Hogging the middle lane: How student performance heterogeneity leads Turkish schools to fail in PISA?

Abstract

This study aims to analyse the relationship between students’ mathematics achievement in Programme for International Student Assessment (PISA) 2012 and the instructional climate-related factors in the index of principals’ perceptions (learning hindrance, teacher morale and teacher intention). As preliminary analysis procedure, the chi-squared automatic interaction detection analysis was performed with relevant independent variables. Teacher’s achievement expectation from students and achievement-oriented behaviours were other significant predictive indicators on PISA mathematics achievement. Based upon these independent variables and standard deviation estimates of PISA mathematics scores, the present research developed a theoretical model by means of confirmatory factor analysis, explaining how students’ PISA mathematics achievement is associated with classroom and within school homogeneity through teachers’ expectation and achievementoriented behaviours. Results showed that the developed model provided a great model-data fit. This model revealed that classroom achievement homogeneity and within school achievement homogeneity were the most important predictors on students’ PISA mathematics achievement.

Keywords: PISA, CHAID, mathematics, homogeneity.