For each speed, the motion coherence threshold was established by using a single-interval, two-option, forced-choice procedure.Patients had to indicate the direction of the coherent motion. The signal-to-noise ratio was one to one at the beginning. A correct answer led to a decrease of the signal-to-noise ratio by 1 dB, whereas an incorrect answer increased the signal-to-noise ratio by 3 dB. For each speed, 128 trials were conducted. Subsequently,the proportion of correct responses was calculated. The level at which subjects are expected to judge 75% correctly is defined as the coherence level. To find the coherence level for each speed and direction, a probit analysis was conducted.