Question

If you are comparing counts of two unrelated nominal variables with a chi-square test, what remaining assumption you need to check before continuing? What options do you have if your assumptions are not met?

Answer 1

The assumptions of a chi-square test are:

• Method of simple random sampling is used
• variable is measured on nominal scale of measurement
• expected value of each of cell is at least 5

Test Statistic:

If the expected cell count of each cell is less than 5 it is preferred to group a few categories of the variable to make the expected count greater than or equal to 5.

Phi coefficient is also used to test the relationship between two variables which are measured by the nominal scale of measurement