The Effect of Computer-Based Feedback in Game-Like Fluid Reasoning Tasks

Abstract

We focused on the effect of various types of feedback in a game-based logical reasoning test called Triton and the Hungry Ocean in elementary school students (grades 3-6, total N = 447). We formed four experimental feedback condition groups: no feedback (A), simple (B), elaborated (C), and learner-controlled feedback (D). We did not see a positive effect on performance in any of the feedback types. However, in group D, students tended to choose elaborated feedback more often with increasing task difficulty (r = .93). Also, group D students using elaborated feedback more often showed better performance without having a higher average IQ (CFT 20-R). Tracing of latencies after incorrect answers did not show differences in response times between groups. Time spent reviewing feedback after the first error compared to the time spent reviewing feedback for all subsequent incorrect answers differed only for elaborated feedback (C) by approximately 1 second.

Keywords

• feedback
• learning
• game-based performance tests
• metacognition