Proc Mianalyze: obtaining combined results for categorical variable (glm/ ANOVA) ?

[SteveDenham]

I believe you will have to open a ticket with Technical Support.  This seems to be an ongoing problem that no one has yet come up with an answer for.

Steve Denham

[TZaihra]

okay thanks a lot Steve.

[rutgers]

Hi TZaihra,

i was wondering if you found a solution to the problem below. i am stuck with the same.

regards

akriti

[Anky_Sri]

I am facing a similar prob. Please let me know if someone was able to get this right and how.

Please the attached code for proc mixed and proc mianalyze and let me know how to fix it. I am using EG 6.1

I am getting the following error:

ERROR: Within-imputation Estimate missing for effect Routine_Technological_Problems in _Imputation_= 1 in the input PARMS= data set.

[BarbClothier]

Not sure if this will help various people on this thread, but I've had some success by looking carefully at the parms and covb datasets that get outputted and making sure all row and column names are exactly named the same (case sensitive) and I believe at times, I've had to manipulate those data sets even more to make sure reference groups and etc... are what they should be.  I vaguely recall at times needing to delete reference group rows that had 0 or missing in the cells??  Be sure things are sorted by _Imputation_ also.  At times, I've had to shorten the variable name also.  Hope this is somewhat helpful.