Training new examiners to competency should incorporate the largest and best amounts of corresponding ridge formation ever recorded in a non-match.  Latent print examiners should research, study, and be tested on these amounts in order to familiarize themselves with the extent similar ridge formations exist between different sources.  The largest and best amounts found in these non-matches will define relative pre-determined working minimum thresholds for amounts of corresponding ridge features in a latent v. exemplar that should be exceeded in order to establish valid basis for sufficiency to infer positive identification. 

For routine casework, and pending formal application of any statistical probability model, the author suggests latent print examiners study the Clark non-match.  Based on independent  statements by John "Dusty" Clark and Christophe Champod, and results from application of the T-Model (see Clark Non-Match), the Clark non-match is one of the, if not the, largest and best fingerprint look-alikes ever recorded.  In order to establish valid basis for sufficiency to infer identification, in general, the amount of matching ridge features present in two impressions should exceed the amount present in the Clark non-match.  This fact requires latent print examiners to be highly familiar with the Clark non-match and consequently should be of primary importance in latent print examiner training.

Blind Proficiency Testing

In order to test an examiner’s ability to not make erroneous identifications, amounts of corresponding ridge formations found in non-matches can be incorporated in proficiency exams in the form of routine casework, e.g., under normal working conditions and without knowledge that a given case is a test. 

Partial known impressions from a person can be searched against AFIS databases to establish look-alike candidates.  The person’s print can then be deposited on an object, processed, and recovered as a normal latent.  The known latent can then be used in a proficiency test in the form of routine casework and the look-alike candidate’s ten-print record used as the exemplar to be compared. 

Example

The below latent impression belongs to the author and the below known impression is a look-alike found as a result of a State of California Department of Justice AFIS search.  These impressions can be used to create a mock latent case and administered to a fingerprint examiner in the form of routine casework.  A supervisor may designate one examiner to submit the latent to another examiner to verify identification.  This format tests the reviewing examiner’s ability to not make erroneous identifications as well as insures independent decision-making since the reviewing examiner is unaware the case is a test.

The below sample test case was made by the author.  A number of different parts of the author’s right middle fingerprint (Figure 14) were automated in the Santa Clara County Automated Fingerprint Identification System database to search for look-alike candidates. 
 

Figure 14

Exemplar of author’s right middle fingerprint.

Candidate exemplar fingerprint impression bearing ridge formations that “correspond” to ridge formations found in the “latent” fingerprint impression made by the author’s right middle finger.



The author deposited his right middle fingerprint on a metal cabinet, dusted it using standard black fingerprint powder, and then examined the developed latent lift.  The non-corresponding ridge formations in the latent were “wiped away” leaving only “corresponding” ridge formations found in the look-alike candidate.  The “latent” fingerprint impression was then lifted using conventional lifting tape and placed on a latent lift card (Figure 16). 

Figure 16

“Latent” fingerprint impression made by the author’s right middle finger.

A fingerprint chart displaying the “corresponding” ridge formations in the latent and exemplar was made using Adobe Photoshop (Figures 17 and 18) displaying the “corresponding” ridge formations.  Even though the test latent v. exemplar contains some minor red flags, the amount of apparent corresponding ridge detail is within tolerance and sufficient for purposes of a proficiency exam. The latent v. exemplar is especially unique since two pores were recorded in relatively similar positions.  If the “latent” fingerprint impression on the latent lift card and the exemplar were submitted to you during the course of routine casework with your supervisor’s request to review his/her identification, is it possible you could erroneously identify it?

 

Figure 17

Charted “latent” fingerprint impression made by the author’s right middle finger.

Figure 18

Charted exemplar containing look-alike amount of ridge detail to author’s right middle fingerprint.

A    Pore

B    Pore

C    Bifurcation

D    Ending Ridge/Bifurcation

E    “V” shaped Edge contour

F    Ending Ridge           

G    Ridge Break

H    Ridge Break                          

I    Ending Ridge 

J    Ending Ridge

K    Ending Ridge

L    Ending Ridge

M   Ending Ridge/Bifurcation

NEXT PAGE >>> 

Top 10 reads about probability and statistics:

1.  The History of Statistics:  The Measurement of Uncertainty before 1900 by Stephen M. Stigler

2.  Introduction to Statistics for Forensic Scientists by David Lucy

3.  Statistics Hacks:  Tips and Tools for Measuring the World and Beating the Odds by Bruce Frey

4.  Understanding Basic Statistics by Brase and Brase

5.  Statistics:  A Self-Teaching Guide by Donald J. Koosis 

6.  A Treatise on Probability by John Maynard Keynes

7.  The Black Swan by Nassim Nicholas Taleb

8.  Fooled by Randomness by Nicholas Taleb

9.  Against the Odds:  The Remarkable Story of Risk by Peter L. Bernstein

10.  Solving Crime With Mathematics by Keith Devlin and Gary Lorden

Fingerprint data is relatively easy to obtain.  The problem is there appears to be little interest in the broad fingerprint community to gather this data and relatively few experts to analyze it.

The next dream job:  Statistician  (video)