Sequential response bias in randomized response sequences: A computer simulation

Abstract

In this study sequential bias in randomized response sequences is simulated by computer. The computer model assumes that response bias is composed of two components. The first is a repetition avoidance tendency, which is simulated by a memory search through the seven most recently produced responses, and the second is a repetition tendency for the two most recently produced responses. This repetition tendency is assumed to be related to the better availability of recently produced responses. Both processes together give a reasonable approximation of human subjects' data. The results can be explained by the assumption that subjects try to control perseveration tendencies by using a comparison-based memory search mechanism.