McNemar test

Command: Statistics
Next selectCrosstabs
Next selectMcNemar test

Description

The McNemar test is a test on a 2x2 classification table when you want to test the difference between paired proportions, e.g. in studies in which patients serve as their own control, or in studies with "before and after" design.

In the McNemar test dialog box, two discrete dichotomous variables with the classification data must be identified. Classification data may either be numeric or alphanumeric (string) values. If required, you can convert a continuous variable into a dichotomous variable using the Create groups tools. The variables together cannot contain more than 2 different classification values.

For example, in a study a test is performed before treatment and after treatment in 20 patients. The results of the test are coded 0 (negative) and 1 (positive). Is there a significant change in the test result before and after treatment?

How to enter the data in the spreadsheet

How to enter data in the spreadsheet for the McNemar test

Required input

The dialog box for the McNemar test is completed as follows:

McNemar test dialog box

Select two discrete variables with related classification data. Classification data may either be numeric or alphanumeric (string) values, but both variables cannot contain more than 2 different classification values.

After you have completed the dialog box, click OK to obtain the classification table with the relevant statistics.

McNemar test results

Classification table

The program displays the 2x2 classification table.

Difference and P-value

The program gives the difference between the proportions (expressed as a percentage) with 95% confidence interval.

When the (two-sided) P-value is less than the conventional 0.05, the conclusion is that there is a significant difference between the two proportions.

In the example, 45% of cases were positive before treatment (Classification A = 1) and 35% were positive after treatment (Classification B = 1). The difference before and after treatment is -10% with 95% CI from -29.1% to 9.1%, which is not significant (P=0.625, n=20).

Note

The two-sided P-value is based on the cumulative binomial distribution.

The 95% confidence interval is calculated according to Sheskin, 2011.

Literature

  • Altman DG (1991) Practical statistics for medical research. London: Chapman and Hall. Book info
  • Sheskin DJ (2011) Handbook of parametric and nonparametric statistical procedures. 5th ed. Boca Raton: Chapman & Hall /CRC. Book info

See also

External links

This site uses cookies to store information on your computer. More info...