Test preparation in a high-stakes setting: experiments with testing and restudy strategies

This study investigated the transferability of the effects of testing (with and without feedback on performance) and restudy strategies on test performance in an authentic test preparation context. Three experiments were conducted in which students from a teacher training college in China reviewed educational psychology concepts using one of the strategies and took a post-intervention test. The design of the three experiments differed only in the settings: two were undertaken in high-stakes test preparation settings (n = 172; n = 257) while the third was conducted in a low-stakes setting (n = 115). Both testing with feedback and restudy were effective test preparation strategies with comparable effect sizes on immediate test performance in the high-stakes preparation conditions (Cohen’s d > 0.57). Testing with feedback had a significant effect on the course test results conducted either two days (Experiment 1, Cohen’s d = 0.58) or seven days (Experiment 2, Cohen’s d = 0.91) after the intervention compared to the control group while restudy did not. These significant effects were only observed in high-stakes preparation conditions, where students prepared for the high-stakes course test. There was no significant test preparation effect in the low-stakes condition. Overall, the findings highlight the critical role of testing with feedback in sustaining learning gains in authentic high-stakes contexts.

Oxford Review of Education. 1-21