This is the public pilot rubric for NZRC. It tells students what matters, gives judges a shared language, and keeps recognition tied to the quality of the paper rather than the confidence of the writer.
ONE PAPER · FOUR LENSES
Each criterion is scored from 0 to 10. Judges score independently, then reconcile significant differences using evidence from the paper. The total is out of 40.
This is a pilot rubric. The judging panel may refine wording after the first cycle, but any change will be documented before it is used for a new competition.
DESCRIPTORS, NOT VIBES
A high score does not mean a paper sounds like a university paper. It means the writer makes their question, reasoning, and evidence visible to a careful reader.
Does the paper create a clear route from question to conclusion?
Purpose, sections, progression, and conclusion are deliberate and easy to navigate.
Overall route is clear, with minor gaps, repetition, or uneven sequencing.
Some useful parts are present, but the reader has to reconstruct the route.
Purpose or organisation is missing enough that the paper cannot be followed reliably.
Do the claims follow from the reasoning rather than from confidence?
Claims are precise, qualified where needed, and connected through explicit reasoning.
Main reasoning works, though some steps are assumed, broad, or underdeveloped.
Ideas are present but claims and reasoning are inconsistently connected.
The paper mainly asserts, describes, or repeats information without a defensible line of reasoning.
Are sources relevant, credible, used accurately, and acknowledged?
Evidence is purposeful, credible, accurately interpreted, and cited consistently.
Most evidence supports the paper, with a few weak, unexplained, or inconsistently cited sources.
Some sources appear, but relevance, quality, accuracy, or citation is uneven.
Evidence is absent, unreliable, misrepresented, or not attributable.
Can a thoughtful reader understand what the writer means?
Language is controlled, specific, and readable; technical terms are explained when needed.
Meaning is generally clear, with occasional awkwardness, ambiguity, or imprecision.
Meaning is sometimes recoverable, but sentence-level or terminology issues interrupt the reader.
Frequent ambiguity or language problems prevent a reliable understanding of the paper.
Where possible, entries should be anonymised. Do not score school reputation, prior awards, polish of the topic, or assumptions about the student.
Do not reward access to equipment, paid databases, or specialist vocabulary. Reward visible reasoning and honest use of available evidence.
Every score should have one short evidence-based comment: what worked, what limited the paper, and what the next revision should try.
Judges must disclose prior supervision, family, school, or other relationships with an entrant and recuse themselves when appropriate.