these contents of the behavioral repertoire would not readily
differentiate among them. However, children may differ in the Rhythmicity of their eating
or sleeping behaviors and in the Activity Level and Quality of Mood associated with these

behaviors.

The question “how” refers mainly to formal characteristics of behavior, such as

reactivity, activity, or self-regulation. According to Thomas and Chess (1977), Rothbart

3
(1981), and Strelau (1987), these temperament characteristics are present since early

childhood and are relatively stable throughout life.

Importance of Temperament

Many studies have tested the utility of temperament by examining its power as a key
characteristic of individuality that contributes essentially to the efﬁciency and adequacy of
human behavior in everyday life. The relevant evidence falls under two main categories.
First, it has been shown that individual differences in temperament are linked to infant and
child psychological health (Chess and Thomas, 1982), resilience to stress (Werner and
Smith, 1982), classroom behavior (Pullis and Cadwell, 1982), and academic achievement
(I. Lerner, 1983).

Second, several naturalistic and experimental studies have shown that the behavioral
characteristics of children have an important effect in determining how other people
respond to them. Children with different temperamental features elicit different behaviors
from those with whom they interact. For example, easily adaptable children tend to be
protected from parental criticism even in a stressful home environment. Similarly, it has
been found that highly active babies are less likely to show developmental retardation in a
depriving institutional environment, probably because they elicit more caretaking (Rutter,
1977). Dunn and Kendrick (1980) found, in addition, that children’s Quality of Mood in
infancy was linked to maternal responsiveness to them. The higher the negativity of the
Mood, the less helpful and attentive the mother.

The temperamental qualities brought by the child to the interactions and situations that
he or she encounters therefore play an important part in determining how that encounter
proceeds and in determining the overall quality of parent-child relations (Crockenberg,

1981). Table 1 below summarizes the above mentioned research.

 

 

 

 

 

 

 

 

 

 

 

 

 

omrn— own 3 Susan 8:88.63
anus—«OE Us ‘00: 3 858080 g
.8233 cows—Eco .
.omrumﬁnoo—m v5 merge 332—8 v5 :23:
ac assesses 5.3 a. use. 3:85
as. «sues £32 .38.. 0......» ”8...; use a 3955.35 38..
5.3 2 can a 2.5..
.3 8.3.8... 8&2. 835.. an- ace mass. 3...... 88:3. 55:85». 5.5 a
3.5.5.8 ac 8.3.050 8.2.. a . n- -an .8 5220 2.3384 2.5 2.5. 8.53
8. nee. : 2.. :1... am. 220 5.3
haze-=52: N as seesaw. La... 9.8.5.... 68:
a use: 03-2 8:2,...
aces. a. sec... 5...; a. 8.25. a- .35 cases 35...»... was...» c.823— a
as: use. ac a 5.8: 8.2.. .N- "3.3533 .2393 5.85 as: use. e8:- :55
SV-noova .23 . mg
2-5 Sat omee 832» m ﬂue 358%
«5:39.35 d “cos-5:8 .802 6:230 0:5 E .368 308 £60.38
:83? scene... .38 .323 2.2. 25 83%... .85.. 828
058.8 e". .n. s ..
see ems: 392.8... 8.28. .n- s? 52 so... a 8.8. 33.8 e as. e .52- 53:
mo «8 .23 32.68% am BEE we. Bevan-5.3m 96..» 0:0 recon £2.23. Eve-.8380...- 8E3
seam 2.5-8. 3.0 0.3....
33 sea: 2 9.8.5 26-.
aw...- owmﬁ“ 82352.5: .323 $8 .22.... ”Sega... .85.. ”garage 38:
Mamas 83.8. 3..- 53 £55..0 18.83 was: s .83» Essa—=2 .1320 e
commemom ecu 8—2: 2. 7 .038 Seuss—m .390 .«o amaze. 3:03.". «co—cam- «=3.—
oam 30cm 35 ole—am Binge-eo—Euem 29¢. beam 03......) 2.33.5 038.5» 38:8; 8:95—3—

 

anauonﬁoh he cues-8.....— ee 5.308% .— Ham—«F

 

Structure of Temperament

Chess and Thomas (1977) identiﬁed nine components of temperament:

1.

Activity Level — refers to descriptions of the quality of the child’s motor

behavior.

Rhythmicity ~— refers to the regularity of biologic functions, such as sleep-
wake cycles.

. Approach-Withdrawal — refers to a positive/negative response to a new sit-

uation, person, or environmental demand.

. Adaptability — refers to the ease or difficulty of adaptability to the require-

ment for change in an established behavior pattern.

Threshold of Responsiveness — refers to the amount of stimulation it takes

to evoke a behavior.

Quality of Mood — rated as the preponderance of positive versus negative

mood expression.

Intensity of Reaction — refers to the intensity of mood expression, irrespec-

tive of whether it is positive or negative.

Distractibility — refers to the ease or difﬁculty of distractibility of an ongo-

ing activity by an extraneous stimulus.

Attention Span/Persistence -— refers to the length of attention span and the

degree of persistence with a difﬁcult task.

6
Based originally on the NYLS, Lerner and Mndle (1986) also describe temperament by

means of nine orthogonal dimensions:

1. Activity Level—General — refers to the level of motor behavior displayed in
daily functioning.

2. Activity Level-Sleep — refers to the level of motor behavior displayed

while asleep.

3. Approach-Withdrawal — refers to the tendency to move toward or away

from stimuli.
4. F lexibility-Rigidity — refers to the ease of adjustment to situations.
5. Quality of Mood — refers to the positive or negative valence of affect.
6. Rhythmicity-Sleep — refers to the cyclicality or regularity of sleep patterns.
7. Rhythmicity-Eating — refers to the regularity or predictability of eating.

8. Rhythmicity-Daily Habits — refers to the regularity or predictability of

daily routines.

9. Task Orientation — refers to the level of persistence on, attention to, and

lack of distractibility from tasks.

Among these nine dimensions, only four—Activity Level-General, Approach-
Withdrawal, Quality of Mood, and Task Orientation—are to some extent identical to
Thomas and Chess’s (1977). The temperament traits, as proposed by Windle and Lerner,
tend to describe behavior characteristics in a very speciﬁc way. This is expressed in the
distinction of three kinds of Rhythmicity: Rhythmicity-Sleep, Rhythmicity-Eating, and
Rhythmicity-Daily Habits.

Categories of Temperament

Thomas and Chess (1977) have identiﬁed, both clinically and by factor analysis, three

temperamental constellations:

1. Easy temperament: Comprises a combination of regularity, positive
approach responses to new stimuli, quick adaptability to change, and a mod-
erately intense positive mood. Children in this group are easy to manage;

hence the term easy temperament.

2. Diﬁicult temperament: Comprises irregularity in biological functions, neg-
ative responses to new situations or people, slow adaptability to change, and
intense mood that is predominantly negative. Parents and teachers often ﬁnd

such children difﬁcult to manage; hence the term diﬁicult temperament.

3. Slow-to-Warm-Up temperament: Comprises negative responses of mild

intensity to novel situations, with slow adaptability aﬁer repeated contact.

It should be noted that not all individuals ﬁt neatly into one of these three temperamental
patterns, because of the varying and different combinations of temperamental traits that are

possible.

Contextual Demands Regarding Temperament

Given the existence of easy and difﬁcult temperament attributes, the question becomes:
What gives a given temperament attribute its particular meaning? Super and Harkness
(1981) point out that the child’s context is structured by three kinds of inﬂuences: The
physical and social setting; the dominant customs in the culture; and the “psychology” of
the caregivers. This psychology is termed an “ethnotheory”. The term refers to caregivers’
preferences and expectations regarding the meaning or signiﬁcance of particular behaviors.

Super and Harkness point out that not all people have the same preferences regarding

8
temperament because every cultural or subcultural group holds different attitudes, values,

and beliefs.

These psychological differences in the meaning of temperament produce diﬁ‘erences in
what is regarded as a wanted or an unwanted attribute. In other words, because speciﬁc
cultural, subpopulation, or ecological groups may differ in how much they want particular
attributes, they may also differ in their ethnotheories (i.e. their attitudes and expectations)
regarding the difﬁculty the possession of a particular temperamental attribute presents for

interaction.

Theoretical Orientation

This study draws heavily on the Goodness of Fit Model developed by Thomas and Chess
(1977), and on the Developmental Contextual Perspective articulated by Lerner and Lerner

(1983). Following is a brief discussion of each of these frames of reference.

Developmental Contextual Perspective

A major theoretical question in the temperament literature is: What variables explain the
process by which temperament is linked to other inter- and intra-individual variables?
From the contextual perspective, temperament has meaning for the person only as a
consequence of the impact it has on the context In order to predict when and how certain
temperamental attributes relate to speciﬁc aspects of psychological functioning, we need to
look at the relationship between person and context (Lerner & Lerner, 1983).

This perspective involves the idea that development occurs through reciprocal relations,
or “dynamic interactions” between organisms and their contexts (Lerner, 1978). A notion
of integrated or “fused” levels of organization is used to account for these dynamic
interactions. Variables from levels of analysis ranging from the inner-biological, through

the psychological, to the sociocultural, all change interdependently across time so that

9
variables ﬁ'om one level are both products and producers of variables from the other

integrated levels (Lerner, 1982). In other words, children are embedded in their families.
There are bidirectional relations between individual development and contextual change.
Child characteristics promote differential reactions from parents, which may feed back to
children and provide a basis for their further development. Schneirla (1957) termed these
relations “circular functions”. Thus, in the context of these person-environment
bidirectional relations, children’s characteristics of individuality allow them to be agents

in, or producers of, their own development (I. Lerner & R Lerner, 1983).

Just as the child brings temperamental characteristics to the parent-child relationship,
the parents bring their own temperamental characteristics and their own expectations to the
parent-child relationship. It is these parental expectations and demands that provide the
functional signiﬁcance for a given temperamental attribute possessed by a child (Thomas
& Chess, 1977).

The Goodness-of-Fit Model

The “goodness of ﬁt” model proposed by Stella Chess and Alexander Thomas (1977) is
a conceptual model of the functional signiﬁcance of temperament for an individual’s
psychological development. It examines the relationship between the individual child's
temperament and the demands and expectations of the parents. According to this model,
there is a “good” ﬁt when an individual’s behavioral style enables him/her to cope
successfully with the demands and expectations of the environment. On the other hand, a
“mismatch” between an individual’s temperament and the demands of the environment
results in a “poor” ﬁt, which leads to unfavorable developmental outcomes. ‘lhus, a child’s
temperamental trait can only have adverse effects on the child’s development if it is

combined with a poomess of ﬁt.

10
The goodness of ﬁt does not necessarily depend upon a similarity of temperament

between mother and infant. Thomas and Chess analyzed the interactions between parents
and their children in their New York Longitudinal Study and found different kinds of
combinations. In some cases, similarity of temperament promoted a goodness of ﬁt; in
others it led to a poomess of ﬁt And the same variation of ﬁt was true of parents and

children of opposite temperaments.

A clear example of goodness versus poomess of ﬁt has been provided by the ﬁndings in
the New York Longitudinal Study sample (NYLS), which is a primarily upper middle-
class, as contrasted to the ﬁndings in a Puerto Rican working-class sample (PRWC) living
in the congested and underprivileged East Harlem section of New York City. Half of the
PRWC children with behavior disorders under 9 years of age were high-activity children,
whereas only one NYLS child displayed these symptoms.

The PRWC families usually had a relatively large number of children and lived in small
apartments with little space for constructive motor activity that highly active children
required. In addition, safe playgrounds and recreational areas were not available in the area
in which these families lived. By contrast, the NYLS families lived in spacious homes with
backyards, with safe streets and playgrounds available. The high-activity children in these

families therefore were able to exercise their need for motor activity.

The differences in the incidence of behavior disorders in the temperamentally high-
active children in these two contrasting populations was cleariy due to the nature of the
environmental restrictions and opportunities, which made for a goodness of ﬁt for the

NYLS children and a poomess of ﬁt for the PRWC children.

The Goodness-of-Fit Model in Temperament Research

Scientists (Jahoda, 1961; Hunt, 1961; Kendall, 1978; Endler, 1975; French et al., 1974;
Harrison, 1978; Smelser, 1961; Amidon and Flanders, 1961; Beach, 1960; Brophy, 1959;

1 1
Pervin, 1967; Kulka, 1979) in different domains of research have been using, and lending

considerable support for, the goodness-of-ﬁt model of adjustment. The consensus is that
the degree of congruence between various aspects of the person’s characteristics and the
person’s environment is more related to the person’s overall performance and adjustment
than either person or environment characteristics alone. However, I will review only those
studies pertinent to children and adolescents’ temperament, in order to provide support for
the idea that the inﬂuence of children’s temperamental characteristics on the behavior of
others (particularly parents and teachers) can best be understood within the framework of
the goodness-of-ﬁt model. Table 2, “The Goodness of Fit Model in Temperament

Research,” on page 19 presents a summary of the studies reviewed below.

Much of the literature supporting the use of the goodness-of-ﬁt model in studying
temperament effects is derived from the Thomas and Chess New York Longitudinal Study
(NYLS) or from independent research that has adopted their conceptualization of

temperament. The contribution of the NYLS is considered ﬁrst.

The Research of Thomas and Chess

Thomas and Chess’s NYLS-core-sample is composed of 133 white, middle-class children
of professional parents. Another smaller sample is composed of 98 New York City Puerto
Rican children of working-class parents. Each sample subject was studied from at least the
ﬁrst month of life onward. Although the distribution of temperamental attributes in the two
samples was not different, the import of the attributes for psychosocial adjustment was

quite different.

For example, the Puerto Rican parents were very permissive in regard to Rhythmicity of
Sleep. They allowed the child to go to sleep any time the child desired, and permitted the
child to awaken any time as well. Because parents placed no demands in regard to

Rhythmicity of Sleep, there were no problems of ﬁt associated with an arrhythmic infant

12
or child. In this sample, arrhythmicity remained continuous and independent of adaptive

implications for the child throughout the ﬁrst ﬁve years of life (Thomas et a1, 1974).

On the other hand, strong demands for rhythmic sleep patterns were placed by white,
middle-class families on their children. Thus, an arrhythmic child did not ﬁt with parental
demands, and consistent with the goodness-of-ﬁt model, arrhythmicity was a major
predictor of problem behaviors during infancy and through the ﬁrst ﬁve years of age
(Thomas et al., 1974). Parents in this middle-class sample took steps to change their
arrhythmic children’s sleep patterns, so that low Rhytlunicity tended to be discontinuous

for most children.

Interestingly, arrhythmicity did begin to predict adjustment problems for the Puerto
Rican children when they entered the school system. Their lack of a regular sleep pattern
impaired their concentration in class, and caused them to be late for school (Thomas et al.,
1974). Thus, it is important to consider ﬁt with the demands of the particular context of

development.

With regard to differences in the demands of the physical contexts of the families, high
Activity Level was highly associated with problem behavior in the Puerto Rican children
and not in the middle-class children because of differences between the physical features
of the respective groups’ homes. In the Puerto Rican sample, the families usually had
several children and lived in small apartments. In addition, parents were reluctant to let
their children out of the apartments because of the dangers involved in playing on the
streets of East Harlem. As a consequence, even average motor activity tended to cause

problems for interaction in the Puerto Rican group.

In the core sample, however, the parents had the ﬁnancial resources to provide large
apartments or houses for their families. There were typically suitable play areas for the
children around the neighborhood. Consequently, high Activity Level did not constitute a

behavioral problem in the core sample children.

13
In sum, the NYLS data set provides results compatible with the goodness-of-ﬁt model.

Data independent of the NYLS also lend some support, and will be reviewed next.

The Research of Super and Harkness

Support for the goodness-of-ﬁt model is provided in a cross-cultural study by Super and
Harkness (1981). They studied infants in a rural farming community in Kokwet, Kenya
named Kipsigis and infants of suburban families living in the metropolitan Boston area. In
all of their Boston families, there was only one person at home with the baby during most
of the day. The infant’s activities are therefore oﬁen temporally scheduled in order to meet
the needs of both the mother and the infant. In contrast, the Kipsigis mother is rarely alone
with the infant. During the day, an average of ﬁve additional people are in the house.
Within this situation, the infant is free to sleep and nurse at will because whenever the

mother needs a break, the infant is attended to by another nearby caregiver.

Super and Harkness (1981) found that an infant who does not show Rhythmicity of
sleeping and eating does not have a problem ﬁtting the cultural demands imposed by the
Kokwet setting and is not likely to evoke negative reactions in the mother. The Kipsigis
mothers did not view characteristics like negative Mood, low Rhythmicity, and low
Adaptability as indicative of long-term problems. However, an infant in the Boston setting
with the same low level of Rhythmicity did not ﬁt well with maternal demands. The Boston
mothers view low Rhythmicity, low Adaptability, and negative Mood as undesirable
characteristics that present immediate and potentially long-tenn problems. Thus, as in the
NYLS data set, the same temperamental characteristic has a different impact on others as a
consequence of its embeddedness in a different cultural context with a diﬂ‘erent set of

demands.

There is evidence for the use of the goodness-of-ﬁt model in contexts other than the

home and in regard to relationships other than that of parent-child. A key context of

l4
childhood and adolescence is the school, and focal relationships in such a setting are those

between children and teachers and between children and peers. A series of studies

investigating these relationships were canied out by Lerner and Lerner (1987).

Research in the Lerner and Lerner Laboratory

J. Lerner (1980) examined the role of congruence between temperamental attributes and
school demands for psychosocial adjustment in young adolescents. Junior high school
students, emailed in four eighth-grade classes of a junior high school located in a large
suburb of New York City, were assessed with regard to their temperamental attributes.
Participants were 75% white, 21% black, and 4% belonged to another racial background.
All classes were taught by the same teacher. The demands of the school-social and the
school-academic contexts in regard to the temperamental attributes were also assessed. In
addition, for each of the two contexts, both actual and perceived demands were assessed.
Finally, as indices of personal and social adjustment, measures of grade point average,
perceived academic and social competence, positive and negative peer relations, general
self-esteem, academic self-esteem, social self-esteem, and overall peer relations were

obtained for all adolescents.

Results indicated that those adolescents whose temperamental attributes were least
discrepant from the demands of the two contexts had scores on the measures used to index
adj ustrnent that were indicative of better adjustment than was the case for those adolescents

whose temperamental attributes were most discrepant from the demands.

Palermo (1982) completed a study in the Lemers’ laboratory assessing ﬁfth graders’
ratings of their own temperaments, the ﬁfth graders’ mothers’ ratings of their children’s
temperaments, and the demands for behavioral style held by the teachers and mothers of
the ﬁfth graders. Outcome measures included each teacher’s ratings of classroom ability

and adjustment, classroom peers’ sociometric appraisals of each child’s positive and

15
negative peer relations, and mother’s reports of problem behaviors shown at home. Again,

children whose temperaments provided a better ﬁt to teachers’ demands, especially with
regards to the dimension of Attention Span/Distractibility, had more favorable scores on
teacher-, peer-, and mother-derived outcome measures than did children for whom the ﬁt

was not as close.

In another study, J. Lerner (1983) measured eighth graders’ temperaments and the
demands for behavioral style in the classroom maintained by each child’s classroom
teacher and peer group. Those children whose temperaments best matched each set of
demands had more favorable teacher ratings of adjustment, better grades, more positive
peer relationships, fewer negative peer relations, and more positive self-esteems than did
children whose temperaments were less well matched with either teacher and/or peer
demands. Speciﬁcally, the eighth graders who displayed high Rhythmicity, low Intensity of
Reactions, positive Quality of Mood, high Approach behaviors, and high Adaptability

were the best adjusted on these outcome measures.

Kacerguis (1983) examined the goodness-of-ﬁt model in the home setting, focusing on
the pre- versus post-pubescent daughter-mother dyads. Studying a group of 53 pre-
pubescent daughter-mother dyads and a group of 42 post-pubescent daughter-mother
dyads, Kacerguis obtained ratings by the mothers of the level of conﬂict in the parent-child

relationship. In addition, all adolescents rated their own temperament.

Kacerguis (1983) speculated that the source of parent-child conﬂict diﬂ‘ered between
pre- versus post-pubescent daughter-mother dyads; parents of prepubescents expect
different behaviors of their children than do parents of post-pubescents. As a consequence
of these different behavioral expectations, temperament should be differentially linked to
parent-child conﬂict in the two puberty groups. In other words, certain temperamental

characteristics that “ﬁt” with mothers’ expectations of their prepubescents and therefore

16
created no mother-daughter conﬂict, should not “ﬁt” with the different set of expectations

that mothers of post-pubescents hold, and thus should result in mother-daughter conﬂict.

Kacerguis found that among the prepubescent daughter-mother dyads higher scores on
three temperamental attributes (activity, Rhythmicity, and reactivity) were signiﬁcantly
related to higher levels of mother-reported-conﬂict, and higher scores on two other
attributes (attention and adaptability) were signiﬁcantly related to lower levels of conﬂict.
The relations between temperament and parent-child conﬂict were markedly diﬂ‘erent
among the postpubescent daughter-mother dyads. First, two significant reversals in
direction of relationship occurred: higher Activity Level and Rhythmicity scores were
associated in this group with lower conﬂict scores. Second, no signiﬁcant relations
between either attention, adaptability, or reactivity and parent-child conﬂict were found,
and all three of these relations diﬂ‘ered signiﬁcantly from the corresponding ones among

the prepubescent daughter-mother dyads.

In another study completed in the Lemers’ laboratory, Hooker, \Vrndle, and Lerner
(1984) studied sixth-grade children’s temperament and its ﬁt with teacher demands.
Speciﬁcally, they assessed whether the relations between temperament and dimensions of
perceived competence, as measured by Harter’s (1982) Perceived Competence Scale,
could be accounted for by the goodness of ﬁt model. Use of this scale was predicated on
the view that children who have positive interactions with the signiﬁcant others in their

context should come to perceive themselves as competent individuals.

Hooker et al. (1984) expected that temperament scores that diverged most from the
teachers’ ethnotheory of temperamental difficulty should be associated with positive
scores on the four components of perceived competence measured by the Harter’s scale:
cognitive, social, physical, and general self-worth. For 20 of the 36 relations for which

predictions were made, signiﬁcant correlations were found in the expected direction.

17
Moreover, the multiple correlations between temperament and each of the competence

domains were signiﬁcant

These results were cross-validated in a study of late adolescent college students
conducted by \Vrndle et a1. (1986). The ethnotheories of both parents (mostly mothers) and
of peers were studied. For 17 of the 28 relations for which predictions were made,
signiﬁcant correlations were found. All of these were in the expected direction. High
general and sleep Activity Level and low Rhythmicity, Flexibility, Approach, and attention
span, and negative Mood constitute temperament levels that are rated to make for difﬁculty
in interaction and varied negatively with competence scores and positively with CES-D
scores. The CES-D was used by Windle et al. (1986) because of the ﬁndings that failure to
meet the demands of one’s context may be associated with feelings of depression

(Seligrnan, 1975).

Further support for the goodness-of-ﬁt model is provided by a study by J. Lerner, R.
Lerner, and Zabski (1985) which looked at the ﬁt between fourth-grade students’ self-rated
temperament and teachers’ demands. Although ﬁt scores related to Adaptability/
Approach-Withdrawal were not related to any outcome measure, ﬁt scores related to
Attention Span/Distractibility were related to teacher ratings of adjustment (with students
meeting or exceeding teachers’ demands having better teacher-ratings). Students who met
or exceeded the teachers’ demands had adjustment ratings which were higher (mean= 4.0)

than those for students who showed less attention than demanded (mean= 3.3).

Moreover, ﬁt scores related to Reactivity were related to both teacher-rated and actual
abilities (with students who met teachers’ demands for Reactivity or who showed even less
reactivity having better outcomes). Speciﬁcally, in comparison to children who exceeded
the level of Reactivity expected by their teachers, children who met or fell below these
demands were rated as more able and more adjusted, and in addition achieved better scores

on the Stanford Achievement Test for Reading (SAT -Reading) and the Comprehensive

18
Test of Basic Skills (CTBS). The means for the ability rating, the adjustment rating, the

CTBS score and the SAT -Reading score for the group that met or fell below the demands
were 4.4, 4.4, 108.8, and 6.2 respectively; the corresponding means for the poorly ﬁtted
group were 3.8, 3.3, 103.5, and 5.2 respectively.

R. Lerner et al. (1986) summarized two studies that looked at temperament and its ﬁt
with contextual demands among early and late adolescents. In these studies, positive
correlations were expected between cognitive, social, and general perceived competence,
on the one hand, and Rhythmicity, Flexibility, Quality of Mood, Attention Span, and
Approach, on the other hand. Conversely, negative correlations between these competence
domains and the two activity attributes were expected. For 14 of the 21 relationships for
which predictions were made signiﬁcant correlations were found in the expected directions
among the early adolescents group. These results were essentially cross-validated within

the late-adolescent sample.

In another study, Talwar, Nitz, and Lerner (1988) found that poor ﬁt with parental
demands was associated with low teacher ratings of scholastic competence and higher
parent ratings of conduct problems. Corresponding relations were found in regard to ﬁt
with peer demands. Moreover, goodness-of-ﬁt scores between temperament and demands
were more often associated with adj ustrnent than were temperament scores alone. This was
true for the parent context at the end of the sixth grade (27% signiﬁcant correlations
between adjustment measures and ﬁt scores vs. 13% signiﬁcant correlations between
adjustment measures and temperament alone), and for the peer context at the beginning of
seventh grade (28% signiﬁcant correlations between adj ustrnent measures and ﬁt scores vs.

11% signiﬁcant correlations between adjustment measures and temperament alone).

In sum, the studies summarized above point to the importance of considering the
demands of the signiﬁcant others in the child’s context in order to understand the ﬁrnctional

signiﬁcance of temperamental individuality for adjustment.

l9

 

2.32:»... {a 2.2“
N. "”838:— Eunice

 

 

 

 

 

 

 

 

 

 

 

e 8.8... 3:323... its. as: 3.2 ages .3-
2838a u omen}? >2 we £33 umbe— u 082 E «cows—o.— bom-
Mﬁﬂwwuseﬁw :52 3:33... .828 «58 one 3:03:60 €583
cove : ride s...2 we. .32 823 39...... a Bee .3286..- .828- 8%..
sum. uses-«- 8.25. an- as... 2.2.5 .23 a8... 25 $6- .§eaa§...- sea.
228m
89.2 358.2 583.3
mo 85-25 v5. 933— “52082 282- 23 C
E 325568 92:52 .8330 gunmen $38232- 8053:
<2 <2 .3. a :52 as»... see seas:- eeeié- 93 sea
8.2: 2
£03502 N .505: .m Eo-c 3:23
AMA—V 2.82 830 wee—83 :82
«v a... as .3 8.... 2.8 2.2... 2- 85.... .a as be .2
“new. ME 5 £5305 now—m- 838 on 59¢ sepia 3:33.835 38. $8:
3... an. .ahﬁmm. eases. a29.2... each .3232 £022.. 3.26...- as
”been... .22.. inseam. .§23 N... H.220 eeﬁﬁessom 22.3323 5.328- 32.5532. 259:
22m .85 new 0.9.5 Eaagsoessm 25 seem 92.2..» .5825 0.2.2.5 38.3... 8:293.

 

guano-ow— .euEaaonth. E .232 x..— ue Boa—x39 2:. .N HAGSH

 

20

 

 

 

 

 

 

 

 

 

 

 

8......
non—co... 3.6... =3...
«.8... «www.mhm 8.2.... .n. u... 58:. .83
8V.— mnr .8832. .328 3.225 E...
u....§ .. .8. 8.... -8... N... .83 28...... ......... 85...... .55....-
Maﬁa”? 8.2.... a... .8... 2.3... a. 86.... .8585.
5.-.... a... .89.... ... .8832. .5883 5...... 5.- 283%... 888.8... .2...
H........................ .o... ..... mm- A8... 8...... a. 2...... 38.88.86 .28.... 3...... 9......- .898...
.o.v..-n8.v.. 5.3
o.-. 5.... .9...
2.3.3 2.25:3... 35.80..
2... $25.50 w... now. :82 2.... E 8.8.. .33.. v5
5943 cacao-c... .0... >2 .52. 2... 8.8.3.... 0.8032. .00....»
2...... 8 3.. .m. o. ..
a... .9... 58...... 8...... 8.. a... 8...... a. 8 8.8. 3...... u... 8.8.... a .3... 3......
.o .8 5.... 38.8.8 n... 8.2.... .m- o.......... 8......» .3.» 25 .8... ..........H a... 8.8... .... 8......
od— uowa :82 mﬁmEaaa
on: ".3838. 8.092 8.— ... 0.52.8.3... .28..-
..oto....-........ ...-. .2583. 28:3. 8.2.... N a. on... a. ”5.2.5.. 5030... 2.9»
.8. Eammmwwﬁm ........... 8......m a. a... 2... .o 5...... .82....- 3.82... .o.
n. «8...... 8...... S - .8... 2.8.5.8.... 2 8 8.8. .858... a. 3...... 85...... 82.8.:- .8...
8...... 8...... a. .28.... 8.2.... 3.. .......... 8.856 8....3 .3... ...o .o m......... 8.8...- .=.............. 2.5. .58....
S...- ...E. .5 .35.... .........u......_.......m 25 33m 2......5 .5825 2......» 38...... 8.8.8....

 

 

3.23% N club.

21

 

8V:
Km... "so... omega

u.::>.h:§n. .838

 

 

 

 

 

 

 

 

 

 

 

Nov: K...” "83...... :. b.5888 «553......
”3.288: 8.... 38.58 a 0:... :. :5 DEA: 35:8:
_ Swwmawmammww 5 8. no? use. .58 25 .o 3.5.: 223. $8:
H 3:53... 8.2: 3. 3:83» 83m 5: .u :28. 8:08...- «Egg ._a .0
>39. 5.28.: Ba- 8.25.. 3- 2888.... .223 9.2m 0:0 8.68 «8...- 3.8....om- .3qu
352.650..
. So: .8 new.
3 v: 8
28:3: .3. n .5 :85 3:88.02
.8. v: s was». $3 as. :5
ugmﬂﬂahwww «3.5.8:: ova. AN. now: :88. 8:82.88 33..
E8»..:m.m $3. $380.. $53 3:88.03 3.80 18.83 .8388... 20:88.8. ._u «o
5.3 .m. n: omEe>< 389% so .2. .5388... 3...? .820 Dim—U- E8833. 2.8.3
35:8:
one» so 8:25. .32.
2.. :. 8:82.88 «8:38:88 ..¢ .0
i <7. <2 ! 5.2.5 <2 .8388“. 220. 188.8:
85 80:”. 8% 9.95m 8.329.535 2.»... maﬁa 033.5» 2:825 2.3.5, 38...»:— 8:95.3—

 

528. N .35..

 

22

2.3.3. 82 8.008.... "S. 0.02

 

 

 

3... "08.0 5.0.:
00.. "000.» 2.350..
”08030....
00:000..-.5000 :82
8.. "08.0 5-..»...
ma "0:05 5-30.
8:800:50
00.0.8008. :8:

00.m 80.5—

 

.8.0... 0...
8.050.. 0a.

  

03m 0......0m

 

20.53.00 3:00.00

 

 

0.... 00.52-

080.00... 00:088.

0.5.2.958. 3032.00 ..0 .00...

£0... 3:88.000 3:00.. .2000... 00:080..
.03 .30 20...... .5003... .80... 0......
.830. 0. 0.0.0.... .3000 .00..- 2.080.880. ..0 .0
05.88%. .223 0502.08. .8500... .85.- .5320; as...
82.0050500308 0...... .305 030...!» 0508.5 030...¢> 00.0.10...— 02.0.0.0:

 

 

 

 

 

.0008. 0 0......

 

23
Additional Research
Additional research, although not directly testing the goodness-of-ﬁt model, lends support
to the idea that the temperament of the child inﬂuences maternal behaviors. Much of this
research is limited by methodological problems because of nonindependence of measures;

nonetheless, the associations reported are worthy of note.

First, temperamental difficulty (i.e., high intensity of response, irregularity in biological
functions, negative Mood, low adaptability, and withdrawal) has been associated with less
responsive caregiving or less stimulating contact from mothers (Crockenberg & Acredolo,
1983; Dunn & Kendrick, 1980; Kelly, 1976; Klein, 1984; Linn & Horowitz, 1983;
Milliones, 1978; Peters-Martin & Wachs, 1984).

Some studies that do employ relatively independent measures of maternal behavior and
infant temperament found that mothers are less engaged with their babies if they are
difﬁcult or irritable (Crockenberg & Acredolo, 1983; Linn & Horowitz, 1983). These
studies are countered by ﬁndings of studies reporting that mothers of difﬁcult or irritable
babies are more engaged with their babies than are mothers of easy, less irritable babies
(Bates, Oson, Pettit, & Bayles, 1982; Crockenberg & Smith, 1982; Fish & Crockenberg,
1981; Klein, 1984; Peters-Martin & Wachs, 1984; Pettit & Bates, 1984).

Crockenberg (1986) noted that several factors may be responsible for the inconsistent
ﬁndings, for example, curvilinear effects-—mothers increasing their involvement in
relation to their infants’ needs and withdrawing only if the infants are extremely difﬁcult;
the inﬂuence of some third variable such as maternal attitudes or the gender of the infant;
or an interaction between temperament and some other characteristic of the caregiver or

caregiving environment.

Generally, even with these inconsistencies, the results of the studies summarized allow

the inference that at a given point in development neither children’s attributes, nor the

24
demands of their settings per se, are the key predictors of their adaptive functioning.

Instead, the relationship between the child and the context seems most important in home,
peer, and school settings. Investigating this relationship involves looking at two
components: an attribute of child individuality, and a feature of the context that is relevant
to the domain of child individuality. In addition, a model that speciﬁes the nature and
implications of the relationship between child and context is necessary. The findings
reviewed here suggest that the goodness-of-frt model is a useful one to employ in studying

these relationship.

Focus of the present Study

The goal of this study was to add to our knowledge of the dynamics of the goodness of ﬁt
model by further documenting the extent to which maternal expectational demands are

similar to/different from maternal behavioral demands (i.e. maternal behavioral style).

It is necessary at this point to clarify why the “Goodness of Fit” literature uses the term
“maternal behavioral demands” to stand for maternal temperament. The work of Lerner
and Lerner (1987) has delineated three possible sets of demands that children may
encounter as they interact with their contexts: a) expectational demands, b) behavioral
demands, and c) demands of the physical setting. These sets of demands are “possible”
because they are part of the child’s environment. For any given child one or more of these
sets of demands may not be relevant. So far, only attitudinal demands have been
investigated in the temperament literature (e.g. Lerner, 1980; Palermo, 1982; Kacerguis,
1983; Hooker et al., 1984). These expectational demands were found to be relevant in the
child’s environment. Thus, our use of the term “maternal behavioral demands” is not built
on the assumption that maternal temperament does constitute an actual or relevant demand.

Rather, this study takes the position that maternal temperament characteristics are only a

25
set of “potential” demands. One of the goals of the present study is to answer the question

of whether maternal temperament is in fact an actual set of demands.

This study also examined the relationship between “ﬁt” and maternal functioning. In

particular the following research inquiries were addressed:

1. The initial purpose of this study was to assess the relationship between
maternal expectational demands and maternal behavioral demands. Do
mothers want what they themselves are? It made sense to predict that moth-
ers’ demands would not closely match their behaviors. Behavioral styles
that are acceptable for adults are not necessarily acceptable for infants.
What parents demand of their children at each age level depends on what
behaviors are deemed valuable and adaptive for the child at that age level,
and could be quite different from the parent’s own behaviors. For example,
A parent may have an irregular sleep pattern but desire for their young child

to have a rhythmic sleeping pattern.

2. A second purpose of this study was to examine the relationship between two
types of “ﬁt” and maternal functioning. The ﬁrst type of ﬁt (or lack of ﬁt) is
between infant behaviors and maternal expectational demands. The second

type of ﬁt is between infant behaviors and maternal behavioral demands.

Literature indicates that a poor infant-mother ﬁt is correlated with less
responsiveness, attention, holding, and help offered by the mother to her
infant (Kelly, 1976; Milliones, 1978; Dunn & Kendrick, 1980; Linn &
Horowitz, 1983). In addition, children’s temperament inﬂuence their par-
ents’ mental health functioning. For example, sleep arrhythmicity in chil-
dren was found to result in stress, anxiety, and anger in the parents (Thomas

& Chess 1976; Thomas et al., 1974).

26
Such an effect of child temperament on the parent’s own level of adaptation

has been reported in other data sets wherein, for instance, infants who had
high thresholds for responsiveness to social stimulation and thus were not
soothed easily by their mothers evoked intense distress reactions in their

mothers (Brazelton, Koslowski, & Main, 1974).

In another study, Wolkind and De Salis (1982), interviewed mothers at four
and fourteen months after the birth of their child. They found that at the
four-month interview, mothers with negative mood/irregular children were
signiﬁcantly more likely to report physical tiredness than other mothers.
And, at the 14-month interview, women with negative mood/irregular
babies were more likely than the remainder to have developed a mental-

health disorder (speciﬁcally depression) during the intervening ten month.

Therefore, it was expected that, in this sample, a mismatch between infant
behaviors and either set of maternal demands would lead to higher levels of

maternal depression and anxiety.

But which of the two types of ﬁt/misﬁt described above is more closely
associated with maternal ﬁrnctioning? Since no available studies have
looked at the match between the mother’s behavioral demands and the

infant’s temperament, this present study attempted to explore this issue.

. A third question addressed by this study was whether the direction of misﬁt
had an inﬂuence on maternal functioning. The concept of the Goodness of
Fit is built around the notion that no temperament attribute is good or bad in
itself. Rather, it is whether this attribute meets the contextual demands that
matters. Therefore, in general, lack of ﬁt in either direction was expected to

be highly associated with depression and anxiety for the new mother. In

27
other words, high absolute values of discrepancy scores between maternal

expectational demands and infant temperament were expected to be associ-

ated with high scores on the depression and anxiety scales.

Some literature, however, seems to indicate diﬂ‘erential outcomes based on
the direction of the misﬁt. For example, Lerner et al. (1985) found that stu-
dents who met or exceeded the teacher’s demands for attention had adjust-
ment ratings which were higher than those for students who showed less
attention than demanded. Therefore, this study explored whether the direc-
tion of the misﬁt matters for some, or all, of the temperament dimensions

investigated.

4. In addition, the degree to which maternal temperament, depression, and
anxiety were stable over time was be assessed. It was expected that these
constructs would be relatively stable before the birth of the child (Le. they
would be stable from Time 1 to Time 2, which are both pre-birth measure-
ment points), but would show instability in the pre-birth to post-birth (Time

2 vs. Trme 3) transition.

5. Finally, this study looked at the change in scores from one measurement
point to the other. These changes were analyzed closely for birth-by-subject
interactions and effects of initial level of a variable on the amount of change

in that variable.

It was predicted that the mother’s Activity Level, Approach, and Flexibility
would increase after the birth of the baby while her Task Orientation would
decrease. No directional hypothesis was proposed with regards to the moth-
er’s Mood, Depression, Anxiety and demands with respect to her infant’s
temperament. In addition, it was predicted that there would be a signiﬁcant

birth-by-subject interaction for all the maternal variables.

28
Finally, it was predicted that differences in initial level of all variables

would explain a signiﬁcant portion of the variation in the response of the
mothers to the birth. Compared to mothers who start out low on any given
dimension, mothers who start out high on that dimension were expected to
have smaller increases in the level of that dimension after the birth of their
child. On the other hand, compared to mothers who start out high on any
given dimension, mothers who start out low on that dimension were
expected to have smaller decreases in the level of that dimension after the
birth of their child.

This study used data from 60 families to assess the above mentioned relationships.
Three waves of questionnaires were employed to collect all information. A more detailed
description of the procedure used to collect the data for this investigation, the sample
examined, the measures utilized, and the design of the study is presented in the following

section.

Chapter 2
Method

Sample

This study uses data from the MSU “Becoming A Parent Study”, a short-term longitudinal
study of psychosocial changes during the transition to parenthood. The Becoming A Parent
Study involved 180 families. Of those, 60 families had completed all pre-pregnancy and

post-pregnancy sets of questionnaires. The present study uses data from these 60 families.

Subjects

The participants were 60 pregnant women, ranging in age from 19-42 years (mean= 27.9,
SD= 4.5), each expecting their ﬁrst child. The sample is 83% Caucasian, 3% African-
American, 2% Asian, 2% Hispanic, and 10% other races. The subjects are primarily
married (85%) with only 13% single and 2% divorced. The mean education level for the
sample is a college degree. 63% of the sample is employed full time, while 20% work part-
time and the remaining 17% are unemployed. This sample includes a variety of
occupations, ranging from unskilled and clerical workers to managers and executives. The
mean level of occupational prestige is 47 on a scale of 15 to 90. The average total family
income is $40,000 per year. In order to control for the possible inﬂuence of having several
children, only ﬁrst-time mothers were sampled. 55% of the women reported that their
pregnancy was planned, 23% reported it was unexpected, and 17% reported it was
expected, though not planned. The majority of the sample (75%) experienced two or fewer
birth complications during the delivery of their babies.

Recruitment

All ﬁrst time expectant mothers were volunteers, recruited through Sparrow Hospital's
Family Care Clinic, Butterworth OB Gyn Clinic, the Obstetric and Gynecology Clinic at
29

30
The Michigan State University Clinical Center, The Physician’s group, and through

Lansing and Grand Rapids area prenatal classes. Butterworth's clinic serves approximately
900 low income mothers per year, while the clinic on the campus of Michigan State
University primarily serves middle-income mothers. Women were recruited into the study
at the time for their ﬁrst prenatal visit. Each participant met the following criteria: 1.
Length of pregnancy < 24 weeks at ﬁrst prenatal visit, 2. Nulliparous, 3. No chronic

diseases, e.g. Diabetes mellitus, hypertension, cancer, etc, and 4. Singleton pregnancy.

Procedure

All women were approached by a nurse in the hospital/clinic waiting room. The nurse
brieﬂy described the research project and gave the expectant mother a letter explaining
what the project involves. Attached to the letter were consent forms, to be ﬁlled out if they
choose to participate. Upon receiving the signed consent forms, we contacted subjects by
phone and arranged for them to receive the questionnaires. Adjustments were made for the

researcher to go to the subject’s home or workplace when difﬁculties arose.

Questionnaires were mailed to the participants’ homes. Women were able to complete
the questionnaires at their convenience within a speciﬁed period of time and return the
completed packets by mail using prepaid return envelopes. Upon receipt of the completed
questionnaire packet, the women were reimbursed $5.00 for the time they spent completing

the questionnaires.

Design

Data were gathered by two different methods. The original project design involved data
collection at three points in time: upon entrance into the study at the ﬁrst prenatal care visit,
at the middle of the third trimester, and eight weeks postpartum. The group that was

administered three sets of questionnaires throughout the pregnancy were named “3-wave”

31
subjects. These subjects constitute 52% (31 out of 60) of the current sample under

investigation. On average, 3-wave women in this sample entered the study during their 20th

week of pregnancy.

Initial recruiting attempts proved difﬁcult and slow, due to the number of women under
the age of 18 who were ineligible for participation. Additional subjects were recruited from
expectant parent organizations. Since most participants do not begin classes until the last
trimester of pregnancy, an adjustment to the original design was made. The second method
of data collection then required administration of packets at two points: during the last
trimester of the pregnancy and eight weeks postpartum. This group of subjects were
labeled “2-wave” subjects. The sample includes 29 (48%) “2-wave” subjects. On average,

they entered the study during the 34‘h week of their pregnancy.

In order to conduct the current analyses, data from all waves was utilized. Data for the
subjects who received three waves of questionnaires was used in analyzing the stability of
the DOTS-R, Depression, and Anxiety scales over the three points in time. In addition, this
data was used in analyzing change scores from Time 1 (pre-delivery) to Time 2 (also pre-

delivery), from Time 1 to Trme 3 (post-delivery), and from Trme 2 to Time 3.

For the subjects who received two waves of questionnaires, data necessary to conduct
change analysis from Time 2 to Time 3 was extracted from the ﬁrst set of questionnaires
and from the second set of questionnaires, administered in the last trimester of pregnancy

and eight weeks postpartum, consecutively.

For all subjects, data from the third trimester of pregnancy (Time 2), as well as data
collected eight weeks after the birth of the babies (Time 3) was used to conduct the
remaining analyses. Data extracted from the set of questionnaires that was administered in
the last trimester of pregnancy consisted of: a) a scale assessing the mother's temperament,
b) a scale assessing her demands and expectations with regards to her infant’s

temperament, c) a scale assessing her depression level, and d) a scale assessing her anxiety

32
level. Data extracted from the set of questionnaires that was administered eight weeks

postpartum consisted of the same four questionnaires speciﬁed above in addition to a scale
assessing the infant's behavioral style or temperament. Table 3 below shows the

questionnaires used and the point(s) in time when they were obtained.

TABLE 3. Questionnaires Used and Point(s) In Time They were Given

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The 1 Time 2 'I’hneJ
Quuﬁou‘m 3W-l 3W4 2W-l 3W4 2W-2

(n-31) (n- 31) (n- 29) (n- 31) (n- 29)
Demographics Yes No Yes No No
DOTS-R Yes Yes Yes Yes Yes
DOTS-RzEthnotheory No Yes Yes Yes Yes
IBQ No No No Yes Yes
CES-D Yes Yes Yes Yes Yes
STAI Yes Yes Yes Yes Yes

Measures

The goodness-of-ﬁt model emphasizes the importance of studying contextual and
individual characteristics. In keeping with the conceptual framework of that model and
following the general procedure established in the Lemers’ studies, infant and mother
temperament were assessed, as well as the expectational demands held by the mother.
Discrepancy scores between each infant attribute and each maternal expectational demand
were calculated. Also, discrepancy scores between each infant attribute and each maternal
behavioral demand, as well as discrepancy scores between each maternal behavioral and
each maternal expectational demand were calculated. Finally, ﬁt scores were related to two

matemal-health-functioning outcome measures: Depression and anxiety.

33

Mother's Behavioral Demands (Temperament)

Each female participant rated her temperament, or behavioral style, using the Dimensions
of Temperament Survey-Revised or "DOTS-R” (VVrndle & Lerner, 1986). This 54-item
questionnaire assesses temperament along nine orthogonal dimensions: Activity Level-
General, Activity Level-Sleep, Approach-Withdrawal, Flexibility-Rigidity, Quality of
Mood, Rhythmicity-Sleep, Rhythmicity-Eating, Rhythmicity-Daily Habits, and Task
Orientation. The response format for each item is “1” = really false for me; “2” = sort of
false for me; “3” == sort of true for me; “4” = really true for me. An example of a DOTS-R
item (indexing Task Orientation) is “I can always be distracted by something else, no

matter what I may be doing.”

Scoring the DOTS-R involves forming attribute scores by summing the scores on
individual items. On the basis of the number of items per attribute the range of possible
scores for each attribute is; 7-28 for Activity Level-General; 4-16 for Activity Level-Sleep;
7-28 for Approach-Withdrawal; 5-20 for Flexibility-Rigidity; 7-28 for Quality of Mood; 6-
24 for Rhythmicity-Sleep; 5-20 for Rhythmicity-Eating; 5-20 for Rhythmicity-Daily
Habits; and 8-32 for Task Orientation. Higher DOTS-R scores indicate higher levels of
General and Sleep Activity, a tendency to Approach, higher Flexibility, a positive Mood,
higher levels of Rhythmicity in sleep, in eating, and in daily habits, and a higher Task

Orientation level.

Internal consistency coefﬁcients (Cronbach alphas) for the above nine DOTS-R
attributes are 75, .81, .77, .62, .80, .69, .75, .54, and .70 respectively for a sample of 224
sixth graders (Windle & Lerner, 1986). Construct validity for the Dots-R has been reported
by Windle et al. (1986) in a study of temperament, perceived self-competence, and
depression among early and late adolescents. Construct validity was also assessed by
Windle (1985) in an inter-inventory study among late adolescent college students. Both

convergent and discriminant validities were found between the DOTS-R attributes and the

34
traits measured by the EASI-II (Buss & Plomin, 1975) and the Eysenck Personality

Inventory (Eysenck & Eysenck, 1969).

Mother's Expectational Demands

To reiterate a point made previously, this study stems from the view that, unless there is
knowledge of the relationship between temperament and the contextual demands of a
particular setting, temperament has little meaning. Thus, useful predictions about the
relationship between infant temperament and maternal adaptive functioning cannot be
made. Therefore, having an infant with particular temperament attribute is not necessarily
in and of itself useful in predicting maternal functioning. Rather, it is more useful to know
if a particular infant attribute matches or mismatches the mothers behavioral or
expectational demands. Therefore, it is the measurement of temperament and the
measurement of contextual demands, and whether there is a goodness of ﬁt between the
attribute and the demand, that should be of prime interest in predicting maternal

functioning.

The Parent's Ethnotheory Scale for Temperament, or ”DOTS-R: Ethnotheory” (Lerner
and Lerner, 1986) was used to assess the mother's expectations and ideas about how she
wants her infant to behave. This scale relies on the concepts of “developmental niche” and
of “ethnotheory”, as formulated by Super and Harkness (1981). DOTS-R items are used to
assess the ethnotheories regarding temperamental difficulty which are maintained by the
parents of young infants. The DOTS-R is used in order to produce, for the DOTS-R:
Ethnotheory, scores for ﬁve temperamental attributes that constitute a subset of the nine

characteristics measured by the DOTS-R.

For example, an item such as ‘My child gets sleepy at different times every night” is
presented to parents who are told to consider the item in terms of how they want their child

to behave. If the item describes the way the parent wants the child to behave, then the

35
behavior would not make it difﬁcult for the parent to interact with the child, even if the

child always showed that behavior. The less the parent wants the behavior, the more
difﬁcult the interactions between the parent and the child will be, if the child always shows
this behavior. Thus, based on the degree to which the parent wants the behavior described
in the item, he or she rates the item in regard to the level of difﬁculty for interaction which
would be associated with the behavior. This performance-based rating reﬂects an

ethnotheory of the demands imposed on infants in their developmental niches.

The DOTS-RzEthnotheory scale consists of 13 items describing the following ﬁve
dimensions of temperament: Activity Level, Approach-Withdrawal, Quality of Mood,
Task Orientation, and Flexibility-Rigidity. As with the DOTS-R, each DOTS-
RzEthnotheory question uses a four-choice format with high scores indicating greater
difﬁculty of interaction. The response alternatives are “4” = really wanted (and therefore
not difﬁcult); “3” = sort of wanted (and therefore only a little diﬁ'rcult); “2” = sort of not
wanted (and therefore somewhat difﬁcult); and “ 1” = really not wanted at all (and therefore
very difﬁcult). Ifthe item was considered not to be important or relevant to the parent then

the corresponding response would be “not difﬁcult.”

The scoring procedures for the DOTS-RzEthnotheory questionnaire are identical to
those described above for the scoring of the DOTS-R scale. And, in correspondence with
the DOTS-R, ﬁve scores are derived, one for each temperament attribute measured by this
scale. These scores, then, represent a parent's ethnotheory about what he/she would ﬁnd
difﬁcult about a child's temperamental style based on the behaviors the parent wants from
the child.

Psychometric properties of the DOTS~RzEthnotheory scale have been reported in
Windle and Lerner (1986). The internal consistency coefﬁcients (Cronbach alphas) for the
subscales range from .65 to .92 with an average reliability of .81.

36
Infant Temperament
The Infant Behavior Questionnaire (Rothbart, 1981) was used to assess mother’s ratings of
infant temperament. This questionnaire assesses temperamental dimensions of Activity
Level, Smiling and Laughter, Fear, Distress and Latency to Approach Sudden or Novel
Stimuli, Distress to Limitations, Duration of Orienting, and Soothability through caregiver
report. It consists of 94 items, but for the purpose of this study only 79 items were used in

order to cut down on the length of time required to ﬁll out the questionnaire.

Each item has seven response choices ranging from never to always. An illustrative item
is the following: “during the past week, when being undressed, how often did your baby
smile or laugh?” Household reliability was assessed using a sample of 22 mothers who
ﬁlled out the questionnaire along with a second adult in the household who spent time

caring for the infant.

Product-moment correlations for agreement of the 22 matched pairs of questionnaires
were: Smiling and Laughter, r = .45, Duration of Orienting, r = .46, Soothability, r = .54,
Fear, r = .66, Distress to limitations, r = .60, Activity Level, r = .69. All correlations were

signiﬁcant at p < .05.

Evidence for the convergence validity of the IBQ was provided by Goldsmith and
Rieser-Danner (1990). They used a sample of 32 mothers and 57 day-care teachers who
reported on infants aged 2.5-8.2 months. They reported convergence among the IBQ, the

Revised Infant Temperament Questionnaire, and the Infant Characteristics Questionnaire.

In addition, Rothbart (1981) reported extensive convergence between IBQ scores and
measures derived from home observations using a longitudinal sample of 46 infants seen
at 3, 6, and 9 months of age. Internal reliability has been reported with correlations
averaging above .50 for all dimensions at three, six, nine, and twelve months. Reported
alpha coefﬁcients exceed .70.

37

Discrepancy Scores

The temperament scores and the demands scores (expectational or behavioral) together
determine the degree of ﬁt between an infant attribute and the context (mother). Following
the work of Lerner and Lerner (1987a), discrepancy scores between each infant attribute
and each maternal expectational demand were calculated. In addition, discrepancy scores
between each infant attribute and each maternal behavioral demand, as well as discrepancy
scores between each maternal behavioral and each maternal expectational demand were

calculated.

The IBQ dimensions were matched up with the DOTS-R and the DOTS-RzEthnotheory
dimensions based on face validity. Five of the DOTS-R and DOTS-RzEthnotheory
dimensions were retained because they seemed to correspond closely to ﬁve of the IBQ
dimensions. All remaining DOTS-R, DOTS-RzEthnotheory, and IBQ dimensions were
discarded. The retained DOTS-R and DOTS-RzEthnotheory dimensions and their

corresponding IBQ dimensions are listed below in Table 4.

TABLE 4. Retained DOTS-R, DOTS-R:Ethnotheory, and IBQ Dimensions

 

 

 

 

 

 

 

 

4:11.:TS-R / DOTS-RzEthnptheory Dignsion CglmpondinggiQ Dimension
Activity Level - General Activity Level
Approach - Withdrawal Distress and Latency to Approach Sudden or
Novel Stimuli
Flexibility - Rigidity Distress to limitations
Quality of Mood Smiling and Laughter
Task Orientation Duration of Orienting

 

 

 

 

Thus, each of the 60 subjects received three sets of discrepancy scores. Each set of
discrepancy scores consists of ﬁve difference scores, one for each of the retained

temperamental attributes. A discrepancy score equal to zero or near zero indicates a match

38
between an infant attribute and the demands (expectational or behavioral) of the mother or,

alternatively, a match between the expectational and the behavioral demands of the mother

with respect to that attribute.

To give an example of how discrepancy scores are calculated, let’s consider the question
of whether there is a “ﬁt” between maternal temperamental attributes (DOTS-R
dimensions) and maternal expectational demands (DOTS-R:Ethnotheory dimensions). In
this case, discrepancy scores are obtained by z-scoring all DOTS-R and DOTS-
RzEthnotheory scores and then subtracting from each mother’s DOTS-R z-score (for each
of the ﬁve retained attributes) the corresponding DOTS-RzEthnotheory z-score (i.e.,
DOTS-R z-score minus DOTS-R: Ethnotheory z-score = discrepancy, or ﬁt, score). These
scores indicate the total amount of discrepancy between maternal attributes (i.e. maternal
behavioral demands) and maternal expectational demands. High scores (positive or
negative) would indicate a maximum amount of mismatch between attributes and
demands, and low scores would indicate a maximum amount of match between attributes
and demands. A discrepancy score of zero would indicate the best ﬁt (i.e., the least

mismatch between temperament and preferences).

Maternal Mental Health Functioning: Depression

To assess each woman's level of stress before and after having their child, the Center for
Epidemiological Studies-Depression Scale (CBS-D) was used. This scale is a self-report
measure of depressive symptomology that was developed for research applications. It
assesses three components of depressive symptoms: 1. behavioral, 2. cognitive, and 3.
happiness-sadness. Subjects are asked to respond to 20 statements describing particular
ways they might have felt during the past week, with answers ranging from “0” == rarely or
none of the time to “3” = most or all of the time. The possible range of total scores is 0-60,

with higher scores reﬂecting greater stress. The CBS-D appears to have adequate

39
psychometric properties (Radloﬂ‘, 1977) with reported coefﬁcient alphas of .84, .85, and

.90.

Maternal Mental Health Functioning: Anxiety

To assess each woman's level of anxiety before and aﬁer the birth of her infant, the State-
Trait Anxiety Inventory Form Y or STAI (Spielberger, 1983) was used. It yields
information on the individual's level of both ”state” anxiety and ”trait" anxiety. The 20
state-anxiety items ask the subject to respond to statements describing how they feel "right
now". The 20 trait-anxiety items ask the subject to respond to statements describing how
they "usually fee ". Items are rated on a 4-point scale ranging from 1 = "not at all” to 4 =
”very much". Higher totals indicate greater anxiety levels. Concurrent validity and test-
retest reliability (range .73 to .86) have been reported by Spielberger (1983). Alpha

coefﬁcients was reported at .93 for state anxiety, and .91 for trait anxiety.

Chapter 3
Results

Data Assessment

Data was coded and entered into a computer ﬁle by undergraduate students trained by two
graduate assistants. Since all the measures consist of Likert-type scales, coding each
question was done through a basic procedure of assigning to it the number of the answer
choice. All data was coded twice by two separate staff members to ensure accuracy of
coding. Next, data was systematically key-punched into the computer using identiﬁcation
numbers. The SPSS Windows statistical package was used for data entry.

All entered data was veriﬁed once to ensure accuracy of key-punching. Veriﬁcation was
done by obtaining a printout of the data in the computer database and checking it against
the actual questionnaires ﬁlled out by the subjects. To preserve statistical power and
maintain equal sample size for all variables, missing values were replaced with the series

mean.
Upon completion of data entry, the following analyses were performed:

1. Descriptive statistics and variable intercorrelations for all of the measures

used.

2. Scale reliability analyses to determine the internal consistency reliability

(alpha) of the measures employed in the study.

3. Stability analysis of DOTS-R, CESD, and STAI scales.

4O

41
4. Change analysis of DOTS-IL CESD, STAI, and DOTS-RzEthnotheory

scales followed by some regression analyses to determine the factors inﬂu-
encing the amount of change in the level of a variable after the birth of the

infant

5. Correlation analysis to assess the predictive relationship between maternal

temperament and maternal demands.

6. Correlation analysis to assess the predictive relationship between discrep-

ancy scores and measures of maternal functioning.

Descriptive Statistics

As described eariier in the sample section, the group of women in the study are highly
educated with a mean age of 28, have fairly prestigious occupations, and are primarily

caucasian career women.

Table 5 presents the mean and standard deviation for each of the scales used in the
present study. Examination of the table reveals that, for each subscale, the mean at Time 1
differed only slightly from the mean at Time 2 and the mean at Time 3. In addition, the
mean scores of the mother’s expectational demands (DOTS-RzEthnotheory) are not much
different from the mean scores of her behavioral demands (DOTS-R). Finally, mean scores
on the Depression and State Anxiety scales seem to decrease slightly from Time 1 to Time

2 as well as from Time 2 to Time 3.

42

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TABLE 5. Descriptives for Scales
Time 1 Time 2 Time 3

Scale Mean SD Mean SD Mean SD
DOTS-R Activity Level: General 2.65 .64 2.60 .57 2.64 .58
DOTS-R Approach-Withdrawal 2.67 .58 2.71 .49 2.73 .57
DOTS-R Flexibility-Rigidity 2.84 .57 2.84 .62 2.87 .61
DOTS-R Quality of Mood 3.42 .56 3.41 .57 3.51 .49
DOTS-R Task Orientation 2.61 .60 2.59 .40 249 .36
DOTS-R:Ethnotheory Activity Level NA NA 3.14 .54 3.11 .56
DOTS-RzEthnotheory Approach-Withdrawal NA NA 2.92 .57 2.84 .59
DOTS-R:Ethnotheory Flexibility-Rigidity NA NA 2.97 .48 2.88 .52
DOTS-R2Ethnotheory Quality of Mood NA NA 2.81 .52 2.73 .57
DOTS-RzEthnotheory Task Orientation NA NA 2.47 .47 2.44 .55
IBQ Activity Level NA NA NA NA 3.66 .74
IBQ Latency to Approach Sudden or Novel Stimuli NA NA NA NA _ 2.32 .66
IBQ Distress to Limitations NA NA NA NA 3.68 .74
IBQ Smiling and Laughter NA NA NA NA 4.63 1.1
IBQ Duration of Orienting NA NA NA NA 3.83 1.0
CESD 15.68 9.7 15.03 9.5 12.82 8.3
STAI State Anxiety 36.14 11 34.5 11 31.49 9.2
STAI Trait Anxiety 51.74 4.5 52.68 5.8 53.37 4.2

 

 

 

 

 

 

 

Note: NA= Not Applicable

 

43
Variable Intercorrelations

Calculation of Pearson product-moment correlation coeﬁicients for all variables used in
the study, including demographic variables, provided information concerning the
interrelationships among these variables for the sample. These intercorrelations are

presented in Table 6 below.

Examination of the intercorrelations revealed that, in the sample, age, degree to which
the pregnancy was planned, and income were all positively correlated with each other. In
other words, the older the mother the more income she has and the more likely she is to

have expected and planned for her pregnancy to occur.

Not surprisingly, depression and state anxiety at all times of testing were negatively
correlated with age, degree of planning, and income. Trait anxiety, on the other hand,
seems to be positively correlated with the three demographic variables. That is, older
women, who were also more likely to have planned their pregnancy and to have higher
income and more prestigious jobs, had higher (but not clinically problematic) levels, of

trait anxiety compared to younger women.

Note also that at each time point, each dimension of the mother’s temperament (DOTS-
R subscale) was positively correlated with itself at the other two time points. For example,
mothers who were high on Activity Level at Time 1 were also highly active at Time 2 and
at Time 3 (after the birth of their babies).

The same was also true of the mother’s demands with respect to her infant’s
temperament. To give an example, on average, a mother who demanded a high level of
Task Orientation in the infant at Time 2, still demanded a high level of Task Orientation in

her infant after he/she was born.

Finally, the last set of correlations that are of interest to note are those between the

mother’s behavioral demands (DOTS-R) and her expectational demands (DOTS-

ICALI: DINOOIA’MICI

1. AﬂI

2. OCCUPATION PREITIOE
3. 'IAMMID PIIGMAMCY?

4. IMCOMI

IIRTM COMPLICNTII

4. ACTIVITY LIVID

7.
0.

APPROACH-MITMDIAMAL
EIIXIEILITY-EIGIDITY

9. Quality o2 Mood

10. TAIE ORIENTATIOM

1:.

12.
13.

ICALE:

14.
15.
14.
17.
10.

DIPIEIIIOM'AT TIMI 1

STATE AMIIETY
TIAIT ANXIETY

ACTIVITY LEVEL

APPROACH-MITHDRAMAL
FLEXIBILITY-RIGIDITY
ﬁnality of Mood
1A8! ORIENTATIOM

DOTI-A AT TIMI 1

ITAI AT TIME 1

DOTS-I AT TIME 2

1.0000
.2554.
.4207“
.3932“

-.0092

-.1240
.1454
.1530
.1159
.2931

-e515‘..

-.3272
.4402"

.1014
.2424
-.0025
.1401
.0402

ICALI: DOTI-IleMMDﬂHEOEY AT TIMI 2

19.
20.
21.
22.
23.

2‘.

ICALE: ITAI AT TIME 2

25.
24.

27.
20.
29.
30.
31.

ICALI:

32.
33.
34.
35.
34.

37.
30.
30.
40.
41.

‘2.

ICALI: ITAI AT TIMI 3

43.
44.

ACTIVITY DEVIL

APPAOACH-MITMDAAMAL
EbEIIEILITY-IIOIDITY
Quality o! Mood

TASK ORIENTATION

EIPEEIIIOM AT TIME 2

ITATE AMRIETY
TIAIT ANXIETY

ACTIVITY LEVEL

APPEOACM-MITMDIAMAL
EbEXIEILITY-IIOIDITY
Quality of Mood

TAIE ORIENTATION

ACTIVITY DEVIL

APPROACH-MITMDIAMAL
ThEXIEILITY-EIOIDITY
Gallity o! Mood

TASK ORIENTATION

ACTIVITY LIVED

A'PIOACﬂ-MITMDIAMAL
ThIIIIILITY-IIOIDITY
Quality of Mood

TAII OIIIMTATIGI

DIPIEIIIOM AT TIME 3

IRATE ANXIETY
TIAIT AMRIETY

DOTS-I TIMI 3

IBQ AT TIME 3

.1474
.0202
.1719
.3492“
.2420.

-.3034"

-.4525"
.3771"

.0924
.0002
.0544
.0449
.0009

DOTI-AzETMNOTﬂEOEY AT TIMI 3

-.0405
-.0754
-.0400
.1775
.0140

-.4019"
-.1174
.1791
-.4592“
-.3209'

-.4470“

'0215‘
.2077.

.2554.
1.0000

.1939

.3255.
-.1497

-,3011
.0740
.2474
.0042

-.0037

-0323‘

-.1739
.2070

-.1020
.1492
.0053
.1002

-.0140

-.1533
.0747
-.0402
.1403
.0907

-.0557

.0124
-.°°51

-.1431
.0339
.0554
.1741

-.0295

-.1171

-.0139

-.0910
.0071
.2452.

-.1595
-.1590
.0409
-.3159‘
-.1775

-02020

-.0301
.1000

44

.4207"

.1939
1.0000

.5410"
-.0400

-.2944
-.0972
-.0340
.1147
.2440

-o“”.

-.2034
.3414

-.1440
.1024
.0411
.3337“

-.1053

-.0410
.0541
.1477
.0041
.3440"

-.3101°

-0‘11‘..
.3929“

-.1457
-.1420
-.0145
.1352
.0509

.0127
.0319
.1244
.1302
.1929

-.2147
-.0000
-.0025
-.1931
-.1994

—.4354“

-03811..
.4559.”

.3932“

.3255.

.5410“
1.0000
-.0314

.0000
—.1440
-.0314

.1509

.2047

-.3740‘

.3950.

.0044
-.0245
-.0474

.1400

.0031

.1310
.1133
.1347
.3575“
.2744‘

-02‘37

-.2.81.
.3150.

.0945
-.2034
-.0042

.1040

.0021

.1453
.0534
-.0425
.2377
.2024

-.2475‘
-.1044
-.0454
-.2044‘
-.1494

-.4441°'

-0238,
.2940.

-.0092
-.1497
-.0400
-.0314
1.0000

-.0720
.2514
.1314
.3900‘

-.0934

-0125‘

-0093?
.3712.

.0019
.0902
.1494
.1922
-.0494

.2044
-.0425
.1105
-.1000
-.1113

-0225?

-0107,
.1140

-.1025
-.0435
.0100
.0434
-.0310

.1509
.0744
.0277
-.1701
-.1240

.0790
.0434
-.1404
-.0549
.0411

.1530

.0710
-.0703

-.1240
-.3071
-.2944

.0000
-.0720

1.0000
.0543
-.1044
-.1309
.0357

.2091

.2140
-.2425

.0012‘°
-.2774
-.3590‘
-.2720
-.0710

.3315
.2040
.0534
.2409
-.1404

.4540.

.4700“
-.3742'

.7540“

.1002

.0495
-.2212
-.2044

.3410.

.1344
-.0000

.2097
-.2092

.0179
.0951
.1454
.0037
.0270

'21”

.3309
-.3570‘

TABLE 6. Intercorrelations Among All Variables Used in the Study

.1454
.0740
-.0972
-.1440
.2514

.0543
1.0000

.4570“

.2501
-.1500

—.2124

-013‘5
.0943

-.0040
.4040“
.2502
.1553

-.0230

.3540

.4320.

.4774“
-.0413
-.1399

-.°°72
-.1925

.1970
.0134"
.7229“
.2019
—.1490

.3202

.4303.

.2744
-.0425
-.1990

.2035
-.0441
-.1753

.0231
-.2035

‘oI‘gl

-.2491
.0744

.1530
.2474
-.0340
-.0314
.1314

-.1044
.4570“

1.0000
.1751
.0559

-.2‘5‘

-.1007
.1447

-.1402
.4437.
.4740”
.0047
.0593

.1504
.2144
.1713
—.0404
-.2703

-.3279

.2144
-00338

.1011
.4037"
.7579“
.0453
-.1702

.1247
.2403
.1433
-.1249
-.1400

.1724
-.1797
-.1471
-.0944
-.3441

“.106‘

-.2512
.1092

VARIABLE

ECALE: DIMOORAPHICE

1. AGE

2. OCCUPATION PRESTIGE
3. PLANNID PREGNANCY?
4. INCOMI

5. IIRTII MICATIQII

ICALI: DOTI-R AT TIMI 1
4. ACTIVITY LEVEL

7. AEPROACN-NITNDRAMRL
0. ELIXIEILITY RIOIDITY
9. Quality 0! Mood

10. TAIE ORIENTATION
11. DIPREIIION AT TIMI 1
ICALE: STAI AT TIMI 1
12. STATE ANXIETY

13. TRAIT ANIIETY

SCALE: DOTS-R AT TIME 2

14. ACTIVITY LEVEL

15. APPRM-NITMDRAMAL

14. ELEXIEILITY-RIOIDITY
17. Quality of ﬂood

10. TASK ORIENTATION

ECALE:
19. ACTIVITY LEVEL

20. APPROACH-MITHDRANAL
21. lLEXIEILITY-RIOIDITY
22. mality of ﬂood

23. TASK ORIENTATION

24. DEPRESSION AT TIME 2
ECALE: STAI AT TIMI 2
25. ETATE ANIIETY

24. TRAIT ANIIETY

”ALE: DWI-R AT TI‘ 3

27. ACTIVITY LEVEL

20. APPROACH-NITHDRAMAL

29. ELERIEILITY-RIOIDITY
30. mility of Mood

31. TASK ORIENTATION

ICALE:
32. ACTIVITY LEVEL

33 . APPRm-MITIIDRHAL
34. TLEXIEILITY-REOIDITY
35. Quality of Mood

34. TASK ORIENTATION

ICALE: IBQ AT TIHI 3

37. ACTIVITY LEVEL

30. APPROACH-NITNDRAMAL
39. TLEXIEILITY-RIOIDITY
40. mality of Mood

41. TASK ORIENTATION

42. DIPREIIION AT TIMI 3
ECALI: ITAI AT TIMI 3
43. ETATEFANRIETY

44. TRAIT ANIIETY

.1159
.0042
.1147
.1509
.3900.

-.1309
.2501
.1751

1.0000

-.4242‘

-.5292°°

-.5797“
.4915“

-.1002
.2071
.1590
.4700"

-.1314

DO‘I'I-Rzl'l'IIIOTHEORY AT TI‘ 2

.0427
.4205.
.1531
.3374
.0500

-0 5179..

.0491
.1041

.0235
.2002
.4510.
.7043“
-.2224

m-RszORY AT TI). 3

-.0347
.3935.

-.0540
.1743
.2040

.2571
-.0400
.0459
.1294
-.2190

-.3170

-.4354'
.5402"

10

.2931
-.0037
.2440
.2047
-.0934

.0357
-.1500

.0559
-.4242'
1.0000

-0 19”

.0215
.0420

.0040
6.1431

.1119
-.0057

.4444"

-.0022
-.1705
-.0391
.0574
.2154

-o 11”

-o“‘2.
.2719

-.0209
-.2322
-.3040
-.4010'
.5934"

.0559
-.1391
-.0075
-.1107

.1012

-.4345‘
-.1053
-.0201
-.3154
-.0474

-.2420

-.2105
.0744

45

Table6 (eont’d)
11 12
-.5151" -.3272
-.3234 —.1739
“0“”. “02°“
-.3740° -.1951
-.1254 -.0937
.2091 .2140
-.2124 -.1345
-.2454 -.1007
-.5292'° -.5797°‘
-.1990 .0215
1.0000 .7159“
.7159.“ 1.0000
-.7154“ -.4551“
.2040 .1454
-.7007“ -.5422“
-.3415 -.0773
-.1592 -.0053
-.1003 -.0334
-.0024 -.0424
-.2001 -.2071
“03‘10 -9211‘
.0354“ .4120.
.3435 .2449
-.2420 -.1037
.2242 .1702
-.1423 -.1207
-0285: -011“
-.345°.. '0‘8‘7..
-.2174 -.1272
.0504 .0079
-.1321 -.0703
.0044 .2509
.0140 .1343
-.3305 -.4547'
.5249“ .3054
.1447 .1570
.1430 -.1592
.1534 -.0224
.5041“ .2147
.4274" .5709“
-.4524‘° 0.4419.

13

.4402"
.2070
.3414
.3950.
.3712‘

-.2425
.0943
.1447
.4915“
.0420

-.7154‘°

-.4551“
1.0000

-.0727
.2147
.2240
.7305“
.1707

.1313
.1740
.0795
.4240.
.2075

-.4052"

.5079"

.0509
.0740
.1011
.4070.“
.0500

-.0420
.1590

-.2223

-.0500
.3047.

-.1500
-.3427

.1110
-.1352
-.2724

-o 52”..

-01‘31..
.7441.“

14

.1014
-.1020
-.1440

.0044

.0019

.0012“
-.0040
-.1402
-.1002

.0040

.2040

.1454

1.0000
-.0307
—.0952
-.0422
-.1505

.2409
.1940
.0371
.2501
-.1100

.1250

.0724
-.0003

.7247“

.0401
-.0570

.0077
-.0700

.3332"

.0127
-.1303

.1191
-.0341

.0201
-.0109
.1037
-.0247
.0153

.1134

. 1‘13
-025‘1.

15

.2424
.1492
.1024
-.0245
.0902

-.2774
.4040“
.4437.
.2071

-.1431

-.1042
.2147

-.0307

1.0000
.4102“
.4039"
.0472

.2444

.3014.

.4294"
-.0313
-.0447

.0 113°

.0 1729
.1212

.1100
.4999“
.5125“
.4957“
.0497

.2907.

.1954

.1009
-.0544
-.1172

.0434
-.1150
-.1102

.1723
-.0144

.0344

14

-.0025
.0053
.0411

-.0474
.1494

-.3590'
.2502
.4740“
.1590
.1119

-.2143

—.1300
.2240

-.0952
.4102“

1.0000
.4741“
.0351

.0392

.3010.

.2971‘
-.0710
-.1937

.2415

.2240
.1791

.0209
.2914.
.5537“
.3299.
-.0794

.2325
.3427.“
.3107.
.1432
.1049

.2497.
.1544
.0000
.2179
.0009

.0003

.0005
- 009‘9

VIRIAELE

SCALE: DEMOCRAPNICS

SCALE: DOTS-R AT TIME 1

4.
7.
0.
9.
10.

11.

SCALE:

12.
13.

SCALE:

14.
15.
14.
17.
10.

OCCUPATION PRESTIGE
PLANNED PREGNANCY?

INCOME

EIRTN COMPLICATIONS

ACTIVITY LEVEL

APPROACH-NITNDRAMAL
ELRIEILITY-RIOIDITY

eucliey of ﬂood

TASK ORIENTATION

DEPRESSION AT TIRI 1

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACN-NITHDRAMAL
FLIXIEILITY-RIGIDITY

Quality of ﬂood

TASE ORIENTATION

STAI AT TIME 1

DOTS-R AT TIME 2

17

.1401
.1002
.3337“
.1400
.1922

.2720
.1553
.0047
.4700“
-.0057

-.7007"

-.5422"
.7305"

-.0422
.4039"
.4741"

1.0000

-.0207

SCALE: DOTS-RzEWORY AT TI' 2

19.
20.
21.
22.
23.

24.

SCALE: STAI AT TIME 2

250
24.

SCALE: DOTS-R AT TIME 3

27.
20.
29.
30.
31.

SCALE:

32.
33.
34.
35.
34.

SCALE'
0

37.
30.
39.
40.
41.

42.

‘3.
44.

ACTIVITY LEVEL

APPROACH-MITHDRAMAL
TISXIEILITY-RIOIDITY

Qua] ity of ﬂood

TASK ORIENTATION

DEPRESSION AT TIII 2

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACH-WITHDRAMAL
ELIEIEILITY-RIOIDITY

Qual icy o! ﬂood

TASK ORIENTATION

ACTIVITY LEVEL

AEPROACN-NITNDRAMAL
TI‘XIEILITY-RIOIDITY

Qual ity of ﬂood

TASE ORIENTATION

ACTIVITY LEVEL

APPROACH-MITNDRAMAL
ELIXIEILITY-RIOIDITY

Quality of ﬂood

TASE ORIENTATION

DEPRESSION’AT TINI 3

STATE ANXIETY
TRAIT ANXIETY

DOTS- R: ETIEUO‘ITIEORY

IEQ AT TIRI 3

STAI AT TIME 3

.2103
.2374
.3059‘
.0994
.0744

.4313“.

-.3104°
.4004“

.0391
.3047.
.4214"
.7049"
-.0040

AT TIﬂE 3
.3400“
.2003
.1399

-.0105
.1794

.1153
-.1457
-.0051

.0453

.0497

- 02‘.7.

-.2053‘
.3402”.

46

Table 6 (eont’d)
10 19 20
.0402 .1474 .0202
-.0140 -.1533 .0747
-.1053 -.0410 .0541
.0031 .1310 .1133
-.0494 .2044 -.0425
-.0710 .3315 .2040
-.0230 .3540 .4320.
.0593 .1504 .2144
-0131. 00‘27 0‘205.
.4444.‘ -.0022 -.1705
-.0773 -.0053 -.0334
.1707 .1313 .1740
-.1505 .2409 .1940
.0472 .2444 .3014‘
.0351 .0392 .3010.
-.0207 .2103 .2374
1.0000 -.0703 -.1123
-.0703 1.0000 .2434.
-.1123 .2434. 1.0000
-.3002° .2400 .3349“
-.0230 .1013 .1435
.1944 -.0020 -.1400
-.1373 -.1541 -.1435
-.0119 -.0972 —.0244
.1734 .0493 -.0770
- 00280 0310’. o 19"
.0029 .2342 .2002.
-.1199 .2354 .3744"
-.1570 .2443 .2205
.4103" .0240 -.0093
-.1070 .4209“. .1453
-01‘5. 03173. 05.76..
-.0475 .0004 .1214
-.2100 .1044 .0705
.2200 -.3374" -.2102
-.0934 .0431 -.0427
- 00.3, - 03230. " 030.1.
.1210 -.2253 -.0992
.0417 ‘ 022,3 -' .2790.
-.0091 -.1019 -.1314
-.0494 -.0250 .0549
.0744 .0532 .0205

21

.1719
.0402
.1477
.1347
.1105

.0534
.4774“.
.1713
.1531
-.0391

-.0024

.0795

.0371
.4294“
.2971“
.3059‘
-.3002‘

.2400

.3349“
1.0000

.0404
-.1329

c.1002
.0072

.1503
.4494"
.4040“
.3039.
-.3414"

.2324

.3040“

.3242.
-.0709
-.1977

.0342
-.1011
-.3043°

.0027
-.0704

-.0597

22

.3492“
.1403
.0041
.3575“
.1000

.2409
-.0413
.0404
.3374
.0574

-.2001

-.2071
.4240‘

.2501
-.0313
-.0710

.0994
-.0230

.1013
.1435
.0404
1.0000
.0149

-.0993

-.0593
.2174

.2974‘
.0775
.0720
.1512
-.1412

.0470
-.0140
-.0420

.4474“

.0775

.0140
.0145
.3245.
.1220
.1415

-.2951‘

.2241
.1013

23

.2420.
.0907
.3440.“
.2744“
-.1113

-.1404
-.1399
-.2703
.0500
.2154

-.2114
.2075

-.1100
-.0447
-.1937
.0744
.1944

-.0020
-.1400
-.1329

.0149
1.0000

-.1124

.1045

-.0340

-.1431

-.1455
.1294
.2944.

-,0530

-.0420

-.1445
.1345
.5231"

-.0992
.0309
.1041

-.1904

-.1247

-.3510"

-031‘3.
.3974"

24

-.3034"
-.0557
-.3101‘
-.2437
-.2257

.4540.
—.2404
-.3279
-.5179“
-.1190

.0354“

.4120.
-.4052“

.1250
-.1730
-.2415
-.4313"
-.1373

-.1541
-.1435
-.0490
-.0993
-.1124

1.0000

.4022“
-.4000°‘

.1050
.0149
-.2440
-.2344
-.0030

-.0902
-.0431
-.1474
-.0052
-.0527

.1353
.1100
.1311
.2421.
.1511

.4302“

.1934
-0‘1"..

VARIABLE

SCALE: DEMOGRAPHICS

1.
2.
3.
4.
5.

4.
7.
0.

ACE
OCCUPATION PRESTIGE
PIJHIIED PREGNANCY
INCOMI

BIRTH CCMPLICATIGNS

9. Quality of ﬂood

10.

11. DEPRESSION’AT TIMI 1

SCALE: STAI

120
13.

SCALE: DOTS-R AT TIME 2

14.
15.
14.
17.
10.

SCALE:

19.
20.
21.
22.
23.

2‘.

SCALE:

25.
24.

TASK ORIENTATION

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACH-NITHDRANAL
TLEXIBILITY-RIGIDITY

Quality of ﬂood
TASK ORIENTATION

ACTIVITY LEVEL

APPROACH-NITHDRANAL
PLEXIBILITY-RIGIDITY

Quality of ﬂood
TASK ORIENTATION

DEPRESSION AT TIME 2

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACH-NITHDRAMAL
ELEXIBILITY-RIGIDITY

Quality of ﬂood
TASK ORIENTATION

DOTS-R AT TIME 1
.ACTIVITY LEVEL
APPROACH-NITHDRAMAL
PLEXIBILITY-RIGIDITY

AT TIMI 1

STAI AT TIMI 2

DOTS-R AT TIME 3

25

-.4525“
.0124
-.4114"
-.2001°
-.1075

.4700"
-.0072
-.2144

.0491
-.4442°

.3435

.2449
-.3155

.0724
-.1729
-.2240
-.3104‘
-.0119

DOTS-R:ETIIIO'IHEORY AT TI!‘ 2

-.0972
-.0244
-.1002
-.0593
-.1921

.4022“

1.0000
-.7143°‘

.1705
.0194
-.1344
—.1534
-.1479

”ALE: m-a:mo‘rrrnorur AT TIDE 3

32.
33.
34.
35.
34.

SCALE:

37.
30.
39.
40.
41.

42.

SCALE: STAI AT TIME 3

43.
44.

ACTIVITY LEVEL

APPROACH-NITHDRAMAL
PLEXIBILITY-RIGIDITY

Quality of ﬂood
TASK ORIENTATION

IBQ AT TIME 3
ACTIVITY LEVEL

APPROACH‘NITHDRANAL
ELEXIBILITY-RIGIDITY

Quality of ﬂood
TASK ORIENTATION

DEPRESSION AT TIMI 3

STATE ANXIETY
TRAIT ANXIETY

-.2431
.0130
-.1144
.0731
-.0427

.2207
.0430
.0377
.2534
.0040

.3040.

.4100"
-.3001"

47

Table 6 (cont’d)
24 27 20
.3771“ .0924 .0002
-.0051 -.1431 .0339
.3929“ -.1457 -.1420
.3150. .0945 -.2034
-031“. 015‘... 01002
-.1925 .1970 .0134"
-.0330 .1011 .4037"
.1041 .0235 .2002
.2719 -.0209 -.2322
-.2420 .2242 -.1423
-.1037 .1702 -.1207
.5079" .0509 .0740
-.0003 .7247" .0401
.1212 .1100 .4999"
.1791 .0209 .2914'
.4004" .0391 .3047.
.1734 -.0200 .0029
.0493 .3109' .2342
-.0770 .1949 .2002.
.0072 .1503 .4494.“
.2174 .2974’ .0775
o 1.65 ’0 03‘. "e 1‘31
-.4SSS.. 01858 001‘9
-.7143" .1705 .0194
1000” -003” -0085?
-.0390 1.0000 .3104.
-.0057 .3104. 1.0000
-.0297 .1055 .4145“.
.2740. .2110 .4753“
.1411 -.1000 -.1370
.2274 .3041" .4193“
-.0911 .0732 .2222
.1794 -.0547 .1292
.0147 .1571 .0302
.1459 -.0250 -.1425
-.1751 .1293 .1045
-.0700 -.0473 -.0105
.1752 .2747. -.0347
-.2012 -.0143 .2102
-.3740“ .0290 .0375
-.3740" .0324 -.1274
.4520“. -.0449 -.0102

29

.0544
.0554
-.0145
-.0042
.0100

.0495
.7229‘°
.7579"
.4510.
-.3040

-.2051

-.1704
.1011

-.0570
.5125“
.5537“
.4214“

-.1199

.2354
.3744"
.4040“
.0720
-.1455

-02‘60

.1055

.4145“
1.0000

.4404“
-.3010‘

.2744.
.4405"
.2244
.0534
-.1700

.3450.“
-.0413
-.1344

.2124
-.0344

.0019

.0 10”
.0349

30

.0449
.1741
.1352
.1040
.0434

-.2212
.2019
.0453
.7043“

-.4010°

-.5450“

.4070“

.0077
.4957"
.3299‘
.7049"
-.1570

.2443
.2205
.3039'
.1512
.1294

-.1534
.2740.

.2110

.4753“

.4404"
1.0000
-.1030

.3701"
.1040
-.0472
.1070
.1714

.0959
-.2707°
-.0149

.1207

.0444

-0 2038

-0 2872.
.3149.

31

.0009
-.0295
.0509
.0021
-.0310

-.2044
-.1490
-.1702
-.2224
.5934“

-.1272
00580

-.0700
.0497

-.0794

-.0040
.4103"

.0240
-.0093
-.3414"
-.1412

.2944.

-005”

.1411

-.1000
-.1370
-.3010‘
-.1030
1.0000

-.0705
-.0455
-.0235
-.1709

.2513

-.1590
-.0970
-.0053
-.0444
-.0233

-.2233

.1740

32

-.0405
-.1171
.0127
.1453
.1509

.3410‘
.3202
.1247
-.0347
.0559

.0504

.0079
.0420

.3332“
.2907'
.2325
.3400“
-.1070

.4209“
.1453
.2324
.0470
-.0530

.0902

.2431
.2274

.3041“
.4193“
.2744.
.3701"
-.0705

1.0000
.2004
.0703
.1275

-.0357

.2140
.1120
-.1223
.0302
.1444

.1240

-003“
-.0200

Table 6 (eont’d)
VARIAlaE 33 34 35 34 37 30 39 40
”LE: UMRAPHICS
1. ME -.0754 -.0400 .1775 .0140 -.4019“ -.1174 .1791 -.4592“
4. moon: .0534 -.0425 .2377 .2024 -.2475‘ -.1044 -.0454 -.2044‘
”ALE: DUES-R AT TIME 1
4. ACTIVITY IEVEL .1344 -.0000 .2097 -.2092 .0179 .0951 .1454 .0037
7. ”am-mum .4303‘ .2744 -.0425 -.1990 .2035 -.0441 -.1753 .0231
9. $1.11!? of ﬂood .3935. -.0540 .1743 .2040 .2571 -.0400 .0459 .1294
10. TMK ORII'TATIM -.1391 -.0075 -.1107 .1012 -.4345‘ -.1053 -.0201 -.3154
11. DDRBSICN AT TI). 1 -.1321 .0044 .0140 -.3305 .0451 .5249“ .1447 .1430
SCALE: STAI AT TIME 1
12. STATE ANXIETY -.0703 .2509 .1343 -.4547‘ -.1044 .3054 .1570 -.1592
13. TRAIT ANXIETY .1590 -.2223 -.0500 .3047. -.1500 -.3427 .1110 -.1352
SCALE: DUES-R AT TIME 2
14. ACTIVITY LEVI. .0127 -.1303 .1191 -.0341 .0201 -.0109 .1037 -.0247
10. TASK ORIUTATIW -.1450 -.0475 -.2100 .2200 -.0934 -.0439 .1210 -.0133
SCALE: DOTS-R: EMORY AT T“ 2
1’. ”In" m 03173. 0060‘ 020“ -0331‘.. 0063‘ -03230. -0225: -0095:
20. mam-urnmm .5074“ .1214 .0705 -.2102 -.0427 -.3047' -.0992 -.0391
21. PLEXIBILITY-RIGIDITY .3040“ .3242. - .0709 - . 1977 .0342 - . 1011 - .3043. .0027
22. Quality of ﬂood -.0140 -.0420 .4474“ .0775 .0140 .0145 .3245. -.1220
24. DDR’IN AT ms 2 -.0431 -.1474 -.0052 -.0527 .1353 .1100 .1311 .2421.
”LE: STAI AT TIME 2
25. STATE ANXIETY .0130 -.1144 .0731 -.0427 .2207 .0430 .0377 .2534
24. TRAIT ANXIETY -.0911 .1794 .0147 .1459 -.1751 -.0700 .1752 -.2012
SCALE: DOTS-K AT TIME 3
27. ACTIVITY LEVEL .0732 -.0547 .1571 -.0250 .1293 -.0473 .2747. -.0143
29. MIBILITY-RIGIDITY .4405“ .2244 .0534 -.1700 .3450“ -.0413 -.1344 .2124
31. no: oaxmauou -.ocss -.ozss -.1709 .2513 '-.1soo -.os'ro -.ooss -.o«s
”LI: WE‘R: WAY AT TI‘ 3
32. ACTIVITY LIVE. .2004 .0703 .1275 -.0357 .2140 .1120 -.1223 .0302
33. macaw-utmmr. 1.0000 .4050“ .1140 -.2744’ .1074 -.0009 -.1145 .0594
34. PLEXIBILITY-RIGIDITY .4050“ 1.0000 .0715 -.3057“ .0959 .0320 -.0049 -.0144
35. ”11%! of M004 .1140 .0715 1.0000 -.0900 .0423 .1279 .1704 -.1250
34. TASK ORIDITATIMI -.2744‘ -.3057“ -.0900 1.0000 .0952 .2509. .2940. .0244
m: I” AT TI‘ 3
37. ACTIVITY LEVEL .1074 .0959 .0423 .0952 1.0000 .2005. .1259 .4130“
30. APPRm-NITHDRAML - .0009 .0320 . 1279 .2509. .2005. 1 .0000 . 1992 . 1455
39. PIIXIBILITY-RIGIDITY -.1145 -.0049 .1704 .2940. .1259 .1992 1.0000 -.1720
40. m1ity of N000 .0594 -.0144 -.1250 .0244 .4130“ .1455 -.1720 1.0000
42. DEPR-IU AT TI). 3 .0440 -.1405 -.1344 -.2549‘ .3750“ .2341 -.1750 .4549“
SCALE: STAI AT TIME 3
43. STATE ANXIETY .0454 .0941 .0099 -.3371“ .0904 .1221 -.0404 .0477
44. TRAIT MISTY -.0755 -.1394 .0703 .3737“ «.2344 -.1333 .0992 -.2417

48

49

Table 6 (eont’d)
“RIABLE 41 42 43 44
“ALE: DWI“
1' n -032”. -o“7°.. -021“ 02.".
3. PM PM? -.1994 -.4354°' -.3017.. .4559“
‘0 I“ -0 2‘96 -o“‘1.. -0 23” o 29“ .
CALI: DOTS-R AT TI- 1
4. ACTIVITY I‘VEL .0270 .2190 .3309 -.3570°
9. Quality of Mood -.2190 -.3170 -.4354‘ .5402“
10. TASK ORIUTATIW -.0474 -.2420 -.2105 .0744
11. “PRBSIG AT TIME 1 .1534 .5041“ .4274“ -.4524°‘
”ALI: STAI AT TIME 1
12. STATE AMIETY -.0224 .2147 .5709“ -.4419‘
13. TRAIT MIETY -.2724 -. 5290“ -.7431“ .7441“
“ALE: MRS-R AT TIME 2
14. ETIVIT'Y LEVI. .0153 .1134 .1413 -.2547‘
15. APPRQGl-NITTIDRMAL -.0144 -. 0901 -. 1934 . 0344
14 . PLEXIBILITY-RIGIDITY - . 0009 . 0003 . 0005 - . 0949
17. molity of Mood .0497 -.2407° -.2053‘ .3402“
10. TASK ORIIITATIQI .0417 -.0091 -.0494 .0744
”ALI: m-n:maon AT TIT. 2
19. ETIVITY LEVI. -.2293 -. 1019 -.0250 .0532
21 . ELIXIBILITY-RIGIDITY -.0704 -.0942 -. 1405 -.0597
23. TASK ORII'TATIQ‘ -.1247 -.3510“ -.3143° .3974“‘
2‘- “SIM!" “T T” 2 01511 0‘3”.. 0193‘ -0‘1,‘..
ME: STAI AT TIME 2
2,0 "‘7' mxm .0040 030‘.. 0‘1”.. -03801..
24. TRAIT ANXIETY -. 1152 -.3740“ -.3740“ .4520“
“ALE: DOTS-R AT TI‘ 3
20. APPRQGI-NITHDRMAL .0042 .0375 -. 1274 -.0102
29. ELIXIBILITY-RIGIDITY -.0344 .0019 -. 1000 . 0349
30. ﬁnality of ﬂood .0644 -.2030 -.2075° .3149‘
31 . TASK ORIUTATIW -. 0233 -. 2233 -. 1302 . 1740
m: DOTS-R: ETTNOTHEORY AT TI‘ 3
33 . APPRM—NITHDRANAL -. 1325 . 0440 . 0454 -. 0755
34 . PLIXIBILITY-RIGIDITY -. 0074 -. 1405 . 0941 -. 1394
35. mdlity of ﬂood -.2151 -.1344 .0099 .0703
34. TASK ORIUTATIW . 1112 -.2549' -.3371“ .3737“
scam: I” AT TI‘ 3
310 ”TIVITY m 021” 037”.. 00’“ -023“
30. ”PRW-NITIDRMAL .3457“ .2341 . 1221 -. 1333
40. 9.1.1in of Mood .5445“ .4549“ .0477 -.2417
‘20 "SEMI“ “T 'm 3 I“31.. 100°” 05“3.. -o.°”..
scans: an: au- rnra 3 N“;
43. era-rs mxm .2003 .5443“ 1.0000 -.0099°° ‘-Signiﬁemtatps.05

«. mn- mrm -.3203° -.eou“ -.4499" 1.0000 ”-Srgniﬁe-aupsm

SCALE:
AGE
OCCUPATIOM’PRIBTIOE
PLANNED PREGNANCY?

neon:
BIRTH COMPLICATIONS

1.

SCALE:

‘S
7.

9. Quality of ﬂood

10.

11.

SCALE: STAI AT TIME 1

12.
13.

”ALE: DOTRS-R AT TIME 2

14.
15.
14.
17.
10.

19.
20.
21.
22.
23.

24.

SCALE:

2:.
24.

”ALE: DOTS-R AT TIIC 3

27.
20.
29.
30.
31.

SCALE: DOTS-RtETHNOTHEORY

32.
33.
34.
35.
34.

SCALE: IBQ AT TIME 3

37.
30.
39.
40.
41.

42.

SCALE: STAI AT TIME 3

‘3.
44.

TASK ORIENTATION

DEPRESSION AT TIME 1

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACH-NITHDRAMAL
TLEXIBILITY-RIGIDITY
Quality of ﬂood

TASK ORIENTATION

ACTIVITY LEVEL

APPROACH-NITHDRAMAL
TLEXIBILITY-RIGIDITY
9aa1 ity of Mood

TASK ORIENTATION

DEPRESSION AT TIME 2

STATE ANXIETY
TRAIT ANXIETY

ACTIVITY LEVEL

APPROACH-NITHDRANAL
ELEXIBILITY-RIGIDITY
Quality of ﬂood

TASK ORIENTATION

ACTIVITY LEVEL

APPROACH-NITHDRAMAL
ILEXIBILITY-RIGIDITY
Quality o! ﬂood
TASK ORIENTATION

ACTIVITY LEVEL

APPROACH-NITNDRAMAL
PLEXIBILITY-RIGIDITY
mal ity of ﬂood

TASK ORIENTATION

DEPRESSION’AT TIME 3

STATE ANXIETY
TRAIT ANXIETY

DOTS-R AT TIME 1
ACTIVITY LEVEL
APPROACH-NITHDRAMAL
0. PLEXIBILITY-RIGIDITY

STAI AT TIMI 2

41

-.3209'
-.1775
-.1994
-.1494
.0411

.0270
-.2035
-.3441
-.2190
-.0474

.1534

-.0224
-.2724

.0153
-.0144
-.0009

.0497

.0417

MS-R:ETINTHEORY AT T198 2

-.2293
—.2790‘
-.0704
-.1415
-.1247

.1511

.0040

-.0944
.0042
-.0344
.0444
-.0233

AT TIME 3
.1444
-.1325
-.0074
-.2151
.1112

.2109

.3457“
-.0509

.5445"
1.0000

.4431“

.2003
-0 3203.

42

-.4470“
-.2020
-.4354“
-.4441’°
.1530

.2190
-.1491
-.1044
-.3170
-.2420

.5041“

.2147
-.5290“

.1134
-.0901

.0003
-.2407‘
-.0091

-.1019
-.1314
-.0942
-.2951°
-.3510‘°

.4302“

.3040.
-.3740°‘

.0290
.0375
.0019
-.2030
-.2233

.1240

.0440
-.1405
-.1344
-.2549‘

.3750“
.2341
-.1750
.4549“
.4431“

1.0000

.5443"
-.4000“

49

Table 4 (cont’d)
43 44
-021“ 02.77.
-.3017" .4559“
-023” 029‘0.
.33” -035”.
-0‘35‘. .3002"
.4274" -.4524°'
.5709" -.4419°
-.7431“ .7441"
a 1‘13 -025‘7.

-.1934 .0344
.0005 -.0949
-02053. 03‘82..

-.0494 .0744
-.0250 .0532
.0549 .0205
'- 1‘05 -0059,
-.2241 .1013
-.3143° .3974"
.1934 -.4194“
.4100“ -.3001°‘
-.3740“ .4520“
-.1274 -.0102
-02015. 031.9.
-.1302 .1740
-.0344 -.0200
.0454 -.0755
.0941 -.1394
.0099 .0703
-.3371“ .3737"
.0904 -.2344
-.0404 00992
.0477 -.2417
.2003 -.3203'
05‘63.. -.4000..
2000” -o“”..
-.4499" 1.0000

Notes:
‘-Si ' atps.05
” -Srp|ifrcantatps .01

50
RzEthnotheory). Examination of Table 6 reveals that, with the exception of Activity Level,

the dimensions of the mother’s temperament (DOTS-R) do not seem to be correlated with

the dimensions of temperament she demands in her infant (DOTS-RzEthnotheory).

Internal Consistency Reliability of Measures

Scale reliabilities, at each of the three points of testing, were examined using coeﬁicient
alpha internal consistency estimates. Reliability is important to assess because it inﬂuences
the degree to which two variables are correlated. Thus, the unreliability with which a
variable is measured might mask its relationship with other variables and thereby affect the

interpretation of the relationship.

Estimates of internal consistency reliability ranged from .60 to .95 across scales. These
alphas are presented in Table 7 below, along with the number of items that make up each
scale. All alphas at Time 1 are based on 31 cases (the number of 3-wave women in the
sample). All alphas at Time 2 and Time 3 are based on 60 cases (the number of 3-wave
women plus the number of 2-wave women in this study). Notice that the DOTS-
RzEthnotheory was measured at Time 2 and Time 3 only, while the IBQ was measured only
after the delivery occurred (i.e. at Time 3). Also, notice that it was not possible to compute
the reliability of the Quality of Mood subscale of the DOTS-RzEthnotheory since this
subscale is made up of only one item. Following is a brief discussion of the reliability

coefficients for each measure used in this study.

Reliability of DOTS-R

As can be seen in Table 7, the reliability coefﬁcients for the ﬁve temperament attributes
measured by the DOTS-R range from .36 to .91 at Time 1, from .76 to .90 at Time 2, and
from .68 to .88 at the Time 3 administration. The majority of the reliability coeﬁicients are
above .8 and therefore the reliability of the DOTS-R dimensions can be considered high.

5 1
TABLE 7. Reliability Estimates for Scales

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Alpha (0.)
Number Time 1 Time 2 Time 3
Scale of Items (N8 31) (N- 60) (N- 60)
DOTS-R Activity Level: General 7 .86 .83 .87
DOTS-R Approach-Withdrawal 7 .83 .76 .88
DOTS-R Flexibility -Ri gidity 5 .76 .84 .81
DOTS-R Quality of Mood 7 .91 .90 .87
DOTS-R Task Orientation 8 .36 .76 .68
DOTS-R:Ethnotheory Activity Level 2 NA .68 .61
DOTS-RzEthnotheory Approach-Withdrawal 2 NA .72 .71
DOTS-RzEthnotheory Flexibility Rigidity 2 NA .40 .53
DOTS-R2Ethnotheory Quality of Mood 1 NA NA NA
DOTS-RzEthnotheory Task Orientation 2 NA -.01 .51
IBQ Activity Level 15 NA NA .73
IBQ Latency to Approach Sudden or Novel Stimuli 12 NA NA .74
IBQ Distress to Limitations 16 NA NA .73
IBQ Smiling and Laughter 11 NA NA .84
IBQ Duration of Orienting 11 NA NA .82
CESD 20 .90 .91 .88
STAI State Anxiety 20 .94 .93 .92
STAI Trait Anxiety 19 .48 .62 .42

 

 

 

 

Note: NA= Not Applicable

52
Reliability of the DOTS-RzEthnotheory

Overall, the alphas for the DOTS-R:Ethnotheory scores are substantially lower than the
alphas for the DOTS-R scores.These results are inﬂuenced by the small number of items
that compose each of the DOTS-RzEthnotheory subscales.

Reliability of the IBQ

As Table 7 indicates, the subscales of the Infant Behavior Questionnaire have reliabilities
that range from .73 to .84. These alphas indicate that the IBQ, like the DOTS-R, is

internally consistent.

Reliability of the Two OutcomeMeasures

With reliabilities around .9, the Center for Epidemiological Studies-Depression scale can
be considered highly reliable. Similariy, the State Anxiety scale, with an average internal
consistency reliability of .93, is highly reliable. The sample’s responses to the Trait
Anxiety scale, however, does not seem to be adequately consistent. This is a surprising
finding given that the Trait Anxiety scale has an adequate number of items and that it

measures a fairly consistent construct.

Overall, it seems that the reliabilities of the scales used in this study were in the

moderate to high range.

Reliability of Discrepancy Between Expectations and Behavior

Table 8 shows the reliability of the three types of discrepancy scores used in this study.
Discrepancy scores were all based on z-scores because the IBQ was scored on a different

scale than that used to score the DOTS-R and the DOTS-R:Ethnotheory.

53
Discrepancy scores between the mother’s behavioral demands (i.e. temperament) and

her expectational demands are calculated by subtracting from the DOTS-R z-scone for each
temperament dimension the corresponding DOTS-RzEthnotheory z-score.

Discrepancy scores between the mother’s expectational demands and her infant’s
temperament are calculated by subtracting from the DOTS-RzEthnotheory z—score for each

temperament dimension the corresponding IBQ z-score.

Finally, discrepancy scores between the mother’s behavioral demands (i.e. the mother’s
temperament) and her infant’s temperament are calculated by subtracting from the DOTS-

R z-score for each temperament dimension the corresponding IBQ z-score.

The reliability of all discrepancy scores was calculated according to the following

formula:
Reliability or no,” = (a’ - r,,) I (1 - rxy)

Where D(x-y) is the discrepancy score between the z-score obtained on scale x and the z-
score obtained on scale y, a- is the average reliability of scales x and y, and r,‘y is the

correlation between scale x and scale y.

Notice that the reliability of most discrepancy scores is lower than the reliabilities of its
component scores. That is, the reliability of a difference score tends to be less than that of
the component scores, unless the components are negatively correlated. Therefore, using
these scores to predict outcomes will sometimes put us in the undesirable position of using

a measure that is more unreliable than the scores comprising it.

Stability Analysis

Several years ago, David R. Heise (1975) proposed a set of equations that can be used to
calculate stability. His method is built on the assumption that the reliability of a scale

54
TABLE 8. Post-Delivery Reliability Estimates for Discrepancy Scores

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Alpha ((1)

at

Time 3

_ _ k Illiscrepancy ‘ (N- 60)
Discrepancy—between DOTS-RzEthnIOtheoryTnd DOTS-WWW Level .58
Discrepancy between DOTS-RzEthnotheory and DOTS-R Approach-“Withdrawal .74
Discrepancy between DOTS-RzEthnotheory and DOTS-R F lexibility-Rigidity .57
Discrepancy between DOTS-RzEthnotheOiy and DOTS-R Quality of Mood NA
Discrepancy between DOTS-RzEthnotheory and DOTS-R Task Orientation .46
Discrepancy between DOTS-RzEthnotheory and IBQ Activity Level .58
Discrepancy between DOTS-RzEthnotheory and IBQ Approach-Withdrawal .75
Discrepancy between DOTS-R2Ethnotheory and IBQ Flexibility-Rigidity .63
Discrepancy between DOTS-R:Ethnotheory and IBQ Quality of Mood NA
Discrepancy between DOTS-RzEthnotheory and IBQ Task Orientation .62
Discrepancy between DOTS-R and IBQ Activity Level .77
Discrepancy between DOTS-R and IBQ Approach-Withdrawal .81
Discrepancy between DOTS-R and IBQ Flexibility-Rigidity .80
Discrepancy between DOTS-R and IBQ Quality of Mood .83
Discrepancy between DOTS-R and IBQ Task Orientation .75

 

 

Note: NA= Not Applicable

 

55
remains stable from one time point to the next. In actual research this assumption is not

necessarily true in all cases. The reliability of some scales may very well vary from one
point of time to the next. Thus, for some of the scales Heise's equations are relevant, but for
some other scales Heise's equations would not be applicable. For this reason, a least
squares path analysis program (Hunter, 1992) that allows us to calculate stabilities even
when the reliabilities are not constant over time was used to produce the stability

coefﬁcients.

The PATH program corrects the test-retest correlations for attenuation due to the
imperfect reliabilities of the scales. Correcting for attenuation is an important feature of the
data analyses. A fundamental psychometric problem in any study is measurement
reliability. That is, whether or not a relationship found between two variables is in part
dependent upon the reliability of the measurements of both variables. If one or more of the
measurements used are unreliable, the unreliability with which a variable is measured

might mask its real relationships with other variables.

Thus, one way to correct for the limiting effects of unreliability is to correct for
attenuation. The correction for attenuation corrects for the degree to which the correlation
is reduced by the unreliability of measurement contained in one or both of the variables. In
other words, the disattenuated correlation indicates what the correlation would be between
two variables if both of their reliabilities are perfect. The process of correcting a correlation

coefﬁcient for attenuation involves the following formula:
Corrected rxy=r,y/(.,/OT,c * Jay.)

Where rxy is the attenuated correlation between scale x and scale y, 0.x is the reliability
of scale x, and my is the reliability of scale y. Notice that the higher the reliability of the two
scales the less the correlation between them would be attenuated and therefore the smaller
the difference between their corrected and their attenuated correlations would be. As the

reliabilities approach zero, however, the correlation between the scales would be greatly

56
attenuated. Correcting for attenuation in this case would significantly increase the

correlation coeﬂicient and contribute to the accuracy of the reported results.

After correcting for attenuation, program PATH then proceeds to estimate stability
coefficients, in light of the scales’ reliabilities at each point in time, using multiple
regression, i.e. “ordinary least squares”. The generic model underlying the stability

analysis of all scales is presented below in Figure l where Lt represents the true (latent)

 

 

 

 

'33

Figure 1: Generic Model Underlying All Stability Analyses

score at Time t, Xt represents the fallible score (determined by responses on a particular
questionnaire measure) at Time t, at represents the reliability of the measure at Time t, the
square root of (11 represents the correlation between the true score and the fallible score,
and et represents the error of measurement at Time t The stability coefﬁcients are the
correlations between latent scores over time. More speciﬁcally, S A represents stability
from Time 1 to Time 2, and SB represents stability from Time 2 to Time 3. The stability
from Time 1 to Time 3 (SC) is the product of S A and SB.

57
Arrows from true scores to the fallible scores mean that a person’s true trait determines

their responses on the questionnaire measure. However, questionnaire responses depend on
a variety of other factors—moods, distractions, and misunderstandings. All such sources of
measuMent error are accounted for by the error term q. Further, the measurement errors

are assumed to be uncorrelated with one another over time (Heise, 1975).

The PATH program assesses the “ﬁt” of the above shown model to the data with a Chi
square (x2) statistic. Results of the stability analyses are summarized below in Table 9
which presents the reliability and stability coefficients, as well as the value of the 12 and its
tail probability (which are indicators of how well the data fits the model presented above),
for each of the DOTS-R, Depression, and Anxiety scales which are all the scales used in

this study that were measured at three different points in time.

TABLE 9. Summary of Stability Analyses

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable sA s, sc a, a; a, x’ :3»me
Scale: DOTS-R

Activity Level .95 .85 .80 .86 .83 .87 .48 .49
Approach-Withdrawal .87 .86 .74 .83 .76 .88 2.53 .l 1
Flexibility-Rigidity .60 .67 .40 .76 .84 .81 7.22 .00
Quality of Mood .74 .89 .66 .91 .90 .87 .86 .35
Task Orientation .99 .86 .99 .36 .76 .68 .26 .61
Scale: Depression (CBS-D) .93 .49 .46 90 .91 .88 1.28 .26
Scale: Anxiety (STAI)

State Anxiety .26 .45 .12 .94 .93 .92 6.72 .00
Trait Anxiety .93 .87 .81 .48 .63 .42 5.54 .02

 

 

 

 

 

 

 

 

 

 

Following is a brief discussion on the stability of each of the scales/subscales assessed
using the PATH program.

58
Stability of Mother’s Behavioral Demands (DOTS-R)

As expected, the mothers' scores on Activity Level, Approach-Mthdrawal, and Task
Orientation were more stable between Time 1 and Time 2 (the two control points) than they
were between Time 2 (pre-delivery) and Time 3 (post delivery). However, contrary to our
hypothesis regarding the stability of measures, on average mothers’ Flexibility-Rigidity
and Quality of Mood scores were more stable from Time 2 (pre-delivery) to Time 3 (post-
delivery) than they were from Time 1 to Time 2.

For all the DOTS-R subscales, with the exception of Flexibility-Rigidity, the tail
probability of the overall 12 indicated that the data ﬁt the model adequately. One possible
explanation for the failure of the Flexibility-Rigidity data to ﬁt the stability model is that
the errors are correlated rather than unrelated as the model suggests. Note that the
imperfect reliability of the scale would not be a factor here since program PATH corrects

for attenuation.

Stability of Depression Scores (CES-D)

Depression scores strongly ﬁt our expectations. They were highly stable from Time 1 to
Time 2 but were unstable (they decreased) from Time 2 to Time 3. This indicates that the
birth event had a clear impact on the mothers’ depression level. Furthermore, the tail

probability of the overall 12 indicates that the depression data ﬁt the model adequately.

Stability of Anxiety Scores (STAI)

State anxiety scores were more stable from Time 2 to Time 3 than they were from Time 1
to Time 2 and therefore our hypothesis regarding the stability of anxiety scores over Time
1 and Time 2 and their instability between Time 2 and Time 3 is rejected. In contrast, trait
anxiety scores seemed to conﬁrm our stability hypothesis—they were more stable over the

two control time points (Time 1 and Time 2) than they were between Time 2 and Time 3.

59
However, the tail probability of the overall 12 for both the trait anxiety and the state

anxiety subscales indicates that the anxiety data for the sample does not ﬁt the stability
model shown in Figure 1. Again, this misﬁt may be caused by the existence of correlated
errors. This is possible if such errors of measurement as misunderstandings, distractions,
and mood are all caused by one event that is exogenous to the model such as divorce,
moving, marriage, giving birth, change in employment status, or any other stressful life

event.

Change Analysis

Changes in scores from Time 1 to Time 2 (A(1_2)), from Time 2 to Time 3 (A(2.3)), and from
Time 1 to Time 3 (A(1,3)) were analyzed using program WITHIN—a program to compute
the analysis for a within-subjects design (Hunter, 1995). When comparing the mother’s
Activity Level at Time 2 to her Activity Level at Time 3, for example, the effect of birth
registers in a diﬂ‘erence score (A (2-3)) which represents the mean change in the Activity

Level raw scores before and after the treatment (birth). This is measured as follows:
A (24) ACIiVlty Level = ACthity Level; - ACﬁVity Level;

Because the scales are not perfectly reliable, the difference score could be caused by
errors of measurement rather than real changes occurring after birth. The program used the
appropriate formulas to correct for the effects of error of measurement caused by the

imperfect reliability of the scales.

For each of the DOTS-R, DOTS-RzEthnotheory, CES-D, and STAI subscales the mean
change in scores (A) from each point of testing to the others as well as the relevant sample
size (N) are reported in Table 10 through Table 12 below. A positive A for any given
subscale indicates an increase, over the speciﬁed time periods, in the level of that scale

while a negative A indicates a decrease in the subscale’s level over the time periods

speciﬁed. The tables also present the mean, standard deviation (SD), and reliability ((1) of

60

each of the subscales along with the sample size (N) at each point of measurement.

TABLE 10. Observed Mean Changes in DOTS-R Scores

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Tim 1 'r 2 Tim 3

SIID- C llllc ] e A0,”

Scale Statistic N-31 N-31 N=60 lN-31 N-6o N831
Mean 2.65 2.72 2.6 2.65 2.64 .07

'3

3 so .64 .61 .57 .64 .58

e

1% a .86 .88 .83 .92 .87

i; = a,

'3 Mean 2.67 2.73 2.71 2.72 2.73 .06

a

8

g SD .58 .45 .49 .59 .57

B

is

it a .83 .74 .76 .91 .88

E.

a.

< w: :L

:- ﬁBrr 2.34 2.67 2.84 2.75 2.87 -.17

:2

go so .57 .51 .62 .64 .61

5‘

a:

:2

5 a 76 74 a4 86 81

ityofMood

Qual

Task Orientation

 

Mean

3.42

3.28 3.4

l

3.52 3.51

 

SD

.56

.58 .57

.49

 

.91

.89

.89

.87

 

SD

.38 .40

.36

.36

 

 

 

.36

.75 .76

 

.71

 

 

 

-.14

A (24)

N'60

 

l

 

.03

0

    

l

 

A (M)

N=31

 
 

0.0

P—m

.02 .05

0

Mean 2.61 2.58 2.59 2.50 2.49 -.03 -.10 -. ll

 

 

 

61
TABLE 11. Observed Mean Changes in DOTS-RzEthnotheory Scores

 

Sub-Scale Statistic N-60 N-60
Activity Level

 

 

 

 

 

 

 

 

 

 

 

ApproachoWithdrawal

Flexibility-Rigidity Mean 2.97 2.88 -.09
so | .48 | 52 |
a. .40 53

Quality of Mood Mean 2.81 2.73 NE
so | .52 .57
a NE NE

Task Orientation ' Mean 2.47 2.44 -.03
so [ .47 .55
a j -.01 51

 

 

 

 

 

Note: NEB Can Not Be Estimated

TABLE 12. Observed Mean Changes in Maternal Functioning Scores

62

 

 

 

: State Anxiety CBS-D (Depression)

STAI: Thit Anxiety STAI

 

Time 1

 

 

SD

9.69

9.17

9.46

9.06

8.33

 

.91

.91

.91

 

 

 

 

 

 

 

 

 

 

so 10.77 10.24 11.12 9.17 9.22
a .94 .93 .93 .93 .92
WW 52.74 53.37
so 4.48 6.31 5.30 4.23 4.17
a .48 .71 .63 .42 .42

 

A 0-2)
N-31

 

 

 

 

30 .69 1.0

 
 
 

 

 

63
Inspection of Table 10 through Table 12 reveals that, overall, there were no big changes

beMeen any two time points. Consistent with the ﬁndings of the stability analyses in the
previous section, most of the scales/subscales are fairly stable over time. Moreover,
contrary to our expectations, the biggest changes in any given scale/subscale did not
necessarily occur between Time 2 (pre-delivery) and Time 3 (post-delivery).

In addition to mean changes in raw scores, birth-by-subject interactions were measured.
The STG is a statistic which represents, in raw score units, the standard deviation of true
(i.e. corrected for attenuation) gain scores. An STG greater than zero signiﬁes a birth by
subject interaction while an STG equal to or close to zero indicates a lack of such

interaction.

Another question that was investigated is: To what extent can we explain the variation
in effect size by a simple interaction? That is, to what extent does the initial (pre-birth)
level of a variable influence the amount of change occurring in that variable after the birth?
Do mothers who start out high on a any given variable (subscale) experience as much
increase/decrease in the level of that variable after the birth of their infant as mothers who

start out low on that variable?

In order to answer the above questions, the correlation between initial levels and change
scores, the “self impact” correlation (ir), was computed. An ir of zero indicates that the

interaction cannot be explained by considering the initial level of activity.

An -l.0 < it < 0.0 indicates that part of the variation in change scores can be explained
by differences in initial level—the higher the initial level the smaller the change score.
However, since the self impact correlation is not perfect, some moderator variable(s) also
contribute to the observed differences in the response to the birth. In other words, the

interaction is complex rather than simple.

64
And, an fr value of -l .0 indicates a perfect self impact correlation. In this case, variations

in change scores are totally explained by diﬁ‘erences in the initial level.

All the above change analyses yield sample statistics which, due to random sampling
errors, will differ from the population statistics by random amounts. The relationship
between a sample statistic and a population statistic is a matter of probabilities. Rather than
rely on the statistical signiﬁcance test, which has a high error rate when the population
statistic is not zero (Hunter & Levine, 1994), Program CONFINT was used in this study to
produce conﬁdence intervals for each statistic and to calculate those probabilities (Hunter,
1992).

For example, given the sample difference score (A) and the sample size (N), we can

compute the following probabilities for the population difference score (A'):
Inference Probability - P (A: > 0) -= P1
Reverse Probability - P (A: s 0) - Q1
The two probabilities are complementary so that:
Q] 8 1 - PI

However, because of sampling error, it is not possible to draw perfectly correct
inferences. Drawing an inference from data can be ,viewed as placing a bet on a
probabilistic outcome. To accept the hypothesis of a positive difference score, for example,
is to bet that A' > 0. To accept the hypothesis that the effect is not positive is to bet that A'
s 0. Given the sample difference score (A) and the sample size (N), program CONFINT
ﬁrst computes the PI and the Q1 then it converts these probabilities into odds. The odds of
the PI and the odds of the Q1 are computed as follows:

Odds of positive difference 8 PI IQI

Odds of non-positive difference - 01 [PI

65
The Odds Ratio is thus a relative measure of PI and Q1 (Hunter and Levine, 1993).

The PI and odds ratio provide additional information when a directional hypothesis is
used. To give an example: IfPI = .93, then QI = .07 and the odds of A' > 0 are nearly equal
to 13:1. To bet on a positive mean difference in the population is a good bet, but there is an
error rate of 7% for that bet

Note that there is a perfect coordination between whether a one tailed test is signiﬁcant
at the 5% level and whether the P1 is at least 95%. In other words, to ask if a statistic is
signiﬁcant by a one tailed test at the 5% level is exactly the same as asking if the P1 (or the

Q1 in the case of negative prediction) is 95% or more.

Notice also that as the sample statistic gets larger, the PI gets larger. This is because the
larger the sample statistic, the more likely it is that the population statistic is greater than
zero. The maximum value for the PI is .99 rather than 1.0 because 1.0 represents 100%

certainty which is seldom true, and not very realistic, in scientiﬁc research.

A ﬁnal note: Ifthe sample statistic is zero, the sampling error for that sample statistic
was either positive or negative with equal probability. It is equally likely that the
p0pulation statistic was less than zero or more than zero. That is, the PI and the Q1 are both
equal to 50%. Thus, the sample statistic of zero is the dividing line. If the sample statistic
is positive, then PI > 50% and Q1 < 50%. Ifthe sample statistic is negative, then PI < 50%
and Q1 > 50%.

Because we are mainly interested in the changes that occur after the birth of the infant,
only analyses relevant to the changes that occurred from Time 2 and Time 3 are discussed.
The results summarized in Table 13 through Table 15 present the point estimate (size) and
the standard error (SE) for the sample statistics (A, STG, and tr). In addition, conﬁdence

intervals for the p0pulation statistics (A', STG', and ir’) are presented.

66
The tables report the 90% conﬁdence interval, the PI and the odds that the population

statistic is > 0 for all cases where a positive directional hypothesis was proposed. For these
cases, our prediction is conﬁrmed if .66 S PI 5 1.0. On the other hand, we conclude that the
prediction is wrong if PI 5 .33. And, we suspend judgement on whether the prediction is
conﬁrmed or not if the results show that .34 5 P1 5 .65.

In the case of a negative directional hypothesis the 90% conﬁdence interval, the Q1 and
the odds that the population statistic is S 0 are reported Our prediction is considered to be
conﬁrmed if .66 5 Q1 5 1.0. It is considered to be wrong if Q1 S .33. And, we suspend
judgement on whether the prediction is conﬁrmed or not if .34 5 Q1 5 .65.

In the case of a non-directional hypothesis, only the 95% conﬁdence interval is reported
The Pl/QI and the odds ratio are not relevant in this case and the decision to reject or accept

the hypothesis is based on whether zero is included in the conﬁdence interval or not.

Following is a brief discussion on the analysis of change in DOTS-R, DOTS-
RzEthnotheory, and the two matemai functioning measures from Time 2 (pre-birth) to Time
3 (post-birth).

67
TABLE 13. Analysis of Change in DOTS-R From Time 2 to Time 3

 

90%/95%

Confidence
Interval

A .04 .06 4,055 4.50.13 .76 3:1

   

       

A’>0

 

Yes

 

STG .29 .05 0.21 S 870 S 0.38 .99 999.1 STG, > 0 Yes

 

if '21 .18 4.49 S if. S 008 .88 7.1 I", < 0 Yes

A .02 064140075430.“ .64 2:1 A’>0 sr

 

 

val " Activity Level

 

 

E

'U

g src .28 .05 020 5 3m. 5 0.36 .99 9991 my > 0 Yes

15.

E if .1123 427 S if' S 0.49 .32 .521 if, < 0 N02 Wrong
3 direction
:-

 

 

if '.44 .12 “0.65 S if. S _024 .99 ”.1 1” < 0 Yes

FIeaihility-Rigidi

 

 

Mood

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘5

.E'

g I" -.54 .13 _0.75 S if. S _0.34 .99 99.1 if < 0 Y“
3 STG .18 .05 010 S 376' S 0.15 .99 999.1 STG, > 0 Yes
5 i" -.53 .15 4.76 S "A S _029 .99 9”] I!" < 0 Yes
Notes:

NA= Not Applicable

SJ= Suspend Judgement

68
TABLE 14. Analysis of Change in DOTS-R:Ethnotheory From Time 2 to Time 3

 

 

 

 

   

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

90% [95%
Confidence
Interval
-0.18$A‘ $0.12 4'“
E 816 .37 .07 0.26 5 sm 5 0.48 .99 999:1 sro’ > 0 Yes
8‘
g ir -44 .15 4.68 S if s -0.19 .99 999:1 ir’ < 0 Yet
<
2 -0.22 s A' s 0.06 4' '* °
3 .
; STG .29 .07 0.17 S 316' S 0.41.99 999.1 STG > 0 Yes
all
9
g if ~25 .20 "0.58 S ".0 S 008 .90 9.1 If, < 0 Ye!
M
E A -.09 .08 4.24 S A' S 006 NA NA A, ¢ 0 NO
5‘
E
E
A NE NE NE NA NA A’ it 0 NE
E STG NE NE NE NE NE STG’ > 0 NE
‘8
.E‘
3 it NE NE NE NE NE it’ < 0 NE
3 -0.16SA' 50.10 NA 4"‘0
g STG 0.0 .05 4.08 S STG' S 0.08 .50 1:1 STG, > 0 SJ
at it 0.0 NE NE I NE NE ir’ < 0 NE
3
1-
Notes:
NA- Not Applicable
NE= Can Not Be Estimated

SJ= Suspend Judgement

69

TABLE 15. Analysis of Change in Maternal Functioning From Time 2 to Time 3

 

iety CES-D (Depression)

 

STAI: mu Anxiety STAI: State All:

 

90%/95%

Conﬁdence
Interval

 

  

 

 

-0.75 S ir' S -0.46

4.65 54' 5023 ”*0
it -.61 .09 .99 999:1 ir’ < 0 Yes

 

STG

10.4

.96

8.77 S 570' S 11.93

999:]

STG’ > 0

Yes

 

-.65

.08

-0.78 S if S -0.52

999:]

ir’<0

 
 

 

Yes

 

 

 

i, '.86 .12 “1.0 S if. S -'0.66 .99 999.1 if, < 0 Yes

 

 

 

 

 

 

 

 

 

Note: NA= Not Applicable

70

Analysis of Change in Maternal Behavioral Demands

As Table 13 indicates and in support of our hypotheses, on average, after the birth of
their infants the mothers' Activity Level increases, their Task Orientation decreases and
their Quality of Mood changes in the positive direction. Note that, for the Activity Level
subscale, a classical test of signiﬁcance would have failed to reject the null hypothesis and
would have concluded that, on average, birth did not have a signiﬁcant effect on the
mother’s Activity Level. However, the odds ratio indicates an approximate 3:1 chance that

A’ was in the predicted direction.

We predicted that the mothers’ Approach and ﬂexibility would increase aﬁer they had
their babies. However, with a PI of .64 for the Approach-Withdrawal A’ conﬁdence interval
and a PI of .65 for the Flexibility-Rigidity A’ conﬁdence interval, we decided to suspend
judgement on whether our predictions regarding these maternal temperament dimensions
are conﬁrmed or not.

Change analysis also revealed that there is a birth-by-subject interaction for all ﬁve
maternal temperament dimensions. As can be deduced from the large FPS and odds ratios
associated with the conﬁdence intervals for STG’, the birth-by-subject interaction is very

strong.

The next question that was investigated was: To what extent can we explain the variation
in effect size as caused by diﬁ‘erent responses to the birth for mothers at different initial
levels on the DOTS-R dimensions. Results indicated that for all the DOTS-R dimensions
except Approach-Withdrawal, part of the birth-by-subject interaction can be explained by
differences in initial levels of the dimensions. Note that in the case of the Activity Level
impact correlation, a traditional test of signiﬁcance would have failed to reject the null
hypothesis and would have concluded that the initial level of activity has no impact on the

amount of change in Activity Level.

. 71
The ir’s, however, are not perfect which means that the interaction is complex rather

than simple. The interaction cannot be explained by considering only the initial level of
activity. Part of the variation in change scores can be explained by diﬁ‘erences in initial
level—the higher the initial Activity, Flexibility, Quality of Mood and task-orientation
levels, the smaller the change score. However, some moderator variable(s) also contributes

to differences in the response of the mother’s temperament to the birth event.

The Approach-Withdrawal Q1 and odds ratio for the ir’ conﬁdence interval do not lend
support to our prediction that for Approach-Withdrawal ir’ is less than zero. In other words,
the mother’s initial level of Approach-Withdrawal is not negatively correlated to the
amount of change in the mother’s level of Approach-Withdrawal after the birth of the
infant. Since a Q] of .32 (with an odds ratio of .5: 1) implies a PI of .68 (with an odds ratio
of 2: 1), it seems that the opposite of our prediction is true. That is, in the case of Approach-
Withdrawal, ir’ is greater than zero. This means that mothers who start out low on
Approach before their infant’s birth experience smaller increases in their tendency to
approach new situations after the birth of their infants than mothers who start out high on

Approach.

Post-hoe correlation analyses were performed in an attempt to discover any moderators
among the demographic variables. The correlations were all corrected for attenuation due
to the imperfect reliability of the change scores, and are presented in Table 16 below along
with the “No-tailed 95% conﬁdence interval associated with each. The table also presents
the decision reached with respect to the null hypothesis:

[Lap-0

Where p represents the population correlation.

72
Inspection of the correlations presented in Table 16 revealed that the number of birth

complications moderates the change in the mother’s Activity Level from pre- to post-

delivery—the fewer the complications the bigger the change in Activity Level scores.

In addition, the correlations indicate that the mother’s level of education, the degree to
which the pregnancy was planned, and the mother’s income level are moderators of the
amount of change in her Approach-Withdrawal level from pre- to post-delivery. The more
educated the mother is, the more planned the pregnancy was, and the higher the income
level, the less the change in the mother’s Approach-Withdrawal score.

The change in Quality of Mood seems to be moderated by the number of birth
complications and by the degree to which the pregnancy was planned. The larger the
number of birth complications and the more the pregnancy was planned, the smaller the
size of the change in the mother’s Quality of Mood score.

No signiﬁcant moderators emerged for the two remaining DOTS-R dimensions—
Flexibility-Rigidity and Task Orientation.

It is important to note that these correlation analyses are all post-hoe in nature and
therefore their results need to replicated in a different study before we conclude that the

relationships found do indeed exist in the population.

73
TABLE 16. Correlations Between Demographics and DOTS-R Change Scores

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l512-31
Approach-
Variable Withdrawal
r. = -.01 r. = -.15 rll = -.05
rc - -.02 rc - -.27 rc - .03 rc - -.22 r.3 8 -. 10
-27 Sps .25 -.42 Sps .08 -.24 Sps .28 -.40 Sps .10 -.30 Sps .20
Fail to Fail to Fail to Fail to Fail to
rejectHo Niacin. wiectHo miectHo rejectHo
Number of r. = -.25 r. = -.18 r. = -.14 r. = -.25 r. = .03
Birth r,= - -.38 rc - -28 r,= - -.18 r,= - -36 re - .06
Complications -.49 Sps -. 10 -.43 Sps .07 -.39 Sps .11 -.49 Sps -.01 -.22 $5 .28
F Reject Ho Fail to Fail to Reject Ho Fail to
'9in niectHo reject Ho
Education r,| = .15 r. = -.33 rII = .04 r. = -.11 r. - -.12
re - .23 rc = -.52 r, -= .05 rc - -.16 rc - -.24
U -. 10 Sps .40 -.56 Sps -.01 -.21 Sps .29 -.36 S95 .14 -.37 Sps .13
1:. Fail to RejectHo m to Failto Fail to
rejectHo rejectHo reject Ho fejectHo
Degree to T r. = -.03 r. = -.34 r. = -.06 r. = -.35 r. = .18
Which rc = -.05 rc = -.54 rc = -.08 rc - -.51 to = .35
Pregnancy -.28 Sps .22 -.57 Sps -.11 -.31 spS .19 -.57 Sps -.13 -.0‘7 sps .43
Was Planned II Fail to Reject [-1o Fail to Reject 1'1o Fail to
reject“. rejectﬁo rejectHo
Income r. == .01 r. 8 -.25 r. = -.02 r. = -. 12 r. = -.00
rc - .02 r,= = -.40 rc - -.03 r,, = -.17 re - .00
-.25 Sps .27 -.49 Sps -.Ol -.28 Sps .24 -.37 Sps .13 -.26 Sps .26
FF Fail to Reject Ho Fail to Fail to Fail to
Rican. I'dectHo reject“. reject“.
Notes:

r. - Attenuated sarn 1e correlation

rc = Corrected samp e correlation
= PIonulation correlation

1,11,, = all hypothesis

74
Analysis of Change in Maternal Expectational Demands
Results of the change analysis of the DOTS-RzEthnotheory scales are presented in
Table 14, “Analysis of Change in DOTS-RzEthnotheory From Time 2 to Time 3,” on
page 67. Program WITHIN was not able to estimate statistics for the Quality of Mood
subscale due to the fact that it is composed of only one item and therefore its reliability can
not be calculated.

As Table 14 indicates, and contrary to our predictions, on average, after the birth of their
infants the mothers' expectational demands with regards to their infants’ Activity Level,
Approach-Withdrawal, Flexibility-Rigidity, and Task Orientation do not change
signiﬁcantly.

Change analysis also revealed that there is a birth by subject interaction for the Activity
Level, Approach-Withdrawal, and Flexibility-Rigidity expectations. As can be deduced
from the large FPS and odds ratios associated with the STG’ conﬁdence intervals for these
subscales, the birth-by-subject interaction was very strong. Judgement was suspended with
regards to the signiﬁcance of the interaction for the Task Orientation demands. With a PI
of .5 and an odds ratio of 1:1 no conclusions could drawn regarding the population birth by

subject interaction.

The next question that was investigated was: To what extent can we explain the variation
in change scores by differences in initial levels. Results indicate that for the Activity Level
and Approach-Withdrawal demands, part of the birth-by-subject interaction can be
explained by differences in initial scores. Note that in the case of the impact correlation for
the Approach-Withdrawal demands, a naditional test of signiﬁcance would have failed to
reject the null hypothesis and would have concluded that the initial level of Approach
demands has no impact on the amount of change in Approach demands after birth occurs.

75
The ir’s, however, are not perfect Part of the variation can be explained by differences

in initial level—the higher the initial levels of demands for Activity Level, Flexibility,
Quality of Mood and Task-Orientation, the smaller the change score. Some moderator
variable(s) also contribute to variations in the changes in mothers’ expectational demands.
These were explored and will be discussed below.

The Flexibility-Rigidity Q1 and odds ratio for the ir’ conﬁdence interval are borderline
and therefore we suspend judgement with regards to our prediction that for Flexrbility-
Rigidity demands, r'r’ is less than zero. Program WITHIN could not estimate the SE, 95%
conﬁdence interval, Q1, and odds ratio for the task-Orientation demands because the
sample data estimated the standard deviation of true change scores—the STG—to be zero.
If this is true in the population, then the self impact correlation—the ir—is undeﬁned In
any case, estimation of the corrected self impact correlation is unstable for this data. And,

the standard error of r'r can not be estimated using the current method (Hunter, 1995).

Post-hoe correlation analyses were performed in an attempt to discover any moderators
among the demographic variables. The correlations were all corrected for attenuation and
are presented in Table 17 below along with the two-tailed 95% conﬁdence interval
associated with each. The table also presents the decision reached with respect to the null
hypothesis: Ho: p = O. The corrected correlations between demographics and the change in
Quality of Mood demands were not possrble to estimate because the reliability of the
Quality of Mood demands is undeﬁned. Program CORRECT was also not able to estimate
the corrected correlations between demographics and the change in Task Orientation
demands because the reliability of the Task Orientation demands at Time 2 is negative.

Inspection of Table 17 revealed that none of the demographic variables moderate the
change in the mother’s demands. Two of the correlations, however, came very close to

being signiﬁcant. For both the correlation between education and the change in Approach-

76

Withdrawal demands and the correlation between age and the change in Task Orientation

demands, zero was far off the center of the conﬁdence interval.

TABLE 17. Correlations Between Demographics and DOTS-RtEthnotheory Change Scores

 

  
   
    

 

 

  

   

 
  
    

    

 

 

 

 

 

 

 

 

 

 

 

 

 

49:31
Approach- Flexibility- Task
Withdrawal Rigidity Orientation
Age r.8-.l7 r.8-.12 r.8-.l9 r.8-.l6 r.8-.23
r0 8 -.27 r, 8 -.2l rc 8 -.41 r6 8 NE re 8 NE
-.42 SpS .08 -.37 SpS .13 -.44 sps .06 -.41 $5 .09 -.47 SpS .01
Fail to Fail to Fail to Fail to Fail to
8596111. miectHo reject“. reiectHo reiedHo
Number of r. 8 -.04 r. 8 .15 r. 8 -.07 r. 8 -.09 r. 8 -.03
Birth rc8-.06 rc8.27 rc8-.15 rc8NE rc8NE
Complications -.29 Sps .21 -. 10 Sps .40 -.32 Sps .18 -.34 Sps .16 -28 $95 22
Fail to F ail to Fail to Fail to Fail to
rejectH. rejectH. rejectH. rciectH. rejectH.
Education r. 8 .07 r. 8 -24 r. 8 -.10 r. 8 -.14 r. 8 -.09
rc8.ll rc8-.42 rc8-22 rc8NE rc8NE
-.18 Sps .32 -.48 Sps .00 -.35 Sps .15 -.39 Sps .ll -.34 $5 .16
Fail to Fail to Fail to Fail to Fail to
rejectH. reject}; rejectH. rejecm. reieaH.
Degree to r. 8 .05 r. 8 -.02 r. 8 «.01 r. 8 .13 r. 8 -.14
Which rc 8 .08 1',= 8 -.04 rc 8 -.02 re 8 NE r, 8 NE
Pregnancy -.20 Sps .30 -.28 Sps .24 -.27 Sps .25 -. 12 SpS .38 -.39 Sps .11
Was Planned Fail to Fail to Fail to Fail to Fail to
rejectH. rejectn. rejectH. .. rejectn. mien”.
Income r. 8 .02 rll 8 -.06 r. 8 -.15 r. 8 -.09 r. 8 -.04
re 8 .03 re 8 re 8 -.33 rc 8 NE r, 8 NE
~24Sps28 -.3l Sp$.19 -.40$ps.10 -.34 $595.16 -29 Sps 21
Fail to Fail to Fail to Fail to F ail to
rejectHo Rican. Niacin. reject“. reject“.
Notes:
r. 8 Attenuated sample correlation
rc 8 Corrected sample correlation
8 Population correlation
$8 Null h esis
8 Not e to estimate

77

Analysis of Change in Maternal Functioning

As Table 15, “Analysis of Change in Maternal Functioning From Time 2 to Time 3,” on
page 68 indicates, on average, after the birth of their infants the mothers' state anxiety
decreases while their trait anxiety scores do not change signiﬁcantly. Maternal depression
seems to decrease after the birth of the baby. However, the change in depression score,
while very close to being signiﬁcant because zero is close to the upper end of the
conﬁdence interval, is not big enough to support our prediction concerning A’.

Change analysis also revealed that there is a birth by subject interaction for all three
maternal functioning dimensions. As can be deduced from the large PI’s and odds ratios
associated with the conﬁdence intervals for STG’, the birth-by-subj ect interaction is very

strong.

The next question that was investigated was: To what extent can we explain the variation
in the size of the change scores for each maternal functioning dimension as caused by
different initial levels of that dimension? As Table 15 indicates, for all the maternal
functioning measures, a large part of the birth-by-subject interaction can be explained by
differences in initial levels. The ir’s, however, are not perfect which means that the
interaction is complex rather than simple. The interaction cannot be explained by
considering only the initial level of activity. A large part of the variation in change scores
can be explained by differences in initial level—the higher the initial depression, state
anxiety, and trait anxiety levels, the smaller the change score. However, some moderator
variable(s) also contributes to differences in the response of the mother’s temperament to
the birth event. '

Post-hoc correlation analyses were performed in an attempt to discover any moderators
among the demographic variables. The correlations were all corrected for attenuation due
to the imperfect reliability of the change scores, and are presented in Table 18 below along

78
with the two-tailed 95% conﬁdence interval associated with each The table also presents

the decision reached with respect to the null hypothesis: Ho: p = 0.

Inspection of the correlations presented in Table 18 revealed that the number of birth
complications moderates the change in the mother’s depression level from pre- to post-

delivery—the more the complications the bigger the change in the depression level scores.

In addition, the correlations indicate that the mother’s age is a moderator of the amount
of change in her state anxiety level from pre- to post-delivery. The older the mother, the

more the change in the mother’s state anxiety score.

No signiﬁcant moderators emerged for the change in the trait anxiety score from Time 2

to Time 3.

To reiterate a point made earlier, it is important to remember that these correlation
analyses are all post-hoe in nature and therefore their results need to replicated in a
diﬁ‘erent study before we conclude that the relationships found do indeed exist in the

population.

TABLE 18. Correlations Between Demographics and CES-D and STAI Scores

 

 

 

 

 

 

 

 

 

 

 

 

 

A<24)
Variable Depression STAI
State Anxiety Trait Anxiety
Age r. 8 -.01 r. 8 .28 r. 8 -.18
r,= 8 -.01 re 8 .30 r6 8 -.46
-.29 Sps .27 .05 Sps .55 -l .0 Sps .17
Fail to reject H0 ..Reject Ho Fail to reject Ho
Number of Birth r. 8 .36 r. 8 .17 r. 8 -.l9
Complications r, 8 .40 re 8 .18 re 8 -.49
.155pS.64 -.08$ps.45 -l.0$ps.l4
Reject I-io Fail to reject Ho Fail to reject H0
Education r. 8 .08 r. 8 .19 r. 8 -.l4
r.= 8 .09 re 8 .20 r6 8 -.36
-.19 Sps .37 -.06 Sps .46 -1.0 sps 28
Fail to reject 1-10 Fail to reject H0 Fail to reject 1-1o
Degree to Which r. 8 -.07 r. 8 .10 r. 8 -.07
Pregnancy Was Planned re 8 -.08 r.= 8 .11 re 8 -.18
-.36 Sps .20 -.16 59$ .38 -.84 Sps .47
Fail to reject H0 Fail to reject H0 Fail to reject H0
Income r.8-.15 r.8 .09 r.8-.ll
r¢8-.l7 rc8 .10 rc8-28
-.44 $195.11 -.17 sps .37 -.93 Sps .37
Fail to reject H, Fail to reject Ho Fail to reject 11(,
Notes:

r. 8 Attenuated sarn le correlation
rc 8 Corrected sarnp e correlation

8 Population correlation
Fl. 8 Null hypothesis

 

80
Relationship Between Behavioral and Expectational Demands

Calculations of Pearson product-moment correlations provided information regarding the
relationship between the mother’s temperament and her demands both before and after the
birth of the infant The correlations were all corrected for attenuation due to the imperfect
reliability of the DOTS-R and DOTS-RzEthnotheory sub-scales. To test the prediction that
maternal expectational demands would be different from maternal behavioral demands, a
two-tailed 95% conﬁdence intervals was constructed for each correlation coefﬁcient. This
information is presented in Table 19 and Table 20 below.

As can be seen in Table 19, contrary to our predictions, maternal behavioral demands are
positively correlated with maternal expectational demands for Activity Level, Approach-
Withdrawal, and Flexrbility-Rigidity. In other words, the more active, approaching, and
flexible the mother, the higher the levels of activity, approach, and ﬂexrbility she demands
ﬁom her infant

TABLE 19. Correlations Between DOTS-R And DOTS-R Ethnotheory at Time 2

 

   
 
     
    

  

 

 

 

 

 

 

 

 

Dimension for which
Relationship was
Assessed

Activity Level No. Behavioral & expectational
demands are positively related.

Approach-Withdrawal .30 .38 .07 5p: .53 No. Behavioral & expectational
demands are positively related.

ﬂexibiity-Rigidity .30 .46 .07 Sps .53 No. Behavioral & expectational
demands are positively related.

Quality ofMood .10 .ll -.15Sps.35 Yes. Behavioral&expectatioual
demands are not related.

Task Orientation 20 .34 -.04 Sps .44 Yes. Behavioral & expectatimal
demands are not related.

 

 

 

otes:
r. 8 Attenuated sam le correlation
re 8 Corrected sarnp e correlation

81
However, our predictions with regards to Quality of Mood and Task Orientation were

supported The mother’s own Quality of Mood and level of Task Orientation do not appear
to be related to her demands for Quality of Mood and Task Orientation.

The above-stated results only pertain to Time 2 (or pre-delivery) data Information about
the relationship between post-delivery behavioral demands and post-delivery expectational
demands is presented in Table 20. It can be seen that the mother’s Activity Level and her
Task Orientation level are both positively related to her expectational demands for activity
and Task Orientation—the more active and task oriented the mother, the higher the level of

activity and the level of Task Orientation she demands in her infant.

In contrast, and in support of our predictions, Approach-Withdrawal, Flexibility-
Rigidity, and Quality of Mood DOTS-R scores are not related to the equivalent DOTS-
RzEthnotheory scores. Mothers who have positive mood, are flexible, or have a high level
of approach do not necessarily demand that their infants also have high levels of positive
mood, approach, or ﬂexrbility.

TABLE 20. Correlations Between DOTS-R And DOTS-R Ethnotheory at Time 3

 

    
 
    

Dimension for which

  

 

 

 

 

 

 

Relationship was Prediction
Assessed Supported?
Activity level .38 .51 .16 Sps .60 No. Behavioral & expectational
‘ i demands are positively related.
Approach-Withdrawal .22 .30 -.02 Sps .46 Yes. Behavioral & expectational
demands are not related.
Flexibility-Rigidity 23 .40 -.01 Sps .47 Yes. Behavioral & expectational
demands are not related.
Quality of Mood .ll .12 -.14 Sps 36 Yes. Behavioral & expectational
demands are not related.
Task Orientation .25 29 .01 Sps .49 No. Behavioral & expectational
demands are positively related.

 

 

 

 

 

 

 

Notes:
r. 8 Attenuated sam e correlation
rc 8 Corrected sarnp e correlation

82
Absolute Difference Scores and Maternal Functioning

A Pearson correlation was used to test the prediction that larger difference scores
between child temperament and maternal demands, regardless of the direction of the
difference, would correlate positively with higher levels of maternal depression and

anxiety. Two types of absolute discrepancy were examined:

1. Discrepancya: refers to the absolute discrepancy between child tempera-
ment (IBQ scores) and maternal behavioral demands (DOTS-R scores).

2. Discrepancyb: refers to the absolute discrepancy between child tempera-
ment (IBQ scores) and maternal expectational demands (DOTS-
R:Ethnotheory scores).

Results of the correlation analyses are summarized in Table 21 through Table 26 below.
Each table presents the relevant attenuated correlation coefﬁcients (ra) followed by the
corrected correlation coefﬁcients (re). the 90% conﬁdence interval, and the inference

probability (PI) associated with each

In addition, the power associated with each result is presented. Whenever conclusions
are drawn from a set of data, there is a possrbility of making one of two mutually exclusive
types of error: a) Type I error—rejecting a true null hypothesis— or b) Type 11 error—
failing to reject a false null hypothesis. While the probability of a Type I error (a) is
speciﬁed by our choice of signiﬁcance level, the exact probability of a Type I] error (B) is
rmknown The best that can be done is to keep B at a reasonably low value (Koppel and
Zedeck, 1989). The statistical concept of power refers to the probability of correctly
rejecting the null hypothesis. It is related to B as follows:

Power 8 1 - B

83

TABLE 21. Correlations of Absolute Discrepancy Between DOTS-R and IBQ with Depression

 

   
 

      
 

Discrepancy
Dimension

 

90%
Confidence
Interval

   

  
 
  
 

Sample
Size

 

 

 

 

 

 

 

 

 

 

 

 

Activity level .27 .33 .07 sps .47 .99 Yes 72% 127
Approach-Withdrawal .16 .19 ~05 59$ .37 .90 I Yes 35% 400
Flexibility-Rigidity .10 .12 ~. 11 Sps .31 .78 Yes 19% 1055
Quality of Mood .14 .16 ~07 $95 .35 .86 Yes 29% 528
Task Orientation .03 .04 ~. 18 SpS .24 .59 SJ 8% 11933

 

TABLE 22. Correlations of Absolute Discrepancy Between DOTS-R and IBQ with State Anxiety

 

 
    

Dismal-Mr
Dimension

  

 

90%
Conﬁdence
Interval

  
 
    

Sample
Size

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~20 ~24 ~.40 Sps .00 .05 No 0% 248
Approach-Withdrawal .08 .09 ~.13 Sps .29 .73 Yes 15% 1660
ﬂexibility-Rigidity .07 .08 ~. 14 Sps .28 .71 Yes 14% 2175
Quality of Mood .03 .03 ~.18 $95 .24 .59 SJ 8% 11933
Task Orientation 22 .26 .02 Sps .42 .96 Yes 55% 202

 

 

34

TABLE 23. Correlations of Absolute Discrepancy Between DOTS-R and IBQ with Trait Anxiety

 

Discrepancy
Dimension

    

90%

Confidence
Interval

 

 

 

 

 

 

 

 

 

 

 

 

Activity Level -.25 S95 .17

Approach-Withdrawal -.02 -.03 -23 Sps .19 .44 s: 4% 26875
Flexibility-Rigidity .06 .10 -. 15 Sps .27 .68 Yes 12% 2967
Quality of Mood .07 .12 -.14 sps .28 .71 Yes 14% 2175
Task Orientation -.15 .27 -.36 sps .06 .12 No 0% 457

 

TABLE 24. Correlations of Absolute Discrepancy Between DOTS-RtEthnotheory and IBQ with
Depression

 

Discrepancy
Dimension

     

90%

Confidence
Interval

 

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~. 17 Sps .25 SJ 9%
Approach-Withdrawal .05 .06 ~.16 Sps .26 .65 SJ 10% 4282
Flexibility-Rigidity ~.O3 ~.04 ~24 Sps .18 .41 SJ 3% 11933
Quality of Mood .13 .14 ~08 Sps .34 .85 Yes 27% 616
Task Orientation .09 .12 ~. 12 SpS .30 .76 Yes 17% 1307

 

 

 

85

TABLE 25. Correlations of Absolute Discrepancy Between DOTS-RzEthnotheory and IBQ with

 

   

   

 

State Anxiety
90%
Dhcrepancy Confidence
Dimension Interval

 
 

 
 

 

 

 

 

 

 

 

 

 

Activity Level ~. 14 ~.19 ~.35 Sps .07 .14

Approach-Withdrawal ~.09 ~.11 ~.30 Sps .12 .24 No 1% 1307
Flexibility-Rigidity ~.08 ~.11 ~29 Sps .13 .27 No 1% 1660
Quality of Mood ~.06 ~.06 ~27 Sps .15 .32 No 2% 2967
Task Orientation .28 .37 .08 Sps .48 .99 Yes 76% 117

 

 

 

 

 

 

 

 

TABLE 26. Correlations of Absolute Discrepancy Between DOTS-R:Ethnotheory and IBQ with

 

 

 

 

 

 

 

 

Trait Anxiety
90% Sample
Discrepancy . Confidence Prediction Statistical Size
Dimension r2 rc Intervil PI Supported? Power Needed
Activity Level .06 .12 ~.15 Sps .27 .68 Yes 12% 2967
Approach-Withdrawal .06 .ll ~.15 Sps .27 .68 Yes 12% 2967
Flexibility-Rigidity .20 .39 ~00 Sps .40 .95 Yes 48% 248
Quality of Mood ~.04 ~.06 ~25 Sps .17 .38 SJ 3% 6703
Task Orientation ~21 ~.41 ~.41 Sps ~.01 .05 No 0% 223

 

 

 

 

 

 

 

 

 

 

86
For example, if B 8 .40, then power 8 .60. This means that if we repeated the same study

over and over, we would reject the null hypothesis 60% of the time. In other words, the
higher the power, the more ability we have to detect a relationship that exists, in the

population, between two variables.

One well-known way of increasing power is to increase the sample size. Table 21
through Table 26 present the sample size that would be needed in order to have a power of
95% for each of the correlations shown. Following is a brief discussion of the ﬁndings
associated with each of the two types of absolute discrepancy.

Correlations Between Discrepancy. and Maternal Functioning

Results summarized in Table 21 through Table 23 indicate that the absolute discrepancy
between the mother’s Activity Level and the infant’s Activity Level was positively related
to depression but not related to state anxiety. Judgement with respect to the relationship
between an Activity Level discrepancy and trait anxiety was suspended because the P1 was
within the .34-.65 range.

The absolute discrepancy between the mother’s Approach-Withdrawal level and the
infant’s Approach-Withdrawal level was positively related to depression and state anxiety.
Notice that the zero point was approximately in the middle of the interval for both
correlations. A classical test of signiﬁcance would have'failed to reject the null hypothesis
at the .05 level. Judgement with respect to the relationship between an Approach-
Withdrawal Level discrepancy and trait anxiety was suspended because the P1 was within
the .34-.65 range.

The absolute discrepancy between the mother’s Flexibility-Rigidity level and the
infant’s Flexibility-Rigidity level was positively related to depression, state anxiety, and

trait anxiety. Notice that the zero point was approximately in the middle of the interval for

87
all of the estimated population correlations. A classical test of signiﬁcance would have

failed to reject the null hypothesis at the .05 level.

The absolute discrepancy between the mother’s Quality of Mood and the infant’s
Quality of Mood was positively related to depression and trait anxiety. Notice that the zero
point was approximately in the middle of the interval for both correlations. A classical test
of signiﬁcance would have failed to reject the null hypothesis at the .05 level. Judgement
with respect to the relationship between a Quality of Mood discrepancy and state anxiety
was suspended because the P1 was within the .34-.65 range.

Finally, the absolute discrepancy between the mother’s Task Orientation level and the
infant’s Task Orientation level was positively related to state anxiety but not related to trait
anxiety. Judgement with respect to the relationship between a Task Orientation discrepancy
and depression was suspended because the P1 was within the .34-.65 range.

Overall, out of the 15 predictions made, 9. were conﬁrmed, 2 were not supported, and 4

were borderline.

Correlations Between Discrepancyb and Maternal Functioning

Results summarized in Table 24 through Table 26 indicate that the absolute discrepancy
between the mother’s Activity Level and the infant’s Activity Level was positively related
to trait anxiety. Notice that the zero point was approximately in the middle of the interval
for this correlation A classical test of signiﬁcance would have failed to reject the null
hypothesis at the .05 level. On the other hand, the absolute discrepancy between the
mother’s Activity Level and the infant’s Activity Level was not related to state anxiety.
Judgement with respect to the relationship between an Activity Level discrepancy and
depression was suspended because the P1 was within the .34-.65 range.

The absolute discrepancy between the mother’s Approach-Withdrawal level and the
infant’s Approach-Withdrawal level was positively related to trait anxiety. Notice that the

88
zero point was approximately in the middle of the interval for this correlation. A classical

test of signiﬁcance would have failed to reject the null hypothesis at the .05 level. On the
other hand, the absolute discrepancy between the mother’s Approach-Withdrawal level and
the infant’s Approach-Withdrawal level was not related to state anxiety. Judgement with
respect to the relationship between an Approach-Withdrawal discrepancy and depression
was suspended because the P1 was within the .34-.65 range.

Similarly, the absolute discrepancy between the mother’s Flexibility-Rigidity level and
the infant’s Flexibility-Rigidity level was positively related to trait anxiety. Notice that the
zero point was approximately in the middle of the interval for this correlation. A classical
test of signiﬁcance would have failed to reject the null hypothesis at the .05 level. This
discrepancy was not related to state anxiety. Judgement with respect to the relationship
between a Flexibility-Rigidity discrepancy and depression was suspended because the P1
was within the .34-.65 range.

The absolute discrepancy between the mother’s Quality of Mood and the infant’s
Quality of Mood was positively related to depression Notice that the zero point was
approximately in the middle of the interval for this correlation. A classical test of
signiﬁcance would have failed to reject the null hypothesis at the .05 level. On the other
hand, the absolute discrepancy between the mother’s Quality of Mood and the infant’s
Quality of Mood was not related to state anxiety. Judgement with respect to the relationship
between a Quality of Mood discrepancy and trait anxiety was suspended because the P1
was within the .34-.65 range.

Finally, the absolute discrepancy between the mother’s Task Orientation level and the
infant’s Task Orientation level was positively related to both depression and state anxiety.
Notice that the zero point was approximately in the middle of the interval for both of these
correlations. A classical test of signiﬁcance would have failed to reject the null hypothesis

at the .05 level. On the other hand, the discrepancy was not related to trait anxiety.

89
Overall, out of the 15 predictions made, 6 were conﬁrmed, 5 were not supported, and 4

were borderline.

Difference Scores and Maternal Functioning

Exploratory correlation analyses were conducted in order to investigate whether the
direction of a discrepancy has an impact on maternal functioning. A Pearson correlation
was used to test the non-directional null hypothesis that correlations between difference
scores and maternal functioning are not signiﬁcantly different than zero. Two types of

discrepancy were examined:

1. Discrepancyc: refers to the discrepancy between child temperament (IBQ

scores) and maternal behavioral demands (DOTS-R scores).

2. Discrepancyd: refers to the discrepancy between child temperament (IBQ
scores) and maternal expectational demands (DOTS-RzEthnotheory scores).

Results of the correlation analyses are summarized in Table 27 through Table 32 below.
Each table presents the relevant attenuated correlation coefﬁcients (1") followed by the
corrected correlation coefﬁcients (re), the 95% conﬁdence interval, and the power and
needed sample size associated with each. Following is a brief discussion of the ﬁndings

associated with each of the two types of discrepancy.

Correlations Between Discrepancyc and Maternal Functioning

Results summarized in Table 27 through Table 29 indicate that, for 6 out of the 15

relationships investigated, the direction of the discrepancy between infant and mother is
temperament characteristics is important

A discrepancy where the child is higher than the mother on Activity Level, Quality of
Mood, or Task Orientation is related to maternal depression.

90
TABLE 27. Correlations of Discrepancy Between DOTS-R and IBQ with Depression

 

  

Discrepancy Dimension

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~26 ~.32 ~.50 sps ~.02 Reject H0 57% 167

Approach-Withdrawal ~.l4 ~.l7 ~.39 sps .11 Fail to reject 20% 636
H(,

Flexibility-Rigidity .17 20 ~08 $05 .42 Fail to reject 27% 423
Ho

Quality of Mood ~.50 ~.59 ~.69 Sps ~.3l Reject Ho 99% 30

Task Orientation ~.47 ~.58 ~.67 Sps ~27 Reject Ho 99% 36

 

TABLE 28. Correlations of Discrepancy Between DOTS-R and IBQ with State Anxiety

 

  
 

Discrepancy Dimension

  
 

  
  
 

95%
Confidence
Interval

   
 

    

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~.04 ~.05 ~295 Sps .21 Fail to reject
I‘Io

Approach-Withdrawal ~.l8 ~21 ~.43 Sps .07 Fail to reject 2% 311
110

Flexibility-Rigidity ~.04 ~.05 ~29 sps .21 Fail to reject 6% 8075
I-10

Quality of Mood ~27 ~31 ~.51 :95 ~03 Reject 1-10 61% 153

Task Orientation ~23 ~28 ~.47 sps .01 Fail to reject 46% 220
Ho

 

 

 

91

TABLE 29. Correlations of Discrepancy Between DOTS-R and IBQ with Trait Anxiety

 

   

Discrepancy Dimension

  

 

95

Conﬁdence
Interval

 

   

 

 

 

 

Activity Level .14 .25 '~. 11 $95 .39 Fail to reject 20% 636
Ho

Approach-Withdrawal .09 . 15 ~. 16 sps .34 Fail to reject l 1% 1575
1-10

Flexibility-Rigidity ~.04 ~.07 ~29 Sps .21 Fail to reject 6% 8075
Ho

Quality of Mood .42 .71 .21 Sps .63 Reject Ho 97% 50

Task Orientation .35 .62 .13 Sps .57 Reject 1-10 87% 82

 

 

 

 

 

 

 

 

 

TABLE 30. Correlations of Discrepancy Between DOTS-RzEthnotheory and IBQ with Depression

 

       

 

 

 

 

 

 

 

 

 

 

 

95% Sample
Conﬁdence Statistical Size
Discrepancy Dimension r. r Interval Decision Power Needed ‘
* Activity Level -20 ~28 -44 SpS .04 Fail to reject 360% 299 ‘
Ho
Approach-Withdrawal ~. 11 ~.14 ~.36 Sps .14 Fail to reject 14% 1046
1'10
Flexibility-Rigidity .02 .03 ~24 Sps .28 Fail to reject 5% 26875
110
Quality of Mood ~.40 ~.43 ~.61 Sps ~.19 Reject 11.0 96% S8
Task Orientation ~.52 ~.70 ~.71 Sps ~.33 Reject Ho 99% 26

 

 

92

TABLE 31. Correlations of Discrepancy Between DOTS-RiEthnotheory and IBQ with State

Anxiety

 

   
  
  
 

   
 

Discrepancy Dimension

95%
Conﬁdence
Interval

  
 
    

7 Sample

  

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~.35 sps .15 Fail to reject
Ho

Approach-Withdrawal ~.05 ~.06 ~.30 Sps .20 Fail to reject 7% 5159
1-10

Flexibility-Rigidity .10 .13 ~.15 Sps .35 Fail to reject 12% 1055
110

Quality of Mood .02 .02 ~24 SpS 28 Fail to reject 5% 32375
Hu

Task Orientation ~.40 ~.53 ~.61 Sps -. 19 Reject H.o 96% 58

 

TABLE 32. Correlations of Discrepancy Between DOTS-R: Ethnotheory and IBQ with Thait

 

   

   

 
 

  
 
 
 

Anxiety
95% Sample .
Conﬁdence Statistical She
Discrepancy Dimension Interval

 

 

 

 

 

 

 

 

 

 

 

Activity Level ~.08 SpS .42 Fail to reject
it,

Approach-Withdrawal .04 .07 -.21sps.29 Fail toreject 6% 8075
ii,

Flexibility-Rigidity -.17 -33 -.42 sps .08 Fail to reject 27% 423
H,

Quality of Mood 21 .32 -.03 sps .45 Fail to reject 39% 269
H,

rerrt Orientation .52 .99 .33 $5 .71 Rejectir, 99% 26

 

 

 

93
Similarly, a discrepancy where the child is higher than the mother on Quality of Mood

is related to state anxiety. On the other hand, a discrepancy where the mother is higher than
her infant on Quality of Mood or Task Orientation is related to trait anxiety.

Correlations Between Discrepancyd and Maternal Functioning

Results summarized in Table 30 through Table 32 indicate that, for 4 out of the 15
relationships investigated, the direction of the discrepancy between infant temperament

and maternal demands is related to maternal functioning.

A discrepancy where the child is higher than the mother’s expectational demands on
Quality of Mood, or Task Orientation is related to maternal depression. Similarly, a
discrepancy where the infant is higher than the mother’s demands on Task Orientation is
related to higher maternal state anxiety.

On the other hand, a discrepancy where the mother’s expectational demand for Quality
of Mood is higher than her infant’s Quality of Mood score is related to trait anxiety.

Overall, it appears that the absolute discrepancy between maternal behavioral demands
and infant temperament, as well as the absolute discrepancy between maternal
expectational demands and infant temperament is more related to maternal functioning
than the discrepancy between maternal behavioral demands and infant temperament, and
the discrepancy between maternal expectational demands and infant temperament.

Following is a discussion of the methodological issues and the limitations of this study
that may have contributed to the present ﬁndings.

94
Limitations of the study

I . The sample was relatively small, due to the sample attrition associated with
longitudinal research. For example, the birth of a baby may entail moving to
a larger home, hence increasing the chance of loosing subjects. The MSU
Becoming A Mother Study, which is the data set from which data for this
study was obtained, originally recruited 119 families- However, only 60
families completed all waves of questionnaires. It is data from these 60 fam-
ilies that was used in this study. The small sample size reduced the power of
the statistical tests and increased the probability of a Type 11 error.

2. The subjects’ demographic measures had limited variability. For example
the majority of the sample was White, married, educated, from the middle
and upper middle classes, and with very few birth complications. Recruit-
ment for the MSU Becoming a Mother Study was aimed at both low and
middle class communities. However, there were limited resources available
to compensate pmticipants, therefore many women tended to decline our
offer to participate in the study. This restriction of range limits the general-
izability of the results to the population. In addition, caution must be taken
in interpreting the low correlation coefﬁcients obtained in this study, as they
may have resulted from the restriction of the range of values of one or both

of the variables involved.

3. The reliability of the DOTS-RzEthnotheory scales, which are composed of
only two items each, was inadequate. This problem was overcome by cor-
recting for attenuation However, the few number of items comprising each
subscale may not be enough to accurately measure maternal demands, and

thus the validity of the questionnaire for our sample may be questionable.

95

4. Discrepancy scores, which are used to predict outcomes, tend to be less reli-

able than their component scores. Although this study corrected for attenu-
ation, the small variation in the mothers’ expectational as well as behavioral
demands limited the ability of the discrepancy scores to predict maternal
functioning measures. Similar problems of minimal variability associated
with demands has emerged in other studies in this area of research (e.g.,

Hooker et at, 1984; Lerner, 1983; Palermo, 1982; Talwar 1988).

. Possrble reporter bias due to the fact that infant temperament was reported

solely by the mother, who also reports her own temperament and her expec-
tational demands. A description of an infant’s temperament by the mother
will possibly reflect the mother’s subjective perception of the infant so that
a discrepancy between maternal demands and infant temperament may be in
reality a discrepancy between the mother’s perception of her demands and
her perception of her infant’s temperament. Future research should attempt
to address the question of how far an observer would agree with the mother

in rating herself and her infant on temperamental traits.

The purpose of the DOTS-RzEthnotheory was to identify and measure
demands regarding the same underlying dimensions of temperament con-
stantly evident across the age groups (Lerner and Lerner, 1986). Thus, it is
possible that demands regarding some of the temperament attributes of the
infant that play an important role in the mother’s functioning were not
assessed by the DOTS-RzEthnotheory instrument On the other hand,
demands regarding some of the attributes actually assessed by the DOTS-
RzEthnotheory are unlikely to be relevant during infancy and therefore a
discrepancy between these demands and the infant would not be expected to

96
play a key role in maternal functioning. Therefore, a measure of maternal

demands needs to be developed which is sensitive to the dimensions of the

infant temperament that are salient to mother-infant interaction

A discussion of the major ﬁndings of the present study will be presented in
the following chapter.

Chapter 4
Discussion

Individual differences in temperament are associated with variation in personal and social
adjustment (Windle et al., 1986). As such, the role of temperament characteristics in
inﬂuencing the quality of psychological adjustment has been a topic attracting increasing
theoretical interest (Chess and Thomas, 1986; Plomin and Dunn, 1986). The
Developmental Contextual Perspective (Lerner, 1982; Lerner and Kauffman, 1985; Lerner
and Lerner, 1983; Thomas and Chess, 1977) stresses the relationship between the

individual and the context as an important factor affecting functioning.

The “Goodness of Fit” model, devised by Thomas and Chess (1977) and developed
further by Lerner and Lerner (1983) adopts this perspective. According to this model
adaptation derives from a congruence or match between the characteristics of individuality
of the infant and the demands of the social environment, as indexed by the expectations of
key signiﬁcant others with whom the infant interacts, for example the mother. When the set
of organismic features of the infant provides a good ﬁt with the demands of the mother,
positive maternal functioning and better mother-child relationships should be found.

Subsequently, healthy child adjustment should occur.

The present study was designed to closely examine the mechanisms of the “Goodness of
Fit” model. In particular, this study was aimed at appraising a) the relationship between
maternal behavioral demands and maternal expectational demands, b) the relationship
between the infant’s ﬁt with each type of maternal demands and maternal functioning, and
c) whether or not the direction of the discrepancy between the infant’s temperament and

each type of maternal demands makes a difference in terms of the mother’s functioning.

97

98
A detailed list of the hypotheses made in this study and whether the data lent them

support or not is presented in Table 33 below. The major results of the present research will

be discussed in the next sections.

TABLE 33. Summary of the Findings of the Present Study

 

Judgement

 

 

 

 

Hypothesis Hypothesis
I'I thesis Supported For Not supported for suspended for
Mothers’ eXpectational Quality of Mood Activity Level (+)
demands would an Tag Orientati AME]. +
closelymatchtheir on ()H
behavioral demands It Flwbnhty 41181410 (+)
Time 2
Mothers’ expectational Approach-Withdrawal Activity Level (+)
“mm“ WWW,“ Flexibility-Rigidity risk Orientation (+)
closely match their ,
behavioral demands at Quality Of Mood
Time 3
Absolute Discrepancy Activity Level Task Orientation
between maternal -

~W thdr l

behavioral demands Approach ‘ , .awa
and infant tempera- Flexrbrlrty-Rrgrdrty
ment correlated with Quality of Mood
higher Depression
Absolute Discrepancy Approach-Withdrawal Activity Level Quality of Mood
between maternal n H1311 -R1Hd
behavioral demands exr 1W _8‘ "y
ment correlated with
higher State Anxiety
Absolute Discrepancy F lexibility-Rigidity Task Orientation Activity Level
W8“ Man" ° {Mood Approach-W'thdrawal
behavioral demands QM", ° '
and infant temper-a~
ment correlated with
higher Trait Anxiety

 

 

 

 

 

 

TABLE 33 (cont’d)

99

Hypothesis Hypothesis Judgement
Hypothesis Supported For Not supported for suspended for
Activity Level (~) Approach—Withdrawal

Direction of Discrep-
ancy between maternal
behavioral demands
and infant tempera-
ment relevant with
respect to the eﬂ‘ect of
the discrepancy on
maternal Depression

Quality of Mood (~)
Task Orientation (~)

Flexibility-Rigidity

 

Direction of Discrep-
ancy between maternal
behavioral demands
and infant tempera-
ment relevant with
respect to the effect of
the discrepancy on the
mother’s State Anxiety

Quality ofMood (-)

All others

 

Direction ofDiscrep-
ancy between maternal
behavioral demands
and infant tempera-
ment relevant with
respect to the effect of

the discrepancy on the
mother’s Trait Anxiety

Quality of Mood (+)
Task Orientation (+)

Activity Level
Approach-Withdrawal
Flexibility-Rigidity

 

Absolute Discrepancy
between maternal
expectational demands
and infant tempera-
ment correlated with

higher Depression

Quality of Mood
Task Orientation

Activity Level
Approach-Withdrawal
Flexibility-Rigidity

 

 

Absolute Discrepancy
between maternal
expectational demands
and infant tempera-
ment correlated with
higher State Anxiety

 

Task Orientation

 

All others

 

 

 

100
TABLE 33 (cont’d)

 

   
   

    

  
   
  

Hypothesis

 

 

 

 

Hypothesis Judgement
Supported For Not supported for suspended for
Absolute Discrepancy Activity Level Task Orientation Quality of Mood
DCIWOCIITIIIICITIBJ Approach-Withdrawal
expectational demands , .. , , ,
and infant tempera- Flesnbrlrty-Rrgrdrty
ment correlated with
higher Trait Anxiety

 

Direction of Discrep- Quality of Mood (~) Activity Level

“Y W“ mm“: Task Orientation (-) Approach-Withdrawal
expectational demands F1 , , , , , ,

and infant tempera- CXIDlllty-ngldlty
ment relevant with
respect to the eﬂ‘ect of
the discrepancy on
Depression

 

Direction of Discrep- Task Orientation (~) r All others
ancy between maternal-
expectational demands
and infant tempera-
ment relevant with
respect to the effect of

the discrepancy on
State Anxiety

 

Direction of Discrep- Task Orientation (+) All others
ancy between maternal-
expectational demands
and infant tempera-
ment relevant with
respect to the effect of
the discrepancy on Trait
Anxiety

 

Maternal temperament Activity Level F lexibility-Rigidity
Wham“ “mud” Approach-Wthdrawal ° ofMood
relatively stable before , ', WW

birth but shows insta- “3" WWW
bility from Time 2 to
Time 3

 

 

 

 

 

 

TABLE 33 (cont’d)

101

 

Hypothesis

Maternal functioning
relatively stable before
birth but shows insta-
bility from Time 2 to
Time 3

Hypothesis
Supported For

Depression
Trait Anxiety

Hypothesis
Not supported for

State Anxiety

Judgement
suspended for

 

Mother’s Activity
Level, Approach, and
Flexibility would
increase after delivery

Activity Level

Approach-Withdrawal
Flexibility-Rigidity

 

Mother’s Task Orienta-
tion would decrease
aﬁer delivery

Task Orientation

 

Mother’s Quality of
Mood would change
after delivery

Quality of Mood (r)

 

Maternal expectational
demands would change
after delivery

All others

Quality of Mood (NE)

 

Maternal functioning
would change after
delivery

State Anxiety (D)

Depression
Trait Anxiety

 

Age correlated with the
change in all DOTS-R,
DOTS-RzEthnothcory,

and maternal function-
ing subscales

State Anxiety (+)

All others

 

 

Number of birth corn-
plications correlated
with the change in all
DOTS—ll, DOTS-It
Ethnotheory, and
maternal functioning
subscales

 

DOTS-It Activity (~)
DOTS-It Mood (~)
Depression (+)

Allothers

 

 

 

 

TABLE 33. (cont’d)

102

 

   
  

  

Education correlated
with the change in all
DOTS-IL DOTS-R:
Ethnotheory, and
maternal functioning
subscales

Hypothesis
Supported For

  

Hypothesis

Not supported for

Judgement
suspended for

 

 

DOTS~Rz Approach (.)

All others

 

Degree of planning cor~
related with the change
in all DOTS~R, DOTS-
R: Ethnotheory, and
maternal functioning
subscales

DOTS-R: Approach (~)
DOTS-It Mood (~)

All others

 

Income correlated with
the change in all
DOTS-R, DOTS-R:
Ethnotheory, and
maternal functioning
subscales

DOTS-R: Approach (~)

All others

 

A signiﬁcant birth-by-
subject interaction
exists for all DOTS-1L
DOTS-RzEthnotheory
and maternal function-
ing subscales

All others

DOTS-R: Ethnotheory:
Task Orientation

DOTS-R: Ethnotheory:
Quality of Mood (NE)

 

A signiﬁcant impact
correlation exists for
all DOTS-R, DOTS-
lL'Ethnotheory, and
maternal functioning
subscales

 

 

All others

 

DOTS-R: Approach (+)

 

DOTS-R: Ethnotheory:
Flexibility-Rigidity

DOTS-R: Ethnotheory:
Quality ofMood (NE)

DOTS-R: Ethnotheory:
Task Orientation (NE)

 

Notes:

(NE)= Can Not Be Estimated

(~)8 Negative Relationship

(+)8 Positive Relationship
(1)8 Increased

 

103
Relationship Between Behavioral and Expectations] Demands

Overall, the prediction that the mother’s expectational demands would not closely match
their behavioral demands received more support at the post-delivery measurement point
(Time 3) than at the pro-delivery measurement point (Time 2). This lack of relationship
between maternal temperament and maternal expectational demands supports the
“Goodness-of-ﬁt” premise that ﬁt does not require similarity between behaviors because
the mother may demand attributes that are different from her own temperament attributes.
Fit would only mean similarity if the mother’s expectations matched her own temperament.
This was not the case, as far as the post-delivery data was concerned, in this study. This
suggests some desire, on the mother’s part, for complementarity between her own

temperament and that of her infant’s.

It is interesting to note that, whereas before birth the mother’s expectational demands for
Approach and Flexibility were positively related to her own levels of Approach and
Flexibility, after the birth of the baby the mother’s own levels of Approach and Flexibility
seem to be different from her expectational demands concerning these attributes. On
another level, results of the change analyses indicated that maternal expectational demands
did not change from Time 2 to Time 3. On the other hand, aﬁer giving birth, the mother’s
levels of Approach and Flexibility seemed to increase, although, due to low power, the

magnitude of the increase was not big enough to reach signiﬁcance.

By combining these two sets of results, we can conclude that aﬁer giving birth, the
mother seems to have higher levels of Approach and Flexibility while her expectational
demands for these attributes remain constant. This makes sense when we consider the fact
that new mothers taking care of their ﬁrst-bom infants quickly learn to be receptive to new

situations and ﬂexible when dealing with them.

104
Absolute Discrepancy Scores and Maternal Functioning

Overall, results indicated that a discrepancy between maternal behavioral demands and
infant temperament was more related to maternal functioning than a discrepancy between
maternal expectational demands and infant temperament Mothers seemed to suﬂ‘er more
Depression and Anxiety when their infant’s temperament was different than their own, than

when the infant’s temperament was different than their expectations.

It would be interesting to investigate these relationships into childhood and adolescence.
With development, child temperament-demands ﬁt could become less important to mother
functioning. For example, with older children, who do not spend as much time with
parents, ﬁt with maternal behavioral demands may not have as much of an impact on the
mother’s functioning as it does in infancy when the mothers are with their infants for most
of the day. In addition, mothers may adjust to their child’s individuality so that the child’s

being “just like them” or not would not have as much impact on their functioning.

By itself, the discrepancy between expectational demands and infant temperament
seems to be, at best, moderately related to maternal functioning. This contradicts the
available literature that asserts the importance of the ﬁt between the expectational demands
of the signiﬁcant others and the child’s temperament for optimal functioning (Kelly, 1976;
Milliones, 1978; Dunn & Kendrick, 1980; Linn & Horowitz, 1983; Thomas & Chess,
1976; Brazelton et al., 1974; Wolkind & De Salis, 1982).

This unexpected pattern could be attributed to the fact that Time 3 data was collected
only two to three months post-delivery. This period of time may have not been long enough
for the inexperienced ﬁrst-time mother to form a clear picture of what her true
expectational demands might be. This interpretation is supported by the change analysis
which showed that, contrary to what we predicted, maternal expectational demands did not
change much after the birth of the baby. In this case, then, maternal temperament might

have served as a more accurate reflection of the mother’s values and preferences.

105
Direction of Discrepancy and Maternal Functioning

Overall, taking the direction of the discrepancy into account when computing correlation
coefficients between discrepancy scores and maternal functioning did not improve the
chances of finding support for the hypotheses. This lends support to the “Goodness of Fit”
premise (Chess and Thomas, 1977) that a child’s temperament trait has adverse eﬁ‘ects if it
results in a poomess of ﬁt with expectational demands. Whether the child is higher or
lower on the trait than expected by the parents is not necessarily relevant. What matters is

the magnitude or the absolute value of the discrepancy.

A closer look at the results, however, reveals that the direction of the discrepancy does
play a role, in the case of certain temperament attributes, with respect to speciﬁc aspects of
maternal functioning. In particular, mothers of infants who exceeded maternal behavioral
demands for Mood (i.e. the mother’s own level of Mood) had higher Depression and State
Anxiety scores, while mothers whose behavioral demands for Mood (i.e. their own level of

Mood) exceeded their infant’s level of Mood had higher Trait Anxiety scores.

These results are along the same lines of Lerner, Lerner and Zabski’s (1985) ﬁnding that
students who exceeded the teacher’s demands for attention had higher adjustment ratings
than students who showed less attention than demanded. More studies need to investigate
this issue in order to ﬁnd out which attributes are more adaptive when they are on target

and which ones are more adaptive when they exceed or fall short of maternal demands.

A similar pattern emerged with respect to maternal expectational demands for Task
Orientation and speciﬁc aspects of maternal functioning. Mothers of infants who exceeded
maternal expectational demands for Task Orientation had higher Depression and State
Anxiety scores. Perhaps when the highly-orienting infant is distracted away from the
mother for a long period of time, the mother interprets this intense orienting in their
newborns as a sign of alienation or lack of interest, on the part of the infant, in forming an

“attachment” to the mother. And, perhaps, when the infant orients towards the mother for

106
a long period of time, the mother perceives this behavior as a sign of over-dependency on

her that would interfere with her various other daily activities that do not involve the infant.

On the other hand, mothers whose expectational demands for Task Orientation exceeded
their infant’s level of Task Orientation had higher Trait Anxiety scores. This could be
interpreted in light of the fact that infants who cannot orient towards any particular
stimulus for as long a duration as their mothers expected them to, tend to “wear out” their

mothers physically and emotionally.

Stability of Maternal Temperament and Maternal Functioning

In support of our hypotheses, the mother’s Activity, Approach, Task Orientation,
Depression, and Trait Anxiety levels were relatively stable before birth but showed
instability from Time 2 to Time 3. However, contrary to what was expected, Maternal
Flexibility, Mood, and State Anxiety seem to be more unstable between Time 1 and Time
2 than between Time 2 and Time 3. This ﬁnding seems to point to pregnancy as an
important, emotionally-laden period during which ﬁrst-time prospective mothers learn to.

become more ﬂexible when faced with the unavoidable changes in routine.

Change in Maternal Variables After Delivery

After delivery, the mother’s Activity Level increased, while her Task Orientation decreased
reﬂecting the expected changes associated with new motherhood. Judgement was
suspended for changes in the levels of maternal Approach and Flexibility. It seems that
with a larger sample size, and therefore more statistical power, the increase in the mother’s

Approach and Flexibility levels could have reached signiﬁcance.

It is interesting to note that the mother’s Quality of Mood improves after delivery. It
could be that the relief and security felt by the new mothers for having safely delivered

their ﬁrst-horns could contribute to this increase in maternal Quality of Mood.

107
On the other hand, maternal expectational demands did not change after delivery. This

is an interesting ﬁnding, given the fact that most new mothers are expected to adjust their
views and preferences regarding children’s temperament, once they experience having a
child of their own. As stated earlier, one reason for this ﬁnding may be that Time 3 testing
was done too soon after delivery. In addition, it is possible that expectational demands
regarding some of the three-month-old infant’s attributes that are salient to a ﬁrst-time
mother were not assessed by the DOTS~RzEthnotheory scale.

It is important to note that, overall, the functioning of the mothers in our study improved
after delivery—Depression and Trait Anxiety did not change, while State anxiety
decreased. This may reﬂect the fact that our sample was composed of relatively well-

adjusted, educated, married, mature women who experienced few birth complications.

The amount of change in most maternal behavioral demands, maternal expectational
demands, and maternal functioning did not seem to be strongly related to demographic
variables, possibly because of the lackiof variability in the sample’s demographics. The
cases where the relationship was signiﬁcant, however, are interesting to discuss. For
example, as would be expected, the number of birth complications was related to smaller
changes (i.e. smaller increases) in maternal Activity and Mood, and to larger changes in
maternal Depression. Mothers also experienced smaller improvements in their quality of
Mood after delivery when the pregnancy was not planned to begin with. Finally, mothers
with less education, less planning, and lower levels of income experienced smaller
increases in their levels of Approach after delivery. This ﬁts in with the vast amount of
research that suggests a link between life stressors and difﬁculty adjusting to the newborn
and to one’s new role as a mother (Nuckolls et al., 1972; Cohen, 1979; Kaplan, 1982;
Norbeck & Anderson, 1989).

108
Future Research Directions

This study added to the “Goodness of ﬁt” literature because it was the ﬁrst to investigate
the difference/similarity between maternal behavioral demands and maternal expectational
demands and to assess the “infant-behaviors” ﬁt Previous research only assessed the
“infant-expectations” ﬁt which, while important, is not sufﬁcient for understanding the
dynamics underlying maternal and infant functioning. Previous research assumed that
maternal expectational demands accurately represent what mothers really want This study
showed that we cannot always assume that mothers express what they ultimately want
through their expectational demands. Sometimes, investigating maternal behaviors, as well

as expectations, can provide valuable insight into actual maternal demands.

In addition, while the “Goodness of ﬁt” literature indicated that the mother-child ﬁt is
important for child functioning, the saliency of the mother-child ﬁt for maternal
functioning was not investigated before. Crockenberg (1986) has found a relationship
between infant soothability and matem-al responsiveness. That research was built on the
assumption that mothers demand high soothability in their infants and that maternal
responsiveness to the infant is an indicator of maternal functioning. The present study built
on and reinforced these previous ﬁndings by directly assessing maternal demands and

maternal functioning.

Although sample limitations in the present study led us to suspend judgement on a few
questions, and to fail to support some interesting hypotheses, future directions to continue

research are indicated from the present study’s ﬁndings.

First, it is evident that future research should include a much larger sample of
individuals from a more diverse population. This would provide suﬁicient power to detect
existing relationships. Moreover, future research should attempt to use a longer, more

reliable instrument to assess maternal expectational demands. Perhaps, an independent

109
observer should also provide ratings of the mother and of the infant’s temperament traits to

overcome reporter bias.

Further research examining the mechanisms behind the Goodness of Fit model is
suggested by the present study. Future research in the area should include a measure of
maternal behavioral demands and should focus on the discrepancy between these
behavioral demands and functioning, as well as the discrepancy between expectational

demands and functioning.

Thomas et al. (1974) found an association between the child’s ﬁt with the demands of
the physical setting and problem behaviors. For example, high activity level tended to
cause problems for interaction for the Puerto Rican children who live in small apartments,
but not for the core-sample children who typically lived in large houses. This present study
did not assess demands of the physical setting, in order to control the number of
questionnaires the subjects were required to ﬁll out. Future research can look at these

demands and investigate the eﬁ‘ects of child-setting ﬁt on functioning.

Furthermore, examination of both real and absolute discrepancy scores would help
provide more information about the role that the direction of the discrepancy plays in
inﬂuencing outcomes. Finally, future studies should attempt to include several
measurement points at six-month intervals aﬁer delivery. This would allow examination of

the processes that may not be apparent shortly after the birth of the baby.

Only in this way can researchers capture the true dynamics involved in the relationship

between infants’ temperament and their subsequent development.

APPENDIX

Appendix

Instruments Used In The Present Study

1. Demographic Information

2. Dimensions of Temperament Survey-Revised

3. Dimensions of Temperament Survey-Revised: Ethnotheory
4. Infant Behavior Questionnaire.

5. Center for Epidemiological Studies-Depression Scale

6. State-Trait Anxiety Inventory: Form Y.

. Because of copyright laws, these questionnaires are not included in the Appendix

110

Today's date
leak of pregnancy

111

wava:‘_‘_
Tori: ,1,Q,1

DEMOGRAPHIC INFORMATION

 

 

 

 

Due Date
1. Gender: Penal. _ ltala __
2. Age (years)
3. Are you currently employed? rull all.
Part tine ___
What is your occupation?
4. Do you do volunteer work? Please describe brlotly
5. what is the highest level of education you have
completed?
nigh School or loan
Technical or trade school
Some college
College degree
Some graduate study
Graduate degree
6. Marital Status
Single
Cohabitating
Divorced, ranarriod
Married
Separated
Divorced
Widowed
widowed, ranarrlad .
7. Have you over fathered a child before?

Yes
No

How many children would you eventually like to have?

 

Do you have plans to: your auto to work during the

pregnancy?

You, full tine
Yea, part tine '
No ,

112

IDI:

10. Do you plan tor your late to return to work after
the birth of your child?
Yes
No

ll. Were you using birth control when you learned your
late was pregnant?
No
Yes What aethod?

 

12. This pregnancy was
___ expected
___ unexpected
___,planned

13. Do you belong to any organisations or groups?
Yes
No

Please list then below, and indicate your level of
involvement in each

a. Paid nenbership only

b. otticer

c. Involvesent in activities] projects

d. Attend neetings

 

 

 

 

 

 

Native Anerican
Other (specify)

Organizations Level of Involvosent
l.
2.
3.
4.
5.
14. Ethnic background
___ Asian
_ Black
___ Caucasian
___ Hispanic

 

Please indicate what y

deductions and taxes.

a. < 85.000

O. $10,000 -
do 915,000 '
I. 930.000 -
to 325.000 -
g. 830,000 -
he $35,000 -
is 3‘0, 000 -
1. 345,000 -
to 350.000 -
1. $55,000 -
.e "Opooo -
0

14.999
19.999
24,999
29,999
34,999
39,999
44,999
49,999
$4,999
99,999
64,999

113

10!:

our total taaily incoae last.year’was betore
Include tasily incose tron all sources.

IIANK YOU VERY Inca FOR YOUR RARIICIPEIIOﬂllli

114

Please circle the answer that best describes how you generally bebavo.
Please circle gn12_gng_nnnhgg for each state-eat.

Ieally Sort of Sort of Ieally
Pull! PILSI runs - runs
{or ne tor'ne {or no (or ne

1. It takes no a long tine to
get used to a new thing in

ny hone. l 2 3 4
2. I can't stay still for long. 1 2 3 4
3. I laugh and snile at a lot of

things. 1 2 3 4

4. Once I an involved in a task.
nothing can distract an Iron

it. i 2 3 4
5. I keep working at a task until

it's tinished. l 2 3 4
6. l nove around a lot. l 2 3 4
7. I can nake nyselt at hone

anywhere. l 2 3 4

S. I can always be distracted by
something else, no natter what
I nay be doing. 1 2 3 4

9. I stay with an activity for
a long tine. l 2 3 4

10. It I have to stay in one
place for a long tine. I get
very restless. l 2 3 4

ll. 1 usually nove towards new ,
objects shown to ne. 1 2 3 4

12. It takes no a long tine to
adjust to new schedules. 1 2 3 4

13. I do not laugh or snile at
nany things. 1 2 3 4

14.

.15.

16.
17.

18.

19.
20.
21.
22.

23.
24.
25.

26.
27.

28.
29.

115

10!: __._____

Wave:

fornx‘-I:§:1
leally Sort of Sort of Seally
IILSS IILSI llﬂl 2308

for’me for me for'me for'me

If I an doing one thing. some-
thing else occurring won't get
me to stop.

My first reaction is to reject

something new or unfaniliar to
me.

Changes in plans make me
restless.

I often stay still for long
periods of time.

things going on around me can
not take me away fron what I
am doing.

Once I take something up. I
stay with it.

Sven when I am supposed to be
still, I get very fidgety after
a few ninutes. ‘

I am hard to distract.

On meeting a new person I tend
to move towards hin or her.

I smile often.
I never seem to stop moving.

It takes me no time at all to
get.used to new people.

I move a great deal in my sleep

I do not find that I laugh
often.

I move toward new situations.

I move a lot in bed.

9

116

IDS:
Weve:
Porn:

1n

Seally Sort of Sort of Seally
PILSS PILSS SIDS ran:
for’me for'me for me for me

30. In the morning, I an still in
the same place as I was when I
fell asleep.

31. When things are out of place.
it takes me a long tine to
'get used to it.

32. I don't move around much at
all in my sleep.

33. Hy mood is generally cheerful.
34. I resist changes in routine.
35. I laugh several times a day.
36. My first response to anything

2:w is to move my head toward
3?. Generally I am happy.

38. I never seem to be in the same
place for long.

39. I wake up at different times
40. I eat about the same amount
for dinner whether I am home,
visiting someone. or traveling.

4l.I take a nap. rest, or break
at the same times every day.

42.I usually get the same amount
of sleep each night.

43. I seem to get sleepy just about
the same time every night.

44.I get hungry about the same
time each day.

r‘ O4 ta on

p

NNNN

N

UUUU

U

bbbb

.

117

IDS: _ _ _ _

Wave:

form 1 r1
Seally Sort of Sort of leally
PILSS PILSS cans l!!!

for’me for'me for me for’me

4S.When I am»away fromnhome I still
wake up at the same time each

norning. l 2 3 4
46.! eat about the same amount at
breakfast from.day to day. l 2 3 4

4?.I feel full of pep and energy at
the same time each day. l 2 .3 4

4S.I eat about the sane amount at
supper from.day to day. l 2 3 4

49.I wake up at the same time on
weekends and holidays as on other
days 'of the week. I 2 3 4

SO.Ky appetite seems to stay the same
day after day. l 2 3 4

Sl.Ihe number of times I have a bowel
movement on any day varies from

day to day. l 2 3 4
52.1 usually eat the same amount each

day. l 2 3 4
53.I have bowel movements at about

the same time each day. l 2 3 4

S4.Wo matter when,I go to sleep. I wake
up at the same time each morning. I

N
hi
ﬁ

118

 

'hIOd
'03
Porn: [:1
SOW I III! I! CIIIS SO SSSIWS
(lother)
SISSCIIOSS

Ill first tine parents have their rsonal ideas about how they went their
children to behave. In this guest we would like to know what
characteristics you would like to see in your child.

Sere is a sample items

leally Sort of Sect ef Seally
true true Pelee false
Ser'ls Per lb Per’ls Per We

Sons parents want children who read

in their spare time rather than play

games. I 2 3 4

for each itsn. please let us know how true the state-eat is by circling the OS!
response that is most appropriate for you. lemember. we are interested in knowing
what SOC honestly want. not what you think all parents in general went.

Sow.'g:eaee complete the fellow items by circling the CW: response that best

reflects t you think you will want run your CIIID.
Seally Sort of Sort of _ Seally
true true Pulse Pelee
Per In Inc We Perils Per’le
l. Some parents want children who
are very zzzsioally active. whernove
around a . l 2 2 4

2. Some parents want children who do not
approach new situations right awey,
who stay back for a while before
getting involved in a new activity
or routine. I 2 3 4

3. Some parents think that the
appearance and attractiveness of
their children is very important.
they prefer that their children
appear attractive. l -2 3 4

4. Some parents want children who are in
a good mood in all situations. who
express positive emotions so all . .
occasions. I 2 3 4

119

Some ents want children who
guic y shift their attention frcn

one activity to another. WHO 48' 04.417
redirected. l 2 3 4

Some parents prefer that their

‘ children be physically nature. that

7.

10.

they look older than other children the
same age. I 2 3

Some parents want children who keep
an established routine. who will not

easily change their style ef doing
things. I 2 3 4

Some parents want children who

approach new situations right aweyo

who do not hesitate to get involved

in a new activity or routine. I 2 l 4

Some parents want children who
are in a eood only in

positive emotions only on
particular occasions. l 2 3 4

.Some parents want children who

are up; very physically active.
who do not move around a lot. 2 2 3 4

Some parents want children who

are apt to change to a new routine.

who can easily change their way

of doing things. I 2 3 4

Sons parents prefer that their

children be less physically

nature. that they look younger

than other children the same age. I 2 3 4

‘Some parents want children who

remain involved in a certain task

or activity no setter what else

happens. who cannot be easily

redirected. l 2 3 4

120

 

I . D. I
"IV.
Porn W75?

Circle the ”be: for each statasont which best describes how often you
felt at behaved this way-0021M I!!! PAST WIIK.

. Occasionally

lately or Sees or a or a host or
Sons of Little of Moderate All of
the-Iise the Iise Amount of the ﬂea

(loss than 'i'iso
I Day) (1-2 Boyd) ”-4 days) (3-7 Days)

MISC nu ”SI 9212:
l. I was bothered by things

that usually don't bother so. 0 1 2 3
2. I did not feel like eating: 0 1 2 3
my appetite was poor.....
3. Ifelt thetIcouldnoteheke o l 2 3
off the blues oven with help
from sy fsnily or friends.
4. Ifoltthathssjustasgood 0 l 2 3
as other people.........
3. I had trouble keeping ny O l 2 3
cind on what I was doing.
6. I felt depressed........ 0 I 2 3
7. I felt that everything I did 0 l 2 3
was an effort...........
I. I felt hopeful about the future. 0 I 2 3
9. Ithougbtmylifohadboonc 1 2 3
“‘1u.OOOOOOOCOOOOOCOCO
1°. I ‘.t: (WOOOOOOOOOO o 1 z 3
ll. lly sleep was restless... 0 1 2 3
11. 1'“ WOOOOOOOOOOOO. o 1 2 3
13. I talked less than usual. 0 l 2 3
14. I felt lensly........... 0 l 2 3
13. People were unfriendly.. 0 I 2 3
16. I enjoyed life.......... 0 I 2 3
17. I had crying spolls..... 0 l 2 3
1's 1 (.1: II‘.............. o ‘ 2 3
19. I felt that people disliked no. 0' l 2 3
20. I could not get 'going'. 0 l 2 3

REFERENCES

References

Amidon, E., & Flanders, N. A. (1961). The effects of direct and indirect teacher inﬂuence

on dependent-prone students learning geometry. Wyoming,
52, 286-291.

Bates, J. (1980). The concept of diﬂicult temperament MexﬁlLEalthQ, 2h, 299-319.

Bates, J., Freeland, C., & Lounsbury, M (1979). Measurement of infant difficulties. Child.
Development, 5.0, 794-803.

Bates, J. ,Olson, S. ,Pettit, G. & Bayles, K. (1982). Dimensions of Ind1v1dua11ty 1n the
Mother-Infant Relationship at Six Months of Age Childmxtlopmcni 53. 446-461.

Beach, L. R. (1960). Sociability and academic achievement 1n various types of learning
situations. lonmalnﬂﬁducationalﬂxcholagy, 5.1, 208-212.

Bell, R Q. (1974). Contnbutions of human infants to caregiving and social interaction In
hvs. lLewis & L. A. Rosenblum (Eds.),1]Jt.Eﬁ'ects_oflhtlnfanLnnlls.Cmer NY:
1 CY

Belsky, J. (1984). The determinants of parenting: A proccess model. W
5.5,, 83-96.

Brazelton, T. B., Koslowski, B., & Main, M. (1974). The origins of reciprocity in
mother-infant interaction. In M. Lewis & L. A. Rosenblum (Eds), W

' InfanLonItsLategixtt NY: Wiley.
Brophy, A. L. (1959). Self, role, and satisfaction. MW, 59.
263-308.

Buss, A., & Plomin, R. (1975). -
Wiley-Interscience.

Carey, W. B., & McDevitt, S. C. (1978). Stability and chan e m individual tem rament
diagnoses from infancy to early childhood 10W
Childi’sxchiatm. 11, 331-337.

Carey,W. B. ,&McDevitt,S. C. (Ed). (1989). '
temperamenucseamhAm Amsterdam/Lissa Swets & Zeitlinger.

Chess, 8., & Thomas, A. (1982). Infant bonding: Mystique and reality. AmdmnlnumaL
ofmhopsxchiany. 52, 211-222.

Chess, 8., & Thomas, A. (1986). Impommcnﬂnﬂinimmeﬁnt New York: Guilford
Press.

 

Cohen, R. (1979).Ma1adaptation to pregnancy. Walden, 3, 15-24.

Crockenberg, S. (1981). Infant Imtability, Mother Responsiveness, and Social Support
Isrgguggges on the Security of Infant-Mother Attachment Chilthnlopmtm, 52.

Crockenberg, S. (1986). Are temperamental differences in babies associated with
predictable differences in care giving? In J. V. Lerner & R M. Lerner (Eds),

121

122
. WWW San Francisco:

Jossey-Bass.

Crockenberg, S, & Acredolo, C. (1983). Infant temperament ratings. A function of mfants
or mothers, or both? WW1, 6, 61-72.

Crockenberg, S., & Smith, P. (1982). Antecedents of Mother-Infant Interaction and Infant

Irritability 1n the First Three Months of Life. Infaderaxiounddmlopmam, 5,
105-1 19.

Dunn, 1., & Kendrick, C. (1980). Studying Temperament and Parent-Child Interaction:
Com son of Interview and Direct Observation. DexalnpmtntaLMcdicinoand.
,22, 484-496.

Endler, N. S. (1975). A person-situation interaction model of anxrety In C. D. S pielberger
&I. G. Samson (Eds. ), Snessandanxittx Washington, DC: Hemisphere Publ.

Eysenck, H. J ., & Eysenek, S. B. G. (1969). Bennnahmmucmraandmmsuremem. San
Diego, CA: Knopp.
Fish, M. ,& Crockenber ,.S (1981). Correlates and Antecedents of N1ne-Month Infant

Bgehsavior and Mo er-Infant Interaction InfanLBehavioLananelopmm A,
6 - 1

French, J. R. P., Jr., Rodgers, W. & Cobb, S. (1974). Adjustment as person-environment

ﬁtInG. VCoelho, D. A. Hamburg,&J. E. Adams(Eds.), Cnpingandadaptation
(pp. 316-333). New York: Basic Books.

Gesell, A. (1937). Early evidences of individuality in the human infant. Scienchomhlx
45., 217-225.

Goldsmith, H, Buss, A, Plomin, R, Rothbart, M, Thomas, A., Chess, 5., Hinde, a, a -
McCall, R. (1987). Roundtable: What is temperament? Four approaches. Child.
Dexelopmem, 5.8, 505-529.

Goldsmith, H. H, & Ricser—Danner, L. (Ed. ). (1990). W. NY:
Guilford

Harrison, R. V. (1978). Person-environment ﬁt and job stress. In R Payne (Eds), SnessaL
monk (pp. 103-131). New York: Wiley.

I-Iarteré 8.9(1982). The perceived competence scale for children MW 53.
7— 7.

Heise, D. R. (1975). CansaLAnalxsis. New York: John Wiley & Sons.

Hooker, K., Windle, M, & Lerner, R. M (1984). '
‘ - . Unpublished manuscript, The
Pennsylvania State University.

Hunt, J. M. (1961). Intelligencaandexperience. New York: Ronald Press.

Hunter, J. E. (1992). WWW Unpublished manuscript, The
Department of Psychology, Michigan State University.

Hunter, J. E, & Levine, R (1993). WWW. Unpublished
manuscript, The Department of Psychology, Michigan State University.

123

Hunter, J. E, & Schmidt, N. (1993). Wm. Unpublished manuscript,
The Department of Psychology, Michigan State University.

Hunter, J. E. (1994). MW. Unpublished manuscript, The
Department of Psychology, Michigan State University.

Hunter, J. E, & Levine, R (1994). nelnfcmncahohahilim Unpublished manuscript,
The Department of Psychology, Michigan State University.

Hunter, J. E. (1995). WWW. Unpublished manuscript, The
Department of Psychology, Michigan State University.

Jahoda, M. (1961). A social-psychological approach to the study of culture Human.
Relations. .14. 23-30

Kacerguis, MA: (1983)
Pennsylvania State University.
Kaplan, S. R. (1982). Medical school and women over 30. .LAJAJILA. 31, 39-50.

Kelly, P. (1976). The Relation of Infant' 3 Temperament and Mother‘s Psychpathology to
Interactions 1n Early Inﬂmcy. In K. F. Rregel & J. A. Meacharn (Eds) 1111'.
Developinglndividualinaﬁhangingﬂorld Chicago: Aldine.

Kendall, P. C. (1978). Anxiety: States traits-situations? mammmmmm
Psychology, 45. 280-287.

Keppel, G., & Zedeck, S. (1989). WWW New York: W. H.
Freeman and Company.

Klein, P. (1984). The relation of Israeli mothers toward infants in relation to infants'
perceived temperament. ChildDexelopmtm, 55, 1212-1218.

Kohnstarnm,G. A., Bates, J. E. ,&Rothbart,M. K. (Ed). (1989). Win.
childhood. Chichester: Wiley.

Kulka, R. A. (1979). Interaction as - non-environment ﬁt. In L. R Kahle (Edsq) Noni.

. Unpublished dissertation, The

 
 

Lerner, J. V. (1980) I.‘

. O O O
I. 10‘! ".’|| OIIIII'I ‘ I' .I- '9 III‘I 'IC .0 ' ""l
1' ‘.’ ‘ A . I v a ‘ u .

Unpublished doctoral dissertation, The Pennsylvania State University.

Lerner, J. V. (1983). The role of temperament in psychosocial adaptation 1n ear
adolescents: A test of "goodness of ﬁt" model. lonmalofﬁmetic Esyghnlihgy, 143,
149-157.

Lerner, J. V., & Lerner, R M. (1983). Tern cut and ada on accross life:
‘ Theoretical and empirical 1ssues. In 1 B.Baltes &J. G. Brim (Eds) Lifozstnn.

DavelopmeniandBthayiorwp. 197-231). NY: AcademicPress.

Lerner, J. V, & Lerner, R. M. (1986). WWW The
Pennsylvania State University: Unpublished Scale.

 

124

Lerner, J. V., Lerner, R. M, & Zabaski, S. (1985). Temperament and Elementary School
Children' s Actual and Rated Academic Performance: A Test of the 'Goodness of F 1t'

Model. loumaLoﬂChﬂdEsYehologYAndhxehiamzo. 125- 136.

Lerner, R M. (1978). Nature, nurture and dynamic interactionism. Walnut,
2]., 1-20.

Lerner, R. M. (1982). Children £11111 gdglgscents as producers of their own development.
Dertelopmentamexiew, - 7

Lerner, R M. ,& Busch-Rossnagel, N. (1981). Individuals as producers of their
development. Conceptual and empirical bases. In R M. Lerner & N. A.

Busch-Rossnagel (Eds. ),
Life-.SpaoﬂetspeetixeNY NY:Academic Press.

Lerner, R. M., & Kauﬂ‘rnan, M. B. (1985). The concept of development in contextualism.
DexelopmentalReYiew, 5. 309-333.

Lerner, R. M. &,Lemer, J. V. (1987). Children 1n their contexts: A goodness-of-ﬁt model.

an. B. Lancaster, J. Altrnan,A. S. Rossi, &L. R. Sherrod(Eds.), EatenﬁnsAemss.
theLifeSmn(pp 377-404). NY. AldineDeGruyter.

Lerner, R. M. ,,Lerner J. V., Windle, M, Hooker, K. ,.,Lenerz,K &East, P. L. (1986).
Children and adolescents 1n their contexts: Tests of a goodness-of-ﬁt model. In R.

Polmin&J.Dunn(Edsq)IheSmd1Lo£IempemmenLChanges,Conﬁnuities.and
ChallengesI-Iillsdale, NJ: Erlbaum.

Lerner, R. M., Skinner, E. A., & Sorell, G. T. (1980). Methodological implications of
contextual/dialectical theories of development. HummDemlopmenL 23. 225-235.

Lerner, R. M. & Spanier, G. B. (1978). A dynamic interactional view of child and family

development In R. M. Lerner & G. B. Spanier (Eds) ChildJnﬂnenoesotLMmitaL
andEamihLlnteraetinmALifejpanBerspeetixe Y. Academic Press.

Lewis, M., & Lee-Painter, S. (1974). An interactional approach to the mother-infant dyad.

lnM Lewis &L. A. Rosenblum (Eds) WW
Wiley.

Lewis, M., & Rosenblum, L. A. (1974). WWW. NY:
Wiley.

Linn, P, & Horowitz, F. (1983). The Relationship Between Infant Individual Differences
and Mother-Infant Interaction During the Neonatal Period. InfantBohaYioLmd.

DevelopmenLQ4IS-427.

Milliones, J. (197 8) Relationship Between Perceived Child Temperament and Maternal
Behaviors. ChildDeYelopment, 49., 1255-1257.

Norbeck, J. S., & Anderson, N. J. (1989). Psychosocial predictors of pregnancy outcomes
‘ in low-income black, hispanic, and white women. Watch. 3.8, 204-209.

Nuckolls, K. B., Cassel, J. ,.&Kap1an,B H. (1972). Psychosocial assets, life crisis, and the
prognosis of pregnancy AmedeanloumaLopridemiology, 95, 431-441.

Palermo, M. (1982)

 
 

Unpublished dissertation, The Pennsylvania State University.

125

Pervin, L. A. (1967). A twenty-college study of Student 11 College interaction using TAPE
(Transactional Analysis of Personalrty and Environment): Rationale, reliability, and
validity. lmrmalofEdueationalBsYeholoa, 5.8, 290-302.

Peters-Martin, P., & Wachs, T. (1984). A longitudinal study of temperament and its
correlates 1n the ﬁrst 12 months. InfantBehaxioundMelopmont. 2, 285-298.

Pettit, G., & J., B. (1984). Continuity of Individual Differences 1n the Mother-Infant
Relationship from 6 to 13 Months. Childdexdopmem. 55.. 729-739.

Pullis, M., & Cadwell, J. (1982). The Influence of Children's Temperament Characteristics

on Teacher's Decision Strategies. AmericanEducationalReseatehloumal. 19(2),
165-1 81.

Radloff, L. (1977). The CBS-D Scale: A self-report depression scale for research 1n the
general population .loumaloprpliedPsyeholomcalMeasures, 1, 355-401.

Rothbagté M8 (1981 ). Measurement of temperament in infancy. ChildDemlomnm 52,
5 -5 7.

Rutter, M. (Ed). (1977a). IndhddnaLdiﬁ'ezem‘os. Oxford: Blackwell Scientiﬁc
Publications.

Rutter, M. (1977b). Individual differences. In M. Rutter & H. L. (Eds.,) W
inadmppzoaohm (pp. 3-21). Oxford: Blackwell Scientiﬁc Publications.

Scarr, S. & McCartney, K. ( 1983) How peo 1e make their own environments. A theory of
genotype-environment eﬂ‘ects. Chldlgexdopment, 54,. 424-435.

Schneirla, T. C. (Ed.). (1957).
Minneapolis: University of Minnesota Press.

Seligman,M. E. P. (1975). Helplessnesaﬂndepmiondexelopmmanddeath San

Francisco: Freeman.

Smelser, W. T. (1961). Dominance as a factor 1n achievement and perception in cooperative
pgoblem solving interactions. JmmalofAhnonnaLandSoeialPsyehology,
5 5-542

Spielbezger, C. D. (1983). '
o Alto, CA: Consulting Psychologists Press.

Strelau, J. (1983). ImperamentEersonahmAcnmx New York: Academic Press.

Strelau, J. (1987). The concept of temperament in personality research EmopeanloumaL
ofﬁersonalinr, 1, 107-117.

Super, C. M. ,&Harkness, S. (1981). WWW
thememdmdual New York: Academic Press.

Talwar, R. ,Nitz,K., &Lemer, R. M. (1988). Relationsamongmhudoleseent.

.Unpublished manuscript, The Pennsylvania State University, University
Park.

Thomas, A., & Chess, S. (1976). Evolution of Behavror Disorders Into Adolescence.
W, 133(1), 539-542.

 
 

126
Thomas, A., & Chess, S. (1977). IemperamentandDeYelopment NY: Bnmner/Mazel.

Thomas, A. ,& Chess, S. (1981). The role of temperament in the contributions of
individuals to their development. In R M Lerner & N. A. .vBusch-Rossnagel. (Eds),

 
 

Academic Press. ‘

Thomas, A. ,,Chess S. ,Sillen, J., & Mendez, O. (1974). Cross-cultural study of behavror 1n
children with special vulnerabilities to stress. In D. F. Ricks, A. Thomas, & M. Roff
(Eds), WWW Minneapolis MN: University of

Werner, E. E, & Smith,R S. (1982).
Resihentﬁhﬂdtenanderth New York: McGraw-Hill.

Windle, M. (1985). W.
Unpublished manuscript, The Johnson O'Conner Research Foundation, Chicago.

Windle, M., Hooker, K., Lenerz, K., East, P. L., Lerner, J. V., & Lerner, R. M. (1986).
Temperament, perceived wniiaggeé 312111 depression in early- and late-adolescents.
DexelopmentalPsxehology. - .

Windle, M. ,& Lerner, R. M (1986). Reassessing the dimensions of temperament and
individuality across the lifespan: The Revised Dimensions of Temperament Survey

(DOTS-R). .lonmalofAdoleseenLRmcareh, 1, 213-230.

Wolkind, S. N. ,& DeSalis, W. (1982). Infant temperament, maternal mental state and child
behavior prob1ems. In R Porter & G. M. Collins (Eds), Qihanundation.
- . ~ . . r . London: Pitrnan.

 
 

MICHIGAN STATE UNIV. LIBRARIES
11|leWWIWWIWIIHUWW““WWW
31293014201655