"Test the hypothesis that there is a significant difference among the treatments using the usual Type I error.
If there is a significant difference, where does it occur? Adjust your Type I error rate accordingly to test these post hoc hypotheses." Guess