Antecedents of examinee motivation during low-stakes tests: examining the variability in effects across different research designs

Abstract

To understand examinee motivation, researchers have used expectancy-value theory, which posits that examinee’s perceived value of the test and self-efficacy in the test domain impact test-taking effort. Conclusions from studies designed to examine these relations may be premature given methodological issues, most notably, use of cross-sectional designs. We illustrate how support for expectancy-value theory is impacted by the research design employed. Additionally, we examine if test emotions (e.g. anger) directly impact examinee effort, as posited by control-value theory. We gathered longitudinal data on perceived test value, test emotions, examinee effort, and test performance from college students (N = 955) at three points during a low-stakes test used for institutional accountability mandates. When variables were measured concurrently, the relation between perceived test value and expended effort was significant, even when including test emotions as predictors of effort. This value-effort relation was also observed when modelling longitudinal data, even when including the direct effects of emotions on subsequent effort. However, when autoregressive effects were estimated, the direct effect of value on effort was nonsignificant, whereas direct effects of some emotions on subsequent effort remained significant. Thus, when trying to understand student motivation during low-stakes testing, their test emotions should be considered.