Variance stabilizing transformations for meta-analysis of binary data

The binary data are often used in the medical field or in social sciences to study an effect between two groups. It can be for example the effect of a treatment between a treated and a control group. The outcome will be the number of successes and the total number of trials in each group. Let us call X1, X2 the successes and n1, n2 the number of trials, then X1 and X2 are binomial random variables with probabilities p1 and p2 respectively. The difference between these two probabilities will attest the difference of the effect between the two groups. Several statistics may be used to compare two probabilities.

Calculation of geological uncertainties associated with 3-D geological models

Ian Pomian-Srzednicki

3-D geological models are built with data collected in the field such as boreholes, geophysical measurements, pilot shafts or geological mapping. Unfortunately, these data are always limited in number. It implies that geological information is sparse and subsurface models are thus always built of both subjective interpretation and mathematical interpolation/extrapolation techniques. These models are therefore uncertain and this uncertainty is rarely pointed out in a geological prognosis. Our study proposes to bring a new methodology for the evaluation of geological uncertainties related to 3-D subsurface models and to test its potential use. The methodology we propose is based on the 3-D subsurface model, which is here considered as the most probable prediction (notion of best guess). The various geological interfaces that compose the subsurface model are handled individually as Gaussian random fields. At each location of an interface, the random function Z(u) describing the position of this interface is composed of a deterministic part m(u) which represents the expected position, and a random part σ(u)ε(u) which describes fluctuations around the predicted position. Then, a model of spatial variability (a variogram function γ(h)) is proposed in order to condition the random field according to available observations. Several structural constraints, such as the shape of folds and the thickness of layers can also be accounted for in this model. At this point, we are able to estimate the local variance all over the study area by the application of the kriging technique. Finally, the variability is converted into three-dimensional information by calculating probabilities, this describes the occurrence of the various rock masses that are present in the study area. The probabilities are calculated according to intersection rules that govern the stratigraphic sequence of the subsurface model, and they allow us to build a probabilistic model of subsurface structures in the form of a three-dimensional probability field. All of this has been incorporated in a computer program.

EPFL2002

Inference about the number of contributors to a DNA mixture: Comparative analyses of a Bayesian network approach and the maximum allele count method

Kjell Konis

In the forensic examination of DNA mixtures, the question of how to set the total number of contributors (N) presents a topic of ongoing interest. Part of the discussion gravitates around issues of bias, in particular when assessments of the number of contributors are not made prior to considering the genotypic configuration of potential donors. Further complication may stem from the observation that, in some cases, there may be numbers of contributors that are incompatible with the set of alleles seen in the profile of a mixed crime stain, given the genotype of a potential contributor. In such situations, procedures that take a single and fixed number contributors as their output can lead to inferential impasses. Assessing the number of contributors within a probabilistic framework can help avoiding such complication. Using elements of decision theory, this paper analyses two strategies for inference on the number of contributors. One procedure is deterministic and focuses on the minimum number of contributors required to 'explain' an observed set of alleles. The other procedure is probabilistic using Bayes' theorem and provides a probability distribution for a set of numbers of contributors, based on the set of observed alleles as well as their respective rates of occurrence. The discussion concentrates on mixed stains of varying quality (i.e., different numbers of loci for which genotyping information is available). A so-called qualitative interpretation is pursued since quantitative information such as peak area and height data are not taken into account. The competing procedures are compared using a standard scoring rule that penalizes the degree of divergence between a given agreed value for N, that is the number of contributors, and the actual value taken by N. Using only modest assumptions and a discussion with reference to a casework example, this paper reports on analyses using simulation techniques and graphical models (i.e., Bayesian networks) to point out that setting the number of contributors to a mixed crime stain in probabilistic terms is, for the conditions assumed in this study, preferable to a decision policy that uses categoric assumptions about N. (C) 2012 Elsevier Ireland Ltd. All rights reserved.

Elsevier2012

Calculation of geological uncertainties associated with 3-D geological models

Ian Pomian-Srzednicki

EPFL2002

Inference about the number of contributors to a DNA mixture: Comparative analyses of a Bayesian network approach and the maximum allele count method

Kjell Konis

Elsevier2012