Document Type
Article
Publication Date
5-31-2024
Keywords
SEMtree; Longitudinal Study; KCYPS 2018; Covariate
Abstract
In this study, we investigated the applicability of the SEMtree model to analyze the longitudinal data analysis in educational studies. SEMtree is one of the advanced approaches to incorporate the advantages of structural equation modeling and decision tree models under confirmatory and exploratory data analysis frameworks. The main purpose of this study is to figure out how the researcher deals with the data conditions of covariates when using the SEMtree model. By analyzing KCYPS 2018 the 4th-grade elementary school students' panel, we found that there were differences of the duplicated selected covariates when splitting the nodes in the SEMtree model between the time-invariant and the time-varying types. The binary covariates showed less duplicated selection of covariates when splitting the nodes, while the Likert-type covariates showed more number of terminal nodes. Thus, the researcher should consider how to deal with covariate types when using the SEMtree model in educational data analysis.
Citation
Hong Minju, Lee Juyeon. A comparisons of the covariate types in applications of SEMtree model to educational studies. Journal of Curriculum and Evaluation 2024; 27(2):279-298. https://doi.org/10.29221/jce.2024.27.2.279
Comments
The article is in Korean and is distributed under a CC-by-NC-SA 4.0 license.