Measuring scientific reasoning – a review of test instruments

Education systems increasingly emphasize the importance of scientific reasoning skills such as generating hypotheses and evaluating evidence. Despite this importance, we do not know which tests of scientific reasoning exist, which skills they emphasize, how they conceptualize scientific reasoning, and how well they are evaluated. Therefore, this article reviews 38 scientific reasoning tests. They used to primarily consist of multiple-choice questions, but since then have become more diverse, even including tests that automatically analyse virtual experiments. Furthermore, this review revealed that the tests focus on the skills of generating hypotheses, generating evidence, evaluating evidence, and drawing conclusions. Additionally, conceptualizations of scientific reasoning have moved towards a domain-specific set of different but coordinated skills over the years. Finally, concluding from gaps in test evaluation, a future focus should be on testing theoretical assumptions, comparing different scientific reasoning tests, and how relevant test results are in predicting criterion variables like academic performance.