Displaying 1 result from an estimated 1 matches for "intercetp".
Did you mean:
intercept
2011 Jul 21
0
gls yields much smaller std. errors with different base for contrasts
...s2
(Intercept)
91.74
1> coefficients(summary(cs1))[1]+coefficients(summary(cs1))[3] #6m:0f
intercept in cs1
(Intercept)
114.195
1> coefficients(summary(cs2))[1] #6m:0f intercept in cs2
(Intercept)
114.195
1> coefficients(summary(cs1))[1]+coefficients(summary(cs1))[4] #6m:6f
intercetp in cs1
(Intercept)
70.24667
1> coefficients(summary(cs2))[1]+coefficients(summary(cs2))[4] #6m:6f
intercetp in cs2
(Intercept)
70.24667
#Curiously, removing the interaction [gls(response~rep +
sex_ratiob...)] gives reasonable numbers for the standard errors
Any suggestions would be gr...