The Development of Children’s Sensitivity to how Boys and Girls Talk

• In Event: Poster Session 04
  In Poster Session: PS 04 Topic - Language, Communication

Abstract

Past studies have shown that adult listeners can classify gender from prepubertal children’s speech (e.g., Perry et al., 2001) – long before anatomical differences between the sexes emerge. This suggests that gender differences in speaking style are learned early in life through socialization. Despite clear evidence that adults are sensitive to gendered speech patterns in children, no study to date has asked whether children also perceive gender cues in child speech. Here, we conduct the first study to investigate how well children classify gender in child speech. Specifically, we ask 1) when sensitivity to gendered speech patterns in age-matched peers emerge in childhood, and 2) whether performance is moderated by children’s and parents’ gender attitudes. We hypothesize that children may vary in how sensitive they are to cues to gender in child speech, and that socialization may play a key role in explaining this variation.

Thus far, we have tested 19 children between the ages of 3 and 6. This age range is chosen to match the age of the child speakers in the stimuli as closely as possible. Our final sample will include a total of 20 children from each age group, and 20 adults as a comparison group by January (final N = 100). Data collection will easily be completed by January because we are collecting data online. In the experiment, children were presented with 14 isolated words produced by the same 12 English-learning children (6 boys and 6 girls) at each of three ages: 2.5, 4, and 5.5 years old, Thus, each child listener heard tokens for all 12 children, each at three different ages over the course of 84 trials. In each trial, children determined the gender of the speaker by clicking on a picture of a boy or a girl. At the end of the experiment, parents were asked to complete a gender socialization scale (Blakemore & Hill, 2008) and a gender identity questionnaire (Johnson et al., 2004) to measure parents’ attitudes about gendered behaviours and their assessment of children’s gender identity, respectively.

We fit a generalized mixed-effects model to our data. We included speaker’s age as a fixed effect, random intercepts for word, and a random slope for speaker’s age by participant. Preliminary results (see Figure 1) show that, overall, children are above chance at classifying gender (β = 0.17, SE = 0.06, z = 2.81, p = 0.005). Upon completion of data collection, we will further examine whether child listeners are better at identifying gender in older children, or they are more accurate with children of their own age (i.e., own-age advantage). Moreover, to investigate whether accuracy is moderated by gender attitudes of themselves and their parents, we will enter children’s scores from these questionnaires as fixed effects in the above model. If results continue to pattern this way, then this will be the first study to show that children can detect gender differences in child speech. It will also shed light on how children’s socialization contributes to this development.

Authors

• Priscilla Fung, University of Toronto
  Presenting Author

• Jessamyn Schertz, University of Toronto
  Non-Presenting Author

• Elizabeth K. Johnson, University of Toronto
  Non-Presenting Author