How our AI scoring works - and how to actually use the feedback to get better
Let’s be honest for a second.
You finish a practice station, the timer runs out, the screen refreshes… and there it is. A score. Sometimes reassuring. Sometimes confusing. Occasionally a bit rude (or so it feels).
So what does that score really mean? And more importantly - how do you turn it into better OSCE performance instead of just a number you glance at and forget?
This post breaks it down. No jargon. No fluff. Just a clear look at how our AI scoring system works and how you should be reading your feedback if your goal is real improvement, not false confidence.
First things first: what your practice score is not
Before we get into the mechanics, let’s clear up a common misunderstanding.
Your practice score is not:
- A prediction of your final OSCE mark
- A pass/fail decision
- A judgment on your intelligence or nursing ability
It’s a learning signal. A snapshot. A directional nudge.
Think of it less like an exam result and more like a coach saying, “Here’s what I’m seeing - now let’s fix it.”
The big picture: how AI scoring actually works
At a high level, our AI is trained to evaluate how you communicate, not whether you memorised clinical textbooks.
During a practice station, the system analyses:
- What you say
- How you say it
- When you say it
- What you miss
It’s listening for patterns that experienced OSCE examiners care about - structure, empathy, clarity, safety, and professionalism.
Not accents. Not fancy words. Not “perfect” sentences.
Just effective communication.
What goes into your score (the real components)
Your overall score is usually made up of several smaller building blocks. You don’t always see them as separate numbers - but they’re working quietly in the background.
1. Structure and flow 🧭
OSCEs love structure. The AI checks whether your consultation has a logical shape, such as:
- Introduction and consent
- Agenda setting
- Information gathering
- Explanation and planning
- Closing and safety-netting
If your conversation jumps around - history, then advice, then back to introductions - the score dips slightly. Not because it’s “wrong,” but because examiners value clarity under pressure.
2. Communication clarity 🗣️
This is about being understood, not being impressive.
The system looks at:
- Simple, patient-friendly language
- Logical explanations
- Avoidance of unexplained medical jargon
Short sentences often score better than long, tangled ones. Pauses are okay. Rambling usually isn’t.
3. Empathy and patient-centred language 💙
This part matters more than many candidates realise.
The AI detects phrases and responses that show:
- Acknowledgement of concerns (“That sounds worrying”)
- Validation (“I can see why you’re concerned”)
- Emotional awareness (“That must have been stressful”)
Candidates often feel empathetic but forget to say it out loud. The score reflects what is communicated, not what is intended.
4. Safety and professionalism 🛡️
This includes things like:
- Checking understanding
- Offering opportunities for questions
- Appropriate safety-netting (“If this gets worse…”)
- Maintaining a respectful, calm tone
Missing safety cues doesn’t always tank your score - but repeated omissions add up.
5. Completeness (what you didn’t say) 📋
This one surprises people.
Sometimes your feedback highlights gaps, not mistakes:
- You explained the condition well but didn’t check understanding
- You gathered history but didn’t summarise
- You gave advice but didn’t confirm consent
These aren’t failures. They’re coaching notes.
Why two similar attempts can get different scores
Yes, this happens. And no - it’s not random.
Small changes can make a big difference:
- Adding one empathetic statement
- Improving your opening
- Ending with a clear safety-net
OSCE marking (human or AI) is cumulative. A few well-placed improvements can noticeably lift your score even if the scenario feels “the same.”
How to read your feedback (without overthinking it)
After each practice, you’ll usually see:
- Strengths
- Areas to improve
- Specific suggestions
Here’s the trick most students miss 👇 Don’t try to fix everything at once.
Instead:
- Pick one communication habit to improve next time
- Practice that deliberately
- Reattempt and check if that area improves
That’s how progress actually sticks.
Turning feedback into real OSCE improvement
Here’s a simple, repeatable approach that works well:
Attempt → Review → Adjust → Repeat
- Attempt the station naturally
- Review feedback calmly (not emotionally - easier said than done)
- Adjust one or two things only
- Repeat with intention
Over time, patterns emerge. You’ll notice the same comments disappearing. New strengths showing up. Confidence creeping in quietly.
That’s the goal.
A quick word on “low scores”
Low early scores are… normal. Very normal.
Most strong OSCE performers didn’t start strong - they became consistent.
Practice scores improve fastest when you:
- Focus on communication habits, not memorised scripts
- Speak like a human, not a checklist
- Reflect instead of rushing into the next attempt
Progress isn’t always linear. Some days feel off. That’s part of it.
Final thought
Your practice score is a mirror, not a verdict.
Use it to:
- Spot blind spots
- Build safer habits
- Sharpen communication under pressure
Do that consistently, and the numbers tend to follow - almost as a side effect.
If you want, next we can walk through a sample feedback report and break it down line by line, or look at how to boost scores specifically for communication-heavy OSCE stations.
Just say the word.