Revisiting Unified Theory of Technology and Use of Technology Using Meta-analytic Structural Equation Modelling

Abstract

The Unified Theory of Technology and Use of Technology (UTAUT) has been widely used in information system studies since its introduction in 2003. The current study synthesizes 40 empirical studies based on UTAUT in educational contexts using the one-stage meta-analytic structural equation modelling method. While the study confirmed the initial findings by Ventakesh et al. (2003), the model in this study underperformed in the explained variance of behavioral intention. However, the explained variance of usage behavior performed better than the original UTAUT. After introducing new direct relationships between the UTAUT constructs, it was found that the construct, facilitating conditions, was a new predictor of behavioral intention. At the same time, effort expectancy and social influence were new predictors of usage behavior. There have been studies on the UTAUT model with many diverse findings since its inception in 2003. The method introduced in this study, a One-Stage combined Meta-Analysis and Structural Equation Modelling (OSMASEM), offers an approach for researchers to use past empirical data to examine the UTAUT framework without relying on replicating similar studies. As more empirical data from UTAUT research are added to train the data model using OSMASEM, researchers can now study how educational technology trends change over time.

Examining Unified Theory of Acceptance and Use of Technology 2 through Meta-analytic Structural Equation Modelling

“…In other words, for utilising the UTAUT2 model for studies, PV should be one of the essential constructs in future research. Or (2023a) argued that since past studies had shown that PV had no significant effect on BI when examining technologies that were free of charge, it was recommended that the original UTAUT model be adopted or extended with added constructs instead of citing it as UTAUT2 research. For studies that included PV as a construct, it has been found that PV was a predictor of BI (Azizi et al, 2020;Farooq et al, 2017;Gengfu & Chotiyaputta, 2019;Meet et al, 2022;Mehta et al, 2019;Moorthy et al, 2019b;Tseng et al, 2019;Xu et al, 2022).…”

“…Some studies were then excluded by applying the following criteria: (1) the studies did not target teachers, lecturers, educators or students in K-12, college or university education; (2) the studies were not based on the UTAUT2 model, but the UTAUT model. Past research using the OSMASEM approach had been conducted previously (Or, 2023a); (3) the studies had insufficient statistical reporting of the correlations between UTAUT2 constructs; (4) correlations between variables were negative where R package, metaSEM, is unable to compute; and (5) UTAUT2 was examined outside of educational contexts. Figure 2 summarises the results of the literature search and screening procedures.…”