Scheffe's test is one of the best adjustments that can used to decrease experimentwise error rates when testing multiple comparisons. Scheffe's test is a very conservative adjustment that some believe is the "safest" method. The F-ratio used in the calculation is unique in that the Mean Square (MS) for only the two groups being compared is used in the numerator and the MS for all respective comparisons is used in the denominator. This means that each pairwise comparison has to have the same significance as the variance for all comparisons when using Scheffe's test.
Despite the strength of this adjustment for multiple comparisons, one way to still achieve significant results with Scheffe's test is to collect larger sample sizes. Larger sample sizes have the added benefit of increased statistical power and more precision and accuracy in measurement. Researchers just need more evidence to meet the stringent demands of Scheffe's tests.
Despite the strength of this adjustment for multiple comparisons, one way to still achieve significant results with Scheffe's test is to collect larger sample sizes. Larger sample sizes have the added benefit of increased statistical power and more precision and accuracy in measurement. Researchers just need more evidence to meet the stringent demands of Scheffe's tests.