Testing the psychometric properties and equivalence of the Czech version of the Satisfaction with Life Scale (SWLS) using Confirmatory Factor Analysis, Item Response Theory and Bayesian modelling
Abstract
Satisfaction with Life Scale (SWLS) is one of the most commonly used instruments for measuring life satisfaction. The aim of this study is to test the psychometric properties of the Czech version of the SWLS using Confirmatory Factor Analysis, Item Response Theory and to test its invariance between social groups in terms of gender, age and education using Bayesian modelling on a representative sample of the Czech online population since the scale has not been tested on representative data in the Czech Republic.
The research sample consists of 960 respondents aged 18 to 69 years. The results confirmed very good psychometric properties of the Czech version of the SWLS, but at the same time, it is obvious that the fifth item shows worse results compared to the other four items.
In terms of dimensionality, CFA and IRT confirmed its modified single-factor structure with correlated residuals between the fourth and fifth items as the most appropriate. Testing for approximate measurement invariance using Bayesian modelling showed that the SWLS measures comparably between groups based on gender, age, and education.
In conclusion, the Czech version of the SWLS is a suitable, verified, and reliable instrument for measuring the life satisfaction of Czech citizens.