' 1‘ .1 11121171111 1‘ 1‘ 1111~11111111111111|1 11411011111111." 111:111111 1111 '1 1111 1111 1111111 ..~_ 4;; m -_ f "* . E: .— - ~,. #_ 6—7.“ <— 7 I. v >0 9 O . ' ..... . -M...” Kn . ‘ - v 13: ,1 1 u: 3:91, 651 . £111 11 11241111“ ‘ ~111. 31 : M1 11111111111'11 11111 1.1111111"| 111:1" 111 ’ 1 1 11) Ah‘ kirk; ‘1' E’ 1 'i: ‘32 {£1.11 ‘111031; ' ‘ (1111?:11111 11'; (:13 t1.vi\ ,1 - 111111111311 J13,- 11.1‘10 r 0‘: 1 "5.19 3.1. I mu. MW“H“HW\\\\||\\\|\\\l\\\l\\\\|\\8\\\\\\\|\\|m : i. .. 3 1039 .. 3129 [iféhigqn 5U” WCFSJ ty This is to certify that the thesis entitled A CROSS-CULTURAL COMPARISON OF ADULTS' PERCEPTIONS OF INFANT SEX AND PHYSICAL ATTRACTIVENESS presented by Olga Nita Hernandez has been accepted towards fulfillment of the requirements for Ph.D Psychology degree in 0/3.“ {9% Major p/ {Ar 6/11/81 Date 0-7 639 OVERDUE FINES: 25¢ per day per item ' RETURNING LIBRARY MATERIALS: ”ace in book return to remove charge from circulation records a a k. + V: "‘ I JUL” 1 1 2000 A CROSS-CULTURAL COMPARISON OF ADULTS' PERCEPTIONS OF INFANT SEX AND PHYSICAL ATTRACTIVENESS By Olga Nita Hernéndez A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1981 ABSTRACT A CROSS-CULTURAL COMPARISON OF ADULTS' PERCEPTIONS OF INFANT SEX AND PHYSICAL ATTRACTIVENESS By Olga Nita Hernandez Previous research has attempted to assess the relative influence of variations in physical attractiveness. The present study was designed to make a cross-cultural comparison of adults' perceptions of infant sex and physical attractiveness. Four hundred eighty subjects equally divided by culture and sex participated in this study. Two sets of 14 photographs of infants faces equally divided by culture, sex, age, and differential cuteness ratings were used as the stimuli. Reactions to the photographs were measured using a 5-point Likert scale for physical attractiveness and a 5-point dimension scale for perceived sex. The sample was recruited from Michigan State University and the Uni- versity of Puerto Rico - Mayagfiez Campus and was divided into four equal sized groups: American and Puerto Rican subjects rating American infants; American and Puerto Rican subjects rating Puerto Rican infants. Within each group, half of the subjects viewed each slide for 8 seconds while the remaining half viewed each slide for 15 seconds. Stimulus presentation varied randomly. Testing was conducted in regular classrooms using mixed sex groups. Repeated measures ANOVA revealed that Puerto Rican raters gave higher cuteness ratings to U.S. infants than U.S. raters did. Nine-month-old P.R. infants received higher cuteness ratings than 9-month-old U.S. infants did. A significant effect was found for infants' differential cuteness, indi- cating that high-cute infants received higher cuteness ratings than low- cute infants did. A significant interaction between exposure time and raters' culture indicated that P.R. raters gave higher cuteness ratings with 15 seconds exposure time than U.S. raters did. Analysis of perceived sex revealed that when infants were divided by sex, raters of both cultures were fairly accurate in guessing the infants' sex. The results are com- pared with those of other studies of adult attraction to infants. A MIS PADRES ii ACKNOWLEDGMENTS I wish to thank, Dr. Hiram E. Fitzgerald, my chairperson, for the assistance and support that he gave me on this dissertation as well as throughout my graduate career. I also wish to acknowledge the other com- mittee members, Dr. Ellen Strommen, Dr. John P. McKinney, and Dr. Elaine Donelson, for their professional insight and assistance. Special thanks are expressed to Gaby Belli and Dr. Jack Condon, for their statistical advice and for their patient assistance in helping with data analysis. Special appreciation is extended to the parents of the infants who served as stimuli for this research. Their cooperation made this study possible. Also I would like to extend sincere thanks to all professors, graduate assistants, and students of Michigan State University and the University of Puerto Rico - Mayaguez Campus who made the successful com- pletion of this work possible. I wish to thank my parents, Olga and Victor Hernandez, for their support, love, and encouragement. To my brother and sister, Vitin and Zeny, thanks for many things. Finally, I wish to express my deepest gratitude to my friend, Aracely, for the endless hours of work put into this project, without her help, encouragement, love and understanding this dissertation would not now have been completed. iii CHAPTER 1. CHAPTER 2. CHAPTER 3. CHAPTER 4. CHAPTER 5. APPENDIX A. APPENDIX B. APPENDIX C. REFERENCES TABLE OF CONTENTS INTRODUCTION REVIEW OF LITERATURE Physical Attractiveness Research The Physical Attractiveness Stereotype Adults Perceptions and Behaviors toward Children Varying in Attractiveness Childrens' Perceptions and Behaviors Toward Peers Varying in Attractiveness Adults' Responsiveness to Infants Effects of Infants' Physical Attractiveness on Adult Responses Summary of Purpose and Hypotheses METHOD Subjects Stimuli U.S. Infant Photographs P.R. Infant Photographs Dependent Variables Design of the Study Procedure RESULTS Cuteness Ohter Significant Findings Perceived Sex Other Significant Findings Summary of Results DISCUSSION Tests of Hypotheses Additional Findings Conclusions and Implications for Future Research iv 13 24 26 26 27 28 29 3O 3O 33 35 37 39 45 47 49 52 52 57 62 63 7O 77 96 A-l A-2 C-1 C-2 C-3 c-a C-lO C-ll LIST OF TABLES Sample Size of Raters' Sex and Culture by Infants' Sex Mean Ages and Standard Deviations of Raters Sex by Infants' Age Culture, Sex, Differential Cuteness and Age Split of Infants' Photographs Overall Cuteness Mean Ratings and Differential Cuteness Ratings of American Infants Variable Codes Design Significant Main Effects on Cuteness Ratings and Perceived Sex Rating Overall Mean Cuteness Ratings Cuteness Ratings of Photographs Chosen for Use in Subsequent Study Analysis of Variance - Cuteness Rating Analysis of Variance - Perceived Sex Rating Means for ANOVA Cuteness Rating Means for ANOVA Cuteness Rating Means for ANOVA Cuteness Rating Means for ANOVA Cuteness Rating Means for ANOVA Cuteness Rating Means for ANOVA - Raters' Culture x Infants' Differential Cuteness - Cuteness Rating Raters' Culture x Infants' Culture - Raters' Culture x Exposure Time - Infants' Culture x Infants' Sex - Raters' Culture x Infants' Age - Infants' Culture x Infants' Age - Means for ANOVA - Infants' Culture x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Raters’ Sex x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Infants' Sex x Infants' Age- Cuteness Rating 26 27 28 29 31 32 36 67 68 77 78 79 79 79 79 80 80 80 80 81 C-13 C-l4 C-15 C-16 C-17 C—18 C-19 C-20 C-21 C-22 C-23 C-24 C-25 C-26 C-27 C-28 C-29 C-30 C-31 C-32 Means for ANOVA- Infants' Sex x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Infants' Age x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Raters' Culture x Infants' Culture x Infants' Age - Cuteness Rating Means for ANOVA - Raters' Culture x Raters' Sex x Infants' Age - Cuteness Rating Means for ANOVA - Raters' Culture x Raters'Sex x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Infants' Culture x Infants' Sex x Infants' Age - Cuteness Rating Means for ANOVA - Infants' Culture x Infants' Sex x Infants' Differential Cuteness- Cuteness Rating Means for ANOVA - Raters' Culture x Infants' Age x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Infants' Culture x Infants' Age x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Raters' Sex x Infants' Age x Infants' Differential Cuteness - Cuteness Rating Means for ANOVA - Infants' Sex x Infants' Age x- Infants' Differential Cuteness - Cuteness Rating Means for Perceived Means for Perceived ANOVA - Raters' Culture x Infants' Culture Sex Rating ANOVA - Infants' Culture x Exposure Time - Sex Rating Means for ANOVA - Raters' Culture x Infants' Sex- Perceived Sex Rating ‘ Means for ANOVA- Exposure Time x Infant's Sex - Perceived Sex Rating Means for ANOVA- Raters' Sex x Infants' Sex - Perceived Sex Rating Means for Perceived Means for ANOVA- Infants' Culture x Infants' Age - Sex Rating ANOVA - Infants' Culture x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA - Raters' Sex x Infants' Differential Cuteness - Perceived Sex Rating Means for Perceived Means for ANOVA- Infants' Sex x Infants' Age- Sex Rating ANOVA- Infants' Sex x Infants' Differential Cuteness - Perceived Sex Rating vi 81 81 82 82 83 83 84 84 85 85 86 86 86 87 87 87 88 88 88 89 89 C-33 C-34 C-35 C-36 C-37 C-38 C-39 C-40 c-41 C-42 C-43 C-44 Means for ANOVA- Infants’ Age x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA- Raters' Culture x Infants' Culture x Infants' Sex - Perceived Sex Rating Means for ANOVA- Raters' Culture x Exposure Time x Infants' Sex - Perceived Sex Rating Means for ANOVA - Infants' Culture x Raters' Sex x Infants' Sex - Perceived Sex Rating Means for ANOVA - Infants' Culture x Raters' Sex x Infants' Differential Cuteness Means for ANOVA - Raters"Culture x Infants' Sex x Infants' Age - Perceived Sex Rating Means for ANOVA - Infants' Culture x Infants' Sex x Infants' Age - Perceived Sex Rating Means for ANOVA - Raters' Sex x Infants' Sex x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA- Raters' Culture x Infants' Sex x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA - Raters' Culture x Infants' Age x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA - Infants' Culture x Infants' Sex x Infants' Differential Cuteness - Perceived Sex Rating Means for ANOVA - Infants' Sex x Infants' Age x Infants' Differential Cuteness - Perceived Sex Rating vii 89 90 9O 91 91 92 92 93 93 94 94 9S 10. 11. 12. LIST OF FIGURES . Raters' Culture x Infants‘ Culture for Mean Cuteness Rating Infants' Culture x Infants' Sex x Infants' Age for Mean Cuteness Rating . Raters' Culture x Exposure Time for Mean Cuteness Rating Raters' Culture x Infants' Culture x Infants' Age for Mean Cuteness Rating Raters' Culture x Raters' Sex x Infants' Age for Mean Cuteness Rating . Raters' Culture x Raters' Sex x Infants' Differential Cuteness for Mean Cuteness Rating Infants' Culture x Infants' Sex x Infants' Differential Cuteness for Mean Cuteness Rating Infants' Sex x Infants' Age x Infants' Differential Cuteness for Mean Cuteness Rating Raters' Culture x Infants' Culture for Mean Perceived Sex Rating Raters' Culture x Infants' Sex for Mean Perceived Sex Rating Raters' Culture x Infants' Culture x Infants' Sex for Mean Perceived Sex Rating Raters' Culture x Infants' Sex x Infants' Age for Mean Perceived Sex Rating viii 37 38 4O 41 42 43 44 45 46 47 48 49 CHAPTER 1 INTRODUCTION Currently, there is an expanding literature concerning factors that influence adults' attributions of infant attractiveness. To date, however, few studies have involved samples other than those with American infants and adults. Therefore, the purpose of the present study was to conduct a cross-cultural comparison of adult's perceptions of infant perceived sex and physical attractiveness. Samples of college students from Michigan State University and the University of Puerto Rico judged the cuteness and perceived sex of photographs of both Puerto Rican and American infants. To provide the rationale and background of the study the following topics will be considered: the importance of physical attractiveness for social interactions, most relevant research studies for this investigation, and the issues to be addressed as well as the need for the study. The importance of studying the effects of physical attractiveness on social interactions has been well documented (Adams, 1977; Berman, 1980; Berscheid & Walster, 1974; Hildebrandt, in press). An examination of reported findings supports the notion of the presence of a physical attrac- tiveness stereotype summarized by the expression, "what is beautiful is good" (Dion, Berscheid & Walster, 1972). In general, research on the physical attractiveness stereotype suggests that physical appearance affects several types of evaluations in adults and children. Recently, developmental psychologists have concentrated on adults' reactions to infants' physical attractiveness ( Hildebrandt & Fitzgerald, 1977, 1978, 1979a, 1979b). In spite of the recency of this field of research there already is considerable diversity in the approaches taken by researchers for studying the impact of physical attractiveness in adult-infant interaction. Most experimental studies have used photographs of infants as stimuli (e.g. Hildebrandt & Fitzgerald, 1979a, 1979b), few studies have used live infants (Hildebrandt, 1980). The interest in infants physical attractiveness and its possible influence on adult—infant relations led Hildebrandt and Fitzgerald (1977, 1978, 1979a, 1979b) to develop a series of studies designed to determine the parameters of perceived infant cuteness and to assess individual dif- ferences in adults' reactions to infant physical attractiveness. Their findings have shown that there are specific physical features which identify a baby as cute and also that behaviors such as looking and smiling are related differentially to the perceived cuteness of the infant. Moreover, adults' cuteness ratings of infants have been found to be related to the infants' facial features, facial expression, birth order, age, and sex, as well as the mode of presentation of the stimuli (i.e. either live or in photographs). Some of these relationships are weak, with variation occurring from study to study. For example, in one study no sex differences were found in average cuteness ratings, although women used the ends of the cuteness scale more than men did (Hildebrandt & Fitzgerald, 1979a). In light of the increasing work done on the effects of infants' physical attractiveness and the possible implications for adult-infant and caregiver-infant interaction, it is necessary to examine the topic in more detail. For example, most investigations in this area have been limited to one culture and one racial-ethnic group (i.e. American). Therefore, generalization to other cultures is constrained. In addition, past studies have focused parametric manipulations on aspects of the physiognomy of the infant's facial features with little attention to other aspects of the stimulus situation such as stimulus frequency or exposure duration. The present study was designed to expand research on infant physical attractiveness in two ways. First, this study examined the effects of culture specificity on adults' reactions to infant physical attractiveness and perceived sex. Second, the study was designed to permit examination of the effects of stimulus exposure time on ratings of physical attractive— ness and perceived sex. The specific hypotheses to be tested in this study are presented in chapter 2. CHAPTER 2 REVIEW OF LITERATURE Physical Attractiveness Research Since Berscheid and Walster (1974) pointed out the importance of physical appearance for interpersonal attraction, numerous investigators have demonstrated the impact of physical attractiveness in a wide variety of settings. For example, there is accumulating evidence that the physical attractiveness stereotype in adults can be generalized across differing samples, contexts, and settings (Adams, 1977). Studies with adults have reported that physical attractiveness influences person perception (Sigall & Landy, 1973), heterosexual liking (Berscheid, Dion & Walster & Walster, 1971), and attribution of personal characteristics (Miller, 1970). The Physical Attractiveness StereOtype The vast majority of studies in this topic have verified the existence of a "physical attractiveness stereotype" (i.e. the idea that physically attractive individuals are attributed more positive traits than unattrac- tive individuals) in both adutls and children (Dion, 1972, 1973; Dion & Berscheid, 1974). Collectively, research findings are consistent with those of Dion, Berscheid and Walster (1972) who found that the physical attractiveness stereotype operates along the lines of "what is beautiful is good". Adults'Perceptions and Behaviors toward Children Varying in Attractiveness In the past, the effects of children's perceived attractiveness on adult-child interactions was not of great concern to developmental psychol- ogists. Only recently have studies been directed to the possibility that adults might perceive and treat children differently as a function of perceived attractiveness ( Hildebrandt, in press). In these investigations, the typical methodology involves showing a photograph of a child who is either high or low in physical attractiveness, to an adult who is asked to rate the child on a variety of characteristics such as cuteness, gender, race, IQ, home background, and school performance. Two studies by Dion (1972, 1974) provide support for the assumption that physical attractiveness of the child may influence adults' behavior. Using a photograph of a child (previously judged to be attractive or unattractive) attached to a behavioral description of a mild or severe transgression by the child, Dion (1972) found that attractive children were rated as less deviant when the transgression was severe, whereas the unattractive children were judged as antisocial and more dishonest. However, in this study there were no differences in the punishment advocated by female college students for the offenses committed by these children. In contrast, a subsequent study by Dion (1974) reported a cross-sex effect in that adult females were more punitive toward attractive girls and unattractive boys than toward attractive boys; the effect did not appear for adult males. Dion's explanation of these results was based on the assumption that women and men differ in their degree of task orientation. She proposed that men may be more task oriented while women may be more interpersonally oriented. Therefore, women may be more influenced than men by such social cues as sex and attractiveness. A number of studies have found that teachers' expectations about a child's performance are related to the child's degree of perceived attrac- tiveness. In general, these investigations have shown that teachers consistently rate attractive children more favorably, although in some cases perceived attractiveness interacted with other child characteristics (Adams, 1978; Adams & Cohen 1976a, 1976b; Adams & Crane, 1980; Clifford, 1975; Keble, Bramble & Mason, 1974). For example, several studies have indicated that such variables as child behavior, may reduce the effects of physical attractiveness on teacher's expectations (Adams & Cohen, 1976b; Adams & La Voie, 1974a, 1974b; La Voie & Adams, 1974). While evidence is accumulating to suggest that adults are likely to develop an evaluative tendency toward children based on a physical attrac- tiveness stereotype, little research has been conducted to determine if parents are influenced by the attractiveness stereotype when interacting with their own children. Since parents are a major influence in child socialization, it seems reasonable to suggest that early development of the "beauty is good" stereotype will affect their interactions with their children. Adams and La Voie (1975) reported that children's physical attractiveness influenced parents predictions of the child's personal- social success. More recently, Adams and Crane (1980) found that parents' expectations of their children's behavior were consistent with a "beauty is good" stereotype in both social attribution and social preference measures. Not all studies have demonstrated that attractive children are perceived and treated differently by adults than are unattractive children. For instance, La Voie and Adams (1974) found that attractiveness appeared to exert little influence on teachers' ratings of a child after teachers' were given information about the child's general behavior. In another study no attractiveness effect was found when teachers rated a child's personality after having read the child's essay (Keble et al., 1974). Moreover, Dion (1972, 1974) found no effects for physical attractiveness on adult's ratings of children who committed mild transgressions compared to those who had not transgressed. In addition, results from a few studies suggest that physically attractive children are treated exactly opposite to expectations derived from a physically attractive stereotype. For example in one study teachers rated the transgressions of attractive children as more severe than the transgressions of unattractive children (Marwit, Marwit & Walker, 1978). In another study teachers rated the work habits of highly attractive children lower than those of moderate or low attractiveness (Adams & La Voie, 1974). These results suggest that there may be other characteristics (e.g. personality trait differences, level of education) which may over- ride any effects due to perceived attractiveness. In summary, literature reviewed up to this point supports the physical attractiveness stereotype in adult-child relationships, although several studies point out that the effect may be reduced by giving more concrete information about the child's behavior, performance, or personality. Children's Perceptions and Behaviors Toward Peers Varying in Attractiveness Studies of peer relations suggest that children's perceptions of unfamiliar peers also are influenced by physical attractiveness. Both Dion (1973) and Dion and Berscheid (1974) have demonstrated that children's preferences for each other are affected by the physical attractiveness stereotype, with attractive children liked better than unattractive ones. Langlois and Stephan (1977) found that children rated attractive peers as smarter, friendlier, nicer and less mean than their less attractive peers. Attractive children are expected to behave more prosocially and are more likely to be chosen as potential friends than are unattractive children (Dion, 1973; Langlois & Stephan, 1977; Styczynski & Langlois, 1977). Particularly interesting is a study conducted by Trnavsky and Bakeman (1976) who attempted to test individual differences in the strength of the physical attractiveness stereotype in children. They found that more attractive children held the attractiveness stereotype more strongly than less attractive children. Although there is little direct observational research concerning children's differential treatment of attractive or unattractive peers several studies have reported that more attractive children have more friends and receive more social acceptance (Cavior & Dokecki, 1973; Cavior, Miller & Cohen, 1975; Dion & Berscheid, 1974; Kleck et al., 1974; Lerner & Lerner, 1977; Salvia, Sheare & Algonzine, 1975). Styczynski and Langlois (1977) demonstrated that acquaintance has a significant effect on preschool Childrens' popularity ratings and behavioral expec- tations in relation to gender as well as attractiveness. They interpreted their results as suggesting that when the effects of physical attractive- ness are examined within real social situations, attractiveness may indeed be a social disadvantage, especially with boys. Results of this study showed that unacquainted boys rated attractive children as more pOpular than unattractive children, although the reverse was true for acquainted boys (i.e. they selected unattractive children as more popular). These investigators suggested that boys may be negatively influenced by attrac- tiveness when they are rating familiar peers, since the attractive child may be allowed special privileges and may behave more antisocially thus reducing the child's popularity with male peers. Adult's Responsiveness to Infants The ethologist Lorenz (1943) proposed a detailed hypothesis dealing with adult's responsiveness to the young. Lorenz suggested that certain specific and typical babyish features of infants elicit caregiving and approach behavior from human adults. This hypothesis, which emphasized the physiognomic features of the infant, stimulated much contemporary research concerning specific facial characteristics that define "babyish- ness". In recent years considerable attention has been directed to the study of adult's responses to infants (Barman, Cooper, Mansfield, Shields, & Abplanalp, 1975; Fullard & Reiling, 1976; Hess & Polt, 1960). As suggested by Feldman and Nash (1978, 1979a, 1979b), responsiveness to infants is an important psychological phenomenon that has cultural significance because a society must nurture its young to maturity in order to survive. Since the prevailing assumption has been that child- bearing responsibilities are linked primarily to females, few attempts have been made to study behaviors in males that may be influential in regulating male-infant interactions (Bem, 1974; Bem, Martyna & Watson, 1976). Berman's (1980) critical review of studies which have examined adults responsiveness to infants, suggest that the literature fails to resolve the question of whether males and females differ in their responsiveness to infants. Furthermore, she points out that such contemporary research depends upon pictorial representation of infants, possibly because of 10 Lorenz' emphasis on the physical features of infants. In contrast, very little work examines adult's responses to a live baby. Adults' responses to infants have been measured in a number of ways. The majority of studies equate interest in infants with adult's self reports such as statements of preference or ratings of attractiveness. Others have used behavioral measures such as observer's records or ratings of participant's overt behaviors. Finally, the least frequent studies are those which have employed physiological measures such as skin conduc- tance, heart rate, pupillary response, and blood pressure. Collectively, the findings reported from numerous studies do not provide straightforward support for sex differences in adults' responsiveness to infants. As emphasized by Berman (1980), studies employing different measures have produced conflicting results. For example, studies using ratings of attractiveness and/or behavioral measures have reported (1) that females are more reactive (Berman, 1976; Frodi, Lamb, Leavitt, Donovan, Neff & Sherry, 1978; Fullard & Reiling, 1976) (2) that males show more interest in babies (Sternglanz, Gray & Murakami, 1977) and (3) that both sexes are equally responsive (Berman, Ablanab , Cooper, Mansfield, & Shields, 1975). Conversely, studies which have measured physiological responses have reported different findings, for instance, Frodi et a1. (1978) did not find sex differences in heart rate and galvanic skin response. In contrast, Hess and Polt (1960) reported that females are more reactive than males when viewing pictures of babies, although in this study the dependent variable was pupillary dilatation. A problem in comparing and interpreting these contradictory findings stems from the fact that one cannot assume that the measures are equivalent. As Berman (1980) has noted, the type of stimuli presented in the studies 11 may contribute significantly to the differences found between studies. In general the most typical methodology for self-report studies involves the presentation of infant pictures which limits the subject to respond only to the infant's physical features and expression. On the contrary, behavioral studies have employed live infants which present the participant with a variety of stimuli such as clothing and behavior. Berman further suggests that adult's responses to infants vary depending on the response required from the participant (i.e. self-report, behavioral and/or phys- iological), the subjects experiential or prior rule relationship to infants (e.g. mothers vs. fathers), as well as physical and social qualities of the situation (e.g. laboratory vs. natural setting and individual vs. group). Although the collective data on adult's responses to infants fails to demonstrate sex differences in responsiveness, the majority of self- report studies consistently support findings indicating that adults prefer pictures of infants, and that these tendencies may be stronger in females than in males (e.g. Berman, 1976; Cann, 1953; Feldman & Nash, 1978; Fullard & Reiling, 1976; Sternglanz, Gray & Murakami, 1977). For instance, Fullard and Reiling (1976) demonstrated that women show stronger verbal preferences for infants than men do. In addition, Cann (1953) reported that regard- less of their marital or parental status women preferred infant pictures more often than men did. However, men who are fathers or whose wives are pregnant prefer baby pictures more often than other men do. These partic- ular findings support the ethological hypothesis that babyishness elicits positive responses from adults. In keeping with the earlier concern regarding the infant's stimulus characteristics, several studies have investigated the parameters of infant physical features which adults define as babyish, attractive, or 12 cute. Considerable attention has been given to facial features which distinguish infants from adults. Thus, for example, Gardner and Wallach (1965) and Hess (1970) found that infantile facial features include a high and protruding forehead, large eyes placed in the middle of the face, a small nose and fat cheeks. Interestingly enough, most infants are babyish in their physical appearance. However, there are individual differences among infants in those characteristics which define babyishness. Consequently an assumption derived from ethological theory is that infants whose features are more infantile or "babyish" should be preferred and might elicit more positive responses from adults than infants whose features are less "babyish" (Sternglanz, Gray & Murakami, 1977). It seems that infants' facial feature variations influence adults' responses to infant attractiveness. Using line drawings of infants, and varying eye position, Brooks and Hochberg (1960) reported that eyes positioned in the center of the face received higher cuteness ratings than eyes positioned higher or lower than the center. Another study of facial feature variations which manipulated several features such as eye width, eye height, and iris size, indicated that college students rated as more attractive the faces which had relatively large eyes and a relatively large forehead (Sternglanz, et al.,1977). However, as suggested by Hildebrandt and Fitzgerald (1978) inter- pretation of Sternglanz et al. (1977) findings requires consideration of the fact that drawings of infant faces were used and also that the range of feature variation was larger than one might anticipate in a group of real infants. Besides, this study failed to consider the inter- relationship among features which one might expect to be very important when judging a real infant's physical attractiveness. 13 In an attempt to determine which facial features college students considered important in their judgments of infant cuteness, Hildebrandt (1976) asked subjects to mention which aspects of the photographs were most important in helping them to decide which photographs were cuter. Facial expression, eyes, hair, fatness, facial proportions, and ears were the features most frequently mentioned. A subsequent study was conducted to investigate if cuteness ratings could be predicted from objectively measurable facial features (Hildebrandt & Fitzgerald, 1979b). In contrast to previous studies which used drawings of infants in order to study how facial feature variation might influence adult behavior, this study used actual infant photographs. Using a multiple correlational approach it was found that facial feature combi- nations were predictive of infant's perceived cuteness. It was reported that a cute infant is likely to have a large forehead, large eyes and pupils and to have short and narrow features. The results of Hildebrandt and Fitzgerald's study served to support Lorenz' hypothesis that there are a number of characteristics identified as babyish which might evoke positive responses from adults. Effects of Infants' Physical Attractiveness on Adult Responses There are theoretical and empirical reasons for hypothesizing that adults apply a physical attractiveness stereotype to infants. Theo— retically, Lorenz (1943) was the first to suggest that particular babyish facial features elicit approach behavior from human adults. Although there is a widespread interest in adults' responses to infants, the amount of available evidence concerning the effects of physical attrac- tiveness is rather limited. Moreover, the importance of an infant's 14 physical attractiveness on adult's responses and behaviors has been questioned by several investigators (Hildebrandt & Fitzgerald 1977, 1978, 1979a, 1979b, 1981; Power, Hildebrandt & Fitzgerald, in press). Nevertheless, a series of studies has been conducted over the past years which have examined the role of infant's physical appearance in eliciting behaviors and attitudes from adults (Hildebrandt & Fitzgerald 1977, 1978, 1979a, 1979b, 1981). The majority of these investigations have employed infant's facial photographs selected from a set of 60 chromatic slides as the stimuli. These photographs were taken by a pro- fessional photographer under controlled conditions when the infants' facial expressions were judged to be relatively neutral. A grey cape covered the infants' shoulders in order to remove clothing cues. In only one study did participants rate live infants (Hildebrandt, 1980). Typically the response required from subjects in most of these studies are ratings of infant cuteness using a 5-point scale in which a 1 = not very cute, 2 = less cute than average, 3 - average cuteness, 4 a more cute than average and 5 = very cute. Also in some cases, participants have been asked to guess the sex of each infant or to rank order subsets of the photographs. In general, variation in cuteness ratings has been linked to particular characteristics of the infants being rated and to the characteristic of the adults doing the rating. One particularly interesting study investigated the relationship between infant's physical sex and adult's perceptions of infants' physical attractiveness (Hildebrandt & Fitzgerald, 1979a). Five experiments were conducted with college students who rated the cuteness and/or sex of female and male infants ranging in age from 3 to 13 months. In all experiments 60 photographs were projected one at a time via slides while participants marked their ratings of cuteness and/or sex on computer 15 answer sheets. In four of the experiments slides were presented for 8 seconds each while in the remaining experiment slides were presented for 15 seconds each. The results of these studies indicated that subjects had little difficulty assigning a gender label to infants although in some cases the assigned label was incorrect. Moreover, it was found that both age and perceived sex of the infant were related to the cuteness ratings. Specifically, it was demonstrated that older infants received higher cuteness ratings than younger infants, showing that cuteness ratings increased with age, reaching a peak at 11 months. However, infant sex interacted with age indicating that female infants received their highest ratings at 11 months whereas male infants received their highest ratings at 13 months. The authors interpreted this difference as being due to the developmental maturity of female infants in comparison to same age male infants. Perhaps two of the most interesting findings in these studies were that when sex was unknown female infants received slightly higher ratings than males and cuter infants were more likely to be guessed as female than were the less cute infants. 0n the other hand, when the sex was known (i.e. experimenter told subjects either correct or incorrect sex), labeled males received higher ratings than labeled females. Interestingly enough, this finding was interpreted as supporting the hypothesis that adults seem to have a sex stereotype regarding infant cuteness which suggests that female infants are expected to be cuter than male infants. For example, subjects are more likely to attribute a female sex to a cute infant, but if they are told that a particular infant is female rather than male, they judge it less cute because they have higher standards for females. Furthermore, results also revealed that there were no sex differences 16 in the average ratings. Nevertheless, the cuteness ratings given by female subjects differentiated among the infants more than males did. This difference was interpreted as possibly being due to the differential use of the cuteness ratings scale. Specifically it was found that female subjects used the ends of the rating scale (1 and 5) more than male subjects and used the middle of scale (3) less than male subjects. To summarize, results clearly support the prOposition that adults expect females to be more physically attractive than males. As previously indicated, this effect was evident in both perceived and actual sex of the infants although the difference in cuteness ratings between perceived males and perceived females was larger than the difference between actual males and females. The effects obtained, however, were not simple ones, since they occasionally varied from experiment to experiment and also with sex of the adult subject. For instance, the results indicating that female infants are expected to be more physically attractive than are male infants, suggested a relationship between perceived sex and perceived cuteness. Given that the difference was larger when comparing by perceived infant sex than when comparing by actual infant sex, this particular result supports the conclusion that there is a relationship between infant's sex and adults' perceptions of an infant's physical attractiveness. However, these results should be viewed with some caution since findings regarding the direction of the impact of infant's sex on per— ceived attractiveness have not been straightforward. For example, although the Hildebrandt & Fitzgerald (1979a) findings suggest that cuter infants are more likely to be perceived as female, there also were a number of highly cute infants who were perceived as male, as well as a number of less cute infants who were perceived as female. As suggested by Hildebrandt 17 and Fitzgerald, probably perceived cuteness most strongly affects per- ceived sex when insufficient information (e.g. lack of hair in an infant) is provided to make a sex attribution. On the other hand, knowing the sex of the infant does not substantially change the cuteness ratings an infant receives. Indeed, high cute infants were similarly rated when sex was both known and unknown. The authors note that actual or perceived sex of an infant is an influential but not absolute determinant of its perceived cuteness. Subsequent investigations by these researchers have tried to determine what other factors may influence adult's perceptions of infant's cuteness. In one study, Power, Hildebrandt and Fitzgerald (in press) used photographs of smiling and crying infants in order to investigate how infant's affective expression may influence adult's judgments of infant's physical attractive- ness. According to their findings, cuteness ratings are also influenced by facial expression with smiling infants receiving higher ratings than crying infants. However, facial expression neither substantially increases the cuteness ratings of an unattractive infant nor substantially lowers the rating of a highly attractive infant. Thus, these results seem to suggest that facial expression alone does not make an infant attractive. Although these results cannot be generalized to other populations without further study, this investigation served to amplify Hildebrandt's and Fitzgerald's conclusion that cuteness ratings of infants might be influenced by a wide variety of factors. As it has been pointed out previously, most of the studies by Hilde- brandt and Fitzgerald have investigated the effects of several variables on adult's responses to variations in static infant physical attractiveness. Recently, Hildebrandt (1980) conducted a study in which both students and 18 parents rated live 3 month old infants rather than photographs of infants. In contrast to previous studies using photographs, college students gave higher ratings to live infants. This difference may be explained by the fact that the photographs of the infants had no clothing cues since a grey cape covered all infants' shoulders and that facial expression, overt behavior as well as vocalization which may have influenced the ratings of the subjects. One surprising result of this study was that birth order was found to influence the ratings of both parents and college students. The findings revealed that first born infants received higher cuteness ratings from their parents and also from college students who did not know the infants' birth order when rating them. As suggested by Hilde- brandt, first born infants tend to be better dressed and groomed than later born infants which may be a possible explanation for this particu- lar finding. Taken together, findings from the previous series of studies by Hildebrandt and Fitzgerald have demonstrated that characteristics of the stimuli such as the infant's age, sex (both actual and perceived) facial features, facial expression and birth order, as well as the mode of presentation (either a photograph or live) are related to the cuteness ratings given to a particular infant. However, research employing college student samples have indicated few relationships between rater characteristics and their responses to infant's physical attractiveness. In general, no sex differences have been shown in the average ratings, although women use the ends of the cuteness rating scale more than men do (Hildebrandt & Fitzgerald, 1979a). In addition, characteristics such as prior contact with infants, marital 19 status, and occupation have not been found to be related to average ratings. Moreover, results also have demonstrated that not all subjects agree on the cuteness of a particular infant. Further attempts have been made to investigate the correlates of individual differences in ratings of infants. TWO investigations have indicated that parents give higher ratings to their own infants than college students do (Hildebrandt, 1980; Hildebrandt & Fitzgerald, 1981). The interpretation given to these find- ings suggests that the increase in parents' ratings may be related to repeated exposure or familiarity to their infants. For instance, similar findings were reported in a study by Corter, Trubub, Boukydis, Ford, Celhaffer and Minde (1978) who found that a sample of nurses gave higher ratings to infants with whom they had had more contact. Nonetheless, it seems that a common definition of attractiveness can be identified as suggested by the fact that ratings of mothers and col- lege students are moderately correlated (Hildebrandt, 1980). Across dif- ferent groups of college students, a high correlation has been found between ratings of infant's photographs ranging from 3 to 13 months and the rank ordering of infant's photographs of 4 and 8 months old infants (Hildebrandt & Fitzgerald, 1978, 1979a). In addition, it has been reported that a substantial agreement exists among college students; judgments of 3-month-old infants (Hildebrandt, 1978). Despite the fact that a common definition of attractiveness may be suggested by the previous findings, further research is needed to determine why some adults rate a particular infant as high in cuteness whereas other adults rate the same infant as low in cuteness. Indeed, to a large extent, attractiveness is in the eye of the beholder. In addition, research by 20 Hildebrandt and Fitzgerald points to several additional factors to be taken into consideration when investigating adult's judgments of an infant's physical attractiveness. One factor not considered by previous investigators concerns the extent to which cultural background may affect judgments of infants' perceived attractiveness. Specifically in the area of adult's responses to infant attractiveness, research studies have focused only in one culture of both rater (adult) and ratee (infant). According to Werner (1979) cross-cultural studies contribute a comparative perspective to development and extend the range of our knowledge and understanding. Furthermore, she suggests that this type of study best illustrates the universal sequences of human behavior that we share as a species as well as the diversity of behavior that is adaptive in a wide range of environments. Several reasons may be cited to justify a cross-cultural comparison in this area of research. First, as emphasized recently by Leiderman, Tulkin and Rosenfeld (1977) a major rationale for conducting a cross- cultural study is its value as an "eye opener" since it is a way of sharpening perceptions and suggesting new ideas and hypotheses. Second, as pointed out by Rohner (1977) one of the main reasons for conducting cross—cultural work is to "test for the level of generality of a theory or proposition". Based on the assumption that the primary goal of cross- cultural research is the testing of and addition to theory, it is necessary that we attempt to determine the limits within which explanatory concepts and theories are applicable as well as the kind of modifications that have to be made in order to make them universal. Therefore, conducting a cross-cultural study is a way of testing how previous findings may be generalized to other cultural groups and, thereby, to the species in general. 21 Given that few studies have examined adult's responses to infant's attractiveness using a cross-cultural approach, it seems necessary to briefly summarize some of the findings of cross-cultural research which have examined childrens' reSponses to infants. As emphasized by Berman, (1980) two cross-cultural studies have been conducted regarding childrens' responsiveness to infants. In one study Konner (1975) reported that in the !Kung San culture older sisters engaged more in face to face interaction with their siblings than brothers did. In another study Whiting and Whiting (1975) examined childrens' responsiveness to infants in several cultures. According to their findings, the proportion of childrens' responses to infants varied widely for different cultures. Although the data in this area remains inconclusive, the few cross- cultural studies cited above have illustrated that there are both simi- larities and differences one might expect when conducting a cross-cultural study. One major approach in the field of cross-cultural psychology had been testing specific hypotheses drawn from the general body of psycho- logical knowledge in order to elucidate the applicability of various generalizaitons. As suggested by Brislin, Lonner and Thorndike (1973) cross-cultural psychology is the empirical study of members of various culture groups who have had different experiences that lead to predictable and significant differences in behavior. Based on Lorenz' theory which emphasizes species typical reactions to "babyishness" features, one would predict no significant differences in cuteness ratings as a function of rater culture. Obviously this prediction requires empirical confirmation or rejection. Nevertheless, my own experiences -- however subjective they may be-- suggest that dif- ferences in cuteness attributions and perceived sex ratings will be found 22 between Puerto Rican and American raters. One reason for this expected differences is that in Puerto Rico white skin type generally is considered more physically attractive than medium or darker skin type. Moreover, in Puerto Rican culture there is a tendency to use earrings, bright color clothing, and hair bows to identify female infants. Therefore, using a grey cape which predominantly is not a "female infant" color and removing earrings (pierced ears) and other female infant's clothing cues may affect the ratings of Puerto Rican female infants by Puerto Rican raters more than it affects American raters. Finally, cultural differ- ences in sex role orientation is another factor that may contribute to cultural differences in physical attractiveness attributions. Generally in Puerto Rico the role of motherhood is greatly emphasized to woman and childrearing is considered to be the female's responsibility. Therefore Puerto Rican women may be more attracted to infants than are Puerto Rican men and consequently one might expect that Puerto Rican female raters will rate infants differently than men will. Based on the previous discussion, and being aware that a series of studies by Hildebrandt and Fitzgerald have been conducted only in one culture, it is the main purpose of this study to extend their research using culture of both adult and infant as a variable in addition to several factors which they have investigated. Therefore, the influence of an infant's culture, sex, age and differential cuteness, as well as the adult's sex and culture were assessed on adult's perceptions of infant perceived sex and physical attractiveness: Comparisons were made across cultures of raters and of infants in order to determine if there were cultural differences among the groups. Furthermore, comparisons also were performed within culture of both raters and ratee in order to 23 determine whether there were significant differences within groups. Finally, a second purpose of this study was to determine whether duration of stimulus exposure significantly affects adult's ratings of infants. Previous research by Hildebrandt and Fitzgerald (1979) used exposure durations of 8 or 15 seconds. It was the aim of this study to add this new variable (i.e. time of exposure) in order to determine if manipulation of exposure time influenced adult's perceptions of infant's perceived sex and physical attractiveness. According to Zajonc (1968) attitudes may change in a positive di- rection by increasing the amount of exposure to a stimulus. Several "mere exposure" hypothesis (Christensen, investigations have supported this 1970; Litvak, 1969; and Saegert, Swamp, & Zajonc, 1973). Harrison and Zajonc (1970) presented 12 novel stimuli (Chinese ideographs) to 66 subjects with varying frequencies of zero to 25 expo- sures. Half of the stimuli were shown for 10 seconds per exposure and half for 2 seconds per exposure. Their findings revealed that response competition decreases with increasing frequency of exposure, while affec- tive ratings increased. On the other hand, response competition also was found to be reduced by an extended duration of exposure but duration did not have an entirely clear effect upon affective ratings. In another study, Marcus and Hakmiller (1975) used three conditions in which they manipulated the total duration of exposure, frequency, and duration of study trial to determine which of these factors affect emo- tional judgments in college students. Results indicated that when total duration of exposure was held constant, there were no differences between the attractiveness ratings of the slides of female nudes when the other two factors were manipulated. On the other hand, when the total duration 24 increased as a function of frequency of exposure, the female subjects showed an increase in liking for the slides. In addition, when the total duration increased as a function of duration of study trial, males showed a change in affective judgments in a positive direction. In conclusion, these researchers indicate that if total duration of exposure increases and duration of study trial or frequency increases concurrently, increases in judgments of attractiveness will occur. Two different conditions were used in the present study to determine whether a "mere" exposure effect was exhibited by adult subjects in the perception of infant's physical attractiveness. The conditions varied the total duration of stimulus exposure in which half of the subjects saw each slide for 8 seconds while the remaining half saw each for 13 seconds. Based on the assumption that an increased exposure time should provide the subjects with a better opportunity to single out distinctive characteristics of the infant, it was expected that significant differences could be found in the infants' ratings as a function of exposure time. Summary of Purpose and Hypotheses The literature investigating adults' perceptions of infants' physical attractiveness and perceived sex is sparce and inconclusive. Based on the literature reviewed, there are important research questions that remain unaswered. For example, do people of different cultural backgrounds give infants similar or different ratings of physical attractiveness and perceived sex? Do infants of different cultural backgrounds receive similar or different ratings from adult raters? Does exposure time sig- nificantly affect the ratings of infants? The present study was designed to investigate these questions. Specifically this study has attempted 25 to test the following hypotheses: Hypothesis I There will be significant differences between U.S. and P.R. raters with respect to cuteness ratings as a function of infants' culture. Hypothesis II There will be significant differences between U.S. and P.R. infants with respect to cuteness ratings as a function of infants' sex and infants' age. Hypothesis III There will be significant differences between low and high cute infant groups (ratings determined by previous research) with respect to cuteness ratings. Hypothesis IV There will be significant differences between 8 seconds and 15 seconds exposure time groups with respect to cuteness ratings. Hypothesis V There will be significant differences between U.S. and P.R. raters with respect to perceived sex ratings as a function of infants' sex. Hypothesis VI There will be significant differences between U.S. and P.R. raters with respect to perceived sex ratings as a function of infants' culture. Subjects CHAPTER 3 METHOD A total of 480 college students equally divided by sex and culture participated in this study. Table 1 shows a summary of the sample size for each group according to the sex and culture of raters as well as culture of infants. 1. U.S. raters 2. U.S. raters 3. P.R. raters 4. P.R. raters Table 1 Sample Size of Raters' Sex and Culture - U.S. - U.S. by Infant's Culture infants infants infants infants Males 60 6O 60 6O Females 60 6O 60 60 Total N = Total 120 120 120 120 480 Subjects were recruited from introductory psychology courses at Michigan State University and at the University of Puerto Rico - Mayagfiez Campus. Table 2 presents mean ages and standard deviations of raters' sex by infants' culture. 26 27 Table 2 Mean Ages and Standard Deviations of Raters Sex by Infants' Culture Raters' Sex Group Males Females U.S. raters - U.S. infants 20.20 19.92 (3.95) (1.23) U.S. raters - P.R. infants 20.45 20.18 (5.29) (1.19) P.R. raters - U.S. infants 20.96 20.46 (1.59) (3.29) P.R. raters - P.R. infants 20.80 19.70 ‘ (1.01) (2.38) Stimuli Color photographs of infant faces were used in the study. There were two sets of 12 photographs equally divided by sex, age and differen- tial cuteness of infant for each culture (See Table 3). Using a Kodak Carousel slide projector, photographs were presented via slides to mixed sex groups. 28 Table 3 Culture, Sex, Differential Cuteness, and Age Split of Infant's Photographs Culture U S. P.R. Age Male Female Male Female Ono.) I Low High Low High Low High Low High 3 9 13 nj 3 one per cell U.S. infant photographs. One set of 12 photographs (6 male, 6 female) was selected from a larger collection of infant photographs on the basis of differential cuteness ratings (i.e. low cute - high cute) obtained in previous studies (Hildebrandt & Fitzgerald, 1977, 1978). An infant with a rating below the mean (i.e. 2.75) was considered a low-cute infant while an infant with a rating above the mean was considered as high cute. Table 4 shows the overall mean of the cuteness rating for each infant as well as its differential cuteness rating. 29 Table 4 Overall Cuteness Mean Ratings and Differential Cuteness Ratings of American Infants Infant # Infant Age and Overall Cuteness Differential Sex Mean Rating Cuteness Rating 1 3 M 2.990 H 3 3 M 2.515 L 4 3 F 3.199 H 2 3 P 2.429 L 6 9 F 3.577 H 8 9 F 2.184 L 7 9 M 3.102 H 5 9 M 2.112 L 12 13 F 3.663 H 10 13 F 2.301 L 9 13 M 2.791 H 11 13 M 2.571 L These pictures were taken by a professional photographer under con- trolled conditions when the infants' facial expression was judged to relatively neutral. Also to eliminate clothing cues the shoulders of the infants were covered with a grey cape. P.R. infant photographs. In order to obtain the set of 12 photo- graphs of Puerto Rican infants and to match the photographs for physical attractiveness across sex and nationality; a procedure similar to that used by Hildebrandt and Fitzgerald (1978) was used to select infants. 30 (Refer to Appendix A for information on how stimuli for this study were developed). Dependent Variables The selection of cuteness and perceived sex ratings for the dependent measures was based primarily upon the work of Hildebrandt and Fitzgerald (1976, 1977, 1978, 1979a). Cuteness rating scale. Students rated infant physical attractiveness according to a five point, Likert-type scale in which: 1 = not very cute, 2 - less cute than average, 3 - average cuteness, 4 = more cute than average and 5 - very cute. Sex rating scale. Subjects rated perceived sex of infants along a dimension ranging from 1 (male) to 5 (female). Design of the study The study was designed to test certain assumptions about the dif- ferences and similarities in cuteness and perceived sex ratings of infants between subjects from Puerto Rico and United States. Tables 5 and 6 show the variable coding used for the data analysis and the design of the study, respectively. In summary, 480 subjects equally divided by cultures participated in this investigation. One half of the total sample (120 U.S., 120 P.R. raters) were asked to judge the physical attractiveness and to guess the sex of U.S. infants while the remaining half rated P.R. infants. In all experiments half of the subjects equally divided by sex viewed the infant's slides for 8 seconds while the other half viewed it for 15 seconds. 31 This procedure yielded a 2 x 2 x 2 x 2 x 2 x 2 x 3 factorial design. The design over subjects was a 2 (culture of rater) x 2 (sex of rater) x 2 (culture of infant) x 2 (time of exposure), whereas the design over measures was a 2 (sex of infant) x 2 (previous cuteness) x 3 (age of infant). Table 5 Variable Codes Stimuli Levels Code Infant's Culture (B) 2 U.S. 1 P.R. 2 Infant's Sex (E) male 1 female 2 3 months 1 0 Infant 5 Age (F) 3 9 months 2 13 months 3 Differential Cuteness (G) 2 low 1 high 2 Subjects Raters' Culture (A) 2 U.S. 1 P.R. 2 Exposure Time (C) 8 seconds 1 15 seconds 2 Raters' Sex (D) 2 male 1 female 2 Dependent Variables Cuteness ratings 1 - 5 Perceived Sex ratings 1 - 5 32 Table 6 Design Male Infants Female Infants Ages Ages 3 9 13 3 9 13 Cuteness Cuteness L H L H L H L H L H L Raters' Culture Infants' Culture Time Exposure Ratersr’Sex U.S. 15 U.S. P.R. 15 U.S. 15 P.R. F P.R. 15 Design over subjects: 2 x 2 x 2 x 2 = 16 N = 480 Design over measures: 2 x 3 x 2 = 12 33 Procedure The experimental procedure was first explained to the subjects, questions were answered, and signatures were obtained using the appropriate version (i.e. English or Spanish) of the standard Department of Psychology Research Consent Form (Refer to Appendix B). All the materials needed for the Spanish speaking samples were translated from the English version using the back translation method (Brislin, 1970). The general procedure for all the experiments was the following: Each participant observed 28 infant photographs each for 8 or 15 seconds, in blocks of 14 with a 5 second interslide interval and a 20 second blank slide interval between blocks. Slides were shown two times in the same random ordered group of 12 with 2 filler slides inserted at the beginning to make a total of 28. The purpose of presenting 2 filler slides at the begin- ning was to familiarize subjects with the type of stimuli to be used in the study. The slides were projected on a screen for mixed-sex groups after the following instructions were given; You will be seeing: 24 photographs of infant faces. The infants range in age from 3 to 13 months. In part I each slide will be on screen for 8 (15) seconds. During the time they are on rate each one according to the following scale: very cute more cute than average average cuteness less cute than average not very cute Nw-L‘U'l 1" There will be 2 seconds between slides. 34 In part II you will be seeing 12 photographs for 8 seconds each and this time you will judge whether the infant is male or female according to the following scale: 1 male 2 3 4 5 female Do not write your name or student number on the computer answer sheet. Participation in this experiment is strictly voluntary. Your responses will be anonymous and you may withdraw from the experiment at any time. Each slide will be on screen for 8 (15) seconds. You should not rate the first two pictures, they are just to show you what the infants look like and how much they vary in cuteness. Please be as quiet as possible during the experiment so you don't influence the ratings of your neighbors. Do you have any questions?. Remember, do not rate the first two pictures you see. I will tell you when to begin. If you sign the informed consent forms and hand them to the front we will get ready to begin (Refer to Appendix B for Spanish version). tn: on; 886 w a :Y' to 11105 the 1'95- Ana CHAPTER 4 RESULTS Pearson correlation analysis indicated that cuteness ratings and perceived sex ratings were not significantly correlated ( r - .0019). Therefore, the data for each dependent variable was analyzed by means of a seven-way analysis of variance with repeated measures (BALANOVA). Comparisons of mean differences were conducted for significant two- and three-way interactions using simple effects tests with p set at <.05. Only those ANOVA components that reached the p <.05 level of significance are reported separately for each dependent variable. Due to the com- plexity of the design and the large number of main effects and interactions only the most meaningful features of the data are reported. As can be seen in Tables C-1 and C-2 (Appendix C) several four-, five-, and six- way interactions reached statistical significance. However, attempts to explain four-, five-, or six-way interactions not only would be tedious, but also would be spurious and lack meaning. For example, it is axiomatic that statistical significance is not necessarily correspondent with scientific meaningfulness (O'Brien & Shapiro, 1968). Therefore only the most salient features of the data are emphasized in order to formulate the most coherent and lucid presentation of these data. The most salient results are those directly related to hypothesis and selected others inso- far as they reflect on cross-cultural comparisons. The results are presented in several sections. A summary of the Analysis of Variance for each dependent measure is presented in Tables 35 36 C-1 and C-2. A posteriori mean comparisons were conducted for both sig- nificant main effects and interaction effects. A least significant dif- ference test was performed on all significant main effects with p set at <.01. Table 7 illustrates the significant main effects mean comparisons for both dependent measures. Means for significant two— and three-way interactions are presented in Tables 0-3 to C-44 (Appendix C). The majority of simple simple main effects for three-way interactions were significant at the .05 level. Table 7 Significant Main Effects on Cuteness Rating and Perceived Sex Rating Variables Cuteness Perceived Sex Raters' Culture (A) U.S. P.R. U.S. P.R. 2.77* 2.82* 2.40* 2.45* Infants' Culture (B) U.S. P R. U.S. P R. 2.89* 2.75* 2.14* 2.71* Raters Sex (D) Males Females Males Females 2.75* 2.90* 2.54* 2.31* Infants' Sex (E) Males Females Males Females 2.89* 2.76* 2.06* 2.79* Infants' Age (F) 3 mo. 9 mo. 13 mo. 3 mo. 9 mo. 13 mo. 2.72* 2.83* 2.91* 2.30* 2.22* 2.76* Infants' Differ- ential Cuteness (G) Low-Cute High-Cute Low-Cute High-Cute 2.73* 2.91* 2.45* 2.40* Note: Those horizontal mean comparisons which are asterisked in each of the variable categories were significant at the .01 level. 37 Cuteness The first hypothesis to be tested predicted differences between U.S. and P.R. raters as a function of infant culture. Support for this hypoth- esis was indicated by a significant raters' culture x infants' culture interaction, F (1, 464) = 19.56, p. <.0005. Sinple effects tests for all comparisons depicted in Figure 1 revealed that subjects from Puerto Rico gave higher ratings to U.S. infants [F (l, 464) - 20.49, p <.01] and lower ratings to P.R. infants than did subjects from the U.S. The second hypothesis tested predicted differences in cuteness ratings for U.S. and P.R. infants as a function of their sex and age. A significant three-way interaction [F (2, 928) = 42.87, p <.0005) ] provided support for this hypothesis (see Figure 2). . mu' CULTURE x llffllll’c' cumin: no ' 3" roe comm Rama 5'1flflikflflflm '*“'PIIUHIRO 3.8 4 2.8 . 920-04" “CHIN-0:0 33M: 0 I / 1.8 INFRIIO' CULTURE 38 no. 2 Ith' cum I: want so: I: Iwm' not run cum Rattle 4 ‘F'900llllihflfllfl3 a ""g:lII1:llhlfl1t. a 3" - ‘F'.PIIHUILIJDI1IUI I c 0 I a. if I I ‘-*- 1 2.8 a I I T .1 ‘ 2 . to! I ' i 8 QUIT". OIIITHO 18 EDITH! "III“? ID! The simple effects analysis for this interaction showed that all mean comparisons were significant except for comparisons across infant' age for Puerto Rican male infants and between sexes for U.S. 13 month-old- infants. In addition, there were no significant differences in how U.S. and Puerto Rican male infants were rated at either 9 months or 13 months. As can be seen in Figure 2, both U.S. male and female 9-month—old were rated lower than were infants at the other two age groups. Overall, 9-month-old U.S. female infants were given higher cuteness ratings than 9-month-old U.S. male infants. In contrast, both male and female P.R. infants received higher cuteness ratings at 9 months than at other ages. Interestingly enough, P.R. males were perceived as cuter than P.R. female infants at all ages. To summarize the interaction depicted in Figure 2, U.S. female 39 infants were seen as cuter at 3 months whereas U.S. male infants were seen as cuter at 13 months. On the other hand, P.R. male and female infants were seen as cuter at 9 months. There was support for the hypothesized relationship between infants' differential cuteness and cuteness ratings, F (1, 464) = 36.777, p <.0005. An a posteriori least significance test was performed for these means indicating that high-cute infants received overall higher cuteness ratings than low-cute infants. Finally, there was no support for the hypothesized effect due to exposure time, F (l, 464) 8 .726, p <.395. Revelant means compared in the ANOVA were 2.80 for 8 seconds and 2.75 for 15 seconds. Other significant findings: Several main effects of interest also are reflected in Table 7. A main effect was found for raters' culture, F (l, 464) = 4.892, p <.027 as well as for infants' culture, F (l, 464) a 7.932, p <.0005. Inspection of cell means indicated that P.R. raters gave the highest ratings and U.S. infants received higher ratings. Moreover, overall sex differences were found for raters' sex, F (1, 464) I 8.806, p <.003 and for infants' sex, F (1, 464) 8 28.998, p <.0005. As indicated by the cell means in Table 7, female raters gave infants higher ratings whereas male infants received the highest ratings. There also was a significant main effect for infants' age , F (2, 298) = 23.101, p <.0005, revealing that 13-month-old infants were rated higher than 3 or 9-month-old infants. Although an exposure time main effect was not confirmed, a raters' culture by time-of—exposure interaction was significant, F (l, 464) = 4.023, p <.045. Sinple effects tests for mean comparisons depicted in Figure 3 40 showed that P.R. raters judged infants to be higher in cuteness, F (l, 464) = 8.89, p <.01 during 15-seconds exposure times than during 8-second exposure times. No such effects were found for U.S. raters. r10. 9 RatrRa‘ entrant x txrcauuc tine run CUIENEBO RBIIIO 4 a-ntua Refer: a l*"PRlUHtR$ 5 3.8 . R c 'r’ s " _. 5 ”1' ' 2.5 . R R r in o z . 1.8 1 I e arccuoa is accouoa EXPONWHETTNE An interesting triple interaction of raters' culture by infants' culture by infants' age was significant (shown graphically in Figure 4), F (2, 928) = 7.768, p <.0005. Sinple simple main effects tests revealed that all except four mean comparisons were significant. The exceptions were tests for (a) raters' culture and 3-month-old P.R. infants; (b) raters' culture and 9-month-old U.S. infants; (c) U.S. raters by U.S. infants across the three age groups, and (d) P.R. raters' by 9-month-old P.R. and U.S. infants. The interaction depicted in Figure 4 primarily reflects a significant difference in how U.S. and P.R. raters evaluated the cuteness of 3- and 41 13-month-old U.S. infants, [F (l, 464) = 21.57, p <.01; F (1, 464) = 34.35, p <.01, respectively]. Inspection of the cell means (Table C-14) revealed that P.R. raters gave higher cuteness ratings at both ages. Moreover, a significant difference was found for Puerto Rican ratings of both U.S. and P.R. 3-month—old infants, F (l, 464) = 48.04, p <.01 (See Figure 4). no.4 Rarm‘ CULTURE x INFRNTS' CULTURE x Iuraurs' HOE ‘ FOR ntnu nurtures RRIIRo Hus RarrRs us Inmate " Hue RRrrRe PR INFBHTB 5 Mn RRrrRs us Iuraurs g 3.5 - twrn RRIrRs PR Iurnuts c U .k d,/”’. T ‘s 0” It! 3 4 ~.~."\ 0":- l"1 a c .«>"’=§~" __- 5 ._____'..¢’—g—-£i A .22-" a M... R 208 .. t a T .5 o 2 4 s 1'5 r* I l : nouns 9 mums 13 mum ‘ Israurs' Roe Another interesting triple interaction was found between raters' culture as a function of raters' sex and infants' age, F (2, 928) = 4.0007, p <.019. Analysis for simple simple main effects depicted in Figure 5 indicated that all possible mean comparisons were significant except for (a) raters' culture for both 3 and 9-month-olds, for male raters, (b) for raters' sex of U.S. raters with 3 and 9-month-old infants, and (c) for 42 raters' sex of P.R. raters with 9-month—old infants. rIo.s RthRo' cULIURr x RntrRa orx x Iurnura’ not run ness cuteness Rntruo 6 HUB MILE MYERS [1 H08 FEHGLE RBTERI E ‘*“PRI¥KIIMHIRS a 3-5 - Hm mm: name It 5 f 3 .. ._. / 5 Nfiw"fj W t -*”" "' _il‘ 8 3'..— 8 208 u: R R T .1 O 2 .. "5 I l ‘1 3 HONTHO 8 HONTHS 13 "001118 Iuraurs' not As shown in Figure 5, this interaction primarily was accounted for by Puerto Rican female raters who gave higher ratings to 3-month-old infants and also by U.S. female ratings of infants at all ages. A significant triple interaction also was found for raters' culture by raters' sex by infants' differential cuteness, F(1, 464) - 9.473, p <.002. Simple effects tests showed that one major component of the interaction was the difference in ratings given to low and high cute infants by U.S. female raters [F (l, 464) a 45.55, p <.01] and between P.R. male ratings of low and high cute infants [F (1, 464) 8 5.17, p <.01]. As can be seen in Figure 6, female and male raters of both cultures gave higher ratings to high cute infants. However, there is a discrepancy in the perception of low-cute infants since P.R. females gave higher ratings to low-cute infants than U.S. female raters. Moreover, a similarity in ratings for high-cute infants is observed for both culture group females. no.0 um’ mm x‘am’ an x m' onmrm ‘ CUHBI3." CUHDIICIHNTIO ‘F"UCIflfl£ MINI" n Hue mu norm g “"PR “NJIIITENI . 3.3 4 mm mm um: I C 'i a s ‘ *— ¢ s——n 3 108 k j I T .‘a 0 I . ". I l IJIICQTI IflIlCflTE Im' 01mm comm Another significant triple interaction was found for infants' culture by infants' sex by infant differential cuteness, F (l, 464) = 115.371, p <.0005. Simple effects tests revealed that there were significant dif- ferences among the ratings of high-cute female infants between infants' culture. Conversely, no significant sex differences were found in the ratings given to P.R. and U.S. low-cute infants. Interestingly enough, this interaction was primarily accounted for by the infant sex effect for high cute P.R. female infants and for the infants' culture effect for high-cute female infants. As can be seen in Figure 7 the high-cute P.R. 44 female infants were rated significantly lower than high-cute P.R. male infants and also significantly lower than high-cute U.S. female infants. '10.? INFBHTB' CULTURE X INFRNTS'SEX X INFHNTS' OIFFERENTIRL CUTENESB FOR BERN CUTENESS RRTINO 4 Hum CUTE rm: INFRNTS n HLOH cur: FEHHLE Iurnurs 5 Hum CUTE MILE Iurnurs : 3-5 . --tniou cure FEflflLE Iurnura c 'r' c a . N r 3 2.5 . R a T .5 o 2 ~ 1.5 turnura' CULIURE Finally, an interesting interaction between infants' sex by infants' age by infants' differential cuteness was found, F (2, 928) 8 170.69, p <.0005. Simple effects tests were significant for most of the mean comparisons except the following: for infants' sex of 3- and 13-month-old high-cute infants, and for differential cuteness of 13-month-old male infants. As represented in Figure 8 a significant pattern emerged for male and female infants across age groups for low and high cute infants. For both low-cute female infants and high-cute male infants, ratings reached a peak at 9 months. In contrast, for low-cute male infants and high-cute female infants ratings were the lowest at 9 months. The comparisons of high-cute male and female infants as well as the comparison 45 of low-cute male and female infants was significant in the 9 month infants' age group. Moreover, a significant difference was found between ratings of 3-month-old low-cute infants wherein low-cute male infants received higher ratings than female low-cute infants. no.0 1m: ex 2: to not I: m‘ oImmL ‘ cu m m cum um» Hm curt m mm" 2 Hunt can an: rm 9 . Hum curt lulu “III" a - - Huron curt m 0mm c ‘r’ c 3 " n z 2 ‘0‘ q R a 1' .1 o 5 - 1.8 T f I a write 0 mm I: name Imu' no: Perceived Sex A significant rater's culture by infants' culture interaction was significant for perceived sex rating, F (l, 464) 8 4.366, p <.037, providing support for the hypothesized difference between U.S. and P.R. raters in regard to perceived sex ratings. Simple effects tests revealed that except for raters' culture by U.S. infant, all the other comparisons were found significant. Cell means for this interaction are shown in Table C-23 and as seen in Figure 9, U.S. infants were perceived as being more male by both raters' cultures. On the other hand, P.R. infants 46 were perceived by both raters' culture as being more female. Within raters' culture it can be observed that U.S. raters considered their own culture infants as being more male while P.R. infants were rated as more female. On the other hand P.R. raters considered their own culture infants as being more female while the U.S. infants were considered more male. no.9 mm‘ mm x Imura' m m mam an an'i'o‘m 4 2 Hot rm: ,1 Ha rm. 9.: .. P c R E I 3 - v F, a R "' r 2.3 . x : .— . r z .. I R o m r l “8 '8 mm‘ cum Hypothesis six was supported by a significant effect for raters' culture by infants' sex, F (l, 464) a 11.385, p <.0001. Simple effects tests indicated that except for the differences between raters' culture for male infants all possible comparisons were found significant. As Figure 10 illustrates there is a similarity between both cultures when guessing the sex of the infants since actual females are seen as more female and actual males are perceived as more male. However, P.R. raters were more accurate in general in their attributions of infant's sex. 47 fTflnfllllflllfl CILUNIEXIHIIINI It! IIIlilllflflnlflflmllflllltnflb "*HIII‘HIMITI 'F'.FEII£IHIWIWI 3.3 .4 3 J ’0‘ .1 Ciro-IS” NRC BM<~MflIMW I'M: 1.8 r P! 8.4 nm‘ cums: Other significant findings- Once again significant main effects were found for infants' culture, [F (l, 464) - 121.045, p <.0005] revealing that U.S. infants were perceived as being more male and P.R. infants as being more female. (Refer to Table 7 for means relevant to main effects). In addition, an overall difference was found for raters' sex, F (l, 464) = 19.618, p <.0005. This overall sex difference indicates that male raters perceived infants to be more male and female raters perceived infants as being more female. Significant main effects also were found for infants' sex, F (l, 464) 8 281.912, p <.0005, and for infants' age, F (2, 928) - 98.110, p <.0005. Looking at the means for infants' sex one can observe that male infants were rated as being more male and that female infants were perceived as being more female. Concerning the means for infants' age 48 overall, younger infants (i.e. 3 and 9 months) were rated as being more male while older infants were perceived as being more female. The triple interaction of raters' culture, by infants' culture by infants' sex was significant, F (l, 464) = 6.011, that all possible comparisons were significant except for U.S. and P.R. raters by U.S. male infants. As illustrated in Figure 11 U.S. male infants as well as P.R. female infants were more accurately identified. FIG-11 MTERC' CULTURE X IlFflNTS' CULTURE X TNFMTS'SEX FOR HER“ PERCEIVED SEX RHTINO 9 H g Hus nets INFRNTB N ”US FEHRLE "‘73."! 3.5 .. MPH ML: INFRNTS P *"‘PR “MILE INFRNTS E c a - ----...._..—c--—-—I-¢-OCD¢. r *"""’ I z 3.0.0 o 2.8 '1 .WM“... me 3 I“ E X 2 . R \I H I I ‘05 -I N 0 l RarcRs' cuLch Finally, the interaction among raters' culture, infants' sex and infants' age also was significant, F (2, 928) = 8.294, p <.0005. (Relevant means for this interaction are shown in Table C-38, Appendix C.). As shown in Figure 12 simple simple main effects revealed that the following points were not significantly different. Within raters' cultures at 3 months, male and female infants were perceived similarly; 9-month-old 49 females and males were perceived similarly by raters from each culture. Summarizing the results of this interaction it seems that older infants are more accurately labeled than 3—month-old infants by raters from each culture. no.1: RnrcRo' outrun: x Inrnnto'srx x Iurnnro' no: son mu PERCEIVEO sex Rnnno Hus RarrRs MILE Inmate I'm-us RnrrRo rEnRLE Inmate ’ 3.5 . em." RntrRa MILE Inmate NPR RRIrRo rrnntr INFANTS 205 q GIN-132 xnu °M