The Effects of Computer-Assisted Instruction on Mathematics Achievement in Mainland China: A Meta-Analysis. (2020)

• Record Type: Journal Article
• Authors: Xie, Chen; Cheung, Alan C.K.; Lau, Wilfred W.F.; Slavin, Robert E.
• Is Part Of: International journal of educational research. Volume 102(2020)
• Publication Date: 2020
• Subjects: Educational technology applications -- Mathematics achievement -- Mainland China -- Meta-analysis; Education -- Evaluation -- Periodicals; Curriculum evaluation -- Periodicals; Education -- Periodicals; Program Evaluation -- Periodicals; Éducation -- Évaluation -- Périodiques; Programmes d'études -- Evaluation -- Périodiques; Éducation -- Recherche -- Périodiques; Curriculum evaluation; Education -- Evaluation; Periodicals; 370.72
• DOI: 10.1016/j.ijer.2020.101565
• Languages: English
• ISSNs: 0883-0355

Abstract

Highlights: Computer-assisted instruction (CAI) produced a moderate positive effect in comparison to traditional instruction in Chinese mainland. Programs fell into four categories: Inquiry-CAI, Cooperative-CAI, Transmission-CAI, and Supplemental-CAI. Among various types of CAI, Cooperative-CAI and Inquiry-CAI had the largest effects. Effect sizes were higher than those from meta-analyses using similar methods in the U.S. and other Western nations. Abstract: This meta-analysis examined the effect of computer-assisted instruction (CAI) on mathematics achievement in mainland China. Using rigorous inclusion criteria, we identified 36 qualifying studies, covering 13, 438 students in grades 1-12. They produced a moderate overall effect (ES=+0.38) in comparison to traditional instruction. Of the various types of CAI, Cooperative-CAI and Inquiry-CAI had the largest effects. This finding seemed to support the logic behind the recent curriculum reform in mainland China, namely that CAI produced better results when combined with new teaching models. Even Supplemental-CAI and Transmission-CAI were more efficient than traditional transmission teaching. The impact of other substantive and methodological study features on the variation of effect sizes by moderator analysis are also discussed.