# Research-to-presentation storyboard

Pretectal workflow template · October 2, 2026
Guide: https://pretectal.com/workflows/research-to-presentation

Start with one audience decision and one source-supported claim. Complete the evidence map before drafting the teaching sequence. Leave unresolved fields marked NOT LOCATED or NOT REVIEWED.

## 1. Define the decision

- Topic:
- Intended audience and existing knowledge:
- One learning objective:
- Misunderstanding this segment addresses:
- Delivery format and available time:
- Presenter / responsible owner:
- Presentation version:
- Applicable organizer or provider requirements:

## 2. Build the evidence map

CSV template: https://pretectal.com/resources/evidence-map.csv

For each claim, record its exact wording, original source URL or DOI, table / figure / page, population, comparison, outcome, time point, value and unit, uncertainty, limitation, reviewer, and unresolved questions.

Label any unsupported item NOT LOCATED. Keep adjusted and unadjusted results distinct. Do not reconstruct unavailable observations.

Before continuing: open every source location used in this segment. Resolve missing support or remove the claim from the presentation.

## 3. Create the teaching sequence

### Scene 1: Question and prediction

- The decision the learner makes:
- Choices and the misconception each choice represents:
- Information available before answering:
- When feedback becomes visible:

### Scene 2: Evidence

- Claim IDs from the evidence map:
- Chart or visual type and why it fits:
- Source values, units, scale, and uncertainty:
- Labels and accessible text alternative:
- Meaningful limitation displayed with the result:

### Scene 3: Interpretation

- What the finding supports:
- What it cannot establish:
- Speaker notes:
- Question that checks the distinction:

### Scene 4: Application

- New fictional scenario or interpretation question:
- Preferred answer and acceptable alternatives:
- Feedback for every choice:
- Source location supporting each explanation:

## 4. Build and check the interaction

- Interaction file or URL and version:
- Source values checked against the evidence map:
- Every answer and expected feedback:
- Actual feedback observed for every answer:
- Chart labels, scales, signs, units, and uncertainty checked:
- Keyboard and touch paths checked:
- Static fallback location and comparison with the interactive version:
- Open mismatch, owner, and correction needed:

Do not progress with a mismatched value or unsupported feedback state. Recheck affected states after each correction.

## 5. Rehearse and hand off

- Rehearsal date and intended-audience participant or role:
- Participant's explanation of the finding and its limit:
- Confusion observed:
- Change made and version retested:

### Production record

- AI tool, version if known, date, and exact task:
- Material supplied and authorization to use it:
- Draft outputs retained:
- Human corrections and their reason:
- Scientific reviewer and scope actually completed:
- Interface / accessibility / delivery checks:
- Unresolved issues and owners:
- Final version and release decision owner:

### Handoff

- Editable presentation and static fallback:
- Evidence map and source links:
- Asset permissions and attribution:
- Accurate AI disclosure:
- Review record and correction contact:

Completion check: a reader can locate each claim, use the static fallback, and identify the owner of the exact version delivered. Revisit the sequence if the rehearsal exposes a misunderstanding of the result or its limitations.

This template organizes work. It is not an approval, accreditation, or guarantee of accuracy. Keep incomplete review tasks clearly marked.