Publication
Pre-submission checklist
This page is generated from the checks themselves rather than written next to them. If it is on the first list, the software checks it, and you will see the same wording inside the editor.
- Title at least 8
Sentence case. Say what the work is about rather than what makes it sound impressive.
- Every author has accepted
Adding somebody is an invitation, not an assignment. Nobody is listed until they say yes.
- Abstract 100 to 350 words
The question, what you did, what you found, and what it means. A reader should be able to stop here and know whether to keep going.
- Keywords 3 to 6
Three to six. The method is a keyword; the field is the discipline, which is a separate thing.
- Discipline
Classify by the domain of the question. Machine learning applied to protein folding is biology, and the method is a keyword.
- Contributions statement at least 10 words
What each author did, and what any mentor did. This is the written form of what judges are probing when they ask how much of this was yours, and it is what the existing archive most conspicuously lacks. How you write it is yours.
- References at least 5
At least five. Work with no prior art behind it is either a first in the field or a literature search that did not happen.
- Introduction and background at least 150 words
The problem, why it matters, and your hypothesis or design goal, stated plainly enough that somebody outside your field follows it.
- Prior work at least 100 words
What is already known and the specific gap you are working in. Not a list of everything you read.
- Methods at least 200 words
Enough that somebody else could repeat it. This is the section reviewers most often send back, and the test is specific: could a stranger run your experiment from this text alone.
- Results at least 150 words
What happened, with the data that supports it. What it means belongs below.
- Discussion and limitations at least 150 words
What it means, and what it does not. Every study has limitations and naming them is what makes the rest credible.
- Conclusion at least 75 words
What you now know that you did not before. Short is fine.
- Future work at least 50 words
What you would do next with more time. Judges ask this out loud, so having written it down is rehearsal.
These are cheaper for a machine to catch than for a reviewer to report, and every one of them has been the reason a real submission bounced.
- Every axis is labeled, with units
A plot with an unlabeled y-axis is the most common single defect at any fair.
- Error is represented, and the text says what kind
Standard deviation, standard error, and a confidence interval are three different claims and look identical on a chart.
- Figures are readable from three feet away
Print one and stand back. This takes a minute and it is never wasted.
- Display and safety rules are met
Check against the current rules for the fair you are entering, not against last year.
A project entry records the work as the fair saw it: a title, an abstract, a category, and what happened. There is no manuscript, so most of the first list does not apply.
- Title
- Every author has accepted
- Abstract
- Keywords
- Discipline
- Contributions statement
- Competition record
Work you finished somewhere else
If the paper is already written, none of the first list stops you. The checks still run and you still see what they say, and they become suggestions rather than a wall. What is asked for is a title, the authors, a date, an abstract, keywords, a discipline, a license, and the file.
One last read
Read your abstract on its own and ask whether it states what you found. An abstract that describes the procedure and stops is the single most common thing a reviewer sends back.