Research map / Probabilistic and scientific modeling / Causal discovery
Position: The Causal Revolution Needs Scientific Pragmatism
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Position paper arguing that causal methods are held back by scientific perfectionism and computer-science/automation biases, and that the field should adopt scientific pragmatism and value pluralism.
Problem
Causal models and methods face slow adoption in knowledge-generating applications because of perfectionist standards and machine learning's preference for assumption-free, automated, predictive-accuracy-driven approaches.
Approach
Views causal models as useful tools for purposes rather than as correct descriptions, and argues for balanced standards between causal and predictive models, human-centric value pluralism, and acceptance of explicit assumptions.
Key results
- Cites meta-science studies showing causal language is widespread in observational research, e.g. 31% of obesity/nutrition papers in 2006 with inappropriate causal language.
- Reports that 34% of academic studies and 48% of media articles in a 2015 social-media sample used causal language reviewers judged too strong.
Surprising findings
- Argues that the causal version of a model is not 'simpler' under parsimony, since it implies more conclusions, so predictive models should not be favored for free.
Limitations
- Largely argumentative; several positions are explicitly speculative without supporting evidence.
- Evidence cited is concentrated in health and medical sciences.
Experimental setup
Builds on: Tukey conclusions vs decisions falsificationism principle of parsimony
Closest papers by meaning
- Causality can systematically address the monsters under the bench(marks) (NeurIPS 2026 · similarity 0.86)
- Position: Causal Machine Learning Requires Rigorous Synthetic Experiments for Broader Adoption (ICML 2025 · similarity 0.85)
- Causal learning with the invariance principle (NeurIPS 2026 · similarity 0.83)
- Systems with Switching Causal Relations: A Meta-Causal Perspective (ICLR 2025 · similarity 0.83)
- Rigor in AI: Doing Rigorous AI Work Requires a Broader, Responsible AI-Informed Conception of Rigor (NeurIPS 2025 · similarity 0.83)
- Causal Order: The Key to Leveraging Imperfect Experts in Causal Inference (ICLR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.