By: Michele Triplett
FDIAI News, Oct-Dec 2022
PNWDIAI The Examiner, Fall 2022
Try this experiment based on 5th grade math.
Ask 10 people to solve the equation 4 + 12 ÷ 2 = _______ . Some people will interpret this information from left to right, arriving at an answer of 8. Others will perform the 12÷2 portion first, 4 + (12 ÷ 2) = 4 + (6) = 10. Which answer is correct?
PROBLEM 1: Without a standard on how to solve a problem, you may get different answers to the same question. These answers are differing opinions, not scientific conclusions.
Scientific conclusions are responses arrived at with a systematic approach that follows rules, not responses arrived at using personal interpretations.
PROBLEM 2: As you can also see from this experiment, which has been tested extensively on facebook pages around the world, independently reproducing a conclusion (verification) would not always ensure a quality result. Both answers, 8 and 10, will be reproduced by others in a large enough sampling.
Reproducibility within itself does not ensure quality results.
SOLUTION: Mathematicians developed standards to ensure information is interpretated consistently among everyone. The standard for how to interpret the question of 4 + 12 ÷ 2 = is known as the order of operations. Multiplication and division are performed first, from left to right, and then addition and subtraction are performed. By using the order of operations, the division portion of the above problem is performed first, followed by the addition portion, making 10 the correct conclusion.
Do these problems sound familiar to Friction Ridge Examiners?
In 2013, the OSAC (Organization of Scientific Area Committees), under the guidance of NIST (National Institute of Standards and Technology), was tasked with developing standards for many of the forensic disciplines. However, the standards being developed still rely on the self-confidence of the practitioner when interpreting data. Self-confidence cannot be standardized. This is one reason the OSAC and ASB (Academy Standard Board) standards for friction ridge comparisons are destined for failure. In lieu of allowing self-confidence, data that is demonstrable to the satisfaction of others (i.e., general consensus interpretation), should be utilized to standardize interpretations. General consensus in science is not the same as consensus by a majority vote. General consensus in science is a term that means that all doubt has been resolved. The interpretation of data can be checked for general consensus prior to developing a conclusion based on that data.
CONCLUSIONS:
- Standards/rules need to be developed for friction ridge comparisons in order to ensure quality results that
are reproducible. - Standards need to promote scientific protocols and discourage personal interpretations and opinions.
- Verification must be used as a review, to ensure proper application/interpretation of rules. Verification
used as a review process is a stronger quality assurance measure than simply having a second expert arrive
at the same conclusion. A second expert independently performing a comparison is time consuming, with
a low amount of effectiveness; many of the past known errors were verified in this manner and did not
prevent the error from occurring. - Reproducibility is a requirement of science; however, it is not a requirement for solving analytical conclusions. Reproducibility is a requirement for physical experiments.
- Those that recognize the value of verification being a review process, also recognize that you cannot
blindly review the work of another expert. Blind verification, under this convention, is not possible.
Once standards (rules) are developed that discourage personal interpretations, many other areas will also be
improved, such as error rate studies, proficiency tests and the ability to transparently articulate information
regarding a comparison to the courts. Current error rate studies measure conclusions against ground truth answers, which is unknown in casework. Under the suggested convention, errors are judged based on correct
application of the systematic approach instead of measured on whether a practitioner arrived at the ground
truth conclusion.
This is only the first step of building a strong foundation; a foundation that follows mainstream science and is
based on scientific rigor. Additional rules that are needed will be identified as the process is flushed out. Some
obvious questions are:
What are the possible conclusions?
What is the risk of error for a particular comparison?
How much documentation is required?
However, the development of additional rules will only be effective if they rest on a solid foundation.
Michele Triplett is the Forensic Operation Manager/Quality
Manager for the King County Regional AFIS Program in
Seattle, WA. She is a Certified Latent Print Examiner and
holds a BS in Mathematics and Statistical Analysis. She has
been employed in the friction ridge identification discipline
since 1991 and is actively involved with several organizations,
committees, and educational events. She currently
serves as the Director of the I A I Certification Programs
and as the secretary for the Academy Standards Board Friction
Ridge Consensus Body. Ms. Triplett teaches classes on
scientific principles, has developed a method for measuring
the complexity of pattern evidence comparisons, and has
authored the ‘Fingerprint Dictionary’.
