Critical analysis of forensic cut-offs and legal thresholds: A coherent approach to inference and decision.

In this paper we critically discuss the definition and use of cut-off values by forensic scientists, for example in forensic toxicology, and point out when and why such values - and ensuing categorical conclusions - are inappropriate concepts for helping recipients of expert information with their questions of interest. Broadly speaking, a cut-off is a particular value of results of analyses of a target substance (e.g., a toxic substance or one of its metabolites in biological sample from a person of interest), defined in a way such as to enable scientists to suggest conclusions regarding the condition of the person of interest. The extent to which cut-offs can be reliably defined and used is not unanimously agreed within the forensic science community, though many practitioners - especially in operational laboratories - rely on cut-offs for reasons such as ease of use and simplicity. In our analysis, we challenge this practice by arguing that choices made for convenience should not be to the detriment of balance and coherence. To illustrate our discussion, we will choose the example of alcohol markers in hair, used widely by forensic toxicologists to reach conclusions regarding the drinking behaviour of individuals. Using real data from one of the co-authors' own work and recommendations of cut-offs published by relevant professional organisations, we will point out in what sense cut-offs are incompatible with current evaluative guidelines (e.g., [31]) and show how to proceed logically without cut-offs by using a standard measure for evidential value. Our conclusions run counter to much current practice, but are inevitable given the inherent definitional and conceptual shortcomings of scientific cut-offs. We will also point out the difference between scientific cut-offs and legal thresholds and argue that the latter - but not the former - are justifiable and can be dealt with in logical evaluative procedures.

Analysing and exemplifying forensic conclusion criteria in terms of Bayesian decision theory.

There is ongoing discussion in forensic science and the law about the nature of the conclusions reached based on scientific evidence, and on how such conclusions - and conclusion criteria - may be justified by rational argument. Examples, among others, are encountered in fields such as fingermarks (e.g., 'this fingermark comes from Mr. A's left thumb'), handwriting examinations (e.g., 'the questioned signature is that of Mr. A'), kinship analyses (e.g., 'Mr. A is the father of child C') or anthropology (e.g., 'these are human remains'). Considerable developments using formal methods of reasoning based on, for example (Bayesian) decision theory, are available in literature, but currently such reference principles are not explicitly used in operational forensic reporting and ensuing decision-making. Moreover, applied examples, illustrating the principles, are scarce. A potential consequence of this in practical proceedings, and hence a cause of concern, is that underlying ingredients of decision criteria (such as losses quantifying the undesirability of adverse decision consequences), are not properly dealt with. There is merit, thus, in pursuing the study and discussion of practical examples, demonstrating that formal decision-theoretic principles are not merely conceptual considerations. Actually, these principles can be shown to underpin practical decision-making procedures and existing legal decision criteria, though often not explicitly apparent as such. In this paper, we will present such examples and discuss their properties from a Bayesian decision-theoretic perspective. We will argue that these are essential concepts for an informed discourse on decision-making across forensic disciplines and the development of a coherent view on this topic. We will also emphasize that these principles are of normative nature in the sense that they provide standards against which actual judgment and decision-making may be compared. Most importantly, these standards are justified independently of peoples' observable decision behaviour, and of whether or not one endorses these formal methods of reasoning.

The meaning of justified subjectivism and its role in the reconciliation of recent disagreements over forensic probabilism.

In this paper we reply to recent comments in this Special Issue according to which subjective probability is not considered to be a concept fit for use in forensic evaluation and expert reporting. We identify the source of these criticisms to lie in a misunderstanding of subjective probability as unconstrained subjective probability; a lack of constraint that neither corresponds to the way in which we referred to subjective probability in our previous contributions, nor to the way in which probability assignment is understood by current evaluative guidelines (e.g., of ENFSI). Specifically, we explain that we understand subjective probability as a justified assertion, i.e. a conditional assessment based on task-relevant data and information, that may be thought of as a constrained subjective probability. This leads us to emphasise again the general conclusion that there is no gap between justified (or, reasonable) subjective probability and other concepts of probability in terms of its ability to provide assessments that are soundly based on whatever relevant information available. We also note that the challenges an expert faces in reporting probabilities apply equally to all interpretations of probability, not only to subjective probability.

The consequences of understanding expert probability reporting as a decision.

In this paper we reiterate that the personalist interpretation of probability is inevitable and as least as informed as any other allegedly more 'objective' definition of probability. We also argue that the problem faced by forensic scientists, the reporting on imperfect personal knowledge, in terms of probabilities, can be reconstructed as a decision problem. Tackling this problem through a rigorous decision theoretic analysis provides further argument in support of the view that optimal probability reporting is in terms of single numbers, not intervals.

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.

Analysis and evaluation of magnetism of black toners on documents printed by electrophotographic systems.

This paper reports on a study to assess the potential of measurements of magnetism, using a proprietary magnetic analysis system, for the routine analysis of toners on documents printed by black and white electrophotographic systems. Magnetic properties of black toners on documents printed by a number of different devices were measured and compared. Our results indicate that the analysis of magnetism is complementary to traditional methods for analysing black toners, such as FTIR. Further, we find that the analysis of magnetism is realistically applicable in closed set cases, that is when the number of potential printing devices can be clearly defined.

The decisionalization of individualization.

Throughout forensic science and adjacent branches, academic researchers and practitioners continue to diverge in their perception and understanding of the notion of 'individualization', that is the claim to reduce a pool of potential donors of a forensic trace to a single source. In particular, recent shifts to refer to the practice of individualization as a decision have been revealed as being a mere change of label [1], leaving fundamental changes in thought and understanding still pending. What is more, professional associations and practitioners shy away from embracing the notion of decision in terms of the formal theory of decision in which individualization may be framed, mainly because of difficulties to deal with the measurement of desirability or undesirability of the consequences of decisions (e.g., using utility functions). Building on existing research in the area, this paper presents and discusses fundamental concepts of utilities and losses with particular reference to their application to forensic individualization. The paper emphasizes that a proper appreciation of decision tools not only reduces the number of individual assignments that the application of decision theory requires, but also shows how such assignments can be meaningfully related to constituting features of the real-world decision problem to which the theory is applied. It is argued that the decisonalization of individualization requires such fundamental insight to initiate changes in the fields' underlying understandings, not merely in their label.

Statistical hypothesis testing and common misinterpretations: Should we abandon p-value in forensic science applications?

Many people regard the concept of hypothesis testing as fundamental to inferential statistics. Various schools of thought, in particular frequentist and Bayesian, have promoted radically different solutions for taking a decision about the plausibility of competing hypotheses. Comprehensive philosophical comparisons about their advantages and drawbacks are widely available and continue to span over large debates in the literature. More recently, controversial discussion was initiated by an editorial decision of a scientific journal [1] to refuse any paper submitted for publication containing null hypothesis testing procedures. Since the large majority of papers published in forensic journals propose the evaluation of statistical evidence based on the so called p-values, it is of interest to expose the discussion of this journal's decision within the forensic science community. This paper aims to provide forensic science researchers with a primer on the main concepts and their implications for making informed methodological choices.

The importance of distinguishing information from evidence/observations when formulating propositions.

The value of forensic results crucially depends on the propositions and the information under which they are evaluated. For example, if a full single DNA profile for a contemporary marker system matching the profile of Mr A is assessed, given the propositions that the DNA came from Mr A and given it came from an unknown person, the strength of evidence can be overwhelming (e.g., in the order of a billion). In contrast, if we assess the same result given that the DNA came from Mr A and given it came from his twin brother (i.e., a person with the same DNA profile), the strength of evidence will be 1, and therefore neutral, unhelpful and irrelevant(1) to the case at hand. While this understanding is probably uncontroversial and obvious to most, if not all practitioners dealing with DNA evidence, the practical precept of not specifying an alternative source with the same characteristics as the one considered under the first proposition may be much less clear in other circumstances. During discussions with colleagues and trainees, cases have come to our attention where forensic scientists have difficulty with the formulation of propositions. It is particularly common to observe that results (e.g., observations) are included in the propositions, whereas-as argued throughout this note-they should not be. A typical example could be a case where a shoe-mark with a logo and the general pattern characteristics of a Nike Air Jordan shoe is found at the scene of a crime. A Nike Air Jordan shoe is then seized at Mr A's house and control prints of this shoe compared to the mark. The results (e.g., a trace with this general pattern and acquired characteristics corresponding to the sole of Mr A's shoe) are then evaluated given the propositions 'The mark was left by Mr A's Nike Air Jordan shoe-sole' and 'The mark was left by an unknown Nike Air Jordan shoe'. As a consequence, the footwear examiner will not evaluate part of the observations (i.e., the mark presents the general pattern of a Nike Air Jordan) whereas they can be highly informative. Such examples can be found in all forensic disciplines. In this article, we present a few such examples and discuss aspects that will help forensic scientists with the formulation of propositions. In particular, we emphasise on the usefulness of notation to distinguish results that forensic scientists should evaluate from case information that the Court will evaluate.

[Individualised parent counselling in paediatric practices for the reduction of second-hand smoke exposure of their children: a feasibility study]. / Individualisierte Beratung von Eltern in Kinderarztpraxen zur Reduktion der Passivrauchbelastung ihrer Kinder: eine Machbarkeitsstudie.

AIM: The aim of this study was to test the feasibility of a web-based programme provided by paediatric practices for counselling parents to reduce second-hand smoke exposure of their children. METHODS: Accompanying persons of children were systematically screened concerning tobacco smoking at their home in 2 Swiss paediatric practices. They were invited for programme participation if they or their partners smoked at home regularly. The web-based programme provided at least 1 computer-tailored counselling letter. Upto 3 additional counselling letters could be requested online by the participants over a period of 3 months. The letters were tailored according to the indoor smoking behaviour of the parents and considered individual barriers and resources for the establishment of a smoke-free home. Additionally, further information and advice could be requested on the programme website. Feasibility indicators were the participation rate, programme use, and programme evaluation by the participants. RESULTS: 3 055 (82.3%) of 3 712 accompanying persons of children in the paediatric practices were screened concerning tobacco smoking at their home. 96 (56.8%) of 169 eligible persons participated in the programme. 68 (70.8%) of the 96 programme participants could be reassessed at post assessment. 9 (15.0%) of 60 participants who provided a valid e-mail address requested more than one counselling letter. The counselling letters and the web-based programme were evaluated positively by the programme participants. CONCLUSION: Systematic screening combined with the provision of individually tailored counselling letters for parents to reduce second-hand smoke exposure of their children was feasible in paediatric practices. Possible strategies to in-crease the use and reach of the programme are -discussed.

Drawbacks in the scientification of forensic science.

This letter to the Editor comments on the article On the limitations of probability in conceptualizing pattern matches in forensic science by P. T. Jayaprakash (Forensic Science International, [10]).

Bayes factor for investigative assessment of selected handwriting features.

This paper extends previous research on the use of multivariate continuous data in comparative handwriting examinations, notably for gender classification. A database has been constructed by analyzing the contour shape of loop characters of type a and d by means of Fourier analysis, which allows characters to be described in a global way by a set of variables (e.g., Fourier descriptors). Sample handwritings were collected from right- and left-handed female and male writers. The results reported in this paper provide further arguments in support of the view that investigative settings in forensic science represent an area of application for which the Bayesian approach offers a logical framework. In particular, the Bayes factor is computed for settings that focus on inference of gender and handedness of the author of an incriminated handwritten text. An emphasis is placed on comparing the efficiency for investigative purposes of characters a and d.

An investigation of the potential of DIP-STR markers for DNA mixture analyses.

The genetic characterization of unbalanced mixed stains remains an important area where improvement is imperative. In fact, using the standard tools of forensic DNA profiling (i.e., STR markers), the profile of the minor contributor in mixed DNA stains cannot be successfully detected if its quantitative share of DNA is less than 10% of the mixed trace. This is due to the fact that the major contributor's profile "masks" that of the minor contributor. Besides known remedies to this problem, such as Y-STR analysis, a new compound genetic marker that consists of a Deletion/Insertion Polymorphism (DIP) linked to a Short Tandem Repeat (STR) polymorphism, has recently been developed and proposed. These novel markers are called DIP-STR markers. This paper compares, from a statistical and forensic perspective, the potential usefulness of these novel DIP-STR markers (i) with traditional STR markers in cases of moderately unbalanced mixtures, and (ii) with Y-STR markers in cases of female-male mixtures. This is done through a comparison of the distribution of 100,000 likelihood ratio values obtained using each method on simulated mixtures. This procedure is performed assuming, in turn, the prosecution's and the defence's point of view.

Decision analysis for the genotype designation in low-template-DNA profiles.

What genotype should the scientist specify for conducting a database search to try to find the source of a low-template-DNA (lt-DNA) trace? When the scientist answers this question, he or she makes a decision. Here, we approach this decision problem from a normative point of view by defining a decision-theoretic framework for answering this question for one locus. This framework combines the probability distribution describing the uncertainty over the trace's donor's possible genotypes with a loss function describing the scientist's preferences concerning false exclusions and false inclusions that may result from the database search. According to this approach, the scientist should choose the genotype designation that minimizes the expected loss. To illustrate the results produced by this approach, we apply it to two hypothetical cases: (1) the case of observing one peak for allele xi on a single electropherogram, and (2) the case of observing one peak for allele xi on one replicate, and a pair of peaks for alleles xi and xj, i ≠ j, on a second replicate. Given that the probabilities of allele drop-out are defined as functions of the observed peak heights, the threshold values marking the turning points when the scientist should switch from one designation to another are derived in terms of the observed peak heights. For each case, sensitivity analyses show the impact of the model's parameters on these threshold values. The results support the conclusion that the procedure should not focus on a single threshold value for making this decision for all alleles, all loci and in all laboratories.

Object-oriented Bayesian networks for evaluating DIP-STR profiling results from unbalanced DNA mixtures.

The genetic characterization of unbalanced mixed stains remains an important area where improvement is imperative. In fact, with current methods for DNA analysis (Polymerase Chain Reaction with the SGM Plus multiplex kit), it is generally not possible to obtain a conventional autosomal DNA profile of the minor contributor if the ratio between the two contributors in a mixture is smaller than 1:10. This is a consequence of the fact that the major contributor's profile 'masks' that of the minor contributor. Besides known remedies to this problem, such as Y-STR analysis, a new compound genetic marker that consists of a Deletion/Insertion Polymorphism (DIP), linked to a Short Tandem Repeat (STR) polymorphism, has recently been developed and proposed elsewhere in literature. The present paper reports on the derivation of an approach for the probabilistic evaluation of DIP-STR profiling results obtained from unbalanced DNA mixtures. The procedure is based on object-oriented Bayesian networks (OOBNs) and uses the likelihood ratio as an expression of the probative value. OOBNs are retained in this paper because they allow one to provide a clear description of the genotypic configuration observed for the mixed stain as well as for the various potential contributors (e.g., victim and suspect). These models also allow one to depict the assumed relevance relationships and perform the necessary probabilistic computations.

Decision-theoretic reflections on processing a fingermark.

A recent publication in this journal [1] presented the results of a field study that revealed the data provided by the fingermarks not processed in a forensic science laboratory. In their study, the authors were interested in the usefulness of this additional data in order to determine whether such fingermarks would have been worth submitting to the fingermark processing workflow. Taking these ideas as a starting point, this communication here places the fingermark in its context of a case brought before a court, and examines the question of processing or not processing a fingermark from a decision-theoretic point of view. The decision-theoretic framework presented provides an answer to this question in the form of a quantified expression of the expected value of information (EVOI) associated with the processed fingermark, which can then be compared with the cost of processing the mark.

E-learning initiatives in forensic interpretation: report on experiences from current projects and outlook.

This paper reports on the purpose, design, methodology and target audience of E-learning courses in forensic interpretation offered by the authors since 2010, including practical experiences made throughout the implementation period of this project. This initiative was motivated by the fact that reporting results of forensic examinations in a logically correct and scientifically rigorous way is a daily challenge for any forensic practitioner. Indeed, interpretation of raw data and communication of findings in both written and oral statements are topics where knowledge and applied skills are needed. Although most forensic scientists hold educational records in traditional sciences, only few actually followed full courses that focussed on interpretation issues. Such courses should include foundational principles and methodology - including elements of forensic statistics - for the evaluation of forensic data in a way that is tailored to meet the needs of the criminal justice system. In order to help bridge this gap, the authors' initiative seeks to offer educational opportunities that allow practitioners to acquire knowledge and competence in the current approaches to the evaluation and interpretation of forensic findings. These cover, among other aspects, probabilistic reasoning (including Bayesian networks and other methods of forensic statistics, tools and software), case pre-assessment, skills in the oral and written communication of uncertainty, and the development of independence and self-confidence to solve practical inference problems. E-learning was chosen as a general format because it helps to form a trans-institutional online-community of practitioners from varying forensic disciplines and workfield experience such as reporting officers, (chief) scientists, forensic coordinators, but also lawyers who all can interact directly from their personal workplaces without consideration of distances, travel expenses or time schedules. In the authors' experience, the proposed learning initiative supports participants in developing their expertise and skills in forensic interpretation, but also offers an opportunity for the associated institutions and the forensic community to reinforce the development of a harmonized view with regard to interpretation across forensic disciplines, laboratories and judicial systems.

Decision-theoretic analysis of forensic sampling criteria using bayesian decision networks.

Sampling issues represent a topic of ongoing interest to the forensic science community essentially because of their crucial role in laboratory planning and working protocols. For this purpose, forensic literature described thorough (bayesian) probabilistic sampling approaches. These are now widely implemented in practice. They allow, for instance, to obtain probability statements that parameters of interest (e.g., the proportion of a seizure of items that present particular features, such as an illegal substance) satisfy particular criteria (e.g., a threshold or an otherwise limiting value). Currently, there are many approaches that allow one to derive probability statements relating to a population proportion, but questions on how a forensic decision maker--typically a client of a forensic examination or a scientist acting on behalf of a client--ought actually to decide about a proportion or a sample size, remained largely unexplored to date. The research presented here intends to address methodology from decision theory that may help to cope usefully with the wide range of sampling issues typically encountered in forensic science applications. The procedures explored in this paper enable scientists to address a variety of concepts such as the (net) value of sample information, the (expected) value of sample information or the (expected) decision loss. All of these aspects directly relate to questions that are regularly encountered in casework. Besides probability theory and bayesian inference, the proposed approach requires some additional elements from decision theory that may increase the efforts needed for practical implementation. In view of this challenge, the present paper will emphasise the merits of graphical modelling concepts, such as decision trees and bayesian decision networks. These can support forensic scientists in applying the methodology in practice. How this may be achieved is illustrated with several examples. The graphical devices invoked here also serve the purpose of supporting the discussion of the similarities, differences and complementary aspects of existing bayesian probabilistic sampling criteria and the decision-theoretic approach proposed throughout this paper.

Learning about Bayesian networks for forensic interpretation: an example based on the 'the problem of multiple propositions'.

Both, Bayesian networks and probabilistic evaluation are gaining more and more widespread use within many professional branches, including forensic science. Notwithstanding, they constitute subtle topics with definitional details that require careful study. While many sophisticated developments of probabilistic approaches to evaluation of forensic findings may readily be found in published literature, there remains a gap with respect to writings that focus on foundational aspects and on how these may be acquired by interested scientists new to these topics. This paper takes this as a starting point to report on the learning about Bayesian networks for likelihood ratio based, probabilistic inference procedures in a class of master students in forensic science. The presentation uses an example that relies on a casework scenario drawn from published literature, involving a questioned signature. A complicating aspect of that case study - proposed to students in a teaching scenario - is due to the need of considering multiple competing propositions, which is an outset that may not readily be approached within a likelihood ratio based framework without drawing attention to some additional technical details. Using generic Bayesian networks fragments from existing literature on the topic, course participants were able to track the probabilistic underpinnings of the proposed scenario correctly both in terms of likelihood ratios and of posterior probabilities. In addition, further study of the example by students allowed them to derive an alternative Bayesian network structure with a computational output that is equivalent to existing probabilistic solutions. This practical experience underlines the potential of Bayesian networks to support and clarify foundational principles of probabilistic procedures for forensic evaluation.