fixing Annotated Bibliography essay
http://www .tandf.co.uk/journals/pp/01650254.html DOI: 10.1080/01650250042000078
Gender stereotyping in infancy: Visual preferences for and knowledge of gender-stereotyped toys in the second year
Lisa A. Serbin, Diane Poulin-Dubois, Karen A. Colburne, Maya G. Sen, and Julie A. Eichstedt Concordia University, Montreal, Canada
Infants’ visual preferences for gender-stereotyped toys and their knowledge of stereotyped toys were examined in two experiments using an adaptation of the preferential looking paradigm. Girls and boys aged 12, 18, and 24 months were tested for their preference for photos of vehicles or dolls, and for whether they associated (‘‘matched’’) these two stereotyped sets of toys with the faces and voices of male and female children. Results of Experiment 1 (N = 77) demonstrated signi� cant preferences for gender stereotyped toys appearing by 18 months of age. In Experiment 2 (N = 58), girls were able to associate the gender-stereotyped toys with girls’ and boys’ faces by 18 months of age, but boys were not. Implications for theories of early gender development are discussed.
Preferences for stereotyped play activities are regarded as one of the earliest, most pervasive, and developmentally consistent manifestations of gender roles in children. In most traditional societies, gender-stereotyped play activities during early child- hood are accepted and encouraged as useful preparation for adult roles. In increasingly complex and technological modern societies, in which work and social roles are rapidly evolving, there is less acceptance of the concept that children should be encouraged to focus their play on traditionally stereotyped activities in preparation for gender-differentiated adult roles. Accordingly, there is both scienti� c and popular interest in identifying and understanding the origins of gender-stereo- typed play.
Girls’ preferences for dolls, and boys’ preferences for cars, trucks, and other vehicles, have been extensively documented in the developmental literature and widely discussed in the popular press. Children have been observed to display toy preferences that are consistent with gender stereotypes as early as 14 to 20 months of age (O’Brien & Huston, 1985; Roopnarine, 1986). It has been suggested that early prefer- ences for stereotyped toys may direct girls’ and boys’ cognitive and social development, via differential practice of the skills associated with different types of play materials and activities (Liss, 1983). Children’s preferences may also impact their social worlds by leading them to play with other children who have shared interests and similar play styles. The ‘‘separate worlds’’ associated with gender across the lifespan may � rst be observable in the early play preferences and styles of toddler girls and boys (see Maccoby, 1998, for a recent review).
The origins of gender differences in toy preferences, however, remain a topic of both scienti� c and popular debate. Toddlers’ toy preferences may relate to the function or use of the toy. For example, there may be a gender difference in interest in toys that elicit motion (e.g., cars) or nurturance
(e.g., dolls), as suggested by Eisenberg, Murray, and Hite (1982). There have been numerous attempts to explain the appearance of preferences for gender-stereotyped toys and activities: Some of these theories suggest a causal link between the development of knowledge of gender stereotypes and the appearance of stereotyped preferences for speci� c types of play materials and activities. That is, children’s preference for stereotyped toys may be due to their knowledge of those stereotypes (Martin & Halverson, 1981). Other approaches suggest that stereotyped toy preferences may precede knowl- edge of gender stereotypes, and result from social learning or biological factors.
Cognitive developmental approaches (e.g., Blakemore, LaRue & Olejnik, 1979; Kohlberg, 1966), and more recently, schema theory (e.g., Bem, 1981; Levy & Carter, 1989; Liben & Signorella, 1987; Martin & Halverson, 1981; Ruble & Martin, 1998; Ruble & Stangor, 1986), suggest that children start to form an associative network, or schema, around the dichot- omous category system of gender during the � rst year of life. Gender categories of ‘‘male’’ and ‘‘female’’ are initially based on obvious perceptual characteristics (e.g., vocal pitch; hair length; Leinbach & Fagot, 1993; Miller, 1983) that differ- entiate the male and female adults in the infant’s world.
The behavioural roles of females and males, as observed by the infant (presumably based on differentiated behaviour of mothers and fathers, female and male siblings, etc.), might also be assimilated into the child’s gender schema within the � rst two years. Objects associated with stereotyped activities (e.g., speci� c types of toys, household items, clothing) may be incorporated into the infant’s gender schema during this process. This knowledge of what is ‘‘for girls’’ and ‘‘for boys’’ becomes incorporated into the cognitive component of the gender schema (Bem, 1981; Martin & Halverson, 1981). Because gender is related to self-identity and because of the
International Journal of Behavioral Development © 2001 The International Society for the 2001, 25 (1), 7–15 Study of Behavioural Development
Correspondence should be sent to Lisa A. Serbin, Centre for Research in Human Development, Department of Psychology, Concordia University, 7141 Sherbrooke Street West, Montreal, Quebec, Canada, H4B 1R6; e-mail: Lserbin@vax2.concordia.ca
The research reported in this paper was supported by grants from the S.S.H.R.C. of Canada, and the FCAR of Quebec. Portions of this paper were
presented at the meeting of the Society for Research in Child Development, Washington, DC, April, 1997. The authors wish to thank the participating babies and parents, and the students and staff of the Child Development Lab, for their help and cooperation. Thanks, also, to la Commission d’Accès a l’Information du Que bec.
8 SERBIN ET AL. / VISUAL PREFERENCES
prescriptive nature of stereotypes, gender-related cognitions are also affect-laden. Accordingly, children are motivated to act in accordance with stereotypes. This motivation forms the affective component of the gender schema (Bem, 1981; Martin & Halverson, 1981). This gender schematic approach suggests that children’s cognitive gender schema plays a formative role in the development of stereotyped preferences and behaviour. Accordingly, cognitive associations of speci� c objects with gender categories could precede or coincide with the appear- ance of stereotyped preferences during the second year.
As yet, there is little empirical evidence regarding the acquisition of gender stereotypes during the � rst two years, or about the sequence in which knowledge of stereotypes and preferences for stereotyped activities appears (Ruble & Martin, 1998). Most cognitive studies of gender role acquisition have focused on children’s response to gender labels (man/woman; boy/girl) and the accurate labelling of the self ‘‘gender identity’’) as signalling that gender categories have been acquired (Etaugh, Grinnel & Etaugh, 1989; Leinbach & Fagot, 1986; Weinraub et al., 1984). Knowledge of the activities and other characteristics associated with gender (i.e., gender stereotypes) is expected to follow, along with motivation to conform to the norms, prescriptions, and proscriptions implicit in the child’s expanding network of gender-related assocations (Kohlberg, 1966; Smetana & Letourneau, 1984).
Contrary to this view, the empirical literature suggests that consistent use of gender labels for self and others does not occur until about 212 years of age (Fagot & Leinbach, 1989), whereas preferences for gender-typed activities and toys can be observed in natural and laboratory settings considerably earlier (O’Brien & Huston, 1985; Roopnarine, 1986). However, other cognitive manifestations of gender knowledge, such as dis- crimination of faces and voices (Fagan & Singer, 1979; Leinbach & Fagot, 1993; Miller, Younger, & Morse, 1982), and intermodal association of faces and voices (Poulin-Dubois, Serbin, Kenyon, & Derbyshire, 1994; Walker-Andrews, Bahrick, Raglioni, & Diaz, 1991) occur within the � rst year. Conceivably, these earlier phenomena may signal the forma- tion of a preverbal gender construct that allows the child to begin associating other objects or activities with gender categories, preceding or coinciding with the development of preferences for gender-stereotyped activities and toys.
Early in the process, it is possible that these preferences and cognitive stereotypes develop independently or in parallel sequence, gradually becoming linked to form a comprehensive cognitive/affective/behavioural schema over the early child- hood years. It is also possible that the sequence or timing of the earliest manifestations differs for girls and boys, due to differential socialisation or rates of maturation. For example, the socialisation pressures on boys to act in accordance with stereotypes are much stronger than for girls, and come from many different sources, such as parents, teachers, and peers (Fagot, 1978; Fagot & Hagan, 1991). Accordingly, a number of studies suggest that boys’ preference for gender-stereotyped toys is stronger than girls’ (Carter & Levy, 1988; Lobel & Menashri, 1993; Sen, 1999) and appears earlier (O’Brien & Huston, 1985). Other studies suggest that the cognitive aspects of gender role acquisition occur at somewhat different ages for girls and boys, with girls showing earlier acquisition of gender labels (Poulin-Dubois, Serbin, & Derbyshire, 1998), and boys showing differential memory for masculine stereotyped events at an earlier age than girls (Bauer, 1993; Boston & Levy, 1991; Sen, 1999).
Although gender-stereotype knowledge, sex-typed prefer- ences, and behavour all appear within the � rst three years of life, the developmental sequence in which these appear has not been established. This may be because knowledge of gender stereotypes cannot be measured easily before the child is able to use language � uently or to comply reliably with verbal test instructions (e.g., by labelling, pointing or sorting). These abilities are required by most established measures of gender identity, gender-stereotype knowledge, and other gender concepts (e.g., Edelbrock & Sugawara, 1978; Fagot, Leinbach, & Hagan, 1986; Martin & Little, 1990; Thompson, 1975). Reports that knowledge of gender stereotypes appears after the second or third birthday (Etaugh & Duits, 1990; Martin & Little, 1990; Perry, White, & Perry, 1984), considerably later than the appearance of stereotyped activity preferences, may re� ect the limitations of the current research paradigms for the testing of preverbal infants.
Researchers have recently adapted experimental paradigms involving measurement of infants’ visual attention (Spelke, 1976) to study the development of gender concepts during the � rst two years of life. Using these procedures it is possible to investigate the development of knowledge about gender stereotypes and gender-stereotyped activity preferences during the preverbal period. Two recent studies using an adaptation of the intermodal preferential looking paradigm have shown that sometime during the second half of the � rst year, infants ‘‘match’’ voices and faces according to gender, as expressed by their longer looking time at the face that corresponds to the gender of the voice they are hearing (Poulin-Dubois et al., 1994; Walker-Andrews et al., 1991). These studies demon- strate that young infants are not only capable of discriminating faces on the basis of gender, as demonstrated by earlier habituation studies (Fagan, 1976; Leinbach & Fagot, 1993; Levy & Haaf, 1994), but that they have typically formed intermodal (auditory/visual) gender categories before their � rst birthday.
The intermodal preferential looking paradigm could also permit the measurement of infants’ knowledge of other aspects of gender categories, in addition to the physical characteristics of males and females. Speci� cally, infants’ association of various activities and objects with gender categories may be examined by a ‘‘match-to-standard’’ adaptation of the pre- ferential looking paradigm. Differential association of speci� c types of toys with male versus female faces and voices would demonstrate that infants associate these objects with gender. The preferential looking paradigm can also be used to examine infants’ preferences for speci� c visual stimuli. Infants’ gaze at two competing stimuli can be compared to obtain a measure of their relative preference for the two visual displays.
The two studies presented in this paper address the development of preferences for stereotyped toys and knowl- edge of the association of toys with gender categories between 12 and 24 months of age, using the preferential looking paradigm. In the � rst experiment, 12-, 18-, and 23-month-old infants were shown paired photos of stereotyped male and female toys (vehicles and dolls). In a series of trials, each pair of toys was preceded by either a male or female face and voice, to determine whether the infants could visually match the toys with the face by looking differentially at the toy that corresponded to the preceding ‘‘standard’’. An interspersed series of preference or ‘‘control’’ trials showed the toy pairs without any preceding gender stimuli, to assess children’s visual preferences for male- and female-stereotyped toys.
Experiment 2 employed a reversal of the paradigm used in Experiment 1 with new groups of 18- and 24-month-olds, and was intended to clarify the infants’ association between toys and gender categories. In both studies, the primary focus was on the early appearance of gender-stereotyped preferences and stereotype knowledge, and on the exploration of gender differences in the affective and cognitive components of the gender schema during the second year.
Experiment 1
Method
Subjects. Seventy-seven children, twenty 12-month-olds (8 boys, 12 girls, M age = 12.6 months, range 12.0–13.43 months), thirty 18-month-olds (15 boys, 15 girls, M age = 18.67 months, range 18.1–19.5 months), and twenty-seven 23-month-olds (14 boys, 13 girls, M age = 23.1 months, range 22.1–24.0 months), were recruited from hospital birth lists. An additional two 12-month-old children were eliminated from the sample due to side bias (65% or more of total looking time, across trials, on either left or right screen), and two 23-month- olds were excluded because of inattention (more than four trials lost in one condition, see criteria for data screening). Eighty-seven percent of the sample was of European–Canadian origin, 10% were of West-Indian descent, and 3% were Asian– Canadian. Children were from middle and lower socio- economic status families, for whom English was the sole or predominant language spoken at home. None of the children had any hearing or visual problems, as reported by the parents.
Stimuli. Photographs of an 8-year-old boy’s face and an 8- year-old girl’s face served as ‘‘standards’’ for matching in the test trials. These photos were obtained from a modelling agency; both children were attractive and readily identi� able by gender, based on hair length. ‘‘Target’’ stimuli were colour photos of toys, six vehicles (tractors, cars, and trains) and six dolls (assorted baby and rag doll types, all clothed). Each photo of a vehicle was paired with a photo of a doll, similar in size and colour, to form six pairs of stereotyped masculine and feminine toys. The photos were digitised for presentation on computer monitors.
Voice prompts were recorded by four school-aged boys and four girls, who were individually recorded saying, ‘‘Where’s my toy? . . . Find my toy! . . .’’. A group of � ve male and six female adult judges rated each voice recording on a 5-point scale, according to how clearly the voice ‘‘sounded like’’ a boy (rated 1), a girl (rated 5), or was gender-neutral (rated 3). One boy’s voice and one girl’s voice with mean ratings of 1.0 and 4.7, respectively, were selected and digitised for computer pre- sentation.
Apparatus and procedure. Each infant was brought into the laboratory and seated in a baby chair attached to a table. The parent (usually the mother) was seated in a chair directly behind the child. Infant and parent faced the front panel of a three-sided testing chamber from a distance of 170 cm. The front panel contained two 12-inch colour monitors, separated by 60 cm. An audio speaker placed midway between the two monitors played the verbal prompts. Above the speaker was a 25-watt blue light that served to direct the infant’s attention to the centre between trials. Trial length and synchronisation of
the auditory and visual stimuli were controlled by linked microcomputers. The testing session lasted about 10 minutes. All sessions were videotaped, with an 8mm video camera, the lens of which was located at the child’s eye level between the two monitors.
Every trial began with blank screens and the light blinking. After 3 s, the light went off and a pair of pictures appeared on the screen for 5 s. There were two types of trials: test and control. During the test trials, identical pictures of either the male or the female child appeared on the two screens for 5 s, accompanied by the corresponding gender child’s voice saying, ‘‘Where’s my toy?’’. When the images disappeared, the blue light was illuminated brie� y to focus the child’s gaze between the two screens. A pair of toys (one doll and one vehicle) then appeared for 5 s, accompanied by the same voice saying, ‘‘Find my toy!’’ (see Figure 1). On control trials, the same stereotyped toys (i.e., a doll and a vehicle) appeared for 5 s, with no preceding faces and no voice prompt. The parent was instructed not to talk to her child or to make any attempt to orient the child toward either screen during the testing. A pause command allowed the experimenter to interrupt the testing if the child became restless or cried.
The 24 trials were divided into 2 blocks of 12 trials each, separated by a short pause. The � rst block of trials started with a warm-up trial during which two pictures of common objects (a chair and a dog) were presented, and the infant was asked by an adult female voice to ‘‘look at the dog!’’. Each block of trials included six control trials, during which a pair of stereotyped toys appeared with no preceding face or voice prompt, and six test trials in which the toy pictures were preceded by either the boy’s or girl’s face, accompanied by corresponding voice prompts. The presentation of test and control trials was alternated within each block. The order of presentation of the pairs of toys was randomised and constant for all subjects. In the test trials, target pictures (i.e., the ones matching the male or female face/voice) were presented equally often on each screen. In both test and control trials, an equal number of vehicles and dolls was presented on the left and right. Each of the six pairs of toys was presented four times: twice in the control trials and twice in the test trials. Each individual toy picture appeared once as a ‘‘match’’ to the preceding child face and once as a ‘‘mismatch’’.
Interobserver agreement, data screening, and scoring. Two primary observers, who were unaware of the order of trials or the position of the photos, coded the videotapes. A third observer coded a random 25% of each primary observer’s tapes to measure interobserver agreement. The amount of time (in seconds) the infant looked at each stimulus, as well as time spent looking away from the computer screens, was recorded from the videotapes. Pearson product-moment correlations were computed between observers’ ratings of time the infant looked to each side during each of the 24 trials. The correlations between each of the two primary observers and the third observer, across all trials, were over .90. Data were excluded from the analyses for any trial in which: (a) � xation to one of the toys was 100% of total � xation time to the pair; or (b) total � xation to both toys was less than 1.25 s (25% of trial duration). These trials were excluded to eliminate unreliable data (Hirsh-Pasek & Golinkoff, 1996). With shorter looking times, determining preferences becomes unreliable. Following these criteria, 9% of trials were eliminated prior to data analysis.
INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2001, 25 (1), 7–15 9
10 SERBIN ET AL. / VISUAL PREFERENCES
Results
Children’s average looking times at the set of dolls and at the set of vehicles were computed for match (e.g., looking time at vehicles following photo of boy), mismatch (e.g., looking times at vehicles following photo of girl), and control (e.g., looking time at vehicles with no voice, and no preceding photo) conditions. For instances of missing trials, the data were averaged across the remaining trials.
These looking times were analysed in a 3 (condition: match, mismatch, and control) 2 (toy type: vehicles vs. dolls) 3 (age group: 12, 18, and 23 months) 2 (child’s sex) mixed- model four-way analysis of variance (ANOVA). Condition and toy type were within-group variables, and age and child’s sex were the between-group variables. This analysis yielded two signi� cant three-way interactions: between age, sex, and toy type [F(2,71) = 4.19, p < .02], and between age, toy type, and condition [F(4,142) = 2.40, p < .05], (see Table 1). To
Figure 1. Examples of ‘‘test trials’’ for Experiments 1 and 2.
Table 1 Experiment 1: Looking time (in seconds)
Age Vehicles Dolls All Pictures
Match Mismatch Control All Match Mismatch Control All Match Mismatch Control All
12 months Boys 1.64 1.67 1.36 1.56 2.10 2.00 1.98 2.03 1.87 1.83 1.67 1.79 n = 8 (.48) (.20) (.28) (.20) (.66) (.63) (.52) (.58) (.24) (.35) (.28) (.24) Girls 1.62 1.70 1.41 1.58 2.15 2.36 2.00 2.17 1.88 2.03 1.71 1.87 n = 12 (.37) (.35) (.47) (.37) (.49) (.47) (.43) (.40) (.34) (.21) (.31) (.26) Total 1.62 1.69 1.39 1.57 2.13 2.21 1.99 2.11 1.88 1.95 1.69 1.84 n = 20 (.40) (.30) (.39) (.31) (.55) (.55) (.46) (.47) (.30) (.28) (.29) (.25)
18 months Boys 2.33 2.23 2.22 2.26 2.01 2.00 1.79 1.93 2.17 2.11 2.01 2.10 n = 15 (.56) (.51) (.55) (.49) (.39) (.44) (.42) (.38) (.28) (.35) (.30) (.26) Girls 1.66 1.79 1.73 1.73 2.47 2.43 2.40 2.43 2.06 2.11 2.06 2.08 n = 15 (.52) (.35) (.44) (.37) (.38) (.34) (.42) (.28) (.31) (.25) (.23) (.19) Total 1.99 2.01 1.98 1.99 2.24 2.21 2.09 2.18 2.12 2.11 2.04 2.09 n = 30 (.63) (.48) (.55) (.50) (.44) (.45) (.52) (.41) (.30) (.30) (.26) (.22)
23 months Boys 2.73 2.87 2.60 2.73 1.46 1.38 1.44 1.43 2.09 2.13 2.02 2.08 n = 14 (.68) (.56) (.49) (.54) (.49) (.51) (.44) (.42) (.35) (.22) (.25) (.21) Girls 2.23 2.12 1.90 2.09 2.05 1.86 1.94 1.98 2.17 2.00 1.92 2.04 n = 14 (.56) (.30) (.37) (.28) (.38) (.41) (.38) (.29) (.36) (.23) (.18) (.15) Total 2.49 2.50 2.25 2.43 1.75 1.61 1.69 1.69 2.13 2.06 1.97 2.06 n = 28 (.66) (.58) (.56) (.53) (.53) (.52) (.48) (.46) (.35) (.23) (.22) (.18)
1) Standard (5 seconds)
``Where’s my toy?’’
2) Test Trial (5 seconds)
``Find my toy!’’
Sequence Voice
Monitor 1 Monitor 2
Voice
Monitor 1 Monitor 2
``See my doll? That’s my doll!’’
``Look at me!’’
interpret these interactions, we subsequently carried out separate three-way ANOVAs (Sex of Child Toy Type Condition), within each of the three age groups.
Response at 12 months of age. At this age, a signi� cant main effect of type of toy indicated that children of both sexes showed a visual preference for dolls over trucks [F(1,18) = 13.31, p < .002], (see Figure 2). In other words, the 12- month-old children, as a group, had a preference for looking at dolls more than vehicles. This may relate to the resemblance of dolls to human faces, which are attractive to infants (Morton & Johnson, 1991). No evidence of gender-stereotyped prefer- ences was found at this age level. That is, there was no signi� cant interaction beween child’s sex and type of toy [F(18) = .17, n.s.] (see Figure 2). As seen in Figure 2, 12- month-old girls looked at the dolls for an average of 2.17 s, while boys looked at the dolls an average of 2.03 s. Girls looked at the vehicles for an average of 1.58 s, boys for 1.56 s. Descriptively, 63% of the boys and 83% of the girls showed a visual preference for the dolls (i.e., these children spent more time looking at the dolls than the vehicles, on over half of the 24 trials). In terms of trials, boys showed a preference for the dolls on 63% of the trials, and girls showed a preference for the dolls on 65% of the trials.
Finally, the 12-month-old children spent more time looking at the screens during the match and mismatch trials (which were preceded by photos of children’s faces and accompanied by verbal prompts), than during the control trials (when the toys appeared without prior visual or auditory cues), [F(2,36) = 10.23, p < .001], (see Table 1). However, the children did not indicate any awareness of stereotyping of the toys. Looking times under match and mismatch conditions did not differ [t(19) = 1.06, n.s.].
Response at 18 months. At this age, the results were substantially different. An interaction of sex of child with type of toy [F(1,28) = 20.88, p < .001], revealed that girls looked
at the dolls longer than the boys did M = 2.43 versus M = 1.93 s, respectively [t(28) = 4.16, p < .001] (see Figure 2). Descriptively, 80% of the girls looked longer at the dolls than at the vehicles on over half of the trials. In terms of trials, girls looked longer at the dolls than at the vehicles on 68% of the trials. Boys at 18 months showed a preference for the vehicles, with looking time averaging 2.26 s versus 1.73 s for the girls [t(28) = 3.38, p < .002]. Sixty percent of the boys looked longer at the vehicles than at the dolls on more than half of the trials. In terms of trials, on 56% of the trials, boys looked longer at the vehicles than at the dolls. No other signi� cant main effects or interactions were found. In other words, at 18 months boys showed more visual interest in the male sex-typed vehicles than girls did, and girls looked longer at the dolls than boys did. As at 12 months, however, there was no evidence of ‘‘matching’’ type of toy to gender categories. The main effect of condition and all interactions between condition and other variables were not signi� cant (see Table 1).
Response at 23 months. At this age, a signi� cant main effect of toy indicated that, in general, the children looked more at the vehicles than at the dolls [F(1,25) = 26.00, p < .001] (see Fig. 2). Sex-typed preferences were still apparent, however, as indicated by a signi� cant interaction of child’s sex by type of toy [F(1,25) = 19.15, p < .001]. When the amount of time spent looking at vehicles was compared for boys and girls, the difference was statistically signi� cant, with boys’ average score higher than girls’, M = 2.74 s and M = 2.09 s, respectively [t(25) = 3.85, p < .001] (see Figure 2). Descriptively, 86% of the boys showed a visual preference for the vehicles on over half of the trials. In terms of trials, boys showed a visual preference for the vehicles on 75% of the trials. For the dolls, this pattern was reversed, with girls’ mean looking time, M = 1.98 s, signi� cantly higher than boys’, M = 1.43 s, [t(25) = 3.94, p < .001]. Fifty-seven percent of girls showed a preference for dolls on over half of the trials. In terms of trials, girls showed a visual preference for the dolls on 49% of the trials.
Finally, as in the two younger groups, there was no indication of ‘‘matching’’ either dolls or vehicles with their respective gender-typed category by the 23-month-old chil- dren. A signi� cant condition by toy interaction [F(2,50) = 4.82, p < .02], only indicated that the 23-month-old children looked more at the vehicles during the two experimental ‘‘voiced’’ conditions than during the control condition and showed no signi� cant difference in looking time at the dolls under the three experimental conditions (see Table 1).
Discussion
In sum, visual preferences for gender-stereotyped toys appeared to emerge between 12 and 18 months of age, and remained evident at 23 months. There was no evidence, however, of knowledge of sex-typing (i.e., matching toys with male/female faces and voices) at any of the three age levels tested. These results seem to indicate that boys’ and girls’ visual preferences for vehicles and dolls, respectively, emerge earlier than their association of these toys with gender categories. The � ndings also suggest that children are unaware of associations between these toys and gender categories until after their second birthday. However, before reaching these conclusions, several methodological issues must be addressed concerning the experimental paradigm used in Experiment 1.
INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2001, 25 (1), 7–15 11
Figure 2. Experiment 1: Girls’ and boys’ mean looking times at vehicles and dolls at 12, 18, and 23 months (based on 5 s trials; bars indicate standard errors).
12 SERBIN ET AL. / VISUAL PREFERENCES
During the presentation of the experimental stimuli on the video screens, the children expressed considerable interest in the toys by looking at the display screens for most of the available time, and also by spontaneously pointing, naming, and talking about the toys (e.g., ‘‘Look! Truck! Truck! My truck!’’ exclaimed repeatedly by one of the 23-month-old boys). As indicated earlier, the children showed preferences for particular types of toys at all ages tested, and generally seemed totally involved in the experience of examining the visual stimuli when the toys were on screen. This degree of interest and involvement may have interfered with the children’s cognitive processing of the ‘‘match’’ between the gender of the child’s face and voice, presented prior to the toys in each trial, and the toys shown.
In other words, strong interest and preferences for toys may have dominated the children’s attention during the trials, reducing the memory trace of the preceding gender category cues (the faces and voices) to an ineffective level. We therefore carried out a second study with a modi� ed paradigm, including a series of trials in which one stereotyped toy (either a doll or a vehicle) was � rst shown as a ‘‘standard’’ for matching, followed by paired faces of boys and girls. That is, Experiment 2 reversed the sequence of faces and toys used in the paradigm for Experiment 1. Based on the children’s responses in Experiment 1, preferences for faces were not expected to be as strong as for toys and hence, might interfere less with the matching response in this revised paradigm. Two age groups, of 18 and 24 months respectively, were included in Experiment 2. Twelve-month-olds were excluded from this experiment because this age group did not show any sex-typed preferences in Experiment 1. We anticipated that, with the revised paradigm, an association between girls’ and boys’ faces and the preceding stereotyped toys might appear towards the end of the second year.
Experiment 2
Method
Subjects. Twenty-eight 18-month-olds (M age = 18.3 months, range 17.5–19.5 months), and thirty 24-month-olds (M age = 24.3 months, range 23.5–25.17 months), equally divided into boys and girls, were recruited from the same sources as in Experiment 1. Ninety-� ve percent of these children were of European–Canadian backgrounds, with the remaining four children from Asian and West-Indian Canadian backgrounds. An additional 13 children were excluded from the sample due to side bias or inattention.
Stimuli. Fifty photos of school-aged children were obtained from a local modelling agency. Six pairs of male and female children were selected from this set, on the basis of similar facial features, facial expression, hair colour, and pose, in order to make the picture pairs similar in salience and attractiveness. These pairs, each containing one girl’s and one boy’s face, became the test stimuli for Experiment 2 (see Figure 1). The toys used in Experiment 1 were used as ‘‘standards’’ in the test trials of Experiment 2. A neutral voice was needed in Experiment 2 because gendered voices might have led to a ‘‘matching’’ of boys’ and girls’ faces to the voices by gender, irrespective of the toy shown as a standard, and obscuring the
meaning of differential attention to the photos (Poulin-Dubois et al., 1994). Accordingly, a recorded child’s voice judged ‘‘gender-neutral’’ (i.e., indistinguishable by sex) by � ve female and � ve male adult judges, was used. Although we were aware that this modi� cation might have changed the information- processing load required for the task, the modi� cation was necessary.
Procedure. In Experiment 2, the test trials reversed the sequence of Experiment 1. That is, during the 12 test trials, identical photos of one of the toys � rst appeared on both screens for 5sec, with a gender-neutral voice saying ‘‘See my car (doll)? That’s my car!’’. The toy pictures were followed by one of the pairs of children’s faces, with a boy appearing on one screen and a girl on the other, and the gender-neutral child’s voice saying ‘‘Look at me!’’. The faces appeared for 5 s. On 12 control trials, the pairs of faces were shown for 5 s without any preceding visual ‘‘standard’’, whereas the gender-neutral voice said ‘‘Look at the people!’’. This voice prompt was added to the control trials in Experiment 2 to better equate the children’s overall level of attention during the control and test trials. Aside from the modi� cation in picture presentation and voices, the protocol (number and sequence of trials) was identical to Experiment 1.
Coding, interobserver agreement, and data screening. The two observers who coded the videotapes were unaware of the positions of the photo stimuli. A Pearson product-moment correlation coef� cient averaging .90 was computed between the two observers’ recorded looking time towards each screen, on a randomly selected 25% of videotaped sessions. In Experiment 2, 15% of the trials were eliminated due to insuf� cient total looking time (less than 1.25 s: 25% of trial), or to visual � xation on one screen exclusively during the trial.
Results
Children’s average looking times at the faces of girls and faces of boys were computed for match (e.g., looking time at girls’ faces following photos of dolls), mismatch (e.g., looking time at girls’ faces following photos of vehicles), and control (e.g., looking time at girls’ faces with no preceding photo) condi- tions.
As in Experiment 1, data were analysed in a four-way ANOVA, with age (18 vs. 24 months), and child’s sex as between-subjects variables, and face (boy or girl) and condition (match, mismatch, or control) as within-subjects factors. As in Experiment 1, the dependent measure was looking time, however in Experiment 2, looking time at children’s faces was used in the test trials instead of looking time at toys. This analysis yielded a signi� cant effect of condition [F(2,108) = 10.48, P < .001], which was quali� ed by an interaction between child’s sex and condition [F(2,108) = 3.73, p < .03], (see Table 2). No other effects or interactions were statistically signi� cant.
To determine the source of the interaction, one-way analysis of variance comparing the match, mismatch, and control conditions were run for girls and boys separately. For the girls, there was a main effect of condition [F(2,56) = 10.97, p < .001]. A Tukey HSD test (p < .01) revealed that girls in both age groups looked signi� cantly more at the faces that ‘‘matched’’ the preceding toy standard than at the ‘‘mismatch-
ing’’ faces, with mean looking times of 2.19 and 1.98 s, respectively, and also looked more at ‘‘matching’’ faces than at the control faces (M = 1.95 s), (see Figure 3). Although the differences in looking times among the three conditions appears small (e.g., 0.21 s), the associated effect size is in fact quite large (partial eta2 = .28), indicating that this � nding is very robust. Looking time at the matching faces was also examined as a percentage score, to see whether the girls’ response pattern differed from chance levels (which would be 50% looking time on each screen). The girls looked at the matching faces an average of 53% of their total looking time, which was signi� cantly different from chance [t(28) = 3.36, p < .002]. Descriptively, 62% of the girls looked at the matching faces (looking time > 51%) more than the mismatched faces. In other words, the girls as a group looked at the faces that corresponded to the gender stereotyping of the toys; they looked more at the girls’ faces after seeing dolls, and looked more at the boys’ faces after seeing vehicles. Under the control condition, when no toy was previously shown as a standard,
they did not show a statistically signi� cant pattern of differential looking at the faces.
For the boys, there was also a main effect of condition [F(2,56) = 3.96, p < .03], which appeared to be due to lower looking times during the control condition (M = 1.97 s) than during the experimental conditions (Mmatch = 2.11 s, Mmismatch = 2.12 s). A follow-up Tukey HSD test, however, failed to � nd any differences in the boys’ looking times between the three conditions. In other words, the boys showed no evidence of differential ‘‘matching’’ of boys’ or girls’ faces to the preceding vehicle or doll standard.
We subsequently carried out a number of follow-up analyses to see whether the sex by condition interaction might be related to less overall attentiveness (i.e., less total time looking at the test stimuli) or to more invalid trials (i.e., trials eliminated according to the constraints of the paradigm) for the boys. Neither of these post-hoc explanations for the sex by condition interaction was supported by the results. We also reanalysed the data including whether or not the child had an older sibling (same or other sex) as an independent variable in the design. Again, the existence of a sibling did not predict or interact with other variables in predicting children’s response to the paradigm.
In summary, with the modi� ed paradigm used in Experi- ment 2, girls showed signi� cant matching of boys’ and girls’ faces with vehicles and dolls, respectively, by 18 months of age. In contrast, the boys as a group did not demonstrate an awareness of the association of the dolls or vehicles with gender at either 18 or 24 months. It should be noted that, unlike the results of Experiment 1, there was no ‘‘inter- ference’’ during the matching trials as a result of visual preferences for the test stimuli. That is, no preferences for either same- or other-sex children’s faces were found using this paradigm. Furthermore, the inclusion of a gender-neutral voice may have changed the level of information processing. It is possible that the ambiguity of the sex of the voice encouraged the children to scan and compare the two faces to � nd a match for the ambiguous voice (Poulin-Dubois et al., 1994). Depth of processing may have therefore been increased in the paradigm used in Experiment 2.
INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2001, 25 (1), 7–15 13
Table 2 Experiment 2: Looking time (in seconds)
Boy Face Girl Face Both Faces
Match Mismatch Control Match Mismatch Control Match Mismatch Control
18 months Boys 2.21 2.16 1.89 2.07 2.09 1.95 2.14 2.13 1.92 n = 14 (.31) (.53) (.40) (.38) (.49) (.40) (.30) (.45) (.34) Girls 2.08 1.92 1.92 2.16 1.95 1.90 2.12 1.93 1.91 n = 14 (.41) (.41) (.38) (.35) (.46) (.36) (.27) (.38) (.27)
24 months Boys 2.06 2.16 2.04 2.09 2.08 2.00 2.08 2.12 2.02 n = 15 (.37) (.43) (.41) (.35) (.34) (.34) (.29) (.24) (.25) Girls 2.11 1.96 1.91 2.40 2.11 2.08 2.25 2.03 2.00 n = 15 (.52) (.36) (.31) (.28) (.39) (.41) (.32) (.23) (.28)
Both ages Boys 2.13 2.16 1.97 2.08 2.08 1.98 2.11 2.12 1.97 n = 29 (.35) (.47) (.40) (.35) (.41) (.37) (.29) (.35) (.29) Girls 2.09 1.94 1.91 2.28 2.03 1.99 2.19 1.98 1.95 n = 29 (.46) (.38) (.34) (.34) (.43) (.39) (.30) (.31) (.30)
Figure 3. Experiment 2: Girls’ and boys’ mean looking times at match, mismatch, and control stimuli (based on 5sec trials; bars indicate standard errors).
14 SERBIN ET AL. / VISUAL PREFERENCES
General discussion
The results con� rm previous reports, based on naturalistic observations, that gender-stereotyped toy preferences appear during the second year, and are established by 18 months of age. In the current study, ‘‘preferences’’ were determined from children’s visual attention patterns to paired stereotyped toy photos. Despite the fact that this experimental paradigm is very different from naturalistic observation of play, the present results correspond to children’s toy choices and behavioural patterns when play is observed in the home or laboratory playroom (Fein, Johnson, Kosson, Stork, & Wasserman, 1975; O’Brien & Huston, 1985; Roopnarine, 1986). These prefer- ences probably relate to the activities involved in using the toys. That is, toddler boys may have a greater interest in vehicles which stimulate high levels of motor activity (or that move when activated by the child, or the child’s imagination) than in play materials that evoke more sedentary play. Conversely, toys evoking nurturance may be of greater interest to girls (Eisenberg et al., 1982). It is interesting that girls’ and boys’ interests diverge between 12 and 18 months, in light of the greater physical mobility children achieve during this period. That is, once they are capable of walking and running, children may express increasing interest in toys that involve physical motion. It should be noted that the girls in Experiment 1 also showed an increasing interest in vehicles over the period from 12 to 24 months, but they also maintained their interest in dolls (see Figure 1).
Regarding children’s awareness that toys are associated with gender categories, in Experiment 2, the girls, but not the boys, matched the gender-‘‘appropriate’’ faces to the preceding doll or vehicle standard. This indicates that the girls were aware that the dolls were associated with girls, and vehicles with boys.
The boys did not demonstrate any awareness of gender- based associations. There are several possible explanations for this gender difference. First, null results (i.e., match and mismatch conditions not signi� cantly different from each other, and not different from chance levels) may mean that the child is unaware of the association being tested, or alternatively, that the child is not responding to the paradigm for some reason (e.g., lack of interest in the display, or overwhelming salience of some part of the display, which distracts the child). Informal observation of the boys’ responses to the toy ‘‘standards’’ suggested they were extremely inter- ested in the vehicles, and may have been distracted by these attractive toys from the faces in the subsequent test trials. Therefore, we cannot conclude that the boys were unaware of the match between gender and toys based on these results. However, there is certainly no evidence from these studies that they were aware of the association.
Second, several previous studies suggest that preschool girls’ knowledge of stereotypes may be greater than boys’. For example, preschool-aged girls tend to score higher than boys do on tests of stereotype knowledge and gender comprehension (Bem, 1989; Signorella, Bigler, & Liben, 1993; Thompson, 1975), although by middle childhood boys may give more stereotyped and less � exible responses on stereotyping tests (Ruble & Martin, 1998; Serbin, Powlishta, & Gulko, 1993). In a recent study on children’s matching of labels (e.g., boy, girl, man, lady) to male and female faces, it was found that girls could match faces with the appropriate labels by 18 months, although the boys did not respond consistently to the gender labels even at 24 months of age (Poulin-Dubois et al., 1998).
Similarly, results of another study using an inductive general- isation procedure demonstrated that 24-month-old girls were likely to select a gender-‘‘appropriate’’ female or male doll to engage in imitation of stereotyped play activities, indicating an awareness of the association of these activities with gender. In contrast, the boys did not differentially select a male or female doll to carry out the activities (Poulin-Dubois, Serbin, Eichstedt, & Beissel, 1999). In other words, the cognitive component of young girls’ gender schemas may be stronger than that for boys.
A third possible explanation for the gender difference is that the affective component of boys’ gender schemas may be stronger than that of girls’. During their second year, boys appear to know less about the categorisation of activities by gender than girls do, but nevertheless have strong preferences for male-stereotyped toys and play activities. It is possible that these sex-typed preferences in� uence or delay their acquisition of some cognitive aspects of gender schemas. That is, it may be that their strongly differentiated interest in male-stereotyped activities causes boys to pay less attention to feminine or nonstereotyped events or stimuli, which they � nd less salient. Consequently, at this early stage, the strong preference and interest in ‘‘boys’ toys’’ and masculine activities may interfere with noticing (and remembering) that males and females have different interests and preferences. In support of this theory, several studies have demonstrated that toddler boys show increased attention to and memory for masculine-stereotyped events and scripts, compared with their memory for feminine- stereotyped events. Girls, in contrast, remember both femi- nine- and masculine-stereotyped events equally well (Bauer, 1993; Boston & Levy, 1991; Sen, 1999). Boys’ differential attention to male-stereotyped items and activities might result from their strong preferences (an affective component of gender schemas), and could interfere with their noticing and remembering the association between speci� c activities and gender categories. Toddler girls, on the other hand, who pay equal attention to male- and female-stereotyped activities, may be better able to acquire knowledge about the association of speci� c activities with male or female gender.
To summarise the present � ndings, it is clear that sex-typed toy preferences appear during the second year. For girls, it is also clear that speci� c types of toys (i.e., dolls and vehicles) are associated with gender at approximately the same point in development as gender-typed preferences emerge. Speci� cally, the present � ndings suggest that girls’ awareness of gender associations with toys (and presumably with the play activities that accompany these toys) emerges by 18 months of age, which is considerably earlier than previously demonstrated. Such gender associations with toys were not yet evident among the boys, however. The reason for this gender difference remains to be determined. One possible explanation is that boys’ affective component may be stronger than girls’. Testing girls’ and boys’ respective levels of motivation to act in accordance with gender stereotypes (e.g., a test of gender salience, see Sen, 1999) and comparing those levels to the amount of gender-stereotype knowledge could provide further information about the underlying cause of this gender difference.
The nature of the relation between the development of gender-typed preferences and the awareness of gender associa- tions of these toys also remains to be determined. The present experiments did not allow us to examine individual patterns to determine if there was a sequence in which gender-typed
preferences and associations emerged because the two ques- tions were examined in separate experiments. Testing the children in Experiment 2 for their toy preferences on a different day was also an option, but the results of Experiment 1 suggested that most of the 18-month-olds would show gender-typed preferences (i.e., there would be little variability within the group concerning gender-typing of toy preferences). In subsequent research, it may be possible to track these two phenomena longitudinally, to see whether there is a predictable sequence in the appearance of gender-related toy preferences versus association of stereotyped toys with gender categories.
The present experiments suggest that both gender-typed preferences and association of objects with gender categories begin during the second year of life. The � ndings also con� rm that gender-related concepts may be acquired before children can respond reliably to previously established methods for studying them, which cannot be used before about 30 months. Whether this means that categorisation of gender-typed objects precedes the emergence of gender identity (which could be conceptualised as ‘‘self-categorisation’’ by gender) is an intriguing issue, which could be pursued directly using comparable methodology. The present studies suggest that a rethinking of cognitive theory regarding the role of gender identity in the early process of gender role learning may be warranted. If the appearance of gender identity occurs during the third year of life, it may be a relatively late aspect of schema formation that appears after gender categories and rudimentary gender-role associations (which include activities, toys, and labels) have already been acquired.
Manuscript received April 1999 Revised manuscript received July 1999
References
Bauer, P.J. (1993). Memory for gender-consistent and gender-inconsistent event sequences by 25-month-old children. Child Development, 64, 285–297.
Bem, S.L. (1981). Gender schema theory: a cognitive account of sex typing. Psychological Review, 45, 155–162.
Bem, S.L. (1989). Genital knowledge and gender consistency in preschool children. Child Development, 60, 649–662.
Blakemore, J.E., LaRue, A.A., & Olejnik, A.B. (1979). Sex-appropriate toy preference and the ability to conceptualize toys as sex-role related. Developmental Psychology, 15, 339–340.
Boston, M.B., & Levy, G.D. (1991). Changes and differences in preschoolers’ understanding of gender scripts. Cognitive Development, 6, 417–432.
Carter, D.B., & Levy, G.D. (1988). Cognitive aspect of early sex-role development: The in� uence of gender schemas on preschoolers’ memories and preferences for sex-typed toys activities. Child Development, 59, 782–792.
Edelbrock, C., & Sugawara, A.I. (1978). Acquisition of sex-typed preferences in preschool-aged children. Developmental Psychology, 14, 614–623.
Eisenberg, N., Murray, E., & Hite, T. (1982). Children’s reasoning regarding sex-typed toy choices. Child Development, 53, 81–86.
Etaugh, C., & Duits, T. (1990). Development of gender discrimination: Role of stereotypic and counterstereotypic gender cues. Sex Roles, 23, 215–221.
Etaugh, C., Grinnell, K., & Etaugh, A. (1989). Development of gender labelling: Effect of age of pictured children. Sex Roles, 21, 769–773.
Fagan, J.F. (1976). Infants’ recognition of invariant features of faces. Child Development, 47, 627–638.
Fagan, J.F., & Singer, L.T. (1979). The role of simple feature differences in infants’ recognition of faces. Infant Behavior and Development, 2, 39–45.
Fagot, B.I. (1978). The in� uence of sex of child on parent reaction to toddler children. Child Development, 49, 459–465.
Fagot, B.I., & Hagan, R. (1991). Observations of parent reactions to sex- stereotyped behaviors: Age and sex effects. Child Development, 62, 617–628.
Fagot, B.I., & Leinbach, M.D. (1989). The young child’s gender schema: Environmental input, internal organization. Child Development, 60, 663–672.
Fagot, B.I., Leinbach, M.D., & Hagan, R. (1986). Gender labelling and the adoption of sex-typed behavior. Developmental Psychology, 22, 440–443.
Fein, G., Johnson, D., Kosson, N., Stork, L., & Wasserman, L. (1975). Sex stereotypes and preferences in the toy choices of 20-month-old boys and girls. Developmental Psychology, 11, 527–528.
Hirsh-Pasek, K., & Golinkoff, R.M. (1996). The intermodal preferential looking paradigm: A window onto emerging language comprehension. In D. McDaniel, C. McKee, & H. Smith-Cairns (Eds.), Methods for assessing children’s syntax (pp. 105–124). Cambridge, MA: MIT Press.
Kohlberg, L. (1966). A cognitive-developmental analysis of children’s sex-role concepts and attitudes. In E.E. Maccoby (Ed.), The development of sex differences (pp. 82–173). Stanford, CA: Stanford University Press.
Leinbach, M.D., & Fagot, B.I. (1986). Acquisition of gender labels: A test for toddlers. Sex Roles, 15, 655–666.
Leinbach, M.D., & Fagot, B.I. (1993). Categorical habituation to male and female faces: Gender schematic processing in infancy. Infant Behavior and Development, 16, 317–332.
Levy, G.D., & Carter, D.B. (1989). Gender schema, gender constancy, and gender-role knowledge: The roles of cognitive factors in preschoolers’ gender- role stereotype attributions. Developmental Psychology, 25, 444–449.
Levy, G.D., & Haaf, R.A. (1994). Detection of gender-related categories by 10- month-old infants. Infant Behavior and Development, 17, 457–459.
Liben, L.S., & Signorella, M.L. (Eds.) (1987). Children’s gender schemata. San Francisco: Jossey Bass.
Liss, M. (Ed.) (1983). Social and cognitive skills: Sex roles and children’s play. New York: Academic Press.
Lobel, T.E., & Menashri, J. (1993). Relations of conceptions of gender-role transgressions and gender constancy to gender-typed toy preferences. Developmental Psychology, 29, 150–155.
Maccoby, E.E. (1998). The two sexes: The implications of childhood divergence for adult relationships. Cambridge, MA: Harvard University Press.
Martin, C.L., & Halverson, C.F. (1981). A schematic processing model of sex typing and stereotyping in children. Child Development, 52, 1119–1134.
Martin, C.L., & Little, J.K. (1990). The relation of gender understanding to children’s sex-typed preferences and gender stereotypes. Child Development, 61, 1427–1439.
Miller, C.L. (1983). Developmental changes in male/female voice classi� cation by infants. Infant Behavior and Development, 6, 313–330.
Miller, C.L., Younger, B.A., & Morse, P.A. (1982). The categorization of male and female voices in infancy. Infant Behavior and Development, 5, 143–159.
Morton, J., & Johnson, M. (1991). CONSPEC and CONLERN: A two-process theory of infant face recogntion. Psychological Review, 98, 164–181.
O’Brien, M., & Huston, A.C. (1985). Development of sex-typed play behavior in toddlers. Developmental Psychology, 21, 866–871.
Perry, D., White, A., & Perry, L. (1984). Does early sex typing result from children’s attempts to match their behavior to sex role stereotypes? Child Development, 55, 1–24.
Poulin-Dubois, D., Serbin, L.A., Eichstedt, J.A., Sen, M.G., & Beissel, C.F. (1999). Men don’t put on make-up: Toddlers’ knowledge of the gender stereotyping of household activities. Manuscript submitted for publication.
Poulin-Dubois, D., Serbin, L.A., & Derbyshire, A. (1998). Toddlers’ intermodal and verbal knowledge about gender. Merrill-Palmer Quarterly, 44, 338–354.
Poulin-Dubois, D., Serbin, L.A., Kenyon, B. & Derbyshire, A. (1994). Infants’ intermodal knowledge about gender. Developmental Psychology, 30, 436–442.
Roopnarine, J.L. (1986). Mothers’ and fathers’ behaviors toward the toy play of their infant sons and daughters. Sex Roles, 14, 59–68.
Ruble, D.N., & Martin, C.L. (1998). Gender development. In W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3: Social, emotional and personality development (5th ed., pp. 933–1016). New York: Wiley.
Ruble, D.N., & Stangor, C. (1986). Stalking the elusive schema: Insights from developmental and social-psychological analyses of gender schemas. Social Cognition, 4, 227–261.
Sen, M.G. (1999). ‘‘Boys don’t make breakfast!’’: Differential memory for own- gender-stereotyped events by two-year-old children. Unpublished doctoral dissertation, University of Minnesota, Minneapolis.
Serbin, L.A., Powlishta, K.K., & Gulko, J. (1993). The development of sex typing in middle childhood. Monographs of the Society for Research in Child Development, 58 (2, Serial No. 232).
Signorella, M., Bigler, R.S., & Liben, L.S. (1993). Developmental differences in children’s gender schemata about others: A meta-analytic review. Special issue: Early gender-role development. Developmental Review, 13, 147–183.
Smetana, J.G., & Letourneau, K.J. (1984). Development of gender constancy and children’ s sex-typed free play behavior. Developmental Psychology, 20, 691–696.
Spelke, E.S. (1976). Infants’ intermodal perception of events. Cognitive Psychology, 8, 553–560.
Thompson, S.R. (1975). Gender labels and early sex role development. Child Development, 46, 339–347.
Walker-Andrews, A.S., Bahrick, L.E., Raglioni, S.S., & Diaz, I. (1991). Infants’ bimodal perception of gender. Ecological Psychology, 3, 55–75.
Weinraub, M., Clemens, L.P., Sockloff, A., Ethridge, T., Gracely, E., & Myers, B. (1984). The development of sex role stereotypes in the third year: Relationships to gender labelling, gender identity, sex-typed toy preference, and family characteristics. Child Development, 55, 1493–1503.
INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2001, 25 (1), 7–15 15
Copyright of International Journal of Behavioral Development is the property of Sage Publications Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.