Research to Story (RTS) Framework
Welcome to Research to Story
Imagine a typical undergraduate research topic. A student arrives with a question about oyster decline in the Chesapeake Bay. But instead of writing a paper about the topic, the student is asked to make something else, something multimodal, let’s stay an audio essay or a podcast episode. The student is at a turning point (and doesn’t know it yet). Is there some way to “present” my findings in a multimodal form that is frictionless. After all, I’ve done all the thinking, now its time to mechanistically present these ideas on some surface.
RTS asks her a different set of questions than a library search tool would:
- What personal connection do you have to this environmental issue? Why does this matter to you specifically, not just academically?
- If you were telling someone about oyster decline over coffee, what would you want them to understand that the data alone can't convey?
- What assumptions are you bringing to this research? What are you hoping to find, or afraid to find?
Why these questions?
These are the types of prompts and questions that can help to open the rhetorical space that makes research storyable. They’re designed to push a student toward re-seeing their inquiry as an experience to shape: perspectives to bring, contradictions to surface, stakes to articulate. A podcast asks what the researcher was willing to inhabit.
Through Movement 1's four reflection rounds, the student engages with AI-generated “Socratic”questions that are built from her own evolving writing. The model draws on what the student has noticed, wondered about, and struggled with. Each round builds on the previous one: the student selects a question, writes a reflection, and the AI uses that reflection to generate deeper questions for the next round. She documents what she observes in her inquiry process, what feels uncertain, and what threads she wants to carry forward. Alongside this core work, she is encouraged to experiment with small audio production tasks that normalize technical friction and help-seeking as part of the work.
After four rounds, she receives an AI-generated synthesis: a reorganization of her own thinking that surfaces patterns, tensions, and narrative possibilities that she developed. This becomes the foundation for a multimodal narrative that is:
- Grounded in research
- Structured for listening
- Carrying new insights
- Surfacing what matters
- Connected to human stakes
She has created a documented trail of her thinking process. Evidence of becoming a researcher who communicates.
This is one way to frame the work I’m trying to do with Research to Story (RTS): a prototype web application wrapped in a pedagogical framework designed to intervene at the metacognitive level of research, helping learners think like researchers who communicate. This site documents what is pedagogically, technologically, and philosophically under the hood.
What RTS Is (and Isn't)
RTS leaves to other tools:
- A search engine or database discovery tool
- An academic research assistant that locates sources
- A literature review generator or citation finder
- A way to outsource the "finding stuff" part of research
RTS is:
- An intervention in habits of mind at the early stages of undergraduate research
- A scaffold for developing research dispositions: curiosity, persistence, rhetorical awareness, comfort with uncertainty
- A system for helping students think like researchers who communicate
- A framework that treats the journey from curiosity to public narrative as intellectually substantive work
If a student enters "I'm researching oyster decline in the Chesapeake Bay," RTS won't provide literature scans. It will generate questions that cultivate dispositions and open up the space for some narrative possibility:
- Personal Connection (Making it storyable through lived experience): What brought you to this topic? What story would you tell if you had to explain why oysters matter using only your own experience? What do you want to understand better?
- Research Potential (Making it storyable through structure and tension): *What tensions or contradictions have you noticed in how this issue is discussed? What are you assuming about causes that you haven't yet verified? Are there perspectives missing from the dominant conversation? Do you have the right kind of evidence?
- Audience Awareness (Making it storyable through sensemaking and conversation): Who needs to care about this beyond environmentalists? How would you make someone who eats oysters understand the stakes differently than someone who doesn't? What might change if people knew this?
These questions do different pedagogical work than library research tools. RTS scaffolds the thinking that happens before, during, and after source-gathering: the metacognitive work of framing, questioning, and narrating that transforms research from compilation into communication. It is designed to be a small intervention that says:
“Hey, by the way, you are not writing a research paper but you have something to say. Prepare yourself for the rhetorical, causal, interpretive labor of constructing how someone could come to find that something themselves.”
The Research-to-Story Bridge
No controversy that research results should be rigorous, dependable, clear, complete, and impactful. Or that compelling stories should capture attention, bring new perspectives, open minds, reveal truths, and connect to lived experience.1
RTS tries to occupy the space where those two sets of qualities converge.
Story is the arrangement of ideas, evidence, compositional strategies, and perspectives - all done in a way that makes “sense” to an audience. Story carries meaning in a journey-shaped vessel from the first question a researcher asks. Story is:
- Sensemaking: organizing complexity for understanding
- Structure: highlighting tension, contradiction, stakes
- Conversation: locating voice in relation to others
- Experience: shaping pacing, tone, mood, surprise for listeners
When it is time to communicate research, how do you prepare a student to reconstruct, for a stranger out there in the world, the journey toward the connections she has made?
The questions that make research rigorous: What assumptions am I making? What contradictions exist? Who benefits from knowing this? These are the same questions that make stories compelling. RTS treats them as interdependent from the start.
A Note on AI Integration
RTS uses generative AI (currently Google Gemini) as a type of orchestration engine. Through intentional system instructions, the model generates questions, reorganizes student reflections, and surfaces patterns in the student's own thinking. For the full technical and philosophical treatment, see A Note on AI Use.
The system instructions are written as instruments focused on cultivating habits of mind. They come from my own reflective practice of working with undergraduate students in research communication courses. They're designed to function the way I often do with students: as a a prompter who asks:
- What do you notice that others might not?
- What feels hard about this, and why might that difficulty be productive?
- How would you explain this to someone who isn't already invested?
- Where's the tension, contradiction, or surprise in what you've discovered?
On the technical side, I’ve tried to log every AI interaction with any “observability” data that is available through Gemini’s API: things like token counts, computational costs, model versions, even traces of the model's intermediate processing (called “thought summaries in the API documentation). The idea is that students would also learn to read AI outputs as other statistical artifacts. When they see a synthesis of their reflections, they also see: "This reorganization required 12,847 tokens of processing. Here's what the model's chain-of-thought looked like. This is a type of pattern-matching and not to read as understanding. That’s your job.”
This attempt at observability, combined with some intentional and interventionist UI language that dethrones the anthropomorphism often deployed in AI discourse, aims to build some critical AI literacy through direct engagement with computational and mechanistic evidence.
- "Generating a summary by identifying the most statistically significant sentences..."
- "This process mimics human synthesis by finding thematic patterns in your writing."
- "The model is generating a summary designed to read like a human interpretation."
- "Read this upcoming summary critically. It's a statistical reflection of your words, not an understanding of your ideas."
About This Site
This site documents the flows, restarts, processes, and pathways of RTS iteratively. It emerges from my experience supporting undergraduate researchers as they navigate the often invisible space between research and storytelling: a space rich with narrative frameworks, audience awareness, rhetorical decision-making, and the challenging of false divides between "making" and "thinking."
In early research communication courses, creativity flourishes when it's tied to inquiry. Students often feel overwhelmed when the instruction is simply "be creative." What works is a structured approach to research-driven storytelling where the story emerges from discoveries the student is actively making. That's where research, writing, and media reinforce each other.
For Developers and Builders: This site also documents the handler trio pattern that makes RTS extensible. I’m writing an Extensibility Guide to show the simple handler trio architectural pattern that powers Movement 1, all Deep Dives, and can be adapted for practically any pedagogical variation.
Take your audience on the same journey of discovery you experienced, making them feel the stakes, understand the complexities, and arrive at the conclusion with a sense of earned insight.
While each module within RTS has been tested formally and improvisationally through coursework and workshops, the current phase of development focuses on what generative AI models can do for orchestration: sequencing and moving inquiry data dynamically across movements, while being explicit about what they cannot do: think, understand, or know.
This project is a learning experiment. I am learning alongside the system: through its gaps, frictions, and possibilities.
Why Research to Story?
Inquiry shapes story. Story reveals inquiry.
RTS scaffolds the research process as a lived, recursive, and relational process: probing, discovering, negotiating medium-specific constraints, and contributing publicly. Students move:
- From what research is about
- Toward what research can do
Through a structured but flexible progression, students are guided to:
- Connect personal curiosity to evolving research questions
- Embrace friction and uncertainty as productive
- Critically engage audiences, infrastructures, and mediums
- Shape inquiry into communicative artifacts that retain complexity, tension, and rhetorical awareness
- See "technology" as another way to structure evidence, sequence ideas, and make rhetorical choices
How the Framework Works: Five Domains of Growth
RTS scaffolds development across five interwoven dimensions that map loosely onto domains of maker growth:
1. RTS Movements
Progressive inquiry stages from personal connection to public storytelling, with prompts that encourage analytical and rhetorical exploration, expansion, and synthesis. For the full pedagogical design of Movement 1, see Movement 1: Spark to Stakes.
In Movement 1, students complete four rounds of reflection where AI generates Socratic questions based on their research topic and previous reflections. Questions like "What personal experience sparked your curiosity about this topic?" or "What contradictions have you noticed that traditional sources don't address?" Students select questions, write reflections, and the cycle repeats, each round building on the last.
After completing Movement 1's four-round cycle, students can choose "deep dive" extensions: focused explorations through specific inquiry lenses like "Spark of Inquiry" or "Puzzles and Unknowns." Each deep dive follows the same 4-round structure but narrows attention to a particular dimension of their thinking. A new module called "The Shape Finder" is currently in development.
Emphasis: structure, pacing, message + insights, evidence, significance
2. Audience and Medium Awareness
Throughout RTS movements, questions continuously refocus students on context:
- Who needs to hear this story?
- Where might your audience encounter this work?
- How does audio change what you can communicate compared to a written essay?
Students engage with questions about sonic affordances, listening practices, and how medium shapes meaning. Audio storytelling is a distinct rhetorical mode with its own affordances.
Emphasis: audience, platform, timeliness, stakes
3. Black Box Micro-Engagements (BBME)
Small production tasks that normalize technical friction, demystify tools, nurture creative resilience, and build digital scholarship habits. The point is simple: early encounters with production tools matter.
Example tasks: Record a 30-second research summary. Layer two audio sources. Create a transition between clips.
After each task, students complete four-part reflections:
- Action Step Completion - What did you make?
- Personal Reflection - Tools used, frustrations encountered, problem-solving approaches
- Relational Reflection - Who or what helped? (YouTube tutorial? Roommate? Library staff?)
- Source Documentation - Cite/link the help you received
This builds habits of understanding production as a social and iterative process.
Emphasis: planning, tools, challenges, problem-solving
4. Reflection Journal Companion
A parallel process-tracing scaffold where students surface what they notice, name tensions, and carry forward living questions: building a visible record of inquiry itself.
At each movement, students respond to three prompts:
- What I Am Noticing - Observations, sparks, emerging insights
- What Feels Hard or Unsettled - Points of discomfort, contradiction, or doubt
- What I Want to Carry Forward - Threads of inquiry, tension, or discovery to nurture in future stages
These accumulate across movements, creating a metacognitive record of how thinking evolves: evidence of risk-taking, response to feedback, and growth as a researcher.
Emphasis: creative choices, feedback response, growth trajectory
5. Generative AI as Orchestration
A generative AI model operates at the API level, configured to interact with structured data stored in a database: a living ledger where every reflection, AI-generated question, synthesis, and tool interaction is recorded. For the complete schema and data architecture, see Database & Living Ledger.
This creates a complete audit trail of computational operations, enabling research into how students interact with AI-scaffolded inquiry. The system generates questions about the student's own thinking. At least, that is the design constraint.
AI scaffolds forward movement through structured prompting, reframing, and tension-surfacing, but never introduces new content. It can only work with what students have already written.
Emphasis: documentation, iteration, traceability
What RTS Does vs. What Other Tools Do
To clarify RTS's distinctive pedagogical position:
| Tool Type | What It Does | Example |
|---|---|---|
| Search/Discovery Tools | Find sources, databases, literature | Google Scholar, library catalogs |
| Research Assistants | Summarize sources, generate citations | Zotero, EndNote, some AI tools |
| Writing Assistants | Generate or improve prose | Grammarly, ChatGPT for drafting |
| RTS | Scaffold metacognitive habits that make research storyable | Generate questions about your thinking that open narrative possibility |
RTS assumes students will use other tools for finding and managing sources. What RTS provides is the thinking infrastructure that makes source work meaningful: the habits of mind that turn information into insight, insight into story, and story into impact.
Who Is This For?
For Students:
- Undergraduates developing research projects into public-facing narratives
- Anyone learning to compose in multimodal formats (RTS is currently focused on audio essays/stories, aka "podcasts")
- Students building critical AI literacy alongside research skills
- Those who want to understand their own inquiry process as it develops
For Instructors:
- Writing program faculty integrating multimodal composition
- Librarians teaching research as iterative inquiry
- Educational technologists exploring AI-enhanced pedagogy
- Anyone interested in scaffolding research as a lived, documented process
For Researchers:
- Those studying human-AI interaction in educational contexts
- Scholars of digital rhetoric and composing processes
- Anyone investigating AI literacy frameworks and pedagogical applications
- Researchers interested in complete audit trails of student-AI interaction
Aligning RTS with the ACRL Framework
RTS lands within the ACRL Framework for Information Literacy in Higher Education's conceptual territory. Rather than treating the six frames as isolated outcomes, RTS designs its movements as lived, recursive practices that help students embody the dispositions and knowledge practices the Framework describes.
Research as Inquiry: RTS treats inquiry as an unfolding posture across every movement. Structured prompts, recursive follow-up questions, and synthesis activities mirror the ACRL emphasis on research as an iterative process of asking, refining, and reframing questions.
Scholarship as Conversation: Through reflective journaling and audio micro-engagements, students see themselves as contributors to ongoing conversations. Movements invite students to situate their voice in relation to others: who has said what, what's missing, what matters.
Authority Is Constructed and Contextual: RTS foregrounds rhetorical stance and audience awareness, helping students notice how credibility and authority shift depending on context, medium, and purpose. Movements that focus on friction, bias, and language framing explicitly ask students to interrogate assumptions about sources and categories.
Information Creation as Process: By integrating Black Box Micro-Engagements, RTS positions tools, media, and production choices as generative parts of research. Students experience firsthand how different modes of creation shape meaning and value.
Searching as Strategic Exploration: Movements that guide keyword expansion, metadata awareness, and source mapping help students recognize searching as a rhetorical and interpretive act, aligning with the ACRL frame that values exploration, flexibility, and discovery.
Information Has Value: RTS emphasizes attribution and acknowledgment practices in its micro-engagements and reflective prompts. Students learn to value the labor of knowledge production, recognize information privilege, and document the relational networks that support their projects.
RTS treats research as an interpretive, rhetorical, and creative process: students are invited to see themselves as learners, makers, and contributors whose work carries both intellectual and public stakes. I think that’s the gist of the Framework.
Current Implementation Status
Fully Operational:
- Movement 1 (4-round Socratic reflection system with AI synthesis)
- 6 Deep Dive categories (focused extensions of Movement 1)
- AI observability features (token analytics, thought summaries, model tracking)
- Instructor Dashboard (course management, student progress monitoring, journey export)
- Export functionality with AI Literacy Lens annotations
- Complete living ledger architecture with Row Level Security
In Development:
- Reflection Journal Companion
- Black Box Micro-Engagement framework
- Scaling Movement 1 patterns to other movements
- Cross-movement synthesis capabilities
- The Shape Finder
The system architecture established in Movement 1: session management, AI interaction patterns, the handler trio, and metadata capture, provides the proven template for scaling to a multi-movement framework.
Footnotes
-
Source: Presenting Research Results: https://shiny.stats4sd.org/PresentingResults_Book/story.html#scientific-results-and-story-telling ↩