The effect of learner-centered methods in foreign language teaching on academic achievement and attitudes towards lesson: A meta-analysis study

This paper aims to analyze the results of the individual studies that examine the effect of learner-centered methods in foreign language teaching on academic achievement and attitudes towards lesson via meta-analysis. The literature review based on inclusion and exclusion criteria yielded a total 234 studies (260 effect sizes) on academic achievement variable and 24 studies (25 effect sizes) on attitudes towards lesson. Random effect model was used in data analysis. The results demostrated that learner-centered methods had a large effect both on academic achievement (g=0,951, 95% CI [0,852-1,049], p<.0.000) and on attitudes towards lesson (g=0,925, 95% CI [0,642-1,209], p<.0.000). Forest plot, Funnel plot, Rosenthal Classic fail-safe N, Orwin fail-safe N, and Duval and Tweedie trim and fill methods were run to check whether the effect sizes calculated for academic achievement and attitudes towards lesson showed any publication bias or not. In this way, the effect sizes distribution were visually interpreted and statistically tested. The results indicated no publication bias. After general effect size analysis, moderator variables- also called study characteristics- were analyzed. Publication year, the study site, publication type, method, type of language skill, level of education, English being the official language/or not at the study site (this moderator was just for academic achievement), length of experiment and size of the participants were defined as moderator for both academic achievement and attitudes towards lesson in the present study. Publication year, type of publication, level of education, English being the official language/or not at the study site, length of the experiment and size of the participants were determined to be effective moderators in significantly differentiating effect sizes. As for attitudes towards lesson, publication year, type of language skills and level of education turned out to be effective in moderating effect sizes. In other words, a significant difference was found between the groups in above mentioned moderators. One of the striking and remarkable results of the present study is the fact that the length experiment turned out to revealed opposite results for academic achievement and attitudes towards lesson. When the length of experiment is short, it increases academic achievement more; on the other hand, the longer the length of experiment, the more effective it is in increasing positive attitudes towards lesson. Key words: Learner-centered approaches, achievement, attitude, cooperative learning, drama, project-based learning, meta-analysis