[Senate Hearing 112-671]
[From the U.S. Government Publishing Office]
S. Hrg. 112-671
THE SCIENCE AND STANDARDS OF FORENSICS
=======================================================================
HEARING
before the
COMMITTEE ON COMMERCE,
SCIENCE, AND TRANSPORTATION
UNITED STATES SENATE
ONE HUNDRED TWELFTH CONGRESS
SECOND SESSION
__________
MARCH 28, 2012
__________
Printed for the use of the Committee on Commerce, Science, and
Transportation
----------
U.S. GOVERNMENT PRINTING OFFICE
77-701 PDF WASHINGTON : 2012
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800;
DC area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC,
Washington, DC 20402-0001
SENATE COMMITTEE ON COMMERCE, SCIENCE, AND TRANSPORTATION
ONE HUNDRED TWELFTH CONGRESS
SECOND SESSION
JOHN D. ROCKEFELLER IV, West Virginia, Chairman
DANIEL K. INOUYE, Hawaii KAY BAILEY HUTCHISON, Texas,
JOHN F. KERRY, Massachusetts Ranking
BARBARA BOXER, California OLYMPIA J. SNOWE, Maine
BILL NELSON, Florida JIM DeMINT, South Carolina
MARIA CANTWELL, Washington JOHN THUNE, South Dakota
FRANK R. LAUTENBERG, New Jersey ROGER F. WICKER, Mississippi
MARK PRYOR, Arkansas JOHNNY ISAKSON, Georgia
CLAIRE McCASKILL, Missouri ROY BLUNT, Missouri
AMY KLOBUCHAR, Minnesota JOHN BOOZMAN, Arkansas
TOM UDALL, New Mexico PATRICK J. TOOMEY, Pennsylvania
MARK WARNER, Virginia MARCO RUBIO, Florida
MARK BEGICH, Alaska KELLY AYOTTE, New Hampshire
DEAN HELLER, Nevada
Ellen L. Doneski, Staff Director
James Reid, Deputy Staff Director
John Williams, General Counsel
Todd Bertoson, Republican Staff Director
Jarrod Thompson, Republican Deputy Staff Director
Rebecca Seidel, Republican General Counsel and Chief Investigator
C O N T E N T S
----------
Page
Hearing held on March 28, 2012................................... 1
Statement of Senator Rockefeller................................. 1
Statement of Senator Boozman..................................... 3
Statement of Senator Udall....................................... 28
Witnesses
Eric S. Lander, Ph.D., President and Founding Director, Broad
Institute of Harvard and MIT; Professor of Biology, MIT;
Professor of Systems Biology, Harvard Medical School; Co-chair,
President's Council of Advisors on Science and Technology
(PCAST)........................................................ 5
Prepared statement........................................... 9
Patrick D. Gallagher, Ph.D., Under Secretary of Commerce for
Standards and Technology, U.S. Department of Commerce.......... 12
Prepared statement........................................... 14
Dr. Subra Suresh, Director, National Science Foundation.......... 16
Prepared statement........................................... 18
Appendix
National District Attorneys Association (NDAA), prepared
statement...................................................... 39
Response to written questions submitted to Eric S. Lander, Ph.D.
by:
Hon. John D. Rockefeller IV.................................. 40
Hon. Amy Klobuchar........................................... 42
Hon. John Boozman............................................ 42
Response to written questions submitted to Dr. Subra Suresh by:
Hon. John D. Rockefeller IV.................................. 43
Hon. John Boozman............................................ 45
Response to written questions submitted to Patrick D. Gallagher,
Ph.D. by:
Hon. John D. Rockefeller IV.................................. 47
Hon. John Boozman............................................ 48
THE SCIENCE AND STANDARDS OF FORENSICS
----------
WEDNESDAY, MARCH 28, 2012
U.S. Senate,
Committee on Commerce, Science, and Transportation,
Washington, DC.
The Committee met, pursuant to notice, at 2:35 p.m. in Room
SR-253, Russell Senate Office Building, Hon. John D.
Rockefeller IV, Chairman of the Committee, presiding.
OPENING STATEMENT OF HON. JOHN D. ROCKEFELLER IV,
U.S. SENATOR FROM WEST VIRGINIA
The Chairman. Welcome, everyone. You don't see a lot of
bodies around here, because we only allowed the two most
intelligent, insightful members to be invited to this committee
hearing, and so if you believe that, it'll be an easy hearing.
I don't get a chance very often to say that the Commerce
Committee is working on truth and justice, but that's what
we're doing today, and it's about using more science in our
criminal justice system. It's about creating standards that
judges, prosecutors, defense lawyers, law enforcement people
and juries can trust.
This is the second hearing we've held. We had John Grisham
at our first hearing, but what was interesting, and, Senator
Boozman, you may remember this, is that--well, I'll get to
that, because there was a more interesting witness who was
here.
And we heard that there were many disciplines in forensic
science, like ballistics, bite marks, even fingerprint
analysis, and they're not based on peer reviewed science. And
we heard that the forensic science community does not have the
resources or sometimes the desire to conduct this type of
research, and maybe it's just the money, but that's what--we
can talk about all those things.
And most disturbing, and we heard that many forensic
scientist disciplines lack what our witnesses call a culture of
science, at the last hearing they said that. Too often,
therefore, the conclusions that are reached are subjective and
lack scientific validation and standards, and, thus, forensics
comes under question.
Without properly analyzing evidence, it's hard for law
enforcement people to apprehend, prosecute criminals, and it's
more likely that our system will wrongfully convict people who
are, in fact, innocent.
What's clear at this point is that we need more research
and better standards in forensic science. Easily said, but
you've got to have the money to do it. And to be credible, this
work needs to be performed by scientific experts outside of the
law enforcement culture.
Today, we're going to talk with three leading scientists--I
mean, you're all rock stars, even though you're not best
selling authors. If you are, raise your hand and I'll duly
recognize you. This is about leveraging the expertise of our
Federal science agencies to force improvement in this process.
If our shared goal is to build a culture of science in the
forensic science disciplines, the National Institute of
Standards and Technology and the National Science Foundation
are the two Federal agencies we will look to for guidance and
for expertise. These two agencies will have become a link
between the forensic science and the broader scientific
research community.
NIST's work focuses on measurement science and standards of
forensics, forensic science, the standards and the measurement.
NIST scientists have decades of collaboration under their
belts, with the FBI, for example, to improve their hardware and
computer programs for fingerprinting screening and all kinds of
things.
The FBI Criminal Justice Information Service--I love these
things, FBI CJIS--which is based in West Virginia, houses the
world's largest biometric database as part of the integrated
automatic fingerprint identification system. You wonder why
more people don't pursue science?
The FBI CJIS also hosts the Department of Defense's
biometrics database that is fully interoperable with the FBI
database and includes a broader array of biometrics data
including fingerprinting, iris, palm, facial, voice and DNA.
This kind of collaboration between our science, those who
practice it, and our criminal justice system will just have to
grow and deepen. You know, the culture of science --many other
cultures get in the way, too----jurisdictional, who's on top,
who's in charge, all the rest of it, which is just classic
behavior in our country and I guess everywhere in the world.
But we have to put our evidence standards on a solid
scientific footing. That's the point. Putting more science into
forensic science is one of the Commerce Committee's top
priorities this year. It is for me and I know it is for Senator
Boozman. I'm working on legislation that I hope to introduce in
April that will apply to this.
My questions today will focus on the best way to apply the
Federal Government's scientific knowledge and the resources to
this problem, and I look forward to hearing your testimony, and
I'm going to introduce you individually.
But I want to call on, now, Senator Boozman and point out
that he is just--it's just like he has entered graduate school
again. He has taken up the subject of forensic science on his
own without any pushing, except from inside of his brain or
soul, and he's interviewing everybody in sight and has kinds of
group meetings and fora and things that are way beyond my
capacity to understand. So I'm really proud that he's here,
that he's our Ranking Member, and of him in general.
STATEMENT OF HON. JOHN BOOZMAN,
U.S. SENATOR FROM ARKANSAS
Senator Boozman. Thank you, Mr. Chairman. I think that we
have had a lot of input from our very competent staffs that are
working so hard on this issue. And it really is a very, very
important issue, and I appreciate you bringing it to the
forefront and allowing us to have these hearings.
As you know, I feel very strongly about this, not only
because the field of forensic science is critically important
to upholding our nation's criminal justice system, but also
because the forensic sciences are a vital element in supporting
homeland security and counterterrorism missions and protecting
the safety of the public.
In our last forensics hearing, we discussed the many
advances that have been made in the field of forensic science
over the last two decades that have led to the prevalence of
forensic evidence in our judicial system in court rooms,
particularly in the realm of DNA technology and medical
identifiers, which are widely relied upon by investigators,
attorneys, judges and jurists throughout the judicial process.
We also discussed, in part, the recommendations made by the
National Academies of Science Report of 2009 and the need for
standardization in many specific forensic science areas, as
well as hearing, as the chairman mentioned, from people like
John Grisham on the tragedy that results from a wrongful
conviction based on faulty forensic evidence.
However, as we are all aware, much of the surge of
attention in forensic science has come from the Hollywood sets
of popular television shows that portray the state-of-the-art
forensic laboratories and the use of forensic evidence often as
a central factor in their ability to solve crimes in a 60-
minute segment.
This, of course, makes for good entertainment, but in
reality, these shows grossly misrepresent what our system can
accurately rely on in terms of the complexity and uniqueness of
the various fields within the broader field of forensic
science.
Today, I'm looking forward to hopefully hearing how we can
bridge this gap between the basic scientific research in these
fields and build a structure that will be accurate and reliable
enough to hold up in a court of law in the most effective and
efficient manner.
And while there is no doubt that greater peer review
research efforts and basic scientific training are necessary to
increase crime laboratory capacity and improve the accuracy,
precision and reliability that are necessary to build this
structure, we must accomplish this by leveraging building upon
existing initiatives and expertise within the forensic
community.
We must set our focus on strengthening forensic science to
ensure reliable findings and improved judicial integrity,
national security and public safety without completely
reinventing the wheel, and I'm confident that we can all work
together and do so.
Just this past week, as the Chairman mentioned, we had the
opportunity to sit down in an informal setting with many of the
leaders in the forensic community, including Dr. Gallagher, the
Director of NIST who's here with us today, and we do appreciate
you very, very much, as well as the representatives from NSF,
and also the Director, Dr. Lob of NIJ, and the Director of the
White House's Office of Science and Technology Policy, Dr.
Holdren. So we truly had an all-star cast and were able to
discuss this, again, with myself and our staff who has worked
so very, very hard.
The purpose of the meeting was to promote an honest
dialogue to address the problems in the field of forensics and
discuss proposed solutions in a manner that is often difficult
to do during a formal committee hearing.
I was very pleased with the conversation. In fact, I base a
large portion of my confidence so we can effectively achieve
improvement in an efficient manner on the remarks during the
meeting.
I think we all agreed that the best path forward was
through better collaboration and coordination of our existing
resources. Therefore, I'm very eager to hear the comments and
suggestions from today's esteemed panel. I would like to thank
all of you for being here today, and so I know your time is
valuable, so let's get started. With that, I yield.
The Chairman. Thank you, Senator Boozman, very much, and
you're too modest about yourself.
As I say, I want to introduce each of you, and, first, as
it turns out, I'm looking at him, is Dr. Eric Lander, who is a
world renowned expert in genomics. Dr. Lander is President and
Founding Director of the--is that Eli Broad?
Dr. Lander. Broad.
The Chairman. Broad, yes. That's him, though.
Dr. Lander. That's him.
The Chairman. Yes, of Harvard and MIT. The Broad Institute
propels the understanding and treatment of human diseases by
studying their genetic underpinnings.
Dr. Lander is also a professor of biology at MIT. This is
kind of impressive. I mean, I don't treat you guys as well, and
I apologize, but, I mean, this is impressive.
Dr. Lander is also a professor of biology at MIT and a
professor of systems biology at Harvard Medical School and Co-
chair of the President's Council of Advisors on Science and
Technology, which Senator Boozman just mentioned. So that's
quite a lot.
And then Patrick Gallagher, right in the middle, is----I
mean, this is an all-star cast. It really is. I mean, I think
the last time I was at NIST was 20 years ago, and shame on me.
And I drive by the National Science Foundation, but do I come
in? No. Shame on me. I mean, the repository of knowledge and
the depths of science and the eagerness in those institutions
is extraordinary.
So Dr. Gallagher is the Under Secretary of Commerce for
Standards and Technology and the Director of the National
Institute of Standards and Technology, NIST, at the Department
of Commerce. Does one beget the other?
Dr. Gallagher. After the COMPETES reauthorization bill of
2010, they are one and the same. Prior to that, I was just
Director of NIST.
The Chairman. You see.
Dr. Gallagher. So this committee----
The Chairman. I helped advance your career.
Dr. Gallagher. You did. Thank you.
[Laughter.]
The Chairman. And, finally, we have Dr. Subra Suresh, who's
the Director of the National Science Foundation, an enormous
repository of knowledge.
And so, Dr. Lander, you're on.
STATEMENT OF ERIC S. LANDER, Ph.D., PRESIDENT AND FOUNDING
DIRECTOR, BROAD INSTITUTE OF HARVARD AND MIT; PROFESSOR OF
BIOLOGY, MIT; PROFESSOR OF
SYSTEMS BIOLOGY, HARVARD MEDICAL SCHOOL; CO-CHAIR, PRESIDENT'S
COUNCIL OF ADVISORS ON SCIENCE AND TECHNOLOGY (PCAST)
Dr. Lander. Thanks very much, Chairman Rockefeller, Ranking
Member Boozman, thanks for inviting me here today to talk about
this issue of insuring quality and consistency in forensic
science.
As you said, I have these other jobs. I direct this Broad
Institute, worked on the Human Genome Project and direct the
President's Council of Advisors on Science and Technology, but
I think the real reason that I was asked to come testify today
has to do with an experience I had 23 years ago in the first
case in which DNA fingerprinting evidence was seriously
examined in our criminal justice system.
I've given you some extended testimony, and I'm just going
to try to summarize and describe the key points here.
The Chairman. Feel free, we're not pressed here.
Dr. Lander. OK. Well, 23 years ago, I got invited to
testify in a case in New York called the People v. Castro, and
in that case--it was a murder case in the Bronx--the defense
asked me, because I was a human geneticist, if I would look at
the DNA fingerprinting evidence. I did this reluctantly and pro
bono, but, in the end, I agreed to look at it and to testify
and, as much as I was a DNA scientist, the evidence itself was
appalling.
There were no standards for declaring when two DNA bands
matched. There were no standards for declaring when some non-
matching band could be ignored or should be counted as a non-
match. There were no real standards for declaring the
probability of a match. And testing labs were giving, blithely,
numbers like one in 10 billion for the chance of a match, when,
in fact, when you probed underneath it, there was no real
evidence to support those claims.
That was just the first year or so of DNA fingerprinting,
but it was a situation where you had a world class technology
developed by molecular biology, but it was being applied in a
way that lacked standards.
Well, it was a fascinating case. It was a 15 week, pretrial
hearing in the Bronx, and near the end of it, something very
unusual happened. Well, unusual for the legal system, maybe not
for science.
All the witnesses who had testified on behalf of the
defense got together with all the witnesses who had testified
for the prosecution, in the middle of the case, without any of
the judges or lawyers, and we spent a day reviewing the
evidence with each other as scientists would do.
We later found out this isn't the sort of thing that
usually happens in the legal system, but it's very typical for
scientists. And, at the end of the day, all the scientists who
had testified for the prosecution agreed that the evidence was
appalling and agreed to switch sides and testify for the
defense.
The end of the day, the judge really had little choice but
to declare that DNA fingerprinting was, in principle, a
powerful technology, but, in practice, had been applied so
sloppily that it couldn't be admitted.
This gave rise to a lot of consequences. I had the
tremendous honor to then work together with the FBI's crime
lab, in particular, a wonderful scientist called Bruce Budowle
at Quantico, on trying to set standards for this.
I served on this National Academy committee that was
organized on DNA fingerprinting in the early 1990s, and within
about 5 years, based on the efforts of many people working
together in the law enforcement community and the scientific
community, and bringing very different perspectives, but
talking to each other, DNA fingerprinting was put on a firm
foundation, the firm foundation it's on today, where it is
amazingly sensitive and highly accurate, but it wasn't an
accident. It was the result of real robust collaboration
between two cultures.
Now, at the beginning, I've got to say, the law enforcement
community had serious worries that we were going to have, you
know, scientists running around and producing, you know, all
sorts of things that would disrupt the use of DNA
fingerprinting in the courts.
What ended up happening was exactly the contrary. That
collaboration made the technology stronger for prosecutors,
made it easy to use for cold cases and for identifying
perpetrators of rapes and murders, and it made it more powerful
for the defense as well.
In the end, it wasn't a question of being a tool for the
prosecution or the defense. It was a tool for truth, and that
made a big difference. When we get the truth wrong, we both
risk convicting an innocent person, and we risk having a
perpetrator still running around on the streets. So the defense
and the prosecution have a common interest in getting this
right.
Well, the power of DNA fingerprinting, as it became a
highly accurate technology, had another unexpected consequence.
For the first time, we could go back and look and we, not that
I myself was involved, but, we, as a community, could go back
and look at past cases and see where we got things wrong.
What we'd call it in science is false positives. Think,
times you thought you'd made a match, but, in fact, DNA now
revealed you were wrong. And you could then begin to ask how
did that happen. How did you make that mistake?
Well, it turns out that in about 60 percent of cases that
were examined forensic science had been used in these cases
that involved wrongful conviction. So people could go back and
ask what was wrong with this forensic science.
On a notable case, there was an honorably discharged
veteran who was convicted of murder in Arizona based on bite
marks. A forensic examiner said his bite marks matched the bite
marks on the neck of the victim, and he was convicted of
murder.
He was later exonerated, based on DNA evidence, and the
real perpetrator was found, who, incidentally, committed
another attack on a girl 20 days after the first case. Had we
been able to finger the right person at the right time, we
would have not only avoided convicting an innocent person, but
caught a guilty person.
Another case in New York, a rape and murder, where someone
was tied to that rape and murder based on hair analysis and
soil analysis and fabric print analysis, where an examiner said
that they were similar. Well, DNA evidence, 20 years later, set
that person free as absolutely innocent, and it points out that
hair analysis can be wrong.
And the FBI has found, in a study it did, about one in
eight of the matches that were detected in the study were later
found to be non-matches, based on DNA evidence, fabric
analysis, bite mark analysis. In all of these cases we often
lack objective standards for declaring what's a match. What
features do you want to declare to be matching? What features
can you ignore? What probabilities can you attach?
Well, as you already said, in 2009, the National Academy of
Sciences issued an important and a thoughtful report about how
to strengthen forensic science, and it pointed out that there
are issues. There's nothing wrong with there being issues about
a science. Everything has problems. We need to know about what
its problems are, so we can make it better.
But the academy recognized that with regard to many of
these technologies, they said, and I quote, ``The simple
reality is that the interpretation of forensic evidence is not
always based on scientific studies to determine its validity.''
Sometimes, like in the case of fingerprints, there's a lot
of information about them, but we still don't have the studies
that tell us about our ability to truly match fingerprints,
when you have variability on the surfaces it's been put on,
when you have partial prints. You get a match based on some
criteria, but there still aren't really objective criteria. And
there are two notable cases, one involving the Madrid bombing,
when fingerprint evidence pointed to the wrong suspect.
So we know that the error rate is not zero. We know it's
not perfect. It is not necessary to have a perfect technology.
The goal is to have a technology where we understand what it's
good for and what its weaknesses are, so we can weigh evidence
appropriately.
So, as I say, the big issue is often having a good method
for declaring whether things really match or don't match and
having a good way to attach probabilities, and that just takes
science.
What's the solution to all this? Based on my experience
with DNA fingerprinting, based on what I saw 23 years ago, I
know what the solution is. It is getting a collaboration
between the scientific community and the law enforcement
community working together.
The National Academy was unambiguous in its report. This
can't be done within the Department of Justice alone. Now,
there's nothing wrong with the Department of Justice. There are
fantastic public servants there, but the people who are
practicing a technology and using it day by day in law
enforcement can't be the people who can stand back and
objectively say what's wrong with it, what are the problems
with it.
You can't have the same community both be the advocacy
users and the skeptics about a technology. It's a marriage of
the users and independent skeptics working together that make
things better.
How do we fix it? Well, look, in my opinion--and I'm going
to emphasize, despite co-chairing the President's Council, I'm
here today as an individual. I'm not speaking on behalf of the
administration, but I'll give you my opinion about it, which is
that we need a partnership between the DOJ and these two
agencies here, NIST and NSF.
With regard to setting the standards in forensic science,
there's no doubt that the DOJ clearly has an essential role in
identifying the most important needs and in promoting the
widespread adoption of standards.
But there's also no doubt in my mind that NIST should
clearly take the lead in identifying where our gaps in research
are, where the weaknesses are, and in developing and proposing
specific standards and best practices for measurement, for
analysis and for interpretation. The two agencies need to work
together, but each needs to lead in its own respective domain.
There are many ways one can organize to do that through
appropriate task forces led in one or the other agency, and I'm
not going to suggest how to micromanage that, but we clearly
need clear and crisp processes that will accomplish those two
distinct but complementary goals.
With regard to research--forensic science research--we need
a robust scientific research agenda to develop the most
important body of empirical evidence to be used, the most
effective technologies to be used.
So the NIJ provides some limited funding for forensic
science research, and that's a good thing, but I think the NSF
has a critical role to play in supporting basic research
underlying forensic science. The setting of that agenda must
surely be a collaboration between the law enforcement community
that says here are the things we desperately need and the
scientific community that says here's how we can find those
things out.
So I'm going to say for both NIST and NSF, I don't want to
create unfunded mandates. I hope, in fact, that both of these
agencies will proceed to do this, and I hope they will have the
additional resources necessary to be able to do these well,
because this serves justice overall, prosecutors, defense, and,
most importantly, the whole American people.
So, in any case, in closing, based on my experiences of a
very successful situation 23 years ago, I think it is possible
to bring together these two cultures. I think we can make
tremendous strides in advancing the quality of forensic
science.
I am sure there are people today in the law enforcement
community who will worry about how will all this science weaken
the tools. I think they may have it backwards. I think if this
collaboration happens it will strengthen the tools. These tools
will become more powerful by being better understood. They will
become cheaper to use. We'll be able to use them in local
jurisdictions. They will make them efficient, will make them
reliable. I think that is a win for everybody.
I'm pleased to see the activity in both the executive
branch that I've gotten to observe in my role as the Chair of
the President's Council, and here with this committee's own
interest and its attention in the legislative branch to these
important problems. And with everybody's continued attention, I
think we can enlist the full power of science in the service of
justice.
Thank you.
[The prepared statement of Dr. Lander follows:]
Prepared Statement of Eric S. Lander, Ph.D., President and Founding
Director, Broad Institute of Harvard and MIT; Professor of Biology,
MIT; Professor of Systems Biology, Harvard Medical School; Co-chair,
President's Council of Advisors on Science and Technology (PCAST)
Chairman Rockefeller, Ranking Member Hutchison, and Members of the
Committee:
Thank you for inviting me here today to speak to you about an issue
of tremendous importance for our nation and our justice system:
ensuring the quality and consistency of forensic science relied upon in
criminal proceedings.
My name is Eric Lander. I am the President and Founding Director of
the Broad Institute of Harvard and MIT, which was the leading
contributor to the International Human Genome Project a decade ago and
works today at the forefront of genomic medicine. I am also the co-
chair of President Obama's Council of Advisors on Science and
Technology (PCAST), which is the external scientific advisory group to
the White House. I want to emphasize, however, that I am not here today
to represent the Administration's position. Rather, I have been asked
to testify based on a longstanding personal interest that traces back
23 years, to my involvement in the earliest days of DNA fingerprinting.
Today, we consider DNA fingerprinting to be the gold standard for
forensic science. It's a staple on television in the fictional crime-
solving on ``CSI'' and on ``Law and Order''; and in reality, it is a
technology with amazing sensitivity and near-flawless accuracy.
But, this wasn't always the case.
In 1989, I participated in one of the first DNA fingerprinting
cases in the United States--a New York case called People v. Castro.
Because DNA fingerprinting was such a new technology and I was a
molecular geneticist with expertise on the human genome, the defense
asked me to review the evidence and to testify in a pre-trial hearing
on the admissibility of the DNA evidence. I did so reluctantly and
insisted on doing so pro bono.
To make a long story short, the evidence turned out to be
appalling. There were no objective standards for declaring when two DNA
bands matched; for deciding when non-matching bands could be ignored as
``noise''; or for calculating the probability of a match. The testing
labs were issuing breathtaking statements that particular DNA patterns
had frequencies of less than one in 10 billion--in effect, asserting
that they were unique, despite the lack of any rigorous support for
these claims.
The pre-trial hearing lasted for 15 weeks. Near the end, the
scientific experts who had testified for the defense and the
prosecution took an unusual step--unusual, at least, for the legal
system. We decided to have a one-day joint scientific meeting to review
the evidence together, without the lawyers or judges.
At the end of the day, the scientific experts for the prosecution
agreed with those for the defense that the DNA evidence was
unacceptable. They decided to switch sides and testify for the defense.
Needless to say, the judge excluded the DNA evidence--deciding that DNA
fingerprinting was reliable in theory but not as practiced.
It was a triumph of the scientific method and the scientific
culture.
Following the case, I worked with others to ensure that we had
reliable standards for DNA fingerprinting. I had the pleasure to work
closely with extraordinary public servants in the FBI's Crime Lab,
including Bruce Budowle, of the FBI's unit at Quantico. And, I served
on the first of two committees assembled by the U.S. National Academy
of Sciences on DNA fingerprinting. Sometime later, I also agreed to
serve on the Board of the Innocence Project.
Within about five years, DNA fingerprinting was put on firm
foundation--through a robust collaboration of law enforcement on the
one hand and independent scientists on the other. It was the alchemy of
rigorous scientific attention that turned DNA fingerprinting from base
metal into the gold standard it is today.
At the beginning, the law enforcement community had serious
concerns about inviting independent scientists to set standards because
they worried that it might weaken DNA fingerprinting as a law
enforcement tool. In fact, DNA became a stronger tool for the police
and prosecutors--making it possible to revive cold cases, to catch
serial rapists and murderers. And, DNA also became a stronger tool for
the defense to protect those who were wrongfully accused.
In the end, DNA became a tool not for the prosecution or for the
defense, but for the truth, which is the main goal. When we fail to
find the truth, we may fail society in two ways--by locking up an
innocent person and by leaving a criminal free to commit more crimes.
The power of DNA fingerprinting had another unexpected and very
important consequence. For the first time, it gave us a way to revisit
old cases and to prove that some people had been wrongfully convicted--
to prove that hundreds of people in jail were actually innocent; to
prove that at least 17 people who had been on death row were actually
innocent; and to infer that, in all likelihood, at least some people
who had been executed were actually innocent.
Because many of these wrongful convictions involved forensic
science, it became important to ask how the forensic science testimony
could have been wrong. The goal here is not to point fingers. The goal
is to identify errors, understand the reasons and improve the science
so that it is accurate. That's how science advances in research labs
and in clinical labs. And, it is how science must advance in the
justice system.
We have learned a lot, both from legal cases and from scientific
studies, about the need for improving forensic science.
A paper by Garrett and Neufeld in 2009 reported that, in 137 cases
where transcripts of forensic testimony were available and a convicted
person was later exonerated by DNA evidence, roughly 60 percent
involved problematic forensic testimony.
The cases included ones like that of an honorably discharged
veteran who was wrongly convicted of murder in Arizona based in part on
a comparison of a Styrofoam impression of his teeth with bite marks on
a murder victim's neck. DNA testing eventually led to the veteran's
exoneration in 2002. (In fact, the actual perpetrator went on to attack
a young girl 20 days after the murder, a crime that might have been
prevented had the police had the right suspect.)
In another illuminating case, a man was convicted of rape and
murder in New York, in part on the basis of hair analysis, soil
comparison, and fabric print analysis. The forensic expert reported
similarities of hair, soil and fabric prints from the man's truck and
from the crime scene and victim. Yet, there were no empirical data on
the frequency of those materials, so no way to know how common such
characteristics or ``matches'' might be. DNA testing eventually
exonerated the man nearly 20 years after his conviction.
In 2009, the National Academy of Sciences issued an important and
thoughtful report about strengthening forensic science. It cited
serious issues with the analysis and interpretation of forensic
evidence.
It cited, for example, an FBI study that found that \1/8\ of hair
samples said to ``be associated'' based on microscopic comparison were
subsequently found to come from different people based on DNA analysis.
It noted serious issues with bite marks, tool marks, and fiber
comparisons, including the lack of objective standards and the lack of
meaningful data and databases from which the probability of matches can
be inferred. It identified issues with fingerprints, whose evidentiary
value depends importantly on the quality of the latent fingerprint
image and for which fully validated analysis methods are still needed.
The report stated that: ``With the exception of nuclear DNA
analysis, however, no forensic method has been rigorously shown to have
the capacity to consistently, and with a high degree of certainty,
demonstrate a connection between evidence and a specific individual or
source. In terms of scientific basis, the analytically based
disciplines generally hold a notable edge over disciplines based on
expert interpretation. But there are important variations among the
disciplines relying on expert interpretation. For example, there are
more established protocols and available research for fingerprint
analysis than for the analysis of bite marks. There also are
significant variations within each discipline. For example, not all
fingerprint evidence is equally good, because the true value of the
evidence is determined by the quality of the latent fingerprint image.
These disparities between and within the forensic science disciplines
highlight a major problem in the forensic science community: The simple
reality is that the interpretation of forensic evidence is not always
based on scientific studies to determine its validity. This is a
serious problem. Although research has been done in some disciplines,
there is a notable dearth of peer-reviewed, published studies
establishing the scientific bases and validity of many forensic
methods.'' [Emphasis added].
I should emphasize that the problem is often not with the
technology per se. As we saw with DNA fingerprinting, it is often that
there is a lack of serious scientific standards for analysis and
interpretation--that is, (1) methods for deciding that two samples are
similar matches and (2) methods and databases for attaching meaningful
probabilities to such similarities. Without scientific standards for
measurement, analysis and interpretation, expert opinion is not
scientific and thus not meaningful in court.
What is the solution? As it was with DNA fingerprinting, the answer
lies in drawing on two cultures--the criminal justice community, which
understands most fully the needs for and uses of forensic evidence, and
the independent scientific community, which understands most fully the
principles of rigorous scientific analysis.
The National Academy of Sciences report was unambiguous that the
task could not be accomplished within the criminal justice community
alone. In particular, it concluded that ``advancing science in the
forensic science enterprise is not likely to be achieved within the
confines of the [Department of Justice]''. The National Academy report
went so far as to recommend the creation of an independent National
Institute of Forensic Sciences, within or associated with a science-
based agency.
For my part, I think that it may be possible to achieve these goals
through a partnership between the DOJ and two science-based agencies,
NIST and NSF. But, it will be important that the partnership have clear
and complementary roles.
[1] With respect to standards for forensic science:
DOJ clearly has a central role in (i) identifying the most
important needs for forensic measurement, analysis and interpretation,
and (ii) promoting the widespread adoption of good standards for
forensic science throughout the justice system.
NIST clearly should take the lead in (i) identifying research gaps
and weaknesses in forensic science and (ii) developing and proposing
specific standards and best practices for forensic measurement,
analysis and interpretation.
The two agencies should actively engage the other in the work, but
it is important that the distinct activities have distinct leadership.
Scientific standards should be based on robust input from the broad
scientific community--not simply the input of forensic scientists or
practitioners. As emphasized in the report from the National Academy of
Sciences, scientific standard-setting should be led by a science-based
agency such as NIST, not units within DOJ. Conversely, the adoption of
standards requires the perspective of practitioners. It should be led
by DOJ.
In my opinion, the partnership between NIST and DOJ should be
formalized through appropriate advisory committees or task forces with
assigned responsibilities.
[2] With respect to forensic science research:
We need a robust scientific research agenda to support the
development of a body of empirical knowledge on the validity of
technologies and methods. This would greatly help the cause of
advancing the status of forensic science.
While the National Institute of Justice (NIJ) provides some support
for forensic science research, the program has very limited funding and
engages a very limited scientific community--both in its grantees and
its peer reviewers.
I believe that NSF has a critical role to play in supporting basic
research underlying forensic sciences. The NSF engages the full breadth
of the U.S. scientific community in both research and peer review.
For both NIST and NSF, I do not want to create unfunded mandates. I
believe that some additional funding will be required to NIST and to
NSF to carry out these roles with respect to forensic science.
In closing, based on my experiences with the evolution of DNA
fingerprinting, I believe it is possible that by bringing together the
two cultures of science and justice, we can make large strides in
advancing the quality of forensic science.
Again, I speak only for myself here. But, I am pleased to see that
both the Executive and Legislative branches have become increasingly
attentive to the issues of ensuring quality and consistency in forensic
science. I am very hopeful about the various activities underway in
both branches--including an ongoing process within the National Science
and Technology Council, discussions in recent months among
representatives of the departments and agencies that have equities in
forensic science, and the interest of this Committee. With everyone's
continued attention, we can enlist the full power of science in the
service of justice.
Thank you.
The Chairman. Thank you, and you've given me some questions
to ask you.
Now, we should go to you.
STATEMENT OF PATRICK D. GALLAGHER, Ph.D.,
UNDER SECRETARY OF COMMERCE FOR STANDARDS AND
TECHNOLOGY, U.S. DEPARTMENT OF COMMERCE
Dr. Gallagher. Thank you very much, Mr. Chairman and
Ranking Member Boozman, for your leadership on this topic, and
I want to include Senator Udall, and----
The Chairman. Yes, I forgot to introduce him. He's from
some state.
Senator Udall. A very important state.
The Chairman. New Mexico, right? And he's just pure gold.
And he was the third person who was admitted to this august
dais today. There'll be no more.
Dr. Gallagher. As an Albuquerque native, I'm a little bit
biased.
The Chairman. Oops.
[Laughter.]
Dr. Gallagher. But I have to agree.
It's a real pleasure to be here today and to discuss the
role of measurement science and forensics. As you know, NIST
has a specific mission and that mission is to define a uniform
scientifically based national system of measurement and to
support those who have to use that system of measurement,
whether it's industry or whether it's other Federal agencies,
or practitioners.
The scientific basis for accurate measurements using the
most rigorous, soundly defensible and universally accepted
science that gives accurate, reproducible and reliable
measurements underpins any system like this.
The hallmark of NIST mission in measurement science is that
there is a scientific basis for every measurement, and a well-
defined system of traceability to that basic unit of
measurement, so that the uncertainty and precision of the
measurement can be defined and understood.
NIST also supports the quality and integrity of the
measurement system, including forensic measurements, and this
includes services like providing validation of methods,
performing primary calibration services, providing calibration
artifacts, such as Standard Reference Material, standards data
and supporting laboratory accreditation programs.
In the context of this mission, NIST has always played a
role in supporting forensic science. In fact, as early as 1913,
NIST was the nation's de facto criminal forensic science
laboratory. And it was NIST, then the National Bureau of
Standards, that the FBI turned to in 1932 to help them
establish their laboratory and train their scientists in the
principles of forensic investigation.
In fact, it was an NBS scientist whose analysis of the
ransom letters from Charles Lindbergh helped lead to the
conviction of that kidnapper.
Today, by comparison, the range of measurements used by our
law enforcement community is extraordinarily broad. There are
nearly 400 forensic laboratories in the United States. More
than 90 percent of those are at the state and local level, not
the Federal level, and they are dedicated to some aspect of
forensic science.
The NIST programs provide a wide range of services to
support these laboratories across a broad range of
measurements, including chemical analysis, biological,
radiological and nuclear detection and analysis, fire and
explosives analysis, gunshot residue, latent fingerprint
analysis, biometrics, digital evidence and many other areas.
Our laboratory program validates the performance of
measurements and provides services to help laboratories and
practitioners assure the equality of those measurements, so
that forensic specialists can reliably, routinely and
repeatedly provide the services they are called upon to
provide.
In the area of forensic science, NIST is perhaps best known
for our work in DNA. One of NIST's researchers, John Butler, is
a leading expert in this area and has worked developing a new
DNA analysis approach which uses smaller fragments of DNA than
ever before, was essential in helping to identify many more of
the victims of the September 11 attacks----
The Chairman. Did he go to Stanford?
Dr. Gallagher. John Butler?
The Chairman. Yes.
Dr. Gallagher. I'd have to look at his bio.
The Chairman. Yes. Because I may know him. I'm just--I'm
sorry. Just struck me.
Dr. Gallagher. The NIST work in DNA profiling, testing
helped establish the methods and support the methods now
routinely used across all crime labs to match individuals to
evidence samples.
The FBI requires that forensic DNA labs use this Standard
Reference Material to calibrate equipment before any of the
data can be entered into the National Criminal DNA Data base.
The National Institute of Justice also requires that crime
laboratories it funds use the same methodologies and tools.
Our work in DNA analysis is the gold standard in forensic
science, and what makes that true is the scientific rigor and
grounds of that work. The measurements are so precise that a
DNA sample for any one individual can be accurate to a very
high level. No other area of forensic science has achieved that
level of precision. And what we stand ready to do is to bring
our measurement science expertise and approach to many other
areas of forensic science.
NIST is working to identify sources and to develop standard
procedures for minimizing the chance of error in impression
analysis, a very difficult area, including fingerprints and
ballistics.
We also have a technical working group on biological
evidence preservation. Some of the resources NIST has developed
in this field include databases such as our latent print
database, our short tandem repeat DNA, Internet database and
the world's largest database of literature related to DNA
research.
We also have expertise in cell phone and computer forensics
including the recovery of deleted files and logs.
Of course, with growing demand in this area, our budget
request for 2013 included a specific increase in this area.
Mr. Chairman, this measurement science and standards role,
our expertise and our dissemination of both the research and
the tools to support the practice of measurement is a key part
of our mission. We look forward to aligning this, so that it
can support the forensic science community.
And I want to thank you, once again, for this opportunity
to discuss this with you.
[The prepared statement of Dr. Gallagher follows:]
Prepared Statement of Patrick D. Gallagher, Ph.D., Under Secretary of
Commerce for Standards and Technology, U.S. Department of Commerce
Chairman Rockefeller Ranking Member Hutchison, and Members of the
Committee, thank you for the opportunity to appear before you today to
discuss the importance of forensic science. The Department of
Commerce's National Institute of Standards and Technology (NIST) has a
long history of collaboration in the area of Forensic Science. In the
Fiscal Year 2013 (FY 2013) budget NIST has requested $5 million for an
initiative that will enable NIST to create a strategic program to
broadly address the most critical issues in Forensic Science today.
NIST's Role in the Forensic Sciences
NIST was founded with a specific mission--to define and advance a
uniform, scientific, national system of measurement to support industry
and other Federal agencies. This system of measurement is underpinned
by NIST's measurement science research. This scientific basis for
accurate measurements using the most rigorous, soundly defensible, and
universally accepted science gives accurate, reproducible, and reliable
measurements. In this context, Forensic Science has always been part of
NIST, since much of Forensic Science is about forensic measurements.
Measurement and forensic scientists are bound by mutual interests
in accuracy and uncertainty, a quantifiable expression of the quality
of our measurements. NIST works to resolve the uncertainty as it
pertains to all types of applied sciences. Resolution of uncertainty
will lead to the accuracy that is necessary in many applications,
including Forensic Science. Some of the other areas NIST has
measurement expertise in that have applicability in Forensic Science
are dimensional analysis, chemical and material analysis, DNA,
structural fire analysis, radiation signatures and digital data.
One of the founding principles for NIST is establishing
traceability in the marketplace for measurement. The work NIST does
with measurement standards and their traceability to NIST research
provides the crucial framework for measurement.
Justice can, in some instances, quite literally hang on a single
thread, or in the parlance of forensic scientists, a single fiber.
Forensic scientists are under a tremendous amount of pressure to not
only get it right but also to explain methodologies and results to a
judge and jury. NIST can and does provide metrics to help define the
resolution of methods and the veracity of the results.
The next piece of the NIST mission is our role in standards. NIST's
measurement research allows NIST to inform the standards function and
make sure that the standards are realistic and scientifically valid, in
this case for use in labs and the field. For NIST to perform our
standards role well, we must have independent measurement science
research in the appropriate disciplines of forensic science.
The Past, Present and Future of Forensic Science Measurement and
Standards at NIST
NIST has supported forensic science throughout our history. In
fact, from 1913 until the Federal Bureau of Investigation (FBI) hired
its first scientist in 1932, NIST was the Nation's de facto criminal
forensic science laboratory. Our involvement in the forensic sciences
originates with Wilmer Souder--one of the Nation's best and least known
criminologists to whom the FBI turned in 1932 to help them establish
their lab and train their scientists in the principles of forensic
investigation.
Souder's interest in forensic science began in 1913, when famed
document examiner Albert Osborn sent some precision measuring devices
to NIST for calibration. By the 1930s Souder had become a pioneering
expert in the identification of questioned documents, handwriting,
typewriting, bullets, cartridge cases, and firearms. In his nearly 40
years at NIST, he assisted almost 1,000 Federal investigations of
crimes, including extortion, forgery, kidnapping, murder, bootlegging,
and theft.
Perhaps most famously, Souder was among the handwriting experts
whose analyses of the ransom letters helped to convict Bruno Richard
Hauptmann for the kidnapping and murder of Charles Lindberg, Jr.
NIST continues this long history of work in support of law
enforcement. We have worked with the Department of Justice, the
Department of Homeland Security, and the Department of Defense toward
the development of standards for body armor, nonlethal weapons, and
explosives detection technologies, among others.
In the area of forensic science, we are perhaps best known for our
work in DNA analysis. One of our researchers, John Butler, Ph.D.,
Leader of the Applied Genetics Group, literally wrote the book
(actually, he wrote four books with another on the way) on forensic DNA
typing. Butler's work, developing a new DNA analysis approach which
uses small fragments of DNA, was essential in helping to identify the
victims of the September 11, 2001 attacks on the World Trade Center.
NIST continues its work to further improve techniques for
identifying severely degraded DNA and advance the state-of-the-art for
forensic DNA typing. NIST also produces Standard Reference Materials
for calibration and quality control for forensic science and genetics
laboratories throughout the United States and the world. NIST's work in
genetic kinship analysis made it possible for police in California to
catch a killer known as the ``Grim Sleeper,'' who had been at large for
more than 20 years.
The FBI already requires that forensic DNA labs use NIST's Standard
Reference Materials (SRMs) for quality assurance before they may enter
their data into the national criminal DNA database. The National
Institute of Justice (NIJ) also requires that the crime laboratories it
funds use these SRMs.
Many broad aspects of NIST's work have applicability in forensic
science. Measurement is the comparison of a known to an unknown, and
NIST's job is to supply forensic science labs with as many knowns as
possible by actively offering our measurement expertise and continually
working with the community to help them do their jobs more effectively.
Some of the resources NIST has developed that are used in the field
include databases such as our mass spectroscopy database and our latent
print database. We also have expertise in cell phone and computer
forensics, including the recovery of deleted files and logs.
Additionally we have fire research and arson investigation expertise
that have provided assistance in major investigations such as the World
Trade Center building collapses, the Rhode Island nightclub fire, and
the Chicago high-rise fire, as well as an extensive array of fire
modeling software. We've been performing fire research for a very long
time. We have provided guidance in fire research by initiating the
compilation of best practices, resulting in the 1980 publication of the
Fire Investigation Handbook, and through this publication, entered into
a close partnership with the National Fire Protection Association.
NIST is working to identify sources and develop standard procedures
for minimizing the chances of error in impression analysis, including
fingerprints and ballistics. We also have a technical working group on
biological evidence preservation.
The work done by the NIST can help to establish a more solid
scientific basis for comparing samples and interpreting the types of
evidence mentioned earlier. A more scientific basis for comparison will
give the forensic science and law enforcement community a better
understanding of how well those interpretations can be trusted. The
goal is to provide a vocabulary that will help define the limits of
certainty so police officers and forensic scientists can testify before
a jury and say this evidence came from that suspect with a quantified
confidence.
There are nearly 400 labs in the U.S. dedicated to some aspect of
forensic science. These labs operate under a variety of standards,
mandated at the state or local level and that may be unique to each
department.
There is, of course, much to be said for expertise, but even
experts can make mistakes. This is why standards are important.
Standards unite our efforts and help us to speak with one voice. They
bolster trust. They set a minimum level of performance, a baseline for
defining success, and a vocabulary for expressing degrees of confidence
with consistency and objectivity.
Measurement Science and Standards in Support of Forensic Science in
Fiscal Year 2013
The $5 million initiative proposed in Fiscal Year 2013 request will
enable NIST, in coordination with DOJ, to create a strategic program to
oversee and manage standard development in forensic science.
Forensic science must deal with an incredibly wide range of
interdisciplinary fields, from DNA sequencing to electron microscopy to
the visual matching of patterns like footprints or tool marks. Often
evidence samples are degraded, incomplete, or available only in very
small amounts, which also presents challenges for developing the full
range of measurement tools required for ensuring confidence in results.
In 2009, a committee of the National Research Council (NRC) made a
number of important recommendations for strengthening the public's
trust in forensic science findings. The recommendations included strong
support for improved measurement and validation methodologies,
development of additional forensic science standards, and dissemination
of best practices to strengthen the precision and reliability of
forensic science analyses.
NIST's work in advancing forensic science led the NRC to explicitly
name NIST as one of several Federal agencies that should collaborate on
developing new forensic science measurements and standards. Working
with NIJ and other agencies through reimbursable funding, NIST has
measurement science research under way in chemical, biological,
radiological, and nuclear detection and analysis; fire and explosives
analysis; gunshot residue, latent fingerprints, and many other areas.
NIST's work in DNA profiling and testing, for example, helped establish
the methods now used by all crime laboratories to match individuals to
evidence samples.
With the requested $5 million initiative, NIST will be able to
develop state-of the art measurement science and standards as the basis
for forensic disciplines and technologies. Working with stakeholders,
NIST has identified critical areas of investment that will be
complementary to current research. It will also provide practitioners
with analyses in disciplines that require more research in the near
term, including areas in which quality control is acknowledged as the
most pressing issue, and in which significant investment in human
capital or equipment is necessary to make an impact. Examples of
priority program areas in this new initiative include: new reference
methods and technologies for understanding crime scenes and identifying
criminals, including the uncertainty and standards associated with
those techniques; improved calibration systems, reference materials and
databases, and technology testbeds for ensuring reliable and accurate
forensic science practices; and development of rigorous training
programs.
A major outcome of this initiative will be to strengthen the
utility and reliability of forensic science evidence in the courtroom.
This work also has the potential for significant cost savings for the
U.S. justice system by reducing the number of mistrials and appeals
related to questions about forensic science analysis. One economic
analysis of cost savings from forensic DNA testing alone estimated a
cost savings of $35 for every dollar invested; the same analysis
predicted that if DNA testing were fully utilized the United States
could expect a $12.9 billion annual savings in prevented crime.\1\
---------------------------------------------------------------------------
\1\ Butler, John. ``Fundamentals of Forensic DNA Typing,'' Academic
Press 2009, p.261.
---------------------------------------------------------------------------
NIST anticipates additional impacts to include new, innovative
forensic science technologies; increased use of documentary standards
and measurement services by the forensic science community; and the
creation of reference materials, reference databases and new
calibration services to improve the consistency of the implementation
of forensic science across the Nation.
In conclusion, public trust in the justice system relies on the
validity and certainty of evidence presented to the courts.
Increasingly, that evidence is gathered and analyzed with innovative
forensic science technologies. Any time a new technology is developed,
accurate measurements, standards, and uncertainty estimates are needed
to ensure that the technology works as intended.
That is where NIST's expertise in the forensic sciences is
critical, and our Fiscal Year 2013 request will build a stronger
forensic science program at NIST.
Thank you again, for the opportunity to testify today, I would be
happy to answer any questions you may have.
The Chairman. Thank you very much.
Dr. Suresh.
STATEMENT OF DR. SUBRA SURESH, DIRECTOR,
NATIONAL SCIENCE FOUNDATION
Dr. Suresh. Thank you, Mr. Chairman, Ranking Member
Boozman, Senator Udall, thank you so much for inviting me to
testify today.
Mr. Chairman, I also want to take this opportunity to thank
you again for your support of science, and also of the National
Science Foundation.
As you well know, NSF supports basic research and education
at the frontiers of knowledge in all fields of science and
engineering and at all levels of education, science and
engineering education.
Many of NSF's activities contribute directly to building
the human capital, the infrastructure and advanced methods
needed to ensure the vigor and vitality of the forensic
sciences.
NSF supports significant basic research that may be applied
in forensic settings. Supported research investigates the
effectiveness of currently employed forensic science approaches
and explores potential applications of cutting edge theory and
technologies.
NSF awards across the foundation support training and
activities and programs which directly address the need of the
21st century forensics workforce.
A search of recent NSF awards shows that the foundation has
supported 147 awards just in the period 2009 to 2011 that
contribute to the strengthening of the forensic sciences. So in
keeping with my One NSF philosophy, each of the foundation's
seven directorates contributes to this effort.
The awards represent many facets of NSF activity including
basic research awards, major research instrumentation, small
business innovation research, student support, as well as
workshops.
Just in this period from 2009 to 2011, more than $50
million of research has been awarded to institutions in 36
states and in the District of Columbia, large and small
colleges and universities, EPSCoR states, minority-serving
institutions, community colleges and small businesses.
Let me provide you with just a taste of our activities in
support of the forensic sciences. Our data analysis also shows
that there are more than 200 current awards that are supported
by NSF.
With support from the Social, Behavioral and Economic
Sciences Directorate, or SBE, researchers at the University of
Arkansas are investigating how to overcome obstacles to the
assessment of likely age changes in facial features.
An award by the Computer and Information Science and
Engineering Directorate is using computer approaches to
handwriting examination, which contributes to the scientific
analysis of documents of questioned authorship.
NSF has long used workshops to identify cutting-edge
opportunities for future directions. In fact, after the NRC
report was published in 2009, NSF-supported workshops including
one on cognitive bias and forensic science, that was at
Northwestern University and another one on nanoscale science
and technology for forensics.
NSF supports activities designed to achieve excellence in
U.S. science education. Students participate in supported
research and thereby gain skills that are transferable to crime
labs.
Some awards specifically expose students to research in a
forensic setting. A project at Tuskegee University, Auburn
University, as well as Mississippi State University provides
occupational training to America's veterans in digital
forensics.
Other awards, including one at Arkansas State University,
capitalize on the popularity of shows such as CSI to engage
students in science.
NSF provides funding for small business innovation research
to stimulate technological innovation in the private sector,
and a number of awards support commercial development of
technologies applicable to forensic settings.
Likewise, investments in infrastructure provide databases
and instrumentation used in forensic applications and research.
NSF also works collaboratively with other agencies. The
award that supports training of veterans was made in
coordination with the Department of Veterans Affairs.
Our science staff serves on the National Science and
Technology Council Subcommittee on Forensic Science, and SBE,
our Directorate on Social, Behavioral and Economic Sciences, is
developing a memorandum of understanding with the National
Institute of Justice to facilitate support of relevant forensic
sciences.
So, in summary, NSF has supported and is committed to
continue supporting the basic sciences that form the foundation
for forensic applications, to collaborate with other mission
agencies and to support science education opportunities
necessary for the 21st century, especially in the area of
forensic sciences.
Thank you, Mr. Chairman. I'll be happy to answer any
questions.
[The prepared statement of Dr. Suresh follows:]
Prepared Statement of Dr. Subra Suresh, Director,
National Science Foundation
Introduction
Chairman Rockefeller, Ranking Member Hutchinson, and distinguished
Members of the Committee, thank you for inviting me to participate in
this hearing on ``The Science and Standards of Forensics.''
I am pleased to have the opportunity to discuss the National
Science Foundation's (NSF) investments that strengthen the forensic
sciences in the United States.
As you well know, NSF supports research at the frontiers of
knowledge across all fields of science and engineering (S&E) and all
levels of S&E education. Its mission, vision and goals are designed to
maintain and strengthen the vitality of the U.S. science and
engineering enterprise. In this role many of NSF's activities
contribute directly to building the human capital, infrastructure and
advanced methods needed to ensure the vigor and quality of the forensic
sciences.
NSF is supporting significant basic research that may be applied in
forensic settings both in the near and longer term. Supported research
investigates the effectiveness of currently employed forensic science
approaches and also explores potential applications of cutting edge
theory and technologies. Activities in NSF's Education and Human
Resources Directorate, as well as in basic science directorates,
support training programs and activities which directly address the
need for a 21st century forensics workforce.
A search of the NSF Awards Abstracts Database identifies 210 active
awards using the search-term ``forensics.'' Each of the Foundation's 7
directorates is represented in this sample of awards. Of these awards,
147 were made in the years 2009-2011 and several awards have been made
thus far in 2012. The awards represent many facets of NSF activity
including basic research awards, Major Research Instrumentation, Small
Business Innovation Research, Doctoral Dissertation Improvement awards,
Research Experience for Undergraduates, and Workshops. For 2009-
2011alone, the awards total in excess of $53 million and awards were
made to institutions in 36 states and the District of Columbia. Awards
were made to large and small universities, state and private
universities, minority-serving institutions, small liberal arts
colleges, community colleges and a number of small businesses. Awards
have also included collaborations with international scholars.
After a brief discussion of background issues I will provide you
with a number of examples of our activities in support of the forensic
sciences. I will also point to several actions currently underway at
the Foundation that should enhance our contribution to this effort.
Background
In 2009, the National Research Council (NRC) published
``Strengthening Forensic Science in the United States: A Path
Forward.'' The report was prompted by the Senate's concern in 2006 that
``. . .there exists little or no analysis of the remaining needs of the
(forensic science) community outside the area of DNA.''
The NRC report goes on to indicate areas where it determined there
to be significant challenges facing the forensic science community:
lack of mandatory standardization, certification,
accreditation
disparities between local, state and Federal laboratories
insufficient funding for instrumentation
unacceptable backlogs
The most significant comment that has direct relevance to NSF is
that
. . . forensic science . . . research, education, and training
lack strong ties to our research universities. The forensic
science system is underresourced also in the sense that it has
only thin ties to an academic research base that could support
the forensic science disciplines and fill knowledge gaps (pg
15).
Further, the report advocates for investment in research:
. . . of the various facets of underresourcing, the Committee
is most concerned about the knowledge base. . . . [There are]
fundamental limitations in the capabilities of forensic science
disciplines to discern valid information from crime scene
evidence (pg 15; emphasis added).
Activities at NSF that contribute to Strengthening Forensic Science
Workshops
NSF has long used workshops and other small gatherings of scholars
and members of relevant communities to discuss cutting edge ideas and
to identify and investigate gaps in knowledge and to propose future
directions. In the area of forensic science, NSF has supported several
workshops in the recent past; the NSF awards database lists 11 active
awards containing the key words ``forensic science'' and ``workshop.''
Two recent examples:
In direct response to a recommendation of the NRC report that
research on human observer bias be encouraged, the Behavioral
and Cognitive Sciences Division of the Social, Behavioral and
Economic Sciences (SBE) Directorate supported ``Cognitive Bias
and Forensic Science'' at Northwestern University in September,
2010. The workshop brought together lawyers, forensic
scientists, and academic researchers in the area of cognitive
bias to examine the role that psychological factors may play in
forensic pattern recognition. The report of the workshop is
available at http://www.law.northwestern.edu/faculty/
conferences/workshops/cognitivebias/. In line with the workshop
goal to ``. . . convert general theories and testable
hypotheses into concrete research proposals'' attendees
continue planning the development of joint research projects.
The Division of Electrical, Communications and Cyber Systems/
Directorate for Engineering (ENG) supported a workshop in
August 2011 on ``Nanoscale Science and Technology for
Forensics'' at the University of Connecticut. ``The workshop
assembled key experts from nanotechnology areas
(optoelectronics, materials, fabrication, engineering and
medicine) to focus on applications in forensic science.'' This
multidisciplinary meeting was designed to advance
identification of future research needs and to promote new
collaborations. The workshop also established recommendations
for the development of programs for training graduate and
undergraduate students to become the next generation of
forensic scientists and engineers. A special effort was made to
include student attendees.
Training Activities
NSF supports numerous activities designed to achieve excellence in
U.S. science, technology, engineering and mathematics (STEM) education
at all levels and in all settings (both formal and informal) in order
to support the development of a diverse and well-prepared workforce of
scientists, technicians, engineers, mathematicians and educators, as
well as a well-informed citizenry. This is certainly the case in the
realm of forensic science. Many students participate in NSF supported
research and thereby gain exposure to the conduct of research and some
of these students ultimately focus their attention and career in a
forensic science. In addition, there are a number of awards that
specifically expose students to research in a forensics setting.
Some awards capitalize on the popularity of shows such as CSI to
engage students in science. One example is an award to Arkansas State
University titled ``CSI: Classroom Student Investigations'' that was
supported by the Division of Research on Learning in Formal and
Informal Settings/Directorate for Education and Human Resources (EHR).
This project ``. . . uses the popularity of the Crime Scene
Investigation television show . . . to train teachers in forensic
science topics and use that training in their science classrooms to
stimulate and encourage middle and upper school students in science
topics generally.'' Upon completion, the project will ``. . . examine
the impact on students' interest in STEM careers in classrooms of
participating teachers and examine how participation in the program
affects participating teacher implementation of reform based pedagogy
and technology.''
Other basic research projects, while not focused on educational
goals, include capacity building components. With support from the
Division of Mathematical Sciences of the Directorate for Mathematical
and Physical Sciences (MPS), researchers at Michigan State University
are investigating modeling and computational issues in fingerprint
analysis. A central question in fingerprint analysis is the
individuality of a person's prints. The whorls, ridges and valleys
present complex data that lead to the assumption of uniqueness. But in
a legal setting, there are significant questions as to what constitutes
a match when comparing a latent print from a crime scene and those of a
defendant. These researchers are developing computational models for
addressing the question of uniqueness and this may significantly
impacts how fingerprint evidence is reported and used for the
identification of individuals. Graduate students working with the
principal investigators will be equipped with the analytic, computing
and methodological skills that are necessary to perform high level
forensic research.
Basic Research
Numerous basic research projects have potential applications in
forensic science. Questioned Documents analysts attempt to extract
information from a document utilizing as many sources as possible,
including handwriting analysis. A recent Early Concept Grants for
Exploratory Research (EAGER) award by the Division of Information and
Intelligent Systems/Computer & Information Science and Engineering
(CISE) to a researcher at SUNY Buffalo is using computer approaches to
``Automatic Identification of Writer Accent and Script Influences in
Handwriting.'' The investigator is testing hypotheses with respect to
handwriting analysis by examining and analyzing the written works of
native and non-native writers of a particular script or alphabet.
One of the most recent awards, supported by the Division of
Behavioral and Cognitive Sciences of the Social, Behavioral and
Economic Sciences (SBE) Directorate, uses GPS to track vultures.
Vultures arrive early in the process of decomposition of human remains
and leave few clues to indicate their scavenging activity. This can
greatly complicate medico-legal death analysis. This doctoral
dissertation improvement award seeks ``to establish a predictability
model of likely vulture scavenging habitats using remote sensing
techniques and spatial and temporal statistics,'' the results of which
could have significant implications for the practice of forensic
pathology.
The Division of Chemistry (MPS) awarded funds to a researcher at
the University of Iowa to use surface enhanced Raman scattering (SERS)
for the detection of small molecules without the use of traditional
receptor--based surface chemistry. If successful, this process could
provide new means of detecting trace levels of drugs and biomolecules
and thereby enhance the sensitivity of forensic investigations.
The Division of Social and Economic Sciences (SBE Directorate)
funded researchers at the University of Arkansas, in collaboration with
researchers at the University of Central Lancashire in the United
Kingdom, to assess current methods of forensic age progression
(assessing likely age changes in facial features), including
identifying factors that influence the accuracy of age progression
methods. The PIs will also explore novel methods for creating and
presenting age progressed images that may improve the accuracy of
forensic identifications.
Data and Scientific Infrastructure
Forensic scientists benefit from access to large databases as they
attempt to analyze and interpret crime scene evidence. NSF has
supported a variety of data infrastructure projects in recent years
that generate valuable resources for forensic practitioners. These
projects also include support for the training of future scholars.
The Division of Behavioral and Cognitive Sciences (SBE) continues
to fund the Allele Frequency Database (ALFRED) at Yale University. This
database currently houses information on human genetic variation on a
global scale -ALFRED now has data on 663,602 genetic polymorphisms, 714
populations and more than 37,000,000 frequency tables (one population
typed for one site). These data are invaluable for investigating human
population structure, migrations and relationships, and can also be
utilized by forensic scientists.
The Biology (BIO) Directorate's Division of Biological
Infrastructure supports The Human Impact Pollen Database at the
University of Massachusetts, Boston. This searchable digital image
database of pollen from plants that are associated with human
activities is critical for investigating both past and current human-
environment interactions. This on-line database provides value for a
variety of disciplines including forensic identification.
The NSF also provides Major Research Instrumentation (MRI) grants
to support the development of specialized laboratories or the
acquisition of cutting-edge equipment that facilitates research and
training opportunities at U.S. institutions. Several such awards have
supported forensic science research and training in recent years.
The Office of Cyberinfrastructure provided MRI support to
Jacksonville State University for the development of a cybersecurity
laboratory facility to facilitate research and training activities in
digital forensic methods of analysis, among other relevant areas.
Civil, Mechanical and Manufacturing Innovation in the Directorate
of Engineering provided MRI support to Southern Illinois University at
Edwardsville for the acquisition of a 3D laser scanner and associated
modeling software that promote research and training on high resolution
photographic and three-dimensional coordinate data which is often used
by forensic scientists in the analysis of crime scenes. The
instrumentation is being used to test new modeling and analytic
approaches for investigating forensic sites in a multidisciplinary
context.
Finally, the Division of Chemistry/Directorate of Mathematical and
Physical Sciences has provided MRI support to Cleveland State
University for the purchase of a triple quadrupole/linear ion trap
liquid chromatograph mass spectrometer system. he instrument will
support a wide range of research and training activities, including the
ante- and post-mortem forensic analysis of drugs and other specimens.
Small Business Innovation Research
The NSF provides funding for Small Business Innovation Research
(SBIR) to stimulate technological innovation in the private sector and
to increase the commercial application of federally supported research
results. A number of awards in recent years have supported the
development of materials, algorithms, and instrumentation that have
significant implications for the practice of forensic science. Two
examples are provided below.
SBIR support was recently provided to DNA Polymerase Technology
Inc. for the development of novel enzymes that can aid the rapid
detection of pathogens via DNA detection and amplification. The
processes explored may aid forensic practice, where the acquisition of
small amounts of DNA in the context of inhibitors can present
challenges to identification.
NSF support was also provided to Deurion, LLC for SBIR development
of Surface Acoustic Wave Nebulization (SAWN) for use with mass
spectrometers. SAWN provides a means of ionization outside of the
laboratory with significant portability and ease of use. This technique
may improve law enforcement's ability to collect and analyze crime
scene materials.
Human Resources
In addition to funded research, training and workshops, NSF has
invested significant human resources in support of the forensic
sciences.
The National Science and Technology Council's (NSTC) Committee on
Science established a Subcommittee on Forensic Science in direct
response to the NRC report. NSF has been represented on the
subcommittee since its inception. The individual who attends the
Subcommittee meetings also co-Chairs the Research, Development, Testing
and Evaluation (RDT&E) Interagency Working Group (IWG). An NSF program
officer also serves on this working group.
NSF has provided input in numerous areas including discussions
about:
conducting merit-based peer review,
identifying and prioritizing research opportunities,
designing survey instruments for assessing current practices
and needs,
judging validity and reliability in laboratory sciences
Over the past several years SBE/BCS science assistants also aided
in the design of the Subcommittee's website (http://
www.forensicscience.gov/iwg.html) and supported the Interagency Working
Group on Outreach and Communication.
Near and Longer Term Activities
Collaborating With Mission Agencies
The NSF continues to work with other agencies to identify
opportunities for advancing the forensic sciences. As the recent past
indicates, many investigator-initiated projects directly address
scientific questions of importance in forensic settings.
NSF continues to be represented on NSTC Subcommittee on Forensic
Sciences and on its RDT&E IWG. The IWG has assessed the state of the
science in a number of forensic settings (e.g., latent print analysis,
questioned documents, fiber analysis, odontology) and is preparing
annotated bibliographies and other documents that will help to
elucidate foundational aspects of the forensic sciences and encourage
further scientific inquiry.
The NSF's Directorate for Social, Behavioral and Economic Sciences
is developing a Memorandum of Understanding with the National Institute
of Justice regarding research, development, and evaluation of social
and behavioral sciences as they pertain to legal and forensic matters.
Activities undertaken via the MOU will foster information-sharing about
the most promising areas of research in the social and behavioral
sciences, and serve as a catalyst to identify synergies and
opportunities for future collaboration.
Coordination within NSF
As clearly documented above, NSF invests significant human and
financial resources in advancing the forensic sciences. In order to
better coordinate our efforts going forward, we are planning to convene
an internal group of appropriate program officers to share information
regarding support for activities with obvious forensic applications.
Another potential activity that could benefit the forensic sciences
would be the issuance of a Dear Colleague Letter (DCL) that notifies
researchers of the Foundation's interests in supporting activities with
potential applications to the forensic sciences. Such a DCL would draw
the attention of academic and forensic communities to the potential for
utilizing forensic settings as test-beds for asking basic research
questions. Some psychology researchers, for instance, have already
begun to utilize forensic laboratories as settings for asking basic
questions about human cognition and decision making. The DCL could be
designed to encourage collaborative, interdisciplinary teams (to
include basic and applied forensic scientists) to develop scientific
proposals around the relevant questions. Likewise the DCL could
encourage the use of forensic settings for development of new
methodologies and instrumentation.
In keeping with the philosophy of OneNSF, the Foundation could
develop a cross-cutting panel to review a set of proposals that focus
on aspects of forensic science emanating from a number of relevant
basic science directorates. Such an interdisciplinary approach that
brings together basic researchers and practitioners would create new
knowledge, stimulate discovery, and address a range of complex
problems.
NSF continues to develop a multi-year plan of Integrated NSF
Support Promoting Interdisciplinary Research and Education (INSPIRE).
This activity responds to issues raised in a variety of publications
and to perceptions in the research community that NSF does not always
provide good opportunities for comprehensive review and support of
unsolicited interdisciplinary research. The current INSPIRE activity
provides funding for high risk/high reward research that brings
together ideas and approaches that cross intellectually distinct areas
of science. Given the strong potential for coordinating the interests
of basic scientists and the forensic science community through such
opportunities, we have encouraged the development of interdisciplinary
partnerships that address forensic science issues under this umbrella.
Summary
In summary, NSF is committed to supporting the basic sciences that
form the foundation for forensic applications. Many of the projects
funded in recent years will strengthen the forensic sciences both
through support of research with obvious application to forensic
settings and, in the longer-term, through as yet unimagined scientific
and technological developments. In keeping with the NRC's
recommendations, research in the behavioral and social sciences will
also inform the forensic community regarding the impact of cognitive
biases on the evaluation and utilization of forensic information.
The forensic sciences are also strengthened by NSF's support of
many other activities. As a general statement, involvement of students
in supported research will help to ensure a skilled scientific
workforce for the 21st century and provide important training
opportunities that will ultimately improve the practice of forensic
science. The Small Business Innovation Research program will help to
spur economic growth with projects that improve the precision and
operability of instrumentation and processes in forensic laboratories.
Major Research Instrumentation and database development activities will
also assist in building infrastructure for pursuing forensic-related
opportunities. And NSF will continue to provide human and financial
resources in the years ahead to coordinate and collaborate with other
Federal agencies as we work to improve the practice of forensic
science.
Mr. Chairman, this concludes my remarks. Once again, thank you for
the opportunity to appear before you today on this topic. I would be
happy to answer any questions you may have.
The Chairman. And thank you, Dr. Suresh.
Let me just start with questions, and this can be
freewheeling. The Mr. Grisham that I mentioned before was at
our December hearing, but I thought, frankly, the best
testimony came from a former Federal prosecutor named Geoffrey
Mearns. You know him? I just met him that one meeting, and he
is a member of the National Academy of Sciences committee that
reviewed the state of forensic research or science.
At the Department of Justice, Mr. Mearns prosecuted many
high-profile cases, including the Oklahoma City bombing.
As a prosecutor, he said, and I just totally identify with
this--he said he always assumed that evidence used in a
courtroom was based on objective scientific analysis. I mean,
why wouldn't he?
But after studying the issue as part of the National
Academy's review, he told us his faith was shaken. Mr. Mearns
testified that he came to realize that there was not nearly
enough genuine science to validate many forensic science
disciplines.
So, Dr. Lander, you mentioned in your statement sort of
bands of human genome.
Dr. Lander. Yes.
The Chairman. And then you also mentioned hair follicles.
Now, let's go to the hair follicles, since I can relate to that
a little more easily. How can one mess that up scientifically?
Dr. Lander. So look under a microscope at two hairs that
might come from the same person or might come from different
people. What are you going to look at? The color? The width?
The curliness or frizziness?
You can think up a whole bunch of features that might
describe a hair. So which ones do you pay attention to? Which
ones might you ignore as having to do with the conditions under
which the hair was found? Is this really the same color of
brown or not?
Well, you could imagine a forensic examiner with complete
honesty saying, look, in my experience, these things are very,
very similar. That's, indeed, the experiment the FBI ran when
they looked at things and asked, when people said they're
really very similar, in my expert judgment.
You can imagine someone testifying in a courtroom. In my
judgment, these really must have come from the same person. But
then the FBI ran an experiment where they got the DNA off the
bottom of the hair and found one time in eight, even when
concentrating on certain features they thought matched, the
hairs didn't match.
Look, the same is true for fingerprints. Now, it's much
better technology, but you run a fingerprint here, there are
all these wiggly patterns. What wiggles are you matching and
what wiggles are you ignoring? Now, it's usually not on a
perfectly flat surface. It's not like when you get your
fingerprints done by the FBI. It's on some funny surface. So
it's never going to match exactly. How close is close enough?
It's actually the same thing as we found 23 years ago with
DNA. When we said, in that case 23 years ago, Do the bands line
up? Well, they don't line up perfectly. They're always a little
different. How different is too much? How much is the inherent
noise in the technology?
That's what NIST is so good at. NIST looks at the same
things a hundred times, a thousand times and says, what's the
inherent variability? And if it's more than that, we shouldn't
really trust it.
Many things that we think are obvious, actually, well,
they're not so obvious, especially if you're going to apply
them thousands of times, and one percent of the time you're
wrong, that's still an awful lot of wrong identifications.
The Chairman. But, now, a lot of that, and going to Senator
Boozman's appropriate reference to NCIS and CSI, et cetera,
actually, in their defense, in West Virginia, at our two
universities, which do a lot of forensic science, I mean,
students are just pouring into those programs. So, you know,
they get some credit for that, if, in fact, they do, but I'll
give it to them for the moment.
But those aren't done by people. Those are done by computer
software, and all of a sudden, match comes up. So it isn't a
human observation. It's a computer observation. Help me
understand that.
Dr. Lander. Well, if you have your samples analyzed on CSI,
the computer does it and it just says match. In reality, it's
often a human who's doing it, a human who's deciding which
features matter. And even when it's a computer doing it,
there's someone who wrote a computer program to determine
whether things match.
The Chairman. But was it for that particular science
decision?
Dr. Lander. Well, it depends. In the case of DNA, that case
I told you about 23 years ago, someone had a computer program.
They measured the bands by computer. They measured the distance
and said it differed by this much, and the computer determined
was it significant. It was how many standard deviations apart
was it.
The problem wasn't the ruler. The problem was whether that
difference was a significant difference or not. The ruler was
fine. The calculation was fine, but the inference of
significance was what's wrong.
So the best computer in the world wouldn't help. There was
no data underlying it to tell you that two things that were the
same must be this close, and if it's farther apart than this
much, they can't be matching. You need a database underlying it
or when there are many features, you need to know which ones to
choose.
We can fool ourselves by using the computer, by pretending
we're being objective, by choosing some things and then
ignoring what's really the heart of the matter. Do we have data
underlying it to tell us whether this match is really
significant?
The Chairman. Well, but then you have left me with two
problems, and then I'll turn it over to Senator Boozman. You
have an apparent imperfection of human decision-making.
Dr. Lander. You do.
The Chairman. Which could go on eternally.
Dr. Lander. Yes.
The Chairman. And as the day wears on or as his or her
years wear on, it could get worse. And then you have a
computer. And most people--it's sort of like robotics at an
automobile plant. You look at the robotics and you say, they
must be doing it exactly right, because when there's a problem,
at least at Toyota, you know, it'll say, problem, and then a
robot will come in and fix the problem, machine to machine.
And so it just buttresses this belief that science has to
be exact. There's some form of way of getting at science,
forensic science which it has to be exact, but you're giving me
no hope.
Dr. Lander. No, no. Let me restore your hope, Mr. Chairman.
The Chairman. OK.
Dr. Lander. I want to restore your hope. I'm saying just
because it's done by a computer or a machine doesn't make it
right, but that doesn't mean it can't be right. That is what we
have places like NIST for. What you do is you measure something
hundreds of times. You empirically validate how much variation
there is, and then you build that into the computer program.
My objection isn't to the computer program. My objection
isn't to the human. My objection is to either of them
proceeding in the absence of data, measurement data. It may
sound boring to measure things carefully under many different
conditions, but it is the heart of accuracy. DNA works because
it's been done so many times that we know the problems that
arise.
Let me fully restore your faith in the ability to get it
right. You've just got to concentrate in advance on getting it
right. So when we have bite marks and someone testifies that
this bite mark is the same as that bite mark, but there are no
studies that show how much human dentition varies, how well a
mark is transferred from your teeth to a neck, there are no
such studies, you should be worried about that.
But if NIST had decided to go into the bite mark business
and had carefully evaluated that across a couple of hundred
subjects under many different necks that were available for
biting, you might be able to put real legs under it.
You can put legs under almost anything if you go to the
trouble of doing it or at least you will know how accurate it
is. It is all about that collaboration.
The Chairman. Yes, and then the sad thing is what we
started with and that is that people tend to believe that if
it's being introduced as evidence in a court, it's just got to
be true or else it just wouldn't be there.
Dr. Lander. So, therefore, it's our job as scientists, as
lawmakers to make sure that it is true by providing that
scientific foundation for it.
The Chairman. Thank you, sir.
Senator Boozman.
Senator Boozman. Thank you, Mr. Chairman.
And so I agree, you have to have the collaboration to get
that done, and then, you have the good science to back it up.
With just the pure science, how do you get that into the
field? You know, this isn't like pure science in the sense of
creating some sort of scientific breakthrough that you're going
to market and perhaps make many, many dollars out of it.
When we're talking about bite marks and things like that,
the commercial aspect would not be very great. How do we get
that from your laboratory where there's a breakthrough made out
in the field to the small town policeman?
Dr. Suresh. OK. Let me take a stab at that. In fact, I want
to go back to the chairman's question to Dr. Lander. You know,
the level of uncertainty that you have in DNA interpretation is
no different from the level of uncertainty we have in any
scientific experimental work. So let me give you an example
that we all know.
Whenever we develop new materials--for example Alcoa, not
too far from West Virginia, designs a new material and Boeing
puts that into a plane. It's a 20 year process.
So what does Alcoa do? They design a material outside of
Pittsburgh in their research center, and they make the material
in Davenport, Iowa. And they do a lot of testing, and they pull
the material, they twist the material, they bend the material,
they break the material, and they give the material to Boeing.
Boeing doesn't believe anybody else's data because human
lives are involved in flying a plane. They do their own in-
house testing. And in order to make sure that the testing is
reliable, and the interpretation of the testing is reliable,
there are standards, which have come into existence thanks to
the work of NIST.
There is a whole organization called American Society for
Testing and Materials that over the course of many, many
decades has established standards. If you want to pull a piece
of metal, what are the standards by which you do your
experiment? Those standards are established by NIST and various
professional societies. And it's that kind of validation of
scientific data that needs to exist for the interpretation of
DNA. That's what is lacking. That's where the scientific method
comes in.
So, historically, what NSF has done is fund the research at
universities that work with industry and create the basic
scientific data. Agencies like NIST come in and help develop
the standards. These, too, are then adopted by industry and
that becomes the bread and butter of how the industry develops
a new material and puts it into service. I think it's that kind
of a scientific method that needs to be established in
forensics.
So to your question, Mr. Boozman, with respect to how do we
bring it to the attention of people, we can, with these
standards, with these new tools and technologies, we have a
variety of things in place. I can only speak for NSF here.
If there are basic scientific discoveries, we can have
engineering research centers that work with industry. We have
small business innovation research. We have partnerships for
innovation. We have innovation research. These are all programs
that NSF supports.
Those kinds of programs, the SBIR program, which is not
just at NSF, it's in nine Federal agencies, can help take the
basic scientific discoveries and help translate them into
commercial practice for small businesses, entrepreneurs, bring
them in touch with venture capital community.
And the program we launched last year, the NSF Innovation
Corps, is another attempt by NSF to bring that kind of thinking
from basic discoveries to the marketplace to the community.
Senator Boozman. Go ahead.
Dr. Gallagher. I didn't want to take your time, but just
very quickly, you asked sort of two questions. One is how do
you set priorities, and that happens at the junction between
the world that's practicing forensics and the scientific world.
It's really at that interface that those priorities merge, work
and science most contribute.
The other part of your question was interesting, because
there's an impression that putting science in forensics is
tantamount to putting scientists everywhere.
We're not talking about putting Ph.D. research scientists
in every criminal jurisdiction across the United States. What
we're saying is that the tools they use should have a basis in
science. And so, in fact, this can be built into the process
they use, into the laboratory tools they use, the technology
they use, and their methodology.
Senator Boozman. No, and I agree, and guess my concern is
that it's going to take money to get that to small town
Arkansas or wherever, and it's just very difficult right now.
Something we need to do.
The Chairman talked about the human factor that comes in
and there's always a human factor, whether it's the judge or
the jury or whomever.
We can take some of that out by accreditation and
certification which, to me, is very important. Can you all
comment about that where you guys think we need to go in that
regard?
Currently, you've got some crime lab certification, but
talk about the need for perhaps some higher degrees in the
science of forensics.
Dr. Gallagher. So the human factor really has, again, a
couple of different elements. One of them has to do with the
methods that you are using, whether making the measurement,
handling the sample or interpreting the information that comes
from your measurement. Those can be standardized, and those
standards are based on science. That's kind of where the
science gets put in.
And so you develop standard operating procedures for
laboratory personnel. You develop specific specifications for
equipment. You develop standardized analysis tools.
What accreditation and certification do is they're
basically the quality control system that's placed on that
system to give the system the assurance that the laboratory is
following those procedures, that those people have the skills
that they need to have to do what they're being asked to do,
that the technology that they bought is compliant with and
meeting the specifications. And that's a very important part of
the system, because it basically shows that we're following our
own process and that it's about quality control.
Senator Boozman. And the things that are in place now,
perhaps with some strengthening, are they adequate as far as
the accrediting agencies that we have?
Dr. Gallagher. The take-home message I got, certainly, from
the National Academy report, is that it's the disaggregation of
our system that's really interesting. So we have areas where
it's quite strong, where, in fact, the data--for example, DNA
data is not put into the CODIS, the national DNA profile
database, unless those standards are met. So there's what we
call conformity assurance. There's a process in place to make
sure that people follow that and that the data has some
integrity.
In other areas, the systems are weaker or they're not
uniform across the U.S. And so one of the big messages in there
was getting much more systematic about the quality control as
well as the methods.
Senator Boozman. One last thing that's related to that, in
talking to some of the crime lab folks, some of them are
advocating for an independent entity, an office of forensic
science within DOJ to coordinate all of this. And I have been
very impressed with the collaboration. I think it's been good,
and it was really much greater than what I thought.
Do we need a group like that combined with maybe a group of
the guys out in the field some sort of panel there to make sure
that the collaboration continues and somebody's responsible?
Dr. Gallagher. Certainly, in my opinion the answer is yes,
that most of the progress we saw in DNA, for example, came from
the intersection of the world of science with the practicing
world of criminal forensics. You can't really have one without
the other. I mean, having a beautiful scientific basis for
something that can't be deployed and implemented and practiced
or has no meaning in the field is not going to be that useful.
So I don't know what the structures look like, but there is
no question that the ability to convene and have the
appropriate mechanisms where the scientific and the standards
deployment and the practitioning community can talk together,
and the communication needs to work both ways, the science
coming into the process and also the priorities coming out of
the practicing field back into the world of science. So that's
a key part, I think of any solution we're talking about.
Dr. Lander. If I can comment very briefly on that, yes, I
think it's a good idea to have such an office in Justice, but
is it enough? No office of forensic science sitting at DOJ will
be a substitute for the kind of scientific work that has to go
on with regard to standard setting. So an office at the DOJ can
play an incredibly important role in promoting the adoption of
standards throughout the country, identifying needs,
accreditation. Only the DOJ can do that.
But it's tasking it with the wrong mission to ask it to be
the independent scientific body that sets those standards. So
as long as there is strength at both places, and they, as Dr.
Gallagher said, communicate about the needs and then the ways
to address those needs, I think we're in fine shape.
Senator Boozman. Thank you, Mr. Chairman. I apologize to
Senator Udall for running a little bit long.
STATEMENT OF HON. TOM UDALL,
U.S. SENATOR FROM NEW MEXICO
Senator Udall. Well, I thank both of you and thank the
panel. This has been a very good panel. I attended the last
hearing, and I thought it was excellent. And I appreciate,
Chairman Rockefeller, you showing such an interest in this.
I remember last time the National Academy of Sciences
report. I think that came out in 2009, is that correct? And
that report, I think it was called Strengthening Forensic
Science in the United States, A Path Forward, had some very
concerning conclusions in it, state of forensic science, the
lack of scientific foundation behind forensic science
disciplines, lack of standards in laboratory techniques, and it
went on and on and on.
I'm wondering, in the opinion of the panel, since the
report came out, and you would think a report like this would
kind of shake things up a little bit and move the ball down the
road, have there been concrete accomplishments that, in your
mind, Dr. Lander, you've followed this for a long time, 23
years it says in your testimony, Mr. Gallagher, same question.
I don't know, Mr. Suresh, if you feel you want to answer on
that, or is this just a report that sits there and gathers a
lot of dust and nobody listens to it? What's the----
Dr. Lander. Well, let me first say, with regard to the 23
years, I became involved 23 years ago in this case and have
been interested in this since then. It's not a field that I
have worked in continuously there. I have other things I do,
but, as an observer, I'll answer your question.
That 2009 report was deeply disturbing. It really did point
out, as the chairman said, that there are a lot of areas where
our evidentiary foundation is a lot weaker than we thought. I
frankly would have thought it might have provoked more action.
I think, though, it has not sat on the shelf and gathered
dust either. I think we're in the process in these couple of
years of digesting a worrying conclusion and figuring out what
to do. I think we've taken long enough to figure out what to
do. I think it's clear that the what-to-do involves this
collaboration with real responsibility tasked in science
agencies and real responsibility tasked in DOJ, and it's time
to act.
Were this to go on for a lot longer without meaningful
structural response, I would be quite worried, but I think it's
been a good process. I know here in the legislature and in the
executive branch and out in the field, both of scientists and
forensic labs trying to think about how to proceed. One
shouldn't go too quickly to jump to a solution, but one
shouldn't go too slowly either. It is time to act.
Dr. Gallagher. So I think the honest answer to your
question is it probably depends upon how you're looking at it.
The academy report had sort of two elements to it. One was a
structural recommendation to form a new agency, and within that
structure it called for a whole list of specific areas to be
addressed.
And, of course, the problem I think we ran into was that
the structural solution was the only one in the report and it
wasn't one that was viable or deemed viable.
So the question is in addressing a structural solution that
answers the problem without falling to a new agency, that is
probably not something that's done yet. It's been very active,
but it's fair to be impatient.
However, if you look at the 15 or 16 areas underneath, they
have spurned probably the most active interagency process I
have seen in my 18 years of government. In fact, what's
striking is it's much broader than just Federal involvement. We
have representatives from state crime labs and other experts
involved directly in the Federal interagency process, and
they've made a lot of progress in addressing certification
requirements and a whole list of other things, so that once the
structural answer is put on the table, we're ready to roll. And
so it's kind of mixed.
Dr. Suresh. I can point to three or four different
activities that are evolved or continuing to evolve in response
to the NRC report. One is the two workshops that I mentioned,
one on cognitive bias. The other one is on nanotechnology and
forensic science. These workshops were organized and supported
by NSF in response to NRC report. So that's the first one.
The second is, I mentioned in my opening remarks the
memorandum of understanding that's in the works between NSF and
NIJ, and that's something that's a direct outcome of the NRC
report.
The third is the activity that is part of the National
Science and Technology Council Subcommittee on Forensic
Science, and there are several possibilities there. One is to
develop a White Paper that summarizes recommendations to
achieve the goals of the NRC report. The other one would be to
create a prioritized national forensic science research agenda.
A third would be to draft a detailed strategy for developing
interoperability standards. At least a discussion is taking
place through NSTC. So those are four tangible outcomes
following the NRC report.
Senator Udall. Thank you.
And I think, Chairman Rockefeller, your efforts here at the
Committee, I think, have spurred things to move along. And I
think we need to get to the point where we get an
organizational part of this, as you just talked about, that's
really going to come to grips with it and take advantage of all
the energy that's going on out there in this respect. Thank you
very much. Thanks for your attention.
The Chairman. You were a prosecutor.
Senator Udall. I was a prosecutor. That's correct, both at
the, at the Federal level, I was Assistant United States
Attorney and prosecuted criminal cases.
I was thinking the same thing that you said. I always had
the impression when we went into court that the judge was the
arbiter over the science. And you had the sense that, you know,
and the rules all say that, that the judge, he makes sure that
the best scientific information comes in, and whenever it's
fingerprint evidence or whatever.
And you get the sense as a prosecutor, well, that's up
there with the judge, and if he lets it in, then it's all going
to be fine. And, as a prosecutor, you're working with the law
enforcement people and they're doing the same thing they've
done day after day and have been allowed to do.
And so it's kind of a shock when you read the kinds of
things in this report that, whoa, this is very different than
the sense of prosecuting a case. I mean, you really need to
look behind.
And I'm glad we're doing this, because I think it's
important, very important. So thank you for your work on this,
Chairman Rockefeller.
The Chairman. No, but that's interesting that you have that
same----
Senator Udall. Yes.
The Chairman. It hinders us enormously as policymakers.
Senator Udall. Yes.
The Chairman. If you, as a prosecutor, had that feeling
about a judge, and then one could start doing an analysis of
juries. Not allowed to do that because that's called the
American system. They have a right to be wrong, right? You
don't have a right to be wrong.
Senator Udall. Well, and the juries, Chairman Rockefeller,
my understanding, talking with some of my old friends that have
stayed in prosecution, is the juries are watching these crime
shows so much now that the crime shows are impacting what
juries think should be produced by prosecutors in the
courtroom. And if they don't produce all the fancy things that
they see on television, they think there's something wrong with
the case and they think there's reasonable doubt there and they
throw the case out. So, I mean, we have another problem there
when it comes to juries.
The Chairman. Is that why you ran for the Senate?
Senator Udall. I got out of all that business. No, it isn't
exactly why I loved it, but it's a tough business, the
prosecution arena, and trying to get focused on, and the
obligation as a prosecutor, different from a defense attorney,
is to do justice. And so you know in the daily activities that
you carry out that's the ethical obligation on you, and the
idea that the science isn't quite there on some of these
techniques is pretty disturbing.
The Chairman. It is. It is.
Senator Udall. Yes.
The Chairman. But that's valuable stuff.
Senator Udall. Yes.
The Chairman. I'm going to ask sort of a weird question, a
catchall, and I hope it comes out the way I hoped it would, but
it may not.
Dr. Suresh, you have a pretty decent budget, and you spend
approximately $50 million over 3 years on forensics? See, now
that's two-tenths of a percent of your budget. I would appear
to be critical, but I'm sort of setting the scene here, OK?
You've done, in fiscal 2010, 13,000 awards, and I'm trying
to figure out each of the three of you or the two of you, how
do we sort of pull this whole thing together? It means that you
have to have the proper funding, but if you have the proper
funding you've got to use the proper amount of funding of that
proper funding for forensics.
And the EPSCoR program, which I'm thoroughly familiar with,
gets it out into the New Mexicos and the West Virginias and the
Arkansases, and before it all went to Harvard, Yale, Princeton,
Stanford. I remember that fight with Dr. Eric Bloch. It was not
pleasant.
Then NIST, NIST, I look upon as NIST is being kind of the
decider. NIST is right. Others can make mistakes, but NIST
doesn't make mistakes, because you do what Dr. Lander said. You
just keep pounding away at the science until you've got the
genome bands or they're too close or they're not close enough,
but you figure that out and you do that. Am I right?
Dr. Gallagher. That's right. And----
The Chairman. Don't answer the question.
Dr. Gallagher. Yes. That's right.
The Chairman. I'm still formulating. I'm formulating my
question. I'm not sure how it's going to come out.
And then I go to Dr. Lander, and anybody who's a deputy to
your boss has to be a perfect person. I think John Holdren's
one of the great people in government. He doesn't have to do
it, you know. He just does it because he loves it. So you're
kind of pointing to those two and saying, well, let's make this
happen.
And then we run into what I've run into so many times when
I was Governor, the county law enforcement system, the city law
enforcement system, the FBI law enforcement system, the
intelligence community. There are 18 different agencies of the
Federal government that collect intelligence.
And, after 9/11, the first law we passed, to our
everlasting shame, but thank God we did it, was to allow the
CIA to talk to the FBI. They were not allowed to talk up until
that time. And in that story is a lot of the 9/11 Commission
tragedy, because the dots were there and they could have been
connected, but they weren't because they couldn't talk.
So that means that people have to give up stovepipes. They
have to be willing to share. Sharing is not a human
characteristic. In government, it's sort of miserable, and I
think in corporate life it probably is, too. I don't know that.
In families, it's often very hard, you know, to share. In other
words, to give up to get to the desired result.
So I want each of you to figure out a way, tell me how we
get to the point. Because, in the meantime, until all of this
is put in place, until the--I mean, I'm still trying to get
over the software thing, because if you can't trust software,
what you're saying is you can't trust people. If you can trust
people, you can trust software. But then I've got to get
through people and software before I can relax, and, in the
meantime, we're sending people to prison or sending them to the
chair or we're not.
Dr. Lander. We're not catching them.
The Chairman. We're not catching them, right.
So how is this puzzle put together in a way which is
practical? Actually, I don't insist on that, because, I mean,
people are just, in the intelligence community, beginning to
share. They're beginning to share. We're finding that now in
cybersecurity, still. In the Senate, we have committees that
won't share with each other because they're jockeying to hold
on to their jurisdiction. So something gets dropped.
I mean, human behavior is not admirable. So how do we put
this thing together?
Dr. Lander. I'm happy to start here.
The Chairman. OK.
Dr. Lander. Happily, I think this is simpler than many of
the situations you referred to, Mr. Chairman. The ability to
light the fires of excitement in the scientific community is
really pretty great.
The Human Genome Project, a clear agenda was put out by
this Congress in the late 1980s. It said, we need to get the
sequence of the human genome. We could somehow get the sequence
of the human genome. It brought into science a generation of
young people, myself included, who said this was exciting.
People identified that agenda as important.
The single most important thing in marshaling science
behind some public purpose is the clear setting of an agenda.
And that isn't as hard as it seems. If the law enforcement
community and these science agencies came together in an
appropriate structure, some advisory committee, some task
force, some something, which was tasked with identifying where
are the biggest gaps in forensic science. What are we missing
right now to write the software that we need or build the
machines? And you state that clearly to the scientific
community, you unleash the minds of a new generation.
So we have problems with hair. Great. Let's get that out in
front of the scientific community. And what you're going to
find is labs in West Virginia and kids in Berkeley, and, you
know, older scientists in Maine who are going to say, oh, is
that really an important problem? Let's get on the web a
database of 10,000 hairs and let's set an X prize, a challenge,
who can get the best program to identify those hairs.
It'll turn out probably to be cheaper than you imaged
because when you unleash that creativity around each of these
problems, we're going to see software, we're going to see
clever new methods.
Right now, what smart young scientist coming along knows to
think that these are really important problems that our
government cares about?
I think if you got NIST and NSF together with DOJ in an
appropriate structure setting that scientific agenda, you would
see tremendous returns on that investment, because it isn't
about government stovepipes. In the end, it's about unleashing
creative energies. They'll write applications to the NSF
saying, oh, I want to work on this. I'd put their energy there.
Get a clear agenda out there about our greatest needs. And
that's where the DOJ is crucial. The DOJ will know. The
prosecutors will know what are our greatest needs right now.
NIST will be able to say, in order to do that, what science
might we have. NSF will be able to talk to its community and
say, we have funding mechanisms for really meritorious
applications. And you, your committee and the Congress, will be
able to supply an appropriate amount of funding to make sure it
gets done, I hope.
The Chairman. But you would have the final definition.
Dr. Lander. I would have a clear agenda out there. If you
state an agenda of the worthy challenge problems, in this
country, the scientific community rises to meet challenges.
The Chairman. But, then, it puts its results, in whatever
fashion, into the hands of NIST as the dispenser.
Dr. Lander. Indeed. Yes. So then the mechanism to go from
scientific discovery to the setting of standards must pass
through NIST. You don't want the kid in Berkeley or the
scientist here or there turning them into the standards. That's
what we have NIST for.
The Chairman. Right.
Dr. Lander. But you need the science underlying it. You
then need NIST to turn it into standards, and you need the DOJ
to be able to turn that into promoting the adoption across our
justice system.
The Chairman. Dr. Gallagher, what forces the community at
large, God, I'm over time, by a lot. What forces the world at
large, the legal world at large?
Senator Udall. You're the chairman.
Senator Boozman. He's never over time.
The Chairman. You can't leave, Tom. You're a prosecutor.
Senator Udall. I will stay here for 5 minutes.
[Laughter.]
The Chairman. I am undeterred.
How do you get the NIST exactitude standard out to where it
will be understood in Albuquerque and Welch, West Virginia, and
accepted in those places? Because they have their own, you
know, the state has its own labs. Maybe the county has its own
labs. I mean, counties do, don't they? They have coroners and
forensics. They have that stuff, larger ones, at least.
Dr. Gallagher. I have two senses about this problem. One is
that this problem is not as difficult, in my opinion, as some
of the ones you alluded to. In fact, I remember during my
confirmation we were talking about cybersecurity, and I told
you NIST had to work with NSA and DHS, and you said, oh, my
gosh. But, in fact, that's working actually quite well. And I
think, in this case, I have to say, in my opinion, a
multiagency answer is probably better than the single agency
answer that was in the academy.
And the reason for that is if you map it across two
agencies, let me just focus on NIST and the Department of
Justice, the focus of NIST will be on the integrity of the
measurement. And we're not really influenced by the application
of it.
Whereas, the whole role of the Department of Justice is to
apply that measurement to prosecute crime and to promote
justice. Those are complementary roles, and, in fact, if you
combine them into one place, they could actually create
tension, which is part of what we see in the system right now.
So this is a case where NIST being the technical non-
regulatory, the nerds, in support of the people who have to
apply it is actually a good construct. And you asked the right
question, which is, ``OK. NIST does its work and let's say
we've articulated that this appears to be the right basis for
doing this measurement, and here's how we recommend that it's
done. How do you drive it into practice?''
And it's a combination of carrots and sticks. You have to
facilitate the adoption. In other words, you have to put it
into the language, bake it into the technology, make it
consumable by the people who have to do these measurements.
The other thing is that you have to force the adoption, in
some sense. Somebody has to be the adopter-in-chief, and, in
this case, we have this very complicated Federal, state and
local problem.
My view is that the Federal adoption actually is a huge
ingredient here, that if the Department of Justice becomes the
adopter for the Federal law that it's going to have an enormous
impact on the states and local jurisdictions. And we can
facilitate that adoption if we've brought the state and local
participants in from the beginning, so they see their own
involvement in the process as well.
And I know the Department of Justice shares that view with
me. So we bring them along from the beginning and then the
Department of Justice can manage the requirement-setting on the
Federal side, and that will, I think, have a very profound
impact on adoption.
The Chairman. OK. Well, in sheer embarrassment I go to
Senator Boozman, hoping that Senator Udall won't leave. He is.
Senator Boozman. And I'll yield to you if you've got a
question or comment. Go ahead.
Senator Udall. No, no. Well, I just want to thank both of
you. I know you're taking a real interest in this, and this is
such an important issue. I think back to doing prosecutions and
you'd call experts. You know, you'd get an expert. Everybody'd
say, well, you need an expert in whatever it was. Let's just
say hypothetically hair. And an expert would go out, and you'd
get the best one, and then everybody would say, well, whatever
he says, it's going to go into evidence and it'll get into
evidence on your criminal case.
Well, what you all are telling me, and what this report is
opening my eyes on is that it may well be he was a very, very
good expert, but the real question is did he really have the
depth of science to back it up? And that's what I think we're
exploring today, and I think it's very important.
So that's just my final comment. I apologize to Senator
Boozman for stepping out on him, but I really appreciate both
of you----
The Chairman. No, he insisted that you ask a question.
Senator Udall. Yes. Yes.
The Chairman. I was the one who was arguing.
[Laughter.]
Senator Udall. Well, you're both very generous. Thank you.
Senator Boozman. Thank you, Tom.
Secretary Gallagher, if I'm understanding, then we
potentially have an entity that gives guidance and helps
collaborate, the scientific community and the Federal
Government, the standards community, all of that gives guidance
out. And then what you're saying is we don't need to federalize
this, but the reality is that as we come up with good science,
if the Federal Government adopts certain things in their
jurisdiction, then it will follow that that will upgrade the
whole level, and so I think that's a good point.
Unlike watching CSI or similar shows where generally, the
first person there is the coroner, and in probably half of our
states, those people are elected with no training whatsoever
and no accreditation. Where does that fall into all of this?
Dr. Gallagher. So I'm going to give you a short answer and
then promise to follow up, because I'm also not an expert in
the medical-legal area, but that has been actually a key
discussion point within the NSTC process. And there's, in fact,
a specific subgroup that's looking at qualification standards
and practices within that community. And I'd prefer to get back
with you on that one, so I don't attempt to----
[The information requested follows:]
In June 2009, Office of Science and Technology Policy (OSTP)
Director Dr. John Holdren signed the Charter of a new Subcommittee on
Forensic Science, under the National Science and Technology Council
(NSTC) Committee on Science, to address the concerns raised by a
congressionally mandated study by the National Research Council (NRC)
of the National Academies on the status of forensic science in the
United States. The NRC report, entitled ``Strengthening Forensic
Science in the United States--A Path Forward,'' was published by the
National Academies Press on February 17, 2009 (NRC Report). Among a
host of other concerns with forensic science practice, the NRC Report
was critical of the status of the medicolegal death investigation in
the United States:
``What also is needed is an upgrading of systems and
organizational structures, better training, the widespread
adoption of uniform and enforceable best practices, and
mandatory certification and accreditation programs. The
forensic science community and the medical examiner/coroner
system must be upgraded if forensic practitioners are to be
expected to serve the goals of justice.'' [NRC Report, p. 15]
The NSTC Subcommittee on Forensic Science Accreditation and
Certification Interagency Working Group has been tasked to analyze the
issue of medicolegal death investigator certification. We anticipate
its recommendations will be submitted to the NSTC's Committee on
Science in the fall for consideration.
Senator Boozman. No, I appreciate it, and, again, you all
can comment if you like, but it does seem like that's an
integral part of the whole thing, that it's something else. And
I think our crime labs would also like some advice and input in
that regard, because it makes it very difficult, especially in
our smaller communities where many times that entity is looked
to, and, yet, in many of our states there's no training at all.
Well--yes, sir.
Dr. Suresh. I just wanted to add a couple of points to the
comment that the chairman made about NSF's ongoing investments.
The 147 projects or so amounting to about $50 million that I
mentioned were identified by doing a search with the term
forensics. So there is a lot of funding that NSF provides which
feeds into this, but it's not directly aimed at forensic
science.
For example, we fund genetics and genomics research in our
Biological Sciences Directorate. The basic discovery there has
a lot of potential implications for forensic science. So that's
background basic research.
Likewise, in the computer and Information Science and
Engineering Directorate, there is a lot of funding that goes
into data analytics, image processing. Those kinds of things
have huge implications for the development of forensic science
within the NSF context.
So if I were to look at basic science funding with the
implication for forensic science, it's likely to be a lot more
than $50 million. So I just wanted to mention that.
Senator Boozman. Good. Again, thank you all for being here.
Your testimony today has been very, very helpful to me. And we
do appreciate the collaboration that we really are seeing.
Again, I was really very pleasantly surprised in the sense that
I knew that some of that was ongoing, but I think we really
are. However, we all have some frustration that we haven't
perhaps outwardly moved as far forward as we'd like to on the
2009 report.
But I think that it appears that there has been a lot of
work behind the scenes that really is moving in that direction.
And this, as the Chairman has pointed out so many times, is
such a high stakes thing, and this is so important that we get
this right. But we have to get it right so that it'll transfer
again down to the small communities in Arkansas and West
Virginia where they just don't have the resources.
In fact, we could probably have another hearing, Mr.
Chairman, just on the backlog now of the crime labs dealing
with stuff that we all agree that they're dealing with in a
very appropriate manner, but the backlog, in some cases, is
tremendous because of lack of resources.
So thank you, Mr. Chairman.
The Chairman. OK. And thank you.
Can I just ask one more question, and that is I've got to
restore some semblance of confidence in something called
software, and, at this point, it's crash landed in my mind,
because it's written by people.
And, on the one hand, I'm thinking of those, you know, when
you Google typography and then you can make valleys disappear
and turn into mountains or go back 1,000 years and get what it
was like then, and just absolutely amazing things, which take
tremendous disciplined brain power, and yet those are written
by humans, but, on the other hand, those aren't case specific.
They're just general information.
So to make software appropriate to individual cases that
come up and so that somebody doesn't get put away for the wrong
reasons, can software be developed for that?
Dr. Lander. Sure. Software is just rules made faster.
There's nothing a piece of software can do that you couldn't do
yourself with a pencil and paper and enough time. They are just
the embodiment of rules in a machine.
The problem is rarely with the software. It's with the
choice of rules you put in there. Put good rules in there, good
search rules in Google, you get good search results. Lousy
search rules, you don't find what you want on the web. Put good
rules there about those matching DNA bands or about hair,
you'll get good matches and a statement about how good those
matches are. Now, there's a 20 percent chance it's wrong or a 1
percent chance it's wrong or a one-in-a-million chance it's
wrong.
Don't worry about the software. I'm confident that good
software can be written. It's all about the rules. Rules come
from the knowledge. If we get the knowledge right, then NIST
will be able to write standards, which are those rules, and
then the software folks will be able to produce the software
you want. We trace it back there to the real source of the
issue.
The Chairman. My confidence was restored, which probably is
a good point to end the hearing on.
Thank you all very, very much for being patient and very
edifying.
[Whereupon, at 4:02 p.m., the hearing was adjourned.]
A P P E N D I X
Prepared Statement of the National District Attorneys Association
(NDAA)
Chairman Rockefeller, Ranking Member Hutchison, members of the
Committee, thank you for allowing us to submit a ``Statement for the
Record'' for this important hearing on behalf of the National District
Attorneys Association (NDAA), the oldest and largest professional
organization representing over 39,000 district attorneys, state's
attorneys, attorneys general and county and city prosecutors with
responsibility for prosecuting up to 95 percent of all criminal cases
in the United States.
During the hearing a question was posed to the effect of the
leverage which may exist whereby any standards adopted federally could
be ``forced'' upon the states. The choice of words may have been
unfortunate. Nevertheless it makes a point about the nature of the
collaboration that must exist if such an effort is to be successful.
The collaboration must not be solely limited to NIST, the NSF and the
Federal forensic community. As pointed out, there are an estimated 400
forensic laboratories in the United States and approximately 380 of
those laboratories are State and local laboratories while the remaining
labs represent Federal and private forensic laboratories. NIST, the NSF
and our Federal partners in the forensic community all have a role to
play in this effort, but none greater than that of state and local
laboratories that are charged with analyzing evidence in what accounts
for over 95 percent of the crimes committed in this country.
This collaboration can be performed successfully, as evidenced by
the DNA Advisory Board within the Department of Justice which had
participants from NIST and other Federal, state and local partners.
That effort and the lessons we learned can be duplicated here as well.
During the hearing, Senator Udall posed the question whether since
the release of the 2009 National Academy of Sciences report there had
been any concrete accomplishments toward improving the state of
forensic science. It should not be overlooked that the legislation
requesting such a study was sought and supported by the forensic
science community. We can be proud to report that their have been
significant accomplishments and those efforts actually predate the
report itself. The National Academy report singled out for particular
criticism three forms of analysis--serology, bite mark and microscopic
hair analysis. Most of the exoneration cases which identify forensic
science as a contributing factor involved those forms of analysis. Most
of those cases occurred prior to the existence of forensic DNA typing
in 1985 or its ready availability in this country around the mid 1990s.
However, their use by the forensic science community has been extremely
limited for a number of years.
Consider, for example, microscopic hair comparison that Dr. Lander
testified to, where certain physical characteristics were compared. He
testified that it was subsequently determined in a study that in
approximately 1 out of 8 comparisons examiners would reach a conclusion
that there was a ``match'' between known and questioned hairs. Using
mitochondrial DNA testing it was determined that in170 hair
examinations, 1 in 8 hairs believed to match did not come from the same
source. That study was conducted by scientists from the FBI and the
Forensic Science Initiative of West Virginia University. That study was
published in 2002. Thereafter, microscopic hair comparison has been
limited to serving as a ``screening'' test for purposes of identifying
cases in which mitochondrial testing of evidentiary hairs would be
appropriate.
Forensic odontology has long been utilized within the forensic
community, most notably as a method of identifying human remains. Bite
mark evidence is another aspect of the work of a forensic odontologist.
The most noteworthy case in which bite mark evidence was used was in
the prosecution of Ted Bundy in the state of Florida. However, with the
advent of DNA profiling, bite mark evidence has been relegated largely
to those cases in which a swab for saliva in the area of the bite mark
has not yielded DNA sufficient for testing. That has been considered
best practice since at least 1997. As reference, please see Manual of
Forensic Odontology, 3d. Ed. American Society of Forensic Odontology
(revised 1997).
Serology has likewise been relegated to use as a screening tool
within the forensic laboratory almost from the day forensic DNA
profiling became readily available within this country. The science of
serology is interesting however because it is well researched and its
limitations are well known within the scientific community. It
continues to be used today in hospitals across the country. It is well
validated.\1\ The reason that serology has been replaced for
evidentiary purposes is that it lacks the powers of discrimination
between individuals that DNA testing provides. A simple blood type, for
instance, may only narrow the range of potential suspects to
approximately 40 percent of the population. The problem with the
serology cases was not the science so much as the competence or
integrity of the scientist, the prosecutor or the defense counsel.
---------------------------------------------------------------------------
\1\ Some confusion exists within the general public with respect to
the terms ``validated,'' ``invalidated'' and ``unvalidated.''
Validated, as Dr. Lander described it, is that process of testing and
retesting in order to identify the limitations of the reliability of a
technique or measuring system. Invalidated means that something has
been tested and the results of testing show the conclusions to be
unreliable. Unvalidated means that there is a lack of sufficient
testing necessary to render a conclusion as to the reliability of a
measurement, test or conclusion. Unvalidated and invalidated are not
synonymous. Testing and being found to be true or false is one thing,
never being adequately tested is something completely different.
---------------------------------------------------------------------------
The commitment of the forensic science community for reliable
science is evidenced by its investment in the accreditation process.
Three organizations currently accredit forensic laboratories within the
U.S. \2\ Virtually all public laboratories are accredited today. ASCLD-
LAB has accredited an estimated 380 such laboratories, Federal, state,
local, private and international. Most of those laboratories were
accredited before 2009. The National Academy report recommended
accreditation to a recognized international standard for accreditation
(ISO 17025). Accrediting bodies in this country were in the process of
accrediting laboratories to that standard before the Academy report was
published. An estimated 180 labs are already accredited to that
standard by ASCLD-LAB with the remainder in the process of becoming so
accredited.
---------------------------------------------------------------------------
\2\ Forensic Quality Services, American Society of Crime Laboratory
Directors-Laboratory Accreditation Board and A2LA.
---------------------------------------------------------------------------
It is NDAA's belief that non-DNA forensic science disciplines have
been demonized in recent years because their reliability is not up to
the ``DNA Standard'' seen on popular television shows like CSI.
Unfortunately, real world examples of cases tried on television are few
and far between. Some cases have DNA, but most cases do not. As stated
on the Innocence Project's website (www.innocenceproject.org), since
1989 there have been 289 post-conviction DNA exonerations in the United
States. While NDAA agrees that even one wrongful conviction of an
innocent person is too many, this number needs to be taken into proper
context to gain an accurate portrayal of the state of forensic science
in America's criminal justice system.
In the United States there are, at minimum, 10 million cases per
year (not including traffic offenses) where serious crimes have been
committed. This means since 1989 there have been at least 220 million
cases in America involving serious crimes: while 289 post-conviction
exonerations are of real concern to NDAA, in reality these wrongful
convictions occurs less than one-hundredth of 1 percent of the time in
America's courtrooms.
Many stakeholder groups point to these 289 post-conviction
exonerations and reactively conclude that America's use of forensic
sciences in the courtroom is suspect and the system is irreparably
broken. NDAA could not disagree more with this notion; it is important
for us to remember that the vast majority of the time during criminal
cases--more than 99.99 percent of the time--the prosecutor properly
serves justice and gets the case right. That said, NDAA fully supports
improvements to forensic science and agrees that Federal resources be
used to improve the quality and reliability across all forensic science
disciplines.
______
Response to Written Questions Submitted by Hon. John D. Rockefeller IV
to Eric S. Lander, Ph.D.
Good Science Leads to Good Law Enforcement
Question 1. The crime writer John Grisham was the most famous
witness at our December hearing. But I thought some of the best
testimony came from a former Federal prosecutor named Geoffrey Mearns,
who was a member of the National Academy of Sciences' committee that
reviewed the state of forensic science. At the Department of Justice,
Mr. Mearns prosecuted many high-profile cases, including the Oklahoma
City bombing.
As a prosecutor, he said he always assumed that evidence used in
the courtroom was based on objective scientific analysis. But after
studying the issue as part of the National Academy's review, he told us
his faith was shaken. Mr. Mearns testified that he ``came to realize
that there was not nearly enough genuine science to validate many
forensic science disciplines.''
In your testimony, you talk about how law enforcement officials
were at first reluctant to work with you on developing standards for
DNA testing. Can you explain how you convinced them that scientific
standards were good for our criminal justice system?
Answer. The change was driven by necessity. Law enforcement
officials had originally rejected as unnecessary a proposed study of
DNA forensics by the National Academy of Sciences. Then a high-profile
case revealed serious flaws in the practice of DNA fingerprinting, with
both prosecution and defense witnesses ultimately agreeing on the
problems. Law enforcement officials then became concerned that these
findings might jeopardize the use of DNA fingerprinting, and then
agreed that setting higher standards was desirable.
The recent NAS report on problems with forensic science should have
been a similar wake-up call. But there has been continuing resistance.
Question 2. Can we have a fair justice system while we are waiting
for the science of forensics to catch up?
Answer. Yes, but . . .
The justice system can be fair provided that the reliability of
forensic testimony is accurately described. For some forensic
disciplines (such as bite marks), this would involves telling juries
that the evidence is scientifically unsupported and thus unreliable.
For others (such as hair analysis), it would require telling juries
that errors occur at an appreciable frequency (10 percent in an FBI
study) and thus the evidence is not definitive.
For too many forensic technologies, we don't know how to evaluate
the evidence: we don't know the rate of false positives and false
negatives.
The integrity and fairness of our justice system is threatened when
we lack standards, but we allow witnesses to tell juries otherwise.
It may be best to exclude certain classes of evidence until the
science and standards catch up.
Scientific Analysis Reduces the Chance of Bias in the Criminal Justice
System
Question 3. The National Academy of Sciences report makes a very
interesting point about how human error can creep into forensic
science. It discusses a concept called ``contextual bias,'' which means
that a forensic analyst's conclusions can be influenced by what he or
she knows about the suspect or the facts of the case. The report isn't
suggesting that anybody is acting with intentional bias, but it does
suggest that a more independent and rigorous scientific approach could
minimize this problem. What steps do scientists take to limit the
possibility of ``contextual bias'' in your work? Can you also explain
why this is an important issue in the field of forensic science?
Answer. In most scientific situations (not involving forensic
testimony), scientists take a variety of precautions to guard against
contextual bias.
These may include ``blinding'' themselves to the identity/
characteristics and to the changes or treatments given to research
subjects. For example, in a clinical trial to evaluate a drug, a
scientist will not know until after concluding an experiment which
group of patients is the control group that did not get the drug and
which group of patients is the treatment group that took the drug. This
will prevent the scientist from being biased toward observing the
responses in patients that the scientist hopes to see. This is so
critical that unblinded clinical trials are considered suspect.
Scientists also follow objective, verifiable methods to test their
hypotheses against available data. They try only to draw conclusions
that are supported by the data, and to be cautious not to overstate
their results.
Their studies, before publication in journals, typically undergo a
peer-review process.
And, it should be possible for other scientists to replicate any
experiment, and they should draw the same conclusion. Only when science
has been independently replicated can it be truly considered sound.
Forensic evidence presented as scientific findings in criminal
trials often differs in important ways from the process that scientists
use in conducting studies and publishing their results in peer-reviewed
journals. For example, forensic experts do not necessarily ``blind''
themselves to the identity of the accused or the facts of the case, and
are not required to demonstrate the accuracy of methods on a larger
sample. In criminal cases, scientists present evidence that is not
typically peer-reviewed.
Prizes and Challenges in Forensic Science
Question 4. The America COMPETES Reauthorization (P.L. 111-358)
gave broad authority to Federal agencies to use prizes and challenges
as drivers for stimulating private industry and individuals to solve
problems of national importance; certainly the forensic science
community is facing concerns of national importance. What specific
problems in forensic science do you think are best suited for a prize
or challenge?
Answer. Prize competitions involving the broader public could be an
excellent method to address the challenge of (i) determining the best
analytical methods to evaluate specific kinds of forensic data and (ii)
determining the accuracy of those methods (which is critical for
evaluating their use in courts).
It might be ideal to focus on the 5 most important forensic
technologies, as determined by DOC and DOJ and reflected in the NAS
report. (Examples could include hair, bite marks, bullet marks, and
fingerprints.)
Question 5. How would you recommend structuring such a challenge so
that we achieve the best possible results in the most timely manner?
Answer. For each type of forensic evidence selected, DOJ/FBI could
contract with an external party to prepare a very large set of samples
(many thousands) including associated digital data that could be shared
without limitation and physical specimens that could be shared with
qualified parties.
The samples should reflect the wide range of conditions under which
evidence is encountered. (For example, for a fingerprinting sample set
and dataset, the FBI should collect and offer fingerprints that include
partial prints from many kinds of surfaces, and the full range of type
of fingerprints collected in criminal cases and beyond.)
Information about which samples actually match would be known but
withheld.
Competitors would be challenged to develop methodologies that would
be tested relative to known ``right answers''. (Crime labs could
participate in the competition, as well as scientists and
technologists.)
A contest advisory board would be selected by the NAS, based on
input from FBI and NIST, to oversee the competitions.
______
Response to Written Questions Submitted by Hon. Amy Klobuchar to
Eric S. Lander, Ph.D.
Question 1. I am concerned about the effect of delays in DNA and
other forensic analysis in criminal cases--delays in forensic analysis
can prevent law enforcement from apprehending criminals or delay
exoneration of innocent persons. Can you comment more on this?
Question 2. From your perspective in the scientific community, what
factors contribute to delays in analyzing forensic evidence?
Question 3. Do you believe standards are a way to reduce the delays
forensic analysis?
Answer. [Dr. Lander believes he has insufficient knowledge to
answer these questions usefully.]
______
Response to Written Questions Submitted by Hon. John Boozman to
Eric S. Lander, Ph.D.
Question 1. Your testimony indicates that DNA forensics was not
widely accepted by the practitioners of forensic science? Today it is
widely accepted in the law enforcement and forensic community. What
specifically changed, in the community of practitioners, to make this
happen?
Answer. The change was driven by necessity. Law enforcement
officials had originally rejected as unnecessary a proposed study of
DNA forensics by the National Academy of Sciences. Then a high-profile
case revealed serious flaws in the practice of DNA fingerprinting, with
both prosecution and defense witnesses ultimately agreeing on the
problems. Law enforcement officials then became concerned that these
findings might jeopardize the use of DNA fingerprinting, and then
agreed that setting higher standards was desirable.
The recent NAS report on problems with forensic science should have
been a similar wake-up call. But there has been continuing resistance.
Question 2. The scientific working groups (SWGs) are currently
active within the DOJ. I understand some SWGs are more active than
others. How do you feel about SWGs with NIST oversight, with the both
practitioners and scientists in this group? How do you think the
individuals should be chosen?
Answer. [Dr. Lander believes he has insufficient knowledge to
answer this question usefully.]
Question 3. Do you feel that current ISO standards for forensics
are inadequate? Please give specific examples.
Answer. [Dr. Lander believes he has insufficient knowledge to
answer this question usefully.]
Question 4. Do you think that we need advanced level degree
programs in the science of forensics? What are your thoughts about the
current education level and accreditation system for practitioners? In
your opinion, what are the current needs in this area?
Answer. [Dr. Lander believes he has insufficient knowledge to
answer this question usefully.]
______
Response to Written Questions Submitted by Hon. John D. Rockefeller IV
to Dr. Subra Suresh
Forensic Science Pipeline
Question 1. A healthy forensic science pipeline supports basic
research, development of practical applications, as well as training of
students and practitioners. I am particularly interested in education
of students because the Bureau of Labor Statistics projects a 20
percent increase in jobs for forensic science technicians alone by
2018.
Dr. Suresh, your written testimony describes activities NSF is
supporting to expose students to forensic science research. However, as
I understand it, there are no Ph.D. programs in forensic science. How
does this contribute to the fundamental problems in forensic science
research?
Answer. The National Science Foundation supports basic research and
education in all fields of fundamental science and engineering. Many of
the research projects that are supported contribute to scientific
advances that ultimately contribute to the scientific basis for
forensic analysis. While many awards are made to senior research
personnel, a significant number of awards are made each year directly
in support of doctoral research by students with forensic science
interests that will advance the application of science in forensic
settings. For instance, an award was made to Louisiana State University
in support of doctoral student's research titled ``Using GPS to Track
Vultures in Texas.'' As the student explains in the application for
funding, understanding the behavior of vultures over time and space and
its relation to decomposition of remains, will improve the efficiency
and accuracy of forensic estimation of time-since-death. Other doctoral
research is less obviously tied to forensics, but nevertheless advances
those sciences. An award to an anthropology doctoral student at Johns
Hopkins University titled ``Environmental Effects on Human Cranial and
Postcranial Sexual Dimorphism'' will contribute valuable data that can
assist forensic scientists in determining the sex of skeletal remains.
Often science and engineering doctoral students conduct research
while supported by awards made in support of their mentors, and many
research projects supported across the foundation are budgeted to
assist in supporting graduate students. Thus, a collaborative research
project on nuclear data measurements and radiation detector
development, funded at Duke and North Carolina A&T State Universities,
contains funds for support of undergraduate, graduate and post-doctoral
students. This research may advance nuclear forensics, an increasingly
important area in countering terrorist threats, and is but one example
of basic research that holds promise for forensic application and
contributes to workforce development. Examination of high impact
publications in the forensic sciences shows that authors of most
frequently cited articles are based in basic science research programs.
Clearly, students and senior researchers in these basic science
settings provide results and methodologies that inform the forensic
sciences.
Question 2. Your written testimony points out that the Foundation's
forensic science awards are spread out over all seven directorates. Is
this a good thing? Would a specific forensic science program at NSF
help to consolidate the work and attract additional worthy research
proposals?
Answer. A crucial strength of the National Science Foundation is
that it supports basic research in all fields of science and
engineering. Scientists submit research proposals to standing programs,
or increasingly to interdisciplinary programs, and the proposals are
reviewed via high quality merit review. Through its gold standard merit
review process, NSF brings research to bear on a number of timely
national problems.
The forensic sciences constitute an exceedingly diverse set of
investigatory areas ranging from anthropology to zoology. While there
certainly are some challenges within forensics that could, or already
do, benefit from collaboration across the traditional academic
stovepipes, many of the forensic sciences are quite distinct in their
theoretical bases and methodological approaches. Creation of a forensic
science program which received competing proposals in, for instance,
analytical chemistry and forensic anthropology would bring very
different research projects into competition with each other.
NSF does utilize a wide array of mechanisms in support of cutting
edge, transformative research and several of these have already been
applied to proposals with forensic science significance. Co-review of
proposals that intersect two or more standing programs is a long-
standing practice within the foundation and awards which contribute to
the forensic sciences have been made with the support of several
programs. Programs in three divisions across two directorates supported
a project (``Cyber-Enabled Chemical Imaging: From Terascale Data to
Chemical Imaging'') that will assist in providing detailed information
on the chemical composition of substances found on surfaces--a
capability of significant utility in forensic analysis. Additionally,
many programs are by their very nature interdisciplinary in character.
The Law and Social Science, a program with roots in a number of social
and behavioral sciences including sociology and psychology, is
supporting research on overcoming obstacles to the successful use of
forensic age progression. This research utilizes information and
theories from anthropology, computer science, developmental sciences
and psychology with an ultimate aim of improving the utility of this
forensic technique.
However, we are aware that more can always be done. Subsequent to
publication of the National Academy's report Strengthening Forensic
Science in the United States: A Path Forward'' (http://www.nap.edu/
catalog.php?record_id=12589), the Social, Behavioral and Economic
Sciences Directorate (SBE) supported a workshop on cognitive biases in
forensic examination. SBE, possibly in conjunction with the Computer
and information Sciences (CISE) directorate, is considering another
workshop in the area of human and computer recognition of patterns in
forensic settings.
As stated in Dr. Suresh's testimony of March 28, 2012 NSF might
also issue a:
. . . Dear Colleague Letter (DCL) that notifies researchers of
the Foundation's interests in supporting activities with
potential applications to the forensic sciences. Such a DCL
would draw the attention of academic and forensic communities
to the potential for utilizing forensic settings as test-beds
for asking basic research questions. Some psychology
researchers, for instance, have already begun to utilize
forensic laboratories as settings for asking basic questions
about human cognition and decisionmaking. The DCL could be
designed to encourage collaborative, interdisciplinary teams
(to include basic and applied forensic scientists) to develop
scientific proposals around the relevant questions. Likewise
the DCL could encourage the use of forensic settings for
development of new methodologies and instrumentation.
We are also considering the establishment of internal communication
mechanisms, such as a Sharepoint site. This would allow staff to share
information on awards, workshops, publications etc. about research with
forensic science significance supported across the foundation. Such
exchanges of information would stimulate coordination and
collaboration. Coordination and collaboration is not confined to
internal activities. Several program staff have served on the NSTC
Subcommittee on Forensic Science. A number have already discussed
opportunities for joint support of research with the National Institute
of Justice, the FBI and Department of Defense.
Thus, there are many current activities, and others which are under
consideration, that are investments in the forensic sciences and which
serve to attract high quality research projects.
Question 3. What are your thoughts about how applying the
scientific method can reduce the possibility of contextual bias in the
criminal justice system?
Answer. The National Science Foundation has begun to address the
issue of contextual bias and supports basic research in many relevant
areas, including human cognition and decisionmaking as it may relate to
the forensic sciences. It is well known within the psychological
research literature that individuals are susceptible to certain biases
and preconceptions in the perception and interpretation of stimuli, and
in the decision processes that lead to such interpretations. The NSF's
Directorate of Social, Behavioral, and Economic Sciences (SBE) has
supported research in this area through its programs in Perception,
Action, and Cognition; Social Psychology; Economics; Decision, Risk,
and Management Sciences; and Law and Social Sciences.
In response to the National Academy's report on forensic science,
the Division of Behavioral and Cognitive Sciences (SBE Directorate at
NSF) sponsored a workshop to explore the role of cognitive bias in
forensic examiner decisionmaking. The goals of this workshop included
stimulating basic research projects among researchers who may not have
considered the interplay between behavioral science and forensic
science, as well as identifying applied research projects that might
improve communication and decisionmaking by forensic examiners.
Participants offered important recommendations for translating basic
research on countering cognitive biases including (i) assessment
procedures that shield investigators from contextual information that
may bias their decisions and (ii) the use of evidence ``lineups'' that
require discrimination of a source from that of distractor samples, or
tasks that require examiners to consider factors that might lead to
decisions supporting both a ``match'' and a ``non-match'' of a sample.
Participants noted (i) that additional research would be required to
successfully translate and refine procedures that might prove most
effective in the forensic science context, and (ii) that collaboration
between behavioral scientists and forensic practitioners would be an
important determinant of success in this context.
______
Response to Written Questions Submitted by Hon. John Boozman to
Dr. Subra Suresh
Question 1. Has the current administration formulated a consistent
policy position regarding the science and standards of forensics? If
yes, please state; if no, when will this policy be announced?
Answer. The National Science Foundation supports basic research in
all areas of science and engineering except for the medical sciences.
Through its gold-standard merit review process, the Foundation
identifies the highest quality research for support, both in
traditional academic areas and novel areas which are `high-risk, high-
reward. Establishing administration priorities in science lies with the
Executive Office of the President's Office of Science and Technology
Policy. Therefore, OSTP would be the best source of information on the
administration's policy in forensic science and standards.
Question 2. It is my understanding that the National Science
Foundation is currently funding research in forensic science, although
spread out in several directorates and not underneath one single
category of `forensic science.' Could you please give me a scope of the
projects that are currently funded? One of the recommendations of the
National Academies Report is more `coordination' of the underlying
science in this area; what practical benefits, from the NSF standpoint,
do you expect to see from this coordination?
Answer. A crucial strength of the National Science Foundation is
that it supports basic research in all fields of science and
engineering except medical science. As stated in Dr. Suresh's testimony
of March 28, 2012, all Directorates have funded basic research with
implications for forensic science. While a description of each award is
beyond the scope of this response, these awards encompass the variety
of mechanisms offered by the Foundation, including standard awards to
scientists at academic institutions, doctoral dissertation research
grants, research experiences for undergraduates, and major research
infrastructure awards. The awards also cover a variety of forensic
sciences, including forensic anthropology, DNA and trace analysis,
digital forensics, and the various comparative forensic sciences such
as fingerprinting and forensic authorship identification.
Within each Directorate, scientists submit research proposals to
standing programs, or increasingly interdisciplinary programs, which
are reviewed via high quality merit review. The forensic sciences
constitute an exceedingly diverse set of investigatory areas ranging
from anthropology to zoology. Some topics could, or already do, benefit
from collaboration across the traditional academic stovepipes. However,
many of the forensic sciences are quite distinct in their theoretical
bases and methodological approaches. Creation of a forensic science
program that receives competing proposals in, for instance, analytical
chemistry and forensic anthropology would bring very different research
projects into competition with each other. This might well result in a
reduction in support of the best ideas and projects across the various
fields that forensic scientists draw upon and inhibit the connection
between the forensic sciences and the underlying basic research in, for
example, chemistry.
Nevertheless, increased coordination of NSF's contribution to
forensic science research is possible. For example, proposals that
intersect two or more standing programs can undergo ``co-review'' by
these programs--a mechanism that has been used to support awards that
contribute to the forensic sciences at NSF. However, we are aware that
more can always be done. Subsequent to the National Academy's report on
forensic science, the Behavioral and Cognitive Science Division of the
Social, Behavioral and Economic Sciences Directorate (SBE) supported a
workshop on cognitive biases in forensic examination. SBE, possibly in
conjunction with the Computer and information Sciences Directorate and
the United Kingdom's Home Office, is considering another workshop in
the area of human and computer recognition of patterns in forensic
settings. As stated in Dr. Suresh's testimony of March 28, 2012, NSF is
also considering the creation of a Dear Colleague Letter (DCL) ``that
notifies researchers of the Foundation's interests in supporting
activities with potential applications to the forensic sciences. Such a
DCL would draw the attention of academic and forensic communities to
the potential for utilizing forensic settings as test-beds for asking
basic research questions.'' NSF is also considering the establishment
of internal mechanisms, such as a Sharepoint site, to inform program
officers of research with forensic science significance supported
across the foundation and thereby stimulate coordination and
collaboration.
Coordination and collaboration is not confined to internal
activities. Several program staff have served on the NSTC Subcommittee
on Forensic Science. Additionally, a number of program officers have
discussed opportunities for joint support of research with the National
Institute of Justice, the FBI, and Department of Defense.
We believe in the importance of NSF's continued contributions to
basic science that informs development of the forensic sciences. The
mechanisms described above could promote further coordination in this
area and provide critical knowledge that could improve the efficacy of
forensic practice in the years ahead.
Question 3. As you know, the fiscal situation in this country is
very tight. The NSF has many scientific priorities, but the demand for
basic forensic science research will increase. With limited resources,
how would you prioritize which areas of forensics will get priority in
terms of basic science research?
Answer. The National Science Foundation recognizes that we face
difficult choices in the face of increased demand for research support
and constrained resources. In making these choices, we will rely on
advice and guidance from our Advisory Committees and community-based
decisionmaking through the merit review process and other mechanisms.
NSF supports basic research in all fields of fundamental science
and engineering. Many of the research projects that are supported
contribute to scientific advances that ultimately contribute to the
scientific basis for forensic analysis. Advances in virtually all areas
of science, ranging from anthropology to zoology, may immediately, or
downstream, lead to important applications in the forensic setting. As
noted in the response to the previous question, the National Science
Foundation identifies the most promising projects through the use of a
merit review system that is held high esteem around the world. Indeed,
last month NSF hosted representatives from more than 50 countries for
the first Global Summit on Merit Review.
NSF review processes provide for submission of research proposals
by individuals or teams of scientists. The proposed activities are then
reviewed and assessed against the criteria of Intellectual Merit and
Broader Impacts, as well as additional, applicable criteria such as the
integration of research and education. The most compelling projects may
then be recommended for funding.
The merit review system ensures that the highest quality projects
are funded. Any specific proposal may have immediate or downstream
applications in one or more areas of forensic science. NSF has
traditionally listened closely to its research communities and
frequently learns through proposals submitted by investigators about
their cutting-edge interests. Projects are supported that best
demonstrate their relevance based on two agency-wide criteria:
intellectual merit and broader impacts.
All NSF scientific staff members are continually involved in
activities such as outreach to scientific communities, attendance at
scientific meetings and conferences, sponsorship of workshops, and
interactions with colleagues in agencies with legal and forensic
missions. Such efforts allow NSF staff to monitor trends, identify
opportunities and set priorities in concert with the many basic
research communities that ultimately inform forensic practitioners. In
addition, workshops may be funded with the purpose of examining trends
in research, identifying gaps and needs, and recommending priorities.
______
Response to Written Questions Submitted by Hon. John D. Rockefeller IV
to Patrick D. Gallagher, Ph.D.
NIST and Forensic Science
Question 1. NIST has an outstanding reputation for doing cutting
edge work in technology, measurement science, and standards, and your
written testimony nicely outlines NIST's long and rich history in
forensic science. My question to you is simple: with NIST's credentials
in forensic science going back almost 100 years, why do we still have a
problem today?
Answer. As outlined in the NAS report, the issues that plague the
forensic science community are complex. The NAS report describes the
practice of forensic science in the U.S. as highly fragmented, both
across jurisdictions and across disciplines. Other contributing factors
include the lack of uniform oversight and the limited funding available
for basic research and standards development.
There are approximately 400 Federal, state and local government
crime laboratories and more than 90 percent of them fall under state
and local jurisdictions. There are great disparities among them with
respect to funding, oversight, personnel certification, and laboratory
accreditation. For example, the accreditation of crime laboratories in
the U.S. is still voluntary for Federal laboratories and for crime
laboratories in 46 of the 50 states. The main barrier to uniformed
practices of crime laboratory practices across the U.S. is state
sovereignty, which limits the role that the Federal Government can play
in the development and enforcement of forensic science regulations.
Each forensic science discipline applies a unique subset of
principles from traditional sciences to draw conclusions about evidence
collected in a case. The quality as well as the degree of foundational
scientific rigor underlying the practice within each discipline has
correctly been described as uneven at best.
Although NIST has a rich history in forensic science, its
contributions have only sought to address the specific needs of other
Federal agencies and industry with which it has partnered to date.
However, as outlined in my testimony, our work has a broad impact on
the forensic science disciplines with which we have worked. As a leader
in measurement science, standards, and technology, NIST is a logical
choice to partner with other agencies and professional organizations on
a national initiative to lead the U.S. toward improving the practice of
forensic science in the United States. We look forward to partnering
with other Federal agencies with complementary core strengths to
collectively contribute to the future of forensic science.
Scientific Analysis Reduces the Chance of Bias in the Criminal Justice
System
Question 2. The National Academy of Sciences report makes a very
interesting point about how human error can creep into forensic
science. It discusses a concept called ``contextual bias,'' which means
that a forensic analyst's conclusions can be influenced by what he or
she knows about the suspect or the facts of the case. The report isn't
suggesting that anybody is acting with intentional bias, but it does
suggest that a more independent and rigorous scientific approach could
minimize this problem. What are your thoughts about how applying the
scientific method can reduce the possibility of contextual bias in the
criminal justice system?
Answer. Applying the scientific method is just the beginning of
resolving the issue of contextual bias in the practice of forensic
science. As we have seen in a number of recent studies, sources of
contextual bias are numerous and the solution is multi-layered. In
addition to the appropriate triage of facts and information selectively
provided to the forensic science practitioner, there are also processes
and procedures for quality assurance that must be built into the
technical review of the first scientist's analytical findings. For
example, the practice of requiring a technical reviewer to reexamine
only those cases where the original examiner has made a ``match''
between questioned evidence and a suspect in a fingerprint case must be
prohibited if we wish to achieve true objectivity. The reviewer who has
the expectation that only cases with ``matches'' will reach her desk
for review inevitably has contextual bias built into the process. One
solution is to forward a variety of outcomes to the reviewer including
cases with exclusions and inconclusive outcomes in addition to cases
with ``matching'' outcomes.
NIST has been actively studying the issue of contextual bias and
other human factors issues. In collaboration with the NIJ, NIST
convened an expert working group to do a scientific assessment of the
effects of human factors on forensic latent print analysis and to
develop recommendations to reduce the risk of error. The recently
published report, Fingerprint Analysis: Improving the Practice through
a Systems Approach, (http://www.nist.gov/manuscript-publication-
search.cfm?pub_id=
910745) is the result of a 2\1/2\ year study. It was composed of
forensic science practitioners, psychologists specializing in
contextual bias, statisticians, prosecutors, defense attorneys,
academicians and accident prevention specialists. The working group
also created a process map (http://nij.gov/nij/topics/forensics/
evidence/impression/latent-print-flowchart.htm) that illustrates the
latent print examination process, and the report details steps in that
process where human error risks could be minimized.
The Expert Working Group on Human Factors in Latent Print Analysis
and its recently published report serve as an excellent model for the
other forensic science disciplines. NIST will begin a new panel
evaluating the human factors issues in questioned documents analysis in
the coming months. Many forensic science disciplines would benefit from
implementing this model to identify and help limit the potential for
contextual bias.
Response to Written Questions Submitted by Hon. John Boozman to
Patrick D. Gallagher, Ph.D.
Question 1. What working relationship between exists between NIST
and DOJ? Do both agencies agree on the clear division of appropriate
responsibilities regarding this problem? Could you outline, to date,
some of these activities and briefly tell us what progress this synergy
has produced? What problems do you anticipate with this type of inter-
agency, ``inter-cultural'' collaboration?
Answer. NIST and DOJ enjoy a decades old history predating World
War II of successful collaborations in criminal investigations and
supporting the development of the original FBI Laboratory. During the
intervening years, DOJ and NIST have partnered on technology and
standards development in many areas of public safety, including
emergency response involving law enforcement agencies, fire
departments, emergency medical teams, corrections and forensic science
communities worldwide. DOJ and NIST have numerous formal agreements in
place articulating the respective roles and responsibilities of the two
agencies. The synergistic relationship between our two agencies is an
excellent example of leveraging the core strengths of each agency to
produce deliverables with benefits to public safety communities that
exceed what either agency could generate on its own.
One example includes the response by the U.S. Attorney General to
the 2003 premature field failure of recently issued body armor that was
penetrated by a standard round from a handgun that the body armor had
been certified to stop. In response, the AG created the Body Armor
Safety Initiative. Under a subsequent agreement, NIST undertook
research to measure and characterize the root causes of these failures
and developed testing methodologies to insure that such failures did
not occur again. The research and collaboration between NIST and DOJ,
with practitioners and relevant stakeholders, led to the revised
standard for body armor (NIJ Standard 0101.06), which included testing
for environmental conditions that had led to the 2003 failure and a
more robust conformity assessment and accreditation program for the
independent laboratories conducting body armor testing and
certification.
The impact of that ongoing collaboration is enormous. Ballistic-
resistant body armor has been credited with saving the lives of more
than 3,000 police officers.
Another example resulted in the research and development of new
technology to permit DNA identification of a substantial proportion of
the human remains recovered from Ground Zero at the site of the World
Trade Center disaster on September 11, 2001. There were more than
20,000 bits of human tissue and bone recovered, some no larger than a
fingertip, and most were badly decomposed or partially incinerated. The
severely degraded condition of these human remains made it extremely
difficult for the forensic biology laboratory of the New York City
Office of the Chief Medical Examiner (OCME) and its collaborators to
obtain interpretable DNA profiles from these human fragments using
conventional DNA methodologies in practice within crime laboratories in
2001. DOJ and NIST collaborated and funded applied DNA research to
develop at NIST a new set of DNA reagent molecules called ``Mini-STRS''
that would enable scientists to go back and identify successfully much
more of the partially degraded DNA samples than ever before. The result
was a dramatic improvement in the proportion of fragments of human
remains that could be identified and associated with the known
reference DNA standards of the victims or members of their families.
The impact of that ongoing collaboration between DOJ and NIST is
also enormous. The result of that giant leap forward in forensic DNA
testing capability contributed significantly to the subsequent
identification of more than 1,600 victims who perished at the WTC
disaster on 9/11, many by DNA testing alone.
There are dozens of other similarly significant synergistic
outcomes of the successful collaboration between NIST and DOJ in the
world of testing, accreditation and standards development including
such forensic science disciplines as fire investigations, drug
detection, biometrics, firearms/ballistics, and genetics (DNA). This
partnership benefits from the mutual exploitation of the core strengths
in each other's agency to the benefit of the entire public safety
community and society.
Question 2. The area of forensics is not only interesting
scientifically, but also very important since the stakes are high. On
the one hand, we have issues of measurement, but on the other hand we
have judicial and legal issues. Could you outline some of the
unintended consequences you and DOJ have encountered so far?
Answer. As noted above, NIST and DOJ enjoy a decades old history of
successful collaborations in criminal investigations and supporting the
development forensic science standards and technology. There has been
much mutual discussion about the current status of forensic science
practice in the United States and the impact on judicial and legal
issues. NIST and DOJ are keenly aware of current issues and challenges
and our partnership affords us an effective view of matters from both
scientific and legal perspectives. NIST has subject matter expertise in
scientific measurement, technology and standards development and DOJ
has subject matter expertise in judicial and legal issues as well as
the practice of forensic science. In response to your request to
characterize the ``unintended consequences'' NIST and DOJ have
encountered so far, it is accurate to describe the collaboration quite
to the contrary--highly attuned to the needs of the forensic science
community, in large measure attributable to the experience and
expertise resident within our agencies as well as both agencies' long
history of outreach with state and local subject matter experts
throughout the Nation. As a matter of current operational practices,
NIST and DOJ routinely collaborate heavily with state and local
agencies in the formation of technology and standards development and
identifying current challenges to forensic science practitioners and
members of the criminal justice community.
Although one cannot foresee all possible further contingencies, the
boots on the ground in both organizations are career professionals and
both camps are keenly sensitive to the needs of the Nation's forensic
science community and value the input of state and local stakeholders
to provide the necessary guidance to inform our day-to-day
decisionmaking.