July 2019
Volume 60, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2019
Word neighborhood size is not a limiting factor of reading speed with central field loss
Author Affiliations & Notes
  • Aurelie Calabrese
    Aix-Marseille University, Marseille, France
    CNRS, Marseille, France
  • Lauren Sauvan
    North Hospital, Marseille, France
  • Carlos Aguilar
    Nice Sophia-Antipolis University, Nice, France
    CNRS, Marseille, France
  • Eric Castet
    Aix-Marseille University, Marseille, France
    CNRS, Marseille, France
  • Footnotes
    Commercial Relationships   Aurelie Calabrese, None; Lauren Sauvan, None; Carlos Aguilar, None; Eric Castet, None
  • Footnotes
    Support  Fondation de France - Berthe Fouassier
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 1062. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Aurelie Calabrese, Lauren Sauvan, Carlos Aguilar, Eric Castet; Word neighborhood size is not a limiting factor of reading speed with central field loss. Invest. Ophthalmol. Vis. Sci. 2019;60(9):1062.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose : Word neighborhood size is a measure of orthographic similarity defined as the number of words that can be produced by changing a letter in a word of the same length. For normal readers, this factor has a facilitator effect on word recognition (the more neighbors, the easier to identify). Given that the presence of a central scotoma hides portions of the text, people with central field loss (CFL) may tend to rely more on inference while reading. Here we investigate the hypothesis that larger word neighborhood size reduces reading speed with CFL because of increased uncertainty.

Methods : 18 individuals with binocular CFL (mean age = 74 ± 15, mean acuity = 0.83 ± 0.27 logMAR) read French sentences displayed with the self-paced reading paradigm, which allows to measure reading time for each individual word during eye-mediated sentence reading. On each trial, a short sentence was presented on the screen with all words masked by chains of x. Participants were asked to read the entire sentence out loud, unmasking words one at a time with a keyboard press. 16 pairs of meaningful sentences were constructed so that each sentence of a pair contained a target word with few neighbors, while the other sentence contained a target synonym with many neighbors. Linear mixed-effects models were used to assess the effect of word neighborhood size (Analysis 1) and by-pair relative neighborhood size (Analysis 2) on word reading time, while controlling for correlated factors such as word length and frequency.

Results : For well-read target words (mean accuracy = 89.4%), reading time was on average 2.97 seconds. Increasing the absolute value of word neighborhood size increased word-reading time, but this effect was found to be small and non-significant (Analysis 1). Similar results were found in a by-pair investigation (Analysis 2).

Conclusions : Contrary to normal reading, word neighborhood size does not have a facilitator effect when reading with CFL. This psycholinguistic factor does not seem to be a predictor of reading performance for CFL patients. These findings are especially relevant in the context of text simplification targeted to low vision, which aims to increase reading performance by reducing the psycholinguistic complexity of text. To be efficient, reading aids using this technique will have to consider the individual contributions of various psycholinguistic factors (e.g. neighborhood size, word frequency, etc.).

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.