CHAPTER 12 Comparing Proportions and Analyzing Cross-Tabulations 159

Chapter 12

Comparing Proportions

and Analyzing

Cross-Tabulations

S

uppose that you are studying pain relief in patients with chronic arthritis.

Some are taking nonsteroidal anti-inflammatory drugs (NSAIDs), which are

over-the-counter pain medications. But others are trying cannabidiol (CBD),

a new potential natural treatment for arthritis pain. You enroll 100 chronic arthri-

tis patients in your study and you find that 60 participants are using CBD, while

the other 40 are using NSAIDs. You survey them to see if they get adequate pain

relief. Then you record what each participant says (pain relief or no pain relief).

Your data file has two dichotomous categorical variables: the treatment group (CBD

or NSAIDs), and the outcome (pain relief or no pain relief).

You find that 10 of the 40 participants taking NSAIDs reported pain relief, which

is 25 percent. But 33 of the 60 taking CBD reported pain relief, which is 55 percent.

CBD appears to increase the percentage of participants experiencing pain relief by

30 percentage points. But can you be sure this isn’t just a random sampling

fluctuation?

IN THIS CHAPTER»

» ^{Testing for association between}

categorical variables with the

Pearson chi-square and Fisher Exact

tests»

» ^{Estimating sample sizes for tests of}

association