Speech Categorization Consistency Predicts Overall Language Abilities
Year: 2025
Type: preprint
Abstract: Recent research in speech perception highlights categorization consistency—the degree of trial-by-trial variability in response—as a more informative metric of individual differences in speech categorization over how gradient underlying category structure individuals have. However, its relevance to broader language functions remains unknown. In Experiment 1, English-speaking adults (n=57) completed a Visual Analog Scaling (VAS) task, rating a speech continuum on a continuous scale, alongside two language assessments. Results showed that a greater consistency in speech categorization strongly predicted better language abilities (accounting for 29% of the variance), whereas categorization gradiency showed no such relationship. Experiment 2 (n=76) added a VAS task in which participants performed the same task with images from a visual continuum, to determine whether the observed effects reflected domain-general categorization mechanisms. Replicating and extending the findings of Experiment 1, speech categorization consistency again emerged as a robust predictor of language outcomes, with an even larger effect size (~40% of the variance). Critically, commonality regression revealed that a substantial proportion of variance in language performance was uniquely attributable to speech, but not visual, categorization consistency. These provide compelling evidence for a speech-specific and robust link between speech categorization consistency and general language function.
Authors: Hyoju Kim, Bruce Tomblin, Bob McMurray
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4411753952