A researcher compares blood pressures of 40 vegetarians and 40 meat-eaters to see if diet affects blood pressure. Which test is appropriate?

A researcher compares blood pressures of 40 vegetarians and 40 meat-eaters to see if diet affects blood pressure. Which test is appropriate?

• A. 2-sample t-test
• B. Chi-square test of independence
• C. Linear regression
• D. Paired t-test

Answer: (A)

To compare averages of a continuous outcome between two independent groups, use a two-sample t-test. Here, blood pressure is a continuous variable and the vegetarians and meat-eaters are two separate, independent groups. The two-sample t-test directly tests whether the average blood pressure in vegetarians differs from that in meat-eaters, assuming the data are roughly normally distributed (or the sample size is large enough for the t-test to be robust) and the observations are independent. Chi-square is used for relationships between categorical variables, not for comparing means of a continuous variable. Linear regression could model blood pressure as a function of diet, but for a straightforward comparison of means between two groups, the two-sample t-test provides the simplest and most direct test. A paired t-test would be for related or matched samples, such as the same individuals measured under two conditions, which isn’t the case here.

To compare averages of a continuous outcome between two independent groups, use a two-sample t-test. Here, blood pressure is a continuous variable and the vegetarians and meat-eaters are two separate, independent groups. The two-sample t-test directly tests whether the average blood pressure in vegetarians differs from that in meat-eaters, assuming the data are roughly normally distributed (or the sample size is large enough for the t-test to be robust) and the observations are independent.

Chi-square is used for relationships between categorical variables, not for comparing means of a continuous variable. Linear regression could model blood pressure as a function of diet, but for a straightforward comparison of means between two groups, the two-sample t-test provides the simplest and most direct test. A paired t-test would be for related or matched samples, such as the same individuals measured under two conditions, which isn’t the case here.