Doctors at a technology research facility randomly assigned equal numbers of people to use standard keyboards in one room and ergonomic keyboards in another to see if ergonomic keyboards increase words per minute. Which test should be used?

Doctors at a technology research facility randomly assigned equal numbers of people to use standard keyboards in one room and ergonomic keyboards in another to see if ergonomic keyboards increase words per minute. Which test should be used?

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

Answer: (A)

The main idea is to compare the average performance (words per minute) between two independent groups. Since participants are randomly assigned to either the standard keyboard group or the ergonomic keyboard group, the two groups are separate and independent, and the outcome is a continuous measure. The two-sample t-test is designed to compare the means of two independent groups on a continuous outcome, making it the right choice here. The other tests fit different situations: a 2-proportion z-test is for comparing proportions (binary outcomes), a paired t-test is for measurements taken on the same individuals or matched pairs (before-and-after on the same person, or twin pairs, etc.), and the McNemar test is for paired binary outcomes. If the two groups have very different variances or a small sample size with potential non-normality, you’d adjust with Welch’s t-test or consider nonparametric alternatives, but the standard approach for this design and variable is the two-sample t-test.

The main idea is to compare the average performance (words per minute) between two independent groups. Since participants are randomly assigned to either the standard keyboard group or the ergonomic keyboard group, the two groups are separate and independent, and the outcome is a continuous measure. The two-sample t-test is designed to compare the means of two independent groups on a continuous outcome, making it the right choice here.

The other tests fit different situations: a 2-proportion z-test is for comparing proportions (binary outcomes), a paired t-test is for measurements taken on the same individuals or matched pairs (before-and-after on the same person, or twin pairs, etc.), and the McNemar test is for paired binary outcomes. If the two groups have very different variances or a small sample size with potential non-normality, you’d adjust with Welch’s t-test or consider nonparametric alternatives, but the standard approach for this design and variable is the two-sample t-test.