Communication Causal Inference
Overview
This Skill starts from the intervention, counterfactual, and estimand. It does not
upgrade an observational regression into a causal design by adding controls.
Required start
- State treatment, outcome, unit, timing, population, comparison, and estimand.
- Draw or describe the causal graph, including selection, measurement, interference,
anticipation, and post-treatment variables.
- Use references/causal-design-protocol.md to
select a design only after its identifying variation is established.
Design routing
- randomized experiments route jointly to
communication-experiment;
- DiD/event studies require a credible untreated trend and explicit treatment timing;
- IV requires relevance, exclusion, independence, and a clear complier estimand;
- RDD requires a deterministic assignment threshold and continuity near the cutoff;
- matching/weighting requires conditional exchangeability, overlap, and pre-treatment
covariates;
- synthetic control requires a defensible donor pool and pre-treatment fit;
- interrupted time series requires a stable pre-period and no concurrent shock;
- natural-experiment language is used only for genuinely as-if-random variation.
Verification
Report identifying assumptions beside diagnostics and falsification tests: pre-trends,
placebos, negative controls, bandwidth/specification sensitivity, overlap, balance,
attrition, spillovers, and measurement stability. Preserve failed diagnostics and
weaken the claim when assumptions are not supported.
Output shape
Return an estimand statement, DAG/assumption map, design and alternatives, data timing,
diagnostics and falsification plan, sensitivity analysis, and the exact population and
intervention to which the causal claim applies.
Hard boundaries
- Temporal order alone is not identification.
- Fixed effects do not remove all time-varying confounding.
- Matching balances measured covariates; it does not create randomization.
- A policy date does not by itself create a valid intervention.
- Statistical significance is not assumption validation or practical importance.
1---2name: communication-causal-inference3description: Design and audit causal inference for communication research using DAGs, natural experiments, DiD/event studies, IV, RDD, matching or weighting, synthetic control, and interrupted time series. Use for 因果推断/政策评估/平台冲击/DiD/IV/RDD/匹配. Do not use when the goal is only description or prediction.4---56# Communication Causal Inference78## Overview910This Skill starts from the intervention, counterfactual, and estimand. It does not11upgrade an observational regression into a causal design by adding controls.1213## Required start14151. State treatment, outcome, unit, timing, population, comparison, and estimand.162. Draw or describe the causal graph, including selection, measurement, interference,17 anticipation, and post-treatment variables.183. Use [references/causal-design-protocol.md](references/causal-design-protocol.md) to19 select a design only after its identifying variation is established.2021## Design routing2223- randomized experiments route jointly to `communication-experiment`;24- DiD/event studies require a credible untreated trend and explicit treatment timing;25- IV requires relevance, exclusion, independence, and a clear complier estimand;26- RDD requires a deterministic assignment threshold and continuity near the cutoff;27- matching/weighting requires conditional exchangeability, overlap, and pre-treatment28 covariates;29- synthetic control requires a defensible donor pool and pre-treatment fit;30- interrupted time series requires a stable pre-period and no concurrent shock;31- natural-experiment language is used only for genuinely as-if-random variation.3233## Verification3435Report identifying assumptions beside diagnostics and falsification tests: pre-trends,36placebos, negative controls, bandwidth/specification sensitivity, overlap, balance,37attrition, spillovers, and measurement stability. Preserve failed diagnostics and38weaken the claim when assumptions are not supported.3940## Output shape4142Return an estimand statement, DAG/assumption map, design and alternatives, data timing,43diagnostics and falsification plan, sensitivity analysis, and the exact population and44intervention to which the causal claim applies.4546## Hard boundaries4748- Temporal order alone is not identification.49- Fixed effects do not remove all time-varying confounding.50- Matching balances measured covariates; it does not create randomization.51- A policy date does not by itself create a valid intervention.52- Statistical significance is not assumption validation or practical importance.
haochengw372-hash