A test of variance homogeneity is given in Section 4.13. If the hypothesis of homogeneity is not rejected, use the pooled variance estimate s2 in (4.34) for estimating the underlying variance 𝜎2. If...

A test of variance homogeneity is given in Section 4.13. If the hypothesis

of homogeneity is not rejected, use the pooled variance estimate s2 in (4.34)

for estimating the underlying variance 𝜎2. If a reliable estimate of the variance is available either from other sources or from (4.34), use the studentized

maximum modulus test in Section 4.14 to test effect significance for ȳ. Its

experimentwise error is controlled at the nominal level 𝛼.