Use of bma/wals on binary dependent variable

[Shamsuddeen Adamu]

Hi there...I have a binary dependent variable (use foxtlogit analysis)...I am now at a point of considering model selection (and perhaps doing a robustness check) and I am wondering if I can use BMA/WALS on my binary dependent variable panel...ll appreciate the help. Thanks

[Maarten Buis]

In your previous post you did not get an answer that satisfied you because you used a non-standard model and did not manage to explain how that gave you trouble implementing standard Bayesian Model Averging. This post does not improve on that. So given this history I will give this only one try.

1) What is -foxlogit-?

2) Give us an extract of your data or some random data with the same structure as your data. Use -dataex- as discussed in the FAQ (In the top black bar on the top of this page there is a link to the FAQ)

3) What have you tried thus far? Use the extract mentioned above to give use code that shows what you have done. This way you can much more precisely convey what you have done.

4) Why do you think that what you have tried is not succesfull?