META-ANALYSIS OF COMPUTER-BASED GRAMMAR INSTRUCTION AND STUDENTS ACHIEVEMENT

The present meta-analysis is a comprehensive investigation of the magnitude of effectiveness of computer assisted language learning (CALL) on learners grammar achievement. To this aim, a total sample of 26 studies published during the two recent decades was analyzed. The overall mean effect size was found to be +0.78, indicating that CALL programs had a moderate, positive, significant effect on students vocabulary achievement. The distribution of effect sizes in this meta-analysis was highly heterogeneous. To further explore the sources of variation between effect sizes across studies, moderator analyses for methodological and substantive features (research design, sample size, proficiency level, age, intensity of program, year of publication) were run. Findings further suggested that heterogeneity between effect size estimates in this meta-analysis was influenced by the difference in learners proficiency level and the publication year of the studies; whereas other variables did not contribute to the production of heterogeneity among effect sizes.

Keywords:

CALL , grammar achievement , meta-analysis , effect size , heterogeneity