No abstract

Multicollinearity is one of the most important issues in regression analysis, as it produces unstable coefficients’ estimates and makes the standard errors severely inflated. The regression theory is based on specific assumptions concerning the set of error random variables. In particular, when errors are uncorrelated and have a constant variance, the ordinary least squares estimator produces the best estimates among all linear estimators. If, as often happens in reality, these assumptions are not met, other methods might give more efficient estimates and their use is therefore recommendable. In this paper, after reviewing and briefly describing the salient features of the methods, proposed in the literature, to determine and address the multicollinearity problem, we introduce the Lpmin method, based on Lp-norm estimation, an adaptive robust procedure that is used when the residual distribution has deviated from normality. The major advantage of this approach is that it produces more efficient estimates of the model parameters, for different degrees of multicollinearity, than those generated by the ordinary least squares method. A simulation study and a real-data application are also presented, in order to show the better results provided by the Lpmin method in the presence of multicollinearity

The COVID-19 pandemic has highlighted the vulnerability of specific population sections, with regards to economic and work conditions, mental and physical well-being, and context-based factors, emphasizing the need for timely policy measures aimed at counteracting the Italian economic framework’s fragility—which poorly adapts to unexpected circumstances. Identifying the most vulnerable groups is, therefore, essential with a view to carrying out targeted measures. Concerning University, the economic downturn caused by COVID-19 could likely result in a decrease in enrollments to both the first and further years of study, with significant consequences on the future of students and the system as a whole. The class of students is of great interest, as it is made up of individuals differing from each other in many ways. Our investigation is aimed at observing anxiety levels filtering the perception of one’s anxiety state in a highly stressful time such as the pandemic from the usual anxiety levels. This evaluation allows us to evaluate the similarity of individual behaviors during the lockdown period with those from the previous period.

In this paper, we deal with the evaluation of Conditional Value-at-Risk in the framework of portfolio theory by using a modified Gaussian Copula -where the modification is obtained by introducing the Generalized Correlation Coefficient -and by assuming a Generalized Error Distribution with properly estimated shape parameter p for the returns of the considered risky assets. In so doing, we add to the connection between standard Copula theory and financial risk assessment. A comparison analysis of our findings with those obtainable through a standard Gaussian Copula-based procedure in a set of real data is also presented.

Types of discrimination are usually distinguished by economic theory in statistical and taste-based. Using a correspondence experiment, we analyze which of the two affects Italian labor market the most. In this respect, we studied the difference in discrimination reserved to first- and second-generation immigrants, taking gender differences into account. Even if we want to admit a rational discrimination based on perceived productivity differences (statistical discrimination) against first-generation immigrants (concerning language and education gaps), the same would not be reasonable for second-generation ones. Since they are born and educated in Italy, where they have always lived, the associated discrimination must be taste-based.