Determination of critical observed frequencies in chi square

Abstract

The chi-square goodness-of-fit test is viewed as restrictive in interpretation potentials. The concept, rationale, formulas, and limitations for a simple test of observed frequencies are offered. This test, respecting limitations, allows the determination of a critical observed frequency value for any cell that would yield a significant deviation from expectations. The test is statistically conservative and allows category- or condition-specific interpretation of chi-square data. © 1981, The Psychonomic Society, Inc.. All rights reserved.