If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

## Statistics and probability

### Course: Statistics and probability>Unit 14

Lesson 2: Chi-square tests for relationships

# Chi-square test for association (independence)

Chi-square test for association/independence.

## Want to join the conversation?

• Just to make sure, calculating Chi-square for association and homogeneity are same but the interpretation is different.
Am I getting this correctly?
• Yes, they are calculated the same however the interpretation is different. One is asking if they are independent or have no association with one another and is done on something with one sample and a test for homogeneity is a test for multiple samples and is asking if there is difference between the different samples.
• Maybe I missed it, but when you go back to check the expected values, why did they have to be at least greater than or equal to 5?
• To meet the condition of Large counts for any X^2 Statistic.
• When specifically does one use a T-test and a chi-square test.
• A t-test is used to determine the difference between two sets of data. A chi-square test involves looking for a relationship (homogeneity, independence, or goodness-of-fit.)
• Where did you get the p value from in the last section of the video?
• You can use chi-squared cdf to calculate the probability you would get such a chi squared value. Set the max to 9E99, min to the chi squared statistic you just found, and the degrees of freedom is (row-1)(col-1) as he shows.
• Why isn't it Row total x Column / Table total? also when I tried to use the formula I mentioned it doesn't add up to the total... I don't really understand what is going on
• At , why is the P value 0.018? According to the Chi squared table, the p value, corresponding to the Alpha level of 0.05 and the Degree of Freedom of 4 should be 9.49.
• That's not how you use the table. The chi-squared value is the input, and then the tail probability is the p-value. Alpha level is just a preset level that needs to be passed in order to reject the H0. Also, how are you going to get a probability higher than 100%?
(1 vote)
• I understand all the calculations and hypothesis testing, but I don't understand what "association" means here. What does it mean that there's an association between hand length and foot length or that they're not independent? Does that mean that if you get another sample, you'll get roughly the same distribution/percentages?