Have you seen the news? The FBI has announced errors in its database which is used across the nation and in Texas. In two prior blog posts (on separate sites), I addressed this issue.
August 7, Just How Accurate is DNA?, HCCLA.org (with memo downloads)
September 18, FBI DNA Calculation Errors, HCCLAtv.com
Today, Fox 26 News visited with me to get a perspective on just how big this problem is.
As I told Andrea Watkins, Fox26 News, the problem will be large simply because each case will have to be identified and then reviewed to see what impact the DNA results may have had on a particular plea-bargain or jury verdict. Ultimately, this will amount to thousands of cases locally as the errors have existed and been perpetuated since 1999.
The problem is that errors in the statistical database could have caused an inflated reliance on a match. For example, instead of the probability of a particular DNA sample matching a suspect or defendant being 1 in 1 billion, it could be that it is only a 1 in 100 chance of belonging to the same person. This certainly changes the landscape and statistical chance of the DNA being left by the same person.
This problem is compounded in “mixture” cases. A mixture case is where two or more persons have contributed to the sample. We often see swabs collected from crime scenes where the analysis reveals the DNA of two individual people. The statistical errors, once corrected, can cause a scientific result which once included a suspect to now exclude that same person because of the lack of strength in the probability of the result matching. (I hope that makes sense. Essentially, science sets limits. If the probability falls below the lower limit, it is no longer considered a scientific match; so if the recalculation falls below, it is no longer considered reliable.)
Of course correcting each DNA statistical calculation will not ultimately help every defendant or suspect. But i will change the scientific reliability in some cases; therefore, each case will have to be analyzed to see whether or not it is affected.
This is a huge undertaking. It will take time. And, it will certainly add to the backlog in crime labs and the overall turn-around on testing old as well as new evidence.
In any event, I’m staying on top of this and will continue to report what I can.