Modeling User Training and Support for Information Technology Implementations: A Bayesian Test of Competing Models

Introduction

Implementations of information technology (IT) innovations in organizations are known to induce changes in user perceptions of the work environment (Barki & Hartwick, 1994; Cooper & Zmud, 1990; Davis, 1989; Doll & Torkzadeh, 1988; Finlay & Mitchell, 1994). User perceptions of the outcomes of an IT implementation are also dependent on the implementation process and the extent to which components of the implementation process such as user training and user support are deemed effective by individuals (Igbaria, Guimaraes, & Davis, 1995; Lee, Kim, & Lee, 1995; Shaw, 2001b; Venkatesh & Davis, 1996).

Prior studies on IT implementation issues have adopted contrasting approaches in modeling the impacts of user training and user support: one approach has been to model user training and user support as direct influences on perceived outcomes of individuals while the other approach has been to model user training and user support as indirect influences on outcomes perceived by individuals (Sabherwal, Jeyaraj, & Chowa, 2006; Speier & Brown, 1997; Taylor & Todd, 1995; Venkatesh & Davis, 1996). While both approaches have received empirical support and yielded useful insights on the underlying relationships, prior literature has not really dwelt on whether one of the two approaches may be more appropriate in understanding the roles of user training and user support in IT implementations.

We construct and employ a Bayesian test for choosing the most plausible model of explanation of the effects of user training and user support when Structural Equation Modeling (SEM) techniques reveal multiple competing models that fit the data equally well. More specifically, we examine the roles of user training and user support in IT implementations by constructing a direct effects model and an indirect effects model, evaluate their explanatory power against a baseline model, and determine one model to be more plausible than the other. Our research provides initial evidence on how these two components of IT implementation processes may impact the outcomes perceived by individuals affected by IT implementations.

The remainder of the paper is organized as follows. The next section provides descriptions of the alternative approaches to modeling user training and user support based on prior literature. The following section presents the research methods employed in this study. The last section presents the results, discussion, and implications.