Abstract
Purpose:
To determine if differences in legibility between the ten letters of the British standard (BS) set contribute to variability in visual acuity (VA).
Methods:
Intra-session test-retest repeatability of best-corrected VA was assessed in 68 visually-normal young adults under bright photopic conditions. Threshold VA was determined with custom Bailey-Lovie charts (five letters per line) in two conditions in random order: (1) Non-Paired Condition. Presentation of two letter charts having non-matched letter legibility across lines. The five letters on each line in a single chart were chosen randomly without replacement from the BS set; and (2) Paired Condition. Presentation of paired letter charts having matched letter legibility across lines. In these chart pairs, each of the ten BS letters appeared exactly once for a given line size across the two charts of a pair.
Results:
The 95% limits of agreement for test-retest repeatability were ±0.053 log MAR for the Non-Paired condition, and ±0.075 log MAR for the Paired condition. Repeatability was not significantly better in the Paired condition (Brown-Forsythe test, one-tailed, p = .97). When three outlying points were removed (> ±3 s.d.), the finding did not change substantially (n = 65, p = .82).
Conclusions:
When visual acuity is based on the average of two chart presentations (each with five letters per line), differences in letter legibility in the British Standard set do not have a significant effect on the repeatability of visual acuity.
Keywords: 754 visual acuity