This thesis comprises four empirical studies dealing with different aspects of educational choice in Sweden. Two studies focus on the hierarchical outcome, namely level of education. The other two are concerned with the horizontal outcome field of study. Large-scale quantitative data from surveys and registers were used to examine the influence of gender, family of origin, and other social contexts on individuals' educational choice.>P> In paper I, Childhood Conditions and Educational Careers two general questions are addressed: 1) To what extent can differences in childhood conditions account for the fact that children from various social class origins choose academic upper secondary school programmes to such differing degrees? 2) Which changes in educational attainment during this century have resulted from the effects of class background and other childhood conditions? Regarding the first question, the parents' level of education, as an indicator of the teaching and cultural environment in the home, proved to be the most important childhood condition in accounting for class differences. Financial difficulties during childhood were less important in comparison. Regarding the second question the results showed that the influence of the parents' social class and educational level, and the number of siblings (as a partial indicator of financial difficulties) were weaker for younger cohorts (1930-1949 and 1950-1973) than for older ones (1892-1929). However, the effects of several childhood conditions examined had not decreased.
Starting in the late sixties Sweden experienced an extensive geographical dispersion of university and college education. Whether these new establishments have had any reducing effect on the social selection into higher education is evaluated in paper II, The Establishment of New Institutes of Higher Education: A Means of Reducing Educational Inequality? The results clearly give a negative answer. The interpretation is that 1) people from different social backgrounds do not differ in their sensitivity to geographical distance to university sites, and 2) the new colleges and universities have not been established in regions with a high proportion of disadvantaged social classes (working classes for example), nor have they managed, on the whole, to increase the number of students from the newly established university and college regions in relation to the rest of the country.
Can family background account for intra-gender differences in the heavily gender-typed choice of study field? Paper III, Parental Role Models, Gender, and Educational Choice, describes several relationships on this topic from a socialization perspective where parents are regarded as role models for their children. It was found that the parents' educational as well as occupational sector increased the likelihood of both boys and girls choosing a similar field of study (vocational and academic programmes in upper secondary school), and that both the mother's and the father's occupation and education were important in this respect. This so- called same-sector effect was somewhat stronger for fathers and sons, while no such same-sex influence was confirmed for girls. Further, a "high" social origin ( measured by class and level of education - was positively associated with gender-atypical choices.
In paper IV, A Multilevel Approach to Gender-Typical and Gender-Atypical Educational Choices: Do Schools and Classrooms Matter? the questions asked in the third paper were extended to include any influences from social contexts above the level of the family, more specifically, classrooms and schools. Logistic multilevel models were used to show significant school and classroom variations in the choice of study field. However, the contextual characteristics that were used in an attempt to account for these (mainly aggregate variables based on pupil characteristics) left the variances unchanged to the greatest extent. For example, the girl/boy ratio in the classroom does not seem to influence gender-typical or gender-atypical educational choices. One classroom effect showed a systematic influence however: the better classmates performed in mathematics and natural science subjects, the lower the propensity to choose the engineering programme, and further, the better classmates performed in language and social sciences subjects, the less probability that a girl or a boy would choose the humanities/social sciences programme.
Stockholm: The Swedish Institute for Social Research, Stockholm University , 1998. , 22 p.