How can we test whether state-of-the-art generative models, such as Blender and GPT-3, are good AI teachers, capable of replying to a student in an educational dialogue? Designing an AI teacher test is challenging: although evaluation methods are much-needed, there is no off-the-shelf solution to measuring pedagogical ability. This paper reports…

In an effort to better align item development to goals of assessing higher-order tasks and decision making, complex decision trees were developed to follow clinical reasoning scripts and used as models on which multiple-choice questions could be built. This approach is compatible with best-practice assessment frameworks like Evidence Centered…