Are verbal transformations sensitive to age differences and stimulus properties?

Verbal transformations are illusory changes that arise from listening to a spoken word repeated over and over for a prolonged period of time. Past research has yielded some evidence of age differences in participants' reports of verbal transformations. The goal of the present study was to examine reports of illusory percepts in young and older adults to determine whether participant characteristics (i.e., age, cognitive integrity, hearing acuity, vocabulary knowledge, and mood) and differences in sensitivity to properties of the stimulus material could account for age-related declines in reports of illusory changes. We observed age-related declines in new forms (illusory percepts) but merely a trend in the same direction for transitions (changes from the veridical sensory experience to illusory percepts and reversals to either the veridical sensory experience or other percepts). There was no evidence that participant characteristics other than age accounted for reports of illusory changes. However, we found age differences in participants' sensitivity to properties of the stimulus material: Young adults but not older adults were sensitive to the frequency of the veridical percepts. In contrast, both age groups were sensitive to the size of the neighborhood of similar-sounding words. These findings are discussed regarding the transmission deficit hypothesis (MacKay & Burke, 1990).

Text familiarity, word frequency, and sentential constraints in error detection.

The present study examines whether the frequency of an error-bearing word and its predictability, arising from sentential constraints and text familiarity, either independently or jointly, would impair error detection by making proofreading driven by top-down processes. Prior to a proofreading task, participants were asked to read, copy, memorize, or paraphrase sentences, half of which contained errors. These tasks represented a continuum of progressively more demanding and time-consuming activities, which were thought to lead to comparable increases in text familiarity and thus predictability. Proofreading times were unaffected by whether the sentences had been encountered earlier. Proofreading was slower and less accurate for high-frequency words and for highly constrained sentences. Prior memorization produced divergent effects on accuracy depending on sentential constraints. The latter finding suggested that a substantial level of predictability, such as that produced by memorizing highly constrained sentences, can increase the probability of overlooking errors.