Use of posterior predictive assessments to evaluate model fit in multilevel logistic regression

Assessing the fit of a model is an important final step in any statistical analysis, but this is not straightforward when complex discrete response models are used. Cross validation and posterior predictions have been suggested as methods to aid model criticism. In this paper a comparison is made between four methods of model predictive assessment in the context of a three level logistic regression model for clinical mastitis in dairy cattle; cross validation, a prediction using the full posterior predictive distribution and two “mixed” predictive methods that incorporate higher level random effects simulated from the underlying model distribution. Cross validation is considered a gold standard method but is computationally intensive and thus a comparison is made between posterior predictive assessments and cross validation. The analyses revealed that mixed prediction methods produced results close to cross validation whilst the full posterior predictive assessment gave predictions that were over-optimistic (closer to the observed disease rates) compared with cross validation. A mixed prediction method that simulated random effects from both higher levels was best at identifying the outlying level two (farm-year) units of interest. It is concluded that this mixed prediction method, simulating random effects from both higher levels, is straightforward and may be of value in model criticism of multilevel logistic regression, a technique commonly used for animal health data with a hierarchical structure.

Residuals for log-Burr XII regression models in survival analysis

In this paper, we compare three residuals to assess departures from the error assumptions as well as to detect outlying observations in log-Burr XII regression models with censored observations. These residuals can also be used for the log-logistic regression model, which is a special case of the log-Burr XII regression model. For different parameter settings, sample sizes and censoring percentages, various simulation studies are performed and the empirical distribution of each residual is displayed and compared with the standard normal distribution. These studies suggest that the residual analysis usually performed in normal linear regression models can be straightforwardly extended to the modified martingale-type residual in log-Burr XII regression models with censored data.

The generalized log-gamma mixture model with covariates: local influence and residual analysis

In a sample of censored survival times, the presence of an immune proportion of individuals who are not subject to death, failure or relapse, may be indicated by a relatively high number of individuals with large censored survival times. In this paper the generalized log-gamma model is modified for the possibility that long-term survivors may be present in the data. The model attempts to separately estimate the effects of covariates on the surviving fraction, that is, the proportion of the population for which the event never occurs. The logistic function is used for the regression model of the surviving fraction. Inference for the model parameters is considered via maximum likelihood. Some influence methods, such as the local influence and total local influence of an individual are derived, analyzed and discussed. Finally, a data set from the medical area is analyzed under the log-gamma generalized mixture model. A residual analysis is performed in order to select an appropriate model.

Log-modified Weibull regression models with censored data: Sensitivity and residual analysis

This paper proposes a regression model considering the modified Weibull distribution. This distribution can be used to model bathtub-shaped failure rate functions. Assuming censored data, we consider maximum likelihood and Jackknife estimators for the parameters of the model. We derive the appropriate matrices for assessing local influence on the parameter estimates under different perturbation schemes and we also present some ways to perform global influence. Besides, for different parameter settings, sample sizes and censoring percentages, various simulations are performed and the empirical distribution of the modified deviance residual is displayed and compared with the standard normal distribution. These studies suggest that the residual analysis usually performed in normal linear regression models can be straightforwardly extended for a martingale-type residual in log-modified Weibull regression models with censored data. Finally, we analyze a real data set under log-modified Weibull regression models. A diagnostic analysis and a model checking based on the modified deviance residual are performed to select appropriate models. (c) 2008 Elsevier B.V. All rights reserved.

Regression models for grouped survival data: Estimation and sensitivity analysis

In this study, regression models are evaluated for grouped survival data when the effect of censoring time is considered in the model and the regression structure is modeled through four link functions. The methodology for grouped survival data is based on life tables, and the times are grouped in k intervals so that ties are eliminated. Thus, the data modeling is performed by considering the discrete models of lifetime regression. The model parameters are estimated by using the maximum likelihood and jackknife methods. To detect influential observations in the proposed models, diagnostic measures based on case deletion, which are denominated global influence, and influence measures based on small perturbations in the data or in the model, referred to as local influence, are used. In addition to those measures, the local influence and the total influential estimate are also employed. Various simulation studies are performed and compared to the performance of the four link functions of the regression models for grouped survival data for different parameter settings, sample sizes and numbers of intervals. Finally, a data set is analyzed by using the proposed regression models. (C) 2010 Elsevier B.V. All rights reserved.

Acute Kidney Injury after Pediatric Cardiac Surgery: Risk Factors and Outcomes. Proposal for a Predictive Model

Objectives: To characterize the epidemiology and risk factors for acute kidney injury (AKI) after pediatric cardiac surgery in our center, to determine its association with poor short-term outcomes, and to develop a logistic regression model that will predict the risk of AKI for the study population. Methods: This single-center, retrospective study included consecutive pediatric patients with congenital heart disease who underwent cardiac surgery between January 2010 and December 2012. Exclusion criteria were a history of renal disease, dialysis or renal transplantation. Results: Of the 325 patients included, median age three years (1 day---18 years), AKI occurred in 40 (12.3%) on the first postoperative day. Overall mortality was 13 (4%), nine of whom were in the AKI group. AKI was significantly associated with length of intensive care unit stay, length of mechanical ventilation and in-hospital death (p<0.01). Patients’ age and postoperative serum creatinine, blood urea nitrogen and lactate levels were included in the logistic regression model as predictor variables. The model accurately predicted AKI in this population, with a maximum combined sensitivity of 82.1% and specificity of 75.4%. Conclusions: AKI is common and is associated with poor short-term outcomes in this setting. Younger age and higher postoperative serum creatinine, blood urea nitrogen and lactate levels were powerful predictors of renal injury in this population. The proposed model could be a useful tool for risk stratification of these patients.

Credit default swap (CDS) prediction model & trading strategy

This project focuses on the study of different explanatory models for the behavior of CDS security, such as Fixed-Effect Model, GLS Random-Effect Model, Pooled OLS and Quantile Regression Model. After determining the best fitness model, trading strategies with long and short positions in CDS have been developed. Due to some specifications of CDS, I conclude that the quantile regression is the most efficient model to estimate the data. The P&L and Sharpe Ratio of the strategy are analyzed using a backtesting analogy, where I conclude that, mainly for non-financial companies, the model allows traders to take advantage of and profit from arbitrages.

Single-function approach to calibrating whatman n.º 42 filter paper based on suction versus water content relationships

Several suction–water-content (s-w) calibrations for the filter paper method (FPM) used for soil-suction measurement have been published. Most of the calibrations involve a bilinear function (i.e., two different equations) with an inflection point occurring at 60 kPasmooth transition between the high and low suctions based on a regression analysis of various previously published calibrations obtained for filter paper Whatman No. 42 (W42) is presented and discussed. The approach is applied herein to data obtained from three establish bilinear calibrations (six equations) for W42 filter paper to determine the two fitting parameters of the continuous function. An experimental evaluation of the new calibration show that the suctions estimated by the contact FPM test using the proposed function compare well with suctions measured by other laboratory techniques for two different soils for the suction range of 50 kPa

Association between LDL-C, Non HDL-C, and Apolipoprotein B Levels with Coronary Plaque Regression

Background:Previous reports have inferred a linear relationship between LDL-C and changes in coronary plaque volume (CPV) measured by intravascular ultrasound. However, these publications included a small number of studies and did not explore other lipid markers.Objective:To assess the association between changes in lipid markers and regression of CPV using published data.Methods:We collected data from the control, placebo and intervention arms in studies that compared the effect of lipidlowering treatments on CPV, and from the placebo and control arms in studies that tested drugs that did not affect lipids. Baseline and final measurements of plaque volume, expressed in mm3, were extracted and the percentage changes after the interventions were calculated. Performing three linear regression analyses, we assessed the relationship between percentage and absolute changes in lipid markers and percentage variations in CPV.Results:Twenty-seven studies were selected. Correlations between percentage changes in LDL-C, non-HDL-C, and apolipoprotein B (ApoB) and percentage changes in CPV were moderate (r = 0.48, r = 0.47, and r = 0.44, respectively). Correlations between absolute differences in LDL-C, non‑HDL-C, and ApoB with percentage differences in CPV were stronger (r = 0.57, r = 0.52, and r = 0.79). The linear regression model showed a statistically significant association between a reduction in lipid markers and regression of plaque volume.Conclusion:A significant association between changes in different atherogenic particles and regression of CPV was observed. The absolute reduction in ApoB showed the strongest correlation with coronary plaque regression.

Co-movements between US and UK stock prices: the roles of macroeconomic information and time-series varying conditional correlations

This paper provides evidence on the sources of co-movement in monthly US and UK stock price movements by investigating the role of macroeconomic and financial variables in a bivariate system with time-varying conditional correlations. Crosscountry communality in response is uncovered, with changes in the US Federal Funds rate, UK bond yields and oil prices having similar negative effects in both markets. Other variables also play a role, especially for the UK market. These effects do not, however, explain the marked increase in cross-market correlations observed from around 2000, which we attribute to time variation in the correlations of shocks to these markets. A regime-switching smooth transition model captures this time variation well and shows the correlations increase dramatically around 1999-2000. JEL classifications: C32, C51, G15 Keywords: international stock returns, DCC-GARCH model, smooth transition conditional correlation GARCH model, model evaluation.

A priori ratemaking using bivariate poisson regression models

In automobile insurance, it is useful to achieve a priori ratemaking by resorting to gene- ralized linear models, and here the Poisson regression model constitutes the most widely accepted basis. However, insurance companies distinguish between claims with or without bodily injuries, or claims with full or partial liability of the insured driver. This paper exa- mines an a priori ratemaking procedure when including two di®erent types of claim. When assuming independence between claim types, the premium can be obtained by summing the premiums for each type of guarantee and is dependent on the rating factors chosen. If the independence assumption is relaxed, then it is unclear as to how the tari® system might be a®ected. In order to answer this question, bivariate Poisson regression models, suitable for paired count data exhibiting correlation, are introduced. It is shown that the usual independence assumption is unrealistic here. These models are applied to an automobile insurance claims database containing 80,994 contracts belonging to a Spanish insurance company. Finally, the consequences for pure and loaded premiums when the independence assumption is relaxed by using a bivariate Poisson regression model are analysed.

Progress Towards to Equity Market Integration in Eastern Europe

The advent of the European Union has decreased the diversification benefits available from country based equity market indices in the region. This paper measures the increase in stock integration between the three largest new EU members (Hungary, the Czech Republic and Poland who joined in May 2004) and the Euro-zone. A potentially gradual transition in correlations is accommodated in a single VAR model by embedding smooth transition conditional correlation models with fat tails, spillovers, volatility clustering, and asymmetric volatility effects. At the country market index level all three Eastern European markets show a considerable increase in correlations in 2006. At the industry level the dates and transition periods for the correlations differ, and the correlations are lower although also increasing. The results show that sectoral indices in Eastern European markets may provide larger diversification opportunities than the aggregate market. JEL classifications: C32; C51; F36; G15 Keywords: Multivariate GARCH; Smooth Transition Conditional Correlation; Stock Return Comovement; Sectoral correlations; New EU Members

Equilibrium Exchange Rate Determination and Multiple Structural Changes

The large appreciation and depreciation of the US dollar in the 1980s stimulated an important debate on the usefulness of unit root tests in the presence of structural breaks. In this paper, we propose a simple model to describe the evolution of the real exchange rate. We then propose a more general smooth transition (STR) function than has hitherto been employed, which is able to capture structural changes along the (long-run) equilibrium path, and show that this is consistent with our economic model. Our framework allows for a gradual adjustment between regimes and allows for under- and/or over-valued exchange rate adjustments. Using monthly and quarterly data for up to twenty OECD countries, we apply our methodology to investigate the univariate time series properties of CPI-based real exchange rates with both the U.S. dollar and German mark as the numeraire currencies. The empirical results show that, for more than half of the quarterly series, the evidence in favour of the stationarity of the real exchange rate was clearer in the sub-sample period post-1980.

A general model to predict individual exposure to solar UV by using ambient irradiance data

Excessive exposure to solar ultraviolet (UV) is the main cause of skin cancer. Specific prevention should be further developed to target overexposed or highly vulnerable populations. A better characterisation of anatomical UV exposure patterns is however needed for specific prevention. To develop a regression model for predicting the UV exposure ratio (ER, ratio between the anatomical dose and the corresponding ground level dose) for each body site without requiring individual measurements. A 3D numeric model (SimUVEx) was used to compute ER for various body sites and postures. A multiple fractional polynomial regression analysis was performed to identify predictors of ER. The regression model used simulation data and its performance was tested on an independent data set. Two input variables were sufficient to explain ER: the cosine of the maximal daily solar zenith angle and the fraction of the sky visible from the body site. The regression model was in good agreement with the simulated data ER (R(2)=0.988). Relative errors up to +20% and -10% were found in daily doses predictions, whereas an average relative error of only 2.4% (-0.03% to 5.4%) was found in yearly dose predictions. The regression model predicts accurately ER and UV doses on the basis of readily available data such as global UV erythemal irradiance measured at ground surface stations or inferred from satellite information. It renders the development of exposure data on a wide temporal and geographical scale possible and opens broad perspectives for epidemiological studies and skin cancer prevention.

Multivariate contemporaneous-threshold autoregressive models

This paper proposes a contemporaneous-threshold multivariate smooth transition autoregressive (C-MSTAR) model in which the regime weights depend on the ex ante probabilities that latent regime-specific variables exceed certain threshold values. A key feature of the model is that the transition function depends on all the parameters of the model as well as on the data. Since the mixing weights are also a function of the regime-specific innovation covariance matrix, the model can account for contemporaneous regime-specific co-movements of the variables. The stability and distributional properties of the proposed model are discussed, as well as issues of estimation, testing and forecasting. The practical usefulness of the C-MSTAR model is illustrated by examining the relationship between US stock prices and interest rates.