Interesting new study and guidelines
The distance between the actual predictive and explanatory power of common CSI procedures and the reality is huge.
It is interesting how much we take things on faith if they are supposedly to be “scientific”. I wonder how many other things we take on faith as scientifically valid without any real world evidence of rigor.
Yep. Even fingerprints, when you have a “partial” is more art than science. They will “match” to say a number of points, but that matching in some cases is subjective, and if the Police say they think the prints match, the tech is pressured to find a match.
It is really upsetting, and I hope that those wrongfully imprisoned on junk science are set free, and better procedures are set up.
Also, it’s upsetting that some people aren’t willing to look at the science:
Scientists should listen to criticism and review evidence and accept if things need to change or ignore or rebut the evidence if it’s incorrect, not rabidly defend and lash out at critics.
The misuse of science here shouldn’t make you doubt science everywhere. But you should be aware of how people can misuse science to further their agenda, either knowingly or unknowingly.
Good point. I wonder why there haven’t been lawyers challenge fingerprint analysis in court by, for example, asking the expert to identify a latent print to a collection of 10 (with the possibility that the source print isn’t even in the lot of 10).
It also reminds me of the TV show, CSI, where, perhaps in the first season, they ID’d a bullet to a specific batch of ammunition based on the precise chemical analysis of elements in the bullet. Then, many years later, they had an episode where they reviewed an old case and suspected the original forensics to have been biased because it was based on ID’ing the ammunition based on the precise chemical analysis of the bullet used in the crime, saying the forensic team at the time should have known such comparisons were bogus, citing a scientific paper that showed that such matching is unreliable. Funny thing was, the date of the cited paper was at least a decade before the first episode.
And then there are all the people who were convicted on “hair analysis”, which has been proven to be even less reliable then an eye-witness identifying the color, make, and model of an automobile involved in a late-night hit and run.
Thank god graphology holds up. :rolleyes:
Is fingerprint matching really not a science these days? FIngerprint recognition on phones seems to work reliably
That’s not fingerprint analysis – that is fingerprint identification (matching)
Comparing a fingerprint or partial print to a known sample and identifying it as “Yes, a match” or “No, not matched” is much easier.
There has been a major and IMO needed correction wrt forensics and its use in the criminal justice system. In the mid 2000s you literally just had to say “science” and the courts were sold. Its an exaggeration, but not by much.
I do hesitate before declaring something “junk.science” though. The question should be, “does this technique provide data which is probative of a fact in issue”. A bite mark might not be sufficiently sensitive to identify someone cold, but it can be important corroborating evidence. For instance a bite mark showing an individual had a chipped tooth can be employed to confirm or refute an identification.
Tha’s a bad example, because bite marks can’t show that the biter had or did not have a chipped tooth. Bite mark analysis is just NFG.
I did wonder about the shoe-print evidence though. There is no evidence that shoe-print evidence is ‘scientific’. Is that evidence that shoe-print evidence is invalid? You might have lots of very clear shoe-, prints, in a surface that does take and retrain shoe-prints, and it seems like it would be as good as eye-witness testimony. That is, not very good at all, but not worthless. Perhaps the only objection is that it is sold to the jury ad definitive, (the same way eye-witness testimony sometimes is).
The Myth of Fingerprints
What I’m starting to wonder is, if point-based fingerprint matching is so inexact, why don’t defense attorneys (with truly innocent clients*) make a big blowup of the defendant’s print, and a big blowup of the evidence print, and ask the tech to match them up? If they really don’t match, it should be obvious, and then the whole point-system should fall out of use.
I used to just assume that you used the computer to find probable matches in the system, but then the techs had to actually compare the partials to the system prints, by eye, to be sure. Maybe that’s not true?
*Maybe that’s the true myth - an innocent defendant! 
These studies that conclude “DNA is the only legitimate forensic science” irk me. Everyone forgets that toxicology and drug chemistry are forensic sciences. Toxicology is the forensic discipline presented in court most (OWIs anybody?). Drug chemists identify components in pills, powders and plant material by chromatographic and spectroscopic methods. Both of these are mature scientific disciplines and have been around since antiquity. The 2009 NAS report itself stated as much. And there’s a reason why you haven’t heard many complaints about these (other than about individual poor practitioners) - it’s because nobody, not even the defense bar, disputes that toxicology and chemistry are actual, you know, science!
So go ahead and bash latent prints, questioned documents and trace evidence analysis if you must. Just don’t say “DNA is the only real forensic science.” It ain’t.