Development of a cognition-priming model describing learning in a STEM classroom

Successful STEM learning depends on the interaction of affect, cognition, and application of ideas. Simply put students who are unwilling to persist in STEM based endeavors do not suddenly develop into scientists, mathematicians, engineers or computer scientists, nor do they seek out STEM related courses or STEM based careers. The purpose of this study is to investigate content, cognitive, and affective outcomes related to STEM integrated curriculum within the K-5 arena. Educational and psychological literature tends to focus one aspect of the other when examining the role of affect and cognition on student outcomes. Current trends in educational measurement and psychometrics have begun to address the artificial disconnect that exists between affect, cognition, and content outcomes within the science education literature. The methods used to develop the results within this study are a mixture of quantitative methods to develop a model of learning occurring in a STEM school. Using ANOVA, structural equation modeling, and model analysis, an understanding of the problems presented becomes clear. Analysis of model fit statistics suggests adequate model fit (χ2(21) = 30.91, p = 0.075, CFI = 0.94, TLI = 0.93, RMSEA = 0.04, SRMR = 0.05). The standardized structural coefficients for the path from group to each of the constructs is statistically significant (p < 0.05) thus indicating that the two groups differ on the constructs of self-efficacy, science interest, spatial visualization, and mental rotation. An estimate of effect size of the mean group difference across the statistically significant constructs reveals self-efficacy (d = 1.27, large), science interest (d = 1.97, large), spatial visualization (d = 1.30, large), and mental rotation (d = 1.42, large). There is considerable evidence that the inclusion, STEM integrated learning at the earlier elementary level becomes critically important for the students as they progress in school. © 2015 Wiley Periodicals, Inc. J Res Sci Teach 52: 410–437, 2015