Explosive Detection Canines in the Field: A Multi-Site Black Box Validation Study

Michelle Karpinsky¹Haylie Browning²Adele Quigley-McBride³Paul Bunker⁴Will Chapman⁵Paola A Prada-Tiedemann^2*Lauryn E DeGreeff^1*

• ¹Florida International University Department of Chemistry and Biochemistry, Miami, United States
• ²Texas Tech University Department of Environmental Toxicology, Lubbock, United States
• ³Simon Fraser University Department of Psychology, Burnaby, Canada
• ⁴Chiron K9, Somerset, United States
• ⁵Noblis, Reston, United States

In 2009, the National Research Council called upon the forensic science community to standardize the best practices and guidelines in the collection and analysis of evidence with the goal of ensuring quality and consistency within the field. In response to this need, the Organization of Scientific Area Committees for Forensic Science (OSAC) was established to coordinate the development of best practices and standards in the forensic sciences. The OSAC Dogs and Sensors subcommittee was part of this initiative focusing on standardizing training and certification protocols for canine detection teams. Though efforts to create and promote such standards are ongoing worldwide, the developed assessments for both training and operational contexts have yet to be empirically validated. As a first step toward addressing this gap, a proof-of-concept black box study was carried out to assess the OSAC explosive canine detection standard based on performance of explosive detection canines. The evaluations were held in three separate geographic locations with a total of 56 canine/handler teams, took place over two days, and included searches recommended within the ANSI/ASB Standard 092 as well as scenarios designed to more closely mimic what the teams might experience in practice. Overall, the results from the individual canine/handler team responses revealed that no team would have passed the OSAC certification; however, the results indicated comparable performance on both assessment types (standard assessments and operational scenarios), suggesting that the performance on Standard 092 certification assessments may predict operational effectiveness. Additionally, canine/handler performance varied significantly across all three trials in both correct alert and false alert rates. Furthermore, detection success rate also varied across the mandatory six different explosive types presented, suggesting a variation in performance attributable to the diversity of training aid material routinely available to the participating teams.