Evaluation of single pixel step sizes in visual acuity assessment

Area of Science:

• Ophthalmology
• Vision Science
• Psychophysics

Background:

• Adaptive thresholding procedures, such as the Z-algorithm (ZEST), improve the precision of visual acuity testing by focusing on estimates near the current threshold.
• Smaller increments in letter size may enhance the accuracy of these adaptive techniques.

Purpose of the Study:

• To evaluate the validity of visual acuity assessment when letter sizes are not restricted to multiples of 5 pixels.
• To determine if non-standard pixel increments affect the precision of adaptive thresholding procedures.

Main Methods:

• Snellen letters were presented with sizes varying by ±5 pixels around the estimated threshold.
• Letter sizes that were multiples of 5 pixels were presented 160 times, while other sizes were presented 40 times.
• Post hoc ZEST procedures and simulations were used to compute and validate acuity threshold estimates.

Main Results:

• No significant differences in acuity thresholds were observed between letters rendered in multiples of 5 pixels and those without this constraint.
• Empirical data from subjects and simulations fell within established bounds, indicating reliable results for both conditions.
• No idiosyncratic results were found for either set of letter sizes tested.

Conclusions:

• Visual acuity scores derived from letter sizes not conforming to a 5-pixel multiple are valid.
• This finding supports the use of adaptive psychometric procedures like ZEST and QUEST, allowing for optimal letter size selection near the estimated threshold.