HerfstkleurenHelpdesk IBM SPSS Statistics 20

For students from Arnhem Business School
Home Codebook Data Data editing Analysis Graphs Settings Links Methods

Data Split File

In our example data file one of the variables indicates the district where the respondents live; a second variable shows the age of the respondents.
If you want to study the ages for the respondents from all districts separately, say by means of a histogram or by using descriptive statistics, you can use Select Cases a number of times. Select a district, perform the analyses and then select the next district.

There is an easy alternative for this by using Split File. It is shown in the example below.

Start by opening data_split.sav.
Next choose from the menu Data > Split File

Complete the dialog box as shown above.
Note that there are two split options: Compare groups and Organize output by group.

Try the other option yourself and study the differences in output for the calculations that we show below.

The result is a data set that is sorted by district. Furthermore, on the bottom right of the screen the status bar shows that a splitting of the file is applied (Split by district).

How does the split effect the analyses?

Let us calculate the average age again; proceed as follows:


By means of the button you can select the statistics that you want to be calculated. We use the default settings.

Here is the result:

Note 1: In the output it is clear now that a split of the data has been applied and the values of the selected statistics are shown per district.

Note 2: If you want to return to a single data file with all cases choose Data > Split File and tick the radio button for "Analyze all cases, do not create groups".


Last modified 30-10-2012

Jos Seegers, 2009; English version by Gé Groenewegen.