From ajps-skills
Defends research design for AJPS manuscripts: causal identification, experimental/survey design, formal-empirical linkage, case-based inference. Strengthens arguments without writing code.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ajps-skills:ajps-research-designThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
AJPS publishes many methods but holds identification and inference to a high standard. The design must
AJPS publishes many methods but holds identification and inference to a high standard. The design must
credibly connect the argument (ajps-theory-building) to evidence and rule out the strongest rival.
This skill is mode-aware: pick the section that matches your work and defend it on its own terms.
ajps-literature-positioningajps-submission).ajps-replication-and-verification, qualitative path).For the single strongest rival explanation, write: "If the rival were true rather than my argument, the data would look like ___; instead they look like ___." If you cannot, the design does not yet identify the contribution.
| Design | What the referee demands | Common desk-reject / reject trigger |
|---|---|---|
| RDD | Density/manipulation test, bandwidth robustness, no sorting at cutoff | Treating a non-discontinuous threshold as sharp |
| DID / event study | Modern staggered-adoption estimator, pre-trend evidence | Naive TWFE with heterogeneous timing |
| IV | First-stage strength, defended exclusion, weak-IV-robust CIs | "Plausibly exogenous" instrument with no defense of exclusion |
| Matching/weighting | Balance + unobserved-confounder sensitivity bound | Selection-on-observables read as clean causation |
A close-election RD on incumbency states the estimand (local effect of barely winning on next-cycle vote share at the threshold) and the continuity assumption that licenses it. The density test shows no sorting (illustrative p = 0.62); the estimate is stable across bandwidths h = 0.08-0.16; a donut-hole spec holds. The adjudication sentence: if incumbency advantage were candidate-quality persistence rather than an officeholding effect, the jump at the bare-win threshold would vanish — instead it is +6 points (illustrative). That sentence converts a quantitatively demanding AJPS referee.
Calibration anchor: AJPS spans American, comparative, IR, theory, and methods, but applies a hard premium on credible identification across all of them; confirm any human-subjects/IRB specifics against the journal's current submission guidelines.
【Mode】quant-causal / experiment / formal-empirical / case-based
【Estimand or claim】what is being identified/shown
【Key assumption(s)】and how each is defended
【Rival ruled out】the adjudication sentence
【Inference】clustering / RI / small-cluster correction
【Robustness/sensitivity】planned checks
【Next】ajps-data-analysis
../../resources/external_tools.md — design/identification packages (R/Stata/Python) and CAQDAS for qualitative work../../resources/official-source-map.md — human-subjects / IRB requirements and submission policynpx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin ajps-skillsDefends research design for APSR manuscripts: causal identification, case selection, process tracing, experimental design, and formal-empirical linkage.
Defends research designs for JOP (Journal of Politics) manuscripts — covering causal identification, experimental/survey design, formal-empirical linkage, and qualitative case selection/process tracing. Does not write code.
Defends research design for BJPS manuscripts: causal identification, case selection, process tracing, experimental design, and formal-empirical linkage.