The Innocence Project analyzes every DNA exoneration to determine what factors contributed to the wrongful conviction and how the criminal justice system can be improved in the future. Most wrongful convictions involve more than one contributing cause; for example, an eyewitness may have misidentified an innocent person, and in the same case a forensic analyst may have testified that hairs from the crime scene match the defendant’s hair. In the jury’s eyes, the eyewitness testimony is strengthened by the forensic evidence (and vice versa).
The Innocence Project has refined how it defines and documents forensic problems as a factor in wrongful convictions that were overturned with DNA testing. We now track cases where unvalidated or improper forensic science contributed to wrongful convictions. The Innocence Project defines unvalidated or improper forensic science as:
• the use of forensic disciplines or techniques that have not been tested to establish their validity and reliability;
• testimony about forensic evidence that presents inaccurate statistics, gives statements of probability or frequency (whether numerical or non-numerical) in the absence of valid empirical data, interprets non-probative evidence as inculpatory, or concludes/suggests that evidence is uniquely connected to the defendant without empirical data to support such testimony; or
• misconduct, either by fabricating inculpatory data or failing to disclose exculpatory data.
Among all DNA exonerations nationwide, unvalidated or improper forensic science was a factor in approximately 50% of the underlying wrongful convictions. Determinations about which cases meet this definition are made based on trial transcripts and other official sources. In many instances, experts in forensic science and the law were consulted in making these determinations.
Previously, the Innocence Project defined this category of cases more broadly to include limited, unreliable and fraudulent forensic science. Under that definition, cases where validated science was used – or unvalidated science was properly used – were included. For example, cases that involved serology (or blood group testing), even where the serology was conducted properly and the analyst testified accurately, were included because serology is “limited” in what it can show.
By changing the definition to unvalidated or improper forensic science, the Innocence Project is able to look closer at what went wrong in each wrongful case – and able to focus on problems that can and should be prevented in the future.
For more on unvalidated or improper forensic science as a factor in wrongful convictions,
. For a chart with every case the Innocence Projects identifies as involving unvalidated or improper forensic science (and more on our methodology),
. For more on reforms that can prevent wrongful convictions based on unvalidated or improper forensic science,