Introduction to ANOVA

[RainerShines]

Thank you for this video! I am a 4th year psychology student who put off taking the mandatory statistics course until the last possible second! So I sit in class, confused, with a bunch of 2nd year students every week. Online tutorials like this have been probably the only reason I've been getting marks in the 70s and low 80s so far. Good luck to every other math-hating student like me who has to take stats to get their degree!

[shalinajoy]

I thought I was smooth sailing as I came to the end of my PhD program, but then I ran into the STATISTICS brick wall! What a nightmare this has been for me. I am so grateful that I have found your channel. You have such a way of explaining things. I am definitely going to subscribe! I have to be sure to watch more videos before dissertation time :(

[shortwhat]

I am studying a degree and my lecturer for research method only speaks with no visual ques which is not my preferred way of learning. You've literally turned a 1 hour lecture that confused me to hell into a 7min video that I 90% understood. thank you so much.

[amanansari]

anova method to compare the means of two or more groups, with t test we compare to more samples, in anova we have factors and level, like gender is a factor and male and females are levels.

Some assumptions in anova is that the we have a normal distribution for the sample means, outliers are removed from the data set.

Hypothesis in anova - Imagine having one factor with three level then is null hypothesis all the level are equal while alternate hypothesis is like not all factor are equal. when dealing with the effect of one variable is called the main effect.

now two factor with three level then there are 3 sets of hypothesis null hypothesis all the level are equal for each factor while alternate hypothesis is like not all factor are equal for each factor. Then null hypothesis is two factor being similar alternate hypothesis is two factor are dissimilar. Interaction effect. So when you do plot and if the line cross each other then you can conclude that it has interaction effect.

types of anova - one vay anova, one factor with at least two independent levels, repeated measure anova is one factor with at least two dependent levels(like measuring the people for something at day 1, 2, 3), factorial anova is two or more factors each having two level and the level can be dependent or independent or mixed(days 1, 2, 3, 4 in columns and M & F as another level in rows ).

When we reject the null hypothesis we are saying that there is a difference, but where is it. For this after anova is a done we have do post hoc analysis to find where the difference lies.


While doing anova we calculate F statistics. If there is no difference between the factor or variable we expect F to be 1 if there is difference we expect F to be greater than 1. F statistics is always positive. so if F ratio< = 1 null hypothesis is true of f ratio is > 1 then null hypothesis is rejected.

[feebee]

This is “short, but powerful” like we say in Dutch. Thank you for this clear video! 👌🏼

[swaminathans]

Beautifully explained! I am taking my Six Sigma Black Belt Certification Course and just saw some of your lectures. You have a great teaching talent!

[mbbk]

thank you so much for your videos. they are a wonderful addition to my biostats class for my masters in family nurse practitioner. so easy to understand and extremely well done. keep up the great work.

[leewayexpressionsave]

I am so thankful for you sir .. You are so brilliant, I learned a lot.. God Bless you. May have long long life ahead...

[MsBRETTB]

Thank YOU, finally a video that gives a simple explanation.

[MrCastleking]

WOW! I have sat through so many lectures not knowing what on earth my lecturer is talking about.... Thank you so much for your straight forward and simple explanation! Stats isn't that hard after all!

[vinitkatariya]

The pdf or ppt of these lectures will be very helpful.

[saryarohit]

Thanks for posting these videos. They are super informative and clear. They are also always well-organized with a good structure and the content is always sufficient. Examples are useful as well.

[lorikaimadison]

That was so good! The organization of it all! Thank you!

[animalslife]

Hey very good lectures. Can I know what reference books are you using in learning statistics? Thanks.

[--Zaki--]

Hi sir, U lectures are superb and easy to comprehend. But I have a small suggestion that your speed is little bit fast. Sir your lectures are learnt by everyone worldwide so there are different people with different pronunciation. Hope you don't mind

[vipashapmehta]

tysm, it was very good and clear video. i understood the overview of the content well.

[garretthaslam]

Oh my GOD, I had no idea that stats could make sense.

[kylmaz]

Hello. If our hypothesis was "Ho: μA1 = 5", what would our solution be? (That is, judging by its equality to any number.)

[ankushbhawal]

Thank you sir so much....what a video....

[ashutoshpatil]

thanks for the introduction thats what i was in need for