Adaptive feedback from artificial neural networks facilitates pre-service teachers’ diagnostic reasoning in simulation-based learning
Michael Sailer, Elisabeth Bauer, Riikka Hofmann, Jan Kiesewetter, Julia Glas, Iryna Gurevych, Frank Fischer
Abstract
In simulations, pre-service teachers need sophisticated feedback to develop complex skills such as diagnostic reasoning. In an experimental study with N = 178 pre-service teachers about simulated pupils with learning difficulties, we investigated the effects of automatic adaptive feedback, which is based on artificial neural networks, on pre-service teachers' diagnostic reasoning. Diagnostic reasoning was operationalised as diagnostic accuracy and the quality of justifications. We compared automatic adaptive feedback with static feedback, which we provided in form of an expert solution. Further, we experimentally manipulated whether the learners worked individually or in dyads on the computer lab-based simulations. Results show that adaptive feedback facilitates pre-service teachers’ quality of justifications in written assignments, but not their diagnostic accuracy. Further, static feedback even had detrimental effects on the learning process in dyads. Automatic adaptive feedback in simulations offers scalable, elaborate, process-oriented feedback in real-time to high numbers of students in higher education.