Understanding through Discussion


Welcome! You are not logged in. [ Login ]
EvC Forum active members: 85 (8925 total)
Current session began: 
Page Loaded: 08-20-2019 5:34 AM
18 online now:
Dredge, PaulK (2 members, 16 visitors)
Chatting now:  Chat room empty
Newest Member: Jedothek
Post Volume:
Total: 860,059 Year: 15,095/19,786 Month: 1,818/3,058 Week: 192/404 Day: 6/73 Hour: 1/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
1
23Next
Author Topic:   Is Fingerprint Analysis a Science?
Trixie
Member (Idle past 1933 days)
Posts: 1011
From: Edinburgh
Joined: 01-03-2004


Message 1 of 38 (644098)
12-15-2011 6:14 AM


The Scottish Fingerprint Inquiry reported yesterday and one of it's findings was that the identification of fingerprints was opinion, not fact and suggests that it can't be considered scientific.

For those of you who know nothing about this case, briefly, a fingerprint at a murder scene was identified as belonging to a police officer attached to the case. Said police officer denied ever having been in the murder house. On this basis, the officer was tried for perjury, but cleared when experts from the USA discovered the fingerprint never belonged to the officer in the first place and a terrible mistake had be made. Meanwhile, David Asbury was sentenced to life in prison for murder on the basis of yet another erroneous fingerprint identification in the same case!!!!!

The Inquiry has determined that the police officer did not make the fingerprint in question and that the fingerprint which led to Asbury being convicted was similarly misidentified.

The whole debacle took place in the 1990s, but we're only getting it cleared up now. The finding of the Inquiry that fingerprint identification is not science, but opinion, and should be stated as such in court proceedings will raise howls of indignation from fingerprint experts around the world. It's akin to declaring that DNA analysis is not science because someone made a mistake.

So where do we draw the line between science and opinion? I've always held that in writing research papers, the only place for opinion is in the discussion of the results and said opinion has to be backed up by data from the paper or previously published data.

Coffee house probably, but not very sure since it's nowt to do with evolution or creation, but does concern the validity of science in general.

Edited by Admin, : Change title.


Replies to this message:
 Message 3 by Percy, posted 12-15-2011 7:03 AM Trixie has responded
 Message 5 by Straggler, posted 12-15-2011 8:38 AM Trixie has not yet responded

  
Admin
Director
Posts: 12618
From: EvC Forum
Joined: 06-14-2002
Member Rating: 3.1


Message 2 of 38 (644100)
12-15-2011 6:43 AM


Thread Copied from Proposed New Topics Forum
Thread copied here from the Science or opinion? thread in the Proposed New Topics forum.
    
Percy
Member
Posts: 18801
From: New Hampshire
Joined: 12-23-2000
Member Rating: 2.2


Message 3 of 38 (644101)
12-15-2011 7:03 AM
Reply to: Message 1 by Trixie
12-15-2011 6:14 AM


I've heard rumblings over the past decade or so that fingerprint analysis isn't as reliable as is commonly thought, but I haven't paid much attention. I just took a look at the Wikipedia article on fingerprints and found this quote from an editor of a fingerprint periodical:

Sandy L. Zabell writes:

In 1995, the Collaborative Testing Service (CTS) administered a proficiency test that, for the first time, was “designed, assembled, and reviewed” by the International Association for Identification (IAI).The results were disappointing. Four suspect cards with prints of all ten fingers were provided together with seven latents. Of 156 people taking the test, only 68 (44%) correctly classified all seven latents. Overall, the tests contained a total of 48 incorrect identifications.

This doesn't seem surprising where people are involved. Latent prints and especially partials probably leave a lot of room for human interpretation.

--Percy


This message is a reply to:
 Message 1 by Trixie, posted 12-15-2011 6:14 AM Trixie has responded

Replies to this message:
 Message 4 by Trixie, posted 12-15-2011 8:08 AM Percy has acknowledged this reply
 Message 6 by Jon, posted 12-15-2011 9:01 AM Percy has acknowledged this reply
 Message 7 by NoNukes, posted 12-15-2011 10:51 AM Percy has acknowledged this reply

    
Trixie
Member (Idle past 1933 days)
Posts: 1011
From: Edinburgh
Joined: 01-03-2004


Message 4 of 38 (644102)
12-15-2011 8:08 AM
Reply to: Message 3 by Percy
12-15-2011 7:03 AM


The Inquiry report goes into detail with regard to confirmation bias and circular reasoning. It's very interesting to see how this sort of thing manifests itself. I see a sort of equivalence in carrying out scientiic experiments without the inclusion of adequate controls. For example, the fingerprint examiners (FE) supposedly worked to a 16 point standard, i.e., they had to find 16 points in agreement and sequence and no unexplained differences before they could call it a identification. However they admitted that sometimes by the time they had found 7 or 8 points in agreement, they had alreay made up their minds and any differences they subsequently encountered must have an explanation, even if that explanation was unknown.

A case was also made by one of them that he would examine only a target area and if the target area had 16 points in agreement he would disregard any differences outwith the target area. Now that worries me greatly because it doesn't matter how many similarities there are, a single, obvious difference precludes identity. That's exactly the same as evidence supporting a scientific theory - it doesn't matter how much corroborative evidence you have, a single piece of evidence contradicting that theory is enough to kill the theory.

My contention is that fingerprint identification is science if it is practiced scientifically. What happened in the McKie case happened because the FE lost sight of that basic requirement of science. Declaring that all fingerprint identification is not science is akin to throwing the baby out with the bathwater.


This message is a reply to:
 Message 3 by Percy, posted 12-15-2011 7:03 AM Percy has acknowledged this reply

  
Straggler
Member
Posts: 10285
From: London England
Joined: 09-30-2006


Message 5 of 38 (644107)
12-15-2011 8:38 AM
Reply to: Message 1 by Trixie
12-15-2011 6:14 AM


Scientific Conclusions and Opinions
Trix writes:

So where do we draw the line between science and opinion?

Well (as a starting point) scientific conclusions are necessarily based on scientific evidence whilst opinions have no such restriction. One can have an opinion that is based on evidence, but one can equally hold an opinion that isn't based on evidence or even an opinion which contradicts evidence.

Conclusions based on fingerprint analysis would seem to qualify as scientific conclusions rather than opinions on this basis. The fact that the analysis was done poorly in this particular example shouldn't detract from that.


This message is a reply to:
 Message 1 by Trixie, posted 12-15-2011 6:14 AM Trixie has not yet responded

  
Jon
Inactive Member


Message 6 of 38 (644108)
12-15-2011 9:01 AM
Reply to: Message 3 by Percy
12-15-2011 7:03 AM


Sandy L. Zabell writes:

In 1995, the Collaborative Testing Service (CTS) administered a proficiency test that, for the first time, was “designed, assembled, and reviewed” by the International Association for Identification (IAI).The results were disappointing. Four suspect cards with prints of all ten fingers were provided together with seven latents. Of 156 people taking the test, only 68 (44%) correctly classified all seven latents. Overall, the tests contained a total of 48 incorrect identifications.

This doesn't seem surprising where people are involved. Latent prints and especially partials probably leave a lot of room for human interpretation.

According to Wiki, polygraphs are even said to have a better accuracy rate, at about 61%:

quote:
Wikipedia on Polygraph:

Polygraphy has little credibility among scientists. Despite claims of 90-95% validity by polygraph advocates, and 95-100% by businesses providing polygraph services, critics maintain that rather than a "test", the method amounts to an inherently unstandardizable interrogation technique whose accuracy cannot be established. A 1997 survey of 421 psychologists estimated the test's average accuracy at about 61%, a little better than chance.


It's disturbing that such horribly inaccurate identification methods are allowed in court, where they can be used to completely alter another person's life.

Jon

Edited by Jon, : No reason given.


Love your enemies!

This message is a reply to:
 Message 3 by Percy, posted 12-15-2011 7:03 AM Percy has acknowledged this reply

Replies to this message:
 Message 9 by Wounded King, posted 12-15-2011 11:33 AM Jon has acknowledged this reply
 Message 10 by Rahvin, posted 12-15-2011 12:17 PM Jon has responded
 Message 17 by caffeine, posted 12-16-2011 4:19 AM Jon has not yet responded

  
NoNukes
Inactive Member


Message 7 of 38 (644119)
12-15-2011 10:51 AM
Reply to: Message 3 by Percy
12-15-2011 7:03 AM


Of 156 people taking the test, only 68 (44%) correctly classified all seven latents. Overall, the tests contained a total of 48 incorrect identifications.

This seems to be a carefully worded conclusion, and I can certainly imagine some situations where a large number of reasonably good technicians would fail to get a perfect score on a test of this sort, while the percentage of correct identifications of finger prints would still be quite high.

Let's also recall that in an adversarial proceeding, we would expect sloppy fingerprint determinations to be challenged by other fingerprint experts.

Edited by NoNukes, : No reason given.


This message is a reply to:
 Message 3 by Percy, posted 12-15-2011 7:03 AM Percy has acknowledged this reply

Replies to this message:
 Message 8 by Dr Adequate, posted 12-15-2011 11:24 AM NoNukes has acknowledged this reply

  
Dr Adequate
Member
Posts: 16097
Joined: 07-20-2006


(5)
Message 8 of 38 (644124)
12-15-2011 11:24 AM
Reply to: Message 7 by NoNukes
12-15-2011 10:51 AM


Let's also recall that in an adversarial proceeding, we would expect sloppy fingerprint determinations to be challenged by other fingerprint experts.

... if you can afford them.


This message is a reply to:
 Message 7 by NoNukes, posted 12-15-2011 10:51 AM NoNukes has acknowledged this reply

  
Wounded King
Member (Idle past 2322 days)
Posts: 4149
From: Edinburgh, Scotland
Joined: 04-09-2003


Message 9 of 38 (644125)
12-15-2011 11:33 AM
Reply to: Message 6 by Jon
12-15-2011 9:01 AM


According to Wiki, polygraphs are even said to have a better accuracy rate, at about 61%

I think your estimation of the accuracy rate for fingerprinting is off, 156 people were to identify 7 fingerprints so there were presumably ~1092 identifications potentially made, of which 48 were incorrect. Only 68 of the respondents correctly identified all the samples, meaning that 88 respondents didn't, but since only 48 incorrect identifications were made the remaining ones are presumably instances where the respondent chose not to make an identification.

There are no figures provided in Percy's quote for how many times the respondents who didn't get all 7 right declined to make an identification, but unless it is very high the chances are slim that the accuracy was below 61%.

CTS appear to run tests like this on a regular basis, see here, and several of these have results where ~90% of the respondents got all of the identifications correct. So perhaps the first year was just an outlier.

TTFN,

WK


This message is a reply to:
 Message 6 by Jon, posted 12-15-2011 9:01 AM Jon has acknowledged this reply

    
Rahvin
Member
Posts: 3964
Joined: 07-01-2005


(1)
Message 10 of 38 (644129)
12-15-2011 12:17 PM
Reply to: Message 6 by Jon
12-15-2011 9:01 AM


It's disturbing that such horribly inaccurate identification methods are allowed in court, where they can be used to completely alter another person's life.

Eyewitness testimony is roughly as bad.

Trial by jury sounds like a great idea, but frankly juries terrify me because the average person will often weigh the least reliable evidence as the most damning. Having been on a jury once served only to make that fear worse.


“The human understanding when it has once adopted an opinion (either as being the received opinion or as being agreeable to itself) draws all things else to support and agree with it.”
- Francis Bacon

"There are two novels that can change a bookish fourteen-year old's life: The Lord of the Rings and Atlas Shrugged. One is a childish fantasy that often engenders a lifelong obsession with its unbelievable heroes, leading to an emotionally stunted, socially crippled adulthood, unable to deal with the real world. The other, of course, involves orcs." - John Rogers


This message is a reply to:
 Message 6 by Jon, posted 12-15-2011 9:01 AM Jon has responded

Replies to this message:
 Message 11 by NoNukes, posted 12-15-2011 1:12 PM Rahvin has not yet responded
 Message 12 by Jon, posted 12-15-2011 1:27 PM Rahvin has not yet responded

  
NoNukes
Inactive Member


Message 11 of 38 (644138)
12-15-2011 1:12 PM
Reply to: Message 10 by Rahvin
12-15-2011 12:17 PM


Eyewitness testimony as a whole is far less accurate than is circumstantial evidence.

In my experience, fingerprint evidence has been extremely reliable. The implication that it is unreliable 44 per cent of the time appears to be a flawed one. Almost every bad finger print ID I've ever read about was based on fraud rather than scientific error.

Dr. Adequate points out the truth that it can be expensive to challenge finger print evidence. I can agree with that, but generally, indigent defendants can force the state to foot the bill for challenging finger print evidence.


This message is a reply to:
 Message 10 by Rahvin, posted 12-15-2011 12:17 PM Rahvin has not yet responded

  
Jon
Inactive Member


Message 12 of 38 (644143)
12-15-2011 1:27 PM
Reply to: Message 10 by Rahvin
12-15-2011 12:17 PM


Eyewitness testimony is roughly as bad.

And I don't think eyewitness testimony should be allowed in court. There was, I think, a thread about a similar topic on the short-lived spin-off forum Politicus Maximus... but the Wayback Machine doesn't seem to recognize the name.

But the issue here, I guess, is the status of fingerprint analysis as a science or not.

I think the whole thing is a bunch of match-up and puzzle-doing; more of a hobby, really, and so even if practiced scientifically, probably not overly reliable since it depends on the abilities of the analysts more than on some rigorous standards that are harder to screw up.

Jon


Love your enemies!

This message is a reply to:
 Message 10 by Rahvin, posted 12-15-2011 12:17 PM Rahvin has not yet responded

Replies to this message:
 Message 13 by Straggler, posted 12-15-2011 1:40 PM Jon has responded
 Message 29 by Perdition, posted 12-16-2011 4:39 PM Jon has not yet responded

  
Straggler
Member
Posts: 10285
From: London England
Joined: 09-30-2006


Message 13 of 38 (644144)
12-15-2011 1:40 PM
Reply to: Message 12 by Jon
12-15-2011 1:27 PM


Jon writes:

I think the whole thing is a bunch of match-up and puzzle-doing; more of a hobby, really, and so even if practiced scientifically, probably not overly reliable since it depends on the abilities of the analysts more than on some rigorous standards that are harder to screw up.

Surely this is just a question of technology. A sufficiently advanced pattern matching algorithm could surely allow computers to do all the matching required and to rigorously determine the statistical likelihood of error based on the quality of any given fingerprint.

If the premise of fingerprint matching is sound but human involvement in matching is the failing here then that can be technologically overcome.


This message is a reply to:
 Message 12 by Jon, posted 12-15-2011 1:27 PM Jon has responded

Replies to this message:
 Message 14 by Rahvin, posted 12-15-2011 2:16 PM Straggler has responded
 Message 15 by Jon, posted 12-15-2011 2:41 PM Straggler has responded

  
Rahvin
Member
Posts: 3964
Joined: 07-01-2005


Message 14 of 38 (644145)
12-15-2011 2:16 PM
Reply to: Message 13 by Straggler
12-15-2011 1:40 PM


Surely this is just a question of technology. A sufficiently advanced pattern matching algorithm could surely allow computers to do all the matching required and to rigorously determine the statistical likelihood of error based on the quality of any given fingerprint.

If the premise of fingerprint matching is sound but human involvement in matching is the failing here then that can be technologically overcome.

I wonder at the actual training received by the fingerprint analysts, and whether they themselves are actual trained scientists.

But I think the real problem is how fingerprints are presented to juries. Like with DNA, a typical jury member is going to just say "well, his prints were there, so he did it," without understanding the relative probabilities involved when dealing with fingerprint analysis. The print card may have a nice set of prints, but the prints you leave on an object through normal contact aren't nearly so clear. When the jury hears "it was a match," I'm not sure that they are actually told how much of a match the prints are.

A positive match on 25% of a thumbprint is useful information, but it's not as significant as a match to multiple different full prints ( > 80% latents for both thumb and forefinger, for example). I don't think the average jury member thinks along those lines unless the defense has an expert who specifically directs them that way.


“The human understanding when it has once adopted an opinion (either as being the received opinion or as being agreeable to itself) draws all things else to support and agree with it.”
- Francis Bacon

"There are two novels that can change a bookish fourteen-year old's life: The Lord of the Rings and Atlas Shrugged. One is a childish fantasy that often engenders a lifelong obsession with its unbelievable heroes, leading to an emotionally stunted, socially crippled adulthood, unable to deal with the real world. The other, of course, involves orcs." - John Rogers


This message is a reply to:
 Message 13 by Straggler, posted 12-15-2011 1:40 PM Straggler has responded

Replies to this message:
 Message 16 by NoNukes, posted 12-15-2011 10:46 PM Rahvin has not yet responded
 Message 18 by Trixie, posted 12-16-2011 5:50 AM Rahvin has not yet responded
 Message 20 by Straggler, posted 12-16-2011 8:10 AM Rahvin has not yet responded

  
Jon
Inactive Member


Message 15 of 38 (644146)
12-15-2011 2:41 PM
Reply to: Message 13 by Straggler
12-15-2011 1:40 PM


A sufficiently advanced pattern matching algorithm could surely allow computers to do all the matching required and to rigorously determine the statistical likelihood of error based on the quality of any given fingerprint.

Perhaps. As I said, though, even if we find a way to practice fingerprint analysis scientifically, can we ever consider the analysts to be reliable?

I brought up polygraphs earlier. Reading the Wiki article further, we find this:

quote:
Wikipedia on Polygraph:

In the 1998 Supreme Court case, United States v. Scheffer, the majority stated that "There is simply no consensus that polygraph evidence is reliable" and "Unlike other expert witnesses who testify about factual matters outside the jurors' knowledge, such as the analysis of fingerprints, ballistics, or DNA found at a crime scene, a polygraph expert can supply the jury only with another opinion..."


How does the testimony of finger print analysts compare?

Even when we make the practice scientific, how reliable, really, are the 'experts' giving the testimony?

Jon


Love your enemies!

This message is a reply to:
 Message 13 by Straggler, posted 12-15-2011 1:40 PM Straggler has responded

Replies to this message:
 Message 19 by Straggler, posted 12-16-2011 8:02 AM Jon has responded

  
1
23Next
Newer Topic | Older Topic
Jump to:


Copyright 2001-2018 by EvC Forum, All Rights Reserved

™ Version 4.0 Beta
Innovative software from Qwixotic © 2019