Bioinspired Communication & Ethics

Module 2: Proposal Writing & Review

This module provides training in research proposal development and evaluation. We begin with the research landscape — how breakthroughs happen and how funding works — then build proposal writing skills through practice, case studies, and peer review, before examining how AI is reshaping the research enterprise.

Student Profile (Mentimeter survey, n≈18): Most students are Year 2–3, preparing for qualifying exams. 10/14 require a written proposal; 8/16 do a formal proposal defense. Top concerns: novelty, clarity, and feasibility. Priority sections: Background & Significance (9), Research Design/Methods (9), Specific Aims/Hypotheses (7). Confidence in proposal writing: 3.4/5.

---

Module Structure: 8 Lectures

#	Lecture	Key Topics
1	The Research Landscape	Nobel Prize discovery patterns, hypothesis vs. serendipity
2	Research Framing & Word Choice	Hypothesis-driven framing, fundamental vs. applied, NSF directorate culture
3	Funding Agencies & Your First Research Narrative	NSF vs. NIH, drafting challenge & objectives
4	Writing the Research Narrative	Field-specific best practices, NIH specific aims, revision workshop
5	Intellectual Merit, Broader Impacts & Case Studies	NSF review criteria, CAREER case study, review panel simulation
6	AI in the Research Enterprise	AI history, hypothesis generation debate, AI detection failure
7	Peer Review, Ethics & Responding to Critique	Mock panel review, AI evaluation limits, ethical framing
8	GCR Team Proposal Workshop	Growing Convergence Research proposal drafting & cross-team review

📝 Running Assignments

Two threads build across the module — one individual, one team-based:

Individual: Research Narrative Draft

Lecture	Milestone
3	Draft challenge statement (While…However) + 3 research objectives
3	Peer review in pairs (cross-discipline)
4	Revision workshop: strengthen challenge & sharpen objectives
4	“Elevator Test” — 90-second pitch of your narrative

This produces a draft challenge/objectives statement students can use for their qualifying exam proposals.

Team: GCR Convergence Proposal

Lecture	Milestone
8	Draft convergent research question + challenge + objectives + IM/BI
8	Cross-team peer review using simplified rubric
Post-L8	Final revised submission on Blackboard

---

Lecture 1: The Research Landscape — How Breakthroughs Happen

Goal: Before writing proposals, understand what the research enterprise actually looks like — how discoveries are made, what patterns exist across fields, and why this matters for how you frame your own work.

📊 Presentation (~35 min)

• 🔬 How Scientific Breakthroughs Really Happen: 125 Years of Nobel Prizes

  Analysis of Physics, Chemistry, and Physiology/Medicine Nobel Prizes (1901–2025). Key findings: Physiology/Medicine has the highest rate of serendipitous discoveries (~50%); Physics is most theory-driven (~45–50% hypothesis-driven); Chemistry has the strongest tool-building tradition. All fields are shifting toward more hypothesis-driven research over time.

🎯 Team Activity: “Classify Your Own Field” (~30 min)

Each team (15 min discussion + 15 min share-out):

1. Pick 3 landmark discoveries in your team members’ fields (can be Nobel-winning or not)
2. Classify each as hypothesis-driven, discovery-driven, or method-driven — and defend your choice
3. Identify one discovery that doesn’t fit neatly into any category. Why not?
4. Discuss: Is your field becoming more or less hypothesis-driven over time? What’s driving the shift?

This exercise forces students to apply the Nobel analysis framework to their own disciplines, rather than treating it as an abstract historical overview.

💬 Mentimeter Discussion (~15 min)

• Where does your own dissertation research fall on the hypothesis–discovery–method spectrum?
• Has your field had its own “penicillin moment” — a breakthrough no one was looking for?

💡 Key Takeaways

• Different fields have different discovery patterns — there’s no single “scientific method”
• The funding system’s preference for hypothesis-driven proposals creates structural tension with how breakthroughs actually happen
• Understanding these patterns helps you position your own work strategically

---

Lecture 2: Research Framing — Hypothesis, Fundamental vs. Applied, & Word Choice

Goal: Learn why “hypothesis-driven” became synonymous with “fundamental research” in grant writing, where that association breaks down, and how to frame your work strategically without misrepresenting it.

📊 Presentations (~30 min)

• 📈 Research Methodology Patterns Across NSF Directorates

  Analysis of 494 funded grant abstracts (2024–2025) across BIO, PHY, CMMI, CHE, and CBET. Physics favors measurement (47%), Biology/Chemistry favor hypothesis testing (~50–60%), Engineering requires interdisciplinary integration (88% CMMI). Open with this to show students what funded proposals actually look like.

• 🎯 Hypothesis, Fundamental Research & Word Choice

  The critical connection between hypothesis-driven and fundamental research. Includes corrected word choice guidance (words aren’t “forbidden” — context is everything), the 2×2 matrix (hypothesis vs. discovery × fundamental vs. applied), and the NSF quote explicitly endorsing optimization and control as research topics.

✍️ In-Class Practice: Paragraph-Level Reframing (~35 min)

Use the interactive exercises built into the Word Choice presentation (Slides 7–8). Teams work on the realistic paragraph-level prompts:

• Round 1 (15 min): Each team takes one exercise passage (biomedical, environmental/AI, or chemistry/materials). Reframe without changing the actual research.
• Round 2 (10 min): Teams swap and critique each other’s reframes — did they change the research or just the framing? Did they fabricate unsupported numbers?
• Share-out (10 min): Each team presents their best reframe. Class votes on which ones successfully led with the knowledge question without losing the applied context.

💬 Mentimeter Quick Poll

• Which of these words would most trigger “not fundamental” for a reviewer in your field? (Revisit the survey from last year — “Optimize” was #1 at 10/16)

💡 Key Takeaways

• “Hypothesis-driven” is a useful framing strategy, not a logical requirement for fundamental research
• Words need careful framing, not avoidance — “design” is fine when discovering design principles
• Don’t fabricate specific predictions without evidence — reviewers will question unsupported claims
• Match your framing to your target directorate’s culture

---

Lecture 3: Funding Agencies & Your First Research Narrative

Goal: Understand the key differences between NSF and NIH (and other agencies), then immediately apply that knowledge by drafting your first challenge–objective statement.

🏛️ Mini-Lecture: Know Your Audience (~20 min)

NSF vs. NIH at a Glance:

Feature	NSF	NIH
Review criteria	Intellectual Merit + Broader Impacts	Significance, Investigators, Innovation, Approach, Environment
Scoring	Qualitative (E/VG/G/F/P)	Numerical 1–9 (lower is better)
Proposal length	15 pages (project description)	12 pages (research strategy, R01)
Specific aims	Integrated into narrative	Separate 1-page document (critical)
Preliminary data	Helpful but not required	Essentially required for R01
Resubmission	No formal response to reviews	1-page Introduction responding to prior reviews
Broader impacts	Required, weighted equally	Not a separate criterion
Fundamental vs. applied	Strongly favors fundamental framing	Accepts translational and clinical framing

Other key agencies (brief overview): DOE Office of Science, DARPA, private foundations (Sloan, Gates, CZI). Fellowships: NSF GRFP, NIH F31, Ford, Hertz.

Growing Convergence Research (GCR): NSF’s emphasis on deep integration across disciplines — not just collaboration, but disciplines reshaping each other. This connects to the team GCR proposal assignment in Lecture 8.

📚 Key References

• 📄 NSF PAPPG — Proposal preparation requirements
• 📄 NSF Merit Review — Intellectual Merit + Broader Impacts criteria
• 📄 NSF Award Search — Search funded awards in your target program
• 📄 NIH Application Guide — Comprehensive NIH guide
• 📄 NIH Review Criteria (Scoring Guidance) — The 1–9 scoring system explained
• 📄 NIH RePORTER — Search funded NIH projects
• 📄 NSF GCR Program — Growing Convergence Research
• 📄 NSF GRFP — Graduate Research Fellowship

✍️ Practice Session: Draft Your Challenge & Objectives (~60 min)

Individual writing exercise — students work on their own research, not hypothetical examples:

Step 1 — Draft a Research Challenge (20 min):

Using the “While…However” template:

While [broad area] is critical for [benefit/goal], a major challenge is [specific knowledge gap], which limits our ability to [achieve something important]. This gap exists because [current state] fails to [explain/account for phenomenon].

Write one paragraph (4–6 sentences) framing the core challenge of your dissertation research. If you don’t have a dissertation topic yet, frame a challenge from your lab’s recent work.

Step 2 — Draft 3 Research Objectives (15 min):

Following the three-objective framework:

1. Foundational: Establish the core tool, method, or framework
2. Mechanistic: Elucidate underlying mechanisms or test core hypotheses
3. Application/Validation: Apply findings to demonstrate utility

Write one sentence each. Use strong action verbs (test, measure, establish, elucidate, characterize).

Step 3 — Peer Review in Pairs (15 min):

Swap with a partner from a different field. Each reviewer answers:

• Can I understand the challenge even though I’m not in this field?
• Do the objectives directly address the stated challenge?
• Is the framing fundamental or applied? Does it need to be shifted?
• Would I fund this? What’s missing?

Step 4 — Revise and Submit (10 min):

Revise based on partner feedback. Submit on Blackboard for instructor review.

Why this matters: Your Mentimeter survey showed 10/14 students need a written proposal for qualifying exams. This exercise produces a draft they can actually use. It’s the most directly career-relevant activity in the module.

---

Lecture 4: Writing the Research Narrative — From Challenge to Objectives

Goal: Deepen narrative writing skills with field-specific best practices, the NIH specific aims page structure, and hands-on revision of the challenge statements drafted in Lecture 3.

📊 Presentation (~25 min)

• ✍️ Writing Compelling Research Challenges & Objectives

  Field-specific best practices from funded NSF proposals (BIO, PHY, CMMI, CBET, CHE). Covers the three-objective framework in depth with real examples, quantifying ambition, the “if-then” logic connecting challenge to objectives, and common mistakes to avoid.

📋 NIH Specific Aims Page (~15 min)

The 1-page specific aims document is arguably the most important page in biomedical research:

Paragraph 1 — The Hook: Open with the problem’s significance. Why does this matter?

Paragraph 2 — The Gap: What is unknown? What has been tried and why did it fail?

Paragraph 3 — Your Solution: Long-term goal, objective of this application, central hypothesis, and its basis.

The Aims (numbered): 2–3 specific, measurable aims with brief rationale for each.

Paragraph 4 — The Payoff: Expected outcomes and significance.

NIH Content — Placeholder for Development: This section will be expanded with:
• 2–3 real NIH specific aims pages (strong and weak) for in-class analysis
• Before/after comparison showing a weak aims page revised to be competitive
• NIH study section simulation exercise (scoring with 1–9 scale)
• Guest lecturer from an NIH-funded lab or study section member (TBD)
• Key differences in how NIH vs. NSF reviewers evaluate “significance” vs. “intellectual merit”
These materials will be developed in consultation with NIH-experienced faculty and may include a dedicated guest lecture session.

✍️ Revision Workshop: Strengthen Your Lecture 3 Drafts (~40 min)

Students receive written feedback on their Lecture 3 challenge/objectives submissions. Working in pairs:

Round 1 — Strengthen the Challenge (15 min):

• Does your challenge use deficit-based language? (“It remains poorly understood…,” “Current methods fail to…”)
• Is the knowledge gap specific enough? (Not “Cancer is a major problem” but “The lack of real-time biomarkers for X limits early intervention in Y”)
• Apply the “While…However” structure if you haven’t already

Round 2 — Sharpen the Objectives (15 min):

• Do your objectives follow the Foundational → Mechanistic → Application sequence?
• Does each objective start with a strong action verb?
• Can you articulate the “if-then” logic? (“If Obj 1 succeeds, then Obj 2 becomes possible, which enables Obj 3”)
• Are success criteria clear?

Round 3 — The “Elevator Test” (10 min):

Each student reads their challenge + objectives aloud in 90 seconds. Partner answers: “What will you learn?” and “Why should I care?” If the partner can’t answer both, revise.

💡 Key Takeaways

• The NIH specific aims page is a complete argument in one page — master its structure even if you target NSF
• The “While…However” template works across agencies and fields
• “If-then” logic between objectives tells reviewers your plan is coherent
• The elevator test reveals whether your narrative communicates to non-specialists

---

Lecture 5: Proposal Components — Intellectual Merit, Broader Impacts & Case Studies

Goal: Understand what makes Intellectual Merit and Broader Impacts compelling through real case studies, then practice evaluating and writing these sections.

📊 Presentations (~30 min)

• 📋 Writing Excellent Intellectual Merit & Broader Impacts

  NSF’s two review criteria in depth. Side-by-side comparisons of weak vs. strong approaches. Key insight: the strongest proposals weave IM and BI together, rather than treating them as separate sections.

• 🏆 NSF CAREER: A Case Study in Proposal Revision

  Two versions of the same CAREER proposal (2018 unfunded → 2019 funded). The “irregular lattice = FEM” proof was the killer preliminary result. Key lesson: scientific substance beats structural polish.

🎯 Team Activity: “Review Panel Simulation” (~40 min)

Teams role-play as an NSF review panel evaluating two short (1-page) proposal excerpts provided by the instructor:

Step 1 — Individual Review (10 min): Each student reads both excerpts and assigns ratings (Excellent / Very Good / Good / Fair / Poor) for Intellectual Merit and Broader Impacts separately. Write 2–3 sentences of justification for each rating.

Step 2 — Panel Discussion (15 min): Teams discuss as a panel. Appoint a “panel chair” who must synthesize the group’s views. Where do you agree? Where do you disagree? What would you tell the PI to improve?

Step 3 — Panel Summary and Share-Out (15 min): Each panel presents their consensus rating and the single most important strength and weakness they identified. Class compares how different panels rated the same proposals.

Why simulate a panel, not just review? Because the panel discussion is where proposals live or die. Students need to experience how individual ratings get negotiated into a group consensus — and how a champion or detractor can swing the outcome.

📚 References

• 📄 NSF CAREER Program — Current solicitation
• 🔍 Writing a Winning NSF CAREER Proposal — Detailed lessons from proposal versions
• 📄 Science of Science Funding — Research on what makes proposals successful

💡 Key Takeaways

• Preliminary data is the single strongest signal to reviewers
• NSF CAREER rewards transformative vision over conservative scope — “improvements” can weaken proposals
• Broader impacts that emerge from the research are stronger than bolted-on outreach
• Panel dynamics matter — one strong advocate can fund a proposal, one strong critic can sink it

---

Lecture 6: AI in the Research Enterprise

Goal: Examine AI’s role in science through its history, capabilities, limitations, and ethical implications — connecting the proposal module to the paper writing module that follows.

🎮 AI Quiz & History (~25 min)

Interactive Mentimeter quiz on AI history and facts (competitive, with leaderboard):

• Who is the most famous figure in AI history? (Students said Sam Altman and Alan Turing — but 0/16 named Donald Hebb)
• First formal model of an artificial neuron? (13/16 correctly identified McCulloch & Pitts, 1943)
• How many citations does ResNet have? (Answer: ~300K — students’ guesses ranged from 10K to 1M)
• “1960s or 2020s?” exercise — students try to date AI claims:
  • “Machines will be capable, within twenty years, of doing any work a man can do” → 1960s (11 correct)
  • “We are at the beginning of a new era… a new reasoning engine” → 2020s (10 correct)
  • “Within a generation… the problem of creating ‘artificial intelligence’ will be substantially solved” → 1960s (7 correct — many students thought this was modern!)
  • “The ultimate goal is to make computers that can learn to solve problems on their own” → 1960s (10 correct)
  • “AI is probably the most important thing humanity has ever worked on” → 2020s (13 correct)
  • “We are building ‘electronic brains’ that will one day be our partners in thought” → 1960s (13 correct)

The lesson: AI hype cycles repeat. The rhetoric of the 1960s is nearly indistinguishable from 2020s discourse. This is a science communication case study in real time.

🧩 AI Winter Jigsaw Discussion (~20 min)

Using the Toosi et al. (2021) paper “A brief history of AI: how to prevent another winter”:

• Groups 1 & 2 (The Analysts): Diagnose the causes. What specific factors led to the loss of faith and funding?
• Groups 3, 4 & 5 (The Strategists): Extract the lessons. What concrete practices should researchers adopt to build sustainable, ethical, and well-communicated progress?
• 15 min in groups → 5 min share-out of key insights

🤖 Mini-Debate: Can AI Generate Research Hypotheses? (~20 min)

Prompt with data: “GPT-4 generated 100 hypotheses in 3 hours; experts rated 40% as ‘plausible.’ A PhD student generates 5–10 hypotheses over 3 years.”

• Last year’s survey: 11/16 students said AI cannot generate research hypotheses
• Teams take assigned positions (pro/con/nuanced) and debate for 15 min
• Close with: What does “generate” mean? Is pattern-matching the same as understanding?

🔎 AI Detection Failure (~15 min)

Brief presentation of the AI detection results from the CAREER proposal reviews:

• 🔎 Can AI Detect AI-Written Text?

  Three AI models achieved 0–20% accuracy. The “angry reviewer” persona (AI-generated) fooled all detectors. Key implication: detection is unreliable; focus on demonstrating understanding instead.

Connect to students’ survey data: trust in AI for funding decisions was 2.7/5; 13/18 preferred human-centric AI role.

🌉 Bridge to Module 4

Close with: “You’ve now seen AI generate hypotheses, write proposal reviews, evaluate proposals, and create research timelines. In the Ethics module, we ask: what are the ethical responsibilities when using these tools? The same tensions between capability and judgment apply, but the stakes are different.”

📚 Reading

• 📄 A brief history of AI: how to prevent another winter — Toosi et al. (2021), arXiv
• Geoffrey Hinton’s and Daniel Kahneman’s perspectives on AI communication (discussed in class)

💡 Key Takeaways

• AI hype cycles repeat — the rhetoric of the 1960s is nearly indistinguishable from 2020s discourse
• AI can generate plausible-sounding hypotheses but cannot evaluate their scientific merit
• AI detection is unreliable — focus on demonstrating understanding, not catching cheaters
• Understanding AI’s history helps you communicate about AI more responsibly in your own proposals

---

Lecture 7: Peer Review, Ethics & Responding to Critique

Goal: Experience the reviewer’s perspective, develop constructive review skills, and engage with ethical questions about AI, framing, and responsible communication in science.

📊 Case Study: AI Over-Values Structure (~15 min)

• 🤖 NSF CAREER: AI Analysis Case Study

  AI models attempted to evaluate which CAREER proposal version was stronger — and got it wrong. AI over-valued structural polish (“modular, reviewer-friendly”) and under-valued the scientific substance (preliminary data, demonstrated feasibility).

Discussion prompt: “If AI can’t reliably evaluate proposals, what CAN it usefully do in the review process?” Connect to the student survey showing 13/18 prefer human-centric AI roles.

✍️ Mock Panel Review (~45 min)

Teams conduct a formal mock panel review of a provided proposal excerpt (different from Lecture 5):

Step 1 — Individual Written Review (15 min): Using simplified NSF criteria, each student writes a structured review:

• Summary of proposed research (2–3 sentences)
• Intellectual Merit: Key strength and key weakness
• Broader Impacts: Key strength and key weakness
• Overall rating (E/VG/G/F/P) with justification
• One specific, constructive suggestion for improvement

Step 2 — Panel Deliberation (15 min): Teams discuss as a panel. The panel chair must:

• Identify where reviewers agree and disagree
• Facilitate discussion of the most contentious points
• Reach a consensus recommendation (fund / fund with revisions / do not fund)

Step 3 — Writing the Panel Summary (15 min): Each team writes a 1-paragraph panel summary that captures their consensus and the key reasons. Submit on Blackboard.

💬 Ethics Discussion (~20 min)

Structured around the Caltech case study (José Andrade’s mechanics course redesign: “When knowledge is instantly available, judgment becomes the differentiator”):

• Is using AI to draft a proposal section ethical? Where is the line between tool and ghost-writer?
• If you can’t detect AI-generated text, should we stop trying? What replaces detection?
• How does framing your research differently for different agencies relate to honesty? Is strategic framing the same as spin?
• Your students will write proposals that AI could write faster. What makes your proposal worth funding?

📚 References

• 📄 Conflict in Research Groups: Four Strategies — Relevant for collaborative review

💡 Key Takeaways

• Constructive peer review is a professional skill — “this is weak” without specifics doesn’t help anyone
• Panel dynamics (champion vs. detractor) matter as much as individual review quality
• Ethical framing isn’t about avoiding strategy — it’s about ensuring your framing honestly represents your research
• AI is a tool for drafting, not a substitute for scientific judgment

---

Lecture 8: GCR Team Proposal Workshop

Goal: Apply everything from the module by developing a Growing Convergence Research (GCR) proposal as a team, then conducting cross-team peer review.

📋 GCR Proposal Assignment Overview (~10 min)

Each team develops a Growing Convergence Research proposal that integrates skills from the entire module:

Requirements:

• A convergent research question where disciplines genuinely reshape each other (not just collaborate)
• A challenge statement using the “While…However” structure
• 3 specific objectives following the Foundational → Mechanistic → Application framework
• Intellectual Merit section (1 paragraph)
• Broader Impacts section (1 paragraph)
• Framing appropriate for a target NSF directorate (specify which one)

What “convergence” means (vs. multidisciplinary or interdisciplinary):

	Multidisciplinary	Interdisciplinary	Convergent
Structure	Disciplines work side by side	Disciplines integrate methods	Disciplines reshape each other
Example	A biologist and engineer share data	An engineer uses biological models	Biology and engineering co-create a new framework neither could conceive alone
GCR standard	✗	Partial	✓

✍️ Team Working Session (~40 min)

Teams draft their GCR proposals. Instructor circulates to provide feedback. Key checkpoints:

• Minute 10: Each team should have their convergent research question. Instructor spot-checks: “Is this truly convergent, or just interdisciplinary?”
• Minute 25: Challenge statement and objectives drafted. Quick pair-check with another team.
• Minute 40: Full draft ready for cross-team review.

🔄 Cross-Team Peer Review (~30 min)

Each team reviews another team’s draft using a simplified rubric:

1. Convergence Quality: Do the disciplines genuinely reshape each other, or is this just collaboration? (1–5)
2. Challenge Clarity: Can I understand the knowledge gap without being in this field? (1–5)
3. Objective Logic: Do the three objectives follow the “if-then” chain? (1–5)
4. Framing Match: Is the language appropriate for the target directorate? (1–5)
5. One specific suggestion for strengthening the proposal

Teams return reviews; original teams have 10 minutes to discuss and plan revisions.

Final submissions (revised based on peer review) are due on Blackboard by the following week.

💡 Key Takeaways

• True convergence means disciplines reshape each other — not just contribute side by side
• The “While…However” structure and three-objective framework apply to team proposals just as they do to individual ones
• Cross-team peer review reveals whether your proposal communicates to non-specialists
• Everything in this module — framing, narrative, review criteria, ethical reasoning — comes together in this exercise

---

---

🏥 NIH Content Development Plan

Current status: The module’s interactive materials are primarily NSF-focused. NIH content exists as structural guidance (specific aims page format, comparison table) but lacks the depth of interactive materials available for NSF.

Planned development (for future iterations):

Priority 1 — Real NIH Specific Aims Pages: Collect 2–3 publicly shared specific aims pages (strong and weak) for in-class analysis. Many funded PIs share these on lab websites or through institutional resources. These would anchor Lecture 4’s NIH section, replacing the current text-only template.

Priority 2 — Before/After Aims Page Revision: Develop a case study showing a weak specific aims page revised to be competitive — parallel to the NSF CAREER 2018→2019 case study. Ideally from a real resubmission, anonymized with permission.

Priority 3 — NIH Study Section Simulation: Design a mock study section exercise where students score a proposal excerpt using NIH’s 1–9 scale across all five criteria (Significance, Investigator, Innovation, Approach, Environment). This would parallel the NSF panel simulation in Lectures 5 and 7.

Priority 4 — Guest Lecturer: Invite an NIH-funded faculty member or current/former study section member to discuss how NIH review actually works in practice — especially the differences from NSF that can’t be captured in a comparison table (e.g., the role of the Scientific Review Officer, triage processes, payline politics).

Priority 5 — NIH-Specific Framing Guide: Develop an interactive presentation on NIH framing conventions — how “significance” differs from NSF’s “intellectual merit,” how to frame translational work, and how the innovation criterion is evaluated.

---

📚 Module Activities & Learning Objectives

Through this module, students develop skills in:

• Research Landscape Awareness: Understanding how breakthroughs happen across fields and how funding agencies evaluate research
• Strategic Framing: Matching proposal language and structure to agency culture and directorate expectations
• Research Narrative Craft: Writing compelling challenges, objectives, and significance statements — including practice with their own dissertation research
• Proposal Evaluation: Conducting mock panel reviews using official NSF criteria
• Peer Review: Providing constructive written feedback and participating in panel deliberation
• AI Literacy: Understanding AI’s capabilities and limitations in research — from hypothesis generation to proposal review to detection
• Ethical Reasoning: Engaging with questions about framing, AI use, and responsible communication

» Detailed assignment instructions, rubrics, and submission portals are available on the course Blackboard site.

---

📚 Additional Resources

• 📄 How to Write a Grant Proposal (Nature) — Step-by-step guide
• 📄 NSF Broader Impacts Network — Practical examples and templates
• 📄 Grant Application Writer’s Workbook — Comprehensive proposal writing guide
• 📄 NIH Funding Mechanisms — R01, R21, R03, F31, K99/R00 explained

---

← Scientific Writing & Peer Review Research Ethics →