The influences of task repetition, napping, time of day, and instruction on the Sustained Attention to Response Task.

INTRODUCTION: The Sustained Attention to Response Task (SART) helps to quantify vigilance impairments.Previous studies, in which five SART sessions on one day were administered, demonstrated worse performance during the first session than during the others. The present study comprises two experiments to identify a cause of this phenomenon.
METHOD: Experiment 1, counting eighty healthy participants, assessed effects of repetition,napping, and time of day on SART performance through a between-groups design. The SART was performed twice in the morning or twice in the afternoon; half of the participants took a 20-minute nap before the second SART. A strong correlation between error count and reaction time (RT) suggested effects of test instruction. Participants gave equal weight to speed and accuracy in Experiment 1; therefore, results of 20 participants were compared to those of 20 additional participants who were told to prefer accuracy (Experiment 2).
RESULTS: The average SART error count in Experiment 1 was 10.1; the median RT was 280 ms. Neither repetition nor napping influenced error count or RT. Time of day did not influence error count, but RT was significantly longer for morning than for afternoon SARTs. The additional participants in Experiment 2 had a 49% lower error count and a 14% higher RT than the participants in Experiment 1. Error counts reduced by 50% from the first to the second session of Experiment 2, irrespective of napping or time of day.
CONCLUSIONS: Preferring accuracy over speed was associated with a significantly lower error count. The data suggest that a worse performance in the first SART session only occurs when instructing participants to prefer accuracy, which is caused by repetition, not by napping or time of day. NOTE: We advise that participants are instructed to prefer accuracy over speed when performing the SART and that a full practice session is included.