STATISTICA & APPLICAZIONI

The statistical techniques that are applied to the analysis and prediction of the failure of business enterprises include regression modelling of hazard functions. Because a large number of financial variables are available as potential  predictors, it is necessary to choose an appropriate variable selection procedure in this modelling. Furthermore, alternative models, for cause-specific and subdistribution hazards, are available when different modes of failure are  being examined as competing risks. We demonstrate the application of both types of hazards to the study of failures from bankruptcy, liquidation or dissolution among 75479 manufacturing firms in seven European countries in  2000-2018, and we examine the results of variable selection by stepwise and lasso methods for both approaches. All analyses resulted in successful prediction, with areas under the ROC curve of 0.90 or just below, but the lasso  approach achieved this with a smaller number of variables selected.

A Comparison of Variable Selection Methods in Competing Risks Models for Business Failures by Francesca Pierri, Marialuisa Restaino, Chrys Caroni is licensed under CC BY-NC-ND 4.0

We show that the standard Kuiper test is inappropriate when testing Benford’s law. The Kuiper cumulative distribution function, indeed, shows unreasonable large deviations, up to about 60%, from the ones observed in Monte Carlo simulations. These deviations are scaled down to a level below 0:3% for moderately large and large numbers of data points (n
100) if an appropriate linear transformation of the argument of the Kuiper cumulative function is performed. The test based on the transformed Kuiper function is then accurate at a level of 1% when testing Benford’s law in data.

On the Kuiper Test for Benford’s Law by Leonardo Campanelli is licensed under CC BY-NC-ND 4.0

In this article we consider Bowley’s skewness measure and the Groeneveld-Meeden b3 index in the context of finite population sampling. We employ the functional delta method to obtain asymptotic variance formulae for plug-in  estimators and propose corresponding variance estimators. We then consider plug-in estimators based on the Ha´jek cdf-estimator and on a Deville-Sa¨rndal type calibration estimator and test the efficiency of the corresponding  variance estimators and normal confidence intervals in a simulation study.

Finite population inference for skewness measures by Leo Pasquazzi is licensed under CC BY-NC-ND 4.0

R shiny is a popular open-source platform for developing web applications using R, and its use in sports has grown rapidly in recent years. The aim of this review is to summarize the current state of the art on the use of R shiny in  sports and to identify their benefits and limitations. From various databases and repositories, a total of 310 applications, created in or after 2014, were detected. Of these, 48.7% have been developed by a US based developer, 91.03% include a database, 70% aim to perform a statistical analysis, and 1% come from a scientific article. The most represented sport is basketball (n = 70; 22.9%). Only 3 apps were specifically referred to female’s sport, while 
79.2% were focused on the professional setting. Little information on used statistical techniques was observed (n = 67; 21.6%), and almost none presented scientific reference (n = 19; 6.1%). Regarding their implementation, 20%  (n = 62) had a help page and 9.7% (n = 30) returned an execution error. The quality of Shiny web applications in sports needs improvement in transparency, scientific rigor, and homogeneity to better support statistical communication.

R Shiny Web Applications in the Sports Field: A Scoping Review by Joan Martinez, Isaac Subirana, José Fernández, Jordi Cortés, Rajitha M. Silva, Martí Casals is licensed under CC BY-NC-ND 4.0