Flattening network data for causal discovery: What could wrong?

Methods for learning causal dependencies from observational data have been the focus of decades of work in social science, statistics, machine learning, and philosophy [9, 10, 11]. Much of the theoretical and practical work on causal discovery has focused on propositional representations. Propositional models effectively represent individual directed causal dependencies (e.g., path analysis, Bayesian networks) or conditional distributions of some outcome variable (e.g., linear regression, decision trees). However, propositional representations are limited to modeling independent and identically distributed (IID) data of a single entity type. Many real-world systems involve heterogeneous, interacting entities with probabilistic dependencies that cross the boundaries of those entities (i.e., non-IID data with multiple entity types and relationships). These systems produce network, or relational, data, and they are of paramount interest to researchers and practitioners across a wide range of disciplines. To model such data, researchers in statistics and computer science have devised more expressive classes of directed graphical models, such as probabilistic relational models (PRMs) [2] and directed acyclic probabilistic entityrelationship (DAPER) models [4]. Despite the assumptions embedded in propositional models, a common practice is to flatten, or propositionalize, relational data and use existing algorithms [5] (see Figure 1, focusing on algorithms that learn causal graphical models). While there are statistical concerns, this process is generally innocuous if the task is to model statistical associations for predictive inference. In contrast, to learn causal structure, estimate causal effects, or support inference over interventions, the effects of flattening inherently relational data can be particularly deleterious. In this paper, we identify four classes of potential issues that can occur with a propositionalization strategy as opposed to embracing a more expressive representation that would not succumb to these problems. We also present empirical results comparing the effectiveness of two theoretically sound and complete algorithms that learn causal structure: PC—a widely used constraint-based, propositional algorithm for causal discovery [11], and RCD—a recently developed constraint-based algorithm that reasons over a relational representation [6].