Abstract

College-level general chemistry courses often have the unfortunate reputation of requiring extensive memorization and recall of disparate and disconnected trivia. However, the chemical education research literature suggests that students in these courses struggle for a variety of reasons unrelated to recall including poor preparation in previous chemistry courses, deficient algebraic skills, and working memory limitations. In educational research, “working memory” is a psychological construct that indicates a person’s ability to simultaneously hold and manipulate information. Although various theories of working memory abound, they share a common prediction with respect to problem solving: if the working memory demand of a question is greater than the problem solver’s working memory capacity, he or she will fail to answer the question correctly. Consequently, questions with high working memory demands tend to assess students’ working memory capacities rather than their content knowledge (Niaz, 1996) (Tsaparlis, Kousathana, & Niaz, 1998). As students’ grades should reflect their content knowledge and not a cognitive trait, it is essential for instructors to monitor the working memory demands of their assessment items. Unfortunately, the historical methods of measuring working memory demand and capacity are unsatisfactory. First, assessments of questions’ working memory demands typically involve performing a task analysis (e.g., counting the discrete steps in solving a problem) from an expert’s perspective. Values determined in this manner assume that the student takes the same problem-solving route of the expert, which is highly unlikely given that experts and novices categorize and approach questions differently (Chi, Feltovich, & Glaser, 1981). Second, working memory capacity is often measured via digit-span recall tests (e.g., recounting a string of digits in reverse order) given immediately before or soon after a problem-solving session. Such tests cannot be given while a student is problem solving as they are disruptive. Recent research has explored the utility of noninvasive monitoring of physiological responses as measures of working memory. Previous research has drawn correlations between pupil diameter and ones load on memory. It has been discovered that there is a direct correlation between the size of ones pupil and their cognitive load (Kahneman, & Beatty, 1966). They found that the greater the number of digits each participant had to recall, the greater the diameter of their pupil was. This is referred to as the task-evoked pupillary response (TEPR). In light of this discovery, eye-tracking research has begun to use pupil diameter as a measure of working memory. The purpose of this study is to validate the method of using pupil diameter as a direct measurement of working memory demand. This study will use existing measurements of working memory (digit-span recall tests) to correlate to working memory demand. This will also allow future work to use pupil diameter as a real-time measure of working memory demand within a variety of tasks.