Hello Javier,

It's not fully clear from your question what you're after if you propose to use a simultaneous (all independent variables entered) approach. What you could do is: (a) run the full model (all IVs in); then (b) run a set of k models, one for each of the k IVs, in which one IV is omitted. The difference in R-squared between (a) and any specific run from (b) will tell you how much explanatory power is added by a specific IV, having first controlled for the other IVs. Those variables for which bigger differences are observed are more influential in explanatory power, for your sample and the set of variables you chose to measure.

Good luck with your work!

Hint: Read about Variance inflation factors (VIF).

good luck.

Please refer to:

1. Landau, S. and Everitt, B. S.(2004). A Handbook of Statistical analyses using SPSS.

2.Howitt, D. and Cramer, D.(2008). Introduction to SPSS.

Regards,

Zuhair

this statistical tool is used when you have a model that with variables dependent and independent in same time and in this case you dont use ordinary least squares (OLS)

I would advice you to start by performing an ordinary least squares (OLS) with each independent variable (IV). This way you will be able to have a preliminary idea as to how each IV relates to the dependent variable. You may now proceed to multiple regression (MR). In most cases, chances are that the IVs are also related in one way or the other, you can test for this by determining the correlation matrix. If this is the case, you now have a problem of multicollinearity and you can quantify it through variance inflation factor (VIF) for each IV. A visual evidence of multicollinearity is in the form of sign reversals and insignificant predictor coefficients in the MR model as compared to under OLS.