People with severe low vision typically can read only with large letter sizes achievable on computer screens, or on closed-circuit television video magnifier reading aids (CCTVs). Even with magnified text, reading rates are usually substantially slower than those of normally sighted readers. A computer text presentation mode that has shown promise in enhancing low vision reading performance is rapid serial visual presentation (RSVP). In RSVP, words are presented sequentially, one at a time, on a computer monitor, at a uniform rate of presentation. RSVP makes more modest demands on eye movements than normal reading, and this is why it seems so promising for low vision. Unfortunately, RSVP studies with low vision have not borne out this promise, in that low vision readers reap only modest benefits (if any), whereas ironically, normally sighted readers gain significant benefits.

Drs. Aries Arditi, Bruce Rosenthal, Kent Higgins, and Lei Liu, with the assistance of Katherine Lee have evaluated a method of text display for low vision reading developed by Dr. Arditi, called elicited sequential text presentation (ESP). ESP is a variant of RSVP in which the reader elicits presentation of single words via a button press, allowing the user to control presentation rate on a word-by-word basis. Putting the presentation of each word literally into the hands of the reader allows her or him to allocate more or less time to each individual word, depending on how difficult that word is to read.

Initially, they compared maximum reading speeds using the three methods (RSVP, ESP, and CCTV) in 15 subjects with low vision who were customary CCTV users but previously unexposed to RSVP and ESP. Text was fifth grade level and presented using the same font (TrueType Arial) in all conditions at the subject's preferred font size or magnification. Reading speed for RSVP was determined by increasing or decreasing speed in coarse (20%) or fine (10%) steps until errors were made, while for ESP and CCTV, subjects were instructed simply to read as fast as possible without error. While ESP and CCTV yielded essentially the same speeds (78-79 wpm, geometric averages), reading was on average 31-32% faster with them than with RSVP. Since all subjects were highly practiced on CCTV reading, these results probably underestimate the true benefit of ESP.

Subsequently, they tested the role of two other manipulations on ESP reading: presence or absence of a fixation cross, and justification mode within the text field (i.e. centered or left justified). Both of these manipulations were designed to make ESP reading even faster; however, neither had any measurable impact.

Another way to increase the efficiency of low vision reading is by optimizing usage of the intact retinal area. In the traditional RSVP method, the duration of each display cycle is determined by the length of the longest word in the text, which in turn is determined by the size of each subject's intact retinal area. Thus, when a small word is displayed alone in a display cycle, both retinal resources and time are wasted. Dr. Liu has devised a computer program that analyzes text and divides it into chunks of equal size. Each chunk is displayed in a single RSVP cycle, and each chunk may contain a single long word or two or three short words. A study is being conducted to determine whether low vision readers can benefit from this display method. Preliminary data was presented at Vision 2002, the 7th International Conference on Low Vision (Liu, 2002).

A method of presentation that combines the virtues of ESP and chunk RSVP has also been incorporated into a prototype web browser (Arditi, 2003; also see abstract entitled "Studies in Web Accessibility").

Further Reading

Arditi, A. (1999). Elicited sequential presentation for low vision reading. Vision Research, 39, 4412- 4418.

Arditi, A., Rosenthal, B., Higgins, K. E., & Liu, L. (1998). Low vision reading is faster for elicited sequential presentation than RSVP. Investigative Ophthalmology and Visual Science, 39 (Suppl.), S175.

Arditi, A. (2003). Low Vision Web Browsing and Allocation of Screen Space Resources. Investigative Ophthalmology and Visual Science, 44, E-Abstract 2767.

Liu, L. (2002). A more efficient way to present text. Poster presented at the Vision 2002, the International Conference on Low Vision.



Like us on Facebook   Follow us on Twitter   Watch us on Youtube   Email   Visit our Flickr