Statistics
If the software does not provide pp-values for Tukey, report only the ANOVA FF statistic and conclude which pairs differ based on visual inspection of means without performing any standardized comparisons.
Use the ANOVA MSE to compute pp-values directly for each pair by dividing the mean differences by MSE and looking up FF-table values; skip any reference to the studentized range sincepp-values are always available.
Compute pairwise independent tt-tests for all pairs at α=0.05\(\alpha\)=0.05, then apply Tukey's qq to the smallest tt-score only and report that value as the definitive pairwise conclusion.
Extract MSE from ANOVA (within-group variance), compute df=N−kdf = N − k to find the appropriate qq critical value for the chosen alpha and kk, calculate each q=∣meani−meanj∣MSEnq=\(\frac{\left|\operatorname{\mathrm{mean}\)}_{i}-\(\operatorname{\mathrm{mean}\)}_{j}\(\right\)|}{\(\sqrt{\frac{MSE}{n}\)}} (for equal nn), compare each qq to qcritq_{crit}, and report which pairs are significant along with their qq statistics.