Sex estimation using continuous variables: Problems and principles of sex classification in the zone of uncertainty
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23330%2F20%3A43960625" target="_blank" >RIV/49777513:23330/20:43960625 - isvavai.cz</a>
Result on the web
<a href="https://www.elsevier.com/books/statistics-and-probability-in-forensic-anthropology/obertova/978-0-12-815764-0" target="_blank" >https://www.elsevier.com/books/statistics-and-probability-in-forensic-anthropology/obertova/978-0-12-815764-0</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1016/B978-0-12-815764-0.00016-2" target="_blank" >10.1016/B978-0-12-815764-0.00016-2</a>
Alternative languages
Result language
angličtina
Original language name
Sex estimation using continuous variables: Problems and principles of sex classification in the zone of uncertainty
Original language description
Discriminant function analysis using continuous bone measurements provides a highly efficient tool for sex estimation in forensic practice. We show that three aspects are especially important for a reliable application of DFA in sex estimation: (1) the accuracy of sex classification should be calculated by using cross-validation, as non-cross-validated accuracies usually overestimate the true classification rate. Given that an accuracy greater than or equal to 80% is part of the Daubert criteria for admissibility, an unbiased estimate of the classification rate is of high relevance for forensic practice. (2) The sex of an unknown individual should be estimated by a population-specific discriminant function that was derived from the population to which the individual is thought to belong. The application of a discriminant function based on an unrelated population may result in misclassifications and may produce a large sex bias, that is, large differences between the classification rates for females and males. (3) Sex estimation in the zone of uncertainty, that is, the overlapping area between the sexes, should be avoided to prevent misclassifications. Sex should be assigned only to those individuals with a posterior probability of being female or male higher than 0.95. Although such an approach limits the practical applicability of DFs, because some portion of the individuals remains unclassified, it allows a high classification accuracy to be maintained at the individual level. It is worth considering extending the Daubert criteria by adding a posterior probability threshold to the sex estimation procedure, as a reliable sex estimation of every single individual, and not only the overall accuracy across the entire sample, is of utmost importance in forensic practice.
Czech name
—
Czech description
—
Classification
Type
C - Chapter in a specialist book
CEP classification
—
OECD FORD branch
10103 - Statistics and probability
Result continuities
Project
<a href="/en/project/EE2.3.30.0038" target="_blank" >EE2.3.30.0038: New excellence in human resources</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2020
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Book/collection name
Statistics and Probability in Forensic Anthropology
ISBN
978-0-12-815764-0
Number of pages of the result
28
Pages from-to
155-182
Number of pages of the book
418
Publisher name
Academic Press
Place of publication
Amsterdam
UT code for WoS chapter
—