Reference health
Why Experimenters Might Not Always Want to Randomize, and What They Could Do Instead
9/9 checkable references clean · checked 2026-09-09
Every reference with a DOI in the deposited reference list resolved to a known work in Crossref or DataCite at the dated check, and none carried a retraction, withdrawal, or removal notice.
10 without a DOI — not checked. A reference deposited without a DOI is never matched by title or guessed at; it stays outside the checked set, and this line discloses that.
The 9 checked references that resolve
resolves10.1111/ajps.12162
The Effect of Fact‐Checking on Elites: A Field Experiment on U.S. State Legislators
resolves10.1111/ajps.12141
Causes of Noncompliance with International Law: A Field Experiment on Anonymous Incorporation
resolves10.7910/DVN/I5KCWI
Matlab implementation for "Why experimenters might not always want to randomize, and what they could do instead"
resolves10.1017/S0003055413000543
How to Promote Order and Property Rights under Weak Rule of Law? An Experiment in Changing Dispute Resolution Behavior through Community Education
resolves10.1111/ajps.12180
Campaign Contributions Facilitate Access to Congressional Officials: A Randomized Field Experiment
The 10 references without a DOI — listed, not checked
no DOI — not checkedThe prior used to calculate this EMSE assumes that C((x 1, d 1), (x2 , d 2)) = 10 * exp(–(||x 1 – x 2||2 – (d 1 – d 2)2)/10); for details, see Section 4.
no DOI — not checkedSquared error loss is the canonical loss function in the literature on estimation. It has a lot of convenient properties allowing for tractable analytic results, and in keeping with the literature we will focus on squared error loss. Other objective functions are conceivable, such as for instance expected welfare of treatment assignments based on experimental evidence; see for instance Kasy (2014).
no DOI — not checkedThis is the ex post bias, given covariates and treatment assignment. This is the relevant notion of bias for us. It is different from the ex ante bias, which is the expected bias when we do not know the treatment assignment yet.
no DOI — not checkedThe code producing Table 1 is available online at Kasy (2016). At this address, we also provide Matlab code for implementing our proposed approach in practice.
no DOI — not checkedConsider for instance the setting of Nyhan and Reifler (2014), where legislators i received letters about fact-checking (Di = 1) or did not (Di = 0). The outcome of interest Y in this case is the future fact-checking rating of these legislators; an important covariate Xi might be their past rating.
no DOI — not checkedThis class of estimators includes all standard estimators of β under unconfoundedness, such as those based on matching, inverse probability weighting, regression with controls, kernel regression, series regression, splines, etc. Linearity of the estimator is unrelated to any assumption of linearity in X; we are considering the posterior BLP of β in Y rather than the BLP of Yi in Xi .
no DOI — not checkedTo be precise, the support of these priors is the closure of the Reproducing Kernel Hilbert Space corresponding to C. For further discussion, see the Online Appendix.
no DOI — not checkedIf experimenters have a preference for randomization for reasons outside the decision problem considered in the present article, a reasonable variant of the procedure suggested here would be to randomize among a set of assignments that are “near-minimizers” of risk. If we are worried about manipulation of covariates, in particular, a final coin flip that possibly switches treatment and control groups might be helpful. I thank Michael Kremer for this suggestion.
no DOI — not checkedKasy M. 2014. Using data to inform policy. Working Paper.
no DOI — not checkedI thank Larry Katz for pointing this out.
What this badge says. CiteStamped means the CHECKABLE references of this
work were clean at the dated check: each resolved to a known work in a public registry, and none
carried a retraction notice at that time. It says nothing about the quality, findings,
or importance of the work itself, and nothing about references deposited without a DOI.
checked 2026-09-09 — re-checked daily as this page is visited; titles and statuses come from Crossref and DataCite and are not part of the signed record
Embed this badge
Both snippets point at the live badge image and link back to this page. The badge re-renders from the daily check, so an embed never goes stale by more than a day of visits.
<a href="https://citestamp.com/citestamped/10.1093/pan/mpw012"><img src="https://citestamp.com/citestamped/10.1093/pan/mpw012/badge.svg" alt="CiteStamped reference-health badge" width="460" height="64"></a>
[](https://citestamp.com/citestamped/10.1093/pan/mpw012)