OVERDUE FINES:

25¢ per day per item

RETURNING LIBRARY MATERIALS:

3‘ — P‘Iace in book return to remove
“"5 , charge from ci rcu1at10n records

{(#n‘fl\\ in

 

DEC 55 I???

 

 

PREDICTING INDIVIDUAL DIFFERENCES.
IN FATHER-INFANT INTERACTION

By
Timothy L. Goth-Owens

A THESIS

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

MASTER OF ARTS

Department of Psychology
1981

G] "j o'.'~"//

ABSTRACT

PREDICTING INDIVIDUAL DIFFERENCES IN
FATHER-INFANT INTERACTION

By
Timothy L. Goth-Owens

The study investigated the relationships between (a)
infant characteristics, (b) father characteristics, (c)
marital characteristics, and (d) maternal behavior charac-
teristics in predicting variations in the frequency and
quality of father-infant interaction. The subjects were
twenty-five families recruited from local childbirth
classes and obstetricians' practices.

Four months after the birth of their infants, parents
completed the Michigan Infant Temperament Survey, the
Temperament Scale-Erman, the Bem Sex Role Inventory, the
Locke-Wallace Scale of Marital Satisfaction and the
Parent Participation Inventory, a measure designed to
assess fathers' participation in caregiving activities.
In.addition, parents' interactions with their infants
were observed in two one-hour home observations.

Fathers' participation in caregiving was predicted
by positive maternal behavior (r = -.53). Fathers'
marital satisfaction was significantly correlated with

negative paternal behavior (-.58). No other relationship

Timothy L. Goth-Owens

between predictors and criterion variables was signifi-
cant. The findings are offered as limited support for
a family systems analysis of early parent-child inter-

action. The lack of significant findings is also

discussed.

To Judy, Amanda, and Caleb or Eliza

ii

ACKNOWLEDGMENTS

I would like to express my appreciation to Dr. Gary
E. Stollak. His support and encouragement were there
when I needed them. His assistance in matters ranging
from the concrete to the intangible aided me immensely
in the completion of this task. My appreciation is
also extended to Dr. Lawrence Messé'for timely and
practical guidance in my attempt to walk the line
between using data fully and capitalizing on chance.

I would like to thank Dr. Lucy Rau Ferguson for her
critical reading, useful comments, and for consistently
offering guidance in a manner demonstrating respect for
my abilities and my interests.

Special thanks go to all those involved in planning
and executing the study from which this research is drawn.
In particular credit is due to Phyllis Watts, Iriet Pesh—
kess and Tirtze Assor, along with Professors Stollak and
Messe’.

Finally, I wish to thank my Spouse, companion,
colleague, and friend, Judy Goth-Owens. Our sharing

has enriched me in countless ways.

iii

TABLE OF CONTENTS

INTRODUCTION .....................................
REVIEW OF THE LITERATURE .........................

Summary and Critique: Comparing Fathers
with Mothers .............................
Individual Differences in Fathers ..........
Determinants of Individual Differences
in the Father-Infant Relationship ......
Infant Characteristics ...................
Sex of Infant ..........................
Infant Temperament .....................
Conclusions ............................
Father Characteristics ...................
Sex-Role Orientation ...................
Temperament ............................
Conclusions .............................
The Father-Mother Relationship ...........
The Mother-Infant Relationship ...........
Summary: IndiVidual Differences in the
Father-Infant RelationShip .............
Limitations of Past Research ...............
The Present Study ..........................
Significance .............................
Variables ................................
Behavioral Measures ....................
Self-Report Measures ...................
Hypotheses ...............................

METHOD ..........................................

Subjects ...................................
Instruments and Observational Data .........
Bem Sex Role Inventory (BSRI) ............
Parent Participation Inventory (PPI) .....
Temperament Scale-Erman (TS-E) ...........
Michigan Infant Temperament Scale (MITS).
Locke-Wallace Marital Scale (L-W) ........
Demographic Questionnaire ................
Observational Data .......................
Reliability ............................
Procedures .................................

iv

RESULTS .......................................... 36
Preliminary Analyses ........................ 36
Data Reduction ............................ 36
Ratings ................................. 36
Observation Periods ..................... 36
Stability of Behavioral Measures ........ 38
Reduction of Behavioral Variables ....... 38
Correlational Analyses ...................... 42
Zero-order Correlational Analyses ......... 42
Multiple Regression Analyses .............. 42
Preliminary Regression Analyses ......... 45
Final Regression Analyses ............... 46
DISCUSSION
The Sample .................................. 49
Parent Behavior Scales ...................... 51
Predicting Individual Differences in the
Behavior of Fathers ....................... 54
Fathers' Participation in Caregiving ...... 55
Positive Paternal Behavior ................ 58
Negative Paternal Behavior ................ 60
Limitations of the Present Study ............ 62
Methodological Limitations ................ 62
Conceptual Limitations .................... 63
Implications for Future Research ............ 65
Summary ..................................... 68
REFERENCE NOTES .................................. 69
REFERENCES ....................................... 70
APPENDIX A. Parent Participation Inventory ...... 76
APPENDIX B. Demographic Questionnaire ........... 78
APPENDIX C. Home Observation Manual ............. 79

APPENDIX D. Descriptive Statistics .............. 85

Table

\lO‘

LIST OF TABLES

Intercorrelations Among Rating Scales..

Stability of Behaviors and Ratings
Across Observations ..................

Intercorrelations Among Father
Behaviors and Ratings ................

Intercorrelations Among Mother
Behaviors and Ratings ................

Scale Compositions, Alpha Coefficients,
and Temporal Stability ...............

Zero-order Correlation Coefficients ......

Regression Summary Table .................

vi

Page
37

39

4O

41

43
44
47

INTRODUCTION

The present study was undertaken in order to expand
our understanding of the nature and determinants of
fathers' interactions with infants. In particular, the
study attempted to identify characteristics of infants,
fathers, marital relationships, and mother-infant inter-
action which influenced the behavior of a sample of
fathers in interaction with their infants.

The attempt to specify influences on fathers' behavior
and involvement with their infants is based on several
assumptions within the family systems approach. This is
approach advocated by a number of developmental psychol-
ogists concerned with elucidating the role of the father
in infant development (Belsky, 1980a, 1980b; Lewis and
Feiring, 1978; Parke, Power, and Gottman, 1979; Pedersen,
1980b).

The family systems framework is based on several
assumptions about the functioning of family members
vis a vis each other. First, it is assumed that
individual family members, including the infant, influence,
and are influenced by, each other member of the family.
At this level of analysis the implication is that
individual family members bring characteristics to

their interactions with family members which influence

1

the process and outcome of those interactions.

A second mode of analysis implied by a family systems
framework involves delineating reciprocal influences of
the various dyads within a family. Thus, in a three
person family the father-infant relationship is presumed
to be influenced both by the father-mother relationship
and the mother-infant relationship.

In the family systems framework several determinants
of differential behaviors of fathers toward infants are
suggested. One expects to find the behavior of fathers
varying as a function of (a) differences in the individual
characteristics of the infants as well as the fathers,
(b) differences in the marital relationship and, (c)
differences in the mother-infant relationship.

The present investigation utilized those sources
of influence in order to predict variations in the degree
to which fathers participate in the care of their infants
and variations in their behavior with the infants. Mul-
tiple regression analyses were used to determine which
combinations of variables accounted for the greatest
variance in measures of fathers' participation and
behavior.

The rationale for selection of the particular variables
utilized in this study is contained in the ensuing review
of the literature. Variables were sought which would

have a conceptual or empirical basis in the limited

literature pertaining to individual differences in father-

infant interaction. Those that emerged as plausible

predictors are as follows:

(1)

(2)

(3)

(4)

Infant Characteristics

A. Gender

B. Temperamental "Difficulty"

Father Characteristics

A. Sex-role Orientation (masculinity, femininity)

B. Temperament (activity, sociability,

emotionality, impulsivity)

Marital Characteristics

A. Fathers Marital Satisfaction

B. Mothers Marital Satisfaction

C. Discrepancy in Parents' Perception of
-Infant Temperament

D. Parents' Temperamental Similarity

Mother Behaviors

A. Positive Maternal Behaviors

B. Negative Maternal Behaviors

In summary, the present study was undertaken to address

the lack of research aimed at identifying determinants

of individual differences in father-infant interaction.

Fourteen predictor variables were identified on the basis

of a family systems framework and available data pertaining

to father—infant relationships. These predictors were

utilized in an attempt to account for variations in

a) the degree to which fathers participate in the routine
care of infants, b) the quality of observed father-infant
interaction and c) the frequency of observed father-

infant interaction.

REVIEW OF THE LITERATURE

A sustained, systematic exploration of the role of
the father in infant development has only recently begun.
As late as 1975, Lamb concluded that fathers were "the
forgotten contributors to child development." In 1978,
Clarke-Stewart observed that the number of publications
devoted to fathers that have appeared since Lamb's 1975
review suggesnsthat they are no longer forgotten. She
concluded that there was, however, little solid, replica-
ted evidence concerning the role of the father or the
father's specific contributions to the behavior and
development of infants.

One particular approach to the study of fathers has
been concerned with identifying (a) similarities in and
differences between, mothers' and fathers' behavior as
caregivers and (b) similarities in, and differences between,
the behaviors of children toward mothers and fathers.
Such comparisons appear useful in several respects. First,
by demonstrating areas in which the behavior of mothers
and fathers is similar, researchers such as Belsky
(1979b), Kotelchuck (1976), Parke and O'Leary (1976)
and Sullivan and McDonald (1979) have been able to argue
against a priori assumptions that there is something

uniquely female involved in the capacity to care for

infants and toddlers.

Second, by establishing similarities in the behavior
of infants toward mothers and fathers, researchers have
been able to offer empirical support for arguments to
the effect that fathers as well as mothers are (a) salient
figures in the infant's social world, (b) capable of
eliciting attachment and affiliative behaviors and
attachment bonds and (c) likely to significantly influence
the course of child development. This line of research,
reviewed elsewhere by Parke (1979), is typified by the
work of Lamb (1980).

Finally, the comparison of mothers and fathers allows
researchers to identify differences in the behavior of
mothers and fathers. Differences so identified may
suggest the manner in which parents contribute differen-
tially to the development of children.

There has been a great deal of inconsistency in the
reporting of differences. One of the moSt consistent
findings is that, in general, mothers spend more time
with infants than do fathers and are more involved in
routine caregiving activities (Clarke-Stewart, 1978;
Kotelchuck, 1976; Pedersen and Robson, 1969). There
have been consistent reports of differences in the reasons
for which parents are likely to pick up or hold their
infants; mothers are more likely to pick up their child

to give care and set limits, whereas fathers tend to

do so for purposes of play (Lamb, 1980). The style

of father-infant play, but not the quantity, appears

to distinguish fathers and mothers (Clarke-Stewart,

1978; Earls and Yogman, 1979; Parke and Sawin, 1980:
Pedersen, Anderson, and Cain, 1980). Fathers appear

to engage in more active, physical, idiosyncratic, and
stimulating play than mothers. Clarke-Stewart (1978)

and Lamb (1979) have suggested that the context of play
may be one in which fathers are likely to have particularly
potent influences.

Summary and Critique:
Comparing Fathers and Mothers

 

 

In general, reviewers of the father-infant literature
have concluded that the similarities between father-
infant and mother-infant relationships far outweigh the
differences (Parke, 1979; Pedersen, 1980a). With the
exception of the few areas noted previously, there has
not been a consistent identification of features dis-
tinguishing mothers from fathers. In part, the lack
of consistency may result from the fact that studies
have used infants from eight to thirty months of age,
observed in home and laboratory, using different
observational approaches that involve different levels
of stress for infant and family, and using different
measures. Lamb (1976a) has argued that all of these

factors are likely to significantly affect results and

make comparisons across studies problematic.

Another difficulty with this line of research is
that it involves an underlying assumption that "maternal
behavior" and "paternal behavior" each represents a
unitary phenomenon (Pedersen, 1980b). Part of the in-
consistency in results may stem from regarding mothers
and fathers as distinct homogeneous groups in analysis
of variance designs, with no further distinction made
between types of mothers and fathers. The search for
between group differences may obscure within-group
differences that are significant and that may contribute
different results in different samples. The assumption
has been that elucidating differences between fathers
and mothers will result in a better understanding of
the father's role in infant development. Differences
so identified may provide direction for theorists and
researchers. However, the ideal test of fathers'
contributions to infant development would be to compare
infant outcome among children with similar personal
characteristics whose mothers behave toward them similarly,
but whose fathers behave toward them (and the mothers)
in different ways. This approach, then, requires an
investigation of individual differences among fathers
as they influence the infant both directly and indirectly.
Such an approach is more likely to identify the range

of family patterns, the range of social experiences

available to infants in their families, and the signifi-
cance of variations in family pattern for the behavior

and development of infants.

Individual Differences in Fathers

 

Though little is known about individual differences
in the behavior of fathers, there is evidence to suggest
that fathering is not a unitary phenomenon. Differences
have been noted in the extent to which fathers are present
in the home (Biller, 1976; Lynn, 1974; Radin, 1976) as
well as in fathers' participation in the day-to-day care
of children (Clarke—Stewart, 1978; Field, 1978; Pedersen
and Robson, 1969; Russell, 1978; Peshkess, 1980; Spelke,
Zelazo, Kangan, and Kotelchuck, 1973). A small number
of studies have identified variations in the behavior
of fathers toward their infants (Belsky, 1980; Clarke-
Stewart, 1978; Peshkess, 1980) and in the beliefs and
eXpectations fathers have regarding their infants (Parke
and Sawin, 1980; Peshkess, 1980).

There also is little evidence available regarding
the effects of variations in paternal behavior on infant
development. Clarke-Stewart (1978) reported that
individual difference in the behavior of fathers predicted
the infants' scores on measures of intellectual and
social competence. Such individual differences also

proved useful in Belsky's (1980) study as predictors

10

of infant exploratory competence. Differences in
fathers' participation in infant caregiving have been
linked with differences in the social behavior of
infants (Pedersen and Robson, 1969; Spelke, et al., 1970).
Such a small number of studies is insufficient for
establishing relationships between father behavior and
infant outcome with any degree of certainty. Moreover,
the correlational nature of those studies permits no
clear statement of direction of effects. The studies
do, however, suggest a working hypothesis. That hypoth-
esis is that factors such as paternal warmth, nurtur-
ance, availability, style of play, verbalizations and
expressions of affect influence the behavior and develop-
ment of infants. Such a formulation reemphasizes the
utility of approaches in which individual differences
in fathers are specified. In addition, it is conceptually
consistent with studies of later childhood linking
paternal involvement, warmth, nurturance, and acceptance
with positive child outcomes (cf. Lynn, 1974)L
Determinants of Individual

Differences in the Father-
Infant Relationship

 

 

 

Corresponding to the lack of research reporting
variations in the father-infant relationship is a lack
of research into the determinants of, or influences on,
variations in the father-infant relationship. There

is very little information regarding factors likely to

ll

influence the fathers' level of involvement in care-
giving, or the quantity, quality, or style of father-
infant interaction. There have been a small number
of studies suggestive of potentially relevant sources

of influences.

Infant Characteristics

 

In an approach assuming that infants influence their
caregivers (Lerner and Spanier, 1978; Lewis and Rosenblum,
1974) individual differences in infants become significant
as variables potentially influencing the organization
and development of the father-infant relationship.

Sex of Infant. The gender of the infant is one

 

characteristic likely to influence the father-infant
relationship. Expectant parents have been reported to
hope for boy infants more frequently than girl infants
(Hoffman, 1977). Parents of newborns describe similar
infants in a sexually stereotyped manner (Rubin, Proven-
zano, and Luria, 1974). In addition to these attitudinal
and perceptual influences, behavioral differences in

the parents' treatment of male and female infants have
been noted (Clarke-Stewart, 1978; Kotelchuck, 1976;
Lamb, 1977a). At the same time, differences in boys'
and girls' behavior toward parents has been reported
(Lamb, 1977a; Pedersen and Robson, 1969; Spelke et al.,
1973). It appears that parents show a preference for

same-sexed infants and that by the second year infants

12

begin to show a preference for the same-sexed parent.

The differential treatment of boys and girls may be cog-
nitively mediated, as evidenced by Fagot's (1974) finding
that fathers of boys see themselves as providing a role
model. In addition, although evidence is conflicting,
there may be biologically linked sex differences in
infants that contribute to the differential treatment

of boys and girls (Fitzgerald, 1977; Korner, 1974;
Maccoby and Jacklin, 1974).

Infant Temperament. Thomas, Chess, and Birch (1968)

 

use the term "temperament" to refer to the behavioral
style of infants. They have found individual differences
in infants in this domain from early in infancy and have
found some continuity in this characteristics over the
course of development. They have found that differences
in infant temperament, in interaction with differences

in caregiving environments, can significantly affect

the course of child development. There has been little
direct research as to the effects of infant temperamental
style on the father-infant relationship. Pedersen (Note
2) reported links between infant temperament and the
marital relationship, but did not report links between
temperament and the father-infant relationship. Rendina
and Dickerscheid (1976) reported that fathers were

more likely to persist in efforts to soothe tempera-

mentally difficult boys than girls. This finding,

13

however, provides very little information regarding the
impact of temperament per se.

In a tangentially related study, Scholom, Zucker,
and Stollack (1979) used adult temperament measures and
parents' retrospective reports of infant temperament
to predict adjustment of four-year-olds. Among the
findings reported was a correlation between father-
son temperamental dissimilarity and the sons' positive
adjustment and between mother-father-daughter tempera-
mental similarity and the daughters' positive adjustment.
While it is not known what sort of family interaction
patterns might have evolved from the period of infancy
through four years, it seems probable that differences
did emerge on the basis of temperamental "fit" between
family members.

Conclusions. Studies of father and infant behavior

 

that have been interactional in design, i.e. that have
focused specifically on the mutually regulated flow of
father and infant behaviors rather than time sampling
of behaviors, have shown that fathers"behaviors are
regulated by infants (Earls and Yogman, 1979; Parke and
Sawin, 1980; Vandrill, 1979). Differences in child and
father behavioral style (temperament) are likely to
influence how this mutual regulation occurs. So also
will the different infant capacities that emerge as a

function of maturation and any differences that might

14

be related to infant gender influence the interaction
process.) It appears necessary to include specification
of infant differnces in attempts to understand the

variations found in father-infant relationships.

Father Characteristics

 

Just as individual differences in infants contribute
to the organization and development of father-infant
relationships, individual differences brought to the
relationship by fathers also are likely to be influential.

Sex-Role Orientation. A number of authors have

 

suggested that the sex-role orientation of fathers may
be a more powerful predictor of paternal behavior than
biological gender (Parke, 1979; Pedersen, 1980b; Russell,
1978). It has been suggested that personality constructs
such as masculinity, femininity, and androgyny (Bem,
1974) may provide useful insight into variations in care-
giving behaviors. There is indirect evidence to support
such a view. Bem, Martyna, and Watson (1976) found that
college students who were classified as androgynous and
feminine displayed more nurturant behavior toward infants
than students classified as masculine, regardless of
the biological gender of the students.

A more direct study of the relationship between sex-
role orientation and caregiving behavior is that of
Russell (1978). He found that fathers classified as

androgynous and feminine were more involved in caregiving

15

activities and play than those fathers rated as
masculine. This relationship held across socioeconomic
status levels and children's age levels. There was also
an interaction with mothers' sex-role orientation. Mas-
culine fathers married to masculine mothers were more
involved in caregiving than masculine fathers married

to androgynous mothers and feminine mothers. The
measure of caregiving used in this study was self-report,
so it is not possible to determine the relationship
between fathers' sex-role orientation and the specific
style or quality of father-infant interaction.

Pedersen (1980b) has noted that labelling fathers
on the basis of differences in their caregiving styles
does not, in and of itself, contribute to understanding
those differences. Whether highly involved fathers
are labelled "androgynous" or "highly involved" does
not seem particularly crucial, except insofar as the
relationship between those dimensions contributes to
the validity of the androgyny construct. From the
perspective of the fathereinfant literature sex-role
orientation instruments that predict paternal behavior)
may prove most useful as sorting and classifying instru-
ments in the early stages of recruiting subjects.

Temperament. The Scholom et al. (1979) study

 

described above suggests that the temperamental similarity

of father and infant may contribute to the father-infant

l6

relationship. Moreover, the father temperamental
factor "energy” was marginally predictive of child
adjustment (p <.lO). This finding suggests that, in
addition to the "fit" between father and infant,
differences in the behavioral style of the father,
irrespective of the infant, may contribute to the
father-infant relationship.

Conclusions. There is minimal knowledge regarding
the characteristics of fathers that might contribute
to individual differences in the father-infant rela-
tionship. What evidence is available suggests that
fathers' level of involvement and the quality and
style of father—infant interaction may be a function
of a range of individual differences in fathers.

Those differences may be in the personality character-
istics (e.g. sex-role orientation) or biobehavioral
characteristics (temperament) as well as other

attitudinal and affective domains.

The Father-Mother Relationship

 

No analysis of the family as‘a social system can
divorce the relationship between two family members
from the relationships between other dyads in the family.
This point has been stressed by a number of authors
(Belsky, 1979a; Feiring and Lewis, 1978; Parke, 1979;
Parke et al., 1979; Pedersen, 1980a, 1980b; Pedersen,

Anderson, and Cain, 1980). There has been more of an

17

emphasis on linking the father-mother relationship with
the mother-infant relationship than with the father—infant
relationship. This emphasis stems from a belief that the
father's role in child development is likely to be largely
indirect, i.e. to involve influences on the child through
the mother (Bowlby, 1951; Clarke-Stewart, 1978; Lewis and
Weintraub, 1976; Parke, Power and Gottman, 1979). This
approach has been concerned with such variables as the
father's emotional and economic support of the mother

and the influence of that support on the mother-infant
relationship. Pedersen's (Note 2) report of a significant
correlation between fathers' positive evaluations of
mothers and mothers' competence in feeding is an

example of this approach.

However, there have been a small number of studies
linking the mother-father relationship with the father-
infant relationship. Pedersen et a1. (1980) compared
parent-infant interaction during periods when spouses
were interacting with each other with periods during
which spouses were not interacting with each other.

They found that parental behaviors requiring greater
focused attention, such as (a) smiling, (b) vocalizing,
(c) eye contact, and (d) active play were inhibited
during periods in which active spousal interaction

was occurring. Behavior requiring less focused

attention, such as holding, rocking, and

18

cuddling were not affected by spousal interaction.

These findings help account for the reports of a number
of investigators to the effect that overall level

of parent-infant interaction diminishes when both parents
are in the room with the infant as opposed to only one
parent (Clarke-Stewart, 1978; Lamb, 1976a; Parke and
O'Leary, 1976; Pedersen et al., 1980).

Using a somewhat narrower focus, Pedersen, Anderson,
and Cain (Note 2) correlated specific behaviors of spouses
toward each other with specific parents' behaviors toward
infants. They reported a correlation between spouses'
expressions of negative affect toward each other and
each parent's expression of negative toward the infant.
They did not find a correlation between spouses'
expressions of positive affect toward each other and
expresssions of positive affect toward the infant.

In the same vein, Belsky (1979a) reported relation-
ships between a number of spouse-spouse behaviors and
parent-infant behaviors in a home observation. Belsky
found that spouses who engaged in conversations unrelated
to their infants were less likely to show active involve-
ment in parenting behaviors. When mothers were frequently
engaged in discussion of the infant, fathers tended to
be highly involved with the infants. However, fathers'
discussion of the baby was not related to wives'

parenting behavior. Spousal relationships marked by

19

a great deal of harmony (careful listening, emotional
warmth, mutual agreement), active attempts to include
all family members in interaction, and shared pleasure
in the infant's behavior were associated with a high
level of parenting activity. This was particularly
true for fathers. Belsky (1979a) concluded that wives
may have a greater infuence on fathering than do husbands
on mothering.

The father-mother and father-child relationships
have been linked less directly in three other studies.
In the Scholom et a1. (1979) study cited previously,
correlations were reported between mother-father tempera-
mental similarity on some dimensions and child adjustment.
Johnson and Lobitz (1974) reported a negative correlation
between marital satisfaction and parental negativity
toward chidren. Finally, Heath (1976) found that fathers
rated by themselves, their wives, and their friends as
competent fathers were involved in marriages marked by
a high degree of marital and sexual satisfaction. The
evidence suggests that, whether behavioral observations
or self-report measures are utilized, links between

the father-mother and father-infant relationship emerge.

The Mother-Infant Relationship
A family systems perspective also implies that the
mother-infant and father-infant relationships influence

each other. As a consequence, an understanding of

20

differences in father-infant relationships requires

an understanding of differences in the corresponding
mother-infant relationships. Pedersen et a1. (Note 2)

are the only investigators who have reported on specific
links between mother-infant and father-infant interaction.
They reported a correlation between fathers' expression
of negative affect toward the infant and mothers'
expressions of negative affect toward the infant in an
observational setting with all three family members
present.

There have been reports of the "fit” between mothers
and fathers, i.e. which types of mothers tend to be
married to which types of fathers. Clarke-Stewart (1978)
found that fathers who engaged in the most social-physical
play were married to mothers who talked and played with
objects with their children the most. She suggested that,
given her previous findings on optimal maternal care
(Clarke-Stewart, 1973), this combination of parenting
styles might be optimal.

Belsky (1980) looked at combinations of maternal
and paternal sytles in his study of infant exploratory
competence. His analysis was concerned with which combi-
nations predicted infant exploratory competence. It
did not provide an analysis that could elucidate the
links between father-infant and mother-infant relation-

ships. However, his results suggest that certain patterns

21

of mother-infant and father—infant interaction are likely
to be more beneficial to children than others.

Another approach to the link between mother-infant
and father-infant relationships has been to assess
discrepancies in a) parents' perceptions of infant
temperamental characteristics, b) evaluations of their
children, c) beliefs about infant abilities, and d)
attitudes regarding child-rearing practices (Clarke-
Stewart, 1978; Note 1), This approach utilizes differences
in parents' cognitions as an index of the differences in
their relationship with their infants. Clarke-Stewart
found such differences to be negatively correlated with
measures of infant competence. Pedersen et a1. (Note 1)
reported a correlation between discrepancies in parents'
perceptions of infant temperament and mothers' negative
affect toward their infants. Discrepancy scores may
reflect differences between parents in information,
attitudes, values, and sensitivities (Note 1). In
addition to providing an index of concordance between‘
the parent-infant relationships, such scores may also
serve as indices harmony and agreement in the marital
relationship. Feldman (cited in Parke et al., 1979)
found a correlation between marital satisfaction and
parents' concordance on child-rearing attitudes.

In summary, there is conceptual justification for

linking the two parent-infant relationships. Whether

22

the link is through modelling, as suggested by Parke and
Sawin (1980) is not clear. Pedersen et a1. (Note 3) have
suggested that those relationships may be related to one
another in a complementary or, in some instances, compen-
satory fashion.

Summary: Individual

Differences in the
Father-Infant Relationship

 

 

 

Although the data are limited, it appears that fathers
do differ from one another in the amount of time they
spend with their infants, the types of activities in which
they engage with their infants, their attitudes and beliefs,
and the quantity and quality of their interactions with
their infants. There is some correlational evidence
consistent with a belief that these differences influence
the cognitive and social develpment of infants.

Fathers are likely to differ from one another as a
function of a number of factors. First, characteristics
of the infant and characteristics of the father and the
"fit" between those characteristics influence the develop-
ment of the father-infant relationship. Second, the
father-mother relationship and the mother-infant relation-
ship affect the father-infant relationship. Finally,
the larger context in which the family interacts, is
likely to influence the organization and development of

the father-infant relationship.

23

Limitations of Past Research

 

Russell (1978) remarked that there has been little
study made of critical factors associated with whether
and how fathers interact with their children. The range
of father-infant relationships and the determinants of
father-infant relationships have not been studied in
a systematic or sustained manner. As a result, the
conclusions reached in the present review must be regarded
as working hypotheses at the present stage of research
in this area. In addition to there being a paucity of
studies of individual differences in the father-infant
relationship, those studies that have been reported have
been characterized as "long on variables and short on
subjects" (Pedersen, 1980b, p. 147). As a result, what
little data are available must be viewed cautiously.

Part of the problem in this area is the lack of past
theory and research to guide in the formulation of
research questions. Significant determinants of indie d
vidual differences have not been clearly identified.

As a result, a large number of exploratory questions have
been asked of a very small number of fathers. Very

few studies (Field, 1978; Russell, 1978; Spelke et al.,
1973) have utilized a priori groupings of fathers. With
no rationale for selecting subjects dictated by theory

or research, it becomes difficult to recruit substantial

numbers of subjects who vary significantly on relevant,

24

meaningful dimensions.

The Present Study

 

Significance

 

The present study was an exploratory attempt to
identify characteristics of fathers, infants, marital
relationships, and mother-infant relationships that con-
tribute to the variance in a) the degree to which fathers
are involved in infant caregiving activities, b) the
quantity of observed father-infant interaction, and c)
the quality of observed father-infant interaction. A
range of variables were utilized in an attempt to describe
fathers more thoroughly than previous studies have done.
A number of these variables, which have been directly
or tangentially related to father-infant interaction
in previous research, were measured using self-report
instruments. Another goal of this study was to identify
self-report measures predictive of variations in father-
infant interaction. Instruments so identified can be
utilized in future research as a means of sorting and
classifying potential subjects. Such sorting procedures
increase the likelihood that subjects chosen for studies
will vary significantly in areas relevant to the father-
infant relationship. This is the type of approach
Pedersen (1980a) has proposed with reference to sex-

role orientation inventories.

25

Variables

 

Behavioral Measures. A time sampling pnocedure was

 

utilized to measure specific parent behaviors in a home
observation. Parent behaviors measured included those
identified as related to infant outcome in previous
studies of parent-infant interaction (Belsky, 1979a;
Clarke-Stewart, 1978; Yarrow, Rubinstein, and Pedersen,
1975; Yarrow, Rubinstein, Pedersen, and Jankowski, 1972).

The quality of parent-infant interaction was also
rated utilizing scales devised by Ainsworth and her
colleagues (Ainsworth, Bell, and Stayton, 1971). The
dimensions rated have been utilized to predict infant
outcome, particularly the quality of infant-to-parent
attachment.

Self-Report Measures. From the preceding review

 

of the literature a number of self-report variables
were identified that bear a relationship to the father-
infant relationship. The following variables were
measured through parents' reports:

1) parents' perceptions of infant temperament

2) fathers' temperament

3) fathers' sex-role orientation

4) parents' marital satisfaction

5) fathers' involvement in caregiving activities.

An additional variable, sex of infant, was also

utilized.

26

Hypotheses

 

In general, it was hypothesized that a) fathers'

involvement in caregiving activities, b) the quantity

of observed father-infant interaction, and c) the

quality of observed father—infant interaction would

be significantly correlated with measures of characteris-

tics of a) the father, b) the infant, c) the father-

mother relationship, and c) the mother-infant relation-

ship.

Specific hypotheses were as follows:

Infant Characteristics.

 

1.

Infant sex will be significantly associated
with measures of a) fathers' involvement in
caregiving, b) quantity of father-infant
interaction, and c) quality of father-
infant interaction.

Measures of infant temperamental "difficulty"
will be significantly correlated with
measures of a) fathers' involvement in
caregiving, b) quantity of father-infant
interaction, and c) quality of father-

infant interaction.

Father Characteristics.

 

3.

Measures of fathers' sex-role orientation
(masculinity and femininity) will be sig-
nificantly correlated with measures of

a) fathers' involvement in caregiving,

b) quantity of father-infant interaction
and c) quality of father-infant inter-
action.

Measures of fathers' temperamental
characteristics (activity, impulsivity,
sociability, and emotionality) will be
significantly correlated with measures

of a) fathers' involvement in caregiving,
b) quantity of father-infant interaction,
and c) quality of father-infant inter-
action.

The

27

Measures of fathers' level of involvement in
caregiving will be significantly correlated
with measures of a) quantity of father-infant
interaction, and b) quality of father-infant
interaction.

Father-Mother Relationship.

 

Measures of fathers' marital satisfaction will
be significantly correlated with.measures of
a) fathers' involvement in caregiving, b)
quantity of father-infant interaction, and

c) quality of father-infant interaction.

Measures of mothers' marital satisfaction
will be significantly correlated with
measures of a) fathers' involvement in.
caregiving, b) quantity of father-infant
interaction, and c) quality of father-
infant interaction.

Measures of the discrepancy between mothers'
and fathers' perceptions of infant tempera-
mental characteristics will be significantly
correlated with.measures of a) fathers'
involvement in caregiving, b) quantity of
father-infant interaction, and c) quality of
father-infant interaction.

Measures of the temperamental similarity
between mothers and fathers will be signifi-
cantly correlated with measures of a)
fathers' involvement in caregiving, b)
quantity of father—infant interaction and

c) quality of father-infant interaction.

Mother-Infant Interaction.

10.

Measures of the quantity and quality of mother-
infant interaction will be significantly
correlated with measures of a) fathers' involve-
ment in caregiving, b) quantity of father-infant
interaction, and c) quality of father-infant
interaction.

METHOD

Subjects

 

Subjects for this study were 25 intact families
consisting of mothers, fathers, and first-born four-
month-old infants. They were recruited by requesting
obstetricians and instructors of childbirth preparation
classes in the greater Lansing, Michigan area to dis-
tribute letters explaining the study and requesting
participation of couples in the last trimester of
the wives' first pregnancy. This study was part of a
larger research project examining parental perceptual
style and infant psychological development (Stollak
and Messéz Note 3). Interested couples volunteered by
returning a postcard included in the introductory
letter.

The sample was relatively homogenous with respect
to demographic variables. The mean age of fathers
was 27.03 years (S.D. = 3.12); mothers' average age was
25.87 years (S.D. = 3.21). These couples had been married
an average of 3.5 years at the time their first child
was born (S.D. = 1.75 years). The mean years of education
for fathers was 15.24 (S.D. = 2.01) while for mothers
the mean was 15.14 years (S.D. = 1.96). Approximately

55 percent of the subjects identified themselves as

28

29

Protestant. Twelve percent were Catholic, 2 percent
were Jewish and 15 percent ascribed to some other
religion. Sixteen percent indicated no religious
affiliation.

A wide range of occupations were represented among
the working mothers and fathers in the sample. Over
half of the working mothers held white collar and
professional positions. The same held true for the
employed fathers. In general, this sample was middle—
class, well educated and had at least one member of the
family engaged in a relatively high-status occupation.

Only seven of the twenty-five families had mothers
who stayed home full time to care for the infant. In
two families the father stayed home full-time. The
remaining 16 families utilized some supplemental care
arrangement. In those families, the mother was viewed
as the primary caregiver when the child was not with

the supplemental caregiver.

Instruments and Observational Data
Data were obtained from parents' completion of six
instruments and from two one-hour home observations.
Instruments used were as follows:

Bem Sex Role
Inventory (BSRI)

 

 

The Bem Sex Role Inventory treats masculinity and

femininity as two separate dimensions rather than as a

30

bipolar continuum (Bem, 1974). Respondents can
characterize themselves as masculine and/or feminine by
endorsement of masculine and feminine personality traits.
The scale consists of 60 adjectives (20 masculine, 20
feminine, and 20 neutral). The scale has been found to
be internally consistent (Bem, 1974) and predictive of
conceptually related behaviors (Bem, 1975; Bem and Lenny,
1976; Bem et al., 1976).

Parent Participation
Inventory (PPI)

 

 

The Parent Participation Inventory is a 17 item
questionnaire designed by Peshkess (1980) to assess each
parent's participation in housekeeping tasks and infant
caregiving activities. A subset consisting of those
ten items on the instrument pertaining specifically to
infant caregiving was utilized. Respondents indicate
whether caregiving tasks are handled by one or the other
spouse completely or mostly, or shared equally. For the
present study the inventory was scored such that a high
score indicates a high degree of father involvement in
caregiving activities. The father score was the average
of each parent's independent completion of the instrument.
The agreement between spouses regarding fathers participa-
tion was determined by calculating the correlation
between fathers' scores and mothers' scores. That correla-

tion was equal to .69. A copy of this instrument can

31

be found in Appendix A.

Temperament Scale-
Erman (TS-E)

 

 

The Temperament Scale-Erman is an 80 item forced-
choice, true-false questionnaire yielding scores on four
adult temperamental characteristics, activity, emotionality,
sociability, and impulsivity (Erman, 1977). The individual
scales have demonstrated adequate discriminant validity,
criterion validity, and internal reliability (Erman, 1977).
Each spouse was given a score for each of the four
dimensions, with a high score indicating greater endorsement
of a characteristic. In addition, the temperamental
similarity between each couple was calculated. The sum
of the absolute value of the differences between spouses
on each scale was used as an index of dissimilarity (cf.
Scholom et al., 1979).

Michigan Infant
Temperament Scale (MITS)

 

The Michigan Infant Temperament Scale is 136 item,
forced-choice, true-false questionnaire (Bonem, 1978).
It yields scores on eight dimensions, 1) activity, 2)
adaptability, 3) intensity, 4) threshhold, 5) mood,
6) approach, 7) distractability, and 8) rhythmicity. All
scales show adequate discriminant validity, temporal
stability, inter-observer agreement, and internal consis-

tency (Bonem, 1978). The average of parents' independent

32

characterizations of the infants on the adaptability,
mood, and rhythmicity scales was calculated. Thomas,
Chess, and Birch (1968) identified low adaptability,
negative mood, and low rhythmicity as characteristics

of "difficult" children. The sum of the average of

the fathers and mothers scores on these three scales was
utilized as an index of infant "difficulty." A low

score indicates a more difficult child. The interobeseive
agreement for that difficulty scale, i.e. the correlation
between mothers scores and fathers was .45.

The Locke-Wallace Marital
Scale (L-W)

 

 

The Short Form of the Locke-Wallace Marital Scale
provides a measure of marital adjustment and satisfaction.
The scale consists of 15 items descriptive of various
aspects of the marital relationship. A high score
indicates a higher level of satisfaction. Locke and
Wallace (1959) reported a split-half reliability of .90

and adequate criterion validity for this instrument.

Demographic Questionnaire

 

Each participant completed a questionnaire providing
information regarding age, level of education, current
occupation, years in occupation, income and years

married. (See Appendix B).

33

Observational Data

 

Two types of data were available from home observations
of parent-infant interaction. The first consisted of
ratings of each parent on three qualitative dimensions
used in previous research by Ainsworth and her colleagues
(Ainsworth, Bell, and Stayton, 1971). Those dimensions
are 1) sensitivity--insensitivity, 2) acceptance--rejection
and 3) accessibility--ignoring/neglecting. A fourth
dimension used by Ainsworth, cooperation--interference,
was viewed as inappropriate for use with four-month-olds
due to their limited mobility and lack of clear evidence
for intentionality.

The second type of data consists of timeésampling
of a number of parent behaviors. Behaviors recorded at
15 second intervals were: 1) tender and careful holding,
2) playful holding, 3) inept holding, 4) instrumental
holding, 5) parent within hearing distance, 6) spontaneous
verbalization, 7) responsive vocalization, 8) negative
verbalization, 9) positive affect, 10) negative affect,
11) social-physical play, 12) object mediated play.
Descriptions of these categories and training manual
are found in Appendix C.

Scores for parent behaviors to be computed were
1) number of 15 second intervals in which each parent
was verbally or physically accessible to the infant, and

2) percentage of accessible segments in which each

34

behavior occurs.

Reliability. Observers were four advanced under-

 

graduate women. They were trained uSing the manuals found
in Appendix C and that developed by Ainsworth et al. for
use with their rating scales. Observers then rated video—
taped segments of parents' interactions with infants
ranging in age from three to twelve months. Training
continued until a 90 percent index of agreement was
reached on rating scales and coding of behaviors. Obser-

vers were blind to the hypotheses of this study.

Procedures

 

Since this study is a part of a much larger project,
many of the other procedures experienced by the couple
are not directly relevant to the present investigation.
However, the entire range of procedures will be described
very briefly. (Descriptions of other parts of the project
can be found in Stollak and Messéfi Note 3; Peshkess, 1980;
and Watts, 1980).

Couples who returned a postcard indicating a willing-
ness to participate in the study were given an appointment
time during the third trimester of pregnancy. At the
time of the appointment, they came to the university,
viewed a videotape, filled out a series of questionnaires
and were interviewed. Each person was then given a

separate packet of questionnaires to be completed

35

independently at home. The couple was interviewed a
second time, this time at home. Questionnaire packets
were collected at that time. At the second interview
couples were paid $60 for their participation. They
were given a stamped, addressed postcard to return when
their child was born. The card indicated the name, birth-
date, and sex of the child.

Approximately three and one half months following
the birth of their infant, each couple was recontacted
by telephone. Two one-hour observations were scheduled
at that time. They were scheduled at a time when all
family members were expected to be home and awake. Each
spouse was mailed a packet of questionnaires to be
completed independently. This packet included all of
the instruments used in the present investigation. The
packets were picked up at the time of the first obser-
vation session. Following the second visit, couples
were paid $25 for their participation.

During each observation period, one observer was
present and an audiotape recorder was operating. The
same observer was present at both of the observation
periods for a particular family. Parents were instructed
to go about their daily routine. Observers attempted

to be as unobtrusive and uninvolved as possible.

RESULTS

Preliminary Analyses

 

Data Reducation

 

A large number of variables relative to sample size
were utilized in this study. Three behavioral ratings
and twelve behavioral categories were coded at two
different observation periods. Various data reduction
strategies were utilized in order to reduce the number
of variables to more manageable proportions. (Descriptive
statistics for all predictor and criterion variables
are contained in Appendix D.)

Ratings. Table 1 illustrates the intercorrelations
among the three rating dimensions for fathers and
mothers. The average intercorrelation for fathers'
ratings was .94. The average for mothers was .80. It
appears that parents tended to be perceived by raters
as high on all dimensions if they were high on any
dimension. This was particularly true for fathers.
Consequently, a global rating scale was derived using
the mean of the ratings on each scale.

Observation Periods. The collapsing of rating scales
left twelve behaviors and one rating measure for each
parent sampled at two points in time, one week apart.

A t test for the global rating and the twelve behaviors

36

37

Table 1

Intercorrelations Among Rating Scales

 

 

 

 

 

Fathers
Sens. Accp. Accs.
Sensitivity 1.00 .93 .97
Acceptance 1.00 .93
Accessibility 1.00
Mothers
Sens. Accp. Accs.
Sensitivity 1.00 .97 .73
Acceptance 1.00 .88

Accessibility 1.00

 

38

for fathers and mothers was utilized to determine

if there were significant mean differences from one
observation period to the next. None of the variables,
for fathers or mothers, differed significantly from one
time to the next (p <.05). The mean of the two samplings
for each of the thirteen variables for fathers and for
mothers was utilized for the subsequent analyses.

Stability of behavioral measures. Table 2 shows the

 

correlation between the variables at the first and second
observation periods. These correlations are an index

of the stability of the behaviors and ratings over a one
week period. The average correlation between time one
and time two was .42 for fathers and .45 for.mothers.

Reduction of behavioral variables. Table 3 presents

 

the intercorrelations among the father behaviors and
ratings. Table 4 illustrates the same information for
mothers. An attempt was made to reduce each of these
matrices through principle components analysis. How-
ever, no solution was found that met the criteria of

a) significant reduction in number of variables, b) simple
structure and c) meaningfulness.

In lieu of reduction of variables into factors, an
attempt was made to derive scales from the twelve
behavioral dimensions. A preliminary inspection of the
intercorrelations in the matrix of father behaviors and

mother behaviors suggested that if one excludes the

39

Table 2

Stability of Behaviors and Ratings
Across Observations

 

 

Father Mother
Global rating .76* .51*
Tender holding .33 . .75*
Playful holding .52* .24
Inept holding .06 ' -.05
Instrumental holding .26 .57*
Within hearing distance .64* .34
Spontaneous vocalization .54* .77*
Responsive vocalizations .62* .54*
Negative vocalizations .61* .12
Positive affect .53* .74*
Negative affect -.12 .12
Social-physical play .14 .74*
Object-mediated play .16 .49*

 

*p <.05

 

40

mmX H

 

no.v a:
oo.a swan
.voz.nno
on. oo.a swan
.snm-oom
NN.: Ho.- oo.H uoowmw
o>aumwoz
wo. om. mm. oo.H uoomwm
m>wuwmom
om.- «0. «mm. on. oo.H .oo>
o>wumwoz
oH. ~o.n wo.u no. mo. oo.H .oo>
* m>wmcoammm
Na. mm. mm. ow. mm. mm. oo.H .oo>
a « mnomcmucoam
mH.u mH.- Ne.- no. mo.- om. Ho.- oo.H mocmumwp
« wcwuwon
annuaz
so. NH.- oH.u no. mo. ma. Ho. mm. oo.H wcwpaos
HmucmenuumaH
mm. NH. no. 0H.u no.. mo.. no. mm.u mo.- oo.H wcwvaoc
unocH
mm. Nu. mo. mm. mm. mm. «o. mo.u «0.; 0H. oo.H wcwvaon
* a * unusuam
mm. 00.: ma. mm. ac. cm. om. oa. Ho.u se.: Hm. oo.H wcfluaon
% Hovcoh
no.1 mo.; om.u oq. mo.u um. um. mm. Hm. oo.: ma. mm. oo.H mcwumu
* « * Hmnoau

 

Mafia mafim..mwms.wmw .oo> oo> oo> Mawummm.uwcH unocH hmam .cohkwcHumm
.hno .oom .woz .mom .wmz mama comm CHSuH3

wcwumx was muoa>mnom panama wcoa< mcowumHouuoouwucH

m manna

41

mo.v as

 

oo.H xmaa
.uoz-.nno

om. oo.H head
mhnmaoom

mo. mo.u oo.H use an
o>Hum oz

no. ma. NN.- oo.H uoommm
o>wuwmom

no. mm. mm. «0. oo.H .oo>
. * m>Humwmz
Ho. No. oa.u me. no. oo.H .oo>
« m>fimcoammm

«o. oo. wH.- om. Ho.n om. oo.H .oo>
« msomcmucoum

am. No.- «N. cm.- ma.u na.u Hm.u oo.H mocmumwp
« wcwummn
Casuaz

as. mo. se.- mm. No.o so.. on. NH.- oo.a maneuon
* * HmucmesuumcH

Ha. mH.- mm». mm.. as. ao.- a~.- ea. o~.- oo.H waaeaon
a quCH

mo.u mo. ma. se.- q~.u mw.- no.n 0H. NN.- mo.u oo.H wcwcaon
Hausmam

mo.n mo.u m~.- cm. OH.- om. ma. om.- nm.u mH.u mo.u oo.H wannaos
mousse

mo.- H«.. mm.- ma. om.- Ha. NH. om. ~H.- aH.- om. «N. oo.H manna“
* * Hmnoao

 

Nmam.>mHm . mm .mmm .oo>.oo> .oo> wflwummm.umcH umocH Noam .aoH mawumm
.nno .uom . oz .mom .wmz ammm.coam :HSuwz

a

wawumm new muow>msmm Honuoz wcoa< msowumHouuoououaH

a manna

42

categories a) inept holding, b) negative affect, c)
negative vocalization and d) time within hearing distance,
the remaining variables were highly intercorrelated. Two
scales for each parent, Positive Behavior and Negative
Behavior were formed. Table 5 lists the composition and
alpha coefficients for each of these scales. Positive
paternal behavior scores correlated .45 with global ratings
of fathers on sensitivity, acceptance, and accessibility.
Negative maternal behavior scores correlated -.49 with
global ratings. Negative paternal behaviors and positive
maternal behavior did not correlate significantly with

global ratings.

Correlational Analyses

 

Zero-order correla-
tional analyses

 

 

Subsequent to data reduction, three dependent variables
remained; father participation in caregiving, positive
paternal behavior and negative paternal behavior. These
were to be predicted by two variables pertaining to infant
characteristics, six pertaining to father characteristics,
five pertaining to marital characteristics, positive
maternal behavior and negative maternal behavior. The
Pearson Product moment correlation between the three
criterion variables and the fifteen predictor variables
are listed in Table 6. That table contains 44 correla-

tion coefficients. (Participation in caregiving is a

 

Scale Composition, Alpha Coefficients,

Scale Name

43

Table 5

Temporal Stability

Scale Composition

Alpha

Stability

 

Positive Pater-
nal Behavior

tender holding
playful holding
spontaneous voca-
lization
responsive voca-
lization
positive affect
social-physical
play
object-mediated
play .78

.42

 

Negative Pater-
nal Behavior

Positive Mater-
nal Behavior

Negative Mater-
nal Behavior

negative vocalization
negative affect .70

tender holding
playful holding
instrumental holding
spontaneous voca-
lization
responsive voca-
lization
positive affect
social physical
play
object mediated
4play .64

inept holding
negative voca-
lization
negative affect .72

.63

.63

.10

Zero-Order Correlation Coefficients

44

Table 6

(two-tailed)

 

 

Father
participa- Positive Negative
tion in paternal paternal
caregiving, behavior behavior 5
Infant characteristics
sex .06 -.02 .05
temperamental E
difficulty .18 -.36* - 13
Father characteristics
masculinity -.05 .13 .29
feminity 05 .13 .30
activity - 23 .37* .03
impulsivity - 35* -.16 .13
sociability 06 .09 .09
emotionality - 15 -.09 .05
part. caregiving 1.00 17 .09
Marital characteristics
father satisfaction -.32 -.35* -.58**
mother satisfaction .02 -.31 -.49**
temperamental
dissimilarity -.03 24 .02
discrepancy in
inf. rating .15 .14 .01
Maternal Behavior
positive -.53** .08 .31
negative .24 -.18 .07
*p <.10

**p <.05

45

predictor in two sets of correlations and a criterion

in the third. Only three of those reach significance

with alpha set at .05. Three significant findings

are about what one could expect on the basis of chance
fluctuations in the data utilized in the computation of

44 correlation coefficients. Those relationships that
were statistically significant are a) negative Correlation
between fathers' marital satisfaction and negative

paternal behavior, b) a negative correlation between

 

mothers' marital satisfaction and negative paternal
behavior and c) a negative correlation between positive
maternal behavior and fathers' participation in care- I

giving.

Multiple Regression Analyses

 

Preliminary. Stepwise multiple regression analyses

 

were performed to see if any cluster of variables would
account for a significantly greater degree of variance

in the criterion variables than would individual predictor
variables. In order to avoid the capitalization on chance
that would result from entering all fifteen predictors
into an equation for each of the criterion variables,

a preliminary series of multiple regressions were per-
formed. For each criterion variables four regression
equations were calculated. For those regressions, the
predictors were grouped into 1) infant characteristics,

2) father characteristics, 3) marital characteristics

 

46

and 4) maternal behavior. A separate forward stepwise
regression analysis was done to predict each of the
criterion variables from each of the four categories

of predictors. Variables were entered into the equations
empirically. In order to be retained for subsequent

analyses, variables had to a) enter their respective

if'f'l-‘F

equations at a significance level less than .10 and b)
be preceded in their entry into the equation only by

predictors entering at the .10 alpha level.

 

Using the above criteria, father impulsivity and
positive maternal behavior were extracted as predictors
of father participation in caregiving. Positive paternal
behavior was predicted by a) infant temperamental diffi-
culty, b) fathers' activity and c) fathers' marital
satisfaction. Finally negative paternal behavior was
predicted only by fathers' marital satisfaction.

Final regression analyses. Regression analyses for

 

three criterion variables are illustrated in Table 7.

As the table indicates, positive mother behavior and
fathers' impulsivity account for 34 percent of the vari-
ance in fathers' participation in caregiving (p <.05).
Thirty-four percent of the variance in positive father
behavior is accounted for by fathers' activity, fathers'
marital satisfaction and infant temperamental difficulty.
Though the overall F test for this equation is signifi-

cant at the .05 level, none of these variables entered

 

47

 

cowuomm

 

 

 

 

 

 

moo. mm.HH mm.n om. om. mm. moo. mm.HH umHumm H HOH>mson
HmuHHmB Hmcumumm
.mumzumm m>Humwoz
Nmo. om.m om.n no. cm. on. NoH. mm.m KAuHSUHm
uMHo.QEou
ucmmcH m
«mo. oo.m mm.u NH. om. Hm. ooo. mm.m COHuomm
anumm
HmuHHmE
umnumm N u0H>m£oH
ooo. Hn.m mm.. «H. «H. Am. ooo. Hn.m muHsuom Hmauoumo
nonumm H m>HuHmom
0H0. sa.m mm.- so. em. mm. omH. mN.N sua>
. nHmHDQEH
Monumh N
wCH>Hw
moo. mm.w mm.- mm. mm. mm. moo. mm.w .nmn H nmumo
Hmcumuma CH cowuma
.mom -HoHuumm
Hmnumm
- mocmo m m mfldmnv shadow m mocmo Houcm wououcm mmum mHHmHum>
-Hmfiamnm HHmum>o mHaeHm mumsem a a «HaHuHaz -Hmficwflm en a mHanum> sumacmama

m mHan

mHan humeadm GOHmmouwmm

48

into the equation with a probability less than .05.
Finally, fathers' marital satisfaction accounts for 34
percent of the variance in negative paternal behavior

(p <.05). No other variable made a significant independent
contribution to the prediction of scores on this dimen-

sion.

 

 

DISCUSSION

The Sample

 

One of the purposes of this research was to identify
those characteristics of infants, fathers, father-
mother relationships, and mother-infant interaction which

might predict variations in the degree to which fathers

 

participate in the day-to-day care of infants, and varia-
tions in the behavior of fathers in interaction with
their infants. Unfortunately, a relatively small, homo-
genous sample of families participated in the study.
Thus, there was a limited amount of variation in either
the predictor variables or criterion measures.

The sample was well educated, white, and almost
entirely middle class. Both parents were in their mid-
to-late-twenties. Most had been married for over three
years at the time of their infants' birth. All but one
of the 50 adults in the sample had completed high school;
40 of the remaining 49 had at least two years of college
education. The sample may have been atypical in the
number of mothers who planned to stay home and care for
the children full time (28 percent). The majority of
mothers were continuing to work and using some degree
of supplemental care.

The homogeneity of the sample with regard to

49

 

50

demographic characteristics was also reflected in responses
to the predictor measures. While normative data are not
available as a point of comparison with the present sample,
the mean responses on the measures used can be compared
to the possible range of responses. On this basis, the
sample can be described as one in which both parents
are quite satisfied with their marriage, parents tend
to agree with each other regarding the characterization
of their infants temperament and they tend to be tempera-
mentally similar to each other. Fathers, while not
generally primary caregivers, were highly involved in
the care of their infants. They seem to share considerably
more of the responsibility for the care of infants than
do the fathers in Clarke-Stewart's (1978), Kotelchuck's
(1976), or Pedersen and Robson's (1969) samples. Finally,
the infants in this sample tended to be high on dimensions
of adaptability, rhythmicity, and mood. In short their
parents appeared to be finding them relatively easy to care
for.

In summary, the small number of subjects and their
homogeneity on a variety of dimensions is relevant in
two ways. First, the homogeneity of the sample is likely
to have decreased the variance of all variables, thereby
attenuating the possible range of regression coefficients.
Secondly, the homogeneity of the sample limits the

degree to which results are generalizable.

 

 

51

Parent Behavior Scales

 

The pattern of intercorrelations among paternal
behavior categories and among maternal behavior categories
(Tables 3 and 4) suggests that there are strong relation-
ships among the specific behaviors sampled. The fact
that these relationships could not be captured through
factor analytic solutions may be much more a function
of sample size than of the manner in which parental
behavior is organized. In the present sample it appears
that when fathers held their babies frequently, they
also talked to them frequently. When they talked to them
frequently, they tended to play with them frequently.
Mothers who held babies frequently tended to verbalize
frequently. Thus, it may be that activity level, as
indexed by the combined frequencies of the behaviors
sampled, is an accurate means of portraying the observa-
tional data.

However, there was reason to view a subset of the
behaviors in isolation from the remaining categories.
Conceptually, the variables a) inept holding, b) negative
vocalizations and c) expressions of negative affect
differ from the remaining categories of behavior. On
the basis of Ainsworth's descriptions of the sensitivity,
acceptance, and accessibility rating scales, one would
expect negative correlations between parents' ratings

and the frequency of inept holding, negative affect and

 

52

negative vocalizations and positive correlations between
the ratings and frequencies of the remaining behaviors.
This expectation was supported for the relationships
between ratings and verbalization categories for fathers
and between ratings and negative vocalizations for
mothers. However, the most striking difference between
this particular subset of variables (negative affect,
negative vocalization, and inept holding) and the
remaining behaviors was their relative infrequency.

For fathers, the mean frequency of each of these cate-
gories was less than,4 percent. Thus, fathers displayed
these behaviors in less than one-half of 1 percent of

the fifteen second periods in which they were within proxi-
mal or distal contact of the infant. The modal frequency
was zero. For mothers, the mean frequency of negative
vocalizations was 1.5 percent and the mean frequency for
the remaining two behaviors was less than 1 percent. Zero
was the modal frequencies for mothers in these three
categories. In short, these behaviors were rarely observed
in those parents who did display them; never observed

in most parents.

The conceptual relationship between inept holdings,
negative affect, and negative vocalization and their
relatively low frequencies led to their isolation from
the remaining nine behavioral categories for the purpose

of scale construction. The subset of each of those

53

groupings within father behavior and within mother
behaviors yielding the highest alpha coefficient were
retained as criterion variables. Their labelling as
positive behavior and negative behavior is not without
justification; nor is it without qualification. One
would expect negative outcome for infants exposed to

a high level of the negative behaviors from mothers
(Ainsworth, et al.. 1971). If similar processes operate
for-father-infant relationships, then one would expect
the same relationship to hold. However, there are
currently no data available regarding those paternal
behaviors predictive of secure attachment in the
Ainsworth paradigm.

With regard to the positive scales, their labelling
as positive is more tentative. They reflect positive
behavior in the sense of demonstrating a high involvement
of parents with infants. However, there are insufficient
data available regarding a) whether there is a differen-
tial impact of those behaviors on child development
relative to gender of parent and infant and whether
b) there may be some curvilinear relationship between
child outcome and parents engaging in these behaviors
designated as positive. Thus, the labels are not applied
out of an empirically supported position regarding the
desirability of these behaviors for optimal child develop-

ment. These behaviors are not grouped together because

54

they are all known to be "good" for babies. However,

one can hypothesize that they are ”good" for babies:

for fathers, they correlate significantly with ratings

on accessibility, acceptance, and sensitivity, and

they are behaviors that are aesthetically pleasing. Thus,
they are designated positive for the present investiga-fi

tion.

Predicting Individual Differences in
the Behavior of Fathers

 

 

Before discussing the implications of the results
of this study for the hypotheses under consideration,
two points merit some attention. First, only three of
44 simple correlations between predictor variables and
criterion variables reached significance at the .05 level.
While the significant correlations will be discussed as
if they provide support for hypotheses, it is important
to bear in mind that these significant correlations
may simply reflect the chance fluctuations in a large
number of measurements made of a small, homogeneous
group of subjects.

Secondly, correlational analysis utilizing a sample
size of 25 does not constitute a rigorous test of the null
hypothesis.' Thus, in those instances where the research
hypotheses were not supported, one cannot assume that
the null hypothesis has received empirical support. With

a small sample, restricted range of measurements, and

55

resulting attenuation of correlation coefficients, the
significance test for r carries an unacceptably high
probability of Type 11 error.

With those two limitations on the interpretation
of the present results in mind, the implications of the
findings for the hypotheses will be presented.

Fathers' Participation in
Caregiving

h -

 

 

‘.

The best predictor of the extent to which fathers
participated in caregiving tasks was the behavior of the
mother in relation to the infant. A negative correlation
between father participation and positive maternal
behavior accounted for 27 percent of the variance in
father participation. In families where fathers were the
most involved in caregiving, mothers showed the lowest
frequency of positive behaviors.

There are a number of possible interpretations for
this relationship. First, it may be that within the
observation periods, high participating fathers were
actively involved with their infants, thereby leaving
less opportunity or necessity for mothers to interact.
However, the lack of significant correlations between
fathers' behavioral frequencies and participation score
suggest that this was not the case at least within the
observation period. However, a high level of participa-

tion from fathers may establish a pattern whereby mothers'

 

56

overall involvement is lessened, as a result of their
expectation that fathers will share some of the respon-
sibilities. In this case, the direction of effects is
from father to mother, with fathers' level of involve-
ment acting to decrease the frequency of mothers'
interactions with infants.

The direction of effects may be the reverse; it may
be the frequency of mothers' interactions influences the
degree to which fathers participate in caregiving.
Mothers who are less active may allow, encourage, and/or
expect more participation from husbands. Fathers who
perceive lower involvement on the part of their spouses
may take a more active role in order to "fill in the
gaps." Again, because the relationship is between over-
all reported participation of fathers and observed
behavior of mothers, these interpretations suggest that
the effects of involvement occur as a result of the
pattern set up between the parents over time. It is
not simply due to the fact that there is a finite number
of 15 second categories in which parents could emit
codable behaviors, with one parent becoming involved
as the other becomes uninvolved.

A second variable, fathers' impulsivity, was nega-
tively correlated with participation scores and
accounted for an additional seven percent of the variance

in the participation scores. However, the significance

57

level of this relationship, with the effects of positive
maternal behavior partialled out, did not reach the
.05 criterion.

In summary, neither of the infant characteristic
(sex, temperamental difficulty), none of the father
characteristics (masculinity, femininity, activity,
impulsivity, sociability, or emotionality) and none
of the marital characteristics (satisfaction, discrep-
ancy scores) predicted the degree to which fathers
would engage in caregiving activities. The only
hypothesis to receive support was that predicting a
relationship between maternal behavior and father
participation.

The failure to find a relationship between participa-
tion in caregiving and scores on the Bem Sex Role Inven-
tory is inconsistent with Russell's (1978) report of mean
differences between androgynous, feminine, and masculine
fathers in participation in child care activities.

There are several differences between Russell's study
and the present investigation that may account for the
differences. First, Russell utilized a larger and more
diverse sample. Secondly, the children in Russell's
sample were not limited to infant age but ranged from
infancy through approximately twelve years of age.
Differences between fathers may emerge more strongly

later in the life span of the child or father. Finally,

58

Russell's larger sample allowed for an analysis of

group differences. The present investigation was

limited to tests of the strength of association between
masculinity and femininity scores and the participation
index. Bem's (1974) procedure for determining the pre-
dominant sex-role orientation results in a range of
androgyny scores that are classified as undifferentiated.
Thus, groupings on the basis of masculine, feminine,

and androgynous, exclude segments of the population who
are not clearly a) high on both masculinity and femininity
or b) significantly different in their masculinity and
femininity scores. The sample for the present study
contained ten masculine fathers, six androgynous fathers,
three feminine fathers and six undifferentiated fathers.
This breakdown does not permit testing of mean

differences between groups.

Positive paternal behavior

 

None of the correlations between predictor measures
and frequency of positive parental behavioral reached
significance at the .05 level. Several approached the
.05 level. Infant temperamental difficulty (r = .36),
fathers' activity score (r = .37) and fathers' marital
satisfaction (r = -.35) were correlated with frequency
of positive paternal behavior with a probability less
than .10. Whle these three variables in combination

account for 34 percent of the variance in frequency of

59

positive paternal behavior, the fact that none enter

the regression equation at a significance level less
than .05 suggests that they must, for the present,

be regarded as chance findings. In summary, the

present study does not support hypotheses that frequency
of positive paternal behavior can be predicted by

a) infant characteristics, b) father characteristics,

c) marital relationship characteristics or d) maternal
behavior.

Belsky's (1979a) report of a significant correlation
between marital harmony and frequency of parenting
activities was not corroborated in the present study.
Belsky observed spousal behaviors and inferred harmony
from those; the present investigation used self-reports
of marital satisfaction, temperamental similarity scores
and differences in the perception of infant temperamental
characteristics as indices of harmony. What Belsky
reported, in essence, was concordance in the behavior
of spouses toward one another and toward their infants
within a specific observation period. The self-report
measures utilized in the present study may reflect a more
global view of the marriage. To the extent that global
harmony is reflected in behavior, differences may not
become manifest in relatively short periods of time

in the presence of an observing stranger.

60

Negative Paternal
Behavior

As mentioned previously, the frequency of negative
paternal behaviors was very low (X'= .272; S.D. = .56).
The very low frequency of such behaviors may account for
the extreme deviation from normality observed in the
distribution of negative paternal behavior scores for
the sample (Kurtosis = 8.79; Skewness = 2.98). The
distribution of this variable was dramatically different
from.that of any of the predictor variables. Regression
coefficients may be calculated for data that are not
normally distributed. However, the possible range of
coefficients is restricted to the extent that distributions
of a criterion differ from distribution of predictors
(Carroll, 1961). One outcome of this phenomenon is that
fewer significant correlations than expected may result
from analyses of such variables. Thus, statistical con-
siderations may account for the finding of so few signifi-
cant correlation between predictors and negative paternal
behaviors.

Both parents' marital satisfaction scores were signifi-

cantly and negatively correlated with the frequency of

negative paternal behavior (fathers' r = -.58; mothers'
r = -.49). Because of the correlation between mothers'
and fathers' marital satisfaction (r = .42), mothers'

satisfaction score did not account for a significant

amount of variance in frequency of negative paternal

61

behavior beyond that accounted for by fathers' marital
satisfaction (r2 = .34). This finding suggests that
fathers who are most satisfied with their relationship
with their spouses are likely to be married to women who
are in turn satisfied with their relationships with their
husbands. These fathers are also least likely to express
negative affect or to verbally prohibit the baby's
activities. It would appear that men satisfied with
their marriages are also satisfied with their infants.
This finding is consistent with that of Pedersen et a1.
(Note 2) who reported a relationship between expressions
of negative affect between parents and between each parent
and the infant. Significant correlations between marital
variables and parenting behaviors provide support for

the type of family systems analysis advocated by Belsky
(1980), Parke et a1. (1979) and Pedersen (1980b).

The present study did not find a significant correla—
tion between a) infant characteristic (sex and tempera-
mental difficulty), b) fathers' masculinity, femininity
and temperament scores, c) discrepancy in parents' ratings
of infants, d) parents' temperamental similarity or e)
maternal behaviors and negative paternal behavior. Thus,
there was little overall support for the ten hypotheses

of this study.

62

Limitations of the Present Study
Perhaps the greatest weakness of the present investi-
gation has already been alluded to. The sample was
small, the range on many of the measures, especially pre-
dictors, was narrow, and therefore only the most potent
of relationships are likely to appear at a statistically
significant level. In short, subjects were too few and

too similar.

Methodolggical limitations

 

The most prominent methodological weakness of this
study lies in the lack of reliability data for the obser-
vational measures. While an initially high level of reli-
ability was obtained (.90), it is not clear how much
"drift" in raters away from that level of agreement may
have occurred in the year in which data collection took
place. While the decision not to place two raters in
homes because of the potentially inhibiting effects of
two observers on parents is valid, it may have been use-
ful to recruit a small number of families who could have
been observed as checks on reliability but who were not
included in the study.

In part, the reason for questioning reliabilities
stems from the observation that the intercorrelation
patterns between father ratings and behaviors differed

from the intercorrelation pattern for mother ratings and

63

behavior. This occurred in spite of the fact the identi-
cal behavioral definitions of sensitivity, acceptance,
and accessibility were utilized during training of
raters. It appears that fathers were rated on the basis
of observed, positive behaviors in accordance with the
training instructions. However, for mothers, frequency
of negative vocalizations and social physical play earned
lower ratings. Different criteria were applied to
mothers. One way of interpreting the data on mothers

is that mothers were "innocent until proven guilty,"

i.e. mothers were rated high unless they demonstrated
negative behaviors. Fathers, on the other hand, had to
demonstrate positive behavior to be rated highly. This
analysis does not speak directly to the reliability of
behavioral measures. It does suggest that mothers' be—
havior may have been perceived differently from fathers'.
In that case, there is reason to suspect that the coding
of behaviors may have been similarly influenced. Attention
to this possibility in training procedures and in sub-

sequent reliability checks seems warranted.

Conceptual Limitations

 

Many of the hypotheses of the present study rely on
a trait approach to parenting behavior. An underlying
assumption was that infants, fathers, and mothers bring
differences in stable traits to the parenting situation

and that these traits directly influence behavior. Bem

64

and Allen (1974) have shown the utility of considering
traits in interaCtion with situations and subjects' con-
ceptions 0f the stability and relevance of traits as pre-
dictors of human behavior.

While infants' sex and temperament and fathers' mas-
culinity, femininity and temperament may be traits that
influence father-infant interaction, the influence may
be less direct than the present investigation assumed.

For example, sex of infant has been reported to influence
the behavior of fathers (Clarke-Stewart, 1978; Kotelchuck,
1976; Lamb, 1977a). Fagot (1974) has also reported that
fathers see their role as providing a model for sons,

more so than for daughters. If one takes the type of
cognitive-mediational perspective recommended by Parke
(Parke, 1979; Parke et al., 1979) an hypothesis is
suggested, namely that those fathers who perceive their
role as providing a role model for sons are likely to
differ from fathers who do not share such a perception.
Moreover, such fathers are likely to account for more

of the mean difference betweeen behavior of fathers and
mothers. To further complicate matters, fathers may differ
in their beliefs about where and how sex-role socializa-
tion should take place. Fathers may differ in regard

to the age at which such concerns become manifest and

in what types of situations in which they see socialization

practices as relevant.

65

In summary, the kinds of traits used as predictors
in this study may well influence the behavior of fathers.
However, it is likely that more information is needed
before the nature of these relationships become clear.
How fathers see their role, how they see their own
characteristics interacting with their role as fathers,
and how mothers see their husbands' characteristics inter-
acting with parenting skills are the kinds of questions
that would help spell out the relationship between the
traits and behavior of fathers. Obviously, such an
approach generates far more information than can be
appropriately analyzed with samples as small as the one

used in the present study.

Implications for Future Research

 

The preceding discussion of the limitations of the
present study suggests directions for future research
aimed at identifying and predicting variations in the
behaviOr of fathers of infants. In addition to larger
samples, it is desirable to insure some heterogeneity
in the sample. This is particularly important for the
independent variables. As a preliminary screening
device instruments could be completed by fathers.

If one were interested in sex-role orientation, then
groups of masculine, feminine and androgynous fathers

could be identified. For other types of measures,

66

fathers falling in the upper and lower quartile on indices
under consideration could be retained for further
study.

Once a sample of sufficiently different fathers has
been identified, couples might be interviewed in enough
depth to determine the manner in which such differences
might influence behavior. With a sufficient sample size,
multivariate methods can be used to predict the manner
in which traits, cognition, and situations interact
to influence behavior.

The most promising type of measurement to emerge from
this present investigation is the Locke-Wallace Scale.
While there were few significant findings in this study,
marital satisfaction of both parents was significantly
correlated with negative paternal behavior (p <.05).
Fathers' satisfaction was positively correlated with
fathers' participation (p <.15) and positive paternal
behavior (p <.10) at levels approaching significance.
These were relationships found in a relatively satisfied
sample of fathers.' Assuming that the present study did
not tap into a narrow band of a non-linear relationship,
one could expect significant behavioral differences
among fathers with divergent levels of marital satisfac-
tion.

The other significant relationship found was between

the behavior of mothers and the degree to which fathers

67

participated in caregiving. This type of finding, in
conjunction with the marital satisfaction findings,
supports the views of a number of authors (Belsky,

1980; Lewis and Feiring, 1978; Parke et al., 1979; Peder-
sen, 1980) suggesting that we look to the subsystem
characteristics of family systems in order to account
for the behavior of family members. If the interest is
in predicting differences in the relationship of fathers
and infants then it may behoove researchers to look for
variations in family patterns in order to account for
such differences. This approach is presented as an
alternative to looking for traits/cognitions/situations
interactions. However, it may well turn out that
fathers and mothers bring differences in personality
and temperamental traits to marriages. This sets off

a series of mutually regulated interactions which carry
over into the realm of parenting (of both parents) once
an infant is born. The infant makes independent contri-
butions to these ongoing interactions. In any case,
future research will need to attend to the manner in
which characteristics of individuals within families
interact with various relationships in the family to
account for significant portions of the variance in the

behavior of fathers or any other family member.

68

Summary

This research attempted to investigate variables
associated with variations in a) the degree to which
fathers participate in caregiving activities, b) fre-
quency of positive paternal behavior and c) frequency
of negative paternal behaviors. The predictors of
these variables were grouped into a) infant characteris-
tics, b) father characteristics, c) marital relationship
characteristics and d) maternal behavior.

Only three of the simple correlations between predic-
tor and criterion variables proved significant. Given
the number of correlations computed these may well be
chance findings. Positive maternal behavior was
significantly, negatively correlated with fathers'
participation in caregiving. Mothers' marital satisfac-
tion and fathers' marital satisfaction were both
significantly negatively correlated with negative
paternal behavior. These findings were offered as
qualified support for a family systems analyses approach
to describing and predicting variations in the behavior
of fathers of infants.

The lack of significant findings was attributed to
a) a small homogeneous sample, b) an oversimplified
conceptualization of the relationship between traits and
behavior and c) possible unreliability of observational

data.

REFERENCE S

REFERENCE NOTES

1. Pedersen, F. A. Mother, father, and infant as an
interactive system. Paper presented at the annual
convention of the American Psychological Association,
Chicago, September, 1975.

2. Pedersen, F. A., Anderson, B. J., and Cain, R. L.
An approach to understanding linkages between
the parent-infant and spouse relationships.
Paper presented at the meeting of the Society
for Research in Child Development, New Orleans,

March, 1977.
3. Stollak, G. E., and Messét L. A. Parent perceptions

and infant development. Unpublished research
proposal, Michigan State University, 1977.

69

REFERENCES

Ainsworth, M. D. 8., Bell, S. M., and Stayton, D. J.
Individual differences in strange-situation behavior
of one-year-olds. In H. R. Schaffer (ed.), The
Origins of Human Social Relations. London and
New York: Academic Press, 1971.

 

Belsky, J. The interrelation of parental and spousal
behavior during infancy in traditional nuclear
families: An exploratory analysis. Journal of
Marriage and the Family, 1979, 16, 7494755 (a).

 

 

Belsky, J. Mother-father-infant interaction: A
naturalistic observational study. Developmental
Psychology, 1979, $5, 601—607 (b).

 

 

Belsky, J. A family analysis of parental influence
on infant exploratory competence. In F. A. Peder-
sen (ed.), The Father-Infant Relationship: Obser-
vational Studies in the Family Setting. New York:
Praeger, 1980. ‘

 

 

Bem, S. L. The measurement of psychological androgyny.
Journal of Consulting and Clinical Psychology, 1974,
42, 155-162.

 

Bem, S. L. and Lenney, E. Sex typing and the avoidance
of cross-sex behavior. Journal of Personality and
Social Psychology, 1976, 33, 48-54.

 

 

Bem, S. L., Martyna, W. and Watson, C. Sex typing
and androgyny: Further explorations of the
expressive domain. Journal of Personality and Social
Psychology, 1976, 34, 1016—1023.

 

 

Bem, D. J. and Allen, 8. On predicting some of the
people some of the time: The search for cross-
situational consistencies in behavior. Psychological
Review, 1974, 81, 506-520.

 

Biller, H. B. The father and personality development:
Paternal deprivations and sex role development. In
M. E. Lamb (ed.), The Role of the Father in Child
Development. New York: Wiley, 1976.

 

 

Bonem, H. Measurement of Infant Temperament. Unpublished
master's thesis, Michigan State University, 1978.

70

71

Bowlby, J. Maternal Care and Mental Health. Geneva:
World Health Organization, 1951.

 

Carroll, John B. The nature of the data, or how to
choose a correlation coefficient. Psychometrica,

1961, g, 374-372.

 

Clarke-Stewart, A. K. Interactions between mothers and
their young children: Characteristics and conse-
quences. Monographs of the Society for Research
in Child Development, 1973, 38 (6-7, Serial No. 153).

 

 

Clarke-Stewart, A. And daddy makes three: The father's

impact on mother and young child. Child Development,
1978, 42, 466-478.

 

Earls, F. and Yogman, M. The father-infant relationship.
In J. G. Howells (ed.), Modern Perspective in the
Psychology of Infancy. New York? Bruner Mazel, 1975.

 

 

Erman, H. A Multi-Trait Multi-Method Study of Adult
Temperament. Unpublished master's thesis, Michigan
State University, 1977. '

Fagot, B. I. Sex differences in toddler's behavior

and parental reaction. Developmental Psychology,
1974, 10, 554-558.

 

Feiring, C., and Lewis, M. The child as a member of the
family system. Behavioral Science, 1978, 23, 225-
233.

 

Field, T. Interaction behavior of primary versus
secondary caretaker fathers. Developmental Psy-
chology, 1978, 14, 183-184.

 

Fitzgerald, H. E. Infants and caregivers sex differences
as determinants of socialization. In E. Donelson,
and J. Gullahorn (eds.), Women: A Psychological
Perspective. New York: John Wiley and Sons, 1977.

 

 

Heath, D. H. Competent fathers: Their personalities
and marriages. Human Development, 1976, 12, 26-39.

 

Hoffman, L. W. Changes in family roles, socialization
and sex differences. American Psychologist,
1977, 32, 644-658.

 

Johnson, 8. M. and Lobitz, G. K. The personal and mari-
tal adjustment of parents as related to observed
child deviance and parenting behaviors. Journal
of Abnormal Child Psychology, 1974, 2, 193-207.

72

Korner, A. The effects of infant's state level of
arousal, sex, and ontogenetic state on the care-
giver. In M. Lewis and L. A. Rosenblum (eds.),
The Effect of the Infant on its Caregiver.

New York: Wiley, 1974.

 

Kotelchuck, M. The infant's relationship to the father:
Experimental evidence. In M. E. Lamb (ed.), TQe
Role of the Father in Child Development. New
York: Wiley, 1976.

 

Lamb, M. E. Fathers: Forgotten contributors to child
development. Human Development, 1975:.19: 245-266.

 

Lamb, M. E. Effects of stress and cohort on mother-
and father-infant interaction. Developmental
Psychology, 1976, TT, 435-443.

 

 

Lamb, M. E. Father-infant and mother-infant interaction
in the first year of life. Child Development, 1977,
38, 168-181 (a).

 

Lamb, M. E. Paternal influences and the father's role:
A personal perspective. American Psychologist,
1979, 33, 938-943.

 

Lamb, M. The development of parent-infant attachments
in the first two years of life. In F. A. Pedersen
(ed.), The Father-Infant Relationship: Observational
Studies in the Family Setting. New York: Praeger,
1980.

 

 

Lerner, R. M. and Spanier, G. B. (eds.). Child Influ-
ences on Marital and Family Interaction: A Life

Span Perspective. New York: Academic Press,
1978.

 

 

 

Lewis, M. and Rosenblum, L. The Effect of the Infant

on Its Caregiver: The Origins of Behavior (Vol. 1).
New York: Wiley, 1977.

 

 

Lewis, M. and Weinraub, M. The father's role in the
infant's social network. In M. E. Lamb (ed.),
The Role of the Father in Child Development.
New York: Wiley, 1976.

 

Locke, H. J., and Wallace, K. M. Short marital adjust-
ment prediction tests: Their reliability and validity.
Marriage and Family Livipg, 1959, TT, 251-255.

 

Lynn, D. B., The Father: His Role in Child Develepment,
Monterey, California: Brooks/Cole, 1974.

73

Maccoby, E. E. and Jacklin, C. N. The Psychology of
Sex Differences. Stanford: StanfOrd University
Press, 1974.

 

 

Parke, R. D. Perspectives on father-infant interaction.
In J. Osofsky (ed.), Handbook of Infant Development.
New York: Wiley Interscience, 1979.

Parke, R. D. and O'Leary, S. E. Father-mother-infant
interaction in the newborn period. Some findings,
some observations and some unresolved issues.

In H. Reigel and J. Meacham (eds.), The Developing
Individual in a Changing World: Vol. II, SoCiali
and’Environmental Issues. The Hague: Mouton, 1976.

 

Parke, R. D., Power, T. G., and Gottmen, J. M.
Conceptualizing and quantifying influence patterns
in the family triad. In M. E. Lamb, S. J. Suomi,
and G. R. Stephenson (eds.), Social Interaction
Analysis: Methodological Issues. Madison: Univer-
sity of Wisconsin, 1979.

 

Parke, R. D. and Sawin, D. B. The family in early
infancy: Social interactional and attitudinal
analysis. In F. A. Pedersen (ed.), The Father-
Infant Relationship: Observational Studies in
the Familnyetting. New York: Praeger, 1980.

 

 

Pedersen, F. A. Overview: Answers and reformulated
questions. In F. A. Pedersen (ed.), The Father-
Infant Relationship: Observational Studies in
the Family Setting. ,New York: Praeger, 1980 (a).

 

Pedersen, F. A. Research issues related to fathers
and infants. In F. A. Pedersen (ed.), The Father-
Infant Relationship: Observational Studies in the
Family Setting. New York: Praeger, 1980 (b).

 

Pedersen, F. A., Anderson, B. J., and Cain, R. L.
Parent-infant and husband-wife interactions
observed at age five months. In F. A. Pedersen
(ed.), The Father-Infant Relationship: Observa-
tional Studies in the Family Setting. New York:
Praeger, 1980.

Pedersen, F. A., and Robson, K. S. Father participation

in infancy. American Journal of Orthopsychiatry,
1969, 32, 466-472.

74

Peshkess, I. The Effects of Parental Participation
in Child Care Giving and of Parental Expectation
and Perceptions of the First Born Infant or
Parenting. Unpublished doctoral dissertation,
Michigan State University, 1980.

Radin, N. The role of the father in cognitive,
academic, and intellectual development. In M.
E. Lamb (ed.), The Role of the Father in Child
Development. New York: Wiley, 1976.

 

 

Rendina, 1., and Dickerscheid, J. D. Father involve-
ment with first-born infants. Family Coordinator,
1976, 33, 373-379.

 

Rubin, J. 2., Provenzano, E. J., and Luria, Z.
The eye of the beholder: Parents' views on sex
of newborns. American Journal of Orthopsychiatry,
1974, 3;: 720-731.

 

Russell, G. The father role and its relation to
masculinity, femininity and androgyny. Child
Development, 1978, 32, 1174-1181.

 

Scholom, A., Zucker, R. and Stollak, G. Relating
early child adjustment to infant and parent
temperament. Journal of Abnormal Child Psychology,

1979, 1, 297-308.

 

Spelke, E., Zelazo, P., Kaga, J. , and Kotelchuck, M.
Father interaction and separation protest.
Developmental Psychology, 1973, 3, 83-190.

Thomas, A., Chess, 8., and Birch, H. Temperament and
Behavior Disorders in Children. New York: New
York University Press, 1968.

 

 

Vandrill, D. A microanalysis of toddlers' social
interaction with mothers and fathers. Journal
of Genetic Psychology, 1979, 134, 299-312.

 

Watts, P. First Pregnancy: It's Impact on Sex Role
Orientation, Career Orientation, and Body Image.
Unpublished doctoral dissertation, Michigan
State University, 1980.

Yarrow, L. J., Rubinstein, J. L. and Pedersen, F. A.
Infant and Environment: Early Cognitive and
Motivational Development. New York: Wiley, 1975.

 

 

75

Yarrow, L. J., Rubinstein, J. L., Pedersen, F. A.,
and Jankowski. Dimensions of early stimulation
and their differential effect on infant develop-
ment. Merrill-Palmer Quarterly, 1972, T3, 205-
218.

 

APPENDIX A
PARENT PARTICIPATION INVENTORY

76

Appendix A
Parent Participation Inventory*
Now that you have lived together with your child, some of your schedules
and your division of labor have become regulated. Please specify which one
of you does the following jobs --imain1y yourself, mainly your spouse, or

both jointly.

l. Feeding my baby is done:

 

 

 

Mainly by my spouse Mainly’by myself Both jointly and’about equally

2. Cooking is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly‘and about equally

3. Washing and drying the laundry is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and about equally

4. Giving water to the baby is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and about equalIy

5. Attending the baby when it starts to cry is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and’about equally

6. Giving the baby a bath is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and about equally

7. Washing the dishes is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and about equally

8. Changing diapers is done:

 

 

 

Mainly by my spouse Mainly by myself Both jointly and about equally

*Items 1, 4, 5, 6, 8, 9, ll, 14, 15 used for Fathers' Participation
in Caregiving Score.

9.

10.

11.

12.

l3.

14.

15.

l6.

17.

77

Playing with the baby is done:

 

 

Mainly by my spouse Mainly by myself

Emptying the garbage is done:

 

 

Mainly by my spouse Mainly by myself

Talking to the baby is done:

 

 

Mainly by my spouse Mainly by myself

Minor household repairs is done:

 

 

Mainly by my spouse Mainly by myséIf

Grocery shopping is done:

 

 

Mainly by my spouse Mainly by myself

Roughhousing with the baby is done:

 

 

Mainly by my spouse Mainly by myself

Putting the baby to sleep is done:

 

 

Mainly by my spouse Mainly by myself

Punishing the baby is done:

 

 

Mainly by my spouse Mainly by myself

General house cleaning is done:

 

 

Mainly by my spouse Mainly’by myself

 

Tab—Eh

jointly

and about

equally

 

Both

jointly

and'about

equally

 

Both

jointly

and about

equally

 

Both

jointly

and about

equaIIy

 

Both

jointly

and about

equally

 

Both

jointly

andfabout

equally

 

Both

jointly

and about

equally

 

Both

jointly

and about

equally

 

Both

jointly

and’about

equally

APPENDIX B
DEMOGRAPHIC QUESTIONNAIRE

Appendix B

 

 

 

Name Age Phone #
Address
Occupation ' Years in occupation
Highest level of education completed (circle one)
Elementary grade: 1 2 3 4 5 6
Junior high school: 7 8 9 »
High school: 10 11 12
College: 1 2 3 Degree Granted
MA Degree:

Ph.D. Degree:
Other Degrees/ Vocational training

 

What is your annual income (not including your spouse's)?

 

Date of Marriage:

 

78

APPENDIX C
HOME OBSERVATION MANUAL

Appendix C
Home Observation Manual
The categories for observation are: holding, accessi-
bility, verbalization, expression of affect, play,

crying, feeding and variability of stimulation. These
each have sub-categories which are as follows:

Holding

tender and careful; playful; inept; instrumental.

Accessibility

 

within sight; within hearing distance; out of contact.

Verbalization

 

spontaneous; response to baby's vocalization; negative
verbalization.

Expression of Affect

 

positive; negative

Play

social and physical; via inanimate stimulation

Crying (this is timed in seconds)

time it takes to respond to cry; length of cry

Feeding (timed in seconds)

breat, bottle; solids.

79

80

The observations are done by one person who goes into
the home at a time when the child is awake and can be
observed in his/her normal routine for that time of
day (both parents are to be present in the home). The
entire observation takes two hours. Two observations
are conducted at one week intervals. No assumption is
made that the presence of another person in the home
will not in some way distort the parent-child inter-
action. Furthermore, no assumption is made that the
degree of distortion will be constant from one home
to the next. However, it is assumed that with an increase
in the time that the parents and observer are together,
it will become progressively more difficult for the
mother to inhibit her ordinary reaction tendencies.

Scoring will be done via checkmarks. On each page
there will be room for 14 lS-second segments. The scorer
will score by observing 15 seconds and scoring the
behaviors, again observing 15 seconds and scoring, etc.
(A tape recorder will be on for the entire visit.)

Specifications for the observations will be that both
parents be home and that the time of the appointment is a

time when the baby is usually awake. If the baby is

asleep, the interview will be rescheduled.

81

Sample Scoring Sheet:
H W or B. H-husband W-wife B-baby
HOLDING ACCESSIBILITY * VERBALIZATION

 

 

 

 

(See Scoring Sheet attached)

Definitions of
Sub-Categories

 

 

Tender and careful. This behavior is characterized
both by a gentle muting down of the parents' usual Speed
and vigor of movement and by a pacing of the temp of
the physical handling of the infant to his/her tempo
of response.

Playful. This behavior is characterized by the
playful aspect of holding--such as throwing the child
into the air, Spinning him/her around.

Inept. The parent handles the child abruptly, roughly
or very inappropriately. No Show of affection or of
playfulness (might be a form of punishment or just
thoughtlessness).

Instrumental. The parent handles the child for the
sake of performing some Specific duty, such as changing
its diaper, or moving the child from one room to another
or from the crib to the floor. This behavior is performed
more for the parent's necessity than because the child
indicated it wanted to be held.

Within sight. The parent can see the child and the

82

child can see the parent.

Within hearing distance. The parent is out of range
of sight but if the child vocalizes or cries, the parent
can hear the child.

Out of contact. The parent can neither see nor hear
the child.

Spontaneous verbalization. Spontaneous vocalization
to the child refers to any sound or words emitted by a

parent. For example, she may say, s-s-S or "sweet
baby" or any random.words or sounds. In order to check
this item the paretn's vocalization muSt have occured
spontaneously as opposed to having occurred in response
to some vocalization by the child. .This then would be
scored by H or W depending on whether the husband or
wife vocalized. Spontaneous vocalization may occur

by the baby vocalizing -- making any form of sound
(e.g., ga, ga).' This would be scored B.

Response for baby's vocalization. Responds to

vocalization of child with a vocal or verbal response.
The parental response may either be a complete word
or words or merely clearly differentiated sounds,
e.g., "ta-ta", "tSk-tsk” or "you talking to Mommy."
The key factor here is that the parent is responding
to the child's vocalization, not ignoring it.

Negative verbalization. Any form of negative

verbalization (not necessarily negative affect),

e.g. "don't touch that"; ”no, no"; etc.

83

Positive affect. The parent praises the child's
behavior verbally by saying such things as "My, wasn't
that clever" when the child does some behavior. This
might be phrased in the negative as well such as "you
really are a terror" but said with a smile on his/her
face and then followed by "you managed to pull that
cup off the table all by yourself" showing obvious
praise would still be checked.

Voice conveys positive feeling. The parent feels
good about his/her child, sounds animated when speaking
to him/her, does not use a flat or querulous tone of
voice.

Negative affect. Voice conveys negative feelings,
anger or annoyance. The parent displays displeasure
with the child and voice conveys flat or querulous tone.

Social and/or physical play. Any form of game
that is not played via inanimate objects such as pat-a-
cake or rough-housing, throwing the child up in the air
and catching him/her. The key is the child's obvious
enjoyment of the activity.

Object-mediated play. Any form of play that makes
use of a toy or inanimate instrument.

Response to cry. The length of time (timed in
seconds) it takes the parent to respond to the child's
cry (the response may be verbal or physical).

Length of cry. The total length of time (timed in

seconds) the child cries from onset to finish.

84

Feeding. Simply Stating the medium used: bottle,
breast or solids and also timed in seconds.
Variability of stimulation. The number of toys

within reach of the child.

APPENDIX D
DESCRIPTIVE STATISTICS

85

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

omuo oH. HH. . oo.mm oo.n mq.no mo.o No.o~ unmawuooama

ucmwaH mo

mwcHumm .mucoumm

uxocwnouomwa

ooHno Nm. no. oo.- oo.m H<.oH o~.o om.~H xchH zuHumHHaHm

Hmucofimuoaaoa

nonuozuuonumm

mono mm. . oH. . oo.~o oo.~m no.Hq Ho.o No.om GOHuommeumm

HmuHumz Honuoz

mono on. - on. n oo.on oo.mm Nq.~m oo.n oo.mo :oHuommeumm

HmuHumz Hannah

oHuo mH. : oo. u n.HH n.H «5.0 oo.~ ~o.m xmoSH coHu

-maHoHuuom Renown

n~-o os.H oa.~ oo.oH oo.s ~H.¢ No.m so.a suaamaoauoam “sauna

nmuo Nm. - mm. n oo.cH oo.o un.o «H.~ om.oH huHHHanoom Hosumm

m~-o so. - an. - oo.~H oo.m ao.m n~.~ so.a mua>HmHsasH panama

nNuo No. . m~.H oo.oH oo.o ~m.o -.~ oo.HH muH>Huo< umnumm

a -H no. - mo.~ we.“ nH.m am. em. so.s auacaaasoa “magma

n -H on.. «H. n oo.o oH.m so. om. nH.n huHSHHSUmmz nonumm

omuo «Ho. no. . on.oo oo.nm -.~m «H.o oH.o~ anuHSUwaHn Hounda

nmuoaBmH ucmwzH

H .o moo. NH.~u H o om. Hm. me. Now ucmwcH
nm>mummo

muz<m NHmHmmom mmmz3uxm mHmOBme SDZHx<S ZDZHZHZ woz<Hm<> .>mn .GHm Z<MZ N2<z mademds

 

moHumHuwum o>wunwuomoa

a vacmon<

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooHuo oo.m Hn.oH on.H o NH. no. oH. uuommm
o>Humwoc weapon
oouo we. «no. mn.om o oo.oNH mm.HH oH.~N uummwm
0>Huamoa uozumm
ooHuo mo.~ no.o oo.m o oo. no. on. cowumuHHooo>
o>Humwoa nonuum
ooHuo no. Hm. -.oH o mo.oH oo.o no.n cowumNHHmoo>
o>wmcooomu Hannah
ooHuo so. no. om.oo o mo.mNH o~.HH on.oH cowumNHHmuo> woos
.ucwucoam Manson
ooHuo om.Ha mu.H oo.ooH NH.o~ oo.mom No.mH oo.~o oucmumHo wcHumon
:Hnuw3 Hannah
ooHno on. No. u mn.om mm.m on.~oH oo.oH mm.n~ waHvHon Hounds
usuuuaH possum
ooa-o s~.~ HH.s m.H o «H. mm. AH. wsHeHon
uaoaH Monumm
ooa-o H¢.H as.H an.o~ o oo.Hm no.n Ho.o waaeHo:
HammoHa Hannah
ooHuo mq.H mo.H «H.5H o mn.o~ on.c om.¢ wSHuHon
Hanson Hannah
o no mo. . «H.u mo.o mo.c on.H «N.H om.n wcHumm
HonoHu Honuoz
o no m.H . oo. oo.m mu.H mm.m oo.H no.o wcHumm
HonoHu season

nu>mmmmo

muz<¢ mHmHmmom mmmz3u¥m mHmOHMD¥ ZDZHx<z SDZHZHZ MUZ<HM<> .>ma .OHm z<mz Midz mHm<Hm<>

 

muwumaumum 0>Huowuowoa

n unecoaa<

87

 

 

 

 

 

 

 

 

 

 

 

 

 

ooHuo oN.H oH. Hfi.m o oo.N oN.H m¢.H :oHumNHHmoo>
m>Humwoc Magoo:
ooH:o mo.H on. on.oN on.H «N.MN Nw.o Hn.o coHuoNHHmoo>
o>Hchomou Honuoz
ooHuo mN.H mH.N MN.oo oo.NH mn.nNH Nm.MH NN.mN :OHumuHHmoo> snow
ncmucoam uozuoz
ooHuo no.Nu on.n oo.ooH oo.nn on.om Nm.o mo.Nm moamumHo mcHuwos
cHnuH3 nonuoz
ooHno on. NH. om.no mH.o No.ooN mm.oH oo.oN wchHo£ Hmucma
unnumcH nonuoz
ooHuo mo.o MN.NN mN.o o «N. om. «N. wchHon
unocH Honuoz
ooHuo so. NH. u mN.o o No.m mm.N oo.N wcHuHon
HnwmmHa Honuoz
.ooHuo oo.H mm.H n~.om o om.HHH mm.oH oo.oH wchHon
nouns» nonuoz
.ooHno om.N on.o Nm.N o Hm. on. NN. uoH>o on
Hwauouma o>Huw oz
ooHuo on. Nm. oo.oN on. Hm.NN Nn.o Nq.o uoH>mnon
. Hmcuouma o>HuHmom
ooH-o mm.H Ha. Hs.HH o OH.NH ms.m kH.m aaHa umumauoa
unconno nonumm
oo~-o mm. on. on.mH o Ha.n~ ao.n sm.m swam Hmosmsaa
uHmHoom nonumm

nm>mmmmo

Nuz<m mHmHmmom mmMZSNfim mHmOHme ZDZHx<Z ZSZHZHZ moz<Hm<> .>mn .nHm z<mz m2<z NHm<Hm<>

 

moHumHumum o>HuoHuomon

a xHocmaa<

88

 

 

 

 

 

 

 

ooHuo oN.H no.H Hm.m o mm.H nH.H om. HOH>m£mn
Hmnhouma m>Humwoz
ooHuo mm. oq. u oo.NN nm.m oo.oN wo.n no.mH u0H>mson
Hmcuouma w>HuHmom
ooa-o «a. oo.H- as.” o o~.m sm.~ m~.m amHa emumaema
:uoonn0nuonuoz
ooa-o o~.N sa.n oa.mH o Ha.aH ss.s AH.s swam Hmuamaaa
:HmHoom Honuoz
ooHno oH.N on.m on.n o mo.N No.H no. uoommm
o>Humwma umnuoz
ooHno on. oN. om.mn NN.¢H Ho.omN om.mH Nn.nm uommmm
o>HuHmoo Honuoz

nm>mmmmo

muz<m mHmHmmom mwmz3mxm meOHMDM XDZHN<Z ZDZHZHZ moz<Hm<> .>mo .QHm z<m2 M242 MHm<HM<>

 

moHumHuwum o>HuoHHomoo

a anCoom<