publishing research

you can publish real research as a teenager. not "student research" in scare quotes — actual papers in actual journals that actual researchers read. the barrier is lower than you think, and the process teaches you more than any class.

how to find a research mentor

the hardest part isn't doing the research — it's finding someone willing to work with you. here are the paths that actually work:

cold email a PI

this is the most direct route. find a professor or researcher whose work interests you, read their recent papers, and send them a specific email about their work. not "I'm a high school student interested in your field" — that's what everyone sends. instead: "I read your 2025 paper on X, tried to reproduce your results, and had a question about Y."

the response rate is low (~5-10%), so send volume. 30-50 thoughtful, customized emails is reasonable. see mentorship-networking for the full cold email playbook.

through a program or internship

research programs like RSI, SSP, Garcia, and Simons pair you with PIs. but you don't need a formal program. I found a researcher at a hospital research lab through my neurotech internship — visited them, discussed category theory, then we collaborated on a CNN binary classifier for anesthetic depth from EEG signals (<400 parameters). the internship opened the door, but the collaboration happened because I showed genuine interest and technical capability.

see summer-programs for research-oriented programs.

through existing connections

teachers, parents' colleagues, community members — anyone in academia can introduce you. one warm intro is worth 20 cold emails.

the process

research isn't magic. it's a learnable process:

read papers. pick a field. read 10-20 papers. use Google Scholar, arXiv, PubMed. you'll be confused at first — that's normal. by paper 10 you'll start seeing the landscape.
find a gap. what hasn't been done? what could be done better? what could be applied to a new domain?
reach out to someone working in that area. propose a specific collaboration. "I noticed no one has tried approach X on dataset Y — I'd like to try it and I think your lab's expertise in Z would be valuable."
do the work. build the model, run the experiments, collect the data. this is the part most people skip — they want the publication without the months of grinding.
write it up. follow the structure of papers in your target journal. abstract, introduction, methods, results, discussion.
submit. get rejected. revise. resubmit. this is normal.

my research

EEG anesthetic depth classifier

collaborated with a researcher at a hospital research lab on a CNN binary classifier for anesthetic depth from EEG signals
the model had fewer than 400 parameters — intentionally small to be interpretable
submitted to Davidson Fellows (submitted Feb 2026) and targeting IEEE TBME
this came from genuine curiosity about consciousness and brain signals, not from wanting a publication

phase change materials (PCMs)

paper on PCMs for AI data center cooling
applied materials science to a real engineering problem

AI bias

offense detection across datasets
examining how bias manifests differently depending on training data

where to publish

journals that accept high school research

International Journal of High School Research (IJHSR) — dedicated to high school student research, multiple issues per year
Journal of Student Research (JSR) — multidisciplinary, faculty-reviewed, accepts high school through grad students
The Concord Review — specifically for history research essays (5,000-9,000 words)
PRESS Journals — high-quality research and review articles across scientific disciplines
Journal of Emerging Investigators (JEI) — peer-reviewed, specifically for middle and high school students
Curieux Academic Journal — student-run, publishes across disciplines

real journals (not student-specific)

arXiv — preprint server, no peer review but gets your work out fast and cited. CS, physics, math, bio.
IEEE student papers — IEEE conferences often have student paper tracks
field-specific journals — many journals don't care about your age, only your work. if the research is good enough, submit to the real venues.

conferences and fairs

JSHS (Junior Science and Humanities Symposium) — present original STEM research, regional → national pipeline, fully funded
Golden Gate STEM Fair / regional science fairs — the ISEF pathway starts here
ISEF — the pinnacle of high school science fairs. $9M+ in prizes.
NeurIPS High School Projects Track — yes, NeurIPS has accepted high school submissions. 330+ submissions in their inaugural track.
Davidson Fellows — 10k,25k, or $50k for significant research. not a journal, but a serious award that validates your work.

see competitions-hackathons for Davidson Fellows and science fair details.

the Davidson Fellows path

Davidson Fellows is worth special mention. it's a 10,000-50,000 scholarship for students 18 or under who have completed a "significant piece of work." the bar is high — they want genuine contribution to a field, not a school project with a fancy title.

the application is essentially: describe your work, its significance, and your process. research you publish can become a Davidson Fellows submission. I submitted my EEG paper (submitted Feb 2026).

practical tips

start reading papers now. even if you don't understand everything. you'll learn the vocabulary, the structure, and what "good research" looks like in your field.
use Semantic Scholar and Connected Papers to find related work and understand the citation graph.
learn LaTeX. every serious paper is written in LaTeX. overleaf.com makes this easy.
keep a research notebook. document everything — your hypotheses, experiments, failures, insights. you'll thank yourself when writing the paper.
research is slow. my EEG paper took months. don't expect to go from "I want to do research" to "published paper" in a few weeks.

the products you ship and the research you publish are the two strongest things you can put in a cold email or a grant application.