Maximum-likelihood estimation of the random-clumped multinomial model as a prototype problem for large-scale statistical computing

Numerical methods are needed to obtain maximum-likelihood estimates (MLEs) in many problems. Computation time can be an issue for some likelihoods even with modern computing power. We consider one such problem where the assumed model is a random-clumped multinomial distribution. We compute MLEs for this model in parallel using the Toolkit for Advanced Optimization software library. The computations are performed on a distributed-memory cluster with low latency interconnect. We demonstrate that for larger problems, scaling the number of processes improves wall clock time significantly. An illustrative example shows how parallel MLE computation can be useful in a large data analysis. Our experience with a direct numerical approach indicates that more substantial gains may be obtained by making use of the specific structure of the random-clumped model.     

Simple subcontracting rules for make-to-order shops with limited subcontractor capacity: an assessment by simulation

Companies that produce a high variety of customised products often suffer from large fluctuations in demand. Subcontracting can be an important means of overcoming resulting capacity shortages during high demand periods. A set of effective subcontracting rules, for determining which jobs to subcontract and which to produce internally, has recently been presented in the literature for this type of company. But evaluations of these rules have assumed subcontractor capacity is infinite. This study examines the impact of limited subcontractor capacity on the performance of the three best-performing subcontracting rules for make-to-order companies using a simulation model of an assembly job shop. Limiting subcontractor capacity inhibits the ability of a subcontracting rule to protect the internal shop from surges in demand, which negatively affects performance. However, significant performance differences between the rules evaluated are maintained, which underlines the importance of choosing the right subcontracting rule. Further analysis reveals that a limit on the work that can be subcontracted leads to less work being subcontracted more often, which requires adequate response by management. Meanwhile, the assumption of infinite capacity results in sporadically subcontracting a large amount of work. The results have important implications for future research and practice.     

Workload Control and Order Release: A Lean Solution for Make‐to‐Order Companies

Protecting throughput from variance is the key to achieving lean. Workload control (WLC) accomplishes this in complex make‐to‐order job shops by controlling lead times, capacity, and work‐in‐process (WIP). However, the concept has been dismissed by many authors who believe its order release mechanism reduces the effectiveness of shop floor dispatching and increases work center idleness, thereby also increasing job tardiness results. We show that these problems have been overcome. A WLC order release method known as “LUMS OR” (Lancaster University Management School order release) combines continuous with periodic release, allowing the release of work to be triggered between periodic releases if a work center is starving. This paper refines the method based on the literature (creating “LUMS COR” [Lancaster University Management School corrected order release]) before comparing its performance against the best‐performing purely periodic and continuous release rules across a range of flow directions, from the pure job shop to the general flow shop. Results demonstrate that LUMS COR and the continuous WLC release methods consistently outperform purely periodic release and Constant WIP. LUMS COR is considered the best solution in practice due to its excellent performance and ease of implementation. Findings have significant implications for research and practice: throughput times and job tardiness results can be improved simultaneously and order release and dispatching rules can complement each other. Thus, WLC represents an effective means of implementing lean principles in a make‐to‐order context.     

Efficient Semiparametric Estimation of the Fama–French Model and Extensions

This paper develops a new estimation procedure for characteristic‐based factor models of stock returns. We treat the factor model as a weighted additive nonparametric regression model, with the factor returns serving as time‐varying weights and a set of univariate nonparametric functions relating security characteristic to the associated factor betas. We use a time‐series and cross‐sectional pooled weighted additive nonparametric regression methodology to simultaneously estimate the factor returns and characteristic‐beta functions. By avoiding the curse of dimensionality, our methodology allows for a larger number of factors than existing semiparametric methods. We apply the technique to the three‐factor Fama–French model, Carhart's four‐factor extension of it that adds a momentum factor, and a five‐factor extension that adds an own‐volatility factor. We find that momentum and own‐volatility factors are at least as important, if not more important, than size and value in explaining equity return comovements. We test the multifactor beta pricing theory against a general alternative using a new nonparametric test.     

Online signal extraction by robust regression in moving windows with data-adaptive width selection

Online (also ‘real-time’ or ‘sequential’) signal extraction from noisy and outlier-interfered data streams is a basic but challenging goal. Fitting a robust Repeated Median (Siegel in Biometrika 69:242–244, 1982) regression line in a moving time window has turned out to be a promising approach (Davies et al. in J. Stat. Plan. Inference 122:65–78, 2004; Gather et al. in Comput. Stat. 21:33–51, 2006; Schettlinger et al. in Biomed. Eng. 51:49–56, 2006). The level of the regression line at the rightmost window position, which equates to the current time point in an online application, is then used for signal extraction. However, the choice of the window width has a large impact on the signal extraction, and it is impossible to predetermine an optimal fixed window width for data streams which exhibit signal changes like level shifts and sudden trend changes. We therefore propose a robust test procedure for the online detection of such signal changes. An algorithm including the test allows for online window width adaption, meaning that the window width is chosen w.r.t. the current data situation at each time point. Comparison studies show that our new procedure outperforms an existing Repeated Median filter with automatic window width selection (Schettlinger et al. in Int. J. Adapt. Control Signal Process. 24:346–362, 2010).     

Multivariate Real-Time Signal Extraction by a Robust Adaptive Regression Filter

We propose a new regression-based filter for extracting signals online from multivariate high frequency time series. It separates relevant signals of several variables from noise and (multivariate) outliers.

Unlike parallel univariate filters, the new procedure takes into account the local covariance structure between the single time series components. It is based on high-breakdown estimates, which makes it robust against (patches of) outliers in one or several of the components as well as against outliers with respect to the multivariate covariance structure. Moreover, the trade-off problem between bias and variance for the optimal choice of the window width is approached by choosing the size of the window adaptively, depending on the current data situation.

Furthermore, we present an advanced algorithm of our filtering procedure that includes the replacement of missing observations in real time. Thus, the new procedure can be applied in online-monitoring practice. Applications to physiological time series from intensive care show the practical effect of the proposed filtering technique.     

Phthalate Exposure Through Food and Consumers' Risk Perception of Chemicals in Food

Phthalates have been detected in various types of retail foods. Consumers' exposure to phthalates is common. Consumers are concerned about chemicals in food. Our aim was to investigate the relationships between consumers' exposure to phthalates through food, consumers' interest in a natural and healthy diet, risk perception of food chemicals, and consumers' diet patterns. We collected data through a mail survey in the adult Swiss-German population ( N  = 1,200). We modeled exposure to di(2-ethylhexyl) phthalate (DEHP), dibutyl phthalate (DBP), benzyl butyl phthalate (BBP), and diethyl phthalate (DEP) based on a food frequency questionnaire and phthalate concentrations reported from food surveys. Using rating scales, we assessed risk perceptions of chemicals in food and interest in a natural and healthy diet. Higher risk perceptions and higher natural and healthy diet interest were associated with higher daily doses of DEHP, BBP, and DEP. No health risk from phthalates in food was identified for the vast majority of the population. Four consumers' diet clusters were discerned, with differences in phthalate exposure, risk perceptions, and interest in a natural and healthy diet. This study shows that even those consumers who express strong interest in natural food and low acceptance of food chemicals, and who try to make respective food choices, are exposed to contaminants such as phthalates.     

Overeducated Workers as an Insurance Device

Empirical findings show that firms frequently employ overeducated workers. Since overeducated workers earn more than adequately educated ones working in similar jobs these findings seem to be puzzling. In this paper, we introduce an insurance approach to explain the employment of overeducated workers. Referring to this approach, overeducated workers are employed by a firm to avoid high losses in the case of a crisis (e.g. when the production process breaks down). Contrary to adequately educated workers, overeducated ones may be helpful in this situation by quickly offering improvisatorial solutions. First, we use a simple model to demonstrate the insurance argument. Second, we test the major implication of the model empirically by using industry panel data: if and only if high‐skilled workers are employed for insurance purposes, the average wage of high‐skilled workers in firms that need insurance will be lower compared with firms that do not need insurance. The data confirm this theoretical result.     

DOES A GOOD CENTRAL BANKER MAKE A DIFFERENCE?

The personalities of central bankers moved center stage during the recent financial crisis. Indeed, several central bankers even became “superstars.” In this article, we investigate whether superstar central bankers have an impact on economic performance. We employ school grades given to central bankers by the financial press, defining as superstars those central bankers receiving the top grade. First, we explain the grades in a probit estimation with measures of economic performance, institutional features, and personal characteristics. Second, we employ a matching approach to account for the endogeneity of grading with respect to economic performance. Using entropy balancing, we identify credible counterfactuals for top‐graded central bankers, that is, nonsuperstar central bankers who face similar situations. Comparing the economic performance of both groups, we find that superstars do indeed matter: a top‐graded central banker faces a significantly more favorable output‐inflation tradeoff than his peers. This effect is driven by outstanding central bankers in both advanced and emerging economies and is especially prevalent in the precrisis subsample. (JEL E52, E58)     

The Field of American Media Sociology: Origins,Resurrection, and Consolidation

This contribution is written against the backdrop of the historic dispersal of early American media sociology out from the core concerns of the discipline and into various importer academic disciplines (including communication, journalism, and media studies) and an ever‐growing pervasiveness of media in everyday life which is reflected by a resurgence of sociological scholarship in the United States since the early 2000s. The article divides the field in works that study media inwards – along the threefold dimensions of production and technologies, communication and discourse, reception and effects – and works that study media outwards. We argue that this latter perspective, examining broader theoretical, methodological, and substantive social implications of mass‐mediated communication, is the most promising one for a mature field of American media sociology. On this basis, we conclude with some suggestions regarding possible new, and as of yet understudied, lines of inquiry for future media sociologists.