The effectiveness of generative AI tools in education depends largely on prompt engineering--the practice of designing inputs and interactions that guide AI systems to produce relevant, high-quality outputs. This systematic literature review examines empirical studies published since the release of ChatGPT in late 2022, identifying two broad…

Providing test questions after an initial study phase is a common instructional technique. In theory, questions that require higher-level (deep) processing should be more beneficial than those that require lower-level (shallow) processing. However, empirical evidence on the matter is inconsistent. To shed light on two potential reasons for these…