Intrinsic and Extrinsic Motivation for University Staff Satisfaction: Confirmatory Composite Analysis and Confirmatory Factor Analysis

Abstract

The hierarchical construction model (HCM) may be used to minimize colinear formative indicators while increasing the statistical power of content-specific constructs. However, the present research discovered a strong correlation between intrinsic and extrinsic motivation and job satisfaction, indicating a lack of discriminating validity that limits the use of HCM. Thus, this research condensed data on job satisfaction for university staff by comparing a consistent partial least square (PLSc) model to a composite model without utilizing higher-order constructs. The sample consisted of 392 individuals working in a Thai university with a total of 1,042 staff. The results show that the composite model performs better than a consistent partial least square, generating bias. Intrinsic motivation is both a direct and indirect effect on job satisfaction. Extrinsic motivation is a complementary model mediator effect. The limitation of the study is an inherent relation between intrinsic and extrinsic motivation and indicators of job satisfaction, which can cause a common factor model bias. Further studies where the partial least square structural equation model (PLS-SEM) is compared early with the covariance-based structural equation model (CB-SEM) are needed, particularly studies using composite indicators. PLS-SEM can now be used to measure both confirmatory composite analysis and confirmatory factor analysis, while CB-SEM can only estimate confirmatory factor analysis. The method of condensing data may help eliminate discriminant validity issues.