thanks a lot for doing this. The problem I think I still have is that I cant base my analysis on threads where all the messages are missing because, as you say, we can't be sure whether they deleted them for management or censorship purposes. I can only use threads where some messages were deleted and others were not - and there appears to be very little evidence of this. What I had wanted to do was pick out all these threads then (if there were lots randomly sample) qualitatively analyse the text of messages to see what the reason was for the deletion and if it was fair. (This could then be cross-checked against standard content analysis rules).

But if I am right, and there are very few of these types of thread then the analysis would, rightly or wrongly, be seen as insignificant.

20371 : 20371 : hey webmaster
32 : M : 20371 : 20395 : RE: hey webmaster
29 : M : 20371 : 20717 : RE: hey webmaster
29 : M : 20371 : 20768 : RE: hey webmaster -1 : M :

Just to clarify, you are saying that all this thread was removed, but that the last message was deleted the same day it was posted which suggests it was censored?

I am gonna start going through this zip in a bit more detail now - I think I had better go get a coffee first though!!!

This journal is worth a look:

http://www.ascusc.org/jcmc/

--
scott