n. in a contingency table, the situation in which only a subset of entries or frequencies are independent or uninfluenced by one another. Entries may not be independent for a variety of reasons: They may be invalid, missing, or not counted in the analysis.