Uncertainty about the true source. A note on the likelihood ratio at the activity level

Area of Science:

• Forensic Science
• Probability Theory
• Statistical Analysis

Background:

• Evaluating fiber evidence in forensic cases is crucial.
• Uncertainty often exists regarding the reference fiber sample's authenticity.
• Existing methods may not fully address this uncertainty.

Purpose of the Study:

• To propose a novel likelihood ratio approach for fiber evidence evaluation.
• To address uncertainty about whether the reference fiber is the one worn during the offense.
• To provide a statistically robust framework for forensic casework.

Main Methods:

• Developed a likelihood ratio (LR) based evaluation framework.
• Incorporated additional parameters with categorical assumptions.
• Analyzed the properties of the proposed LR approach via sensitivity analyses.

Main Results:

• The proposed LR approach converges to existing formulas under specific assumptions.
• Sensitivity analyses revealed the behavior of the LR under varying conditions.
• Demonstrated the practical implications of the approach in forensic arguments.

Conclusions:

• The new likelihood ratio method offers a refined approach to evaluating fiber evidence.
• The method provides a framework for handling uncertainty in reference samples.
• This research contributes to more rigorous statistical interpretation in forensic science.