Applied Stochastic Models in Business and Industry

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.723

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.725

Recent reforms intended to promote more accountable and responsive government have increased public attention to performance analysis and accelerated the production and use of information on agency performance and public program outcomes. Drawing from cases and empirical studies, this presentation considers questions about what should count as evidence, how it should be communicated, who should judge the quality and reliability of evidence and performance information, and how to achieve a balance between processes that produce rigorous information for decision making and those that foster democratic governance and accountability. Promising directions are suggested for efforts to improve government effectiveness through the use of more rigorous information in decision making, along with acknowledgment of the limitations and risks associated with such efforts. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.726

Random assignment experiments are discussed by drawing parallels between issues in performance management studies and in clinical trials. In addition, the need for statistical rigour and for measures of uncertainty in performance management tools is highlighted. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.724

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.732

Quantile regression is an emerging modelling technique; we examine an approach allowing this technique to model binomial variables in a Bayesian framework and illustrate the value of this advanced technique on a set of local government performance indicators from England and Wales. In U.K. local government, there is currently particular interest in assessing performance relative to ‘top’ and ‘bottom’ quartiles; all authorities are expected to match the current best quartile performance within 5 years, any authority in the ‘bottom’ quartile is assumed to be significantly below par. By its very nature, quantile regression lets us to explore relationships between various covariates and these particular levels of performance. Additionally, by examining a number of other percentiles, we demonstrate how quantile regression gives a much fuller insight into the apparent behaviour of the system we are modelling. Rather than relying on asymptotic results, we use Bayesian methods that allow us to explore the uncertainty implicit in our model building and predictions. We suggest that this is most important when analysing data that are used to make managerial and administrative decisions. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.727

Benchmarking plays a relevant role in performance analysis, and statistical methods can be fruitfully exploited for its aims. While clustering, regression, and frontier analysis may serve some benchmarking purposes, we propose to consider archetypal analysis as a suitable technique. Archetypes are extreme points that synthesize data and that, in our opinion, can be profitably used as benchmarks. That is, they may be viewed as key reference performers in the comparison process. We suggest a three‐step data driven benchmarking procedure, which enables users: (i) to identify some reference performers, (ii) to analyze their features, (iii) to compare observed performers with them. An exploratory point of view is preferred, and graphical devices are adopted throughout the procedure. Finally, our approach is presented by means of an illustrative example based on The Times league table of the world top 200 universities. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.728

Structural equation models (SEMs) make it possible to estimate the causal relationships, defined according to a theoretical model, linking two or more latent complex concepts, each measured through a number of observable indicators, usually called manifest variables. Traditionally, the component‐based estimation of SEMs by means of partial least squares (PLS path modelling, PLS‐PM) assumes homogeneity over the observed set of units: all units are supposed to be well represented by a unique model estimated on the overall data set. In many cases, however, it is reasonable to expect classes made of units showing heterogeneous behaviours to exist. Two different kinds of heterogeneity could be affecting the data: observed and unobserved heterogeneity. The first refers to the case of a priori existing classes, whereas in unobserved heterogeneity no information is available either on the number of classes or on their composition. If a group structure for the statistical units is given, the aim of the analysis is to search for any differences in the behaviours of the a priori given classes. In PLS‐PM this would mean studying the effect of directly observed moderating variables, i.e. estimating as many (local) models as there are classes. Unobserved heterogeneity, instead, implies identifying classes of units (a priori unknown) having similar behaviours. Such heterogeneity is captured by an unobserved (latent) discrete moderating variable defining both the number of classes and the class membership. A new method for unobserved heterogeneity detection in PLS‐PM is proposed in this paper: response‐based procedure for detecting unit segments in PLS‐PM (REBUS‐PLS). REBUS‐PLS, according to PLS‐PM features, does not require distributional hypotheses and may lead to local models that are different in terms of both structural and measurement models. An application of REBUS‐PLS on real data will be shown. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.730

In this paper we use the Rasch model (RM) as a tool to measure the perceived quality of student services at the Reception Office of an Italian university faculty. This paper has both a substantive and a methodological aim. The former is concerned with measuring the service quality of the Reception Office, while the latter concerns the definition and validation of an instrument for measuring perceived quality. The sample comprised 273 students enrolled at the Faculty of Economics at the University of Palermo (Aiello F. Il modello di Rasch per la costruzione di uno strumento di misura della qualità di un Servizio. Ph.D. Thesis, University of Palermo, Palermo, 2005). The RM is applied to produce specific measurements of the perceived quality for each service feature in order to conduct a critical analysis of the results. We sought to verify some important assumptions and desiderata of the RM: (i) Unidimensionality: in the user's opinion, are the single features of the service ‘good’ at defining the whole construct of global quality? (ii) Targeting: is the range of the measurements provided by the RM for each item able to cover all the possible satisfaction levels required by the students? (iii) Measurement of the perceived quality of each feature: is it possible to rank the items according to their level of perceived quality (which may be a good way of identifying the weaknesses of the service in order to allocate additional resources to make due adjustments)? (iv) Item separation: are the service features (items) redundant with respect to the satisfaction levels required by the students? (v) Differential item functioning: is the item‐difficulty hierarchy invariant across the student factors? In our opinion the last issue was of great interest as it can identify assessment differences. Indeed, there appeared to emerge a different severity of judgment towards the single features of the service from the students, according to their length of stay in the system. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.729

Unlike the value at risk, the expected shortfall is a coherent measure of risk. In this paper, we discuss estimation of the expected shortfall of a random variable Yt with special reference to the case when auxiliary information is available in the form of a set of predictors Xt. We consider three classes of estimators of the conditional expected shortfall of Yt given Xt: a class of fully non‐parametric estimators and two classes of analog estimators based, respectively, on the empirical conditional quantile function and the empirical conditional distribution function. We study their sampling properties by means of a set of Monte Carlo experiments and analyze their performance in an empirical application to financial data. Copyright © 2008 John Wiley & Sons, Ltd.

D'Esposito, Maria Rosaria; Tenenhaus, M.

2008 Applied Stochastic Models in Business and Industry

doi: 10.1002/asmb.731

We test the underinvestment hypothesis (the so‐called Furubotn–Pejovich effect) from the specialized literature on co‐ops by comparing the shadow price of capital and the dual capacity utilization index for a panel of Italian cooperative and conventional firms, 1996–2003. The results do not show any difference between co‐ops and conventional firms in this respect. Copyright © 2008 John Wiley & Sons, Ltd.