Chi-Square Test [Simply explained]

The chi-square test is a hypothesis test used for categorical variables with nominal or ordinal measurement scale. The chi-square test checks whether the frequencies occurring in the sample differ significantly from the frequencies one would expect. Thus, the observed frequencies are compared with the expected frequencies in the Chi2-Test and their deviations are examined.
Chi-Square test calculator:
datatab.net/statistics-calcul...
More Information about the Chi-Square test:
datatab.net/tutorial/chi-squa...
Chi-squared distribution:
datatab.net/tutorial/chi-squa...

[@datatab]

If you like, please find our e-Book here: datatab.net/statistics-book 😎

[@akackon]

For those wondering how to get the expected frequencies: For each cell ((Sum of row)*(Sum of columns))/Total number of observations

[@datatab]

Many many thanks!!!

[@reyanshnambiar1524]

hi, thanks for the formula, May i kno the total number of observations is for the entire table or count of rows and column used in this section alone.

[@akackon]

@reyanshnambiar1524 not so sure if I understand the question well but total number of observations should be sum of column totals or sum of row totals

[@nikhilkenvetil1594]

Example: Consider the "Observed Frequency" table at 7:00. For "Female without graduation", you calculate the Expected frequency like this: ((6+7) * 43)/(43+49). Similarly, for "Male without graduation", Expected frequency would be ((6+7) * 49)/(43+49) and so on or and so forth. Just adding to what @akackon said.

[@frankl1]

The null hypothesis is that there is no relationship between those two variables, so there should be as many females as males for each educational level. So why do we have those differences in the expected frequencies?

[@markfidelix]

By far the best quick video I've seen about the Chi-square test!

[@omardiaaeldessouky7215]

I love explaining concepts well and marketing for the product in the same time. Very creative and informative approach :)

[@datatab]

Glad you liked it and many thanks for the nice feedback! Regards Hannah

[@moritzwolle2141]

It´s a trap

[@zohrehschuessler394]

You are one of the best statisticians I have known via KZbin.

[@krishnaiahkonda5916]

Chi-Square Test is a hypothetical test to determine the relationship between two categorical variables like gender and highest qualification. Well explained with online and offline examples. Thank you DATAtab.

[@anujsangal3886]

The best explanation of Chi square test by a software company

[@datatab]

Many thanks!!!

[@Becky71610]

This video was amazing, thank you so much! you make it so easy to understand!

[@datatab]

Glad it was helpful!

[@thabangramashapa1094]

This video might've helped me obtain my qualification, thank you!!!

[@datatab]

Glad I could help!

[@germanslobo]

Amazing explanation! So simple and easy to understand. Thanks.

[@datatab]

Glad it was helpful!

[@danteperez5347]

The best explanation on the Internet!

[@datatab]

Many many thanks for the nice feedback! Regards, Hannah

[@danjarupath]

Agreed 💯

[@asma7918]

What a clear explanation. Thank you.You are helping many..👏👏🙏

[@datatab]

It's my pleasure

[@dstan16224]

Thank you for the video! Finally understood the concept.

[@datatab]

Glad it helped!

[@PediaTricks007]

Simplest video explanation on chi square test. Thanks

[@HemSingh-rt5bp]

Splendid explanation with examples thank you so much

[@JagooTheBoss]

best explanation of chi-square so far

[@RameshKhanal16]

Really Helpful.. Thank you for the quality content!

[@datatab]

Glad you liked it and many thanks for the nice feedback! Regards Hannah

[@Vivek-np9vm]

Thank you so much for the calculator. This is exactly what I was looking for. Love from India.

[@datatab]

Thanks!

[@notabotbutarealperson]

I passed my whole test thanks to this one video.

[@datatab]

Glad it helped!

[@NK-iy6if]

God bless you!!! Your videos are very helpful, and your explanation is very clear and easy to follow. Thank you so much!!!

[@datatab]

Many thanks for your nice feedback! 🙂 Regards, Hannah

[@Hannah-brien]

no ur wrong

[@sajjadabdullah]

Very informative and easy. Thank you

[@emmanuilovich]

Thank you very much for such a clear explanation

[@dr024]

Consider that this is the ChiSq test of independence. There are other chisq tests like chisquare test for variance, chi-sq test for goodness of fit, etc.

[@datatab]

Many thanks for your feedback!!! Yes this is true! Ragards, Hannah

[@ShadowRonan]

Thanks so much for explaining this clearly

[@datatab]

Glad it was helpful!

[@ray2day354]

Amazing video, you have explained everything very well

[@datatab]

Glad you liked it!

[@ma.cecillemanabat5617]

a very well explained for the CHI squared.help me alot

[@datatab]

Many thanks!

[@rumipromitsett4588]

I can't believe it can be this good

[@datatab]

You're very welcome! Many thanks for your nice Feedback! Regards Hannah

[@ajmirasheed2385]

Do more videos. It was very understanding

[@datatab]

Many thanks!!! We will do : ) Regards, Hannah

[@LindaInHiding]

Thank you so much for this, it really helped me

[@datatab]

Glad it helped! Regerds, Hannah

[@queenisforever1]

very good, keep up the good work

[@datatab]

Thanks!

[@thutrang638]

it's so useful. thank you very much

[@datatab]

You are welcome!

[@jygliocam7448]

Thank you for this, how did you get the expected frequencies for the perfectly independent variables?

[@datatab]

Hello, you have to calculate the expected frequencies first!

[@jygliocam7448]

@@datatab Hello thank you so much. Do you have video on how to solve?

[@boosted91]

I have the same question - how were the expected frequencies calculated? Is that using obtained data in previous studies?

[@user-wr9ks3tf4n]

@@boosted91 the expected frequencies for female without education , for example , is 43/(49+43)*(6+7).

[@laradomingos6852]

@@user-wr9ks3tf4n I don't understand, following your steps the results were 27.814 when the should be 6.08! I am doing something wrong, just don't know what!

[@Griffindor21]

Great explanation👍👍, but....what was the null hypothesis in the first place?

[@nesmamohamed4488]

very simple, thanks

[@datatab]

You're welcome!

[@TR-rx5ej]

Can I use mann kendall test for normally distributed data set? Mann kedall test is used for non normally distributed data set.

[@shadialhag6616]

Thank you very much for the wonderful way of explaining this topic, is there a video on SPSS topic.

[@datatab]

No only with DATAtab, sorry!

[@chamanthawickramarathne2029]

You executed this test for two nominal variables only. Can we use this for two ordinal variables as well and can we use this to find the relationship between nominal and ordinal variables?

[@abdullahallmasudbhuiyan4540]

You should go for spearmens or kendall's tau-b corelation

[@cidc5067]

Hi! Is the significance level calculated also? or is it only based on judgement?

[@datatab]

Yes, the p-value is also calculated. Regards, Hannah

[@CoachPegasus]

Thank you

[@datatab]

YOu are welcome! Regards, Hannah

[@vaishalijadhav862]

Thank u for wonderful explaination. I want to know wheather we have to purchase datatab software or it is freely available

[@datatab]

With datatab we earn money, so we have to charge something, but it is very cheap! datatab.net/premium/packages

[@muhammadwaseemassistantdea6073]

very informative video

[@myobnp9684]

Constructive feedback. This is missing an explanation of the null hypothesis at the beginning. Then p values and chi-squared values are not clearly distinguished at the end.

[@newmercies1]

Hi, nice video. Please clarify what is your null hypothesis (Ho) and your alternative hypothesis - it is significant to mention it before you run the test as it defines the area for rejection or FTR (fail to reject). Please clarify.

[@datatab]

Many thanks for your feedback!

[@kenken578]

Same question, may I know for clarification of what's the specific null hypothesis and the alternative one for this prob?

[@EranM]

@DATAtab Is this a bot writing the answers?? Just write the null hypothesis..

[@datatab]

@@EranM ​ @Beverly Cepeda I am sorry! No it is not a bot but there are so many comments that it is hard for me to read them carefully, sorry for that! We have just created a new Video for chi2 Test, I hope we explain the hypotheses better in this video: kzbin.info/www/bejne/e5vXiZ5jrqp-bNk. For the example in the "old" video: Null hypothesis: there is no relationship between gender and highest educational attainment. Alternative hypothesis: There is a correlation between gender and the highest educational attainment. Regards Hannah & Mathias

[@malaikaimtiaz6446]

I want to ask a question. Will yes and no valued as 1 and 0 respectively be categorical?

[@World_Exploror]

Good Explanation

[@datatab]

Thanks!

[@micahmicz0917]

I thought your stat-calculator was free but we need to buy it.

[@datatab]

otherwise, unfortunately, the calculator would not exist, as I would not be able to finance it! But it is very cheap for Staitsitk software solutions!

[@swatijadhav686]

hey how did we get those expectd frequencies values?

[@tofunmialayode1927]

Very useful 👌

[@datatab]

Glad to hear that

[@marcelmaurice4586]

Excellant tutorial without irrelevant peripheral presenter's ego. Could categorical variables be equated to discrete variables?

[@datatab]

Hi many thanks for your nice feedback! No actually, a discrete variable is e.g. how many people watched this video! In that case it is not possible but purely theoretically with discrete variables an infinitely large number can come out. Categorical variables include a finite number of categories or unique groups and there does not have to be a ranking. Regards Hannah

[@applyalone]

Perfect

[@datatab]

Many thanks!

[@applepakistan3164]

In this video (chi square) at 6:21 time span, if p value is smaller than .05 then we should accept the null hypothesis ( false ) or we reject the null hypothesis (true), please clarify 🙏🙏🌹thanks

[@datatab]

if p value is smaller than .05 then we reject the null hypothesis

[@suranjanasikdar6224]

@@datatab if p value is exactly 0.5 then it means rejection of null hypothesis too?

[@Sushi_10000]

@@suranjanasikdar6224 p-value equal to 0.05 also results in rejecting null hypothesis

[@ajaya9355]

whether Chi square test is used for Likert Scale questions Thank you

[@SmartLife123]

Is Datatab like spss, an analysing tool?

[@sanjay8925]

What if critical value for .01 & .05 level of significance is given and our chi square value falls in between them, like if chi square value is lesser at .01 level of significance and more at .05 level of significance, so do we accept null hypothesis or reject??

[@datatab]

You define you significance level before the test and only one value, so either 0.01 or 0.05. Then the question is, is it smaller or greater.

[@datadoctor10]

Thank you mam

[@raylapikeri8766]

Good !

[@applepakistan3164]

Madam, what is the difference between Chi-square, and Chi-square Goodness of Fit test ? Thanks

[@datatab]

There are several scenarios when using a Chi2 test, the two most common being that when two variables are tested to see if they are independent, the observed frequency is compared to the frequency distribution that would occur if both variables were completely independent. The second way is to compare the observed frequency with a given frequency, that would be goodness of fit.

[@rasmitasharma4133]

Difference between spearman rank and chi square test?

[@saidfawzy]

لGreat explination but how did you get the values in the table of expected frequency ?

[@datatab]

Many thanks for your feedback! Please have a look here datatab.net/tutorial/chi-square-test and scroll a little bit down! Regards Hannah

[@saidfawzy]

@@datatab Thanks for your concern and kindness . The world is so beautiful ,only because there is someone like you.

[@ivysonmat]

thanks a lot =)

[@datatab]

You're welcome!

[@sindilesira1117]

how did u get the p value

[@noraelsamachi3102]

Hello please I want to run a quistionnaire but I dont know exactly how to put my data in spss? Can you please help me? I have 2 categories of men , 75 smokers - no smokers 25 and 15 smokers women and 95 no smokers!

[@dr.kamrankundi9819]

Dear researchers, I have MSR (meta cognitive self regulation) learning strategy as independent variable while Students' Score (Students'GPA) is my dependent variable. MSR is on five point Likert scale while for Students'GPA each student put his own GPA on the questionnaire in form of digits like 2 , 2.1, 2.7, 2.3, 3, 3.6, 4 etc. The students score is in random form. Each student put his own score whatever he got in the previous semester. Now my question is which kind of correlation should I use? Can I go for the Pearson correlation? How to handle with the Students'GPA variable? The sample size is 300 students which means that I have 300 Students'GPAs. It's a confusing whether I should consider the Students'GPA as ordinal, interval or ratio variable? Which kind of correlation should I go for? Please, guide.

[@pghjmuhammadfirdauspghjrus8870]

Use Correlation, you already have two sets of data MSR (Ordinal) and GPA (Ratio). But, check the correlation assumptions first, either to use mean or median type analysis. Correct me if im wrong, thanks.

[@applepakistan3164]

Madam, we apply chi square to check relationship of two variables, we apply linear correlation also to check relationship, what is the difference between these tests , 🙏🙏

[@datatab]

In a correlation, your two variables must have a rank order. With the Chi2 test, you only need two nominal variables. Regards Hannah

[@applepakistan3164]

@@datatab rank means likert scale n ordinal also ??🤲🙏🌹

[@datatab]

@@applepakistan3164 Yes for a correlation you need two ordinal variables!

[@applepakistan3164]

@@datatab thank you for your help 🤲🤲🤲🙏🌹

[@miscelleneoustubes]

@@datatab For correlation, isn't that we need two scale variables?

[@MedicalDoctors-4-life]

How to get expected frequency

[@glentox1563]

Can this be used if I want to measure the relationship between gender and level of satisfaction?

[@datatab]

Without knowing the details I would rather say there fits the Mann-Whitney U-Test: Have a look at this video: kzbin.info/www/bejne/iqjOZ594naefbpo

[@matilderasori3251]

where do you take the expected frequency from?

[@datatab]

The expected frequency is given by you (by the research question)!

[@maykhaeunicebacalso4551]

I got df=39 and the table doesn't have a value for 39 T_T how do you calculate that?

[@Ssoonneess]

in 6:50 how do you get the "Expected Frequencies"?

[@datatab]

Many thanks for your feedback! Please have a look here datatab.net/tutorial/chi-square-test and scroll a little bit down! Regards Hannah

[@PHALANGE1931]

I love mathematics 😍💞

[@datatab]

: )

[@queensterlartey]

How did you get the expected frequency.

[@datatab]

Hi please find the information here: datatab.net/tutorial/chi-square-test at Chi-Square Test of Independence!

[@ibrahimahmedtache3347]

Why is chi-square test employed in categorical data analysis?

[@datatab]

Hello, sorry, I do not quite understand your question! If you have two categorical variables and want to check if there is a relationship, you can do this with a Chi 2 test. Assuming all conditions are met.

[@RobertaMontesdeOca]

is is possible to do this test for a Liker data?

[@datatab]

Hi, without knowing your hypotheses exactly, you would rather use a Spearman rank correlation in this case. kzbin.info/www/bejne/jofCiGStabOqpcU

[@monsuratyusuf3252]

is this video for Alevel biolgy math skill?

[@datatab]

Hmm yes you can certainly use it for that!

[@Intorduceingtoself]

To Good

[@datatab]

Glad it was helpful! Thanks for your nice feedback! Regards Hannah

[@pade9732]

This video is a commercial, not an explanation...

[@tk006]

Still a good explanation

[@victoriadamisa]

But at least, it provides a solution

[@Ilemaurice687]

What if we have only one sample plz

[@datatab]

Then you can also use the CHi 2 test!

[@abolajioluwadamilola400]

I tried calculating the observed and expected frequencies all together just like you did, but I didn't get the same answer,pls can u make it clear to me🤲, and also the part where u used the chi square table,it feels like I'm the only one that doesn't understand here😭,I need ur answers pls

[@akaezesamuel4688]

Using DATAtab, you failed to reject the null hypothesis, but when you calculated by hand, you rejected the null hypothesis, please which is it? So confusing

[@datatab]

Hi, by hand we also don't reject the null hypothesis! The calculated chi2 value ist smaler than the critical one and we don't reject it! Or did I say it wrong once in the video? We also have a new version of this video: kzbin.info/www/bejne/e5vXiZ5jrqp-bNk Regards mathias

[@yash1152]

nice (:

[@esqinmehemmedli5611]

Done

[@datatab]

👍

[@Ilemaurice687]

One group

[@datatab]

Then you can also use the CHi 2 test!

[@yimenugardie2923]

excuse, the display lecture is not visible

[@pelas91]

6:30

[@ImTheCitizenInsane]

"Whether gender has an influence on education" suggests causality, which is wrong here

[@leewilliam3417]

Mmmmmmmm😊

[@MarcLL]

But I was told there were an infinite number of genders 🤔

[@davidejibia3930]

you were misinformed im afraid

[@JacSilverMoon]

Too bad its not free, asks you to pay to get results!

[@theliterunner]

i wish dislike button was still working for youtube

[@datatab]

😥