A Meta-Analytical Review of the Effectiveness of Artificial Intelligence on Writing Instruction for K–12 Students

• In Event: Educational Technology/Computational Thinking

Abstract

The meta-analysis extends previous reviews about educational technology’s impact on students’ writing performance by focusing on the application of educational technology with artificial intelligence (ETAI) in writing instruction. Thirty-one studies from 27 publications incorporating 6236 students were included; 82 effect size estimates were pooled using robust variance estimation to account for statistical dependencies. The results indicated a moderate effect size of ETAC program and showed that genre of writing and students’ grade level did not significantly moderate the effect size. However, publication year, students’ language background , IA programs with/without human tutors, post-test samples, setting of experiment as well as mode of writing significantly affect the effect sizes of the writing outcomes. Implications for both practice and research are provided.

Authors

• Xueyan Hu, China Three Gorges University

• Zhihong Xu, Texas A&M University - College Station

• Gilbert Ramirez, Texas A&M University - College Station

• L. Quentin Dixon

• Jing He, Old Dominion University