Purpose: To assess whether case-based questions elicit different thinking processes from factual knowledge-based questions.
Method: 20 general practitioners (GPs) and 20 students solved case-based questions and matched factual knowledge-based questions while thinking aloud. Verbatim protocols were analysed. Five indicators were defined: extent of protocols; immediate responses; re-reading of information given in the stem or case after the question had been read; order of re-reading information, and type of consideration, i.e. 'true-false' type or 'vector', that is, a deliberation which has a magnitude and a direction.
Results: Cases elicited longer protocols than factual knowledge questions. Students re-read more given information than GPs. GPs gave an immediate response on twice as many occasions as students. GPs re-ordered the case information, whereas students re-read the information in the order it was presented. This ordering difference was not found in the factual knowledge questions. Factual knowledge questions mainly led to 'true-false' considerations, whereas cases elicited mainly 'vector' considerations.
Conclusion: Short case-based questions lead to thinking processes which represent problem-solving ability better than those elicited by factual knowledge questions.