Research Information System
Quality and Quantity, vol.57, pp.537-552, 2023 (Scopus)
© 2021, The Author(s), under exclusive licence to Springer Nature B.V.Partial Least Squares (PLS) path modeling is suitable for predictive research and can also handle both reflective and formative measurement models. On the other hand, when the data derive from a common factor model population, PLS-SEM’s parameter estimates differ from the prespecified values. This trait is PLS-Structural Equation Modeling (SEM) bias, which is a controversial issue among many researchers. Bearing that in mind, the study's ultimate aim is to evaluate PLS-SEM bias at a relatively large sample size through regular and consistent PLS-SEM in the mobile shopping context. The subsidiary goal is to assess word-of-mouth concept, which is rarely used in the mobile context, within Technology Acceptance Model by employing PLS path modeling. Data were collected from 560 consumers via questionnaires and analyzed via SmartPLS 3. Findings show that regular PLS-SEM bias does not seem to diminish at a relatively large sample size when estimating data from common factor population. This study, also, offers to prefer PLSc in reflectively structured models in marketing, and also put forward that word-of-mouth is a substantial determinant in the acceptance of mobile shopping.