Abstract
A new model is presented to account for the reaction time fluctuations in concentration tests. The model is a natural generalization of an earlier model, the so-called Poisson-Erlang model, published by Pieters & van der Ven (1982). First, a description is given of the type of tasks for which the model has been developed. Next, the new model, called the Poisson inhibition model, is described. Each reaction time is considered as series of alternating processing times and distraction times. During processing, the transition rate from work to distraction is assumed to be constant. Therefore, the number of distractions has a Poisson distribution. During distraction, the transition rate from rest to work is inversely related to the level of inhibition. The model is a limiting case of a model in which inhibition is assumed to oscillate between a lower and an upper limit. The present model is described in such a way that computer simulations of the reaction times can be made. Furthermore, the moments of the reaction times are derived. At the end of the paper it is shown that a description of the actual time series in terms of the underlying inhibition process is in complete agreement with Spearman's theory about the universal factors, which are the general factor and the factors oscillation and perseveration.
