Optimal Methods of Input for Pedagogic Engineering Software

Abstract

Computer simulations are commonly employed to teach intuitive causal engineering relationships, yet there is little research regarding what aspects make such simulations maximally pedagogically effective. This paper investigates how the discrete vs. continuous nature of user input, such as moving a dial vs. typing a parameter, affects actual learning as well as the user-perceived learning.  The N = 91 cohort size was not sufficient to establish statistical differences at the \(\alpha\) = 0.05 level, but at \(\alpha\) = 0.1 it was observed that simulations using continuous input caused students to learn more effectively, and that students who used continuous input methods believed, following their use, that continuous input learning methods were more effective than discrete input learning methods. Surprisingly, an inverse correlation was found between objectively-measured student understanding and subjectively-rated student belief of their own subject mastery. In other words, students who used continuous input simulations believed they were better teaching tools in general, even though they believed their specific learning was inadequate.