Introduces PIV as the probability that a rejected null hypothesis on observed data would also be rejected when including counterfactual outcomes, serving as a robustness measure for causal claims.
For objective causal inference, design trumps analysis
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.AP 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On the probability of a causal inference is robust for internal validity
Introduces PIV as the probability that a rejected null hypothesis on observed data would also be rejected when including counterfactual outcomes, serving as a robustness measure for causal claims.