Defense Propositions Research Articles

Reporting likelihood ratio for casework in firearm evidence identification

Likelihood ratio (LR) plays an important role in estimating the weight of evidence in firearm evidence identifications. LR is computed from a statistical model including the distribution of the known-matching (KM or within) and known-nonmatching (KNM or between) comparison scores. Current LR procedures rely on KM/KNM scores from existing reference firearm toolmark data sets or alternatively from generating a set of test fires using multiple firearms. Both procedures may contain theoretical or practical issues which may hinder the LR procedures from reporting an unbiased LR estimation in casework. In this paper, a reference data set was established from a set of firearms, each test-fired two cartridge cases, resulting in a basic data set and a control data set. The congruent matching cells (CMC) method was used to generate CMC scores that are used to fit in the KM/KNM statistical distributions for LR estimation. In the initial test, 130 firearms from eight manufacturers were used for generating a reference data set consisting of 260 cartridge cases representing 130 KM and 8385 KNM pairwise breech face images. Test results showed that the KM and KNM distribution intersect at CMC=2, which is equivalent to LR=1 (equally to support both the prosecutor and the defense propositions). When the CMC threshold is increased to 6 or more, the LR values are higher than a million, which can provide extremely strong support to the conclusion of the same firearm (or the prosecutor's proposition) in the casework of firearm evidence identification.

A likelihood-ratio framework for evaluating results of forensic gunshot-residue analysis

When reporting results of Gunshot Residue (GSR) analysis from a person suspected to be involved in a recent shooting, most forensic experts only provide the court with the raw results (i.e. the number of GSR particles found) and a disclaimer that a positive finding does not prove that the suspect was involved in a firearm shooting incident whilst a negative finding does not prove that he was not. Probabilistic analysis of the GSR results provides more value to the court, so the present study calculated likelihood ratio (LR) values for finding 0–8 characteristic GSR particles (containing Lead, Barium and Antimony) on a suspect's hands, based on the available GSR data from the published literature as well as studies by the authors. Defense propositions, i.e. modes for GSR acquisition other than involvement in a shooting event, were divided into three broad categories: low, medium and heavy background. For each background level and number of GSR particles found, minimal and maximal LR values were calculated. Thus, for each proposition the defense provides for the presence of GSR on the defendant's hands, the forensic expert can provide a possible set of minimal and maximal LR values, leaving the court to examine the defendant's contention and decide which of the three background modes is more plausible according to the circumstances of the specific case.

The use of Bayesian Networks and simulation methods to identify the variables impacting the value of evidence assessed under activity level propositions in stabbing cases

This paper presents a methodology allowing identification of the variables associated with transfer, persistence or recovery of DNA traces that have the most significant impact on the result of an evaluation measured through a likelihood ratio (LR). It builds on a case scenario involving trace DNA recovered from knife handles where the prosecution alleges that the person of interest (POI) stabbed a victim whereas the defence claims that the POI has nothing to do with the incident and the victim was stabbed by an alternative offender (AO). The defence proposition will also be refined to account for the possibility of secondary transfer. The variables having a significant impact on the LR are identified taking advantage of a graphical probabilistic environment (using Bayesian Networks, BN), coupled with simulation techniques. The paper presents (a) a BN, based on previous work Taylor et al. [5]; (b) its parametrization based on the current literature that represents the current state of knowledge used to inform the conditional probability tables of the BN and; (c) the implementation of the simulation methods. Results show that, regardless of the DNA outcome obtained, the most impacting variable is the “DNA match probability” when the defence alleged that the POI has nothing to do with the incident. It means that, given the current state of knowledge, such cases can easily be interpreted considering activity level propositions as they would not require any further data acquisition. When secondary transfer is alleged under the defence’s perspective, the LRs are generally much lower than for the previous case. The DNA match probability has less impact and variables associated with the donor will take the lead on the ranges observed on the LRs. Overall, once extraction and sampling efficiency have been set, the remaining variables that have an impact on the value of the evidence are the DNA quantity on hands and the background. With the most impacting variables so identified, it becomes manageable to direct further data acquisition if so required. Generally, the background that could be present on the knife handle, the environmental conditions are not critical due to their limited impact on the LR value. We note, however, that this identification of the significant variables depends on the obtained DNA results and this selection may be refined on a case by case basis. To allow one to explore all possibilities a dedicated Shiny application has been designed (https://lydie-samie.shinyapps.io/DNA_Activity/).

Reframing the Debate: A Question of Probability, Not of Likelihood Ratio

Evidential value is measured by a likelihood ratio. This ratio has two components, the probability, or probability density, of the evidence if the prosecution proposition is true and the probability (density) of the evidence if the defence proposition is true. It takes the form of a single value, even if these probabilities are subjective measures of belief of the reporting forensic scientist.

Reframing the debate: A question of probability, not of likelihood ratio

Evidential value is measured by a likelihood ratio. This ratio has two components, the probability, or probability density, of the evidence if the prosecution proposition is true and the probability (density) of the evidence if the defence proposition is true. It takes the form of a single value, even if these probabilities are subjective measures of belief of the reporting forensic scientist.

A Practical Guide for the Formulation of Propositions in the Bayesian Approach to DNA Evidence Interpretation in an Adversarial Environment.

The interpretation of complex DNA profiles is facilitated by a Bayesian approach. This approach requires the development of a pair of propositions: one aligned to the prosecution case and one to the defense case. This note explores the issue of proposition setting in an adversarial environment by a series of examples. A set of guidelines generalize how to formulate propositions when there is a single person of interest and when there are multiple individuals of interest. Additional explanations cover how to handle multiple defense propositions, relatives, and the transition from subsource level to activity level propositions. The propositions depend on case information and the allegations of each of the parties. The prosecution proposition is usually known. The authors suggest that a sensible proposition is selected for the defense that is consistent with their stance, if available, and consistent with a realistic defense if their position is not known.

Automatic Comparison and Evaluation of Impressions Left by a Firearm on Fired Cartridge Cases

Recent years have been characterized by a series of publications in the field of firearms investigation questioning the reliability and objectivity of such examination. This research investigates new solutions to decrease the subjective component affecting the evaluation that follows the comparison of impressions left by a firearm on the surface of spent cartridge cases. An automatic comparison system based on 3D measurements has been developed and coupled to a bivariate evaluative model allowing assigning likelihood ratios. Based on a dataset of 79 pistols (all SIG Sauer 9 mm Luger caliber), the system shows a very high discriminating power and the LRs that it provides are very indicative of the true state under both the prosecution and the defense propositions. For example, likelihood ratios exceeding a billion are predominantly obtained when impressions originating from the same source are compared. The system is also characterized by relatively low rates (≤ 1%) of misleading evidence depending on the firearm considered.

Integration of Pore Features into the Evaluation of Fingerprint Evidence*

Fingerprint practitioners rely on level 3 features to make decisions in relation to the source of an unknown friction ridge skin impression. This research proposes to assess the strength of evidence associated with pores when shown in (dis)agreement between a mark and a reference print. Based upon an algorithm designed to automatically detect pores, a metric is defined in order to compare different impressions. From this metric, the weight of the findings is quantified using a likelihood ratio. The results obtained on four configurations and 54 donors show the significant contribution of the pore features and translate into statistical terms what latent fingerprint examiners have developed holistically through experience. The system provides LRs that are indicative of the true state under both the prosecution and the defense propositions. Not only such a system brings transparency regarding the weight to assign to such features, but also forces a discussion in relation to the risks of such a model to mislead.

Understanding DNA results within the case context: importance of the alternative proposition

There is a growing awareness of the dangers of reporting DNA profiling results in criminal investigations without consideration of the implications of the finding within the case context. A recent article in the New York Times (July 24 2013) described a robbery that resulted in the death of Raveesh Kumra. Foreign DNA on the victim's fingernails corresponded with the profile of a local man, Lukis Anderson who was charged with murder. Following 5 months in prison, it was found that Mr. Anderson could not have committed the crime as he was in hospital at the time of the robbery. This and other cases demonstrate that reporting the DNA profile results alone can be misleading. The investigators and courts may be impressed by the probity of the DNA result in isolation and not think about other issues such as the possibility of secondary transfer. The Association of Forensic Science Practitioners, UK and Ireland (2009) attempted to address this problem for trace evidence through the introduction of Standards for the Formulation of Evaluative Forensic Science Expert Opinion. These standards require that the scientific finding is considered relative to two mutually exclusive propositions, one from the prosecution and one from the defence [based on the work of Evett et al. (2000)]. Within the case context, the probability of the evidence given the prosecution proposition divided by the probability of the evidence given the defence proposition produces the Likelihood Ratio (LR). The magnitude of the LR indicates the degree of support for one proposition vs. the other. This approach allows the scientist to help the court to understand the implications of the findings for the particular circumstances of each case. Since the publication of the AFSP standard, EFE (Forensic Science Laboratory, Dublin, Ireland) has worked at applying the criteria to its casework. The following examples are drawn from EFE casework and illustrate how the alternative proposition can significantly affect the impact of the DNA finding.

Expressing evaluative opinions: A position statement

The judgment of the Court of Appeal in R v T [1] raises several issues relating to the evaluation of scientific evidence that, we believe, require a response. We, the undersigned, oppose any response to the judgment that would result in a movement away from the use of logical methods for evidence evaluation. A paper in this issue of the Journal [2] re-iterates logical principles of evidence interpretation that are accepted by a broad range of those who have an interest in forensic reasoning. The divergence between those principles of interpretation and the apparent implications of the R v T ruling are epitomised by the following issues that represent our collective position with regard to the evaluation of evidence within the context of a criminal trial. 1) The interpretation of scientific evidence invokes reasoning in the face of uncertainty. Probability theory provides the only coherent logical foundation for such reasoning. 2) To form an evaluative opinion from a set of observations, it is necessary for the forensic scientist to consider those observations in the light of propositions that represent the positions of the different participants in the legal process. In a criminal trial, the propositions will represent the positions of prosecution and defence, respectively. 3) It is necessary for the scientist to consider the probability of the observations given each of the stated propositions. Not only is it not appropriate for the scientist to consider the probability of the proposition given the observations, there is a danger that in doing so the jury will be misled. 4) The ratio of the probability of the observations given the prosecution proposition to the probability of the observations given the defence proposition, which is known as the likelihood ratio, provides the most appropriate foundation for assisting the court in establishing the weight that should be assigned to those observations. 5) A verbal scale based on the notion of the likelihood ratio is the most appropriate basis for communication of an evaluative expert opinion to the court. It can be phrased in terms of support for one of a pair of clearly stated propositions. 6) Not only are phrases such as “could have come from” or “is consistent with” ineffective for communicating the scientist's opinion with regard to the weight that should be assigned to a set of observations, but there is also a danger that they may be misleading. 7) Probabilities should be informed by data, knowledge and experience. All data collections are imperfect and incomplete and it necessarily follows that different experts might legitimately assign different probabilities to the same set of observations. 8) The logical approach to evaluating evidence implicit in the foregoing points has come to be known as the “Bayesian approach”. The ideas behind this approach are not novel. Indeed, they were first applied to resolving a serious miscarriage of justice in the Dreyfus case in 1908. 9) It is regrettable that the judgment confuses the Bayesian approach with the use of Bayes' Theorem. The Bayesian approach does not necessarily involve the use of Bayes' Theorem. 10) While we are fully in agreement with the principle of disclosure, candour and full disclosure in court can undermine comprehensibility when scientific evaluations involve technicalities. Pre-trial hearings should be used to explore the basis of expert opinions and to resolve if possible any differences between experts.

The calcuLatoR—software for evaluation of evidence value of continuous type data

The role of the forensic chemist can be to evaluate physicochemical data (evidence— E) in the context of the prosecution proposition H p and the defence proposition H d . From a forensic point of view, the most suitable form of evaluating the evidence value of physicochemical data is by calculation of the likelihood ratio (LR). There are many LR models which could be used for measuring the evidence value of results of analysis of glass samples. A disadvantage of these models is that there is a lack of commercially available software that is suitable for use by forensic experts, who may lack experience in programming. Therefore such users are required to write their own case specific routines, e.g. in R software ( www.r-project.org). The Graphical User Interface software named “calcuLatoR” for the calculation of the LR value for univariate data (such as refractive index data) has been developed. The LR model assumes two sources of variation, i.e. between replicates within the same object (within object variability) and between replicates between various objects (between object variability). It was assumed that the within object distribution is normal with constant variance. The between object distribution was modelled by a univariate kernel density estimator, i.e. using Gaussian kernels. The calcuLatoR could also be used for evaluating the evidential value of multivariate data but it must be assumed that the considered variables are independent. In this situation, LR should be calculated for each variable separately and a final value of LR is equal to their product. Validation of the calculations performed by the calcuLatoR was carried out by comparing results obtained by routines written in R.

Calculating likelihood ratios for a mixed DNA profile when a contribution from a genetic relative of a suspect is proposed

This technical note describes a practical method for evaluating evidence in the case of a two person conditioned DNA mixture where the defence proposition is that the unknown contributor is genetically related to the suspect. A conditioned mixture is one where the presence of DNA from one of two individuals is accepted by both prosecution and defence. A typical example would be a vaginal swab in an alleged rape case, where the presence of the complainant's DNA would be expected and samples have been taken from the complainant and a suspect. Much has been written about the interpretation of such mixtures and the calculation of the conditional genotype probabilities that must be carried out. In general, such treatments assume that the unknown contributor, under the defence proposition, is unrelated to the known individuals. In this paper, we consider the case where the defence proposition is that the unknown contributor is genetically related to the suspect. We describe a method, incorporating a flow chart and reference tables that facilitate manual calculations of the likelihood ratio for several postulated genetic relationships.

Computation of Likelihood Ratios in Fingerprint Identification for Configurations of Three Minutiæ

Recent challenges to fingerprint evidence have brought forward the need for peer-reviewed scientific publications to support the evidential value assessment of fingerprint. This paper proposes some research directions to gather statistical knowledge of the within-source and between-sources variability of configurations of three minutiae on fingermarks and fingerprints. This paper proposes the use of the likelihood ratio (LR) approach to assess the value of fingerprint evidence. The model explores the statistical contribution of configurations of three minutiae using Tippett plots and related measures to assess the quality of the system. Features vectors used for statistical analysis have been obtained following a preprocessing step based on Gabor filtering and image processing to extract minutia position, type, and direction. Spatial relationships have been coded using Delaunay triangulation. The metric, used to assess similarity between two feature vectors is based on an Euclidean distance measure. The within-source variability has been estimated using a sample of 216 fingerprints from four fingers (two donors). Between-sources variability takes advantage of a database of 818 ulnar loops from randomly selected males. The results show that the data-driven approach adopted here is robust. The magnitude of LRs obtained under the prosecution and defense propositions stresses upon the major evidential contribution that small portions of fingermark, containing three minutiae, can provide regardless of its position on the general pattern.