Judging is a conversation, not a recital. These twelve come up again and again.
- "Explain your project in one minute." Practise this until it is automatic. Question, method, finding, meaning.
- "Why this question?" They are testing whether you own the project. A specific origin story beats "it seemed interesting".
- "What was your control?" Know all of them and why each exists.
- "How many trials, and why that many?" Have a reason beyond "that is what fit".
- "What would you do differently?" Never "nothing". Name one real limitation and its fix.
- "What surprised you?" Shows you were paying attention rather than confirming an expectation.
- "What does this number mean?" Pointing at a figure. Know your axes, your units, and your error bars.
- "Could something else explain this?" They want to see whether you have considered alternatives. Name one and how you ruled it out.
- "What did your mentor do?" Be precise and honest. Vagueness reads badly.
- "What is the next experiment?" A real one, specific enough to run.
- "Who would use this?" Applications, hedged appropriately.
- "What did you not understand at the start?" An honest answer here is disarming and memorable.
The best possible answer to a question you cannot answer is "I do not know — here is how I would find out." Judges have heard every kind of bluff.