Perbandingan Metode Pengujian Teori TAM Pada Penerimaan Teknologi E-Money di Pontianak

Abstract

As a researcher, one could analyze Technology Acceptance Model (TAM) by utilizing several methods. Such methods are summated scales regression, factor analysis score regression, covariance-based SEM, and PLS-based SEM. However, there exists less effort to compare the difference in estimates of these methods in a single dataset. The differing estimates could therefore lead to type I statistical errors. This research purpose was to compare these methods objectively with a single dataset. The dataset tested was derived from e-money research in Pontianak, with 280 data collected at refueling stations during May, June, July 2020. This research contributes to proofing how method choices will produce a differing interpretation even though tested on the same dataset. Summated scale regression and PLS-based SEM produced similar estimation results but differed from factor analysis score regression and covariance-based SEM. Further testing implies that the different estimates are due to the elimination of indicators, which is method-specific. Therefore, method justification, completeness of research report, and research context are crucial for accounting limitation of the method chosen. PLS-based SEM was a suitable method for data utilized with 50,3% variability in usefulness, 58% variability in intention to use is accounted for from the research model.