Study Finds Bloodstain Bias

by Joe Palazzolo, The Wall Street Journal (Aug. 1, 2014)

This short article summarizes the findings of a 2014 study. (The study itself is included as optional reading.)

Bloodstain pattern analysis dates back to the 1800s, but only recently has it been subject to rigorous validation.

"At this time very little is known about this beyond the instincts of experienced instructors and investigators who have observed the reproducibility of bloodstain patterns over many crime scenes and practical sessions in the classroom," begins a study on BPA (forgive the acronym) funded by the U.S. Department of Justice and released by the agency late Thursday.

DNA analysis is now used to identify who bled at a crime scene, but BPA is still a vital tool. For example, it's common for a suspect to claim that he was stained by a victim's blood while trying to render aid. In that instance, using BPA to figure out how the blood got on him is more important than determining whose blood it is.

The study, by researchers in the U.S. and New Zealand, focused on how context influences bloodstain pattern analysts. They asked 27 experts from Australia, Europe and North America to look at various stains and identify them using terminology from the Working Group on Bloodstain Pattern Analysis, the international group that promotes best practices.

The researchers engineered the stains, on rigid and fabric surfaces, in imaginative (and morbid) ways.

Blunt force: produced by the impact of a hammer on a pool of blood containing 1 to 6 drops of blood.
So-called "cast-off patterns": created from the swinging of a blood-covered wrench or a small bloodied knife, swung towards and across the target surfaces.
"Expirated patterns": created by an experimenter blowing air gently through blood-covered lips or by coughing a small volume of blood from the mouth onto nearby vertical target surfaces.
Firearms-related bloodstain patterns: created by shooting a .22 caliber bullet though a blood soaked sponge.

The contextual information was presented in the form of a short vignette that provided background information about how the bloodstain pattern was found and what was known about the case. Each scenario supplied information that either supported the correct answer, was neutral or pointed to the wrong answer.

The researchers found significant bias. Misclassification was much lower (8% for rigid surfaces, 14% for fabric surfaces) when the vignette supported the right answer than when it was neutral (11% for rigid surfaces, 26% for fabric surfaces). When the background information pointed toward the wrong classification, the error rate was higher still (20% for rigid surfaces, 30% for fabric surfaces).

That's bad, because analysts are often exposed to case information such as medical findings and even witness testimony before they classify blood stains. The study proves that such information influences them. The researchers suggest police departments take greater care to wall off bloodstain pattern analysts, at least while they are engaged in pattern recognition, so they can operate as close to bias-free as possible.

This book, and all H2O books, are Creative Commons licensed for sharing and re-use with the exception of certain excerpts. Any excerpts from the Restatements of the Law, Principles of the Law, and the Model Penal Code are copyright by The American Law Institute. Excerpts are reproduced with permission, not as part of a Creative Commons license.