Bloodstains are commonly encountered at scenes of violent crime. Forensic practitioners known as bloodstain pattern analysts seek to answer questions about the manner and sequence of events of a crime by examining the bloodstains left behind. Their methods are largely based on subjective expert opinion. A recent paper by NIJ-funded researchers reports the error rates observed when bloodstain pattern analysts examine bloodstains that mimic casework.[1] The error rates from so called “black box” studies like this are critical for the courts to understand the limits of the testimony given by forensic experts.
In one of the first large error-rate studies of the discipline, this project, led by Austin Hicklin at Noblis, Inc., recruited 75 practicing bloodstain pattern analysts. Participants were asked to respond to classification prompts or questions about bloodstains that had been produced under known conditions or taken from casework. The researchers then examined how often the analysts’ conclusions matched the known cause and how often they didn’t. They found that, on average, about 11% of the time the analysts’ conclusions were wrong. And while the average group response as a whole — the consensus — was seldom wrong, any two analysts’ conclusions contradicted each other at an overall rate of about 8%. However, when focusing specifically on erroneous responses, those same errors were reproduced by a second analyst from 18% to 34% of the time. Those rates are particularly concerning because technical review of operational casework by a second analyst is a common method intended to prevent errors.
Bloodstain pattern analysis has been under increased scrutiny since the landmark 2009 National Academies report Strengthening Forensic Science in the United States: A Path Forward, which criticized this and other forensic disciplines on issues of accuracy, reliability, and validity.[2] A more recent report from the President’s Council of Advisors on Science and Technology (PCAST),[3] while not specifically examining bloodstain pattern analysis, broadly recommended well-designed error rate studies of forensic analyst performance, as well as the development of objective, quantitative methods of analysis.
While the error rates found in this study are higher than for other forensic disciplines like latent fingerprints or firearm identification, they are in line with previous bloodstain pattern studies.[note 4] The researchers note the lack of uniformly applied terminology and classification standards for bloodstain pattern analysis as potentially contributing to the high variability in analysts’ responses. They go on to make recommendations for the development and implementation of bloodstain pattern analysis methodology and terminology standards for the field to move forward on a common foundation.
NIJ continues to support error rate studies like this for bloodstain pattern analysis and other forensic disciplines (e.g., latent prints, footwear, firearms, tire tread). Additionally, NIJ maintains a research portfolio applying experimental and computational fluid dynamics to understand the physical processes involved in bloodstain pattern formation. That research may lead to the development of objective tools to support bloodstain pattern analysts.
About This Article
The research in this article was funded by National Institute of Justice award 2018-DU-BX-0214, awarded to Noblis, Inc. This article is based on “Accuracy and reproducibility of conclusions by forensic bloodstain pattern analysts,” published in Forensic Science International, August 2021.
Browse all projects in NIJ’s Bloodstain Pattern Analysis portfolio.