[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: trend test of OR in matched case control studies
Sanny Chen wrote:
> How do I perform a trend test of OR in a matched case control study?
> For an unmatched study, TABODDS does the trend test, but what is the
> equivalent command for a matched case control study?
> Thanks in advance,
Wouldn't -symmetry , trend- do what you want?
Otherwise, you could set up the linear component of an orthogonal polynomial
contrast for the predictor using -desmat- and then use -clogit-/-xtlogit-:
char predictor[pzat] orp(1)
desmat: xtlogit response predictor, i(clusterID)
* For searches and help try: