ASSESSING VALIDITY OF SOURCES OF INNOVATION, INNOVATION AND FIRM PERFORMANCE FOR VARIANCE BASED STRUCTURAL EQUATION MODELLING

Abstract

In recent years, PLS-SEM is preferred by many researchers from different fields due to the three main distinctive justifications which are related to non-normal data, small sample size and formatively measured constructs. There are two stages in PLS-SEM analysis which include the assessment of measurement model and structural model. The former is conducted to evaluate the validity and reliability of the variables while the latter is performed to examine the causal relationships between variables. The main focus of this paper is to reveal the stages assessing validity of some sources of innovation, innovation and firm performance, started with explanatory factor analysis (EFA) and followed by confirmatory analysis (CFA). A total of six common used validity tools in PLS-SEM will be presented: outer loadings, average variance extracted (AVE), composite reliability, Fornell-Larker criterion, cross-loadings and HeterotraitMonotrait Ratio (HTMT)