On the question of proportionality of the count of observed Scrapie cases and the size of holding

Background: The present paper investigates the question of a suitable basic model for the number of scrapie cases in a holding and applications of this knowledge to the estimation of scrapie-ffected holding population sizes and adequacy of control measures within holding. Is the number of scrapie cases proportional to the size of the holding in which case it should be incorporated into the parameter of the error distribution for the scrapie counts? Or, is there a different - potentially more complex - relationship between case count and holding size in which case the information about the size of the holding should be better incorporated as a covariate in the modeling? Methods: We show that this question can be appropriately addressed via a simple zero-truncated Poisson model in which the hypothesis of proportionality enters as a special offset-model. Model comparisons can be achieved by means of likelihood ratio testing. The procedure is illustrated by means of surveillance data on classical scrapie in Great Britain. Furthermore, the model with the best fit is used to estimate the size of the scrapie-affected holding population in Great Britain by means of two capture-recapture estimators: the Poisson estimator and the generalized Zelterman estimator. Results: No evidence could be found for the hypothesis of proportionality. In fact, there is some evidence that this relationship follows a curved line which increases for small holdings up to a maximum after which it declines again. Furthermore, it is pointed out how crucial the correct model choice is when applied to capture-recapture estimation on the basis of zero-truncated Poisson models as well as on the basis of the generalized Zelterman estimator. Estimators based on the proportionality model return very different and unreasonable estimates for the population sizes. Conclusion: Our results stress the importance of an adequate modelling approach to the association between holding size and the number of cases of classical scrapie within holding. Reporting artefacts and speculative biological effects are hypothesized as the underlying causes of the observed curved relationship. The lack of adjustment for these artefacts might well render ineffective the current strategies for the control of the disease.

PopABC: a program to infer historical demographic parameters

PopABC is a computer package for inferring the pattern of demographic divergence of closely related populations and species. The software performs coalescent simulation in the framework of approximate Bayesian computation (ABC). PopABC can also be used to perform Bayesian model choice to discriminate between different demographic scenarios. The program can be used either for research or for education and teaching purposes.

Statistical inferences in phylogeography

In conventional phylogeographic studies, historical demographic processes are elucidated from the geographical distribution of individuals represented on an inferred gene tree. However, the interpretation of gene trees in this context can be difficult as the same demographic/geographical process can randomly lead to multiple different genealogies. Likewise, the same gene trees can arise under different demographic models. This problem has led to the emergence of many statistical methods for making phylogeographic inferences. A popular phylogeographic approach based on nested clade analysis is challenged by the fact that a certain amount of the interpretation of the data is left to the subjective choices of the user, and it has been argued that the method performs poorly in simulation studies. More rigorous statistical methods based on coalescence theory have been developed. However, these methods may also be challenged by computational problems or poor model choice. In this review, we will describe the development of statistical methods in phylogeographic analysis, and discuss some of the challenges facing these methods.

abctools: an R package for tuning approximate Bayesian computation analyses

Approximate Bayesian computation (ABC) is a popular family of algorithms which perform approximate parameter inference when numerical evaluation of the likelihood function is not possible but data can be simulated from the model. They return a sample of parameter values which produce simulations close to the observed dataset. A standard approach is to reduce the simulated and observed datasets to vectors of summary statistics and accept when the difference between these is below a specified threshold. ABC can also be adapted to perform model choice. In this article, we present a new software package for R, abctools which provides methods for tuning ABC algorithms. This includes recent dimension reduction algorithms to tune the choice of summary statistics, and coverage methods to tune the choice of threshold. We provide several illustrations of these routines on applications taken from the ABC literature.

Sensitivity of simulated regional Arctic climate to the choice of coupled model domain

The climate over the Arctic has undergone changes in recent decades. In order to evaluate the coupled response of the Arctic system to external and internal forcing, our study focuses on the estimation of regional climate variability and its dependence on large-scale atmospheric and regional ocean circulations. A global ocean–sea ice model with regionally high horizontal resolution is coupled to an atmospheric regional model and global terrestrial hydrology model. This way of coupling divides the global ocean model setup into two different domains: one coupled, where the ocean and the atmosphere are interacting, and one uncoupled, where the ocean model is driven by prescribed atmospheric forcing and runs in a so-called stand-alone mode. Therefore, selecting a specific area for the regional atmosphere implies that the ocean–atmosphere system can develop ‘freely’ in that area, whereas for the rest of the global ocean, the circulation is driven by prescribed atmospheric forcing without any feedbacks. Five different coupled setups are chosen for ensemble simulations. The choice of the coupled domains was done to estimate the influences of the Subtropical Atlantic, Eurasian and North Pacific regions on northern North Atlantic and Arctic climate. Our simulations show that the regional coupled ocean–atmosphere model is sensitive to the choice of the modelled area. The different model configurations reproduce differently both the mean climate and its variability. Only two out of five model setups were able to reproduce the Arctic climate as observed under recent climate conditions (ERA-40 Reanalysis). Evidence is found that the main source of uncertainty for Arctic climate variability and its predictability is the North Pacific. The prescription of North Pacific conditions in the regional model leads to significant correlation with observations, even if the whole North Atlantic is within the coupled model domain. However, the inclusion of the North Pacific area into the coupled system drastically changes the Arctic climate variability to a point where the Arctic Oscillation becomes an ‘internal mode’ of variability and correlations of year-to-year variability with observational data vanish. In line with previous studies, our simulations provide evidence that Arctic sea ice export is mainly due to ‘internal variability’ within the Arctic region. We conclude that the choice of model domains should be based on physical knowledge of the atmospheric and oceanic processes and not on ‘geographic’ reasons. This is particularly the case for areas like the Arctic, which has very complex feedbacks between components of the regional climate system.

Performance and prediction: Bayesian modelling of fallible choice in chess

Evaluating agents in decision-making applications requires assessing their skill and predicting their behaviour. Both are well developed in Poker-like situations, but less so in more complex game and model domains. This paper addresses both tasks by using Bayesian inference in a benchmark space of reference agents. The concepts are explained and demonstrated using the game of chess but the model applies generically to any domain with quantifiable options and fallible choice. Demonstration applications address questions frequently asked by the chess community regarding the stability of the rating scale, the comparison of players of different eras and/or leagues, and controversial incidents possibly involving fraud. The last include alleged under-performance, fabrication of tournament results, and clandestine use of computer advice during competition. Beyond the model world of games, the aim is to improve fallible human performance in complex, high-value tasks.

Diagnosis of variability and trends in a global precipitation dataset using a physically motivated statistical model

A physically motivated statistical model is used to diagnose variability and trends in wintertime ( October - March) Global Precipitation Climatology Project (GPCP) pentad (5-day mean) precipitation. Quasi-geostrophic theory suggests that extratropical precipitation amounts should depend multiplicatively on the pressure gradient, saturation specific humidity, and the meridional temperature gradient. This physical insight has been used to guide the development of a suitable statistical model for precipitation using a mixture of generalized linear models: a logistic model for the binary occurrence of precipitation and a Gamma distribution model for the wet day precipitation amount. The statistical model allows for the investigation of the role of each factor in determining variations and long-term trends. Saturation specific humidity q(s) has a generally negative effect on global precipitation occurrence and with the tropical wet pentad precipitation amount, but has a positive relationship with the pentad precipitation amount at mid- and high latitudes. The North Atlantic Oscillation, a proxy for the meridional temperature gradient, is also found to have a statistically significant positive effect on precipitation over much of the Atlantic region. Residual time trends in wet pentad precipitation are extremely sensitive to the choice of the wet pentad threshold because of increasing trends in low-amplitude precipitation pentads; too low a choice of threshold can lead to a spurious decreasing trend in wet pentad precipitation amounts. However, for not too small thresholds, it is found that the meridional temperature gradient is an important factor for explaining part of the long-term trend in Atlantic precipitation.

Initial results from Phase 2 of the international urban energy balance model comparison

Urban land surface schemes have been developed to model the distinct features of the urban surface and the associated energy exchange processes. These models have been developed for a range of purposes and make different assumptions related to the inclusion and representation of the relevant processes. Here, the first results of Phase 2 from an international comparison project to evaluate 32 urban land surface schemes are presented. This is the first large-scale systematic evaluation of these models. In four stages, participants were given increasingly detailed information about an urban site for which urban fluxes were directly observed. At each stage, each group returned their models' calculated surface energy balance fluxes. Wide variations are evident in the performance of the models for individual fluxes. No individual model performs best for all fluxes. Providing additional information about the surface generally results in better performance. However, there is clear evidence that poor choice of parameter values can cause a large drop in performance for models that otherwise perform well. As many models do not perform well across all fluxes, there is need for caution in their application, and users should be aware of the implications for applications and decision making.

Dichotomous-choice contingent valuation with 'don't know' responses and misreporting

A new approach is presented that simultaneously deals with Misreporting and Don't Know (DK) responses within a dichotomous-choice contingent valuation framework. Utilising a modification of the standard Bayesian Probit framework, a Gibbs with Metropolis-Hastings algorithm is used to estimate the posterior densities for the parameters of interest. Several model specifications are applied to two contingent valuation datasets: one on wolf management plans, and one on the US Fee Demonstration Program. We find that DKs are more likely to be from people who would be predicted to have positive utility for the bid. Therefore, a DK is more likely to be a YES than a NO. We also find evidence of misreporting, primarily in favour of the NO option. The inclusion of DK responses has an unpredictable impact on willingness-to-pay estimates, since it impacts differently on the results for the two datasets we examine. Copyright (C) 2009 John Wiley & Sons, Ltd.

Model selection for the mixed logit with Bayesian estimation

In this paper, the mixed logit (ML) using Bayesian methods was employed to examine willingness-to-pay (WTP) to consume bread produced with reduced levels of pesticides so as to ameliorate environmental quality, from data generated by a choice experiment. Model comparison used the marginal likelihood, which is preferable for Bayesian model comparison and testing. Models containing constant and random parameters for a number of distributions were considered, along with models in ‘preference space’ and ‘WTP space’ as well as those allowing for misreporting. We found: strong support for the ML estimated in WTP space; little support for fixing the price coefficient a common practice advocated and adopted in the environmental economics literature; and, weak evidence for misreporting.

Mixed Logit Model Performance and Distributional Assumptions: Preferences and GM foods

Using mixed logit models to analyse choice data is common but requires ex ante specification of the functional forms of preference distributions. We make the case for greater use of bounded functional forms and propose the use of the Marginal Likelihood, calculated using Bayesian techniques, as a single measure of model performance across non nested mixed logit specifications. Using this measure leads to very different rankings of model specifications compared to alternative rule of thumb measures. The approach is illustrated using data from a choice experiment regarding GM food types which provides insights regarding the recent WTO dispute between the EU and the US, Canada and Argentina and whether labelling and trade regimes should be based on the production process or product composition.

Statistical methods for ordered categorical data based on a constrained odds model

The proportional odds model provides a powerful tool for analysing ordered categorical data and setting sample size, although for many clinical trials its validity is questionable. The purpose of this paper is to present a new class of constrained odds models which includes the proportional odds model. The efficient score and Fisher's information are derived from the profile likelihood for the constrained odds model. These results are new even for the special case of proportional odds where the resulting statistics define the Mann-Whitney test. A strategy is described involving selecting one of these models in advance, requiring assumptions as strong as those underlying proportional odds, but allowing a choice of such models. The accuracy of the new procedure and its power are evaluated.

Diagnosis of variability and trends in a global precipitation dataset using a physically motivated statistical model

A physically motivated statistical model is used to diagnose variability and trends in wintertime ( October - March) Global Precipitation Climatology Project (GPCP) pentad (5-day mean) precipitation. Quasi-geostrophic theory suggests that extratropical precipitation amounts should depend multiplicatively on the pressure gradient, saturation specific humidity, and the meridional temperature gradient. This physical insight has been used to guide the development of a suitable statistical model for precipitation using a mixture of generalized linear models: a logistic model for the binary occurrence of precipitation and a Gamma distribution model for the wet day precipitation amount. The statistical model allows for the investigation of the role of each factor in determining variations and long-term trends. Saturation specific humidity q(s) has a generally negative effect on global precipitation occurrence and with the tropical wet pentad precipitation amount, but has a positive relationship with the pentad precipitation amount at mid- and high latitudes. The North Atlantic Oscillation, a proxy for the meridional temperature gradient, is also found to have a statistically significant positive effect on precipitation over much of the Atlantic region. Residual time trends in wet pentad precipitation are extremely sensitive to the choice of the wet pentad threshold because of increasing trends in low-amplitude precipitation pentads; too low a choice of threshold can lead to a spurious decreasing trend in wet pentad precipitation amounts. However, for not too small thresholds, it is found that the meridional temperature gradient is an important factor for explaining part of the long-term trend in Atlantic precipitation.

Backward elimination model construction for regression and classification using leave-one-out criteria

A fundamental principle in practical nonlinear data modeling is the parsimonious principle of constructing the minimal model that explains the training data well. Leave-one-out (LOO) cross validation is often used to estimate generalization errors by choosing amongst different network architectures (M. Stone, "Cross validatory choice and assessment of statistical predictions", J. R. Stast. Soc., Ser. B, 36, pp. 117-147, 1974). Based upon the minimization of LOO criteria of either the mean squares of LOO errors or the LOO misclassification rate respectively, we present two backward elimination algorithms as model post-processing procedures for regression and classification problems. The proposed backward elimination procedures exploit an orthogonalization procedure to enable the orthogonality between the subspace as spanned by the pruned model and the deleted regressor. Subsequently, it is shown that the LOO criteria used in both algorithms can be calculated via some analytic recursive formula, as derived in this contribution, without actually splitting the estimation data set so as to reduce computational expense. Compared to most other model construction methods, the proposed algorithms are advantageous in several aspects; (i) There are no tuning parameters to be optimized through an extra validation data set; (ii) The procedure is fully automatic without an additional stopping criteria; and (iii) The model structure selection is directly based on model generalization performance. The illustrative examples on regression and classification are used to demonstrate that the proposed algorithms are viable post-processing methods to prune a model to gain extra sparsity and improved generalization.