ICAR AIEEA-PG - Group 07 Statistical Sciences (Master's admission and PG scholarship) · India · National Testing Agency for ICAR
Original practice questions written to the published syllabus · not past papersICAR AIEEA-PG - Group 07 Statistical Sciences (Master's admission and PG scholarship) is a postgraduate admission exam run by National Testing Agency for ICAR, for candidates in India. Computer-based test, 120 minutes, single paper of 120 MCQs, +4 / -1, English. 2026 exam on 4 July 2026. Fills ICAR's all-India quota of Master's seats at agricultural universities.
Source: cdnbbsr.s3waas.gov.in.
Checked against the exam body's published information. Formats and fees change; confirm the current edition there before you book.
Basic principles of experimental design, the analysis of variance, CRD and RBD: degrees of freedom, sums of squares, F, standard errors and critical differences on small data sets.
Fisher's three principles and how they shape a field or animal trial: uniformity trials, plot size and shape, the CRD, RCBD and Latin square with their analysis and efficiency, missing plots, transformations, multiple comparisons and analysis of covariance.
Our 38 ICAR AIEEA-PG questions in this subject: 25 easy, 13 medium, 0 hard, across 12 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 ICAR AIEEA-PG and are not shown on any other exam's page.
What is the main purpose of randomisation in a field experiment?
A. To remove systematic bias and make the error estimate valid
Random allocation gives every treatment the same chance of landing on any unit, so no treatment is systematically favoured and the errors behave as independent, which justifies F and t tests. Reducing error by grouping is local control; error degrees of freedom depend on the design, not on randomisation.
Before contour lines are drawn on a fertility contour map, the basic-unit yields are usually:
D. Smoothed by moving averages of neighbouring units
Individual basic-unit yields are noisy, so each is replaced by the average of a small group of neighbouring units before equal-yield lines are drawn. Logs, ranks or CVs are not the standard step.
As plot size increases in a uniformity trial, the coefficient of variation of plot yields:
C. Decreases, but at a diminishing rate
Larger plots average over more soil so the CV falls, but neighbouring soil is correlated, so each extra unit of area buys less reduction. Only with uncorrelated units (b = 1) would variance per unit fall in proportion to size.
In a CRD with 5 replications per treatment the error mean square is 2.375 on 16 df. Taking t(0.05, 16) = 2.120, what is the critical difference between two treatment means?
A. 2.07
SEd = sqrt(2 MSE / r) = sqrt(2 x 2.375 / 5) = 0.9747; CD = 2.120 x 0.9747 = 2.07. Omitting the 2 gives 1.46; using 1.96 gives 1.91; dividing the root by r instead of putting r inside gives 0.92.
For which experiment is a completely randomised design most appropriate?
A. A laboratory test of fungicides on agar plates kept in one incubator
The CRD puts all unit-to-unit variation into error, so it suits homogeneous material such as plates in one incubator. A slope, varied animals or two soil types call for blocking (or a covariate for initial weight).
How are treatments randomised in an RCBD?
D. Separately and independently within each block
Each block is a complete replicate, and a fresh randomisation in each block keeps treatment and position independent. Once-over randomisation is a CRD; row-and-column balance is a Latin square; repeating one order in every block is a systematic layout.
How many error degrees of freedom does a 6 x 6 Latin square have?
D. 20
Error df = (t - 1)(t - 2) = 5 x 4 = 20. 25 would be (t - 1)^2, which is the RCBD formula with t blocks; 30 = t(t - 1); 15 subtracts the treatment df twice.
An RCBD with 6 treatments and 4 blocks has one missing plot estimated by the missing-plot formula. How many error degrees of freedom remain?
D. 14
Normal error df = (4 - 1)(6 - 1) = 15; one df is lost for the estimated value, leaving 14. 23 is the total df of the complete experiment.
Which procedure is designed specifically to compare each treatment with a single control?
C. Dunnett's test
Dunnett's critical values allow for t - 1 comparisons against the same control and are more powerful for that question than procedures built for all pairs (Tukey, Duncan) or all contrasts (Scheffé).
If the difference to be detected is halved while everything else stays the same, the required number of replications:
D. Becomes four times as large
r is proportional to 1/d^2, so halving d multiplies r by 4. Equivalently, the SE of a difference falls only with the square root of r.
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.
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.