Which of the following best defines the Tukey post hoc test?

A regression-based method that uses covariates to adjust group means and then runs independent t-tests on the adjusted means without altering significance thresholds.

A nonparametric ranking method intended for ordinal outcomes that compares medians pairwise without reference to any ANOVA framework and therefore does not require an omnibus test.

A follow-up procedure to ANOVA that performs all pairwise comparisons while controlling the familywise Type I error rate using the studentized range.

A procedure that tests only a single contrast chosen before data collection and ignores multiple comparisons to keep calculation simple and identical to a two-sample t-test approach.