ICAR AICE-JRF/SRF (PhD) — Environmental Sciences (subject 71) and Agricultural Meteorology (subject 60) · India · National Testing Agency for ICAR
Original practice questions written to the published syllabus · not past papersDesigning, analysing and publishing environmental research: hypotheses, field and laboratory designs, long-term experiments, mixed methods, meta-analysis, reproducibility and ethics.
In the exams: UGC NET Paper 1 (research aptitude) and Unit 9; ICAR AICE-JRF/SRF; PhD coursework
How our questions for ICAR AICE-JRF/SRF spread across the units of Research Methods in Environmental & Sustainability Science. Difficulty: 5 easy, 29 medium, 13 hard.
| Mixed-effects models and repeated measures | 8 | |
| Field, plot and landscape-scale experimental designs | 7 | |
| Spatial statistics and geostatistics | 6 | |
| Systematic review and meta-analysis | 6 | |
| Research questions and hypotheses | 4 | |
| Long-term experiments and monitoring plots | 4 | |
| Multivariate methods in ecology | 4 | |
| Before-after-control-impact designs | 3 | |
| Open data, reproducibility and research ethics | 3 | |
| Qualitative and participatory methods | 1 | |
| Scientific writing and peer review | 1 |
Choose an answer, then open the working. Each explanation says why the right answer is right and why the others are not.
A test has β = 0.25. What is its power?
B. 0.75
Power = 1 − β = 1 − 0.25 = 0.75. 0.95 and 0.05 confuse power with 1 − α and α.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
What are the error degrees of freedom in a 5 × 5 Latin square design?
A. 12
Error df = (t − 1)(t − 2) = 4 × 3 = 12. Total df is 24; rows, columns and treatments use 4 each.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
For studying the effect of nitrogen deposition on lichens across a region, a gradient design (regression along deposition levels) is often preferred over a few-level ANOVA design because:
A. It yields the shape of the response and more information per sampled site
Sampling many levels lets the response curve and thresholds be estimated. Confounding along gradients must still be handled with covariates or design.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
Which is the main scientific value of long-term experiments that short projects cannot provide?
B. Detecting slow changes, lag effects and responses to rare extreme events
Slow processes (soil carbon), lags and rare events need long records. Years are not substitutes for spatial replication of treatments.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
Monthly soil respiration is measured on the same chambers for two years, with correlation decreasing as months get further apart. Which covariance structure fits this pattern?
D. First-order autoregressive, AR(1)
AR(1) sets correlation ρ at lag 1, ρ² at lag 2 and so on, decaying with distance in time. Compound symmetry assumes equal correlation at all lags; identity assumes none.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
A grouping factor has only three levels (three regions). What is the usual advice?
B. Treat it as a fixed effect, as three levels give a poor variance estimate
Variance among very few levels is estimated poorly; a common rule of thumb is about five or more levels for a random effect.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
A meta-analysis reports a mean lnRR of −0.105 for soil carbon under a practice. What percentage change does this represent?
B. -10%
% change = (e^−0.105 − 1) × 100 = -10.0%, about −10%. Reading lnRR directly as a percentage is close here but not exact; +11% inverts the sign.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
A meta-analysis of 20 studies gives Cochran's Q = 40. What is I²?
C. 52.5%
df = k − 1 = 19; I² = (40 − 19)/40 × 100 = 52.5%, moderate heterogeneity. Using k instead of k − 1 gives 50%.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
Why is vote counting (tallying significant versus non-significant studies) discouraged in evidence synthesis?
D. It ignores effect size and precision, penalising small studies
Vote counting treats a small underpowered study's non-significance as evidence against an effect; meta-analysis weights by precision.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
A funnel plot of effect size against precision is strongly asymmetric, with small studies showing only large positive effects. This suggests:
C. Possible publication bias or other small-study effects
Missing small studies with null or negative results produce asymmetry; heterogeneity and poor quality in small studies can also do so, so it is suggestive, not proof.
Further reading: See the manual sources: Hurlbert 1984; Quinn and Keough; Zuur et al.; Legendre and Legendre; Underwood 1994; CEE Guidelines; Koricheva et al.
Ten here, 47 in the app. The Laisha app has every practice question we have written to ICAR AICE-JRF/SRF in this subject, marked as you go, with timed papers, a jump-to-unanswered list and a certificate when you pass at 70 per cent.
Free for ten days, then one subscription of US$10 a month. We would rather say so here than have you find it later.
An AI operating system for agriculture: free guides, manuals and tools written for every country.