Mapping and explaining the gender gap in students’ second language proficiency across skills, countries and languages

This study discusses the generalizability of gender differences in the second language competences of European adolescents across three languages, three skills and fourteen countries. In most cases, females do better than males but the effect sizes are small or medium at best (Cohen's ds < −0.46). However, English appears rather gender-neutral on average, with males sometimes outperforming females. We also found evidence for cross-skill variation: writing turns out to be more prone to gender differences than listening or reading. Thirdly, we found cross-country variation in the gender gap, which supports the hypothesis that gender is primarily a social factor rather than a biological factor in learning. Multilevel mediation analyses with constituents of motivation indeed show that in most cases where females outperform males, up to 60% of this advantage can be explained by the differential appeal of the students' L2 course and by the instrumental, integrative and intrinsic value that students attribute to the L2.