Proceedings of the Annual Meeting of the Cognitive Science Society

This study uses latent space network modeling to explore how the structure of the mental lexicon might change from middle childhood (ages 7 to 11 years) to young adulthood. English-speaking children and adults (N = 21 per group) completed a repeated word association task where they generated the first word that came to mind in response to a list of 48 cues (24 nouns, 24 verbs) repeated three times. We applied mixed-effects models to examine how response characteristics, derived from robust metrics (e.g., WordNet, BERT), varied by group and list repetition. Semantic and phonological distance between responses and cues increased over list repetitions. Adult responses exhibited stronger semantic (word embedding and taxonomic) similarities, while child responses showed stronger phonological relatedness. To evaluate factors associated with producing specific responses, we modeled the corpus using latent space networks. This confirmed a stronger influence of distributional statistics in adults and phonological relatedness in children.