Optimizing testing feedback in introductory chemistry: a multi-treatment study exploring varying levels of assessment feedback and subsequent performance

Providing students with feedback on their performance is a critical part of enhancing student learning in chemistry and is often integrated into homework assignments, quizzes, and exams. However, not all feedback is created equal, and the type of feedback the student receives can dramatically alter the utility of the feedback to reinforce correct processes and assist in correcting incorrect processes. This work seeks to establish a ranking of how eleven different types of testing feedback affected student retention or growth in performance on multiple-choice general chemistry questions. These feedback methods ranged from simple noncorrective feedback to more complex and engaging elaborative feedback. A test-retest model was used with a one-week gap between the initial test and following test in general chemistry I. Data collection took place at multiple institutions over multiple years. Data analysis used four distinct grading schemes to estimate student performance. These grading schemes included dichotomous scoring, two polytomous scoring techniques, and the use of item response theory to estimate students’ true score. Data were modeled using hierarchical linear modeling which was set up to control for any differences in initial abilities and to determine the growth in performance associated with each treatment. Results indicated that when delayed elaborative feedback was paired with students being asked to recall/rework the problem, the largest student growth was observed. To dive deeper into student growth, both the differences in specific content-area improvement and the ability levels of students who improved the most were analyzed.