The impact of school and classroom characteristics on educational choices by boys and girls: A multilevel analysis

Does enrolment in comprehensive schools affect girls' and boys' choices of educational field at the upper secondary school level! For example, do girls in some comprehensive schools choose the engineering programme (gender-atypical for girls) to a significantly higher degree than girls do in other schools! These questions were studied in the context of a 25 per cent random sample of Swedish comprehensive schools in 1991 and 1992. Multilevel analyses were carried out on pupils who chose one of the academic programmes. Three hierarchical levels are included in the models: pupils, classrooms and schools. The results show that pupils' choices of study field vary significantly between schools and classrooms when pupil-level characteristics have been controlled for, The results also show that both boys' and girls' choices correlate with those of classmates of the same sex, but not with those of the opposite sex. However, the contextual variables in this study that characterize schools and classrooms (such as the percentage of boys/girls in classrooms, etc.) do not account for the school and classroom variances, although some school/classroom characteristics are significant. The results give especially strong support for the so-called 'frog pond effect': for example, if classmates do well in mathematics and natural sciences subjects, the likelihood that they will enter the humanities/social sciences increases (controlling for the pupils' own marks in these subjects). When classmates do poorly in these subjects, their chances of choosing engineering increase.