A quasi-experimental study of a peer-assisted strategy-based reading intervention in elementary school

Abstract

Promising methods of reading instruction for elementary school students incorporate peer-assisted learning routines and reading strategies. In addition, models of reading comprehension point to the importance of various determinants of reading competence such as reading fluency and vocabulary knowledge. Multicomponent reading intervention programs need to be evaluated to determine IF and HOW they unfold their effects on the reading competence of elementary school students on the basis of such theoretical and empirical models. Accordingly, the present study was designed as a quasi-experimental study of a 20-lesson peer-assisted and strategy-based multicomponent intervention for whole-class instruction in elementary school. Linear mixed models and latent growth models were used to analyze the longitudinal data (pre-, post- and follow-up test) on the reading competencies (reading fluency, vocabulary knowledge, reading strategy competence, reading comprehension) and intrinsic reading motivation of students in the intervention (N = 187) and control group (N = 177). The results showed an interaction between the groups and the change in reading comprehension, indicating a significantly increased score in the intervention group at the posttest (d = 0.15) but not at the follow-up test (d = 0.12). The results of the latent growth model point to the importance of designing interventions that explicitly integrate reading strategies, reading fluency and vocabulary knowledge and also foster intrinsic reading motivation. In addition, reading fluency was revealed to be the strongest predictor of reading comprehension and the change in fluency over time was closely linked to reading comprehension development.