Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A two-level model for evidence evaluation.

Colin G G Aitken1, Grzegorz Zadora, David Lucy

  • 1School of Mathematics and The Joseph Bell Centre for Forensic Statistics and Legal Reasoning, The King's Buildings, The University of Edinburgh, Mayfield Road, Edinburgh EH9 3JZ. c.g.g.aitken@ed.ac.uk

Journal of Forensic Sciences
|February 24, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Elemental profiling of processed foods using ICP-MS and chemometric evaluation of elemental patterns.

Food chemistry·2026
Same author

Glass microtraces originated from mobile phones - Interpretation of their elemental composition for forensic purposes.

Forensic science international·2025
Same author

The impact of time and storage conditions on samples containing OC pepper sprays on the quantitative ratios of capsaicinoids.

Forensic science international·2024
Same author

Determination of time since deposition of bloodstains through NIR and UV-Vis spectroscopy - A critical comparison.

Talanta·2024
Same author

Breaking with trends in forensic dating: A likelihood ratio-based comparison approach.

Forensic science international·2023
Same author

Differentiation of oleoresin capsicum sprays based on their capsaicinoid profiles.

Forensic science international·2021
Same journal

A novel relationship between time offsets in capillary electrophoresis and DNA sequence variations in short tandem repeats.

Journal of forensic sciences·2026
Same journal

A 4-zone model to determine fentanyl overdose probability.

Journal of forensic sciences·2026
Same journal

Determining the utility of radio frequency identification (RFID) technology for disaster victim identification (DVI).

Journal of forensic sciences·2026
Same journal

Stakeholders' perspectives on integrating point-of-care diagnostics into forensic death investigations in South Africa.

Journal of forensic sciences·2026
Same journal

Vehicle-assisted suicidal strangulation: A case report and literature review.

Journal of forensic sciences·2026
Same journal

A feasibility study of chromatographic methods with mass spectrometry, flame ionization detector (FID), and micro electron capture detector (μECD) for laboratory adoption of organic gunshot residue analysis at trace levels.

Journal of forensic sciences·2026
See all related articles

This study introduces a novel random effects model for comparing multivariate observations, crucial for forensic science. The method accurately determines if samples share a common origin, improving evidence analysis in casework.

Area of Science:

  • Forensic Science
  • Statistical Modeling
  • Materials Analysis

Background:

  • Comparing sets of replicated multivariate continuous observations is challenging when their common origin is unknown.
  • Existing methods may struggle with the complexity of hierarchical data and multivariate dependencies.

Purpose of the Study:

  • To develop and validate a statistical model for comparing multivariate observations from potentially common origins.
  • To improve the accuracy of source attribution in forensic casework, specifically using glass fragment analysis.

Main Methods:

  • Applied a two-level hierarchical random effects model to calculate a likelihood ratio for comparing sample sets.
  • Modeled within-group distributions with a Normal distribution and between-group distributions using kernel density estimation.

Related Experiment Videos

  • Utilized a graphical method to reduce the multivariate problem by estimating variable dependency structures.

    • Main Results:

    • The model demonstrated strong performance on a large glass fragment database (200 objects, 4 fragments each, 8 elements).
    • Achieved a 15.2% false-positive rate and a 5.5% false-negative rate, outperforming previous approaches.
    • Successfully applied to casework examples comparing crime scene glass with suspect-associated glass.

    Conclusions:

    • The developed random effects model provides a robust solution for comparing multivariate observations in forensic contexts.
    • The method enhances the reliability of source attribution, particularly for trace evidence like glass fragments.
    • This approach offers a statistically sound framework for evaluating evidence in criminal investigations.