LAISHA
Open Laisha

CSIR-UGC NET Nonparametric & Rank-Based Methods: 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
37written to this exam in this subject
11syllabus 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.

Nonparametric & Rank-Based Methods in CSIR-UGC NET

Methods that hold without a normal model: the empirical distribution, goodness-of-fit, sign and rank tests for one, two and many samples, rank correlation, and tests used for scored or ranked agricultural data such as disease ratings and sensory panels.

Our 37 CSIR-UGC NET questions in this subject: 0 easy, 12 medium, 25 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 CSIR-UGC NET and are not shown on any other exam's page.

Question 1 · hard

For x ≤ y, what is Cov[Fₙ(x), Fₙ(y)] for the empirical distribution function of a random sample of size n?

  1. F(x)[1 − F(y)]/n
  2. 0, since different points use different observations
  3. F(y)[1 − F(x)]/n
  4. F(x)F(y)/n
Answer and working

A. F(x)[1 − F(y)]/n

Cov[1{X ≤ x}, 1{X ≤ y}] = F(min(x, y)) − F(x)F(y) = F(x)[1 − F(y)] for x ≤ y; averaging n iid terms divides by n. The values are positively correlated because they use the same observations.

Question 2 · hard

Normality of residuals is tested with the KS test, using the sample mean and variance as the parameters of F₀ and the standard KS table. What is the problem?

  1. The statistic must then be referred to χ² with 2 df instead
  2. The test becomes too liberal; it rejects true normality far too often
  3. Too conservative; Lilliefors critical values are needed
  4. Nothing; the KS test is distribution-free whatever is estimated
Answer and working

C. Too conservative; Lilliefors critical values are needed

Estimating parameters from the same data pulls F₀ towards Fₙ, so Dₙ tends to be small and the standard critical values are too large: the true size is well below α. Lilliefors' table, or Shapiro-Wilk, should be used.

Question 3 · hard

A sequence contains 3 symbols A and 3 symbols B, all 20 arrangements equally likely. What is the probability of exactly 2 runs?

  1. 0.20
  2. 0.3333
  3. 0.10
  4. 0.05
Answer and working

C. 0.10

Two runs occur only as AAABBB or BBBAAA: 2 of C(6, 3) = 20 arrangements, probability 0.10. 0.05 counts only one of the two orders.

Question 4 · medium

Which estimator of the centre of symmetry is associated with the Wilcoxon signed-rank test?

  1. The sample mean of the differences
  2. The mode of the differences
  3. The median of the Walsh averages
  4. The midrange of the differences
Answer and working

C. The median of the Walsh averages

The Hodges-Lehmann one-sample estimator is the median of all (dᵢ + dⱼ)/2, i ≤ j. It inverts the signed-rank test and is robust to outliers.

Question 5 · hard

Under H₀ in Friedman's test with b blocks and k treatments, what are the mean and variance of a single treatment's rank sum Rⱼ?

  1. Mean b/2, variance b/4
  2. Mean k(k + 1)/2, variance b(k² − 1)/12
  3. Mean b(k + 1)/2, variance b(k + 1)/12
  4. Mean b(k + 1)/2, variance b(k² − 1)/12
Answer and working

D. Mean b(k + 1)/2, variance b(k² − 1)/12

In each block a treatment's rank is uniform on 1, ..., k, with mean (k + 1)/2 and variance (k² − 1)/12; summing over b independent blocks multiplies both by b.

Question 6 · hard

In Cochran's Q, why can blocks in which every treatment gives the same response (all 0 or all 1) be dropped?

  1. They would make the statistic negative and must therefore be removed
  2. They violate the binary assumption and invalidate the test
  3. They double-count successes and inflate the degrees of freedom
  4. They contribute nothing to differences between treatments, and the statistic is unchanged
Answer and working

D. They contribute nothing to differences between treatments, and the statistic is unchanged

A block with Rᵢ = 0 or Rᵢ = k adds the same amount to kN and ΣRᵢ², and the same to every treatment total, leaving Q unchanged. Only blocks with mixed responses carry information.

Question 7 · hard

For a bivariate normal population with correlation ρ = 0.5, what is the population value of Kendall's tau?

  1. 0.2500
  2. 0.4826
  3. 0.3333
  4. 0.5000
Answer and working

C. 0.3333

τ = (2/π)arcsin ρ = (2/π)(π/6) = 1/3 = 0.3333. (6/π)arcsin(ρ/2) = 0.4826 is the population Spearman correlation.

Question 8 · medium

Several observations are tied in rank. How should Spearman's coefficient be computed?

  1. With the shortcut 1 − 6Σd²/[n(n² − 1)] unchanged
  2. As Pearson's correlation of the mid-ranks
  3. By breaking ties at random and averaging the results
  4. After deleting all tied observations
Answer and working

B. As Pearson's correlation of the mid-ranks

The shortcut assumes no ties; with ties, give mid-ranks and compute Pearson's r on them. Deleting ties discards data and can bias the estimate.

Question 9 · hard

How is the Pitman asymptotic relative efficiency of test A relative to test B defined?

  1. The limit of n_A/n_B needed for equal power against a fixed distant alternative
  2. The ratio of the sizes of A and B at equal sample size and fixed critical values
  3. The limit of n_B/n_A needed for equal power at equal level against alternatives approaching H₀
  4. The ratio of the powers of A and B at a fixed alternative with equal n
Answer and working

C. The limit of n_B/n_A needed for equal power at equal level against alternatives approaching H₀

Pitman efficiency uses a sequence of local alternatives converging to H₀; fixed alternatives lead to other concepts (Bahadur efficiency). The ratio is of sample sizes, not powers.

Question 10 · hard

How can the Kruskal-Wallis statistic be written in terms of sums of squares of the ranks?

  1. H = N × SS_within(ranks)/SS_total(ranks)
  2. H = SS_between(ranks)/SS_within(ranks)
  3. H = (N − 1) × SS_between(ranks)/SS_total(ranks)
  4. H = SS_total(ranks)/(N − 1)
Answer and working

C. H = (N − 1) × SS_between(ranks)/SS_total(ranks)

Without ties, SS_total of the ranks is N(N² − 1)/12, and H = (N − 1)SS_between/SS_total, a one-way ANOVA computed on ranks and scaled to be approximately χ²_(k−1). The F-type ratio between/within is the rank-transform ANOVA, a different statistic.

A study plan for Nonparametric & Rank-Based Methods 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 (46% of the 37 questions): Empirical distribution function and its properties (5, 14%); Asymptotic relative efficiency (4, 11%); Kruskal-Wallis test (4, 11%); Mann-Whitney U test and two-sample Kolmogorov-Smirnov test (4, 11%).

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

  2. Then the middle band: Sign test and Wilcoxon signed-rank test (4, 11%); Spearman's and Kendall's rank correlations (4, 11%); Cochran's Q test (3, 8%); Goodness of fit: chi-square and Kolmogorov-Smirnov tests (3, 8%).

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

  3. Then the lighter units: Analysing ordinal scores: disease ratings and sensory panels (2, 5%); Friedman's test for randomised blocks (2, 5%); Run test for randomness (2, 5%).

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

  4. Leave time for the hard ones: Analysing ordinal scores: disease ratings and sensory panels (2 hard of 2), Friedman's test for randomised blocks (2 hard of 2), Run test for randomness (2 hard of 2).

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

  5. 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 Nonparametric & Rank-Based Methods

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.