Finite mixtures of quantile and M-quantile regression models

In this paper we define a finite mixture of quantile and M-quantile regression models for heterogeneous and /or for dependent/clustered data. Components of the finite mixture represent clusters of individuals with homogeneous values of model parameters. For its flexibility and ease of estimation, the proposed approaches can be extended to random coefficients with a higher dimension than the simple random intercept case. Estimation of model parameters is obtained through maximum likelihood, by implementing an EM-type algorithm. The standard error estimates for model parameters are obtained using the inverse of the observed information matrix, derived through the Oakes (J R Stat Soc Ser B 61:479–482, 1999) formula in the M-quantile setting, and through nonparametric bootstrap in the quantile case. We present a large scale simulation study to analyse the practical behaviour of the proposed model and to evaluate the empirical performance of the proposed standard error estimates for model parameters. We considered a variety of empirical settings in both the random intercept and the random coefficient case. The proposed modelling approaches are also applied to two well-known datasets which give further insights on their empirical behaviour.     

Market research with the aid of a smartphone application – a case study

We discuss using the presumed ‘wisdom of the crowd’ to reduce the bullwhip effect in supply chains by enhancing the accuracy of demand forecast. Our case study describes a joint project of a leading European technology company (We chose RAGD as a pseudonym for the company throughout the study.) and the University of Kassel on a smartphone application for end customers to gather early information for this producer’s forecast. Our results confirm the ‘wisdom of the crowd’ hypothesis: A group of experts, in our case technicians who install and maintain RAGD-products, is capable of anticipating market fluctuations six weeks in advance. This only holds true if the ‘crowd’ of technicians is large enough. Our business climate index outperforms the company forecast in the first six months, when on average 22 technicians took part per week. We discuss successes and limitations of the cooperation and provide recommendations for similar projects.     

Strategic monitoring of port authorities activities: Proposal of a multi-dimensional digital dashboard

Abstract

The consequence of port evolution is that port authorities (PA) around the world are modifying their nature and their role, acquiring more and more an active role in the governance of logistic systems and are often adopting managerial and entrepreneurial behaviours. This new role of PA requires the adoption of a complex approach to monitoring the performance of the single stakeholders and of the whole port. Existing studies on port performance consider only specific indicators of port performance such as operational and financial efficiency neglecting aspects like customer satisfaction, port security and port sustainability. This paper tries to fill this literature gap: (i) outlining an integrated multi-dimensional dashboard to monitor port performance in a strategic perspective; (ii) analysing the role of digital technology to support the PA in the implementation of the monitoring dashboard; (iii) defining the evolutionary stages of tools implementation in the strategic monitoring of PAs.     

Sleepiness,choice consistency,and risk preferences

We investigate the consistency and stability of individual risk preferences by manipulating cognitive resources. Participants are randomly assigned to an experiment session at a preferred time of day relative to their diurnal preference (circadian matched) or at a non-preferred time (circadian mismatched) and choose allocations between two risky assets [using the Choi et al. (Am Econ Rev 27(5):1921–1938, 2007), design]. We find that choices of circadian matched and mismatched subject are statistically similar in terms of satisfying basic requirements for preference consistency. However, mismatched subjects tend to choose riskier asset bundles.     

The relationship of maternal work characteristics to childcare type and quality in rural communities

Drawing on data from the Family Life Project collected in North Carolina and Pennsylvania, this paper examines the relationship between maternal work characteristics and childcare type and quality in rural communities. Research is limited on the childcare experiences of rural families. Rural areas have less access to formal childcare and families often commute long distances for work, restricting childcare options. Employed mothers using childcare were selected (n=441). Logistic and OLS regression was used to examine which characteristics, including workplace support, objective occupational measures, hours, wage, and shift, predicted care type and quality. Results indicated that most families were using informal care. Those with more hazardous work conditions and working night shifts were less likely to use centers. Higher quality care was related to more workplace support, center use, and higher wages. Implications for social policy and practice are discussed.     

An Integrated Approach to Managing Change

Complex systems often involve a huge number of variables whose interdependence must be taken into account. This makes the adoption of an integrated approach essential for the proper management of a wide range of aspects related to human life, from social matters to the environment. In particular, this paper takes into consideration three different experiences in which the use of an integrated approach represented the most effective response for managing the evolving realities we took into consideration. We focused on the issue of land management, with particular attention to the problem of monoculture in Brazil. Then we analyzed the attempted cooperation between the States of the Aral Sea area for efficient water management of the basin and, finally, we dealt with the matter of energy efficiency and how the production of home integrated systems could lead to a better rationalization of energy consumption in the near future.      

On boosting kernel regression

In this paper we propose a simple multistep regression smoother which is constructed in an iterative manner, by learning the Nadaraya–Watson estimator with _L2$L_2$ boosting. We find, in both theoretical analysis and simulation experiments, that the bias converges exponentially fast, and the variance diverges exponentially slow. The first boosting step is analysed in more detail, giving asymptotic expressions as functions of the smoothing parameter, and relationships with previous work are explored. Practical performance is illustrated by both simulated and real data.     

Metabolic impact of shift work

In developing countries, shift work represents a considerable contingent workforce. Recently, studies have shown that overweight and obesity are more prevalent in shift workers than day workers. In addition, shift work has been associated with a higher propensity for the development of many metabolic disorders, such as insulin resistance, diabetes, dislipidemias and metabolic syndrome. Recent data have pointed that decrease of the sleep time, desynchronization of circadian rhythm and alteration of environmental aspects are the main factors related to such problems. Shortened or disturbed sleep is among the most common health-related effects of shift work. The plausible physiological and biological mechanisms are related to the activation of the autonomic nervous system, inflammation, changes in lipid and glucose metabolism, and related changes in the risk for atherosclerosis, metabolic syndrome, and type II diabetes. The present review will discuss the impact of shift work on obesity and metabolic disorders and how disruption of sleep and circadian misalignment may contribute to these metabolic dysfunctions.     

Finding disturbances in on-farm biogas production

When implementing innovations, disturbances are very likely to take place. Disturbances are undesirable because they can lead to unwanted outcomes, such as economic losses and work overload to workers. However, they can be powerful opportunities for learning and re-designing innovations. Here, we will present activity theoretical tools for analyzing disturbances in a way that they could be used as learning opportunities. We illustrate the proposed tools by analyzing a disturbance that took place during the implementation of a project of biogas production. By interpreting the disturbance process with a network of activity systems, we found that on-farm disturbances were formed as ruptures, innovations and asynchronies originated in other activity systems. This finding suggests that disturbances are outcomes of the functioning of networks, rather than simple results of failure of individuals or technical devices. The proposed tools could be used in interventions to help practitioners and ergonomists to recognize the systemic and networked nature of problems, and therefore, realize that they may require the collaboration of actors from different activities. In this sense, disturbances may be turned into opportunities for learning and developing innovations. We conclude by discussing how the method could be used in ergonomic design and intervention.     

Random coefficient models for binary longitudinal responses with attrition

We extend the approach introduced by Aitkin and Alfò (1998, Statistics and Computing, 4, pp. 289–307) to the general framework of random coefficient models and propose a class of conditional models to deal with binary longitudinal responses, including unknown sources of heterogeneity in the regression parameters as well as serial dependence of Markovian form.Furthermore, we discuss the extension of the proposed approach to the analysis of informative drop-outs, which represent a central problem in longitudinal studies, and define, as suggested by Follmann and Wu (1995, Biometrics, 51, pp. 151–168), a conditional specification of the full shared parameter model for the primary response and the missingness indicator. The model is applied to a dataset from a methadone maintenance treatment programme held in Sydney in 1986 and previously analysed by Chan et al. (1998, Australian & New Zealand Journal of Statistics, 40, pp. 1–10).All of the proposed models are estimated by means of an EM algorithm for nonparametric maximum likelihood, without assuming any specific parametric distribution for the random coefficients and for the drop-out process.A small scale simulation work is described to explore the behaviour of the extended approach in a number of different situations where informative drop-outs are present.