I'm working on a thesis to determine the effect of independent variables(four independent variables, denoted by X1,X2,X3 and X4) on the dependent variable (one dependent variable, denoted by Y). I used 40 observation data from 10 companies in the period of 4 years.
I'm using linear regression with SPSS 15. When I regress the four independent variables together on the dependent variable, the result is one of the independent variables that are not included in the Coefficient’s table but included in the table of the Excluded Variable (X2 variable).

After I find information on the internet regarding the excluded variable in the SPSS results, some articles / discussions related to excluded variable say it is associated with symptoms of multicollinearity. The solution is to use more data.

Then I added the data observation from the original observational data consist of 40 to 112 (from 28 companies over 4 years). However, the result remains the same, ie there is one independent variable (X2 variable) that is included in the table on the Excluded Variable.

Because it still fails, I try to regress the four independent variables separately or individually so one independent variable regressed with one dependent variable and the result was not one of independent variable included in the table of Excluded Variables although two of it remains not significant.

What I want to ask is whether the way I use was correct ?, where each independent variables regressed separately or individually on the dependent variable, in this case means that there are four regression model.
If that is not the right way then how’s the solution to overcome the problem of the excluded variable?

Thank you .