Strong primary cue weighting is linked to robust dimension-selective attention

Speech perception is supported by multiple redundant acoustic cues. How do listeners decide how to prioritise these cues? Although computational models suggest that cue weighting reflects the statistical reliability of cues in language input, listeners show stable individual differences in weighting strategies. These differences may arise from variation in perceptual acuity (affecting the perceived reliability of specific dimensions) or from selective attention to the most informative cue. Here, we tested these two competing explanations by examining how perceptual cue weighting, dimension-selective attention, and perceptual acuity affected prosodic categorization in 54 speakers of English as a first language (L1) and 60 L1 Mandarin Chinese speakers. We found that individuals vary in whether they tend to rely on a single cue or integrate across cues when they use prosodic information to make categorical decisions about speech content. Cue weighting strategies correlated across prosody categorisation tasks, but not along acoustic dimensions, suggesting that individuals are not globally biased toward using one cur over another (e.g., pitch versus durational information). Instead, stronger selective attention to acoustic dimensions was associated with greater reliance on primary cues, indicating that attention might drive cue prioritisation. This relationship held across both English and Mandarin speakers, although cross-task correlations were weaker in the Mandarin group, possibly reflecting ongoing adaptation to second-language prosodic patterns. Overall, these findings suggest that individual cue weighting strategies are shaped less by perceptual acuity and more by the ability or propensity to selectively attend to acoustic dimensions, and moreover that speech perception involves active, goal-directed selection of relevant information.