This post hoc test can be used to determine the significant differences between group means in an analysis of variance setting. The
Unequal N HSD test is a modification of the
Tukey HSD test, and it provides a reasonable test of differences in group means if group n's are not too discrepant (for a detailed discussion of different post hoc tests, see Winer, 1985, pp.140-197). For more details, see the
General Linear Models chapter. See also,
Post Hoc Comparisons . For a discussion of statistical significance, see
Elementary Concepts .