Department of Psychology & Neuroscience, The University of North Carolina at Chapel Hill, Chapel Hill, NC, USA.
Department of Psychology, CB#3270, The University of North Carolina at Chapel Hill, Chapel Hill, NC, 26599-3270, USA.
Atten Percept Psychophys. 2023 Oct;85(7):2475-2487. doi: 10.3758/s13414-023-02762-w. Epub 2023 Aug 2.
Skilled reading involves processing the upcoming word in parafoveal vision before it is fixated, leading to shorter fixations on that word. This phenomenon, parafoveal preview benefit, is a key component of theoretical models of reading; it is measured using the invisible boundary paradigm, in which reading times on a target word are compared for instances when preview is accurate and when the target word is masked while in the parafovea. However, parafoveal masks have been shown to induce unintentional processing costs, thereby inflating measures of preview benefit. The degraded mask has been explored as a potential solution to this problem, leading to mixed results. While previous work has analyzed the preview effect by comparing mean reading times on the target word, the present study provides a more comprehensive analysis by examining the distribution of the preview effect across target word fixation times for unrelated and degraded masks. Participants read sentences containing target words whose preview was either identical, unrelated, or degraded, and their eye movements were recorded. Analyses revealed that although there were no mean differences between reading times for the unrelated and degraded conditions, the pattern of the effects varied as a function of target word fixation times. Unrelated masks resulted in positively sloped generally linear delta plots, while degraded masks resulted in relatively flat delta plots for fixations longer than 200 ms. These differences suggest that different cognitive mechanisms are involved in the processing of the two mask types. Implications for understanding and measuring preview benefit are discussed.
熟练阅读涉及在注视前处理视区外即将到来的单词，从而减少对该单词的注视时间。这种现象，即视区外预览优势，是阅读理论模型的关键组成部分；它通过不可见边界范式来衡量，在这种范式中，目标词的阅读时间会在预览准确和目标词在视区外被掩蔽的情况下进行比较。然而，视区外掩蔽已被证明会导致无意识的处理成本，从而夸大了预览优势的测量结果。退化掩蔽已被探索作为解决此问题的潜在方法，但结果喜忧参半。虽然之前的研究通过比较目标词上的平均阅读时间来分析预览效果，但本研究通过检查无关和退化掩蔽的目标词注视时间分布，提供了更全面的分析。参与者阅读包含目标词的句子，目标词的预览要么相同，要么无关，要么退化，同时记录他们的眼动。分析表明，尽管无关和退化条件下的阅读时间没有差异，但效应的模式随目标词注视时间的变化而变化。无关掩蔽导致一般线性 delta 图呈正斜率，而退化掩蔽导致注视时间超过 200 毫秒的 delta 图相对平坦。这些差异表明，两种掩蔽类型的处理涉及不同的认知机制。讨论了对理解和测量预览优势的影响。
