Zastosowanie modelowania PLS-SEM do badania innowacyjności gospodarek krajów Unii Europejskiej

Abstract

Purpose – The article aims to present the possibility of using PLS-SEM method in research on the innovativeness of EU economies. Research method – Structural equation modeling is a method that allows the study of relationships between variables that are unobservable. The SEM model consists of two sub-models: structural and measurement. Covariance-based (CB) or partial least squares (PLS) methods are used to estimate the model. The choice depends on the empirical context, research objectives, sample size and statistical properties of the data. Due to the small size of the population, the PLS method was used in the study. The models were estimated on the basis of cross-sectional data for 2018 and 2021. Results – The article presents the stages of construction of the PLS-SEM model, from model specification, through estimation, statistical verification, to interpretation of the results. Particular attention is paid to the possibility of formulating conclusions based on various elements of the estimated model. Originality / value / implications / recommendations – PLS-SEM is a relatively new econometric method belonging to the group of second generation multivariate analysis methods. Its dynamic development can be observed especially in management sciences. In the discipline of economics, the method is not widely known and used, but it has great potential and can be useful in the study of relationships between unobservable categories. The empirical research presented in the article fills the cognitive gap in the use of advanced econometric methods in economic research.