Split Function in SPSS

[Dr Neeraj Kaushik]

There's a wonderful concept of Split in SPSS.
The purpose of Split is to divide the file on the basis of some
variable, then whatever analysis we'll perform, it'll be split for
those categories.

Where's it?
Data Menu --> Split File

Example:
There's dependent variable Height
Independent variables are Age, Nutrition value in diet, Height of
parents
and we want to run Regression

Here one may be interested to see whether the all three variables are
equal important for boys & girls both or Can it be like for girls one
indep var is more important & for boys its some other var? Here we can
use split file on the basis of gender and we'll have 2 regression
equation separately for boys & girls.

How to do?
Make a dummy file with 5 var Gender, Height, Age, Diet & Parent's
height
Run regression with Height as dep var and Age, diet & parent's height
as indep.
Note which var are important by looking at Standardized coefficients
(Beta) in Coefficients table of Regression output. Var with highest
value of Beta is considered most important.

Now go to Data--> Split
It'll ask for 3 options
1. Analyse all cases, don't create groups
2. Compare groups
3. Organise output by groups

1st option is default setting which doesnt take any split.
2nd option is to compare the results of different groups (e.g. male/
female in our case) but in same table
3rd option is used for obtaining altogether different tables for
males & females

Choose 2nd option & see whether same indep var are important for males
as well as females in predicting dep var height.

CAUTION:
Once split all output will be given on male/female until & unless we
remove split by clicking on 1st option of
Analyse all cases, don't create groups

At times we get wonderful results even in frequency, cross-tabulation
where split gives deep insight into the data.

Regards
Neeraj