Time filter

Source Type

Port Adelaide, Australia

Taylor D.,Forensic Science South Australia | Taylor D.,Flinders University | Buckleton J.,ESR
Forensic Science International: Genetics | Year: 2015

A set of low template mixed DNA profiles with known ground truths was examined using software that utilised peak heights (STRmix™ V2.3) and an adapted version that did not use peak heights and mimicked models based on a drop-out probability [1,2] (known as semi-continuous or 'drop' models) (STRmix™ lite). The use of peak heights increased the LR when Hp was true in the vast majority of cases. The effect was most notable at moderate template levels but was also present at quite low template levels. There is no level at which we can say that height information is totally uninformative. Even at the lowest levels the bulk of the data show some improvement from the inclusion of peak height information. © 2014 Elsevier Ireland Ltd. All rights reserved. Source

Taylor D.,Forensic Science South Australia | Bright J.-A.,ESR Ltd. | Buckleton J.,ESR Ltd. | Curran J.,University of Auckland
Forensic Science International: Genetics | Year: 2014

A typical assessment of the strength of forensic DNA evidence is based on a population genetic model and estimated allele frequencies determined from a population database. Some experts provide a confidence or credible interval which takes into account the sampling variation inherent in deriving these estimates from only a sample of a total population. This interval is given in conjunction with the statistic of interest, be it a likelihood ratio (LR), match probability, or cumulative probability of inclusion. Bayesian methods of addressing database sampling variation produce a distribution for the statistic from which the bound(s) of the desired interval can be determined. Population database sampling uncertainty represents only one of the sources of uncertainty that affects estimation of the strength of DNA evidence. There are other uncertainties which can potentially have a much larger effect on the statistic such as, those inherent in the value of Fst, the weights given to genotype combinations in a continuous interpretation model, and the composition of the relevant population. In this paper we model the effect of each of these sources of uncertainty on a likelihood ratio (LR) calculation and demonstrate how changes in the distribution of these parameters affect the reported value. In addition, we illustrate the impact the different approaches of accounting for sampling uncertainties has on the LR for a four person mixture. Source

Stephan C.N.,POW Inc | Stephan C.N.,University of Queensland | Simpson E.K.,Forensic Science South Australia | Byrd J.E.,POW Inc
Journal of Forensic Sciences | Year: 2013

Several methods that have customarily been used in craniofacial identification to describe facial soft tissue depths (FSTDs) implore improvement. They include the calculation of arithmetic means for skewed data, omission of concern for measurement uncertainty, oversight of effect size, and misuse of statistical significance tests (e.g., p-values for strength of association). This paper redresses these limitations using FSTDs from 10 prior studies (N = 516). Measurement uncertainty was large (>20% of the FSTD), skewness (≥0.8) existed at 11 of the 23 FSTD landmarks examined, and sex and age each explained <4% of the total FSTD variance (η2 calculated as part of MANOVA). These results call for a new and improved conceptualization of FSTDs, which is attained by the replacement of arithmetic means with shorths and 75-shormaxes. The outcomes of this implementation are dramatic reduction in FSTD complexity; improved data accuracy; and new data-driven standards for casework application of methods. © 2013 American Academy of Forensic Sciences. Source

Taylor D.,Forensic Science South Australia | Bright J.-A.,ESR Ltd. | Buckleton J.,ESR Ltd.
Forensic Science International: Genetics | Year: 2013

A method for interpreting autosomal mixed DNA profiles based on continuous modelling of peak heights is described. MCMC is applied with a model for allelic and stutter heights to produce a probability for the data given a specified genotype combination. The theory extends to handle any number of contributors and replicates, although practical implementation limits analyses to four contributors. The probability of the peak data given a genotype combination has proven to be a highly intuitive probability that may be assessed subjectively by experienced caseworkers. Whilst caseworkers will not assess the probabilities per se, they can broadly judge genotypes that fit the observed data well, and those that fit relatively less well. These probabilities are used when calculating a subsequent likelihood ratio. The method has been trialled on a number of mixed DNA profiles constructed from known contributors. The results have been assessed against a binary approach and also compared with the subjective judgement of an analyst. © 2013 Elsevier Ireland Ltd. All rights reserved. Source

Taylor D.,Forensic Science South Australia | Taylor D.,Flinders University
Forensic Science International: Genetics | Year: 2014

Continuous DNA interpretation systems make use of more information from DNA profiles than analysts have previously been able to with binary, threshold based systems. With these new continuous DNA interpretation systems and a new, more powerful, DNA profiling kit (GlobalFiler) there is an opportunity to re-examine the behaviour of a commonly used statistic in forensic science, the likelihood ratio (LR). The theoretical behaviour of the LR has been known for some time, although in many instances the behaviour has not been able to be thoroughly demonstrated due to limitations of the biological and mathematical models being used. In this paper the effects of profile complexity, replicate amplifications, assuming contributors, adding incorrect information, and adding irrelevant information to the calculation of the LR are explored. The empirical results are compared to theoretical expectations and explained. The work finishes with the results being used to dispel common misconceptions around reliability, accuracy, informativeness and reproducibility. © 2014 Published by Elsevier Ireland Ltd. All rights reserved. Source

Discover hidden collaborations