Open-book tests: search behaviour, time used and test scores

Because of the increasing medical knowledge and the focus of medical education on acquiring competencies, the use of open-book tests seems inevitable. Dealing with a large body of information, indicating which kind of information is needed to solve a problem, and finding and understanding that knowledge at the right moment are behaviours that cannot be assessed during closed-book tests.

To examine whether there is a relationship between students' search behaviour - using references or not when answering a question - during open-book tests and their test scores.

Second- (N = 491) and third-year medical students (N = 325) participated in this study. Search behaviour was operationalized as the number of questions for which students consulted their references. Furthermore, we collected data on the time students spent on answering all open-book questions and their test scores. To determine the relations, we calculated Spearman's and Pearson's correlations.

Second- and third-year students consulted their references for 87% and 73% of the questions and spent 5.0 and 4.3 min on answering an open-book question, respectively. We did not find significant correlations between search behaviour and test scores.

Both 'well' and 'poorer performing' students often consulted their references. Spending almost 5 min per open-book question in multiple choice format seems to be too much. More research is needed to establish optimal open-book test time and to explore how 'well performing' students use their references during open-book tests.