# Jru Writing Style > Use when the prose of a Journal of Risk and Uncertainty (JRU) manuscript must convey a decision model and its measurement with precision. Sharpens abstract, intro, and exposition for a risk/uncertainty audience; it does not invent evidence or citations. - Skill: `thedixitjain/jru-writing-style` (Agent Skill) - Install (CLI): `npx skillmds add thedixitjain/jru-writing-style` - Raw SKILL.md: https://api.skillmd.com/api/skills/thedixitjain/jru-writing-style/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: thedixitjain (https://skillmd.com/u/thedixitjain) - Updated: 2026-09-08 - Page: https://skillmd.com/skills/thedixitjain/jru-writing-style --- # Writing Style (jru-writing-style) ## When to trigger - The abstract says "we study risk preferences" without naming the primitive, the method, or the headline number - The introduction motivates with a policy anecdote but never states the decision-theoretic claim - Notation drifts: the same symbol means utility in one section and value in another, or w(p) is never formally introduced - The paper reads as either pure theory or pure empirics when JRU readers expect the bridge between them ## The JRU voice JRU prose is **precise about the decision model and honest about measurement**. The reader is a risk-and-uncertainty specialist who wants, fast: which primitive the paper moves, how it was elicited or estimated, and how big and how precise the effect is. The house register is technical but not ornamental — the model earns its symbols, and every claim is calibrated to what the identification supports. ### Abstract (treat each sentence as load-bearing) 1. The primitive and the question ("How does ambiguity attitude shape insurance takeup?"). 2. The method in a phrase (incentive-compatible elicitation; hedonic-wage estimation; an axiomatization). 3. The headline result as a *number with a direction* (a magnitude, not "significant"). 4. The decision-theoretic interpretation — what it implies for EU vs. its alternatives. Keep within the journal's abstract limit (待核实 — verify exact word cap on the official guidelines). ### Introduction - Open on the **decision-theoretic puzzle**, not a news hook. The motivating tension should be a gap in how we represent or measure a risk primitive. - State the contribution as a property of a primitive in the first page (mirror `jru-literature-positioning`). - Preview the identification in one or two sentences — JRU readers want to know *what pins the parameter* before the results. - Tell the reader what the paper does **not** claim; over-claiming is the fastest route to a skeptical referee. ### Exposition - Introduce the representation formally before interpreting it; define u, w(p), the prior set, the reference point once and use them consistently. - Report effects as **magnitudes with precision**, in interpretable units (a VSL in dollars, an elasticity, a λ). - Keep the experiment's incentive structure and the estimand visible in the main text, not buried in a procedures appendix. ### Calibrating claims to the evidence JRU's referees are quick to flag prose that runs ahead of the identification. A few habits keep the register honest: - **Match the verb to the design.** Write "is associated with" for a correlation, "raises" only where the identification supports causation, "is consistent with" when a model fits but is not uniquely identified. - **Quantify, then interpret.** Lead with the number and its precision; follow with the decision-theoretic reading. Avoid an interpretation the confidence interval cannot bear. - **Name the scope of the claim.** A lab estimate of loss aversion is a statement about the elicited population and stakes, not a universal constant — say so. - **Reserve "robust" for the robustness section.** Do not assert robustness in the intro before the checks in `jru-robustness` are reported. ### Section-by-section register - *Model section:* terse and formal — definitions, then results, then one paragraph of interpretation. - *Design / data section:* concrete — the mechanism, the stakes, the estimand, the sample, in plain sequence. - *Results section:* numbers with precision first, decision-theoretic meaning second. - *Discussion:* the boundary of the claim and what it implies for EU vs. its alternatives, without re-litigating the introduction. ## Checklist - [ ] The abstract names the primitive, the method, a signed magnitude, and the EU-vs-alternatives interpretation - [ ] The intro opens on a decision-theoretic puzzle and states the contribution within the first page - [ ] The identification is previewed before the results - [ ] Notation for u, w(p), priors, and reference point is defined once and used consistently - [ ] Effects are reported as magnitudes in interpretable units, with precision - [ ] The paper explicitly states what it does not claim - [ ] Abstract/length conform to the journal limits (verify exact caps; mark 待核实 if unconfirmed) ## Anti-patterns - An abstract that says "significant effects" without a number or a direction - A policy-anecdote opening that never reaches the decision model - Symbol drift (u vs. v, w(p) introduced only in a footnote) - Hiding the incentive mechanism or estimand in an appendix while the intro makes strong claims - Prose that over-states generality beyond what the elicitation/estimation supports ## Worked vignette (illustrative) A draft abstract reads: "We run an experiment on decision-making and find significant behavioral effects." The JRU rewrite: "We elicit ambiguity attitudes with an incentive-compatible matching-probabilities task (N illustrative) and estimate an α-MEU model. Ambiguity aversion (α illustrative 0.62) raises the willingness to pay for an unambiguous prospect by [magnitude], a wedge expected utility cannot generate. The result implies insurance demand responds to *ambiguity*, not just to risk." Same study — but the primitive, method, magnitude, and theoretical stakes are all visible. ## Terminology a JRU reader expects you to use precisely Sloppy vocabulary signals a non-specialist. Use these distinctions exactly: - **Risk** (known probabilities) vs. **uncertainty / ambiguity** (unknown probabilities) — never interchangeable. - **Risk aversion** (utility curvature) vs. **probability weighting** (distortion of p) vs. **loss aversion** (asymmetry around a reference point) — three different things, not synonyms for "cautious." - **Value** v(·) (prospect-theory carrier of utility over gains/losses) vs. **utility** u(·) (EU carrier over final wealth) — keep the symbols distinct. - **Certainty equivalent**, **risk premium**, **willingness to pay** — defined precisely, not used loosely. A paper that uses these terms cleanly reads as written by an insider; one that blurs them invites the referee to doubt the rest. ## Output format ```text 【Journal】Journal of Risk and Uncertainty 【Skill】jru-writing-style 【Verdict】lands / tighten / rebuild abstract 【Abstract has】primitive + method + signed magnitude + EU-vs-alt interpretation [Y/N] 【Intro】decision-theoretic puzzle + contribution on page 1 [Y/N] 【Notation】consistent u / w(p) / priors / reference point [Y/N] 【Limits】abstract/length within journal caps (待核实) [Y/N] 【Next skill】jru-replication-package ``` --- **Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `Journal-of-Risk-and-Uncertainty-Skills/skills/jru-writing-style/SKILL.md`