Maximum Likelihood Estimation (MLE) of students’ understanding of vector subtraction

In this paper, we report on the impact that slight changes in question format have on student response to one‐dimensional vector subtraction tasks. We use Maximum Likelihood Estimation (MLE) analysis to analyze students’ responses on six very similar questions which vary in context (physics or mathematics), vector alignment (both pointing to the right or opposed), and operation (left‐right subtraction or right‐left subtraction). Responses on all questions are generally correct and do not vary by instructional week or even by course. Context and specific operation do not show significant differences. Vector alignment is significantly different, indicating that perception or heuristic thinking is a bigger cause of failure than conceptual deficit. The emphasis in this paper is an introduction to likelihood estimation.