Research Paper Deconstructor

Deconstruct the paper below. Audience: a smart reader in an adjacent field who needs to know whether to build on this work, cite it, or ignore it. Faithfulness rule: distinguish sharply between what the paper SAYS and what the paper SHOWS; never fill gaps with what's typical for the field. **Output:** **1. The claim ladder** — - What the title/abstract imply (the claim as it will be tweeted) - What the authors actually claim in the discussion - What the data, as reported, supports These three are rarely the same; the gaps between rungs are the most important output. Quote the paper at each rung. **2. Method in plain language** — what they actually did (design, sample, N, measures, controls) in under 150 words, no jargon survives. Then the three method choices that most constrain interpretation (e.g., convenience sample, proxy outcome, short follow-up, self-report) — stated as constraints, not accusations. **3. Results with effect sizes** — the key findings as magnitudes in real units, not just significance ("treatment group scored 4 points higher on a 100-point scale, p<.05" — and say whether 4 points matters in practice). Flag: any place where the abstract's number and the results-table number differ; subgroup findings not pre-registered; figures whose axes flatter the effect. **4. The caveat graveyard** — limitations the authors buried in the discussion or omitted entirely. Include the ones that only appear when you compare methods to conclusions. **5. Trust calibration** — considering design strength, sample, effect size, and field context: is this "strong evidence", "interesting if it replicates", or "hypothesis-generating only"? One paragraph, committed, with the single follow-up study that would settle it. **6. If I cite this** — the one-sentence version of this paper that is actually defensible. Paper (or the sections I have): {{paper_text}}