ABSTRACT
BACKGROUND: The TEMPS-A has been validated in 8 languages, the original being American English, and includes among others such Latin languages as Italian, French, and Spanish-Buenos Aires. This is the first Portuguese-Lisbon validation. METHODS: The sample included 1173 students from six different universities and representing most disciplines (such as medicine, law, humanities, engineering, etc.), both sexes (67% female), and ages between 17 and 58 (x+/-SD=21+/-4). Standard psychometric tests were used for internal consistency, validity, and factor analysis. RESULTS: The study upheld the 5 Factor proposed structure of TEMPS-A. Cronbach alpha varied from 0.67 for the depressive and 0.83 for the anxious, with the others in-between. We could retain all 110 items of the Interview Schedule. The highest mean scores were found for the hyperthymic, and the lowest for the irritable. As expected, depressive and anxious subscales had strong correlations, followed by the cyclothymic and anxious, and cyclothymic and irritable; in exploratory factor analysis, these subscales constituted Factor I, contrasted to the depressive and the hyperthymic as a biphasic continuum (Factor II). Females scored higher on the depressive, cyclothymic and anxious, and the males on hyperthymic and irritable. Overall, however, no temperament was "dominant" in this population, all temperaments z-scores being 3.3-4%! LIMITATIONS: Study limited to university students of young age. CONCLUSIONS: TEMPS-A Lisbon is a reliable and valid instrument. The only relatively weak factor is the depressive, which is similar to other language versions. Gender differences and correlations of temperaments are generally similar to other countries. What appears relatively special to the Portuguese is the relatively "balanced" mix of temperaments in this university student population.
