DNA Evidence—New Tactics to Use in Challenging

Printer-friendly versionPrinter-friendly versionPDF versionPDF version
Friday, July 10th, 2015
DNA Evidence—New Tactics to Use in Challenging

Who, in defending a case involving DNA evidence, has not heard testimony from the State’s expert to the effect that there is a one-in-3 billion chance of a match to someone other than the defendant? It does not matter that all 13 alleles do not match; the testimony is essentially the same. Invariably.

Powerful evidence. It instantly creates the impression in the mind of the jury that it would take over 3 billion people being tested before a match occurred. But, as you will see, it can be a false impression. A very false impression.

Or worse, the State’s expert offers testimony to the effect that: “The frequency of an unrelated African American (both appellant and Franklin are African American) having the same DNA pattern as that found on the cigarette butt consistent with the major contributor’s DNA was approximately one in 118 billion.”1

Worse yet, the State’s expert offers testimony to the effect that: “The complainant could not be excluded as a contributor to the bloodstain on the bill. The chances of another contributor are 1 in 30 quintillion Caucasians, 1 in 56 quintillion African-Americans, and 1 in 343 quadrillion Hispanics.” And, “The chances of another contributor are 1 in 191 quintillion Caucasians, 242 quintillion African-Americans, and 1 in 273 quintillion Hispanics.”2

Never mind that some of these chances are “derived” after testing alleles at only 6 of the 13 loci that are normally tested!3,4 Or that European agencies usually use two extra markers, D2 and D19, to make 15 loci (16 if you include AMEL) that are tested.5 Or that Taiwan now tests allelles at 23 markers because of matches at 13 alleles that turned out to be exclusions when 23 markers were tested.6 Or that there is no uniform methodology for determining those chances within the field—not even within the same lab!7 Or that the policies of some agencies requires their employees to testify—when only two alleles match—that the person cannot be excluded.8

So, how do you challenge this testimony?

Before listening to the presentation of Greg Hampikian, PhD, at the National Child Abuse Defense and Resource Center’s seminar last October, we were relegated to using the tried (tired?) tactics to challenge this one-in-however-many billion, trillion, quadrillion, or quintillion chance testimony. These tactics include: (1) challenging the methodology used in testing the sample(s), especially when a mixture of DNA was tested; (2) challenging chain of custody or errors in sampling; (3) challenging the technician on his or her methodology in light of known errors; (4) challenging the degradation of the DNA specimen; (5) challenging the methodology of calculating the chances; (6) challenging the underlying study that is used to calculate these chances; (7) challenging the possibility of contamination of the sample; (8) challenging the lab based on its history of past errors, etc.9 This is not an exhaustive list, but those of us who have challenged DNA know them all.

Wouldn’t it be nice to have some new “bullets” to put into your cross-examination “gun”?10 Or to have some additional aces (bullets) to lay down when playing poker with the State’s experts on DNA?11 Or to be able to force the State’s expert to have to bite the bullet and concede that their opinions are flawed, if not outright wrong?12

Dr. Hampikian gave everyone who attended the NCADRC seminar several new aces (bullets) to use when playing poker with the State’s expert and trying to trump the State’s expert’s testimony on DNA evidence.13

Dr. Hampikian’s First Bullet

The first bullet was to get us lawyers to understand the difference between chance and probabilities (statistics).14

For instance, take the question of what is the number of people you have to have in a room to have a better than 50% probability of two of them sharing the same birthday? Call this the birthday problem. People will guess at anywhere from 366 to 183 and they will be wrong. This is a simple example of how people (jurors) confuse chance with probabilities (statistics).

The birthday problem asks how many people you need to have at a party so that there is a better-than-even chance that two of them will share the same birthday. Most people think the answer is 183, the smallest whole number larger than 365/2. In fact, you need just 23. That’s right, 23.

The answer 183 is the correct answer to a very different question: How many people do you need to have at a party so that there is a better-than-even chance that one of them will share your birthday?

If there is no restriction on which two people will share a birthday, it makes an enormous difference. With 23 people in a room, there are 253 different ways of pairing two people together, and that gives a lot of possibilities of finding a pair with the same birthday.

Here is the precise calculation. To figure out the exact probability of finding two people with the same birthday in a given group, it turns out to be easier to ask the opposite question: what is the probability that no two will share a birthday—i.e., that they will all have different birthdays? With just two people, the probability that they have different birthdays is 364/365, or about .997. If a third person joins them, the probability that this new person has a different birthday from those two (i.e., the probability that all three will have different birthdays) is (364/365) x (363/365), about .992. With a fourth person, the probability that all four have different birthdays is (364/365) x (363/365) x (362/365), which comes out at around .983. And so on. The answers to these multiplications get steadily smaller. When a twenty-third person enters the room, the final fraction that you multiply by is 343/365, and the answer you get drops below .5 for the first time, being approximately .493. This is the probability that all 23 people have a different birthday. So, the probability that at least two people share a birthday is 1 - .493 = .507, just greater than ½.15

Carrying this out to 30 people and the answer you get drops to approximately .293. So the probability that at least two people share a birthday with 30 people in the room is 1 - .293 = .707, or greater than 70%. Carrying this out to 35 people and the answer you get drops to approximately 0.185. So the probability that at least two people share a birthday with 35 people in the room is 1 - .185 = .815, or greater than 80%. Carrying this out to 40 people and the answer you get drops to approximately 0.108. So the probability that at least two people share a birthday with 40 people in the room is 1 - .108 = .892, or almost 90%. Carrying this out to 44 people and the answer you get drops to approximately .007. So the probability that at least two people share a birthday with 44 people in the room is 1 - .007 = .993, or greater than 99%.

Changing the birthday problem slightly, ask how many people you need to have at a party so that there is a virtual certainty that two of them will share the same birthday. The answer to this question is 45, because with 45 people in the room, the probability that at least two people share a birthday is greater than 100%!16

A similar problem is presented by the “Children Puzzle.” I tell you that a couple has two children and that (at least) one of them is a boy. I ask you what is the probability that their other child is a boy. Most people think the answer is 1/2, arguing that it is equally likely that the other child is a boy or a girl.17 But that’s not the right answer for the question I have asked you. Here’s why. In terms of order of birth, there are four possibilities for the couple’s children: BB, BG, GB, GG. When I tell you that at least one child is a boy, I rule out the possibility GG. That leaves three possibilities: BB, BG, GB. With two of these, the other child is a girl; so the probability of the other child being a girl is 2/3. Leaving the probability of the other child being a boy at 1/3.18

A similar problem is presented by these questions: What is the probability of tossing a coin and having it come up heads 10 times in a row versus what is the chance (probability) that on the tenth flip of the coin, it will come up heads? The first probability is (½)10—one-half to the tenth power. The second probability is ½. But most people will answer both questions as ½.

These are just three examples of how people (jurors) confuse chance with probabilities. Getting jurors to understand the difference between chance and probabilities (statistics) is very important. Getting the State’s expert to talk in terms of probabilities (statistics) instead of chances is even more important.

Dr. Hampikian’s Second Bullet

Which brings us back to DNA in the courtroom and the second bullet that Dr. Hampikian gave us. The State’s DNA expert is going to testify that there is a one-in-3 billion chance that there would be a match on the alleles that were tested and which matched.19 Or, as there are fewer than 13 alleles that match, that your client cannot be ruled out.20

What do you do?

First, if it’s a one-in-3 billion chance testimony, we would recommend that you file a Rule 702 challenge to any such testimony. Why? Because chance is not probability.

You need to reframe the question or the expert’s statement. The question or statement is, more properly, Is a coincidental match to the DNA database possible? And if so, what is the probability of that coincidental match?

If a profile has a random match probability of one-in-3 billion, how big does the database have to be before a “random match” is expected (over 50% chance)? The answer is about one and one-half billion. Using the “birthday problem” above, you see that this is the “How many people do you need to have at a party so that there is a better-than-even chance that one of them will share your birthday?” answer.

Second, you need to ask the State’s expert whether he has ever examined the FBI’s DNA database or even the Texas DNA database to see if there were any random matches and, if so, on how many alleles the profiles matched. Being the cynics that we are, we would expect the expert to announce that, in fact, he had done so. Which then leads to how did he get access to the database when no one else has been able to do so, were those results were published in a peer-reviewed scientific article, etc.

Of course, this leads to the fertile ground of cross-examination: How many DNA profiles are in the FBI database? Or the Texas database?

Which brings us back to, what is the chance (probability) that there is a random match somewhere in the DNA database?

If a DNA database has a number of profiles that each has about a one-in-3 billion random match probability, the question becomes: How big does that database have to be before you expect (more than 50% chance) a match between two profiles in the database?

Remember, the FBI has the world’s largest DNA database, but it has never made its database available to independent scientists to examine. The authors have searched online and have been unable to find any definitive answer to even the question of how many DNA profiles it has in its database.21

So, how big does that database have to be before you expect (more than 50% chance) a match between two profiles in the database?

According to Dr. Hampikian, the answer is 65,493. That’s right—65,493. Not one and one-half billion people. Not a billion people. Not five hundred million people. Not a million people. Slightly more that 65,000 people!

So, where did Dr. Hampikian come up with that number? Arizona. That’s right—Arizona.

Well, actually, it was the examination of the Arizona DNA database that was performed by Steven P. Meyers, MS, with the California DOJ Jan Bashinski DNA Lab. Dr. Hampikian showed those slides to the audience at the NCADRC Seminar in October 2014.22

You see, among all of the states that have DNA databases, only Arizona has made its DNA database available to scientists to examine. In that examination, the scientists were able to find the following matches:

  • 122 pairs match at 9 of 13 loci
  • 20 pairs match at 10 of 13 loci
  • 1 pair matches at 11 of 13 loci (full siblings)
  • 1 pair matches at 12 of 13 loci (full siblings)

And that’s in a database of only 65,493 profiles!

Which suggests to the authors that if you are defending someone on a crime where DNA is being used to “finger” your client, it might be time to ask for discovery of the State of Texas’ DNA database, so that your expert can examine it to determine whether Texas has similar matches that have not been disclosed.23 This alone could be critical in showing that the State’s expert’s pontifications as to chances of a match are nothing other than something that the expert has pulled out of an orifice somewhere.

And, in light of what was uncovered in the Arizona database, it could be argued that the prosecutors are withholding Brady material for at least two separate reasons. First, the Texas DNA database is supposedly larger than the Arizona database, so one could presume that there are matches in the database that are similar to those found in Arizona. Second, the Arizona matches will have been submitted to CODIS, which means that those matches are in CODIS. This is impeachment evidence that the prosecutor has access to and it should be turned over under Brady.

Dr. Hampikian’s Third Bullet

Which brings us to the third bullet that Dr. Hampikian gave those in attendance at the NCADRC Seminar last October. It is the case of Chen Long-Qi out of Taiwan.24

The facts of that case are as follows: On March 24, 2009, two escorts were raped between 4 to 6 a.m. in a warehouse that Chen and his friend rented for agricultural products distribution. The victims failed to identify the assailants due to alcohol intoxication.25

Chen always maintained his innocence during the investigation and trial. He claimed that he left before the crime to pick up his wife, Ko, at her workplace. Ko’s timesheet corroborated Chen’s words. An eyewitness also testified that Chen was not at the scene. Despite no testimony linking Chen to the crime, the district court and high court found him guilty of gang rape with the other two co-defendants. The decision was solely based on a DNA test concluding that Chen “cannot be excluded” from the semen stain found on one of the victims’ underwear. Chen was convicted of gang sexual assault and was sentenced to 4 years in March 2013.26

With help from the Taiwan Association for Innocence, Chen filed a motion for retrial in June 2013 seeking to retest the DNA evidence. The court authorized a 23 loci STR test on the original mixture DNA sample. The new test result showed that Chen “can be excluded” from the DNA sample. Based on this new piece of evidence, the court granted his motion in December 2013.27

According to Dr. Hampikian in his presentation at the NCADRC Seminar, Chen was acquitted on April 15, 2014. As Dr. Hampikian explained: “Last year the Taiwan Association for Innocence director showed me the case of a man convicted of gang rape through DNA evidence. While that first DNA test was accurate, it was a complex mixture, and newer testing is more discriminating. Through a court hearing the National Crime Lab agreed to do further testing with newer kits, and they were able to exclude Chen Long-Qi.28

Dr. Hampikian’s Fourth Bullet

Dr. Hampikian’s fourth bullet dealt with the problem with statistics. He used the case of Donny Denman to illustrate.

Who is Donny Denman? Donny Denman is the man who the FBI pronounced dead after they examined the DNA in some bones found in New Mexico. Since Donny Denman had been missing for years and since they did not have Donny’s DNA, they used Donny’s siblings to test the mitochondrial DNA. And the FBI concluded that the DNA matched and the bones were Donny’s.

Donny had a funeral. The Pastor gave the eulogy. A death certificate was issued in Donny’s name. There was only one problem: Donny was still alive.29

Granted that using mitochondrial DNA is not as effective in distinguishing individuals as the more common nuclear DNA process, there was a coincidental match, nonetheless. What’s really interesting to the authors is that the FBI said that Denman’s case was the first time the FBI lab has had a “coincidental match.”

The reason that statement is so interesting is that the Arizona DNA database, supra, would have been submitted to CODIS. Are we to believe that the DNA samples that led to the matches that Steven P. Meyers, MS, found in the Arizona database, supra, somehow did not make it into CODIS?

Or are we to believe that FBI is telling the truth when they state that they have never had a “coincidental match.” Remember, these are the same people who tell us that fingerprints are unique and who incorrectly identified Brandon Mayfield, a lawyer from Portland, Oregon, as the Madrid train station bomber.30

Related to this is that experts say there is no way to tell what the odds are for a coincidental match. But courtesy of the coincidental matches that Steven P. Meyers found in the Arizona DNA database, we know that the odds of a coincidental match at 9 of 13 alleles is 122/65,493. We also know that the odds of a coincidental match at 10 of 13 alleles is 20/65,493. We also know that the odds of a coincidental match at 11 of 13 alleles is 1/65,493. And we know that the odds of a coincidental match at 12 of 13 alleles is 1/65,493.31

That’s nowhere near a one-in-3 billion chance! Not even in the same ballpark. Not even on the same planet.

Dr. Hampikian’s Fifth Bullet

Dr. Hampikian’s fifth bullet dealt with the problem of contamination of the samples in the laboratory.

Dr. Hampikian gave several examples of cases where people were identified as the perpetrator but the identification was flawed by contamination occurring in the laboratory.

One such case was the case of Carlton Gary, the so-called Columbus Stocking Strangler. He spent almost 30 years on death row in Georgia, and in 2009, hours before he was to be executed, the Georgia Supreme Court ordered DNA testing. Ultimately, the Georgia Bureau of Investigation laboratory conducting the tests reported it had tainted the DNA evidence.32

The interesting corollary is that when the DNA was re-tested, it did not match anyone in the CODIS database. Two years later, a gun crime was committed in Georgia and the DNA from the suspect in that gun crime was submitted to CODIS. A match was found, so that suspect was interviewed and he was excluded from the Columbus stocking murder cases due to his age—he couldn’t have committed those crimes back in the ’70s. It turns out that the samples in both cases were contaminated at the Georgia Bureau of Investigation Crime Lab with the same DNA evidence. And it turns out that the DNA that contaminated both samples was from a semen sample produced by someone who works in that lab—a sample produced as a quality control!33

Dr. Hampikian’s Sixth Bullet

Dr. Hampikian’s sixth bullet dealt with the problem of contamination of the samples that occurs outside the laboratory.

In his presentation at the NCADRC Seminar in October 2014, Dr. Hampikian talked about the need for crime scene technicians to change their gloves between each piece of evidence they handle so as to avoid transferring DNA from one piece of evidence to another. This, alone, presents a fertile ground for cross-examination.

Dr. Hampikian also talked about the “phantom of Heilbronn.”34 This is the debacle suffered by the German police when they spent 16 years chasing a woman who never existed. The unnamed woman was suspected of being a serial killer who over 16 years carried out a string of six murders, including strangling a pensioner. It turns out the misidentification was caused by swabs used to collect DNA samples having been contaminated by an innocent woman working in a factory in Bavaria.

Conclusion

The take-away from Dr. Hampikian’s two presentations is this: There are forensic DNA errors; there are statistical and interpretative errors; and there are contamination errors. Now you have six new aces to lay down on the table when you want to trump the State’s expert in your quest for justice for your clients who are being “fingered” by DNA.

Endnotes

1. Brown v. State, 163 S.W.3d 818 (Tex. App.—Dallas 2005, pet. ref’d).

2. Owolabi v. State, 448 S.W.3d 148 (Tex. App.—Houston [14th Dist.] 2014, no pet.).

3. Brown v. State, 163 S.W.3d at 825—826.

4. CODIS identifies genetic markers at 13 STR loci, plus Amelogenin (AMEL) to determine sex. See http://www.dnaconsultants.com/Default.aspx?PageID=5813864&A=SearchResult... (last accessed January 22, 2015).

5. Id.

6. http://wrongfulconvictionsblog.org/2014/02/07/taiwan-association-for-inn... (last accessed January 22, 2015).

7. Presentation by Greg Hampikian, PhD, at TEDx Boise 2015. Dr. Hampikian’s presentation can be found at: http://tedxtalks.ted.com/video/Forensic-DNA-Mixups-|-Greg-Hamp . There’s lots of great information, and this presentation is recommended to all who read this article.

8. Dr. Hampikian notes that the experts from the Georgia Bureau of Investigation will testify, when only two alleles match, that the defendant cannot be ruled out. Using statistics (probabilities), this is an absolute exclusion. You should be aware of similar testimony from the State’s experts in Texas. Never forget that the State’s expert testified that the DNA in the rape kit was an exact match to Josiah Sutton. But only three alleles matched, and that was an absolute exclusion! See http://www.voiceforthedefenseonline.com/story/taint-question-reliability... .

9. In his presentation at TEDx Boise 2015, Dr. Hampikian discusses the problems associated with contamination of the samples, including the possible sources of contamination.

10. Two the authors are attorneys, licensed in Texas, proud possessors of a CHL, who can often be found exercising the privileges that come with a CHL. It seemed appropriate to them to use the term “bullet,” in the sense of the metal cartridge that one inserts into a pistol. This term was not suggested by Dr. Hampikian, and to the authors’ knowledge, he has never referred to his points as bullets—not even when he was using a PowerPoint presentation containing what would otherwise be called bullet points.

11. In cards, an ace is referred to as a bullet. Random House Dictionary, Random House Inc., 2015. The NCADRC seminar is usually held in Las Vegas, and the authors have been known to participate in the games of chance offered in the casinos. There has been more than one occasion when they each would have been more than happy to have had one more of these bullets to play.

12. “Bite the bullet”: to force oneself to perform a painful, difficult task or to endure an unpleasant situation. Random House Dictionary, Random House Inc., 2015.

13. The authors are grateful to Dr. Hampikian for his assistance in the prep­a­ra­tion of this paper and his providing the slides referred to in this paper together with the link to his TEDx 2015 presentation. His assistance was limited to check­ing the paper for errors, and he did not have any input into the final draft or its terminology or the words used.

14. Again, the term “bullet” is used in the sense defined in note 11—an ace.

15. http://www.npr.org/templates/story/story.php?storyId=4542341 (last accessed January 3, 2015).

16. As you will see in this paper, this question can be rephrased: How many people’s DNA profiles do you have to have before you are virtually certain to have two people who match at 9, 10, 11, or even 12 loci.

17. That answer is the answer to the question of what is the chance that the other child is a boy.

18. Id.

19. Or in a trillion, a quadrillion, a quintillion, or in a whatever chance. For simplicity, the authors will keep it to a one-in-3 billion chance.

20. As noted by Dr. Hampikian, the experts from the Georgia Bureau of Investigation will testify that when only two alleles match, the defendant cannot be ruled out. Using statistics (probabilities), this is an absolute exclusion. You should be aware of similar testimony from the State’s experts in Texas. Never forget that the State’s expert testified that the DNA in the rape kit was an exact match to Josiah Sutton. But only three alleles matched, and that was an absolute exclusion! See http://www.voiceforthedefenseonline.com/story/taint-question-reliability... .

21. The authors are not, by this statement, claiming to be the absolute best online researchers.

22. The authors have tried to attach Dr. Hampikian’s PowerPoint presentation on this study but have been unable to do so. If you will email L. T. Bradt at ltbradt@flash.net, he will be happy to share the PowerPoint presentation that Dr. Hampikian shared with him.

23. This raises an interesting Brady issue. But that is for another time and another article.

24. http://wrongfulconvictionsblog.org/2014/02/07/taiwan-association-for-inn... (last accessed January 22, 2015).

25. Id.

26. Id.

27. Id.

28. See also http://news.boisestate.edu/update/2014/04/21/greg-hampikian-58/ (last accessed January 27, 2015).

29. http://www.abqjournal.com/news/metro/302718metro04-25-08.htm (last accessed January 24, 2015).

30. http://seattletimes.com/html/localnews/2001937794_mayfield25m.html (last accessed January 24, 2015).

31. As to both the 11 of 13 alleles matching and the 12 of 13 alleles matching in the Arizona database, these profiles involved full siblings. Remember, the RMP assumes that people are unrelated, so if you use these examples, the State’s expert may throw it back in your face.

32. http://www.ledger-enquirer.com/2014/02/23/2970172/stocking-strangler-com... (last accessed January 24, 2015).

33. Greg Hampikian’s presentation at TEDx Boise 2015. Dr. Hampikian’s presentation can be found at http://tedxtalks.ted.com/video/Forensic-DNA-Mixups-|-Greg-Hamp . There’s lots of great information, and this presentation is recommended to all who read this article.

34. See Dr. Hampikian’s presentation at TEDx Boise 2015, found at http://tedxtalks.ted.com/video/Forensic-DNA-Mixups-|-Greg-Hamp . See also http://news.bbc.co.uk/2/hi/europe/7966641.stm (last accessed January 25, 2015).