R help - Jan 2014 - MuMIn Z-Test

Dear list, 
    
    I'm using MuMIn for model averaging and I had a question about the   
z-test that MuMIn performs when using the summary call. e.g.:

    > model.1 <- subset(model.sel(model.1.glmm.list, rank = AICc),
delta<3)
    > summary(model.avg(model.1))
    > 
    > ...
    > 
    > Model-averaged coefficients: 
    >                       Estimate Std. Error Adjusted SE z value
Pr(>|z|)
    > 0|1                   0.094832   0.358649    0.360952   0.263 0.792760
    > 1|2                   1.193099   0.366761    0.369124   3.232 0.001228
**
    > age                   1.163131   0.416735    0.419355   2.774 0.005544
**
    > distribution          0.063250   0.321025    0.323071   0.196 0.844784
    > left                  0.158854   0.247816    0.249294   0.637 0.523985
    > right                -1.317514   0.328174    0.330271   3.989 6.63e-05
***
    > income                 0.705691   0.190689    0.191843   3.678 0.000235
***
    > population           -1.026989   0.319914    0.321969   3.190 0.001424
**
    > weight               -0.178042   0.135962    0.136761   1.302 0.192967
    > 
    > Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 


    What exactly is the z-test testing? I read a post saying that p<0.05
means the confidence intervals    don't span zero, however you can get the
confidence intervals using    the confint() call. What is the z-test telling me
that the    confidence intervals are not?
    
    Also - the MuMIn vignette says that the p-value assumes a normal    error
distribution. A normal distribution of what? The models used    in model
averaging? Or the original data?
    
    Any clarification would be much appreciated. 
    
    Tom
    
 
	[[alternative HTML version deleted]]