9th Conference European Sociological Association

RN10 Sociology of Education

2009-09-05 13:30:00 2009-09-05 15:00:00 Saturday, 5 September 13:30 - 15:00 Special Purposes / Teachers I Building I, 2E8

Female or Male Teacher: A Determinante of Gender-specific Differences in Primary School Achievement

In the meantime, gender gaps in school performance are an empirically confirmed and therefore widely accepted fact. Particularly regarding the reading competences, girls outperform boys in virtually all countries participating in large-scale international studies like PIRLS at the primary and PISA at the secondary educational level. Investigating the causes of these gender-specific differences comprehensively requires besides individual determinants the inclusion of contextual influences, containing amongst others teacher characteristics. Therefore our exploratory focus highlights the effect of teacher gender on the reading performance of girls and boys.

The existing literature suggets that a match between student and teacher gender has a positive impact on the educational outcome, which can be explained by role-model, unintended behavioral and active preference effects. Vice versa, a gender-mismatch can lead to a phenomenon called sterotype threat which can retard the academic identification of students and subsequently their performance (Dee, 2005) as well as unintended and intended discrimination by the teacher. However, it has to be mentioned that the results of the available empirical literature are decidedly mixed.

Methodologically, most of the studies apply OLS regression to model a student-teacher interaction effect while including a number of control variables from different levels of aggregation (student, class, school). By using the Austrian 2006 PIRLS dataset, we follow a multilevel modeling approach to account for the hierarchical structure of the data (students nested within classes nested within schools) and further integrate the relevant variables of interest at the appropriate level to avoid statistical and conceptional problems (Hox, 2002: 3). Consequently, we create a cross-level interaction effect between student and teacher gender by building a random intercept random slope model with student gender at the individual and teacher gender at the class level as well as a number of control variables, selected by theoretical aspects.

To conclude, our aim is to shed light on the effect of teacher gender on the reading competences of students at the end of primary schooling by (1) trying to provide a coherent model based on a solid theoretical foundation and (2) using the appropriate methodological tools for analysing hierarchical data.