An associate law professor at UBC says evidence based on sniffer dogs, firearms analysis and bloodstain pattern work can be less reliable than courts assume.
The Story
An associate professor of law at the University of British Columbia says Canadian courts and lawyers often treat several forensic and investigative techniques as reliable without enough empirical support, despite concerns about scientific validity and bias.
Sara Gordon, whose findings are based on a series of papers, examined how courts handle evidence tied to sniffer dogs, ballistics analysis and blood spatter patterns. She said courtrooms tend to “just assume” that the techniques are reliable when police officers testify to their results.
Gordon said many forensic methods were not originally developed in university or laboratory settings but in police departments, and that only in the last 10 to 20 years has their validity been tested empirically. She said, in many cases, the evidence does not support the types of conclusions examiners are expected to reach.
Gordon pointed to potential bias in all three areas.
In sniffer-dog work, she said bias can occur when officers are sufficiently confident a person has narcotics that they may unintentionally signal it to the dog, including by unconsciously changing their behaviour as they pass.
In ballistics and bloodstain pattern analysis, Gordon said contextual bias is a risk because many examiners are police officers who have access to information about the crime scene that can shape interpretation. She said an examiner who knows details such as who was involved or the relationship between the victim and accused may approach observations with expectations about what they should find.
The professor said major U.S. reports about a decade ago highlighted the lack of scientific support for several forensics techniques used across North America, which led to further research. She said courts have largely not caught up to the science and that her papers are meant to help judges and attorneys make informed decisions about whether evidence should be admitted.
In discussing sniffer dogs specifically, Gordon noted that the Supreme Court of Canada last reviewed the constitutionality of dog searches in 2013, when it found police could conduct such searches with less cause than for other types because they are minimally intrusive and accurate.
Gordon said subsequent studies across jurisdictions have found sniffer dogs to be less reliable than believed. She cited findings from one of her papers in the Alberta Law Review saying successful dog alerts leading to illicit drugs are often “no better than chance.” She also referenced an ombudsman’s review in New South Wales that examined more than 10,000 alerts by 17 drug detection dogs and found false alerts occurred 74 per cent of the time.
Gordon also said there are no national standards in Canada for the training, testing and deployment of police sniffer dogs, nor meaningful transparency about how those functions are carried out across the country.
She added that dog searches can be embarrassing, humiliating and scary, particularly for racialized people and those subject to enhanced police scrutiny, and argued the same reasonable and probable grounds standard should apply to dog searches to comply with the Canadian Charter of Rights and Freedoms.
Forensic firearms testimony has also drawn Gordon’s scrutiny. She said Canadian courts have admitted ballistics evidence for nearly a century based primarily on examiners’ testimony without meaningful scrutiny of whether conclusions are limited to what is empirically supported.
Gordon said she could find only one Canadian case where a judge refused to admit forensic firearms testimony due to concerns over scientific validity. She said the case was heard in Ontario five years earlier and involved testimony from a firearms examiner at the province’s Centre of Forensic Sciences, including a discussion of the absence of required specific “agreements” or matching microscopic markings and concerns including a lack of proficiency testing.
Gordon said validation studies ballistics examiners often rely on can have methodological flaws, including unrepresentative samples and testing conditions that do not match real working conditions. She also said examiners may present conclusions in “highly certain terms” without disclosing limitations, including testimony she said can overstate error rates.
On bloodstain pattern analysis, Gordon said recent studies show high error rates, citing levels up to 32 per cent in one of her upcoming papers for the Canadian Bar Review.
She said courts have not applied the scrutiny required for novel or contested scientific evidence and have instead focused on examiners’ training and experience, “taking it for granted that the field is sufficiently reliable.”
Gordon said bloodstain pattern work is based on fluid dynamics and involves complex physics, but that some police training courses may be short and many officers lack scientific training to apply those principles. She said the technique can become focused on reconstructing crime scenes in extreme detail, which she said the science does not support and that it is vulnerable to contextual bias.
She said crime scene reconstruction based on bloodstain pattern analysis should generally be inadmissible, and that a forensic technique with error rates exceeding 30 per cent should not be relied upon to establish guilt beyond a reasonable doubt.
← More stories