Abstract for: Providing structural transparency when exploring a model’s behavior: Effects on performance and knowledge acquisition

Prior exploration is an instructional strategy which has improved performance and knowledge acquisition in system-dynamics based learning environments, but only to a limited degree. This study investigates whether model transparency, showing users the internal structure of models, can extend the prior exploration strategy and improve learning even more. In an experimental study, participants in a web-based simulation learned about and managed a small developing nation. All participants were provided the prior exploration strategy but only half received prior exploration embedded in a structure-behavior diagram intended to make the underlying model’s structure more transparent. Participants provided with the more transparent strategy demonstrated better knowledge acquisition of the underlying model on an objective measure (multiple-choice posttest) but no difference on a subjective measure (open-ended verbal protocols based on short essay questions). Furthermore, their performance (managing the nation) was the equivalent to those in the less transparent condition. Combined with our previous studies, the results suggest that while prior exploration is a beneficial strategy for both performance and knowledge acquisition, making the model structure transparent in this way (with structure-behavior diagrams) is more limited in its effect and may depend on the participants’ level of expertise.