How to calculate SE from categorical covariates?

[karlo]

Hi,
When we parametrize using the number of categories minus 1 colum of covariates whe get the beta values relative to the "reference" one. So, the estimation of one beta is the sum of its value plus the "reference" category. Is the standard error calculate on the same way?

eg. seasonal detection
(matrix)
P[1-1] 1 claro 1 sub
P[1-2] 1 claro 1 sub
P[2-1] 1 claro 0 sub
P[2-2] 1 claro 0 sub

(estimates)
D1 :detection P[1-1] 1.690528 (0.463752)
D2 :detection P[1-1]claro 1.015061 (0.258583)
D3 :detection P[1-1] -0.921363 (0.302432)
D4 :detection P[1-1]sub -0.511395 (0.159412)

(result)
season 1 beta0=1,69 + -0,92
season 2 beta0=1,69

is the SE of season 1 beta0 the sum of SEs?

thanks,
karlo