LAISHA
Open Laisha

ASRB NET Linear Models & Regression Analysis: practice questions

ASRB NET - Agricultural Statistics (and Computer Application / Bioinformatics) · India · Agricultural Scientists Recruitment Board (ICAR)

Original practice questions written to the published syllabus · not past papers
10worked here, free
70written to this exam in this subject
13syllabus units

About ASRB NET

ASRB NET - Agricultural Statistics (and Computer Application / Bioinformatics) is a professional certification run by Agricultural Scientists Recruitment Board (ICAR), for candidates in India. NET: one objective paper of about 150 questions, 2 hours, negative marking, online, Master's level (pattern as reported for other disciplines).

Official site: asrb.org.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.

Linear Models & Regression Analysis in ASRB NET

Applied regression and linear-model questions: which model, which test, which diagnostic.

The general linear model in matrix form: estimability, the Gauss-Markov theorem, least squares with full and less-than-full rank, the analysis of variance as a linear model, model checking and selection, and regression diagnostics. It is the theory under every design in Phase 3.

Our 70 ASRB NET questions in this subject: 13 easy, 39 medium, 18 hard, across 13 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 ASRB NET and are not shown on any other exam's page.

Question 1 · medium

When is lambda' beta estimable in y = X beta + e?

  1. When lambda' lies in the null space of X
  2. When lambda' lies in the row space of X
  3. Only when X has full column rank
  4. When lambda' has all elements equal to one
Answer and working

B. When lambda' lies in the row space of X

lambda' beta is estimable iff lambda' = a'X for some a. Full rank makes every function estimable, but rank deficiency does not stop contrasts being estimable.

Question 2 · hard

Yield correlates strongly with rainfall across districts, but the partial correlation given irrigated area is near zero. What is the most likely explanation?

  1. Irrigated area has no relation to yield or rainfall at all
  2. Rainfall has no effect on yield anywhere in the country
  3. The partial correlation must have been calculated wrongly
  4. Irrigation relates to both and drove the simple correlation
Answer and working

D. Irrigation relates to both and drove the simple correlation

When a third variable is related to both, the simple correlation can be largely explained by it; the partial correlation removes that common association.

Question 3 · medium

The lag-one autocorrelation of residuals in time order is about 0.4. What is the approximate Durbin-Watson statistic?

  1. 2.0
  2. 1.2
  3. 0.4
  4. 2.8
Answer and working

B. 1.2

d is approximately 2(1 - rho-hat) = 2 x 0.6 = 1.2, below 2, indicating positive autocorrelation.

Question 4 · medium

A residual plot against fitted values shows a clear U-shaped curve. What is indicated?

  1. A missing curvature term such as x squared
  2. Heteroscedastic errors
  3. Too many predictors
  4. Normally distributed errors
Answer and working

A. A missing curvature term such as x squared

Systematic curvature in residuals means the mean function is mis-specified.

Question 5 · medium

Using the 2p/n rule with p = 4 and n = 40, which leverage should be flagged?

  1. 0.05
  2. 0.25
  3. 0.15
  4. 0.10
Answer and working

B. 0.25

2p/n = 0.20; only 0.25 exceeds it. 0.10 is the average leverage.

Question 6 · easy

Regressing predictor x_j on the other predictors gives R_j^2 = 0.9. What is its variance inflation factor?

  1. 0.9
  2. 10
  3. 1.11
  4. 9
Answer and working

B. 10

VIF = 1/(1 - R_j^2) = 1/0.1 = 10, usually taken as serious collinearity.

Question 7 · medium

A subset model with p = 4 parameters has RSS = 120; the full model's residual mean square is 5; n = 30. What is Mallows' Cp?

  1. 2.0
  2. -2.0
  3. 32.0
  4. -14.0
Answer and working

A. 2.0

Cp = RSS_p/s^2 - n + 2p = 24 - 30 + 8 = 2.0. Values near or below p indicate little bias.

Question 8 · medium

What does the Gauss-Newton method do at each iteration?

  1. Linearises by Taylor expansion, then solves least squares
  2. Fits a polynomial of degree equal to the number of parameters
  3. Transforms the response to logs and fits by OLS
  4. Randomly searches the parameter space for a smaller RSS
Answer and working

A. Linearises by Taylor expansion, then solves least squares

Gauss-Newton updates estimates by solving a linear least squares problem in the derivative matrix around the current values.

Question 9 · hard

Fitting log y by OLS instead of fitting y = a exp(bx) directly by non-linear least squares implies which assumption?

  1. Multiplicative errors with constant variance on the log scale
  2. Additive errors with constant variance on the original scale
  3. No errors at all in the response variable
  4. Errors whose variance falls as the mean increases on the log scale
Answer and working

A. Multiplicative errors with constant variance on the log scale

The log fit treats errors as proportional; the direct fit assumes additive errors. They give different estimates.

Question 10 · easy

Which transformation linearises y = a exp(bx)?

  1. sqrt(y) = a + bx
  2. log y = log a + bx
  3. y^2 = a + b x^2
  4. 1/y = a + bx
Answer and working

B. log y = log a + bx

Taking natural logs gives a straight line in x with intercept log a and slope b.

A study plan for Linear Models & Regression Analysis in ASRB 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 (53% of the 70 questions): Fixed, random and mixed effects; nested and crossed classifications (8, 11%); Non-linear regression: estimation and inference (8, 11%); Examination of residuals and model adequacy (7, 10%); Multiple and partial correlation (7, 10%); Selecting the best regression: stepwise, all-subsets, AIC and BIC (7, 10%).

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

  2. Then the middle band: Influence and leverage: Cook's distance and hat values (6, 9%); Multicollinearity and ridge regression (6, 9%); Weighted and generalised least squares (6, 9%); Estimable functions and the Gauss-Markov theorem (5, 7%); Less-than-full-rank models and generalised inverses (5, 7%).

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

  3. Then the lighter units: Distribution of quadratic forms and the Fisher-Cochran theorem (2, 3%); Testing linear hypotheses; R-squared and adjusted R-squared (2, 3%); The general linear model and least squares in matrix form (1, 1%).

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

  4. Leave time for the hard ones: Distribution of quadratic forms and the Fisher-Cochran theorem (1 hard of 2), Estimable functions and the Gauss-Markov theorem (2 hard of 5), Influence and leverage: Cook's distance and hat values (2 hard of 6).

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

  5. Bank the easy marks: Weighted and generalised least squares (2 easy of 6); Examination of residuals and model adequacy (2 easy of 7).

    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 — ASRB NET Linear Models & Regression 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.