For Students, Ages 13 and Up
Student Researcher Section
A real peer-reviewed section of the journal for school and undergraduate researchers. Your name, your paper, your citation. Reviewed by working scientists who will tell you what they actually think.
Why This Section Exists
Most academic journals will not read a paper written by a sixteen-year-old. The few that will are almost all in North America, most charge a submission fee, and nearly none of them care whether your research is about the place you live. So a generation of Caribbean students learns machine learning, builds something genuinely interesting with it, shows it at a science fair, and then watches the work disappear into a folder.
This section is the fix. It is not a competition, a showcase or a school project gallery. It is a peer-reviewed section of a real journal, with an editor, reviewers, revision rounds, a permanent record and a citation you can put on a university application, a scholarship form or a CV for the rest of your life.
The Caribbean Artificial Intelligence Association funds the journal, which is why this section costs nothing. Comparable student journals abroad charge around forty-five US dollars to submit. We charge nothing, because a submission fee decides who gets published on the basis of who can pay, and that is precisely the thing this section exists to undo.
Who Can Submit
You are eligible if, on the day you submit, you are:
- 13 or older, and
- enrolled in secondary school, high school, sixth form or an equivalent programme, or studying for a first undergraduate degree, and
- working with a mentor who will countersign the submission. A mentor is any adult who has supervised the work: a teacher, a lecturer, a club leader, a lab supervisor, a parent with relevant expertise, or a researcher who agreed to advise you.
You do not need to live in the Caribbean, and you do not need to be Caribbean. Your research does. Every student paper must apply machine learning or artificial intelligence to a question about the region, its people, its economy, its environment or its institutions, or to a small-island condition the Caribbean shares with places like the Pacific and the Indian Ocean states.
Teams are welcome, up to six student authors. Everyone listed must have done real work on the project and must be able to explain it. Your mentor may be acknowledged, and may be a co-author only if they meet ordinary authorship criteria.
What Makes a Good Student Paper
Reviewers are not looking for a breakthrough, and they are not marking you against a professional researcher. They are looking for four things.
- A real question. Something specific enough to answer. "Can a model predict which bus routes in Kingston run late, using published timetable and traffic data?" is a research question. "AI in transport" is a topic.
- Real data, honestly described. Where it came from, how much of it there is, what is wrong with it, and what you had to throw away. A small dataset you collected yourself and understand completely beats a large one you downloaded and never inspected.
- A method a reader could repeat. Which model, which settings, how you split training from testing, what you compared against. A baseline as dull as "always predict the average" is worth more than none.
- Limitations you found before we did. Saying what your result does not prove is not a weakness in a paper. It is the single clearest signal that you understand your own work.
Negative results are publishable here. If you built the model, ran it properly and it did not work, that is a finding, and writing it up honestly is worth more than quietly changing the question.
What You Can Submit
| Format | What it is | Length |
|---|---|---|
| Student Research Article | A full study with a question, data, method, results and limitations | 1,500 to 4,000 words |
| Data Story | A dataset you collected or assembled about the region, described so others can use it | 1,000 to 2,000 words |
| Replication Note | You reran someone else's published result on Caribbean data and reported what happened | 1,000 to 2,500 words |
| Build Note | A working model, tool or app, with the design decisions, the evaluation and the failure modes | 1,200 to 3,000 words |
Every format needs a plain-language summary of 150 words that a classmate outside your subject could follow. Write it last. It is the part reviewers read first.
How Review Works
The process mirrors the main journal, with the difference that feedback is the point rather than a by-product. Every submission receives written comments, including the ones we cannot publish.
Mentor countersigns and submits
Your mentor confirms the work is yours, that consent is in place if you are under 18, and that the data was gathered safely and lawfully.
Editorial check, within 10 days
The student track editor checks scope, eligibility, safety and completeness. If something is missing you are asked for it rather than rejected for it.
Two reviewers, plus a student reader
Two working researchers review the science. A postgraduate or undergraduate reader checks that the writing is followable. Reviewers see your work without your name on it.
Written feedback, whatever the decision
You receive the reviewers' comments in full, written to be useful to someone learning. Nobody is told only that their paper was rejected.
Revise, up to two rounds
Most accepted student papers are revised twice. That is normal, it happens to professional researchers, and it is where the real learning sits.
Publication and citation
Accepted papers appear in the student section of the edition, open access under CC BY 4.0, with a permanent identifier and a citation in your name.
Safety, Ethics and Data
These rules are not paperwork. They are the difference between research and a problem, and reviewers check every one of them.
- Other people's data needs consent. If you surveyed classmates, ran an experiment on people, or used photographs or recordings of anyone, say how you asked permission. Never publish anything that identifies a person who did not agree to it.
- No scraping private or restricted data. Public, licensed and open data only. If you are unsure whether a source counts, ask your mentor, and if the answer is still unclear, ask us before you submit.
- No medical, legal or financial advice. You may publish a model that predicts a health outcome. You may not tell readers what to do about their health.
- Give credit. Every dataset, tutorial, repository, model and person that helped gets named. Copying text from anywhere without attribution ends the submission.
- Be careful with your own privacy. Do not put your home address, phone number or your classmates' full names into a paper that will be public forever.
Using AI Tools on Your Own Paper
You are studying AI, so you will use AI. That is expected and it is allowed. The rule is simple: tell us exactly what you used it for.
Fine, when disclosed: writing and debugging code, checking grammar, translating, explaining a concept you did not understand, generating figures from your own results, summarizing papers you then read yourself.
Not fine, ever: having a model write the paper and submitting it as your own, generating results you did not run, inventing citations, or describing an experiment you did not do. Fabricated references are the easiest thing in the world for a reviewer to catch, and they end a submission immediately.
Every submission includes an AI use statement. Nobody has ever been rejected here for honestly describing how they used a tool. People are rejected for hiding it.
The CAIA Young Researcher Award
In each edition, the student track editor and the review panel name one paper the outstanding student contribution. The award carries a certificate from the Caribbean Artificial Intelligence Association, a feature on the journal homepage, and an invitation to present the work to the Association. There is nothing to enter. Every published student paper is considered automatically.
For Teachers and Mentors
Your role is to confirm three things: the student did the work, the data was obtained safely and lawfully, and a parent or guardian has consented where the student is under 18. You are not certifying that the science is correct. That is what the reviewers are for.
You do not need a doctorate, a research background or an institutional affiliation. A secondary school teacher supervising a CAPE or A-level project is exactly the mentor this section was designed around. If you are supervising several students, submit each paper separately, and write to students@caribbeanaijournal.org if you would like a briefing session for your class or club. We run them free.
Student Submission
For the inaugural edition, submissions close 14 September 2026. This form is completed by the student and countersigned by the mentor, whose email address is required. Read the safety and ethics rules before you begin.