On this page
In a viva the examiner asks why you did each step, and a memorised aim can fall apart by the second question. You can rehearse at home by asking an AI chatbot to play the examiner, if you set the rules.
In CBSE's 2026-27 physics practical evaluation scheme, "viva on experiments, activities and project" carries 5 of the 30 practical marks in both Class XI and Class XII. In the Class XI chemistry scheme, "class record and viva" together carry 4 of 30 marks. Check your own board's current scheme.
What do examiners ask in a physics or chemistry viva?
The same physics curriculum, in its guidelines for visually impaired students, says viva questions may cover basic theory or principle, apparatus and chemicals, procedure, precautions and sources of error. That list makes a ladder for your practice:
Principle: what law or reaction is the experiment built on?
Apparatus: what is each piece for, and what is its least count?
Procedure: why this order of steps?
Precautions: what goes wrong if you skip one?
Sources of error: which push your result up, which push it down, and how did you reduce them?
A memorised answer can often carry you through the first two rungs. The last two separate understanding from memory.
Why practise out loud instead of rereading the lab manual?
In a 2006 study by Henry Roediger and Jeffrey Karpicke, students who took recall tests on a passage remembered more of it 2 days and 1 week later than students who reread it, even though rereading made them feel more confident. A 2013 review led by John Dunlosky compared ten study techniques and rated practice testing as one of only two of high utility, alongside spacing practice over time; rereading was rated low.
A mock viva is practice testing with follow-ups.
How do you set up an AI mock viva?
Paste a prompt like this into an AI chatbot, changing the experiment:
"Act as a strict CBSE practical examiner for the metre bridge experiment. Ask me one question at a time and wait for my answer. After each answer, ask one follow-up that tests why, not what. Cover principle, apparatus, procedure, precautions and sources of error. Give no answers until I say stop, then score me with the rubric below and list what to revise."
Then follow three rules:
Answer out loud first, then type. The real viva is spoken.
Say "I don't know" when you don't. A guess the AI accepts teaches you nothing.
Check every correction against your NCERT lab manual or your teacher. AI can be confidently wrong.
To rehearse the hands-on steps too, see how to rehearse practical exam procedures with AI, and use AI to check your lab record before the examiner reads it.
What does a good follow-up chain look like?
Here is a metre bridge chain, with the unknown resistance X in the right gap, the known resistance R in the left, and the balance length l measured from the left end:
What is the principle? The Wheatstone bridge: at the balance point no current flows through the galvanometer.
Then why is X = R(100 − l)/l? At balance the resistances are in the ratio of the two wire lengths, because the wire is uniform.
Why aim for a balance point near the middle? The bridge is most sensitive there, and end errors matter less.
Why interchange R and X for a second reading? Averaging the two reduces the end errors from the copper strips.
Why press the jockey gently and briefly? A steady current heats the wire and changes its resistance; sliding scrapes the wire and makes it non-uniform.
For a chemistry titration of sodium hydroxide against standard oxalic acid, the precaution and error rungs sound like this: why can't you make a standard NaOH solution by weighing (it absorbs moisture and carbon dioxide from air), why rinse the burette with the solution but not the conical flask, and why phenolphthalein suits a weak acid against a strong base.
How should the AI score you?
Give the AI this rubric. It is ours, made for practice, not CBSE's marking scheme. One mark each:
Principle stated correctly in one or two sentences.
Each piece of apparatus named with its purpose.
Procedure explained with reasons, not recited.
Two precautions, each linked to what it prevents.
Two sources of error, with the direction each pushes the result and how you reduced it.
Ask for the score only at the end, with your weakest answer named. Repeat the experiment a few days later rather than five times in one evening.
If a follow-up shows a gap in the idea behind the experiment, Fermi's AI tutor works out what you already know, builds a plan for the session and checks your understanding as it goes. For ideas easier to see than to read, the Fermi Simulation Lab has interactive simulations.
For speaking calmly on the day, read how to prepare for an oral test or viva in school. Tonight, pick one experiment and run a five-question mock viva.
