S9-DISC4-2 - The Affect Effect: Integrating Student Emotions into the Design of Engineering Technology Courses with Optimization Method

1. Innovative Practice Full Paper
HAIFENG WANG1 , Laura Cruz1, Makayla Shank1
1 Penn state university

Abstract—This Innovate to Practice Full Paper presents a study that enhanced both cognitive and non-cognitive learning outcomes through optimizing science, technology, engineering and math (STEM) students’ affective responses. Much of engineering education has focused on designing courses and curriculum to maximize both cognitive and, increasingly, non-cognitive learning outcomes.  These latter, including traits such as persistence, curiosity, and hope, have been identified as particularly salient to the retention and success of under-represented populations in STEM (Sultana, Kahn, & Abbas, 2017; Scheidt et al, 2017). The role of affective outcomes, such as feelings or values, while often linked to non-cognitive traits, have been comparatively under-studied in the engineering context (Alias et al, 2014; Kort et al, 2001). This despite the fact that there is promising research in both educational psychology and computer science that links positive affect with enhanced learning outcomes (Baker et al, 2010; McCann et al, 2020; Shen, Wang, & Shen, 2009).  

 Part of the challenge in studying affect is that the constructs are complex, changeable, and highly varied (D’Mello & Graesser, 2012; Scherer, 2009).  To design a study based on affect, an engineering professor partnered with an education professor, and an advanced undergraduate psychology student to develop a model for capturing and applying affective outcomes in applied engineering courses. To measure affect, we collected weekly surveys of the students’ affective responses to both the mode of delivery and nature of the content using categories such as boredom, surprise, and confidence, each of which have been identified as potentially significant by other researchers.  The surveys are based on a modified version of the CAP perceived learning scale, which consists of nine (9) validated measures of cognitive, affective, and psychomotor perceptions of learning (D’Mello & Graesser, 2012; Scherer, 2009). 

To strengthen the replicability of our results, these surveys were administered to students in both upper and lower division courses (offered as part of a four-year Engineering Technology curriculum) over the course of an academic year.  We then integrated these responses into a predictive linear recursion model, which was, in turn, used to make curricular decisions periodically throughout the semester.  In other words, the students’ affective responses were used to influence the content and the delivery of the course as it was being taught.  

Our findings suggest that using this predictive model to optimize affective responses also served to significantly enhance both cognitive and non-cognitive learning outcomes at all levels of instruction.  The study has implications for the further study of the role of affective outcomes in engineering education; as well as the advancement of co-created (with students) models of instructional and curricular design that incorporate affective variables.

Keywords—Affect, Linear Recursion Model, STEM, Optimization