A murder in Boston.

Area of Science:

• Forensic Science
• Statistics
• Genetics

Background:

• A 2009 conviction for a 1990 murder case involving defendant D and victim C is under review.
• DNA evidence found under the victim's fingernails was attributed solely to D and C.

Observation:

• The interpretation of this DNA evidence at the activity level was debated during trial and in subsequent reviews.
• An alternate hypothesis suggested DNA transfer occurred during a social meeting 2-4 weeks prior to the crime.

Findings:

• A Bayesian network (BN) was employed to evaluate the DNA evidence under two propositions: D was the attacker (Hp) versus D was not the attacker (Ha).
• The BN analysis yielded a Likelihood Ratio (LR) of approximately 800, favoring the proposition that D was the attacker.
• This suggests the DNA evidence, despite potential informality in original testimony, was not misleading to the jury.

Implications:

• Bayesian network analysis provides a quantitative framework for evaluating forensic evidence in legal contexts.
• This methodology can help clarify the probative value of DNA evidence, especially when activity level interpretations are contested.
• The findings support the original conviction by demonstrating the DNA evidence is highly consistent with D being the perpetrator.