NAS report: Many forensic disciplines prone to error

The National Academy of Sciences has published its long-awaited report critiquing forensic science titled "Strengthening forensic science in the United States: A path forward." Go here for a preview and ordering information. According to the accompanying press release:

Rigorous and mandatory certification programs for forensic scientists are currently lacking, the report says, as are strong standards and protocols for analyzing and reporting on evidence. And there is a dearth of peer-reviewed, published studies establishing the scientific bases and reliability of many forensic methods. Moreover, many forensic science labs are underfunded, understaffed, and have no effective oversight.

Forensic evidence is often offered in criminal prosecutions and civil litigation to support conclusions about individualization -- in other words, to "match" a piece of evidence to a particular person, weapon, or other source. But with the exception of nuclear DNA analysis, the report says, no forensic method has been rigorously shown able to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual or source.

What's more:

there has been little rigorous research to investigate how accurately and reliably many forensic science disciplines can do what they purport to be able to do. In terms of a scientific basis, the disciplines based on biological or chemical analysis, such as toxicology and fiber analysis, generally hold an edge over fields based on subjective interpretation by experts, such as fingerprint and toolmark analysis. And there are variations within the latter group; for example, there is more available research and protocols for fingerprint analysis than for bitemarks.

Nuclear DNA analysis enjoys a pre-eminent position not only because the chances of a false positive are minuscule, but also because the likelihood of such errors is quantifiable, the report notes. Studies have been conducted on the amount of genetic variation among individuals, so an examiner can state in numerical terms the chances that a declared match is wrong. In contrast, for many other forensic disciplines -- such as fingerprint and toolmark analysis -- no studies have been conducted of large populations to determine how many sources might share the same or similar features. For every forensic science method, results should indicate the level of uncertainty in the measurements made, and studies should be conducted that enable these values to be estimated, the report says.

Not particularly cheap, at $36.66 plus shipping, nor, one imagines, will it be a light read, but I've already ordered my copy.

Related Grits posts:

• USDOJ should embrace, not fight forensics based science
• US justice system commonly relies on shoddy forensic 'science'