I am using VGAM package to model a multinormial outcome with spline
function. But from the following result, I only can get p-value for variable
with spline function, such as s(As). However, there is no p-value available
for variable with linear function, such as SEX, Parity. Does anybody know
why?
> summary(fit1)
Call:
vgam(formula = mr~ s(As) + s(Pb) + SEX + parity,
family = "multinomial", data = paper)
Number of linear predictors: 3
Names of linear predictors: log(mu[,1]/mu[,4]), log(mu[,2]/mu[,4]),
log(mu[,3]/mu[,4])
Dispersion Parameter for multinomial family: 1
Residual Deviance: 3292.768 on 6423.567 degrees of freedom
Log-likelihood: -1646.384 on 6423.567 degrees of freedom
Number of Iterations: 10
DF for Terms and Approximate Chi-squares for Nonparametric Effects
Df Npar Df Npar Chisq P(Chi)
(Intercept):1 1
(Intercept):2 1
(Intercept):3 1
s(As):1 1 2.2 1.5061 0.51558
s(As):2 1 2.8 3.7633 0.26187
s(As):3 1 3.8 1.8600 0.74002
s(Pb):1 1 2.4 7.7796 0.03172
s(Pb):2 1 2.8 8.5664 0.03091
s(Pb):3 1 3.3 5.6044 0.16036
SEX:1 1
SEX:2 1
SEX:3 1
parity:1 1
parity:2 1
parity:3 1
Thanks a lot.
Weichao
[[alternative HTML version deleted]]