Statistics 101: ANOVA Post Hoc Test (Fisher's LSD)

Рет қаралды 50,235

Күн бұрын

In this Statistics 101 video, we explore the basics of post hoc analysis in ANOVA. While this video uses Fisher's LSD as an example, there are others such as Tukey HSD, Scheffe's Method, and Bonferroni that have similar aims. Post hoc analysis can be controversial but this is simply an introduction to the topic using an example problem. Enjoy!
My playlist table of contents, Video Companion Guide PDF documents, and file downloads can be found on my website: www.bcfoltz.com
#statistics #regression #machinelearning #datascience

Пікірлер: 38

@michaelgeorgemcdonald5927 3 жыл бұрын

Can't wait to use ANOVA and Test on some data I'm tracking. Thanks for the videos. They're great! Very well presented and concise. Well done to you and thanks again!

@scottsara123 4 жыл бұрын

I love your video, whenever i need clarity about statistical inference i follow you. Thanks.

@eldrigeampong8573 4 жыл бұрын

You are the best Brandon. Your videos are excellent and you make statistics very easy to grasp. Keep it up

@JiancarloUlloa 4 жыл бұрын

Thank you for this video, and to present the information so clearly!!

@VenkataSubramaniRamachandra 4 жыл бұрын

As usual a very neat and clear presentation. Thanks a lot.

@juliesampson162 4 жыл бұрын

thanks for the aside about the coasters. the wooden ones have an extra "scare" factor. giant dipper in santa cruz is one of those, but the King is over a mile?!? I love it. Thanks for these videos. you rock.

@teresarogers2353 4 жыл бұрын

Thanks! You make this easier to understand

@Muuip 4 жыл бұрын

Great presentation, very helpful. Much appreciated.

@aiikaiik 3 жыл бұрын

Incredibly Informative!!! Thanks Brandon, I subscribed.

@faraesthetics7161 3 жыл бұрын

I love your video. Really help me in my test! Thank you

@Mina-xo6ii 4 жыл бұрын

I love your videos, they helped me alott! will you please make one about PLS (partial least squares regression)

@ChalaMosisa1bestsong 4 жыл бұрын

your videos helps a lot

@luciehronova1584 4 жыл бұрын

Thank you very much for your videos! :)

@ahmadabul4883 4 жыл бұрын

I highly recommend watching this tutorial

@sylviaemejuru1956 3 жыл бұрын

Thank you very much. You are the best!

@vaishnavisingh7175 Жыл бұрын

Thanks Brandon!

@ElimelchRT 3 жыл бұрын

very good explained!

@alejandrocrates 4 жыл бұрын

thanks ever so much man

@joyousmomentscollection 4 жыл бұрын

I love King's Island.. :) I have been there too.. 45 min drive from where I used to live .. lol

@michaelodonnell7844 4 жыл бұрын

A brilliant video like all your other videos. Thanks very much. One question. In this video, you compared the groups pairwise to identify which differences between the groups was significant. But what about the Type 1 error multiplicative effect for 3 groups. Why is this not a factor here?

@liranzaidman1610 3 жыл бұрын

I love these videos

@tracykakyoalexis2155 2 жыл бұрын

Thank you 😊

@heshammohamed2524 4 жыл бұрын

Excellent work..... please Foltz what can I do if my data is non-normally distributed? what is the best method can I use instead of two-way ANOVA.

@jemiyardenglem6160 2 жыл бұрын

Thanks

@MS-xb8pd 4 жыл бұрын

8:51 when I put the same table in excel and run one way ANOVA: Single Factor, it gave me different results. I am facing problems with these statistical techniques when I perform them in excel. Please help me

@tonmoydas3520 4 жыл бұрын

Sir, please make tutorials for other post hoc analysis methods.

@BobTheZealot 4 жыл бұрын

9:18 "The alternative hypothesis that they are all not equal" should be "The alternative hypothesis that they are not all equal"

@brunog.campos3236 3 жыл бұрын

4:24 what if the 3º value is not relevant for the comparison between 1 and 2, in addition to be completely independent from the 2º value? And the variance of these 3º value is masking the difference between 1 and 2?

@Magmatic91 3 жыл бұрын

I find it hard to understand the part with error compounding. I mean, I know that in probability if two events are independent and have let's say a probability of 0.05 of happening (In our case Type One Error). Then if we want to repeat the experiment let's say 3 times, then we set P(Error) = 0.05 and then we have 0.05 ^3 = 0.0125%. So at the end we decreased our P(E) by repeating the experiment 3 times. I'm I getting something wrong here, can your please help me understand this concept? Thanks.

@karannchew2534 3 жыл бұрын

"Fisher’s LSD, which is the F test, followed by ordinary t-tests among all pairs of means, but only if the F-test rejects the null hypothesis. The F-test provides the overall protection against rejecting H0 when it is true."

@selousscout9664 3 жыл бұрын

please explain when the 0.5 significance level is compounded, and when not. I see you doing t tests comparing each one individually in this video. Please help