Bonferroni Correction and Holm Method (Post Hoc Tests) after one-way ANOVA in Excel

[@TheJazzMan333]

Excellent video! I enjoyed the visualization, the color coding, and the pacing. Thanks for creating this content!

[@DrGrande]

Thank you!

[@miguelsuarez475]

Hello Prof. Grande. I think it is better to say the adjusted alpha would 0.05/number of comparison rather than the numbers of levels. Because if you have 4 levels you would have 6 comparisons, and the p-value should be divided by 6. Am I correct? Now I am a little confused since journal reviewers are demanding Bonferroni for everything. including in Multivariable logistic regression, (alpha divided by the number of explanatory variables) Which from my perspective it does not make sense. Have you heard something about it? Thank you very much

[@artgarr8664]

I agree with you Miguel

[@donaldeaton5556]

Great video! Dr. Grande, for the Holm Method formula, should we subtract "rank+1" from the number of levels?... or number of total comparisons? I tried it with an 8 level factor, but 28 total comparisons, and ended up with negative P values in the Holm column (for those post tests with ranks greater than 8). In your example the number of levels and number of comparisons were the same, so not a problem. Thanks much, Don

[@hagegco]

Many thanks Dr. Todd

[@DrGrande]

You're welcome!

[@lappaz100]

Great video! But i still do not understand which is the best method? We only say there is a difference between them? Ex. say i have 4 methods, A -> D. i calculate an ANOVA and find that there is a ssd and then use the Bonferoni - holm and T - test to do 6 more test. I then come up with 2 cases that are significant. Ex A-B and B-C but which is considered the "best" method?

[@samkab677]

Hi..between TAU and REBT ..which is best? How to know? What test to use? Or mean and Stdv are enough to use?

[@AnasaktSadhak]

Hello Professor Granade, Thank you for sharing this informative video. It is really helpful. However, I have a questions mainly abur Holm-Bonferroni (HB) corrections regarding my research statistics. as following: 1) Is this HB correction is applied over only on t-test at post-hoc analysis level? 2) If no, how can/should I apply this HB correction over my mixed way ANOVA analysis, considering main effects of each factor and interaction among several factors? 3) Finally the most important question for me, how to report this HB corrected values in my manuscript/paper? Should I report the original/undjusted p values or this HB correccted p values for above mentioned variables? I look forward to hearing from you. Thank you. With Regards Satish

[@whereIstood0705]

Great video! You make anova on Excel very easy! I do have one question.... I ran bonferroni on excel and spss and got different results for the same dataset ... what am I doing wrong?

[@olameekkvalvaag3269]

How do you get Adjusted p-values form holm?