Free science fair judging sheet: a six-row rubric out of 24
A free science fair judging sheet: six rows scored 1 to 4 for 24 points, what a 4 and a 1 look like on each, how many judges a project needs, and a no-names tally.
A good science fair judging sheet has six rows, question, method, data, conclusion, board and interview, each scored 1 to 4 for a total of 24, with a line saying what a 4 and a 1 look like so two judges score the same project the same way. Add a box for one kind line to the child, and put the entry's code at the top instead of the child's name. That's the whole sheet.
I built that sheet into Peabody's free tools for whoever is running the fair. The fair organizer prints it on paper. The fair-night kit puts the same six rows on each judge's phone and adds them up. The judges' printables make the sign-up sheet, the schedule and a table tent for every board. All of it runs in your browser, and no child's name or score is ever sent.
The six rows
| Row | A 4 looks like | A 1 looks like |
|---|---|---|
| Question | Testable: one thing changed, one thing measured | A topic or a demonstration, nothing tested |
| Method | Controls named and kept; three or more tries a level | Changed several things at once, or one try |
| Data | A clear table with units, means where expected, a graph | No numbers, or numbers with no table |
| Conclusion | Uses the numbers; honest when the hypothesis was wrong | Restates the hypothesis, no numbers |
| Board | In order, readable from a step back, the child's own words | Hard to follow, or a grown-up's words |
| Interview | Explains the test and the results in their own words | Can't say what was changed or measured |
A 3 and a 2 sit in between. Judges score faster and agree more often when only the ends are written out.
"Means where expected" matters for mixed-grade fairs. Peabody's sheets add a mean column from 4th grade, so a 2nd grader with three tries per level and no mean still scores a 4 on data.
What the sheet doesn't score
- How it looks. Glitter and 3D letters earn nothing. The board row asks whether it's in order and readable, and whose words they are.
- Whether the hypothesis was right. A wrong guess with honest numbers can score 24. The conclusion row rewards saying it plainly.
- How much it cost. Most of Peabody's projects cost nothing, and the rubric doesn't know the difference.
How many judges, and for how long
Society for Science's Standard for Judging, one of its standards for science fairs, says "a science fair should have a minimum of two judges reviewing any project per round of judging; more, if possible" (read 31…
Society for Science's Standard for Judging, one of its standards for science fairs, says "a science fair should have a minimum of two judges reviewing any project per round of judging; more, if possible" (read 31 October 2026). At one district's STEM fair whose judging procedures are public (Alpine School District, Utah, read 31 October 2026), every project is judged twice in a 45-minute period, and a judge with ten projects has four to five minutes for each.
So for a school fair, two judges a project and five minutes a board is a good plan. That's about ten to twelve boards an hour for a pair. The judges' printables plan it for you: pick how many judges, how many to a team, the minutes a board (4, 5, 6, 8 or 10) and a start time, and it prints who sees which codes and when, with anyone left over as a standby for a no-show.
Codes, not names
Each project gets an entry code: a letter for each grade band and a number, A1, A2, B1. The teacher writes each child's name beside their code on a key that stays on paper.
Each project gets an entry code: a letter for each grade band and a number, A1, A2, B1. The teacher writes each child's name beside their code on a key that stays on paper. Judges only ever see the code, on a table tent by the board and on their sheet.
Codes do two jobs. Judges score the work, not a name they recognise from the school run. And nothing with a child's name in it goes into a phone or a file.
Fair night, on phones
If you'd rather not add up paper sheets at nine at night:
If you'd rather not add up paper sheets at nine at night:
- Set the codes and print the key. One letter a band, as many numbers as entries.
- Each judge scans their own QR code with their phone's camera. It opens a judge card set up for this fair: tap a code, tap 1 to 4 on each of the six rows, and pick one line for the child if they like.
- Gather the scores. Each judge sends their file (AirDrop, a message or email: their choice), or reads you their lines.
- The tally averages each entry's judges, so a board seen by two judges isn't behind one seen by three, and ranks within each band. Ties are shown and never broken: that's the judges' call.
- Print the results by band and the certificates with the code and the place. You write the names.
The lines a judge can pick for a child are short and kind: "A great question, and you tested it", "Your table and graph were clear", "Next time: put the units on your numbers", and five more.
The judge's question card
Most of the interview row comes from three or four questions. " The judging questions post gives a strong and a weak answer for each, which is useful for a judge who's never done it.
Most of the interview row comes from three or four questions. The organizer prints a card of nine, in a child's words, starting with "What made you want to ask this question?" and "What did you change, and what did you measure?" The judging questions post gives a strong and a weak answer for each, which is useful for a judge who's never done it.
Try tonight
Open the organizer, put in your fair date, and print the rubric and the judge's card. If you have more than about twenty projects, open the fair-night kit too and try scoring three made-up codes on your own phone, so…
Open the organizer, put in your fair date, and print the rubric and the judge's card. If you have more than about twenty projects, open the fair-night kit too and try scoring three made-up codes on your own phone, so you know how it works before the judges arrive.
Questions grown-ups ask
No, not at an elementary fair. The sign-up sheet says so: judges ask questions and listen.
Do judges need a science background?
No, not at an elementary fair. The sign-up sheet says so: judges ask questions and listen. The rubric's ends do the rest.
Should kindergarten be judged on the same sheet?
Use the same six rows, kindly. A kindergartner's data is counted in steps or blocks and their interview is a sentence or two, and a 4 means a 4 for their age. Some schools give every K–2 child a certificate rather than a place.
Can we change the rows?
Yes. The rubric post has versions for K–2, 3–5 and 6–8. Keep it to six rows or fewer so judges can score in five minutes.
Peabody (peabody.kids) is free through the end of 2026, with no ads and no sign-up. The library has free stories by reading level for the classroom shelf.
Checked against how Peabody works on 31 October 2026. Written with help from AI. Peabody is a reading app, not a reading assessment: if you're worried about your child's reading, talk to their teacher. Questions, or something wrong? Write to hello@peabody.kids.