LAISHA
Open Laisha

GATE ST Statistical Inference II: Testing of Hypotheses & Sequential Analysis: practice questions

GATE - Statistics (ST) · India · IITs and IISc on rotation (2027: IIT Madras)

Original practice questions written to the published syllabus · not past papers
10worked here, free
61written to this exam in this subject
11syllabus units

About GATE ST

GATE - 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.

Statistical Inference II: Testing of Hypotheses & Sequential Analysis in GATE ST

The logic and optimality of tests: errors and power, the Neyman-Pearson lemma, uniformly most powerful and unbiased tests, likelihood-ratio tests and their large-sample forms, and Wald's sequential probability ratio test.

Our 61 GATE ST questions in this subject: 19 easy, 29 medium, 13 hard, across 11 units.

10 practice questions, worked

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.

Question 1 · medium

A single Poisson count X tests H₀: λ = 2 against H₁: λ = 4. P(X ≥ 5 | λ = 2) = 0.0527 and P(X ≥ 6 | λ = 2) = 0.0166. Which non-randomised MP test has the largest size not exceeding 0.05?

  1. Reject when X ≥ 6, with size 0.0166
  2. Reject when X ≥ 4, with size 0.1429
  3. Reject when X ≥ 5, with size 0.0527
  4. Reject when X ≤ 1, with size 0.4060
Answer and working

A. Reject when X ≥ 6, with size 0.0166

The MP test rejects for large X (λ₁ > λ₀). X ≥ 5 has size 0.0527 > 0.05, so the largest admissible non-randomised region is X ≥ 6, size 0.0166. Rejecting for small X points the wrong way.

Question 2 · hard

A single observation X has density 1 on (0, 1) under H₀ and 2x on (0, 1) under H₁. What is the power of the most powerful test of size 0.05?

  1. 0.1000
  2. 0.0975
  3. 0.9975
  4. 0.0500
Answer and working

B. 0.0975

Λ = 2x is increasing, so reject when X > c with P(X > c | H₀) = 1 − c = 0.05, c = 0.95. Power = P(X > 0.95 | H₁) = 1 − 0.95² = 0.0975. 0.9975 uses the wrong tail.

Question 3 · easy

Which family has a monotone likelihood ratio in T = ΣXᵢ?

  1. Uniform (θ, θ + 1)
  2. A mixture of two normals with unknown weight and means
  3. Poisson(λ)
  4. Cauchy with unknown location
Answer and working

C. Poisson(λ)

The Poisson is a one-parameter exponential family with natural parameter log λ increasing, so it has MLR in ΣXᵢ. The Cauchy location family has no MLR in any statistic.

Question 4 · easy

A level-α test is unbiased when:

  1. Its power is at least α at every θ in the alternative
  2. Its test statistic is an unbiased estimator of θ
  3. Its power equals 1 − α at every alternative
  4. Its size is exactly α at every θ in the null set
Answer and working

A. Its power is at least α at every θ in the alternative

Unbiasedness of a test means it rejects a false H₀ at least as often as a true one. It has nothing to do with unbiased estimation.

Question 5 · hard

In exponential families with nuisance parameters, how are UMPU tests for the parameter of interest constructed?

  1. By replacing the nuisance parameters by their MLEs and using normal critical values
  2. By testing at level α/2 for each nuisance parameter separately
  3. By integrating the nuisance parameters out against a flat prior distribution
  4. By conditioning on the sufficient statistic for the nuisance parameters
Answer and working

D. By conditioning on the sufficient statistic for the nuisance parameters

With the nuisance parameters fixed, their sufficient statistic is complete; conditioning on it removes them and gives tests that are similar on the boundary. The UMPU test is the conditional UMP or UMPU test.

Question 6 · easy

The generalised likelihood-ratio statistic for H₀: θ ∈ Θ₀ is Λ = sup_Θ₀ L(θ)/sup_Θ L(θ). When is H₀ rejected?

  1. When Λ exceeds 1, so the null fits better
  2. When Λ equals 0.5, the midpoint of its range
  3. When Λ is small, i.e. when −2 log Λ is large
  4. When Λ is large, close to its maximum of 1
Answer and working

C. When Λ is small, i.e. when −2 log Λ is large

Λ lies in [0, 1]; values near 1 mean the restricted model fits nearly as well as the full one, small values mean H₀ fits poorly. Λ cannot exceed 1 because the full maximum is at least the restricted one.

Question 7 · easy

Which large-sample test of H₀: θ = θ₀ needs only the fit of the model under H₀?

  1. The likelihood-ratio test
  2. The score test
  3. None of the three
  4. The Wald test
Answer and working

B. The score test

The score test uses U(θ₀) and I(θ₀), both evaluated under H₀. The Wald test needs the unrestricted MLE; the LRT needs both fits.

Question 8 · medium

Four ordered p-values are 0.004, 0.011, 0.020 and 0.040. How many hypotheses does Holm's step-down procedure reject at family-wise level 0.05?

  1. 3
  2. 1
  3. 2
  4. 4
Answer and working

D. 4

Holm compares p₍ⱼ₎ with 0.05/(m − j + 1): 0.004 ≤ 0.0125, 0.011 ≤ 0.0167, 0.020 ≤ 0.025, 0.040 ≤ 0.05, so all four are rejected. Bonferroni (all at 0.0125) would reject only two.

Question 9 · easy

What does the false discovery rate measure?

  1. The probability that a non-rejected hypothesis is in fact false
  2. The probability of making at least one false rejection in the family
  3. The proportion of true null hypotheses among all hypotheses tested
  4. The expected share of false rejections among all rejections
Answer and working

D. The expected share of false rejections among all rejections

FDR = E[V/R] (0 when R = 0), with V false rejections out of R rejections. The probability of at least one false rejection is the FWER.

Question 10 · medium

After a non-significant trial, a scientist computes 'observed power' from the observed effect size. What is wrong with this?

  1. Observed power must be computed at α = 0.10, not at the α used in the original trial
  2. Observed power is always exactly 0.5, whatever the trial found
  3. Observed power requires the non-central F, which applies only to ANOVA
  4. Observed power is a function of the p-value and adds no information on the effect
Answer and working

D. Observed power is a function of the p-value and adds no information on the effect

Post-hoc power at the observed effect is a monotone transformation of the p-value; a non-significant result always has low 'observed power'. Power is a planning quantity, computed for a pre-specified effect.

A study plan for Statistical Inference II: Testing of Hypotheses & Sequential Analysis in GATE ST

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.

  1. First, the heavy units (48% of the 61 questions): Neyman-Pearson lemma and most powerful tests (9, 15%); Monotone likelihood ratio and uniformly most powerful tests (7, 11%); Wald and score (Rao) tests (7, 11%); Likelihood ratio tests and Wilks' theorem (6, 10%).

    Read the manual chapters for these first and do every worked question you can find on them.

  2. Then the middle band: Power and sample-size calculation for common tests (6, 10%); Hypotheses, the two types of error, size and power (5, 8%); Multiple testing: family-wise error rate and false discovery rate (5, 8%); Relation between tests and confidence intervals (5, 8%).

    Aim for one sitting per unit, then mixed questions across them.

  3. Then the lighter units: Unbiased tests and UMPU tests for exponential families (5, 8%); Sequential probability ratio test: OC and ASN functions (4, 7%); Tests based on Z, t, chi-square and F distributions (2, 3%).

    Learn the definitions, classifications and standard numbers; these units are more often recalled than worked.

  4. Leave time for the hard ones: Unbiased tests and UMPU tests for exponential families (2 hard of 5), Likelihood ratio tests and Wilks' theorem (2 hard of 6), Monotone likelihood ratio and uniformly most powerful tests (2 hard of 7).

    Come back to these after a first pass, untimed, and read every explanation, including for the options you rejected.

  5. Bank the easy marks: Likelihood ratio tests and Wilks' theorem (3 easy of 6); Power and sample-size calculation for common tests (3 easy of 6).

    Do these early in revision so they are fast on the day.

  6. Finish timed. Sit the subject as a timed set without notes, mark it, and send each wrong answer back to its unit above.

Last updated · How Laisha writes and checks its pages · Report a mistake

Ask Laisha about this

Opens Laisha's chat with your question ready to send. Sign in or start free.

Unlock every question — GATE ST Statistical Inference II: Testing of Hypotheses & Sequential Analysis

US$10 a month · one price worldwide · cancel any time

  • Every practice question in this subject, with full working
  • Timed exam papers and progress by unit
  • Ask Laisha to explain any answer — up to 1,000 questions a month
Free or Smarter
FreeSmarter
Grow guides, pest pages, articlesRead allRead all
Ask Laisha5 a dayUp to 1,000 a month
Exam practiceThe sample on each pageEvery question, timed papers
Crop calendar and pest alerts—For your own fields
Calculators and tools—All
Uploads and reports—100 a month, real files
10 days free for every new account, no card.