Persisting in STEM into university: The role of emotions in physics and chemistry

Purpose. The present study examined junior college students’ anxiety and enjoyment as predictors of their intrinsic motivation and intention to enroll in science, technology, engineering, and mathematics (STEM) programs in university. Promoting academic achievement and persistence in STEM has been a focal point of recent government directives (e.g., Let’s Talk Science, 2015) and motivation research (e.g., Perez et al., 2014), yet empirical research on academic emotions as potential predictors of continued persistence in STEM is lacking, especially among junior college students.

Theoretical Framework. Building on expectancy-value approaches to academic behaviour (e.g., Eccles et al., 1983), the control-value theory of achievement emotions (Pekrun, 2006) provides a framework for examining the antecedents and outcomes of students’ emotions in educational contexts. Students’ cognitive appraisals of control (e.g., self-efficacy) and value (e.g., utility value) are theorized to predict their discrete emotions that in turn predict their motivation to learn and persistence. The current study tested this predictive model with an underexplored population taking courses in multiple STEM domains.

Method & Measures. 778 students (18- to 20-years-old, 48.2% female) enrolled in a two-year science program at one of four junior colleges in Montreal, Canada, completed paper and pencil questionnaires in their first (T1), second (T2), and final (T3) semesters. Questionnaires assessed science self-efficacy (T1; Pintrich et al., 1991), utility value of science courses (T1; Halloun & Hestenes, 1998), learning-related anxiety and enjoyment for physics and chemistry (T2; Pekrun et al., 2002), intrinsic motivation (T2; Vallerand et al., 1992), and persistence (T3; What is the likelihood you will continue to study in a science/engineering program at a university within the next two years?).

Results & Scholarly Significance. Structural equation modelling was conducted to examine the hypothesized relations among the study variables. Prior achievement and gender were controlled as covariates for both domains. The hypothesized model fit the data well (CFI = .964, .964; TLI = .955, .954; RMSEA = .036, .035, for physics and chemistry, respectively). Direct effects of self-efficacy (β = .19, .28, p < .001) and utility value (.31, .28) positively predicted students’ intrinsic motivation. Higher levels of self-efficacy negatively predicted students’ anxiety (-.39, -.31). Self-efficacy (.27, .10) and utility value (.22, .23) were also found to positively predict students’ domain-specific enjoyment. Moreover, higher levels of enjoyment were found to positively predict intrinsic motivation (.44, .42), which in turn positively predicted students’ intention to persist in STEM in university (.25, .22). Finally, intrinsic motivation fully mediated the effect of enjoyment on persistence (physics: Sobel z = 4.63, p < .001; chemistry: Sobel z = 5.60, p < .001). The results are in line with control-value theory (Pekrun, 2006) and suggest that fostering junior college students’ learning-related enjoyment (via cognitive appraisals of control and value) in chemistry and physics may have positive effects on their persistence in STEM as they pursue more advanced degrees prior to entering the workforce. Moreover, our model suggests that fostering learning-related enjoyment, in both domains, may generate greater persistence relative to efforts to reduce learning-related anxiety. Copyright © 2016 AERA.