Purchase this article with an account.
Susana T L Chung, Jean-Baptiste Bernard; Independent contributions of letter, word and sentence information to reading speed for people with macular disease. Invest. Ophthalmol. Vis. Sci. 2019;60(9):1064. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Reading involves the recognition of individual letters and words, and making sense of the sentence context. Previously, Pelli & Tillman (2007) showed a triple dissociation (i.e. independence) of the contributions of letter (L), word (W) and sentence (S) information to reading speed for people with normal vision. Here, we investigated whether or not the L, W and S information also contribute independently to reading speed for people with macular disease.
Seven observers with macular disease (age: 51–93, logMAR acuity: 0.34–1.1) read aloud sentences (word length: 8–14) with words presented one at a time, each for a fixed word exposure duration, using rapid serial visual presentation (RSVP). Reading speed was defined based on the RSVP word exposure duration that yielded 80% of words read correctly. Text was rendered in Courier. Two print sizes (1.4× & 0.8× the critical print size) were tested. We evaluated the contributions of L, W and S information to reading speed by selectively knocking out each of them — L was removed by substituting letters in words with similar ones based on a confusion matrix; W was removed by alternating upper- and lowercase letters within words; and S was removed by scrambling the word order within a sentence. Reading speeds were measured for each of these conditions and various combinations of them, and were subsequently normalized to that obtained for text with all sources of information present.
Averaged across observers, the contributions of L, W and S information to reading speed were 50%, 18% and 20%, respectively, for the larger print size; and 39%, 25% and 24% for the smaller print size. Observers were still able to read when L, W and S were absent (13–17%). For both print sizes, the contribution of each source of information to reading speed was the same whether or not the other information was available, implying independence of the L, W and S processes in reading.
The independent contributions of L, W and S information to reading speed, along with the higher reliance on L information in reading, suggest that methods to improve reading for people with macular disease should focus more on providing a reliable source of letter information. However, the additive contributions of W and S mean that there are still additional benefits in providing reliable word and sentence information, especially for small print sizes.
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