Applying posterior probability informed thresholds to traditional cranial trait sex estimation methods.

Area of Science:

• Forensic Anthropology
• Bioarchaeology
• Biological Anthropology

Background:

• Traditional cranial nonmetric trait analysis for sex estimation often lacks standardized statistical confidence reporting.
• Subjective qualitative assessments of confidence in sex estimation are common but not uniformly based on statistical data.

Purpose of the Study:

• To evaluate the impact of posterior probability-informed thresholds (PPITs) on sex estimation accuracy.
• To assess the trade-off between sample inclusion and accuracy with PPITs.
• To provide recommendations for using and interpreting PPITs in forensic casework.

Main Methods:

• Utilized a sample of 292 U.S. Black and White individuals (females and males).
• Examined two methods employing nonmetric cranial traits to infer sex.
• Applied PPITs to standardize decision-making in sex estimation.

Main Results:

• PPITs of 0.85 or higher increased accuracy to over 92% while maintaining high sample inclusion.
• PPITs below 0.75 did not yield accuracy significantly above chance.
• An accuracy rate of 76% was observed for the 0.75-0.84 PPIT interval, statistically significant but lower than higher thresholds.

Conclusions:

• Implementing PPITs of at least 0.85 is recommended for reliable sex estimation from cranial nonmetric traits.
• Cases with posterior probabilities below 0.75 should be classified as "indeterminate" in sex estimation.
• The 0.75-0.84 PPIT range offers acceptable accuracy but with lower confidence compared to higher thresholds.