How to handle questions where your statistics knowledge is weak?
Honesty beats bluffing in statistics questions, every time.
The IIM Udaipur panel already knows your transcript; admitting "I'm a bit weak here, but let me try" signals self-awareness, which is itself an MBA competency they screen for. At ₹16 LPA average placements and ₹16L fees, IIM Udaipur is not hiring statisticians; they want managers who know their limits.
What Actually Happened in That Interview
The OP said exactly "A bit weak, but I can try" when asked about statistics strength. The panel then probed with Eigenvalues, a topic most non-engineers haven't touched since graduation.
The OP gave a partially correct answer, named the concept, stumbled on the formula, but reasoned aloud. The panel moved on without penalizing.
That outcome is instructive: partial correctness plus visible thinking beats confident nonsense.
The Concepts You Need to Know
IIM Udaipur's interview panel tests MBA-foundational statistics, not actuarial depth. Prepare these and you cover roughly 90% of what gets asked:
| Topic | What to know |
|---|---|
| Descriptive stats | Mean, median, mode, standard deviation, variance |
| Probability | Conditional probability, Bayes theorem |
| Distributions | Normal, binomial, Poisson (shape, use-case) |
| Hypothesis testing | H0/H1, p-value, Type I and II errors |
| Regression | Simple linear, R-squared interpretation |
Two focused hours the evening before your interview is enough to refresh these. Don't go deeper than this table; the panel won't.
How to Handle a Question You Can't Answer
When an advanced question lands, a three-step response works well. First, name what you do know: "I know Eigenvalues relate to variance decomposition in PCA, though I haven't worked through the math recently." Second, offer to reason from first principles: "Would it help if I approached it from what I remember about matrix transformations?" Third, if you're genuinely blank, say so directly and pivot: "I'd need to revisit this, but here's how I'd approach learning it quickly."
Panels at IIM Udaipur and other IIMs in the 85-95 percentile CAT range schools use these moments to test composure under uncertainty, not to catch you on a textbook definition. Faking a formula is the only answer that actually hurts you.
When Your UG Was Quantitatively Weak
If your undergraduate major was non-quantitative (History, Literature, Commerce without statistics), don't volunteer the weakness in your opening narrative unless it fits naturally. But if the panel asks directly, own it without apology: "My undergraduate work was in [X], so my formal statistics exposure is limited to self-study." Then name the specific topics you did cover. This framing shows initiative rather than absence.
Candidates from engineering backgrounds sometimes over-explain advanced concepts and lose the panel in jargon. A clear, honest, partial answer from a non-engineer often lands better than a rambling answer from someone who technically knows more but communicates poorly.
One Practical Drill
The night before, write definitions of each row in the table above by hand, without looking. If you blank on p-value, that's your gap. Fix that one thing rather than skimming everything shallowly.
Pro Tip: Before your IIM Udaipur interview, prepare one real-world example for each statistical concept (e.g., "A/B testing in e-commerce uses hypothesis testing") so you can anchor abstract definitions in business context, which is exactly what the panel wants to see.