Previous research has demonstrated that practice testing, or practice retrieval, is an effective learning technique. This method can be implemented in two ways: either through mandatory tests like class quizzes, or through voluntary self-testing without formal instruction. Regarding the latter, Liu and colleagues conducted a study to explore whether test anxiety and test stakes affect self-testing usage and, in turn, influence learning performance.
The study included one survey and four experiments. Data for the survey study were collected from 459 Grade 11 students at a school in northwestern China. It used a self-report questionnaire to measure trait test anxiety and one item to ask self-testing frequency. Academic performance was accessed with school exams results. Mediation analysis indicated that students with high test anxiety not only directly lowered exam scores but also indirectly did so through reducing the frequency of self-testing.
In laboratory experiments, participants underwent three stages: learning material, selecting a revision strategy (restudying or self-testing), and taking a test. Conducted at Beijing Normal University, the first experiment with 70 undergraduates showed that high trait test anxiety reduced the likelihood of choosing self-testing. The second experiment (n=25) revealed that self-testing was less common for high-stakes than low-stakes tests. The third (n=36) demonstrated that high stakes significantly impaired test performance by reducing self-testing.
The final quasi-experiment combined measures of test stake, testing anxiety, self-testing usage, and test performance. It randomly assigned two classes of students at Tianjin Vocational Institute to high- or low-stakes test conditions. State test anxiety was measured three time points: at the beginning of the experiment, after students were informed of the test stake condition, and after the study stage. Statistics revealed that the increase of state anxiety was higher in the high-stakes group compared to low-stakes group following test stakes manipulation. Chain mediation analysis showed that high stakes increased state anxiety, decreased self-testing selection, and subsequently hindered test performance.
This study supports the authors’ proposed mechanism whereby test anxiety negative impacted learning performance through discouraging of self-testing practice.
Source: Liu, S., Zhao, W., Shanks, D. R., Hu, X., Luo, L., & Yang, C. (2024). Effects of test anxiety on self-testing and learning performance. Educational Psychology Review, 36(2), 59. https://doi.org/10.1007/s10648-024-09882-1