Underlying structure and measurement invariance by sex of the state trait anxiety inventory: A psychometric analysis in Ecuador

Abstract

Anxiety is currently one of the most prevalent and investigated psychological symptoms requiring precise and reliable instruments for its assessment. The current study evaluates the psychometric properties of the State-Trait Anxiety Inventory (STAI) in an Ecuadorian sample, focusing on the factor structure and potential methodological artefacts, especially those introduced by reverse-scored items. Employing exploratory and confirmatory factor analyses (EFA and CFA), along with structural equation modelling (SEM), the research examines the presence of two distinct factors: genuine anxiety and a secondary factor related to reversed items. Findings indicate that the expected two-factor model fits the data more accurately than a unidimensional model. Notably, cross-loadings suggest the second factor may represent a substantive positive, anxiety-free state rather than solely a methodological effect. Additionally, multigroup CFA confirmed strict measurement invariance across sex, supporting the scale’s utility for group comparisons. These results suggest that the STAI’s total score represents a composite of anxiety and well-being. While the STAI remains a robust tool for evaluating anxiety in clinical contexts, researchers seeking pure measures of anxiety symptomatology should consider analysing the subscales separately.