Pearson's Chi Square Test (Goodness of Fit)

[UoCvodka]

i dont have enough words to thank



[Trackman2007]

Now i'm much smarter when buying a restaraunt

[talkatoo93]

Your work is literally amazing...you've taught me so much ...thank you sooo much ....you're a Good example to learn from

[whyfightnow]

I understand a little better than my math statistic class I struggled in. I'm taking my exam tomorrow. I hope I do good.

[s0nika]

@whyfightnow goodluck!

[whyfightnow]

@s0nika Thank you so much for your support. I got a B on my exams (math statistic & chemistry) Those classes was very tough for me. I'm so glad it's over with. But, at least I know more now than before.

[plausib13]

what programme are you using to demo this? thanks and fantastic video, really helped out with population genetics at uni!!

[Search4Knowledge]

How amazing that I pay 3.400 euro's at university and get bored to death with useless info when Khan explains this in about 11 minutes...

[MusicConfidential]

how do get the pvalue?

[kelleyal1]

This is the first video that actually helped me understand this.

[bliyra]

You have a sexy voice



[kazuyalegrey]

you are AMAZING

[cinthiai]

Turns out I actually do get Chi-Square Test after watching this.

Thank YOU so much!

[JizzontheToilet]

Its Funny how I pay my professors/school thousands of dollars per semester and yet Khan always teaches me better for free.

[Nickiller6969]

it's funny how you always repeat yourself but that makes your video's so easy to understand! thanks for your help!

[Simistar3179]

How are the values in that table calculated?

[shantanudas]

Sal, another clear video. However, may be you could have talked a little bit as to why chi-sq is appropriate for this particular type of problem (or similar type) - how would a student know when to select the same distribution for a problem at hand.

[justin0981]

i love you



[Neueregel]

nice one

[stupidpoems]

Great video, great pace of explaining. Thank you!!

[1991LATIFA]

thank you .. that really helped me :)

[Mustang030988]

Amazing Help, 5/5.

[bartcase]

It's quite funny how every time you say "statistic" you have to slow down and make an effort to get it right. But seriously, I've watched most of your videos from the stats playlist and it's been a great deal of help. Thanks Sal!

[footballmanager221]

So if my chi squared value worked out at 1.86 and my S.L (at 5%) was 3.84, then i would accept my null? How can i word that using the words 'chance' and 'probability'?

[DRXY1]

That helped me a lot. Thanks for posting.

[sharmin801]

all right, so I feel real excited whenever I search video for homework help and find Khan's videos.

[mrvrabie]

OK... I think this example is a little bit wrong... because a Chi-squared distribution (k dof) is the squared sum of k normally random distributed x variables.

So for your example to be true, besides extracting the Expected amount, you would need to devide by the Standard deviation... u need to normalize it.

[CogitoErgoCogitoSum]

Would you mind making a video on creating a confidence interval from a sample, to determine the populations standard deviation

[CogitoErgoCogitoSum]

What if instead of using H0 probabilities to compute expected counts (customers), you instead translated the observed patronage into percentages?

[fattyfetz]

@CogitoErgoCogitoSum this test is a non-parametric statistical test therefore you would use it when you cannot obtain the population parameters μ and σ

[CogitoErgoCogitoSum]

For this example to work dont you have to assume each variable (number of guests each day) is normally distributed?

How do you prove a distribution is normal? How do I match observation to an estimated normal curve? I mean, to create the estimated normal curve for hypothetical reasons, you would need to know its mean and standard deviation... how might you derive those parameters?

[CogitoErgoCogitoSum]

In your first video of Chi-square, the introduction, you said that when you sum the squares of n normally distributed variables you get a chi-square distribution with n degrees of freedom.

In this video you are telling us to subtract one to determine degrees of freedom.

Which is it Sal? Are you contradicting yourself or am I missing something important here?

[creynolds790]

@CogitoErgoCogitoSum its n-1

[CogitoErgoCogitoSum]

@creynolds790 Perhaps it is. Frankly Im less concerned about what "is" and more concerned about the persistence of inconsistency and contradiction I see in Sals videos

[zencoweb]

great teaching skills

[neoaeonian]

This doesn't seem right. I think the first term should be ((30-20)/20)^2, not ((30-20)^2)/20 as you've shown it.

Proof: If he had 300, 140, 340, 450, 570 and 200 customers the chi-sq value should be identical, but your method would have been much higher. Note: the 1st term alone becomes 10,000/200 or 50, and that's just the first term, so it will end up far higher than 11.44.

[techtata]

@neoaeonian

well, it is actually correct, because chi^2= Sum [ (Observed-Expected)^2 / Expected ] Note that the denominator is not squared.

[neoaeonian]

@techtata Yep I was wrong. Contrary to my intuition, ((30-20)^20) meters/20 meters = 5 meters but ((3000-2000 centimeters)^2)/2000 centimeters = 5000 centimeters. My intuition was that the numbers would be different, but both are identical (5K cm = 50m).

Can't say that I yet intuitively understand this test, but thanks for replying, as it made me double check my math (and look it up). Khan was right.

[mishuga]

@neoaeonian This confuses me too. I would love him to explain his normalization a bit more.

[Nandine2]

Hey can you do some videos on error propagation to go with your statistics videos?

thanks

[Leonar12345]

I HAD A TEST ON THIS YESTERDAY AND I WAS FRANTICALLY LOOKING FOR THIS VIDEO WHEN I WAS STUDYING BUT U DIDNT HAVE IT UP =[

[JESSFALLINGDOWN]

Thank you so much for this!!!!

[pum88]

Great!

[brad9899]

*get

[brad9899]

When do you think you will ge to linear regression?

[AA10Megaviv]

@brad9899 I think he's done that already (recently, however) assuming you indeed referred to the topic I think you did by "linear regression".

[brad9899]

can you do, the tests for the variance?

[brad9899]

your such a good teacher, im looking forward to more stats videos, very usefull