---
name: research-communication
description: Translate research into an audience-appropriate talk, summary, explanation, or public narrative while preserving the evidence, uncertainty, ownership, and practical limits.
---

# Research communication

Use this Academic Research Assistant skill for conference talks, public summaries, research pitches, interviews, and explanations outside a manuscript's usual specialist form. Follow the [research contract](../../docs/RESEARCH-CONTRACT.md). Accessibility changes the language and route through the evidence; it must not strengthen the claim beyond what the work supports.

## Inputs and audience

Obtain the source research, audience, purpose, medium, available time or word limit, required content, and the speaker's role in the work. Establish what the audience already knows, what they need to understand, and what action or judgment the communication is intended to support. If sources or results are missing, develop the structure with explicit placeholders rather than inventing a compelling finding.

Distinguish scientific outcomes from applications, proposals, and personal motivations. A laboratory capability is not automatically a demonstrated application; a numerical result is not an observation of nature; a promising hypothesis is not an established mechanism. Preserve collective and individual contributions accurately.

## Develop the narrative

1. Write the central message in one or two plain sentences and identify the evidence required to justify it. Choose a narrative based on the audience's question: a problem and its resolution, a surprising observation and competing explanations, or a capability and its limits.
2. Select the minimum background needed to understand the result. Introduce terms when they become useful. Retain essential definitions and distinctions, while moving optional derivations and technical contingencies to supporting material.
3. Build a clear chain from question through approach to finding and implication. A listener should know why a method was used, what it measured or established, and what the result changes. Avoid a chronology of everything the project team did unless chronology explains the discovery.
4. Translate technical quantities carefully. Preserve units, denominators, comparison groups, scales, and uncertainty when they affect interpretation. Use concrete analogies only within a stated range of correspondence; identify where an analogy stops working.
5. Show evidence with readable visuals. Choose an appropriate chart or diagram, label axes and comparisons, and explain what the audience should notice. Do not remove inconvenient data, truncate scales misleadingly, or imply precision absent from the underlying evidence. Use [scientific visuals](../scientific-visuals/SKILL.md) when producing research figures.
6. Write transitions and a speakable script when requested. Estimate timing from actual content and presentation style, then recommend or perform rehearsal where possible. A word-count estimate is a planning tool, not proof of live duration.
7. Prepare likely questions and answers that separate established knowledge, the project's evidence, unresolved issues, and future work. Keep deeper material accessible without allowing it to overwhelm a hard presentation limit.

## Research-mode distinctions

For quantitative research, explain effect magnitude and uncertainty rather than using statistical significance as a synonym for importance. For qualitative research, convey context, meaning, and variation without presenting selected quotations as a representative survey. Preserve consent and anonymity when choosing examples.

For theoretical work, explain what assumptions make the result true and why the result matters; an intuitive picture should not be represented as a proof. For computational work, say what was modeled, tested, or benchmarked and where real-world validation remains incomplete. Make clear when a visual is schematic, simulated, observed, or an artist's illustration.

## Deliverable and verification

Deliver the requested format: summary, script, explanation, narrative outline, audience-specific abstract, or talking points, with an evidence map for consequential claims. Include source links or notes at a level suited to the medium. Route actual slide or poster production to [presentations and posters](../presentations-posters/SKILL.md).

Verify names, numerical comparisons, source attributions, and ownership statements against the supplied evidence. Check readability, accessibility, medium constraints, and whether caveats remain near the claims they qualify. End with the intended message or action grounded in the work, without guarantees of societal impact, clinical benefit, novelty, or future success. Clearly label a written timing estimate, a rehearsed script, and a delivered talk as different completion states.