Causal Inference
Definition
Causal inference investigates whether and how a change in a cause alters an outcome – not just whether both occur together.
It uses experiments or explicit assumptions and models to estimate counterfactuals: What would have happened under a different intervention? Observational data alone require additional identification assumptions for this.
Simply explained
Like not only seeing that two things occur together, but asking what would change if we deliberately altered one.
Why it matters in practice
Relevant for proof of effects, marketing, process improvement, and decisions where correlation is insufficient.
Example from business
A company tests whether a new process rule actually reduces throughput time, instead of just noting that fast teams use it more often.
Limits and pitfalls
- Mistaking correlation for causation; ignoring confounders; choosing fitting models post hoc; not documenting identification assumptions.
Origin & history
Modern causal inference combines statistics, econometrics, and graphical causal models. Judea Pearl significantly shaped formal work with structural causal models and interventions.
Use in mid-sized companies
- Possible application: Investigate the additional effect of marketing measures, process changes, or rules instead of comparing only simultaneous changes.
Frequently asked questions
Does causality always require an experiment?
Randomization is powerful, but under justified assumptions observational designs are also possible. Is correlation worthless? No, but it alone does not identify cause.
Sources
- Glymour & Jewell, Causal Inference in Statistics: A Primer · Pearl, Glymour & Jewell, Causal Inference in Statistics: A Primer
- Causal Inference in Statistics: An Overview · Pearl (2009), Causal Inference in Statistics: An Overview
- Autorenkopie · Pearl (2009), Autorenkopie
Source verification: 10./11.10.2026.
Editorially reviewed on: