Interpreting Multiple Regression Output in SPSS
Running a linear regression in SPSS is straightforward, but the output contains several tables, and it is not always clear which numbers matter. In this blog, we'll go through the main tables in order and explain what to look for and what to report.
Running the Analysis:
- Go to Analyze > Regression > Linear.
- Move your outcome variable into Dependent and your predictors into Independent(s).
- Under Statistics, tick Estimates, Model fit and Collinearity diagnostics. Under Plots, you can request residual plots to check assumptions.
Model Summary Table:
- R is the correlation between the observed and predicted values of the outcome.
- R Square shows the proportion of variance in the outcome explained by the predictors together. For example, 0.35 means 35%.
- Adjusted R Square corrects for the number of predictors and is often preferred when comparing models.
ANOVA Table:
- This tests whether the model as a whole predicts the outcome better than no model at all.
- Report the F value, the degrees of freedom and the significance (p) value. A significant result, commonly p < .05, indicates that the model is useful.
Coefficients Table:
- Unstandardised B shows how much the outcome changes for a one-unit increase in the predictor, holding other predictors constant.
- Standardised Beta allows you to compare the relative strength of predictors measured on different scales.
- The t value and Sig. column show whether each predictor makes a significant unique contribution.
- Collinearity statistics (Tolerance and VIF) help you check whether predictors are too highly related to each other. Very high VIF values signal a problem, although guidelines on the exact cut-off differ.
Checking Assumptions:
- Look at residual plots for signs of non-linearity or unequal variance.
- Check that residuals are approximately normally distributed.
- Look for influential cases that may be distorting the results.
Reporting the Results:
- Report R Square (or adjusted R Square), the F test, and for each predictor the B, Beta and significance.
- Interpret the results in plain language, for example: "Workload was a significant positive predictor of stress, after controlling for age and experience."
- Avoid causal language unless your design supports it.
The key to reading regression output is to work through it in order: model fit, overall significance, then individual predictors, with assumption checks alongside. Report the main statistics clearly and explain what they mean for your research question.