Why were results from PROC MIXED same as PROC GLIMMIX?

Question

I am running a simple code comparing insect abundance on a vegetated and unvegetated surfaces. Design is RCBD. 1 factor with 2 level (vegetated and unvegetated). 10 blocks. Distribution of residuals is non normal and variance of the two populations are unequal. I could log transform but wanted to explore using GLIMMIX instead. I obtain the same p values, standard erros and variance using either Proc mixed or proc glimmix. Any idea why? Anything to do with robustness of proc mixed assumptions? I am a lowly environmental student with 2 semesters of stats under my belt...enough to make me confused! Any ideas appreciated.

data abundancevisit3;
input blk trt$ y;
lines;
1 g 5.867
1 b 0
2 g 6.933
2 b 0.444
3 g 0.8
3 b 0
4 g 5.2
4 b 0.667
5 g 56.267
5 b 1.333
6 g 14.933
6 b 0
7 g 54.133
7 b 0.444
8 g 5.026
8 b 0
9 g 4.8
9 b 1.333
10 g 11.733
10 b 0
;
options nocenter;
proc print data=abundancevisit3;
run;

proc glimmix data=abundance visit3;
CLASS blk trt;
MODEL Y=trt;
random blk;
lsmeans trt;
run;

Answer 1

You're using two different SAS procedures to fit the same model: a linear mixed model with a random intercept for block. There is some overlap between the capabilities of mixed and glimmix.

See SAS support at http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_glimmix_a0000001459.htm for a comparison between the procedures.

Answer 2

I am not sure if this is another way of putting your answer but I just learned about the link function in glimmix...the default distribution in glimmix is gaussian so in order to use a different distribution I need to use the link statment...

Thanks for expedient input.