Images and Play

Child and Mother Play in Three U.S. Cultural Groups: Comparisons and Associations

Linda R. Cote Eunice Kennedy Shriver National Institute of Child Health and Human Development and Marymount

University

Marc H. Bornstein Eunice Kennedy Shriver National Institute of Child

Health and Human Development

Child and mother play (n � 113 20-month-olds) among South American Latino immigrants, Japanese immigrants, and European Americans in the United States was investigated. Culturally universal patterns of play dominated the findings. For example, no cultural differences in the prevalence of exploratory or symbolic play were found for either children or their mothers. Regardless of their culture, boys engaged in significantly more exploratory and less symbolic play than did girls when they played by themselves. Few relations were found between child play in the two play sessions. Across cultural groups, children’s exploratory play was significantly positively related to both maternal demonstrations and solicitations of exploratory play. The results identify which realms of child growth, parenting, and family function call for special attention and cultural sensitivity, as well as which do not, in the dynamics of immigrant families.

Keywords: play, immigrant, mother, child

Observation of mothers and children engaged in univer- sal, developmentally appropriate activities can help to re- veal cultural influences on children’s development (Rogoff, 2003). Play constitutes such an activity during the 2nd year of life. Cross-cultural research on children’s play has high- lighted the importance of culture-specific practices in struc- turing and organizing the environment in which children’s social interactions and play activities take place (see Born- stein, Haynes, Pascual, Painter, & Galperı́n, 1999; Farver, 1999; Göncü, Tuermer, Jain, & Johnson, 1999; Haight, Wang, Fung, Williams, & Mintz, 1999).

During the 2nd year of life, children’s play normatively moves from being exploratory and object oriented to being

symbolic and pretense oriented. During the sensorimotor period, children’s initial interactions with objects are ex- ploratory in nature, and their schemas characterize objects (e.g., a ball) as something simply to be physically manipu- lated (Piaget, 1951/1962). However, as children begin to engage in symbolic play, they use familiar objects in new ways, for example, by pretending that a ball is an orange that they are eating. Although there are individual differ- ences in play development and related cognitive abilities, across cultures achieving representational competence is one of the principal developmental tasks of the 2nd year of life (Gönçü & Gaskins, 2007; Haight et al., 1999; Piaget, 1951/1962). However, cultural variation in play is also evident (Bornstein, 2007). Immigrant families may find culture-specific aspects of play in their culture of destination particularly challenging or perplexing during the process of acculturation, and culture-specific aspects of play from the immigrants’ culture of origin may be interpreted by clini- cians, teachers, or others as problematic simply because they differ from those of the culture of destination.

The Importance of Studying Immigrant Children’s Play

The study of immigrant children’s play is important for a variety of reasons. First, immigration is a central social and psychological concern today. For example, nearly 25% of children under the age of 18 in the United States are either immigrants themselves or the children of immigrants (Her- nandez, Denton, & Macartney, 2008), and the International Organization for Migration (n.d.) has estimated that approx- imately 192 million people currently live outside the coun- try of their birth or citizenship. Immigration is also a trans-

Linda R. Cote, Child and Family Research Program in Devel- opmental Neuroscience, Eunice Kennedy Shriver National Insti- tute of Child Health and Human Development, National Institutes of Health, U.S. Department of Health and Human Services, Bethesda, Maryland, and Department of Psychology, Marymount University; Marc H. Bornstein, Child and Family Research Pro- gram in Developmental Neuroscience, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Na- tional Institutes of Health, U.S. Department of Health and Human Services, Bethesda, Maryland.

This research was supported by the Intramural Research Pro- gram of the Eunice Kennedy Shriver National Institute of Child Health and Human Development. We thank O. M. Haynes, J. Lampard, M. Ogino, N. Okazaki, K. Painter, L. Pascual, D. Putnik, K. Schulthess, W. Smith, and S. Toda for assistance.

Correspondence concerning this article should be addressed to Linda R. Cote, Department of Psychology, Marymount University, 2807 North Glebe Road, Arlington, VA 22207. E-mail: linda.cote- [email protected]

Journal of Family Psychology In the public domain 2009, Vol. 23, No. 3, 355–363 DOI: 10.1037/a0015399

355

formational force in parents and children, affecting all aspects of their lives. For its part, play is a basic develop- mental milestone in child development and a key venue of early parent– child relationships. Finally, the achievement and sophistication of children’s pretend play has been linked to a variety of cognitive and social outcomes. Be- cause pretend play is one manifestation of the development of symbolic thought (Piaget, 1951/1962; Vygotsky, 1967), it is related to children’s representational competence, in- cluding language development (e.g., vocabulary, linguistic complexity, semantic diversity, and verbal fluency; e.g., Tamis-LeMonda & Bornstein, 1990). Language develop- ment, of course, is one area of particular concern for immi- grant children because many are being reared in bilingual homes and thus have the added challenge of mastering two languages before schooling (Hernandez et al., 2008). In the social domain, pretend play fosters children’s perspective- taking ability, including their ability to understand the thoughts, feelings, and intentions of others (Rubin & Howe, 1985; Youngblade & Dunn, 1995). Such interpersonal un- derstanding is critical to success in interpersonal relation- ships, including those with peers, which become increas- ingly important for all children as they enter social groups outside the home and also begin schooling.

Rationale for Selection of Cultural Groups

We chose for this study mother– child dyads from three cultural backgrounds that shared commonalities as well as dissimilarities with respect to general child-rearing goals and views about children’s play. Specifically, European American mothers socialize their children to be independent and autonomous (Harkness, Super, & Keefer, 1992; Har- wood, Leyendecker, Carlson, Asencio, & Miller, 2002; Richman, Miller, & Solomon, 1988), and one way they do so is by engaging in object-directed activities, such as exploratory play, that allow their children to interact with and gain control over the physical environment (Bornstein, Azuma, Tamis-LeMonda, & Ogino, 1990). In contrast, a major socialization goal of Latino families is for children to learn to act appropriately and respectfully toward other people to develop properly (become bien educado, or well educated; Harwood et al., 2002) rather than asserting their own wants and needs. Similarly, in Japan infants and young children are encouraged to interact with other people to develop properly (Caudill & Weinstein, 1969), relate to others empathically (omoiyari; Lebra 1994), and interact with others appropriately (DeVos, 1993; Morelli & Roth- baum, 2007; Shand, 1985). In both Latin America and Japan, socialization to develop interpersonal skills is em- bodied in person- rather than object-directed interactions (Bornstein, Haynes, et al., 1999; Caudill & Weinstein, 1969; Valenzuela, 1997), including engaging in pretend play and role play with others.

Consistent with these cultural differences in child-rearing goals, previous researchers have reported cultural differ- ences in play between these cultural groups. For example, European American children engage in more exploratory or object-directed play (both relative to symbolic play and in

comparison with children in Argentina and Japan), whereas demographically similar children in both Argentina and Japan engage in more symbolic, person-directed play (see Bornstein, Haynes, et al., 1999; Tamis-LeMonda, Born- stein, Cyphers, Toda, & Ogino, 1992). Although compari- sons of play among immigrant families are scarce, research has suggested that the play behaviors of Latino immigrants from South America and those of Japanese immigrants tend to differ from those of mothers and children in their respec- tive cultures of origin and appear to be similar to European American mothers and children (Cote & Bornstein, 2005). No research has yet compared the play of Latin American immigrants from South America with that of Japanese American immigrants from Japan, as we do here.

Relations Between Child and Mother Play

In addition to exploring intracultural differences in the mean levels of exploratory and symbolic play, we also investigated relations between child and mother play in each group. Because play is an activity through which parents socialize their children in the norms and values of their culture (Tamis-LeMonda, Katz, & Bornstein, 2002), we expected similarities in relations between mother and child play. Relations between children’s and mothers’ explor- atory play and children’s and mothers’ symbolic play have been found among mother– child dyads in Japan, Argentina, and European Americans in the United States (Bornstein, Haynes, et al., 1999; Tamis-LeMonda et al., 1992). On the basis of these findings, and because play is an activity through which cultural values are transmitted from mother to child (Tamis-LeMonda et al., 2002), we hypothesized that similar child–mother play would be positively related in all three U.S. cultural samples.

Cultural Differences in the Initiation of Child Play and Modality of Mothers’ Encouragement of Play

When European American mothers encourage children’s play by demonstrating or soliciting play, children’s play is more sophisticated than when children play alone or when children initiate play with their mothers (O’Connell & Bretherton, 1984; O’Reilly & Bornstein, 1993; Slade, 1987). However, cross-cultural differences characterize the extent to which children versus mothers initiate play (Haight et al., 1999). Specifically, European American mothers have been observed to follow their children’s lead and interests, which is consistent with European American mothers’ child-rearing goal of fostering their children’s autonomy (Harkness et al., 1992; Harwood et al., 2002; Richman et al., 1988) rather than dictating the educational agenda (Farver, 1999). Thus, we predicted that European American children would be more likely to initiate play interactions during the joint play session than would their mothers (i.e., European American children would engage in more child-directed than mother-directed play during joint play). In contrast, Latino parents in South America believe that caregivers need to direct children’s behavior if they are to learn obedience and become bien educado (Bornstein,

356 COTE AND BORNSTEIN

Haynes, et al., 1999; Valenzuela, 1997) and believe that they are assisting children’s development when they direct their children’s actions. Consistent with these child-rearing goals, Latino parents in South America have been observed to direct interactions with their children (Halgunseth, Ispa, & Rudy, 2006; Harwood et al., 2002). Previous research comparing European American and Japanese (sojourner) mothers’ interactions with their children in a shape-fitting task showed that Japanese mothers were more likely than European American mothers to change their children’s fo- cus of attention and keep them on task than were European American mothers, which is consistent with the Japanese child-rearing goal of fostering interdependence in the mother– child dyad (DeVos, 1993; Morelli & Rothbaum, 2007; Shand, 1985). Thus, we predicted that both Latino immigrant mothers and Japanese immigrant mothers would be more likely to initiate play interactions during the joint play session than would their children (i.e., children would engage in more mother-directed than child-directed play during joint play).

In addition to these within-culture differences in terms of whether mothers or children take the lead in play interac- tions, we hypothesized cultural differences with respect to play initiation. Specifically, because mothers in Japan view the mother– child relationship as reciprocal and prize em- pathy over directiveness (Bornstein et al., 1992), we hy- pothesized that Japanese immigrant and European Ameri- can children might engage in more child-directed and less mother-directed play than would Latino immigrant children.

Because cultural differences in the emphasis on adult verbal and nonverbal behaviors have been reported, we coded mothers’ play as (physical, nonverbal) demonstra- tions or (verbal) solicitations. Specifically, U.S. American and Latino adults value verbal assertiveness, and Japanese adults value verbal restraint (Caudill & Frost, 1972). In- deed, previous research with Americans of Latino and Jap- anese ethnicity has reported that Japanese American moth- ers speak less to their children than do South American Latinos (Bornstein & Cote, 2001). Thus, we hypothesized that Latino immigrant mothers would engage in more play solicitations and fewer play demonstrations than would

Japanese immigrant mothers. These differences in mother- ing styles would also have clinical implications for under- standing differential dynamics of interactions in immigrant families of different cultures of origin.

This is a follow-up study to Cote and Bornstein (2005), in which play among child–mother dyads in Argentina, Latino immigrants to the United States, and European Americans was compared, and play among child–mother dyads in Japan, Japanese American immigrants to the United States, and European Americans was compared. In that study, we found no differences between European Americans and either Latino immigrants or Japanese immigrants; however, the two immigrant samples were not compared with each other. Thus, this study extends previous work by (a) com- paring Latino and Japanese immigrants’ play, (b) analyzing relations between child play in two play situations (solitary and joint play), and (c) analyzing relations between child and mother play when engaged in joint play.

Method

Participants

One hundred thirteen 20-month-olds (ranging from 18.89 to 22.51 months) and their mothers from three U.S. Amer- ican cultural groups participated: South American Latino immigrants (n � 37), Japanese immigrants (n � 36), and European Americans (n � 40). All children were firstborn, healthy, and term, and approximately equal numbers of girls and boys participated. All mothers were married to their baby’s father, and the families were middle class. As can be seen in Table 1, participating families were demographi- cally similar to each other, regardless of cultural back- ground.

Participants were from a large East Coast U.S. metropol- itan area and are representative of middle-class mothers in their particular cultural group in that area (U.S. Census Bureau, 2001). Families were recruited by means of a variety of methods common to research with young children who are not yet school age (i.e., they were recruited from hospital birth notifications, patient lists of medical groups,

Table 1 Sociodemographic Characteristics of the Participants

Participant Latino immigrant

(n � 37) Japanese immigrant

(n � 36) European American

(n � 40) Differences

Child Gender (girls:boys) 16:21 17:19 19:21 �2(2, N � 113) � 0.17, ns Agea (days) (M[SD]) 626.11 (22.57) 614.94 (18.93) 612.78 (6.10) F(2, 110) � 6.56, p � .01a

Mother Age (M[SD]) 33.31 (4.67) 33.42 (4.05) 31.05 (5.39) F(2, 110) � 3.06, ns Educationb (M[SD]) 5.94 (0.79) 5.69 (0.79) 5.92 (1.07) F(2, 109) � 0.87, ns Hours of work per week (M[SD]) 18.25 (19.46) 12.18 (19.07) 16.17 (18.05) F(2, 107) � 0.95, ns

Family Nuclear family (%) 78.4 91.7 97.5 �2(2, N � 113) � 7.70, p � .05

Note. Sample sizes for some analyses are smaller because of missing data. a Latino immigrant children were older than Japanese immigrant or European American children. b Because differences exist between countries in the duration, quality, and content of schooling, bicultural researchers adjusted mothers’ years of schooling so that the scales were equivalent to the 7-point Hollingshead (1975) index.

357SPECIAL ISSUE: CHILD AND MOTHER PLAY

newspaper birth announcements, and advertisements in newspapers, but primarily from mass mailings). Before the conduct of these studies, parents of all participants were informed about the study and provided written consent. Treatment of human participants complied with the ethical standards defined by the American Psychological Associa- tion and was subject to oversight by the National Institute of Child Health and Human Development Institutional Review Board.

Immigrant mothers participating in this study self- identified as South American or Japanese American; this is an important methodological issue for research with ethnic minorities (Marı́n & Marı́n, 1991). Both sets of mothers were immigrants (using Hernandez et al.’s [2008] definition of immigrants). South American immigrant mothers’ first language was Spanish, and they were primarily from Ar- gentina, Colombia, and Peru. (In areas of the United States that lack a large concentration of a single Latino group, immigrants tend to identify themselves by their regional affiliation rather than by their country of origin; see Winn, 1992. Furthermore, empirically there were no differences within the South American immigrant group on any depen- dent variables.) Japanese immigrant mothers’ first language was Japanese. Latino immigrant and Japanese immigrant mothers were bicultural (as indicated on the 5-point South American Acculturation Scale or Japanese American Ac- culturation Scale; Bornstein & Cote, 2001): Latino immi- grant, M � 2.36 (SD � 0.43), and Japanese immigrant, M � 2.20 (SD � 0.66), t(69) � 1.22, ns. They were either first- or second-generation Americans (33:4 and 33:3, respec- tively), �2(1, N � 73) � 0.13, ns, and they had lived in the United States for a mean of 10.37 years (SD � 6.75) and 7.54 years (SD � 3.35), t(35.69) � 1.93, ns, respectively. European American participants were either fourth or fifth generation (i.e., most or all grandparents were born in the United States).

Procedures

Children and mothers were visited in their homes when the children were 20 months of age. Visits were scheduled at a time when the children were awake and alert, and no one was present except the child, mother, and observer. A standard set of toys that allowed for a variety of different play behaviors was placed on the floor in front of the child. After a 20- to 30-min acclimation period (McCune-Nicolich & Fenson, 1984), the child was videorecorded playing alone for 10 min and then the child and mother were videore- corded playing together for 10 min. Pilot testing undertaken before data collection ensured that the observational context was ethnographically valid (van de Vijver & Leung, 1997). The toy set (i.e., doll, blanket, tea set, telephone, train, two picture books, foam rubber ball, and nesting barrels) was chosen because it is both ethnographically valid and allows for a variety of different play behaviors (including explor- atory and symbolic play; McCune-Nicolich & Fenson, 1984). Mothers also completed a demographic question- naire.

Play Coding

Child and mother exploratory and symbolic play were coded using a mutually exclusive and exhaustive coding system derived from research on the early development of play (Bornstein & O’Reilly, 1993). The frequency and duration of the following levels of play were recorded: 0 � default (no play); 1 � unitary functional activity (e.g., throw the ball); 2 � inappropriate combinatorial activity (e.g., put the ball into a teacup); 3 � appropriate combinatorial activity (e.g., put the cup on the saucer); 4 � transitional play (e.g., put the telephone to the ear without speaking); 5 � self-directed pretense (e.g., drink from a cup); 6 � other-directed pretense (e.g., pretend to feed the doll); 7 � sequential pretense (e.g., dial and speak into the telephone); and 8 � substitution pretense (e.g., pretend the block is a telephone receiver and speak into it). Children’s play in the mother– child joint play session was also coded as either child initiated or mother initiated. Mothers’ play demonstra- tions (i.e., modeling an action for the child; e.g., pouring from the tea kettle into a teacup) and solicitations (i.e., verbally encouraging the child to participate in a specific play activity; e.g., suggesting that the child pour tea from the kettle into a teacup) were coded separately. Play was coded continuously by noting the start time and play level, and play at a given level continued until there was a 10-s or longer break or the player touched another toy when the end time was noted (to 1 s). Minimum play time was set to 1 s.

Coders were unaware of the hypotheses and were trained to reliability (�s � .60; intraclass correlation [ICC] � .80) on a standard set of videorecordings before they were al- lowed to code the videorecordings for this study. Addition- ally, interrater reliability was checked every 10 videorecord- ings to guard against coding drift. Interrater agreement was computed separately for each cultural group using kappa for behaviors and ICC for frequency of maternal verbal solic- itations based on approximately 15% of the videorecord- ings: child exploratory play, �s � .77; child symbolic play, �s � .76; initiation of child play, �s � .67; mother explor- atory play demonstrations, �s � .75; mother symbolic play demonstrations, �s � .71; mother exploratory play solici- tations, ICCs � .89; and mother symbolic play solicitations, ICCs � .92. Play reliability was not systematically related to cultural group.

Computing the Dependent Variables

Frequency and duration scores for the eight play levels were aggregated into exploratory (sum of Levels 1– 4) and symbolic (sum of Levels 5– 8) play scores. Four indices of play were computed, separately for play session (solitary, joint), separately for play partner (child, mother), for who initiated child play in the joint play session (child, mother), and for exploratory and symbolic: raw frequency of play bouts, the proportion of play bouts, raw duration of play, and the proportion of duration of play. These four indices (raw frequency, proportional frequency, raw duration, and proportional duration) have consistently been found to be positively and significantly correlated, including with the

358 COTE AND BORNSTEIN

current samples (e.g., Bornstein, Haynes, Legler, O’Reilly, & Painter, 1997; Cote & Bornstein, 2005). Therefore, the four indices were aggregated into mean standard scores, which resulted in eight dependent variables: mean z scores of child exploratory and symbolic play in the solitary play session, mean z scores of child exploratory and symbolic play in the joint play session that was child initiated or mother initiated, and mothers’ demonstrations of explor- atory and symbolic play. In addition, for maternal solicita- tions of child play, two indices of exploratory and symbolic play were developed: the raw frequency of play bouts that were exploratory or symbolic and the proportion of play bouts that were exploratory or symbolic. The two (raw and proportional) scores of frequency of play were then aggre- gated into two measures of maternal solicitations by taking the mean z scores separately for exploratory and symbolic play. This resulted in two additional dependent variables: maternal solicitations of exploratory and symbolic play.

Results

Examination of univariate and multivariate distributions of the dependent variables revealed that all were normally distributed. Because differences in girls’ and boys’ play have sometimes been reported (Bornstein, Haynes, et al., 1999; Tamis-LeMonda et al., 1992), we used child gender as a factor in the multivariate analyses of variance (MANOVAs). Furthermore, because child age differed among groups (Ta- ble 1), and we wanted to rule out child age as a possible confound (because older children are more likely than younger children to engage in symbolic play), we used child age as a covariate in all analyses. Wilks’s lambda is reported for all multivariate tests, and pairwise comparisons are t tests with Bonferroni’s correction. Because interaction ef- fects supersede main effects, we report only interaction effects. Because of space limitations, we report only signif- icant results; however, the tables contain all data—that is, means and standard deviations for each dependent variable controlled for child age presented separately by cultural background and child gender.

Child Solitary Play

We performed a repeated-measures analysis of variance (ANOVA) with two between-subjects factors, cultural background (Latino immigrant, Japanese immigrant, and European American) and child gender; one covariate (child age); and two dependent variables (child exploratory and symbolic play) to investigate cultural and gender differ-

ences in children’s engagement in exploratory and symbolic play when they played by themselves. There was a signif- icant multivariate interaction of Play Type � Gender, F(1, 107) � 4.57, p � .05, �p

2 � .04. Univariate ANOVAs revealed that boys engaged in significantly more explor- atory play (M � .14, SD � .83) and significantly less symbolic play (M � �.14, SD � .91) than did girls (ex- ploratory, M � �.19, SD � .83; symbolic, M � .20, SD � .92): exploratory, F(1, 107) � 4.51, p � .05, �p

2 � .04; symbolic, F(1, 107) � 3.84, p � .05, �p

2 � .04. Neither the multivariate main effect of cultural background, nor any of the interactions involving cultural background, was signif- icant. (See Table 2 for full data.)

We computed a post hoc power analysis to determine whether the sample size of 113 provided sufficient power to detect a medium-sized effect in an ANOVA design with two between-subjects factors with two and three levels and two dependent variables. With an effect size of .25 for between- subjects effects (Faul, Erdfelder, Lang, & Buchner, 2007), an alpha of .05, and a sample size of 113, the power estimate for this design was .99, indicating sufficient power to detect a medium or large Within � Between effect.

Child Joint Play

We performed a MANOVA with two between-subjects factors, cultural background (Latino immigrant, Japanese immigrant, or European American) and child gender; one covariate (child age); and four dependent variables (child- initiated exploratory and symbolic play, mother-initiated exploratory and symbolic play) to investigate cultural and gender differences in the kinds of play in which children engaged when they played with their mothers and whether play differed depending on who initiated play. A significant multivariate interaction of Cultural Background � Play Type � Initiator emerged, F(2, 106) � 3.89, p � .05, �p

2 � .07. Japanese immigrant children engaged in more explor- atory play (M � .41, SD � .73) and less symbolic play (M � �.39, SD � .83) when they initiated the play inter- action than when their mothers initiated play (maternal exploratory, M � �.05, SD � .94; symbolic, M � .12, SD � .83): exploratory, F(1, 106) � 7.04, p � .01, �p

2 � .06; symbolic, F(1, 106) � 8.49, p � .01, �p

2 � .07. Both Latino immigrant and Japanese immigrant children engaged in more exploratory than symbolic play when they initiated play: Latino immigrants, F(1, 106) � 9.44, p � .01, �p

2 � .08 (exploratory, M � .46, SD � .75; symbolic, M � �.24, SD � .85), and Japanese immigrants, F(1, 106) � 12.32,

Table 2 Cultural and Gender Differences in Child Solitary Play

Play type

Latino immigrant Japanese immigrant European American

Girls Boys Girls Boys Girls Boys

Exploratory play �.28 (.67) .06 (.72) �.09 (.66) .24 (.68) �.21 (.68) .12 (.69) Symbolic play .24 (.74) �.10 (.79) .18 (.73) �.16 (.75) .20 (.75) �.15 (.76)

Note. Data are means (standard deviations), controlling for child age.

359SPECIAL ISSUE: CHILD AND MOTHER PLAY

p � .001, �p 2 � .10 (means for Japanese immigrants are

reported above). No (cultural) differences emerged with respect to whether mothers or children initiated play overall (Table 3).

Mothers’ Play

We performed a MANOVA with two between-subjects factors, cultural background (Latino immigrant, Japanese immigrant, or European American) and child gender; one covariate (child age); and four dependent variables (mater- nal demonstrations and solicitations of exploratory and symbolic play) to investigate cultural and gender differ- ences in the modality that mothers used to encourage dif- ferent types of play. There was a significant interaction of Cultural Background � Play Modality, F(2, 108) � 5.54, p � .01, �p

2 � .09. Although univariate tests revealed a significant difference across cultural groups for demonstra- tions (wherein Japanese and South American immigrant mothers demonstrated more than European American moth- ers), F(2, 108) � 3.60, p � .05, �p

2 � .06, pairwise com- parisons did not reach statistical significance (p � .05). Univariate tests also showed that European American moth- ers solicited play significantly more (M � .04, SD � .39) than they demonstrated (M � �.13, SD � .28), F(1, 108) � 7.55, p � .01, �p

2 � .07 (Table 4). We computed a post hoc power analysis to determine

whether the sample size of 113 provided sufficient power to detect a medium-sized effect in a MANOVA design with two between-subjects factors (one factor having two levels and the other having three levels) and two within-subjects factors, each with two levels (i.e., four dependent variables). With an effect size of .25 for between-subjects effects (Faul et al., 2007), an alpha of .05, and a sample size of 113, the power estimate for this design was .84, indicating sufficient power to detect a medium or large Within � Between effect.

Relations Between Child Solitary and Joint Play

We performed Pearson correlations to explore relations between children’s play in the two play settings. As Table 5 shows, few relations between children’s play in the two play settings emerged.

Relations Between Child and Mother Play

We performed Pearson correlations to investigate rela- tions between similar kinds of child and mother play. As Table 6 shows, with one exception, all of these relations were significantly positively correlated in every cultural group (all rs � .34, ps � .05).

Discussion

Children’s Play

Our findings support species-general developmental views of mother– child object and representational play (Piaget, 1951/1962) and also point to the existence of species-general relations in child–mother play interactions (Bornstein, 1995). Taken together, our findings of few cul- tural differences in children’s play, but differences in play type (exploratory or symbolic) depending on whether moth- ers or children initiate play, confirm previous reports that children’s play tends to be more sophisticated when their mothers encourage them than when children initiate play with their mothers or play alone (O’Connell & Bretherton, 1984; O’Reilly & Bornstein, 1993; Slade, 1987). Moreover, our results extend these findings to two immigrant samples. Developmentally, children first engage in exploratory play, and over the course of the 2nd year of life, as their repre- sentational capacities advance, they begin to engage in symbolic play (Piaget, 1951/1962). From an applied or clinical vantage, these results suggest that the play of Jap- anese and South American immigrant children is similar to that of European American children in the sense that it is more sophisticated when play is mother rather than child initiated and that by playing with their mothers, children’s play develops.

Also concerning the universality of play, child gender emerged as a significant factor. Regardless of their culture, boys engaged in significantly more exploratory and less symbolic play than did girls when they played by them- selves (Lytton & Romney, 1991). These results confirm previous research on gender differences in children’s play that was conducted with European Americans, Latino fam- ilies in Argentina, and dyads in Japan (Bornstein, Haynes, et al., 1999; Tamis-LeMonda et al., 1992). It is noteworthy that these gender differences were found only for children’s

Table 3 Cultural and Gender Differences in Child Joint Play

Dependent variable

Latino immigrant Japanese immigrant European American

Girls Boys Girls Boys Girls Boys

Child initiated Exploratory play .33 (.57) .60 (.62) .28 (.58) .55 (.60) �.30 (.63) �.04 (.64) Symbolic play �.23 (.65) �.26 (.71) �.37 (.66) �.41 (.68) �.16 (.71) �.19 (.73)

Mother initiated Exploratory play .21 (.73) .24 (.79) �.06 (.74) �.04 (.77) .03 (.81) .06 (.82) Symbolic play �.13 (.65) �.18 (.70) .14 (.65) .10 (.68) �.18 (.71) �.23 (.73)

Note. Data are means (standard deviations), controlling for child age.

360 COTE AND BORNSTEIN

solitary play and not in child–mother joint play. It could be that girls, on average, experience more early caregiving socialization regarding interpersonal skills than do boys or that girls are more susceptible to early caregiving socializa- tion regarding interpersonal skills than are boys (Maccoby, 1988).

We found few relations between child exploratory play in the solitary and joint play sessions and few relations be- tween child symbolic play in the two play sessions. This suggests that across cultures, mothers influence their chil- dren’s play and that children play differently when their mothers are involved than when they play alone.

Mothers’ Play

European American mothers engaged in significantly more play solicitations than demonstrations (regardless of whether their play was exploratory or symbolic). This find- ing is consistent with previous research that has shown European American mothers stress verbal interactions (Bornstein, Tamis-LeMonda, & Haynes, 1999; Caudill & Schooler, 1973), although not necessarily in the play set- ting. Again, the parenting styles of European American mothers appear to differ from those of immigrant mothers, as do styles of family interaction and expectation.

Relations Between Child and Mother Play

As expected, each of the cultural groups evidenced a consistent pattern of interrelations between child and mother play. For example, children’s exploratory play (whether it was child initiated or mother initiated) was significantly positively related to both maternal demonstra- tions and solicitations of exploratory play in all three cul- tural groups. These concordances indicate that children and mothers in each of these cultural groups coordinate in their play. Mothers who played more in pretense at 13 and 20 months had children who played more in pretense (Vibbert & Bornstein, 1989). Relations between children’s and moth- ers’ exploratory and symbolic play have also been found in Argentine, Japanese, and European American families (Bornstein, Haynes, et al., 1999; Tamis-LeMonda et al., 1992). These findings broaden those results to exploratory play and also to acculturating groups.

Although the direction of effect in these correlational results is not clear, research has shown that when mothers encourage children’s play by demonstrating or soliciting, children’s play is more sophisticated than when children initiate play with their mothers or play alone (Bornstein & O’Reilly, 1993; O’Connell & Bretherton, 1984; Slade, 1987). Children are likely taught to play in different ways by their mothers and in ways valued by the culture in which their mother was reared. The finding that when boys played alone, they engaged in more exploratory than symbolic play but that this gender difference was absent when they played with their mothers lends further support to this conclusion. From a clinical or applied point of view, these findings suggest that regardless of ethnicity, parents might foster the

Table 4 Cultural and Gender Differences in Mothers’ Play

Dependent variable

Latino immigrant Japanese immigrant European American

Girls Boys Girls Boys Girls Boys

Demonstrations Exploratory play .13 (.60) .29 (.65) �.06 (.60) .10 (.62) �.08 (.62) .09 (.63) Symbolic play �.12 (.63) �.17 (.68) .04 (.63) �.01 (.65) �.23 (.65) �.28 (.66)

Solicitations Exploratory play .20 (.75) .52 (.81) �.07 (.75) .25 (.77) .02 (.76) .34 (.78) Symbolic play �.11 (.66) �.20 (.71) �.26 (.66) �.35 (.67) �.05 (.67) �.14 (.69)

Note. Data are means (standard deviations), controlling for child age.

Table 5 Relations Between Children’s Play in the Two Play Situations

Joint play

Solitary play

Exploratory Symbolic

Child-initiated exploratory Latino immigrant .33�

Japanese immigrant .01 European American .04

Mother-initiated exploratory Latino immigrant .12 Japanese immigrant �.07 European American .00

Child-initiated and mother-initiated collapsed, exploratory

Latino immigrant .29†

Japanese immigrant �.02 European American .04

Child-initiated symbolic Latino immigrant .15 Japanese immigrant .11 European American .30†

Mother-initiated symbolic Latino immigrant .04 Japanese immigrant �.17 European American .11

Child-initiated and mother-initiated collapsed, symbolic

Latino immigrant .12 Japanese immigrant �.02 European American .31�

Note. Data are Pearson correlations. † p � .10. � p � .05.

361SPECIAL ISSUE: CHILD AND MOTHER PLAY

character and quality of their children’s cognitive growth by structuring their play interactions according to those char- acteristics they hope to foster in their children. Parental goals should always be taken into account when assessing immigrants’ parenting behavior.

Conclusions

With respect to child development, parenting, and accul- turation, our study suggests that immigrant samples of chil- dren and their mothers in the United States tend to resemble each other and European American children and mothers in their culture of destination. This suggests that there may be cultural universals in the expression of (at least some) developmentally important processes, such as the develop- ment of child and mother play. Future research should further explore the extent to which play and other important developmental processes are culturally universal or culture specific. For clinicians and others who work with immigrant parents and children, our results point to realms of child growth, parenting, and family function that require special attention and cultural sensitivity.

References

Bornstein, M. H. (1995). Form and function: Implications for studies of culture and human development. Culture & Psychol- ogy, 1, 123–137.

Bornstein, M. H. (2007). On the significance of social relationships in the development of children’s earliest symbolic play: An ecological perspective. In A. Gönçü & S. Gaskins (Eds.), Play and development: Evolutionary, sociocultural, and functional perspectives (pp. 101–129). Mahwah, NJ: Erlbaum.

Bornstein, M. H., Azuma, H., Tamis-LeMonda, C., & Ogino, M. (1990). Mother and infant activity and interaction in Japan and in the United States: I. A comparative macroanalysis of naturalistic

exchanges. International Journal of Behavioral Development, 13, 267–287.

Bornstein, M. H., & Cote, L. R. (2001). Mother-infant interaction and acculturation I: Behavioral comparisons in Japanese Amer- ican and South American families. International Journal of Be- havioral Development, 25, 549 –563.

Bornstein, M. H., Haynes, O. M., Legler, J. M., O’Reilly, A. W., & Painter, K. M. (1997). Symbolic play in childhood: Interper- sonal and environmental context and stability. Infant Behavior and Development, 20, 197–207.

Bornstein, M. H., Haynes, O. M., Pascual, L., Painter, K. M., & Galperı́n, C. (1999). Play in two societies: Pervasiveness of process, specificity of structure. Child Development, 70, 317– 331.

Bornstein, M. H., & O’Reilly, A. W. (Eds.). (1993). The role of play in the development of thought. San Francisco: Jossey-Bass.

Bornstein, M. H., Tal, J., Rahn, C., Galperin, C., Lamour, M., Ogino, M., et al. (1992). Functional analysis of the contents of maternal speech to infants of 5 and 13 months in four cultures: Argentina, France, Japan, and the United States. Developmental Psychology, 28, 593– 603.

Bornstein, M. H., Tamis-LeMonda, C. S., & Haynes, O. M. (1999). First words in the second year: Continuity, stability, and models of concurrent and predictive correspondence in vocabu- lary and verbal responsiveness across age and context. Infant Behavior and Development, 22, 65– 85.

Caudill, W., & Frost, L. (1972). A comparison of maternal care and infant behavior in Japanese-American, American, and Jap- anese families. In U. Bronfenbrenner (Ed.), Influences on human development (pp. 329 –342). Hinsdale, IL: Dryden.

Caudill, W., & Weinstein, H. (1969). Maternal care and infant behavior in Japan and America. Psychiatry, 32, 12– 43.

Caudill, W. A., & Schooler, C. (1973). Child behavior and chil- drearing in Japan and the United States. Journal of Nervous and Mental Disease, 157, 323–338.

Cote, L. R., & Bornstein, M. H. (2005). Child and mother play in cultures of origin, acculturating cultures, and cultures of desti-

Table 6 Relations Between Child and Mother Play

Child

Mother

Exploratory demonstrations

Exploratory solicitations

Symbolic demonstrations

Symbolic solicitations

Child-initiated exploratory Latino immigrant .54��� .46��

Japanese immigrant .54��� .49��

European American .60��� .44��

Mother-initiated exploratory Latino immigrant .46�� .42��

Japanese immigrant .54��� .58���

European American .50��� .35�

Child-initiated symbolic Latino immigrant .58��� .36�

Japanese immigrant .37� .34�

European American .22 .44��

Mother-initiated symbolic Latino immigrant .61��� .51���

Japanese immigrant .37� .65���

European American .37� .64���

Note. Data are Pearson correlations. � p � .05. �� p � .01. ��� p � .001.

362 COTE AND BORNSTEIN

nation. International Journal of Behavioral Development, 29, 479 – 488.

DeVos, G. A. (1993). A cross-cultural perspective: The Japanese family as a unit in moral socialization. In P. A. Cowan, D. Field, D. A. Hansen, A. Skolnick, & G. E. Swanson (Eds.), Family, self, and society: Toward a new agenda for family research (pp. 115–142). Hillsdale, NJ: Erlbaum.

Farver, J. M. (1999). Activity setting analysis: A model for exam- ining the role of culture in development. In A. Göncü (Ed.), Children’s engagement in the world: Sociocultural perspectives (pp. 99 –127). Cambridge, England: Cambridge University Press.

Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G�Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavioral Research Methods, 39, 175–191.

Gönçü, A., & Gaskins, S. (Eds.). (2007). Play and development: Evolutionary, sociocultural, and functional perspectives. Hills- dale, NJ: Erlbaum.

Göncü, A., Tuermer, U., Jain, J., & Johnson, D. (1999). Children’s play as cultural activity. In A. Göncü (Ed.), Children’s engage- ment in the world: Sociocultural perspectives (pp. 148 –170). Cambridge, England: Cambridge University Press.

Haight, W. L., Wang, X., Fung, H. H., Williams, K., & Mintz, J. (1999). Universal, developmental, and variable aspects of young children’s play: A cross-cultural comparison of pretending at home. Child Development, 70, 1477–1488.

Halgunseth, L. C., Ispa, J. M., & Rudy, D. (2006). Parental control in Latino families: An integrated review of the literature. Child Development, 77, 1282–1297.

Harkness, S., Super, C. M., & Keefer, C. H. (1992). Learning to be an American parent: How cultural models gain directive force. In R. A. Paul & R. A. Shweder (Series Eds.) & R. G. D’Andrade & C. Strauss (Vol. Eds.), Publications of the Society for Psycho- logical Anthropology: Human motives and cultural models (pp. 163–178). Cambridge, England: Cambridge University Press.

Harwood, R., Leyendecker, B., Carlson, V., Asencio, M., & Miller, A. (2002). Parenting among Latino families in the U.S. In M. H. Bornstein (Ed.), Handbook of parenting: Vol. 4. Applied parenting (2nd ed., pp. 21– 46). Mahwah, NJ: Erlbaum.

Hernandez, D. J., Denton, N. A., & Macartney, S. E. (2008). Children in immigrant families: Looking to America’s future. Social Policy Report, 22(3).

Hollingshead, A. B. (1975). The four-factor index of social status. Unpublished manuscript, Yale University.

International Organization for Migration. (n.d.). About migration. Retrieved July 10, 2008 from http://www.iom.int/jahia/Jahia/ lang/en/pid/3

Lebra, T. S. (1994). Mother and child in Japanese socialization: A Japan-U.S. comparison. In P. Greenfield & R. Cocking (Eds.), Cross-cultural roots of minority child development (pp. 259 – 274). Hillsdale, NJ: Erlbaum.

Lytton, H., & Romney, D. M. (1991). Parents’ differential social- ization of boys and girls: A meta-analysis. Psychological Bulle- tin, 109, 267–296.

Maccoby, E. E. (1988). Gender as a social category. Developmen- tal Psychology, 24, 755–765.

Marı́n, G., & Marı́n, B. V. (1991). Applied social research meth- ods series: Vol. 23. Research with Hispanic populations. New- bury Park, CA: Sage.

McCune-Nicolich, L., & Fenson, L. (1984). Methodological issues in studying early pretend play. In T. D. Yawkey & A. D. Pelligrini (Eds.), Child’s play: Developmental and applied (pp. 81–124). Hillsdale, NJ: Erlbaum.

Morelli, G., & Rothbaum, F. (2007). Situating the child in context:

Attachment relationships and self-regulation in different cultures. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psy- chology (pp. 500 –527). New York: Guilford Press.

O’Connell, B., & Bretherton, I. (1984). Toddlers’ play, alone and with mother: The role of maternal guidance. In I. Bretherton (Ed.), Symbolic play: The development of social understanding (pp. 337–368). San Diego, CA: Academic Press.

O’Reilly, A. W., & Bornstein, M. H. (1993). Caregiver-child interaction in play. In M. H. Bornstein & A. W. O’Reilly (Eds.), The role of play in the development of thought (pp. 55– 66). San Francisco: Jossey-Bass.

Piaget, J. (1962). Play, dreams, and imitation in childhood (C. Gattegno & F. M. Hodgson, Trans.). New York: W. W. Norton. (Original work published 1951)

Richman, A. L., Miller, P. M., & Solomon, M. J. (1988). The socialization of infants in suburban Boston. In W. Damon (Series Ed.) & R. A. LeVine, P. M. Miler, & M. M. West (Vol. Eds.), New directions for child development series: No. 40. Parental behavior in diverse societies (pp. 65–74). San Francisco: Jossey- Bass.

Rogoff, B. (2003). The cultural nature of human development. London: Oxford University Press.

Rubin, K. H., & Howe, N. (1985). Toys and play behaviors: An overview. Topics in Early Childhood Special Education, 5(3), 1–10.

Shand, N. (1985). Cultural influence in Japanese and American maternal role perception and confidence. Psychiatry, 48, 52– 67.

Slade, A. (1987). A longitudinal study of maternal involvement and symbolic play during the toddler period. Child Development, 58, 367–375.

Tamis-LeMonda, C. S., & Bornstein, M. H. (1990). Language, play, and attention at one year. Infant Behavior & Development, 13, 85–98.

Tamis-LeMonda, C. S., Bornstein, M. H., Cyphers, L., Toda, S., & Ogino, M. (1992). Language and play at one year: A comparison of toddlers and mothers in the United States and Japan. Interna- tional Journal of Behavioral Development, 15, 19 – 42.

Tamis-LeMonda, C. S., Katz, J. C., & Bornstein, M. H. (2002). Infant play: Functions and partners. In A. Slater & M. Lewis (Eds.), Introduction to infant development (pp. 229 –243). New York: Oxford University Press.

U.S. Census Bureau. (2001). Profile of the foreign-born population in the United States: 2000 (Current Population Reports, Series P23–206) [Electronic version]. Washington, DC: U.S. Govern- ment Printing Office.

Valenzuela, M. (1997). Maternal sensitivity in a developing soci- ety: The context of urban poverty and infant chronic undernutri- tion. Developmental Psychology, 33, 845– 855.

van de Vijver, F. J. R., & Leung, K. (1997). Methods and data analysis for cross-cultural research. Thousand Oaks, CA: Sage.

Vibbert, M., & Bornstein, M. H. (1989). Specific associations between domains of mother-child interaction and toddler refer- ential language and pretense play. Infant Behavior and Develop- ment, 12, 163–184.

Vygotsky, L. S. (1967). Play and its role in the mental develop- ment of the child. Soviet Psychology, 5(3), 6 –18.

Winn, P. (1992). Americas: The changing face of Latin America and the Caribbean. Berkeley: University of California Press.

Youngblade, L. M., & Dunn, J. (1995). Individual differences in young children’s pretend play with mother and sibling: Links to relationships and understanding of other people’s feelings and beliefs. Child Development, 66, 1472–1492.

Received April 30, 2008 Revision received January 9, 2009

Accepted January 12, 2009 �

363SPECIAL ISSUE: CHILD AND MOTHER PLAY