Using random forest analysis to identify student demographic and high school-level factors that predict college engineering major choice

Background: Given the importance of engineers to a nation's economy and potential innovation, it is imperative to encourage more students to consider engineering as a college major. Previous studies have identified a broad range of high school experiences and demographic factors associated with engineering major choice; however, these factors have rarely been ranked or ordered by relative importance. Purpose/Hypothesis: This study leveraged comprehensive, longitudinal data to identify which high school-level factors, including high school characteristics and student high school experiences as well as student demographic characteristics and background, rank as most important in terms of predictive power of engineering major choice. Design/Method: Using data from a nationally representative survey, the High School Longitudinal Study of 2009, and the random forest method, a genre of machine learning, the most important high school-level factors in terms of predictive power of engineering major choice were ranked. Results: Random forest results indicate that student gender is the most important variable predicting engineering major choice, followed by high school math achievement and student beliefs and interests in math and science during high school. Conclusions: Gender differences in engineering major choice suggest wider ranging cultural phenomena that need further investigation and systemic interventions. Research findings also highlight two other areas for potential interventions to promote engineering major choice: high school math achievement and beliefs and interests in math and science. Focusing interventions in these areas may lead to an increase in the number of students pursuing engineering.