< Back to previous page

Project

On track? A study on the effects of tracks in secondary education

Tracking is the grouping of students with similar abilities in the same classroom or school during secondary education. For each group, a tailored educational program is provided, which is called a ‘track’. Many education systems use this practice, based on the assumption that fitting educational environments to more homogeneous student groups will result in more efficient education. Tracking also allows for the development of specialized skills in the different tracks. This skill specialization helps the education system to meet the needs of the labor market. However, it remains unclear whether and to what extent a track affects a student’s outcomes. Therefore, in this dissertation we assessed the effects of going to school in a certain track on several outcomes.

In Flanders, students must choose a track at the end of primary education, when they are 12 years of age. Tracks attract students based on their academic abilities (e.g., academic performance in mathematics and reading comprehension). Therefore, each track differs in the average academic performance of its students at the start of secondary education. Based on these differences, the tracks can be hierarchically ranked. Because each track attracts students with different abilities and socioeconomic backgrounds, differences between student outcomes across tracks are partially attributable to these initial differences. Hence, to estimate the effects of tracks, the methodological challenge was to account for differences between students prior to initial track allocation.

Track choice in Flanders is generally free at the start of secondary education. However, a student who does not meet the level of his or her track will be forced to go to a ‘lower’ track. That many students change from a higher to a lower track over time is considered a unique characteristic of Flemish education. Students who change track also differ from students who remain in their track. Hence, a second methodological challenge was to account for differences between students who change track and students who remain in their track.

We used two longitudinal datasets of secondary education, the contemporary LiSO-dataset and the older LOSO-dataset. To account for the differences between students, both matching methods and G-methods were used. Matching was done by using propensity scores, but we also used Mahalanobis distances and coarsened variables. G-methods either used propensity scores or directly used the estimation of track allocation probabilities. Note that these methods only work if comparable students exist across tracks. Therefore, the effect of a track was estimated for three pairs of tracks, with each pair consisting of hierarchically consecutive tracks. The effects of tracks were assessed for academic performance, unemployment, academic self-concept and engagement.

The results showed that being allocated to a higher track always positively affects academic performance. Being allocated to the higher track also reduced the probabilities of becoming unemployed, although the differences were generally small. For academic self-concept it was usually best to be allocated to a lower track, but this was not the case for the classical and modern track comparison. In which track a student goes to school mattered little for engagement.

Date:1 Sep 2014 →  8 Mar 2019
Keywords:onderwijseffectiviteit, secundair onderwijs
Disciplines:Education curriculum, Education systems, General pedagogical and educational sciences, Specialist studies in education, Other pedagogical and educational sciences
Project type:PhD project