Prediction in forensic science: a critical examination of common understandings

In this commentary, we argue that the term 'prediction' is overly used when in fact, referring to foundational writings of de Finetti, the correspondent term should be inference. In particular, we intend (i) to summarize and clarify relevant subject matter on prediction from established statistical theory, and (ii) point out the logic of this understanding with respect practical uses of the term prediction. Written from an interdisciplinary perspective, associating statistics and forensic science as an example, this discussion also connects to related fields such as medical diagnosis and other areas of application where reasoning based on scientific results is practiced in societal relevant contexts. This includes forensic psychology that uses prediction as part of its vocabulary when dealing with matters that arise in the course of legal proceedings.

Being Small for Gestational Age: Does it Matter for the Neurodevelopment of Premature Infants? A Cohort Study.

BACKGROUND: Whether being small for gestational age (SGA) increases the risk of adverse neurodevelopmental outcome in premature infants remains controversial. OBJECTIVE: to study the impact of SGA (birthweight < percentile 10) on cognition, behavior, neurodevelopmental impairment and use of therapy at 5 years old. METHODS: This population-based prospective cohort included infants born before 32 weeks of gestation. Cognition was evaluated with the K-ABC, and behavior with the Strengths and Difficulties Questionnaire (SDQ). Primary outcomes were cognitive and behavioral scores, as well as neurodevelopmental impairment (cognitive score < 2SD, hearing loss, blindness, or cerebral palsy). The need of therapy, an indirect indicator of neurodevelopmental impairment, was a secondary outcome. Linear and logistic regression models were used to analyze the association of SGA with neurodevelopment. RESULTS: 342/515 (76%) premature infants were assessed. SGA was significantly associated with hyperactivity scores of the SDQ (coefficient 0.81, p < 0.04), but not with cognitive scores, neurodevelopmental impairment or the need of therapy. Gestational age, socio-economic status, and major brain lesions were associated with cognitive outcome in the univariate and multivariate model, whereas asphyxia, sepsis and bronchopulmonary dysplasia were associated in the univariate model only. Severe impairment was associated with fetal tobacco exposition, asphyxia, gestational age and major brain lesions. Different neonatal factors were associated with the use of single or multiple therapies: children with one therapy were more likely to have suffered birth asphyxia or necrotizing enterocolitis, whereas the need for several therapies was predicted by major brain lesions. DISCUSSION: In this large cohort of premature infants, assessed at 5 years old with a complete panel of tests, SGA was associated with hyperactive behavior, but not with cognition, neurodevelopmental impairment or use of therapy. Birthweight <10th percentile alone does not appear to be an independent risk factor of neurodevelopmental adverse outcome in preterm children.

Populating ACT-R's Declarative Memory with Internet Statistics

Decision situations are often characterized by uncertainty: we do not know the values of the different options on all attributes and have to rely on information stored in our memory to decide. Several strategies have been proposed to describe how people make inferences based on knowledge used as cues. The present research shows how declarative memory of ACT-R models could be populated based on internet statistics. This will allow to simulate the performance of decision strategies operating on declarative knowledge based on occurrences and co-occurrences of objects and cues in the environment.

Conditioning of multiple-point statistics facies simulations to tomographic images

Geophysical tomography captures the spatial distribution of the underlying geophysical property at a relatively high resolution, but the tomographic images tend to be blurred representations of reality and generally fail to reproduce sharp interfaces. Such models may cause significant bias when taken as a basis for predictive flow and transport modeling and are unsuitable for uncertainty assessment. We present a methodology in which tomograms are used to condition multiple-point statistics (MPS) simulations. A large set of geologically reasonable facies realizations and their corresponding synthetically calculated cross-hole radar tomograms are used as a training image. The training image is scanned with a direct sampling algorithm for patterns in the conditioning tomogram, while accounting for the spatially varying resolution of the tomograms. In a post-processing step, only those conditional simulations that predicted the radar traveltimes within the expected data error levels are accepted. The methodology is demonstrated on a two-facies example featuring channels and an aquifer analog of alluvial sedimentary structures with five facies. For both cases, MPS simulations exhibit the sharp interfaces and the geological patterns found in the training image. Compared to unconditioned MPS simulations, the uncertainty in transport predictions is markedly decreased for simulations conditioned to tomograms. As an improvement to other approaches relying on classical smoothness-constrained geophysical tomography, the proposed method allows for: (1) reproduction of sharp interfaces, (2) incorporation of realistic geological constraints and (3) generation of multiple realizations that enables uncertainty assessment.

Dysménorrhée: patience, pilules ou bouillotte [Dysmenorrhea: patience, pills or hot-water bottle?].

Which treatments are used for dysmenorrhea and with what reported outcome? A questionnaire was sent to 2400 students and apprentices, following the "retrospective treatment-outcome" method. The response rate was 22%. Most frequent treatments used are ibuprofene (53%), paracetamol (51%), hormonal contraception (40%), hot-water bottle (or hot pad) (35%), food supplements or medicinal plants (23%). Physicians only discuss a tiny proportion of dysmenorrhea treatment in their consultation, because it is mostly a matter of self-treatment, with the family as the source of information in 80% of the cases. Rather surprising because not mentioned in most official guidelines, hot-water bottle (or hot pad) appears as the treatment followed by the best reported outcome (satisfactory in 92% of users).

Two-dimensional probabilistic inversion of plane-wave electromagnetic data: Methodology, model constraints and joint inversion with electrical resistivity data

Probabilistic inversion methods based on Markov chain Monte Carlo (MCMC) simulation are well suited to quantify parameter and model uncertainty of nonlinear inverse problems. Yet, application of such methods to CPU-intensive forward models can be a daunting task, particularly if the parameter space is high dimensional. Here, we present a 2-D pixel-based MCMC inversion of plane-wave electromagnetic (EM) data. Using synthetic data, we investigate how model parameter uncertainty depends on model structure constraints using different norms of the likelihood function and the model constraints, and study the added benefits of joint inversion of EM and electrical resistivity tomography (ERT) data. Our results demonstrate that model structure constraints are necessary to stabilize the MCMC inversion results of a highly discretized model. These constraints decrease model parameter uncertainty and facilitate model interpretation. A drawback is that these constraints may lead to posterior distributions that do not fully include the true underlying model, because some of its features exhibit a low sensitivity to the EM data, and hence are difficult to resolve. This problem can be partly mitigated if the plane-wave EM data is augmented with ERT observations. The hierarchical Bayesian inverse formulation introduced and used herein is able to successfully recover the probabilistic properties of the measurement data errors and a model regularization weight. Application of the proposed inversion methodology to field data from an aquifer demonstrates that the posterior mean model realization is very similar to that derived from a deterministic inversion with similar model constraints.

Glutathione deficit impairs myelin maturation: relevance for white matter integrity in schizophrenia patients.

Schizophrenia pathophysiology implies both abnormal redox control and dysconnectivity of the prefrontal cortex, partly related to oligodendrocyte and myelin impairments. As oligodendrocytes are highly vulnerable to altered redox state, we investigated the interplay between glutathione and myelin. In control subjects, multimodal brain imaging revealed a positive association between medial prefrontal glutathione levels and both white matter integrity and resting-state functional connectivity along the cingulum bundle. In early psychosis patients, only white matter integrity was correlated with glutathione levels. On the other side, in the prefrontal cortex of peripubertal mice with genetically impaired glutathione synthesis, mature oligodendrocyte numbers, as well as myelin markers, were decreased. At the molecular levels, under glutathione-deficit conditions induced by short hairpin RNA targeting the key glutathione synthesis enzyme, oligodendrocyte progenitors showed a decreased proliferation mediated by an upregulation of Fyn kinase activity, reversed by either the antioxidant N-acetylcysteine or Fyn kinase inhibitors. In addition, oligodendrocyte maturation was impaired. Interestingly, the regulation of Fyn mRNA and protein expression was also impaired in fibroblasts of patients deficient in glutathione synthesis. Thus, glutathione and redox regulation have a critical role in myelination processes and white matter maturation in the prefrontal cortex of rodent and human, a mechanism potentially disrupted in schizophrenia.

When does charisma matter for top-level leaders? Effect of attributional ambiguity

One stream of leadership theory suggests leaders are evaluated via inferential observer processes that compare the fit of the target to a prototype of an ideal (charismatic) leader. Attributional theories of leadership suggest that evaluations depend on knowledge of past organizational performance, which is attributed to the leader's skills. We develop a novel theory showing how inferential and attributional processes simultaneously explain top-level leader evaluation and ultimately leader retention and selection. We argue that observers will mostly rely on attributional mechanisms when performance signals clearly indicate good or poor performance outcomes. However, under conditions of attributional ambiguity (i.e., when performance signals are unclear), observers will mostly rely on inferential processes. In Study 1 we tested our theory in an unconventional context-the U.S. presidential election-and found that the two processes, due to the leader's charisma and country economic performance, interact in predicting whether a leader is selected. Using a business context and an experimental design, in Study 2 we show that CEO charisma and firm performance interact in predicting leader retention, confirming the results we found in Study 1. Our results suggest that this phenomenon is quite general and can apply to various performance domains.