Regularized proportional odds models

The proportional odds model (POM) is commonly used in regression analysis to predict the outcome for an ordinal response variable. The maximum likelihood estimation (MLE) approach is typically used to obtain the parameter estimates. The likelihood estimates do not exist when the number of parameters, p, is greater than the number of observations n. The MLE also does not exist if there are no overlapping observations in the data. In a situation where the number of parameters is less than the sample size but p is approaching to n, the likelihood estimates may not exist, and if they exist they may have quite large standard errors. An estimation method is proposed to address the last two issues, i.e. complete separation and the case when p approaches n, but not the case when p>n. The proposed method does not use any penalty term but uses pseudo-observations to regularize the observed responses by downgrading their effect so that they become close to the underlying probabilities. The estimates can be computed easily with all commonly used statistical packages supporting the fitting of POMs with weights. Estimates are compared with MLE in a simulation study and an application to the real data.     

Effects of key enabling technologies for seru production on sustainable performance

Demand uncertainty, economic globalization, and environmental deterioration force factories to innovate their manufacturing systems for achieving sustainable performance. Seru production, which is the latest manufacturing mode developed in Japan, attracts broad attention from both academia and practitioners. The overwhelming majority of existing works on seru production focus on economic performance unilaterally, while neglecting the environmental and social performance. This paper investigates the effects of key enabling technologies for seru production on sustainable performance. Firstly, four key enabling technologies for seru production are identified through systematic review, and an evaluation indicator system of sustainable performance in the context of seru production is developed. Then, the hypotheses about the effects of the identified key enabling technologies for seru production on sustainable performance are proposed on the basis of previous research achievements, theoretical analysis, and practical observations. Finally, the hypotheses are tested through structural equation modeling. Except for two hypotheses which are not supported and one which is indirectly supported, all other hypotheses are verified. The research results show that the four key enabling technologies for seru production have different effects on the three dimensions of sustainable performance. The achievements of this work are of significance to improve the comprehensive understanding of seru production, as well as to develop practical methods to implement sustainable operations.     

Threshold Estimation for Stochastic Processes with Small Noise

Consider a process satisfying a stochastic differential equation with unknown drift parameter, and suppose that discrete observations are given. It is known that a simple least squares estimator (LSE) can be consistent but numerically unstable in the sense of large standard deviations under finite samples when the noise process has jumps. We propose a filter to cut large shocks from data and construct the same LSE from data selected by the filter. The proposed estimator can be asymptotically equivalent to the usual LSE, whose asymptotic distribution strongly depends on the noise process. However, in numerical study, it looked asymptotically normal in an example where filter was chosen suitably, and the noise was a Lévy process. We will try to justify this phenomenon mathematically, under certain restricted assumptions.     

On pseudo-values for regression analysis in competing risks models

For regression on state and transition probabilities in multi-state models Andersen et al. (Biometrika 90:15–27, 2003) propose a technique based on jackknife pseudo-values. In this article we analyze the pseudo-values suggested for competing risks models and prove some conjectures regarding their asymptotics (Klein and Andersen, Biometrics 61:223–229, 2005). The key is a second order von Mises expansion of the Aalen-Johansen estimator which yields an appropriate representation of the pseudo-values. The method is illustrated with data from a clinical study on total joint replacement. In the application we consider for comparison the estimates obtained with the Fine and Gray approach (J Am Stat Assoc 94:496–509, 1999) and also time-dependent solutions of pseudo-value regression equations.     

Faking it: Personality and individual difference predictors of willingness to buy counterfeit goods

Counterfeiting is now widely regarded as a serious social, economic, and political issue. This study examined demographic, personality, and individual difference predictors of willingness to buy counterfeit goods (WBCG) in a community sample of British adults. Two-hundred and thirty-seven participants completed a questionnaire that assessed their willingness to buy 19 types of counterfeit goods, attitudes towards counterfeiting, material values, Big Five personality traits, and demographics. Structural equation modelling showed that attitudes towards counterfeiting were the strongest predictors of WBCG. In addition, material values predicted both WBCG directly and indirectly through attitudes towards counterfeiting. Older participants showed lower WBCG, although this effect was moderated by participants’ conscientiousness, material values, and attitudes towards counterfeiting. There were no sex differences in WBCG once participants’ age and conscientiousness had been taken into account. These results are discussed in relation to the extant literature on the profiles of consumers who buy counterfeit products.     

Estimation of odds ratios under order restrictions

A new method for estimating a set of odds ratios under an order restriction based on estimating equations is proposed. The method is applied to those of the conditional maximum likelihood estimators and the Mantel-Haenszel estimators. The estimators derived from the conditional likelihood estimating equations are shown to maximize the conditional likelihoods. It is also seen that the restricted estimators converge almost surely to the respective odds ratios when the respective sample sizes become large regularly. The restricted estimators are compared with the unrestricted maximum likelihood estimators by a Monte Carlo simulation. The simulation studies show that the restricted estimates improve the mean squared errors remarkably, while the Mantel-Haenszel type estimates are competitive with the conditional maximum likelihood estimates, being slightly worse.     

Measurement Error Correction by Exploiting Gene–Environment Independence in Family‐Based Case–Control Studies

Abstract. Family‐based case–control designs are commonly used in epidemiological studies for evaluating the role of genetic susceptibility and environmental exposure to risk factors in the etiology of rare diseases. Within this framework, it is often reasonable to assume genetic susceptibility and environmental exposure being conditionally independent of each other within families in the source population. We focus on this setting to explore the situation of measurement error affecting the assessment of the environmental exposure. We correct for measurement error through a likelihood‐based method. We exploit a conditional likelihood approach to relate the probability of disease to the genetic and the environmental risk factors. We show that this approach provides less biased and more efficient results than that based on logistic regression. Regression calibration, instead, provides severely biased estimators of the parameters. The comparison of the correction methods is performed through simulation, under common measurement error structures.     

An additive subdistribution hazard model for competing risks data

The cumulative incidence function plays an important role in assessing its treatment and covariate effects with competing risks data. In this article, we consider an additive hazard model allowing the time-varying covariate effects for the subdistribution and propose the weighted estimating equation under the covariate-dependent censoring by fitting the Cox-type hazard model for the censoring distribution. When there exists some association between the censoring time and the covariates, the proposed coefficients’ estimations are unbiased and the large-sample properties are established. The finite-sample properties of the proposed estimators are examined in the simulation study. The proposed Cox-weighted method is applied to a competing risks dataset from a Hodgkin's disease study.     

Concept mapping as an approach for expert-guided model building: The example of health literacy

Concept mapping served as the starting point for the aim of capturing the comprehensive structure of the construct of ‘health literacy.’ Ideas about health literacy were generated by 99 experts and resulted in 105 statements that were subsequently organized by 27 experts in an unstructured card sorting. Multidimensional scaling was applied to the sorting data and a two and three-dimensional solution was computed. The three dimensional solution was used in subsequent cluster analysis and resulted in a concept map of nine “clusters”: (1) self-regulation, (2) self-perception, (3) proactive approach to health, (4) basic literacy and numeracy skills, (5) information appraisal, (6) information search, (7) health care system knowledge and acting, (8) communication and cooperation, and (9) beneficial personality traits. Subsequently, this concept map served as a starting point for developing a “qualitative” structural model of health literacy and a questionnaire for the measurement of health literacy. On the basis of questionnaire data, a “quantitative” structural model was created by first applying exploratory factor analyses (EFA) and then cross-validating the model with confirmatory factor analyses (CFA). Concept mapping proved to be a highly valuable tool for the process of model building up to translational research in the “real world”.     

Universal versus contextual effects on TQM: a triangulation study using neural networks

The objective of this study is to extend previous research on total quality management (TQM)-context-performance relationships and ‘fit’ using multiple methods. We combine artificial neural networks (ANNs) with structural equation modelling (SEM) to analyse several hypotheses and propositions. This is the first study in this area of research that utilises ANNs and a triangulation technique in the presence of several contextual factors. The SEM analyses suggest that company size and industry type may have contingency effects on some of the TQM practices and/or TQM-performance relationships. However, the ANN models have shown that these two contingency factors do not moderate TQM outcomes, implying that all organisations can benefit from TQM regardless of size and type. As well, these models show that formal TQM implementation and/or ISO certifications do not add any predictive power to the ANN models except in one case: TQM implementation and/or ISO certification added to organisational effectiveness and customer results to predict financial and market (F&M) results. The results further indicate that even though implementing TQM alone has a bigger impact on F&M results than obtaining ISO certification alone, combining the two will have an even greater impact on these results. Joint implementation leads to greater improvements in organisational effectiveness, which, in turn, has a positive effect on customer results and consequently F&M results. This is a unique finding within the context of moderator effects on TQM-performance relationships.