Serbian translation of the 20-item Toronto Alexithymia Scale: psychometric properties and the new methodological approach in translating scales

• Published in: Srpski arhiv za celokupno lekarstvo Vol. 141; no. 5-6; pp. 366 - 370
• Main Authors: Trajanović, Nikola N, Djurić, Vladimir, Latas, Milan, Milovanović, Srdjan, Jovanović, Aleksandar A, Djurić, Dusan
• Format: Journal Article
• Language: English
• Published: Serbia Serbian Medical Society 2013
• Subjects: Adult; Affective Symptoms - diagnosis; alexithymia; back-translation; Female; Humans; Linguistics - methods; Male; Middle Aged; psychometric scale; Psychometrics - methods; Psychometrics - standards; Quality Improvement; Reproducibility of Results; Serbia; TAS-20; Translating; translation methodology; Translations; Visual Analog Scale; Serbia
• ISSN: 0370-8179; 2406-0895
• DOI: 10.2298/SARH1306366T

Since inception of the alexithymia construct in 1970's, there has been a continuous effort to improve both its theoretical postulates and the clinical utility through development, standardization and validation of assessment scales.
The aim of this study was to validate the Serbian translation of the 20-item Toronto Alexithymia Scale (TAS-20) and to propose a new method of translation of scales with a property of temporal stability.
The scale was expertly translated by bilingual medical professionals and a linguist, and given to a sample of bilingual participants from the general population who completed both the English and the Serbian version of the scale one week apart.
The findings showed that the Serbian version of the TAS-20 had a good internal consistency reliability regarding total scale (alpha=0.86), and acceptable reliability of the three factors (alpha=0.71-0.79).
The analysis confirmed the validity and consistency of the Serbian translation of the scale, with observed weakness of the factorial structure consistent with studies in other languages. The results also showed that the method of utilizing a self-control bilingual subject is a useful alternative to the back-translation method, particularly in cases of linguistically and structurally sensitive scales, or in cases where a larger sample is not available. This method, dubbed as 'forth-translation' could be used to translate psychometric scales measuring properties which have temporal stability over the period of at least several weeks.