Abstract
Consider a collection of data in the form of success ratios, exhibited for an array of combinations of two or more experimental variables. The simplest effect of changing the value of a single variable, and the most natural one to look for in many situations, is a change in the same direction of all success probabilities. A statistic can be constructed that is sensitive to such consistent changes, while having a X2 distribution under the null hypothesis. While the statistic assumes its simplest form if the number of trials on which the success ratios are based is the same for all combinations of variables, it may be adapted to more general cases.