Self-timing in Memory and Visual Search Tasks

We investigated the ability of people to time themselves as they perform cognitive tasks. One group did a memory search task, another did a visual search task. After each response, participants estimated the duration of their own reaction time. In both tasks, correlations between reaction times and temporal judgments were significant, showing that people can provide precise quantitative estimates of mental processes when they perform memory and visual search tasks. Increasing load lengthened reaction times and temporal estimates in both tasks. Accuracy of temporal judgments increased across blocks of experimental trials, showing considerable improvement in self-timing with practice although in visual search, improvement was more pronounced under low load conditions. Results demonstrate excellent ability for selftiming of memory and visual search, but suggest that in visual search, self-timing is influenced by stimulus conditions. Corallo, Sackur, Dehaene & Sigman (2008) showed that reaction times in number comparison and in tone discrimination are accessible to deliberate report: people were quite accurate at providing introspective temporal estimates in those tasks, with significant correlations between reaction times and temporal estimates of those reactions times. The present study examines whether processing time of item sets in memory search and visual search tasks is accessible also to introspective duration estimation. The interference effect, one of the most common findings in human timing research (Brown, 1997), suggests that temporal processing should be perturbed when tasks requiring attention or working memory executive resources are carried out concurrently with timing. This leads to general temporal underestimation. However, a previous study showed that although time interval production revealed general temporal underestimation concurrent to both memory search and visual search, higher load increased underestimation in memory search, but not in visual search (Fortin, Rousseau, Bourque, & Kirouac, 1993). In another study, longer visual processing in color comparison led to greater underestimation (Gaudreault, Fortin & Macar, 2010). A recent study showed that visual discrimination led to temporal underestimation in concurrent timing, but only when the markers of the interval to be timed were spatially distant (Cicchini & Morrone, 2009). Overall, nontemporal processing requiring attention affects performance in concurrent timing tasks, but the effect seems to vary in the case of visual processing. In the present study, people estimated their reaction times (RTs) in memory (Exp. 1) and visual (Exp. 2) search. General underestimation was expected, but effects of search load might affect differently temporal estimates (TEs) in the two tasks. To see whether temporal accuracy improved with practice, TEs were examined across blocks of trials in both experiments. Experiment 1: Self-Timing in Memory Search