“…Because it is virtually impossible to visit Madeira as a tourist without incurring some expenditure, the dependent variable in this study is necessarily non-zero. It has widely been argued in the literature that in such circumstances, OLS regression is inappropriate (Hung et al, 2012;Santos and Vieira, 2012;Salgado-Barandela et al, 2018;Sharma et al, 2020;Thrane, 2015;Wu et al, 2013), although a large number of papers still rely on OLS (Brida and Scuderi, 2013). One of the most frequently employed econometric methods has, therefore, been the Tobit model (Barquet et al, 2011;Brida et al, 2013aBrida et al, , 2013bBrida et al, , 2013cBuning et al, 2016;Disegna and Osti, 2016) which, as a censored method, is considered to be better suited to handling dependent variables constrained by both lower and upper limits.…”