The Effect of Test Priming on Subsequent Reading Behavior

Abstract

To effectively design learning, instructional designers must understand how a specific intervention has what effect on a learner. One common intervention whose effect is multifaceted and often contested is testing. While tests are an important tool in understanding learners and fostering learning, not all its effects are positive. Many studies have investigated the learning gains of testing though another test. However, less attention has been given to why these gains may occur, or what effects it may have on the learning process.

Thus, the current study aims to understand the effect the existence and type of testing has on one crucial learning process, reading. There are research with similar intentions that have depended on learners self-reports, but the current study aims to produce precise, quantifiable results through eye-tracking methodologies. In the experiment, eye-tracking, interview and test result data was collected on participants randomly assigned to a memory-based test, higher-order thinking-based test or no test/control condition. The reading patterns of participants before and after testing were recorded and analyzed together with qualitative observations.

Results confirm that tests alter reading strategies for subsequent text. Key findings include a decrease in overall reading time and fixation for the no test condition while no such change was found for both test conditions. Readers in the memory test condition showed an increased tendency to pay more attention to definitions, keywords, and relations, and more regressive reading behavior. Readers in the higher-order thinking test condition paid more attention to relations. Overall, readers in the memory test condition seem to show the most change in reading behaviors. Possible explanations for this phenomenon were given, including preexisting expectancies, affective reactions to memory-based questions, and the nature of higher-order thinking questions used in the present study. Future directions for research are suggested based on the findings.