ASRB Agricultural Research Service (ARS) - Agricultural Statistics scientist · India · Agricultural Scientists Recruitment Board (ICAR)
Original practice questions written to the published syllabus · not past papersASRB Agricultural Research Service (ARS) - Agricultural Statistics scientist is a government recruitment exam run by Agricultural Scientists Recruitment Board (ICAR), for candidates in India. Combined NET/ARS/SMS/STO recruitment: preliminary (objective), mains (descriptive, discipline paper) and interview for ARS.
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.
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 27 ASRB Agricultural Research Service questions in this subject: 7 easy, 19 medium, 1 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 ASRB Agricultural Research Service and are not shown on any other exam's page.
From 80 farm yield records, 18 are below 2.5 t/ha. What is the EDF estimate of the probability that a farm yields more than 2.5 t/ha?
B. 0.7750
Fₙ(2.5) = 18/80 = 0.225 estimates P(yield ≤ 2.5), so the exceedance estimate is 1 − 0.225 = 0.7750. Dividing by n + 1 is a plotting-position convention, not the EDF.
A Shapiro-Wilk test on 2,000 plot residuals rejects normality at p = 0.001, but the normal Q-Q plot is nearly straight. What is the sensible conclusion?
C. The departure is real but small and matters little for ANOVA at this sample size
With very large n, goodness-of-fit tests detect trivial departures; F-tests are robust to mild non-normality in large balanced samples. Graphical checks show whether the departure is practically important.
A row of 20 plants (10 diseased, 10 healthy) shows only 6 runs. What is the normal-approximation Z and the interpretation?
D. Z = −2.30: too few runs, indicating clustering of disease
E(R) = 11, Var(R) = 4.737, Z = (6 − 11)/2.176 = −2.30, beyond −1.96: diseased plants occur in clumps, as with plant-to-plant spread. Dividing by the variance instead of the SD understates |Z|.
In 12 on-farm comparisons, a new variety out-yields the local one on 10 farms. What is the exact one-sided sign-test p-value?
A. 0.0193
Under H₀, S ~ binomial(12, 1/2): P(S ≥ 10) = (66 + 12 + 1)/4096 = 0.0193. 0.0161 is P(S = 10) only; 0.0386 is two-sided; the normal approximation without continuity correction gives 0.0105.
Paired percentage changes in yield are strongly skewed, with a few very large gains. Which test of a zero median difference is most defensible?
D. The sign test
The signed-rank test assumes symmetry, which strong skew violates, and the t-tests assume near-normality. The sign test needs only independence and a continuous distribution.
Six treated and six untreated plants are scored for disease; the Mann-Whitney U for the treated group is 4. What is the normal-approximation Z, ignoring ties and continuity correction?
C. −2.24
E(U) = 36/2 = 18, Var(U) = 36 × 13/12 = 39, Z = (4 − 18)/√39 = −2.24. Using 36/4 as the variance gives −4.67.
A Kruskal-Wallis test across four nitrogen rates is significant. Which follow-up identifies which rates differ, with multiplicity control?
A. Dunn's test with a Holm or Bonferroni adjustment
Dunn's procedure compares mean ranks using the pooled ranking and is adjusted for multiple comparisons. Tukey's HSD on raw means is the parametric follow-up.
Why is Friedman's test attractive when blocks are fields of very different fertility, or panel judges who use the scale differently?
A. Ranking within blocks removes block effects of any size or form
Only the order of treatments within each block is used, so block differences in level or scale do not affect the test, with no additivity assumption needed.
Four farmers each try 3 seed treatments and record satisfactory germination (1) or not (0). Treatment totals are 4, 2 and 0; farmer totals are 2, 2, 1 and 1. What is Cochran's Q?
A. 6.0
N = 6, ΣCⱼ² = 20, ΣRᵢ² = 10. Q = (k − 1)[kΣCⱼ² − N²]/[kN − ΣRᵢ²] = 2 × (60 − 36)/(18 − 10) = 6.0, with 2 df; it exceeds 5.991. Omitting the factor k − 1 gives 3.0.
What does the proportional-odds (cumulative logit) model assume for an ordinal response with categories 1 to J?
D. log[P(Y ≤ j)/P(Y > j)] = θⱼ − x′β, with β shared across all cut-points
The cumulative logit model has category-specific intercepts and one set of effects for all cumulative splits. Separate βⱼ for each category is the multinomial (baseline-category) logit; the linear model treats codes as interval data.
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.