Testing the “holy grail” of industrial psychology as a non-recursive bow pattern model in higher education using the PLSe2 method

Abstract

PurposeThe direction of causality between job satisfaction and job performance (known as the holy grail of industrial psychologists) is undetermined and related research findings in different organizational contexts are mixed. Based on the ample literature, mainly from Western countries, on the relationship between job satisfaction and job performance, a non-recursive bow pattern model was utilized to investigate the direct relationship between these two variables in an Asia–Pacific higher education system.Design/methodology/approachThis study is quantitative in approach and survey in design. Additionally, to meet the statistical requirements of non-recursive bow pattern analysis, the authors added welfare as a theory-driven instrumental variable to introduce exogenous variability. Using the efficient partial least squares (PLSe2) estimator, the authors fitted the model to the data collected from 2008 academics affiliated with Malaysian public universities and polytechnics.FindingsThe results showed that while job satisfaction is considerably influenced by welfare, it is not a significant predictor of job performance directly. In addition, a meaningful positive correlation between the disturbance terms of job satisfaction and job performance was observed, suggesting the existence of other factors that could increase both job satisfaction and job performance. The findings' theoretical and practical implications are discussed, and a list of theory-driven evidenced-based policies in this regard is provided.Originality/valueThis is the first study to test a non-recursive bow pattern model and examine the holy grail of industrial psychology based on the PLSe2 methodology, as a parametric approach to partial least squares (PLS), in a higher education context. This study also provides higher education researchers with the advantages of the PLSe2 method, especially in causal-predictive modeling, in the context of applied higher education research.