Gender Differences on Preference for Science : Reanalysis of a Nationwide Junior High School Student Survey

リカ　センコウ　ノ　ジェンダー　サ　ハ　ドノヨウニ　ケイセイ　サレルカ　ゼンコク　チュウガクセイ　チョウサ　ノ　サイブンセキ　ニ　ヨル　ケントウ

タナベ, カズヒコ

Tanabe, Kazuhiko

オオサカ ダイガク ダイガクイン ニンゲン カガク ケンキュウカ キョウイクガクケイ

Department of Education, Graduate School of Human Sciences, Osaka University

Although there are no wide-ranging gender differences in science and mathematics performance between Japanese junior high school students, there are clear differences in aspirations and attitudes toward these subjects. However, it has not been empirically confirmed as to why and how Japanese boys’ and girls’ preferences for science and mathematics differ. Through a reanalysis of a Japanese nationwide junior high school student survey on science education conducted in 1999, this study examines the effects of students’ family environment and the gender bias prevalent in school, on the gender differentiation of science preference among junior high school students. A binary logistic regression analysis using science preference as a dependent variable confirmed that the family socio-economic status, the scientific experience at home, the division of roles in science experiments, and expectations from teachers regarding science were related to junior high school students’ science preference. Mediation analysis by the KHB（ Karlson, Holm, Breen） method also confirmed that school factors partly mediate the gender effect on science preference. On the other hand, familial factors did not significantly mediate the gender effect on science preference. Therefore, the importance of focusing on the gender bias prevalent in school was emphasized. Because the Average Marginal Effect of women did not improve dramatically after considering all factors in the analysis, further research is desired to identify another factor that well accounts for the gender differentiation of science preference.