Table 1.
Cronbach’s alpha values computed on each variable.
Table 2.
Mean (standard deviation in parentheses) values of age and psychological measures for the Italian, U.S., and Singaporean sample.
Table 3.
Correlation matrices p < .05, ** p < .01, *** p < .001.
Table 4.
Mean absolute error (MAE) with [90% confidence intervals] of the three ML models estimating the alexithymia scores (TAS-20; BVAQ-Cognitive; BVAQ-Affective) on the training subset (Italian Train Dataset) and on the three test subsets: The Italian Test Dataset, the U.S. Test Dataset and the Singaporean Test Dataset.
Fig 1.
Ranking of ML predictors for each target variable (TAS-20, BVAQ-Cognitive, BVAQ-Affective), derived from the coefficients of the linear SVM model.
The ranking ranges from 1 (best predictor) to 12 (worst predictor).