DISTRIBUTION INEQUALITIES FOR PARALLEL MODELS WITH UNLIMITED CAPACITY

Inequalities on reaction time distribution functions for n-channel parallel models with an unlimited capacity assumption are presented, extending previous work on serial models. First, the first-terminating and the exhaustive cases are considered without any assumption on the distribution functions. Later, certain general stochastic dependence properties are introduced, and appropriate distribution inequalities are derived and illustrated by specific parametric examples. Moreover, the inequalities are extended to finite mixture distributions and to censoring. (C) 1994 Academic Press, Inc.