Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/amr-data-analysis
Use when stress-testing the LOGIC of an Academy of Management Review (AMR) theory manuscript — checking logical coherence, running thought experiments and counterfactuals, addressing alternative explanations and disconfirming cases, and verifying each proposition follows from its argument. This is ARGUMENT DEVELOPMENT, NOT data analysis; AMR publishes no datasets, no statistics, and no empirical results.
Passed 17 of 22 test cases with the skill, and 6 of 22 without it.
and 17 more in the full report
12 fixed5 passed either way4 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 12 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am working on an AMR theory paper extending resource dependence theory to space exploration ventures. P1 states that resource dependence drives cross-sector partnerships, but P3 claims that resource dependence leads firms to isolate themselves from cross-sector partners. Evaluate internal consistency across my theory.
| Checked | Without skill | With skill |
|---|---|---|
| Flags contradiction between P1 and P3 under 【Coherence】. | Missed | Passed |
| Uses double bracket headers such as 【Coherence】. | Missed | Passed |
The instructions your model reads when the skill loads.
AMR publishes NO empirical data. There is nothing to estimate, plot, or test. The "analysis" in an AMR paper is the analysis of the argument itself: does each proposition follow logically from the constructs and mechanisms? At AMR, logical soundness plays the role that statistical rigor plays at empirical journals.
This skill replaces an empirical "identification + robustness" stage. The mapping:
| Empirical sibling (AMJ/ASQ/SMJ) | AMR theory analog |
|---|---|
| Identification strategy (IV, DiD, RD, matching) | Generative mechanism — the why (Whetten 1989, DOI 10.5465/amr.1989.4308371) |
| Robustness checks / alternative specifications | Internal consistency + counterfactual probes on premises |
| Ruling out confounders | Engaging and bettering the strongest rival theory |
| Replication package (data + code) | Transparent reasoning — premises and derivations a reader can re-derive |
| "Estimates are significant and robust" | Propositions are falsifiable in principle (AMR's "testable knowledge-based claims") |
There is no instrument, no parallel-trends test, no placebo here; their presence signals a misfiled empirical paper.
Run every proposition through these before drafting.
For each Pn, write the chain explicitly: premise → premise → mechanism → conclusion. If any step is missing, the proposition is asserted, not derived. Use a Toulmin frame: claim / grounds / warrant / backing / rebuttal. The warrant (the mechanism that licenses the inference) is where most theory papers are thin.
test cases passed, vs 6 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.