Luau h UH,
K: aims. 813° 3

UNI": .m "V

—-v,. V"

 

 

This is to certify that the i
dissertation entitled

EXPLORING THE COGNITIVE BASIS
OF INDIVIDUAL DIFFERENCES
IN INTEREST IN INFANTS

presented by

Rodrigo Andres Cérdenas

has been accepted towards fulfillment
of the requirements for the

Ph. D. degree in PsychologL

 

 

Major Professor’s Signature I

W: 51%:10/0
(T

Date :

MSU is an Affirmative Action/Equal Opportunity Employer

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProyAcc&Pres/CIRCIDateDue.indd

EXPLORING THE COGNITIVE BASIS
OF INDIVIDUAL DIFFERENCES
IN INTEREST IN INFANTS
By

Rodrigo Andres Cardenas

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Psychology

2010

Abstract

EXPLORING THE COGNITIVE BASIS OF INDIVIDUAL DIFFERENCES
IN INTEREST IN INFANTS

By
Rodrigo Andrés Cardenas

Because infant care is essential to human survival and reproductive success, it has
been hypothesized that humans have evolved biological and cognitive mechanisms to
facilitate adults’ interest in and responsiveness to infants in ways leading to care-giving.
Evidence for such mechanisms comes from studies showing that adults have attentional-
emotional biases, and make distinctive physiological, neural, and hormonal responses, to
infants, specifically to those physical features and behaviors that distinguish them from
older humans. These biases also are normally stronger in women than in men, consistent
with the hypothesis that women, due to their central role in infant care, have evolved a
greater and more stable sensitivity to infants. However, despite evidence that infants
constitute a special stimulus category, the nature of the cognitive mechanisms underlying
interest in infants is still largely unexplored. This dissertation examines whether one such
mechanism is visual attention toward infants, and, if so, how it is expressed (e.g., in
overall looking time, face-recognition, and initial deployment of attention) and whether
and how it is associated with adults’ self-described interest in infants. Seven studies with
young adults were conducted to address these questions. Studies 1-5 were preliminary
and assessed the reliability/validity of three questionnaire-based measures of interest in
infants and the characteristics of the sample of faces used in the main studies. Study 6,

the first main study, used eye-tracking to measure subjects’ visual attention while they

viewed an adult face (a man or woman) paired with an infant face (a boy or girl).
Subjects also completed the interest-in-infants questionnaires and a face recognition test.
As indexed by the number and length of fixations, the results showed that women looked
more at infant than at adult faces of either sex, whereas men looked more at infant faces
only when paired with an adult male face. Women also reported greater interest in infants
than men. The interest scores, however, were consistently associated with visual attention
toward infants only for men (possibly due to a range restriction in women’s interest
scores). The results thus suggest that women’s interest in infants as indexed by visual
attention and self-report is more stable and higher than men’s. Face recognition scores
were not associated with interest-in-infants scores, perhaps because of the difficulty of
the recognition task. Study 7, the second main study, examined whether infant faces have
a greater effect than adult faces on the initial deployment of attention — the so-called
“prior-entry effect”. The results suggested that infant faces do not have this effect or at
least that it is smaller than that reported for other emotional stimuli, such as angry faces.
There also was no association between individuals’ point of subjective simultaneity and
their self-reported interest in infants, consistent with the negative evidence from the prior
entry task. The results thus suggest that the attentional and neural effects reported in other
studies are unlikely to appear during the initial deployment of attention. Overall, the
results suggest that women’s interest in infants is more stable than men’s, consistent with
the hypothesis about the evolution of sex differences. They also show that oculomotor
behavior can be successfully used to assess individual differences, including sex

differences, in interest in infants.

Acknowledgments

I would like to express my deepest gratitude to my advisor, Lauren Harris, for his
invaluable mentoring, help throughout my graduate studies, and for providing an
excellent research environment. I admire his generosity and dedication toward his
students, and I owe an enormous intellectual debt to his expertise and sharp thinking. It is
an honor to work with him and now, for my future students, I have very big shoes to fill.

I also would like to thank the rest of my committee members, Jason Almerigi,
Mark Becker, Thomas Getty, Zack Harnbrick, whose advice and guidance has also been
fundamental in this work. I am especially grateful to Mark Becker for kindly granting me
access to his lab, equipment and software, without which I would not have been able to
conduct this research.

I also would like to thank Mark Eden and Andrew Larsen for their assistance in
data collection and for all their insightful input (lining the study. Alena Patsenka also
generously lent me eye-tracking software without which this work would have been
delayed substantially.

I want to thank my wife, Karen, for all her support and encouragement throughout
my graduate studies and especially during the completion of this dissertation. My son,
Nico, and daughter, Sofi, have been a continuous inspiration for this work. I will always
be grateful to my parents, Pedro and Marisol, my grandparents, Tadeo and Edith, and my
uncle, Jorge, for all their efforts, continuous support and unconditional love.

I was able to complete and write this dissertation in a timely manner thanks to the

Dissertation Completion Fellowship from the Graduate School.

iv

Table of Contents

List of Tables ................................................................................................................ vii
List of Figures ................................................................................................................ ix
Chapter 1: Introduction .................................................................................................... 1
Chapter 2: Individual Differences in Interest in Infants .................................................... 7
High Costs of Care for Human Infants ..................................................................... 7
Role of Cooperative Breeding .................................................................................. 8
Infanticide ............................................................................................................. 10
What Makes Infants Appealing? ............................................................................ 10
Generalization Effect to Other Species ................................................................... 14

Sex Differences ..................................................................................................... 14
Hormones .............................................................................................................. 1 5
Neural Mechanisms ............................................................................................... 18
Summary ............................................................................................................... 19
Chapter 3: Psychosocial Measures of Individual Differences in Interest in Infants ......... 21
Studies 1 and 2 .......................................................................................................... 21
Interest in Infants ................................................................................................... 22

Job Preference Questionnaire ................................................................................. 23
Desire for Parenthood ............................................................................................ 24
Validity of Interest in Infants and Job Preference questionnaires ............................ 24
Interaction with Infants .......................................................................................... 25
Revised Sociosexual Orientation Inventory ............................................................ 28
Self-Perceived Mate Value Scale ........................................................................... 29
Chapter 4: Composition and Characteristics of Facial Stimuli ........................................ 30
Facial Stimuli ............................................................................................................ 30
Face Ratings .............................................................................................................. 34
Study 3 ...................................................................................................................... 34
Valence, Arousal, and Dominance Ratings ............................................................ 34
Subjects ................................................................................................................. 35
Procedure .............................................................................................................. 35
Results ................................................................................................................... 37
Study 4 ...................................................................................................................... 41
Masculinity/femininity, Attractiveness ratings ....................................................... 41
Subjects ................................................................................................................. 41
Procedure .............................................................................................................. 41
Results ................................................................................................................... 42
Study 5 ...................................................................................................................... 44
Distinctiveness and sex identification .................................................................... 44
Subjects ................................................................................................................. 45
Procedure .............................................................................................................. 45

Results ................................................................................................................... 46

Chapter 5: Study 6 -- Individual Differences in Visual Processing of Infant and Adult

Faces ............................................................................................................................. 51
Visual attention and eye-movements ...................................................................... 51
Subjects ................................................................................................................. 53
Procedure .............................................................................................................. 53

Results ....................................................................................................................... 56
Questionnaires ....................................................................................................... 56
Eye movements ..................................................................................................... 58
Face recognition .................................................................................................... 68

Discussion ................................................................................................................. 73

Chapter 6: Study 7 -- Visual prior entry for infant faces ................................................. 77
Subjects ................................................................................................................. 79
Stimuli ................................................................................................................... 79
Procedure .............................................................................................................. 79

Results ....................................................................................................................... 82
Questionnaires ....................................................................................................... 82
Prior Entry ............................................................................................................. 84

Discussion ................................................................................................................. 85

Chapter 7: Conclusions, Limitations, and New Directions ............................................. 87
Limitations of Studies, Further Analyses, and Directions for New Research .......... 90
Use of Undergraduates ........................................................................................... 90
Further analyses of current data ............................................................................. 91
Facial Expression ................................................................................................... 92
Hormonal Changes ................................................................................................ 92
Refinement of Oculomotor Measures ..................................................................... 93

References ..................................................................................................................... 96

vi

List of Tables

Table 1. Zero-Order Correlations Across Questionnaires by Sex of Subjects ................. 25

_ Table 2. Means and (Standard Deviations) of Positive Behaviors of Men and Women for
Interacting with an Infant Under Three Different Emotional States ........................ 27

Table 3. Zero-Order Correlations Between the Job Preference and Interaction with Infants
Questionnaires ....................................................................................................... 28

Table 4. Pixel Area of Each Face Category .................................................................... 31

Table 5. Means and (Standard Deviations) of Apparent Contrast for Each Frequency
Band by Sex of the Face ........................................................................................ 33

Table 6. F-statistics, Effect Sizes, and Significant Group Differences for Unfiltered and
Filtered Frequency Bands ...................................................................................... 33

Table 7. Means and (Standard Deviation) for Valence, Arousal, and Dominance Ratings
for Each Face Category by the Sex of the Subject .................................................. 37

Table 8. Means and (Standard Deviations) of Valence, Arousal, and Dominance Ratings
by Face Category and Sex of the Subject ............................................................... 39

Table 9. Mean and (Standard Deviations) of Attractiveness and Masculinity Ratings by
Face Category and Sex of Subjects ........................................................................ 42

Table 10. Mean (and Standard Deviation) of Distinctiveness Ratings and Sex
Classification Scores for Each Face Category by Sex of the Subject ...................... 46

Table 11. Means and (Standard Deviations) of A’ and B” Scores for Sex Classification
of Infant and Adult Faces by Sex of Subject .......................................................... 48

Table 12. Descriptive Statistics and Sex Differences of the Interaction with Infants
Questionnaire ........................................................................................................ 57

Table 13. Descriptive Statistics and Sex Differences of the Sociosexuality Measures 58

Table 14. Mean and Standard Deviation of Infant Viewing Duration for each
Experimental Condition by Sex of Subject ............................................................. 60

Table 15. Summary of a 2-level Random-Intercepts Model of Subjects Looking Time
Toward Infant Faces .............................................................................................. 64

Table 16. Summary of a 2-Ievel Random-Intercepts Model of Subjects’ F ixations Toward
Infant Faces ........................................................................................................... 66

vii

Table 17. Summary of a 2-level Random-Intercepts Model of Subjects’ F ixations Toward
Infant Faces as Predicted by Their Interest in Infants ............................................. 68

Table 18. Mean and Standard Deviation A’ Scores by Face category and Sex'of the
Subjects ................................................................................................................. 69

Table 19. Proportion of Trials, by Face Category, in which a Face was Classified as Seen
or as a New Face .................................................................................................... 70

Table 20. Summary of a 2-level Random-Intercepts Model of Subjects’ Face Recognition
Responses .............................................................................................................. 71

Table 21. Descriptive Statistics and Sex Differences of the Interaction with Infants
Questionnaire ........................................................................................................ 83

Table 22. Descriptive Statistics and Sex Differences of the Sociosexuality Measures 84

viii

List of Figures

Figure 1. The Self-Assessment Manikin (SAM) Used to Acquire Ratings of Valence,

Arousal, and Dominance ........................................................................................ 36
Figure 2. Predicted Viewing Duration for Infant Faces. ................................................. 64
Figure 3. Trial Sequence Used in Experiment 7. ............................................................ 8O

ix

Chapter 1: Introduction

All animal species with complex nervous systems face a fundamental and
complex information-processing problem: their environment contains vast amounts of
information that cannot be exhaustively processed by the animal’s limited capacity
cognitive system. As a result, only some information receives further processing, other
information is partially processed, and still other information is ignored altogether. The
problem is that not all information has the same survival/reproductive value; in other
words, the animal faces a complex landscape of inherent costs and benefits every time it
processes some information and ignores other information. Making the landscape even
more complex is that the costs and benefits are not necessarily stable across time, and
sometimes even information that is ignored can be important (e.g., by minimizing
informational interference). Thus, not only is cognition effortful (e.g., in energy and time
spent and in risks taken), it also requires an allocation of limited resources, an allocation
that can be extremely difficult because of the complexity of the decision-rules. For all
these reasons, it has been hypothesized that through evolution, certain information-
processing mechanisms have been selected that are more efficient at processing
information that is recurrently relevant (beneficial) for a species; that is, certain attention
and memory systems became tuned to attend to and to remember information most
relevant for survival and reproductive success (e. g., Naime, Pandeirada, & Thompson,
2008; New, Cosmides, & Tooby, 2007). Given the survival/reproductive value of
attending to and remembering information about infants, the possibility thus arises that

there are attentional/memory biases toward infant information that could facilitate a

caregiver’s response, particularly in time-sensitive situations that can place an infant at
risk.

Human infants are secondarily altricial (Martin, 2007; Portmann, 1941), which
means that their survival depends on intensive parental and alloparental care, with
mothers and other females typically having the more central roles (e. g., Clutton-Brock,
1991; Hewlett, 1992; Hrdy, 1999). Because infant care is essential to human survival, it
has been hypothesized that, over the course of human evolution, a variety of
biobehavioral mechanisms have evolved to facilitate adults’ interest in and
responsiveness to infants in ways normally leading to care-giving (e.g., Babchuk, Hames,
& Thompson, 1985; Lorenz, 1971). Evidence for such mechanisms comes from studies
showing that adults have attentional-emotional biases, along with physiological,
neurological, and hormonal responses, to features and behaviors that make infants look
“cute” and “attractive,” and that these biases are normally stronger in women than in men
(see Chapter 2).

Despite ample evidence that infants constitute a special stimulus category for
humans and that infants are more salient for women, the nature of the cognitive
mechanisms underlying interest in infants is still largely unexplored; that is, there has
been little work to demonstrate experimentally that there are cognitive mechanisms tuned
to detect, encode/store, and process infant-related information and how they may vary as
a fimction of individual differences in interest in infants.

The question raised in this dissertation is whether one of the cognitive
mechanisms is expressed through visual attention to faces, and, if so, does it behave as a

function of a variety of individual differences indexed by self-report questionnaires?

Seven studies were designed to address this question. The first five were preliminary, the
last two were the main studies: Studies 1 and 2 tested the reliability and estimated the
validity of the self-report measures of interest in infants; Studies 3, 4, and 5 assessed
certain potentially important characteristics of a sample of infant and adult faces to be
used in the main studies, including their perceived arousal value, dominance,
masculinity/femininity, distinctiveness, and attractiveness; and the main studies, 6 and 7,
examined whether specific oculomotor behaviors are associated with the visual
processing and subsequent recognition of infant faces, whether infant faces capture
attention over adult faces using a visual prior entry task, and, finally, whether the
behavioral data (eye movements, recognition, and attentional capture) are associated with
psychosocial indices of interest in infants.

What follows is an overview of the dissertation. Following this Introduction
(Chapter 1), Chapter 2 reviews the psychosocial and biological correlates of interest in
infants in humans and nonhuman primates, and identifies factors that affect individual,
especially sex-related, differences in this interest. It therefore highlights the
idiosyncrasies and commonalities of human infant care compared to other mammals and
nonhuman primates. It also provides illustrations of factors that are related to adults’
interest in infants, including, but not limited to the adult’s sex, age, parental status,
hormonal status, relationship to the infant, and mating and reproductive strategies (e. g.,
whether one is interested in having children in the short-term vs. long-term or simply not
interested). Thus, although sex differences in interest in infants are expected, with women
being more interested than men, there also should be variability within each sex, and

possibly more in males than in females.

Chapter 3 describes three questionnaires that were developed to measure subjects’
interest in and willingness to interact with infants. It includes a discussion of the
psychometric properties of these measures, along with preliminary evidence of their
construct validity based on how the interest in infants questionnaires relate to other
questionnaires (e.g., desire to have children). Finally, it describes additional
questionnaires that were used to measure other variables, such as mating strategies, that
might contribute to individual differences in performance in the studies and in interest in
infants.

Chapter 4 describes the face database used in the main studies (studies 6 and 7). It
explains how the faces were obtained, the criteria used for their selection, and what
computer graphic techniques were used to make them suitable for use. The chapter also
discusses the limitations of standardization of a database that includes infants and adult
faces of both sexes; Methods and procedures used for data collection are described, and
the results are presented and discussed.

Chapter 5 describes Study 6, on individual differences in visual processing and
memory of infant and adult faces. It discusses how eye movements can be used as
indexes of attention, in particular, how the location and duration of fixations tend to
cluster in the more interesting and informative regions of the visual stimuli. Taking
advantage of the intrinsic and intricate relation between oculomotor behavior, attention,
and how visual information is represented in real time, the study used eye-tracking to
measure the looking times and fixation locations of men and women, all nulliparous-
heterosexual undergraduates, while they viewed a sequence of pairs of faces, with each

pair consisting of one adult face (either a man or a woman) and one infant face (either a

boy or a girl). Subjects then completed questionnaires designed to measure their interest
in infants and subsequently received a recognition memory task at the end. Faces were
presented in pairs consisting of an adult and an infant face in order to measure
oculomotor behavior in contexts where infant faces are competing for attentional
resources. The chapter discusses the results in terms of sex differences in oculomotor
behavior and face recognition and in how the differences are related to the psychosocial
measures used in the study.

Chapter 6 describes Study 7 — on visual prior entry for infant faces. There is
evidence that people are predisposed to attend to emotional stimuli (e. g., a fearful face),
even when told not to do so. Recent studies, using the dot-probe-task, show that infant
faces can produce covert shifts of spatial attention; however, although the task is sensitive
enough to detect shifts of attention (e. g., Brosch, Sander, Pourtois, & Scherer, 2008;
Brosch, Sander, & Scherer, 2007), its temporal resolution at face onset (when the faces
are first presented) is insufficient for determining precisely when and how infant faces
capture attention. To allow measurement of the time course of initial attentional
deployment, Study 7 uses a prior entry task, a method with better temporal resolution
during face onset. The task assumes that those items that capture attention are perceived
as appearing earlier than items that do not capture attention. The chapter describes the
methods used and the results, along with a discussion of the results in terms of the
psychosocial measures used in the study.

The last chapter, Conclusions, returns to the general issues presented in the
Introduction and expanded in subsequent chapters. It highlights the value of studying the

cognitive basis of individual differences in interest in infants using evolutionary theory as

a heuristic tool, and it summarizes the major findings of Studies 6 and 7 and discusses
their contributions to the literature along with their limitations. It also discusses the
importance of understanding the factors that contribute to individual differences in
interest in infants from an applied science perspective. Finally, directions for future

studies are proposed.

Chapter 2: Individual Differences in Interest in Infants

As described earlier, given that offspring survival is critical for reproductive
success and that human infants require intensive and extensive care, it is likely that
humans are endowed with specialized cognitive mechanisms attuned to facilitate infant
care. This chapter reviews the psychosocial and biological correlates of interest in infants
in humans and nonhuman primates, and identifies factors that affect individual, especially

sex-related, differences in this interest.

High Costs of Care for Human Infants

In general, primate infants are costly to rear because they require high amounts of
energy and care (e.g., feeding, grooming, carrying, protecting from predators) and
because the primate’s slow pace of maturation requires care over an extended period of
time (Kaplan & J. Lancaster, 2003; Kramer, 2005; J. Lancaster & C. S. Lancaster, 1987;
Lee & Kramer, 2002). By these criteria, human infants and children are the most costly
by far. At birth, they are about 3 times fatter than expected for a mammal of their size
(Hrdy, 2009, 1999) and, compared to closely related primates of similar size, take twice
as long as to reach maturity and to produce children on their own (Hrdy, 2009; Kramer,
2005). In addition to these costs, there are costs in mating opportunities with others.
Despite all these costs, humans reproduce at significantly faster rates, as indexed by
interbirth intervals, than the average for Great Apes. For orangutans, the interbirth
interval is about 8 years, for chimpanzees over 5, and for gorillas about 4 years, whereas
for humans living in populations with natural fertility, it is closer to 3 years (Galdikas &

Wood, 1990; Sear & Mace, 2008; Sellen, 2007).

Role of Cooperative Breeding

The fact that humans can reach higher than expected reproductive rates despite
the high cost of rearing human infants highlights the critical role of c00perative breeding
in human populations, that is, allomatemal care with provisioning (e.g., share suckling).
This is another characteristic that sets humans apart from the Great Apes. All four —
chimpanzees, gorillas, orangutans, and gibbons -- have continuous-care-and-contact
mothering; in other words, after birth the mother is the sole provider of direct and
continuous care for the infant, with whom she remains in continuous contact — a behavior
due largely to her possessiveness and not to the lack of interest of potential care givers
(Hrdy, 2009). In contrast, humans are cooperative breeders, meaning that a substantial
portion of infant care is provided by someone other than the mother or the father, usually
a close relative. Another indication of the species difference is in the mother’s
willingness to let other individuals hold her infant. The earliest this has been observed in
wild chimpanzees is when the infant is 3 and one-half months old, and in wild
orangutans, 5 months; in human mothers it can be observed immediately after birth
(Hrdy, 2009). Still another difference is in shared suckling which is not observed in wild
apes but is common in humans, being documented in about 87% of foraging societies
described in the Human Relations Area Files (Hewlett, 1989a, cited by Hrdy, 2009). And
still another example is the wide-spread practice of human wet-nursing, a wide-spread
practice that was institutionalized in France in the 18th and 19th centuries (e. g., in Paris,
there was a formal registry of wet-nurses, Sussman, 1977).

Although alloparental care is uncommon among Great Apes, it is common across

the order Primates, with about 45% of species showing it in some form. However, only

about 20% of species, including humans, Show both alloparenting and provisioning
(cooperative breeding) (Hrdy, 2009).

In the vast majority of primates, fathers do not provide direct infant-care; instead,
they protect the troop from other males attempting to kill the infant. In this sense,
primates follow the norm for mammalian species, since only about 5 percent of mammals
Show male parenting (Clutton-Brock, 1989). Thus, in most alloparental primates, mothers
rely on other females (typically close kin) for infant caregiving. In primate species that
show cooperative breeding, such as Callitrichidae (most marmosets and tamarins),
fathers play a substantial role in parental care (F ernandez-Duque, Valeggia, & Mendoza,
2009). By contrast, in humans, across cultures, mothers are the greatest contributors to
direct infant care —about 50% of the direct care (Kramer, 2005; Marlowe, 2005), with
siblings, grandmothers, and the father contributing to lesser extents (Marlowe, 2005). In
fact, by some analysis, fathers have little or no effect on infant survival (Sear & Mace,
2008). The variable impact of male parenting in humans indicates that male parental
investment is facultative rather than obligate; that is, although not required for the
offsprings’ survival, it seems to enhance their survival and reproductive success (Geary,
2008). Several conditions affect the likelihood of male parental investment, including the
degree to which it contributes to offspring survival and reproductive success, certainty of
paternity, and the cost of loss of mating opportunities with other females. Thus, although
men on average show less interest in infants than women, under certain conditions such
differences might be reduced or even absent (e.g., in new fathers with a monogamous

reproductive strategy).

Infanticide

Another fundamental difference between humans’ and other apes’ parental
behavior is that only humans are known to commit parental infanticide. In other primates,
although infanticide is one of the most common causes of infant mortality, it typically is
carried out not by the parent but by a male trying to increase his mating opportunities by
speeding up the sexual availability of the mother or by another mother trying to reduce
rivals of her own offspring. Indeed, not only is parental infanticide not found in
nonhuman great Apes, maternal investment and care are almost guaranteed however
unviable an infant appears to be, whereas in humans, as well as in other cooperative
breeders, infanticide can be carried out by the mother when the infant is suspected not to
be viable or when sufficient alloparenting is unavailable (Daly & Wilson, 1984;
Hausfater & Hrdy, 1984; Hrdy, 1999, 2009). This suggests that although women are
more likely to be more interested in infants than males, their interest may depend on the
socioecological conditions that the woman faces. The implication is that since human
infants are born without the certainty of parental care, they have to be especially
appealing to the mother (or allomothers) in order to elicit that care; in particular, they

must show that they are viable and therefore worth the investment (Hrdy, 2009, 1999).

What Makes Infants Appealing?

The evidence of parental infanticide notwithstanding, most human adults in fact
find human infants appealing and attractive. In a seminal paper published in (1943;
reprinted in 1971), the Austrian ethologist Konrad Lorenz(1971) introduced the concept
of Kindchenschema (or baby schema), according to which certain physical and behavioral

characteristics of infants attract and induce care-giving behaviors in adults. Lorenz

10

intuitively suggested that these characteristics included “a relatively large head,
predominance of the brain capsule, large and low-lying eyes, bulging cheek region, short
and thick extremities, a springy elastic consistency, and clumsy movements” (Lorenz,
1971, p. 154). He believed that infant features trigger “innate releasing mechanisms,” that
is, neurophysiologically hardwired pattern of behaviors associated with affection and

nurture (Gould, 1980; Lorenz, I971; Tinbergen, 1951).
Subsequent studies have generally supported Lorenz’s intuition. For instance,

most adults rate images of infants as cuter and more attractive than images of adults (e. g.,
Fullard & Reiling, 1976; Maestripieri & Pelka, 2002), and their ratings have been
independently linked to specific infant features, including the Size and position of the

eyes, size of the forehead, length of the nose, and shape of the head (Alley, 1981; Brooks

& Hochberg, 1960; Glocker, Langleben, Ruparel, Loughead, Gur, et al., 2009;
Hildebrandt & Fitzgerald, 1979b; Hijckstedt, 1965; Stemglanz, Gray, & Murakami,
I 1977)

Along with attractiveness ratings, several behavioral indicators and other self-
report measures (such as mood) are also associated with exposure to the infant schema.
Adults smile more when looking at infant images (Bradley, Codispoti, Sabatinelli, &
Lang, 2001; Hildebrandt & Fitzgerald, 1978), look longer at cuter than at less cute infants
(Hildebrandt & Fitzgerald, 1981, 1978), are more likely to use babytalk (e.g., high-
pitched vocalizations accompanied by the use of diminutives and simplified short-
sentence structures) when interacting with children with more infant-like features

(Zebrowitz & Brownlow, 1992), and are more willing to take care of cuter than less cute

11

infants (Glocker, Langleben, Ruparel, Loughead, Gur, et al., 2009). Cuter infants also

have been found to receive more caregiver-attention in a group program for infants
(Hildebrandt & Cannan, 1985).

Studies have also examined specific psychophysiological correlates of the infant
schema, in particular, autonomic activity such as pupil dilation, heart rate, and Skin
conductance. A pilot-study of pupillary dilation (greater dilation indicating greater arousal)
in Six nulliparous adults (2 women) found that the women had greater pupil dilation than
the men when looking at infant pictures compared to looking at “control patterns” (Hess
& Polt, 1960). Another study with a large sample of children and adolescents (Bemick,
1966) also found greater pupil dilation when subjects viewed infant pictures, but this
time, although girls expressed higher preference for infant pictures, boys showed greater
pupil dilation. The discrepancy between studies might be due to differences in sample
size and subject age, but it might also be related to the imperfect measuring techniques
used in both studies (Bradley, Miccoli, Escrig, & Lang, 2008) (e.g., pupils were
photographed and sampled at a lower rate than is now possible using modern, video-
based, infi'ared eye-tracking systems).

Personality characteristics also affect physiological responses and sensitivity to
infants. When watching videos of infants who are either smiling, crying, or quiescent,
young nulliparous women with high empathy (in contrast to women with lower empathy
based on their scores on the Mehrabian and Epstein, 1972, empathy questionnaire) show

higher skin conductance responses, a trend toward greater cardiac responsiveness, and a

12

higher expressed desire to pick up and hold the infants (Wiesenfeld, Whitman, &
Malatesta, 1984).

Physiological responses also change depending on the infant’s emotional status.
When watching videos of babies smiling or crying, children and parents show cardiac

deceleration when the babies are smiling and acceleration when they are crying (Frodi &
Lamb, 1978; Frodi, Lamb, Leavitt, & Donovan, 1978).
Physiological responses are also affected by parity (parenthood status) and the

filial relationship between the viewer and the infant. For instance, mothers show higher
skin conductance responses (Hildebrandt & Fitzgerald, 1981; Wiesenfeld & Klorman,
1978) and cardiac acceleration (Wiesenfeld & Klorman, 1978) when looking at pictures of
their own newborns than at pictures of unfamiliar infants.

In a more cognitive domain, the infant schema also appears to produce covert
shifts of spatial attention (Brosch et al., 2008; Brosch et al., 2007). These studies used the
dot-probe-task, in which two faces (in this case an infant face and an adult face, both with
neutral expression) are presented simultaneously for a short time (e.g., 100 milliseconds)
and then turned off, after which a small dot appears briefly in the location previously
occupied by one of the faces. Adults are typically faster (Brosch et al., 2007) or more
accurate (Brosch et al., 2008) at detecting the dot when it appears in the location
occupied by the infant face; that is, people seem to improve because their attention was
directed, or captured, by the infant face. Given that infant faces are more emotionally
arousing than adult faces (e. g., Brosch et al., 2007), this finding is consistent with reports
that emotional stimuli have important effects on perception and attention (Ohman, Flykt,

& Esteves, 2001; Phelps, Ling, & Carrasco, 2006).

13

Generalization Effect to Other Species

Lorenz (1971) suggested that the infant schema also affects our perception of
other species. Indeed, adults, especially women, tend to prefer images of infant animals
over images of adult animals (Berman, Cooper, Mansfield, Shields, & Abplanalp, 1975;
Fullard & Reiling, 1976). This preference for the infant schema has also shaped the
appearance of cultural artifacts such as cartoons (Eibl-Eibesfeldt, 2007; Gould, 1980;
Pittenger, 1990), the teddy bear (Hinde & Barden, 1985; Morris, Reddy, & Bunting,
1995), and perhaps the domestication of some animals (Byme, 2005; Lorenz, 1971). It
even appears to affect one’s ability to perform care-giving acts: viewing images of cute
animals (puppies and kittens) enhances fine-motor dexterity on a task demanding
carefulness (the game Operation) (Sherman, Haidt, & Coan, 2009). However, although
infant-animal faces are appealing, they do not induce the attentional capture effect found

for human infant faces when compared to adult human faces (Brosch et al., 2007).

Sex Differences

Consistent with the female’s more central role in infant-care (Babchuk, Hames, &
Thompson, 1985; Clutton-Brock, 1991; Hewlett, 1992; Hrdy, 1999), women and girls
normally Show higher interest in and responsiveness to infants than do men and boys
(Blakemore, 1981; Bradley, Codispoti, Sabatinelli, & Lang, 2001; F eldman, Nash, &
Cutrona, 1977; Feldman & Nash, 1978; Fullard & Reiling, 1976; Maestripieri & Pelka,
2002). For example, among the behaviors listed earlier, women are more likely to use
baby talk than men (Zebrowitz & Brownlow, 1992), girls and women tend to prefer
images of human infants and baby animals and to give them higher cuteness ratings (e. g.,

Fullard & Reiling, 1976; Maestripieri & Pelka, 2002; Sherman et al., 2009), girls,

14

especially if older than 5 years, are more willing than boys to interact with a baby
(Berman & Goodman, 1984; F rodi & Lamb, 1978), women report to have more
experience in childcare (Gilpin, 1988; Gilpin & Glanville, 1977, 1985), be more willing
to take care of infants (Glocker, Langleben, Ruparel, Loughead, Gur, et al., 2009), are
more likely to look at infants (Robinson, Lockard, & Adams, 1979), to pick up and hold
them (Harris, Spradlin, & Almerigi, 2007), are more sensitive to differences in cuteness in
infant faces (Lobmaier, Sprengelmeyer, Wiffen, & Perrett, 2010; Sprengelmeyer et al.,
2009), are faster and more accurate in recognizing infant facial expressions (Babchuk et
al., 1985), and, for mothers compared to fathers, Show better auditory recognition of their
infants cries (Green & Gustafson, 1983).

The sex differences described above, although generally robust, also show
inconsistencies developmentally; that is, they are not always found in different age
groups. The reason is possibly due, in part, to complex interactions between
biobehavioral and sociocultural variables and to differences in methods and research
designs (Berman, 1980; Fitzgerald, Mann, & Barratt, 1999). For example, in some
domains, the sex differences occur less regularly when fathers are involved, since
fatherhood seems to enhance men’s interest in infants. Indeed, for face recognition of

their newborn infant, fathers may be even more accurate than mothers (Kaitz, Good,

Rokem, & Eidelman, 1988).

Hormones

Hormones have also been examined to understand the psychobiological
underpinnings of human infant care, and, by implication, the adult hormonal responses to

infants. In mammals, several hormones are involved in parental care, including oxytocin,

15

arginine-vasopresin, prolactin, estrogen, and progesterone (Insel & Young, 2001). Such
endocrine changes regulate pregnancy, birth, and lactation as well as parental care (Flinn,
Ward, & Noone, 2005). The specific role of these hormones, however, varies across
species, sexes, and the organism’s experience and context (Fleming & Gonzalez, 2009;
Flinn et al., 2005). Given that some hormones are involved in parental behavior, it is not
surprising that some appear to affect interest in infants and the salience of infant cues.

In humans, most studies have assessed the role of hormones in parental behavior
during pregnancy, birth, and the postpartum period, mainly because endocrine changes
related to infant care are most evident during those reproductive phases. The evidence
shows that over this period, some infant cues become more salient for caregivers, and that
these changes sometimes are linked to hormonal changes, although the effects are
complex and dependent on many factors, such as sex, experience, parity (prior maternal
experience), age, and early life experiences (Fleming & Gonzalez, 2009).

Despite the substantial individual differences among mothers in the pattern of
parental feelings and in behavioral and hormonal responses, studies Show that throughout
pregnancy and from pregnancy to postpartum, maternal responsiveness increases
(Fleming, Ruble, Krieger, & Wong, 1997). Mothers, for instance, are able to recognize
their infants through multiple sensory cues, even after little experience with them. For
example, even if fathers do outperform mothers in the visual face recognition of their
newborns (Kaitz et al., 1988), mothers can recognize their newborn through a variety of
modalities, from seeing their photographs (Kaitz et al., 1988), hearing their cries (Green
& Gustafson, 1983), touching their skin and hand (Kaitz, Lapidot, Bronner, & Eidelman,

1992), and smelling their odor (Kaitz, Good, Rokem, & Eidelman, 1987; Porter, Cemoch,

l6

& McLaughlin, 1983; Russell, Mendelson, & Peeke, 1983). Mothers also give higher
hedonic ratings to infant odors than nonmothers and nonfathers (Fleming et al., 1993);
fathers (Russell et al., 1983) and mothers with higher levels of postpartum cortisol also
show more attraction toward and better recognition of their infant’s odors (Fleming,
Steiner, & Corter, 1997).

Changes in autonomic function that are correlated with hormonal changes also are
affected by multiple variables, such as parity and age. For instance, new mothers,
compared to nonpostpartum women, show greater empathy for infant cries; more
empathic mothers also show increased heart rate and elevated levels of cortisol (Stallings,
Fleming, Corter, Worthman, & Steiner, 2001). These are results for adults. In teen-age
mothers, the changes are minimal (Giardino, Gonzalez, Steiner, & Fleming, 2008).

The association between cortisol and maternal behavior appears to be curvilinear;
in other words, meditun levels of cortisol seem to be positive, low and high levels
detrimental, for maternal behavior. Any such effects also are evident only at early
postpartum stages (Fleming & Gonzalez, 2009). The involvement of cortisol, however,
suggests a potential functional role for cortisol because cortisol is activated during high
arousal states and has been shown to enhance attention, perception, memory, and
emotional processing (Erickson, Drevets, & Schulkin, 2003).

In males, married men with children tend to have lower levels of testosterone than
single men and married men without children (Berg & Wynne-Edwards, 2001; Gray,
Parkin, & Samms-Vaughan, 2007; Muller, Marlowe, Bugumba, & Ellison, 2009; Storey,

Walsh, Quinton, & Wynne-Edwards, 2000).

17

Neural Mechanisms

Our understanding of the neuroanatomy of parental behavior has been advanced
fundamentally through studies of rodents, sheep, and nonhuman primates, and, most
recently for humans, through neurofunctional imaging. The literature on nonhumans
highlights the critical role of basal forebrain structures, such as the media preoptic area
(MPOA), the ventral bed nucleus of the stria terrninalis (vBNST) and its projections into
the midbrain (ventral tegmental area [VTA]) and hindbrain (periaqueducal gray [PAG]),
and sensory, Iimbic, and cortical system that project into MPOA/vBNST (Fleming &
Gonzalez, 2009; Numan, 1994; Swain, Lorberbaum, Kose, & Strathearn, 2007). What the
literature shows is that lesions in these regions and projections induce disruptions in
maternal behavior.

Neuroimaging studies on humans are still in their formative stages, since the
number of studies and samples are small, have used a mix of methods, stimuli (e.g.,
infant vocalizations, photographs of infant faces, videos of infants, depictions of a variety
of emotional valences, and stimuli from the subjects’ own or someone else’s infant), and
subjects (e. g., parents, non-parents, only women). However, although activation patterns
vary depending on the task, methods, and population sampled, the literature shows that
some structures are recurrently involved when people assess infant stimuli. The structures
include the cingulate gyrus (Bartels & Zeki, 2004; Glocker, Langleben, Ruparel,
Loughead, Valdez, et al., 2009; Lorberbaum et al., 2002; Swain et al., 2007), thalamus
(Leibenluft, Gobbini, Harrison, & Haxby, 2004), insula (Glocker, Langleben, Ruparel,
Loughead, Valdez, et al., 2009; Seifritz et al., 2003), orbitofrontal cortex (Glocker,

Langleben, Ruparel, Loughead, Valdez, et al., 2009; Kringelbach et al., 2008; Nitschke et

18

al., 2004), amygdala (Leibenluft et al., 2004; Ranote et al., 2004; Zebrowitz, Luevano,
Bronstad, & Aharon, 2009), precuneus (Glocker, Langleben, Ruparel, Loughead, Valdez,
etal., 2009; Ranote et al., 2004), and fusiform gyrus (Bartels & Zeki, 2004; Kringelbach
et al., 2008; Ranote et al., 2004; Zebrowitz et al., 2009). These structures are found to be
involved, among other functions, in reward, arousal and motivation, planning and social
empathy/altruism, and attention (Fleming & Gonzalez, 2009; Glocker, Langleben,
Ruparel, Loughead, Valdez, et al., 2009; Swain et al., 2007).

Some studies also have found individual differences in brain activity, depending
on sex and parental status (Seifritz et al., 2003), and whether the infant stimuli (face or
vocalization) belong to one’s own infant or someone else’s.

Since the structures involved also have been found to vary depending on the
specific tasks and stimuli used, it may be appropriate to frame the results in terms of a
distributed neural network akin to recent models of face perception (Gobbini & Haxby,
2007; Haxby, Hoffman, & Gobbini, 2000), where structures are co-dedicated to visual
analysis (dynamic and static facial features for recognition), the representation of person
knowledge about the face (e. g., episodic memories, attitudes, mental states), and the

emotional response that the face elicits.

Summary

In sum, comparative and human data reveal a number of factors that affect adults’
interest in infants, including but not limited to the adult’s sex, age, parental status,
hormonal status, relationship to the infant, and mating and reproductive strategies (e. g.,
whether interested in having children in the short-term vs. long-term or simply not

interested). Thus, although in the 7 new studies to be described here, sex differences are

19

expected, with women typically being more interested in infants than men, the possibility
of finding considerable variability within each sex, perhaps even more in males than in
females, should not be discounted. When interpreting the data, we therefore must keep in
mind that the convenience sample used (nulliparous undergraduate students) is not
representative of all the reproductive stages in humans. The results could be different if

parenthood and other variables were to have been included.

20

Chapter 3: Psychosocial Measures of Individual Differences in Interest in Infants

This chapter describes the first two studies of the dissertation. Each was designed
to assess the content and psychometric properties of the three questionnaires used in
Studies 6 and 7 to quantify factors that could contribute to individual differences on the
visual attention tests. The three questionnaires, Interest in Infants, Job Preference, and
Interaction with Infants, were designed to measure how subjects perceive their interest in

infants and their willingness to interact with them.

Studies 1 and 2

Studies 1 and 2 were designed to test the reliability and to estimate the validity of
the new instruments. In Study 1, performed on-line, subjects completed the Interest in
Infants, the Job Preference questionnaire, and the Desire for Parenthood questionnaire,
which was included to assess the validity of the interest in infants questionnaires. In

Study 2, performed in the laboratory, subjects completed the Interaction with Infants and
Job Preference questionnaires].

Subjects for Study 1 were 217 women (Mean age = 19.49, SD = 1.58 years) and
112 men (M = 19.94, SD = 2.26 years) Michigan State University undergraduates. The

sample included 303 Whites (215 women), 32 Black, or African-Americans (26 women),

8 Hispanic or Latinos (6 women), 30 Asians (18 women), and 9 subjects with “Other”

 

1 Statistical analyses reported in this dissertation were conducted with R (R Development Core Team,
2009). Descriptive statistics and reliabilities were computed using the psych package (Revelle, 2009).
Signal detection theory scores were computed with the sdtalt package (Wright, 2009; Wright, Horry, &
Skagerberg, 2009). Multilevel models were calculated using the [me-t package (Bates & Maechler, 2009),
and the regression coefficients’ p—values based on the t-statistic were estimated with the IanguageR
package (Baayen, 2009), which reports the values by implementing a Markov chain Monte Carlo sampling
(10,000 by default). Mixed-models ANOVAs were computed with IBM SPSS Statistics (version I801).

21

ethnic background (6 women). One woman and one man each reported having 1 child;
the remaining subjects were childless.

Subjects in Study 2 were 63 women (Mean age = 19.08, SD = 1.22 years) and 27
men (M = 19.93, SD = 1.27 years) Michigan State University undergraduates. The
sample included 72 Whites (49 women), 6 Blacks, or African- Americans (5 women), 1
Hispanic or Latino man, 4 Asian women, 1 American Indian or Alaska Native woman,

and 6 subjects with “Other” ethnic background (4 women). All subjects were childless.

Interest in Infants

The Interest in Infants questionnaire consists of 21 items about infants, 11 of
which asked how intellectually interesting they are (e. g., “it is fascinating to try to
communicate with babies”) and 10 that asked how physically attractive they are (e.g.,
“only parents find their babies beautiful”). Despite this two-dimensional
conceptualization, a confirmatory factor analysis did not show the expected factor
structure for both men and women; instead the two extracted factors corresponded to
non-reversed-scored and reversed-scored items respectively; therefore, the questionnaire
was treated as one-dimensional. The questionnaire originally had 22 items, but I
intellectual-interest item (“babies are pretty easy to figure out”) was removed because it
showed poor association with the other items (i.e., correlation M = -.03, SD = .06).

To minimize the effect of the infant’s age on perceived attractiveness
(Hildebrandt & Fitzgerald, 1979b), subjects were told that the questionnaire items
referred to infants of about 6 months of age. The 6 months age was chosen as a midpoint
between newborns, many of whom are not particularly attractive, and l-year-olds, who

often are at peak attractiveness. Subjects rated how much they agree with each statement

22

on a 7-point scale (from 1: “strongly disagree” to 7: “strongly agree”). The
questionnaire was presented on a computer, and subjects responded at their own pace.
The questionnaire showed good internal consistency (women: or = .92; men: or =

.9). A one-way between-subjects ANOVA showed a main effect of sex, F (1,381) =

54.89, p < .001, npz = 0.13, indicating that women had significantly higher scores (M =

5.33, SD = 0.83) than men (M: 4.63, SD = 0.88).

Job Preference Questionnaire

The Job Preference questionnaire assessed the subjects’ preference for an infant-
care job compared to their average preference for 6 other jobs at similar wages (e. g., child
care, house keeper, waiter/waitress), according to the Occupational Employment
Statistics of the US Bureau of Labor Statistics (2008).

Subjects were told that all jobs required the same time and effort and were asked
to rate their preference for each job on a seven-point scale (from 1= “strongly dislike” to
7 = “strongly like”).

As part of the same online study, subjects who answered the Interests in Infants
questionnaire also answered the Job Preference questionnaire.

Preference ratings were examined with a mixed-design ANOVA with job type
(infant care, other jobs) as a within-subjects factor and sex of the subject (male, female)
as a between-subjects factor. The analysis showed that job type was not significant,

F (1,381) = 3.47, p = .06, indicating that when men’s and women’s preference for the
infant care job and other jobs were combined, their preferences were similar. The overall

similarity, however, can be attributed mostly to their similar preferences for non-infant-

23

carejobs (Men: M= 3.99, SD = 0.82; Women: M= 4.12, SD = 0.86) because a

2
significant interaction betweenjob type and sex, F(l,381)= 61.10, p < .001, up = 0.13,

indicated that women had higher preference for the infant-care job (M = 5.08, SD = 1.78)

than men (M= 3.4, SD = 1.88).

Desire for Parenthood

The Desire for Parenthood questionnaire was based on the Desire to Have
Children questionnaire (Rholes, Simpson, Blakely, Lanigan, & Allen, 1997). Because the
latter, provided by Rholes (personal communication), was written for married couples, it
was modified for administration to unmarried subjects. The new questionnaire included 8
items from the original questionnaire about the importance of having children (e. g.,
“Without children, I would feel unfulfilled”), and 1 new item, which compared the
importance subjects’ give to having children over having a stable partner. Subjects rated
each statement on a 7-point scale (from 1 = “strongly disagree” to 7 = “strongly agree”).

The questionnaire showed good reliability (women: or = .86; men: or = .77). A

one—way between-subjects ANOVA showed a small effect for sex, F (1,381) = 3.89, p =

.049, “p2 = 0.01, indicating that women had slightly higher scores (M = 4.75, SD = 1.19)

than men (M= 4.5, SD = 0.97).

Validity of Interest in Infants and Job Preference questionnaires

The validity of the questionnaires was assessed by examining the correlations
across questionnaires. Higher positive correlations were expected between the Interest in

Infants, Job Preference, and Desire for Parenthood questionnaires; lower correlations

24

were expected between these measures and the average preference for non-infant-care
jobs (since non-infant—care jobs included jobs that involved “care,” i.e., adult— and child-
care jobs. As summarized in Table 1, the correlation matrix confirmed these expectations,
with at least some of these relations (e. g., the desire for parenthood and the preference for

an infant-care job) being stronger in women.

Table l. Zero-Order Correlations Across Questionnaires by Sex of Subjects

 

 

 

 

Women (N =27l) interest claiajldb 33::
infant-care job .53 ***

Otherjobs .18 ** .28 ***

Desire for parenthood .62 *** .47 *** .23 ***
Men (N=112) interest $331; ml“
infant-care job .53 ***

Otherjobs .29 ** .28 an:

Desire for parenthood .34 *** .25 ** -.003

 

"“p < .05; **p < .01; ***p < .001

Interaction with Infants

The Interaction with Infants questionnaire was modeled after one designed by
Maestripieri and Pelka (2002). The original questionnaire was modified in order to find
out whether interest varies as a function of the infant’s emotional state, that is, whether
individual differences are more evident when the subject is asked to interact with an
infant in distress as opposed to an infant in a positive mood. Three emotional states were
described: negative (crying), neutral (lying quietly), or positive (happy). The following
scenario was presented:

“Several of your friends and some people you do not know are getting together

and have invited you to join them. As you enter the room where they have gathered, you

25

notice that in the corner of the room there is a baby lying in a car seat. The baby appears
to be about 6 months old. After you enter the room, how likely is it that you would do
each of the following?”

Three different Situations were described (e. g., “if the baby is lying quietly, I
would”), each followed by 6 descriptions of possible behaviors that subjects may engage
in as a reaction to the infant’s presence (e.g., “ignore the baby”, “go over and look at the
baby”). Subjects were asked to indicate, on a 7-point scale (from l=“extremely unlikely”
to 7: “extremely likely”), how likely they are to respond in these ways, with higher
scores indicating more positive behaviors towards the infant.

Subjects answered the questionnaire, at their own pace, on a computer in a
laboratory setting.

After deleting one item from the Negative emotion subscale because of its low
and negative correlation with other items (correlation M = .14, SD = .15), the total
questionnaire internal consistency was good (women: or = .93; men: or = .90) as was the
reliability for each of the subscales (Negative: women: or = .73, men: or =.71; Neutral:
women: or = .90, men: or = .88; Positive: women: or = .90, men: or = .83).

Ratings were examined with a mixed-design ANOVA with emotional state of the

infant (negative, neutral, positive) as a within-subjects factor and sex of the subject (male,

female) as a between-subjects factor. The analysis showed a main effect for emotional

state, F (2,176) = 1.06, p < .001, npz = 0.11, indicating that subjects reported more

positive responses as the infant’s emotional state became more positive (see Table 2), a

pattern not affected by the sex of the subject, F (2,176) = 0.22, p = .8. Although women’s

26

scores were higher than men’s, the main effect for sex was not significant, F (1,88) =

2.35, p = .13, perhaps due to a smaller sample size for men.

Table 2. Means and (Standard Deviations) of Positive Behaviors of Men and
Women for Interacting with an Infant Under Three Different Emotional States

 

 

Infant’s emotional Men Women
state (N=27) (N=63)
Negative 4.88 (0.92) 5.13 (1.01)
Neutral 5.07 (1.14) 5.43 (1.33)
Positive 5.31 (0.82) 5.70 (1.01)

 

Subjects also answered the Job Preference questionnaire, so that its relationship
with infant-care jobs and non-infant care jobs could be assessed. If the measure is valid, it
was expected to show stronger correlations with infant-care jobs than non-infant-care
jobs.

A mixed-design ANOVA with job type (infant care, other jobs) as 3 within-
subjects factor and sex of the subject (male, female) as a between-subjects factor showed
no main effect for job type, F (1,88) = 1.15, p = .29, indicating that for the combined
scores for men and women, subjects did not Show a preference for either type of jobs.

However, this was mostly due to the men’s and women’s similar preferences for non-

infant-carejobs (Men: M= 4.02, SD = 0.77; Women: M= 3.95, SD = 0.86) as shown in
the significant interaction between job type and sex, F(1,88) = 11.86, p = .001, npz =

0.12, indicating that women preferred infant-care jobs (M = 4.90, SD = 1.99) while men

prefered non-infant care jobs (infant-care job preference: M = 3.52, SD = 1.67). There

was a small main effect of sex, F(1,88) = 6.20, p = .015. ”p2 = 0.07, indicating that,

27

overall, women gave higher ratings than men (largely due to their higher preference for
infant care jobs). The results thus replicate those obtained with the larger sample of the
online study.

The overall pattern of correlations, summarized in Table 3, confirmed the
expectation that the Interaction with Infants questionnaire would correlate more strongly
with the infant-care job than with non-infant-care jobs for both men and women.
Subscales of the Interaction with Infants questionnaire followed the same pattern in

women, but less clearly in men.

Table 3. Zero-Order Correlations Between the Job Preference and Interaction with
Infants Questionnaires

 

Women (N=66) Infant care other jobs Negative Neutral Positive

 

 

 

Other jobs .27 **

Negative .58 *** .05

Neutral .56 *** .13 .61 ***

Positive .51 *** .03 .67 *** .79 ***

Total .62 *** .08 .84 *** .92 *** .92 ***
Men (N=27) Infant care other jobs Negative Neutral Positive
Other jobs .41 *

Negative .36 .19

Neutral .53 ** .35 .64 ***

Positive .35 -.13 .40 * .56 **

Total 51 ** .20 .82 *** .91 *** .75 ***

 

*p < .05; "p < .01; ***p < .001

Revised Sociosexual Orientation Inventory

Sociosexuality is generally understood as the willingness to engage in
uncommitted sex and, therefore, provides an indicator of mating tactics and thus a way to

assess how individual differences in mating tactics relate to interest in infants (e. g.,

28

people with higher willingness to engage in uncommitted sexual relationships may Show
lower interest in parental investment and therefore lower interest in infants). Sociosexual
orientation was assessed with the Revised Sociosexual Orientation Inventory (SOI-R)
(Penke & Asendorpf, 2008). Based on the original inventory (Simpson & Gangestad,
1991), it includes three subscales (with three items each) for assessing past sociosexual
behavior, sociosexual attitudes, and sociosexual desire. The inventory has good
psychometric properties, including construct and predictive validity (see Penke &

Asendorpf, 2008 for details).

Self-Perceived Mate Value Scale

The Self-Perceived Mate Value Scale (Langton, Honeyman, & Tessler, 2004)
consists of 8 items that ask subjects to rate on a 7-point scale (from 1 = “strongly
disagree” to 7 = “strongly agree”) the reactions they usually receive from the opposite sex
in terms of their mate value (e. g., “Members of the opposite sex are attracted to me”).

The scale shows good reliability (e.g., men: or = .91, M: 3.23, SD = 1.01; women: or =

.93, M = 3.84, SD = 1.13) (Penke & Asendorpf, 2008) and provides an indicator of

information that people might be using to adjust their mating tactics. For instance, men
who perceive themselves as having higher mate value tend to have higher scores in the
SOI (Clark, 2006), but this is not the case for women (Clark, 2006; Mikach & Bailey,
1999). This information therefore could help to explain the relationship, if any, between

sociosexuality and interest in infants.

29

Chapter 4: Composition and Characteristics of Facial Stimuli

This chapter describes the face database used in Studies 6 and 7. It explains how
the faces to be used in these studies were obtained, the criteria used for their selection, the
computer graphic techniques that were used to make them suitable for use, and then the 3
studies (3-5) that assessed whether the faces retained their age, sex, and affective

characteristics after being standardized.

Facial Stimuli

The face images were frontal views of the faces of young adults and infants, all
Whites with neutral expressions. The young adult faces (24 men, 24 women) were
selected from the Productive Aging Lab Face Database (Minear & Park, 2004) and the
FERET database (Phillips, Moon, Rizvi, & Rauss, 2000). The infant faces (24 boys. 24
girls, 3 to 6 months old) were taken from the website F lickr (wwwflickrcom). The age
of the adult faces was determined from information in the databases. The age of the infant
faces was determined from information in the website (for some, the age was specifically
given; for others, it was estimated by comparing the date when the picture was taken with
the date of pictures of the infant’s birth). In the few cases where information about an
infant’s or adult’s face was unavailable, an independent rater confirmed that these faces
were similar in age to other faces in the pool. Faces also were selected so that, within
each age and sex category, they showed similar levels of masculinity/femininity and
attractiveness (independent raters, as described below, rated the faces on these and other

dimensions).

30

The adult faces appeared without glasses or jewelry, and, for male faces, without
substantial facial hair. Non-facial salient marks that may affect recognition, such as moles
and blemishes, were removed using the clone stamp tool in Adobe Photoshop 9.0.2. All
color images were transformed to grey-scale.

Faces were standardized to have identical orientation and similar inter—pupil
distance, and only the face outline was visible (hair, ears, and original background were
removed). Although inter-pupilar distances were standardized across faces, natural age-
and sex-related craniofacial morphological variations (e.g., Ferrario, Sforza, Poggio, &
Schmitz, 1998; Ursi, Trotman, McNamara, & Behrents, 1993; Weston, Friday, & Li(‘),
2007) led to Size variations in the faces. Average area measures, calculated with ImageJ

1.42q (Rasband, n.d.), are summarized in Table 4.

Table 4. Pixel Area of Each Face Category

 

 

 

Area
Face category N M (SD)
Boys 24 24733.17 (1236.61)
Girls 24 24436.88 (1340.20)
Men 24 31429.29 (913.73)
Women 24 24857.7] (1185.56)

 

A one-way ANOVA was used to examine pixel area group differences and

showed that the groups differed, F(3,92) = 197.20, p < .001, ”p2 = .865. Tukey HSD

post-hoe comparisons showed that the difference was due to men’s faces having a larger
area than boys’, girls’, and women’s faces (all p values < .005), with the scores for the

last three groups not being significantly different from each other (p > .05).

31

Consequently, area measures will be included in models attempting to fit the behavioral
data of Studies 6 and 7.

. To minimize the effect of image statistics in the results for Studies 6 and 7, the
luminance and relative contrast of faces were also adjusted to approximate the average
luminance-contrast of adult faces. Adjustments were made by creating an average face of
the adult faces and applying its image statistics to each face in the dataset. Any non-face
area of the image region (188 x 250 pixels) was filled with a grey background (RGB:
192,192,192)

Given that humans are more sensitive to the contrast of middle spatial
frequencies, the faces also were filtered to extract 8 spatial frequency bands (< 4, 4-8, 8-
16, 16-32, 32-64, 64-128, 128-256, >256 pixels/cycle band) in order to examine their
relative contribution to the behavioral results of Studies 6 and 7. The subjective/apparent
contrast was calculated by dividing the standard deviation of the luminance by the mean
luminance of each filtered image (Delplanque, N'diaye, Scherer, & Grandjean, 2007).
Filtered images were produced and measurements were made with ImageJ 1.42q.
Measurements are summarized in Table 5.

A MANOVA showed that the apparent contrast varied across groups at all
frequency bands (including the unfiltered image). The results are summarized in Table 6.
Post hoc Tukey HSD tests showed that at lower frequency bands (less than 64
pixels/cycle), adult faces, and especially male faces, have greater apparent contrast than
infant faces, and that the reverse is true, especially for boy faces, at higher frequency

bands (64 and higher pixels/cycle).

32

Table 5. Means and (Standard Deviations) of Apparent Contrast for Each
Frequency Band by Sex of the Face

 

Frequency Bands

 

. Boys Girls Men Women

(pixels/cycle)

Unfiltered 0.213 (0.019) 0.215 (0.017) 0.225 (0.007) 0.217 (0.008)
< 4 0.042 (0.007) 0.045 (0.007) 0.053 (0.005) 0.053 (0.004)
4 — 8 0.037 (0.004) 0.039 (0.005) 0.047 (0.004) 0.046 (0.003)
8 - 16 0.052 (0.005) 0.055 (0.006) 0.064 (0.003) 0.060 (0.003)
16 - 32 0.064 (0.005) 0.068 (0.005) 0.075 (0.004) 0.069 (0.004)
32 - 64 0.060 (0.007) 0.060 (0.005) 0.068 (0.006) 0.063 (0.005)
64 - 128 0.053 (0.015) 0.045 (0.018) 0.037 (0.009) 0.040 (0.009)
128 - 256 0.032 (0.008) 0.030 (0.008) 0.022 (0.005) 0.025 (0.004)
> 256 0.008 (0.003) 0.007 (0.004) 0.006 (0.002) 0.005 (0.002)

 

Table 6. F-statistics, Effect Sizes, and Significant Group Differences for
Unfiltered and Filtered Frequency Bands

 

 

Frequency Bands F 71p2 Significant group

(pixels/cycle) differences
TukeyI HSD

Unfiltered 3.476 * 0.10 m > b

< 4 24,079 ** 0.44 m,w > b,g

4 - 8 33.576 ** 0.52 m,w > b,g

8 - 16 30.160 ** 0.50 m,w > b,g

l6 - 32 23.951 ** 0.44 m > g,w> b

32 - 64 9.250 ** 0.23 m > b,g,w

64 - 128 6.957 ** 0.19 b > m,w

128 - 256 10.894 ** 0.26 b,g > m; b > w

> 256 4.397 * 0.13 b > m

 

* p < .05, ** p < .001
b=boys, g=girls, m=men, w=women

It is important to keep in mind that although image standardization is commonly
used in visual perception studies to unconfound the effect of low-level visual processing,

a potential problem is that making the image-set more homogenous could reduce the

33

number of cues available for discriminating sex and age (e.g., color, which normally is
darker for adult male faces). If so, image standardization of, for instance, infant images
could diminish their potential for eliciting emotional responses by reducing the salience

of one or another of the characteristic infant features.

Face Ratings

Along with computer graphic processing and image statistics, faces were rated by
independent judges on dimensions known to affect face recognition and attention, such as
attractiveness, distinctiveness, masculinity/femininity, arousal, valence, and dominance
(e.g., Lang, Bradley, & Cuthbert, 2008). Some of these ratings were collected to confirm
that faces retained their age, sex, and affective characteristics after standardization; other
ratings were collected to confirm that faces elicited some of the known effects in the face

recognition task, such as distinctiveness (see Chapter 5).

Study 3

Valence, Arousal, and Dominance Ratings

The purpose of Study 3 was to collect valence, arousal and dominance ratings.
Valence is understood as the dimension that distinguishes positive (pleasant) from
negative (unpleasant) emotional states, arousal distinguishes highly exciting states from
relaxed states, and dominance distinguishes situations of being in control from situations
of being dominated.

Although valence and arousal have received more study, dominance ratings were

included, since they are part of the rating system of the International Affective Pictorial

34

System or IAPS (Lang et al., 2008), one of the most widely used sets of pictorial stimuli

in studies that examine how emotion affects cognition (e.g., memory).

Subjects

The subjects, 100 women (mean age = 20.31, SD = 1.57 years) and 40 men (M =
20.9, SD = 2.27 years), all Michigan State University undergraduates, participated for
course credit. The sample included 107 Whites (74 women), 19 Blacks or African
Americans (14 women), 2 Hispanic or Latino women, 9 Asians (7 women), 1 Native

Hawaiian or Other Pacific Islander woman, and 2 women with “Other” ethnicity. Three

women and one man reported having 1 child; the rest of the subjects reported being

childless.

Procedure

Ratings were collected through an online study that used the Self-Assessment
Manikin or SAM (Lang, 1980). It consists of three series of figures that illustrate the
intensity of the emotional dimensions valence, arousal, and dominance. The valence
dimension is measured with figures that range from happy (9 = “positive”) to unhappy (1
= “negative”) as shown by facial expression; in arousal, from excitement (9 = “high
arousal”) to relaxed (I = “low arousal”), as shown by open vs. closed eyes and by
“activity” in the mid-body; and in dominance, from 1 = “dominated” to 9 = “in control,”

as shown by overall size. Examples of each dimension are shown in Figure 1.

35

mew:

cfifliifl

Figure 1. The Self-Assessment Manikin (SAM) Used to Acquire Ratings of (A)
Valence, (B) Arousal, and (C) Dominance (Lang et al., 2008). Images were
obtained from PXlab website (Irtel, 2007) [Note: Dominance images are not
identical to the original scale].

Subjects were first told how to use the SAM scales followed by a surprise quiz to
ensure that they understood (only those who answered the quiz correctly were allowed to
continue; otherwise, they were given the instructions again and then re-tested).
Subsequently, on each trial, subjects were first shown the message, “Get ready to rate the
next face” (presented for 150 ms), which then was replaced by a face (presented for 5000
ms), and then was followed by the SAM scales (presented until the subject responded).

To minimize fatigue effects, the total pool of 96 faces was pseudo-randomly
divided into two groups, each group containing the same number of faces from each sex-

age category (i.e., 12 boys, 12 girls, 12 men, 12 women), and subjects were randomly

36

assigned to rate one of these two groups. Faces were presented in random order to each
subject.

Because all faces displayed a neutral facial expression, it was not expected that
they would elicit high arousal or valence ratings; it was expected, however, that infant
faces would be more emotionally salient than adult faces and therefore would receive

higher arousal and valence ratings (e.g., Brosch, Sander, & Scherer, 2007).

Results

Scores from 7 subjects (6 women) were removed because they failed to provide

ratings on most faces. Scores for the remaining subjects are summarized in Table 7.

Table 7. Means and (Standard Deviation) for Valence, Arousal, and Dominance
Ratings for Each Face Category by the Sex of the Subject

 

 

 

Face Valence Arousal Dominance
Category Men Women Men Women Men Women
Boys 5.37 5.20 5.20 5.31 5.13 4.77
(0.70) (0.47) (1.19) (1.17) (0.65) (0.52)
Girls 5.30 5.33 4.77 4.89 4.63 4.36
(0.41) (0.41) (0.68) (0.89) (0.73) (0.53)
Men 3.97 4.55 4.92 4.85 3.96 3.79
(0.51) (0.47) (0.71) (0.71) (0.57) (0.47)
Women 4.22 4.58 5.19 4.98 4.29 3.81
(0.33) Q32 (0.71) (0.69L (0.47) (0.40)

 

Valence ratings were examined with a mixed-design AN OVA with sex of the
subject (male, female) as a within-subjects factor and sex (male, female) and age (infant,

adult) of the face as between-subj ects factors. The analysis showed a main effect for age

of the face, F (1,92) = 136.30, p < .001, "p2 = .60, indicating that infant faces received

higher valence ratings (M = 5.30, SD = 0.45) than adult faces (M = 4.33, SD = 0.36). A

37

. . 2
main effect for sex of the subject, F(1,92) =1 8.1 l, p < .001, np = .16, showed that

women’s ratings were higher (M = 4.92, SD = 0.55) than men’s (M= 4.71, SD = 0.80).
This effect was qualified by a significant interaction between sex of the subject and the

age of the face, F (1 ,92) = 31.92, p < .001, npz = .26, showing that although women’s and

men’s valence ratings for infant faces were similar, for adult faces men’s ratings were
lower than women’s. All other main effects and interactions were not significant (all p
values > .05).

Similarly, arousal ratings were examined with a mixed-design ANOVA with sex
of the subject (male, female) as a within-subjects factor and sex (male, female) and age
(infant, adult) of the face as between-subjects factors. The analysis showed that men’s
and women’s ratings across face categories were not significantly different from each
other, F (1,92) = 0.87, p = .77. However, as indicated by a small but significant

interaction between sex of the subject and age of the face, F (1,92) = 8.11, p = .005,

. 2 . . .
partial up = .08, women gave higher arousal ratings to Infant faces than to adult faces

and men did the reverse. All other main effects and interactions were not significant (all p
values > .05).

Finally, dominance ratings were examined with a mixed-design ANOVA with sex
of the subject (male, female) as a within-subjects factor and sex (male, female) and age

(infant, adult) of the face as between-subjects factors. The analysis showed a main effect

for sex of the subject, F (1,92) = 39.23, p < .001, “p2 = .3, indicating that, across faces,

men felt more dominant (M= 4.5, SD = 0.75) than women (M= 4.18, SD = 0.63). There

was also a main effect for age of the face, F (1,92) = 56.98, p < .001, “p2 = .38, with

38

subjects feeling more dominant when judging infant faces (M = 4.72, SD = 0.59) than

adult faces (M = 3.96, SD = 0.43). There also was a small interaction between sex and
age of the face, F (l ,92) = 9.89. p = .002, “p2 = .09, indicating that subjects felt more

dominant when judging boys’ faces than girls’ faces, whereas for adult faces, subjects felt
slightly more dominant when judging women’s faces than men’s faces. All other main
effects and interactions were not significant (all p values > .05).

It is difficult to make a meaningful comparison between the results of these
ratings and the normative ratings on the AIPS because the images in the AIPS and the
faces used in this dissertation were not standardized in the same way and because not all
face categories are represented in the IAPS (e. g., there are no suitable images of White
women with neutral facial expression). In addition, subjects rating IAPS images are
exposed to a much wider variety of emotional content (e. g., images of mutilated bodies
and erotic images). Nonetheless, a sample of images from the IAPS was selected to
compare with some of the ratings. Images were of 9 infants with relative neutral
expression and 4 images of men with neutral expression. The normative values given to
those images by men and women in the IAPS were extracted, and 3 mixed-design
ANOVAs were conducted to compare the ratings for valence, arousal, and dominance.

Table 8 summarizes the results.

Table 8. Means and (Standard Deviations) of Valence, Arousal, and Dominance
Ratings by Face Category and Sex of the Subject

 

 

 

 

Valence Arousal Dominance
Face category Men Women Men Women Men Women
Infants (n = 9) 8.01 6.97 4.79 4.08 6.82 6.46
(0.62) (0.74) (0.55) (0.44) (0.51) (0.66)
4.71 5.63 2.65 4.10 5.84 5.59

Male (‘1 = 4) (0.17) (1.16) (0.37) (0.61) (0.59) (0.71)

 

39

The analysis showed that across faces, men and women had similar valence

ratings, F (l ,1 1) = 0.13, p = .73; however, a significant interaction between sex of subject

and age of the face, F(1,1 l) = 31.08, p < .001, “p2 =.74, showed that men’s ratings were

more dissimilar for infant and adult faces than women’s ratings. Valence ratings also

were higher for infant faces, F(1,l l) = 33.81,p < .001, npz = .76.

For arousal, women gave higher ratings than men, F(l,11) = 4.82, p = .051, "p2 =

.31. In addition, a significant interaction between sex of subject and age of the face,

,
F(l,11) = 40.73, p < .001, 11p“ = .79, showed that men’s ratings were more dissimilar for

infant and adult faces than women’s. Finally, arousal ratings were higher for infant faces,
F(1,11)=18.00,p=.001,np2 = .62.

For dominance, men and women gave similar ratings, F (1,1 1) = 1.78, p = .21;

they also reported feeling more dominant when looking at infants’ faces than men faces,
7
F(1,11)=10.76,p = .007, m," = .49.

Thus, there are some similarities in the overall pattern of ratings between the
IAPS and the standardized face stimuli used in this dissertation, with infant faces
receiving higher ratings across dimensions. This result is similar to the results of another
study (Brosch et al., 2007). However, in the current study (Study 3), although infant faces

received higher arousal ratings, the difference was not significant.

40

Study 4

Masculinity/femininity, Attractiveness ratings

Attractiveness can affect attention because attractive faces are looked at longer
than less attractive faces (Aharon et al., 2001; Landolt, Lalumiere, & Quinsey, 1995), and
more often induce covert shifts of attention (Sui & Liu, 2009). Because the masculinity-
femininity dimension is an important component of facial attractiveness (Rhodes, 2006),
in Study 4, ratings also were collected on attractiveness and masculinity-femininity of

faces.

Subjects

F orty-nine women (Mean age = 19.94, SD = 1.66 years) and 46 men (M = 20.37,
SD = 2.22 years), all Michigan State University undergraduates, participated for course
credit. The sample included 74 Whites (34 women), 7 Black or African American
women, 3 Hispanic or Latinos (1 woman), 8 Asians (5 women), and 1 American Indian
or Alaska Native man, 1 Native Hawaiian or Other Pacific Islander woman, 1 woman

who reported “Other” for ethnic background. All subjects were childless.

Procedure

Subjects rated the attractiveness of each face on a 7-point scale from 1 = “very
unattractive” to 7 = “very attractive”, and rated the femininity/masculinity on a 7—point
scale from 1 = “very feminine” to 7 = “very masculine”.

The procedure was identical to the one used for valence, arousal, and dominance
ratings, except that instead of the Self-Assessment Manikin scales, subjects saw and

answered, after each face, the attractiveness and masculinity scales.

41

Results

Average ratings are summarized in Table 9.

Table 9. Mean and (Standard Deviations) of Attractiveness and Masculinity
Ratings by Face Category and Sex of Subjects

 

 

 

Men Women
Face category Attractiveness Masculinity Attractiveness Masculinity
Boys 3.92 (0.48) 4.15 (0.43) 4.33 (0.71) 4.14 (0.62)
Girls 3.87 (0.52) 4.00 (0.46) 4.28 (0.72) 4.03 (0.47)
Men 3.57 (0.49) 5.95 (0.40) 3.89 (0.72) 6.11 (0.41)
Women 3.51 (0.70) 2.74 (0.59) 3.65 (0.80) 2.56 (0.62)

 

Attractiveness ratings were examined with a mixed-design ANOVA with sex of
subject (male, female) as a within-subjects factor and sex (male, female) and age (infant,

adult) of face as between-subjects factors. The analysis showed a main effect for sex of

subject, F (1,92) = 57.02. p < .001, ”p2 = .38, indicating that women gave higher
attractiveness ratings (M = 4.04, SD = 0.78) than men (M = 3.72, SD = 0.57). A small
interaction between sex of subject and age of face, F (1,92) = 4.68, p = .033, npz = .05,

indicates that men’s and women’s ratings were slightly more similar for adult faces (men:
M = 3.54, SD = 0.60; women: M = 3.77, SD = 0.76) than for infant faces (men: M = 3.90,

SD = 0.49; women: M= 4.31, SD = 0.71). A main effect of age of face, F(l,92) =12.50, p

= .001, npz = .12, showed that infant faces received higher ratings (M = 4.1, SD = 0.56)

than adult faces (M = 3.66, SD = 0.66). All other main effects and interactions were not

significant (all p values > .05).
Masculinity ratings were also analyzed with a mixed-design ANOVA with sex of

subject (male, female) as a within-subjects factor and sex (male, female) and age (infant,

42

adult) of face as between-subjects factors. The analysis showed that there was not a
significant main effect for sex of subject, F (1,92) =0.01, p = .92; that is, women and men

gave similar masculinity ratings in general. However, a small interaction between sex of
face and sex of subject, F (1,92) =5.44, p = .022, npz = .06, indicated that men gave
slightly higher ratings for female faces (men: M = 3.37, SD = 0.83; women: M = 3.30, SD

= 0.92) but did the reverse for male faces (men: M = 5.05, SD = 0.99; women: M = 5.13,

SD = 1.12). An additional small interaction between sex of subject and sex and age of
a

face, F(l,92) =8.47, p = .005, np" = .08, indicated that although men and women gave

similar ratings for boys’ and girls’ faces, for women’s faces, men gave slightly higher

masculinity ratings than women did, and for men’s faces, women gave slightly higher

ratings than men did (see Table 9).

A small effect of age of face, F(l,92) =7.06, p = .009, ”p2 = .07, showed that

adult faces received higher ratings than infant faces (adults: M = 4.34, SD = 1.78; infants:
M = 4.08, SD = 0.47). Most of the variance was accounted for, as expected, by the main
effect of sex of the face, F (1,92) =319.89, p < .001, “p2 = .78, with male faces receiving
higher masculinity ratings (M = 5.09, SD = 1.05) than female faces (M = 3.33, SD =
0.86), and by an interaction between sex of face and age of face, F (1,92) =273.83, p <
.001, "p2 = .75, with infant faces receiving similar masculinity ratings (boys: M = 4.14,
SD = 0.5; girls: M = 4.01, SD = 0.45) whereas adult faces did not (men: M = 6.03, SD =
0.38; women: M= 2.65, SD = 0.58).

In sum, although face standardization removed cues used in sex recognition (e. g.,

hair, color), the faces retained a sufficient number of features that identified them as male

43

or female, which allowed subjects to correctly perceive higher levels of masculinity for
male faces than for female faces. In addition, the analyses confirmed that male and
female faces did not differ significantly in attractiveness, thereby minimizing possible

attractiveness effects in how people respond to adult faces in Studies 6 and 7.

Study 5

Distinctiveness and sex identification

Because Study 6 includes a face recognition task and because distinctiveness
affects face recognition (e.g., Cohen & Carr, 1975; Going & Read, 1974; Valentine &
Bruce, 1986), in Study 5, the faces’ distinctiveness ratings were collected to control
statistically for this variable. According to face space models of face recognition
(Valentine, 1991), this effect is produced by our knowledge (experience) with a
population of faces. Individual faces are understood as points, normally distributed in a
Euclidean multidimensional space defined by various physiognomic features used to
encode faces (e.g., nose size, face shape, race). Typical faces are more commonly
encountered and closer to the central tendency of the normal distribution, while
distinctive faces are more distant and infrequent. A typical face therefore is harder to
remember because it is encoded in a space with a high density of points, making it harder
to distinguish from other points, whereas atypical faces are encoded in a region with a
lower density of points, and therefore can be recognized more accurately and faster. For
instance, caricatures, which exaggerate the distinctive features of faces, are more easily
recognized than the original faces (e.g., Benson & Perrett, 1991; Rhodes, Brennan, &

Carey, 1987).

44

In addition, although masculinity/feminity ratings are a close approximation of
how people perceive the sex of faces, the two judgments are not identical (e. g., Hoss,
Ramsey, Griffin, & Langlois, 2005), since male and female faces can each have different
levels of masculinity/femininity. Therefore, in Study 5, subjects also were asked to
classify the sex of the face to ensure that, for instance, a male face was perceived as a

man and not as a women with high levels of masculinity.

Subjects

Eighty-four women (mean age = 19.67, SD = 1.32 years) and 88 men (M= 19.49,

SD = 1.21 years), all undergraduate students at Michigan State University, participated
for course credit. The sample included 147 Whites (71 women), 16 Black or African
Americans (11 women), 7 Asians (1 woman), 1 American Indian or Alaska Native
woman, and 1 man who reported “Other” for ethnic background. Two of the women had

children.

Procedure

Ratings were collected following a similar procedure to the one used in the other-
face ratings, except that after each face was presented, a distinctiveness and a sex-
classification scale appeared.

The distinctiveness scale was adapted for the age of the face, Since, for instance,
an infant face could be more distinctive in a row of adults. For each adult face subjects
were asked, “How easy it is to spot this face in a crowd of adults?”, and for each infant
face, “How easy it is to spot this face in a crowd of babies?” Subjects responded on a 7-

point scale from 7= “very easy” to 1 = “very difficult”.

45

For sex classification, subjects were asked, “Is this face male or female?”, where

1= “male”, 0: “female”.

Results

Distinctiveness and sex scores for each face were averaged across subjects

according to the sex of the subject. Table 10 summarizes the results by each face

category.

Table 10. Mean (and Standard Deviation) of Distinctiveness Ratings and Sex
Classification Scores for Each Face Category by Sex of the Subject

 

 

 

 

Men Women
Face category Distinctiveness Sex Distinctiveness Sex
Boys 3.91 (0.36) 0.63 (0.21) 3.93 (0.43) 0.68 (0.24)
Girls 3.77 (0.48) 0.54 (0.17) 3.78 (0.50) 0.57 (0.23)
Men 4.68 (0.35) 0.97 (0.04) 4.52 (0.38) 0.97 (0.04)
Women 4.35 (0.43) 0.11 (0.15) 4.36 (0.40) 0.08 (0.14)

 

Distinctiveness and sex ratings were examined in separate mixed-design
ANOVAs, with sex of subject (male, female) as a repeated-subjects factor and sex of face
(male and female) and age of face (infant and adult) as between-subjects factors.

For distinctiveness ratings, there was a main effect of sex of face, F (1,92) = 6.14,

p = .015, "p2 = .06, showing that male faces in general received slightly higher

distinctiveness scores. There was also a main effect of age of face, F (1,92) = 62.08, p <

.001, “p2 = .4, showing that adult faces received higher distinctiveness ratings than infant

faces. Men’s and women’s ratings were not significantly different from each other, and

all other interactions were not significant.

46

For sex ratings, the analysis showed a main effect of sex of face, F (1,92) =

215.24, p < .001, npz = .70, indicating, as expected by how the scale was scored, that

male faces received higher sex scores than female faces. There also was a main effect of

age of face, F (1,92) = 5.15, p = .025, “p2 = .05, indicating that adult faces received lower

scores (M = 0.53, SD = 0.45) than infant faces (M = 0.61, SD = 0.21). There was no main
effect of sex of subject, F (1,92) = 1.63, p = .21, indicating that, in general, men and

women gave similar scores. The analysis also showed an interaction between sex of face

and age of face, F (1,92) = 135.96, p < .001, npz = .60, indicating that adult faces have

more dissimilar sex scores (men: M = 0.97, SD = 0.04; women: M = 0.09, SD = 0.14)
than infant faces (boys: M = 0.66, SD = 0.22; girls: M = 0.56, SD = 0.19). In addition,

there was an interaction between age of face and sex of subject, F (1,92) = 9.00, p = .003,

"p2 = .09, indicating that women gave slightly higher sex scores to infant faces (M =

0.63, SD = 0.24) than men (M = 0.59, SD = 0.19), but for adult faces, woman (M = 0.52,
SD = 0.46) and men (M = 0.54, SD= 0.45) gave Similar sex scores. No other interactions
were significant or approached significance.

Sex discriminability was calculated for each subject with A ’, a bias-free (i.e., it
controls for a possible tendency of guessing one sex over another when unsure) measure
of discriminability, like d’, except that it also can be computed when subjects have hits or
false alarms of 1 or 0. A ’ of 0.5 indicates chance performance and values closer to 1
indicate higher discriminability. A ’ was computed by arbitrarily defining “hits” as the
response “male” for male faces, and “false alarms” as the response “male” for female

faces. Average scores are summarized in Table 11.

47

Table 11. Means and (Standard Deviations) of A’ and B” Scores for Sex
Classification of Infant and Adult Faces by Sex of Subject

 

 

 

Men Women
Measure Infant faces Adult faces Infant faces Adult faces
(n=71) (n=72) (n=74) (n=74)
A’ 0.57 (0.16) 0.96 (0.04) 0.59 (0.15) 0.97 (0.04)
B” -0.06 (0.16) -0.43 (0.59) -0.12 (0.19) -0.36 (0.71)

 

Four two-tailed one-sample t-tests showed that both men and women had A’
scores significantly above chance, that is, higher than 0.5 (Men: infant faces, t(70) = 3.68,
p < .001, adult faces, I (71) = 90.02, p < .001; Women: infant faces, t(73) = 5.29, p <
.001, adult faces, I (73) = 110.51,p < .001).

A mixed design ANOVA with age of face (infant, adult) as a within-subjects

factor and sex of subject (male, female) as a between-subjects factor, showed a main

effect of age of face, F (1, 143) = 895.11, p < .001, npz = .86, indicating that subjects

were better at correctly discriminating the sex of adult faces than the sex of infant faces.
Neither sex of subject, F (1, 143) = 1.12, p = .292, nor its interaction with age of face,
F (1 , 143) = 0.18, p = .674, were significant, indicating that for both adult and infant
faces, women and men had similar sensitivities.

Response biases were calculated for each subject with B” and are summarized in
Table 11. B ’ ’ values range from -1 to 1, where 0 indicates no response bias (i.e., no more
likely to guess “male” or “female”) values greater than 0 indicate a conservative bias
(i.e., more willing to guess “female”) and values lower than 0 indicate a liberal bias (i.e.,
more willing to guess “male”). For adult faces, 45 subjects (25 women) had perfect

scores so their B ” were assumed to be 0.

48

Four one-sample two-tailed t-tests showed that both men and women had biases
to classify faces as males when uncertain, that is, their scores were significantly lower
than 0 (for men: infant faces, t(70) = 3.03, p = .003, adult faces, t(71) = 6.26, p < .001; for
women: infant faces, t(73) = 5.15,p < .001, adult faces, t(48) = 5.3, p < .001).

A mixed design ANOVA with age of face (infant, adult) as a within-subjects

factor and sex of subject (male and female) as a between-subjects factor showed a main

effect of age of face, F(1, 143) = 36.65, p < .001, npz = 0.20, indicating that subjects had

a stronger guessing bias for adult faces. Neither sex of subject, F(1, 143) = 0.006, p =
.937, nor its interaction with age of face, F (l, 143) = 1.48, p = .225, were significant,
indicating that for both adult and infant faces, women and men had similar guessing
biases. The stronger guessing bias for adult faces, however, must be interpreted with
caution because most subjects performed at ceiling for adult faces; any small false alarms
therefore will produce high B ’ ’ (e. g., subject 104 had a hit rate of .416 and a false alarm
of .583 for infant faces and a hit rate of 1 and false alarm of .083 for adult faces, which
led to a B ” close to 0 for infant faces but a value of -1.0 for adult faces). An alternative
way of dealing with hit rates of l or false alarms of 0 is to add a flattening constant of .5
and then compute C (a more commonly used measure of bias). This method showed
again significant biases to guess “male” for each face category: the bias, however, was
higher for infant faces than adult faces.

The analysis also showed that one woman’s face was problematic because it
tended to be classified as male (i.e., sex classification responses were coded as 1 for male
responses and 0 for female responses, and the mean sex score for this face was .66 for

men and .67 for women). It also received the highest masculinity scores among women.

49

In general, however, the ratings for faces drawn from the face-database are
comparable to several findings in the face research literature, namely, that subjects
reacted to our face stimuli in ways similar to how other subjects have reacted to other
face sets. For instance, the bias to respond “male” has also been found with adults’ and
children’s faces for both adult and child subjects (Cheng, O'Toole, & Abdi, 2001;
Rossion, 2002; Wild et al., 2000). The same has been found for infant faces (Hildebrandt
& Fitzgerald, 1977; Nagy, Ne’meth, & Molnar, 2000). Thus, even though several
dimensions of the faces were standardized (e. g., removal of hair) and a relatively narrow
range of variability was chosen (e. g., attractiveness and emotional expressions), the faces

conserved many of the properties of natural (unprocessed) faces.

50

Chapter 5: Study 6 -- Individual Differences in Visual Processing of Infant and

Adult Faces

Visual attention and eye-movements

Because visual acuity declines abruptly outside the fovea, our eyes move
continuously, in the form of discrete saccades, across visual stimuli so that, with each
fixation, we can extract fine detailed information from the visual array, information that
we then encode into short- and long-terrn memory (Henderson, 2007). Although attention
and fixation can be dissociated, as demonstrated in the laboratory (e.g., Posner, 1980), in
everyday life they are tightly linked (e.g., Corbetta et al., 1998; Findlay & Gilchrist,
2003; Henderson, 2003, 2007; Shepherd, Findlay, & Hockey, 1986). AS we move our
eyes, our attention is directed to the location of the current fixation and to the location of
the next fixation (Henderson, 2007 ; van Diepen & d'Ydewalle, 2003). Such patterns of
saccades and fixations are determined by a complex interaction of top-bottom and
bottom-up processes, consisting of the physical properties of the visual world, such as
color and spatial frequency, and cognitive factors, such as our knowledge, experience,
and intentions (Becker, Pashler, & Lubin, 2007; Torralba, Oliva, Castelhano, &
Henderson, 2006; Yarbus, 1967). The study of eye movements, therefore, provides a
window into how we perceive and represent visual information in real time.

In general, the location and duration of a fixation tend to cluster in the more
interesting and informative regions of the visual stimuli, such as regions with high spatial
frequency or regions that are semantically relevant for a given task (Henderson, 2003).

Thus, regions recruiting more and longer fixations are assumed to receive more cognitive

51

processing (Rayner, 1998) and, in turn, to affect how visual stimuli are represented and
remembered (Henderson, Williams, & Falk, 2005).

Eye movements also can be used to study individual differences, since eye
movement studies have found reliable individual differences in complex visual tasks
(e.g., Andrews & Coppola, 1999; Castelhano & Henderson, 2008; Rayner, Li, Williams,
Cave, & Well, 2007). For example, in a free-viewing task, persons more interested in
infants might be expected to look at infant images more often and for longer periods of
time, and even, perhaps, to fixate on certain facial features relatively longer. To the extent
that such persons attend more to infant images, they may also be able to remember them
better.

The goal of Study 6 was to find out whether individual differences in interest in
infants is associated with differential exploratory oculomotor behavior for faces of
different ages (infants vs. adults) and different sexes (male vs. female). Although it is
possible that adult and infant faces elicit distinctive oculomotor responses from people
with different interest in infants, the direction of the effect may be task-specific. Thus, on
free-viewing tasks, people with more interest would be expected to fixate longer and
more often on infant faces than adult faces; on other more cognitive]y-demanding-tasks,
however, such as memory tasks, the direction of the effect may be diminished or even
reversed. For instance, if infant faces are more salient to people with more interest in
infants, it is possible that the greater salience will facilitate processing efficiency, so that
infant faces will require fewer and/or briefer fixations for them to be remembered.

Study 6 was designed to test this hypothesis by examining subjects’ eye

movements while they performed a free-viewing task of several pairs of faces, one of an

52

adult, one of an infant (the design, therefore, assumes that both stimuli are competing for
attentional resources). Subjects then were given the interest-in-infants andother
questionnaires, and a face-recognition test. The prediction was that people with more
interest in infants will Show more fixations, longer viewing durations, and have higher

recognition scores for infant faces than people less interested in infants.

Subjects

The subjects were 32 men (mean age = 19.38, SD=1.18) and 31 women
(M=l9.29, SD = 1.68 years), all undergraduates at Michigan State University, who
participated for course credit. All were White and reported having normal or corrected to
normal vision. None were parents, and 1 woman reported that she was expecting a child.
The sample was selected from a slightly larger sample (69 subjects) based on their
descriptions of their sexual feelings and sexual fantasies on two 7-point Kinsey scales
(Kinsey, 1948). The 63 selected reported being completely heterosexual (N=58) or
predominantly heterosexual (N =5). The intent was to simplify the interpretation of any

potential differences in the attentional effects of same vs. opposite-sex-adult faces.

Procedure

The study was divided in three parts. In the first part (free-viewing task), subjects
saw a sequence of pairs of adult and infant faces while their eye movements were
recorded. In the second part (delay period), subjects answered a series of self-report
questionnaires. In the third part (face recognition), subjects’ memory for the faces they
saw in the free-viewing task was tested. All testing was performed in dimly lit, sound-

attenuated booths.

53

Free-viewing Task. In the free-viewing task, subjects’ eye movements were
monoculary recorded at a sampling rate of 500Hz using the head mounted video-based
eye tracking system EyeLink II (SR Research Ltd., Mississauga, Ontario, Canada).
Before starting the experimental trials, the manufacturer’s procedures were used to
calibrate the eye tracker and to validate the eye positions. For calibration, subjects were
asked to fixate on a sequence of 9 points that appeared at different locations in the
display; the eye tracker was adjusted if necessary until the average tracking error was less
than 0.4 degrees of visual angle. The validation procedure assessed the accuracy of the
system in predicting gaze position from pupil position using an identical random
sequence of 9 points in the display. Once the eye positions were validated, subjects were
shown the sequence of faces. Throughout the eye tracking part of the study, measurement
accuracy was reassessed before each trial by asking subjects to fixate on a dot in the
center of the screen. If accuracy was low, the eye tracker was recalibrated.

Subjects were told that “we are interested in learning about how people look at
faces. You will see a series of pairs of faces while we record your eye movements. You

do not have to do anything, just look freely at the faces and the program will change the
pictures every 6 seconds.”

Subjects saw a sequence of 24 pairs of faces, each consisting of one adult face (a
man or woman) and one infant face (a boy or girl). To minimize face-specific-effects
(e.g., that certain faces, independent of age, are more visually salient), adult and infant

faces were randomly paired and randomly presented to each subject. For each pair, one

face (~ 8 o X 10.50) was shown 6.50 above the fixation point, the other 6.50 below

fixation. These locations were chosen to minimize any possible confound with known

54

laterality effects for emotional stimuli; that is, emotional stimuli tend to be more salient
in the left visual hemispace (e. g., Borod, Zgaljardic, Tabert, & Koff, 2001). This
decision, however, also led to discarding pupilometry as a measure of interest in infants
because of the substantial error associated with the use of head-mounted systems,
namely, larger average pupil size for fixations at the bottom than at the top of the screen
(Pomplun, Sunkara, Fairley, & Xiao, n.d.).

Each face pair was presented for 6 s, meaning that each face could be looked at
for 3 5, assuming that subjects divided their looking times equally between the members
of a pair (3 s also is comparable to the presentation time given in other face recognition
studies; e.g., Hsiao & Cottrell, 2008). It was assumed that limiting the presentation time
to 3 s per face made the face recognition task difficult, and that a difficult task was more
likely than an easy task to show individual differences. The images display location
(above or below fixation), age (adult or infant) and sex (male or female) were
counterbalanced across the 24 trials.

The stimuli were presented on a 19.7-inch CRT monitor placed 30 cm from the
subject. The display resolution was set to 1024 x 768 pixels with a refresh rate of 85 Hz.
Subjects completed this part of the study in about 10 minutes.

Delay Period. During the delay period, subjects answered all questionnaires:
Demographics, Self-Perceived Mate Value, Sexual Orientation Inventory-Revised,
Interest in Infants, Desire for Parenthood. The questionnaires were presented in the same
order for all subjects in a computer station adjacent to the eye-tracker station, and were

told to answer them at their own pace. This part was completed in about 13 minutes.

55

Face Recognition. In the test phase, subjects were shown the 48 faces presented
in the free-viewing task plus 48 foils. To further minimize face-effects, faces were
divided into two sets, with the study set for half the subjects corresponding to the foils for
the other half.

The test phase began 20 minutes after the free-viewing task was completed. Faces
were presented individually in random order. Subjects were instructed to press “Y” if
they had seen the face and to press “”N if they had not, and to respond as quickly and
accurately as possible. Each trial began with the presentation of a central fixation cross,
for 1000 ms, followed by a face, which remained on the screen until the subject

responded. This part of the study was completed in about 5 minutes.

Results

Questionnaires

Interest in Infants. The interest in Infant questionnaire showed good internal
consistency for both sexes (men: or = .91; women: or = .95). Women’s scores (M = 5.46,
SD = 0.98) were higher than men’s (M = 4.73, SD = 0.82) as shown in a two-tailed
independent samples Welch t-test, t(58.58) = 3.23, p = .002.

Job Preference Questionnaire. Preference ratings were examined with a mixed-
design ANOVA with job type (infant care, other jobs) as a within-subjects factor and sex
of subject (male, female) as a between-subjects factor. The analysis showed that the main
effect of job type was not significant, F (1,61) = 0.64, p = .427, mostly because of

women’s and men’s similar preferences for non-infant-care jobs (Men: M = 4.28, SD =

0.80; Women: M = 4.15, SD = 0.77). On the infant-care job, however, women’s

56

preferences (M = 5.29, SD = 1.87) were higher than men (M = 3.5, SD = 1.93) as

indicated by the significant interaction between job type and sex, F (1,61): 17.93, p <
001,11,2 = .23.

Interaction with Infants. The results of the Interaction with Infants questionnaire
are summarized in Table 12. In general, the questionnaire showed good internal
consistency for both sexes. The only scale with reliability lower than the standard figure
of 0.7 was, for women, the subscale of interaction when the infant’s emotional state was
negative. Because the departure from the standard was minimal, the subscale was

included in the multivariate analysis.

Table 12. Descriptive Statistics and Sex Differences of the Interaction with
Infants Questionnaire

 

Men Women Sex difference
a M SD G M SD t df p
Positive .88 4.97 1.03 .93 5.99 1.19 3.62 59.117 <.001
Neutral .88 4.61 1.29 .9 5.65 1.22 3.28 60.96 <.001
Negative .77 4.33 1.2 .67 5.5 0.88 4.45 56.852 <.001
Total .92 4.64 1.02 .92 5.71 0.92 4.40 60.686 <.001

Desire for parenthood. The Desire to Have Children questionnaire also showed
good internal consistency for both sexes (men: or = .86; women: or = .82). As indicated by
a two-tailed independent samples Welch t-test, t(60.99) = 2.5, p = .015, women had a
slightly higher desire to have children (M = 5.24, SD = 1.04) than men (M = 4.57, SD =

1.09).

Self-Perceived Mate Value. The Self-Perceived Mate Value scale showed good

internal consistency for both sexes (women: or = 0.94; men: or = 0.91 ). A two-tailed

57

independent sample Welch t-test showed no significant difference between men (M =
5.04, SD = 1.07) and women (M= 5.09, SD = 1.14), t(60.47) = 0.16, p = .87.

The Revised Sociosexual Orientation Inventory. Table 13 summarizes the
inventory statistics for the Revised Sociosexual Orientation Inventory. The inventory
showed good reliabilities on all subscales and significant sex differences, with men
scoring higher than women on the total scale as well as on the attitude and desire
subscales. These results are comparable to those previously reported for this

questionnaire (Penke & Asendorpf, 2008).

Table 13. Descriptive Statistics and Sex Differences of the Sociosexuality
Measures

 

Men Women Sex difference
0t M SD or M SD t (If p
Behavior .86 2.47 1.36 .86 2.05 1.12 -1.35 59.50 =.019
Attitude .93 5.85 2.36 .87 2.89 1.98 -5.39 59.78 <.001
Desire .76 5.58 1.74 .84 2.44 1.19 -8.38 54.92 <.001
Total .85 4.64 1.42 .83 2.46 1.09 -6.82 57.96 <.001

Eye movements

Eye movements were analyzed by creating a region of interest for each face in the
display. Three measures of interest were considered: location of first fixation, duration of
viewing time for each face, and total number of fixations that each face received.

First fixation location. Since more “attentionally grabbing” objects are more
likely to be looked at first, the analysis asked whether the first fixation was to the infant
or the adult face. The results showed neither: instead, most subjects looked first at the
image located above the fixation point (average proportion of trials for men = .78, and for

women = .87).

58

A multilevel logistic regression therefore was used to examine the probability of
subjects looking at the infant image first based on three variables: the image’s top-bottom
location (0: bottom, 1: top), the pairs shown (boy and man, boy and woman, girl and
man, girl and woman), and sex of subject, and the interaction between these variables. A
variety of models showed that the only significant predictor (in logits) was whether the
image was located on top, ,8 = 2.94, se = .13, p < .001. To further measure the effect of
image location, a subsequent analysis was performed with a subsample of 12 subjects (10
men) who did not show a strong bias to look at the top first. None of the models reached
statistical significance. Thus, given the strong top-down bias of the first-fixation measure,
no further analyses with this variable were conducted.

Total viewing duration for infant faces. The next analysis examined each
subject’s total viewing duration to infant faces compared to adult faces, and more
specifically, how the presence of a given adult face (male or female) affected viewing
duration to infant faces. Total viewing duration was defined as the sum of duration scores
across all fixations on each face; total duration for infant faces therefore was calculated
by subtracting, on each trial, the viewing time for adult faces from the viewing time for
infant faces. Thus, longer viewing durations for infant faces are indicated by positive
scores, longer viewing durations for adult faces by negative scores.

Table 14 summarizes the mean viewing durations for each condition for men and

women separately.

59

Table 14. Mean and Standard Deviation of Infant Viewing Duration for each
Experimental Condition by Sex of Subject

 

 

 

Men Women
Condition M SD M SD
boy with man 405.13 1318.40 429.80 1203.45
boy with woman -27.84 995.95 744.93 1014.88
girl with man 716.16 1250.88 346.20 1063.42
girl with woman -245.62 793.30 407.36 986.80

 

The average infant viewing duration scores for each condition were analyzed with
a mixed-design ANOVA with experimental condition (boy with man, boy with woman,
girl with man, girl with woman) as a within-subjects factor and sex of subject (man,

woman) as a between-subjects factor. A Mauchly’s test showed a violation of the

sphericity assumption (x2(5) = 21.94, p = .001); therefore, degrees of freedom were

corrected using Greenhouse-Geisser estimates of sphericity (a = 0.78). The results

showed a small main effect for condition, F(2.34, l35.7)= 3.84, p = .018, "p2 = .062,

indicating that infant viewing varied across conditions, being particularly lower for the
girl with woman condition (boy with man: M = 417.47, SD = 1251.55; boy with woman:
M = 358.54, SD = 1070.34; girl with man: M= 531.18, SD = 1166.08; girl with woman:

M = 80.87, SD = 946.76). This effect, however, was qualified by an interaction between

condition and sex of subject, F (2.34, 135.7) = 7.6, p < .001, “p2 = .12, indicating that

men looked longer at the infant faces only when they were paired with an adult male

face; when the infant was presented with an adult female face, men look longer at the

adult face. Women, on the other hand, had similar infant viewing times across conditions.
Although there was no main effect for Sex of subject, F (1 , 58) = 1.44, p = .235,

after removing 3 outliers who were consistently above or below +/- 2.5 SD, there was a

60

2 . . . .
small effect of sex, F (1,55) = 5.73, p =.02, np = .09, With women havrng longer vrewmg

times than men (men: M=135.9, SD = 680.51; women: M = 510.59, SD = 480.61). The
removal of the outliers did not change the significance or direction of the other effects.

Relation between viewing duration for infant faces and image statistics. A series
of multilevel regressions were performed to determine whether there was an association
between the total viewing duration for infant faces and the image statistics (e. g., area,
apparent contrast, spatial frequency). Trials represented the first level in all models while
subjects represented the grouping variable.

In all models, differences scores were used as predictors (e.g., the apparent
contrast difference between the infant and adult faces was used to predict the total
viewing duration for infant faces). None of the image statistics or their interactions with
sex and experimental condition reached statistical significance on likelihood ratio tests
(all p > .05). Therefore, image statistics were not further included in multilevel models.

Relation between total viewing duration for infant faces and questionnaires.
Graphic examination of scores showed that two women consistently departed from the
mean distribution of scores. These “outliers,” however, behaved in the direction
predicted: one women, the one who was pregnant, scored high on the interest in infants
questionnaire and had very long infant viewing times; the other woman scored low on the
questionnaire and had very short infant viewing times. Because these two subjects
nevertheless had a significant effect on several of the regression models, it was decided to
exclude them from the multivariate analysis in order to better represent the results of the
other 29 women in the sample. The 2 “outliers” also suggest the possibility that the

sample was affected by a range restriction, at least for women.

61

A series of multilevel regression models were fitted to the data in order to identify
questionnaire variables significantly associated with the total infant viewing duration,
that is, above and beyond the already established relationship between experimental
condition and sex of subjects. Thus, based on the mixed-model ANOVA, a reference
multilevel regression model was defined with infant view duration as dependent variable
predicted by condition (boy-man, boy-woman, girl-man, girl-woman), sex of subject, and
their interaction. In contrast to the ANOVA analysis, data were analyzed at the trial level
(first level) for each subject (subject as grouping second level variable); that is, trial
scores were not averaged (without additional covariates, a trial level analysis will not
change the results provided by the mixed-design ANOVA).

This first step analysis showed that, compared to the base-model, adding the

preference for infantjobs was significant 080) = 5.54, p = .019), as well as the
interaction between preference for infant jobs and the experimental conditions (x2(3) =
10.63, p = .014), the interaction with preference for infant jobs and sex of subject 080)
= 4. 10, p = .043), and the three-way interactions (x2(3) = l 1.94, p = .007). Adding
subjects’ preference for other non-infant care jobs to the base model (x20) = 0.96, p =

.326) or their interactions (780) = 10.78, p = .148) was not significant.

Adding the Interaction with Infants scores to the base-model was not significant

080) = 0.45, p = .502); however, adding the interaction between experimental condition

7
and Interaction with Infants questionnaires showed marginal significance ()(“(3) = 7.62, p

= 0.054). Finally, adding the behavioral scale of the sexual orientation inventory (S01-

62

Behavioral) to the base-model was not significant (x20) = 1.25, p = 0.26), but its

interaction with experimental condition was significant (x2(3) = 13.62, p = 0.003).

A second step in the analysis defined a reference model with infant jobs as a

predictor. Adding the interaction-with-infants scores to the model did not significantly

7
improve the model ()("(1) = 1.46. p = .227) or its interactions (x205) = 12.26, p = .659).

Likewise, adding SOI-Behavioral to the model did not significantly improve the model

(78(1) = 0.30, p = .581) or its interactions (x20 5) = 18.40, p = .242). Therefore, the final

model included only the scores on the preference for an infant care job. The model, a 2-
level random-intercepts model in which trials are nested within subjects, is summarized
in Table 15. In the model, men and condition 1 (boy-men) are used as the reference
categories. To facilitate the interpretation of coefficients, infant job scores were centered
at the mean preference for men (therefore unit changes are understood as change above or
below the men’s mean scores). Figure 2 illustrates the predicted viewing duration for
infant faces of subjects with a preference for the infant job equivalent to the average
preference of male subjects (average interest) and subjects whose preference is 1 unit

above the average preference of male subjects (higher interest).

63

Table 15. Summary of a 2-level Random-Intercepts Model of Subjects Looking

Time Toward Infant Faces

 

 

 

 

 

 

 

 

 

 

 

boy-man

boy-woman

girl-man

Figure 2. Predicted Viewing Duration for Infant Faces.

64

girl-woman

Variable b SE t
(Intercept) 405.133 157.196 2.577
Condition 2 -432.978 179.370 -2.414
Condition 3 311.022 179.370 1.734
Condition 4 -650.756 179.370 -3.628 **
Women 9.454 284.783 0.033
Condition 2 x Women 826.951 324.954 2.545 *
Condition 3 x Women -3l3.118 324.954 -0.964
Condition 4 x Women 599.172 324.954 1.844 .
Infantjob 368.665 83.055 4.439 **
Condition 2 x Infant job -270.171 94.771 -2.851 **
Condition 3 x Infantjob -68.347 94.771 -0.721
Condition 4 x Infantjob -390.123 94.771 -4.116 **
Infantjob x Women -360.253 126.062 -2.858 **
Condition 2 x Infant job x Women 227.379 143.844 1.581
Condition 3 x Infant job x Women 46.337 143.844 0.322
Condition 4 x Infant job x Women 445.247 143.844 3.095 **

.p = .06. *p < .05. **p <. 01.

1000 j
a 1
8 800
«'9
H l
C l
e 600 l
.E I
L
8 400 l
C l
0 ,
25:3 200 g
‘8 a
3 1
U)
.E - + -men (average interest)
03) -200 . +men (higher interest)
5 - -A- -women (average interest)

400 --><-— women (higher interest)

The models shows that, as noted in the ANOVA model, men look longer at
infants only when their faces are paired with same-sex faces, that is, faces of adult males.
Women also look longer when infant faces are presented with same-sex faces (adult
female faces), although these changes were not always significant and less pronounced
than those shown by men. The coefficients also indicate that the effects of interest in
infants (as indicated by scores on the infant job questionnaire) are more evident for men
when the infant faces are presented with an adult male face. The effects on self-reported
interest in women are modest across conditions, with the exception of condition 4.

Total number of fixations for infant faces. As another indicator of interest, each
subject’s total number of fixations to infant faces compared to the number of fixations to
adult faces was examined on each trial, using a binomial model in which fixations on
infant faces were scored as “successes” and fixations on the adult faces as “failures.”

A multilevel binomial regression with fixations on infant faces as the dependent variable

predicted by experimental conditions and sex of the subject, showed a significant effect

of experimental conditions (x2(3) = 17.05. p < .001). There was no main effect of sex of
subject (980) = 3.22, p = .07), but its interaction with experimental conditions was

significant (x2(3) = 42.29, p < .001). The coefficients, summarized in Table 16, show that

men were more likely to fixate on the infant face when it was paired with an adult male
face. Women, in contrast, were more likely to fixate on the infant face than the adult face

in all experimental conditions.

65

Table 16. Summary of a 2-level Random-Intercepts Model of Subjects’ Fixations

Toward Infant Faces

 

 

Variable b SE 2
(Intercept) 0.109 0.051 2.14 *
Condition 2 -0.179 0.053 -3.35 **
Condition 3 0.112 0.054 2.08 *
Condition 4 -0.221 0.054 -4.12 **
Women -0.012 0.072 -0.17
Condition 2 x Women 0.310 0.075 4.14 **
Condition 3 x Women -0.110 0.075 -1.47
Condition 4 x Women 0.244 0.075 3.25 *

 

.p = .06. *p < .05. **p <. 01.

Relation between total fixations for infant faces and image statistics. Binomial

multilevel regressions were used to determine whether there was an association between

the total fixations for infant faces and the image statistics (e. g., area, apparent contrast,

spatial frequency). Trials represented the first level in all models while subjects

represented the grouping variable. In all models, none of the image statistics or their

interactions with sex and experimental condition reached statistical significance on

likelihood ratio tests (all p > .05). Therefore, image statistics were not further included in

multilevel models.

Relation between total fixations for infant faces and questionnaires. A series of

binomial multilevel regression models were fitted to the data in order to identify

questionnaire variables significantly associated with the total number of fixations for

infant faces that were above and beyond the already established relationship between

experimental conditions and sex of subject. A reference model was defined with fixations

in infant faces vs. fixations on adult faces as the dependent variable predicted by

66

experimental condition (boy-man, boy-woman, girl-man, girl-woman) and sex of subject,
and with subject as a random effect.

Compared to the reference-model, adding the preference for the infant care job

7

almost reached statistical significance ()("(1) = 3.742, p = .053), but not its interaction

with sex of subject (x20) = 2.26, p = .132), with the experimental conditions (x2(3) =
7

3.84, p = .278), or the three-way interactions (x"(3) = 3.17, p = 0.366). Adding subjects’

7
preference for other non-infant-care jobs was not significant ()(“(1) = .09, p = .758). In

addition, compared to the reference model, adding the Interaction with Infants scores was

7
not significant (x“(l) = 0.1 1, p = .742) and neither was its interaction with sex of subject
080) = 0.02, p = .887). However, its interaction with experimental conditions almost

reached significance (x2(3) = 7.68, p = .053). None of the other questionnaires showed a

significant relationship with the number of fixations toward infant faces.

A second step in the analysis combined the effects of preference for the infant
care job and the interaction with infant questionnaire. In this model only the preference
for infant care job remained a significant predictor. Therefore, the more parsimonious
model was chosen and the interaction with infant questionnaire was removed. The model

coefficients are summarized in Table 17.

67

Table 17. Summary of a 2-level Random-Intercepts Model of Subjects’ Fixations
Toward Infant Faces as Predicted by Their Interest in Infants

 

Variable b SE "

 

(Intercept) 0.108 0.050 2.162 *
Condition 2 -0.179 0.053 -3.351 **
Condition 3 0.112 0.054 2.084 *
Condition 4 -0.221 0.054 -4.1 19 **
Woman -0.063 0.075 -0.847
Infantjob 0.029 0.015 1.969 *
Condition 2 x Woman 0.310 0.075 4.141 **
Condition 3 x Woman -0.110 0.075 -1.464
Condition 4 x Woman 0.244 0.075 3.252 **

 

.p = .06. *p < .05. **p <. 01.
Along with the already noted effects of experimental condition and its interaction
with sex of subject, the models shows that people with a higher preference for the infant

job are more likely to fixate more often on the infant face than the adult face.

Face recognition

The face recognition data were first analyzed using traditional methods from
signal detection theory (e. g., Stanislaw & Todorov, 1999) by calculating A ’ scores for
each subject on each face category (boys, girls, men, women). Data are presented in this
form first because it facilitates comparing the results with other face recognition studies
and because a multilevel model (without covariates) showed similar results.

A ’ scores, summarized in Table 18, were analyzed with a mixed-model ANOVA
with age of face (adult, infant) and sex of face (male, female) as within-subj ects factors,

and sex of subject (man, woman) as a between—subjects factor. The analysis showed a

main effect of age of face, F(1,59) = 8.56, p = .005, “p2 = .13, indicating that both men

and women have greater signal sensitivity for adult faces (M = .69, SD = .11) than infant

faces (M = .64, SD = .10). An interaction between sex of the face and age of the face,

68

F ( 1,59) = 11.59, p = .001, ”p2 = .16, showed that the effect of age was mostly due to

female faces; that is, men and boys’ faces showed similar A ’ scores (men: M = .66, SD =
.16; boys: M = .67, SD = .1), and woman and girls’ faces showed very different scores

(women: M = .72, SD = .14; girls: M = .60, SD = .16). An interaction between age of face

and sex of subject, F (1,59) = 4.29, p =.043, "p2 = .07, showed that, for men, A ’ scores

were similar for adult and infant faces (adults: M = .67, SD = .1; infants: M = .65, SD
=.1), but for women, adult faces had higher scores than infant faces (adult: M = .71 , SD =
.1; infant: M = .62, SD = .11). Other main effects (sex of subject, sex of face) and their

interactions were not significant (all p > .253).

Table 18. Mean and Standard Deviation A ’ Scores by Face category and Sex of
the Subjects

 

 

 

 

Men Women
Face M SD M SD
Boys .68 .11 .66 .10
Girls .63 .16 .58 .16
Men .65 .15 .67 .18
Women .69 .14 .75 . 14

 

To examine whether the experimental condition affected the likelihood that an
infant face was recognized (i.e., if recognition changed when the face was paired with a
man’s face compared to when it was paired with a woman’s face, only those trials were
examined on which the subject actually saw the face in the free-viewing task. Logistic

multilevel regressions showed that neither the main effects of experimental condition

69

(x2(3) = 7.24, p = .065), sex of subject (x20) = 0.35, p = .553), nor their interaction

(x20) = 2.54, p = .467) were significant.

Table 19. Proportion of Trials, by Face Category, in which a Face was Classified
as Seen or as a New Face

 

 

Response Boys Girls Men Women
New face .43 .50 .51 .51
Seen face .57 .50 .49 .49

 

A multilevel logistic regression showed that subjects were more likely to classify
boy faces as seen faces (fl = 0.298, 2 = 3.18,p = .001) than girls’ faces (fl = -0.298, 2 = -
2.70, p = 0.007). The proportion of trials in which a face was classified as seen is shown
in Table 19. An additional model showed that adult faces of both sexes, in contrast, were
equally likely to be classified as seen (p > .05).

Additional multilevel regression analyses assessed whether the distinctiveness of
faces (see Chapter 4) will contribute to subjects likelihood of saying “seen” to a given
face. The reference model included as predictors whether the face was seen (not-seen as a
reference category), the face type (boys as a reference category), and their interaction.
The first level included trial variables and the second level included subjects and the

variable “seen” as a random slope. Adding distinctiveness significantly improved the

model (x20) = 37.59, p < .001) and its interaction with face type (x2(4) = 41.35, p <

.001). The model, summarized in Table 20, shows an effect of distinctiveness, indicating
that people are more accurate with more distinctive faces, that is, are less likely to say

“not-seen” if a more distinctive face was not seen and to say “seen” if the face was seen.

70

It also shows, consistent with the average A ' scores, that subjects are less accurate on
girls faces and that the effect of distinctiveness was strongest for more distinctive girl
faces. Adding sex of subject or its interaction with other variables did not significantly

improve the model (all p > .05).

Table 20. Summary of a 2-level Random-Intercepts Model of Subjects’ Face
Recognition Responses

 

 

Variable b SE z
(Intercept) -0.779 0.109 -7.182 * *
Girls -0.314 0.139 -2.256 *
Men 0.231 0.154 1.500
Women -0.351 0.135 -2.608 **
Seen face 1.113 0.122 9.122 **
Distinctiveness -0.295 0.091 -3.228 **
Seen face x Distinctiveness 0.403 0.079 5.076 **
Seen face x Girls -0.182 0.161 -1.127
Seen face x Men -0.593 0.191 -3.098 **
Seen face x Women 0.014 0.176 0.079
Distinctiveness x Girls -0.339 0.106 -3.199 **
Distinctiveness x Men -0.036 0.122 -0.297
Distinctiveness x Women -0.119 0.1 13 -l .048

 

.p = .06. *p < .05. **p <. 01.

Face recognition and oculomotor behavior. The relation between subjects’
responses during the face recognition test and oculomotor behavior during the free-
viewing task was examined by selecting subjects’ responses only for those faces they
actually saw in the free-viewing task. Multilevel logistic regression models were used to
predict the likelihood that subjects will classify an infant face as seen. Predictors included
the number of fixations to the infant face, number of fixations to the adult face, a

difference score between these numbers, whether the face was paired with a man’s or

71

woman’s face, the sex of the subject, and a variety of interactions between these
variables. None of these models reached statistical significance (all ps < .05). Similar
models were tested that incorporated the viewing duration times, but because these
models failed to converge, formal statistical analysis could not be performed.

Similar models were implemented to predict the “recognition” likelihood of adult

faces. These models showed that the number of fixations to infant faces (x20) = 4.52, p
= .033), to adult faces (x20) = 20.35, p < .001), or a difference score between the two

(x2(l) = 15.31, p < .001), significantly changed the likelihood of subjects reporting that

they had seen the adult face in the free-viewing task. For instance, the more fixations
made to an infant face compared to an adult face (i.e., a positive difference score), the
lower the likelihood that the adult face was classified as seen in the recognition test (,6 = -
0.045, 2 = -3.92, p < .001). Adding the experimental conditions, sex of subject, or their
interactions as predictors, did not significantly improve these models (all p > .05).

Models that examined the relationship between viewing duration for adult faces
and the likelihood of classifying adult faces as seen failed to converge.

Face recognition and questionnaires. The relation between subjects’ face
recognition and their questionnaire responses was examined by analyzing the likelihood
that a subject reported having seen a face based on whether it was presented in the free-
viewing task, whether it was an adult or an infant face, its distinctiveness, and the
questionnaire response. The focus of this analysis was on the interaction between these
factors, that is, whether the recognition of infant faces was affected by the subject’s

questionnaire score. The results showed that the only questionnaire scores that were

72

 

related to the dependent variable was the Interaction with Infants questionnaire; however,
removing the interaction of Interest with Infants from the model (i.e., whether the

Interaction with Infant scores affected the accuracy for recognition of infant faces) did

not significantly affect the model (78(2) = 1.08, p = .582). Therefore, no further analyses

of this relationship were conducted.

Discussion

The results showed that all the three interest-in-infants questionnaires had good
psychometric properties and behaved as expected, with women expressing more interest
in infants than men as shown by their higher scores on all three questionnaires.

The results also showed that oculomotor behavior could be successfully used to
assess individual differences, including sex differences, in interest in infants, as some
measures of interest were consistently, and in the predicted direction, associated with
oculomotor behavior. Thus, women looked longer at and fixated more often on infant
than at adult faces of either sex, whereas men looked longer at, and fixated more often on
infant faces only when paired with an adult male face. The two measures of oculomotor
behavior, although not perfectly correlated with each other, therefore showed similar
patterns of results. Notably, none of these patterns were accounted for by low level
measures, such as relative contrast or spatial frequency.

The results also were in the predicted direction with respect to the relation
between oculomotor behavior and the Interest in Infants questionnaires. Subjects
reporting higher interest in infants were more likely to look longer and to fixate more
often at infant faces. The association between questionnaire scores and the different

oculomotor measures also showed some qualifications in how these variables are related

73

to each other. For infant viewing time, although women reported greater interest in
infants than men, their scores were not consistently associated with their looking times
toward infant faces, possibly because of the restricted range of their interest scores, with
most scores concentrated in the higher end of the scale. When they were significantly
associated, however, as in condition 4, it was in the predicted direction, that is, women
with higher interest in infants looked longer at infant faces. In contrast, men’s scores,
consistent with their eye-tracking results, were significantly associated with their looking
times toward infants only when the infant face was paired with the face of an adult male.
For the measure of the number of fixations toward infant faces, the association with
interest in infants questionnaires was weaker than the association with looking times. The
only questionnaire variable showing some degree of association with fixations was the
preference for the infant job on the Infant Job questionnaire, with subjects with higher
preference being more likely to fixate more often on the infant face, an effect not
qualified by significant interactions between interest in infants measures, experimental
conditions, and sex of subject.

In sum, although oculomor behaviors toward infant faces showed some reliable
degree of association with scores on the Interest in Infants questionnaires, there are
inconsistencies in the pattern of results. One possible reason may simply be that some
oculomotor measures are more directly associated with interest in infants than others.
Another possibility is that the regions of interest used in the study were insufficiently
precise, so that not all measures gave the same results. For instance, it could be that

people with higher interest in infants are more likely to fixate on those facial regions that

74

are more diagnostic of the infant’s emotional state, such as the mouth and eyes. A finer-
grained analysis will be needed to assess this possibility.

Another question about the association between interest in infants and oculomotor
behavior has to do with the fact that only some Interest in Infants questionnaires were
associated with eye movements. A possible reason is that the questionnaires targeted
different dimensions of interest. For example, one important difference between the Job
Preference, Interaction with Infant, and Interest in Infants questionnaires is that only the
first two assess the subject’s willingness to interact with an infant. Nonetheless, only the
Job Preference questionnaire showed a more consistent association with oculomotor
behavior. One possibility for this discrepancy is that the Interaction with Infant
questionnaire asks people to imagine interacting with a stranger’s infant. This could
introduce a confound, since some people, despite being interested in infants, may worry
that the parent would see this particular interaction as overly intrusive. In contrast, for the
Job Preference questionnaire, interaction is not only socially acceptable, it is expected as
a part of the job.

Overall, then, two of the three measures of oculomotor behavior were associated
with interest in infants, both of which were consistently associated in the predicted
direction.

On the face recognition test, some of the results also were as predicted, namely
that nulliparous adults will be better at recognizing adult faces than infant faces, given
their greater experience and higher degree of expertise with adult faces and given that
adult faces are more distinctive, and that more distinctive faces are typically more

memorable. The association, although weak, between fixations and recognition for adult

75

faces was also expected as the number of fixations is positively correlated with face
recognition. Several aspects of the results, however, are problematic. For infant faces, no
association was found between fixations and recognition accuracy and, overall, the
recognition scores were low compared to other studies (e.g., Hsiao & Cottrell, 2008,
reported A ' recognition scores of .8 when subjects were allowed to fixate only twice on
faces). Thus, it appears that the task was more difficult than intended. Several factors are
likely to have contributed to the subjects’ low performance, including the decision to not
present the faces individually in the free-viewing task, to present them for only a brief
time (meaning that the task was too difficult), to standardize the faces to minimize any
effects of low level vision, and to not instruct the subjects to study the faces in
preparation for a recognition task. The absence of an association between face
recognition scores for infant faces and interest in infants therefore may have been a
product of the overall low recognition performance, that is, of a floor effect. A different
task, in which subjects are actually instructed to study the faces, may help to expand the
range of scores and increase the likelihood of finding the predicted association between

performance and interest in infants.

76

Chapter 6: Study 7 -- Visual prior entry for infant faces

We come finally to the last study, Study 7,on visual prior entry for infant faces.
There is evidence that motivationally significant stimuli can capture attention, in other
words, that people are predisposed to attend to emotional stimuli (e.g., a fearful face)
even when told not to do so (e.g., Eastwood, Smilek, & Merikle, 2003; Ohman, Flykt, &
Esteves, 2001; Vuilleumier & Schwartz, 2001) and that emotional stimuli affect early
vision, that is, that people Show better contrast sensitivity after briefly (75 ms) seeing
fearful faces than neutral faces (Phelps et al., 2006).

Recent studies Show that infant faces, another type of emotionally significant
stimulus, can produce covert shifts of spatial attention (Brosch et al., 2008; Brosch et al.,
2007). These studies used the dot-probe-task, in which two faces (an infant face and an
adult face, both with neutral expression) are presented simultaneously for a short time
(100 ms) and then turned off, after which, following a short delay period of variable
length (100-300 ms), a small dot briefly appears in the location previously occupied by
one of the faces. Participants are typically faster and more accurate at detecting the dot
when it appears in the location occupied by the emotional face (in this case the infant
face); that is, people seem to improve their performance because their attention was
directed, or captured, by the emotional stimulus. Although the task is sensitive enough to
detect shifts of attention, its temporal resolution at face onset (when the faces are first
presented) is insufficient for determining precisely when and how infant faces capture
attention because participants are allowed to respond only after about 200 milliseconds

following face offset. To allow measurement of the time course of initial attentional

77

deployment, Study 7 therefore used a method with better temporal resolution during face
onset.

According to the doctrine of prior entry (Titchener, 1908), stimuli that are
attended immediately at onset are perceived prior to unattended stimuli because attended
stimuli receive processing priority. Although this doctrine has a long history in
experimental psychology, only recently have methods been developed that convincingly
Show the existence of the effect (e. g., Shore, Spence, & Klein, 2001; West, Anderson, &
Pratt, 2009), that is, that attended stimuli receive priority in perceptual processing and
that the effect is not simply due to a response bias (e. g., subjects choosing one response
option when unsure). Visual prior entry is typically measured by a temporal order
judgment (TOJ) in which the subject reports which of two items appeared first on the
display. Items are generally presented in asynchrony; that is, one item appears first, and
after a brief and varied delay (stimulus onset asynchrony, SOA) the second item appears.
Since items that are more attentionally salient are assumed to facilitate perception,
attended items will be perceived as appearing first, and, by varying the SOA, it is
possible to estimate the length of time that the less salient image must precede the more
salient image in order for them to be perceived as appearing simultaneously (point of
subjective simultaneity, PSS). To minimize the impact of response biases (i.e., the
tendency to choose one type of stimulus when unsure which one appeared first), typically
half of the subjects are asked to report which item appeared first and the other half are
asked which item appeared second.

Recently,West et al., (2009), using schematic faces and photographs of faces,

found a prior entry effect for angry faces, such that angry faces were perceived earlier

78

than neutral faces. The PPS for the angry faces ranged from 7.85 to 18.26 ms, with
photographs showing a larger effect than schematic faces.
In Study 7, West et al., 2009, procedure was adapted to see whether infant faces,

compared to adult faces, also Show prior entry effects.

Subjects

The subjects were 26 men (mean age = 19.81, SD = 1.74 years) and 29 women (M
= 19.2, SD = 1.24), all undergraduate students at Michigan State University, who

participated for course credit. All were White and reported having normal or corrected to

normal vision. None were parents.

Stimuli

Eight faces (2 boys, 2 girls, 2 men, and 2 women) were chosen from the stimulus
pool so that the faces in each category had Similar attractiveness. In addition, a mask was
created by producing an average face of the 4 adult faces with a morphing software. To
compensate for the difference in the length of adult and infant faces, the face length for
the mask was resized to represent the average between them. The resizing was done so
that all facial features were also slightly shrunk and shifted towards the center of the face,

making their size and location closer to the average for adult and infant features.

Procedure

The study was in two parts: the prior entry task in part 1, the questionnaires in

part 2.

79

Stimuli were presented on a CRT monitor. The display resolution was set to 1024
x 768 pixels with a refresh rate of 85 Hz. Subjects sat at a distance of 44 cm from the
display.

A typical sequence is shown in Figure 3. Each trial began with a fixation cross

(0.60 X 0.60) with two placeholder boxes (3.2 o X 4.550) centered at 4.90 from fixations.

Subjects were told to fixate on the cross throughout the experiment and to allocate their

attention equally to both boxes.

+ Prestimulus interval (1000 ms)

 

 

 

 

 

SOA (0 -108 ms)

 

Until response

Figure 3. Trial Sequence Used in Experiment 7.

After 1,000 ms, an infant (boy or girl) and an adult face (man or woman)
appeared on the screen, each in one of the two placeholders. On 9% of the trials the faces

appeared simultaneously (stimulus onset asynchrony (SOA) of 0 ms); on the remaining

80

91%, one face preceded the presentation of the other face by one of 5 stimulus onset
asynchronies (12, 24, 48, 60, or 108 ms). Face categories (boy or girl, man or woman),
order of appearance (adult or infant appearing first), SOA, and location (left, right) were
counterbalanced across trials. After the faces appeared, they remained on the screen for
62 ms and then were masked by the average face. The mask remained on screen until
subjects responded by pressing one of two keys on the keyboard.

To control for response biases, subjects were divided into two groups (group 1
with 13 men and 15 women; group 2 with 13 men and 14 women). Subjects were
instructed to indicate which face appeared first on the screen (group 1) or which face
appeared second on the screen (group 2). Subjects responded “the face at the left” by
pressing the 2 key and “the face at the right” by pressing the / key. After the subject’s
response, a white background was displayed for 1,000 ms before the next trial began.

Before the experimental trials began, subjects completed 16 practice trials with
SOA of 170 and 220 ms. Only adult faces (and none of those used in the experimental
trials) were used in the practice trials. The experiment consisted of 352 trials divided into
8 blocks of 44 trials. Subjects were allowed to take small breaks between blocks. Subjects
completed the experiment in about 16 minutes.

In part 2, subjects answered, individually, all questionnaires: Demographics, Self-
Perceived Mate Value, Sexual Orientation Inventory—Revised, and Interest in Infants. All
subjects answered the questionnaires in the same order and in the same computer station,
and were instructed to answer at their own pace. This part was completed in about 13

minutes.

81

Results

Questionnaires

Questionnaire results are presented in combined form for Group 1 and Group 2.
Because two subjects experienced computer problems while answering the
questionnaires, the degrees of freedom vary slightly across questionnaires.

Interest in Infants. The Interest in Infants questionnaire showed good internal
consistency for both sexes (men: or = .90; women: or = .90). Women’s scores (M = 5.08,
SD = 0.77) were higher than men’s (M = 4.59, SD = 0.84) as shown in a two-tailed
independent samples Welch t-test, t(50.21) = 2.18,p = .033.

Job Preference Questionnaire. Job Preference ratings were examined with a
mixed-design ANOVA with job type (infant care, other jobs) as a within-subjects factor
and sex of subject (male, female) as a between-subjects factor. The analysis showed that
the main effect of job type was not significant, F (1,51) = 0.01, p = .922, mostly because
of women’s and men’s similar preferences for non-infant-care jobs (Men: M = 4.27, SD =
0.65; Women: M = 4.04, SD = 0.88). On the infant-care job, however, women’s
preferences (M = 5.26, SD = 1.85) were higher than men (M= 3.0, SD = 1.55) as

indicated by the significant interaction between job type and sex, F (1,51) = 20.29, p <

2
.001, up = .29.

Interaction with Infants. The results of the Interaction with Infants questionnaire
are summarized in Table 21. The questionnaire showed good internal consistency for

both sexes, with the exception of the negative subscale for men. All subscales, except the

positive subscale, showed sex differences, with women scoring higher than men. Based

82

on these results only the total scale was used as it has better reliability and shows a clear

sex difference.

Table 21. Descriptive Statistics and Sex Differences of the Interaction with
Infants Questionnaire

 

Men Women Sex difference
0t M SD or M SD t df P
Positive .90 5.21 1.07 .91 5.72 1.02 1.77 50.64 .082
Neutral .88 4.48 1.27 .85 5.40 1.43 2.89 48.07 .006
Negative .62 4.68 0.88 .71 5.30 0.87 2.57 50.86 .013
Total .92 4.79 0.96 .93 5.47 0.87 2.72 50.03 .009
Desire for parenthood. The Desire to Have Children questionnaire also showed
good internal consistency for both sexes (men: or = .83; women: or = .83). As indicated by
a two-tailed independent samples Welch t-test, t(50.86) = 0.50, p = .61, women (M =
4.57, SD = 1.01) and men (M = 4.43, SD = 1.03) had similar desire to have children.
Self-Perceived Mate Value. The Self-Perceived Mate Value questionnaire also
showed good internal consistency (men: or = .89; women: or = .92). A two-tailed
independent sample Welch t-test showed no significant difference between men (M =
4.49, SD = 1.17) and women (M= 5.06, SD = 1.21), t(51)=1.73,p = .089.
The Revised Sociosexual Orientation Inventory. The Revised Sociosexual
Orientation Inventory also showed good internal consistency on all subscales and as well

as significant sex differences, with men scoring higher than women on the total scale as ‘

well as on the attitude and desire subscales. The results are summarized in Table 22.

83

Table 22. Descriptive Statistics and Sex Differences of the Sociosexuality
Measures

 

Men Women Sex difference
a M SD a M SD t df p
Behavior .78 1.94 1.07 .87 2.07 1.05 0.47 50.86 =.637
Attitude .85 4.04 2.27 .87 2.88 1.95 -1.99 49.27 =.051
Desire .85 4.76 2.12 .77 2.83 1.62 -3.72 46.79 <.001
Total .80 3.58 1.36 .91 2.59 1.40 -2.60 50.99 <.001

 

Prior Entry

On the prior entry task, trials with response-times greater than 2,500 ms were
removed from the analysis (3.2% of group 1 data and 4.6 % of group 2 data).

Given that prior entry effects occur at the short SOAs, subjects’ accuracy was
assessed only for the longer SOAs, which were not expected to be difficult. For SOA 60,
average-accuracy was 91 % for group 1 (87% for men, 94% for women) and 91% for
group 2 (87% for men, 94% for women). For SOA 108, group 1 accuracy was 96% (95%
for men, 96.9 % for women) and for group 2 was 93.3% (96.5% for men, 90.4% for
women). One woman from group 2 had below 70% accuracy, so her data were removed
from further analyses.

Logistic regressions were fitted to each subject’s responses in order to calculate
the PSS, that is, the interval needed for both stimuli to be perceived as appearing
simultaneously. The PS8 was computed as the point at which the subject was equally
likely to report seeing each target as appearing first (or second). Positive PSS indicate
that the infant face is perceived earlier; negative values indicate that the adult face is

perceived earlier.

84

For group 1, a one-sample two-tailed t-test, t(27)= 0.90, p = .3738, showed that
the average PSS (M = 1.003, SD = 5.87) was not significantly different from zero.
Although men had higher PSS values (M = 3.25, SD = 5.55) than women (M = -0.94, SD
= 5.59), the difference was not significant based on a two-sample two-tailed Welch t-test,
t(25.49l) = -1.986, p = .059. By contrast, for group 2, a one-sample two-tailed t-test,
t(25)= -3.39, p = .002, showed that the average PSS (M = —6.07, SD = 9.12) was
significantly smaller than zero, indicating a prior entry effect for adult faces. Although
men had slightly smaller PSS values (M = -3.72, SD = 5.78) than women (M = -8.43, SD
= 11.31), this difference, likewise, was not significant based on a two-sample two-tailed
Welch t-test, t(17.86) = -l .335, p = .198. Given that the PSS values were in opposite
directions for both groups, that is, positive for group 1, negative for group 2, the results,
instead of showing a prior entry effect, therefore showed a response bias effect.

Even though a prior entry effect was absent for the entire sample, a series of linear
regressions were fitted to the data on the possibility that PSS values were associated with
interest in infants or the other individual differences. None of the models reached

statistical significance (all p > .05).

Discussion

Experiment 2 examined whether infant faces Show visual prior entry and therefore
early attentional capture. The results did not show this effect for infant faces. They
instead showed evidence of a response bias; that is, subjects chose one of the face
categories more often when unsure which face appeared first. In addition, and more
critical to our hypothesis, there was no association between PSS and measures of interest

in infants and other individual differences questionnaires. It is unlikely that the negative

85

results reflected insufficient statistical power. Although West et al., (2009) did not report
standard deviations or effect sizes, effect sizes can be roughly estimated based on the
magnitude of the standard errors graphed in their figure 4. Given their sample sizes (12-
14 subjects), the effect sizes were likely to be approximately 0.5 for schematic faces and
closer to 1.0 for photographs. Since the sample in the current study was 28 subjects for
group 1 and 27 subjects for group 2, it should have achieved a power of at least .8 for
similar effects. If there are any visual prior entry effects for infant faces, they therefore
are likely to be much smaller than the ones found for angry adult faces.

Until such effects are found, the results suggest that the reported attentional
orientation effects of infant faces (Brosch et al., 2008; Brosch et al., 2007) are not likely

to occur during the initial attentional deployment but rather after 50 ms.

86

Chapter 7: Conclusions, Limitations, and New Directions

As summarized in Chapter 1, human infant care is essential to human survival but
also shows, compared to other mammals, some of the most remarkable idiosyncrasies of
the human species. Even though rearing human infants is far more costly than in most
mammals, humans are able to reach higher than expected reproductive rates, a fact that
highlights the role of cooperative breeding in human populations and, more specifically,
of allomatemal care with provisioning. Because of the high costs of infant care, the
idiosyncracies of human parental behavior, and the vital benefits of infant survival, it has
been hypothesized that, over the course of human evolution, a variety of biological and
cognitive mechanisms have evolved that facilitate adults’ interest in and responsiveness
to infants in ways normally leading to care-giving. As summarized in Chapter 2, evidence
for such mechanisms comes from studies indicating that adults Show attentional-
emotional biases, along with physiological, neural, and hormonal responses, toward
specific features and behaviors that make infants look “cute” and “attractive.” Despite
evidence that infants constitute a special stimulus category for humans, however, the
nature of cognitive mechanisms underlying interest in infants is still largely unexplored.

The goal of this dissertation was to find out whether one such cognitive
mechanism is revealed in visual attention, and whether and how visual attention is related
to subjects’ self-reports of interest in infants and other subject characteristics. Studies 1
and 2 assessed the reliability and validity of three questionnaires designed to estimate
subject’s interest in infants, each measuring a different dimension of interest, including
the willingness to interact with infants. The results Show that these questionnaires have

good reliability and validity. Evidence of their validity was seen in their association with

87

other related constructs, such as willingness to have children, but also with the behavioral
measures used in Study 6; that is, some of the questionnaires were associated, and in the
predicted direction, with oculomotor behaviors that are also seen as reflecting interest in
infants. Studies 3-5 assessed the characteristics of the face-stimuli used in the
dissertation, in particular, whether the faces retained their age, sex, and affective
characteristics after being standardized. The results confirmed that the faces conserved
many of the properties of natural (unprocessed) faces and therefore that they were
suitable for being used in Studies 6 and 7.

Study 6 examined the exploratory oculomotor behavior toward infant and adult
faces, and Study 7 examined whether infant faces have a greater effect than adult faces on
the initial deployment of attention. Both studies investigated how such behaviors
correlate with subjects’ self-reports of interest in infants and other subject characteristics.

Study 6 showed a clear difference in men’s and women’s reactions to infant and
adult faces. Women looked longer at and fixated more often on infant than at adult faces
of either sex. Men, in contrast, looked longer at and fixated more often on infant faces
only when they were paired with an adult male face. On the self-reports measures of
interest in infants, women also reported greater interest in infants than men; their scores,
however, were not consistently associated with their looking times toward infant faces. In
contrast, men’s scores, consistent with their eye-tracking results, were significantly
associated with their looking times toward infants only when the infant face was paired
with the face of an adult male. Overall, the results suggest that women’s interest in
infants as indexed by their oculomotor behavior and self-reports is more stable and higher

than men’s.

88

As was speculated in Chapter 2, men could be more likely to show the effects of
individual differences inasmuch as their interest in infants and parental investment are
even more conditional than women’s. As reviewed in Chapter 2, an abundance of
evidence, including the results of Studies 1 and 2, shows that women are typically more
interested in infants and have a greater role in infant care than men. Although fatherhood
typically increases men’s interest and role in infant care, some studies Show that fathers
nonetheless have little or no effect on infant survival (Sear & Mace, 2008), which further
supports the hypothesis that male parental investment is facultative rather than obligate,
that is, although not required for the offsprings’ survival, it can enhance their survival and
reproductive success (Geary, 2008), and suggests that men’s interest and parental
investment are more likely to be more affected by a large range of factors (e. g.,
reproductive strategies) than women’s. Study 6 also showed that oculomotor behavior
can be successfully used to assess individual differences, including sex differences, in
interest in infants, as subjects with greater interest looked longer and made more fixations
than subjects with less interest.

Study 7 showed that infant faces do not appear to Show a prior entry effect or at
least that the effect is likely to be smaller than the effect reported for other emotional
stimuli, such as angry faces. Study 7 results also failed to show an association between
individuals’ point of subjective simultaneity (PSS) and their self-report interest in infants,
further indicating the lack of evidence that interest in infants affects the early deployment
of attention toward infant faces. The results thus suggest that the attentional and neural
effects reported in other studies, effects typically recorded 100 ms after stimulus

presentation, are not likely to be evident during the initial deployment of attention.

89

Overall, the results further support the finding that women are more likely than
men to be interested in infants and suggest that they deploy their attention in a manner
consistent with this sex difference. Hrdy (2009) argues that because a woman’s decision
to have and raise a child is highly conditional on whether ecological conditions are
favorable for having a viable offspring, it is inaccurate to characterize women as having a
“maternal instinct” or to assume that their responses to infants come in the form of fixed
response patterns. Although our results do not directly speak to this issue, we agree with
Hrdy’s point that women involvement in infant care is conditional. At the same time, we
also want to stress that under favorable ecological conditions women are more involved
and more interested than men in infant care. In our sample, for instance, although both
men and women expressed similar desire to have children, only women showed a
consistently higher interest in infants across the self-report and behavioral measures.
Thus, although there are important individual differences within each sex, sex differences
in interest in infants seem particularly robust. In other words, although I agree that
women’s behavior toward infants should not be seen as fixed response patterns, the
results of the current studies also agree with the view that women, compared to men, are
in general more attuned to infants and thus more likely to interact with and respond to
infants. The results therefore are consistent with the hypothesis that women, due to their

central role in infant care, have evolved a greater and more stable sensitivity to infants.
Limitations of Studies, Further Analyses, and Directions for New Research

Use of Undergraduates

A probable limitation of all studies was their use of a convenience sample of

undergraduate students. Undergraduates, of course, are not representative of all the

90

reproductive stages in humans. Future studies therefore should include parents and
persons planning or expecting to be parents. Because both groups would expect to have
greater interest in infants than nulliparous subjects, they ought to provide for a more
powerful test for estimating the validity of the oculomotor and other measures used in the
current studies. In this regard, it may be relevant that in the scores for the one subject in
the sample who reported to be expecting a child, there was an even stronger association
between the oculomotor and questionnaire measures than was found for the rest of the

sample.

Further analyses of current data

Several additional analyses of the current data are planned. For instance, to
improve data reliability, factor analysis will be used to merge all of the interest-in-infants
questionnaires into a single measure that will be used in latent variable modeling through
Structural Equation Modeling.

In addition, although common measures of oculomor behavior were used in this
dissertation, the eye-tracking data collected contain a wealth of data warranting further
study. For example, the first fixation location data will be revisited by examining whether
subjects who fixate initially at the top face fixate longer (or have more fixations) before
fixating on the bottom image when the top face is an infant than when it is an adult, and
whether any such differences are related to the sex of the subject and interest-in-infants

scores.

91

Facial Expression

Future studies also should assess the role of facial expressions in the perception of
and attention toward infant faces. Since the ability to respond appropriately and
effectively to an infant’s needs depends on how well the infant’s emotional state can be
understood, varying the emotional state of infant faces (e. g., neutral vs. distressed) could
be a way to enhance some of the sex-related and other individual differences found in
Study 6. As a first step in this direction, in the near future we plan to reanalyze the data
by using smaller regions of interest that are able to Show eye movements on specific
facial features (e.g., eyes, mouth) that are typically more diagnostic for determining the

face’s emotional state.

Hormonal Changes

Given the substantial hormonal changes that women experience during pregnancy
as well as the endocrine changes documented for fathers, it will be of interest to ask
whether these oculomotor behaviors are also affected by endocrine fluctuations. For
example, how do hormonal changes, such as those occurring during pregnancy or during
the menstrual cycle, affect cognitive mechanisms associated with interest in infants, and
what are the effects in persons with endocrine disorders or with normal endocrine
changes associated with aging? Recent work has shown that oxytocin enhances the
recognition of adult faces (Rimmele, Hediger, Heinrichs, & Klaver, 2009). Given its role
in maternal behavior, the possibility arises that oxytocin will also enhance the recognition

I of infant faces.

92

 

Refinement of Oculomotor Measures

A goal of further studies will be to refine these oculomotor measures so as to
enhance their sensitivity to these individual differences. For Study 6, two aspects of the
design need improvement. One of the measures of attention used was the location of the
first fixations. Subjects showed a strong bias to fixate first on the top face of the display,
independently of whether the face was of an adult or an infant. Most likely, this bias is a
product of, or generalization from, reading practices, which, whether the text is read from
left to right or right to left, begin in all alphabets at the top of the page. Similarly,
computer content (e.g., intemet content) typically begins at the top of the screen.
Because, as mentioned earlier, Study 6 used a top-bottom layout to control for laterality
(left-right) effects, the top-down scan bias greatly diminished the usefulness of the first
fixation measure. A possible improvement would be to make the location of faces
unpredictable by having them appear in random locations following an imaginary ring.

Another oculomotor measure that had been planned for use was pupil dilation.
The top-down layout, however, prevented its use because pupil measures with head
mounted systems, especially when the camera is located below the eye, are highly
susceptible to effects of the subject’s gaze angle. In addition, because the pupil diameter
usually requires about 4 s to stabilize after display luminance changes, reliable use of
pupilometry would require a specialized experimental design. One possibility, for a
future study, is to use specialized calibration algorithms, such as neural-networks, that
allow the eye tracker to calibrate its measures based on the subject’s gaze angle (e. g.,

Pomplun, Sunkara, Fairley, & Xiao, n.d.).

93

It is clear that this research is only the first step toward adequately characterizing
the cognitive mechanisms involved in individual differences in interest in infants. It is
important, however, to note that understanding these cognitive mechanisms is not only
worthwhile from a basic science perspective but also from its potential significance for
application and remediation. For instance, there is substantial evidence that physical
appearance affects social interactions (Jackson, 1992; G. Rhodes & Zebrowitz, 2002),
including interaction with infants and children. Thus, on the premise that physical
attractiveness signals infant health, studies find that parental investment is affected by the
infant’s physical appearance (J. Mann, 1992; Waller, Volk, & Quinsey, 2004), and, in
fact, across cultures, “deformed” infants are at greater risk of infanticide (Craig, 2004;
Daly & Wilson, 1984; Kellett, 1992). Relatedly, physically more attractive children
receive less severe punishments and more affective maternal responses than do less
attractive children (Dion, 1972; Langlois, Ritter, Casey, & Sawin, 1995). Indeed, the risk
of maltreatment increases in children with cranial/facial proportions that make them look
older than their actual age, possibly because adults expect them to show socioemotional
behavior of older children (McCabe, 1984). For another example, women’s mental health
is critically important for the mother-infant interaction, as shown, for example, by the
association between postpartum depression (Wisner, Parry, & Piontek, 2002) with a
diminished interest in infants, which presumably is among the factors with negative
consequences for mother-infant interaction and, ultimately, for the infant’s cognitive and
socioemotional development (Murray, Fiori-Cowley, Hooper, & Cooper, 1996). From
these examples, the possibility thus arises that a better understanding of the cognitive and

biobehavioral mechanisms underlying interest in infants not only can improve our

94

 

understanding of how these mechanisms translate into actual care-giving behavior, but
how, through the development of new screening and assessment tools, persons at risk for
poor care-giving can be identified earlier so that effective interventions can be instituted

before care-giving fails.

95

 

References

Aharon, I., Etcoff, N. L., Ariely, D., Chabris, C. F., O'Connor, E., & Breiter, H. C.
(2001). Beautiful faces have variable reward value: MRI and behavioral
evidence. Neuron, 32(3), 537-551.

Alley, T. R. (1981). Head shape and the perception of cuteness. Developmental
Psychology, 17(5), 650-654.

Andrews, T. J ., & Coppola, D. M. (1999). Idiosyncratic characteristics of saccadic eye
movements when viewing different visual environments. Vision Research, 39(17),
2947-2953.

Baayen, R. H. (2009). languageR: Data sets and functions with "Analyzing Linguistic
Data: A practical introduction to statistics ".

Babchuk, W. A., Hames, R. B., & Thompson, R. A. (1985). Sex differences in the
recognition of infant facial expressions of emotion: The primary caretaker
hypothesis. Ethology and Sociobiology, 6(2), 89-101.

Bartels, A., & Zeki, S. (2004). The neural correlates of maternal and romantic love.
Neurolmage, 21(3), 1155—66.

Bates, D., & Maechler, M. (2009). lme4: Linear mixed-eflects models using S4 classes
(Version 0. 9993 75-32). Retrieved from http://cran.r-
project.org/web/packages/lme4/index.html

Becker, M. W., Pashler, H., & Lubin, J. (2007). Object-intrinsic oddities draw early
saccades. Journal of Experimental Child Psychology: Human Perception and
Performance, 33(1), 20-30.

Benson, P. J., & Perrett, D. I. (1991). Perception and recognition of photographic quality
facial caricatures: Implications for the recognition of natural images. European
Journal of Cognitive Psychology, 3(1), 105-135.

Berg, S. J., & Wynne-Edwards, K. E. (2001). Changes in testosterone, cortisol, and
estradiol levels in men becoming fathers. Mayo Clinic Proceedings, 76(6), 582-
592.

Berman, P. W. (1980). Are women more responsive than men to the young? A review of
developmental and situational variables. Psychological Bulletin, 88(3), 668-695.

Berman, P. W., Cooper, Mansfield, Shields, & Abplanalp. (1975). Sex differences in
attraction to infants: When do they occur? Sex Roles, 1(4), 311-318.

96

Berman, P. W., & Goodman, V. (1984). Age and sex differences in children's responses
to babies: effects of adults' caretaking requests and instructions. Child
Development, 55(3), 1071-1077.

Bemick, N. (1966). The development of children 's preferences for social objects as
evidenced by their pupil responses. (Unpublished doctoral dissertation).
University of Chicago, Chicago, Illinois.

Blakemore, J. E. (1981). Age and sex differences in interaction with a human infant.
Child Development, 52(1), 386-388.

Borod, J., Zgaljardic, D., Tabert, M., & Koff, E. (2001). Asymmetries of emotional
perception and expression in normal adults. In F. Boller, G. Gainotti, & J.

Grafman (Eds), Handbook of neuropsychology: Emotional behavior and its
disorders (pp. 181-205). Amsterdam: Elsevier.

Bradley, M. M., Codispoti, M., Sabatinelli, D., & Lang, P. J. (2001). Emotion and
motivation 11: sex differences in picture processing. Emotion, 1(3), 300-19.

Bradley, M. M., Miccoli, L., Escrig, M. A., & Lang, P. J. (2008). The pupil as a measure
. of emotional arousal and autonomic activation. Psychophysiology, 45(4), 602-7.

Brooks, V., & Hochberg, J. (1960). A psychophysical study of "cuteness". Perceptual
and Motor Skills, 11 , 205-205.

Brosch, T., Sander, D., Pourtois, G., & Scherer, K. R. (2008). Beyond fear: rapid spatial
orienting toward positive emotional stimuli. Psychological Science, 19(4), 362-
70.

Brosch, T., Sander, D., & Scherer, K. R. (2007). That baby caught my eye... attention
capture by infant faces. Emotion, 7(3), 685-9.

Byrne, R. W. (2005). Animal evolution: Foxy friends. Current Biology, 15(3), 86-87.

Castelhano, M. S., & Henderson, J. M. (2008). Stable individual differences across
images in human saccadic eye movements. Canadian Journal of Experimental
Psychology, 62(1), 1-14.

Cheng, Y. D., O'Toole, A. J ., & Abdi, H. (2001). Classifying adults' and children's faces
by sex: computational investigations of subcategorical feature encoding.
Cognitive Science, 25(5), 819-838.

Clark, A. P. (2006). Are the correlates of sociosexuality different for men and women?
Personality and Individual Differences, 41(7), 1321-1327.

Clutton-Brock, T. H. (1989). Mammalian mating systems. Proceedings of the Royal
Society of London: Series B, 23 6(1285), 339-72.

97

Clutton-Brock, T. H. (1991). The evolution of parental care. Monographs in behavior and
ecology. Princeton, NJ: Princeton University Press.

Cohen, M. E., & Carr, W. J. (1975). Facial recognition and the von Restorff effect.
Bulletin of the Psychonomic Society, 6(4A), 383-384.

Corbetta, M., Akbudak, E., Conturo, T. E., Snyder, A. Z., Ollinger, J. M., Drury, H. A.,
Linenweber, M. R., et al. (1998). A common network of functional areas for
attention and eye movements. Neuron, 21(4), 761-773.

Craig, M. (2004). Perinatal risk factors for neonaticide and infant homicide: can we
identify those at risk? Journal of the Royal Society of Medicine, 97(2), 57-61.

Daly, M., & Wilson, M. (1984). A sociobiological analysis of human infanticide. In G.

Hausfater & S. B. Hrdy (Eds), Infanticide: comparative and evolutionary
perspectives (pp. 487-502). New York: Aldine Pub. Co.

Delplanque, S., N'diaye, K., Scherer, K., & Grandjean, D. (2007). Spatial frequencies or
emotional effects?: A systematic measure of spatial frequencies for IAPS pictures

by a discrete wavelet analysis. Journal of Neuroscience Methods, 165(1), 144-
150.

van Diepen, P. M. J., & d'Ydewalle, G. (2003). Early peripheral and foveal processing in
fixations during scene perception. Visual Cognition, 10(1), 79-100.

Dion, K. K. (1972). Physical attractiveness and evaluation of children's transgressions.
Journal of Personality and Social Psychology, 24(2), 207-13.

Eastwood, J. D., Smilek, D., & Merikle, P. M. (2003). Negative facial expression
captures attention and disrupts performance. Perception & Psychophysics, 65(3),
352-358.

Eibl-Eibesfeldt, I. (2007). Human Ethology. New Brunswick, NJ: Aldine Transaction.

Erickson, K., Drevets, W., & Schulkin, J. (2003). Glucocorticoid regulation of diverse
cognitive functions in normal and pathological emotional states. Neuroscience
and Biobehavioral Reviews, 27(3), 233-246.

F eldman, S. S., Nash, S. C., & Cutrona, C. (1977). The influence of age and sex on
responsiveness to babies. Developmental Psychology, 13(6), 675-676.

F eldman, S. S., & Nash, S. C. (1978). Interest in babies during young adulthood. Child
Development, 49(3), 617-622.

Fernandez-Duque, E., Valeggia, C. R., & Mendoza, S. P. (2009). The biology of paternal

care in human and nonhuman primates. Annual Review of A nthropology, 38(1),
115-130.

98

F errario, V. F., Sforza, C., Poggio, C. E., & Schmitz, J. H. (1998). Facial volume changes
during normal human growth and development. The Anatomical Record, 250(4),
480-487.

Findlay, J. M., & Gilchrist, I. D. (2003). Active vision: the psychology of looking and
seeing. Oxford; New York: Oxford University Press.

Fitzgerald, H. E., Mann, T., & Barratt, M. (1999). Fathers and infants. Infant Mental
Health Journal, 20(3), 213-221.

Fleming, A. S., Corter, C., Franks, P., Surbey, M., Schneider, 3., & Steiner, M. (1993).
Postpartum factors related to mother's attraction to newborn infant odors.
Developmental Psychobiology, 26(2), 115-132.

Fleming, A. S., & Gonzalez, A. (2009). Neurobiology of human maternal care. In P. T.
Ellison & P. B. Gray (Eds), Endocrinology of Social Relationships (pp. 294—318).
Cambridge, Mass: Harvard University Press.

Fleming, A. S., Ruble, D., Krieger, H., & Wong, P. Y. (1997). Hormonal and experiential
correlates of maternal responsiveness during pregnancy and the puerperiurn in
human mothers. Hormones and Behavior, 31(2), 145-58.

Fleming, A. S., Steiner, M., & Corter, C. (1997). Cortisol, hedonics, and maternal
responsiveness in human mothers. Hormones and Behavior, 32(2), 85-98.

Flinn, M. V., Ward, C. V., & Noone, R. J. (2005). Hormones and the human family. In D.
M. Buss (Ed.), The handbook of evolutionary psychology (pp. 552-580).
Hoboken, NJ: John Wiley & Sons Inc.

F rodi, A. M., & Lamb, M. E. (1978). Sex differences in responsiveness to infants: a
developmental study of psychophysiological and behavioral responses. Child
Development, 49(4), 1182-8.

Frodi, A. M., Lamb, M. E., Leavitt, L. A., & Donovan, W. L. (1978). F athers' and
mothers' responses to infant smiles and cries. Infant Behavior and Development,
1(1), 187-198.

Fullard, W., & Reiling, A. M. (1976). An Investigation of Lorenz's "Babyness". Child
Development, 47(4), 1191-1193.

Galdikas, B. M. F., & Wood, J. W. (1990). Birth spacing patterns in humans and apes.
American Journal of Physical Anthropology, 83(2), 185-191.

Geary, D. C. (2008). Evolution of fatherhood. In C. Salmon & T. K. Shackelford (Eds),
Family relationships: An evolutionary perspective (pp. 1 15-144). Oxford: Oxford
University Press.

99

 

Giardino, J., Gonzalez, A., Steiner, M., & Fleming, A. S. (2008). Effects of motherhood
on physiological and subjective responses to infant cries in teenage mothers: A

comparison with non-mothers and adult mothers. Hormones and Behavior, 53(1),
149-158.

Gilpin, A. R. (1988). Playing vs. caring: Differential socialization for childrearing in
older adolescents. Early Child Development and Care, 34(1), 27-39.

Gilpin, A. R., & Glanville, B. B. (1977). Measuring experience with young children.
Journal of Genetic Psychology, 131(1), 159-60.

Gilpin, A. R., & Glanville, B. B. (1985). Parents' child care experience: effects of sex and
parity. Early Child Development and Care, 18(3), 161-173.

Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W., Gur, R. C., & Sachser,
N. (2009). Baby Schema in Infant Faces Induces Cuteness Perception and
Motivation for Caretaking in Adults. Ethology, 115(3), 257-263.

Glocker, M. L., Langleben, D. D., Ruparel, K., Loughead, J. W., Valdez, J. N., et al.
(2009). Baby schema modulates the brain reward system in nulliparous women.
Proceedings of the National Academy of Sciences, [06(22), 9115-9119.

Gobbini, M. 1., & Haxby, J. V. (2007). Neural systems for recognition of familiar faces.
Neuropsychologia, 45(1), 32-41 .

Going, M., & Read, J. D. (1974). Effects of uniqueness, sex of subject, and sex of
photograph on facial recognition. Perceptual and Motor Skills, 39(1), 109-110.

Gould, S. J. (1980). A biological homage to Mickey Mouse. In The Panda's Thumb:
more reflections in natural history (pp. 95-107). New York: Norton.

Gray, P. B., Parkin, J. C., & Samms-Vaughan, M. E. (2007). Hormonal correlates of
human paternal interactions: a hospital-based investigation in urban Jamaica.
Hormones and Behavior, 52(4), 499-507.

Green, J. A., & Gustafson, G. E. (1983). Individual recognition of human infants on the
basis of cries alone. Developmental Psychobiology, 16(6), 485-493.

Harris, L. J ., Spradlin, M. P., & Almerigi, J. B. (2007). Mothers’ and fathers’ lateral
biases for holding their newborn infants: A study of images from the World Wide
Web. Laterality: Asymmetries of Body, Brain and Cognition, 12(1), 64-86.

Hausfater, G., & Hrdy, S. B. (1984). Infanticide: Comparative and evolutionary
perspectives. Aldine Pub. Co.

Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The distributed human neural
system for face perception. Trends in Cognitive Sciences, 4(6), 223-233.

100

Henderson, J. M. (2003). Human gaze control during real-world scene perception. Trends
in Cognitive Sciences, 7(11), 498-504.

Henderson, J. M. (2007). Regarding Scenes. Current Directions in Psychological
Science, 16(4), 219-222.

Henderson, J. M., Williams, C. C., & Falk, R. J. (2005). Eye movements are functional
during face learning. Memory & Cognition, 33(1), 98-106.

Hess, E. H., & Polt, J. M. (1960). Pupil size as related to interest value of visual stimuli.
Science, 132, 349-50.

Hewlett, B. S. (Ed.). (1992). Father-child relations: Cultural and biosocial contexts.
Foundations of human behavior. New York: Aldine De Gruyter.

Hildebrandt, K. A., & Fitzgerald, H. E. (1977). Gender bias in observers' perceptions of
infants' sex: It's a boy most of the time. Perceptual and Motor Skills, 45, 472-474.

Hildebrandt, K. A., & Fitzgerald, H. E. (1978). Adults' responses to infants varying in
perceived cuteness. Behavioural Processes, 3(2), 159-172.

Hildebrandt, K. A., & Fitzgerald, H. E. (1979a). Facial feature determinants of perceived
infant attractiveness. Infant Behavior & Development, 2(4), 329-339.

Hildebrandt, K. A., & Fitzgerald, H. E. (1979b). Adults' perceptions of infant sex and
cuteness. Sex Roles, 5(4), 471-481.

Hildebrandt, K. A., & Fitzgerald, H. E. (1981). Mothers' responses to infant physical
appearance. Infant Mental Health Journal, 2(1), 56- 61.

Hinde, R. A., & Barden, L. (1985). The evolution of the teddy bear. Animal Behaviour,
33(4), 1371-1373.

Hoss, R. A., Ramsey, J. L., Griffin, A. M., & Langlois, J. H. (2005). The role of facial
attractiveness and facial masculinity/femininity in sex classification of faces.
Perception, 34(12), 1459-1474.

Hrdy, S. B. (1999). Mother nature: a history of mothers, infants, and natural selection.
New York: Pantheon Books.

Hrdy, S. B. (2009). Mothers and others: The evolutionary origins of mutual
understanding. Cambridge, Mass: Belknap Press of Harvard University Press.

Hsiao, J. H., & Cottrell, G. (2008). Two fixations suffice in face recognition.
Psychological Science, 19(10), 998-1006.

Hiickstedt, B. (1965). Experimentelle Untersuchungen zum "Kindchenschema.".
Zeitschriftfu'r Experimentelle und A ngewandte Psychologie, 12(3), 421 -450.

101

Insel, T. R., & Young, L. J. (2001). The neurobiology of attachment. Nature Reviews
Neuroscience, 2(2), 129-136.

Irtel, H. (2007). PXLab: The psychological experiments laboratory (Version 2.1.11)
[computer software]. Mannheim, Germany: University of Mannheim. Retrieved
from Available from <http://www.pxlab.de>

Jackson, L. A. (1992). Physical appearance and gender: Sociobiological and
sociocultural perspectives. Albany: State University of New York Press.

Kaitz, M., Good, A., Rokem, A. M., & Eidelman, A. I. (1987). Mothers' recognition of
their newborns by olfactory cues. Developmental Psychobiology, 20(6), 587-591.

Kaitz, M., Good, A., Rokem, A. M., & Eidelman, A. I. (1988). Mothers' and fathers'
recognition of their newboms' photographs during the postpartum period. Journal
of Developmental and Behavioral Pediatrics, 9(4), 223-226.

Kaitz, M., Lapidot, P., Bronner, R., & Eidelman, A. I. (1992). Parturient women can
recognize their infants by touch. Developmental Psychology, 28(1), 35-39.

Kaplan, H., & Lancaster, J. (2003). An evolutionary and ecological analysis of human
fertility, mating patterns, and parental investment. In K. W. Wachter & R A.
Bulatao (Eds), Offspring: Human fertility behavior in biodemographic
perspective (pp. 170-223). Washington, DC: National Academies Press.

Kellett, R. J. (1992). Infanticide and child destruction--the historical, legal and
pathological aspects. Forensic Science International, 53(1), 1-28.

Kinsey, A. C. (1948). Sexual behavior in the human male. Philadelphia: W. B. Saunders
Co.

Kramer, K. L. (2005). Children's help and the pace of reproduction: Cooperative breeding
in humans. Evolutionary Anthropology, 14(6), 224-23 7.

Kringelbach, M. L., Lehtonen, A., Squire, S., Harvey, A. G., Craske, M. G., Holliday, I.
E., Green, A. L., et a1. (2008). A specific and rapid neural signature for parental
instinct. PLoS ONE, 3(2), el664.

Lancaster, J ., & Lancaster, C. S. (1987). The watershed: Change in parental-investment
and family-formation strategies in the course of human evolution. In J. Lancaster,
J. Altmann, A. S. Rossi, & L. R. Sherrod (Eds), Parenting across the life span:
Biosocial dimensions (pp. 187—205). New York: Aldine de Gruyter.

Landolt, M. A., Lalumiere, M. L., & Quinsey, V. L. (1995). Sex differences in intra-sex
variations in human mating tactics: An evolutionary approach. Ethology and
Sociobiology, 16(1), 3-23.

102

Lang, P. J. (1980). Behavioral treatment and bio-behavioral assessment: Computer
applications. In J. Sidowski, J. Johnson, & T. Williams (Eds), Technology in
Mental Health Care Delivery Systems (pp. 119-137). Norwood, NJ: Ablex Pub.
Corp.

Lang, P. J., Bradley, M. M., & Cuthbert, B. (2008). International affective picture system
(IAPS) .' A jfective ratings of pictures and instruction manual. Technical Report A-
8. Gainesville, Florida: University of Florida.

Langlois, J. H., Ritter, J. M., Casey, R. J., & Sawin, D. B. (1995). Infant attractiveness
predicts maternal behaviors and attitudes. Developmental Psychology, 31(3), 464-
472.

Langton, S. R. H., Honeyman, H., & Tessler, E. (2004). The influence of head contour
and nose angle on the perception of eye-gaze direction. Perception &
Psychophysics, 66(5), 752-71.

Lee, R. D., & Kramer, K. L. (2002). Children's economic roles in the Maya family life
cycle: Cain, Caldwell, and Chayanov revisited. Population and Development
Review, 475-499.

Leibenluft, E., Gobbini, M. 1., Harrison, T., & Haxby, J. V. (2004). Mothers' neural
activation in response to pictures of their children and other children. Biological
Psychiatry, 56(4), 225-232.

Lobmaier, J. S., Sprengelmeyer, R., Wiffen, B., & Perrett, D. I. (2010). Female and male
responses to cuteness, age and emotion in infant faces. Evolution and Human
Behavior, 16-21.

Lorberbaum, J. P., Newman, J. D., Horwitz, A. R., Dubno, J. R., Lydiard, R. B., Hamner,
M. B., Bohning, D. E., et al. (2002). A potential role for thalamocingulate
circuitry in human maternal behavior. Biological Psychiatry, 51(6), 431-445.

Lorenz, K. (1971). Part and parcel in animal and human societies. In Studies in animal
and human behaviour (pp. 115-195). Cambridge, Mass: Harvard University
Press.

Maestripieri, D., & Pelka, S. (2002). Sex differences in interest in infants across the
lifespan - A biological adaptation for parenting? Human Nature, 13(3), 327-344.

Mann, J. (1992). Nurturance or negligence: maternal psychology and behavioral
preference among preterm twins. In The adapted mind: evolutionary psychology
and the generation of culture (pp. 367-390). New York: Oxford University Press.

Marlowe, F. W. (2005). Who tends Hadza children. In B. S. Hewlett & M. E. Lamb
(Eds), Hunter-gatherer childhoods: Evolutionary, developmental, and cultural
perspectives (pp. 177-190). New Brrmswick, NJ: Aldine de Gruyter.

103

Martin, R. D. (2007). The evolution of ' human reproduction: a prirnatological perspective.
American Journal of Physical Anthropology, Suppl 45, 59-84.

McCabe, V. (1984). Abstract perceptual information for age level: A risk factor for
maltreatment? Child Development, 55(1), 267-276.

Mehrabian, A., & Epstein, N. (1972). A measure of emotional empathy. Journal of
Personality, 40(4), 525-543.

Mikach, S. M., & Bailey, J. M. (1999). What Distinguishes Women with Unusually High
Numbers of Sex Partners? Evolution and Human Behavior, 20(3), 141-150.

Minear, M., & Park, D. C. (2004). A lifespan database of adult facial stimuli. Behavior
Research Methods, Instruments, & Computers, 36(4), 630-633.

Morris, P. H., Reddy, V., & Bunting, R. C. (1995). The survival of the cutest: who's
responsible for the evolution of the teddy bear? Animal Behaviour, 50(6), 1697-
1700.

Muller, M. N., Marlowe, F. W., Bugumba, R., & Ellison, P. T. (2009). Testosterone and
paternal care in East African foragers and pastoralists. Proceedings of the Royal
Society: Series B, 276(1655), 347-354.

Murray, L., Fiori-Cowley, A., Hooper, R., & Cooper, P. (1996). The impact of postnatal
depression and associated adversity on early mother-infant interactions and later
infant outcome. Child Development, 67(5), 2512-2526.

Nagy, E., Németh, E., & Molnar, P. (2000). From unidentified to 'misidentified' newborn:
male bias in recognition of sex. Perceptual and Motor Skills, 90(1), 102-104.

Naime, J. S., Pandeirada, J. N. S., & Thompson, S. R. (2008). Adaptive memory: the
comparative value of survival processing. Psychological Science, 19(2), 176-80.

New, J., Cosmides, L., & Tooby, J. (2007). Category-specific attention for animals
reflects ancestral priorities, not expertise. Proceedings of the National Academy of
Sciences, USA, 104(42), 16598-603.

Nitschke, J. B., Nelson, E. E., Rusch, B. D., Fox, A. S., Oakes, T. R., & Davidson, R. J.
(2004). Orbitofrontal cortex tracks positive mood in mothers viewing pictures of
their newborn infants. Neurolmage, 21(2), 583-92.

Numan, M. (1994). A neural circuitry analysis of maternal behavior in the rat. Acta
Paediatrica (Supplement), 397, 19-28.

Ohman, A., F lykt, A., & Esteves, F. (2001). Emotion drives attention: detecting the snake
in the grass. Journal of Experimental Psychology: General, 130(3), 466-478.

104

Penke, L., & Asendorpf, J. B. (2008). Beyond global sociosexual orientations: A more
differentiated look at sociosexuality and its effects on courtship and romantic
relationships. Journal of Personality and Social Psychology, 95(5), 1 1 13-1 135.

Phelps, E. A., Ling, S., & Carrasco, M. (2006). Emotion facilitates perception and
potentiates the perceptual benefits of attention. Psychological Science, 17(4), 292-
9.

Phillips, P. J., Moon, H., Rizvi, S. A., & Rauss, P. J. (2000). The FERET Evaluation
Methodology for Face-Recognition Algorithms. IEEE Transactions on Pattern
Analysis and Machine Intelligence, 22(10), 1090-1104.

Pittenger, J. B. (1990). Body proportions as information for age and cuteness: animals in
illustrated children's books. Perception & Psychophysics, 48(2), 124-130.

Pomplun, M., Sunkara, S., F airley, A., & Xiao, M. (n.d.). Using pupil size as a measure
of cognitive workload in video-based eye-tracking studies. Retrieved from
http://www.cs.umb.edu/~marc/publications.htrn

Porter, R. H., Cernoch, J. M., & McLaughlin, F. J. (1983). Maternal recognition of
neonates through olfactory cues. Physiology & Behavior, 30(1), 151-154.

Portmann, A. (1941). Die Tragzeiten der Primaten und die Dauer der Schwangerschaft
beim Menschen: ein Problem der vergleichenden Biologic. Revue Suisse de
Zoologie, 48, 511—518.

Posner, M. I. (1980). Orienting of attention. The Quarterly Journal of Experimental
Psychology, 32(1), 3-25.

R Development Core Team. (2009). R: A language and environment for statistical ’
computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved
from http://www.R-project.org

Ranote, S., Elliott, R., Abel, K. M., Mitchell, R., Deakin, J. F. W., & Appleby, L. (2004).
The neural basis of maternal responsiveness to infants: an MRI study.
Neuroreport, 15(11), 1825-9. '

Rasband, W. (n.d.). ImageJ (version 1.42q) [Computer software]. U. S. National
Institutes of Health, Bethesda, Maryland.

Rayner, K. (1998). Eye Movements in Reading and Information Processing: 20 Years of
Research. Psychological Bulletin, 124(3), 372-422.

Rayner, K., Li, X., Williams, C. C., Cave, K. R., & Well, A. D. (2007). Eye movements
during information processing tasks: individual differences and cultural effects.
Vision Research, 47(21), 2714-26.

105

 

Revelle, W. (2009). psych: Procedures for Psychological, Psychometric, and Personality
Research (R package version 1. 0-85).

Rhodes, G. (2006). The evolutionary psychology of facial beauty. Annual Review of
Psychology, 5 7, 199-226.

Rhodes, G., Brennan, 8., & Carey, S. (1987). Identification and ratings of caricatures:
Implications for mental representations of faces. Cognitive Psychology, 19(4),
473-497.

Rhodes, G., & Zebrowitz, L. A. (2002). Facial attractiveness: Evolutionary, cognitive,
and social perspectives. Advances in visual cognition, Westport, Conn.: Ablex.

Rholes, W. R., Simpson, J. A., Blakely, B. S., Lanigan, L., & Allen, B. A. (1997). Adult
attachment styles, the desire to have children, and working models of parenthood.
Journal of Personality, 65(2), 357-385.

Rimmele, U., Hediger, K., Heinrichs, M., & Klaver, P. (2009). Oxytocin makes a face in
memory familiar. The Journal of Neuroscience, 29(1), 38-42.

Robinson, C. L., Lockard, J. S., & Adams, R. M. (1979). Who looks at a baby in public.
Ethology and Sociobiology, 1(1), 87-91.

Rossion, B. (2002). Is sex categorization from faces really parallel to face recognition?
Visual Cognition, 9(8), 1003.

Russell, M. J ., Mendelson, T., & Peeke, H. V. S. (1983). Mother's identification of their
infant's odors. Ethology and Sociobiology, 4(1), 29-31.

Sear, R., & Mace, R. (2008). Who keeps children alive? A review of the effects of kin on
child survival. Evolution and Human Behavior, 29(1), 1-18.

Seifritz, E., Esposito, F., Neuhoff, J. G., Ltithi, A., Mustovic, H., Dammann, G., von
Bardeleben, U., et a1. (2003). Differential sex-independent amygdala response to

infant crying and laughing in parents versus nonparents. Biological Psychiatry,
54(12), 1367-1375.

Sellen, D. W. (2007). Evolution of infant and young child feeding: implications for
contemporary public health. Annual Review of Nutrition, 27, 123-148.

Shepherd, M., Findlay, J. M., & Hockey, R. J. (1986). The relationship between eye
movements and spatial attention. The Quarterly Journal of Experimental
Psychology Section A, 38(3), 475-491.

Sherman, G. D., Haidt, J ., & Coan, J. A. (2009). Viewing cute images increases
behavioral carefulness. Emotion, 9(2), 282-286.

106

Shore, D. I., Spence, C., & Klein, R M. (2001). Visual prior entry. Psychological
Science, 12(3), 205-212.

Simpson, J. A., & Gangestad, S. W. (1991). Individual differences in sociosexuality:
evidence for convergent and discriminant validity. Journal of Personality and
Social Psychology, 60(6), 870-83.

Sprengelmeyer, R., Perrett, D. 1., Fagan, E. C., Comwell, R. E., Lobmaier, J. S.,
Sprengelmeyer, A., Aasheim, H. B. M., et al. (2009). The cutest little baby face: a
hormonal link to sensitivity to cuteness in infant faces. Psychological Science,
20(2), 149-154.

Stallings, J., Fleming, A. S., Corter, C., Worthman, C., & Steiner, M. (2001). The effects
of infant cries and odors on sympathy, cortisol, and autonomic responses in new
mothers and nonpostpartum women. Parenting, 1 (1&2), 71-100.

Stanislaw, H., & Todorov, N. (1999). Calculation of signal detection theory measures.
Behavior Research Methods Instruments and Computers, 31 , 137-149.

Stemglanz, S. H., Gray, J. L., & Murakami, M. (1977). Adult preferences for infantile
facial features: An ethological approach. Animal Behaviour, 25(1), 108-115.

Storey, Walsh, Quinton, & Wynne-Edwards. (2000). Hormonal correlates of paternal
responsiveness in new and expectant fathers. Evolution and Human Behavior,
21(2), 79-95.

Sui, J ., & Liu, C. H. (2009). Can beauty be ignored? Effects of facial attractiveness on
covert attention. Psychonomic Bulletin & Review, 16(2), 276-281.

Sussman, G. D. (1977). Parisian infants and Norman wet nurses in the early nineteenth
century: A statistical study. Journal of Interdisciplinary History, 7(4), 637-653.

Swain, J. E., Lorberbaum, J. P., Kose, S., & Strathearn, L. (2007). Brain basis of early
parent-infant interactions: psychology, physiology, and in vivo functional
neuroimaging studies. Journal of Child Psychology and Psychiatry, and Allied
Disciplines, 48(3-4), 262-87.

Tinbergen, N. (1951). The study of instinct. Oxford: Clarendon Press.

Titchener, E. (1908). Lectures on the elementary psychology of feeling and attention.
New York: Macmillan.

Torralba, A., Oliva, A., Castelhano, M. S., & Henderson, 1. M. (2006). Contextual
guidance of eye movements and attention in real-world scenes: the role of global
features in object search. Psychological Review, 113(4), 766-86.

US. Bureau of Labor Statistics. (2008). Occupational Employment and Wages. Retrieved
December 30, 2009, from http://www.b1s.gov/news.release/ocwage.toc.htrn

107

Ursi, W. J., Trotman, C. A., McNamara, J. A., & Behrents, R. G. (1993). Sexual
dimorphism in normal craniofacial growth. The Angle Orthodontist, 63(1), 47-56.

Valentine, T. (1991). A unified account of the effects of distinctiveness, inversion, and
race in face recognition. Quarterly Journal of Experimental Psychology: Human
Experimental Psychology, 43, 161—204.

Valentine, T., & Bruce, V. (1986). The effects of distinctiveness in recognising and
classifying faces. Perception, 15(5), 525 — 535.

Vuilleumier, P., & Schwartz, S. (2001). Emotional facial expressions capture attention.
Neurology, 56(2), 153-158.

Waller, K. L., Volk, A., & Quinsey, V. L. (2004). The effect of infant fetal alcohol
syndrome facial features on adoption preference. Human Nature, 15(1), 101-117.

West, G. L., Anderson, A. A. K., & Pratt, J. (2009). Motivationally significant stimuli
Show visual prior entry: evidence for attentional capture. Journal of Experimental
Psychology: Human Perception and Performance, 35(4), 1032-1042.

Weston, E. M., Friday, A. E., & Lio, P. (2007). Biometric evidence that sexual selection
has shaped the hominin face. PLoS ONE, 2(1), e710.

Wiesenfeld, A. R., & Klorman, R. (1978). The mother's psychophysiological reactions to
contrasting affective expressions by her own and an unfamiliar infant.
Developmental Psychology, 14(3), 294-304.

Wiesenfeld, A. R., Whitman, P. B., & Malatesta, C. Z. (1984). Individual differences
among adult women in sensitivity to infants: evidence in support of an empathy
concept. Journal of Personality and Social Psychology, 46(1), 1 18-124.

Wild, H. A., Barrett, S. E., Spence, M. J., O'Toole, A. J., Cheng, Y. D., & Brooke, J.
(2000). Recognition and sex categorization of adults' and children's faces:

Examining performance in the absence of sex-stereotyped cues. Journal of
Experimental Child Psychology, 77(4), 269-291.

Wisner, K. L., Parry, B. L., & Piontek, C. M. (2002). Postpartum Depression. The New
England Journal of Medicine, 34 7(3), 194-199.

Wright, D. B. (2009). sdtalt: Signal detection theory and alternatives (R package version
1. 0-1).

Wright, D. B., Horry, R., & Skagerberg, E. M. (2009). Functions for traditional and
multilevel approaches to signal detection theory. Behavior Research Methods,
41(2), 257-267.

Yarbus, A. L. (1967). Eye movements and vision. New York: Plenum Press.

108

 

Zebrowitz, L. A., Luevano, V. X., Bronstad, P. M., & Aharon, I. (2009). Neural
activation to babyfaced men matches activation to babies. Social Neuroscience,
4(1), 1-10.

Zebrowitz, L. A., & Brownlow, S. (1992). Baby talk to the babyfaced. Journal of
Nonverbal Behavior, 16(3), 143-158.

109