Introduction

The goal of the current study is to examine differences in communication styles between mothers and fathers engaged in conversation with a 9-year-old son or daughter. Both working class and middle class families were included in the study. This research adds to current literature by examining the communication styles of both mothers and fathers in conversation with girls and boys, by studying children in middle childhood, by including working class as well as middle class families, and by describing children’s communication styles as well as those of parents.

The study is based on ecocultural theory (Bronfenbrenner and Morris 1998), which proposes that everyday, frequently repeated interactions between parents and children, referred to as proximal processes, contain many layers of messages that are conveyed by parents to children. In addition to the explicit content that is transmitted when parents talk to children, the structure and style of parents’ speech convey implicit messages to children, particularly regarding gender and social expectations. These implicit messages, embedded into conversational styles, are the focus of the current research.

It is widely accepted that adult men and women use consistently different styles of communication (Tannen 1990). For example, men tend to be direct and forceful in their speech and to interrupt others, whereas women express support for others’ ideas and are collaborative in their speech. These same characteristics have been described in parents’ speech to their children (Leaper et al. 1998; Malone and Guy 1982). Thus, the conversations between parents and children can be seen as a subtle but pervasive socialization mechanism by which gender roles and their associated status differentials are conveyed to children. Prior research has also shown evidence that parents’ speech to children varies by social class. Middle class parents not only talk to their children more than do working class parents (Hart and Risley 1995; Hoff et al. 2002) but also are more likely to encourage independent thinking as opposed to obedience and acceptance of parental authority (Hoff et al. 2002; Kohn 1977). The present study examines the conversations of working class and middle class mothers and fathers with their boys and girls in order to determine the extent to which parents’ communication styles convey status differences linked to gender and social class.

One approach to operationalizing gender differences in communication styles, adopted in the present study, is the characterization of men’s communication style as assertive and women’s communication style as affiliative (Leaper 1991). An assertive communication style directs attention to the self, is controlling, and attempts to influence the partner’s ideas or actions; this style is considered to reflect the higher status and influence of men in our society. Assertive speech also involves interrupting the conversational partner in order to focus the conversation on one’s own ideas (Anderson and Leaper 1998; Leman et al. 2005) An affiliative communication style concentrates on the other person in the conversation, is encouraging, and attempts to elicit ideas and involvement from the partner; this style reflects the relatively lower social position and lack of power of women (West and Zimmerman 1977). Based on past research, we expected higher levels of assertive speech and interruptions for the fathers in this study and higher levels of affiliative speech for the mothers.

Children’s acquisition of gender-based communication styles has not been thoroughly examined. To date, most of the research examining child conversations with parents has focused on preschool children and their mothers (Leaper and Smith 2004; Stafford 2004). Little research has been conducted on the acquisition or understanding of gendered communication styles throughout childhood. In one study, story characters’ dialogue was written to exemplify male and female communication styles; children in fifth and sixth grades did not link these styles to gender whereas those in seventh and eighth grades did (Hibbard and Buhrmester 1998). These results suggest that it may not be until early adolescence that sex-differentiated speech is explicitly recognized. Leman et al. (2005), however, found that 8-year-old boys used more assertive speech and girls more affiliative speech in peer interactions, suggesting that gender-typed patterns of conversation are present prior to adolescence. In the present study, we examine gendered communication styles during middle childhood to gain further insight into the age at which children begin to use adult-like gender-based communication styles. We expected to see more assertive speech among boys and more affiliative speech among girls.

The extent to which the gender of the partner in a conversational dyad affects gender-differentiated communication styles is not clear. Theoretically, communication styles reflect both one’s own social status and that of the conversational partner (Berger et al. 1977). According to this theory, when interacting with someone of high status, an individual will be more deferential. Thus, both males and females interacting with males should be more affiliative in their conversational style. Studies of children’s interactions with peers suggest a different pattern, however. In one study, both boys and girls observed in same-age peer interactions used more affiliative speech with same-gender peers than in cross-gender interactions (Leman et al. 2005), and in another, both boys and girls used a more affiliative style when the conversational partner was female (Leaper et al. 1999). The most consistent findings of studies examining sex of child differences in parent speech are that mothers talk more with daughters than with sons and show more differentiation in their speech with daughters and sons than fathers do (Lanvers 2004). Because most prior studies have examined parental communication styles with infants and very young children, little is known about child gender effects on parental speech with older children. Based on theory regarding social dominance, we predicted that both mothers and fathers would use a more affiliative style in conversations with sons than daughters, and both boys and girls would use a more affiliative style in conversations with fathers than mothers. We also anticipated that mothers would talk more than fathers, especially in interactions with daughters.

Research on class differences in communication styles is very limited (Laursen and Collins 2004); however, there has been substantial research demonstrating class-based behavioral differences in parent–child interactions that suggest potential communication style differences. Working-class parents believe it is important to teach their children to take orders whereas middle-class parents believe it is important to teach their children to be independent thinkers (Hoff et al. 2002; Kohn 1977). Kohn suggests that these differences reflect the occupational experiences of the parents in that middle-class individuals in professional occupations are expected to make their own decisions, to think creatively, and to be self-directed, but working-class employees are told what to do and how and when to do it. It is not surprising, then, that working-class parents tend to use an authoritarian parenting style whereas middle-class parents use authoritative, or democratic, parenting (Hoff et al. 2002). In a study not specifically focused on class differences, Leaper et al. (1995) found that more authoritarian parents tended to use assertive communication styles whereas more authoritative parents used affiliative communication styles. Thus, we hypothesized that parents from working class families would use more assertive speech than parents from middle class families, who would use more affiliative speech. Further, because earlier research has indicated that middle class parents talk more with their children than do working class parents, we predicted longer conversations for middle class dyads.

Hypotheses

The focus of the present study is gender and social class differences in the conversations of mothers and fathers with sons and daughters. The specific hypotheses tested are: (1) Overall, mothers and daughters will use more affiliative speech than fathers and sons, and fathers and sons will use more assertive speech and interrupt their conversational partner more often than mothers and daughters; (2) mothers and fathers will use more affiliative speech in conversations with sons and more assertive speech in conversations with daughters; (3) sons and daughters will use more affiliative speech in conversations with fathers and more assertive speech in conversations with mothers; (4) mothers will talk more overall and this difference will be intensified with daughters; (5) working-class parents will use more assertive speech than middle-class parents, who will use more affiliative speech; and (6) middle class parents will talk more with their children than working class parents. No specific hypotheses were made regarding social class effects on children’s speech due to the lack of prior literature examining social class and children’s communication styles.

Method

Participants

The 40 parent–child dyads included in the analyses for this study were a subset of the participants from a Midwestern US study site of the National Institute of Child Health and Human Development Study of Early Child Care and Youth Development. Criteria for inclusion in this sub-study were: (a) two-parent family, (b) social class clearly identified as working class or middle class based on Hollingshead (1975) scores (≤40 = working class; ≥48 = middle class), and (c) child observed and videotaped in an interaction with at least one of the parents during the third grade home visit. A total of 64 families from the study site met these criteria. Videotapes were selected at random in order to place five parent–child dyads into each of eight cells (parent gender [2] × child gender [2] × social class [2]). Thus, the final sample included ten mother–son, ten mother–daughter, ten father–son, and ten father–daughter dyads, with half the families in each dyad type identified as working class and half middle class. Only one observation per family was included; that is, if a child was observed with both mother and father, only one of the sessions could be selected. All families were European American. At the time the interactions were collected, the children were 9 years old (M = 9.05 years, SD = 3.02 months)

Procedure

As part of a home visit scheduled during the child’s third-grade year, parents and children were videotaped in a Rules Discussion Task that included rules for children, rules for parents, and difficult situations (see Table 1 for a list of the questions). Discussions lasted 3:30 to 16:00 min; the median time was 6 min.

Table 1 Parent–child rules discussion questions.
Full size table

Coding of Parent–Child Conversations

For the larger study, the Rules Discussion Task was coded for qualitative dimensions of the parent-child relationship (complete information about the procedures and coding system used can be found at http://secc.rti.org). For the purposes of the sub-study reported here, the conversations of the 40 selected parent–child dyads were transcribed verbatim. Utterances were defined by content and natural pauses in the conversation, and interruptions of one speaker by the other were marked on the transcripts. Using both the transcript and the video record, each parent and child utterance and obvious nonverbal communicative behavior (e.g., shrugging shoulders, shaking head) was coded using an adaptation of the Psychosocial Processes Coding Scheme (Leaper 1991) that focused on assertive and affiliative speech. Assertive speech included utterances that were direct and clearly stated the speaker’s ideas or opinions: statements about how things are, opinions, disagreements with the partner or rejection of the other’s ideas, or telling the other person what to do. Affiliative speech included utterances that were indirect or tentative, invited the partner to express ideas or opinions, or supported the ideas of the partner. Utterances that did not fit either of these categories (repetitions, asking the partner to repeat something, and conversational filler [e.g., “well...”, “um”]) were classified as “other” and were not analyzed (these made up 22% of the parents’ utterances and 26% of the children’s utterances).

A second observer independently coded 25% of the interaction sessions across all dyad types. Inter-rater agreement percentages, calculated by dividing the total number of agreements by the number of agreements plus disagreements, for assertive and affiliative speech of parents and children were all above 87%.

Analyses

Initially, the total number of utterances for each member of each dyad and the number of interruptions recorded for each participant were counted. To control for differences in the overall amount of speech across dyads, data on affiliative and assertive speech categories were calculated as proportions of the total number of utterances. Prior to conducting analyses, proportion data were transformed using an arcsine transformation in order to meet the normality assumption for parametric statistical testing. All data were then analyzed using parent gender (2) × child gender (2) × social class (2) analyses of variance (ANOVAs; the data on total utterances and number of interruptions were also analyzed using a Poisson regression approach that is appropriate for count data which may not be normally distributed; the results of these analyses were identical to the ANOVA results. Because the ANOVA results are more familiar and more readily interpretable, they are reported here).

Results

Gender Differences in Parents’ Conversational Styles

The proportion of parent utterances coded as assertive and affiliative are shown in Table 2 (data were transformed for analyses but the proportion scores are shown in the tables for ease of interpretation). Overall, mothers used approximately the same proportion of affiliative and assertive speech in their conversations (M = .38, SD = .09 for affiliative; M = .39, SD = .07 for assertive), whereas fathers used a significantly higher proportion of assertive than affiliative speech (M = .30, SD = .12 for affiliative; M = .48, SD = .11 for assertive; t(19) = 3.69, p = .002). The results partially supported Hypothesis 1, in that mothers used a higher proportion of affiliative speech than fathers, F(1,32) = 7.30, p = .011, and fathers a higher proportion of assertive speech than mothers, F(1,32) = 10.14, p = .003. The effect size for the difference in affiliative speech, calculated using partial eta2, was .19 and for assertive speech, .24, indicating that parent gender accounted for almost 20% of the variance in parent affiliative speech and almost one-quarter of the variance in parent assertive speech. Results of analyses of the number of interruptions did not support the prediction in Hypothesis 1 that fathers and sons would interrupt their conversational partners more often than mothers and daughters. The number of times parents and children interrupted each other during their conversations did not vary by parent gender or child gender. Parents interrupted children an average of 14.5 times (SD = 10.3) and children interrupted parents an average of 15.1 times (SD = 12.4).

Table 2 Mean proportions (SD) of affiliative and assertive speech for parents.
Full size table
Table 3 Mean proportions (SD) of affiliative and assertive speech for children.
Full size table

The test of Hypothesis 2, that mothers and fathers will use more affiliative speech with sons and more assertive speech with daughters, indicated support for this prediction. Both parents used a higher proportion of affiliative utterances when talking with sons (M = .38, SD = .09) than daughters (M = .31, SD = .11), F (1,32) = 6.12, p = .019, partial eta2 = .16, and both mothers and fathers showed more frequent use of assertive speech in conversations with daughters (M = .47, SD = .11) than sons (M = .43, SD = .09), F(1,32) = 6.54, p = .016. The partial eta2 for this effect was .17.

Gender Differences in Children’s Conversational Styles

The prediction in Hypothesis 1 that child conversational styles would also differ by gender was not supported (see Table 3). Unlike their parents, children’s use of assertive and affiliative utterances did not vary by their gender. Across all dyads, 51.0% of children’s utterances were assertive (SD = .08) and 23.3% were affiliative (SD = .08). Similarly, Hypothesis 3, that children’s use of assertive and affiliative speech would differ by the gender of the parent, was not supported in that there was no main effect for parent gender and no significant interaction between child gender and parent gender. Furthermore, children’s assertive and affiliative speech did not mirror the speech of their parents; correlations between the proportions of speech categories for parents and children were not significant, assertive r(40) = -.23, p = .15; affiliative r(40) = .11, p = .52.

Gender Differences in Amount of Speech

Based on prior research, we anticipated (Hypothesis 4) that mothers would talk more than fathers, especially in conversations with daughters. This prediction was not supported. Analysis of the total number of parent utterances indicated a significant difference by child gender, F(1,32) = 7.98, p = .008, but no parent gender differences and no interaction between parent gender and child gender. The effect size was .20, indicating that 20% of the variance in amount of speech was accounted for by child gender. In contrast to our prediction that daughters would elicit more speech than sons, both mothers and fathers talked more with sons (M = 118.5 utterances, SD = 42.2) than with daughters (M = 87.0, SD = 23.1). Total child utterances did not differ by child or parent gender (boys M = 90.0, SD = 48.1; girls M = 75.9, SD = 16.8). Not surprisingly, there was a strong correlation between the amount of speech for parents and children, r(40) = .86, p < .0001; when parents talked more, so did their children.

Social Class Differences

Hypothesis 5 addressed the question of social class differences in parent speech, predicting higher proportions of affiliative speech among middle class parents and higher proportions of assertive speech among working class parents. No specific predictions were made for social class differences in child speech. The hypothesis for parent speech was only partially supported; middle-class parents used significantly more affiliative speech than working-class parents, F(1,32) = 6.15, p = .019, partial eta2 = .16, but there was only a marginally significant trend for working-class parents to use more assertive speech than middle-class parents, F(1,32) = 3.40, p = .075 (Table 4). The partial eta2 for the effect of social class on assertive speech was .10. Finally, Hypothesis 6, regarding social class differences in amount of speech, was not supported. There were no significant social class main effects or interactions for total utterances, nor for child affiliative or assertive speech or interruptions by either parents or children.

Table 4 Mean proportions (SD) of assertive and affiliative speech for parents and children by social class.
Full size table

Discussion

In this study, we found gender differences in parent speech but not in child speech. The results supported our major hypotheses regarding parental communication styles: Fathers used more assertive speech than mothers, and mothers were more affiliative than fathers. In addition, for parents, as predicted, child gender appeared to serve as a stimulus for the frequency of use of each of the communication style categories. Both parents used more assertive speech with girls and more affiliative speech with boys. The findings for parents’ communication style differences based on child gender support the theoretical proposition that the status of the conversational partner matters (Berger et al. 1977). Because males in our society, including male children, have higher status than females, this theory predicts that both males and females will tend to use a more affiliative communication style when interacting with males. This idea was supported for parents in the present study as both mothers and fathers used more affiliative speech with sons than daughters. No parent gender differences were found for interruptions, often considered to be a characteristic of assertive speech styles. In a meta-analysis of the literature on interruptions, Anderson and Leaper (1998) found substantial gender differences only when interruptions were defined as intrusive. In the present study, we did not differentiate intrusive from other types of interruptions (defined by Anderson and Leaper as “affiliative overlaps,” “unsuccessful interruptions,” and “minimal responses” (p. 232) and therefore our coding may not have captured the assertive aspect of intrusive interruptions.

Contrary to our hypotheses, both mothers and fathers talked more with sons than daughters. These results do not support earlier findings that mothers talk more with daughters than sons and that fathers talk more with sons than daughters (for a review, see Lanvers 2004; Leaper et al. 1998) nor that mothers talk to children more than do fathers (Leaper et al. 1998). Much of the prior research that examines mother-father differences has been conducted with very young children. The present results suggest developmental changes in parents’ communication with children that deserve further investigation.

Although it was expected that children’s assertive and affiliative speech would also vary by child gender, there were no significant differences in communication style between boys and girls. Earlier studies reporting gender differences in child speech have included only preschool-age children (Leaper and Smith 2004; Stafford 2004). The lack of gender differences found in the present study may be due to the fact that children in middle childhood have moved away from early stereotypes of gender roles but have not yet experienced gender intensification (Alfieri et al. 1996; Hibbard and Buhrmester 1998).

The children in this study did not show differentiation in their speech styles based on parent gender. Children in general used an assertive rather than affiliative communication style in their conversations with parents. Perhaps children at this age are predominantly self-focused and not yet skilled at the kind of other-focused communication that characterizes affiliative speech.

Social class effects were also found for parent speech, and they fit prior descriptions of class differences in parental values (Kohn 1977). The middle-class parents in this study talked in a more affiliative way, encouraging children to express their ideas and responding positively to children’s opinions. This interactive style is believed to foster independent thought whereas a more assertive style, which was marginally more often used by working class parents in this study, is considered to promote the value of obedience to parental authority. Some prior research (Hart and Risley 1995; Hoff et al. 2002) has reported that middle-class parents in general talk more to their children than working-class parents, but we did not find any differences by social class in parents’ amount of speech. It may be that the situation presented to the parents, in which they were asked to discuss a specific set of topics, may have minimized differences in the amount of talking parents did.

No significant differences in child speech based on social class were found. Aspects of social class that emerge in adult speech patterns may not be salient to children in middle childhood. If, as Kohn (1977) suggests, social class exerts an influence on behavior through employment, then it is logical to find that social class differences are reflected in parents’ speech but not in that of children. The school settings where children spend a lot of time may not differ by social class in the ways occupational settings do. Overall, children did not appear to adopt communication styles that were imitative of their parents as might be suggested from a social learning perspective. There were no significant correlations between parent communication style categories and those of children. Further research to identify the developmental pattern of acquisition of gender- and social class-based communication styles throughout middle childhood and adolescence would aid in our understanding of the processes of social stereotype development and the enactment of social roles.

Several limitations of the current study must be noted. The total number of participants was low, with only 40 dyads represented. A larger sample would have increased our power to detect subtle gender and class differences. The effect sizes for the differences found are substantial, however, and suggest the results are robust. We did not have enough diversity among the participants at this study site of the larger project to include racial or ethnic background of the families as an additional factor. Examining variations related to ethnicity and cultural background, in conjunction with parent and child gender and social class, would add another important dimension to the research on parent-child communication styles. Our focus on only two aspects of communication, assertion and affiliation, helps to provide some consistency within the literature in that these categories have been used extensively in prior research (Leaper and Smith 2004), but they are not exhaustive. Other aspects of parents’ and children’s communication styles might reveal interesting variation by gender and/or social class. In addition, we did not examine parental beliefs and socialization goals; thus, we cannot know what factors underlie the differences we observed in parent communication styles.

The results of the present study support the conclusions from previous research in this area that emphasize the importance of everyday interactions between parents and children in conveying information about social status, particularly with regard to gender. Further research with larger, more diverse samples and examining variation in parents’ and children’s speech over time would give a more complete picture of the nature of gendered communication in families.