Reset Button and Instructional Advice in Computer Simulations

Two experiments involving 107 (Study 1) and 98 (Study 2) students investigated how a computer simulation should be designed and how students should be supported in order to optimize their learning outcome. Thereby, a reset button that served to reset the visualization to its initial state and enabled a more systematic exploration as well as instructional advice were used in the simulation. An introductory text and a simulation were given to students in order to learn fundamental concepts about self-organizing maps, which is a special kind of neural network. Learners who received a reset button in the simulation did not perform better on succeeding retention or transfer tests than learners who did not receive a reset button. Furthermore, nearly 60% learners did not use the button at all (Study 1). Therefore, instructional advice was added before the simulation that asked the student to examine the different parameters in the simulation systematically through the use of the reset button (Study 2). Students who received that advice used the button more often and performed better on transfer than students not receiving that advice. The results can be explained with the extended Scientific Discovery as Dual Search model and with the zone of proximal development.