Assessing Alexithymia: Psychometric Properties and Factorial Invariance of the 20-Item Toronto Alexithymia Scale in Nonclinical and Psychiatric Samples

David Preece, Rodrigo Becerra, Ken Robinson, Justine Dandy

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

The 20-item Toronto Alexithymia Scale (TAS-20) is a self-report questionnaire designed to measure the three components of alexithymia; difficulty identifying feelings in the self (DIF), difficulty describing feelings (DDF), and externally orientated thinking (EOT). We examined the scale’s psychometric properties in Australian nonclinical (N = 428) and psychiatric (N = 156) samples. In terms of factorial validity, confirmatory factor analyses found the traditional 3-factor correlated model (DIF, DDF, EOT) to be the best and most parsimonious solution, but it did not reach adequate levels of goodness-of-fit in either sample. Several EOT items loaded poorly on their intended factor, and a reverse-scored item method factor was present; the factor structure of the scale was invariant across both samples. A higher-order factor model (with a single higher-order factor) was slightly inferior to the correlated models, but still tenable. The total scale score and DIF and DDF subscales displayed sound internal consistency, but the EOT subscale did not. We conclude that the TAS-20 has, for the most part, adequate psychometric properties, though interpretation should focus only on the total scale score and DIF and DDF subscales; we recommend the EOT subscale score not be used. Implications for clinical use and future revision of the scale are discussed.

Original languageEnglish
Pages (from-to)276-287
Number of pages12
JournalJournal of Psychopathology and Behavioral Assessment
Volume40
Issue number2
DOIs
Publication statusPublished - 1 Jun 2018
Externally publishedYes

Fingerprint

Dive into the research topics of 'Assessing Alexithymia: Psychometric Properties and Factorial Invariance of the 20-Item Toronto Alexithymia Scale in Nonclinical and Psychiatric Samples'. Together they form a unique fingerprint.

Cite this