Study Buried For Four Years Shows Crime Lab DNA Testing Is Severely Flawed
from the doing-what-works-rather-than-doing-it-right dept
DNA is supposed to be the gold standard of evidence. Supposedly so distinct it would be impossible to convict the wrong person, yet DNA evidence has been given far more credit than it's earned.
Part of the problem is that it's indecipherable to laypeople. That has allowed crime lab technicians to testify to a level of certainty that's not backed by the data. Another, much larger problem is the testing itself. It searches for DNA matches in samples covered with unrelated DNA. Contamination is all but assured. In one stunning example of DNA testing's flaws, European law enforcement spent years chasing a nonexistent serial killer whose DNA was scattered across several crime scenes before coming to the realization the DNA officers kept finding belonged to the person packaging the testing swabs used by investigators.
The reputation of DNA testing remains mostly untainted, rose-tinted by the mental imagery of white-coated techs working in spotless labs to deliver justice, surrounded by all sorts of science stuff and high-powered computers. In reality, testing methods vary greatly from crime lab to crime lab, as do the standards for declaring a match. People lose their freedom thanks to inexact science and careless handling of samples. And it happens far more frequently than anyone involved in crime lab testing would like you to believe.
An op-ed about the failures of crime lab DNA testing at the New York Times -- written by Boise State Professor of Biology Greg Hampikian -- discusses this ongoing problem using some science of his own: a recently-released NIST study. (h/t Grits for Breakfast)
Researchers from the National Institute of Standards and Technology gave the same DNA mixture to about 105 American crime laboratories and three Canadian labs and asked them to compare it with DNA from three suspects from a mock bank robbery.
The first two suspects’ DNA was part of the mixture, and most labs correctly matched their DNA to the evidence. However, 74 labs wrongly said the sample included DNA evidence from the third suspect, an “innocent person” who should have been cleared of the hypothetical felony.
This is already a problem. People's lives are literally on the line and crime lab testing is more likely to make the wrong call on evidence than the correct one. What's truly disturbing is this study was completed in 2014, but the report was apparently buried by the scientists it implicated. As Dr. Hampikian states in his op-ed, the study's results may still be unpublished if it weren't for forensic scientists publicly complaining about the burial.
Four years have passed since the study's completion and it appears no improvements have been made. The study notes testing protocols vary widely and very little effort is being made to improve error-prone procedures. In addition, the study [PDF] comes with a disclaimer meant to dissuade litigants from challenging DNA evidence by quoting the study's findings.
The results described in this article provide only a brief snapshot of DNA mixture interpretation as practiced by participating laboratories in 2005 and 2013. Any overall performance assessment is limited to participating laboratories addressing specific questions with provided data based on their knowledge at the time. Given the adversarial nature of the legal system, and the possibility that some might attempt to misuse this article in legal arguments, we wish to emphasize that variation observed in DNA mixture interpretation cannot support any broad claims about “poor performance” across all laboratories involving all DNA mixtures examined in the past.
This certainly doesn't raise the reader's confidence in crime lab DNA testing. Instead, it gives the impression the four-year delay between completion and public release was for wagon-circling purposes as crime lab forensic scientists looked for ways to mute the impact of the study's findings.
But there are also problems with the study itself. The authors of the study appear far too willing to cut crime labs slack for their failures. Rather than point out the problems originating from a lack of standardized processes, the study uses them to excuse the failures, as if unintentionally nailing the wrong person for the crime was somehow worthy of gold stars for effort. Here's Hampikian's take:
It is uncomfortable to read the study’s authors praising labs for their careful work when they get things right, but offering sophomoric excuses for them when they get things wrong. Scientists in crime labs need clear feedback to change entrenched, error-prone methods, and they should be strongly encouraged to re-examine old cases where such methods were used.
The study confirms much of what has been exposed earlier: DNA evidence may be based on hard science, but any small variable -- including the inevitable tainting of DNA samples -- has the ability to throw things off. And when it's used as evidence in criminal trials, it has the potential to destroy lives. This study shows -- at least indirectly -- the labs handling DNA evidence aren't taking it nearly as seriously as they should.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Reader Comments
Subscribe: RSS
View by: Time | Thread
"about 105 American crime laboratories and three Canadian labs"
"most labs correctly matched their DNA"
"74 labs wrongly said"
108 - 74 = 30
30 out of 108 got it right but
"most labs correctly matched their DNA"
Do they have a different definition of most?
"However, 74 labs wrongly said the sample included DNA evidence from the third suspect"
So they got it right and wrong?
As far as the disclaimer if I were up on charges stemming from DNA evidence you bet you ass that I would bring this up. You're damn right I'd demand the samples be tested by multiple facilities and scrutinize the collection method and chain of custody.
[ link to this | view in thread ]
Not even a pretense of a fair test
In real cases, with real pressure to get a conviction, it probably would have been 100%. Can't let the "bad guys" get away.
[ link to this | view in thread ]
Re:
No, you misread. The "most labs correctly matched" part was specifically referencing the first two suspects who's DNA was included in the sample. Most labs matched them correctly.
But the same can't be said for the third suspect. That person's DNA was not included in the sample, yet 74 said it was there.
[ link to this | view in thread ]
How soon they forget...
108 labs, "most" (say 90) correctly matched two out of three in the blend. 74 of the labs incorrectly "found" the no-match control in the blend.
This study was about *blended* samples, not *discrete* samples. Send them a cheek swab and they'll probably get it right. Send them a swab from a coin with 30 different DNA traces on it, they *might* get it right.
In any event, ALL DNA evidence can be challenged and a sample of the evidence provided to the Defense for 3rd party testing. There is a caveat, where if a sample is too small for more than one test, the Judge MAY allow it. Or toss it out.
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
say it with me boys and girls
[ link to this | view in thread ]
Hmmmm.....
DNA was scattered across several crime scenes before coming to the realization the DNA officers kept finding belonged to the person packaging the testing swabs used by investigators.
Real life Dexter? Can't believe I'm the person that has to connect the dots here. :)
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
[ link to this | view in thread ]
Re:
A retest at another lab that returns a different result is something the Prosecution does NOT want to see come up.
There's a huge difference between enough evidence for an arrest and evidence to be presented at trial.
DNA testing is still fairly expensive. Prosecutors don't care, it's tax money paying for it. So they'll routinely have DNA evidence retested before trial.
[ link to this | view in thread ]
[ link to this | view in thread ]
Re:
The show where everyone who has ever worked in a lab or overseen one yells at the TV over bad procedure if they have the ill luck to see a lab scene on TV?
That CSI?
:)
[ link to this | view in thread ]
Re: Re:
That is a rather lame excuse for the prosecuting attorneys who use dna evidence in this manner. I do blame the prosecution for bullshitting the court. Presenting false evidence could get one disbarred if judicial review boards were not corrupt.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
Two (or more) positive results may be why the Defense opts not to have a third test done.
In the rare cases where DNA evidence *is* challenged, it's often because the sample was so small it was destroyed in the first test. When a Prosecutor is faced with that situation, they make sure the *rest* of their evidence is as solid as possible.
There's no massive conspiracy here.
[ link to this | view in thread ]
Gold stars
That's the way of the modern world: the worst possible products still get one gold star.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
I gave the only exception to that - single sample destroyed by testing.
[ link to this | view in thread ]
Re: Re: Re: Re:
Sorry, I was referring to this statement. It is one thing for the Prosecutor to retest using public money, it is another thing for a defendant to pony up for a test (just how expensive is it?). Either way, the Prosecutor has a responsibility to get it right.
[ link to this | view in thread ]
Re: Re: Re: Re:
May not reach the level of conspiracy but it may clear the collusion hurdle.
[ link to this | view in thread ]
Re: Re: Re: Re:
That is why I read stories about BFE LEOs going after law students working to free innocent prisoners of the state.
[ link to this | view in thread ]
Re: Re:
In the US, the defense rarely gets an opportunity, as prosecutors make a plea bargain the best option available to the accused, even if they are innocent.
[ link to this | view in thread ]
Re: Re: Re: Re: Re:
Which is why the Defense often doesn't request a third (or fifth) testing.
Cost? Last I looked, 8-10 years back, a clean blood sample ran around $3,500 for a matchable stain.
Frankly, the bulk of cases where DNA retesting is demanded by the Defense are either Public Defender or expensive private attorneys. In those cases, if it's NOT requested, there's usually a reason - if that third test ALSO matches, the Defense is sunk.
[ link to this | view in thread ]
Re: Re: Re:
DNA testing and retesting is only done if it goes to Trial.
For those wondering... Get a lawyer to watch ten minutes of any cop or "court" show. They'll spend the whole ten minutes yelling "OBJECTION!". It's ALL BS - NOT how the system actually works.
[ link to this | view in thread ]
Re: Re:
Step 2. Boast of 100% success rate in finding matches.
Step 3. Profit!
[ link to this | view in thread ]
The customer is always right.
Crime Lab to cop: Sure. Who do you want it to match?
[ link to this | view in thread ]
Math Fail
Duh. And 47% of the time, those two tests will both return false positives for an innocent party. And that's without any LE pressure applied.
[ link to this | view in thread ]
Re: Re:
Yes, the defense's job is to challenge the evidence, but that is extremely expensive. You have to pay for a attorney who is not overworked (aka not a public defender) and pay for a lab to perform the work.
This creates a divide between those that have the resources and those that do not. We need a system where all persons are equal and the current system where the burden is on the defense to challenge bad science is not fair.
We should hold prosecutors more accountable for junk/bad science used at trial.
[ link to this | view in thread ]
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Re:
[ link to this | view in thread ]
Re: Not even a pretense of a fair test
Prosecutors don't like DNA tests in 'sure win' cases because of the problem of DNA contamination. If two or more people's DNA are found it often opens up a pandora's box of things that Defense can raise to pin the blame on the mystery other person.
[ link to this | view in thread ]
Re: Re: Re: Re:
[ link to this | view in thread ]
Re: Re: Re: Re:
That doesn't mean there isn't a serious problem.
[ link to this | view in thread ]
[ link to this | view in thread ]
[ link to this | view in thread ]
Always was.
But but but theoretically it is one in a million... and that would give us how many viable suspects in the bay area alone?
[ link to this | view in thread ]
Re: Re:
Those first several points fall under what is known as "conservation of characters", or a similar phrase; it isn't meant to indicate that things are actually like that, it's just a way to let the show present all the interesting parts without needing to either expect the audience to be familiar with a dizzying array of regular cast members (how many would be needed, to do everything shown as part of the investigation in a typical "CSI" episode?) or actually hire and pay enough people to fill all those roles on a regular basis.
The bad-lab-procedure point is considerably less defensible, and I don't generally bother to try.
[ link to this | view in thread ]
Re: Re:
So how do we know that the labs 'correctly' identified the first 2? Couldn't they have also completely screwed up the tests and still identified the first 2 as suspects even if they hadn't been... I don't think so, at best we can say they ended up with the 'right' answer, but we can't tell if they arrived at it 'correctly'
[ link to this | view in thread ]