The paper aims to evaluate and compare across a large range of countries the impact of gender diversity on the overall job satisfaction of lower-secondary education teachers. It also seeks to examine whether the effects of gender similarity are asymmetrical for men and women. The empirical evidence is based on the estimation of multilevel models that control for individual characteristics, work-related factors, and school-based variables. The results may be suggestive for policy makers and educational planners who are initiating interventions designed to promote diversity within the education system and to remasculinize the teaching profession 10 Except when the data are missing completely at random, which is rarely the case. 11 Namely, replacing missing data by sample means of complete observations. 12 That is, replacing missing values on an explanatory variable with a zero (or, more generally, with any constant) and including an indicator of missingness as an additional predictor in regression models. 13 According to Schafer (1997), three to five imputations yield excellent results 14 CABA, Italy, Singapore, and Spain are excluded from the pooled sample of countries because data for some variables are not available for all respondents or are not applicable. teachers in school j, is a dummy variable accounting for 1 if teacher i in school j is female and zero otherwise, is an interaction term between teacher gender and the proportion of female teachers in school j, is a vector of personal characteristics, denotes a vector of school characteristics, and and are error terms at the school and teacher level, respectively. Significance levels: *** p<0.01; ** p<0.05; * p<0.10; + p<0.20. Source: Authors' estimates based on the TALIS 2018 international database.
