UC Irvine

Individual differences significantly impact learning outcomes, yet research has been limited in predicting training outcomes based on these differences and in examining how they interact with intervention types. Understanding these dynamics is crucial for developing effective training programs. This dissertation addressed these gaps through three studies.

Study 1 examined how individual differences predict variations in learning trajectories during working memory training in undergraduates. Participants were clustered into three training patterns, and machine learning models predicted these patterns based on baseline cognitive abilities, personal characteristics, environmental and experience-related factors. Openness and pre-existing working memory capacity were key predictors of high performance, while openness and video game background predicted learning persistence among lower performers. This study demonstrates the potential to predict individual training performance using participant characteristics before training.

Study 2 explored how baseline abilities interacted with three interventions: working memory-only, numerical knowledge-only, and combined working memory + numerical knowledge training in kindergarteners. Overall, the study found that combined training led to the most pronounced long-term gains in math skills. Moreover, participants with high initial math skills showed the greatest improvements in both working memory and math with combined training, whereas numerical knowledge-only training was more effective for children with lower initial math skills. This suggests that combined training is generally efficient, but baseline abilities also matter.

Study 3 built on Studies 1 and 2 by integrating executive function training within a math invention for children. Using early learning data to predict training improvement, the study found that performance in the first three sessions could predict overall training improvement and identified two learning patterns: a group with higher initial scores and greater improvement, and a group with lower initial scores and slower improvement. Within the lower improvement group, some participants still showed significant progress, catching up to the high improvement group, likely due to completing more training sessions.

Overall, these studies highlight the importance of considering individual differences and the type of cognitive intervention in designing optimally tailored programs. They also illustrate the potential of using pre-training or early individual differences in learning trajectories to predict training outcomes, which has implications for personalized educational strategies.