After some prompting from a recent summary post from Glenn Reynolds, John Lott has posted some clarifications. John admitted that the coding errors that Ayres and Donohue pointed out do exist.
Before posting to this list I checked Tim Lambert's weblog to see what his take on this was. Though accepting the fact that Lott has admitted the errors exists he states that Lott doesn't address what the implications for his results are. I have a hunch that that was posted by Tim before Dr. Reynolds posted an update to his most recent Lott Post. In that update Dr. Reynolds links to this:
http://www.johnlott.org/files/statement.html
The second paragraph seems to be Dr. Lott's statement on the impact those coding errors have on the results. Dr. Lott states that the errors have been corrected in the downloadable data sets on his website for those inclined and adept at confirming his statements.
For those of you interested, the two posts from Dr Reynolds in chronological order are: http://www.instapundit.com/archives/011099.php http://www.instapundit.com/archives/011118.php
Tim's post referred to above: http://cgi.cse.unsw.edu.au/~lambert/cgi-bin/blog/2003/08#0821
-Gene Hoffman
