Influence of stimulus sentence characteristics on speech intelligibility scores in hypokinetic dysarthria

This study was designed to investigate the effects of signal independent sentence characteristics on speech intelligibility scores. Six adult speakers with untreated hypokinetic dysarthria were audio recorded while orally reading randomly selected sentences from the Sentence Intelligibility Test (SIT). These audio recordings were then utilized in a transcription task to obtain speech intelligibility scores. The stimulus material was further analyzed to provide information regarding certain sentence characteristics that could potentially influence speech intelligibility scores (i.e., average number of syllables per word, total number of words per sentence, total number of independent vs. dependent clauses, ratio of function words to content words, and percent predictability). Multiple regression analysis revealed that sentence length and sentence predictability had a significantly unique influence on speech intelligibility scores for these selected SIT sentences. Analysis of variance further indicated that the mean intelligibility scores for short sentences were significantly greater than for long sentences. Mean intelligibility scores for sentences with high predictability were also significantly greater than for sentences with low predictability. There was no interaction of predictability with sentence length for these sentences.