Is semantic preview benefit due to relatedness or plausibility?

Abstract

There is increasing evidence that skilled readers of English benefit from processing a parafoveal preview of a semantically related word. However, in previous investigations of semantic preview benefit using the gaze-contingent boundary paradigm the semantic relatedness between the preview and target has been confounded with the plausibility of the preview word in the sentence. In the present study, preview relatedness and plausibility were independently manipulated in neutral sentences read by a large sample of skilled adult readers. Participants were assessed on measures of reading and spelling ability to identify possible sources of individual differences in preview effects. The results showed that readers benefited from a preview of a plausible word, regardless of the semantic relatedness of the preview and the target. However, there was limited evidence of a semantic relatedness benefit when the plausibility of the preview was controlled. The plausibility preview benefit was strongest for low proficiency readers, suggesting that poorer readers were more likely to program a forward saccade based on information extracted from the preview. High proficiency readers showed equivalent disruption from all nonidentical previews suggesting that they were more likely to suffer interference from the orthographic mismatch between preview and target. (PsycINFO Database Record

Parafoveal preview benefit in sentence reading: Independent effects of plausibility and orthographic relatedness

“…The results replicated earlier evidence that a plausible preview benefits first-pass reading, relative to an implausible preview (Veldre & Andrews, 2016a). We also observed robust effects of preview-target orthographic relatedness on all reading measures.…”

“…As elaborated by Veldre and Andrews (2016a), the effects of preview plausibility are compatible with the mechanisms that Schotter et al (2014) proposed to account for semantic preview benefits in E-Z Reader 10 (Reichle, Warren, & McConnell, 2009). Central to this account is the assumption that the coordination between lexical processing and saccadic planning operates equivalently under normal reading conditions and in the boundary paradigm.…”

“…The significant orthographic relatedness effect in the present data contrasts with the lack of a semantic relatedness effect in our earlier work (Veldre & Andrews, 2016a). This is consistent with the sublexical-integration account described above.…”

“…Schotter et al's (2014) simulations estimated that processing of parafoveal words had advanced to the L 2 stage (lexical access) on only 8 % of trials, and the likelihood of completing the postlexical processing stage (I) required to generate contextually based semantic preview effects would be even lower. The rarity of the sequence of events necessary to observe parafoveal semantic processing in E-Z Reader contrasts with the robust effects of preview plausibility observed by Veldre and Andrews (2016a) and in the present data. The mechanisms accounting for postlexical effects in E-Z Reader are relatively rudimentary and unexplored.…”

“…To provide insight into semantic parafoveal processing, Veldre and Andrews (2016a) investigated whether semantic preview benefit is caused by facilitation from overlapping semantic features shared by the preview and the target word or from the contextual plausibility of the preview word, two factors that have previously been confounded. The results showed that, when the plausibility of the preview word in the sentence was controlled, no effect of preview-target semantic relatedness on fixation durations emerged.…”