Perceived Ease of Use: Isn’t it a necessary condition?

This study deals with Perceived Ease of Use(PEU), which is one of two independent variables in Technology Model(TAM). Most empirical study employing TAM have not paid attention to the idiosyncratic nature of PEU, even though they found significant differences in testing results between PEU and Perceived Usefulness(PU). For example, Legreis et al.(2003) performed a meta-analysis on empirical TAM studies and concluded that PEU and PU are equally supported in the TAM research.
To reveal the problem in Legris et al.(2003), this study begins with a brief meta-analysis with 47 empirical TAM studies from 1987 to 2005, and shows that the relationship between PU-Behavioral Intention(BI) and PEU-BI could be different rather than same. To show the unique role of PEU in TAM, this study proposes that PEU is a necessary condition in TAM, and tests the proposition using the heteroscadasticity of error terms in regression analysis. Two systems, Hangul word processing program and a department website, were selected for the empirical testing 152 college students in a private Korean university responded to the survey questionnaire. The statistical analyses support the hypothesis that PEU is a necessary condition. Considering some limitations of this study, relative easy systems and cross-section data, further studies using different systems with longitudinal data is expected.