In a randomized trial comparing AZT during pregnancy to placebo, researchers compared the proportion of births in which the baby was HIV-positive between the two groups. Which statistical test is most appropriate for comparing these two proportions?

In a randomized trial comparing AZT during pregnancy to placebo, researchers compared the proportion of births in which the baby was HIV-positive between the two groups. Which statistical test is most appropriate for comparing these two proportions?

• A. 2-proportion z-test
• B. Paired t-test
• C. One-sample z-test
• D. ANOVA

Answer: (A)

When you’re comparing two proportions from independent groups with a binary outcome, you use a test designed for proportions. The two-proportion z-test evaluates whether the observed difference in HIV-positive birth rates between the AZT group and the placebo group could occur by chance if the true proportions were the same. It relies on a normal approximation to the distribution of the difference between the two sample proportions, using their observed values to compute the standard error. This is appropriate here because the outcome is binary and the groups are independent. The other options don’t fit: a paired t-test is for paired continuous data, a one-sample z-test compares a single sample proportion to a known proportion, and ANOVA compares means across groups for continuous outcomes. So the two-proportion z-test is the correct choice.

When you’re comparing two proportions from independent groups with a binary outcome, you use a test designed for proportions. The two-proportion z-test evaluates whether the observed difference in HIV-positive birth rates between the AZT group and the placebo group could occur by chance if the true proportions were the same. It relies on a normal approximation to the distribution of the difference between the two sample proportions, using their observed values to compute the standard error. This is appropriate here because the outcome is binary and the groups are independent. The other options don’t fit: a paired t-test is for paired continuous data, a one-sample z-test compares a single sample proportion to a known proportion, and ANOVA compares means across groups for continuous outcomes. So the two-proportion z-test is the correct choice.