«N
— iv ‘
+7.”: wnwflfimwmfim.
.anhfi wwwcx. 152.4
4,1: .. . .V. ....
4&3 .m 3””
Fr... , . V a. n
L 5%., a?

o»! 9.. , . . . . . , , v ,
.é , . ‘ ‘ . , , V a... g.

h

3 ‘, . . _ 1..
fig ,,.,

a
:
t.

z 7

1
fi‘
.5.
’1‘
: .31.“ , . r
i . .. .i
...n.....mz\n,n_z‘n. ..
any? . :8
,r‘ 1

t :5. .L
‘3“. .

 

THESIS

2cm

This is to certify that the

thesis entitled

The Relationship Between Maternal Physical Abuse
Potential and Infant Development in the First Year
of Life

presented by

Laura Nathans

has been accepted towards fulfillment
of the requirements for

M,A. degree in __SJLCh_QJ__Q.Yp 0

 

KW»! QM

Major professor

Date 11!?100

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

 

 

LEBRAR‘?

Michigan State

Unlveratty

 

 

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/Ol cJCIFlC/DataDue.p65«p.15

 

 

THE RELATIONSHIP OF MATERNAL PHYSICAL ABUSE POTENTIAL TO
INFANT DEVELOPMENT IN THE FIRST YEAR OF LIFE
By

Laura Nathans

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF ARTS in
Psychology

2000

ABSTRACT
THE RELATIONSHIP OF MATERNAL PHYSICAL ABUSE POTENTIAL TO
INFANT DEVELOPMENT IN THE FIRST YEAR OF LIFE
By

Laura Nathans

This study examined a sample of 151 mothers and their infants during the infants’
first year of life in order to determine the efl‘ects of maternal physical abuse potential in
the absence of substantiated abuse incidents on infant deveIOpment. Belsky (1993)’s
ecological model for child maltreatment was employed as the theoretical formulation of
child abuse potential-related factors. This model postulates causal influences in
maltreatment situations that stem from individual parent and child traits, familial
interactions, and community supports. Data was collected as part of a home visiting
parenting intervention soon after the infant’s birth (Time 1) and one year later (Time 2).
Factor scores from Milner (1986)’s Child Abuse Potential Inventory (CAPI) as well as a
home visitor report of parenting skills (KCHS Clinical Judgments Form) were used as
measures of the constructs tapped by Belsky’s model, and the Denver 11 (1990) assessed
infant development. Results illustrated significant effects of parent-level individual
factors at all time frames. Also, significant relationships were determined for child-level
individual factors at Time 2, and for change scores between Times 1 and 2. Finally,
significant correlations were found for community-level factors at Time 1. Thus, it was
concluded that individual parent and child traits produce the most salient impacts on

infant development in families with a high potential for maternal physical abuse.

ACKNOWLEDGMENTS

I would first like to thank my committee chair, Dr. Robert Caldwell, for his help, support,
and encouragement throughout this project. I would also like to thank my committee
members—Du. Hiram Fitzgerald and Ellen Whipple—for their valuable comments and
suggestions. Lastly, I would like to thank my friends and family for their constant
encouragement and warmth in helping me grow both personally and professionally

throughout the pursuit of my masters’ degree.

iii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................ vi
CHAPTER 1: INTRODUCTION ...................................................... 1
Overview ............................................................................. 1
Role of Maternal Psychopathology ............................................... 9
Role of Infant Temperament ...................................................... 15
Role of Marital Support ........................................................... 19
Role of External Support Systems ............................................... 22
Role of Ethnicity/Culture ......................................................... 24
Role of Teenage Parenting ......................................................... 27
Kent County Healthy Start Program .............................................. 28
Hypotheses ......................................................................... .30
CHAPTER 2: METHOD ................................................................ .31
Subjects .............................................................................. .31
Measures ............................................................................ .35
Plan of Analysis .................................................................... .38
CHAPTER 3: RESULTS ................................................................ 39
Overview ........................................................................... .39

The Relationship between CAPI Factor Scores and Infant Development..43
The Relationship between KCHS Clinical Judgment Form Scores and

Infant Development ....................................................... 47

The Role of Ethnicity and Age in Potentially Abusive Environments ...... 49
CHAPTER 4: DISCUSSION ........................................................... 51
The Individual Level: Parenting Component .................................. 51

The Individual Level: Child Component ....................................... 55

The Familial Level ................................................................. 56

The Community Level ................................................. . ............ 57

The Role of Ethnicity and Age .................................................... 57
Summary ............................................................................. 58

LIST OF REFERENCES ................................................................. 61

APPENDIX ................................................................................. 69

iv

Table 1:
Table 2:
Table 3:
Table 4:
Table 5:
Table 6:
Table 7:
Table 8:

Table 9:

LIST OF TABLES

Demographic Characteristics of Studied Sample ............................ .32
Descriptive Statistics for Measm‘es Used in this Study ..................... 41
Correlations between Time 1 CAPI Factor Scores and Denver Scores. . .43

Correlations between Time 2 CAPI Factor Scores and Denver Scores. . .44
Correlations between Changes in CAPI Factor Scores and Changes in

Denver Scores from Time 1 to Time 2 ........................................ 46
Correlations between Time 1 KCHS Clinical Judgments Form Scores

and Denver Scores ............................................................... 47
Correlations between Time 2 KCHS Clinical Judgments Form Scores

and Denver Scores ............................................................... 48
Correlations between Changes in KCHS Clinical Judgments Form Scores
and Changes in Denver Scores from Time 1 to Time 2 ..................... 49
Secondary Analysis Correlations for Significant Results from Broad
Denver Scales (Advances, Cautions, and/or, Delays) ........................ 69

C_haDter 1: Introduction

Overview.

A great deal of research in the past three decades has evidenced numerous
detrimental effects of child abuse on children (e.g., Wekerle & Wolfe, 1996). Similar
harms have been demonstrated in abused infants (e. g., Zeanah, Borris, & Larrieu, 1997),
although this time frame is less well-investigated. A specific avenue necessitating further
research is that of the mechanisms through which abuse produces its effects, as it is
unclear whether such harms are engendered by the abuse itself or the environments in
which abusive interactions occur. Further examination of child abuse risk factors’ effects
on developmental pathways of infants would help to elucidate how such trajectories can
be beneficially altered through such interventions as child abuse prevention programs, as
well as more accurately identify factors responsible for such programs’ efficacy. This
study will entail conducting such an investigation with families of infants in their first
year of life at high, risk for becoming child abusers—yet who have not been documented
as abusing. It is expected that parental, child, familial, and contextual/community risk
factors will produce harmful effects associated with abuse, even in its absence.

In the past decade, researchers have begun to explore the detrimental effects of
parental abusive behaviors during infancy. Such investigations are comparatively new in
the field of child abuse research, which has primarily focused on the massive harms child
abuse victims accrue across the developmental traj ectory—particularly in the years
following infancy and toddlerhood (e.g., Aber & Allen, 1987; Friedrich, Einbender, &

Luecke, 1983).

Concerning infancy, sparse yet revealing literature has illustrated that infant abuse
is related to maladaptive infant functioning in the motor, cognitive, and particularly the
social area, where the most dramatic examples of the harmful effects of infant abuse can
be found (Zeanah et al., 1997). For example, maltreated infants have difficulty forming
secure bonds with their parents, evincing disorganized attachments to their mothers, as
well as distress in interactions with them (Zeanah et al., 1997). They also show
internalizing difficulties, such as affective withdrawal, anhedonia, and less positive self-
perceptions, as well as externalizing behaviors, including a greater propensity towards
anger in response to life situations and increased peer-related aggression (Zeanah et al.,
1997). However, the paucity of studies concerning infant abuse and surrounding
environmental conditions—cg, poor parenting skills, depleted community support
networks, and variegated cultural contexts of infant abuse—necessitates further
exploration of these issues. When a more holistic understanding of situational
concomitants of such abuse-engendering situations is obtained, more effective efforts can
likely be undertaken to prevent child abuse as well as developmental harms related to
these dysfunctional environments.

Potential extended effects of physical child abuse are varied, including
psychological, social, and cognitive manifestations. Concerning psychological effects,
many child abuse victims Show such emotional difficulties as post-traumatic stress
disorder, suicidal and destructive behaviors, a low self-esteem, and a more generally
negative outlook towards life events in general (Wekerle & Wolfe, 1996).
Socioemotional deficits are also likely to be present throughout childhood. For instance,

physically abused children frequently Show insecure attachments to mothers; more

avoidance and approach-avoidance behavior with other caretakers; responses to distress
in peers characterized by fear, physical attack, and anger; and hampered capacities to be
sensitive to social cues and discriminate emotions in others (Aber & Allen, 1987;
Wekerle & Wolfe, 1996). In the area of cognitive impairments, physically abused
children have demonstrated significantly lower scores on measure of overall intelligence
(Wekerle & Wolfe, 1996) as well as on more specific verbal skills, such as memory for
words and sentences and ability to recite words (Friedrich et al., 1983). The above
findings paint a bleak picture of future functioning for children reared in abusive
environments.

While it may seem obvious in light of the above evidence that child abuse is
detrimental to healthy child development, the mechanism for such effects is not clear, as
stated above. To further clarify, there are two possible pathways leading to the negative
effects of abusive home environments on children. First, the physical abuse itself may be
responsible for the outcomes outlined above. Second, deficiencies in the ecological
context in which such infants are reared, including poor parenting skills, lack of social
support for parents, a cultural context facilitative of abuse-related behaviors, and
perceived traits of the infants themselves, may be responsible for these results—even if
such conditions do not escalate to the level of actual abuse.

A more comprehensive exploration of the dynamics surrounding abusive
situations—as opposed to merely the abusive incidents viewed in isolation—allows for
elucidation of the means by which abuse influences child development. The
characteristics of abusive individuals and the environments they interact with as well as

engender have been proven to correlate with maladaptive infant social, cognitive, and

motor development (e.g., Calkins & Fox, 1992; Field, 1995; Zeanah et al., 1997),
findings suggestive of significant negative impacts independent of the abuse itself. A
more fme-grained analysis of which specific infant developmental difficulties are
associated with abuse-evocative familial situations is needed to enhance preventive and
treatment approaches. However, a theoretical framework within which to conceptualize
aspects of putatively maltreating contexts is needed through which to conduct such
analyses. Belsky’s ecological model of child maltreatment provides such a formulation.

Throughout his career, Jay Belsky has developed a model of factors involved in
the etiology of child abuse that incorporates several theoretical models. The conceptual
basis for this theory was first outlined in Belsky (1977) and was based on Bronfenbrenner
(1977)’s ecological framework for conceptualizing multiple, interacting levels of
influence on human development. In his initial formulation, Belsky cited three
etiological models for abuse: a) the psychiatric model, which referred to “deep seated
emotional disturbances” (p. 117) present in the abuser that cause him/her to behave in
abusive ways; b) the sociological model, which emphasized both cultural attitudes
towards violence and societal stressors as abuse-engendering; and c) the effect of child on
caregiver model, which figured the effects of child traits perceived to be noxious or
unattractive as being elicitors of parental abusive behaviors. Belsky (1977) argued that as
abuse has been documented in families exposed to stressors inherent in all three of the
theoretical models, a “dynamic interplay between adult, societal, and child characteristics
is at work in the abusive process” (p. 120). For example, Belsky’s model asserts to
examine interactions—such as the positive effects of social support and growth-

enhancing work environments on self-esteem—as opposed to only main effects at each

level (Belsky & Vondra, 1989). It is only when familial risks exceed assets that abuse is
likely (Belsky, 1993; Belsky & Vondra, 1989). Such theorized interactions between
parent, child, and contextual factors have been explored in much greater depth in
Belsky’s later work. The final formulation of this theory—as outlined in Belsky
(1993)—breaks contextual factors down, positing effects at familial, community, cultural,
and evolutionary levels. A discussion of relevant components will help clarify the
theoretical underpinnings of this study.

To begin with, Belsky’s ecological model of child maltreatment addresses the
parent’s role in abusive familial situations as part of the individual—meaning
characteristics manifesting themselves intraindividually—level of his formulation.
Several parental traits empirically validated to be correlated with abusive behaviors have
been factored into the parental component of this model (e.g., Belsky, 1993). First,
parental abuse histories have been cited as being significant predictors of child
maltreatment, as the “intergenerational transmission of abuse” hypothesis has received
some research support (Belsky, 1984, 1993). Parents’ psychological maturity,
complexity, and well-being (e. g., poor self-esteem in such traits as a poor image, greater
levels of aggression/anger, external locus of control, inactive coping style, and
irnpulsivity) are factored into this stage of the ecological model (Belsky, 1984, 1993;
Belsky & Vondra, 1989; Holden et al., 1992). Another individual trait-level factor
incorporated in Belsky’s ecological model of child maltreatment is that of characteristics
of the maltreated children themselves. Belsky (1993) highlighted that—particularly for
physical abuse and neglect—children between the ages of three and eight are most likely

to be maltreated, purportedly due to their mixed dependency and strivings for autonomy

as well as their lack of ability to regulate their emotions. In addition, literature has also
illustrated relationships between abuse and infants and children with “non-optimal”
physical characteristics, such as premature infants, handicapped children, colicky infants,
asthmatic infants, sleep-disordered infants, and infants needing intensive neonatal care
(Belsky, 1993; Culbertson & Schellenbach, 1992; Kaufinan, Johnson, Cohn, &
McCleery, 1992). Moreover, temperamental traits, such as increased initability,
fussiness, dependency, and overactivity have been associated with a greater likelihood for
physical child abuse (Belsky, 1984; Culbertson & Schellenbach, 1992). However, this
body of literature yields inconsistencies, which Belsky (1993) marshaled as supporting
the need to incorporate other levels of the ecological model to determine when such
equivocally-demonstrated risk factors evoke abuse.

A first contextual level that Belsky has hypothesized to play a role in the
maltreating environment is that of the nature of familial interaction as a whole, which
Belsky has termed the “microsystem” (Belsky & Vondra, 1989). Such abuse-
engendering behaviors are demonstrated in several ways, one of which is parent-child
interaction. In general, a lack of adequate parent-infant bonding in the first year has been
found to increase risk for physical abuse (Kaufman et al., 1992). Also, many studies
have demonstrated that physically abusive parent-child dyads are more reciprocally
negative in their interactions (Belsky, 1993). Abusive parents are less supportive of their
children, direct fewer positive behaviors towards them, are less responsive to children’s
initiation of social interactions, have less fim with them in the postpartum period, show
less affection, are more hostile, controlling, and punitive than nonabusive parents

(Belsky, 1993; Kaufman et al., 1992). In addition to parent-child interactions, marital

tensions as well as overall familial negativity levels have been shown to be predictive of
abuse (Kaufman et al., 1992). In sum, Belsky has claimed that such unhealthy
interactions can facilitate child maltreatment if combined with characteristics from other
levels in his model.

In addition to individual and familial variables, the “broader context” in which
abuse occurs is a key level in Belsky (l993)’s formulation. The “broader context” level
of the ecological model consists of a) the community context in which the abusive family
exists (termed the “exosystem”), b) the cultural context in which abusive behaviors occur
(termed the “macrosystem”), and c) the evolutionary context in which the human species
has developed (Belsky, 1993; Belsky & Vondra, 1989). Only the community context will
be discussed, as cultural and evolutionary contextual variables are unilaterally applicable
to all subjects in this study and are thus not variable. Concerning the community context
in which abuse takes places, a great deal of literature has shown that social isolation and a
lack of social support in general are contributory factors to child maltreatment (Belsky,
1993). Such families’ lack of social support has been evidenced in several ways—
through their smaller networks of friends, less fi'equent contact with relatives, less use of
community resources, less involvement in community activities, and smaller likelihood to
own a telephone (Belsky, 1993). Belsky and Vondrda (1989) have shown that less total
support is associated with more punitive child-rearing styles that could lead to abuse.
Moreover, it has been shown that social support through friends in the community
functions to provide guidelines and feedback concerning appropriate parenting behaviors.
Support also leads to better understanding of neighborhood social risks and provides

nurturance that facilitates emotionally responsive child-rearing and a maternal sense of

competence in parenting (Belsky, 1984; Culbertson & Schellenbach, 1992). In sum, it
appears that encapsulating the individual and familial levels of child maltreatment is an
overarching layer of networks families are enmeshed in that interacts with factors above
and below it to engender abuse.

In sum, Belsky’s conceptual model relies upon an examination of a multiplicity of
factors impacting the dynamics of abusive households, postulating influences resonating
at individual, dyadic, familial, community, societal, and evolutionary levels. It is
theorized that it is only when there are more inherent stressors than strengths contained
within the composite of all levels that abuse is likely to be present.

An investigation examining the above-outlined child abuse risk factors’ effects on
infant development is highly relevant to the field of child abuse prevention, which is
geared towards amelioration of putatively problematic life situations inherent in
potentially abusing families. Willis, Holden, and Rosenberg (1992) have highlighted that
it is only through understanding of etiological and risk factors responsible for
development of problems such as child abuse that prevention efforts can be effective. By
clarification of the factors responsible for maladaptive infant development, the routes
through which child abuse prevention programs achieve effects can be better
comprehended and related risk factors more clearly identified. The utilization of a
sample at this age also allows for exploration of the earliest contributory factors to
abusive situations, thereby potentiating researchers’ and social service agencies’ abilities
to derail possibly abusive families’ unhealthy developmental paths very early on.

Prominent researchers have stressed the need for research with this age range, as infants

and young children are disproportionately found to suffer the most severe and fatal abuse
(Culbertson & Schellenbach, 1992).

Child abuse prevention is increasingly being seen as a viable means of addressing
the above-cited difficulties, particularly efforts aimed at families at high risk for abuse
(Wekerle & Wolfe, 1996). Prevention efforts have been viewed as particularly useful
means of tackling such problems as the greater need for and societal cost of mental health
services that maltreated children require (Willis et al., 1992). Such programs aim to alter
the developmental trajectory from “pathological outcome to normative development”
(Willis et al., 1992). These efforts rely upon a growing base of knowledge of risk
indicators for abuse found at individual, familial, and societal levels (Wekerle & Wolfe,
1996) such as those captured by Belsky’s model. Thus, this study will attempt to better
clarify individual, familial, and community-level risk factors’ roles in such situations,
anticipating significant roles for factors at all levels in infant social, language, and motor

development during the first year of life.

Role of Maternal Psychopathology.

Belsky’s ecological model of child maltreatment has implications for the study of
infant maltreatment, as each of its levels contains elements associated in research
literature with specifically infant parenting skills. First, the parenting component of the
individual level of the model can be linked to literature on maternal psychopathology and
infants, as maternal emotional stability/ mental health is conceptualized as being

particularly irnpactful at that level. Evidence that details the relationship between

maternal psychopathology and infant development can first be obtained from research
addressing overall levels of maternal psychopathology.

Scott, Musick, Clark, and Cohler (1983) conducted the seminal study of the
relationship between maternal psychopathology and infant cognitive and social
functioning. In order to obtain a bigger picture of the effects of maternal mental illness,
these researchers grouped several types of mental illnesses together. They examined
women with various psychopathological diagnoses—the mental illness group—as well as
controls in feeding and play situations, in addition to testing their levels of cognitive and
motor development. Results supported that infants of mentally ill mothers displayed
significant cognitive and social, but not motor, deficits. For example, these infants
showed less reciprocal and connected behavior, such as decreased attention-seeking and
responsiveness to mothers, as well as were described as sending less clear and
interpretable social signals than infants with mothers who were considered well. Infants
of mentally ill mothers also evinced affective disturbances, such as more emotional
lability in short periods of time, as well as greater overall negative affect, than infants
with psychologically healthy mothers. Thus, in general, maternal psychopathology
appears to engender immediate, maladaptive consequences for infant cognitive and social
development.

Sameroff, Seifer, and Barocas (1983) found that these deficits are maintained
across the developmental trajectory of infancy. They compared mentally ill mothers with
a variety of psychiatric diagnoses and their infants to nonpathological mother-infant
dyads on measures of motor, cognitive, and social development when infants were

newborn, four months, 12 months, 30 months, and 48 months old. Results supported

10

consistent patterns of deficits in all three major areas of infant development. At four
months, infants of mentally ill mothers displayed more social deficits, such as difficult
temperaments and less adaptive social behavior, as well as cognitive and motor
impairment. At one year, these infants showed less spontaneous social responsiveness, as
well as less overall mobility (a motor deficit). Lastly, at 30 and 48 months, infants with
mentally ill mothers earned lower scores on measures of cognitive development and were
reported to exhibit maladaptive social behaviors. Thus, this study supports stability of
deficits across the developmental traj ectory—with only motor deficits disappearing
between 12 and 30 months.

In addition to the effects of maternal mental illness in general, specific maternal
disorders have been shown to be detrimental to infant social, cOgnitive, and motor
development. First, a great deal of research has been conducted regarding the effects of
depressed mothers’ interactions on their infants (Cohn & Tronick, 1988; Pickens & Field,
1993). This literature has supported that such mothers Show less sensitivity to their
infants and provide them with less stimulation, spending less time looking at, touching,
and talking to their infants (Field, 1995; Zeanah et al., 1997). Depressed mothers also
have been found to exhibit more negative facial expressions and less positive affect while
interacting with their infants, as well as view their infants’ behaviors more negatively
than do more psychologically healthy mothers (Field, 1995; Field, Morrow, & Adlestein,
1993; Zeanah et al., 1997). In addition to socioemotional parenting deficits, research has
demonstrated that these mothers’ speech to their infants, which infants model in their
own language development, is different from that of nondepressed mothers. For

example, depressed mothers make shorter utterances, take longer to respond to infant

ll

cues, fail to adjust their speech to their infants’ speech, and use fewer rising and falling
intonations in interactions with their infants (Bettes, 1988). Overall, depressed mothers’
difficulties are reflected in their perceptions of and social exchanges with their infants.

In turn, infants of depressed mothers manifest social deficits, mirroring their
mothers’ behaviors and affect (Cohn & Tronick, 1988; Pickens & Field, 1993). Such
infants demonstrate lower activity levels and produce fewer vocalizations in interactions
with their mothers than infants of well mothers, as well as look away from their mothers
more, in response to the lack of maternal stimulation and engagement provided them
(Field, 1995; Sameroff et al., 1983). In addition, infants of depressed mothers match
their mothers’ negative affect, spending more time in negative than playful states as well
as less time in positive states overall than infants of nondepressed mothers (Field, 1995;
Field, Healy, & LeBlanc, 1989; Field et al., 1993). Moreover, maternal affect has been
illustrated to frame infant affective states. Cohn and Tronick (1998) and Pickens and
Field (1993) have reported that positive affect causes infants to cycle between affectively
positive and neutral states, as reflected in both facial and behavioral expressions, while
mothers’ depressed affect engenders infant expressions that are patterned alternatively
between neutral and averted gazes. Maladaptive infant social behaviors associated with
difficult infant temperaments have been strongly correlated with maternal negative affect,
as well (Levitt, Weber, & Clark, 1986). Thus, in general, depressed mothers’ emotional
state is harmful to that of their infants in various ways.

These problematic communication patterns result in the formation of disorganized
and insecure attachments between the infant and the depressed mother (Levitt et al.,

1986; Zeanah et al., 1997). Infants who became anxiously attached at some point during

their first 18 months have been found to have mothers who showed such depressive traits
as lacking confidence, irritability, increased negative affect, and decreased interest in
engaging their infants in reciprocal social interactions (Egeland & Farber, 1984; Levitt et
al., 1986). Thus, in general, such infants’ ability to use their mothers as a secure base
from which to explore and master their environment is disrupted by maternal depression.

In addition to the problems infants of depressed mothers have in interactions with
their mothers, such infants utilize the social behaviors learned from these exchanges
when interacting with nondepressed adults (Field, 1995). For example, Field et al. (1989)
found that three- to six-month-old infants of depressed mothers did not vary their
behavior significantly when interacting with mothers vs. strangers. These infants
exhibited similar deficits in both interactions, including problems with vocalization,
fussiness, and decreased physical activity. Moreover, when interacting with depressed
infants, strangers received lower ratings on measures of state, physical activity,
vocalizations, contingent responding, and game playing. Thus, such infants’ social
deficiencies propagate a repetitive cycle of negative interactions across varying social
environments. However, these conclusions must be qualified by the finding that infants
do exhibit less impaired social behavior with familiar nondepressed adults at three
months in such areas as head orientation, gaze behavior, positive facial expressions, and
fussiness (Pelaez-Nogueras, Field, Cigales, Gonzales, & Clasky, 1994).

In addition to its immediate effects on infants’ social interactions, maternal
depression is predictive of infant developmental delays in motor, cognitive, and social
functioning that do not appear until months after mothers begin to exhibit depressive

symptomatology. For example, research has illustrated predictive relationships between

13

maternal depressive symptoms and maternal reports of both externalizing and
internalizing behavior problems in boys at age 3, use of less adaptive behaviors at home
at this age, and infant anxious attachment (Del Carmen, Pedersen, Huffman, & Bryan,
1993; Radke-Yarrow, Cummings, Kuczynski, & Chapman, 1985; Sameroff et al., 1983;
Shaw et al., 1994). Importantly, greater duration of maternal depression has been
determined to enhance the manifestation of cognitive, motor, and social difficulties
(Field, 1995). Predictive relationships have also been shown between levels of overall
depressive and anxious symptomatology and infant anxious attachment (Del Carmen,
Pedersen, Huffman, & Bryan, 1993; Radke-Yarrow, Cummings, Kuczynski, & Chapman,
1985). In sum, maternal depression can result in developmental delays in all areas of the
infant’s functioning, as supported by both objective and subjective perspectives. These
effects appear to be moderated by duration of depressive symptomatology.

Research has also illustrated that others forms of maternal affective illness—such
as heightened levels of maternal anxiety and aggression——are associated with hampered
infant development. For instance, F eldman, Greenbaum, Mayes, and Erlich ( 1997) found
change in maternal trait anxiety levels between three-month and nine-month evaluations
was the most significant predictor of quality of mother-infant interactions. In addition,
Egeland and Farber (1984) determined that changes from secure to insecure attachments
(both avoidant and resistant) between 12 and 18 months were predicted by maternal
social desirability (nonaggression) scores as well as by maternal aggression and
suspiciousness scores. Thus, anxious and aggressive maternal traits predictively

influence infant development.

14

Implications for this study.

In light of all of the above evidence in support of relationships between maternal
psychopathology and infant development, a link between the parenting component of the
individual level of Belsky’s model and infant development can surely be postulated. This

hypothesis will be tested in this study.

Role of Infant Tempera_rne_nt.

A great deal of research has also been conducted that assesses issues relevant to
the relationship between infant development and the child component of Belsky’s
individual level. As child traits that render them hard-to-manage are theorized to be part
of the abusive situation, infant traits’ effects on their own development are relevant to
examine when linking this model to this particular age group. This literature has
supported that infant traits that facilitate parenting stress, such as a difficult temperament,
are related to developmental delays and deficits in infant social and cognitive
functioning. For example, ratings of infant activity level have been shown to be
correlated with how responsive three- to six-month-old infants were in feeding
interactions, as well as to how long interactions were sustained (Hahn, 1989). Thus, the
social stimulation infants received was clearly shaped by their own temperaments.

The link between infant difficultness and problems in infant social functioning is
further elucidated in literature that addresses the relationship between infant temperament
and infants’ abilities to form secure attachments to their mothers. Infant temperamental
traits have been shown to affect responses in such attachment-measuring research

paradigms as the Strange Situation (Calkins & Fox, 1992; Wachs & Desai, 1993). These

15

results are likely found because such characteristics as infant irritability may lead the
infant to experience greater distress at separation and thus desire quick reunion (Calkins
& Fox, 1992). In samples of one-year-olds, maternal temperamental ratings have been
determined to predict attachment classifications in the Strange Situation (Rieser-Danner,
Roggrnan, & Langlois, 1987), as well as specific infant behaviors during the test such as
orientation to people and toys (Braungart & Stifter, 1991). These results have been
replicated with another measure of infant attachment—Waters’ revised Attachment Q-
Sort—thus illustrating that this relationship holds up across different measurement
instruments (Wachs & Desai, 1993). Thus, in general, infant temperament affects the
overall quality of the attachment relationship as well as behaviors fundamental to this
relationship.

In addition to concurrent effects of infant temperament on attachment, research
has also supported predictive associations between these variables. For example,
significantly more infants who have been rated as having a difficult temperament at three
months were also labeled as insecurely attached at one year in a mixed-social-class
Caucasian sample (F rodi, Bridges, & Shonk, 1989). Moreover, specific infant
temperamental traits have been directly linked to their later attachment classifications.
For instance, Del Carmen et a1. (1993) discovered that infant negativity levels at three
months predicted attachment classification at one year in a Caucasian middle-class
sample. Lower autonomic stability and less person orientation at birth, as well as
sensitivity in feeding and play situations and expressiveness at six months, have also been
shown to be significantly correlated with anxious resistant attachment at one year (Belsky

& Rovine, 1987; Egeland & Farber, 1984). In addition, levels of fussiness/irritability at

16

three, six, and nine months as well as a lower activity level at both birth and five months
have been correlated with more insecure attachments overall at one year (Belsky, Rovine,
& Taylor, 1984; Calkins & Fox, 1992; Egeland & F arber, 1984). Changes in
temperament are also predictive of later attachment, as Belsky et a1. (1991) evinced that
infants who decreased in emotionality between three and nine months were more likely to
be classified as insecurely attached at one year. Thus, in general, infant temperament
appears to be strongly related to infant social development and relational attachments in a
predictable fashion.

In order to further support the role that problematic infant traits play in impacting
infant development, the literature that examines the relationship between infant
temperament and maternal behaviors must be explored. Several researchers have posited
that a link between difficult infant temperament and problems in social, cognitive, and
motor functioning is the behavior that infants with difficult temperaments elicit from their
mothers (Calkins & Fox, 1992). Mothers have been shown to provide less social
stimulation to infants with difficult temperaments, both through direct interactions and
provision of appropriate play materials (Hahn, 1989; Klein, 1984). For instance, Klein
(1984) determined that greater perceived levels of difficult temperament effected changes
in mothers’ provision of appropriate stimulation, such as giving infants fewer age-
appropriate toys to play with, being less involved overall with infants, providing them
with less positive social stimulation, and being less responsive to infant cues.
Importantly, the temperamental dimensions assessed in this study resulted in consistently
negative patterns of interaction over time, as results were virtually identical for 6- and 12-

month-old infants. In light of the wealth of evidence cited above, it can be asserted that

17

infants appear to play a significant role in concurrently shaping the interactions they take
part in, as well as the level of care they receive.

The interaction between temperament and maternal sensitivity as well as other
parent-related variables has predictive effects on infant development directly, in addition
to the aforementioned indirect effects through its role in molding growth of parenting
skills. For example, Susman-Stillrnan et a1. (1996) discovered interactive, predictive
relationships for specific levels of infant temperament and maternal sensitivity. The
interaction between low irritability and high maternal sensitivity at three months
predicted attachment status at 12 months, and maternal sensitivity at six months mediated
the effects of irritability at Six months on attachment status at 12 months. Thus, the
researchers concluded that maternal sensitivity develops in response to infant behaviors
such that it becomes the pivotal mechanism through which irritability influences the
attachment relationship. In addition, the interaction between maternal physical contact—
a component of maternal sensitivity subsuming such variables as emotional touch and
physical responsiveness to infant cues—and infant activity level has been determined to
predict infant attachment status at 12 months (Bohlin, Hagekull, Gerrner, Andersson, &
Lindberg, 1989). Overall, the social development literature supports concurrent and
predictive relationships between infant temperament and infant social development.

Infant behavior that influences attachment status has also been shown to
selectively impact infant cognitive development. For example, Klann-Delius and
Hofineister (1997) carried out research evincing that infants rated as securely attached at
12 months produced more vocalizations, as well as responded more quickly to their

mothers’ utterances, in play interactions at between 18 and 36 months. In contrast, when

18

separated from their mothers during the play situation, insecurely attached infants’
language patterns differed from those of securely attached infants, containing more
utterances that showed signs of disruption upon separation from their mothers than did
securely attached infants. Thus, attachment status affects infants’ ability to possess
adequate language Skills and appropriately express distress.

Implications for thl_S_§tudY.

In general, this body of literature pointedly illustrates how infants significantly
contribute to their own cognitive and social development in the first year of life in ways
related to the maladaptive parenting domains factored into the child component of the
individual level of Belsky’s ecological model. This study will test this hypothesis by
determining the relationship between an index of overall infant development and the

child component of the individual level of Belsky’s ecological model.

Role of Marital Supmrt.

Several studies have been conducted that provide evidence that unhealthy familial
interactions in general that contribute to maladaptive parenting are related to detrimental
infant-rearing behaviors. Such literature is relevant to the familial level of Belsky’s
ecological model, which incorporates marital/parenting couples’ dyadic interactions as
part of the overarching abusive situational dynamics. The first major relationship this
literature supports is between the quality of marital interactions in intact families and
infant social development. Positive correlations have been found between marital
conflict and infant and toddler intrusive behavior, as well as conduct problems (Zeanah et
al., 1997). For example, Shaw et al. (1994) explained that for a low socioeconomic status

sample, infant difficulties in social development might manifest themselves only after

19

marital dissatisfaction is present. They discovered that for girls only, low mother ratings
of marital satisfaction at one year predicted maternal ratings of behavior problems on the
Child Behavior Checklist for ages two and three. In addition, Belsky et al. (1991)
reported a relationship between marital conflict and infants’ abilities to show positive
affect in social interactions. Thus, concurrent and predictive relationships exist between
marital conflict and maladaptive infant social behaviors.

Parenting couples’ dissatisfaction and conflict have also been shown to influence
infants’ abilities to form secure attachments with their mothers. A linear relationship has
been determined between increasing infant exposure to parental conflict and infant
behaviors characteristic of insecure attachment (Zeanah et al., 1997). For example,
Goldberg and Easterbrooks (1984) discovered that the less adjusted the married couples
they tested were in terms of agreement, affection, and overall satisfaction with the
marriage, the more likely their 20-month-old infants were to be insecurely attached.
Moreover, the most disorganized attachment behavior has been reported in families
where domestic violence is present (Zeanah et al., 1997). Thus, the greater the level of
marital/parental couples’ conflict in a family, the greater the infant attachment
difficulties.

In addition to parenting couples’ conflict and overall level unhappiness, levels of
support mothers obtain from their partners has been found to relate to infant social
development and attachment status. Zeanah et al. (1994) explained that parents who
support one another and portray a sense of closeness raise infants who show more
positive behaviors in social situations. An example of this effect comes from a study

conducted by Crnic et al. (1984). They examined the relationship between infant social

20

development as exhibited in a dyadic home interaction at 4, 8, and 12 months and
maternal spousal support as measured at one month in a middle-class sample. Findings
highlighted that greater spousal support at one month contributed significantly to
prediction of greater infant responsiveness to mothers, as well as to more positive affect
shown towards them at four months.

This body of literature has also delineated how spousal support impacts infant
social behaviors in the form of attachment status. For example, Jacobson and Frye
(1991) determined that in a lower-class sample, maternal satisfaction with spousal
support immediately prior to the infant’s birth predicted infant attachment ratings at 13
months. Lastly, Spousal support has also been shown to predict mothers’ behavior
towards their infants, which in turn affects infant social development by eliciting and
influencing infant behaviors. For instance, Crnic et al.’s study determined that maternal
spousal support at one month significantly predicted how much positive affect mothers
showed in interactions with their four-month-old infants. In general, this body of
research suggests that infant social development is linked to spousal/partner support
directly, as well as through its effects on the mother.

Implications for this study.

Overall, the predictive relationship between the parenting dyad, and familial
component of Belsky’s ecological model, is well-supported, particularly in the area of
social development. This hypothesis is tested in this study. It adds a dimension to the
findings above, which notably only examined predictive effects of parenting couples’

interactions on infant development. In contrast, this study also looks at concurrent

21

relationships and immediate effects of family problems on infant development, expecting

to discover that such correlations, in fact, exist.

Role of External Support Systems.

The last domain relevant to maladaptive parenting that Belsky’s ecological model
incorporates is that which addresses parents’ levels of external social support.
Researchers have found that lack of an external support system can impede infant
development, particularly in the area of social development. Such literature is applicable
to the community level of Belsky’s model, which postulates a role for social support in
potentiating abusive familial dynamics. Crockenberg (1981) conducted the seminal study
of the predictive relationship between social support and infant social development in the
area of attachment. Crockenberg found that low social support from such individuals as
extended family, friends, and neighbors significantly predicted infant classifications of
anxious-avoidant and anxious-resistant attachment even after the effects of maternal
responsiveness to infants in a play interaction at three months were partialled out. In
addition, low social support mediated the relationship between maternal responsiveness
and security of attachment. In general, these results suggest that the effects of low social
support longitudinally shape the effects of parenting behaviors on infant social
development.

Social support has also been shown affect infant attachment status in interaction
with infant temperament variables. For example, Levitt et al. (1986) found that in a
middle-class sample of mothers and their 13-month-old infants, temperamentally difficult

infants with mothers who reported receiving unsatisfactory social support from their own

22

mothers were likely to be classified as anxious-avoidant. Thus, particular types of social
support can interact with innate infant traits to engender specific typologies of attachment
style.

The finding that infant temperament interacts with social support has also been
illustrated in research that addresses maternal responsiveness to infant behaviors. As
mentioned earlier, these effects can be hypothesized to indirectly affect the infant’s social
development, as greater social stimulation facilitates this development. For example,
Crockenberg and McCluskey (1986), in reanalyzing data from the Crockenberg (1981)
study cited above, found interactions between an irritable temperament at birth and
maternal social support to predict maternal sensitivity at one year. Thus, the
development of parenting skills relevant to handling innately difficult temperamental
traits is impacted by social support.

Social support has also been shown to directly impact maternal behaviors outside
of temperamental variables, in turn affecting infant development. For example, Hahn
(1989) assessed relationships between social support as measured by the Community
Interaction Checklist and mothers’ responsiveness to their three- to six-month old infant’s
cues in a play observation. They determined that mothers were more responsive to their
infants when mothers experienced greater levels of social support. In addition, social
support contributed unique variance to the rating of the overall quality of mother-infant
interaction. Thus, social support directly impacts the quality of stimulation the infant

receives.

23

Implications for this study.

Overall, the above evidence strongly supports the relationship between infant
development—especially in the social sphere—and the social support component of

Belsky’s community level. This study will test this hypothesis.

Role of Ethnicig/Culture.

This study is based on an ethnically and culturally mixed sample of Caucasian,
Hispanic, and African American mother-infant dyads. Thus, it is possible for this
research to examine the potential role of ethnicity and culture in shaping the hypothesized
relationships between Belsky’s ecological model and infant motor, cognitive, and social
development. It is important that any discussion of minority group parenting styles be
prefaced with an acknowledgement of the fact that such parents potentially espouse world
views that differ from that of the majority culture. In effect, infants reared in these homes
are being prepared to function in different familial and social contexts than Caucasian
infants. Therefore, value-laden assessments of the merits of minority parenting practices
should be eschewed. If commonalities are found between minority infant outcomes and
those obtained when examining infants with “likely-to-abuse” parents, it is important that
researchers not draw inaccurate as well as damaging associations fiom such findings or
assume that minorities employ abusive parenting practices. Rather, these results should
be viewed as reflective of parenting conditions inherent in the daily lives of both types of
parents (stressors, parental psychopathology, etc.), conditions that could engender

developmental harms irregardless of culturally-based factors. This study will thus

24

elucidate how racially- and culturally-generated parenting stereotypes may, in fact, be
associated with mediating variables such as those mentioned above.

The similarity between outcomes for infants in the literature alluded to above and
ethnicity’s role in infant development literature is first seen in research conducted by
Sameroff et al. (1983). In their study of general maternal psychopathology cited above,
Sameroff et al. found effects for such variables in interaction with socioeconomic status.
For example, low socioeconomic status Afi'ican-American infants earned lower Bayley
scores—both cognitive and motor—at four months, as well as displayed less social
responsiveness at one year and more maladaptive social behaviors at 30 and 48 months,
than Caucasian subjects. Thus, ethnicity was implicated as potentially impacting infant
development in all major areas across the developmental trajectory.

Research has also been conducted that specifically details how ethnicity, in
interaction with social class, directly affects infant social behaviors, as well as indirectly
affects them through elicitation of unhealthy maternal behaviors. Seifer et al. (1992), for
example, found that minority infants interacted with their environment less positively
than did Caucasian infants. First, when infants were four months old, Afi'ican-American
mothers were less spontaneous in feeding and caretaking situations than were Caucasian
mothers of all socioeconomic classes. These maternal behaviors had direct impacts on
infant social development, as low-SES Afiican-American four-month-old infants were
less spontaneous than Caucasian infants in caretaking situations, yet were more
spontaneous in situations where they were distant from their mothers. At 12 months of
age, similar effects were exhibited, with African-American infants being less verbally

responsive in caretaking situations, as well as showing less ability to actively engage their

25

environments in the presence of their mothers, in comparison with Caucasian infants.
Thus, racial variables differentiated quality of mother-infant interactions. However, this
study will examine if such delineations were secondary effects of the parenting
deficiencies cited in the above literature review that are not etlmically-based.
Ethnic/cultural variables have also been illustrated to affect infant development,
engendering similar outcomes to those found in the literature cited in other sections of
this paper. For instance, Field and Widmayer (1981) rated a wide range of mother-infant
dyadic interaction variables in a sample of Cuban, South American, Puerto Rican, and
Afiican-American three-to-four-month old infants and their mothers. Results showed
differences between infants from different ethnicities that paralleled interaction style
differences their mothers exhibited. For example, discrepancies in the amount of time
mothers spent speaking to their infants were found between cultures, with Cuban mothers
vocalizing the most, followed by South American, Puerto Rican, and lastly African-
American mothers. Mothers’ mean length of utterances was rank-ordered in the same
order. Interestingly, infant gaze scores were in the reverse order—with infants who were
spoken to the most averting gazes the most. Thus, infants’ social behaviors were directly
tied into cultural differences——with African-American infants once again earning lowest
scores. Such a pattern was also exemplified in direct mirroring of rankings for overall
physical activity of mothers and infants during interactions as well as quality of feeding
interactions—with both Puerto Rican mothers and infants earning highest scores,
followed by Cuban, then South American, then Afiican-American infants. This study is
illustrative of how cultural differences—in addition to the racial variables cited above—

may affect the manner in which infants develop.

26

Mtions for this studv.

Overall, differences in parenting outcomes found in both Seifer et al. (1992)’s and
Field and Widmayer (198 l )’s studies may simply reflect overarching parenting problems
characteristic of specific parenting conditions endemic to these cultures’ situational
challenges. This study will assess this hypothesis, as well, purporting that the above

theorization will be upheld.

Role of Teenage Parenting.

As was the case for infants from minority-status cultural backgrounds, literature
examining infants born to teenage parents has supported that such infants experience
similar developmental difficulties to those of parents with a strong likelihood to become
physical child abusers (Elster, McAnarney, & Lamb, 1983). These findings once again
beg the question of whether the adolescent parents’ ages are the primary contributing
factor in engendering infant developmental harms—or if such effects are due to general
parenting deficiencies that create an unhealthy child-rearing environment regardless of
parents’ ages (as is hypothesized for ethnic/cultural effects). As this study examines
comparable samples of adolescent and adult mothers and their infants, it allows for this
question to be explored.

As was described for infants of high-abuse-potential parents, recent research has
shown that infants of adolescent mothers experience developmental delays and deficits in
social development. Concerning social deficits, studies have illustrated that such infants
are more likely to be insecurely attached to their mothers (Elster et al., 1983; Teberg,

Howell, & Wingert, 1983) as well as initiate less social contact with them (McAnarney,

27

Lawrence, Ricciuti, Polley, & Szilagyi, 1986), at around one year of age. In addition,
Teberg et al. (1983) reported that these infants are less likely to maintain exploratory
behaviors, show decreased activity levels, and have a restricted range of available search
behaviors in mother-infant play interactions between 12 and 19 months. Similarly,
Mercer, Hackley, and Bostrom (1984) discovered that 12-month-old infants born to teen
mothers Show less advanced social development overall than do infants of the same age
with mothers in their twenties.

Implications for this study.

In sum, a wide array of problems in establishing and maintaining positive social
interactions have been observed for infants of teenage mothers—difficulties that parallel
those described for parents likely to be physical child abusers that thus could result from
similar causative processes. This study will test this hypothesis. However, whether
cognitive and motor deficits noted during the first year of life result from age-related
behaviors or more generally-employed inappropriate child-rearing strategies has not been
examined prior to this study. This investigation will test the proposition that such effects

result from maladaptive parenting techniques.

Kent County Healthy Start Prom1KCHS).

Recruitment for this project was accomplished by approaching a random selection

 

of new mothers at a major metropolitan hospital to see if they would be interested in
program participation. Assenting mothers were screened and placed in either high- or
low-risk categories following risk assessment procedures. High-risk mothers were

randomly assigned to intervention or control groups. Family support workers collected

28

data for this project beginning immediately after the families’ enrollment in the program
starting in June of 1995. This wave of data collection constituted what will be termed
“Time 1” throughout the rest of this paper. These data were collected to provide a
benchmark of what the family was like prior to intervention. One year after initial
testing, data were collected again to determine intervention effects. This wave of data
collection will be referred to as “Time 2” throughout the rest of this paper. At both Time
1 and Time 2, self-report measures of parental functioning such as the Child Abuse
Potential Inventory (CAPI) were collected to quantify parental functioning. In addition,
as part of family social workers’ evaluations of families’ progress, the Kent County
Healthy Start Clinical Judgments form was developed. This measure was used to provide
corroboration of parenting abilities as assessed by the CAPI from an outside observer’s
perspective, as many of the above studies did not assess perspectival divergence between
subjective and objective reporting of infant behaviors (e.g., Shaw et al., 1994). Solely
subjective examinations rendered results vulnerable to such biases as the maternal
psychopathologies cited above, prior parental history with the infant, and past child-
rearing experiences in general. Thus, the comparative perspective provided by objective
observer ratings remedies this bias.

For purposes of this study, analyses were collapsed across control and
intervention groups because no intervention effects were found in preliminary data
analyses. The following results are a secondary data analysis. In order to assess both
parent and infant outcome variables, assessments were conducted at both the first and

second visits and were used in this study.

29

Hypotheses.

Hypothesis 1: a.) The number of problems in infant development will be positively and
significantly related to individual, familial, and community-level measures of parent and
child dysfunctionality, both in the months immediately following the infant’s birth and a
year after this initial testing. b.) Increases in infant developmental problems between the
months immediately following the infant’s birth and a year after this initial testing will be
significantly correlated with increases in these same measures of dysfunctionality.
Hypothesis 2: a. More problems in infant development will be significantly related to
higher scores on the KCHS Clinical Judgments form—an observational measure of
parent-related individual factors. b. Increases in infant developmental problems between
time one and time two will be significantly correlated with increases in scores on the
KCHS Clinical Judgments form.

Hypothesis 3: Any significant relationships between race/culture and parents’ age
(teenage vs. adult), and infant development scores at times one and two—as well as
changes in these scores between the months immediately following the infant’s birth and
a year after this initial testing —will be an artifact of the relationship between measures

of all components of Belsky’s ecological model and infant development scores.

30

fighter 2: Metho_d

Subjects.

Data obtained from the Kent County Healthy Start Evaluation Project was utilized
for this study. The study tested a sample of 151 infants and their mothers, yielding a total
of 302 subjects altogether. The sample was limited to those families for whom data was
collected at both data points. Demographic characteristics of the sample are presented in

the following table:

31

Table l: Demomphic Characteristics of Studied Sample

Percent

Caucasran
Amencan

Status at Time 1:

Status at Time 1:
loyed
U

Status at Time 1:
o

fest.) at Time 1:

< ,000
5, - 4,
,000 - 34,

9

>

 

32

Table 1, continued: Demogpgphic Characteristics of Stpdied Sample

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Freguency Percent Sample Size (N)
MoLher’s Marl—'ta!
Status at Time 2: 139
Married 55 39.6
Single 78 56.1
Divorced 3 2.2
Separated 3 2.2
Mother’s Job
Status at Time 2: 139
Employed 65 46.8
Unemployed 74 53.2
Father’s Job
Status at Time 2: 121
Employed 96 79.3
Unemployed 25 20.7
Household Income
at Time 2 (est): 140
< $15,000 64 45.7
$15,000 - $24,999 33 23.6
$25,000 - $34,999 14 10.0
$35,000 - $49,999 5 3.6
> $50,000 13 9.3
Mean Standard Sample Size (13)
Deviation
Mother’s Age at
Time 1 24.24 5.71 140
Father’s Age at
Time 1 28.20 6.92 113
Years of Education
aLt Timgl—Mother 11.24 2.33 133
Year of Education
flTifime l—thgp 12.07 2.57 106
Number of
Children at Time 1 2.09 1.20 135
Infant’s Age in
Mos. at Time 1 5.40 4.32 150

 

 

33

 

Table 1, continued: Demographic Characteristics of Studied Sample

 

 

 

 

 

 

 

 

 

 

Mean Standard Sample Size (I!)
Deviation

Years of Education
at Time 2—Mother I 1.32 2.57 136
Years of Education
at Time 2—Father 1 1.96 2.76 114
Number of
Children at Time 2 2.22 1.21 141
Infant’s Age in
Mos. at Time 2 18.72 5.28 150

 

34

 

Measures.
Child Abuse Potential Inventory (CAP!)

The CAPI abuse scale and its six component factor subscale scores were used to
measure maternal physical abuse potential as theorized by Belsky’s ecological model for
child maltreatment. The CAPI is a 160-item self-report inventory in an agree/disagree,
dichotomous-choice format that breaks down into six component factors discussed above

(Milner, 1986). The Distress and Unhappiness factor scores have relevancy for

 

measuring the parental component of Belsky’s individual level, as they tap into levels of

maternal psychopathology. Items for the Distress scale include “I am sometimes very

 

sad,” “I find it hard to relax,” and “I am often angry inside.” Items for the Unhappiness
scale include “I do not laugh very much,” “I am an unlucky person,” and “I have a good
sex life”—reversed scored (Milner, 1986). The Problems with Child and Self subscale is
a representative measure of the child component of Belsky’s individual level, as it
assesses difficult-to-parent child traits. Items include “I have a child who is slow,” “I
have a child who gets into trouble a lot,” and “I have a child who is bad” (Milner, 1986).
The Problems with Family scale is an adequate measure of the marital and familial
relationship components of Belsky’s familial level, as it contains items tapping into
martial and familial conflict (items include “My family fights a lot,” “My family has
problems getting along,” and “My family has many problems”) (Milner, 1986). Lastly,
the Problems with Others factor score is a useful measure of the social support
component of Belsky’s community level, as it taps into the extent of the mother’s

perceived support system. Items include “Other people have made my life unhappy,”

“These days a person doesn’t really know on whom one can count,” and “Other people

35

have made my life har ” (Milner, 1986). Thus, the CAPI can be viewed as capturing
major components of the individual, familial, and community levels of Belsky’s
ecological model for child maltreatment and can be used to estimate relationships
between different levels and infant social, language, and motor development.

The CAPI has been demonstrated to possess adequate reliability and validity.
Milner (1994) reported that split-half reliabilities range from 0.96 to 0.98, Kuder
Richardson-20 reliabilities range from 0.92 to 0.95, and test-retest reliability scores have
been obtained of 0.91 for one day, 0.90 for one week, 0.83 for one month, and 0.75 for
three months. The CAPI has shown adequate construct validity across a wide array of
sampled groups (e.g., depressives, parents of children with emotional difficulties,
individuals with low social support) (Kolko et al., 1993; Milner, 1994). Accurate
classification rates for this scale have been calculated to be in the 90th percentile across a

wide range of diverse samples (Milner, 1994).

Denver Developmental Screening Test 11

The Denver Developmental Screening Test 11 was employed to rate infant motor,
language, and gross and fine motor development across the infant’s first year of life.
This measure is a 1990 revised version of the Denver Developmental Screening Test.
Scores are determined by observing infants’ abilities to carry out basic, age-appropriate
tasks. Scorers then rate the number of Advances, defined as items an infant passed that
were passed by less than 25% of the initial standardization sample at the same age;
Delays, defined as items an infant failed that were passed by 90% of the standardization

sample at the same age; and Cautions, defined as items an infant failed that were passed

36

by 75% of the standardization sample at that age. Reliability estimates for the Denver
have been found to be adequate, with interrater reliability estimates between trained
observers and trained examiners of 98.7%. Test-retest reliability after a lO-minute
interval has been determined to be 0.91 and after a seven-to-lO-day interval to be 0.89
(Mirenda, 1996). No construct or concurrent validity estimates have been calculated, as
the authors assert that the measure does not tap into unitary constructs as well as that

there are no extant diagnostic tests that measure similar constructs (Mirenda, 1996).

KCHS Clinical Judgments Form

The KCHS Clinical Judgments Form’s administration entailed the home visitor
rating several parenting dimensions on a five-point Likert scale, such as knowledge of
child development and anger management (Items 1 to 12). The home visitor also
assessed on a four-point Likert scale the likelihood for mothers to engage in abusive or
neglectful behaviors such as failure to provide adequate food and shelter as well as
corporal punishment (Items 14 to 18). Items 1 to 12 and 14 to 18 were summed to
produce two separate scores. The alpha coefficient for the first 12 items was calculated
to be 0.92 and for items 13 to 18 to be 0.77. Item 13 was removed from analyses of this
measure, as it did not load on either factor in preliminary factor analyses. The subscales
can be seen as relevant to the parenting component of the individual level of Belsky’s
ecological model for child maltreatment, which, as mentioned above, taps into unhealthy
parenting practices. It is also important to note that this measure was only given to
families who received home visitor intervention. As the study collapsed intervention and

control groups during analyses, the number of subjects who are used in determining

37

results with this measure will be lower than the number used to calculate results

involving the Child Abuse Potential Inventory, which was given to all subjects.

Plan of Analysis.

Hypothesis 1: a. Correlations were calculated to determine the nature of relationships
between Denver infant developmental scores and CAPI factor scores. b. Correlations
between changes in Denver infant developmental scores and changes in CAPI factor
scores were calculated to ascertain relationships between them.

Hypothesis 2: a. Correlations were calculated to determine the nature of relationships
between Denver infant developmental scores and KCHS Clinical Judgments form scores.
b. Correlations between changes in Denver infant developmental scores and changes in

KCHS Clinical Judgment scores were calculated to ascertain relationships between them.

 

Hypothesis 3: CAPI factor scores (D_is_t;e§, Unhappiness, Problems with Child and Self
Problems with Family, Problems with Others en total) were partialled out of significant
correlations between infant development measures and mothers’ age status (teenage vs.
adult) to determine if such relationships still remain significant, as were KCHS subscale
scores composite totals in a separate analysis. For ethnic category, two AN COVA’s were
calculated with the 5 CAPI subscales listed above as covariates in one analysis and the 2
KCHS subscales in another, the ethnic category as the independent variable, and Denver
scores as the dependent variable. This procedure was used to determine if development-
ethnicity relationships are still significant outside of measures of components of Belsky’s

ecological model if the initial AN OVA’s without covariates are significant.

38

Chapter 3: Results

Overview.

The purpose of the present study was to determine the association of abuse-
engendering environments with maladaptive infant social, cognitive, and motor
development in homes lacking substantiated abuse incidents. It was hypothesized that all
developmental harms that have been illustrated in prior research to be concomitants of
physical abuse would also be significantly related to critical components of potentially—
but not actually—abusive environments. To determine the relationships between
parenting variables and infant development, in order to avoid alpha inflation, significant
correlations were calculated between measures of parenting environment (CAPI factor
scores and KCHS Clinical Judgments F orrn scores) and overall Denver Advances,
Cautions, and Delays scores at each age. See Table 2 below for descriptive statistics
regarding all measures analyzed. Limiting the number of correlations run provided
greater assurance that significant results were not chance effects. To further clarify, a
total of 21 analyses were run for each time frame (7 measures of Belsky model
variables—5 CAPI factor scores and 2 KCHS Clinical Judgments subscale scores X 3
measures of infant development—Denver Advances, Cautions, and Delays). At a
significance level of .05, it wOuld be expected by chance that one correlation would be
significant (21 X .05 = 1.05). As all results presented below demonstrated hypotheses
supported with more than one correlation, the results can be seen as significant beyond
what would be expected by chance. Correlations between subcategories of Advances,

Cautions, and/or Delays (gross motor, fine motor, social, and language development) and

39

parenting environment measures were then calculated for those Denver-parenting
measure pairings that yielded significant correlations in the first set of analyses. This
procedure was used to determine which areas of infant development most accounted for
the first results (see Appendix for all secondary-analysis correlations). For example, after
a significant correlation between Time 1 CAPI Distress and Denver Cautions scores was
found, correlations between Time 1 CAPI Distress and Denver Social Cautions,
Language Cautions, Gross Motor Cautions, and Fine Motor Cautions were calculated.

This procedure yielded the results discussed in this section.

40

Table 2: Descriptive Statistics for Measures Used in this Study

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Size Minimum Maximum Mean Standard
(111) Deviation

Time 1:
CAPI Distress 151 0.00 261.00 94.37 75.11
CAPI
Unhappiness 150 0.00 62.70 12.83 12.55
CAPI Problems
with Child and
Self 151 0.00 25.20 3.61 6.09
CAPI Problems
with Family 151 0.00 38.00 11.00 12.31
CAPI Problems
with Others 151 0.00 25.50 11.85 7.67
KCHS Parent
Functioning
Problems 84 1.00 4.25 2.33 0.82
KCHS Parent
Problems
(Abusive) 74 1 .00 4.00 1.62 0.83
Denver
Advances 146 0.00 21.00 1.31 2.29
Denver
Cautions 146 0.00 5.00 0.54 0.95
Denver Delays 146 0.00 2.00 0.14 0.41
Time 2:
CAPI Distress 151 0.00 255.09 79.91 71.33
CAPI
Unhappiness 151 0.00 56.22 12.68 10.26
CAPI Problems
with Child and
Self 151 0.00 25.20 3.33 5.48
CAPI Problems
with Family 150 0.00 38.00 11.10 12.95
CAPI Problems
with Others 151 0.00 24.00 11.67 7.81
KCHS Parent
Functioning
Problems 78 1.00 3 .88 2.08 0.70
KCHS Parent
Problems
(Abusive) 72 1 .00 3 .80 1.45 0.60
Denver
Advances 147 0.00 13.00 1.53 1.83

 

 

 

 

 

 

41

 

Table 2, continued: Descriptive Statistics for Measures Used in this Study

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Size Minimum Maximum Mean Standard
(11) Deviation

Denver
Cautions 147 0.00 5.00 0.75 1.07
Denver Delays 146 0.00 13.00 0.44 1.44
Change Scores:
CAPI Distress 151 -150.18 170.00 14.46 55.90
CAPI
Unhappiness 150 -35.00 45.60 0.10 12.22
CAPI Problems
with Child and
Self 151 -20.60 21.00 0.28 5.79
CAPI Problems
with Family 150 -38.00 38.00 -0.14 12.48
CAPI Problems
with Others 151 ~14.00 19.00 0.18 6.10
KCHS Parent
Functioning
Problems 69 -1.33 2.23 0.30 0.68
KCHS Parent
Problems
(Abusive) 56 -1.30 . 3.00 0.38 0.87
Denver
Advances 142 -13.00 10.00 0.20 2.57
Denver
Cautions 142 -5 .00 5.00 -0.20 1 .32
Denver Delays 141 -1 1.00 2.00 -0.30 1.30

 

 

 

 

 

 

42

 

The Relationship Between CAPI Factor Scores and Infant Development.

Hypothesis 1: a.) The number of problems in infant development will be positively and
significantly related to individual, familial, and community-level measures of parent and
child dysfunctionality, both in the months immediately following the infant’s birth and a
year after this initial testing. b.) Increases in infant developmental problems between the
months immediately following the infant’s birth and a year after this initial testing will be

significantly correlated with increases in these same measures of dysfunctionality.

Time 1 Analyses.

Table 3 shows the relationships between CAPI factor scores reflective of
components of Belsky’s ecological model for child maltreatment and Denver Advances,
Cautions, and Delays at Time 1.

Table 3: Correlations between Time 1 CAPI Factor Scores and Denver Scores

Bold = Correlations significant at p < .05 level

 

 

 

 

 

 

 

Denver Advances Denver Cautions Denver Delays

CAPI Distress r=.06 r=.20 r=.06

p=.50 p=.02 p=.46
CAPI Unhappiness r=.05 r=.08 r=-.08

p=.53 p=.37 p=.35
CAPI Problems r--.01 r=.12 r=.08
with Child and Self p=.95 p=.14 p=.35
CAPI Problems r=.06 r=.13 r—-.05
with Family p=.51 p=.13 p=.58
CAPI Problems r=.10 r=.21 r=.02
with Others p=.24 p=.01 p=.8l

 

 

 

 

 

The above table illustrates significant relationships between CAPI Distress scores as well

 

as CAPI Problems with Others scores and Denver Cautions. Additional correlations

43

between these CAPI factor scores and Denver Gross Motor Cautions, Fine Motor
Cautions, Language Cautions, and Social Cautions were then calculated. For the CAPI
m factor score, significant correlations were found with Denver Social Cautions (r
= .19, p < .05) and Denver Language Cautions (r = .20, p < .05). For the CAPI Problems
with Others factor score, a significant correlation was found with Denver Fine Motor
Cautions (r = .20, p < .05). Thus, hypothesis 1a was partially supported for Time l—as
significant relationships between the parenting component of the individual level and the
social support component of the community level of the Belsky model and infant

development were found.

Time 2 Analyses.

Table 4 illustrates the relationships between CAPI factor scores representing
components of Belsky’s ecological model for child maltreatment and Denver Advances,
Cautions, and Delays at Time 2.

Table 4: Correlations between Time 2 CAP] Factor Scores and Denver Scores

Bold = Correlations significant at p < .05 level

 

 

 

 

 

 

 

 

Denver Advances Denver Cautions Denver Delays

CAPI Distress r=-.l6 r=.08 r=.05

p=.05 p=.33 p=.57
CAPI Unhappiness r=-.09 r=.08 r=.01

p=.27 p=.37 p=.94
CAPI Problems r=-.11 r=.18 r=.20
with Child and Self p=.21 p=.03 p=.02
CAPI Problems r=-.06 r=.ll r=.13
with Family p=.49 p=.18 p=.12
CAPI Problems r=-.05 r=.12 r=-.02
with Others p=.55 p=.16 p=.85

 

 

 

44

 

The above table first shows significant relationships between CAPI Distress factor scores

 

and Denver Advances. In secondary analyses, no significant correlations were found

between specific Denver Advances in infant development areas and CAPI Distress factor

 

scores. The table also illustrates significant relationships between CAPI Problems with
Child and Self factor scores and Denver Cautions. In secondary analyses, no significant
correlations were found between specific Denver infant development areas and CAPI
Problems with Child M factor scores. Finally, a significant relationship between
CAPI Problems with Child and Self factor scores and Denver Delays was determined.
Secondary analyses showed significant correlations between this factor score and Denver
Gross Motor Delays (r = .21, p < .05) as well as Fine Motor Delays (r = .17, p < .05).
Overall, hypothesis 1a for Time 2 was partially supported—as significant relationships
between the parent and child components of Belsky’s model and infant development

were found.

Change Score Analyses.

Table 5 reports correlations between changes in CAPI factor scores between times

1 and 2 and changes in Denver Advances, Cautions, and Delays between times 1 and 2.

45

Table 5: Correlations between Changes in CAPI Factor Scores a_nd Changes in

Denver Scores from Time 1 to Time 2 (all calculated so larger scores corresponded to more
problems)

 

Bold = Significant correlations at p < .05 level

 

 

 

 

 

 

 

 

 

 

Denver Advances Denver Cautions Denver Delays

CAPI Distress r=.09 r=.24 r=.1 l

p=.28 p=.00 p=.20
CAPI Unhappiness r=.03 r=.13 r=.03

p=.76 p=.12 p=.71
CAPI Problems r=.020 F20 r=.05
with Child and Self p=.82 p=.02 p=.57
CAPI Problems r=.01 r=.05 r=. 12
with Family p=.89 p=.59 p=.15
CAPI Problems r=.04 r=.07 r=.01
with Others p=.64 p=.42 p=.94

 

This table first shows a significant relationship between CAPI Distress factor scores and

 

Denver Cautions. Further analyses yielded that this correlation partially resulted from
Significant correlations between changes in this factor score and changes in Denver
Social Cautions (r = .20, p < .05) and changes in Denver Language Cautions (r = .21, p <
.05). In addition, there was a significant relationship between CAPI Problems with Child
an_dS_elf factor scores and Denver Cautions, partially made up of a significant correlation
between this factor score and Denver Social Cautions (r = .21 , p < .05). Thus, hypothesis
lb was partially supported for change scores—with significant relationships between the

parenting and child components of the individual level of Belsky’s model being evinced.

46

 

The Relationship Between KCHS Clinical Judgments Form Scores and Infant

Develonment.
Hypothesis 2: a. More problems in infant development will be significantly related to
higher scores on the KCHS Clinical Judgments form—an observational measure of
parent-related individual factors. b. Increases in infant developmental problems between
Time 1 and Time 2 will be significantly correlated with increases in scores on the KCHS
Clinical Judgments form.

Time 1 An_alvses.

Table 6 details results for correlations between KCHS Clinical Judgments Form

subscale as well as total scores and Denver Advances, Cautions, and Delays at Time 1.

Table 6: Correlations between Time 1 KCHS Clinical Judgments Form and Denver

Scores

Bold = Correlations significant at p < .05 level

 

 

 

 

Denver Advances Denver Cautions Denver Delays
Clinical Judgments r=.28 r=.24 r=. 15
Parent Functioning p=.01 p=.03 p=.19
Problems
Clinical Judgments r=-. 15 r=.02 r=.09
Problems (Abusive p=.21 p=.85 p=.45
Behaviors)

 

 

 

 

The table first shows a significant correlation between the KCHS Clinical Judgments
Parent Functioning Problems score and Denver Advances. Secondary analyses illustrated
that this correlation was partially accounted for by significant relationships between this
KCHS subscale and Denver Social Advances (r = -.22, p < .05) as well as Denver Fine
Motor Advances (r = -.42, p < .05). In addition, a significant correlation was ascertained
between the KCHS Clinical Judgments Parent Functioning Problems score and Denver

Cautions. Secondary analyses demonstrated that a significant relationship between this

47

 

KCHS subscale and Denver Gross Motor Cautions (r = .27, p = .05) was a salient part of
the preliminary finding. In general, these results partially support hypothesis 2a for Time
1. Individual parent functioning problems that were not found in overtly abusive homes
as tapped by the KCHS Clinical Judgments form were found to relate to infant
development.

Time 2 Analyses.

Table 7 outlines the findings for KCHS Clinical Judgments Form subscales at Time 2.

Table 7: Correlations between Time 2 KCHS Clinical Judgments Form Scores and
Denver Scores

Bold = Correlations significant at p < .05 level

 

 

 

 

Denver Advances Denver Cautions Denver Delays
Clinical Judgments r=.28 r=.22 r=. 1 3
Parent Functioning p=.01 p=.05 p=.25
Problems
Clinical Judgments r=-.09 r=.03 r=.05
Problems (Abusive p=.48 p=.81 p=.70
Behaviors)

 

 

 

 

The above table illustrates one significant correlation for Time 2: the relationship
between KCHS Clinical Judgments Parent Functioning Problems and Denver Advances.
Secondary analyses supported that this KCHS subscale’s relationships with Denver
Social Advances (r = -.29, p < .05) and Gross Motor Advances (r = -.23, p < .05)
accounted for a significant part of its overarching relationship with Denver Advances.
Thus, hypothesis 2a was partially supported for Time 2, as nonabusive parenting
problems measured by the KCHS Clinical Judgments Form were related to whether

infants exceeded developmental standards for their age.

48

 

Change Score Analyses.

Table 8 shows correlations between changes in KCHS Clinical Judgments Form
subscale scores between times 1 and 2 and changes in Denver Advances, Cautions, and
Delays scores between times 1 and 2.

Table 8: Correlations betweep Changes in KCHS Clinical Judgments Form Scores

and Changes in Denver Scores from Time 1 to Time 2 (all calculated so larger scores represent
more problems)

Bold = Correlations significant at p < .05 level

 

 

 

 

Denver Advances Denver Cautions Denver Delays
Clinical Judgments r=-.01 r=.17 r=. 12
Parent Functioning p=.91 p=.19 p=.34
Problems
Clinical Judgments r—-.02 r--.02 r=.15
Problems (Abusive p=.89 p=.89 p=.28
Behaviors)

 

 

 

 

The above table illustrates that, as measured by the KCHS Clinical Judgment Form
subscale scores, there were no significant relationships between parenting and infant

development. Thus, hypothesis 2b was not supported.

The Role of Ethnicity and Age in Potentially Abusive Environments.
Hypothesis 3: Any significant relationships between ethnicity/culture and parents’ age,
and infant development scores at times one and two—as well as changes in these scores
between the months immediately following the infant’s birth and a year after this initial
testing—will be an artifact of the relationship between measures of all components of
Belsky’s ecological model and infant development scores.

The initial premise upon which these analyses were based—the existence of

significant relationships between both ethnicity as well as age and Denver Advances,

49

 

Cautions, and Delays—was not supported beyond what would occur by chance.
Analyses of time 1, time 2, and change score correlations yielded only one significant
relationship: that between ethnicity and time 2 Denver Advances. No significant
relationships were determined between teenage status and Denver variables at any time.
As one significant correlation would be expected out of 18 analyses with an alpha of .05
(18 x .05 = .9), the significant correlation for the ethnicity variable can be ruled out as

“Statistical dust.” Thus, hypothesis 3 was not supported due to it not being testable.

50

Chapter 4: Discussion

The results of this study provided partial support for its hypothesis that infant
development in the first year of life is related to specific individual-, familial-, and
community-level components of Belsky’s ecological model for child maltreatment, as
assessed by the CAPI factor scores and the KCHS Clinical Judgments form. The
individual level of the Belsky model, in particular, was salient across all time frames as
playing a crucial role in infant development outside of actual abusive incidents. In order
to best elucidate the extent to which the Belsky model was supported in this study, results

will be discussed in relation to each level.

The Individual Level: Parenting Commnent

This study provided substantial support for the constructs tapped by the parenting
component of the individual level of child maltreatment as impacting infant development
in the first year of life outside of actual abusive incidents. This evidence was obtained
from two veins: a.) the significant correlations found at times 1, 2, and for change scores

between the CAPI Distress score and Denver variables and b.) the significant

 

relationships determined at times 1 and 2 between the KCHS Clinical Judgments Parent
Functioning Problems subscale and Denver infant development scores. Interestingly,
there is a degree of overlap between the constructs assessed by the CAPI _I_)_istr_es_§ factor
score and the KCHS Clinical Judgments Parent Functioning Problems subscale. For

example, both the CAPI Distress and KCHS Clinical Judgments Parent Functioning

 

Problems subscale contain items addressing maternal ability to manage anger,

51

stress/anxiety, and feelings of depression. Thus, it can be postulated that such affects
engender delays in infant development in potentially—but not actually—abusive
environments. However, notably, the KCHS Clinical Judgments Parent Functioning
Problems subscale also assesses parenting variables not measured by the CAPI Dimes;
factor score, primarily those related to parental knowledge of appropriate child-rearing
practices (e.g., "Excessive need for child to obey or comply," "Inaccurate sense of child's
needs," etc.). Therefore, this study's results implicate mothers' understanding of their
roles as parents as an important individual-level variable to factor into assessments of
parents' contributions to infant development in potentially abusive environments.

Another interesting finding is that there were no significant correlations between
KCHS Clinical Judgments Problems (Abusive Behaviors) subscale scores and infant
development found for any time frames assessed. These findings strongly implicate
maternal behaviors outside of actual abusive/neglectful actions as being responsible for
negative effects on infant development. This result should be examined in actually
abusive homes, however, before it can be fillly stated as fact.

The pattern of specific Denver subscales with which the above maternal parenting
measures were significantly associated lends credence to the study's initial claims that
effects of maternal problems on infant development are manifested across different
developmental tasks. For instance, at Time 1, significant relationships between Denver

Cautions, Social and Language Cautions and the CAPI Distress factor score, as well as

 

between the KCHS Clinical Judgments Parent Functioning Problems subscale and
Denver Advances, Cautions, Social Advances, Fine Motor Advances, and Gross Motor

Cautions are suggestive of impacts of individual maternal variables across all significant

52

infant developmental areas the Denver measures. The fact that there were correlations
between parenting measures and broad-based Denver Advances and Cautions variables—
which incorporate gross and fine motor, language, and social development totals—
implicates these variables as generally having a holistically negative effect on infant
development. Especially detrimental effects are supported for the areas determined
through secondary analyses to be significant. Thus, for Time 1, the above-mentioned
specific correlations with social, language, and gross and fine motor development further
advance the hypothesis that impacts are found in all major infant developmental areas.

Similarly, the significant relationships found at Time 2 between CAPI Distress

 

factor scores and Denver Advances overall, as well as between KCHS Clinical
Judgments Parent Functioning Problems subscale scores and Denver Advances, support
wide-ranging effects on infant development. The secondary-analysis correlations
between KCHS Clinical Judgments Parent Functioning Problems subscale scores and
Denver Social and Gross Motor Advances suggest particularly strong effects on these two
areas. Thus, social and gross motor difficulties should be targeted at around one year in
families with significant maternal problems.

Finally, the significant relationships between changes in CAPI Distress factor

 

scores and changes in Denver Cautions between times 1 and 2 imply that effects of
affective maternal traits are both major and intertwined across the first year of life.
Secondary analyses bring out that especially detrimental effects on social and language
development are associated with these variables during this time frame. In sum, results
for Times 1, 2, and change scores lend support to effects of the parenting component of

Belsky's individual level on infant development, both holistically and most saliently for

53

certain specific areas of infant development that vary depending on the time frame
assessed.
Another important point for exploration is that different significant correlations

were found between the CAPI Distress factor score vs. the KCHS Clinical Judgments

 

Parent Functioning Problems subscale score and Denver measures. These discrepancies
are suggestive of differential effects of parenting variables on varying aspects of infant
development—as there are different parenting variables tapped by the two measures, as
mentioned above. In addition, no significant relationships were found at times 1, 2, or for
change scores between CAPI Unhappiness factor scores—another designated measure of
the parenting component of Belsky's individual level—and infant development. Thus, it
appears that, during the first year of life, maternal depressive symptomatology assessed in
isolation from other parenting variables is not related to maladaptive infant development
in homes with strong physical abuse potential. However, the fact that significant

relationships were found between CAPI Distress as well as KCHS Clinical Judgments

 

Parent Functioning Problems scores—which notably incorporate depressive
symptomatology as part of the constructs they measure—and infant development
suggests that depression may still affect infant development in such homes as those tested
in this study. Its impact may not occur outside of its interaction with other parenting
problems, such as anxiety, anger management, and knowledge of child development,
however. Further research is needed to partial out in more detail and with more measures
from multiple informants how separate components of the parenting aspect of the
individual level of Belsky's model affect various aspects of infant development in the first

year of life, however, before such suppositions can be accepted as empirical fact.

54

The Individupl Level: Child Component.

The results elucidating the infant's contribution to his/her own development—the
child component of the individual level of Belsky's ecological model—are illustrative of
significant impacts at Time 2 and throughout the first year (as measured by change scores
between times 1 and 2), but not at Time 1. The significant correlation between CAPI
Problem; with Child a_rm_Se_lf factor scores and Denver Delays at Time 2 is suggestive of
broad-ranging delays in infant development after a year being related to difficult-to-
manage child traits. This effect would seem to be particularly salient for gross and fine
motor development, as secondary analyses yielded significant results for delays only in
these areas. Regarding change score results, the significant relationship found between
increases in CAPI Problems with Child mm factor scores and more Denver Cautions
across the first year reflects infant characteristics' general effects on their own
development. Secondary analyses showed that this effect is particularly attributable to
infant traits' impacts on their own social development during the first year. Thus, overall,
this aspect of Belsky’s model garnered support in this study. Finally, the fact that
correlations were found between social and gross and fine motor development—but not
language development—implicate the former three areas as most impacted by hard-to-
manage infant traits, as well as that this variable has broad-ranging effects across multiple
developmental tasks.

In general, the above findings provide evidence that in potentially abusive homes,
problematic infant traits contribute to an environment facilitative of developmental

deficiencies. However, the fact that no effects were found at Time 1 warrants firrther

55

explanation. Previous literature supports such results. For instance, several researchers
have found that temperament levels during the few months predict infant development at
one year (e.g., Belsky, Rovine, & Taylor, 1984; Egeland & Farber, 1984), as well as
concurrent effects of infant temperament on development at one year (e.g., Calkins &
Fox, 1992; Wachs & Desai, 1993). These studies suggest that temperamental effects will
not be seen unless infant development at one year is factored into analyses—which is
exactly what this study's findings evinced. The results for Time 2 mirror findings
regarding concurrent effects at one year. Change score analyses are illustrative of how
interrelated temperament and infant development are during the first year, which
predictive analyses implicate, as well. Thus, this study's results are reflective of the

larger infant temperament knowledge base.

The Familial Level of Belsky's Ecological Model.
No significant correlations were found between Problems with Family CAPI

factor scores and infant development for times 1, 2, or change scores. These findings are
surprising in light of the preponderance of evidence supporting unhealthy familial
dynamics' effects on infant development cited above. Such results imply that overarching
familial dynamics factor into the dynamics of abusive situations to a level facilitative of
impairment only in situations where actual abuse occurs, a conclusion contrary to the

basic hypothesis of this study.

56

The Communig Level of Belsky's Ecological Model.
Results gleaned from Problems with Others CAPI factor scores—the primary

measure of social support/community-level influences utilized in this study—evidenced
significant impacts only at Time 1. Findings showed that lack of an external support
system affects infant development generally, as represented by its significant correlation
with Denver Cautions. The fact that secondary analyses yielded a correlation with Fine
Motor Cautions implicates especially harmful impacts of unsatisfactory social support on
newborn infants' ability to control fine motor movements. It would appear that fine
motor development’s relationship with social support is especially strong—as no other
areas of infant development yielded significant findings. Overall results for this factor
score are particularly salient in that correlations were found with Denver Cautions,
illustrative of halted development. Interestingly, the results contrast the above-cited body
of literature, which mainly reported impacts of social support at around one year (e.g.,
Crockenberg, 1981). Thus, it appears that more research is needed to further explicate
the role of social support in infant development—particularly in the fine motor area—
early in infancy. Also, these findings lend support to the supposition that social support
is associated with maladaptive infant development during and at around one year only in

homes where actual abuse occurs, a finding contrary to expected results.

The Role of Ethnicig and Age.

As stated above, the absence of significant relationships between race as well as
teenage status and Denver scores reflects that such variables did not have tangible effects

on infant development in this sample. These results are surprising in light of the

57

heterogeneous sample employed in this study, both regarding age and ethnicity. A
potential reason for such findings is that the Denver Developmental Screening Test 11,
which functions as a screening measure for developmental dysfunction, was not as adept
at picking up ethnic and age effects. Such findings may have been present, but not in
such a manner that they impacted infant development to the level that infants showed
delays that placed them behind 75 to 90% of their peers, or prevented them from
advancing ahead of more than 75% of them (e.g., they may have been held back behind
50% of their peers, etc.). As literature examining the relationship between the Denver
Developmental Screening Test II’s relationship to age and ethnicity is virtually
nonexistent, this hypothesis is difficult to corroborate. Such research is needed in the
future to better elucidate the nature of these variables’ effects on infant development in

the infant’s earliest years, when such impacts can be most salient.

S my.

In sum, the pattern of results obtained from this study lend partial support to its
overarching hypothesis that problems inherent in the individual, familial, and community
levels of Belsky's ecological model for child maltreatment have negative effects on infant
social, motor, and language development in homes with a high potential for physical
abuse. The parenting component of the individual level was determined to have effects
on all salient areas of infant development at all investigated time frames, and the child
component demonstrated impacts in changes during the first year and upon completion of
the infant's first year of life. Therefore, it appears that individual characteristics—with

the notable exception of depression, as highlighted above—can produce harmful effects

58

on infants in homes where the potential for abuse does not escalate to the level of actual
abusive behaviors. Similar results were found only for the time frame near birth for
Belsky's community level, and not at all for his familial level. Thus, this study suggests
that in order to ameliorate infant development in potentially abusive homes, mothers'
individual emotional and child-rearing difficulties and the extent of their external support
systems should be addressed at around the time of infants' birth, and mothers' difficulties
as well as hard-to-manage infant traits should be targeted throughout the first year.

Also notable was the pattern in which significant results were found. When
viewed fiom a developmental perspective, individual parent variables’ and social support
variables’ effects were salient from birth, while child variables’ effects began to manifest
themselves as the infant developed across the first year. Thus, it appears that infants’
influence becomes more significant, while that of external support systems becomes less
significant, as infants grow and develop into more autonomous individuals. Studying this
population for a longer time frame might potentially evidence effects for familial-level
variables that had not had developmental impacts during the first year. Thus, further
research should explore potential interactions between lack of adequate parenting skills
as well as support systems and the time as well as severity of the beginning and
manifestation of hard-to-manage infant traits to ascertain if such traits may appear in
response to the harms such infants’ development begins with.

In general, assisting mothers with these life difficulties should help infants
develop in a more healthy fashion. Such aid will benefit infants in all developmental
areas, as secondary analyses across all individual-level variables and time frames showed

significant relationships with gross and fine motor, social, and language development.

59

As familial factors in general and community-level factors after birth did not evidence
effects, it seems likely that these levels of Belsky's model are more salient and should be
addressed in homes where physical abuse is present. However, testing a sample with
documented physical abuse is needed to verify such a hypothesis.

There are several other shortcomings that warrant further research to remedy.
First, this research was conducted only with mothers. Exploration of similar variables
with fathers would allow for a fuller picture of the dynamics of potentially abusive
households and how these relate to infant development to be obtained. Second, although
infants were tested at approximately birth and again at approximately one year, there was
variation in infants' ages around these time frames. In the rapidly-changing world of
infant development where a month is associated with vast developmental gains, a sample
with less variance in age at each time frame would make it easier to attribute effects to
specific developmental stages. Third, utilizing more measures of parenting variables and
particularly infant development (where the Denver was the only measure employed)
would further substantiate the findings reported here. In general, this study has
demonstrated how certain traits of potentially abusive homes—particularly of the
individuals in them—can engender the same unhealthy developmental outcomes during

the first year of life as actually abusive homes can.

60

References

Aber, J .L. & Allen, J .P. (1987). Effects of maltreatment on young children’s
socioemotional development: An attachment theory perspective. Developmental
Psychology, 23, 406-414.

Altpeter, T.S. & Walker, CE. (1992). Prevention of physical abuse of children through
parent training. In DJ. Lewis, E.W. Holden, & M. Rosenberg (Eds), Prevention
of child maltreatment: Developmental and ecological perspectives (pp. 226-248).
New York: John Wiley and Sons.

Belsky, J. (1977). A theoretical analysis of child abuse remediation strategies. Journal
of Clinical Child Psychology, 7, 1 17-121.

Belsky, J. (1984). The determinants of parenting: A process model. Child
Development, 55, 83-96.

Belsky, J. (1993). Etiology of child maltreatment: A developmental-ecological analysis.
Psychological Bulletin, 114, 413-434.

Belsky, J ., & Rovine, M. (1987). Temperament and attachment security in the strange
situation: An empirical rapprochement. Child Development, 58, 787-795.

Belsky, J ., Rovine, M., Taylor, DO. (1984). The Pennsylvania infant and family
development project, 111: The origins of individual differences in infant-mother

attachment: Maternal and infant contributions. Child Development, 55, 718-728.

61

Belsky, J., & Vondra, J. (1989). Lessons from abuse: The determinants of parenting. In
D. Cicchetti & V. Carlson (Eds), Child maltreatment: Theory and research on
the causes and consequences of child abuse and neglect (pp. 153-202).
Cambridge: Cambridge University Press.

Bettes, B,A, (1988). Maternal depression and motherese: Temporal and intonational
features. Child Development, 59, 1089-1096.

Bohlin, G., Hagekull, B., Gerrner, M., Andersson, K., & Lindberg, L. (1989). Avoidant
and resistant reunion behaviors as predicted by maternal interactive behavior and
infant temperament. Infant Behavior and Development, 12, 105-1 17.

Braungart, J.M., & Stifter, CA. (1991). Regulation of negative reactivity during the
strange situation: Temperament and attachment in lZ-month-old infants. Infant
Behavior and Development, 1 4, 349-3 64.

Bronfenbrenner, U. (1977). Toward an experimental psychology of human
development. American Psychologist, 22, 513-531.

Calkins, S.D., & Fox, NA. (1992). The relations among infant temperament, security of
attachment, and behavioral inhibition at twenty-four months. Child Development,
63, 1456-1472.

Cohn, J .F., & Tronick, E. (1989). Specificity of infants’ response to mothers’ affective
behavior. Journal of the American Academy of Child and Adolescent Psychiatry,
242-248.

Crnic, K.A., Greenberg, M.T., Robinson, N.M., & Ragozin, AS. (1984). Maternal stress
and social support: Effects on the mother-infant relationship from birth to

eighteen months. American Journal of Orthopsychiatry, 54, 224-235.

62

Crockenberg, SB. (1981). Infant irritability, mother responsiveness, and social support
influences on the security of infant-mother attachment Child Development, 52,
857-865.

Crockenberg, S.B., & McCluskey, K. (1986). Change in maternal behavior during the
baby’s first year of life. Child Development, 5 7, 746-753.

Culbertson, J.L. & Schellenbach, CJ. (1992). Prevention of maltreatment in infants and
young children. In DJ. Lewis, E.W. Holden, & M. Rosenberg (Eds.), Prevention
of child maltreatment: Developmental and ecological perspectives (pp. 47-77).
New York: John Wiley and Sons.

Del Carmen, R., Pedersen, F.A., Huffman, L.C., & Bryan, Y.E. (1993). Dyadic Distress

 

management predicts subsequent security of attachment. Infant Behavior and
Development, 16, 131-147.

Egeland, B., & Farber, EA. (1984). Infant-mother attachment: Factors related to its
development and changes over time. Child Development, 55 , 753-771.

Elster, A.B., McAnarney, E.R., & Lamb, ME. (1983). Parental behavior of adolescent
mothers. Pediatrics, 71, 494-502.

Feldman, R., Greenbaum, C.W., Mayes, L.C., & Erlich, SH. (1997). Change in mother-
infant interactive behavior: Relations to change in the mother, the infant, and the
social context. Infant Behavior and Development, 20, 151-163.

Field, T. (1995). Infants of depressed mothers. Infant Behavior and Development, 18,
1-13.

Field, T., Healy, B., Goldstein, S., Perry, S., Bendell, D., Schanberg, S., Zimmerman,

63

.11“ 9””?

E.A., & Kuhn, C. (1988). Infants of depressed mothers show “depressed”
behavior even with nondepressed adults. Child Development, 59, 1569-1579.

Field, T., Healy, B., & LeBlanc, W.G. (1989). Sharing and synchrony of behavior states
and heart rate in nondepressed versus depressed mother-infant interactions. Infant
Behavior and Development, 12, 357-376.

Field, T.M., & Widmayer, SM. (1981). Mother-infant interactions among lower SES
black, Cuban, Puerto Rican and South American immigrants. In Tiffany Field et
al. (Eds.), Culture and Early Interactions, (pp. 41-62). Hillsdale, N.J.: Erlbaum
Associates.

Friedrich, W.N., Einbender, A.J., & Luecke, W.J. (1983). Cognitive and behavioral
characteristics of physically abused children. Journal of Consulting and Clinical
Psychology, 51, 313-314.

Frodi, A., Bridges, L., & Shonk, S. (1989). Maternal correlates of infant temperament
ratings and of infant-mother attachment: A longitudinal study. Infant Mental
Health Journal, 10, 273-288.

Goldberg, W.A., & Easterbrooks, MA. (1984). Role of marital quality in toddler
development. Developmental Psychology, 20, 304-314.

Hahn, D. (1989). A systems conceptualization of the quality of mother-infant
interaction Infant Behavior and Development, 12, 251-263.

Holden, E.W., Willis, D.J., & Corcoran, M.M. (1992). Preventing child maltreatment
during the prenatal/perinatal period. In DJ. Lewis, E.W. Holden, & M.
Rosenberg (Eds.), Prevention of child maltreatment: Developmental and

ecological perspectives (pp. 17-46). New York: John Wiley and Sons.

64

 

Jacobson, S.W., & Frye, KR (1991). Effect of maternal social support on attachment:
Experimental evidence. Child Development, 62, 572-582.

Kaufman, K.L., Johnson, C.F., Cohn, D., & McCleery, J. (1992). Child maltreatment
prevention in the health care and social service system. In DJ. Lewis, E.W.
Holden, & M. Rosenberg (Eds.), Prevention of child maltreatment:
Developmental and ecological perspectives (pp. 193-225). New York: John
Wiley and Sons.

Klein, RS. (1984). Behavior of Israeli mothers toward infants in relation to infants’
perceived temperament. Child Development, 55, 1212-1218.

Kolko, D.J., Kazdin, A.B., Thomas, A.M., & Day, B. (1993). Heightened child physical
abuse potential: Child, parent, and family dysfunction. Journal of Interpersonal
Violence, 8, 169-192.

Levitt, M.J., Weber, R.A., & Clark, MC. (1986). Social network relationships as
sources of maternal support and well-being. Developmental Psychology, 22, 310-
3 16.

McAnarney, E.R., Lawrence, R.A., Ricciuti, H.N., Polley, J ., & Szilagyi, M. (1986).
Interactions of adolescent mothers and their 1-year-old children. Pediatrics, 78,
585-590.

Mercer, R.T., Hackley, K.C., & Bostrom, A. (1984). Adolescent motherhood:
Comparison of outcome with older mothers. Journal of Adolescent Health Care,
5, 7-13.

Milner, J .S. (1986). The Child Abuse Potential Inventory manual (2"" edition). DeKalb,

IL: Psytec.

65

Milner, J .S. (1994). Assessing physical child abuse risk: The Child Abuse Potential
Inventory. Clinical Psychology Review, 14, 547-583.

Mirenda, P. (1996). Review of the Denver 11. In J .C. Conoley, J .C. Impara. & L.L.
Murphy (Eds.), The T welflh Mental Measurements Yearbook (pp. 265-266).
Lincoln, NE: University of Nebraska Press.

Pederson, D.R., Moran, G., Sitko, C., Campbell, K., Ghesquire, K., & Acton, H. (1990).
Maternal sensitivity and the security of infant-mother attachment: A Q-Sort
study. Child Development, 61, 1974-1983.

Pelaez-Nogueras, M., Field, T., Cigales, M., Gonzalez, A., & Clasky, S. (1994). Infants
of depressed mothers show less “depressed” behavior with their nursery teachers.
Infant Mental Health Journal, 15, 358-3 66.

Pickens, J ., & Field, T. (1993). Facial expressivity in infants of depressed mothers.
Developmental Psychology, 29, 986-988.

Radke-Yarrow, Cummings,E.M., Kuczynski, L., & Chapman, M. (1985). Patterns of
attachment in two- and three-year-olds in normal families and families with
parental depression. Child Development, 56, 884-893.

Rieser-Danner, L.A., Roggrnan, L., & Langlois, J .H. (1987). Infant attractiveness and
perceived temperament in the prediction of attachment classifications. Infant
Mental Health Journal, 8, 144-155.

Sameroff, A.J., Seifer, R., & Barocas, R. (1983). Impact of parental psychopathology:
Diagnosis, severity, or social status effects?. Infant Mental Health Journal, 4,

236-249.

66

Seifer, R., Sameroff, A.J., Anagnostopolou, R., & Elias, PK. (1992). Mother-infant
interaction during the first year: Effects of situation, maternal mental illness, and
demographic factors. Infant Behavior and Development, 15, 405-426.

Shaw, D.S., Vondra, J .I., Hommerding, K.D., Keenan, K., & Dunn, M. (1994). Chronic
family adversity and early child behavior problems: A longitudinal study of low
income families. Journal of Child Psychology, Psychiatry, and Allied
Disciplines, 35, 1109-1122.

Stott, F .M., Musick, J .S., Clark, R., & Cohler, B.J. (1983). Developmental patterns in
the infants and young children of mentally ill mothers. Infant Mental Health
Journal, 4, 217-235.

Susman-Stillman, A., Kalkoske, M., Egeland, B., & Waldman, I. (1996). Infant
temperament and maternal sensitivity as predictors of attachment security. Infant
Behavior and Development, 19, 33-47.

Teberg, A.J., Howell, V.V., & Wingert, WA. (1983). Attachment interaction behavior
between young teenage mothers and their infants. Journal of Adolescent Health
Care, 4, 61-66.

Wachs, T.D., & Desai, S. (1993). Parent-report measures of toddler temperament and
attachment: Their relation to each other and to the social microenvironment.
Infant Behavior and Development, 16, 391 -396.

Wekerle, C.W. & Wolfe, DA. (1996). Child maltreatment. In E.J. Mash & R.A.
Barkley (Eds.), Child Psychopathology (pp. 492-537). New York: Guilford
Publications.

Willis, D.J., Holden, E.W., & Rosenberg, M. (1992). Child maltreatment prevention:

67

Introduction and historical overview. In DJ. Lewis, E.W. Holden, & M.
Rosenberg (Eds.), Prevention of child maltreatment: Developmental and
ecological perspectives (pp. 1-16). New York: John Wiley and Sons.
Zeanah, C.H., Boris, N.W., & Larrieu, J .A. (1997). Infant development and
developmental risk: A review of the past 10 years. Journal of the American

Academy of Child and Adolescent Psychiatry, 36, 165-178.

68

PKtfi-fii‘.‘ .4

APPENDIX

Table 9: Secondary Analysis Correlations for Significant Results from Broad
Denver Scales (Advances, Cautions, and/or, Delays) @ld correlations are signiicant at g < .05

Ml

 

 

1" Set of Significant

Correlations

2W5“ of Significant

Correlations

 

 

Time 1:

 

CAPI Distress

DEW!
Social r = .19
Fine Motor r = .08
Language r = .20
Gross Motor r = .09

 

CAPI Problems with
Others

Denver Cautions:
Social r = .13
Fine Motor r = .20
Language r = .09
Gross Motor r = .15

 

Clinical Judgments
Parent Functioning
Problems

Denver Advances:
Social r = -.22
Fine Motor r = -.42
Language r = -.06
Gross Motor r = -.12

Denver Cautions:

Socialr = .17
Fine Motor r = .13
Languager = .13

Gross Motor r = .27

 

 

 

Time 2:

 

CAPI Distress

Denver Advances:
Social r = -.15
Fine Motor r = -.08
Language r = -.04
Gross Motor r = -.10

 

CAPI Problems with
Child and Self

Denver Cautions:
Social r = .05
Fine Motor r = .07
Language r = .16
Gross Motor r = .13

Denver Delays:
Social r = .10

Fine Motor r = .17
Language r = .08
Gross Motor r = .21

 

Clinical Judgments
Parent Functioning
Problems

 

Denver Advances:
Social r = -.29
Fine Motor r = -.12
Language r = -.11
Gross Motor r = -.23

 

 

69

 

Table 9 continued: Seconda Ana sis Correlations for Si ificant Results from

Broad Denver Scales (Advances, Cautions, and/or, Delays)

 

1'“Set of Significant 2"“ Set of Significant
Correlations Correlations

 

Change Scores:

 

CAPI Distress Denver Cautions:
Social r = .20
Fine Motor r = .13
Language r = .21
Gross Motor r = .12

 

 

CAPI Problems with Denver Cautions:
Child and Self Social r = .21
Fine Motor r = .08
Language r = .06

Gross Motor r = .10

 

 

 

70

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

iwilllilywmulwMilli