SPSS for newbies: K or K-1 dummy variables for regression? 

That's up to you. Either is fine - only the interpretation of the coefficients are different

Thank you so much Phil, I have the last question is, without intercept what is the meaning of t value, because we will look at the sig first right?

The t-statistic is always the same on output. It tests the individual significance.

+Khamphonemisay PHOMMATHAT With k-1 dummies, the coeff on a dummy tells you the mean change in the DV of that group compared to the reference group. If you use k dummies the coeff now tells you the mean value of the DV for that group.

+Phil Chan (statisticsmentorcom) Thank you , that's a better way for interpretation with K dummies. It is easy to understand the effect of each dummy contribute to dependent variable because I though that the sign ( negative or positive ) of Beta is important , not the intercept ( beta zero ).