--- On Wed, 5/8/09, Maarten buis wrote:
> What I would suggest in that
>case is to do a sensitivity analysis, to see whether these
>potential issues are actually big enough problems to care
>about. I have done one such a study, where the results
>proved to be robust. This is just one study, so I will give
>no guarantees that this is a general outcome. However,
>-seqlogit- has tools to perform just such an analysis, so
>you can easily do such a sensitivity analysis on your own
>model and convince yourself whether there is or is not a
>problem in your case.
Dear mr Buis,
Thank you very much for your reply.
