From restat-skills
Stakes the marginal contribution of a REStat manuscript against prior work when positioning is fuzzy or oversold. Focuses on precise deltas rather than literature tours.
How this skill is triggered — by the user, by Claude, or both
Slash command
/restat-skills:restat-literature-positioningThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- The intro's "we contribute to several literatures" reads as a citation dump
REStat referees are applied economists who know the closest five papers cold. Positioning must be a precise delta against named work, not a tour of a literature. Because REStat prizes measurement and credible identification, the most persuasive deltas are of two kinds: (1) identification delta — your design credibly estimates a parameter prior work could only correlate or estimate under weaker assumptions; (2) measurement delta — you measure the object better (new data, less error, a validated index) and that changes an applied conclusion. "First to study" is rarely the contribution; "first to credibly identify / cleanly measure, and it overturns/sharpens [conclusion]" is.
| Pattern | Example shape |
|---|---|
| Identification upgrade | "Prior work used OLS/TWFE; we use [RD / IV / het-robust DID] and the estimate changes" |
| Measurement upgrade | "Prior work proxied [X] with [noisy measure]; our better-measured [X] reverses the sign/magnitude" |
| Reconciliation | "Estimates disagree across studies; we show the gap is [selection / measurement / specification]" |
| Better-powered / wider data | "We bring [bigger / cleaner / linked] data that resolves an under-powered debate" |
| Applied use of a method | "We apply [estimator] to [setting] where its assumptions hold, recovering [parameter]" |
A draft opens: "A large literature studies the minimum wage's employment effects (A, B, C…). We add to this literature." A REStat referee reads that as no contribution. The fix is to name the single closest paper and state the delta: "Paper B estimated this with TWFE across staggered state increases and found a small disemployment effect; we re-estimate with a heterogeneity-robust estimator on the same variation and a better-measured teen-employment series, and the effect is statistically indistinguishable from zero — the earlier negative estimate was driven by forbidden comparisons and measurement error." That is a delta (identification + measurement) that reconciles a contested estimate, with stakes a referee can weigh.
【Closest prior work】[paper1 / paper2 / paper3 (/4/5)] by author-year
【Delta vs each】p1: [...]; p2: [...]; p3: [...]
【Delta type】identification | measurement | setting | reconciliation
【Stakes】what changes if we are right: [...]
【Scope NOT claimed】[...]
【Next step】restat-identification
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin restat-skillsPositions a REStud manuscript against the closest related work by confronting nearest papers and stating the marginal contribution precisely.
Positions the marginal contribution of an AEJ: Applied manuscript precisely against prior applied-micro work. Use when a contribution is fuzzy, undersold, or risks reading as a replication.
Sharpens the contribution-vs-frontier claim for EER manuscripts when the novelty gap is fuzzy, oversold, or under-cited. Helps name the closest papers and state a checkable marginal contribution.