Many experiments have shown that the retrieval of already learning material helps to consolidate memory. Recent research, however, revealed the complexity of the text-material as a boundary condition for this effect. One of few experiments in favor of this assumption was published by de Jonge and colleagues (2015, EPR). In a series of two experiments, they showed that the benefit of testing compared to restudying occurred only for a “less complex” (incoherent, sentences in a scrambled order) but not for a “complex” (coherent, sentences in an inferential order) version of a science text about black holes. The dependent variable was measured by a cloze test. We present a direct replication of their original studies’ critical conditions, differing in terms of statistical power and with respect to methodological design aspects: 2 Experiments varying final test delay and learning strategy in de Jonge et al.’s study; 1 Experiment varying learning strategy and text complexity here. In contrast to the original study, we found a main effect of retrieval practice increasing the long-term memory performance (one-week delay) independent from the text-material complexity.

Reference:
de Jonge, M., Tabbers, H. K., & Rikers, R. M. (2015). The effect of testing on the retention of coherent and incoherent text material. Educational Psychology Review, 27, 305–315. https://doi.org/110.1007/s10648-015-9300-z