Measuring students’ conceptual understanding of wave optics: A Rasch modeling approach

Even graduate physics students have many misconceptions about basic wave optics phenomena. This suggests that there is much room for improvement of the traditional wave optics curriculum. An effective way for initiating a curriculum change is to reconsider and revise the expected learning outcomes and corresponding assessment instruments. By systematically enriching our wave optics instruction and assessment with conceptual tasks, we may increase the probability of students actively engaging in learning the conceptual aspects of wave optics. In this paper, we present the process of developing an item bank for measuring understanding of wave optics in typical introductory physics courses at universities. Thereby, the Rasch modeling approach has been used. The development of the item bank has been guided by results from multiple expert and student surveys, as well as from group interviews and think aloud interviews. Altogether 65 multiple-choice items with a single correct answer and three distractors have been prepared for field testing. Until now, 35 out of 65 items have been field tested by means of a paper and pencil survey which included 188 participants from five universities in Bosnia and Herzegovina, Croatia, and Slovenia. The field test showed that 32 out of 35 items have good psychometric characteristics and that they may be very useful for uncovering students’ misconceptions in wave optics.

Figure 1

Examples of item characteristic curves for the Rasch model.

Figure 2

Distribution of coded segments with respect to the main coding categories—a summary for the pool of 24 items.

Figure 3

Distribution of solving strategy categories—a summary for the pool of 24 items.

Figure 4

Scatterplots for full-test-based versus subtest-based WOT item estimates. Identity line passes through the origin and is red in color. Cross hairs represent 68% confidence intervals. (a) Scatterplot for items related to basic wave optics concepts. (b) Scatterplot for items related to applications of wave optics concepts.

Figure 5

Evidence for item difficulty invariance. Numbers represent items and a 95% confidence band is shown.

Figure 6

Wright’s map for inspecting person-item targeting and empirical hierarchy of item difficulties.

Figure 7

The test information function for the 32-items scale.

Figure 8

Proportion of correct answers for all 35 field-tested items.