ICILS 2018 컴퓨터·정보 소양 예측 변수 탐색

Abstract

ICILS(International Computer Information Literacy Study) assesses secondary school students’ability to use computers and manage information. Among ICILS variables, CIL(computer and information literacy) is considered essential to those living in information society. Compared to previous research that modeled a small number of variables, this study explored hundreds of ICILS 2018 variables to predict CIL via machine learning. In particular, glmmLasso, a penalized regression method, was employed to consider the multilevel structure of ICILS in variable selection. As a result, 12 out of 398 variables were found to be important variables to predict CIL, which included computational thinking(CT), gender, ICT self-efficacy regarding the use of general applications, positive perceptions of ICT for society, home literacy index, principals’ use of ICT for general school-related activities. To our knowledge, this is one of the first empirical studies to apply glmmLasso in social science. Future research directions are discussed.