The Influence of Observed Heterogeneity on Path Coefficient Significance: Technology Acceptance Within the Marketing Discipline

Abstract

The technology acceptance model (TAM) is arguably one of the most widely used models for studying user adoption in the information systems discipline and has started to be used increasingly within the marketing area. While two of its three main hypotheses have received consistent empirical support, the same is not true for the remaining relationship (i.e., the influence of perceived ease of use on the behavioral intention to adopt a system; PEOU-BI). Previously, this empirical contradiction has been explained by introducing the concept of task motivation borrowed from Davis, Bagozzi, and Warshaw (1992). Our paper provides a different explanation. We show that for the same task (and, hence, the same task motivation), the significance can also depend on observed population heterogeneity. We do this by applying partial least squares (PLS) structural equation modeling (SEM) to data stemming from a survey among approximately 2,000 individuals regarding their intention to adopt a customized newspaper. Our findings result in the following three contributions: first, our results provide an alternative explanation for the inconsistent empirical support of the PEOU-BI link within the TAM. Second, we provide a methodological contribution by proposing an approach to control for gamma change when analyzing moderating effects using PLS analysis. Third, our study highlights the importance of conducting a statistical power analysis in order to determine critical t-values in the context of a PLS path analysis.