Differences in question reading and responding behaviour across reading achievement skills: An eye-tracking study

Abstract

Background: Reading comprehension (RC) tests are meant to assess students' ability to read and process information within passages but also require students to read and understand the questions. Just as students of different reading achievement use varying compensatory skills to develop comprehension of a passage, we suspected skill level might also be related to different behaviours as they read questions and response options. Methods: We analysed data of 174 students, with a mean age of 9.06 years. All students attended schools in the southeastern region of the United States and were assessed in English. Students were asked to read passages and respond to multiple-choice questions as their eye movements were recorded. Results: Compared to lower-achieving students, higher-achieving students (1) needed less time to read the passages and questions, (2) were sensitive to question type in that they took more time to read inferential-question stems compared to literal stems and (3) were less likely to re-read when answering questions, but if they did re-read, they were more likely to answer literal questions correctly. Conclusions: Just as lower-achieving students experience greater challenges reading passages and have fewer skills to overcome those challenges, they likewise expend more effort reading questions and searching passages for answers. Unfortunately, their searches are less likely to result in the correct answer choices. Findings emphasise teaching evidence-based test-taking strategies that prioritise comprehension and highlight eye-tracking technology's potential to inform RC test design and evaluation.