Chi-squared Test for Independence! Extensive video!

[@jacobm7026]

Seriously, man, keep at it. Your viewership is going to blow up pretty soon. One of the best channels ive seen so far on statistics

[@LuisIIIBanez-gb2dn]

Agreed, been using his videos to understand stat

[@isaacliu896]

Coolest stats teacher on the web! So glad I found your channel, thank you.

[@randall.chamberlain]

Mate, you are the only one who ties the knots all along the conceptual backgrounds which bring about the methods blindly used by most of us. Superb. Couldn't help but subscribe and comment here. Keep at it. This channel is pure gold.

[@varunsagartheegala]

This is probably the best explanation I have come across for a chi square test. Very easy and intuitive examples to follow. Bravo. I wish all professors could make understanding stats as easy as you

[@GregThatcher]

Thanks! I've wanted to really understand this ever since college (decades ago), and now I do thanks to you.

[@zedstatistics]

Re-uploaded with corrected formula. Thanks Henna for spotting it!

[@hennahaavisto]

No problem. And luckily just that was changed because the video is great! Liked especially the singing at 18:55 and the little judgment analysis after that.

[@7justfun]

@@hennahaavisto quick Q ...8:46.. shouldn't it be 35/120 * 50/120. 120 squared in the denominator?

[@Competent00]

@@7justfun no. 120 is the frequency and the last 120 is multiplied with the whole term. I.e. it is in the numerator. So it cancels out one of the 120 from denominator

[@lbognini]

@@7justfun Listen at 08:00. He explains how we came up with 35*50/120

[@Danni.A]

When you watch one video and everything you have been studying makes sense!!! Thank you!!!!

[@wamebuilds]

this is the best video I've watched on chi-square.

[@sagnikrana7016]

This channel is underrated.

[@AlfredEssa]

Definitely among the best channels, if not the best channel, on statistics.

[@elittlehumans6217]

Thank you for making these videos. You have taught me for my whole term when the teacher wouldn’t take the time to do it. You are the best!

[@chandramoulisarkar2935]

Loved how you linked Chi-Sq Independence Test effectively boils down to a Goodness of fit test. Great Insights

[@namgaydorji7224]

Loved how you explained degrees of freedom! Thank you very much.

[@cmu1919]

5 mins in and this video has already cleared up 90% of my confusion

[@kaitlynlacy853]

man, I wish I didn't find you during finals week. I wish I had found your channel at the beginning of the semester. Thankyou!

[@bhavanisankarlenka]

One of the ultimate explanations, I have just watched all ur videos on distributions i was amazed at how u explained each of them butter smooth with proper intuition😍😇.

[@elenaguzman9413]

I love! Love! the way you explain things! Thank you so much for your amazing teaching abilities - you definitely have a gift.

[@fuluo3189]

Thanks for the videos. It's very intuitive and explanatory. The best statistic videos I've seen so far. Really like them!

[@monikarozkiewicz8326]

This video is amazing! I understand Chi-squared test for independence now :)

[@annatolochko7633]

Thank you for such clear and interesting explanations on statistics topics! I need stats for my data science job and your videos absolutely make it all clear for me! You are an awesome teacher 🙂

[@waytolegacy]

I literally repeated 4 times your intro to listen to the music... ttttt2222tttttt222222tttttt22222.... go on man!

[@qinghuafeng1705]

Thanks for the clear explanation.

[@stephenday4834]

Amazingly cleanly explained. Great video.

[@sharonwilliams5240]

All I can say is wow and thank you. You have broken down a subject that completely confuses me. My professor's notes do nothing but confuse me even more. Thank you for breaking it down step by stipend explaining formulas. I am someone that needs to know the why!

[@composmentis5493]

I love the thumbnail. The colors used are so beautiful. This video is informative also.

[@ThepExcel]

Really good explanation. Thx a lot

[@srinivasakalyankakumanu7484]

Thanks for the fantastic video! Please make a video on the Weibull distribution as well.

[@hemamalinirath2984]

Excellent explanation. Very much informative.you are an excellent teacher.

[@tensor131]

Another brill video thank you. May I suggest an alternative way to calculate the E-table? If there is no association then the 35 who opted FB would distribute in the same ratio 15:20, as the INST 65 and the TT 20; this only requires elementary school arithmetic to find the E's. You can equally do it the other way: the 50 males would distribute in the same ratio 35:65:20 as the 70 females. Of course it is the same formula but I feel this is a more intuitive way to think about it and I know you like intuition!!

[@TheBlueFluidBreathe]

Thanks man, you have the best explaining!

[@keshavbansal5148]

Amazing video Zed!!! Thank you so much!!

[@Dupamine]

Why is it mindblowing that expected resemble normal distribution ?

[@sandeepbade9743]

Brilliant..... Love from India

[@vothien1768]

I love it, so much easier to understand. thank you

[@unuundrakh6237]

I love you❤️❤️❤️, you actually helped to do my homework😘

[@ronaklathiya4081]

how you will divide this into the sum of squares of two standard normal random variables. please share material regarding that if you have.

[@sawsanyousif1570]

Thank you for explaining better than my doctor 🙏🏻🤍

[@sneha_more]

How does he still not have at least a million subscribers 🤔

[@dharuns7026]

Great explanation. Wanted to clarify an info on the video. In the end you say Chi square test has rejected the null hypothesis which means not enough evidence to suggest that they are dependent. Shouldn't it be " not enough to suggest that they are independent" cause that is the H0.

[@AnilKumar-bq3kd]

Nice explanation. Your videos are really interesting and helpful.

[@harpingalong5420]

Is there any way I can learn how the two formulae for Chi square are related…. I mean summation of z^2 and summation of (o-e)^2/e

[@KevinReid-x1o]

Thank you so much for all of the videos you have created. I have learned so much! I am interested in doing a chi-squared test that looks at the clicks generated by 8 different ad sizes across three devices: desktop, tablet and mobile. Is this possible? This would tell me which ad sizes are working the best (or worst) on which devices, right? I am not sure if I am using the right statistical test. Thank you for any help.

[@kyleevalencia1827]

I wonder about this independence test if we can turn it into dependence (not same as not independence) test by modifying the expected distribution value ?

[@dr.kingschultz]

very good videos! I hope you keep doing it

[@piaprenrose8510]

Your video helps a lot!

[@Pickett1312]

Brilliant! Thank you.

[@daniyaniazics]

Finally understand what is degree of freedom 😍🤩

[@AGABIROSEMARY]

Good explanation and interesting

[@suryahr307]

Initially you have uploaded videos randomly and then in the form of playlists which is nice. Could you create a playlist (from the uploaded videos) in an order which is followed in any university please? I am a new PhD student (in IITG) who didn't register for course work of research analytical methods and now I'm sorry for myself.

[@Mimoborg]

Thank you! It's very easy to understand.

[@samhijazi8841]

Good work, my young friend. Dr. H.

[@zedstatistics]

Oh to still be young... [brews dandelion tea]

[@sinaabedini7487]

Wow you're good; keep up the great work. Thank you so much

[@karannchew2534]

Chi Sq Test of Independence Ho: X and Y are independent. So, if chi sq is low, delta is low, distribution of x (e.g. social media number of user) does not change with y (gender), x and y are independent, reject Ho. Chi Sq is lower than the critical value. So, I can't say gender influence choice of app. But I can't say gender does not influence the choice either. Therefore we can't say Gender and App Preference are dependent. We can't say gender influences app preference. However, 19:06, we also can't conclude that Gender and App Preference are independent. We can't say gender does not influence app preference. 19:16 There's not enough evidence that one is guilty. But it doesn't mean he is proven innocent.

[@nithyajacob9106]

Such a good video...thank you very much

[@Ronaldnyawade]

Nice...quite easy understanding...thanks a lot.

[@aimanali9868]

Greatt explanation. 👏👏 Though i am still confused about some things. 😢 But your videos are way too understandable

[@nkosivumilezhou8374]

I like this guy🔥

[@hemantdas9546]

Hi Justin! Can we not take null hypothesis as ho: social media preference is Not independent of sex. How do you decide which one is h0. Please explain.

[@FitnessAndHealth4You]

Isn’t H1 (alternate hypothesis) that for which we are seeking evidence for? Why are you assigning H0 for “Social media preference ISD independent of sex”? I thought it would be H1. Am I missing something here?

[@keyzaragoza3904]

Hi! What method can I use in testing the relationship between social media and mental health of people? TIA

[@kccchiu]

Loving this series, will you make a gamma distribution?

[@maungmyint2056]

Very useful lessons for me, thanks!

[@abhisekbehera9766]

what if few of the expected frequencies are not grater than 5 (in a very big contingency table, let's say of dimension 15x51 ); in that case is chi-square test still valid ?

[@nwalaobinna7224]

How do someone calculate for the p-value

[@JamesPavlaine]

thats good work man good for you and everyone else i guess

[@jongcheulkim7284]

Thank you. I really appreciate it.

[@andreahategan1061]

Thank you for creating for this video. If you have time, it would be great to add few words about the effect size. For example, how to calculate Cramer's v value and how to interpret the value, i.e. when is the value telling is that is a big effect or a small effect? Are the threshold values change based on the degrees of freedom?

[@EigenA]

Great video and cool background too. Are you an EE?

[@howsUkraine]

Coool just keep posting ull get more views

[@rishavchowdhury8426]

very well explained

[@aymanissa6722]

Can we use Chi-squared when we have more than 2 classifications such as Male Females and Trans? So the degrees of freedom will be (3-1)*(3-1) = 4 Is that right?

[@alexlove8086]

Yes,you can

[@diamondblack581]

I understand it better now thank you, can you show though how to do it excel one day.

[@jaybhanushali5362]

Why we have this rule that "All expected frequencies are > 5" in chi square test ? Edit: I have seend the first video on chi square and the second video on goodness of fit as well (without skipping any part in both videos) but I still couldn't understand why we have this rule of expected values being > 5 Edit 2 : If you come approx my comment by chance... one year after the upload of video ... Please consider answering my query by giving reference to any blog / other video in which that rule is explicitly explained Edit3 : If Justin sir is not free or dosent come accross my comment but if any other viewer sees it and knows answer to my query plz put that in reply to my comment. (I am sorry for so many edits but I my curiosity is forcing me to do all that... )

[@Darkwolf9280]

Too good - 🇮🇳

[@MarilynColvinBoon]

Thank you for an easy to understand explanation of Chi-Squares. Simply addressed.

[@PunmasterSTP]

Thanks for another extensively awesome video! Also, X/Twitter > Facebook, Instagram and TikTok...

[@yasminfatima5948]

In exp probability of male and female why you took 35 & 50 in numerator

[@hamzakaina8579]

You are amazing

[@saiyedmohdmisbahuddin3679]

Great....really helpfull

[@Matibonilla]

Thank you! So clear! I just have a question... If your example DID reach a significance level... What should i do to know wich group is the most far away from the expected frecuency? For example, can you say something like: The Tiktok gruop differs the most from de Facebook group?. Ty!! Great videos, keep up at it! Greetings from Argentina

[@Backpackers9898]

What if at least one expected frequency is less than 5???

[@tensor131]

You use Yates' correction, although this is currently in dispute as apparently it over-corrects the error. Of course everything in stats is approximate since real data only follows theoretical distributions approximately. I am guessing (I don't know!!) that Yates correction is related to when you use a continuity correction when calculating binomial probabilities using the normal approximation - for large n it hardly affects the answer so it can be neglected ( i have only just figured this out after watching zedstat's explanation of the link between chi-squared and the (O-E)^2/E formula). I think the vast majority of people who use stats every day do so having no idea why any of it works - unfortunately this can have the undesired consequence that they have used it inappropriately and their conclusion may be invalid.. I understand that this is becoming a bugbear in many recent scientific papers. Kudos for zedstat for teaching the world how it all works!!

[@davidclements9701]

saved my math ia thank you

[@maungmyint2056]

The researchers should study these lessons, esp. ecologist.

[@engincangocer9962]

Thanks...

[@elenabottiglieri]

Amazinggggg

[@korman9872]

tx sir

[@saymetahiruglo3243]

how did he get the 2.84? helppppppppppppppppppppppppppppppppppp

[@kopimo333]

same , i am getting 2.79

[@ScientificallySpeaking-ys3vo]

me too 2 years later@@kopimo333

[@ricardoafonso7563]

. Thank you... .

[@geozone2]

Good

[@Competent00]

Videos like this informative have only 28k views while videos on tiktok have millions and millions of likes. Thats the irony😬

[@redcat7467]

Cool.

[@HERO3530]

Seriously , I would pay college tuition to you instead of some incapable professors.

[@karannchew2534]

There's not enough evidence to suggest they are dependent. They are not dependent, based on the given data. They don't affect each other, based on the given data. BUT There are not necessarily independent. Because we need a more rigorous test and more data to draw such conclusion. There's not enough evidence that he is guilty. Not enough proof that he is guilty. BUT he us not necessarily innocent. Need much more proof and different criteria to conclude he is innocent.

[@eshitasonawane1391]

Wow

[@adityanjsg99]

I hate Equations.. Only a person longing for true understanding will tell that!

[@xiaoweidu4667]

tik-tok