The testing effect is moderated by experimental design

Publication date: October 2016 Source:Journal of Memory and Language, Volume 90 Author(s): Neil W. Mulligan, Jonathan A. Susser, S. Adam Smith A number of encoding variables are moderated by experimental design, such that an effect (e.g., the generation effect) on free recall is larger in mixed-list than pure-list designs. Experiments 1 through 4 consistently found that the testing effect is moderated by experimental design across differences in materials, list structure, list length, presence of feedback, and rates of retrieval-practice success. In these experiments, each study list was followed by its own free recall test. However, when multiple lists are followed by a single test, the testing effect was equivalent for mixed and pure lists (Experiment 5), a limiting condition on the design effect. Overall, the results represent an important similarity between the testing effect and the other encoding variables moderated by design. More specifically, the results suggest an important parallel between the mnemonic effects of retrieving information from episodic (testing effect) and semantic (generation effect) memory.
Source: Journal of Memory and Language - Category: Speech Therapy Source Type: research