UBC scholar Sara Gordon argues courts overrate sniffer dogs, bloodstain analysis and bullet matching

The three papers are legal scholarship. Gordon reviewed the scientific evidence on each technique and compared it with how North American courts admit it. The figures she cites come from the research she reviewed ^[1][4][11]. "Just because a technique has been used in court for decades doesn't mean we should take its reliability for granted, especially when someone's liberty is at stake in a criminal trial," Gordon said ^[3]. She is an associate professor at the Peter A. Allard School of Law at UBC ^[2]. "My hope is that this research will help judges and lawyers better understand the limitations of these techniques," she said ^[20].

The Australian dog figure is the easiest to misread. Turned around, it means roughly one alert in four led to drugs ^[21]. The count runs over alerts, meaning the people a dog picked out. Anyone carrying drugs who walked past without an alert falls outside it. Residual odour, one of the causes Gordon lists ^[14], is hard to score: the dog may be reacting to a real trace of drugs that are no longer there. The thing this doesn't tell you is how the misses divide between the dog's nose, its training and the handler.

In Canada, accuracy is part of the legal rule itself. The Supreme Court gave two other reasons as well: that dog searches are minimally intrusive and narrowly targeted ^[10]. Both are empirical claims too. Gordon's review of research published since the court's latest rulings in 2013 points the other way ^[11]. "They're not nearly as reliable as many courts seem to assume. And in fact, they can be quite invasive," she said ^[12]. Her second sentence answers the court's intrusiveness reasoning directly. I think the dog argument is the strongest of the three, on one condition. The headline field number is Australian ^[13], and the argument holds best if dogs and handlers in Canada perform similarly.

Her proposed reforms include clearer standards for training, certification and performance monitoring of the dogs ^[16]. A scientist would ask for performance monitoring first. It produces a running count of alerts and finds for each animal, so a judge could ask about the specific dog in the case.

The ballistics argument is about calibration. Canadian courts have accepted the evidence for more than a century, and Gordon found that the research does not support the confidence with which examiners often testify ^[5]. The phys.org account does not describe the practical testing behind the 37% misidentification figure ^[6]. From that account alone, a reader cannot check how many comparisons were made or what counted as a miss. Her firearms paper concentrates on Canada, where the evidence is still routinely admitted ^[7].

The bloodstain paper argues from the field's history. Canadian courts have long treated bloodstain pattern analysis as a physical science ^[17]. Gordon traces the discipline to a chemist experimenting in his basement. He later helped establish its scientific journal and early textbooks, which gave the discipline the institutional markers of a scientific field ^[18].