Programming the statistical procedures from SAS

binomial unequal group sizes

Accepted Solution Solved
Reply
Occasional Contributor DDK
Occasional Contributor
Posts: 16
Accepted Solution

binomial unequal group sizes

Dear all,

 

If you have binomial data for which the group sizes differ, how do you correct for it in proc glimmix?

 

For example:

 

proc glimmix data=name;

class var1 var2;

model y/n=var1 var2 / dist=binomial link=logit;

random randomvariable;

run;

 

y is for example the number of affected individuals per group that differs in size (n). I have been told that you can correct for this by adding an overdispersion parameter 'random _residual_;' but I am not too confident about this. Anyone have any ideas? Thanks in advance.


Accepted Solutions
Solution
‎10-16-2015 05:23 PM
Valued Guide
Valued Guide
Posts: 673

Re: binomial unequal group sizes

You don't need to correct for this -- it is automatically taken care of. Do not use the 'random _residual_" statement for this purpose (this is for overdispersion: high variability than one would obtain with a binomial distribution).

View solution in original post


All Replies
Solution
‎10-16-2015 05:23 PM
Valued Guide
Valued Guide
Posts: 673

Re: binomial unequal group sizes

You don't need to correct for this -- it is automatically taken care of. Do not use the 'random _residual_" statement for this purpose (this is for overdispersion: high variability than one would obtain with a binomial distribution).

Super Contributor
Posts: 271

Re: binomial unequal group sizes

Agree. And further, since the random-statement is not neccessary, it may be better to use a procedure (proc logistics) that is specialized for logistic regression.
Occasional Contributor DDK
Occasional Contributor
Posts: 16

Re: binomial unequal group sizes

[ Edited ]

Thanks, thought it was a bit weird to correct for unequal sizes using an overdispersion parameter. Good to hear a confirmation that it is not necessary to do so.

☑ This topic is SOLVED.

Need further help from the community? Please ask a new question.

Discussion stats
  • 3 replies
  • 331 views
  • 5 likes
  • 3 in conversation