LAISHA
Open Laisha

CSIR-UGC NET Statistical Inference I: Estimation: practice questions

CSIR-UGC NET - Mathematical Sciences (statistics option) · India · National Testing Agency for CSIR

Original practice questions written to the published syllabus · not past papers
10worked here, free
36written to this exam in this subject
10syllabus units

About CSIR-UGC NET

CSIR-UGC NET - Mathematical Sciences (statistics option) is a PhD, fellowship and lectureship exam run by National Testing Agency for CSIR, for candidates in India. Computer-based, 3 hours, Parts A (general aptitude), B and C (subject); statisticians answer from the statistics unit.

Official site: csirnet.nta.ac.in.

We have not been able to confirm these details against the exam body's own specification, so fees, dates and pass marks are left out on purpose. Check the current format, eligibility, fees and dates on the official site.

Statistical Inference I: Estimation in CSIR-UGC NET

How to estimate well and know it: sufficiency and completeness, unbiasedness and minimum variance, the Cramer-Rao bound, the Rao-Blackwell and Lehmann-Scheffe theorems, the methods of moments, maximum likelihood and minimum chi-square, and confidence intervals.

Our 36 CSIR-UGC NET questions in this subject: 3 easy, 15 medium, 18 hard, across 10 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 CSIR-UGC NET and are not shown on any other exam's page.

Question 1 · hard

Which criterion identifies a minimal sufficient statistic T (the Lehmann-Scheffé criterion)?

  1. f(x; θ)/f(y; θ) is free of θ exactly when T(x) = T(y)
  2. E[g(T)] = 0 for all θ implies g(T) = 0
  3. f(x; θ) is free of θ whenever T(x) = T(y) holds
  4. T is unbiased for θ and has the smallest variance at every θ
Answer and working

A. f(x; θ)/f(y; θ) is free of θ exactly when T(x) = T(y)

The ratio criterion characterises minimal sufficiency. The last option defines completeness, a different property; minimum variance among unbiased estimators defines a UMVUE.

Question 2 · medium

Basu's theorem is used to show that X̄ and s² are independent for normal samples. Which facts does the argument use, for a fixed σ²?

  1. X̄ and s² are uncorrelated, so they are independent
  2. X̄ and s² are both unbiased, so they are independent
  3. s² is complete sufficient for σ² and X̄ is ancillary for σ²
  4. X̄ is complete sufficient for μ, while s² is ancillary for μ
Answer and working

D. X̄ is complete sufficient for μ, while s² is ancillary for μ

With σ² fixed, X̄ is complete sufficient for μ, and s² (a function of the deviations Xᵢ − X̄) has a distribution free of μ. Basu then gives independence for every (μ, σ²). X̄ is not ancillary for σ², and zero correlation alone does not give independence outside joint normality.

Question 3 · hard

X₁, ..., X₅ are iid uniform on (0, θ). Using Basu's theorem (X₍ₙ₎ is complete sufficient and X₍₁₎/X₍ₙ₎ is ancillary), what is E(X₍₁₎/X₍₅₎)?

  1. 0.8333
  2. 0.50
  3. 0.1667
  4. 0.20
Answer and working

D. 0.20

Independence gives E(X₍₁₎) = E(X₍₁₎/X₍₅₎) E(X₍₅₎), so E(X₍₁₎/X₍₅₎) = [θ/(n + 1)]/[nθ/(n + 1)] = 1/n = 0.20. 1/6 = 0.1667 is E(X₍₁₎)/θ, and 5/6 is E(X₍₅₎)/θ.

Question 4 · hard

Which implication holds in general (assuming a minimal sufficient statistic exists)?

  1. An ancillary statistic is complete
  2. A minimal sufficient statistic is complete
  3. A complete sufficient statistic is minimal sufficient
  4. A sufficient statistic is independent of every ancillary
Answer and working

C. A complete sufficient statistic is minimal sufficient

Bahadur's result: complete sufficient implies minimal sufficient. The converse fails (uniform (θ, θ + 1)). Independence from ancillaries needs bounded completeness (Basu), not sufficiency alone.

Question 5 · easy

What does the Lehmann-Scheffé theorem state?

  1. An estimator attaining the Cramér-Rao bound is a function of a complete statistic, and hence unique
  2. An unbiased estimator that is a function of a complete sufficient statistic is the unique UMVUE
  3. The MLE is the UMVUE whenever a complete sufficient statistic exists
  4. Any function of a sufficient statistic is the UMVUE of its own expectation
Answer and working

B. An unbiased estimator that is a function of a complete sufficient statistic is the unique UMVUE

Completeness ensures uniqueness of the unbiased function of T, and Rao-Blackwell ensures it beats every other unbiased estimator. Sufficiency without completeness is not enough, and the MLE is often biased (uniform maximum, normal σ̂²).

Question 6 · medium

For a random sample from the Laplace (double exponential) density (1/2)e^(−|x − θ|), what is the MLE of θ?

  1. The sample mode
  2. The midrange (X₍₁₎ + X₍ₙ₎)/2
  3. The sample mean
  4. The sample median
Answer and working

D. The sample median

Maximising L means minimising Σ|xᵢ − θ|, which is achieved by the median. The mean minimises squared deviations (normal case); the midrange is relevant for uniform location.

Question 7 · hard

θ̂ = X₍ₙ₎ is the MLE of θ for uniform (0, θ). Which statement about its large-sample behaviour is correct?

  1. √n(θ̂ − θ) converges to N(0, θ²/12), as for any MLE
  2. √n(θ̂ − θ) converges to N(0, 1/I(θ)), with I computed from the density in the usual way
  3. θ̂ is inconsistent, because it is biased for every n
  4. n(θ − θ̂)/θ converges to a standard exponential; the rate is 1/n, not 1/√n
Answer and working

D. n(θ − θ̂)/θ converges to a standard exponential; the rate is 1/n, not 1/√n

P(n(θ − X₍ₙ₎)/θ > x) = (1 − x/n)ⁿ → e^(−x). The regularity conditions fail, so the usual √n normal limit does not hold. θ̂ is biased but consistent.

Question 8 · hard

For a generalised linear model with canonical link, such as logistic regression for germination data, which statement is correct?

  1. Fisher scoring converges but Newton-Raphson diverges, because the observed Hessian is random
  2. Neither method applies; the EM algorithm must be used for every GLM
  3. Fisher scoring and Newton-Raphson coincide, each step being a weighted least-squares fit
  4. Newton-Raphson needs no starting values, because the likelihood is quadratic
Answer and working

C. Fisher scoring and Newton-Raphson coincide, each step being a weighted least-squares fit

With the canonical link the observed and expected information are equal, so the two algorithms are the same and take the iteratively reweighted least squares form. Starting values are still needed (usually from the data themselves).

Question 9 · hard

θ̃ is a √n-consistent estimator of θ. What can be said of the one-step estimator θ̃ + ℓ′(θ̃)/[nI(θ̃)]?

  1. It is no better asymptotically than θ̃ itself
  2. It is unbiased for θ in samples of every size
  3. It is asymptotically efficient, with the same limiting distribution as the MLE
  4. It is exactly equal to the MLE whenever n is large enough for the central limit theorem to apply
Answer and working

C. It is asymptotically efficient, with the same limiting distribution as the MLE

A single scoring (or Newton) step from a √n-consistent start yields an estimator with √n(θ₁ − θ) → N(0, 1/I(θ)), the MLE's limit. It is not exactly the MLE, and not unbiased in finite samples.

Question 10 · medium

Newton-Raphson for the Cauchy location MLE, started far from the data, converges to a value that is not the global maximum. What is the reason and the standard remedy?

  1. Newton-Raphson always converges globally for this model, so the error must lie in the data entry
  2. The score equation has multiple roots; restart from the median and compare log-likelihoods
  3. The Cauchy MLE does not exist; use the sample mean instead
  4. The information is zero; switch to the method of moments
Answer and working

B. The score equation has multiple roots; restart from the median and compare log-likelihoods

The Cauchy log-likelihood is not concave and can have several local maxima. A consistent start (the median) plus a comparison of log-likelihoods at the roots finds the global maximum. The sample mean is useless for the Cauchy, and no moment estimator exists.

A study plan for Statistical Inference I: Estimation in CSIR-UGC NET

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 (58% of the 36 questions): Completeness, ancillarity and Basu's theorem (7, 19%); Iterative maximum likelihood: Newton-Raphson and Fisher scoring (6, 17%); Cramer-Rao inequality and Fisher information (4, 11%); Maximum likelihood: properties, invariance and asymptotic normality (4, 11%).

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

  2. Then the middle band: Minimum chi-square and other methods (3, 8%); Minimum variance unbiased estimation; Rao-Blackwell and Lehmann-Scheffe theorems (3, 8%); Sufficiency, the factorisation theorem and minimal sufficiency (3, 8%); Interval estimation: pivotal quantities and coverage (2, 6%).

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

  3. Then the lighter units: Large-sample confidence intervals (2, 6%); Point estimation: unbiasedness, consistency and efficiency (2, 6%).

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

  4. Leave time for the hard ones: Cramer-Rao inequality and Fisher information (3 hard of 4), Maximum likelihood: properties, invariance and asymptotic normality (3 hard of 4), Sufficiency, the factorisation theorem and minimal sufficiency (2 hard of 3).

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

  5. Bank the easy marks: Minimum variance unbiased estimation; Rao-Blackwell and Lehmann-Scheffe theorems (1 easy of 3); Iterative maximum likelihood: Newton-Raphson and Fisher scoring (1 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 — CSIR-UGC NET Statistical Inference I: Estimation

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.