IBM Support

Standardized regression coefficients outside (-1,1)

Troubleshooting


Problem

I have run a linear regression (Analyze->Regression->Linear) with several predictors in SPSS/PASW Statistics. I was surprised to see that the standardized coefficients, labelled BETA, for some predictors had values which exceeded the bounds of (-1,1). Are standardized coefficients outside these bounds valid?

Resolving The Problem

If there is a single predictor or multiple predictors that are uncorrelated, then the Beta values will be confined to the bounds of (-1,1). However, if there are 2 or more predictors that are correlated, positively or negatively, then the Beta values may exceed those bounds.

Deegan, J.Jr. (1978). On the Occurrence of Standardized Regression Coefficients Greater Than One. Educational and Psychological Measurement, 38, No. 4, 873-888.

The topic is discussed in the context of structural equation modeling by Karl Joreskog at http://www.ssicentral.com/lisrel/advancedtopics.html . Click the link for "How Large Can a Standardized Coefficient be?"

[{"Product":{"code":"SSLVMB","label":"SPSS Statistics"},"Business Unit":{"code":"BU001","label":"Analytics Private Cloud"},"Component":"Not Applicable","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"19.0","Edition":""}]

Historical Number

55212

Document Information

Modified date:
16 June 2018

UID

swg21494173