Nonparametric multiple-comparison methods for simply ordered medians

Nonparametric tests for ordered quantiles

“…Note that the proposed test statistics use estimator of pooled standard deviation in the denominator. This is on similar lines as followed by Nashimoto and Wright (2007) in their test statistic for testing for ordering of medians. We reject H 0 in favour of the alternative for large values of T E (u) and T K (u).…”

“…Remark The critical points given by Hayter and Liu (1996) can be used even if the variances (quantile densities) are not equal,. The statistics proposed in this case are analogs of the statistic given by Nashimoto and Wright (2007), where the denominator is an estimator of pooled standard deviation of ranks. They also showed that such type of statistics are more powerful for right skewed data.…”

“…Nonparametric comparison procedures of several populations for their scale parameters have been discussed by Stevens (1989). Nashimoto and Wright (2007), suggested a rank-based method to test the ordering of medians for non normal data. Nashimoto et al (2013) suggested a new two-stage method of multiple comparison of proportions.…”

Testing for order among K populations: theory and examples

“…Correcting for multiple comparisons can be done using any procedure, for example, Holm—Bonferroni. Preliminary simulation results indicate that this approach is more powerful than the approach described in Hayter (1990) and Nashimoto & Wright (2007).…”