The Associative Learning Model versus the Hypothesis Testing Model: Using Error Patterns to Understand Preschoolers’ Word Learning

Abstract

A key question in childhood language development is how children are able to learn new words in spite of ambiguous real-world situations. Two potential explanations have dominated the literature to date. The classic model suggests that children catalog multiple possible referents while the competing model theorizes that children instead fixate on one possible referent. This research compares and contrasts these models by analyzing preschoolers' selection accuracy after they’ve made an error during a word-learning task. Findings suggest that neither model fully captures the complex problem of word learning during disambiguation tasks where linguistic cues can influence children’s learning. Instead, the results found here may, in fact, be best explained more complex models of learning that incorporate elements from both models.