Long-term retention of temporal information.

The main purpose was to determine the characteristics of long-term retention of temporal information. Visual durations of 1, 4, and 8 sec. were estimated by 120 subjects under the method of reproduction. Four retention intervals were used, viz., immediate reproduction, 2 days, 14 days, and 28 days. The percentage absolute and percentage variable errors were used to evaluate effects of forgetting. When subjects hold time lengths of 1, 4, and 8 sec. in memory for a period of 14 or 28 days, they become less accurate and more variable than if they recall the item immediately or after 2 days. One explanation for the nature of forgetting was suggested. The percentage constant error was used as an index of bias. Subjects had a tendency to overestimate the 1-sec. and to underestimate the 4- and 8-sec. time durations. The hypothesis that the longer-term memory of perceptual estimates of temporal information follows a pattern similar to that of other continuous dimensions was not confirmed over these intervals.