Escala de Estrés Percibido: análisis psicométrico desde la TCT y la TRI

Abstract

Introduction and objectives

The Perceived Stress Scale (Cohen et al., 1983) (PSS) is one of the scales most frequently used to measure stress. Its psychometric properties have been studied with samples from different countries, although there is no complete agreement on its dimensionality. In addition, few studies have used IRT. Taking into account the need to obtain evidence on the psychometric properties of the scale (extended and short version) in our context, we analyzed the structural validity and precision of the measurements from CTT and IRT in a sample of citizens from Córdoba, Argentina.

Materials and method

Participated 279 citizens (n = 279, M = 36.61 years old, SD = 13.97). The Spanish version of the Perceived Stress Scales was used.

Results

Confirmatory factor analysis showed that the model of two correlated factors and the bifactor model showed adequate global adjustments in both versions of the PSS, although in the second one some items showed non-significant loads. The analysis of internal consistency showed adequate indexes in both versions. The results derived from the IRT indicate that the items of the two dimensions of the scale fit appropriately to a graduated response model, the estimated parameters show that the items estimate a wide range of values of the latent traits and that they discriminate between people with different levels of trait. In turn, high correlations between statistical derived from CTT and IRT were observed.

Conclusions

The evidence obtained makes the scale a useful and reliable instrument to be used in our context.