The perceived stress scale: Reliability and validity study in the Czech Republic

Abstract

Objectives. The aim of this study was to create a Czech translation of the Perceived Stress Scale (PSS), to assess its psychometric properties on a representative sample of the Czech general adult population, and to compare the original 14-item version (PSS-14) with the shortened 10-item (PSS-10) and four-item (PSS-4) versions. Sample and setting. Two pilot studies were conducted to create the final Czech translation of the scale (n = 365 and n = 420). The final version of the Czech PSS was administered to a sample of the Czech general adult population (n = 1725 of whom 981 were women, M = 44.32, SD = 12.8). The Beck Depression Inventory (BDI-II) or the State-Trait Anxiety Inventory (STAI) were administered alongside the PSS to a part of the sample. A retest measurement after 14 days was conducted (n = 159). Statistical analysis. Using the confirmatory factor analysis, the one-factor, two-factor and bifactor models were compared. The internal consistency, stability in time, and convergent validity of the scale, as well as the known-group differences, were assessed. The three versions of the PSS were compared. Results. The confirmatory factor analysis supported the bifactor model of the PSS-14 and PSS-10, and the two-factor model of the PSS-4. All versions of the scale showed good internal consistency and stability in time. There was a moderate to strong positive correlation between the PSS and the BDI-II and STAI. Differences based on age, sex, education level, and situational factors were found. Overall, the PSS-10 showed the best psychometric properties of all three versions of the scale. Study limitation. The sample consisted mostly of highly educated respondents.