Axiomatic Theories and Improving the Relevance of Information Systems Research

Abstract

This paper examines the fact that a significant number of empirical studies in behavioral information systems (IS) theory research engage in confirmative testing of self-evident axiomatic theories without yielding highly relevant knowledge for the IS community. To measure how pervasive such testing of axiomatic theories is, we conducted a horizontal analysis using 666 hypotheses from 72 representative behavioral IS theories and discovered that more than 60% of the hypotheses could be regarded as axiomatic theory elements. To further investigate the pervasiveness of repetitive testing of axiomatic theories, we vertically analyzed 1,301 hypotheses from 148 articles in three theory categories: technology acceptance model, diffusion of innovation theory, and institutional theory. These analyses revealed that 68.1% of these hypotheses were axiomatic and that 74.6% of them were inherited from general truths beyond the IS domain. In order to shift the research emphasis toward enhancing the relevance of IS research outcomes without sacrificing methodological rigor, we propose four complementary IS research approaches: (1) identifying disconfirming boundary conditions, (2) measuring the relative importance of axiomatic causal factors, (3) measuring the stage of progression toward visionary goals when the nature of the axiomatic theory can be extended to future visions, and (4) engaging in the conceptual design of visionary axiomatic goals. We conclude with a discussion about why so many scholars devote substantial effort to reconfirming axiomatic theories and suggest avenues for more relevant research outcomes.