Questioning the Evidence

March 1, 2016


BUFFALO, NY - Forensic science is a vital crime-fighting tool in today’s criminal justice system. But it can also lead to false convictions, according to H. David Sheets, PhD. Landmark research by the Canisius physics professor proves that bite-mark analysis is “far from an exact science.”

Bite-mark analysis compares the teeth of crime suspects to bite mark patterns on victims. Historically, forensic odontologists (dentists who provide forensic dental identifications in criminal investigations and mass disasters) operate under two general guidelines when interpreting bite-mark evidence: That everyone’s dental impression is unique, “similar to fingerprints,” and that human skin reliably records an individual’s dental impression.

But these notions “have never been validated scientifically,” says Sheets.

He and his colleagues, Mary A. Bush, DDS and Peter J. Bush, from the University at Buffalo School of Dental Medicine, sought to do just that.

They examined more than 1,000 human dentitions and studied hundreds of bite marks in cadaver skin. With the help of computer analysis and applied statistics, the team then worked to match its database of bite marks to the correct dental impressions.

“When the dental alignments were similar, it was difficult to distinguish exactly which set of teeth made which bites,” Sheets says. “That tells us that a single bite mark is not distinct enough to be linked to a specific individual. It can actually point to many different individuals.”

The Canisius-UB study is among the largest conducted on bite-mark analysis and the first to use human-skin models (rather than animal models, wax or Styrofoam). More notable is that its findings call into question criminal convictions that rested entirely on bite-mark evidence.

Since 2000, at least 25 people convicted on bite-mark evidence have been exonerated due to advances in DNA testing, according to the Innocence Project. The national organization, which works to free wrongfully convicted individuals, is now using the Canisius-UB study to help build a case for having bite-mark evidence cast out of criminal proceedings.

“This is an example of where science can help prevent future wrongful convictions and perhaps even provide some social justice to those already convicted,” Sheets concludes.

The National Institute of Standards and Technology (NIST) and the National Institute of Justice (NIJ), which provide oversight to the federal forensic community, recently appointed H. David Sheets, PhD, to its expert working group on human factors in handwriting analysis. Sheets is charged with identifying the human factors that affect the outcomes of forensic analyses and developing best practices that reduce the likelihood of future errors.