Data split procedure:
1. click on the 'Data menu'
2. choose 'Split file'
3. click on 'compare groups' and specify your grouping variable (e.g. gender, age etc.)
4. click on 'ok'
Three main steps for regression (if my brain is working properly):
1. Check the assumptions: multicollinearity. Check correlations between dependent and independent variables. There should be some association between the variables, however, it may also be problematic if the correation is too high/low. 'Collinearity diagnostics' needs to be checked as well. The checking of other assumptions is also important, including outliers, independence of residuals and so on.
2. Evaluating the model: Model summary (R2 and adjusted R2). Statistical significance of the regression: ANOVA.
3. Assessing each of the independent variables (Coefficients).
Good luck with your analysis.