THE RELATIONSHIP OF SEX AND TEMPERAMENT TO
MUTUAL GAZE BEHAVIOR IN EARLY
JOANN HOWARD PATRAY
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
My memories of my own children and my pediatrician husband's particular
ability to talk to babies with his eyes suggested the basic hypothesis
for this study. Along came Becca and Elyse and the evidence from both of
them was convincing--that for the most part, infants do communicate with
I am particularly indebted to Dr. Ira Gordon for many things, but
especially for his geniune enthusiasm for and expertise in infant research
and his provision that I have the opportunity to be a part of the Parent-
Infant Transaction and Infant Competence Project from its inception.
Appreciation is also extended to Bill Huitt, Lee Cohen, Karen Long, Pat
Schlenker, Louise Stephenson, and Dr. and Mrs. Robert Soar for keeping
the project going. It is the only place I know where things get done ahead
of time. And, of course, a special thanks goes to the fathers, mothers,
and babies of the project. Without their generous contribution of time
and effort, there would be no study.
There are so many special people who expected me to muddle through
even when I was certain I couldn't. I don't laugh anymore at the long
lists on the acknowledgements pages of my peers' dissertation. I know
only too well what they are attempting to say. There is grateful acknow-
ledgement for the contributions of the following.
Pete and Helen Alderman Dr. Vynce Hines
Mac and Amanda Baker Judith Tate
Barbara Bowers Frank and Bobbie Walrath
Dr. Deborah Coggins The "Monday Night Group"
Marsha Granoff and Elyse The "Friday Noon Group"
Gil and Pauline Hoffman
Two people require special recognition. I suspect they know the words
on these pages by memory almost as well as I do. My thanks to Marian
Tillotson, for it is a rare friend indeed that reads another's dissertation
over and over and continues to venture suggestions--sometimes at consider-
able cost to the peacefulness of the surroundings. In a real sense Marian
helped me appreciate what precision and order in language--and statistics--
really mean. Linda Drake was not merely the typist par excellence in this
seemingly impossible task, but she wanted "to get it right" as badly as I
did. And in the final crunch that is particularly encouraging and admirable.
There is no way to adequately acknowledge the contributions of a busy
husband (who always found time) and my three busy teenagers (one of whom
graduated from high school and left for college in the interim of this
study). They know, I guess, that they share in large measure whatever I
have accomplished. I only hope it compensates for the irascible behavior,
vacant stares, forgotten phone calls, etc., ad infinitum. Other members
of my family who have also contributed immeasureably are my mother--
Juliette Howard, my sisters--Charl Cappellini (mother of Becca) and Topsy
Sims, and Art and Barbara Patray.
Finally, members of the graduate committee know only too well of my
debts to them. There's a long trip between being born in McAlpin, Florida,
practicing nursing, and writing a dissertation (and that last hurdle--the
dissertation--was no doubt one of the biggest). Dr. Pat Ashton assumed
the chairmanship of this committee in midstream (when Dr. Gordon left the
University to become Dean of the University of North Carolina School of
Education). I don't think I have ever enjoyed working with a more sup-
portive person (and that includes the times we were discussing the 4th -
5th revision of some paragraphs). I never left her office that I didn't
know more or feel better able to accomplish whatever was necessary at
the time. Dr. Bill Ware provided the statistical expertise and the
measured challenge of "that's your problem" and also his well-timed
sympathetic "it's really difficult to write up a study with nonsignificant
results." Dr. Polly Barton prior to graduate study has been my mentor,
friend, and cohort. She provided valuable insights at critical junctures
all through this graduate experience ("like be sure and take the time to
enjoy what you're learning"). Two additional people have attended
several graduate committee meetings and have made very valuable contri-
butions and suggestions to the writing of the dissertation. They were
Dr. Dorris Payne and Dr. Pat Clune.
In closing, there is one final note of special thanks to Mac Baker.
At long last, Mac, I have planned my work and worked my plan.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS .................. . . .. ii
LIST OF TABLES . . . . . . . . .. ... .. vii
ABSTRACT . . . . . . . . ... .. . . . viii
1 INTRODUCTION . . . . . . . . . . . 1
Statement of the Problem .... . . . . . 4
Purpose of the Study . ... .... . . . . 4
Definition of Terms ............... .. ... 5
Need for the Study ..... . . . . . ... 6
2 REVIEW OF RELATED RESEARCH .... . . . . . 7
Importance of Early Visual Experiences to Infant
Development and Parent-Infant Relationships . . . 7
Factors that Affect Early Parent-Infant
Gaze Interactions ...... . .... . .. 17
Methodological Considerations in the Studies Reviewed . 25
Summary . . ................. . 29
3 METHODOLOGY . . . . . . . . . .31
Sample .................. ....... 32
Procedure . . . . . . . . ..... 34
Instrumentation . . . . . . . ... . . 36
Design .. . . ... . . . . . . .. 41
4 ANALYSIS AND INTERPRETATION OF DATA . . . . .... 44
Analysis of the Data . . . ...... . . .. 45
Interpretation of the Data .. ... .. . . 48
Summary . . .... ....... . . . 52
5 SUMMARY, CONCLUSIONS, AND IMPLICATIONS . . . .... 54
Summary ........... . . . . . . 54
Conclusions . . . . . . . . .. . . 57
Implications . ... .. ............... 58
APPENDICES . . . . . . . . . . ... 61
Appendix A: Questionnaire for Parents . . . . . 61
Appendix B: Parent-Infant Transaction Project -
Home Interview . . . . . . . 62
Appendix C: Informed Consent, University of Florida . 64
Appendix D: Directions for Parents for Games ...... 66
Appendix E: Perception of Baby Temperament Scales
and Coding Technique . . . . .... 67
BIBLIOGRAPHY .. ........ . . . . . . 69
BIOGRAPHICAL SKETCH . . . .... ........... 76
LIST OF TABLES
1. Parent-Infant Sex Dyadic Gaze Behavior: Transformed
Means, Standard Deviations, and Within Group
Variance Covariance Matrix . . .. ......... 46
2. The Relationship of Gaze Temperament: Means, Standard
Deviations, Correlation Coefficients . . . .... .49
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THE RELATIONSHIP OF SEX AND TEMPERAMENT TO THE
MUTUAL GAZE BEHAVIOR OF EARLY
JoAnn Howard Patray
Chairman: Dr. P. T. Ashton
Major Department: Foundations of Education
This study explored the relationship of sex and temperament to the
mutual gaze behavior of parent-infant dyads. It was hypothesized that
there are differences in the mutual gaze encounters among father-son,
mother-son, father-daughter, and mother-daughter pairs. It was further
hypothesized that there is a relationship between the parental perceptions
of their infant's activity and mood level and the mutual gaze behavior of
The study sample consisted of 42 professional and white collar couples
and their firstborn infants, of which 26 were boys and 15 were girls. Data
were collected when the infants were 13 weeks of age. Structured play
interactions between the fathers and the infants and between the mothers
and the infants were videotaped separately in a laboratory setting. Raters
coded brief and prolonged mutual gazes by viewing the videotaped parent-
infant interactions. Parental responses to the Perception of Baby Tempera-
ment Scale were used as estimates of infant activity and mood temperament
In this study, Hypothesis I and II were analyzed separately. The
first hypothesis dealt with the parent and infant sex effect on parent-
infant mutual gaze behavior. The research design for the analysis was a
split-plot, repeated measures, two-way factorial. Brief and prolonged
gaze were the dependent variables and parent sex and infant sex were the
independent variables. A multivariate analysis of variance revealed no
significant differences due to sex in the parent-infant mutual gaze
behavior. Hypothesis II postulated a relationship between infant tempera-
ment and parent-infant mutual gaze behavior. Eight Pearson Product Moment
correlations tested the relationship of fathers' and mothers' perception
of their infants' activity and mood temperament with brief and prolonged
mutual gaze. The correlation coefficients for all these relationships
were nonsignificant. Conclusions for this study included: there was
limited gaze behavior among the parent-infant sex dyads in the structured
laboratory setting of this study, and the mutual gaze behavior that was
observed was not systematically related to parent sex, infant sex, or
the parents' perception of their infants' temperament.
These findings should be interpreted within the limitations of this
research situation. The limitations that were identified were a single
abbreviated period of data collection, a homogeneous, volunteer sample,
the laboratory setting, and the increased probability of error due to the
multiple tests of correlation. It was also possible that maturational
level of the infant and reactivity of the laboratory setting may have
functioned to decrease mutual gaze behavior between parents and their
infants. In addition, a lack of sensitivity in the measurement of the
temperament variables may have produced misleading non-significant findings.
Results from this study indicated that parents and their 13 week old
infants in a structured situation exchanged very few mutual gazes. Follow-up
studies are needed which describe the evolutionary development of the
important parent-infant gaze process. These studies should include multi-
variate approaches to the study of interactive gaze data in many contexts
and with a wide variety of cultures over time. These studies could pro-
vide a more rational basis for selection of variables to be examined in
the more contrived laboratory investigations of the detailed sequential
effects of gaze behavior. Previous reputable studies have also reported
no sex differences in gaze frequency. However, a few of the studies have
documented that sex differences do appear in the long range effects of
gaze behavior. Further study is necessary to document the long range
transactional effects of sex and gaze behavior on infant development.
Additional systematic study of methods to measure infant temperament is
also indicated. Perhaps through more precise estimates of variation in
infants' styles of responding, the role of infant in early parent-infant
relationships can be better described.
Recent research has shown that infants from birth are actively in-
volved in interactions with their parents or caregivers. In this early
period, parents and infants are establishing formative interactive
patterns that will have a lasting effect on their attachment to one
another and the infant's subsequent development. Yet little research
data are available concerning the factors which affect the give-and-take
of early interactions, particularly those before the infant is six months
of age. This study was designed to examine the relationship of parent-
infant mutual (reciprocal) gazing to four variables--parent sex, infant
sex, infant activity level, and infant mood quality.
The gaze encounter has been called the first system of communication
available to both parents and infants, the basis for human sociability,
and the cardinal attachment behavior (Rheingold, 1961; Robson, 1967; &
Stern, 1974). Long before human infants can cling to their parents or
follow after them, babies elicit their parents' attention by looking at
them (Rheingold, 1961). In a recent study, Klaus, Kennell, Plumb, and
Zuehlke (1970) documented that mothers immediately after birth express a
marked interest in their babies' eyes. Stern (1974) found that 3 to 4
month old infants in a play situation with their mothers initiated and
terminated 94% of all the mutual gaze episodes in the mother-infant play
interactions. According to Robson (1967) and the studies he cited,
mothers report strong positive feelings for their babies and caregiving
becomes less routine and more enjoyable when eye-to-eye contact develops.
Numerous studies have confirmed that there is an important relationship
between early parent-infant gaze behavior and subsequent parent-infant
attachment (Blehar, Lieberman, & Ainsworth, 1977; Brazelton, Koslowski,
& Main, 1974) and infant social, affective, and cognitive competency
(Beckwith, Cohen, Kopp, Parmalee, & Marcy, 1976; Clarke-Stewart, 1973;
Goldberg, 1977; Gordon, 1974; Yarrow, Rubenstein, & Pedersen, 1975).
Conversely, a lack of parent-infant gaze behavior has been documented
in the following conditions--childhood depression (Spitz, 1965), autism
(Hutt & Ounsted, 1966), children who "fail to thrive" (Brazelton, 1974),
and neglected and battered children (Gil, 1970).
Very little information is available concerning the specific effects
of sex differences of parents and infants in reciprocal gaze interactions.
For reasons that are not too clear, the father's role in early infancy
has been remarkably neglected in the research study of parent-infant
interactions (Kotelchuck, 1974; Lamb, 1974, 1975, 1977). However, there
are a few contemporary studies that have examined the looking behavior
of infants, fathers, and mothers. Lamb (1977) studied infants 6 through
15 months of age and their parents and found that the infants looked at
their fathers more than their mothers, and Parke (1974) reported that
mothers and fathers gazed at their newborn boys more than they gazed at
their newborn girls. In none of these studies, however, was the recipro-
cation of the partner's gaze discussed, thus limiting interpretation in
terms of the interactive gaze process. There is some evidence that
looking behavior varies systematically in the first two years of parent-
infant relationship due to both parent and infant sex (Gordon, 1974, 1976;
Lewis & Weinraub, 1974). Further study of the sex effect on reciprocal
gaze of parents and infants is indicated and may help clarify the differ-
ence between mother and father roles during the early formative period
of the infant's development. It may also help to confirm the presence
or absence of early sex differences in infants.
Temperament characteristics, such as activity level and usual mood
quality, may also affect the infant's looking behavior. It is uncertain
how the infant's style of responding (temperament) may, in turn, affect
the parent's interactive gaze behavior. For example, Escalona (1968)
examined the relationship of infant activity level to a number of
variables, one of which was the visual behavior of the infant. She found
that inactive infants are more likely to use the visual modality than
the active infants are, and that inactive infants recognize their mother
at an earlier age. Stern's finding (1974)--that when mothers become
aware of their infant's gaze, they respond quickly and in a consistent
pattern by gazing back at their infant--suggests that perhaps the in-
active babies' increased visual behavior may affect parental visual
response. However, additional research of the reciprocal gaze process
is necessary to validate this suggested relationship.
The influence of infant mood temperament on parent-infant mutual
gazing is more difficult to predict due to the paucity of research
findings related to infant mood temperament. Clarke-Stewart (1973)
found that infants who are more positive in their responses (which
specifically, but not exclusively, included looking behaviors) are
more likely to participate in an increased number of social interchanges
with their mothers. Brody and Axelrand (1970) documented that babies
who are frequently crying, irritable, and uncomfortable are less likely
to be available for socialization than babies who tend to be more
pleasurably aroused. Conversely, Robson, Pedersen, and Moss (1969)
found that mothers of the more irritable boy babies in their study were
more likely to participate in eye-to-eye contact with their sons than
the mothers of the self-quieting or less labile girls. These data suggest
that the reciprocal gaze process, like all other human interactions, may
be affected by many complex interrelated variables. Research efforts
designed to study this process must recognize this complexity. To date
very few researchers have done so.
Statement of the Problem
Although there are considerable data which support the central role
of the interactive gaze in early parent-infant attachment and subsequent
infant competency, very little is known about the factors which affect
this important formative process. There are data using static gaze
criteria (such as parent or infant looks in direction of the other
participant in the interaction) which suggest that sex of the parent and
the infant and characteristics of the infant, such as activity and mood
temperament, are factors which contribute to parent and infant gaze
variation. Further research which specifically includes the study of
reciprocal gaze effects would help to clarify the role of these factors
in the give-and-take of the earliest parent-infant interactions.
Purpose of the Study
The purpose of this study was to examine the relationship of four
variables to the mutual gaze behavior of parents and infants. The first
two variables of the study were sex of the parent and sex of the infant.
The four possible parent-infant dyads--father-son, mother-son, father-
daughter, and mother-daughter--were compared to determine if there were
differences in the number of brief and prolonged mutual gazes. The
second two variables examined in the study were temperament variables.
The parents' perceptions of their infant's activity level and general
mood quality were studied to determine their relationship to mutual gaze
The study examined the following two questions:
1. Are there differences in mutual gaze behavior patterns among
father-son, mother-son, father-daughter, and mother-daughter dyads?
2. Are there relationships among the parents' perceptions of
infant activity level and quality of mood and the mutual gaze behavior
of parent-infant dyads?
Definition of Terms
For the purpose of this study, the terms listed below were defined
Mutual Gaze: Sustained mutual eye contact between the child and a
parent. The episode was defined as brief if eye contact was sustained
for at least one second. Episodes which last for more than three seconds
were defined as prolonged.
Activity Level: The tempo, intensity, and frequency of the motor
component of the child's behavior. Parental responses to the Perception
of Baby Temperament Scale for Activity (Pedersen, Anderson, & Cain, 1976)
were used to estimate activity level in this study.
Prevailing Quality of Mood: The amount of pleasant, joyful, and
friendly behavior as contrasted with unpleasant, crying, and unfriendly
behavior of the infant. Parental responses to the Perception of Baby
Temperament Scale of Mood (Pedersen, Anderson, & Cain, 1976) were used
to estimate mood in this study.
Need for the Study
There is general acknowledgement in education that infancy plays a
critical role in human educability; however, few researchers have been
able to identify the connecting links that would help to explain this
assumption. In the last few years with some consistency, investigators
have found a significant relationship between early gaze behavior of
parents and infants and the subsequent competence of the infant (Beckwith,
et al, 1976; Clarke-Stewart, 1973; Gordon, 1974). As a result, further
study of this early influential process is indicated.
Although there is general agreement at the present time that infants
actively contribute to early interactions and that the father's role is
an important one in the formative period of development, little is
actually known about the infant and father's specific contribution to
reciprocal relationships. The exploration of these issues through the
examination of the manner in which the factors--parent and infant sex
and infant temperaments-relate to early mutual gaze behavior of parents
and infants would help to provide timely and valuable information. It
is assumed that this study could provide initial data in these particular
areas which would be especially useful to investigators of more compre-
hensive projects in the future and to educators and counselors involved
in programs which involve similar parent-infant groups.
REVIEW OF RELATED RESEARCH
Research has shown that the reciprocal or contingent gaze behavior
of parents and infants plays a central role in the development of infant
sociability, cognitive competence, and the parent-infant attachment
bond. However, little is known about the factors which affect the early
interactive gaze process. There is some limited evidence that gender
of the parent and the infant, as well as the infant's style of responding
in relation to activity and mood temperament, may influence this early
formative process. The purpose of this review is to survey the available
data concerning these issues. The specific topics included in this dis-
cussion are infant gaze capability, the meaning of the infant's visual
experiences, parent-infant gaze reciprocity and attachment, the contingent
gaze and infant competence development, and parent and infant sex and
infant temperament as variables in parent-infant mutual gaze behavior.
In the final section of the review, methodological considerations in
interactive gaze research are discussed.
Importance of Early Visual Experiences to Infant
Development and Parent-Infant Relationships
The visual experiences of infants play a pivotal role in early
development. Humans are visually oriented organisms. This is particu-
larly true of the immature infant. Long before the infant can deliber-
ately smile, reach out, grasp, or crawl towards parents or significant
others, the baby is capable of using gaze to explore the world or to
communicate his or her interest or disinterest (Appleton, Clifton, &
Infant gaze capability
Research has shown that infants can see at birth. They can dis-
tinguish form, contour, brightness, and color (Appleton, et al., 1975;
Haith & Campos, 1977). Their ability to focus and to track (follow with
their eyes) moving stimuli develops rapidly in the first 3 months of
life (Kalninis & Bruner, 1973; White, 1971; White, Castle, & Held, 1964).
Babies can reduce their state of arousal by turning away from a stimulus
that is too intense or too discrepant (Stern, 1974) and can seek a new
stimulus if they become bored (Fantz, 1967). The onset of the infant's
gaze can signal a readiness to engage in an interaction, or gaze aversion
can "cut off" or terminate a social exchange (Hutt & Ounsted, 1966).
The meaning of the infant's
There is a subtle trap, according to Nunnally and Lemond (1973), in
assuming that the hedonistic motive of adult gazing applies to infant
gazing. With adults, one can freely interchange the terms "stimulus
selection," "preference," and "pleasantness." These authors reported
that studies to date indicate that the "emotionality principle" is
operant in children. An infant is more likely to look at stimuli in the
following order of preference--highly pleasant, highly unpleasant
(noxious, but novel), neutral; whereas, in an adult, the order is more
likely--highly pleasant, neutral, highly unpleasant.
In a review of the theories of infant visual perception, Bower (1966)
reported that, in general, the nativists believe infants can see what
adults see, and the empiricists think infants perceive "buzzing confusion."
Fantz (1967) contended that the evidence that from birth infants have
visual preferences indicates that infants have the capacity to receive
and discriminate patterned stimuli. Both Bower (1974) and Fantz (1967)
seem to agree that infants are probably aware of the same information
that adults are, but infants lack the ability to process and utilize all
the information that they visualize. With maturation and exposure to
diverse stimuli, these theorists believe babies gradually become better
able to integrate and discriminate the various patterns in stimuli and
to attend to increasing numbers of stimuli simultaneously.
Piaget (1970) postulated that more than a "simple copy of external
objects" or a "mere unfolding of structures" (maturation) is involved
in infant awareness and understanding of visual experiences. He con-
tended that the "establishment of cognitive . relations . .
involves a set of structures progressively constructed by continuous
interaction between the subject and his external world" (p. 703).
Piaget theorized that at first the infant "looks for the sake of
looking" (1932, p. 170). Objects and people are luminous spots. In
the first few weeks of life, these images have little meaning other than
brief stimuli, for they are not coordinated with sucking, grasping, or
anything. Gradually people and objects are "assimilated"; the infant
recognizes them and smiles. This, according to Piaget (1967), does not
mean the infant conceptualizes objects separate from himself; they are
instead tangible apparitions. The visual images become progesssively
connected with hearing, grasping, and touching. The baby begins to see
form, prominence, depth, and distance; he or she begins to see things in
perspective and to "look in order to see" (Piaget, 1952, p. 70). At
about 3 to 4 months, the child begins to give things "meanings" and to
act on things with "procedures to make interesting sights last" (Piaget,
1952, p. 153).
In summary, young infants have competence comparable to their
parents in their control of eye behavior (Stern, 1974). Through pro-
longation of gaze or by looking away, infants can communicate their
interest or disinterest. Then too, they are able to elicit the attention
of their parents by engaging them in reciprocal gazing. Taking this
relatively mature function of the infant into consideration, researchers
have found the study of reciprocal gaze interactions a profitable mode
in which to study the earliest parent-infant relationships.
Parent-infant gaze reciprocity
Until the last decade the human infant was considered a passive
organism whose development was largely determined by environmental
shaping and maturational forces. In 1971 Bell argued that infants should
be viewed as active, dynamic contributors to the earliest social en-
counters. Later studies have documented the powerful effect that in-
fants have in early socialization. One study that demonstrated that the
infant's influence begins early--often with the first fleeting gazes
exchanged between parents and infants--was Klaus, Kennell, Plumb and
Zuehle's (1970) investigation of the mother-infant relationship in the
first few minutes of life. They videotaped 12 mothers from different
backgrounds during their initial contact with their infants and found
that within the first 5 minutes all of the mothers had placed their
babies in the "en face" (eye-to-eye) position and had become remarkably
interested in the eyes of their infants. An intense interest in waking
the infant was verbalized by 73% of the mothers; the researchers quoted
one mother as saying "If you open your eyes, I will know you are alive"
(p. 190). The investigators concluded that the results of their study
revealed strong evidence for species-specific behavior in human mothers
and that early eye-to-eye contact of mothers with their babies may be one
of the "innate releasers of maternal caretaking responses" (p. 191).
Researchers who have studied mother-infant bonding (the strong
attachment between parents and infants) suggest that there may be a sen-
sitive period in which eye-to-eye contact should occur for the development
of optimum attachment of the mother (Bowlby, 1969; Brody & Axelrand, 1970;
Klaus & Kennell, 1976). In addition, there are descriptive studies which
suggest that there may be an early critical period for father-infant
bonding as well (Lamb, 1975; Nash, 1976; Parke, 1974).
Stern (1974) carefully documented the sequential reciprocal effects
of gaze behavior between mothers and their 3- to 4-month-old infants
The presence of the infant's gaze, especially in combination
with other expressive behavior, intends to elicit unique maternal
facial and vocal behaviors and tends to hold the mother's gaze
on the infant. The mother then modulates the stimulus config-
uration of her facial and vocal performance, using as cues the
infant's state of arousal and affect and the quality of his
visual attention. (p. 207).
Stern (1974) concluded that the import of these early social exchanges is
that the infant acquires experience in regulating his or her state of
arousal on the basis of another's interpersonal behavior.
Further documentation of the infant's dynamic role in early gaze
interactions can be found in Brazelton, Koslowski, and Main's (1974)
study of infant attention cycles. These researchers compared one minute
observations of babies interacting with their mothers and one minute
observations of babies playing with a toy. The 5 babies in this study
were followed from 5 to 20 months. There was a marked variation in
visual attention of the infants in both degree and duration for the two
conditions over this period of time. When viewing the toy, infants
systematically exhibited a build up of intense interest, then abrupt
withdrawal with gaze aversion. In the sequences with the mothers,
infants had periods of attention and nonattention to which the mothers
developed a sensitivity. The mothers seemed to realize that the infant's
looking away reflected a need for the baby to maintain control over the
amount of stimulation received. As a result, mothers modulated the
intensity of the stimulus or allowed the infant time to recover before
continuing the activity. Through this process, the researchers posited,
the infant learns to master internal or physiological needs and to attend
to the external world. The mother becomes sensitive to her infant's
capacity to receive and utilize stimuli. Gradually, after establishing
"communication," she begins to "teach" the infant to expand this ability
and to attend to new stimuli.
In addition, conclusions from this study (Brazelton et al., 1974)
help to clarify the probable relationship between early gaze interactions
and the subsequent attachment of mothers and infants. The authors stated
that in the mother-infant sequences, "no actor's behavior was ever
independent of the expectancy of interaction" (p. 69). The cycle being
used by each partner was approach, then withdrawal to wait for a response
from the other participant. When the ongoing behavior of the interaction
was sympathetic to the needs of each member of the dyad, there was a
sense of rhythmic interchange in the interaction which was sensed as
"positive." If, on the other hand, one of the members was out of phase
with the other, there seemed to be a negative quality to the interactions.
The researchers concluded--
The smoothness with which these dyads made such adjustments
(to each other's needs) reflected the depth of their attachment
and probably contributed a further opportunity for learning about
each other member. (p. 74)
The relationship between early gaze behavior and attachment was
examined in two studies of the first year of life (Blehar, Lieberman,
& Ainsworth, 1977; Clarke-Stewart, 1973). In the Blehar et al. study,
mother-infant pairs were rated on quality of attachment, and judged to
be "securely" or "insecurely" attached. The securely attached mothers
were found to be more contingently responsive in pacing their interactions,
and they participated in longer face-to-face encounters with their
infants. The infants in the securely attached group were judged to be
more positive in their expressions. Conversely, in the insecurely attached
mother-infant group, encounters were more brief, and mothers in this
group were more likely to initiate interactions with silent impassive
faces and often failed to respond to their babies' attempts to elicit
interactions. Oddly enough, the number of face-to-face encounters did
not differ significantly for the two groups. Duration of the episode
and positive responsivity of the baby seemed to be the more critical
variables in measuring this process. In another study of mother-infant inter-
actions, Clarke-Stewart (1973) found a significant relationship between
the child's looks, smiles, and verbalizations and the mother's expression
of affection and attachment.
The lack of eye contact in attachment disorders has been well
documented. Robson (1967) described mothers who failed to participate
in eye contact with their babies during caregiving as "psychologically
absent" in their ministrations. Conditions that have been associated
with a lack of gaze behavior or with an increased incidence of gaze
aversion have included childhood depression (Spitz, 1965), autism (Hutt
& Ounsted, 1966) and neglected and battered children (Gil, 1970). The
bizarre eye movements of children who "fail to thrive," according to
Barbero (1974), disappear much earlier in the treatment of the children
than do the physical and neurological problems. He states that it is
only after "normal eye contact" develops that the children begin to
Recent studies with blind babies and their parents have also pro-
vided insight concerning the importance of vision in early development
(Stern, 1974). Fraiberg (1974) reported that mothers found the early
behavior of their sightless infants disconcerting. The infants seemed
"depressed," "bland," and "had no expression or affect." In Fraiberg's
study, parents were taught to look for expressive hand movements of
their infant, rather than facial brightening or joyful smiles, as their
baby's recognition of them. If the parents learned this nonvisual and
alien vocabulary, according to Fraiberg, they came to know and respond
appropriately to their infant and a dialogic interaction began.
In summary, data from studies of attachment and/or parent-infant
gaze behavior suggest that there is an important link between the visual
interplay of mothers and infants and their subsequent attachment to one
another. There is evidence that suggests that mothers' sensitivity and
responsiveness to infant elicitations play important roles in the
formative months. The findings also suggest that duration of gaze may
be a more sensitive predictor of the mutual gaze and attachment relationship
than frequency of gaze alone.
The contingent gaze and infant
A varied environment, freedom for exploration, an interested and
responsive caregiver, and contingent stimulation all appear to be con-
ducive to the development of infant competence (Appleton, et al., 1975).
In the first few months of life, the visual experiences of the infant
play an immensely important role in the infant's exploration of the
environment and contingent elicitation of and responsiveness to care-
givers. Stern (1976) claimed that visual interactions provide the
primary basis for competence development during the early period, for
the infant's comprehension of the environment at this time strongly
rests on non-verbal cues (see also Appleton, et al., 1975; Stern, 1974).
Goldberg (1977) has recently described a model of infant competency
and parent-infant interactions which emphasized the importance of the
role of the contingency experience in development. She contended that
infants are instrumental in establishing the social conditions which
provide for their competence development. Infants are equipped, she
stated, with a repertoire of behaviors, such as sucking, rooting, crying,
visual regard and fixation, smiling, and vocalizing which effectively
capture adult attention and facilitate adult-infant interactions. When
adults respond promptly and appropriately to the infant's elicitations,
both the infant and the adult experience what Goldberg called the mutual
reinforcement of "feelings of efficacy" (p. 163). The infant learns he
or she is effective (i.e., can control or influence the environment).
This expectation of effectiveness then enhances the infant's exploration
and practice of new skills, and thereby facilitates development.
Goldberg argued that, just as Watson found that "people are important
to the infant because they play the (contingency) game" (Watson, 1972, p. 340),
parents continue to play the game with infants because of the "contingency
experience provided by infant attention, smiles, and vocalization" (p. 165).
She cited the studies of Lewis and Goldberg (1969) and Ainsworth and Bell
(1974) which demonstrated that infants with responsive, attentive mothers
were more advanced developmentally than infants whose mothers were
unresponsive. She also cited the study by Ramey, Starr, Pallas, Whitten,
and Reed (1975) in which the researchers concluded that the provision
of a contingency experience in a program for infants "who failed to
thrive" significantly influenced the infants' developmental progress.
Goldberg documented developmental outcomes of the contingent gaze behavior
of parents and infants in particular. Thesewere an increase in strong
affectional ties at 3 months (Robson, 1967) and increased social play
and learning (Ambrose, 1963).
Other studies of the first year of life have linked mutual gaze
behavior of parents and infants to later infant competence. Beckwith,
Cohen, Kopp, Parmalee, and Marcy (1976) studied 51 mothers and their
premature infants in caregiving and play situations in their home. The
analysis of their data revealed a significant relationship between
mother-infant mutual gaze behavior at 3 months and the infants' competency
scores on tne Gesell Developmental Schedule at 9 months. Clarke-Stewart
(1973) analyzed longitudinal data over the first 2 years of infancy to
determine "causal directions" of mother-infant interactions and the
infant's later performances. She stated that the single most plausible
causal hypothesis was that the "amount of maternal attention (looking)
influenced the child's later performance on comprehensive tests of
intellectual competence and motivation" (p. 86). This does not suggest
that it was visual attention alone, she stated, but merely that this
behavior plays a prominently causal role in early competence development.
In another study, Gordon (1974) reported that, in a sample of 121 low
income mothers and their infants,mutual gaze behavior in the early months
was significantly related to the boy infants' 12 month competency scores
on the Bayley Scales of Infant Development. However, this relationship
was not significant for the girl infants of the study, suggesting the
possibility of an infant sex effect in the relationship of contingent
gaze and competence development.
The evidence presented in this section of the review reveals that
the infant's early visual experiences play a remarkably important role
in helping the infant to explore and interact with the world, seek new
experiences, and develop skill in interacting with parents. Mothers
of infants are, in turn, mutually reinforced as a result of the infants'
active participation and responsiveness. Noticeably missing from the
literature are data which describe the father's role in these formative
interactions. A study of the factors which may influence the gaze
process could help to explain the specific early contributions of
fathers, mothers, and infants.
Factors that Affect Early Parent-Infant
It is probable that mutual gaze, like most other human interactions,
may be affected by many different factors that may have a variable effect
over time. The infant research literature analyzed for this review pro-
vided some evidence that infant maturational effects, parent and infant
sex, and infant temperament characteristics play influential roles in
the early gaze behavior of parents and infants. There are only limited
data, however, concerning the effect of these factors in the give-and-
take of actual gaze transactions.
Maturational effects in
Findings from the following studies have documented the infant's
active involvement in the gaze process at 6 to 15 weeks (Blehar, et al.,
1977), 5 to 20 weeks (Brazelton, et al., 1974), 3 to 12 weeks (Moss &
Robson, 1968), and 3 months (Beckwith, et al., 1976; Robson, 1967;
Stern, 1974). Escalona (1973) conceptualized mutual gaze as the most
primitive of the sustained reciprocal interactions between parent and child.
These reciprocal interactions, according to her, are characterized by both
partners playing a transmissive, as well as a recipient role. They cannot
occur unless each partner does his or her share. She found in her
intensive study of 2 children for the first two years of life, that
sustained reciprocal events do occur, even in the early weeks of life.
Analysis of her data supports Sander's (1976) conclusion that there is
a "heightened period of reciprocal exchange" between parents and infants
during the 4 to 6 month period.
Sex as a variable in parent-infant
Recent reviews reflect the general lack of consensus in parent-
infant sex difference research (Birns, 1976; Block, 1976; Korner, 1974b;
Maccoby & Jacklin, 1974; Yarrow, Rubenstein, & Pedersen, 1975). Perhaps
Korner's (1974b) recent statement accurately describes the current status
of this issue.
Although we know in general terms that genetic, hormonal,
and experiential factors all contribute to the development of
behavioral sex differences and the formation of gender identity,
we know little about the relative contribution of each to any
given sex linked trait. We also have very little knowledge of
when behavioral sex differences begin to emerge and under what
influences. (p. 197)
One explanation for the lack of infant sex data, according to Korner,
is the failure of many infant researchers to analyze their data for sex
Research concerning changes of mother-father behaviors related to
the attachment process in the first year of life may shed some light on
the dynamics of sex differences in gaze behavior. Lewis and Weinraub
(1974) divided attachment behaviors into two primary categories--
proximal (touching, holding) and distal (looking and vocalizing). They
found proximal behaviors were more prevalent in early attachment
(especially the first months); then, over the course of the first year,
distal behaviors became more prevalent. The changes occur, they
reasoned, due to the emergence of competing infant motives, such as
exploration and increasing competence in communication and mobility.
Specific findings from their study were that boys received more
proximal stimulation from mothers than girls did in the first few months.
However, by 6 months,mothers decreased their physical interaction with
boys, so that by 1 year, girls received more proximal attention from
mothers than boys did. Father-son and father-daughter relationships did
not differ remarkably in this study. The father-infant relationships
were similar to the mother-son relationships; that is, proximal stimu-
lation at first, then distal prominence in the second half of the first
year. These sex variations in parent-child attachment, according to
Lewis and Weinraub, are probably related to the cultural expectations of
early autonomy in sons, acceptability of dependency in daughters, and
decreased proximal behaviors in father-infant relationships.
Some of the research evidence that has been reported concerning
infant sex differences in gaze behavior in particular seems to support
the Lewis and Weinraub explanation. At birth, girls respond faster to
visual stimuli (Korner, 1974a). Boys at 3 weeks and at 3 months have
significantly longer fixation time (Moss & Robson, 1969). At around 3
nonlis of age boys are "crankier," according to Moss (1974), and, as a
result, mothers use more visual and physical contact in soothing and
stimulating them than they do girls. Two recent studies which included
fathers as well as mothers found that, for the most part, infants look
at their fathers more than they do their mothers. Lamb (1977) studied
6 month old infants and found that both boys and girls look at their
fathers significantly more than they do their mothers. In the Moss
(1974) report, 7 month old boys looked at their fathers significantly
more than their mothers. The 7 month old girls in this study
looked at their mothers more often than they did their fathers but not
Very few studies are available which compare father-infant and mother-
infant looking behavior. In one study of the first 48 hours of life,
Parke (1974) found both fathers and mothers looked at their male infants
more than their female infants. And in studies which included only
mothers and their infants in the first year of life, Yarrow et al. (1975)
and Gordon (1974) found variable results. Yarrow et al. (1975) studied
41 black mothers or primary caregivers and their infants and reported that
the mothers (caregivers) interacted more frequently with their infant
boys at higher intensities with richer, more diverse stimuli than they
did their infant girls. In Gordon's (1974) study, the frequency of the
mutual gaze interactions of low income mother-son and mother-daughter
pairs did not differ significantly. Curiously enough, the early
mutual gaze behavior of mother-son dyads correlated positively and
significantly with male infant competency scores at one year, where, on the
other hand, there was no significant relationship between gaze and com-
petence for the mother-daughter pairs. Gordon suggested that further
study of the variables is indicated and that fathers should be included
in future studies. He hypothesized that the significant relationship of
mutual gaze and infant competency of mother-son dyads could possibly be
an important one for father-daughter dyads.
Moss (1974) postulated a complex relationship between infant sex
differences, temperamental characteristics, and infant gaze variability.
He combined data from a number of mother-infant mutual gaze studies
(Moss & Robson, 1968, 1970; Moss, Robson & Pedersen, 1969; Robson,
Pedersen & Moss, 1969) and found that social experiences for the infant
girls and physiological state conditions for infant boys seem to affect
their respective visual behavior.
In summary, sex difference research findings related to early mutual
gaze behavior are varied and inconclusive. Tests of visual fixation
reveal boys fixate on stimuli for longer periods than girls do. Gaze
directional studies indicate that father-infant dyads are more active
than mother-infant dyads. For the most part, studies that have analyzed
reciprocal gaze behavior have failed to include fathers. Findings pre-
sented from mother-infant mutual gaze studies have included the following--
black caregivers are more actively involved with male infants, and in a
study of a large sample of low income families, there were no sex dyadic
differences in gaze encounters, but the gaze behavior had different
developmental outcomes. For the boys, there was a significant relation-
ship between gaze and infant competence; however, this relationship was
not significant for the girl infants. The survey of these studies sug-
gests that the sex effect on parent-infant reciprocal interactions may
be a complex one and may involve an interaction between infant sex and
Infant temperament as a variable
in parent-infant interactions
It is certainly conceivable that the infant temperament or behavioral
style may affect the tempo, rhythm, valence, and intensity of infant gaze
patterns. There is empirical and theoretical evidence that two
temperament characteristics in particular--activity level and mood
quality--play important roles in infant gaze behavior. Thomas, Chess,
and Birch (1968) have defined these qualities as
Activity level: The motor component present in a given child's
functioning and the daily proportion of active and inactive
Quality of mood: The amount of pleasant, joyful, or friendly
behavior as contrasted with unpleasant, unfriendly behavior,
or crying present in a given child's functioning.
Even though the Thomas et al. studies (1963, 1968) and many later studies
(Sameroff, 1975) have concluded that temperament characteristics of the
infant have an important transactional effect on parents, very little is
understood about how these traits specifically affect parent-infant
Infant activity level
Infant activity level, according to Murphy and Moriarity (1976), is
far more than locomotion. According to them, the proportional energy
level of the infant varies as a reflection of the infant's internally
generated impulses and also as a result of the infant's tendency to react
to external stimuli by using motor activity. Moss and Robson (1970)
found that the quiet alert state of infants was correlated with an increase
in visual scanning and fixation time. These findings lend further support
to Escalona's (1968) earlier conclusion that inactive babies use the visual
modality more than active babies do. Data from Stern's (1974) study
indicated that gaze-to-gaze intervals--or the frequency of a given in-
fant's gaze behavior--may be a physiologically controlled function of
the infant. The finding seems to support indirectly a relationship between
activity level of the infant and infant gaze behavior. Stern stated
that gaze duration on the other hand, is dependent on infant interest,
and therefore, it is not dependent on biological givens (such as activity
level) of the child.
There are some data that suggest a curvilinear relationship between
the activity level of the infant and gaze behavior. Both hyperactive
and anaclitically depressed children fail to participate in reciprocal
gazing (Pretchl, 1963; Spitz, 1965). Withdrawn, apathetic children in
institutions often stare at strangers for a prolonged period. However,
this looking behavior has been called "vacuous," empty," "dull." It
lacks the reciprocity described in the contingent gaze process (Provence
& Lipton, 1962; Spitz, 1965).
In summary, data from normal infant studies indicate that the activity
level of infants may determine infant availability for gaze or the fre-
quency of gaze-to-gaze intervals of the infant. The evidence suggests
that the infant who is considered inactive (but within a normal range of
activity) or in a quiet state participates in increased visual fixation
or prolonged gaze behavior. It is uncertain, however, how this effect
influences gaze exchange behavior of parents and infants.
Infant mood temperament
The available research data concerning infant mood temperament and
parent-infant gaze behavior are both elusive and difficult to interpret.
Haviland (1976) cautioned that many investigators in infant research fail
to differentiate between the two qualities, visual responsivity and
affective responsivity. In a recent publication, she used Piaget's
observations of his 2 month old daughter to illustrate the difficulty
involved in separating these areas.
For example, Lucienne sees Piaget at the extreme left of her
visual field and smiles vaguely; she then looks away but con-
stantly returns to the place in which she sees him and dwells on
it. If the infant had merely turned her head with a vacant gaze
and no facial mobility, Piaget probably would have found it
difficult to say she recognized him or was "bringing to himself
/sic7 the image of her desires" (Piaget, 1972, p. 2). There is
something in the smooth turn of the head accompanied by a
searching gaze, the stopping of the head, the fastening of the
gaze on the "desired" object, and the looking with interest and
even pleasure that arouses in the heart of the trained genetic
epistemologist a belief that the child has desires and recognizes
a particular object . . (pp. 360-361)
Haviland contended that it is possible to isolate and study infant affect
and that this variable has a particularly meaningful place in the inter-
pretation of infant behaviors. In addition, DeCarie (1965), Escalona
(1973), Gordon (1974, 1975), Spitz (1965;, Stechler and Carpenter (1967),
and Yarrow et al. (1975) have strongly advocated the need for the study
of infant affect.
Martin (1975), in writing about parent-child relationships in general,
stated that once an infant begins to look, smile, coo, or show other signs
of delight, the mother's responsive behavior is further reinforced and a
strong dyadic system maintained by mutual reinforcement is likely to
develop. Recently, Clarke-Stewart (1973) reported that, over the period
of the first year of life, the child's expression of positive emotion to the
mother was related to the mother's expression of more positive emotion
at a later time. Although Brody and Axelrand (1970) did not specifically
study infant affectivity and gaze variability, their explanation of the
effect of "unpleasure" on infant behavior in general offers some insight
in this area. They postulated that infant states that include irritability
or discomfort interfere with the infant's awareness of external stimuli.
In these states, Brody and Axelrand explained, infants become "preoccupied"
with internal (physiological) sensations, and thereby are less likely to
attend to stimuli from the environment. Stechler and Carpenter (1967)
reported that the baby's expression of positive emotion encouraged or
elicited the approach of caregivers and increased the probability of
continuation of a concurrent interactional pattern. Conversely, they
stated that the negative affect of infants inhibited the concurrent
behavior of the caregiver and caused the caregiver to change his or her
Spitz (1965) theorized that both pleasure and displeasure are
important in infant developmental research. He posited that infant
affect markedly influences infant perception; it makes the perceived
experience seem important or unimportant; and it endows the infant's
perception of the event with a valency that enhances or discourages
the repetition of the experience (p. 85). He stated that the affective
investment of the infant insures the storing of the experience in the
infant's memory. The study of infant affect is particularly important,
according to Spitz, for it is during infancy that "affective processes
are statu nascendi, observable without contamination" (p. 140).
In summary, the research evidence presented to describe the rela-
tionship between infant mood (affect) and parent-infant gaze behavior
consists primarily of indirect data. There are data which support the
contention that infant affect influences the infant's availability for
participation in gaze behavior, the approach of mothers in parent-infant
relationships, and the mother's positive response to the infant on
Methodological Considerations in the Studies Reviewed
Studies selected for this review were topically representative (not
inclusive) of the vast amount of information available in the area of
patient-infant gaze behavior. Investigations with obviously poor
research designs, inappropriate statistical techniques, invalid and
unreliable conclusions were excluded. It was assumed that most of the
sampling procedures for these studies were "convenient" since explanations
of sampling techniques were often not given. Few studies described any
loss of subjects or control measures for the highly variable infant
state (e.g., drowsy, fussy, or crying). All of the studies were concerned
with American parents and infants and represented a range of socio-
economic groups. Contextual settings for the studies ranged from single
observations of highly controlled, standardized laboratory settings to
weekly observations of "familiar friends" in naturalistic settings over
an extended time. Criteria and techniques for measurement of gaze
variables, sex effects, and infant temperament varied remarkably from
study to study. Statistical methodology ranged from simple descriptive
percentages to complex multivariate methods.
Many of the studies selected for review dealt with the elements of
gaze behavior, or its static qualities such as "baby looks in direction
of mother." As a result, the findings from these studies have limited
value in predicting the effect or relationship of various factors on
the mutual exchange involved in gaze interactions. Findings from the
studies suggested father-infant combinations look at each other more than
mother-infants (Lamb, 1977; Moss, 1974; Parke, 1974). Infants who are
inactive or who may be in a quiet alert state are more likely to use the
It is important to note that, in the studies that did deal with
interactive effects, parent-infant mutual gaze behavior proved to be
an effective predictor of more positive response in mothers (Clarke-
Stewart, 1973) and in babies (Blehar, Lieberman, & Ainsworth, 1977),
secure attachment (Blehar, et al., 1977), increased sensorimotor skills
(Appleton, Clifton, & Goldberg, 1975), and advanced cognitive competence
(Beckwith, Kopp, Cohen, Parmalee, & Marcy, 1977; Clarke-Stewart, 1973;
Gordon, 1974). Fathers were not included in the mutual gaze studies.
Although findings from a preponderance of the studies were based on
behavior from naturalistic home settings, studies such as Brazelton,
Koslowski, and Main (1974) and Gordon (1974) were not. The significant
contributions of these two studies suggest that, with careful planning,
laboratory settings can provide an opportunity for indepth investigations
of phenomena that reflect natural behavior. Gewirtz and Boyd (1976) admit
that it is important to map mutual behavioral regularities as they occur
in the natural environment; however, they say it is possible to experi-
mentally simulate and approximate natural settings by simplifying routinely
occurring events in laboratory settings. After careful, extensive reviews
of mother-infant observational studies, Cox (1975) and Lytton (1971)
also advocated the use of experimentally contrived interactions for
indepth study of highly complex interactive data and also for increasing
the probability of comparability among subjects. However, both of these
theorists suggested replication of such studies in naturalistic settings
before findings of the studies were widely disseminated.
Historically there has been an overdependence on intercorrelation
techniques in parent--nfant research (Bell, 1971). It was not unusual,
according to Bell, for investigators in the past to be biased in
interpreting these results. It was not uncommon for significant findings
of relationships between parent and infant behaviors to be interpreted
as "parent caused," rather than to recognize the bi-directional
relationship inferred by the correlation. Bell considered the recent
recognition of the infant's contribution to early relationships particu-
larly important, for it has provided an impetus for the use of more
sensitive research designs and their appropriate statistical methodology
in the study of parent infant relationships.
Statistical methodology utilized by the studies reviewed demonstrated
a wide-range in choice of techniques. Klaus, Kennell, Plumb, and Zuehle (1970)
expressed the data related to mother eye-to-eye contact in simple per-
centages. Stern's (1974) study was enhanced by his use of Markovian
chain to describe sequential effects of gaze, and Escalona (1973) developed a
technique which she called prominence of behaviors at developmental
periods which emphasized the likenesses rather than variability of be-
havior between the two subjects of her study. Moss' (1974) retrospective
look at his extensive data on mother-infant mutual gaze demonstrated that
there may be complicated interrelations between gaze, sex, and tempera-
ment variables. McCall (1970; 1977) has recently advocated the use of
multivariate techniques in studying developmental issues. (See also
Kagan, 1976; Yarrow, Rubenstein, & Pedersen, 1975.) According to him
the more global display offered by these techniques prevents excising
variables out of their natural context. In these procedures, the
interrelated, often complementary, multiple effects of interactive behavior
can be studied simultaneously and thereby provide for better explanations
of natural phenomena. Several of the studies included in this summary
utilized multivariate statistical techniques: Beckwith et al., 1976;
Gordon, 1974; Lamb, 1977; Yarrow et al., 1975. Yarrow et al. (1975)
cautioned future researchers to carefully interpret findings from studies
whose variable number is greater than subjects studied. He suggests the
results of these studies may be misleading and have limited generalizability.
Methodological issues considered as a result of this review of
literature suggest further study of the sex and infant temperament effects
on parent-infant mutual gaze behavior is indicated. There are no
available studies that have examined the relationship of father-infant
and infant temperament to the parent-infant reciprocal gaze process.
There were data available that suggested contrived experimental situations
provide valid data for indepth study of complicated processes such as
mutual gaze interactions. Multivariate research designs were advocated
for the study of social phenomena which have multidimensional inter-
Critical analysis of the research studies in this review supports
the view that early parent-infant reciprocal gaze behavior plays a
critical role in the development of infant sociability and cognitive
competence and parent-infant attachment. It is uncertain, however, what
factors affect the give-and-take of these early interactions. Data from
studies that, for the most part, have measured the direction of parent
and infant gaze behavior (but not its mutual effects) have provided
evidence that parent and infant sex and the infant's activity and mood
temperament affect this process. Fathers have not been included in
studies that have examined reciprocal effects of the gaze process. Some
of the data presented suggested that the relationship of sex and tempera-
ment factors and parent-infant gaze behavior may be a complex one and
that the study of these factors requires the use of multivariate research
designs. In addition, the duration of gaze interactions may be a
particularly sensitive measure for the study of parent-infant relation-
In summary, the major conclusions from this review of parent-infant
interactive research literature were:
1. Reciprocal gaze behavior of parents and infants plays a
critical role in their early formative relationships. It has been
shown to have a significant relationship to infant competence and
socialization and parent-infant attachment.
2. There is some evidence that inactive babies use the visual
mode of behavior more than active babies, but it is uncertain what
this means in terms of the transactional gaze process.
3. Positive affect of the infant functions to make the infant more
available for socialization, and it influences mutual reinforcement of
parent-infant interactions. Its specific influence on gaze interactions
is not known.
4. Very little is known about early father-infant reciprocal inter-
actions. There is some evidence that suggests that father and infants
may look at each other more than mother and infants do.
5. There are data which suggest that there may be a complex
interrelationship between parent sex, infant sex, and infant temperament.
6. Multivariate studies of the sex and temperament factors and
their relationship to the frequency and duration of reciprocal gaze
behavior are warranted.
This study explored the relationships between mutual gaze behavior
of parent-infant interactions in a structured play situation and the
following factors: the sex of the infant, the sex of the parent, and
the parents' perception of their infant's activity level and prevailing
The hypotheses tested were:
1. There are differences among father-son, mother-son, father-
daughter, and mother-daughter dyads in the frequencies of observed
brief and prolonged mutual gaze encounters in a structured play situation.
2. There is a relationship between parental perceptions of the
infant's activity and mood and mutual gaze behavior within structured
a. The parents' perception of infant activity has a positive
relationship with brief gaze and an inverse relationship with prolonged
b. The parents' perception of infant mood has a positive re-
lationship with brief and prolonged gaze.
Data collected as a part of a more comprehensive project, "The
Parent-Infant Transactions and Infant Competency Project" (i.e., the
PITIC Project), were used in this study. The PITIC Project was sponsored
by the National Institute of Mental Health and was based at the Institute
for the Development of Human Resources, University of Florida, Gainesville,
Florida. The period of study funded for the project was March, 1976 to
August, 1978. This study ("The Relationship of Sex and Temperament to
the Mutual Gaze Behavior of Early Parent-Infant Interactions") used the
following data from the PITIC Project: videotapes of the parent-infant
structured play interactions and parental responses to the Perception
of Baby Temperament Scales.
Families were recruited through public media, local maternity
clinics, obstetricians, pediatricians, and prepared childbirth classes.
In order to participate, families had to expect to be in the Gainesville
area until the baby was one year old; the parents had to be married
and had to agree to attend all taping sessions. (For the PITIC Project,
there were 7 taping sessions, 1 every 6 weeks while the baby was 13 to
49 weeks of age.) Recruting efforts were directed at recruiting white,
middle class families with firstborn infants. In all 42 families were
chosen as volunteers for the project. One family withdrew from the
project, leaving a final sample number of 41.
The Hollingshead Two Factor Index of Social Position (Hollingshead & Redlich,
1958) was used to assess the social status of the sample. The advantage
in using this scale is that it considers both the occupational and edu-
cational background of respondents in the calculation of social position.
Using factor analyses techniques, Hollingshead and Redlich (1958) demon-
strated a relationship between index class position on this scale and
actual social behavior of the respondents in their study. At that time
their index position was based on data from fathers only. For this pro-
ject, both father and mother scores were averaged to produce the family
social index. A summary of the results are listed below.
Social Status Classification
Number of Number of Number of
fathers mothers families
in class in class in class
index index index
Class I (major professionals) 11 5 11
Class II (lesser professionals ) 12 18 11
Class III (white collar workers) 16 13 19
Class IV (skilled manual laborers) 2 5 1
Class V (unskilled manual laborers) 0 0 0
The majority of families in this sample were rated as major and lesser
professionals and white collar workers. In this study use of family
rating instead of the father rating produced little change in the final
description of the sample.
All of the parents were Caucasian, native-born Americans. The mean
age for the fathers was 28.6, standard deviation, 4.04. The mean age
for the mothers was 26.4, standard deviation, 2.94.
The babies were firstborn. There were 26 boys and 15 girls. Each
baby was determined to be healthy by a pediatrician's examination pro-
vided with project funds. One child was diagnosed as having congenital
hip dysplasia at a later date but was not excluded from the sample.
There were no known complications during the mothers' pregnancies
as reported by the parents. Childbirth classes were attended by 37 of
the 41 fathers and mothers. Twenty of the fathers were present in the
delivery room during the birth of their baby. One family chose to have
their baby delivered at home. None of the families reported attendance
at parenting or child development classes since the birth of their baby.
The infant feeding methods reported by the parents were 30 of the babies
were being breast fed, 4 were being bottle fed, and 7 of the families
were using a combination of both bottle and breast feedings. There were
serious infections in two of the babies during the newborn period. No
significant periods of separation of any of the family members were re-
ported except for one mother who was hospitalized several times during
the first month following the birth of her baby.
The initial contact with the parents was usually a phone conversa-
tion. A series of questions was asked to screen the families (see
Appendix A). If the family was recruited for the project, a home visit
was made. At this time, the project staff member outlined the project's
purpose and answered the parents' questions (see Appendix B). During
this visit, the mother and father signed informed consent forms (see
Appendix C). Arrangements were made with the family for a visit to a
pediatrician of their own choosing for the baby's examination, and the
first observation session was scheduled.
All of the observations were taped in the same studio under uniform
conditions. Upon their arrival parents were welcomed and asked to read
the directions for the "game" they were to play with their infants during
that session. The area used for videotaping was enclosed on all sides with
four partitions. The television camera lenses were placed in specially cut
portholes in the partitions. During the filming no one but the parent
participating in the play activity was visible to the infant. No in-
structions were given to the parents regarding an optimum position for
taping for themselves or the infant. The parent who was not being taped
remained outside the partitioned room. A television monitor was nearby
on which this parent could watch the encounter.
One of the three cameras was placed in position to afford wide view
of the episode at all time. The other two cameras were used for close-up
shots and alternative views of the interaction. The three images were
viewed simultaneously on adjacent monitors to determine which images
should be recorded. The operator could record one, two, or all three
images (simultaneously) if they contributed to the view of the interac-
tions. Project staff members operated the cameras and the switcherr"
which determined the camera image or images that were recorded.
Each taping segment included a 3 minute segment with the mother and
infant and a 3 minute segment with the father and infant. (For the
PITIC Project a third videotaped segment which included both of the
parents and the infant together for 3 minutes was also taped. This
segment was not used in this study.) In 21 of the taped observations,
the mother interacted with the infant first and in 21 the father inter-
acted with the infant first; this order was randomly assigned.
In the early plans for this study, the 13th, 19th and 25th play
sessions were chosen as data for this study. However, choice of the "games"
used in these sessions resulted in making the data from only the 13th
session appropriate for this analysis. The standard game for the 13 week
observation was called "Dialogue." In this game, the parents attempted
through various means to get the infant to respond to their actions. When
the infant responded, the parent was encouraged to respond positively
and/or mimic the response. (See Appendix D for parent directions for the
game.) In the games chosen for the 19th and 25th week play sessions, the
parents' and infants' attention was often focused on toys. Thus, only
data from the 13th week play session was used in this study.
Temperament data for this study were based on parental responses to
the Perception of Baby Temperament Scales for Activity and Mood. Prior
use of this particular scale has been restricted to parents with babies
20 weeks old or older (Pedersen, Anderson, & Cain, 1976). The research
literature supports the notion that temperament qualities are stable
traits of the infant (Chess, 1973; Scarr-Salapatek, 1976), but that the
behaviors used by the infant to express these characteristics may change
with development. Hence, the researcher felt justified in analyzing gaze
data from the 13 week session in conjunction with the 19th week temperament
This study used two instruments: one for mutual gaze and one for
measuring parents' perception of infant temperament. The mutual gaze cate-
gories of the study were based on those of Escalona's (1973) Scale of Social
Interaction. The Perception of Baby Temperament Scale (Pedersen, Anderson,
& Cain, 1976) was used to estimate infant activity and mood.
Mutual gaze data
Criteria for mutual gaze were based on the gaze categories described
by Escalona (1973) for the Scale of Social Interaction. She called mutual
gaze behavior the most primitive of the sustained reciprocal behavior,
which according to her definition were "prolonged social exchanges in
which both partners play a transmissive, as well as a recipient role . .
and which cannot take place unless each partner does his or her share"
(p. 210). The PITIC Project staff revised the Escalona Scale to make it
appropriate for use with videotaped data. For this study, only two
categories of mutual gaze were coded, brief and prolonged.
1. Mutual gazing, brief. Mutual gaze was defined as sustained
eye contact between the child and parent. It excludes a casual meeting
of glances. In infants younger than 5 months, brief mutual gaze was
defined as lasting at least one second, and mutual gazing must have been
the primary activity on the child's part. Some body movement by the
infant was acceptable if the infant's gaze was riveted to the parent's
2. Mutual gazing, prolonged. Mutual gaze was defined as sustained
eye contact between the infant and parent. Prolonged mutual gaze
differed from brief mutual gaze in the duration of the episode in that
it lasted for more than three seconds.
Two raters were trained by the PITIC Project staff. Both were University
of Florida undergraduate students, one majoring in medical technology
and the other in psychology. The raters judged the duration of the mutual
gaze episode by verbalizing "one-one thousand," "two-one thousand," etc.
Both raters were asked to code all of the play sessions (82 in number--
41 father-infant, 41 mother-infant), and a mean was calculated using
both raters' observations for each play session to produce father-infant
and mother-infant mutual gaze observations.
Reliability of mutual gaze data
An intraclass correlation using the Bartko formula (1966; 1976) was
used to test interrater agreement reliability for coding mutual gaze
behavior. Both raters coded all 82 play sessions (41 father-infant and
41 mother-infant). For brief gaze, the intraclass coefficient was .65,
and for prolonged gaze, it was .64.
In a previous project, Gordon (1974) reported reliability measures
which were based on 14 of the Escalona items. He reported 0.84 percentage
of interrater agreement. Escalona did not report reliability or validity
estimates at the time the scale was published nor in subsequent research
employing the scale.
Infant temperament data
The second instrument used in the study was the Perception of Baby
Temperament Scales (PBT), developed recently by Pedersen, Anderson, and
Cain (1976) to measure the parents' perception of their infants' tempera-
ment. It was one of the few scales developed that could be used with
equal facility by fathers as well as mothers. Nine temperament qualities
were first described by Thomas, Chess, Birch, Hertzig, and Korn in 1963.
Only two of these qualities--activity and mood--were selected for this
The PBT has six items for each of the temperament characteristics.
Two versions of the items for each dimension were used, one for male
babies using masculine pronouns and a second for female babies using
feminine pronouns. The inclusive list of items that measured the patients'
perception of the activity level of a female infant were--
1. During a bath, she kicks, splashes, and wiggles. She is
full of activity at these periods.
2. Her play with toys is active; she often kicks her legs and
waves her arms.
3. During diapering and dressing, she squirms and kicks much
of the time. She is so active that I sometimes have trouble
doing these tasks.
4. She usually lies still during diapering and dressing. She
rarely squirms and kicks during these activities.
5. She usually lies fairly still during sleep. She awakens
in just about the same position as when she fell asleep.
6. When I feed her solid food, she tends to sit quiety, she
rarely squirms or kicks. (Pedersen, Anderson, & Cain, 1976, p. 1)
The items that measured the parents' perception of the mood quality of a
female infant were--
1. When I bathe her, she usually smiles or laughs. She seems to
enjoy bathing times.
2. Sometimes people come over whom the baby has been around
fairly often. She generally is friendly and laughs or smiles
3. When she wakes up for a nap, she almost always smiles and
4. When she wakes up from a nap, she often is a bit fussy.
5. When I feed her and I need to interrupt the feeding for such
things as burping, she seems to fuss for a bit when these inter-
5. She almost always has a fussy period each day. (Pedersen,
Anderson, & Cain, 1976, p. 5)
Each of the items was written on a separate card. To reduce social
desirability bias, the behaviors described were likely to be considered
"normal" by the parents. Half of the items in each scale were stated in
a positive direction (i.e., the respondent must agree to score highly) and
half in a negative direction (i.e., the respondent must disagree to score
One item of the activity scale, item number 6, was deleted for this
study. Very few of the babies were being fed solid food at 19 weeks of
age. So this item was considered inappropriate for use in this investi-
The PBT was administered in a format similar to a Q-sort. Each parent
was tested separately. They were asked to sort the items into the response
categories which represented the degree to which the statements described
their infants. The response categories were--
Very much like my baby
Sometimes or occasionally like my baby
Not like my baby at all
Have no experience with this
Numerical values of scoring these categories ranged from 0 to 3. For
items that were stated in a positive direction, the following were corres-
pondent values: "very much like my baby"--3 points; "somewhat like my
baby"--2 points; and "not like my baby at all"--I point. For the items
that were stated in a negative direction, the following were correspondent
values: "very much like my baby"--l point; "somewhat like my baby"--2
points; "not like my baby at all"--3 points. (See Appendix E for addi-
tional information concerning scoring of this instrument.)
Data from mothers and fathers who used the category "have no experience
with this" were deleted from this part of the study. The parent's zero
rating would have been misleading, for it would have implied a zero level
of the trait, rather than the lack of the parent's experience. A mean was
calculated for each parent for activity (5 items) and mood temperament
(6 items). When the ratings of fathers and mothers who used the category
"have no experience with this" were deleted, the remaining sample number
for temperament data was 34 fathers and 38 mothers.
Reliability of the temperament data
Several measures were used to assess the reliability of the activity
and mood temperament data. They were internal consistency of the items,
correlation of each parent's responses for the 19th session and for the
25th session, and correlation of agreement between father and mother
responses. The Cronbach alpha coefficient (Cronbach, 1970) was used to
estimate how well the scores obtained at the 19th week session represented
the universe scores for the temperament quality. The fathers' alpha co-
efficient for activity was 0.39 and for mood was 0.49. The mothers' alpha
coefficient for activity was 0.25 and for mood was 0.36. When the fathers'
scores at 19 weeks were compared with fathers' scores at 25 weeks, the
resulting Pearson product moment correlation coefficient for activity was
0.41 for activity and 0.40 for mood. For the mothers' scores at 19th and
25th week sessions, the correlation coefficients were 0.53 for activity
and 0.55 for mood. Fathers' and mothers' 19th week scores were compared
using Pearson product moment correlations. For activity the coefficient
was 0.49, and for mood it was 0.34. These data suggest a moderate degree
of homogeneity among these psychological trait items, a lack of stability
in parental responses over a 6 week period, and a lack of agreement between
independent observers' responses.
Pedersen, et al. (1976) used split-half reliability estimates and
father-mother agreement to analyze the reliability of the PBT. Their use
of the split-half technique in measuring this scale is inappropriate.
Equivalent halves of each temperament trait scale would be improbable
due to the limited number of behavior items and their limited range of
variability. These authors also report correlation of agreement between
fathers' and mothers' scores; they were 0.43 for activity and 0.57 for
mood. There was no description of the distribution of the scores re-
ported in their study of the scale.
In the initial research design for this study, the plan was to use a
multivariate model which would have permitted the simultaneous consider-
ation of the gaze, sex, and temperament variables. A review of litera-
ture related to these variables indicated that they may be a multi-level
and interrelated effect among these variables. At the beginning of the
PITIC Project, the estimated sample number was 80 families; however, only
42 families were recruited. As a result and in order to assure adequate
statistical power, the hypotheses of this study were tested separately.
The hypotheses were--
Hypothesis I--there are no differences among parent-infant sex dyads
in a structured play situation in brief and prolonged mutual gaze behavior.
Hypothesis II--there is no relationship between parental perceptions
of infant activity and mood temperament and the mutual gaze behavior of
parents and infants in a structured situation.
Hypothesis I postulated no differences among the four parent-infant
sex dyads in brief and prolonged mutual gaze behavior. This required a
repeated measures design for the father-child and mother-child play
interactions, and it required a design which would allow simultaneous
consideration of both the gaze and sex variables. The research design
chosen was a split-plot, two-way factorial, and the statistical procedure
selected was a multivariate analysis of variance (MANOVA). The split-
plot design is an appropriate one for the study of variables which may
involve increased levels of variability within individuals and in which
there are repeated measures for each subject (Kirk, 1968). A two-way
factorial was indicated for there were two levels of parent and infant
sex and a possible interaction between parent and infant sex. This
hypothesis asked the question, are there differences among the group
mean vectors. When this research questions exists, McCall (1970)
stated that the MANOVA procedure was the appropriate statistical analysis.
The MANOVA technique was selected for several specific reasons.
First, with two dependent variables (brief and prolonged gaze), MANOVA
provided a single test of each contrast, rather than repeated univariate
tests. Also, MANOVA examines the differences while considering the
interrelationship among dependent variables (Kerlinger & Pedhazur, 1973).
The classical regression method for MANOVA (Woodward & Overall, 1975)
was used to adjust each effect for all other model effects. Olsen
(1976) has advocated the use of Pillai's Trace as the multivariate
statistic of choice when there is diffuse variability on many variables.
thus making it an appropriate choice for this analysis. Embedded
in the MANOVA are the univariate analysis of the dependent variable
(Hummel & Sligo, 1971). If the multivariate tests of this analysis
were significant, tests of the univariate or single variable measures
would be indicated. Non-significant multivariate tests would
indicate a lack of systematic co-variation of the main effects of this
model. No univariate tests would be indicated due to the inability to
control for error in the univariate tests. The Statistical Analysis
System's (Barr, Goodnight, Sall, & Helwig, 1976) general linear model
(MANOVA) computer program was used for the calculation of this data.
Computer facilities of the Northeast Regional Data Center on the University
of Florida campus were utilized for this study.
Hypothesis II investigated the relationship between mutual gaze and
the parents' perception of the infant's activity level and mood quality.
Since a bivariate correlation coefficient summarizes the relationship
between two variables by indicating the degree to which variation in one
variable is related to variation in another, correlation coefficient was
selected as the statistical technique to test Hypothesis II. The Pearson
product-moment correlation was selected as the appropriate statistic as it
measures the strength of the linear relationship between two interval level
variables. Scatter plots of the raw data for activity and brief and pro-
longed gaze were examined for possible curvilinear relationships.
Eight Pearson correlations were considered. The fathers' perception
of infant activity level, the mothers' perception of infant activity level,
the fathers' perception of prevailing infant mood, and the mothers' per-
ception of prevailing infant mood each were correlated with the number of
brief mutual gazes and prolonged mutual gazes. Significance tests derived
from the uses of the Student's t with N-2 degrees of freedom were calcu-
lated for each coefficient.
ANALYSIS AND INTERPRETATION OF DATA
This study explored the relationship of sex and temperament to
mutual gaze behavior in parent-infant interactions. The study sample
consisted of professional and white collar families with firstborn
infants. There were 26 boy and 15 girl infants. The data were collected
when the infant was 13 weeks of age. Structured play interactions between
father-infant and mother-infant pairs in a laboratory setting were video-
taped separately. Raters were asked to code brief and prolonged mutual
gaze behavior. Infant activity and mood temperament data were based on
each parent's responses to the Perception of Baby Temperament Scale.
There were two hypotheses to be tested. One hypothesis examined the
effect of parent and infant sex on brief and prolonged parent-infant
mutual gaze, and the second hypothesis examined the relationship of the
parents' perception of infant activity and mood temperament with brief
and prolonged mutual gaze.
The first hypothesis stated that there would be no differences
among the parent-infant sex dyads' (father-son, mother-son, father-daughter,
and mother-daughter) brief and prolonged mutual gaze behavior. The
research design for this hypothesis consisted of a split-plot, repeated
measures, two-way factorial model. Dependent measures were brief and
prolonged gaze and the independent variables were parent sex and infant
sex. Infant sex was nested under a variable named family number. The
statistical procedure selected to test this hypothesis was multivariate
analysis of variance (MANOVA).
For the second hypothesis a relationship was postulated between
parent-infant gaze behavior and infant temperament. The specific rela-
tionships that were predicted were
a. The parents' perception of infant activity has a positive rela-
tionship with brief gaze and an inverse relationship with prolonged gaze.
b. The parents' perception of infant mood has a positive relation-
ship with brief and prolonged gaze.
Each of these 8 relationships (father-activity, mother-activity, father-
mood, and mother-mood with brief and prolonged gaze) were tested statis-
tically with Pearson product moment correlations.
Analysis of the Data
Gross inspection of the data revealed that.in many of the individual
parent-infant pairs, there were no mutual gaze observations. Frequency
tables were constructed and revealed that in 52% of the parent-infant
pairs there were no brief gazes. In 21% of the families there were no
Analysis of the means and standard deviations also indicated a lack
of mutual gaze observations. There was evidence of a skewed distribution;
in 5 of the 8 parent-infant sex dyads (4 for brief gaze and 4 for pro-
longed gaze) the standard deviation exceeded the value of the mean.
As a result, the raw data were transformed by taking the square root of
each observation, and a scalar of one was added to each observation to
prevent the occurrence of singular matrices in the multivariate analysis
of variance due to the abundance of zero observations. In Table 1, the
transformed means and standard deviations for the sex effect on mutual
gaze are reported.
Parent-Infant Sex Dyadic Gaze Behavior: Mean Transformation*
Standard Deviations, and Within Group Variance
square root of rater mean gaze observation
Within Groups Variance Covariance Matrix
l gaze (BMG) 2.654 -C
mutual gaze (PMG) -0.125
OF = 38
The means for all the parent-infant sex dyads are remarkably similar.
Mean range was 1.24 to 1.52. The smallest mean (1.24) was for mother-son
brief gaze, and the largest mean (1.52) was for mother-son prolonged gaze.
In all parent-infant sex dyads, prolonged gaze means exceeded those of
The MANOVA for Hypothesis I revealed a within group variance covariance
matrix which indicated 2.654 sum of the cross-products for the sex variables,
-0.125 for the gaze variables, and the total sum of the cross-products for
all variables in the model was 3.094. The most noticeable finding of the
tests for the parent sex and infant sex effect on brief and prolonged gaze
and the interaction parent sex x infant sex were the remarkably small multi-
variate test values for these effects. For the parent sex effect on the
gaze variables, Pillai's Trace was .1009 (F (2,38) = 2.3, p 7.05). For
infant sex effect on the gaze variables, Pillai's Trace was 0.175 (F (2,38) =
0.16, p>.05). The parent sex and infant sex interaction effect results
was a Pillai's Trace of .008 (F (2,38) = 0.34, p,.05). The hypothesis of
early parent-infant interactions was not rejected at the .05 level of
significance. In other words, tests of this hypothesis revealed that, in a
structured play situation, the parent-infant dyads (father-son, mother-son,
father-daughter, and mother-daughter) of this sample did not differ signifi-
cantly in the frequency of their brief and prolonged mutual gaze behavior.
Further tests of the univariate hypotheses were not warranted, given the
non-significant multivariate findings.
Hypothesis II postulated a relationship between the parents' per-
ception of infant temperament and parent-infant gaze behavior. Pearson
product moment correlations were used to test the relationship of father-
activity, mother-activity, father-mood, and mother-mood with brief and
prolonged mutual gaze. Positive relationships were predicted for the
correlations between brief and prolonged gaze and activity and between
brief and prolonged gaze and mood temperament. An inverse relationship
was predicted for activity and prolonged gaze.
Analysis for the data for Hypothesis II revealed that the means
for the parents' temperament ratings were remarkably similar for all
fathers and mothers on both activity and temperament scales. Standard
deviations suggested little variability for the parents' ratings. In
Table 2 these data are summarized; the eight correlation coefficients
can also be found on Table 2. Scatter plots of the gaze and activity data
were constructed; no curvilinearity was indicated. The Pearson Product
Moment Correlations coefficients were small; none was significant at
the .05 level of significance. Hypothesis II was rejected. In this
study parent-infant mutual gaze behavior did not related significantly
to the infant temperament measures.
Interpretation of the Data
An extensive review of the literature for this study provided
substantial support for the selection of the attributes sex and infant
temperament as viable influences on gaze variation in early parent-infant
reciprocal relationships. However, results from this present study
suggest that these two factors have no statistically significant rela-
tionship to mutual gaze behavior of parents and infants. Careful inter-
pretation of the findings from the literature and from this study is
necessary to provide some explanation for these unexpected outcomes.
The Relationship of Gaze Temperament: Means, Standard
Deviations, Correlation Coefficients
Correlation Coefficients for
Temperament temperament and gaze
estimates with with
n mean sd brief gaze prolonged gaze
Father-activity 34 2.5 0.39 0.21 0.01
Mother-activity 38 2.4 0.50 -0.15 0.18
Father-mood 34 2.4 0.31 0.15 0.20
Mother-mood 38 2.5 0.44 0.14 -0.15
It was decided arbitrarily at the beginning of this study to limit
the sample to white middle class participants. As a result, recruitment
measures and careful screening produced a relatively homogeneous sample.
In addition, this sample of parents showed evidence of having more common
experiences related to early child care than those of the general popu-
lation. A convincing majority, 74 out of the 82 mothers and fathers,
attended prepared childbirth classes; half of the fathers in the sample
were present at the birth of their child, and 30 out of 41 babies were
breast fed. Unusual commitment of these parents to parenting is further
evidenced by their willingness to participate in the 12 month long PITIC
Project which involved 7 taping sessions. This suggests that control of
the sampling procedures for this project could have functioned to limit
variability in gaze behavior.
The age of the infant may have remarkably influenced the gaze be-
havior of parents and infants of this sample. Rapid change of states
(crying, alert, sleepy) is not unusual for infants at 13 weeks. There
was no control for infant state in this study. Taping of the episode
began at the ascribed time regardless of infant state. Brazelton, Koslowski,
and Main (1974) used only behavior that evidenced that the interaction
"was going well" in their study. Mothers in Stern's (1974) study felt
"programmed to maintain the positive arousal of their infant." Both of
these studies used several periods of observation for data collection.
The single short period for this study may have resulted in distorted
findings as a result of infant age and state.
Mutual gaze behavior proved to be a difficult variable to assess
by viewing "canned" videotaped recordings. The project's switcherr"
operator may not have always chosen the image that portrayed the best
perspective for judging the eye-to-eye contact of the parent and the
infant. Also the research conditions of this study allowed the parents
to move about and assume whatever positions they chose in playing with
their infant; this often limited camera perspectives. Because of these
conditions, the coded gaze observations may not reflect the true number
of gazes exchanged.
Previous studies have anticipated these difficulties. Brazelton
et al. (1974) used videotaped observations, but they placed two mirrors
in strategic positions (out of camera range) so that all of the inter-
active effects would be simultaneously recorded on one camera image.
Stern (1974) also used one camera, but he placed two observers, one
behind the mother and one behind the infant to record the mutual gaze
sequences. In retrospect, it seems that reciprocal gaze behavior studies
require specially engineered situations in order to record this complex
data accurately for detailed analysis.
The reactive effects of the laboratory setting with the videotape
equipment may have influenced parent-infant gaze behavior. This may have
been particularly true of the 13th week session which was the family's
first visit to the studio. There was general agreement among the PITIC
Project staff, however, that if anything, the game "Dialogue" (13th week
game) optimized the probability of increased parent-infant interaction.
Behavior on the tapes suggests that the game was more familiar to some
parents and infants than others and might have, as a result, provided
additional source of unexplained variance.
Results from this study suggest that the PBT scale may not be a
sensitive indicator of the temperament attributes. The remarkable
similarity in parent mean ratings could suggest that some scale items
may reflect more socially desirable behavior than others, thus inter-
fering with the assessment of the parents' accurate description of their
A prevalent problem in parent-infant interactive research studies
is how to provide for the study of the multidimensional variables of
parent-infant interactive data within the limitations imposed by the
small and limited sample populations often characteristic of infant
research. Because of the limited sample size (42 families), the
theoretical construct implied in the summary of the review of the
literature--that there was a complex interrelationship of sex, infant
temperament, and parent-infant mutual gaze--was not tested. The sample
number was adequate, however, for the MANOVA of no relationship due to
a sex effect on mutual gaze, and the statistical results of this test
should be considered conclusive of the sex effect. However, the multiple
correlation procedures used to statistically test the gaze and temperament
relationships should be viewed with caution. Due to the lack of ability
to estimate error, some of these findings could be due to chance.
In a laboratory setting, in which American professional and white
collar parents were asked to play a 3-minute standardized game with their
firstborn infants, there were no significant differences in the brief and
prolonged mutual gaze behavior of the parent-infant sex dyads--father-son,
mother-son, father-daughter, and mother-daughter. The homogeneity of the
volunteer sample, the age of the infant, and reactivity of the laboratory
setting were considered to be factors which may have influenced the
mutual gaze observations used to assess this hypothesis. The small
Pearson correlation coefficients resulting from multiple tests of the
relationship of the parents' perception of the infant's temperament and
parent-infant gaze variability suggest that there was little practical
significance in the relationship of temperament and gaze measures as it was
assessed in this study. The Perception of Baby Temperament Scale may
lack sensitivity as a measure of infant activity and mood levels; and may
have, thereby, prevented valid tests of the relationship of temperament
and gaze variables.
SUMMARY, CONCLUSIONS, AND IMPLICATIONS
Purpose of the Study
This study explored the relationship of sex and temperament to the
mutual gaze behavior of parent-infant dyads. It was hypothesized that
there are differences in the mutual gaze encounters among father-son,
mother-son, father-daughter, and mother-daughter pairs. It was further
hypothesized that there is a relationship between the parental percep-
tions of their infant's activity and mood level and the mutual gaze
behavior of parent-infant interactions.
Need for the Study
Early mutual gaze behavior between parents and infants plays a key
role in the initial formative parent-infant relationships. It has been
shown that the mutual gaze of mothers and infants is significantly
related to infant cognitive and social competence. Yet, surprisingly
enough there have been very few studies of early parent-infant reciprocal
gaze interactions, particularly those of the father and the infant. This
investigation was designed to produce initial data concerning the rela-
tionship of sex and temperament to mutual gaze behavior of parents and
infants. Data from this study could provide information for future
researchers involved in more complex studies and for educators and
counselors involved in programs with similar parent-infant groups.
This study was designed to study the relationship of four factors--
parent sex, infant sex, infant activity, infant mood--to the brief and
prolonged gaze of parents and infants. Structured 3-minute play inter-
actions between the father and the infant and the mother and the infant
were videotaped separately in a laboratory setting when the infant was
13 weeks of age.
There were 42 parents and their firstborn infants who volunteered
and were selected to be included in this study. There were 26 infant
boys and 15 infant girls. All of the participants were white, native
born Americans; the social status approximation for the families ranged
from higher professional to white collar workers.
Mean age of the fathers was 28.6 and for the mothers 26.4 There
were no known complications during pregnancy or childbirth, and all of
the babies were healthy.
Two raters coded the total 82 videotaped play interactions for mutual
gaze behavior using the following criteria. Both the parent and the in-
fant had to have their eyes fixated on each other. Gazes which lasted for
more than 1 second were called brief, and gazes which lasted for more than
3 seconds were called prolonged. Final gaze observations represented the
mean score for both raters.
Infant activity and mood temperament levels were based on their father
and mothers' separate responses to the Perception of Baby Temperament
Scales. At the 19th week session, parents were asked to respond to this
scale's Q-sort questionnaire. There were 5 items for activity and
6 items for mood. Final score for each category was based on each
parent's mean rating.
Analysis and interpretation of the data
Two hypotheses were formulated for this study:
Hypothesis I--there are differences among father-son, mother-son,
father-daughter, and mother-daughter dyads in the frequencies of observed
brief and prolonged mutual gaze encounters in a structured play situation.
Hypothesis II--there is a relationship between parental perceptions
of the infant's activity and mood and mutual gaze behavior within struc-
tured parent-infant interactions.
a. The parents' perception of infant activity has a positive
relationship with brief gaze and inverse relationship to prolonged gaze.
b. The parents' perception of infant mood has a positive relation-
ship with brief and prolonged gaze.
A split-plot, two-way factorial, repeated measures research design
was used to analyze Hypothesis I data. The multivariate analysis of
variance was the statistical procedure chosen to test this hypothesis.
Pillai's Trace, a robust multivariate statistic, tested the effects of
parent sex and their interaction in the brief and prolonged mutual gaze
behavior of parents and infants. The hypothesis of no significant differ-
ences due to the effect of parent and infant sex was not rejected at the
0.05 level. No univariate test effects were anlayzed due to the non-
significant multivariate findings.
Eight Pearson product moment correlations were used to test
Hypothesis II. Correlations were calculated for brief and prolonged
mutual gaze behavior and the following--father-activity, mother-activity,
father-mood, and mother-mood temperament mean scores. None of the
predicted relationships between temperament and gaze was statistically
significant at the .05 level.
A number of factors were discussed as possible reasons for the lack
of statistical evidence of a relationship between the sex and temperament
attributes and mutual gaze behavior as they were assessed in this study.
Similarity in the background of sample families, age of the infant, and
the reactivity of the laboratory setting were considered as possible
influences on the variability mutual gaze behavior of the participants.
Parental perceptions of activity and mood temperament as it was measured
in this study may not have related in a systematic fashion to the actual
infant activity and mood traits.
Present day theorists (Kagan, 1976; McCall, 1977; Scarr-Salapatek,
1976) have advocated an increased emphasis in the study of what has been
called the canalizations of development--developmental milestones that have
a pervasive influence over the entire human lifespan. Rheingold (1961)
called the visual function of an infant a powerful, influential evolu-
tionary behavior. Further support for this view can be found in Piaget's
(1952) extensive descriptions of the visual system's major role in
coordinating early organismic developmental phenomena and the infant's
response to the world. An extensive review of current research litera-
ture emphasized the critical role that interactive gaze plays in the
development of the parent-infant attachment process. It also emphasized
the need for further study of this important process.
The present study was designed to investigate the relationship of
sex and infant temperament to the early mutual gaze behavior of parents
and infants. Results of this study revealed that, within limitations
imposed by this investigation, there were no significant differences in
parent-infant mutual gaze behavior due to sex or temperament factors.
In light of these findings, a retrospective critical review of research
literature was done and the following implications for research were
1. There area lack of substantial data which describe the evolu-
tionary development of the influential parent-infant mutual gaze process.
McCall (1977) has recently pointed out the need for investigations that
emphasize studies of what children do, rather than what they can do.
There is probably no question in the minds of those who have studied parent-
infant gaze behavior that infants at very young ages can participate in
interactive gaze behavior, but ultimately, if the necessary and sufficient
causes of this area of development are to be understood, the examination
of what parents and infants actually do in terms of mutual gaze behavior
must be systematically studied. Studies in naturalistic settings which
examine the evolution of the mutual gaze process in children and their
parents from varied backgrounds in many contexts over time are needed.
2. Sample numbers in these studies should be ample to provide for
use of multivariate statistical procedures. Through the use of multi-
variate techniques in analyzing naturalistic data, the nature and magni-
tude of multidimensional factors of gaze variation can be described as
they appear in complex, real situations. Extensive descriptive studies
could provide for more rational selection of variables to be studied in
laboratory settings and, thus, provide for increased effectiveness in the
generalization of findings from laboratory studies. It is rare in infant
research studies to find an investigator that does not point out the need
for larger samples. Yet sample numbers continue to be too small to examine
the complex multivariate questions that are being asked. Rheingold (1967)
attributes this to the lack of public awareness. She suggested that
better dissemination of the results of studies to lay parent groups would
be helpful in stimulating their interest and concern and would serve to
create more potential research subjects.
3. The finding of this study of no difference due to sex in parent-
infant mutual gaze behavior is an interesting one. Generalization of
this finding must be restricted to homogeneous middle class samples in
laboratory settings. Further study of the relationship between sex and
gaze in naturalistic settings with large and varied sample populations
is indicated. Other studies from the literature have reported no infant
sex difference in mutual gaze of mothers and infants, but have found sex
differences in developmental outcomes of the transactional effects of
gaze across time. Future studies should assess the long range transactional
developmental effects of parent and infant sex and mutual gaze behavior.
4. There are some data that suggest that the study of the sequential
effects of the gaze process (the flow of the interaction and participants'
specific contributions) may help to better identify the sex differences
in reciprocal gaze behavior. Further study in this area is advocated.
Care in the design of the research procedure should be exercised so that
the critical behaviors are recorded without distortion. Research instru-
mentation that utilizes mechanical recording, such as videotape equipment,
should include additional observers, use of mirrors, or stabilization of
the positions of the participants in the interaction to accurately record
the details of the interactions.
5. Further study of infant temperament could provide a profitable
mode for defining the specific contributions of the infant to early
interactions. Studies which validate measures of infant temperament
would provide timely contributions. Subsequent studies could then des-
cribe the relationship between parental perceptions of the infant and
infant characteristics. With these data, the effect of parental per-
ceptions (expectations) on subsequent infant development and parent-
infant relationships could be examined.
In summary, visual interactions are important components of all
social exchanges. In large measure they may provide a means for communi-
cating with young infants and for evaluating the effects of a variety
of factors in infant development and parent-infant relationships.
Further extensive study of this important and pervasive canal of develop-
ment is strongly advocated.
Questionnaire for Parents
Spouse: __Phone #:
How did you hear about the project?
Radio Ad Poster
TV Ad Other
What do you know about the project? (Make sure the following are covered.)
Studying mother/father/child transactions (interested in way
parents and their children interact).
Must be married.
In Gainesville area for one year.
Mother/father/child must come to studio in Health Center
once every 6 weeks beginning at 13 weeks for one half hour.
Only parents of normal firstborn children are eligible.
SChild will receive free check-up at 3 months (we will pay
At one year detailed observation of baby's mental and
How long to remain in area? Race:
Occupation: Spouse Occupation:
Education: Spouse Education:
Any previous children:
Child: If born: Name:
Date of birth: Where:
If not: Expected date:
Will you sign statement:
Because of research design, not everyone who wants to will be able to
participate. We will call back in a few days to tell you if you are
eligible. If so, we will come to visit you when the baby is about 10
weeks old to gather some additional information.
Parent-Infant Transaction Project Home Interview
FATHER'S NAME MOTHER'S NAME
BABY'S NAME WEIGHT AT BIRTH LENGTH
BABY'S SEX PLACE OF BIRTH
1. Were there any complications during pregnancy?
2. Was this a planned pregnancy?
3. Were any medications used during delivery?
4. If so, what kind of medication was used?
5. How much?
6. What kind of delivery was it?
7. Was the father present during delivery?
8. How are they feeding the baby?
9. Has the baby had any illnesses since birth?
10. What was it?
11. Was the mother working or going to school prior to the
delivery or pregnancy?
12. Is the mother going back to work or school in the
13. If so, when is she going back to work or school?
14. If she is going back to work or school, what are the
plans for the care of the infant?
15. Does the father share in the chores related to the
16. Who will be the major caretaker for the baby?
17. Have either of you had any courses in child
18. Did either of you participate in a training
course for the baby's birth?
19. Are either of you at present participating in
an infant related group or activity?
20. Is there anyone else besides the parents and the infant
living in the house?
21. If so, who?
22. Who else beside the mother and the father care for the
23. How long will this caretaker(s) be helping out?
24. Does each parent babysit so the other parent can go out?
INITIAL INTERVIEW FORM: JULY, 1976
UNIVERSITY OF FLORIDA
Title of Project: Parent-Child Transactions and Infant Competence
Principal Investigator: Ira J. Gordon
For this study, "Parent-child transactions and infant competence,"
we, the parents of understand that we will be
asked to play with our infant at specific times. These play sessions
will be videotaped at the Learning Resource Center, J. Hillis Miller
Health Center, once every 6 weeks beginning at child's age 13 weeks
until 7 sessions have been completed at 49 weeks. We will be given
specific activities (such as hide and seek with a small object) with
him/her. Each session will include 3 minutes play with the father and
baby; 3 minutes play with the mother and baby; and 3 minutes play with
both parents and baby.
We understand the purpose of this study is to acquire information
on the behavior of parents with babies and the way such behavior might
be connected to baby performance on scales of infant development at age
The information obtained in this study will be used for scientific
analysis only. The videotapes will be maintained at the Institute for
Development of Human Resources, College of Education, University of
Florida. Further written permission will be requested if these materials
should be used for any purpose other than scientific analysis, such as
We understand that we can withdraw from the study at any time.
When our baby is one year old, his performance will be evaluated
using the Bayley Scales of Infant Development. (This is not an intelli-
gence test.) Individual scores of our infant will not be used in reports
of the study but will be available, with interpretation, to us upon request.
Subject's Signature Date
Signature of parent or guardian, if Date
subject is a minor (where applicable)
I, the undersigned, have defined and explained this study to the
The proposed research has been approved by the University of Florida
Committee for the Protection of Human Subjects. If there are any further
inquiries, they may be addressed to the Chairman of the UCPHS or to the
Director of the Division of Sponsored Research for the University, and
your inquiry will be given prompt attention.
DIRECTIONS FOR PARENTS FOR GAMES
(Game for 13 Week Session)
This is probably a familiar activity. The purpose of this activity
is to engage in a conversation with your baby. Try to get your baby
to make a sound. When your baby does make a sound like a coo or gurgle,
respond to him by a combination of stroking his stomach, moving your
head close enough and smiling so you're sure he sees you, and repeating
the sound s/he makes. S/he may then coo some more and you have a "con-
versation" in progress.
Please feel free to sit in one of the chairs or on the floor or you
may stand. If you have an infant seat, it's quite all right to use it.
The most important thing is for you and your baby to be as comfortable
and natural as possible.
Perception of Baby Temperament Scales and Coding Technique
Very much Somewhat Not at
Temperament like my like my all like Don't
Dimension Items baby baby my baby know
I. Activity 1. During a bath she kicks, splashes, and wiggles. 3 2 1 X
She is full of activity at these periods.
2. Her play with toys is active; she often kicks 3 2 1 X
her legs and waves her arms.
3. During diapering and dressing, she squirms and 3 2 1 X
kicks much of the time. She is so active that
I sometimes have trouble doing these tasks.
4. She usually lies still during diapering and 1 2 3 X
dressing. She rarely squirms and kicks during
5. She usually lies fairly still during sleep. 1 2 3 X
She awakens in just about the same position as
when she fell asleep.
6. When I feed her solid food, she tends to sit 1 2 3 X
quietly, she rarely squirms or kicks.
II. Mood 1. When I bathe her, she usually smiles or laughs. 3 2 1 X
She seems to enjoy bathing times.
2. Sometimes people come over whom the baby has 3 2 1 X
been around fairly often. She generally is
friendly and laughs or smiles at them.
3. When she wakes up from a nap, she almost always 3 2 1 X
smiles and seems happy.
4. When she wakes up from a nap, she often is a 1 2 3 X
Very much Somewhat Not at
Temperament like my like my all like Don't
Dimension Items baby baby my baby know
5. When I feed her and I need to interrupt the 1 2 3 X
feeding for such things as burping, she seems
to fuss for a bit when these interruptions occur.
6. She almost always had a fussy period each day. 1 2 3 X
(A separate set of items with masculine pronouns is used for male infants.)
Ainsworth, M. D. S., & Bell, S. M. Mother-infant interaction and the
development of competence. In K. Connolly & J. Bruner (Eds.),
The growth of competence. New York: Academic Press, 1974.
Ambrose, A. The concept of a critical period for the development of
social responsiveness in early human infancy. In B. Foss (Ed.),
Determinants of infant behavior II, London: Methuen, 1963.
Appleton, T., Clifton, R., & Goldberg, S. The development of behavioral
competence in infancy. In F. Horowitz, E. Hetherington, S. Scarr-
Salapatek, G. Siegle (Eds.), Review of child development research
(Vol. 3). Chicago: University of Chicago Press, 1975.
Barbero, G. Failure to thrive. Paper presented at a conference
sponsored by Johnson & Johnson Baby Products Company, Sausalito,
California, October 18 & 19, 1974.
Barr, A. J., Goodnight, J., Sall, J., & Hetwig, J. A users guide to
SAS 76. North Carolina, Raleigh: S.A.S. Institute, Inc., 1976.
Bartko, J. The intraclass correlation coefficient as a measure of
reliability. Psychological Reports, 1966, 19, 3-11.
Bartko, J. On various intraclass correlation reliability coefficients.
Psychological Bulletin, 1976, 83(5), 762-765.
Beckwith, L., Cohen, S. E., Kopp, C. B., Parmalee, A. H., & Marcy, T. G.
Caregiver-infant interaction and early cognitive development in
preterm infants. Child Development, 1976, 47, 579-587.
Bell, R. Q. Stimulus control of parent or caretaker behavior by off-
spring. Developmental Psychology, 1971, 4, 63-72.
Birns, B. The emergence and socialization of sex differences in the
earliest years. Merrill-Palmer Quarterly, 1976, 22(3), 229-254.
Blehar, M. C., Lieberman, A., & Ainsworth, M. D. S. Early face-to-face
interaction and its relation to later infant-mother attachment.
Child Development, 1977, 48, 182-194.
Block, J. H. Issues, problems, and pitfalls in assessing sex differences:
a critical review of the Psychology of sex differences. Merrill-
Palmer Quarterly, 1976, 22(4), 283-308.
Bower, T. G. R. The visual world of infants. Scientific American, 1966,
Bower, T. G. R. Development in infancy. San Francisco: W. H. Freeman
and Company, 1974.
Bowlby, J. Attachment and loss. New York: Basic Books, 1969.
Brazelton, T. B., Koslowski, B., & Main, M. The origin of reciprocity:
the early mother-infant interaction. In M. Lewis & L. Rosenblum
(Eds.), The effect of the infant on its caregiver. New York:
Wiley and Sons, 1974.
Brody, S., & Axelrand, S. Anxiety and ego formation in infancy. New
York: International Universities Press, Inc., 1970.
Chess, S. Temperament in the normal infant. In S. G. Sapir & A.
Nitzburg (Eds.), Children with learning problems: readings in a
developmental-interactional approach. New York: Brunner/Mazel
Clarke-Stewart, K. A. Interactions between mothers and their young
children: characteristics and consequences. Monographs of the
society for research in child development, 1973, 38(6-7, Whole
Cronbach, L. J. Essentials of psychological testing (3rd ed.). New York:
Harper and Row Publishers, 1970.
Cox. A. The assessment of parental behavior. Journal of Child
Psychology and Psychiatry, 1975, 16, 255-259.
DeCarie, T. G. Intelligence and affectivity in early childhood. New
York: International Universities Press, 1965.
Escalona, S. K. The roots of individuality. Chicago: Aldine Publishing
Escalona, S. K. Basic modes of social interaction: their emergence
and patterning during the first two years of life. Merrill-Palmer
Quarterly, 1973, 19, 205-232.
Fantz, R. L. Visual perception and experience in early infancy. In H.
Stevenson, H. Rheingold, & E. Hess (Eds.), Early behavior: compara-
tive developmental approaches. New York: Wiley and Sons, 1967.
Fraiberg, S. Blind infants and their mothers: an examination of the
sign system. In M. Lewis & L. Rosenblum (Eds.), The effect of the
infant on the caregiver. New York: Wiley and Sons, 1974.
Gewirtz, J. L. & Boyd, E. F. Mother-infant interaction and its study.
In H. W. Reese (Ed.), Advances in Child Development and Behavior,
1976, 11, 142-163.
Gil, D. G. Violence against children. Cambridge, Mass: Harvard Univer-
sity Press, 1970.
Goldberg, S. Social competence in infancy: a model of parent-infant
interaction. Merrill-Palmer Quarterly, 1977, 23(3), 163-177.
Goldberg, S., & Lewis, M. Play behavior in the one year old infant:
early sex differences. In Child Development, 1969, 40 (1-2), 21-31.
Gordon, I. J. An investigation of the social roots of competence.
Unpublished final report to the National Institute of Mental Health.
Gainesville, Florida: Institute for Development of Human Resources,
University of Florida, 1974.
Gordon, I. J. Parent-infant transactions and infant competency.
Unpublished research proposal to the National Institute of Mental
Health. Gainesville, Florida: Institute for Development of Human
Resources, University of Florida, 1975.
Gordon, I. J., & Jester, R. Instructional strategies in infant stimulation.
Final report to National Institute of Mental Health on project
#5 ROI NIH734-02. Gainesville, Florida: Institute for Development
of Human Resources, University of Florida, 1972.
Haith M. M., & Campos, J. J. Human infancy. In Rosenzeig & Porter
(Eds.), Annual Review of Psychology (Vol. 28). Palo Alto,
California: Annual Review, Inc., 1977.
Haviland, J. Looking smart: the relationship between affect and intelli-
gence in infancy. In M. Lewis (Ed.), Orgins of intelligence: infancy
and early childhood. New York: Plenum Press, 1976.
Hollingshead, A., & Redlich, F. Social class and mental illness. New
York: Wiley and Sons, 1958.
Hummel, T., & Sligo, I. Empirical comparison of univariate and multi-
variate analysis of variance procedures. Psychological Bulletin,
1971, 76(1), 49-57.
Hutt, C, & Ounsted, C. The biological significance of gaze aversion with
particular reference to the syndrome of infantile autism. Behavioral
Science, 1966, 11, 346-356.
Kagan, J. Three themes in developmental psychobiology. In L. Lipsitt
(Ed.), Developmental psychobiology: the significance of infancy.
New YorT Wi ley and Sons, 1976.
Kalninis, I., & Bruner, J. The coordination of visual observation and
instrumental behavior in early infancy. Perception, 1973, 2, 307-314.
Kerlinger, F. N. Foundations of behavioral research (2nd ed.). New
York: Holt, Rinehart, and Winston, Inc., 1973.
Kerlinger, F. N., & Pedhazur, E. J. Multiple regression in behavioral
research. New York: Holt, Rinehart, and Winston, Inc., 1973.
Kirk, Roger E. Experimental design: procedures for the behavioral
sciences. Belmont, California: Brooks/Cole Publishing Company, 1968.
Klaus, M., Kennell, J., Plumb, N., & Zuehlke, S. Human maternal behavior
at the first contact with her young. Pediatrics, 1970, 46(2), 187-192.
Klaus, M. H., & Kennell, J. A. Parent-to-infant attachment. In V. C.
Vaughan, III & T. B. Brazelton (Eds.), The family--can it be saved?
New York: Medical Yearbook Publishers, Inc., 1976.
Korner, A. F. The effect of the infant's state, level of arousal, sex,
ontogenetic stage on the caregiver. In M. Lewis & L. Rosenblum
(Ed.), The effect of the infant on its caregiver. New York: Wiley
and Sons, 1974a.
Corner, A. F. Methodological consideration in studying sex differences
in the behavioral functioning of newborns. In R. Friedman, R.
Richart, & R. Vandewiele (Eds.),/Sex differences in behavior. New
York: Wiley and Sons, 1974b.
Kotelchuck, M. The infant's relationship to the father: experimental
evidence. In M. E. Lamb (Ed.), The role of the father in child
development. New York: Wiley and Sons, 1974.
Lamb, M. E. The role of the father: an overview. In M. Lamb (Ed.),
The role of the father in child development. New York: Wiley and
Lamb, M. E. Fathers: forgotten contributors to child development. Human
Development, 1975, 18, 245-266.
Lamb, M. E. Father-infant and mother-infant interaction in the first
year of life. Child Development, 1977, 48, 167-181.
,Lewis, M., & Weinraub, M. Sex of parent x sex of child: socioemotional
development. In R. Friedman, R. Richart, & R. Vandwiele (Eds.),
Sex differences in behavior. New York: Wiley and Sons, 1974.
Lytton, H. Observation studies of parent-child interaction: a metho-
dological review. Child Development, 1971, 42, 651-684.
Maccoby, E., & Jacklin, C. The psychology of sex differences. Stanford,
California: Stanford University Press, 1974.
Martin, B. Parent-child relations. In F. Horowitz, E. Hetherington,
S. Scarr-Salapatek, & G. Siegel (Eds.), Review of child development
research (Vol. 4). Chicago: University of Chicago Press, 1975.
McCall, R. Addendum to Chapter 18. In P. H. Mussen (Ed.), Carmichael's
manual of child psychology (Vol. I) (3rd ed.). New York:
Wiley and Sons, 1970.
McCall, R. Challenges to a science of developmental psychology. Child
Development, 1977, 48, 333-344.
/Moss, H. A. Early sex differences and mother-infant interaction. In
S R. Friedman, R. Richart, & R. Vandewiele (Eds.), Sex differences in
behavior. New York: Wiley and Sons, 1974.
Moss, H., & Robson, K. Maternal influences in early social-visual
behavior. Child Development, 1968, 39, 401-408.
Moss, H., & Robson, K. The relation between the amount of time infants
spend at various states and the development of visual behavior.
Child Development, 1970, 41, 509-517.
Moss H., Robson, K., & Pedersen, F. Determinants of maternal stimulation
of development and consequences of treatment for later reaction of
strangers. Developmental Psychology, 1969, 1, 239-246.
Murphy, L., & Moriarity, A. Vulnerability, coping, and growth: from
infancy to adolescence. New Haven: Yale University Press, 1976.
Nash, J. Historical and social change in the perception of the role of
the father. In M. E. Lamb (Ed.), The role of the father in child
development. New York: Wiley and Sons, 1976.
Nunnally, J., & Lemond, L. Exploratory behavior and human development.
In H. Reese (Ed-.), Advances in child development and behavior,
1973, 8, 59-109.
Olson, C. L. On choosing a test statistic in multivariate analysis of
variance. Psychological Bulletin, 1976, 83(4), 579-586.
Osofsky, J., & O'Connell, E. Parent-child interaction: daughters'
effects upon mothers' and fathers' behaviors. Development Psychology,
1972, 7(2), 157-168.
Parke, R. D. Father-infant interaction. Paper presented at a conference
sponsored by Johnson and Johnson Baby Products Company, Sausalito,
California, October 18 and 19, 1974.
Pedersen, F. A., Anderson, B. J., & Cain, R. L. A methodology for assessing
parental perception of infant temperament. Paper presented at the
Fourth Biennial Southeastern Conference on Human Development,
Piaget, J. The origins of intelligence in children. New York: Inter-
national Universities Press, 1952.
Piaget, J. Six psychological studies. New York: Random House, 1967.
Piaget, J. Piaget's theory. In P. H. Mussen (Ed.), Carmichael's manual
of child psychology (Vol. I) (3rd ed.). New York: Wiley and
Piaget, J. The construction of reality. New York: Ballantine Books,
Prechtl, H. F. R. The mother-child interaction in babies with minimal
brain damage. In B. M. Foss (Ed.), Determinants of infant behavior
(Vol. II). London: Methuen, 1963.
Provence, S., & Lipton, R. C. Infants in institutions. New York:
International Universities Press, 1962.
Ramey, C., Starr, R., Pallas, J., Whitlen, C., & Reed, V. Nutrition,
response--contingent stimulation, and the maternal deprivation
syndrome: results of an early intervention program. Merrill-
Palmer Quarterly, 1975, 21(1), 45-53.
Rheingold, H. L. The effect of environmental stimulation upon social and
exploratory behavior in the human infant. In B. M. Foss (Ed.),
Determinants of infant behavior. New York: Wiley and Sons, 1961.
Rheingold, H. L. A comparative psychology of development. In H. Stevenson,
E. Hess, & H. Rheingold (Eds.), Early behavior. New York: Wiley &
Robson, K. S. The role of eye-to-eye contact in maternal-infant attach-
ment. Journal of Child Psychology and Psychiatry, 1967, 8, 13-25.
Robson, K. S., Pedersen, F. A., & Moss, H. A. Developmental observations
of diadic gazing in relation to fear of strangers and social approach
behavior. Child Development, 1969, 40, 619-627.
Sameroff, A. J. Early influences on development: fact or fancy?
Merrill-Palmer Quarterly, 1975, 21(4), 267-294.
Sander, L. Issues in early mother-child interaction. In E. Rexford, L.
Sander, & T. Shapiro (Eds.), Infant psychiatry. New Haven: Yale
University Press, 1976.
Scarr-Salapatek, S. Genetic determinants of infant development: An over-
stated case. In L. Lipsitt (Ed.), Developmental psychobiology. New
York: Wiley & Sons, 1976.
Spitz, R. A. The first year of life: a psychoanalytic study of normal
and deviant development of object relations. New York: Inter-
national Universities Press, Inc., 1965.
Stechler, G., & Carpenter, G. A viewpoint on early affective development.
In J. Helmuth (Ed.), Exceptional infant (Vol. I). New York:
Brunner/Mazel, Inc., 1967.
Stern, D. A microanalysis of mother-infant interaction. In E. Rexford,
L. Sander, & T. Shapiro (Eds.), Infant psychiatry. New Haven:
Yale University Press, 1976.
Stern, D. Mother and infant at play: the dyadic interaction involving
facial, vocal and gaze behaviors. In M. Lewis & L. Rosenblum (Eds.),
The effect of the infant on its caregivers. New York: Wiley and
Thomas, A., Chess, S., Birch, H., Hertzig, M., & Korn, S. Behavioral
individuality in early childhood. New York: University Press, 1963.
Thomas, A., Chess, S., & Birch, H. Temperament and behavior disorders
in children. New York: University Press, 1968.
Watson, J. Smiling, cooing, and the "game." Merrill-Palmer Quarterly,
1972, 15, 323-340.
White, B. L. Human infants: experience and psychological development.
Englewood Clifts: Prentice-Hall, 1971.
White, B. L., Castle, P., & Held, P. Observations on the development
of visually-directed reaching. Child Development, 1964, 35, 349-364.
Woodward, J. A., & Overall, J. E. Multivariate analysis of variance by
multiple regression methods. Psychological Bulletin, 1975, 82(1),
Yarrow, L. J., Rubenstein, J. L., & Pedersen, F. Infant and environment:
early cognitive and motivational development. New York: Wiley and
JoAnn Howard Patray was born in McAlpin, Florida on September 7,
1934. A summary of her previous educational experiences includes
graduation from Columbia High School (Lake City, Florida) in 1952,
Florida State University (Tallahassee) with a Bachelor of Science in
Nursing in 1956, and the University of Florida with a Master's in Nursing
in 1967. In 1976, she entered the doctorate program in Foundations of
Education at the University of Florida as a full-time student.
Her professional work experience from 1956-1975 includes staff
nursing in Pediatrics, nursing administration, public school teaching
(7th and 8th grade), and teaching nursing in community college and in
an undergraduate and graduate program in a university. Employment for
these experiences was in the following Florida agencies--St. Vincent's
School of Nursing, Jacksonville; Shands Teaching Hospital, Gainesville;
Bradford High School, Starke; Lake City Community College, Lake City;
and the College of Nursing, University of Florida, Gainesville.
Mrs. Patray is married to James W. Patray, a physician, and they
have three children, Briar, Jim, and Beth. The family is residing in
Keystone Heights with exception of Briar, who is attending Florida State
University as a first year student. Jim and Beth attend Bradford High
School in Starke, Florida. The permanent address of Mrs. Patray is
P. 0. Box 145, Keystone Heights, Florida.
I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of Doctor
Patricia Teag46 AsMon, Chairman
Assistant Professor of Education
I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of Doctor
William B. Ware
Professor of Education
I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of Doctor
Paurine H. Barton
Professor of Nursing
This dissertation was submitted to the Graduate Faculty of the Department
of Foundations of Education in the College of Education and to the
Graduate Council, and was accepted as partial fulfillment of the require-
ments for the degree of Doctor of Philosophy.
Chairman, Foundations of Educat-
Chairman, Found~ations of Education
Dean, Graduate School