A Study on the Decision to Continue Using a Modeling Grammar

Conceptual modeling is the process of building a representation of selected phenomena in a problem domain for the purpose of understanding and communication among stakeholders (Wand and Weber 2002) in the process of information systems analysis and design. A crucial element in the context of conceptual modeling is the `conceptual modeling grammar', i.e., the set of (graphical) constructs and the rules that show how to combine the constructs for modeling real-world domains (Wand and Weber 2002). In line with the reported popularity of conceptual modeling and despite an observable proliferation of modeling grammars, only a few have been widely accepted by practitioner communities. While IS research has shown that modeling grammars differ quite significantly in their features and characteristics, e.g., correctness and ease of use (Batra et al. 1990) or support for domain comprehension and problem solving tasks (Agarwal et al. 1996), actual practice informs us that, seemingly independent from such intrinsic characteristics, certain modeling grammars have achieved higher levels of adoption and dissemination in modeling practice than others. Hence it appears that the findings from prior IS studies provide only little explanation of actual acceptance and usage patterns. Acceptance and usage studies are quite popular in IS research in general (Davis 1989; Bhattacherjee 2001b) but the question of the continuance decision, viz., the decision of an individual to continue using an artifact (Bhattacherjee 2001b), has only to a small extent been addressed in the modeling community. Accordingly, the imperative of this research is to develop an understanding of the continued use of modeling grammars by individual modelers after its initial adoption. This focus of the study can be justified in referral to the observation that, often, the initial adoption of a modeling grammar is an organizational decision. Yet, prior studies, e.g., (Brown et al. 2002), have suggested that individual acceptance in the post-adoption phase would have significant implications for the long-term viability and eventual success of modeling grammars. In particular, this study seeks to explore whether certain intrinsic characteristics of modeling grammars, such as their capabilities to provide faithful representations of real-world domains (Wand and Weber 1993), pose an impact on an individual's intention to continue using a modeling grammar. The unit of analysis in this study is the `conceptual modeling grammar', following the definition presented above. The ultimate dependent variable of interest is the `intention to continue to use' (Bhattacherjee 2001b), and as independent variables the `representational capabilities' of conceptual modeling grammars (Wand and Weber 1993; Weber 1997) are studied. The study design employs multiple research methods, in particular semi-structured interviews as part of case studies (to explore the independent variables, i.e., to build the model), and surveys (to measure the effect of these variables on the dependent variables, i.e., to test the model). The remainder of this paper is as follows. The next sections describe research model and empirical study design. Then, the contributions to-date are described and a discussion of study limitations is given