GATE - Statistics (ST) · India · IITs and IISc on rotation (2027: IIT Madras)
Original practice questions written to the published syllabus · not past papersGATE - Statistics (ST) is a postgraduate admission exam run by IITs and IISc on rotation (2027: IIT Madras), for candidates in India. Computer-based, 3 hours, 100 marks (15 General Aptitude + 85 subject). MCQ, MSQ and NAT; negative marking on MCQ only; on-screen virtual calculator.
Source: gate2026.iitg.ac.in.
Checked against the exam body's published information. Formats and fees change; confirm the current edition there before you book.
Probability from its axioms to the limit theorems: random variables, expectation, generating functions, inequalities and modes of convergence, ending with the laws of large numbers and the central limit theorem that justify most of applied statistics.
Our 77 GATE ST questions in this subject: 20 easy, 38 medium, 19 hard, across 11 units.
Choose an answer, then open the working. Each explanation says why the right answer is right and why the others are not.
These 10 were written to GATE ST and are not shown on any other exam's page.
X and Y are independent exponential variables with the same rate λ. Let U = X + Y and V = X/(X + Y). Which statement is correct?
B. U is gamma with shape 2, V is uniform on (0, 1), and U and V are independent
Inverse: x = uv, y = u(1 − v), |J| = u. Joint density λ²e^(−λu) u for u > 0, 0 < v < 1 factorises into a gamma(2, λ) density in u times 1 on (0, 1) in v: independent, V uniform (= beta(1, 1)). Rate 2λ describes the minimum, not the sum.
X and Y are independent standard normals and R = √(X² + Y²). What is P(R > 2)?
D. 0.1353
R² = X² + Y² is chi-square(2), i.e. exponential with mean 2: P(R² > 4) = e^(−4/2) = e^(−2) = 0.1353. e^(−4) forgets the mean of 2; 0.0455 is P(|Z| > 2) for a single normal.
Lyapunov's inequality states that (E|X|ʳ)^(1/r) is non-decreasing in r > 0. Which consequence follows?
D. If E|X|³ is finite then E(X²) is finite
(E|X|²)^(1/2) ≤ (E|X|³)^(1/3), so a finite third absolute moment forces a finite second. The converse fails (a t-distribution with 3 df has finite variance but infinite third moment). The correct direction is E|X| ≤ √E(X²).
X has mean μ and standard deviation σ, with no other information. By Cantelli's one-sided inequality, what is the sharpest bound on P(X − μ ≥ 2σ)?
D. 0.20
Cantelli: P(X − μ ≥ kσ) ≤ 1/(1 + k²) = 1/5 = 0.20. Chebyshev's two-sided 1/k² = 0.25 is valid but weaker; halving it to 0.125 is not justified because the distribution need not be symmetric; 0.0228 is the normal tail.
Xₙ converges in distribution to N(0, 1), and Yₙ converges in probability to 2. To what does XₙYₙ + 1 converge in distribution?
D. N(1, 4)
Slutsky: XₙYₙ → 2Z with Z ~ N(0, 1), i.e. N(0, 4); adding 1 gives N(1, 4). N(1, 2) forgets that the multiplier squares in the variance.
∫₀¹ g(x) dx is estimated by (1/n)Σg(Uᵢ), where U₁, U₂, ... are iid uniform on (0, 1) and ∫₀¹|g(x)| dx < ∞. Which theorem guarantees that the estimate converges to the integral with probability one?
A. Kolmogorov's strong law of large numbers
g(Uᵢ) are iid with mean ∫₀¹g and finite E|g(U)|, so by the SLLN their average converges almost surely. The CLT describes the size of the error (order 1/√n) and needs a finite variance; Chebyshev alone gives only convergence in probability.
Which conditions does the Lindeberg-Lévy central limit theorem require of X₁, X₂, ...?
A. Independent, identically distributed, with finite positive variance
Lindeberg-Lévy: iid with 0 < σ² < ∞. The non-identical case with a (2 + δ)-th moment condition is Liapounov's theorem. Normal summands need no limit theorem at all, since their sum is exactly normal.
N is Poisson with mean 3 and, given N, S = X₁ + ... + X_N where the Xᵢ are iid Bernoulli(0.4), independent of N. What is the distribution of S?
C. Poisson with mean 1.2
G_S(s) = G_N(G_X(s)) = exp(3((0.6 + 0.4s) − 1)) = exp(1.2(s − 1)): Poisson with mean 3 × 0.4 = 1.2 (Poisson thinning). A binomial needs a fixed number of trials; the mean 3 ignores the thinning.
An analyst back-transforms the mean of log-transformed yields by exponentiating it. Compared with the arithmetic mean yield, the result is:
A. Never larger, because E(log Y) ≤ log E(Y) by Jensen's inequality
exp(mean of log Y) is the geometric mean. Since log is concave, E(log Y) ≤ log E(Y), so the geometric mean never exceeds the arithmetic mean; they are equal only if Y is constant. Being inverse functions does not let exp and log pass through an expectation.
X and Y are independent gamma variables with a common scale and shapes 3 and 2. What is the distribution of X/(X + Y)?
D. Beta(3, 2), independent of X + Y
For independent Gamma(α) and Gamma(β) with a common scale, X/(X + Y) ~ Beta(α, β) independent of X + Y ~ Gamma(α + β). Here α = 3, β = 2. Beta(2, 3) swaps the parameters; (X/6)/(Y/4) on chi-square scale would give an F, which is a different ratio.
This plan follows where our questions actually fall, so the time goes where the marks are likely to be. It is our reading of the syllabus, not the exam body's weighting.
Read the manual chapters for these first and do every worked question you can find on them.
Aim for one sitting per unit, then mixed questions across them.
Learn the definitions, classifications and standard numbers; these units are more often recalled than worked.
Come back to these after a first pass, untimed, and read every explanation, including for the options you rejected.
Do these early in revision so they are fast on the day.
Last updated · How Laisha writes and checks its pages · Report a mistake
US$10 a month · one price worldwide · cancel any time
| Free | Smarter | |
|---|---|---|
| Grow guides, pest pages, articles | Read all | Read all |
| Ask Laisha | 5 a day | Up to 1,000 a month |
| Exam practice | The sample on each page | Every question, timed papers |
| Crop calendar and pest alerts | — | For your own fields |
| Calculators and tools | — | All |
| Uploads and reports | — | 100 a month, real files |
An AI operating system for agriculture: free guides, manuals and tools written for every country.