The aim of this investigation was to analyze dual-task interference in
the so called Psychological Refractory Period (PRP) paradigm. In this
paradigm subjects have to carry out two choice reaction tasks that ov
erlap in time. A well known result is that reaction time on task 2 (Rt
2) increases with decreasing overlap of the two tasks. Thereby reactio
n time on task 1 is described to be independent of the size of overlap
(Rt1). Usually, this result is explained by the assumption of a PRP w
hich arises in processing of task 2, when serial processing is ongoing
in both tasks. It was asked, 1.) whether the PRP is located before or
after response selection in the first task and 2.) how the second tas
k influences first task processing. In the experiment subjects had to
carry out two choice reaction tasks together. In different conditions
the difficulty of response selection in task 2 was systematically incr
eased by varying the number of response alternatives (0, 1, 2, 3). Dif
ficulty of response selection in task 1 was held constant. Overlap bet
ween both tasks was varied. This experimental design allows different
hypotheses about the sources of interference in both tasks to be exami
ned by use of Schweickert's Critical Path Technique (Schweickert, 1983
). Contrary to the results of Karlin and Kestenbaum (1968) the effects
of number of response alternatives and size of overlap on Rt2 indicat
e a localization of the PRP before response selection. The results sup
port models which assume a serial processing in response selection (We
lford, 1952). They are contrary to models of parallel processing in th
is stage (Keele, 1973). The influence of the number of alternatives in
task 2 on Rt1 can be explained by a mechanism of grouping both motor
responses.