Robust Trust in Expert Testimony

TLDR

The criminal‑trial standard of proof demands robust evidence, meaning it must be unlikely that additional information would alter the probability of guilt, a requirement that judges find hard to assess. This article proposes a method for evaluating the robustness of expert testimony. The method assesses robustness across competence, motivation, external strength, internal strength, and relevance. Illustrating the danger of non‑robust testimony, the authors analyze the Thomas Quick case, where a mentally ill patient was wrongfully convicted of eight murders.

Abstract

The standard of proof in criminal trials should require that the evidence presented by the prosecution is robust. This requirement of robustness says that it must be unlikely that additional information would change the probability that the defendant is guilty. Robustness is difficult for a judge to estimate, as it requires the judge to assess the possible effect of information that the he or she does not have. This article is concerned with expert witnesses and proposes a method for reviewing the robustness of expert testimony. According to the proposed method, the robustness of expert testimony is estimated with regard to competence, motivation, external strength, internal strength and relevance. The danger of trusting non-robust expert testimony is illustrated with an analysis of the Thomas Quick Case, a Swedish legal scandal where a patient at a mental institution was wrongfully convicted for eight murders.