Predictability effects and parafoveal processing in older readers.

Abstract

Normative aging is accompanied by visual and cognitive changes that impact the systems that are critical for fluent reading. The patterns of eye movements during reading displayed by older adults have been characterized as demonstrating a trade-off between longer forward saccades and more word skipping versus higher rates of regressions back to previously read text. This pattern is assumed to reflect older readers' reliance on top-down contextual information to compensate for reduced uptake of parafoveal information from yet-to-be fixated words. However, the empirical evidence for these assumptions is equivocal. This study investigated the depth of older readers' parafoveal processing as indexed by sensitivity to the contextual plausibility of parafoveal words in both neutral and highly constraining sentence contexts. The eye movements of 65 cognitively intact older adults (61-87 years) were compared with data previously collected from young adults in two sentence reading experiments in which critical target words were replaced by valid, plausible, related, or implausible previews until the reader fixated on the target word location. Older and younger adults showed equivalent plausibility preview benefits on first-pass reading measures of both predictable and unpredictable words. However, older readers did not show the benefit of preview orthographic relatedness that was observed in young adults, and showed significantly attenuated preview validity effects. Taken together, the data suggest that older readers are specifically impaired in the integration of parafoveal and foveal information but do not show deficits in the depth of parafoveal processing. The implications for understanding the effects of aging on reading are discussed.

How do task demands and aging affect lexical prediction during online reading of natural texts?

“…Zhang et al (2022) note that previous eye-movement research on aging has been dominated by studies of the reading of isolated sentences. Indeed, we have observed typical aging effects in an overlapping sample of older adults reading isolated sentences for comprehension (Veldre et al, 2022). This may suggest that the characteristic eye-movement signature of older readers that has been attributed to the risky reading strategy does not generalize to paragraph reading.…”

“…The major eye-movement signatures claimed to index this strategy were slower reading times due to longer fixations and more regressions (attributed to slower lexical processing), and increased skipping, longer forward saccades, and enhanced word frequency and predictability effects (which were attributed to a compensatory increased reliance upon prediction to "guess" upcoming words; Rayner et al, 2006). Although the risky reading hypothesis has motivated further research (e.g., Paterson et al, 2020) and computational modelling (McGowan & Reichle, 2018; see also Laubrock et al, 2006), questions have been raised about the strength and consistency of the evidence underpinning it (Payne & Silcox, 2019;Veldre et al, 2021Veldre et al, , 2022.…”

“…This reliance on probabilistic predictions may reflect a resource allocation strategy that harnesses older adults' richer crystallized linguistic knowledge to enhance lexical processing and the construction of meaning in order to compensate for declines in processing speed and working memory (Stine-Morrow et al, 2008). This interpretation may be extended to account for evidence enhanced predictability effects among older adults in studies that typically use isolated high-constraint sentences (e.g., Veldre et al, 2022). Although Zhang et al's (2022) meta-analysis did not find significant age differences in predictability effects for alphabetic scripts due to relatively few studies meeting the inclusion criteria, there were numerically larger predictability effects for older adults on gaze and total duration, but not skipping.…”

How do task demands and aging affect lexical prediction during online reading of natural texts?

“…Zhang et al (2022) note that previous eye-movement research on aging has been dominated by studies of the reading of isolated sentences. Indeed, we have observed typical aging effects in an overlapping sample of older adults reading isolated sentences for comprehension (Veldre et al, 2022). This may suggest that the characteristic eye-movement signature of older readers that has been attributed to the risky reading strategy does not generalize to paragraph reading.…”

“…The major eye-movement signatures claimed to index this strategy were slower reading times owing to longer fixations and more regressions (attributed to slower lexical processing), and increased skipping, longer forward saccades, and enhanced word frequency and predictability effects (which were attributed to a compensatory increased reliance on prediction to "guess" upcoming words; Rayner et al, 2006). Although the risky reading hypothesis has motivated further research (e.g., Paterson et al, 2020) and computational modeling (McGowan & Reichle, 2018; see also Laubrock et al, 2006), questions have been raised about the strength and consistency of the evidence underpinning it (Payne & Silcox, 2019;Veldre et al, 2021Veldre et al, , 2022.…”