Unit 4 · Relationships

How to run a multiple regression in SPSS

Interpret multiple regression output in SPSS: Adjusted R Square, Beta weights, significance and VIF.

Pro lessonAbout 5 minutes7 practice questions

Multiple regression output

Add several predictors to the Independent(s) box and SPSS tests them together and one by one. Here: revision hours, attendance (%) and anxiety.

Output1 [Document1] – SPSS Statistics Viewer

Output

Log

Regression

Model Summary

ANOVA

Coefficients

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	.565	.319	.3021	9.64

Predictors: (Constant), anxiety, hours, attendance

Model	B	Std. Error	Beta2	t	Sig.	VIF5
(Constant)	38.99	6.99		5.58	.000
hours	1.33	0.25	.414	5.32	.000	1.03
attendance	0.09	0.08	.091	1.15	.2533	1.06
anxiety	−0.644	0.15	−.338	−4.31	.000	1.05

Dependent Variable: score

1
Adjusted R SquareR Square always rises when you add predictors, even useless ones. Adjusted R Square corrects for that. Report both.
2
BetaCompare predictors on the same scale. The biggest absolute Beta is the strongest unique predictor.
3
Sig. for each predictorShows whether that predictor adds anything once the others are controlled. Attendance (.253) doesn't.
4
Negative BEach extra point of anxiety predicts 0.64 fewer exam points, holding the other predictors constant.
5
VIFFrom Statistics… › Collinearity diagnostics. Values near 1 are fine. Above 10 (some say 5) means predictors overlap too much.

7 interactive practice questions are part of Pro, along with every other unit and unlimited hearts.