Binocular advantages for parafoveal processing in reading

Mirela Nikolova, Stephanie Jainta, Hazel Blythe, Simon Liversedge

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)

Abstract

During reading, binocular visual input results in superior performance and is particularly important in the pre-processing of parafoveal text prior to direct fixation. It is not yet clear whether binocular vision in the parafovea is necessary for accurate saccadic targeting, or for efficient pre-processing of upcoming text, prior to direct fixation. In the present sentence reading experiment, we used a dichoptic gaze-contingent moving window paradigm in order to establish 1) how much parafoveal binocular input is necessary for fluent reading and 2) which aspect of parafoveal processing is more reliant on binocular vision. Eye movement measures revealed that reading was disrupted unless word N + 1 was entirely binocular in the parafovea, though no additional benefit was observed when word N + 2 was also binocular. Additionally, while fixation durations and reading times were clearly affected by the manipulation, similarly pronounced changes in binocular saccadic parameters such as accuracy, speed, amplitude and velocity were not observed. We concluded that the disruption to reading caused by presenting monocular text to the right of fixation cannot be attributed to difficulties in targeting binocular saccacdes, but instead results from a decreased efficiency in the pre-processing of parafoveal text. These results provide further demonstration for the importance of binocular vision during written text processing.
Original languageEnglish
Pages (from-to)56-63
Number of pages8
JournalVision Research
Volume145
Early online date19 Mar 2018
DOIs
Publication statusPublished - Apr 2018

Keywords

  • Reading
  • Binocular vision
  • Eye movements
  • Parafoveal processing

Fingerprint

Dive into the research topics of 'Binocular advantages for parafoveal processing in reading'. Together they form a unique fingerprint.

Cite this