Delays but not mistakes in dual-task performance: evidence for a sequential integrator

JUAN E KAMIENKOWSKI; MARIANO SIGMAN

Mar del Plata, Argentina

Congreso; MEDYFINOL ‘06; 2006

When two tasks are presented simultaneously or at a short stimulus onset asynchrony (SOA), a systematic delay in the execution of the second task is observed while response times (RT) to the first task are unaffected. This phenomenon, referred as Psychological Refractory Period (PRP), has been explained based on the dependence of RTs with different experimental variables. On the other hand, models of the decision-making process within each single task have relied on a simultaneous analysis of RTs and error rates. Indeed, understanding the covariance of these two variables has been fundamental to understand the stochastic mechanisms leading to decision making. Here we set up to understand this covariation in a dual-task scenario. Subjects performed in rapid succession a non-symbolic number comparison and an auditory discrimination task. Error rates in both tasks could be manipulated experimentally by varying the evidence (correspondingly numerical or frequency distance) yielding performances going from chance to almost perfect. We observe that, while SOA has a strong effect on RT, as has been described previously, dual-task interference does not affect the error rates, nor the distribution of errors as a function of the distance. This indicates that the decision process and the integration time to reach the decision are unchanged when the two tasks are presented simultaneously providing evidence that central processing stages are dedicated sequentially to each task.