PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
July 26, 20250 citations

Unified Causal Field Theory: A Proof of Geometric Subsumption and Extension of Causal Inference Methods Into Unified Framework

View Full Paper
SLSamuel L Leizerman

Key Points

  • MAIN FINDING: The study establishes a unified framework for causal inference using field-theoretic methods in higher-dimensional spacetime.
  • KEY EVIDENCE: It validates the framework by applying it to established methods like Difference-in-Differences and Dynamic Structural Equation Modeling.
  • APPROACH: A geometric-dynamical framework is introduced, leveraging tensor calculus to explore unobserved variable bias.
  • SIGNIFICANCE: This framework enhances understanding of causality, offering better detection of biases than traditional methods in lower-dimensional spaces.

Abstract

In this proof, I extend and apply field-theoretic framework in a generalized (nD + T) spacetime, withsignature (−, +, ..., +), to formally subsume, reinterpret, and unify causal inference methods. I illustrate this unity by demonstrating its applicability to Difference-in-Differences (DiD), regression inter-action effects, potential outcomes, Dynamic Structural Equation Modeling (DSEM), and propensity scoring. Critically, I prove that Regression Discontinuity Design (RDD) and Regression KinkDesign (RKD) are field-theoretic boundary phenomena differentiated by the order of causal curvature discontinuity—interpretable as a collision angle within the manifold. This geometric-dynamicalparadigm offers a novel mechanism for the detection of unobserved variable bias, leveraging tensorcalculus and linear algebra. The framework’s predicted performance stems from its foundation inLorentzian spacetime, the inherent mathematical structure of causality, providing a native environment for causal relationships that is inaccessible to methods operating on degraded, flattened projections of reality. True mathematical discontinuities are shown to be approximations arising when thecharacteristic time scale of a causal transition falls below the observational time step, analogous tothe Planck time limit in physics. The empirical success of existing low-dimensional causal inferencemethods provide a compelling validation for the superior performance predicted out of mathematical necessity and of this higher-dimensional spacetime framework. NOTE: This is a first draft, expect errors and an additional section or two.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Samuel L Leizerman (2025) studied this question.

synapsesocial.com/papers/689a0933e6551bb0af8ce5achttps://doi.org/10.31234/osf.io/bms6a_v2
Ask AI
Helpful
Bookmark
Share
View Full Paper