ABSTRACT
This article presents a taxonomic system for generated disablers (based on Elio, 1998) and generated alternatives. Based on the taxonomy, we distinguish three different types of knowledge that are advocated during generation tasks (1) situations that are semantically strongly related to the content of the premises (2) more remote situations and (3) the invalid or low quality counterexamples. Second, we look at the effect of working memory capacity on the nature of generated counterexamples. We found that participants with a high working memory capacity can generate more counterexamples and are flexible in their search process. Participants with low working memory generate less counterexamples and restrict themselves to the first type of counterexamples.
