www.cmm.bristol.ac.uk/forum

Hi,
is it possible to run survival analysis with runmlwin? I cannot find any information in the manuals.
Thanks,
Miguel

Yes, it's very much possible to do discrete-time survival analysis with runmlwin.

Hi Paatrick,
Can you elaborate more about the model and the possible codes for discrete time survival analysis. Please share your thoughts for the benefit of the readers in forum.

Have a look the readings and slides (and the Mlwin practical) of Fiona Steele on this website: http://www.ccsr.ac.uk/qmss/summer/Paris09/info4.shtml. But note that for most basic survival analysis applications the Stata capabilities are sufficient, see for instance the books by Blossfeld et al. (2007) and Mills (2011).

Blossfeld, Hans-Peter, Katrin Golsch, and Götz Rohwer. 2007. Event History Analysis with Stata. Lawrence Erlbaum.
Mills, Melinda. 2011. Introducing Survival and Event History Analysis. Sage. doi: 10.4135/9781446268360

Hi,
Thanks for the answer. The material you refer by Fiona seems to be related to MLwin. The question was specifically if it is possible to run a multilevel survival analysis with runmlwin.
Any other suggestion?
Regards,
Miguel

Dear Miguel,

Any model that can be specified in MLwiN can also be specified via runmlwin.

However, we have not yet created any specific discrete time survival analysis examples.

A two-level discrete time survival analysis is basically a two-level logistic regression where periods are nested within persons. The response for each period is a 0 if the person survives (or is right censored) and 1 if the person dies.

So if you look at the examples of how to specify a two-level logit model using runmlwin, you should be able to figure out how to specify your survival analysis model.

Sorry not to be of more help (am in an airport)

George

Thanks, I will try.
Regards,
Miguel