Reliance on Visible Speech Cues During Multimodal Language Processing: Individual and Age Differences

Abstract

The current study demonstrates that when a strong inhibition process is invoked during multimodal (auditory-visual) language understanding: older adults perform worse than younger adults, visible speech does not benefit language-processing performance, and individual differences in measures of working memory for language do not predict performance. In contrast, in a task that does not invoke inhibition: adult age differences in performance are not obtained, visible speech benefits language performance, and individual differences in working memory predict performance. The results offer support for a framework for investigating multimodal language processing that incorporates assumptions about general information processing, individual differences in working memory capacity, and adult cognitive aging.

Notes

*p < .005 between groups.

Note. Correlation coefficients for young adults (N = 40) are above the diagonal and those for older adults (N = 40) are below the diagonal.

*p < .05;

**p < .01.

Note. Self-reported hearing loss was a non-significant predictor of variance in both analyses, and age did not significantly predict recall in the speech − alone condition, p values > .05.

****p < .00001.

Note. L and R refer to left and right ears. A and B refer to the two female speakers.

*Side participants in that condition were instructed to shadow. SA = speech + articulatory condition. S = speech (visual information on the screen was inconsistent with the words shadowed).

Note. The t values for backward digit span, reading span, and self-reported hearing loss were non-significant predictors of variance in this analysis, p < .05.

****p < .0001.