Exploring the factors that influence academic stress among elementary school students using a LASSO penalty regression model

Abstract

The main objective of the study was to explore the main predictors that influence academic stress among fourth-grade elementary school students using data from the Panel Study on Korean Children via a least absolute shrinkage and selection operator (LASSO) regularized penalty regression model. The study examined 280 explanatory variables using the LASSO model. After preprocessing the data, it finally selected 21 variables. Out of them, the study found that children’s persistence, peer attachment, bullying, parental achievement pressure, and subjective socioeconomic status as significant predictors, which is consistent with previous studies. However, several variables, including children’s overall happiness, frequency of using slang, preference for mathematics, school life preference, daily time spent for homework and study, average monthly cost of private education, parents’ participation in school events, marital conflict, and residential area, which were newly explored in the study, were also significant. Finally, we presented the significance and implications of the results in relation to decreasing academic stress among fourth-grade elementary school children.