I am confused on when to use chi-squared test for homogeneity or chi-squared test for independence. Can you tell me some tips or how to distinguish those two?
The main difference between those two tests is whether you are testing for an association between two categorical variables or you are comparing one variable between multiple populations.
If you are testing for association, you use chi-squared test of independence.
If you are comparing one variable between multiple populations, use chi-squared test of homogeneity.
For example, if the question asks for whether there is an association between age and height, use chi-squared test of independence. On the other hand, if the question asks whether the distribution of height is different for different nations (so you are comparing distributions), use chi-squared test of homogeneity.
If you fail to understand what the question is really meaning, one key distinction you can use is based on the number of samples taken.
If only one sample was taken, use chi-squared test of independence.
If two or more samples were taken, use chi-squared test of homogeneity.
Hi,
Thank you for your question.
The main difference between those two tests is whether you are testing for an association between two categorical variables or you are comparing one variable between multiple populations.
If you are testing for association, you use chi-squared test of independence.
If you are comparing one variable between multiple populations, use chi-squared test of homogeneity.
For example, if the question asks for whether there is an association between age and height, use chi-squared test of independence. On the other hand, if the question asks whether the distribution of height is different for different nations (so you are comparing distributions), use chi-squared test of homogeneity.
If you fail to understand what the question is really meaning, one key distinction you can use is based on the number of samples taken.
If only one sample was taken, use chi-squared test of independence.
If two or more samples were taken, use chi-squared test of homogeneity.
Does this answer your question?