Multiple regression coefficient value 0

Hi,

I am trying to calculate the coefficients through regress function for linear equation with 37 predictors and 58 data points.

10 of 37 coefficients are equal to 0 and the statistics is also not estimated. I was just wondering what is the reason for that and how can I estimate the multiple linear regression by including all coefficient?

Thanks

Kamil

[dpb]

All depends on the design matrix and the data -- it's probable some variables simply don't have changes independent of others such that there is a nonzero derivative for them...but can't say anything specific w/o the data.

How were the design points generated...just happenstance or by some sampling scheme?