The visual evoked responses: reliability revisited.

To examine the reliability of VER data, responses were obtained from 9 normal subjects. For steady state and transient displays, 7 consecutive VERs were obtained using 10, 25, 50, 75 and 100 ensemble averaged 384 ms activity samples each. Data from the 7 VERs were averaged by considering the first 3, the first 5, and all 7 VERs. For each display, maximum reliability was obtained by averaging at least 5 VERs with each consisting of 50 or more ensemble averaged samples. The steady state display produced more reliable data than did the transient display, but when each display system was used in a refraction format, the transient display produced refractions with narrower uncertainty ranges.