contextual differences in parentpeelearlyyears.com/pdf/contextual differences in parent.pdf ·...
TRANSCRIPT
Contextual Differences in Parent-Child Play: Implications for Children's Gender Role Development - Statistical Data Included
Sex Roles: A Journal of Research, Feb, 2001 by Eric W. Lindsey, Jacquelyn Mize
Save a personal copy of this article and quickly find it again with Furl.net. It's free! Save it.
Eric W. Lindsey [1]
Parent--child play behavior of 33 preschool children (18 boys, 29 European-American,
middle- and upper-middle-class families) was videotaped in separate pretend and
physical play sessions. Children's play behavior with a same-sex peer also was
observed. Analyses focus on contextual differences in parent-child play behavior, as
well as associations between parent-child play and child-peer play. During the pretense
play session parent-daughter dyads, particularly mother-daughter dyads, engaged in
more pretense play than did parent-son dyads. During the physical play session father-
son dyads engaged in more physical play than did father-daughter dyads. These data
suggest that context may play an important role in gender differentiated patterns of
parent- child play behavior. As for children's peer play behavior, consistent with
previous evidence, girls were more likely than boys to engage peers in pretend play and
boys were more likely than girls to play physically with peers. Children whose parents
engaged in m ore pretense play engaged in more pretense play with a peer, whereas
children's whose parents engaged in more physical play engaged in more physical play
with a peer. These findings suggest that parents may contribute to children's gender-
typed play behaviors with peers.
In the search for the origins of children's gender role development, research has focused
on parents' differential pattern of interactions with sons and daughters (see Fagot, 1997,
for recent review). Examination of this literature reveals that often no significant child
gender effects are found in parents' socialization behavior (Lytton & Romney, 1991;
Maccoby & Jacklin, 1974). One explanation for the lack of evidence concerning gender
effects in parents' socialization comes from contextual interaction models of gender-
typed behavior (Beall, 1993; Deaux & Major, 1987). These models suggest that the
salience of gender as a determinant of behavior varies from one situation to another.
According to such models, parents' differential treatment of boys and girls may depend
upon the particular context of parent--child interaction. Empirical support for this
proposition comes from studies in which gender based variations in parent and child
behavior were observed during certain play activities, but not others, and during play
with particular toys (Caldera, Huston, & O'Brien, 1989; Leaper & Gleason, 1996;
Leaper, Leve, Strasser, & Schwartz, 1995; O'Brien & Nagle, 1987). In addition, a
recent meta-analysis by Leaper, Anderson, and Sanders (1998), focusing on studies
comparing mothers' and fathers' use of language with their children, found that the
magnitude of gender differences observed in parents' use of language varied as a
function of the toys available to parent--child dyads across studies. This evidence
suggests that failure to consider the role of context may account for the lack of gender
effects in studies of gender-differentiated patterns of parent--child interaction.
Consequently, previous research may have underestimated parent's contribution to
children's gender role development.
Continue article
Advertisement
Context may play a role in parents' socialization of children's gender role development
in several ways. For instance, parents' choice of context in which to interact with
children may represent a subtle form of gender role socialization. That is, by choosing
to interact with boys and girls in different contexts, parents may lead children to
develop gender schema linking female and male roles to particular contexts. For
example, a father who participates with his son in yardwork, but not interior
housework, may teach his son to associate yardwork with the male gender role, but not
interior housework. In this way, parents may pass on their own gender schema to
children by defining certain contexts as being more appropriate for females or males.
Another way in which context may affect parent--child interaction is by determining
the behaviors parents and children display toward one another. That is, certain behavior
patterns may be more prevalent in some contexts than in others. For example, parents
may use more indirect verbal strategies during teaching activities and may participate
in teaching activities more often with daughters than with sons, providing girls with
more exposure to indirect verbal strategies. Thus, by interacting with boys and girls in
different contexts, parents may be constraining the types of behaviors boys and girls
observe and practice. Studies are needed to determine the extent to which parents'
interaction with sons and daughters is more likely to occur in particular contexts, and
how mothers' and fathers' behavior may differ across contexts, in order to elucidate
how children may acquire gender-typed behavioral patterns from parents.
When examining the role of context in gender-typed patterns of socialization, the
question "What are the most salient contexts for possible gender-typed interaction
patterns between parents and children?" must be considered. Turning to the literature
on parental socialization of gender-typed behavior, the most consistent evidence to
support gender role socialization involves parents' encouragement of gender-typed play
activities in their children (see Lytton & Romney, 1991). Specifically, evidence
suggests that parents may influence gender-based differences in children's involvement
in pretense and physical play. Parents, particularly fathers, engage in more physical
play with sons than with daughters (Jacklin, DiePietro, & Maccoby, 1984; MacDonald
& Parke, 1986), and mothers engage children in more pretense play (Langlois &
Downs, 1980), particularly with daughters (Tamis-LeMonda & Bornstein, 1991). This
play preference pattern may encourage children to view pretense play as being more
appropriate for the fe male gender role and physical play as being more appropriate to
the male gender role. Consistent with this proposal, studies examining children's peer
play behavior indicate that rough-and-tumble forms of play are more characteristic of
boys than of girls (DiPietro, 1981; Maccoby, 1990), whereas pretense play is more
characteristic of girls than of boys (Howes, Unger, & Matheson, 1991). Together this
body of evidence suggests that parents may contribute to boy's preference for physical
play and girl's preference for pretense play. However, no study to date has examined
gender differences in parent--child play during separate pretense and physical play
contexts. Thus, it is not clear to what extent parents may display different levels of
involvement with boys and girls across pretense and physical play.
Interestingly, there is evidence to suggest that parent--child pretense and physical play
may also affect children's gender-role socialization by promoting or constraining the
expression of certain types of behavior by parent and child. Gender-role theory
suggests that parents encourage self-assertive behavior in boys and relationship
enhancing behaviors in girls (Block, 1983). Empirical support for such claims is
equivocal in that some evidence suggests that parents engage in behaviors that promote
independence and self-assertion in sons and behaviors that encourage cooperation and
relationality in daughters (Cherry & Lewis, 1976; Dunn, Bretherton, & Munn, 1987;
Frankel & Rollins, 1983), whereas other studies find no difference in parents use of
assertive behavior with sons and daughters (Crockenberg & Litman, 1990; Gralinski &
Kopp, 1993). It may be that the context in which parent--child interaction was observed
accounts for the discrepancies between studies, however, the extent to which parent and
child b ehaviors are dependent upon particular contexts is not clear. There is some
evidence to suggest that physical play represents a context that facilitates the use of
directive and self-assertive behavior, in that parent-child physical play has been linked
to parents' use of directives and power assertive control strategies (MacDonald, 1987;
MacDonald & Parke, 1984). In contrast, pretense play has been associated with
cooperative and dyadic forms of play between parent and child (Dunn & Brown, 1994).
Such evidence supports the possibility that gender-typed patterns of parent--child
interaction may represent behavioral characteristics that are dependent upon the context
of observation. Alternatively, these patterns may be present across contexts. In order to
explore the possibility that particular gender-typed behaviors are more prevalent in
pretense or physical play, it is important to compare parent and child behaviors across
pretense and physical play contexts.
Contextual Differences in Parent-Child Play: Implications for Children's Gender Role Development - Statistical Data Included
Sex Roles: A Journal of Research, Feb, 2001 by Eric W. Lindsey, Jacquelyn Mize
Save a personal copy of this article and quickly find it again with Furl.net. It's free! Save it.
Continued from page 1.
A basic premise of research on parent-child gender role socialization is that the gender-
typed behaviors learned in the family are carried by children to other social contexts.
The majority of studies, however, have examined gender-typed differences in parent--
child interaction or examined children's gender-typed behaviors with peers or
individuals outside the family. Few studies have directly compared gender-typed
behavior in the parent--child dyad with children's behavior with peers (Bussey &
Bandura, 1999). In the present study, children were observed playing with a same-sex
peer, using the same toys as during the parent--child play session. It was predicted that
gender-typed behavior patterns between parent and child would be mirrored in
children's interactions with peers. That is, girls would engage in more pretense play
than boys and that boys would engage in more physical play than girls. In addition, it
was predicted that girls would use cooperative and facilitative initiation strategies with
peers, w hereas boys would use assertive initiation strategies with peers.
In summary, the present study examines the role of context in gender-typed patterns of
interaction between parent and child. Specifically, observations of parent--child
interaction were conducted in both a pretense and a physical play session. Parent--child
joint involvement in pretense and physical play was assessed across the two contexts. It
was hypothesized that parent-daughter dyads would engage in more pretense play
during the pretense play session, whereas parent--son dyads would engage in more
physical play during the physical play session. In addition, parent--child initiation
strategies were assessed across the two sessions. It was predicted that less assertive and
more polite initiation strategies would be made during the pretense play context,
whereas more assertive and forceful initiations would be made during the physical play
context. Furthermore, it was predicted that parents would display more polite and
relationship enhancing forms of initiations with girls than with boys, and more
instrumen tal and assertive initiation strategies with boys than with girls, across both
play contexts. In addition, child gender-typed behavior was assessed during a play
session with a same sex peer. Similar gender-typed patterns of interaction were
expected to be observed in children's behavior during peer play. That is, boys would
engage in more physical play than girls and girls would engage in more pretense play
than boys, and that boys would use more assertive initiation strategies than girls,
whereas girls would use more polite, egalitarian initiation strategies than boys.
METHOD
Participants
Participants in this study were involved in a larger investigation examining links
between parents' child-rearing beliefs, parent-child interaction, and children's
competence with peers. Parents of children attending a university sponsored preschool
program were asked to participate. Of the 92 intact families available for participation,
81 mothers and fathers agreed to participate in the larger study and completed
questionnaire packets distributed at the beginning of the school year. Subsequently, 54
mothers and 36 fathers were able to be scheduled for the laboratory phase of data
collection. Of these, there were 33 mother and father pairs and it is these families that
are the focus of the current report. There were 18 boys and 15 girls between the ages of
43 and 80 months (M = 60.36). Twenty-nine of the families were European-American,
2 were African-American, and 2 were of other ethnic origin. The families were
primarily from the middle- and upper-middle class, with 82% of fathers being
employed in profess ional occupations (based on Entwisle & Astone, 1994, Total-
Based Socioeconomic Index).
Parent-Child Interaction
During the spring of the school year, mothers and fathers made separate visits to the
laboratory with their child (in counter-balanced order), typically separated by a 2-3-
week interval. While staying in a comfortably furnished room (containing a sofa,
tables, and magazines) parent-child pairs were videotaped from behind a one-way
mirror during a 1-hr period composed of four interaction sessions: a puzzle task, book
reading, pretense play, and physical play. The puzzle task and book reading session
were designed to assess parents' teaching behavior as part of the larger research project.
The present report focuses on parent-child interaction during the third and fourth
session, which involved 10-mm of pretense play and 10-mm of physical play,
respectively. [2] The toys used in both the pretense and physical play session were
selected based on pilot testing and previous research suggesting that they were
conductive to parent-child pretense and physical play (Lindsey, Mize, & Pettit, 1997a,
1997b; MacDonald & P arke, 1984; Stevenson, Leavitt, Thompson, & Roach, 1988).
At the beginning of the third session (pretense play) the researcher entered the room
with a laundry basket full of toys designed to elicit pretense play (e.g., a set of blocks, a
set of zoo animals, a pretense cooking set, people figures, and a set of plastic vehicles)
and asked the parent to "play with your child as you would if you and your child were
home together." The researcher informed the parent--child pair that they had 10 mm to
play with the toys, after which he/she would return to ask the parent and child to clean
up the toys. Once 10 min elapsed, the researcher reentered the room to notify the parent
and child that they had about 5 min to clean up. The researcher began the next session
as soon as the toys were returned to the laundry basket or after 5 min, whichever came
first.
At the beginning of the fourth session (parent--child physical play), the researcher
reentered the room with a basket full of toys designed to elicit physical play (e.g., a set
of Nerf bats and balls, a pair of animal puppets, and a large Slo-Mo ball) and again
asked the parent to "play with your child as you would if you and your child were home
together." The researcher informed the parent and child that they had about 10 min to
play with the toys, after which he/she would return to end the session. Parent and child
behaviors in the pretense and physical play sessions were coded for a variety of play
style and interaction behaviors.
Form of Play
Based on definitions adapted from Stevenson et al. (1988), each 10-s segment of
parent--child interaction was coded for the type of play in which child was engaged.
The six play categories used in this study were functional play, intentional
manipulation of objects to elicit their properties (e.g., shaking, rolling); physical play,
any playful contact or gross motor activity between partners (e.g., tickling, running,
playing with bat and ball); instructive play, naming or requesting naming of objects,
colors, or numbers; pretense play, using play objects to represent other objects and
assuming play roles or both including verbal relabeling of objects or role
transformations; construction, building, stacking, arranging of objects or arranging
objects within, or on a construction made of blocks; and other play, any play activity
not fitting into one of the above categories (e.g., singing, drawing). In addition, there
were two nonplay categories: other interaction, engagement in any interaction that does
not inv olve play (e.g., discussing events of the day or the location of another family
member); and nonplay/noninteraction, any activity that does not involve play and that
does not involve partner (e.g., reading magazine alone, looking out the window alone).
Because it was possible that children might engage in multiple forms of play at the
same time (e.g., combining both instruction and construction play by counting the
blocks while building a tower; combining both physical and pretense play by
pretenseing the large ball is a giant rock being lifted and thrown) coders could assign
multiple play codes to a particular 10-s episode. In addition, coders recorded the
specific type of toy or toys with which children were playing. (3)
Coders also identified parents' level of engagement in their children's activity in each
10-s interval based on the following scale: 1 = parent is not involved in the same
activity as his or her child (i.e., parent is engaged in some other form of play, but not
with child, and does not interact with or attend to the child for the entire 10-s segment);
2 = parent is watching and attending to child's activity but does not actively participate
or talk to the child during the 10-s segment; 3 = parent talks about the child's activity
but does not actively participate in the activity at any time during the 10-s segment; 4 =
parent is actively engaged in the same activity as the child at least once during the 10-s
segment (e.g., parent has puppet/dinosaur/doll in hand or talks in different voice, parent
knocks down blocks or plays ball with child).
Mother--child and father--child videotapes were coded separately by two assistants,
with the first author serving as reliability check on six tapes (17%) for both mothers
and fathers. The overall interrater reliability (K) for play form was .83 and .80 for
mothers and fathers, respectively, and for parent involvement .45 (90% agreement) and
.53 (96% agreement) for mothers and fathers, respectively. The K for parent
involvement was low because parents were highly involved with children in most
intervals (i.e., received a rating of 4), thus the distribution of scores were uneven in the
computational formula, which lowers the K statistic.
For the purpose of this study, only parent--child pretense play and parent-child physical
play were examined. In addition, because our focus was on pretense and physical play,
doubled coded intervals that included pretense play during the pretense play session
were counted as pretense play and double coded intervals that included physical play
during the physical play session were counted as physical play. [4] Variables were
created to represent joint parent--child pretense and physical play based on the
proportion of intervals in which the child was engaged in pretense or physical play and
parents received a rating of 4 (i.e., actual engagement with the child in the child's
activity), divided by the total number of 10-s intervals in which child was coded as
being engaged in physical or pretense play. Thus, each parent-child dyad received two
measures of joint parent-child play (i.e., mother-child joint pretense play, mother-child
joint physical play, father-child joint pretense play, and father-child joint p hysical
play) for each play session.
Parent and Child Initiations
Videotapes of parent-child play sessions were also coded using an event-based coding
scheme for the occurrence of initiations, and responses to initiations, for both parent
and child. The particular coding scheme used in this study was a modified version of
one used previously by Lindsey et al. (1997a, 1997b; see also Lindsey & Mize, in
press). Coders noted the exact time a statement, question, or other behavior intended to
initiate interaction or influence the behavior of the partner was made, who made the
initiation, and recorded both the initiation and the play partner's response verbatim.
Coders were judged to have agreed when an event was coded by both observers and the
recorded time differed by no more than 10 s. In 86% of the cases recorded times for the
same event differed by no more than 5 s. Overall reliability for initiations and
responses was K = .90 and .83, respectively. Initiation codes are described in detail in
the following paragraphs.
Initiations were coded into one of five categories: leads, an initiation that offers the
partner a choice of whether to comply or not comply (e.g., "Wanna wrestle?"); requests
for permission, an initiation that is phrased in the form a question that asks the partner's
permission to perform some action (e.g., "Do you want me to help?"); requests for
information, an initiation that simply asks the listener for information, that provides
information to the listener, or that serves as a teaching question (e.g., "What is this
dinosaur called?"); polite commands, an initiation that offers the partner no choice but
to comply, but is phrased in a polite and courteous way (e.g., "Why don't you get the
other bat?"); and imperatives, an initiation that offers the partner no choice but to
comply and is power assertive (e.g., "Don't move that truck").
Each initiation was also identified as being either a play initiation or nonplay initiation,
depending upon its intent and the action in which it was embedded. Play initiations
were identified as those given with the intent to change or influence a partner's play
behavior and those given while the dyad was involved in some play activity. Nonplay
initiations were identified as those that were not related to ongoing play, that were
given with the intent to influence the partner's behavior with regard to nonplay
materials (e.g., furnishings of the room, clothing, etc.), and those that were given
during the clean-up period between the pretense and physical play session (e.g., "Will
you sit by me?" "Can I use the bathroom?" "Pick up the blocks, please," "Get away
from the mirror").
In addition to initiations, coders recorded the responses parent and child made to each
initiation they received. However, partners' responses were not considered in the
present report and thus are not described or discussed further.
From the raw frequencies of the event level codes, scores were created representing the
rate of parent and child play initiations. Specifically, five measures of play initiation
type were created for both parent and child (rate of play leads, rate of requests for
information, rate of requests for permission, rate of polite commands, and rate of
imperatives) by dividing the frequency of each type of play initiation for that individual
(parent or child) by the number of minutes in the observation session.
Child-Peer Interaction
From 3 to 5 weeks after being observed with their parents, each child was videotaped in
a 15-min play session with a friend from his/her class. Same-sex friendship dyads were
identified on the basis of children's mutual nominations as preferred playmates during
sociometric interviews conducted in each classroom (see Lindsey & Mize, in press). To
be identified as friends, both children nominated one another as a preferred playmate
during their sociometric interview. If there was no mutually nominated friend,
children's ratings of their classmates were used to identify a same-sex peer who was
liked "a lot" and gave a similar "like a lot" rating for the target child. In cases where
both children in a mutual friendship pair had parents who participated in the parent-
child play session, one child was chosen at random as the target child and the other
child was observed in a play session with a child who had given reciprocal "like a lot"
ratings, but whose parents did not participate in the study. In this way, dat a from child-
peer play session were used only once for either member of the child-peer dyad. Of the
33 children whose parents participated in the study, 29 were willing to be videotaped
while playing with a friend. Four children declined to participate in the child-peer
observation sessions. Seven of the 29 children who were observed were paired with a
playmate, using the less stringent peer rating criteria of friendship.
Videorecords of children's friendship dyad play session were coded by two research
assistants who had not participated in any other phase of the data coding using the same
play form coding scheme used to code parent-child play. However, only the target
child's behavior was coded. The first author served as reliability check on six tapes
(20%). The overall interrater reliability (K) for play form was .86. Videotapes of child-
peer play were also coded by a research assistant who had not participated in any other
phase of the data coding using the same event-based coding scheme used for the
parent-child play session. The first author served as reliability check on six tapes.
Overall reliability for the child-peer initiations was K = .78. Both children's behaviors
were relevant for the initiation codes, however, only data for the child who was
designed as the target child, or whose parents also participated in the study, are
considered in this report.
RESULTS
Arc sine-transformed variables were used in all analyses involving proportions,
however, for ease of interpretation, all means presented in the tables and text are the
raw, untransformed proportions. Preliminary correlation analyses were conducted to
determine if parent and child behaviors varied as a function of child age or
demographic variables. Results revealed a significant negative association between
father involvement in children's pretense play during the pretense play session and
children's age (r = -.47, p [less than] .01) indicating that fathers of older children were
less involved in their children's pretense play. In addition, mother's use of play leads, in
both the pretense and physical play session, was negatively associated with children's
age (r = -.45, p [less than] .01 and r = -.47, p [less than] .01, for pretense and physical
play session, respectively), indicating that mothers were more likely to use play leads
with younger children than with older children. There were no significant co rrelations
between children's peer play behavior and family demographic variables. In addition,
there were no demographic differences between families with boys and families with
daughters, thus there was no need to control for demographic characteristics in
subsequent analyses.
Sex Differences in Parent-Child Joint Play Within and Across Play Contexts
In order to examine the hypothesized gender differences in parent--child joint pretense
and joint physical play, measures of joint parent--child play were subjected to a 2 x 2 x
2 x 2 (sex of child x sex of parent x play-form x play session) repeated measures
multivariate analysis of variance (MANOVA, see Table I). Parent sex, play form
(pretense vs. physical), and play context (pretense vs. physical) were within subjects
variables, whereas child sex was a between subjects variable. This MANOVA revealed
a significant main effect for play session, as well as significant play form x play
session, and sex of child x sex of parent x play form x play session, interactions.
Follow up, one-way analysis of variances (ANOVAs, see Table II) indicated that (1)
more joint play occurred in the physical play session than in the pretense play session,
F(1, 33) = 12l.13, p [less than] .001, (2) joint pretense play occurred more often in the
pretense play session than in the physical play session, F(1, 33) = 89.42, p [less than]
.001, and (3) joint physical play occurred more often in the physical play session than
in the pretense play session, F(1, 33) = 156.32, p [less than] .001. Consistent with
hypothesized gender differences involving parents' engagement in pretense play with
sons and daughters, during the pretense play session mothers were more likely to be
involved in pretense play than fathers, F(1, 33) = 5.14, p [less than] .05, and both
mothers and fathers were more likely to be involved in pretense play with girls than
with boys, F(1, 33) = 4.02, p [less than] .05. Furthermore, partial support was found for
the hypothesized gender differences in physical play, in that during the physical play
context fathers were more likely to be involved in physical play with boys than with
girls, F(1, 33) = 7.2l, p [less than] .01.
Sex Differences in Parent--Child Play Behavior Within and Across Play Contexts
Parent Initiations
In order to examine the hypothesized influence of context and parent--child sex on
parents use of initiations, measures of parent initiations were subjected to a 2 x 2 x 2 x
5 (sex of child x sex of parent x play session x initiation type) repeated measures
MANOVA (see Table I). Parent sex, play context (pretense vs. physical), and initiation
type (play lead, request for permission, request for information, polite command, and
imperative) were within subjects variables, whereas child sex was a between subjects
variable. This MANOVA yielded significant main effects for sex of parent, and play
context, as well as significant sex of parent x initiation type, play session x initiation
type, and sex of child x initiation type x sex of parent, interactions.
Follow up, one-way ANOVAs (see Table III) indicated that (1) fathers gave more
initiations than mothers, F(1, 33) = 4.23, p [less than] .05, and (2) parents gave more
initiations during the physical play context than during the pretense play context, F(1,
33) = 6.42, p [less than] .05. Furthermore, compared to mothers, fathers gave more play
imperatives, F(1, 33) = 8.24, p [less than] .01, and polite commands, F = 4.04, p [less
than] .05, whereas mothers gave more play leads than fathers, F(1, 33) = 4.03, p [less
than] .05. This finding is consistent with the proposition that fathers use more assertive
initiation strategies with children than do mothers. Support for the hypothesis that
different types of initiation strategies would characterize pretense play and physical
play was found in that parents gave more play leads, F(1, 33) = 7.11, p [less than] .05,
and requests for permission, F(1, 33) = 9.14, p [less than] .01, during pretense play than
during physical play, but gave more imperatives during physica l play than during
pretense play, F(1, 33) = 11.17, p [less than] .01. In addition, fathers gave more
imperatives to boys than to girls, F(1, 33) = 10.32,p [less than] .01, and gave more play
leads, F(1,33) = 6.03, p [less than] .05, and polite commands, F(1, 33) = 4.31, p [less
than] .05, to girls than to boys. Mothers gave more play leads, F(1, 33) = 7.06, p [less
than] .05, polite commands, F(1, 33) = 3.87, p [less than] .05, and imperative, F(1, 33)
= 8.58, p [less than] .01, to boys than to girls. These findings provide some support for
the hypothesis that parents display more polite and relationship enhancing form of
initiations with girls than with boys, and more instrumental and assertive initiation
strategies with boys than with girls.
Child Initiations
The MANOVA examining child initiations in relation to parent sex, child sex, and play
context (see Table I) revealed a significant main effect for initiation type, and play
session. Univariate analyses (see Table III) revealed that children gave more
imperatives than any other type of initiation, M F(1, 33) = 4.64, p [less than] .05, and
gave more play leads than requests for permission, F(1, 33) = 9.11, p [less than] .01,
requests for information, F(1, 33) = 5.21, p [less than] .05, and polite commands, F(1,
33) = 7.67, p [less than] .05. Children gave more initiations during the physical play
session than in the pretense play session, F(1, 33) = 6.18, p [less than] .05. No other
main effects or interactions were significant.
Child-Peer Play
Gender Differences in Play Form
A 2 x 2 (gender x play form) ANOVA was used to examine differences between boys
and girls engagement in pretense and physical play with peers (see Table IV). There
was a significant main effect for gender, F(1, 27) = 5.22, p [less than] .05, which was
qualified by a significant gender x play form type interaction, F(1, 27) = 8.14, p [less
than] .01. Boys engaged in more joint play with peers than did girls. However, as
hypothesized, girls engaged in more joint pretense play with peers than did boys, where
as boys engaged in more physical play with peers than did girls.
Gender Differences in Initiations to Peer
A 2 x 5 (gender x initiation type) ANOVA was used to examine differences between
boys and girls use of initiations with peers (see Table IV). There was a significant main
effect for gender, F(1, 27) = 6.7l, p [less than] .05, which was qualified by a significant
gender x initiation type interaction, F(1, 27) = 4.04, p [less than] .05. Boys gave
significantly more initiations to peers than girls. In particular, boys gave significantly
more imperatives to peers than did girls. Boys and girls did not differ on any other type
of initiation. These findings partially support the hypothesized differences between
boys and girls initiation strategies.
Associations Between Parent--Child Play Behavior and Child--Peer Play
Correlations were used to examine connections between parent--child play behavior
and children's peer play behavior. There was a significant positive correlation between
mother--child joint pretense play during the pretense play session and children's
pretense play with peer, r .3l, p [less than] .05, and between father-child joint pretense
play during the pretense play session and children's pretense play with peer, r = .28, p
[less than] .05. Similarly, there was a positive association between mother--child and
father--child joint physical play during the physical play session and children's physical
play with peer, r = .32, p [less than] .05 and r = .40, p [less than] .01, for mother--child
and father--child, respectively. There were no significant one-to-one correlations
between parent and child initiations and children's initiations to peer, in either the
pretense or physical play session.
DISCUSSION
Although interpretations should be made with caution due to the small, homogenous
sample, the results of the present study suggest that context, sex of parent, and sex of
child combine in a complex pattern to shape parent--child interaction. Differences in
parent--child interaction were observed across the pretense and physical play contexts.
Specifically, more joint parent--child play occurred in the physical play context than in
the pretense play context, and parents gave more initiations to children in the physical
play context than in the pretense play context. In contrast, children gave more
initiations to parents during the pretense play context than during the physical play
context. In addition, differences were observed between mothers' and fathers' behavior
with sons and daughters, and the nature of these gender differences varied depending
upon the play context in which parent--child interaction occurred. Consistent with
theoretical and empirical arguments for a contextual-interactive model of the
socialization of gender-typed behavior (Beall, 1993; Deaux & Major, 1987; Leaper et
al., 1998), the results of the present study suggest that the incidence of gender
differences in parent--child play is a complex interaction of context and sex of partner.
Consistent with hypotheses concerning parent--child joint play, both mother--daughter
and father--daughter dyads displayed more joint pretense play than mother--son and
father--son dyads. Moreover, mother--child dyads engaged in more joint pretense play
than father--child dyads. Also consistent with hypotheses, during the physical play
session, father--son dyads displayed more joint physical play than did father--daughter
dyads. Contrary to expectations, however, there was no difference in the amount of
physical play in mother--son and mother--daughter dyads. These findings are consistent
with previous studies that suggest that pretense play is more characteristic of parent--
daughter dyads, particularly mother--daughter dyads, than parent--son dyads (Langlois
& Downs, 1980; Lindsey et al., 1997a, 1997b), and that physical play is more
characteristic of father--son dyads than father--daughter dyads (Lindsey et al., 1997a,
1997b; MacDonald & Parke, 1984; Russell & Russell, 1987). The fact that mothers
demonstr ated no difference in the amount of physical play with boys and girls is also
consistent with previous evidence (MacDonald & Parke, 1984). These authors suggest
it may be that mothers are less influenced by gender-stereotypes regarding play forms
than are fathers. However, this argument is contradicted by the finding that there were
differences in the level of joint pretense play between mother--daughter and mother--
son dyads. Again, it may be that the interaction between gender and context determines
mothers' behavior pattern, in that mothers may consider pretense play to be more
appropriate for daughters than for sons, whereas physical play is equally appropriate
for sons and daughters. Additional research, which assesses parents gender stereotyped
beliefs in relation to gender socialization behavior is needed to elucidate this issue.
Support was also found for hypotheses concerning contextual and gender differentiated
patterns of initiations between parents and children. As predicted, parents gave more
polite initiation strategies, in the form of play leads and requests for permission, during
the pretense play session, and more assertive initiations, in the form of imperatives,
during the physical play session. Also as predicted, fathers made more initiations to
children than mothers. In particular, fathers gave more imperatives and polite
commands to children than mothers, whereas mothers gave more play leads to children
than fathers. Mothers' and fathers' initiations also differed on the basis of child gender,
with fathers giving more imperatives to boys than to girls, and more play leads and
polite commands to girls than to boys. Mothers, on the other hand, gave more play
leads, imperatives, and polite commands to boys than to girls. Although mothers and
fathers did not use the exact same types of initiations with boys and girls, the overall
pattern of initiations used by parents suggest that boys were exposed to more assertive
play behavior strategies than girls. In contrast, girls received more examples of
egalitarian and polite play initiations from fathers than did boys. These findings are
consistent with propositions that parents may contribute to children's gender-typed
behavior through a process of modeling instrumental behavior for boys, and facilitative
or cooperative behavior for girls (Block, 1983; Gilligan, 1982; Huston, 1983).
Although purely speculative, the pattern of gender-typed play behavior observed in this
study is also consistent with the proposal that parents may contribute to children's
formation of gender schema regarding particular play forms and play behaviors. That
is, to the extent that children use the different frequencies of father and mother
behavior as examples of what constitutes male-appropriate and female-appropriate
behavior (Perry & Bussey, 1989), parents may be teaching children to link certain
behaviors to male and female gender roles. Specifically, the fact that mothers engage in
more pretense play than fathers and that parents engage in more pretense with
daughters may lead children to link pretense play to the female gender role. In turn,
because physical play is most characteristic of father-son dyads, children may link
physical play to the masculine gender role. In addition, because fathers are more
directive with children than mothers, and both parents use more assertive strategies
with boys than w ith girls, children may develop a schema identifying assertiveness as
a component of the masculine gender role. Future research using multimethod designs,
including interviews to assess children's gender schema regarding pretense and
physical play, as well as assertive and cooperative behavior, may help elucidate
connections between parent-child play and children's formation of gender schema.
As hypothesized, the gender differentiated patterns of parent--child joint pretense and
physical play were mirrored in observations of children's play behavior with a peer.
Specifically, girls engaged in more pretense play with peers than did boys, whereas
boys engaged in more physical play with peers than did girls. These results are
consistent with previous studies of young children's peer interaction (DiPietro, 1981;
Howes et al., 1991). Such differences in children's preference for particular play forms
have been hypothesized to contribute to the pronounced gender segregation found in
children's social groups (Maccoby, 1990). The fact that these differences were also
observed in children's play with parents suggest that children's play form preferences
may be the result of socialization influences. That is, by encouraging girls to engage in
pretense play and encouraging boys to engage in physical play, parents may be
reinforcing gender-typed play preferences in their children. Alternatively, the findings
could represent a child effect in that girls prefer pretense play and boys prefer physical
play, regardless of partner. Evidence that hormonal imbalances account for differences
in boys' and girls' involvement in rough-and-tumble play (Hines & Kaufman, 1994)
support biological models of the emergence of sex differences in play form preference
(Hutt, 1972; Wilson, 1975). Longitudinal research that tracks the emergence of parent--
child pretense and physical play is needed to better understand the possible direction of
effect in parents' and children's preference for particular play forms.
The fact that boys gave more initiations, particularly imperatives, to peers than did girls
is consistent with previous evidence that boys' peer interactions are characterized by
more controlling and directive behavior than girls' peer interactions (Leaper, 1991;
Serbin, Sprafkin, Elman, & Doyle, 1994). Given that parents used more assertive
initiation strategies with boys than with girls, as discussed above, it seems reasonable
to conclude that boys may learn to be more assertive through the process of modeling.
However, no associations were observed between parents' use of initiations with
children and children's use of initiations with peers. This suggests that some process
other than modeling maybe at work. It may be that, because they are used to receiving
highly assertive initiations from parents, boys become accustomed to assertive
strategies and their behavior is more difficult to influence by partners using
nonassertive strategies. This argument is consistent with the proposal that gender-
segregated play patterns among children are the result of different behavioral styles
(LaFreniere, Strayer, & Gauthier, 1984; Maccoby, 1990), and with evidence that the
behavior of boys is more easily influenced by other boys than by girls (Fagot, 1985).
Future research should compare the initiation strategies parents use with children to the
initiation strategies given to and received by children when playing with both sameand
opposite-sex peers.
It is important to note that the parent--child play contexts in this study were contrived,
laboratory settings, and thus the parent--child behavior observed may not represent
patterns of parent--child interaction in more natural settings. Nevertheless, the results
are intriguing when considered against the backdrop of previous, more naturalistic
investigations. For example, evidence suggests that boys and girls have different types
of play environments in the home (Rheingold & Cook, 1975), and that parents and
children engage in both pretense and physical play at home (Haight & Miller, 1992;
Russell & Russell, 1987). Moreover, mothers report a preference for engaging in
pretense play with children, whereas fathers report a preference for engaging in
physical play with children (Haight, Parke, & Black, 1997). Together this evidence
suggests that parent--child dyads do engage in pretense and physical play in more
natural settings, providing support for the assumption that the gender-typed interactions
patterns observed in this study may extend to parent--child interaction in the home.
However, more naturalistic investigations of parent--child physical and pretense play
are needed to confirm the sex differences in parent-child play observed in this study.
In conclusion, the present study joins a growing body of evidence supporting a
contextual-interactive model of the socialization of sex-typed behavior (Beall, 1993;
Deaux & Major, 1987; Leaper et al., 1998). The observed contextual differences in
parent-child behavior suggest that boys and girls may be exposed to different
socialization settings that contribute to gender-segregated patterns of peer interaction.
Specifically, the findings suggest that boys may be socialized to prefer physical forms
of play and assertive influence strategies, whereas girls may be socialized to prefer
pretense play and cooperative forms of interaction. A major implication of this study is
that research that fails to consider the context of parent--child interaction may overlook
an important component in the determinants of gender differentiated patterns of
behavior between parents and children.
ACKNOWLEDGMENTS
This investigation was supported by grants to the second author from the National
Institute of Mental Health (MH49869) and the Alabama Agricultural Experiment
Station (10-004). The authors thank Mandy Parks, Malinda Colwell, Audra Sundberg,
Ashley Main, and Laynee McCeever for their help in various phases of data collection
and coding. We are grateful to the children, parents, and teachers of the Auburn
University Child Study Center for their time and participation.
(1.) To whom correspondence should be addressed at Department of Human
Development and Family Studies, P.O. Box 41162, Texas Tech University, Lubbock,
Texas 79409-1162.
(2.) Piolet testing revealed that many parent-child dyads were exhausted by their
activity during the physical play session. Thus, in order to avoid a fatigue effect with
some parents the order of the pretense and physical play sessions was maintained
across all dyads.
(3.) Examination of the toys parent--child dyads played with in each play session
revealed no differences between mothers and fathers, and no sex-of-child differences.
The majority of joint pretense play during the pretense session occurred with the
kitchen toys (M = 0.24 and 0.21, for mother--child and father-child dyads,
respectively), followed closely by the animal figures (M = 0.18 and 0.20, for mother-
child and father-child dyads, respectively). The majority of joint physical play during
the physical play session occurred with the Nerf bat and ball set (M = 0.38 and 0.42, for
mother-child and father-child dyads, respectively), followed closely by the large Slo-
Mo ball (M = 0.32 and 0.37, for mother--child and father--child dyads, respectively).
(4.) Examination of the play form data revealed that only a small portion of parent-
child joint play was spent engaged in multiple play forms. In the pretense play session,
parent-child dyads spent approximately 9% and 12% (for mother-child and father-child
dyads, respectively) of joint play engaged in both pretense and another play form. In
the physical play session, approximately 5% and 8% (for mother-child and father-child
dyads, respectively) of parent-child joint play was spent engaged in physical play and
another play form, with the majority of joint play composed of pretense and physical
play. Given the fact that rate of pretense-other play was significantly associated with
total pretense play in the pretense play session (r = .42, p [less than] .01, and r = .35, p
[less than] .05, for mother-child and father-child dyads, respectively) and physical-
other play was significantly associated with total physical play in the physical play
session (r = .32, p [less than] .05, and r = .38, p [less than] .01, f or mother-child and
father-child dyads, respectively), instances of joint pretense-other play were included in
the joint pretense play totals and instances of joint physical-other play were included in
the joint physical play totals.
REFERENCES
Beall, A. E. (1993). A social constructionist view of gender. In A. E. Beall & R. J.
Sternberg (Eds.), The psychology of gender (pp. 127-147). New York: Guilford Press.
Block, J. H. (1983). Differential premises arising from differential socialization of the
sexes. Child Development, 54, 1335-1354.
Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and
differentiation. Psychological Review, 106, 676-713.
Caldera, Y. M., Huston, A. C., & O'Brien, M. (1989). Social interaction and play
patterns of parents and toddlers with feminine, masculine, and neutral toys. Child
Development, 60, 70-76.
Cherry, L., & Lewis, M. (1976). Mothers and two-year-olds: A study of sex-
differentiated aspects of verbal interaction. Developmental Psychology, 12, 278-282.
Crockenberg, S., & Litman, C. (1990). Autonomy as competence in 2-year-olds:
Maternal Correlates of child defiance, compliance, and self-assertion. Developmental
Psychology, 26, 961-971.
Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of
gender-related behavior. Psychological Review, 94, 369-389.
DiPietro, J. A. (1981). Rough and tumble play: A function of gender. Developmental
Psychology, 17, 50-58.
Dunn, J., & Brown, J. (1994). Affect expression in the family, children's understanding
of emotions, and their interactions with others, Merrill-Palmer Quarterly, 40, 120-137.
Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states between
mothers and their young children. Developmental Psychology, 23, 132-139.
Entwisle, D. R., & Astone, N. M. (1994). Some practical guidelines for measuring
youth's race/ethnicity and socioeconomic status. Child Development, 65, 1521-1540.
Fagot, B. I. (1997). Parenting boys and girls. In M. H. Bornstein (Ed.), Handbook of
parenting: Vol. 1. Children and parenting. Mahwah, NJ: Lawrence Erlbaum Associates.
Fagot, B. I. (1985). Beyond the reinforcement principle: Another step toward
understanding sex roles. Developmental Psychology, 21, 1097-1104.
Frankel, M. T., & Rollins, H. A., Jr. (1983). Does mother know best? Mothers and
fathers interaction with preschool sons and daughters. Developmental Psychology, 19,
694-702.
Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard University Press.