Skip to content

Marking Rubric

This to align exactly with the System Card and Pitch Deck deliverables. It uses a 1-10 point scale per section for easy calculation, which is then weighted to get the final score. If you are a student please feel free to read through this, it will help you understand where you may need to work on your current idea.

Judge’s Scorecard: AI for Good Schools Challenge

Team Name: __________

School: __________

Age Group: [ ] Junior (Yrs 7-9) [ ] Senior (Yrs 10-13)

Part 1: The Core Criteria

Please rate each section on a scale of 1 to 10.

A. Social Impact (Weight: 30%)

Does this solution solve a meaningful problem?

  • 1-3 (Low): The problem is vague or the solution doesn’t really address it.
  • 4-7 (Med): Good idea, but the impact is small or hard to measure.
  • 8-10 (High): Clear problem with a high-impact solution. Shows deep empathy for the user (School, Community, or Country).

Score: ______ / 10

B. Innovation & Creativity (Weight: 25%)

Is the idea original?

  • 1-3 (Low): A standard idea that already exists (e.g., a basic calendar).
  • 4-7 (Med): An improvement on an existing idea or a clever twist.
  • 8-10 (High): A unique approach. Uses AI in a way we haven't seen before in this context.

Score: ______ / 10

C. Technical Feasibility (Weight: 20%)

Does the "System Card" make sense? Is it grounded in reality?

  • 1-3 (Pure Compute) Screen-Only Solution:
    • The project is a standard website or chatbot running on a laptop.
    • Input is typed manually.
    • Example: A website where you type in your energy bill numbers.
  • 4-7 (Hybrid) Live Data Connection.
    • The solution connects to live real-world data feeds (APIs) or existing systems.
    • Example: An app that pulls live bus times from the Council API or uses the phone’s GPS.
  • 8-10 (Physical AI) ** touches the Real World.**
    • Hardware/Sensors: Uses cameras, microphones, motion sensors, or IoT devices (Micro:bit, Raspberry Pi).
    • Action: The AI triggers a physical change (e.g., turns on a light, sorts a bin, alerts a specific person).
    • Example: A camera that spots a pothole and automatically logs the GPS coordinate.

Note to Judges: Give higher marks to teams that use the sensors they already have (e.g., the Camera or GPS on a phone) over teams that just build a text interface.

Score: ______ / 10

D. Ethics & Safety (Weight: 15%)

Did they consider the risks outlined in their System Card?

  • 1-3 (Low) Ignored the Risks.
    • Uses AI blindly without questioning the data.
    • Inputs personal/private data without warnings.
    • Assumes AI output is always factually correct.
  • 4-7 (Med) Basic Awareness.
    • Data: Acknowledges they don't know exactly what the model was trained on or provides how the data is curated..
    • Input: Includes a privacy warning (e.g., "Don't upload real names").
    • Output: Mentions that users should double-check the AI's answers.
  • 8-10 (High) Full Responsibility Cycle.
    • Data: Discusses potential bias in the training set (e.g., "This model might favor Western accents").
    • Weights: Identifies if they used Open Source or Proprietary models.
    • Input: Designed the system to never store user data and understand the limitations this imposes.
    • Output: Built a safeguard or "human review" step to catch bad AI suggestions.

Score: ______ / 10

E. Communication & Pitch (Weight: 10%)

Quality of the Slide Deck.

  • 1-3 (Low): Hard to read, disorganized, or missing key slides.
  • 4-7 (Med): Clear, readable, and follows the structure.
  • 8-10 (High): Compelling storytelling. Visuals explain the concept perfectly without needing text.

Score: ______ / 10

Part 2: Calculation Table

Section Raw Score (1-10) Multiplier Final Weighted Score
Social Impact ______ x 3.0 ______
Innovation ______ x 2.5 ______
Feasibility ______ x 2.0 ______
Ethics ______ x 1.5 ______
Communication ______ x 1.0 ______
TOTAL ______ / 100

Part 3: Judge’s Feedback (Mandatory)

This will be shared with the students. Please provide one compliment and one area for improvement.

🌟 What impressed me most:

💡 One thing to consider for next time:

Judge's Cheat Sheet (For Reference)

Keep this handy if you aren't an AI expert.

  • Feasibility Check: If a student claims their AI can feel emotions or think like a human, score them low on Feasibility. Current AI is good at Pattern Matching (Predicting, Classifying, Generating), not "Thinking."
  • Junior vs. Senior:
  • Juniors (Yr 7-9): Judge them on Creativity and Empathy. (Is the idea kind? Is it fun?)
  • Seniors (Yr 10-13): Judge them on Feasibility and Detail. (Did they research the data sources? Is the logic sound?)