Laras S. Yuniarto, Amanda M. Seed, Juan-Carlos Gómez School of Psychology & Neuroscience, University of St Andrews, St Andrews, Scotland, UK
Many physical causal tasks do not allow mistakes to be corrected. According to Karmiloff-Smith’s Representational Redescription Model, this may inadvertently impair children’s learning, because self-correction may play a crucial role in progressing from more implicit and procedural representations to more explicit and abstract ones. Our previous study found that in 2.5-year-olds, self-correction in a trap-task may trigger partial representational change. However, questions still remain about the role of self-correction in representational change and the transition between implicit and explicit knowledge. The present study adapts the trap-task into a novel digital game about a hedgehog searching for food. Using the online science platform Gorilla, 2.5- to 3-year-olds (projected N = 150) either practiced the task while correcting their mistakes, without self-correction, or without self-correction but with the opportunity to achieve as much success as the self-correction group. All children then completed a transfer task in which a previous trap became a support, plus an optional interview. Explicit responses were measured using button-presses and interview answers; implicit responses were measured using webcam recordings of gaze direction. Data collection is ongoing, but we plan to conduct GLMM analyses to test the prediction that self-correction leads to more flexible learning of objects’ trap/support roles, creating a significant difference in transfer performance between the children who self-corrected and the other conditions. We will also conduct exploratory analyses of implicit looks and their relationship to explicit performance, with the expectation that implicit looks may not differ across conditions in the same way as explicit measures.
