What's Actually on a Judge's Scoring Sheet? A Line-by-Line Look at a Real Science-Fair Rubric

Most students assume judges score on impression. In reality, a judge usually has a rubric in hand, and every row spells out exactly what counts as "excellent" and what counts as "needs improvement."

Learn to read that sheet and you're essentially holding the answer key. Below I use the publicly available 2024–2025 preliminary-review rubric from the Junior Science and Humanities Symposium (JSHS) — the official document they hand to participants — and walk through it row by row, with a few examples so you can feel the difference between a strong and a weak version of the same thing.

First: how does the scoring work?

A judge scores every row from 0 to 5, in three bands:

✅ Excellent 4–5 🟡 Acceptable 2–3 ⚠️ Needs work 0–1

Add the rows together and you have the project's total — the more rows there are, the more it tests whether you can avoid dropping any of them. Here's our simplified overview so you can see the shape of it:

Row	✅ Excellent (4–5)	⚠️ Needs work (0–1)
Abstract	Under 250 words; purpose, hypothesis, methods, results, conclusion — all clear and logically ordered	Too long or short; the five elements are vague or jumbled
Introduction	States the research problem, the purpose, and why it matters to the field	Problem or purpose never clearly stated
Methods	Materials, hypothesis, study design, statistics and steps all spelled out	Missing materials, no hypothesis, no statistics
Results	Data trends analyzed; tables and figures stand on their own	Trends not addressed; figures poor or missing
Discussion / Conclusion	Ties back to the question, discusses significance, and names the study's limits	Conclusion untethered from the data; limits ignored
Citations	Correct APA; sources significant, published and relevant	No APA; sources irrelevant or unpublished
Integrity (outside assistance / acknowledgements)	States honestly who helped and how	Outside help not disclosed

▲ Our simplified English summary, made to show the structure at a glance. The complete original (including the "Acceptable 2–3" descriptions) is published by the Junior Science and Humanities Symposium (JSHS, jshs.org).

1. The abstract: five things inside 250 words

The rubric is blunt about it: a good abstract stays under 250 words and clearly covers purpose, hypothesis, methods, results, and conclusion — in an order that makes sense.

Two abstracts on roughly the same topic are below. If you were the judge (0–5), what would you give each? Think first, then open it up.

📄 Abstract A: "This study explores the effect of air pollution on human health. The results found the two are related, and further research is warranted."

📄 Abstract B: "To test whether commuter traffic affects street-level PM2.5, we measured a single road segment during morning and evening peaks for 14 days (methods), hypothesizing peak concentrations would be significantly higher (hypothesis). Peaks averaged 38% higher (results), with wind speed the strongest predictor (conclusion)."

Open to see how a judge scores these

Abstract A · about 1 point. Only half the elements are there — no hypothesis, no methods, and a conclusion that says nothing. The judge finishes reading and still doesn't know what you did.

Abstract B · about 5 points. All five elements, and specific enough to be checked.

The judge's number is largely set within the 30 seconds it takes to read your abstract.

2. Methods and results: does it hold up?

Under "Methods," the rubric looks for all materials, the hypothesis or research questions, and the study design — plus clear statistics if you used any. Under "Results," it looks for whether data trends are analyzed and whether the figures can stand on their own.

Keep one thing in mind: the judge is scoring whether your method can support your conclusion. How impressive the conclusion sounds is secondary.

What would you give each of these?

📄 Version A: "We ran the experiment and found A was better than B." (No control group, no repetitions, no figures; trends glossed over.)

📄 Version B: Variables and control group spelled out, repetitions and error stated; results shown in one figure that stands alone, noting "the difference only becomes significant after week 3."

Open to see how a judge scores these

Version A · about 1 point. With no control, no repetition and no figures, the judge has no way to tell whether the conclusion holds.

Version B · about 5 points. Method and results can both be checked; even an ordinary conclusion scores well.

3. Discussion: the students who name their limits score higher

This row separates projects more than any other. Tucked into the definition of "excellent" is a line many people miss: you must be able to point out the limits of the research, and speak to its practical or theoretical implications.

Many students fear that admitting a weakness costs points, so they overstate the conclusion. To a judge, a student who writes an honest "limitations" section reads as far more mature than one who states everything as settled fact.

4. Citations and formatting: the "boring" rows are scored too

The rubric gives APA style and the bibliography their own row: is the citation format right, and are the references significant, published and relevant? Grammar and spelling get a row as well — "professional tone, writing that flows" earns full marks, while "casual tone, hard to follow" sits at the bottom.

These look like small things, but they carry a message: how seriously do you take your own research?

5. The row that matters most in the age of AI: integrity

Two rows of the JSHS rubric are devoted to integrity: the Statement of Outside Assistance (SOA) and the Acknowledgements. The standard asks you to state honestly who helped you and how, signed by both the student and a teacher.

Why does this matter so much now? Because everyone writes with AI. Judges aren't afraid of you using tools — they're wary of a student who can't say clearly which parts were their own and which came from someone (or something) else.

A student who states plainly "here's the tool I used, here's the advice my teacher gave" comes across as more credible, not less. Integrity is a place to gain points.

So the rubric is really a checklist for "what professional looks like"

Read the rows together and the sheet stops being mysterious. It keeps asking one question: did you study something carefully and honestly, and can you explain it?

Abstract: are all five elements clear?
Methods/results: does it hold up? Can the figures stand alone?
Discussion: are you willing to name the limits?
Format/citations: is it professional?
Integrity: is it clear who helped with what?

When we mentor students, this is exactly the work — sitting with them and making every row on that sheet genuinely solid.

Want your child's next paper to hold up against that sheet? We offer a free consultation to help plan a research project that's truly their own.

👉 Learn more: worldscienceacademy.org

Note: This article uses the publicly published 2024–2025 preliminary-review rubric from the Junior Science and Humanities Symposium (JSHS) as a teaching example. All strong/weak examples are the author's own inventions and involve no real student work.