Info Intervention
This provides a novel framework for causal analysis, particularly useful for domains with non-manipulable variables like social sciences or genetics.
The paper tackles the limitations of do intervention in causal diagrams for non-manipulable variables and counterfactual conditions by introducing a new info intervention that intervenes on input/output information of causal mechanisms, viewing causality as information transfer.
Causal diagrams based on do intervention are useful tools to formalize, process and understand causal relationship among variables. However, the do intervention has controversial interpretation of causal questions for non-manipulable variables, and it also lacks the power to check the conditions related to counterfactual variables. This paper introduces a new info intervention to tackle these two problems, and provides causal diagrams for communication and theoretical focus based on this info intervention. Our info intervention intervenes the input/output information of causal mechanisms, while the do intervention intervenes the causal mechanisms. Consequently, the causality is viewed as information transfer in the info intervention framework. As an extension, the generalized info intervention is also proposed and studied in this paper.