[R] Off-topic: looking for a categorical, NMAR dataset

[Andrew Robinson]

Dear Colleagues,

apologies for this off-topic posting.

A Ph.D. student here at U of Melb. is trying to find a dataset to use
to demonstrate a technique that he is developing.  He needs a binary
response and ideally a categorical predictor, although the latter can
of course be induced from a continuous predictor.  The data should
also have missing values (ideally, NMAR, not missing at random) in the
response and in the predictor.

Of course we could generate such a dataset but it would be preferable
to use a dataset in which handling the missingness is an integral part
of the analysis.

The data set could be in any discipline, ideally already published.

If you have any suggestions, please respond directly to Ken at

Kheang Ken Lim <k.lim11 at pgrad.unimelb.edu.au>

Thanks!

Cheers

Andrew

-- 
Andrew Robinson  
Program Manager, ACERA 
Department of Mathematics and Statistics            Tel: +61-3-8344-6410
University of Melbourne, VIC 3010 Australia               (prefer email)
http://www.ms.unimelb.edu.au/~andrewpr              Fax: +61-3-8344-4599
http://www.acera.unimelb.edu.au/