T  H  E      D  E  T  A  I  L

Monday, September 30, 2002

Good morning via the "Detail," a weekly e-mail newsletter that greets latent print examiners around the globe every Monday morning. The purpose of the Detail is to help keep you informed of the current state of affairs in the latent print community, to provide an avenue to circulate original fingerprint-related articles, and to announce important events as they happen in our field.

Last week, we took a break from our 3-part series on the Critics to look at the importance of policy and procedure in the comparison process.  This week we continue with the final chapter in Glenn Langenburg's series:
 

“DEFENDING AGAINST THE CRITIC’S CURSE”: CHAPTER 3, DR. DAVID STONEY
 

I.  INTRODUCTION

Kasey Wertheim and a few others have asked if I would attempt to put into writing some of the issues I discussed during my presentation at the ABFDE (American Board of Forensic Document Examiners) entitled “Defense Against the Dark Arts: Defending Against the Critic’s Curse”.  I have agreed to do so and will attempt to address the three most vocal critics: Dr. Simon Cole, Professor James Starrs, and Dr. David Stoney.  The purpose of these writings, as was the purpose of the original presentation, was not an attack on these three individuals, but rather an objective examination of who these individuals are, what are their major arguments and tactics, and then most importantly how to defend against their attacks and where to obtain the information to support your defense against their “curses”.  This is hardly all encompassing, and any additional information that you can provide would be most helpful.  Any comments that you may have please email me at Glenn.Langenburg@state.mn.us.  Any anonymous comments that you wish to make, please send to Kasey at clpex.com and he will forward them to me without any identifiers.  Please be critical if you see an error.

The third of these writings is focused on Dr. David Stoney.  I first had the opportunity to see Dr. Stoney present at DePaul University at a Daubert symposium in Chicago, Illinois on April 15, 2002, along with Simon Cole (see “DEFENDING AGAINST THE CRITIC’S CURSE”: CHAPTER 1, DR. SIMON COLE).  I was most impressed with Dr. Stoney and his insightful, though critical, views on the science of friction ridge skin identifications.  He raised valid issues and concerns which many in this profession agree need to be addressed.  I personally found Stoney to be distinctly different from Starrs and Cole, not only in his background, but also in his tactics, concerns, and opinions of friction ridge skin identifications.

Unlike the previous two “expert” critics, Stoney is an entirely different beast, and there is no simple defense.  He does not testify as an expert critic as often as Starrs and Cole do.  There are no terribly grievous errors in his testimony.  He has valid, professional criticisms against this discipline and understands the foundation and methodology.  He has contributed research and material towards the advancement of this profession.  The best defense against Dr. Stoney is a firm education in science and the fundamentals of friction ridge skin sciences and methodology AND the ability to articulate them.  Know your science, as most assuredly, he does.

II. BACKGROUND

Dr. Stoney earned a Bachelor of Science (BS) in chemistry and criminalistics from the University of California, Berkeley—a program established by the late great Paul Kirk.  From this same institution, Stoney earned a Master of Science in Public Health and a doctorate (Ph.D.) in forensic science (1).  His thesis work was based on quantitative statistical fingerprint modeling, resulting in various publications in texts and journals (2).  Dr. Stoney worked at the Institute of Forensic Sciences Criminalistic Laboratories in Oakland, California (an independent crime lab) (3).  While there, he performed various forensic examinations, including latent print comparisons (4).  Afterwards he served as an associate professor and director of the Forensic Science Program at the University of Illinois, Chicago (5).  Dr. Stoney is currently the director of the McCrone Research Institute in Chicago, Illinois and has served in this capacity since 1993 (6).  Stoney has published approximately two dozen articles on various aspects of forensic science, including fingerprints (7).

Unlike Cole and Starrs (see previous Chapters 1 and 2), Stoney is a forensic scientist, with the education, training, and experience to support that claim.  He also has limited (academically derived and apprenticeship under John Thornton, previously of the Contra Costa County Criminalistic Laboratory) training in the analysis, comparison, and evaluation of latent prints (8).  However, he has not had intensive, formalized (modular or otherwise) training in the identification of latent prints.  Furthermore, the number of comparisons he claims to have performed is less than 1000 (9).

Dr. Stoney’s main attacks are as follows:

1)     ACE-V methodology has elements of subjectivity and the evaluation is ultimately not scientific

2)     No objective criteria or measurements to measure individuality

3)     Reliability of examiners practicing the ACE-V method has not been sufficiently tested

4)     Error rate is meaningless without a standardized objective method of measurement

5)     Ultimately, fingerprint identification works and it’s good evidence, but it isn’t science and it doesn’t meet Daubert requirements

III.  ISSUE 1

ACE-V methodology has elements of subjectivity and the evaluation is ultimately not scientific.

Attack:

The result of ACE methodology is a subjective opinion.  A subjective opinion based on subjective standards is not scientific (10)

      Defense:

This is a difficult statement to defend against because there are elements of truth to it, allowing for various interpretations.  While it is true that ultimately whether a print matches or does not match is a subjective conclusion, it is not necessarily true that the steps to arrive there are completely devoid of any objective criteria.  Steve Meagher, a Unit Chief for the Latent Print Division of the FBI, stated quite succinctly that in fact our criteria for an identification is very exact: complete agreement of all ridge detail present between known and unknown with no unexplainable differences (11).  Furthermore, Pat Wertheim has drawn excellent analogies to the process of latent print comparison using scientific method (e.g. observation, hypothesis, testing, conclusion, and reliable predictability) to demonstrate the stages of analysis, comparison, and evaluation (12). 

It can also be argued that many aspects of science incorporate subjective decisions, criteria, and conclusions.  Taxonomy is an excellent example.  The classification and identification of species based on various quantitative and qualitative criteria is a very similar process.  In pathology and toxicology, there are many subjective interpretations a scientist must make.  Is this product causing class 2 or class 3 edema and rash on this rabbit’s skin?  Are these red blood cells deformed?  Is a correlation factor of 0.65 a strong or a weak indicator of a causal relationship?  To say that subjectivity has no place in science is not consistent with all the myriad aspects of science.  And contrary to some critics’ opinions, there is no consensus and standard definition among all the various sciences regarding exactly what defines “science” (13,14).

Personally, when I listen to this debate, there appear to be two steadfast camps:  Stoney calling for entirely objective standardized measurements and the pure ridgeologists that say what we currently do is acceptable and scientific.  I personally feel, as a scientist, that the answer lies somewhere in the middle.  This profession would perhaps benefit greatly by further defining various objective criteria and attempting to incorporate standardized measurement into the identification process.  Research is desperately needed here!  What are the frequencies of spurs, short ridges, dots, trifurcations and the like?  What would those frequencies tell us about weighing the various minutiae?  With what frequency do open fields (continuous ridge series with no minutiae) occur?  How can one calculate tolerance ranges for various types of distortions?  When we say total agreement between known and unknown, what does that mean?  Are there ways to measure all three levels of detail, using similar technology that the U.S. Postal Service uses for the analysis of handwriting, and formulate an actual correlation value between a known and unknown (16)?  Would the inclusion of these types of measurements increase the uniformity of examiner conclusions? These are all valid questions of our profession and we owe it to ourselves to at least examine their potential.

IV.  ISSUE 2

No objective criteria or measurements to measure individuality.

Attack:

How much correspondence between known and unknown prints is sufficient to conclude that they originated from the same unique source?  In other words, how much is enough?  Currently fingerprint examiners do not and cannot (17):

·        objectively quantify and measure the amount of detail in a fingerprint (including all three levels)

·        measure the correspondence of the detail between known and unknown

·        objectively interpret the meaning of a given correspondence between known and unknown (i.e. what does total agreement between two prints mean?)

      Defense:

These issues are in the same vein as Issue 1 above.  This does not mean that what we do is not acceptable and valid or does not work, but more importantly it raises the question: can we do it better and more uniformly? Also it logically follows that if one can measure the correspondence between two prints, then one can also measure the disparity between two prints.

V.  ISSUE 3

Reliability of examiners practicing the ACE-V method has not been sufficiently tested.
 

Attack:

The ACE-V methodology has not been objectively tested through controlled, scientific testing and validation procedures (18).  [It is interesting to note that unlike the other critics, Stoney does not attempt to support his argument with the results of various Collaborative Testing Service (CTS) proficiency examinations.  Perhaps he recognizes that these CTS exams are not scientific controlled studies.]

Defense:

Unfortunately, I cannot agree more with Stoney.  When compared to the types of validation studies that exist for analytical methods and analysis (e.g. EPA, FDA, GLP, ISO standards for validating methods) ACE-V has not been tested in a scientific and controlled environment.  This issue is one of concern and interest for me personally, and already several studies are being initiated by myself and others (19).  I hope that other scientists will also contribute to this need.

It is true that proficiency testing and “training to competency” encompass and measure individual performance and application of the methodology, which is an important and necessary factor for qualifying in court.  However individual proficiency and competency testing do not represent controlled scientific studies, nor are those data published, reviewed, and available to the latent print examiner community.  As one researcher warned, “If it isn’t published, it doesn’t exist (20).”

VI.  ISSUE 4

 
Error rate is meaningless without a standardized objective method of measurement.

Attack:

It is meaningless to enter into a discussion concerning error rate until an objective, standardized methodology exists which utilizes objective criteria and measurements.

Defense:

The standard defense against this line of attack is to differentiate between the error rate of the science (or theoretical error rate) and error rate of the scientist.  For purposes of court this is an effective answer.  However once again, I find myself agreeing with Stoney.

If an examiner declares an identification, but a second examiner opines that though the prints are in agreement, there is insufficient evidence to support the identification, is there an error?  When one is making standardized measurements there is always a degree of uncertainty and an error rate is calculable.  Stoney’s statement says just that: you need a standardized objective measurement to calculate an error.  If your target keeps moving from print to print (as we would expect based on a continuum of clarity and quantity of ridge detail), then it is impossible to define the target and calculate how often the target is missed (21).

What I feel most comfortable with is:  we cannot define an error rate with this current methodology, therefore we cannot calculate one.  This is clearly a complex issue and open for further discussion and debate.

VII.  ISSUE 5

Ultimately, fingerprint identification works and it’s good evidence, but it isn’t science and it doesn’t meet Daubert requirements (22).

Attack:

This statement nicely summarizes Stoney’s perspective.  It does not meet the requirements of science and Daubert because of the issues previously discussed.  However it works.  As he stated, “At some point the quantity and the quality of ridge information is great enough to make an identification.  The problem is no one knows at what point that is true and at what point does that becomes reliable”(23). 

Defense:

It is this type of statement that reduces the effectiveness of Stoney as an expert critic, because ultimately he agrees and admits that it does work and can be valuable, crucial evidence.  He admits to having made absolute identifications (24).  However he points out, and in some ways rightly so, that the profession needs to further scrutinize it’s methods, training, and standards and perform valuable research and testing.

I firmly disagree with Stoney’s statement that friction ridge skin identification is not a science.  I believe it is a science, the method is analogous to scientific method, and the resulting conclusions are falsifiable.  It can also be argued that the courts disagree with Stoney’s statement that is does not meet Daubert guidelines for reliability, because it has met various Daubert and modified Frye challenges, successfully, in over 40 instances (25).

SOURCES

(1)   U.S. v.Byron Mitchell, No. 96-407, PA 1999, Days 4 of Daubert hearing, p. 36.

(2)   Stoney, David; Thornton, John.  “A Critical Analysis of Quantitative Fingerprint Individuality Models” and “A Method for the Description of Minutia Pairs in Epidermal Ridge Patterns,” Journal of Forensic Sciences, 31 (4), Oct. 1986, p. 1187-1216; p.1217-1234; Stoney, David. “Measurement of Fingerprint Individuality,” Advances in Fingerprint Technology, 2^nd ed. (Lee and Gaensslen), CRC Press, 2001, p.327-387.

(3)   U.S. v.Byron Mitchell, No. 96-407, PA 1999, Day 4 of Daubert hearing, p. 40.

(4)   Ibid.

(5)   Ibid. p. 39.

(6)   Ibid. p. 37.

(7)   Ibid. p. 41.

(8)   Ibid. p. 58-63.

(9)   Ibid. p. 62.

(10)                      DePaul University Daubert Symposium notes, Chicago, IL, April 15, 2002.

(11)                      International Assoc. for Identification 87^th International Educational Conference, SWGFAST Panel Discussion, personal notes, Las Vegas, NV, August 7, 2002.

(12)                      Wertheim, Pat. “Advanced Ridgeology Comparison Techniques” Training Course, Santa Barbara, CA, October 16-20, 2000.

(13)                      DePaul University Daubert Symposium notes, Chicago, IL, April 15, 2002.

(14)                      Meyer, Carl.  Expert Witnessing:  Explaining and Understanding Science.  CRC Press, 1999.

(15)                      Stoney, David. “Measurement of Fingerprint Individuality,” Advances in Fingerprint Technology, 2^nd ed. (Lee and Gaensslen), CRC Press, 2001, p.329-330.

(16)                      Srihari S., Cha S., Arora H., Lee S. “Individuality of Handwriting,” Journal of Forensic Sciences;2002, 47 (4), p. 1-17.

(17)                      Stoney, David. “Measurement of Fingerprint Individuality,” Advances in Fingerprint Technology, 2^nd ed. (Lee and Gaensslen), CRC Press, 2001, p.329-330.

(18)                      Ibid. p. 330 and U.S. v.Byron Mitchell, No. 96-407, PA 1999, Day 4 of Daubert hearing, p. 87.

(19)                      Langenburg, G.  “A Pilot Study Statistical Analysis of the ACE-V Methodology Analysis Stage”, unpublished study at the time of this writing, expected date: winter 2002.

(20)                      American Board of Forensic Document Examiners Daubert Symposium, Las Vegas, NV, June 2002; personal notes.

(21)                      DePaul University Daubert Symposium notes, Chicago, IL, April 15, 2002.

(22)                      Ibid.

(23)                      Ibid.

(24)                      U.S. v.Byron Mitchell, No. 96-407, PA 1999, Day 4 of Daubert hearing, p. 55.

(25)                      World wide web: onin.com, clpex.com

This concludes the three part series focusing on the three prominent critics of the friction ridge identification discipline: Dr. Simon Cole (Chapter 1), Professor James Starrs (Chapter 2), and Dr. David Stoney (Chapter 3).  Please contact the author if you have further questions, concerns, or criticisms.

Glenn Langenburg, Forensic Scientist, Latent Print Examiner
Minnesota Bureau of Criminal Apprehension
1246 University Ave
St. Paul, MN  55104-4197
(651) 642-0700

About the Author:

Glenn Langenburg has been with the Minnesota Bureau of Criminal Apprehension since January of 2000, serving as a latent print examiner and crime scene investigator.  He graduated from Michigan State University in 1993 with a BS in Criminalistics under the esteemed Dr. Jay Siegel. In 1999, he earned a Master of Science degree in Analytical Chemistry under Dr. Peter Carr (a highly respected chromatography expert) at the University of Minnesota.  Currently he is a PhD candidate in the Toxicology program at the University of Minnesota, but is considering switching to a PhD program in forensic science to continue research involving the statistical analysis of the ACE-V methodology.

The informal CLPEX.com message board is available for banter about the Detail:
http://www.clpex.com/phpBB/viewforum.php?f=2
And the onin.com forum (http://onin.com/fp/wwwbd/) is also available for more formal latent print-related discussions.