RSI essays are unusual. They are not asking about your identity, your activities, or your life story. They are asking you to show that you think like a scientist and that you can explain your thinking clearly to someone who has read thousands of similar files.
Every applicant has grades and scores that clear the bar. What separates admits is how the essays read: specific, curious, and honest about what is known and what is not.
The two prompts
The exact wording shifts year to year. The substance does not.
- Essay 1. Pick a scientific or mathematical concept that interests you. Explain it, and explain why it draws you. Roughly 500 words.
- Essay 2. Describe your most meaningful research or intellectual experience, what you did, and what you learned from it. Roughly 500 words.
The first is a writing sample. The second is a research sample. Readers use the pair to triangulate whether you can both think and do.
Essay 1: the concept essay
Pick depth over prestige
The topic does not need to sound impressive. It needs to be something you actually understand past the first layer. A rigorous treatment of a small, concrete idea beats a vague tour of a famous one. Readers recognize both immediately.
Show the path, not the summary
A textbook explanation is easy to find. What is not easy to find is the moment the idea stopped making sense to you, the question you had to chase down, the approximation that felt like cheating until you understood why it worked. Put that on the page.
Connect outward
After the explanation, show where the idea leads. What remains open? What does it rule out? What would a skeptical reader push back on? This is what turns a definition into a piece of thinking.
Write for a smart non-specialist
Clarity is the scarce resource. Cut every sentence that uses a technical term to impress. If a careful reader with no background in your area cannot follow the argument, the essay is not ready.
Essay 2: the research essay
Say what you did
"We analyzed data" is useless. "I wrote the script that pulled the sequences, and noticed a pattern my advisor had not flagged" tells the reader something. Make the verbs yours.
Include the failure
Research is mostly the thing not working. Describe a dead end, a hypothesis that fell apart, or an approach you had to redo. This is the most reliable signal that the work was real.
Reflect, briefly
What do you understand now that you did not at the start? What would you do differently? A paragraph is enough. Reflection that runs long reads as filler.
Do not beg
Do not explain why RSI is the obvious next step. The readers know what RSI is. Let the essay make the case implicitly by showing that you are ready for the work.
What a weak essay looks like
- The Wikipedia essay. Factually correct, tonally generic, indistinguishable from any other strong student.
- The prestige topic. Quantum computing, CRISPR, or AI, handled at a layer any motivated high school reader could write.
- The resume essay. A list of project titles and awards with no actual thinking between them.
- The sanitized research essay. Every step worked, every experiment confirmed the hypothesis, nothing surprised anyone. Readers do not believe it.
- The rehearsed essay. Heavy polish, no voice, no sign of the person behind it.
A working process
Draft early and badly.
Your first attempt will not be good. Accept that and get the ideas on the page. Everything useful happens in revision.
Read it to a scientist.
Someone in the field should push on the technical accuracy. If they correct you, you were not deep enough to write it.
Read it to a non-scientist.
A smart reader with no background should still follow the argument. If they cannot, the clarity needs work.
Read it out loud.
Sentences that sound wrong are wrong. This is also the fastest way to catch the rehearsed voice that essays drift into.
Cut 15 percent.
Almost every draft is too long. The version that survives a hard cut reads sharper than the one you thought was done.
The short version
Write the essay only you could have written. Pick a concept you have actually wrestled with, and a research experience you can describe in specifics. Do the work on the page that you would do in a lab. The readers are looking for evidence that you are ready to do it for six weeks in Cambridge.
Helpful resources
- CEE RSI - Official Research Science Institute page
- NSF REU - National Science Foundation research programs
- NIH Research Training - NIH research training opportunities