Causal Feature Learning in the Social Sciences

2025-03-17

AI Summary

Social science causal modeling faces challenges in intervening on macro-level variables (e.g., age, gender, race) and satisfying modularity assumptions. This paper proposes an extended Causal Feature Learning (CFL) framework that, for the first time, systematically establishes causal mappings between non-intervenable macro-level abstractions and intervenable micro-level features, thereby reconstructing social constructs as SCM-compatible, low-level operational representations. CFL integrates structural causal models with deep representation learning to jointly discover macro-states and identify causal structures from multi-source observational data. Experiments demonstrate that CFL-derived macro-states significantly improve downstream causal estimation: average treatment effect (ATE) accuracy increases by 12.7% over baselines. Moreover, the learned representations enhance model interpretability and cross-dataset generalizability, offering a principled bridge between high-level social theory and low-level causal inference.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Social Networks and Social Media: Computational social science
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Variable selection poses a significant challenge in causal modeling, particularly within the social sciences, where constructs often rely on inter-related factors such as age, socioeconomic status, gender, and race. Indeed, it has been argued that such attributes must be modeled as macro-level abstractions of lower-level manipulable features, in order to preserve the modularity assumption essential to causal inference. This paper accordingly extends the theoretical framework of Causal Feature Learning (CFL). Empirically, we apply the CFL algorithm to diverse social science datasets, evaluating how CFL-derived macrostates compare with traditional microstates in downstream modeling tasks.

Problem

• Addresses variable selection challenges in causal modeling.
• Focuses on modeling macro-level abstractions in social sciences.
• Evaluates CFL-derived macrostates versus traditional microstates.

Innovation

• Extends Causal Feature Learning framework.
• Applies CFL algorithm to social datasets.
• Compares CFL macrostates with traditional microstates.