Date of Award
Fall 12-2019
Document Type
Thesis
Degree Name
Master of Science (MS)
Department/Program
Forensic Science
Language
English
First Advisor or Mentor
Mechthild Prinz
Second Reader
Nicholas D.K. Petraco
Third Advisor
Adele Mitchell
Abstract
Complex DNA mixtures can be very probative evidence, but comparisons to a person of interest can be affected by allelic drop-out and uncertainty regarding the number of individuals having contributed DNA to a sample. Scientific organizations such as the International Society of Forensic Genetics (Gill et al., 2006) recommend that likelihood ratios should be used to provide a statistical weight when a positive association is made between the DNA profile of a person of interest and an evidentiary DNA sample. To this effect the New York City Office of Chief Medical Examiner (OCME) developed a software program, Forensic Statistical Tool (FST), which calculates likelihood ratios for different scenarios taking into account empirically developed drop-out and drop in rates for different types of mixtures. The FST software was used to explore the effect of underestimation of a contributor’s true drop-out rate and effect of the incorrect estimation of the number of contributors on LR calculations. It was found that underestimating the allelic dropout rate for a true contributor almost always led to an either equal or lower LR than when the original dropout rate was used. It was also found that when the number of contributors was misspecified, there was an increase or decrease in LR values for true contributors. Variation of resulting LRs was higher for more complex mixtures. Finally, LRs for comparisons to individuals, whose DNA was known to not be present in the test mixtures, were lower when using the lower drop-out rates than when using the true drop-out rates.
Recommended Citation
Lucero, Felicia M., "DNA Mixture statistics using a likelihood ratio software tool: effect of variations in drop-out rates and number of contributors" (2019). CUNY Academic Works.
https://academicworks.cuny.edu/jj_etds/137