Follow Up ANOVA - Tukey’s HSD Post Hoc Test (12-5)

[@dimasaad1521]

You’re a lifesaver, you explained it so much better than the slides. Awesome work!

[@AlexandriaRodgers-p8f]

I love this video. Witty, humourous, and the perfect explanation.👍

[@ResearchByDesign]

Thank you so much for the kind words. Glad that the video was helpful!

[@FruitBasketyay]

Thank you for the clear and concise explanation. Loved the wizard of Oz analogy

[@ResearchByDesign]

You're very welcome. I like those fun examples like Wizard of Oz for the lectures and then I can customize my classroom examples for whatever field the students are learning. Thanks for watching!

[@wreckelite]

thank you so much for this! My professor was doing it in excel and was going too fast for me to keep up with her. Your ANOVA and Tukey videos have been a life (and grade) saver for me this semester.

[@ResearchByDesign]

Wonderful, glad to hear that it was helpful. Thanks for the comment

[@larrymcjones]

Man I wish we were able to use excel in our stat class

[@tlhomotsemoteme2423]

Thank you so much - You've explained the test quiet well. We thank you so much.

[@mr.e.kelleher4659]

This was honestly good, clear and concise! Thank you :)

[@tonyerickson1702]

So helpful, thank you!!!

[@gebreyesgurmu2733]

Excellent explanation. Would you please show us how to calculate Tukey's HSD in Minitab?

[@isabellesmith963]

How do you find the actual P-value for each pair? not just whether its significant or not

[@kburgess47]

What is the hsd formula when doing a repeated measures anova design?

[@NarwedH]

Good, coherent explanation. Thanks!

[@bridgettsmith7206]

Thanks for the video

[@nLPondiong]

I have a question, where did you get the listed means? M1=1 , M2= 1.40 , M3=3.60 , M4= 4.20 , please do explain where did you got that , I'll be reporting this topic on Friday in our classroom.

[@IsisMarquesDesign]

Thank you so much for the help! How it changes if the number of subjects in each group is different? It will artificially change the differences in the means? Or the formula automatically accounts for differences in the number of subjects in each condition?

[@ResearchByDesign]

When doing Tukey's by hand, you must have equal numbers in each group. However, in reality, that rarely happens, so SPSS has a correction to adjust for unequal group size in their formula. Not sure, but it may actually be a different test, but the SPSS output will still say "Tukey"

[@woodchuk1]

Research By Design I believe that the test is called the Tukey-Kramer test when modified to incorporate unequal sample sizes.

[@TheLifeHero]

Just a question. So they Tukey doens't classify as significant differences between those that are very concentrated around their means and thus doesn't find effects that cause small changes?

[@m.bouanane4455]

Can we use alpha=.05 for pairwise comparisons or should we use alpha=.05/#group and why? Thx.

[@samirkhan6195]

as number of comparisons increases , Family wise Error rate also increases , you will have to refine your Alpha in order to control Family wise error rate , However Tukey's test does this by taking account of k ( number of treatment levels ) during calculation of Critical score , as you can see in Tukey's q table , larger the number of levels , larger the Critical score the table will yield , larger the Test statistics should be in order to Reject the Null Hypothesis ....

[@sortileges5474]

I am confusion. ANOVA shows that my treatments have a significant difference. But this test says none of them are significantly different to each other. Is this acceptable?

[@samirkhan6195]

One reason could be that Variance of your groups are not similar , try testing with Leven's test or Bartlett's test or maybe Fligner's test for equality of Variance ... If thats not the case then you will have to take consideration that Tukey's test is more conservative even as compared to ANOVA , so your results are expected sometimes , However if i were you i would be more relying on Tukey's results ...

[@takudzwamandizvo8693]

Well explained

[@briannacac]

Is a Tukey HSD necessary if ANOVA proves there is no significance?? Thank you so much!!

[@ResearchByDesign]

The Tukey HSD is a follow-up test used when the ANOVA shows that significant differences exist somewhere in the model. The HSD (post-hoc test) shows you specifically which groups are truly different from others. Therefore, if the ANOVA is ns, then there are no significant differences to find with a post-hoc. (i.e. no, HSD is not necessary if the ANOVA is non-significant)

[@briannacac]

@@ResearchByDesign Thank you so much! that is what I thought! Thank you for responding so quickly!!

[@elizabethryan5464]

Where did you get the 10.49?

[@ResearchByDesign]

The 10.49 is the F value from the ANOVA...it is in the ANOVA summary table in your SPSS output or the last step (MSb/MSw) if you are working by hand in ANOVA

[@linel6262]

Where the question..i dont get it how u get 4 as k treatment value

[@ResearchByDesign]

The k refers to the number of groups. In this case, there are 4 regions, so k = 4

[@Ladieejass]

I’m confused on how you got 5 for nk

[@dafunktrunk]

It's the number of samples, or the sample size per group

[@dafunktrunk]

For instance, if he has 5 different numbers of wicked witch complaints for each region, then the sample size would be 5

[@samibekele9330]

thanks!

[@ResearchByDesign]

You bet!

[@nickkyivy67]

I'm so excited. I finally got rid ofl am so glad that I've finally been confirmed negative from this virus After so much misleading advice from people stating that this virus does not have a cure, this made me live a long time with this virus Herpes taking antiviral medications to manage it whenever I have an outbreak. I really thought there is no cure when I went to my doctor during an outbreak and he said I can only use medications to control and treat the sore blisters of outbreak's but this was proved wrong with [e-mail] or whatsApp mobile number [phone]’’’'

[@mishley2816]

hєllσ,cαn чσu tєll mє hσw tσ αвσut α pσѕt hσc tєѕt whєn єαch grσup hαѕ α díffєrєnt numвєr σf σвѕєrvαtíσnѕ,plєαѕє.????

[@vue597]

Tukey rau Lub ntuj rau txom tau tukey kiag no vim turkey npam they txwv tsis pub ob Lub teb chaws yuav los nkag rau nato turkey yog tus nom thuiskijkery yog tus tsis lub los nkag nato nyuam qhuav ntxim ntuj tig ko dua rau Ukraine zaum no lauj yawg Putin yuav uas tsuaj tas zaum no lauj ok zoo siab rau Lub ntuj rau txim rau tukey raug npam nawb zaum no os

[@syamimizainal736]

💀💀💀 im still stuck

[@samibekele9330]

wow

[@serdardursun3038]

Fear the turks 🇹🇷🇹🇷🇹🇷🇹🇷😎😎😎😎

[@jalapenoandbanana]

🦃

[@nLPondiong]

I have a question, where did you get the listed means? M1=1 , M2= 1.40 , M3=3.60 , M4= 4.20 , please do explain where did you got that , I'll be reporting this topic on Friday in our classroom.