Because a fundamental attribute of a good theory is causality, the information systems (IS) literature has
strived to infer causality from empirical data, typically seeking causal interpretations from longitudinal,
experimental, and panel data that include time precedence. However, such data are not always obtainable and
observational (cross-sectional, nonexperimental) data are often the only data available. To infer causality from
observational data that are common in empirical IS research, this study develops a new data analysis method
that integrates the Bayesian networks (BN) and structural equation modeling (SEM) literatures.
Similar to SEM techniques (e.g., LISREL and PLS), the proposed Bayesian networks for latent variables
(BN-LV) method tests both the measurement model and the structural model. The method operates in two
stages: First, it inductively identifies the most likely LVs from measurement items without prespecifying a
measurement model. Second, it compares all the possible structural models among the identified LVs in an
exploratory (automated) fashion and it discovers the most likely causal structure. By exploring the causal structural
model that is not restricted to linear relationships, BN-LV contributes to the empirical IS literature by
overcoming three SEM limitations (Lee, B., A. Barua, A. B. Whinston. 1997. Discovery and representation of
causal relationships in MIS research: A methodological framework. MIS Quart. 21(1) 109–136)—lack of causality
inference, restrictive model structure, and lack of nonlinearities. Moreover, BN-LV extends the BN literature by
(1) overcoming the problem of latent variable identification using observed (raw) measurement items as the
only inputs, and (2) enabling the use of ordinal and discrete (Likert-type) data, which are commonly used in
empirical IS studies.
The BN-LV method is first illustrated and tested with actual empirical data to demonstrate how it can help
reconcile competing hypotheses in terms of the direction of causality in a structural model. Second, we conduct
a comprehensive simulation study to demonstrate the effectiveness of BN-LV compared to existing techniques
in the SEM and BN literatures. The advantages of BN-LV in terms of measurement model construction and
structural model discovery are discussed.