>I am doing a factor analysis on 14 scale variables. The scale is a likert
>scale from 1 to 10. Iím doing the factor analysis for the purpose of using
>the factors in a regression. When I use the following:
> PROC FACTOR DATA=test METHOD=P PRIORS=ONE ROTATE=V RE SCREE RES;
> var tst1-tst14;
> I get only one factor. Iíve looked through various texts and am not
>sure how to explain all 14 variables going into one factor. The
>questionnaire does not have any classification of the scale questions.
>They are all asked together.
I would recommend that you not do this.
Try using PROC PLS if what you really want is something like this.
PROC PLS does partial least squares regression for you.
For that matter, given the nature of your data, I would be
really concerned that your residual plots and diagnostics might
indicate that ordinary regression is inadequate for this task.
David L. Cassell
3115 NW Norwood Pl.
Corvallis OR 97330
FREE online classifieds from Windows Live Expo Ė buy and sell with people