More Questions

I continue to have a cordial correspondence with the Lancet authors. Special thanks to Gilbert Burnham for answering my questions. For those who care, my latest set of questions was as follows.

Hope all is well and thank you for sharing some of the data from Burnham et al (2006) with me.

1) Is data from Roberts et al (2004) available as well? It is a great idea (as you do the the 2006 article) to try and match up data from the two separate studies. I would like to do the same.

2) Is a copy of the interview form available, either in English or Arabic? This may have been posted at some point but I am unable to find a copy. I would like to get to the bottom of the whole not-enough-time-to-do-the-survey dispute (which I don't attach a lot of credence to) by examining just how long the survey is.

3) It is my understanding that none of the interviewers spoke Kurdish. Is that correct?

4) Is a copy of the studies precise methodology available? I am looking for whatever written material were given to the interviewers which specified the exact procedures to be used in selecting clusters, dealing with security issues and the like. It is impossible to evaluate the merit (if any) of the main street bias critique without understanding the steps used by the interviewer teams.

5) In his talk at MIT in February, Dr. Burnham stated (with regard to Roberts et al (2004):

"Now this is what the confidence intervals would look like. There is a 10% probability that it was less than 44,000 and only a 2.5% chance that it was less than 8,000. If we put Falluja into it, the top end of the confidence interval would be infinity. It really skewed things so badly that we decided that we should just leave it out and be conservative."

Is that an accurate statement about the effect of including the Falluja data?

As always, thanks for your time.

Dave Kane

I won't be printing any replies without the express permission of the senders, of course. If any readers know the answers (surely the survey form has been made public somewhere?), please contact me.