A study where participants received assistance from various embodiments of a Misty robot while solving nonogram puzzles to determine if the physical embodiment of a robot has an effect on learning rates

This paper replicated the study The Physical Presence of a Robot Tutor Increases Cognitive Learning Gains by Leyzberg et al. to explore the robustness of the results that a physically embodied robot tutor increases the cognitive gains of the participants. To measure this effect, participants completed 3 nonogram puzzles under three conditions: (1) personalized visual advice with no robot, (2) personalized advice from a video representation of a robot, and (3) personalized advice from a physically present robot. ANOVA tests were conducted to determine if there were significant differences in the improvement times from puzzle to puzzle between the three conditions. There was a significant improvement from the first puzzle to the last puzzle, but we did not have enough evidence to conclude that any difference exists in the improvement times between the conditions. This study was conducted in person during the COVID-19 pandemic, limiting the total number of participants to 15. Our participants were also limited to a relatively homogeneous population, mainly male undergraduate students at Carnegie Mellon University. Although this study was not able to show conclusive evidence of differences between cognitive learning gains as an effect of robot embodiment, robot embodiment is still an area worthy of continued research.