Applying LPA to classify students based on learning outcomes and self-efficacy in STEM

Abstract

In the context of the increasingly vital STEM fields, guiding students’ careers and implementing workforce training streams in these areas are gaining attention and being implemented early in secondary schools. The main purpose of this article is to provide guidance on Latent Profile Analysis (LPA), a quantitative method to determine the number of latent student profiles based on STEM learning outcomes and self-efficacy in STEM among high school students in Ho Chi Minh City. The analysis results on a sample of 1,074 high school students indicate the potential division into three latent student groups (two majority and one minority group) with significant differentiation in STEM learning outcomes and self-efficacy. Additionally, combined with logistic regression analysis, a T-test for independent samples, and Cohen’s d index evaluation, the study also highlights the influence of gender and frequency of STEM experiences on different profiles, as well as the impact of these profiles on the career orientation of high school students.