*2.4. Statistics*

Continuous data were expressed in mean ± standard deviation, while categorical and ordinal data in frequencies, percentages, and 95% confidence intervals. To compare the continuous variables among countries, we performed one-way ANOVAs with Sche ffe test as post hoc when Levene's test suggested homogeneity of variances (*p* > 0.05). Otherwise, one-way ANOVAs with Welch's correction and Dunnett T3 test as post hoc were performed (Microsoft ® Excel ® 2013, Microsoft Corporation, Redmond, WA, USA). To compare the categorical and ordinal variables among countries, we used the χ2 test of independence (GraphPad Prism 7.05, GraphPad Software Inc., La Jolla, CA, USA) and multiple *t*-tests for proportions with Bonferroni correction as post hoc (Statistics calculator v4.0, StatPac, Northfield, MN, USA). All comparisons were deemed significant at a *p*-value ≤ 0.05. E ffect sizes were calculated for ANOVAs (omega squared, ω2) and χ2 tests of independence (phi statistic, ϕ). Both e ffect size statistics are dimensionless and range between 0 and 1. An e ffect size was considered small, medium, or large if ω<sup>2</sup> reached 0.01, 0.06, or 0.14, respectively; the respective values for ϕ were 0.1, 0.3, 0.5 [30,31]. E ffect sizes below the "small" cut point were considered trivial [32]. The di fference between the highest mean and the lowest mean was reported along with the e ffect size. The comparisons among countries were carried out for the whole sample and separated by sex.
