Temporal-signal detection and individual differences in timing.

Abstract

Signal-detection procedures were used in three experiments to examine sensitivity and bias in time judgments and to evaluate individual differences in timing. The task required subjects to judge whether visual stimuli were presented for a certain target duration (the 'signal') or for a slightly longer duration. In experiment 1, subjects performed versions of the task involving both short (2 s) and long (12 s) target stimuli. Analyses of sensitivity and bias measures (d' and beta) provided evidence for consistency in timing performance within individuals. In experiment 2, subjects were tested on a detection task with 5, 10, or 15 s targets, followed by a temporal-reproduction task involving stimulus durations ranging from 3 to 17 s. Subjects with high temporal sensitivity showed less error in their reproductions than subjects with low temporal sensitivity. In experiment 3, subjects were pretested on a detection task with a 12 s target and then performed a temporal-production task where they attempted to generate a series of 12 s intervals under either control or informational feedback conditions. Feedback improved accuracy and reduced variability in temporal productions. However, the low-temporal-sensitivity subjects were more variable in their responses under both conditions than were the high-sensitivity subjects. The results point to the utility of a temporal-signal-detection task both as a means for studying individual differences in timing and as a pretesting technique for assigning subjects to high-sensitivity and low-sensitivity groups to reduce error in time-judgment data.