Chi-square test of independence? - Answers

This tests whether two categorical variables are related, meaning if they affect each (whether they are independent or associated. Your null hypothesis would be that these two variables are independent. Your alternate hypothesis would be that these two variables are dependent. To carry out this test, you must make sure that all the expected counts are greater than 1 and that 80% of these data are greater than 5. Moreover, you must make sure that the data was received by SRS and that the sample is independent. Afterwards, you can plug it in to your graphing calculator in a matrix and use the x^2test. However, if you do not have a graphing calculator, you must calculator the expected value of each value. You do this by multiplying the total counts in the row * total counts in the column/ total counts. Then for each value, you take: (Observed - Expected)^2 / (Expected). After you receive different values, you add them up to make up your x^2. Afterwards, you find the P-Value but looking at a chi distribution curve/table and find the area that is greater than x^2 value. If it is small, you can reject the null hypothesis (like less than 0.05). If not, you fail to reject the null hypothesis and therefore conclude that these two variables are independent.