Estimating Causal Effects of Treatments in Randomized and Nonrandomized Studies.¶
Rubin, D. B. (1974). Estimating Causal Effects of Treatments in Randomized and Nonrandomized Studies. Journal of Educational Psychology, 66(5), 688-701.
Cited by¶
8 citations across 8 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Blocking (In Experimental Design)
- T3 — Design-based control versus model-based control. Blocking achieves variance reduction through how units are assigned; covariate adjustment in regression achieves it through how the model is fit
This sourceFoundational potential-outcomes framework defining causal effects as comparisons under hypothetical treatments holding background conditions fixed, contrasting design-based and model-based (covariate-adjustment) control.
Supported in partVerified against the publisher's abstract
Rubin's abstract frames design-based versus model-based control only as randomization versus carefully controlled nonrandomized data; it never mentions blocking or regression covariate adjustment.
“Presents a discussion of matching, randomization, random sampling, and other methods of controlling extraneous variation. The objective was to specify the benefits of randomization in estimating causal effects of treatments.”
- T3 — Design-based control versus model-based control. Blocking achieves variance reduction through how units are assigned; covariate adjustment in regression achieves it through how the model is fit
- Counterfactual Subtraction
- The reasoning generalises to any decomposition of an observed quantity into "what happened" minus "what would have happened," which is why the prime sits one level below causality as a general method of operationalising a causal claim into a number.
This sourceFoundational potential-outcomes framework defining a causal effect as the difference between observed and counterfactual potential outcomes, with randomisation/exchangeability as the identifying assumption.
Supported in partVerified against the publisher's abstract
Rubin's abstract supplies the general principle that randomisation controls extraneous variation when estimating causal effects, but not the article's enumerated shared factors or its 4-point residual.
“Presents a discussion of matching, randomization, random sampling, and other methods of controlling extraneous variation. The objective was to specify the benefits of randomization in estimating causal effects of treatments.”
- The reasoning generalises to any decomposition of an observed quantity into "what happened" minus "what would have happened," which is why the prime sits one level below causality as a general method of operationalising a causal claim into a number.
- Counterfactuals
- A statement like "if I had studied harder, I would have passed"
This sourceFoundational potential-outcomes framework: causal effects as comparisons of Y(1) and Y(0) under hypothetical treatments holding background fixed; underlies the fundamental problem of causal inference (only one potential outcome observed per unit). Verified support.
Borderline supportVerified against the publisher's abstract
The abstract concerns estimating causal effects of treatments — the framework in which counterfactuals live — but never states a counterfactual conditional or potential outcome.
“The objective was to specify the benefits of randomization in estimating causal effects of treatments.”
Two runs of the same check disagreed on this citation; the later verdict is shown and the citation is queued for review.
- A statement like "if I had studied harder, I would have passed"
- Experimental Design
- Experimental design makes this counterfactual reasoning concrete by randomization or matched design, formalized in Rubin's (1974) potential-outcomes framework.
This sourceFoundational potential-outcomes framework defining causal effects as comparisons of outcomes under hypothetical treatments holding background conditions fixed, making counterfactual reasoning concrete via randomization/matched design.
Supported in partVerified against the publisher's abstract
Rubin's abstract backs randomization and matched/controlled comparison as routes to estimating causal effects of treatments, but it does not itself state the potential-outcomes or counterfactual formalization the claim attributes to it.
“Presents a discussion of matching, randomization, random sampling, and other methods of controlling extraneous variation. The objective was to specify the benefits of randomization in estimating causal effects of treatments.”
- Experimental design makes this counterfactual reasoning concrete by randomization or matched design, formalized in Rubin's (1974) potential-outcomes framework.
- Factorial Design
- T4 — Controlled-factor clarity versus observational realism. Factorial designs require manipulating each factor to pre-specified levels, which maximizes causal clarity but sometimes requires artificial or unrealistic factor combinations
This sourcePotential-outcomes framework defining causal effects via comparisons under hypothetical treatments holding background fixed; grounds the controlled-manipulation vs observational-realism tension (factorial designs require manipulating factors to pre-specified levels for causal clarity).
- T4 — Controlled-factor clarity versus observational realism. Factorial designs require manipulating each factor to pre-specified levels, which maximizes causal clarity but sometimes requires artificial or unrealistic factor combinations
- Minimal Modification Principle
- A randomized controlled trial applies minimal modification implicitly: randomization holds background variables constant while varying treatment, a logic Rubin (1974) formalized through the potential-outcomes framework for causal inference.
This sourceFoundational potential-outcomes framework: defines causal effects as comparisons of outcomes under hypothetical treatments holding background conditions fixed; formalizes minimal modification implicit in randomized controlled trials and observational designs.
Supported in partVerified against the publisher's abstract
Rubin's abstract supports the framing half of the claim -- randomization controlling extraneous variation for estimating causal effects -- but says nothing about a potential-outcomes formalization of minimal modification.
“Presents a discussion of matching, randomization, random sampling, and other methods of controlling extraneous variation. The objective was to specify the benefits of randomization in estimating causal effects of treatments.”
- A randomized controlled trial applies minimal modification implicitly: randomization holds background variables constant while varying treatment, a logic Rubin (1974) formalized through the potential-outcomes framework for causal inference.
- Preparatory Field Conditioning
- If the focal act would have gone the same way on untouched ground, the upstream work was something else, whatever the schedule called it.
This sourceDefines the effect of a treatment as a comparison between the outcome under treatment and the outcome the same unit would have had untreated, making the untouched condition the reference against which any act is measured.
- If the focal act would have gone the same way on untouched ground, the upstream work was something else, whatever the schedule called it.
- Randomization
- 6. The credibility-restoration property in causal claims — randomization is the unique procedure that, if properly implemented and preserved through analysis, permits credible causal attribution of observed differences to treatment effects rather than to pre-existing selection bias, rendering even single-arm observational associations interpretable as causal within the randomized population
This sourceFoundational potential-outcomes framework: defines causal effects as comparisons of outcomes under hypothetical treatments holding background conditions fixed; formalizes minimal modification implicit in randomized controlled trials and observational designs.
Supported in partVerified against the publisher's abstract
Rubin's abstract supports randomization's value for estimating causal effects but explicitly denies that it is the only sound procedure, so it backs the frame and refutes the word 'unique'.
“The objective was to specify the benefits of randomization in estimating causal effects of treatments. It is concluded that randomization should be employed whenever possible but that the use of carefully controlled nonrandomized data to estimate causal effects is a reasonable and necessary procedure in many cases.”
- 6. The credibility-restoration property in causal claims — randomization is the unique procedure that, if properly implemented and preserved through analysis, permits credible causal attribution of observed differences to treatment effects rather than to pre-existing selection bias, rendering even single-arm observational associations interpretable as causal within the randomized population
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 6 of 8 citations: 5 supported in part, 1 borderline. Each verdict is shown under its citation below, with what in the work backs the sentence.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Links previously used in the corpus¶
Before the registry existed this work was also linked 1 other way.
Registry ID ref:fe1731e00d32 · see in the full table