Is a Conservative Decision in the Lab Always Conservative?

For better or worse, forensic science in the United States has focused on isolating the forensic laboratory from the case. Yet judgements made in a context-blind laboratory can have unexpected, case-specific implications. Here, I'd like to share an example in which a laboratory judgement that was intended to be a conservative, defense-favoring decision had unintended implications to the administration of justice.

In this case, the hands of two suspects had been sampled for primer Gunshot Residue (pGSR). The kit from Suspect A was reported as positive for pGSR, while the kit from Suspect B was reported as negative. As is typical, no detail about the number or type of particles detected in either kit was provided in the report.

It is broadly accepted in the forensic community that pGSR evidence cannot be used to identify the shooter of a firearm. However, it is not a stretch to expect that a prosecutor might use these findings to charge one of the two suspects, that a scientist might be called by the prosecution to testify to these results, and that a juror might be swayed toward the conclusion that Suspect A was more involved in the shooting than Suspect B.

In this case, the defense counsel intervened, and our laboratory was asked to conduct a review. Since a typical GSR report contains only one or two sentences specific to the actual result, a full discovery packet was obtained. A review of the raw data showed, a bit surprisingly, that the same suite of particles was detected on the hands of both Suspect A and Suspect B by the automated analysis routine: a single tricomponent particle (characteristic of pGSR) along with numerous mono- and bi-component particles.

Before going further, a bit of background is provided for those who are not fluent in pGSR analysis. Since a pGSR conclusion may hinge on the presence (or absence) of a single, microscopic tricomponent particle, the composition and shape of each auto-identified tricomponent particle is manually reviewed by the examiner. Based upon its elemental composition and morphology, a given particle may be demoted from or elevated to the tricomponent category. Some of these interpretation criteria are subjective and can differ from laboratory to laboratory and examiner to examiner.

During the manual review, the original analyst confirmed the only tricomponent particle from Suspect A, but rejected the sole tricomponent particle that was autodetected on the hands of Suspect B. The elimination of the only tricomponent particle flipped the report conclusion from positive to negative. Thus, the kit from Suspect A was reported as a positive, and the Suspect B kit was reported as a negative.  

Our review of the composition and morphology of the demoted tricomponent particle from Suspect B disagreed with that of the prosecution's lab. More specifically, we found that the particle fell into an analytical gray area, leaving open the possibility that the demoted particle could have been a tricomponent particle that arose from the discharge of a firearm. The properties relied upon to reject this particle, were not, in fact, settled topics in the GSR literature.

From the perspective of a government crime laboratory, the demotion of a tricomponent particle may be considered a conservative, defense-favoring decision. In many instances, this is the case; however, in the broader context of this case, a negative result that is conservative from the perspective of Suspect B could also serve to further implicate Suspect A. In our opinion, the population of particles present on the hands of both suspects were similar; therefore, the pGSR results were not probative to the question of whether one of the suspects was not involved in the shooting. While I never learned how this case played out, I am confident that the original report did not convey an appropriate message to the criminal justice system.

The nuances of this case lead to some considerations that extend beyond pGSR analysis, to topics of potential interest to the broader forensic community:

1. Case context. Decoupling a forensic scientist from the case circumstances gives some stakeholders a degree of comfort. Yet, detaching a scientist from the context of their samples would be unheard of in any other realm of science. Here, a case scenario that involves a second suspect adds a complexity that was not considered in either the laboratory's SOP or in the preparation of pre-written, boilerplate conclusion statements available for reporting. The case-independent decision, that was intended to err in the favor of the defense, turned out to be neither conservative nor case independent, and the reported pGSR results were likely detrimental to the administration of justice.
    
2. Reporting.  Certainly, a brief and clear summary of forensic results is as important as an abstract in a peer-reviewed journal; however, the minimalistic, boilerplate approach to reporting found in most forensic cases can omit basic data and subjective judgements that may be of critical importance to the evaluation of alternative hypotheses. As forensic science heads increasingly toward more prescriptive methods and stricter reporting requirements, I believe it would be wise to explicitly leave open a pathway that permits the application of less prescriptive, more flexible exploratory approaches and free-form report preparation intended to address case-relevant questions. Standardization, while helpful in certain instances, can also do a disservice to both science and the criminal justice system.