September 2016
Volume 57, Issue 12
Open Access
ARVO Annual Meeting Abstract  |   September 2016
Top-down aspects of letter recognition in the near periphery
Author Affiliations & Notes
  • Manfred MacKeben
    Low Vision Research, The Smith-Kettlewell Eye Res Inst, San Francisco, California, United States
  • Footnotes
    Commercial Relationships   Manfred MacKeben, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science September 2016, Vol.57, 1950. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Manfred MacKeben; Top-down aspects of letter recognition in the near periphery
      . Invest. Ophthalmol. Vis. Sci. 2016;57(12):1950.

      Download citation file:

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

  • Supplements

Purpose : Purpose: Letter recognition in the near periphery is important for patients with a central scotoma, so that they can read using “eccentric viewing”. Previous work has shown that a “bottom-up” approach can improve letter recognition by modifying typographic features of easily confused letter pairs (MacKeben, VIR, 2000). However, Pilot data show that subjects can perceive certain characters irrespective of typographic features, presumably based on idiosyncratic “preferences”.

Methods : Methods: There were 14 normally sighted subjects (Ss) either in their twenties (22-29 yrs) or in their fifties (50-58 yrs), not including the author. The experiments used a sustained attention paradigm: Ss were asked to fixate a center mark on a screen and to direct their attention to a location indicated by a 1 s red dot at 8 deg eccentricity. Then a letter (36 arcmin) was shown tachistoscopically for 66 - 148 ms, adjusted to be near the individual threshold. Letters belonged to the Sloan set (Z, C, R, V, H, K, N, O, D, S - font: Arial Bold). Ss had to verbally report which letter they perceived. Each subject performed between 12 and 36 trial blocks of 20 trials each = 360 – 720 trials (median = 480). Results from all together 6420 such trials are reported here.

Results : Results: The 3 most frequently identified letters were V, D, C. The 3 least identified letters were N, H, R. Ss got between 12.4% and 35.6% of the presented letters wrong (median = 26.3%). The age of Ss was not correlated with error rate. When the most frequently misidentified letter N was shown, Ss most frequently perceived one of 5 other Sloan letters most frequently (V, H, K, R, D), + one outside of the Sloan set. When the second most frequently misidentified letter H was shown, Ss perceived one of 6 other Sloan letters most frequently (D, V, N, R, O, K), + one outside of the Sloan set.

Conclusions : Conclusions: When viewed in the near periphery, the letters in the Sloan set are not equally recognizable. The actually perceived letter may share typographical features with the shown letter, but it also may not (N seen as R or D, H seen as D or O). Tentatively, this could be explained by an expectation, possibly based on top-down mechanisms of visual perception in the near periphery, which would be in contrast to findings using foveal vision (Majaj et al., VR 42(9), 2002).

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.


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.