A binocular moving window technique to study the roles of the two eyes in reading

Abstract

Readers utilize parafoveal information about upcoming words and read less well when this information is denied. McConkie and Rayner (1975) enabled this issue to be explored by developing the moving window paradigm in which the experimenter varies the amount or the quality of the parafoveal information available around the current fixation point. We present a novel binocular version of the moving window technique to study the roles of the two eyes in reading, and we describe a basic experiment allowed by this technique. In the binocular moving window paradigm, each eye contributes its own window to a composite binocular window onto the text. We studied the reading of single lines of text in three conditions: no windows, a symmetrical 8-letters-left and 8-letters-right window for each eye, and a leftward-skewed 14-letters-left and 2-letters-right window for each eye. Note that both eyes saw the composite window onto the text. We tested the hypothesis that readers could be encouraged to generate a greater binocular disparity to augment their window onto the text and to provide a greater preview for one eye. The data offered limited support for this prediction. We observed considerable individual differences in both baseline fixation disparity and in readers' response to the critical asymmetric [14,2] window.

Keywords:

• Reading
• Binocularity
• Fixation disparity
• Moving windows
• Parafoveal preview
• Individual differences

We thank Wolfgang Einhäuser-Treyer for advice on the low-pass filtering and Mateo Obregón for advice on data analysis along with Barbara Juhasz, Simon Liversedge, Kevin Paterson, and Keith Rayner for comments on an earlier version of the manuscript.

We thank Wolfgang Einhäuser-Treyer for advice on the low-pass filtering and Mateo Obregón for advice on data analysis along with Barbara Juhasz, Simon Liversedge, Kevin Paterson, and Keith Rayner for comments on an earlier version of the manuscript.

Notes

¹ Note that this terminology was created independently of the preexisting terminology of crossed and uncrossed retinal disparities; the two should not be confused.

² The authors used an EyeLink 1000 system, which was also used in the present study. Note that earlier studies (e.g., Nuthmann & Kliegl, 2009) used EyeLink I and II systems.

³ This suggestion was made by Kevin B. Paterson.