Beyond instructional practices: Characterizing learning environments that support students in explaining chemical phenomena

Abstract

AbstractMany conversations surrounding improvement of large‐enrollment college science, technology, engineering & mathematics (STEM) courses focus primarily (or solely) on changing instructional practices. By reducing dynamic, complex learning environments to collections of teaching methods, we neglect other meaningful parts of a course ecosystem (e.g., curriculum, assessments). Here, we advocate extending STEM education reform conversations beyond “active versus passive learning.” We argue communities of researchers and instructors would be better served if what we teach and assess was discussed alongside how we teach. To enable nuanced conversations about the characteristics of learning environments that support students in explaining phenomena, we defined a model of college STEM learning environments which attends to the intellectual work emphasized and rewarded on exams (i.e., assessment emphasis), what is taught in whole‐class meetings (i.e., instructional emphasis), and how those meetings are enacted (i.e., instructional practices). We subsequently characterized three distinct chemistry courses and qualitatively examined the characteristics of chemistry learning environments that effectively supported students in explaining why a beaker of water warms as a white solid dissolves. Furthermore, we quantitatively investigated the extent to which measures of incoming preparation explained variance in students’ explanations relative to enrollment in each learning environment. Our findings demonstrate that learning environments that effectively supported learners in explaining dissolution emphasized how and why salts dissolve in‐class and on assessments. Changing teaching methods in an otherwise traditionally structured course (i.e., a course organized by topics that primarily assesses math and recall) did not appear to impact the sophistication of students’ explanations. Additionally, we observed that learning environment enrollment explained substantially more of the variance observed in students’ explanations than measures of precollege math preparation. This finding suggests that emphasizing and rewarding the construction of causal accounts for phenomena in‐class and on assessments may support more equitable achievement.