Association analysis searches data for potentially useful patterns in the form of rules. The aim of this paper is to analyze thoroughly the influence of imprecision or uncertainty in data on the uncertainty of the rule’s quality measure, which is called the confidence. An experiment is conducted that allows to analyze the amount of uncertainty propagated from data to the resulting rule’s confidence. For that, artificial data are generated with various characteristics and uncertainty settings. Results of the experiment show the dependency of the amount of uncertainty in a rule’s confidence on the rate, type and position of uncertainty in data. Also,theeffectofat-normselectedinthedefinitionofconfidenceisevaluated.