The testing effect and the complexity of the learning material: metacognition and self-regulation as keys to lasting learning

One method that has been strongly recommended to yield lasting learning is to engage learners in testing, for example, in terms of practice tests. This benefit of testing over restudying, known as the testing effect, has yielded medium effect sizes in meta-analyses but it has been questioned whether testing is effective for learning with complex materials such as complex expository texts. Moreover, final tests typically take place no later than two weeks after learning and little is known about how best to implement testing in the classroom. In the first project phase, we have investigated the testing effect in school classrooms (11th grade German classes) with low- and high-complexity versions of a schoolbook text and found a testing effect for lasting learning eight weeks after initial learning not only for low- but also for high-complexity texts. The current project aims at increasing integration of the studies in the lessons and more closely aligning the measures with further learning objectives in class. An additional focus will be on metacognition and self-regulated learning. Here, studies have found that learners rarely test themselves. This has been attributed to learners underestimating the utility of testing as indicated by low judgments of learning. However, in the first project phase, we demonstrated that learners appreciate the effectiveness of testing for a low-complexity text when their judgments are made offline, that is, when they only imagine learning with practice testing versus restudying and do not get negative performance feedback from a practice test. The second project phase aims at investigating whether this finding generalizes to more complex texts and to final test delays of eight weeks. Moreover, we assume that learners’ rare use of testing is also due to its perceived cognitive costs, since desirable difficulties such as testing are characterized not only by high effectiveness but also by high effort. Cost-benefit assessments could speak even more against testing when learning with complex texts. To test this assumption, we will focus on the self-regulated use of practice tests and restudying to gain insight on how text complexity affects students’ metacognitive judgments, effort ratings, and study decisions and how these factors impact lasting learning. Given that we found robust testing effects even with complex texts (and assuming that learners indeed test themselves too rarely with complex texts), we will further focus on increasing the use of practice tests by investigating the impact of an informed intervention on study decisions and lasting learning. The second project phase will encompass five highly powered, preregistered classroom experiments with established materials from the first project phase as well as newly developed materials in collaboration with experts on didactics of German.

Keywords: testing effect; complexity; element interactivity; learning with texts

Click here to see our team.