Gender differences in children’s social skills growth trajectories

Abstract

At school entry, girls are rated by teachers as more competent on measures of social skills than boys. It is less clear if this higher rating is stable or grows over time. To address this question, multiple group curve of factors models investigated gender-specific growth trajectories across seven waves of measurement in a large, longitudinal sample (N = 1024, NICHD SECCYD). Results showed that girls’ social skills were consistently rated higher from kindergarten to sixth grade, and the effect size was moderate (latent Cohen’s d = .37 to .62). Boys demonstrated greater heterogeneity in social skills at nearly every grade with the gender difference in variability stable after second grade. An examination of gender differences in growth trajectories showed that boys demonstrated a linear decrease over time, whereas girls’ social skills did not significantly change over time after accounting for initial level of social skills in kindergarten.

Acknowledgements

This study was conducted by the NICHD Early Child Care Research Network supported by NICHD through a cooperative agreement that calls for scientific collaboration between the grantees and the NICHD staff. The authors thank the NICHD for granting permission to use the data. The authors disclose no potential conflicts of interest in the design or conduct of this study. The authors wish to thank Drs. Timothy Keith and Todd Little for helpful feedback regarding the methodology used in the current study as well as Reese Mason for feedback on early drafts.

Notes

1 A recommended correction to the RMSEA was applied in the multiple group latent variable models to account for the number of groups (RMSEA multiplied by the square root of the number of groups; Steiger, 1998).

2 The mean of the quadratic components is not statistically significant, and the quadratic factors may have overfit the data by absorbing random variability (Preacher et al., 2008). While methodologists have differing opinions about which model to accept in this situation, we erred on the side of caution and accepted the linear model (Muthen, 2012; Preacher et al., 2008).

3 Among girls and boys the sixth grade Self-Control subscale was not statistically significantly correlated with any measurements from kindergarten through fourth grade (at a p < .05 level). For girls specifically kindergarten and third, kindergarten and fourth, second and third, and fifth and sixth Self-Control measurements were also not statistically significantly correlated. For boys specifically kindergarten and fifth grade Self-Control measurements were also not statistically significantly correlated.