Summary
Highlights
Dan Krane introduces the topic of subjectivity in DNA testing interpretation, explaining that while some DNA profiles are clear, others are ambiguous and can lead to biased conclusions. He references online resources at bioforensics.com where viewers can access presentation slides and additional materials.
Krane presents an actual evidence sample from a California sexual assault case. The DNA profile is incomplete, showing only blue electropherogram data, and none of the peaks match the victim, suggesting foreign DNA. The focus is on analyzing this ambiguous sample with different potential suspect profiles.
The first suspect, Tom, has a 17-17 D3 locus. The evidence sample shows a 17 and a 12. Analysts, knowing Tom's profile, might dismiss the 12 as 'noise' or a technical artifact due to its low height and unusual shape, thereby including Tom as a possible contributor because his 17 matches. Similar justifications are made for other loci (VWA and FGA) despite discrepancies.
When considering suspect Dick, who has a 12-17 D3 locus, analysts might now interpret the 12 in the evidence sample as real, attributing its small size and odd shape to the trace nature of the sample. Additionally, a 'blob' in the FGA locus of the evidence sample (where Dick has a 20 allele) might be used to mask Dick's 20, leading to his inclusion as well. Dick, despite the ambiguous evidence, was ultimately convicted in the real case.
Suspect Harry has a 14-17 D3 locus. The 14 allele is not clearly visible in the evidence sample, presenting a new challenge: allelic dropout. Analysts might argue that due to the small quantity of DNA, Harry's 14 allele simply wasn't detected, allowing for his inclusion as a possible contributor.
The concept of a DNA mixture is introduced with hypothetical suspect Sally. If Sally (15-15 at VWA) is considered alongside a 17 allele in the evidence sample, the explanation could be that the sample is a mixture from two individuals. This scenario further broadens the criteria for inclusion, showing how multiple individuals with different profiles can all be deemed non-excludable.
Krane explains that the ability to include multiple individuals with differing DNA profiles from the same evidence highlights the issue of 'observer effects' or 'context effects'. These psychological biases cause analysts to interpret ambiguous data in a way that aligns with prior expectations or theories, often influenced by external information about the suspect.
Two case file excerpts illustrate how forensic analysts are often privy to extraneous, biasing information about suspects (e.g., criminal history, gang affiliation, victim's fear, death penalty case status). This information, while irrelevant to the scientific analysis, can subconsciously motivate analysts to resolve ambiguities in a way that supports a particular conclusion.
Krane proposes 'sequential unmasking' as a solution to combat examiner bias. This method involves analysts interpreting evidence samples blindly, without any prior knowledge of suspect or victim reference profiles. Only after a preliminary interpretation is made is the biasing information revealed, allowing for a transparent revision process.
Blind testing forces analysts to be more conservative and objective in their interpretations. While some crime labs, like the Virginia Department of Forensic Science, have adopted this approach, others resist, claiming that knowing the suspect's profile helps them reach the 'right' answer. Krane argues that this defeats the purpose of objective scientific analysis and extends the potential benefits of blind interpretation to all forensic sciences.