Reading development in a tracked school system: A longitudinal study over 3 years using propensity score matching

Background. Assigning students to different school tracks on the basis of their achievement levels is a widely used strategy that aims at giving students the best possible learning opportunity. There is, however, a growing body of literature that questions such positive effects of tracking. Aims. This study compared the developmental trajectories of reading comprehension and decoding speed between students at academic track schools that typically prepare students for university entrance and students at non-academic track schools that usually prepare students for vocational education. Sample. In a longitudinal design with three occasions of data collection, the authors drew on a sample of N= 1,508 5th graders (age at T1 about 11 years, age at T3 about 14 years) from 60 schools in Germany. The academic track sample comprised n= 568 students; the non-academic track sample comprised n= 940 students. Method. Achievement measures were obtained by standardized tests of reading comprehension and decoding speed. Students at the different tracks were closely matched using propensity scores. To compare students’ growth trajectories between the different school tracks, we applied multi-group latent growth curve models. Results. Comparable results were recorded for the complete (unmatched) sample and for the matched pairs. In all cases, students at the different tracks displayed a similar growth in reading comprehension, whereas larger growth rates for students at academic track schools were recorded for decoding speed. Conclusions. Our findings contribute to an increasing body of literature suggesting that tracking might have undesired side effects.