## Overview This registered report examines how different types of input variability affect 7-8 year old children's learning of Japanese spatial postpositions (above/below). We tested discriminative learning theory predictions using a computerized training game. The present repository contains the submitted (not reviewed yet) Stage 2. ## Key Findings - High variability input significantly improved generalization to novel nouns compared to low variability input - No clear benefit was found for skewed distributions over either uniform condition - Individual differences in working memory modulated variability benefits - Online vs school testing revealed substantial environmental effects ## Materials Available - **[Stage 1 Registered Report][1]**: Pre-registered hypotheses and methods linked to this repository - **Stage 2 Final Report**: Complete with results and discussion uploaded with track-changes version and clean version. - **Data**: De-identified participant data - **Analysis Scripts**: R code for all analyses and simulations ## File Structure - `/Data/` - De-identified participant data - `/R/` - R scripts and output - `/Manuscript/` - Stage 2 clean and track-changes reports ## License This work is licensed under CC BY 4.0 [1]: https://osf.io/3q4yg/