From Popperian to Bayesian

Last Monday, I reminisced about my early training in fingerprint classification and comparison. In those early days, the only formal classes available to me in Texas were the Texas DPS Academy courses and the FBI courses. In Kerrville where I worked, there was no on the job (OJT) training. The IAI did not teach classes. There were no published training standards. #SWGFAST was still almost 20 years in the future. There was no laboratory accreditation. The annual #IAI and Division conferences did not have breakout sessions or workshops, but there were always lectures of interest and, more importantly, wonderful networking opportunities and senior latent print examiners eager to answer a rookie’s questions.

When I took the position of supervisor of the ID Unit at Plano PD in 1986, I began taking inked and latent prints from other employees and putting together comparison exercises to train my ID Technicians. I wanted to give them something more than I had in the way of training. I put together a weeklong series of comparison exercises, starting with easy full pattern comparisons and ending with latents having minimal features and no focal points. I had no “train the trainer” instruction, only my instincts. I made a short, rough outline of topics I thought were important. My first class was in 1987. I had a total five students from Plano, Garland, and Richardson PDs.

Word spread and I began teaching outside my own department. The North Central Texas Regional Police Academy asked me to teach the class there in 1988. I continued teaching there for a number of years. One of the people I had met at IAI, Ann Punter, asked me to teach my class at Ontario, CA PD for her employees in 1992 and other PDs in California followed. When the IAI Board created a training division, Jim Murray was the first coordinator and I was the first IAI instructor. My first class in that role was in Conyers, GA in 1996.

I have refined that original class through almost 90 sessions in the US, Europe, and Australia/New Zealand. I continue to teach a much-evolved version of it for the IAI through their training arm, #TriTechForensics. Other latent print training, such as the FBI Academy used to teach, has declined, but dozens of other instructors have started teaching, both for TriTech Forensics (over 100 classes offered this year!), for other forensic training organizations, and on their own.

The concept of forensic science is transitioning from the Popperian ideas of my early days to the Bayesian approach being developed today. The class, as I teach it now, includes a look backwards at the historical perspectives of Sir Francis Galton, Edmund Locard, Karl Popper, Thomas Kuhn, and Carl Hempel. Deborah Smith and I sometime co-teach, and she teaches a latent print comparison class with emphasis on quality measures and SOPs to meet accreditation standards. The field has come a long way since I first got into it and the training opportunities are now richer and more varied.