When grading responses in Vervoe, it's important to remember that scores are ultimately determined by the employer. While this guide offers a helpful framework, grading should be tailored to your organisation’s expectations and role-specific requirements.
Vervoe uses a 0–10 scoring scale, categorised into three levels: Poor, Average, and Excellent. As you grade, Vervoe’s machine learning will suggest a category based on previous data. However, manual grading always overrides AI predictions and helps train the model for future assessments.
Defining the Score Categories
Use this guide as a starting point, but align your scoring with what excellence looks like in your team.
Poor (0–3)
- The candidate misunderstood or misinterpreted the question.
- The answer lacks clarity, substance, or relevance.
- Doesn’t meet minimum role expectations.
Average (4–7)
- The response is adequate but lacks depth or insight.
- Meets the basic requirements without exceeding them.
- Shows competence, but not mastery.
Excellent (8–10)
- Provides a complete, thoughtful, and well-supported response.
- Demonstrates a strong grasp of the subject matter.
- Goes beyond expectations—insightful, strategic, or highly practical.
Tip: Your sample "ideal" response should closely reflect what you'd rate a 10/10. This helps guide both manual grading and train the AI.
Why Use a 0–10 Grading Scale?
-
More Precise Evaluation
Easily distinguish between similar responses within the same category. -
Flexibility Across Roles
Adaptable for different seniority levels or skill requirements. -
Customisable to Your Standards
Define what "excellent" means in your context—technical detail, creativity, clarity, etc. -
Clearer Differentiation
Highlight subtle differences between "acceptable" and "outstanding." -
Consistency Across Evaluators
A standardised scale promotes fairer, more reliable assessments. -
Better Candidate Feedback
Provide more meaningful insights into performance and areas for improvement. -
Reflects Real-World Performance
Recognises the performance spectrum, not just binary outcomes.
For more on Vervoe grading and machine learning, check out our grading guide article or reach out to our support team.