ADOLESCENT GENDER IDEOLOGY SOCIALIZATION: DIRECT AND MODERATING EFFECTS OF FATHERS' BELIEFS

Abstract

Previous research examining the intergenerational transmission of gender ideology focuses generally on the influence of mothers' beliefs. This article extends the understanding of gender ideology construction and transmission in two important ways. Utilizing data from the child sample of the National Longitudinal Survey of Youth 1979 (N = 206), we examine the construction of adolescent gender ideology via mothers' and fathers' gender beliefs. Further, we consider the interaction between maternal and paternal ideologies as they influence adolescent ideology. Findings suggest that paternal ideology plays a strong role in adolescent ideology formation, both directly and as a moderator of maternal influence.

The data collection for the National Longitudinal Surveys of Labor Market Experience of Youth and the corresponding Child and Young Adult surveys was sponsored by the Bureau of Labor Statistics. The data were collected by the Center for Human Resource Research at The Ohio State University. The analyses and interpretations herein are those of the authors. This research was supported by a 2008 Faculty Research and Development Award from George Mason University. The authors thank Catherine Zimmer for her statistical assistance. An earlier version of this manuscript was presented at the 2007 Annual Meeting of the American Sociological Association.

Notes

¹The literature upon which this research is based uses a variety of phrases to describe individuals' level of support for a division of paid work and family responsibilities that is based on the notion of separate spheres: “gender ideology,” “gender role attitudes,” “attitudes about gender,” “gender-related attitudes,” and “gender egalitarianism.” We have chosen primarily to use the phrase “gender ideology,” although the terms are interchangeable. While “gender ideology” as a concept may evoke notions of beliefs about gender relations broadly within a culture (e.g., see Kane 2000), our use of “gender ideology” follows Davis and Greenstein's (2009) lexicon in studying individuals' level of support for a gendered division of work and family responsibilities.

²Initially our sample included 216 adolescents. We were concerned that any interaction effect could be unduly influenced by outliers, especially among fathers. As such, we removed all adolescents whose fathers' gender ideology fell outside two standard deviations from the sample mean. After removing those ten adolescents, our final sample size became 206. Although the substantive results of the analyses described herein are not substantively different with the two samples, we have chosen the more conservative path of publishing results using only the 206 adolescents whose fathers' ideologies fall within two standard deviations of the mean.

^a All measures were recoded such that a larger value represented a more egalitarian attitude. Parameters are standardized indicator loadings on the ideology constructs. Model fit: RMSEA = .026. All parameters are statistically significant at p < .05.

³In this and subsequent stages, we model maternal and paternal gender ideology as predicted by race/ethnicity, religious affiliation, religious service attendance, age, hours of paid employment, and educational attainment. These results are not presented here so as not to distract from the paper's main focus: the construction of adolescent gender ideology. However, these results are available upon request.

⁴Following Klein and Moosbrugger (2000) and Mplus documentation, the coefficients presented here are standardized parameter estimates. The indicators for parental and child gender ideology were transformed into standardized scores (with a mean of zero and a standard deviation of one) for ease of interpretation and comparison across models. All models take into consideration statistically significant correlations among predictors as were determined through preliminary stages of the analysis. That is to say, any significant correlations between predictors of child, maternal, and paternal ideology, though not shown in the tables, have been included in the analysis. Models 3 and 4 also include the statistically significant correlation between maternal and paternal gender ideology (r = .18, p < .05).

^a Analyses include robust standard errors to control for clustering by family. All models include controls for year of adolescent interview as well as for correlations among all predictors. Adolescent gender ideology is a latent variable as shown by the measurement model in Table 1. The scale of the latent variable is such that a higher score indicates a more egalitarian gender ideology. N = 206.

^b Reference category is white.

^c Reference category is Protestant.

^d Not able to be calculated due to estimation technique.

*p < .05, two-tailed tests.