. . . . . in 2001 an LAPD document examiner erroneously concluded that the note was written by Nyle Brenner, Berman's manager and friend. . . . . .document examiner William Leaver concluded that it was "highly probable" that Brenner wrote the note, eight months before he analyzed Durst's handwriting. . . . . . Leaver changed his conclusion in June 2003, writing that it was "probable" that Durst wrote the murderer's note to police, and "highly probable" that Brenner wasn't the author. . . . . Durst's attorney, Dick DeGuerin, said the conflicting handwriting analysis just "shows you what junk science really is."
Can anybody explain what went wrong in this examination?
"Man was born free, but he is everywhere in chains." -- Jean-Jacques Rousseau
I should start by saying I'm not a document examiner; although I do have some basic knowledge of the discipline. From the article you provided I'm not sure anything "went wrong" with the initial examination. It sounds like the examiner was first only given Berman's handwriting samples to compare to the note. His conclusion of 'highly probable' isn't the same as an individualization so he never conclusively reported the Berman was the one that wrote the note. Document examiners have varying levels of conclusions and highly probable falls just short of an identification. One definition I have found for highly probable is "the evidence is very persuasive, yet some critical feature or quality is missing so that an identification is not in order;" from the Standard Terminology for Expressing Conclusions of Forensic Document Examiners at afde.org. After the examiner was given Durst's handwriting sample to compare he concluded that it was probable that the note was written by Durst and highly probable that Brennen did not. This is not unusual in science. Given more information it's not uncommon for a conclusion to change as our conclusions should rest on the information available to us at the time of our examination. For example, say you are doing a comparison where you have about 6-7 features in agreement with no discrepancies/dissimilarities but the known wasn't rolled properly so there is more information in the latent than in the known. If that is the only card on file you may decide to conclude either that the comparison is Inconclusive or that you need better exemplars. A couple weeks later, you are given additional exemplars for comparison that have been rolled properly. Once you have the area you were missing you see that there are very clear discrepancies that could not be caused by distortion. I would imagine, your conclusion would then change to an exclusion given this additional information. Does that mean that you made an error in your initial examination because you didn't originally exclude the individual?
What I find more concerning in the article is that the Supervisor says she rubber stamped the examiner's original conclusion. Did she say that because she truly didn't scrutinize the examiner's work and failed in her duty as a verifier? Or did she say that to throw the examiner under the bus so to speak because the media has made this out to be an error and she doesn't want people to think she also made an error?
My understanding from other document examiners is that his initial conclusion of 'highly probable' was inappropriate. At least part of the reason is that the unknown writing sample was block printing.
Thanks for the reply. I thought the block printing may have been an issue as some of those letters are similar to my own when I write in all caps. It would be interesting to see Brenner's handwriting sample. I read a couple articles on this case but none have shown his sample.