Effects of age and hearing loss on recognition of unaccented and accented multisyllabic words

Abstract

The effects of age and hearing loss on recognition of unaccented and accented words of varying syllable length were investigated. It was hypothesized that with increments in length of syllables, there would be atypical alterations in syllable stress in accented compared to native English, and that these altered stress patterns would be sensitive to auditory temporal processing deficits with aging. Sets of one-, two-, three-, and four-syllable words with the same initial syllable were recorded by one native English and two Spanish-accented talkers. Lists of these words were presented in isolation and in sentence contexts to younger and older normal-hearing listeners and to older hearing-impaired listeners. Hearing loss effects were apparent for unaccented and accented monosyllabic words, whereas age effects were observed for recognition of accented multisyllabic words, consistent with the notion that altered syllable stress patterns with accent are sensitive for revealing effects of age. Older listeners also exhibited lower recognition scores for moderately accented words in sentence contexts than in isolation, suggesting that the added demands on working memory for words in sentence contexts impact recognition of accented speech. The general pattern of results suggests that hearing loss, age, and cognitive factors limit the ability to recognize Spanish-accented speech.

Figures

FIG. 1.

Mean pure-tone thresholds (in dB HL, re: ANSI, 2010) of the three listener groups.

FIG. 2.

Mean recognition of one-, two-, three-, and four-syllable words presented in isolation by three listeners groups. The three panels show data for stimuli recorded by three talkers varying in accent (panel a = no accent, panel b = mild accent, panel c = moderate accent).

FIG. 3.

Mean recognition of one-, two-, three-, and four-syllable words presented in sentence contexts by three listeners groups. The three panels show data for stimuli recorded by three talkers varying in accent (panel a = no accent, panel b = mild accent, panel c = moderate accent).

FIG. 4.

Recognition of moderately accented words presented in isolation vs sentence contexts as a function of length of syllables by three listener groups.

FIG. 5.

Recognition of unaccented, mildly accented, and moderately accented words of varying length of syllables, averaged across conditions with fixed length of syllables (“average”) vs a condition with trial-to-trial variation in length of syllables (“mixed”) by three listener groups.