April 13th, 2011
In the past few years national attention has increasingly focused on the fallibility of forensic science. Forensic disciplines once touted as scientifically accurate such as bite mark analysis have not been subject to peer review or scientific testing. In addition, bias among law enforcement and crime labs in favor of convictions too often leads to misuse of forensics.
Bias can be caused by medical examiners or crime labs reporting directly to the attorney general or other state law enforcement official and influences the entire process of collecting and analyzing scientific evidence. A 2002 study by Michael Risinger, a law professor at Seton Hall, identified five different parts of scientific analysis susceptible to influence by unintentional bias. They are: how the analyst observes the initial data, how he records the data, how he makes calculations, and how he remembers and reinterprets his notes when preparing for trial. For example, researchers at the University of Southampton in the UK published a study in 2006 that found “the error rate of fingerprint analysts doubled when they were told the details of the case they were analyzing”. Balko makes recommendations for how to improve the system by creating different incentives, namely rewarding analysts for doing accurate work and penalizing them for errors.
Read more here
Balko, Radley. “Getting Forensics Right”. March 14, 2011