Different types of input frequency distribution: Their effects on construction learning by elementary English learners in Korea

Following a usage-based approach, this study examined whether different types of input frequency distribution (skewed first, skewed random, and balanced) would have differential effects on construction learning. Four classes of fifth-year elementary school students (n = 65) were randomly assigned to one of three treatment groups or one control group. Three tests were utilized to measure the learning effects and generalizability of the learned knowledge (a video clip selection task, an argument selection task, and a sentence selection task), immediately after the treatment and again one week later. Additionally, two aptitude measures were administered to assess the students’ language analytic ability and working memory capacity. The results showed that only the group exposed to a skewed first distribution retained the learned knowledge until the delayed posttest session. In terms of the aptitude measures, only working memory performance weakly correlated with the posttest comprehension scores. These findings suggest that English teachers could enhance the learning of grammatical rules by providing students with appropriate input that optimizes some important features such as frequency, input distribution, and students’ aptitude.