Quantitative evaluation of saccadic and smooth pursuit eye movements. Is it reliable?

PURPOSE The authors evaluated the reliability of the coefficients of the (1) amplitude/duration and (2) amplitude/peak velocity relationships of the mean precision values and the mean latency values (saccadic eye movements) and the coefficients of the target velocity/gain relationship (smooth pursuit eye movements). They computed test-retest maximum variability limits for these parameters. METHODS After a 1-week interval, saccadic and smooth pursuit eye movements were recorded twice from 20 healthy subjects; 12 of these subjects underwent a third recording session. The estimate of the intraclass coefficient of reliability, R, was adopted to evaluate the reliability of eye movement quantitative analysis. RESULTS The data demonstrated that the reliability was fairly good for the amplitude/peak velocity relationship, was good for the precision, and was excellent for the amplitude/duration, the target velocity/gain relationships, and the latency. CONCLUSIONS Quantitative analysis of both saccadic and smooth pursuit eye movements is reliable. One statistic used to estimate reliability, ie, the within-subjects mean square value, also enables the determination of test-retest normal variability values for both the variances and the differences of measurements.