Data Analysis with Microsoft Excel: Updated for Office 2007

266 Fundamentals of Statistics

The Mann-Whitney test relies on ranks; this lessens the effect of outliers.

The downside is that under certain situations, the Mann-Whitney test will

not be as effi cient as the two-sample t test in detecting differences between

samples. Also, some researchers may not be familiar or comfortable with

this nonparametric approach.

Let’s apply the Mann-Whitney test to the nursing home data. Our hypoth-

eses are

H 0 : Median population revenue for rural nursing homes 5

Median population revenue for nonrural homes

Ha: Median population revenue for rural nursing homes 2

Median population revenue for nonrural homes

To apply the Mann-Whitney test to the nursing home data:

1 Return to the Nursing Home Data worksheet and click Two Sample

Tests from the StatPlus menu and then click 2 Sample Mann-Whitney

Rank test.

2 Verify that the Use column of category values option button is

selected.

3 Click the Data Values button and select Revenue from the list of

range names.

4 Click the Categories button and select the range name Location.

5 Click the Output button and send your output to a new worksheet

named MW-test.

6 Click OK to generate the Mann-Whitney analysis for the nursing

home revenue. See Figure 6-23.

95% confidence

interval

median difference value

between the two samples

Figure 6-23
Mann-Whitney
analysis of the
revenue data

p