Explaining the Heterogeneous Effects of Enjoyment on Academic Achievement: An Interpretable Double Machine Learning Approach (Poster 41)

• In Event: AERA Poster Session Saturday 8:00 a.m.
  In Poster Session: Learning and Motivation in STEM

Abstract

While previous research has established a positive association between enjoyment and academic achievement in science, the complexity of this relationship across individual characteristics and contexts remains underexplored. Utilizing data from 9,764 students in B-S-J-G (China) from PISA 2015, this study adopted interpretable double machine learning approach to reveal significant heterogeneity in the enjoyment-achievement relationship. The results indicated the comprehensive moderators, such as student engagement, teaching environment, and ICT resources and use. The study proceeded with person-level analysis, identifying multiple groups of students with disparate enjoyment effects through interpretation of multi-feature interactions. Overall, the study reveals a more complex picture that extends beyond the basic tenets of early theories, and paves the way for more targeted and context-aware educational interventions.

Authors

• Yingke Xu, Beijing Normal University

• Zhentao Lin, Beijing Normal University

• Wei Tian, Beijing Normal University

• Panpan Sun, Shangqiu Normal University