Applied Linguistics Tech

Mastering Regression Analysis
in SPSS

A step-by-step guide to modeling relationships between variables, interpreting diagnostics, and ensuring robust research results.

🎯 Learning Goals

  • Learn the step-by-step procedure to perform regression analysis in SPSS.
  • Know how to specify dependent and independent variables correctly.
  • Request additional statistics and plots needed for proper interpretation.
1

Open the Dialog

Navigate through the menus to start your analysis.

Click Analyze → Regression → Linear…

  • Move Dependent Variable to Dependent box.
  • Move Independent Variable(s) to Independent(s) box.

Tip: Start with a simple regression (one variable), then extend to multiple regression.

2
Selection: Enter

This forces all predictors into the model at once. Other methods (Stepwise, Forward) exist, but Enter is best for transparency in standard assignments.

Choose the Method

Select "Enter" as your default method.

3

Request Statistics

Don't rely on the defaults. Get the full picture.

Click the Statistics… button and tick:

Estimates
Model Fit (R²)
Collinearity
Durbin–Watson
4

Scatterplot

Y-axis: ZRESID | X-axis: ZPRED

Normality

Tick Normal probability plot (P-P Plot)

Add Diagnostic Plots

Visual checks for linearity, homoscedasticity, and normality.

5

Save Residuals Optional

Useful for deeper diagnostics or specific assignments.

Click Save... to store standardized residuals in your dataset.

Output Generated:

  • Model Summary R, R²
  • ANOVA Table F-test
  • Coefficients B, Beta, t, p
  • Diagnostics Plots

Run Analysis

Click OK to generate your data.

🧠 Quick Check-in

Test your understanding of the procedure.

1. Which method should you usually start with?

Stepwise
Enter
Backward

2. Which option gives VIF values?

Model Fit
Collinearity diagnostics
Estimates

3. Where do you check normality?

Scatterplot
Normal P-P Plot
Durbin-Watson
Download Handout Hướng dẫn phân tích hồi quy (PDF)