Problem Idents

Errors vs. Idents... Posted Here for Your Review

New Case - US
Madrid Bombing

First Case - US

Second Case - Scotland

Third Case - Scotland

Opinion Page

Updated 15 Jun 05
by Ed German, CLPE, FFS

The Japanese have a proverb that applies to experts who think they are above making mistakes:

Saru mo ki kara ochiru gif

"saru mo ki kara ochiru"

The phrase literally translates as "even monkeys fall from trees." In English, the closest adage may be "pride comes before a fall."

Because in every human endeavor there will be oversights, quality control and quality assurance are important to assure reliable results. It is also important that persons incriminated through latent print identifications should have ready access to competent experts to reexamine alleged identifications.

On rare occasions, a fingerprint or palmprint identification error occurs, even among experts trained to competency. In fingerprint identification, DNA identification, firearms identification and other relatively mature forensic sciences, erroneous individualizations are so rare that they cause international news headlines. That extreme rarity benefits society and justice.

These are real cases. This page exists to demonstrate errors so that more than just the experts involved in the cases might learn from mistakes.

In the words of a famous Scotsman,

"There's nane ever fear'd that the truth should be heard but they whom the truth would indite."

If you possess clearer images demonstrating better ridge detail for these cases, please submit them. I will gladly post such images, pointing out correlations they demonstrate which are missing from images on these pages.

-
There is zero tolerance for friction ridge identification error in forensic science.

Ignoring standards causes errors. A minimum number of Galton detail "points" does NOT equal an identification standard. The standard for identification is not a number.

As demonstrated below, relying on a specific number of allegedly matching "points" can be a formula for disaster. The misleading comfort of a "number" can lead Examiners (and Police administrators) to ignore more important factors which insure reliability.

American national standards are certainly not the only watermark by which to measure reliability. References herein are not meant to cast doubt on the value of expert standards established outside America for insuring the reliability of the friction ridge identification process. I write here about what I know best... the identification process in the United States. This page does not purport to represent the opinion of any professional organization or government entity. All the good parts I borrowed from other forensic scientists, any faults in logic or explanation are mine. At the bottom of this page I have posted my personal reasons for building it. --Ed German

These prints are the best images of the Madrid Erroneous Identification submitted to the webservant to date. These are not official photos from any government agency. These images do not allege to accurately represent the spatial resolution, radiometric resolution or other variables of the images produced in Spain and examined by the FBI. They are presented here for the limited evaluation possible from such unofficial images. As time permits, the webservant will prepare additional demonstrative images. Should any of the concerned parties desire to provide images more closely representing the actual evidence, your webservant will post them.

Click the small images to see larger images.

Madrid Error Latent Print

Madrid Error Record Print

Madrid Error Latent Print
(rotated 17 degrees CCW)

Madrid Error Record Print
(not rotated, cropped to match corresponding latent area)

Coming immediately after his trial: the inked print of Algerian Ouhnane Daoud.

The above erroneous identification is acknowledged by all experts involved and (despite early news reports about FBI statements that the latent print was not suitable for identification) no challenge to the Daoud identification is now occurring among fingerprint experts. Publication of the Daoud images here is withheld pending adjudication in Spain.

Click here to post your comments about the above error.

FIRST CASE:

Chart Used at Two Separate Court Hearings

This "identification" was effected by an IAI Certified Latent Print Examiner employed by a small American police department in the Midwest. The examiner had a four year college degree and passed the IAI CLPE certification exam. The Examiner's certification was revoked by the IAI Latent Print Certification Board because of this erroneous identification.

The Examiner never apprenticed under a competent examiner, but had a self-trained Latent Print Examiner "mentor" 90 miles away who occasionally reviewed some identifications. The "mentor" (employed by the State Police laboratory system) was "charged" with making a separate erroneous identification three months after the error displayed here was discovered. The "mentor" had never been proficiency tested, had previously retired from a law enforcement agency in another state as a crime scene technician and was basically self-trained... having never apprenticed under a Latent Print Examiner. The chart scanned here was used during testimony at both a preliminary hearing and a parole revocation hearing. Legal prosecution ceased when the error was revealed. The Examiner's agency did not require identification verifications... and also did not participate in annual proficiency testing.

Click this Picture for an Enlarged Image

Click here to see an enlargement of the Latent Print

Click here to see an enlargement of the Record Print

SECOND CASE:

LATENT FINGERPRINT

Click the Images to see enlargements of the Charts

INKED FINGERPRINT

This erroneous "identification" was effected and verified by four Latent Print Examiners of the Scottish Criminal Records Office and presented as evidence in the trial of Policewoman Shirlie McKie.

American Latent Print Examiners Pat Wertheim and David Grieve presented testimony that it is NOT an identification.

Shirlie McKie was found not guilty and no indication has been made public to date as to whether the jury concurred with the foreign experts or merely considered their testimony as negating the Scottish experts.

On January 19th, 2000, the first UK police expert publicly denounced the SCRO "ident" (first to the author's knowledge). Allan Bayle, a senior fingerprint expert (Ridgeology Specialist) with New Scotland Yard's Scientific Support College posted:

I looked at the McKie case in September. The mark is not identical. I have shown this mark to many experts in the UK and they have come to the same conclusions. I am willing to assist the Mckies, on this mark should they require it.

This display of these small charts is of limited quality. Click on the pictures to see enlargements which show each pixel of the digital image court chart hard copies.

Scan of police crime scene photo of the latent print.

Scan of photo made at the crime scene by Pat Wertheim in March 1999. The light colored scratch through the ridge detail was apparently due to inadvertent contact after the police photos were made.

Scan of original inked print.

If you are a fingerprint expert looking at these images, you should already be able to see that the latent print depicted in the court chart, the scan of the police crime scene photo and the March 1999 scan all depict the same mark from the crime scene.

If you are not a fingerprint expert and this all sounds like Greek to you, click here to see the three latent print images in the same relative size and same orientation. The blue outline on the two larger images corresponds with the latent print area in the court chart.

The court chart has image sampling pixel interference (mottled square or step-like artifacts) and very high contrast.

The SCRO crime scene photo is slightly blurred, but includes all the ridge detail.

The March 1999 crime scene photo has the best clarity or sharpness of the three images, but includes only ridge detail that is also present in the slightly blurred SCRO crime scene photo.

Click this picture to see a larger image

Click on the picture above to see an enlargement.

The green dot at left was identified by the Latent Print Examiners (who effected the "ident") as point number 14 in the above record and latent print charts.

The red dots below the green dot are points 4 and 5 in the above record and latent print charts.

The red dots with yellow arrows were NOT charted by the SCRO experts, but are two of the most obvious discrepancies (neither is present in the other print).

Click here to download a 600 Kb (big file - 15 minutes via some phone modems) uncompressed TIF file scanned directly from the negative exposed at the crime scene. This TIF file has greater detail than the compressed JPG images on this page, but will not open in many web browsers and must instead be viewed with imaging software (or viewed by inserting into a Microsoft Word or other word processing document).

Click here to download a 2.5 Mb (really big file - 60 minutes via some phone modems) uncompressed TIF file scanned directly from the negative exposed at the crime scene. This TIF file has greater detail than the compressed JPG images on this page, but will not open in many web browsers and must instead be viewed with imaging software (or viewed by inserting into a Microsoft Word or other word processing document).

Click these images to see additional enlarged inked prints from Shirlie McKie (the smaller image is a 292 Kb file and the larger is 493 Kb). These eight inked prints were scanned from one piece of paper... the right impression of the smaller image and the left impression from the larger image is the common area of overlap for the two scans.

BBC Article and Audio Interview

Shirley McKie Case Update

Click here for the latest details at
www.shirleymckie.com

Friction ridge identification is a reliable science.

Since the mid-1970's the International Association for Identification (IAI) and its Latent Print Certification Board have worked with state and regional chartered IAI Divisions throughout the U.S. toward the goal of improving professional abilities. Since 1995 the Federal Bureau of Investigation has sponsored the Scientific Working Group on Friction Ridge Analysis, Study and Technology (SWGFAST) for developing consensus standards which preserve and improve the quality of service provided by the latent print community nationwide.

During 1995 through 1997 American Latent Print Examiners established minimum qualification, training and quality assurance SWGFAST guidelines through review, discussion and feedback at local and national meetings, and through peer review in professional journals.

Latent Print Examiners are able to reliably make positive identifications:

1. With appropriate training (SWGFAST training content guidelines);
2. With appropriate experience (CLPE requirements and SWGFAST training guidelines);

3. With appropriate ability (initial competency testing for CLPE, and annual proficiency testing as recommended by SWGFAST QA guidelines);

4. When using the scientific procedure of ACE-V as follows:

Analysis – the qualitative and quantitative assessment of Level 1, 2 and 3 details to determine their proportion, interrelationship and value to individualize.
Comparison – to examine the attributes observed during analysis in order to determine agreement or discrepancies between two friction ridge impressions.

Evaluation – the cyclical procedure of comparison between two friction ridge impressions to effect a decision, i.e., made by the same friction skin, not made by the same friction skin, or insufficient detail to form a conclusive decision.

Verification – an independent analysis, comparison and evaluation by a second qualified examiner of the friction ridge impressions. (SWGFAST QA guidelines)

Friction ridge identification reliability is insured by Training, Experience and Ability coupled with ACE-V to effect a positive identification when sufficient quality (clarity) and quantity of corresponding Level 1, 2 and 3 detail exists. Quality (clarity) and quantity vary with each friction ridge impression.

Analysis is objective.

Comparison uses objective observations.

The final identification decision is subjective and is reached when sufficient quality (clarity) and quantity of corresponding Level 1, 2 and 3 friction ridge details are present.

The quality (clarity) and quantity of details necessary to effect the identification can vary based on training, experience and ability of the Latent Print Examiner.

Even though each latent print can have different quality (clarity) and quantity of Level 1, 2 and 3 details... Latent Print Examiners with similar training, experience and ability should be able to identify the same latent and inked prints.

Quality assurance and quality control guidelines exist to detect erroneous friction ridge impression identifications. The Examiner making the error is immediately suspended from casework activity. Remedies to insure operational reliability include the following:

Return to friction ridge identification duties only after retraining, evaluation and tested ability indicate the expert can operate at zero error rate;

Permanent transfer to other duties;

Employment termination;

David Ashbaugh's important Ridgeology publication defines Level 1, 2 and 3 details and explains in plain language the scientific methodology called ACE-V in use by Latent Print Examiners worldwide.

Return to Page Top

The Acceptable Error Rate for Friction Skin Identifications is ZERO
If you are a Latent Print Examiner, you know the above statement to be true. If you are an attorney, police administrator or other non-expert, this may sound unrealistically strict. Scientists generally concur that a reliability of 95% constitutes scientific certainty ^{(1), (2)}. In many areas of science and many professions, such reliability serves society well. For example, the acceptable error rate for Chicken Sexers (persons who examine the backsides of one-day-old chicks to sort males from females) is between 2.3 and 5 percent when examining 700 to 800 chicks per hour.
A latent finger or palm print "identification" is the determination that two corresponding areas of friction skin impressions originated from the same person to the exclusion of all others. This absolute identification decision can carry great weight in police decisions about investigations, and court decisions about guilt or innocence.

Latent Print Examiner errors involving evaluation (whether marks are of value or not) and elimination (marks were not made by a person) are serious, but do not normally carry the same potential for harm as erroneous identifications. An example of the degree of importance afforded identification decisions can be seen in scoring procedures for friction ridge impression comparison tests. Universally, even if an exam has hundreds of latent prints and thousands of record (inked) prints, some mistakes involving evaluation or elimination are expected, but a single erroneous identification is automatic failure. The acceptable finger or palm print identification error rate is zero.

Because in any field of human endeavor there will always be errors, quality control and quality assurance standards are necessary to protect society. Standards for initial "training to competency," case by case quality control, and long term quality assurance including life cycle training and periodic proficiency testing are essential. Periodic proficiency testing can reveal bad habits developed since initial training to competency. It may also reveal deteriorating eyesight or other human deficiencies impacting casework.

One set of standards is set forth in the Quality Assurance Guidelines of the Scientific Working Group on Friction Ridge Analysis, Study and Technology (SWGFAST).

When erroneous identification decisions (not just clerical errors and not just failing to find a "match") are detected, most agencies immediately suspend the examiner(s) from further casework activity. Remedies to ensure a zero error rate include the following:

Return of the examiner(s) to friction ridge identification duties only after retraining and evaluation indicate the expert(s) can operate at zero identification error rate;
Permanent transfer to other duties;

Employment termination.

(1) THE USES OF SCIENTIFIC UNCERTAINTY, Peter Montague, 1999
(2) STANDARDS OF EVIDENCE NECESSARY TO DEVELOP POLICY, C. Sage, 1997

Return to Page Top

Third Case (David Asbury Case) - another problem identification from the SCRO


Original Evidence Surface	Latent Print	Record Print

Click on the images above to see larger jpg images.

If you are an expert:

Click here to download a large tif file of the latent print (approx. 2300K file size).
-
Click here to download a large tif file of the record print (approx. 2800K file size).

As can be seen from the left and middle images above, the luminescing latent print was photographed on a curved surface.

Fingerprint experts presented testimony about the latent and record prints, charting many "matching points" to prove the "identification." Four of the alleged "matching points" are indicated with red and green dots on the image below.

Click on the above image to see an even larger image. The blue lines on the images match the same relative position as red lines used on court chart images by the fingerprint experts.

The images presented above are much sharper than charted images used in court with this "identification." You can see the dots of the "coach scene" background pattern in the latent print photos on this page, versus the blurred court chart images of the latent print.

Click above to see a larger image.

Click above to see a larger image.

	Close-up of seven of the inked print points on the fingerprint experts' court chart.
	Close-up of seven of the latent print points on the fingerprint experts' court chart.
	Close-up of the same area of the latent print in much sharper focus. The blue lines correspond to the fingerprint experts' charted red lines.
	Close-up of the same area of the latent print in much sharper focus (here without the blue lines).

David Asbury Case
by Pat Wertheim

In January 1996, the body of Marion Ross, a middle-aged spinster, was found in her small home in Kilmarnock, Scotland. The investigation, as well as the search of the crime scene, was undertaken by the police agency with jurisdiction, the Strathclyde Police. They sealed off the scene, established perimeter control, and began the task of collecting evidence.

Among other physical evidence, a number of crime scene marks (latent prints) were collected. A suspect developed, a young handyman named David Asbury, who had done some maintenance repairs in Miss Ross’ home a couple of years earlier. Evidence was collected from his apartment, including a small decorative tin that originally came with individually wrapped candy pieces in it. The tin had a scene painted on it of a horse drawn streetcar or carriage.

Subsequently, one mark on the tin was identified as having been made by the victim, Marion Ross, and one mark from the scene, found on a gift tag on an unopened Christmas gift, was identified as having been made by David Asbury. He was arrested and charged with murder. The only physical evidence used against him in court was the fingerprint evidence: his print inside the scene and the victim’s print on an item in his home. The Strathclyde Police do not have any fingerprint examiners and rely on the services of the Scottish Criminal Records Office (SCRO). Four SCRO experts signed the report making the two identifications.

When the SCRO was eliminating the “outstanding” (unidentified) marks, one mark from inside the crime scene was identified as having been made by Detective Constable Shirley McKie. The fingerprint evidence in the Asbury trial was thrown into doubt when DC McKie testified that she had not been inside the Ross home, in spite of the “fact” that her fingerprint had been identified there. (See “Scotland Case” on this web site). The same four SCRO experts who had signed the report identifying David Asbury inside the scene and Marion Ross in Asbury’s apartment had also signed the report naming DC McKie as the source of the elimination mark inside the scene.

In spite of the doubt DC McKie threw on the fingerprint evidence, the jury convicted David Asbury of murder and sentenced him to life in prison.

This site attempts to give fingerprint examiners around the world a look at the mark on the candy tin and the inked print of Marion Ross, which was “identified” by the SCRO. The mark used in court was not the clearest photograph available, but was one that appears to have been taken slightly out of focus. A crisp, sharply in focus photograph of the mark was found in other court exhibits. In addition, the image of the mark used for charting purposes appears to be digitally degraded, rather than enhanced. Compare the sharp photograph of the mark with the out-of-focus photograph of the mark, and then with the image in the charted enlargements.

Conduct your own comparison and reach your own conclusion. Do you see points in the mark that are not there in the inked? Do you see points in the inked that are not there in the mark? Can you see each of the sixteen points where they supposedly exist in the latent? Is the mark of no value? Is it an erroneous identification? Assuming that the mark on the gift tag was a correct identification of David Asbury inside the scene, as outside experts have easily agreed, should the conviction stand when it was based in part on this “identification,” or should David Asbury be granted a new trial? You be the judge and please post your thoughts on the comments page.

What is your expert opinion?

What do you think about...

The 16 Points charted as matching?
Is an identification supported by the ridge detail present?

Comments for posting on a page about this "identification" are welcome. Postings will not purport to represent the position or opinion of any professional organization or government entity, just of individual experts.

Click here to post your opinion.

As an expert working in fingerprints since 1971, I have seen some of the skeletons in my discipline's closet. In the below paragraphs you can read my personal reasons for building this page. --Ed German

In 1984, a friend of mine (a Latent Print Examiner in the Midwest) brought me the charts shown above in the "first case." That same day I (and coworker Steve Kasarsky) caused an incarcerated prisoner to be freed and charges dropped because we pointed out the error. The wrongfully incarcerated prisoner was on parole from prison. The Latent Print Examiner, being relatively new in the business, had not previously caused anyone's incarceration based upon fingerprint evidence and the Prosecutor decided that no future warrants would be issued based on just the local examiner's work. The "first case" Latent Print Examiner was decertified and continued working as a policeman and crime scene technician. He is still employed by the same agency and to my knowledge has since always submitted fingerprint identifications to outside agencies for verification. I do not reveal the city or state here because I am proud of his (and his department's) integrity and professionalism.

In 1987, a Latent Print Examiner from North Carolina brought me a crime scene mark for "computer enhancement" because he was having trouble charting it for upcoming court. The person "identified" was an active duty US Army staff sergeant who at the time had been in jail for over a year awaiting trial on murder charges. The latent fingerprint was in blood, from the doorknob of the bedroom where the body was found. I told the North Carolina Latent Print Examiner that no enhancement in the world would ever change the fact that the bloody print was not made by the suspect. He said that he wanted a second opinion and I sent him immediately to Marty Ludas in Raleigh, North Carolina who also told him that it was a "bum ident." Despite assurances to both me and Marty Ludas that he would immediately tell the prosecutor, we later learned that the expert allowed the incorrectly identified sergeant to stay in jail another two weeks, until the morning of the start of the murder trial. On the morning jury selection commenced for the trial, the Latent Print Examiner told the Prosecutor that although there was enough similarity in the latent print to make a tentative identification... he had determined just recently that it was NOT a positive identification. This was a ludicrous remark (tentative identification upon which to base arrest), one I have never heard since. The charges were dropped on the spot and the Sergeant freed right there in the courtroom.

When I learned of the sequence of events, I immediately contacted the Prosecutor and Chief of Detectives in North Carolina to relate that the error had been revealed two weeks earlier. Also, because the erroneous "expert" was self-trained and was the trainer for the only other "expert" in his office (the blind leading the blind regarding expertise), I urged the Chief of Detectives to have 100% of all their prior casework reviewed by outside experts. The local police balked at my suggestion and told me that their "expert" said it was a once in a lifetime error and that I was just attacking him and his coworker because they refused to participate in the IAI's certification testing program.

I continued to assert that where there is smoke, there is fire. After several calls and formal letters to the local police and Prosecutor, I finally "won" and it was decided that no new warrants would be issued based on the "LP Examiners'" reports until completion of an outside review of all their casework. The North Carolina state crime lab declined to participate, but Robert Hazen, Senior Expert at the FBI's Latent Fingerprint Section, agreed to supervise the 100% review by his subordinate experts. The review revealed a number of serious errors, including a couple of erroneous fingerprint identifications involving "elimination" prints of police and victims... and also revealed that valid identifications against suspects in serious crimes had been missed. Nobody else was freed from jail, but I still felt vindicated from the claims I was persecuting the two "experts." The last I heard, both "experts" had left the agency and the senior of the two was teaching forensic science at the local community college.

I know that fingerprint (latent print examination) experts perform an invaluable service to society. I also believe that quality control/assurance and other checks and balances such as international peer review must exist to insure the discipline's reliability.

In the "first case," the wrongfully accused was immediately released. In the North Carolina case, I have always regretted that the Army sergeant spent an additional two weeks in jail because I trusted the expert to notify the Prosecutor immediately. It taught me a lot about "expert" egos and recalcitrance related to exposing serious error.

If you are an expert and have knowledge of an erroneous identification impacting a person's liberty or reputation you must "do what has to be done" to inform the investigators in charge of the case... and verify timely notification to the Prosecutor and/or Defense/Court.

If you are an expert and have additional examples of erroneous identifications you think might benefit the discipline through Internet posting, please e-mail me.

--Ed German, CLPE, FFS

Return to Page Top

The webservant feels compelled to assure readers that the content herein does not purport to represent the position or opinion of the US Government, US Department of Defense, US Army, US Army Criminal Investigation Command, Intelligence Division of the HQ, US Army Criminal Investigation Command, SWGFAST, ASCLD/LAB, or any other entity or organization with which the webservant is associated.