My regression has me confused!

#1
Can someone help me explain this?

I did a stepwise multiple regression using the three factors. In Step 3, it shows the Bs in the order I expected, but the Betas come out weighted differently, showing Factor Y to be most important. My questions…

1. Why is this?

2. How would I report Step 3? Do I say “When all the significant variables were incorporated and the results standardized, Factor Y was the most important.”

Many thanks!



B Std. Error Beta
Step 1
Factor X 0.68 0.08 0.52*

Step 2
Factor X 0.58 0.07 0.44*
Factor Y 0.33 0.06 0.35*

Step 3
Factor X 0.44 0.08 0.34*
Factor Y 0.38 0.05 0.39*
Factor Z 0.31 0.08 0.24*

Note. R2 = .268 for Step 1; ∆R2 = .114 for Step 2; ∆R2 = .048 for Step 3 (ps<.001). *p<.001
 

JohnM

TS Contributor
#2
This can be caused by multi-collinearity - in other words, the factors are correlated with each other to some degree (they're not totally independent of each other, like you would see in a designed experiment).
 
#3
Thanks and a further question...

I appreciate the info on multicollinearity. What would be the best way to explain this in the text?

Thanks again!