Challenges in operationalizing conceptual models in aetiological research

Abstract

Conceptual or theoretical models are crucial in developing causal hypotheses and interpreting study findings, but they have been underused and misused in aetiological research, particularly in dentistry and oral epidemiology. Good models should incorporate updated evidence and clarify knowledge gaps to derive logical hypotheses. Developing models and deriving testable hypotheses in operational models can be challenging, as seen in the four examples referred to in this commentary. One challenge concerns the theoretical validity of the model, while another relates to difficulties in operationalizing abstract concepts. A third challenge refers to the lack of sufficient information in the dataset to test partially or even the whole model. Finally, a common challenge is the application of a conceptual model to different contexts. Among the existing methodological approaches to operationalize conceptual models, causal graphs may be helpful, especially when combined with approaches from diverse disciplinary fields via triangulation.