The influence of feedback and shifts in the probability of faulty objects on quality inspection

Abstract

The aim of the present investigation was to prove in which way the decisions of quality inspectors are influenced by a change in the probability of faulty products, the kind and regularity of feedback. 72 participants took part in the experiment. The proportion of faulty objects was manipulated in three ways. The rate of a faulty object constantly amounted to 14% or increased from 11 to 17% or decreased from 17 to 11%. During task execution there was either no feedback or information about the number of missed faulty objects or about the number of missed faulty objects as well as the number of false alarms. Feedback was given either constantly after inspection of 20 objects or after a varying number of objects which however amounted to 20 on average. The so-called c-index was calculated describing the response bias. The experimental results demonstrate that the decision behavior is changing in dependence on shifts of the number of faulty objects when feedback takes place at irregular intervals and when the number of missed faulty objects is given as feedback. The results are discussed with respect to possible work design measures adequate for the stabilization of the decision behavior.