It has been a decade since the US National Research Council produced a damning report on the state of forensic science in the US (and by extension the international community). As such, it is worth asking: has forensic science improved?
The report made a myriad of findings in regard to the forensic science community, including that:
- Criminal laboratories lacked clear standards across jurisdictions.
- Forensic practitioners had inconsistent accreditation requirements across jurisdictions.
- Many forensic techniques raised issues of interpretation. Some methodologies (such as DNA analysis) were capable of expression statistically whilst others (such as blood splatter analysis) were up to expert interpretation.
- Many forensic techniques were under researched with little evidence establishing reliability as well a little understanding of the limits and measures of performance of techniques.
- Criminal laboratories were under resourced and under staffed. This issue gave rise to perceptions that labs were under political sway to ‘produce’ for prosecutors.
The most damning aspect of the report was the suggestion that many forensic techniques currently being used in criminal prosecutions may not be reliable. Some of the biggest causes for concern were comparative techniques subject to contextual bias such as fingerprint comparison, striation analysis (toolmark, ballistics etc), microscopic hair analysis and bitemark analysis.
So, after ten years – has the evidence base improved?
In regard to fingerprint comparison, several studies have been conducted to determine the extent to which cognitive bias can impact analysis. However, many of these studies have suffered from a lack of ‘real world’ scenarios.
Nevertheless, the National Institute of Standards and Technology have developed a variety of protocols and standards which limit the likelihood of human error. Moreover, technology is greatly improving the ability to conduct fingerprint comparisons without expert interpretation.
Striation analysis remains more of an art than a science. There have been a few attempts to understand contextual bias in striation analysis (mostly in regard to ballistics). There have also been attempts to express comparisons statistically. Nevertheless, more research is needed in this area.
Comparative hair analysis (as opposed to toxicological hair analysis) remains under-researched and poorly understood. My attempt at finding any comprehensive evaluation of the technique turned up no results. However, this may be due to the fact that the increasing effectiveness of DNA analysis has rendered the technique largely irrelevant.
The real failure in the years following the report is bite mark analysis, with study after study after study showing the unreliability of the technique for forensic identification. This includes very obvious errors in the techniques used, inconsistencies in the approach of practitioners and a very real risk of contextual bias.
So, has forensic science improved?
Somewhat. It is still incredibly worrying that all of the techniques above remain admissible in most criminal jurisdictions globally, without sufficient legal scrutiny. This is particularly worrying in the cases of bite mark analysis, which has been identified as a possible source of wrongful convictions.
The standards and regulation of forensic labs has also recently been raised as a cause for a concern. Last year, a UK forensic toxicology lab was a cause of incredible controversy due to alleged data manipulation by employees.
If the forensic community wishes to ensure its legitimacy within the legal system there must be a consistent and very high standard for the evidence base of forensic techniques used.