Estimating Bayesian decision problems with heterogeneous priors

In many areas of economics there is a growing interest in how expertise andpreferences drive individual and group decision making under uncertainty. Increasingly, we wish to estimate such models to quantify which of these drive decisionmaking. In this paper we propose a new channel through which we can empirically identify expertise and preference parameters by using variation in decisionsover heterogeneous priors. Relative to existing estimation approaches, our \Prior-Based Identification" extends the possible environments which can be estimated,and also substantially improves the accuracy and precision of estimates in thoseenvironments which can be estimated using existing methods.

Further insights on the French WISC-IV factor structure through Bayesian structural equation modeling

The interpretation of the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) is based on a 4-factor model, which is only partially compatible with the mainstream Cattell-Horn-Carroll (CHC) model of intelligence measurement. The structure of cognitive batteries is frequently analyzed via exploratory factor analysis and/or confirmatory factor analysis. With classical confirmatory factor analysis, almost all crossloadings between latent variables and measures are fixed to zero in order to allow the model to be identified. However, inappropriate zero cross-loadings can contribute to poor model fit, distorted factors, and biased factor correlations; most important, they do not necessarily faithfully reflect theory. To deal with these methodological and theoretical limitations, we used a new statistical approach, Bayesian structural equation modeling (BSEM), among a sample of 249 French-speaking Swiss children (8-12 years). With BSEM, zero-fixed cross-loadings between latent variables and measures are replaced by approximate zeros, based on informative, small-variance priors. Results indicated that a direct hierarchical CHC-based model with 5 factors plus a general intelligence factor better represented the structure of the WISC-IV than did the 4-factor structure and the higher order models. Because a direct hierarchical CHC model was more adequate, it was concluded that the general factor should be considered as a breadth rather than a superordinate factor. Because it was possible for us to estimate the influence of each of the latent variables on the 15 subtest scores, BSEM allowed improvement of the understanding of the structure of intelligence tests and the clinical interpretation of the subtest scores.

Recent misconceptions about the "database search problem": a probabilistic analysis using Bayesian networks

This paper analyses and discusses arguments that emerge from a recent discussion about the proper assessment of the evidential value of correspondences observed between the characteristics of a crime stain and those of a sample from a suspect when (i) this latter individual is found as a result of a database search and (ii) remaining database members are excluded as potential sources (because of different analytical characteristics). Using a graphical probability approach (i.e., Bayesian networks), the paper here intends to clarify that there is no need to (i) introduce a correction factor equal to the size of the searched database (i.e., to reduce a likelihood ratio), nor to (ii) adopt a propositional level not directly related to the suspect matching the crime stain (i.e., a proposition of the kind 'some person in (outside) the database is the source of the crime stain' rather than 'the suspect (some other person) is the source of the crime stain'). The present research thus confirms existing literature on the topic that has repeatedly demonstrated that the latter two requirements (i) and (ii) should not be a cause of concern.

Statistics in archaeology: New directions

Connections between Statistics and Archaeology have always appeared veryfruitful. The objective of this paper is to offer an outlook of somestatistical techniques that are being developed in the most recentyears and that can be of interest for archaeologists in the short run.

Nonparametric bounds on the returns to language skills

This paper applies the theoretical literature on nonparametric bounds ontreatment effects to the estimation of how limited English proficiency (LEP)affects wages and employment opportunities for Hispanic workers in theUnited States. I analyze the identifying power of several weak assumptionson treatment response and selection, and stress the interactions between LEPand education, occupation and immigration status. I show that thecombination of two weak but credible assumptions provides informative upperbounds on the returns to language skills for certain subgroups of thepopulation. Adding age at arrival as a monotone instrumental variable alsoprovides informative lower bounds.

Forecasting and turning point predictions in a Bayesian panel VAR model

We provide methods for forecasting variables and predicting turning points in panel Bayesian VARs. We specify a flexible model which accounts for both interdependencies in the cross section and time variations in the parameters. Posterior distributions for the parameters are obtained for a particular type of diffuse, for Minnesota-type and for hierarchical priors. Formulas for multistep, multiunit point and average forecasts are provided. An application to the problem of forecasting the growth rate of output and of predicting turning points in the G-7 illustrates the approach. A comparison with alternative forecasting methods is also provided.

Worst-case bounds for the logarithmic loss of predictors

We investigate on-line prediction of individual sequences. Given a class of predictors, the goal is to predict as well as the best predictor in the class, where the loss is measured by the self information (logarithmic) loss function. The excess loss (regret) is closely related to the redundancy of the associated lossless universal code. Using Shtarkov's theorem and tools from empirical process theory, we prove a general upper bound on the best possible (minimax) regret. The bound depends on certain metric properties of the class of predictors. We apply the bound to both parametric and nonparametric classes ofpredictors. Finally, we point out a suboptimal behavior of the popular Bayesian weighted average algorithm.

On the accuracy of Latin American trade statistics: A nonparametric test for 1925

This paper proposes a nonparametric test in order to establish the level of accuracy of theforeign trade statistics of 17 Latin American countries when contrasted with the trade statistics of the main partners in 1925. The Wilcoxon Matched-Pairs Ranks test is used to determine whether the differences between the data registered by exporters and importers are meaningful, and if so, whether the differences are systematic in any direction. The paper tests for the reliability of the data registered for two homogeneous products, petroleum and coal, both in volume and value. The conclusion of the several exercises performed is that we cannot accept the existence of statistically significant differences between the data provided by the exporters and the registered by the importing countries in most cases. The qualitative historiography of Latin American describes its foreign trade statistics as mostly unusable. Our quantitative results contest this view.

Inequalities for a new data-based method for selecting nonparametric density estimates

We continue the development of a method for the selection of a bandwidth or a number of design parameters in density estimation. We provideexplicit non-asymptotic density-free inequalities that relate the $L_1$ error of the selected estimate with that of the best possible estimate,and study in particular the connection between the richness of the classof density estimates and the performance bound. For example, our methodallows one to pick the bandwidth and kernel order in the kernel estimatesimultaneously and still assure that for {\it all densities}, the $L_1$error of the corresponding kernel estimate is not larger than aboutthree times the error of the estimate with the optimal smoothing factor and kernel plus a constant times $\sqrt{\log n/n}$, where $n$ is the sample size, and the constant only depends on the complexity of the family of kernels used in the estimate. Further applications include multivariate kernel estimates, transformed kernel estimates, and variablekernel estimates.

Fixed and random effects in Classical and Bayesian regression

This paper proposes a common and tractable framework for analyzingdifferent definitions of fixed and random effects in a contant-slopevariable-intercept model. It is shown that, regardless of whethereffects (i) are treated as parameters or as an error term, (ii) areestimated in different stages of a hierarchical model, or whether (iii)correlation between effects and regressors is allowed, when the sameinformation on effects is introduced into all estimation methods, theresulting slope estimator is also the same across methods. If differentmethods produce different results, it is ultimately because differentinformation is being used for each methods.

Therapeutic Drug Monitoring (TDM) of Imatinib: Effectiveness of Bayesian dose adjustment for CML patients enrolled in the Imatinib Concentration Monitoring (I-COME) study

Introduction: As imatinib pharmacokinetics are highly variable, plasma levels differ largely between patients under the same dosage. Retrospective studies in chronic myeloid leukemia (CML) patients showed significant correlations between low levels and suboptimal response, as well as between high levels and poor tolerability. Monitoring of trough plasma levels, targeting 1000 μg/L and above, is thus increasingly advised. Our study was launched to assess prospectively the clinical usefulness of systematic imatinib TDM in CML patients. This preliminary analysis addresses the appropriateness of the dosage adjustment approach applied in this study, which targets the recommended trough level and allows an interval of 4-24 h after last drug intake for blood sampling. Methods: Blood samples from the first 15 patients undergoing 1st TDM were obtained 1.5-25 h after last dose. Imatinib plasma levels were measured by LC-MS/MS and the concentrations were extrapolated to trough based on a Bayesian approach using a population pharmacokinetic model. Trough levels were predicted to differ significantly from the target in 12 patients (10 <750 μg/L; 2 >1500 μg/L along with poor tolerance) and individual dose adjustments were proposed. 8 patients underwent a 2nd TDM cycle. Trough levels of 1st and 2nd TDM were compared, the sample drawn 1.5 h after last dose (during distribution phase) was excluded from the analysis. Results: Individual dose adjustments were applied in 6 patients. Observed concentrations extrapolated to trough ranged from 360 to 1832 μg/L (median 725; mean 810, CV 52%) on 1st TDM and from 720 to 1187 μg/L (median 950; mean 940, CV 18%) on 2nd TDM cycle. Conclusions: These preliminary results suggest that TDM of imatinib using a Bayesian interpretation is able to target the recommended trough level of 1000 μg/L and to reduce the considerable differences in trough level exposure between patients (with CV decreasing from 52% to 18%). While this may simplify blood collection in daily practice, as samples do not have to be drawn exactly at trough, the largest possible interval to last drug intake yet remains preferable to avoid sampling during distribution phase leading to biased extrapolation. This encourages the evaluation of the clinical benefit of a routine TDM intervention in CML patients, which the randomized Swiss I-COME trial aims to.

A fast Bayesian reconstruction algorithm for emission tomography with entropy prior converging to feasible images

The development and tests of an iterative reconstruction algorithm for emission tomography based on Bayesian statistical concepts are described. The algorithm uses the entropy of the generated image as a prior distribution, can be accelerated by the choice of an exponent, and converges uniformly to feasible images by the choice of one adjustable parameter. A feasible image has been defined as one that is consistent with the initial data (i.e. it is an image that, if truly a source of radiation in a patient, could have generated the initial data by the Poisson process that governs radioactive disintegration). The fundamental ideas of Bayesian reconstruction are discussed, along with the use of an entropy prior with an adjustable contrast parameter, the use of likelihood with data increment parameters as conditional probability, and the development of the new fast maximum a posteriori with entropy (FMAPE) Algorithm by the successive substitution method. It is shown that in the maximum likelihood estimator (MLE) and FMAPE algorithms, the only correct choice of initial image for the iterative procedure in the absence of a priori knowledge about the image configuration is a uniform field.