To introduce a rank-based statistical test that evaluates differences across experimental treatments without requiring the assumption of a normal distribution.
Converted continuous measurement data within randomized blocks or repeated measures into ordinal ranks.
Derived a test statistic based on the sums of ranks across treatment conditions.
Calculated the sampling distribution of the statistic under the null hypothesis and established its asymptotic approximation using the chi-square distribution.
Established a distribution-free test statistic that remains valid regardless of the underlying population distribution.
Demonstrated that using ranks reduces sensitivity to extreme outliers and departures from normality compared to standard analysis of variance.
Abstract
(1937). The Use of Ranks to Avoid the Assumption of Normality Implicit in the Analysis of Variance. Journal of the American Statistical Association: Vol. 32, No. 200, pp. 675-701.