Relations between the worked example and generation effects on immediate and delayed tests

The contradiction between the worked example effect that occurs when learners presented with more instructional guidance learn more than learners presented with less guidance and the generation effect that occurs when the reverse result is obtained can be resolved by the suggestion that the worked example effect is obtained using materials high in element interactivity, whereas simpler, low element interactivity materials result in the generation effect. A 2 (guidance: low vs. high) × 2 (element interactivity: low vs. high) × 2 (expertise: low vs. high) experiment investigated this hypothesis with high school trigonometry learners. On an immediate test, high guidance reflecting a worked example effect was found for novices, but a generation effect was obtained for more knowledgeable learners. In contrast, on a delayed test, a three-way interaction between guidance, element interactivity and expertise was found. This interaction was caused by a worked example effect for material high in element interactivity and a generation effect for material low in element interactivity for novices while for more knowledgeable learners, a generation effect was obtained for both low and high element interactivity materials. These results suggest firstly, that both the worked example and generation effects may be more likely on delayed than immediate tests and secondly, that the worked example effect relies on high element interactivity material while the generation effect relies on low element interactivity material.