McNemar's test is used to...

McNemar's test is used to...

• A. Test differences between two independent groups with nominal data.
• B. Test differences between two related dichotomous variables and changes in response.
• C. Test correlations between continuous variables.
• D. Compare means of ordinal data.

Answer: (B)

It focuses on whether there is a change in the proportion of a binary outcome across two related observations. McNemar's test uses paired nominal data, looking at two related dichotomous variables (for example, before and after measurements on the same subjects) to see if responses flip from one category to the other more in one direction than the opposite. Data are organized in a 2x2 table that highlights discordant pairs—those that change from yes to no versus no to yes. The test examines whether those discordant counts are equal; a significant difference suggests a real change in response over time or due to pairing. This is why it’s the right choice for two related dichotomous variables and changes in response. It isn’t appropriate for independent groups, correlations between continuous variables, or comparing means of ordinal data.

It focuses on whether there is a change in the proportion of a binary outcome across two related observations. McNemar's test uses paired nominal data, looking at two related dichotomous variables (for example, before and after measurements on the same subjects) to see if responses flip from one category to the other more in one direction than the opposite. Data are organized in a 2x2 table that highlights discordant pairs—those that change from yes to no versus no to yes. The test examines whether those discordant counts are equal; a significant difference suggests a real change in response over time or due to pairing. This is why it’s the right choice for two related dichotomous variables and changes in response. It isn’t appropriate for independent groups, correlations between continuous variables, or comparing means of ordinal data.