If we want to test proportions, one wrong way to do so is to
A) put the proportions themselves directly into the chi-square formula.
B) convert proportions to frequencies.
C) take the difference in proportions and turn it into a z score.
D) You can't do any of these things.

Question

Quizplus · Accepted Answer

Putting the proportions directly into the chi-square formula is incorrect because the chi-square test requires observed and expected frequencies, not proportions. Converting proportions to frequencies and taking the difference in proportions to turn it into a z score are both valid approaches for testing proportions, depending on the context and the specific statistical test being used.

If We Want to Test Proportions, One Wrong Way to Do