Capstone — Design a Research Study

You Made It Here

If you completed all the previous modules in this track, you now have the conceptual foundation that most undergraduate marine biology majors don't fully have until their third year. You can read a scientific paper, evaluate data critically, understand what's happening in an ecosystem from the producers up, and explain how policy decisions shape what gets studied and what gets protected.

This module is different from the others. There's no new content to learn. Instead, you'll do what working scientists do — design a real research study. The goal isn't to actually conduct the research (though you can if it's feasible). The goal is to walk through the full intellectual exercise of going from curiosity to research plan.

This is a capstone, not a quiz. There are no right answers — only well-designed and badly-designed projects. Take it seriously.

---

Step 1: Find a Question

The first move in any research project is finding a question that's:

Specific. "How do oceans work?" isn't a research question. "How does the timing of phytoplankton blooms off the California coast respond to El Niño events?" is a research question.

Tractable. A question is tractable if there's some realistic method that could answer it. "What is consciousness?" is not currently tractable. "Does eDNA from harbor seal scat show evidence of recent salmon consumption?" is tractable.

Original. Your question shouldn't have an obvious answer already in the literature. The Foundations F3 module gave you the tools to assess this — search PubMed, Google Scholar, and relevant databases to verify your question hasn't already been answered.

Personally interesting. Research takes time. Months at minimum, years if you go deeper. If you're not actually curious about the answer, the work will feel like drudgery and the quality will suffer.

Some example question types to consider:

• Descriptive — "What species are present in [location] using eDNA methods?"
• Comparative — "Do MPAs in Region X show higher biodiversity than nearby unprotected sites?"
• Mechanistic — "What environmental factors predict the timing of coral spawning at Site Y?"
• Predictive — "Can satellite chlorophyll data predict sardine population trends?"
• Applied — "Which restoration coral genotypes show the highest survival in current Bay Z conditions?"
• Policy-relevant — "How effective is the current quota system at protecting Atlantic cod populations?"

Pick one question. Not three. One. Write it down in a single sentence.

---

Step 2: Background and Hypothesis

Once you have a question, the next step is grounding it in existing literature and forming a specific hypothesis.

Background section (write 200–400 words):

• What is already known about your topic?
• What is the gap in current understanding that your question addresses?
• Why does answering this question matter? (Scientifically? For conservation? For policy?)

This is the Introduction equivalent in a real paper. Cite at least 5 relevant sources — actual papers, not blog posts. Use PubMed, Google Scholar, or your library's databases.

Hypothesis section (write 50–150 words):

State your specific, testable, falsifiable hypothesis. Remember from F6: a hypothesis must make a prediction that could be wrong. "Bleaching will affect coral" isn't a hypothesis. "Coral colonies in the X reef will show a 50% bleaching response when sea surface temperatures exceed Y°C for more than Z weeks" is a hypothesis.

State your null hypothesis as well — the boring default that there's no effect.

---

Step 3: Methods

This is the most important section. Bad methods sink everything else. Spend the most time here.

Write a detailed methods plan covering:

Study area. Where will the work happen? Why this location? Coordinates if applicable.

Sampling design. How many sites? How many samples per site? How many replicates? What time scale (one season, multiple years, daily, etc.)? Justify every number.

Variables.

• Independent variable(s) — What you're varying or measuring as a predictor
• Dependent variable(s) — What you're measuring as an outcome
• Controlled variables — What you're holding constant
• Potential confounds — What could mess up your interpretation, and how you'll address it

Specific protocols. For each measurement, exactly how will it be done? Cite the protocols you're following or describe them in detail. Examples:

• For eDNA sampling — water volume, filter type, preservation method, DNA extraction kit, primer sets, sequencing platform
• For temperature monitoring — sensor type, deployment depth, sampling frequency
• For fish surveys — transect length, observer effort, identification standards

Statistical analysis. What tests will you use? Why? What sample size do you need to detect a meaningful effect? (This is called a power analysis and is often required by ethics boards and grant reviewers.)

Ethics and permits. Will you need IRB or IACUC approval? Permits from agencies (NOAA, state fish and wildlife, country-level fisheries departments)? Indigenous community consultation? Make a list.

Timeline. How long will this take? Sketch a calendar. Realistic research projects almost always take longer than initially estimated.

Resources. Equipment, lab access, boat time, software. What do you actually need? What does it cost? How would you get it?

---

Step 4: Expected Results and Interpretation Plan

Before you collect any data, write what you think will happen.

This sounds backward. It isn't. Pre-registration of expected outcomes is now standard in good science (F5) because it prevents you from rationalizing whatever you happen to find as "what you expected all along."

Write:

Expected primary result — Based on your hypothesis and background, what do you expect to see? Be specific. A graph sketch is great here.

Alternative scenarios — What would it mean if you saw the opposite? What if the result was weaker than expected? What if it was stronger? Walk through 2–3 scenarios.

What would falsify your hypothesis? This is a key test of good hypothesis design. If you can't articulate a result that would prove your hypothesis wrong, your hypothesis isn't really testable.

Broader implications — If you found what you expect, what would it mean for the field? For conservation? For policy?

---

Step 5: Potential Problems

Every project has weak points. Acknowledging them is a hallmark of good science. List at least three:

Technical risks. What could go wrong methodologically? eDNA contamination? Equipment failure? Insufficient sample size?

Logistical risks. What could go wrong with the project itself? Inability to access sites? Permit denial? Funding shortfall?

Interpretive limits. What's outside the scope of what your data can tell you? Confounds you can't fully control? Generalizability limits?

For each problem, write one sentence on how you'd mitigate or address it.

---

Step 6: Putting It Together

Your final deliverable is a research proposal — roughly 1,500–2,500 words — structured like a real grant application or thesis prospectus:

1. Title (~10 words, specific and informative)
2. Abstract (~200 words summarizing the entire proposal)
3. Background and Significance (~400 words)
4. Hypothesis and Specific Aims (~150 words)
5. Methods (~500–1,000 words)
6. Expected Results and Interpretation (~300 words)
7. Potential Problems and Mitigation (~200 words)
8. References (at least 8 cited sources)

This is roughly the format of a real undergraduate thesis proposal or a small grant application. If you do this work well — even on a topic you'll never actually study — you've demonstrated graduate-level thinking about research design.

---

What to Do With Your Capstone

Once you've written your proposal, you have several options for what comes next:

Treat it as practice. The exercise itself is valuable, even if you never go further. You now understand how research projects are conceived.

Find a mentor. Email a researcher whose work is related to your proposal. Ask if they'd give you 15 minutes of feedback. Most researchers — especially academics — are flattered by serious student inquiries and will respond if your email is professional and specific.

Apply to research programs. Many high school and early college research opportunities accept proposals like this as application material — MIT BWSI, Simons Summer Research Program, RSI, regional university research programs, Sea Education Association courses.

Pursue a science fair or competition. Regeneron ISEF, NHD STEM categories, and many state science fairs accept research projects in marine science.

Actually do it. If your project is genuinely tractable with the resources you have — and many small-scale eDNA, citizen science, and observational projects are — you can run it. Many of the best high school marine science projects involve nothing more than a smartphone, a sampling kit, and consistent effort over a single summer.

Publish or present. Smaller research projects can be presented at high school science conferences, regional symposia, or even submitted to journals that accept undergraduate research (such as the American Journal of Undergraduate Research).

---

You Finished the Marine Biology Track

This is the end of the track. By completing all 12 modules, you've covered roughly the same conceptual ground as a strong introductory marine biology course at a research university — with the added benefit of being grounded in the molecular biology, ecology, chemistry, and statistical foundations that most introductory courses can't assume.

You also know more about marine policy and biotechnology than most ocean conservation organizations expect from their entry-level hires. That's not flattery. That's the practical truth of what most institutional knowledge looks like at the early career level.

What you do with this knowledge is up to you. The ocean is in trouble. It needs people who understand it. You now do.

---