ASRB NET - Agricultural Statistics (and Computer Application / Bioinformatics) · India · Agricultural Scientists Recruitment Board (ICAR)
Original practice questions written to the published syllabus · not past papersLooking at data before modelling it: summaries, robust statistics, graphs that show distributions and relationships honestly, outlier detection, and the tidy-data and reporting habits that make an analysis checkable. Deliberately practical and software-based.
In the exams: Not an explicit unit in the catalogued exams; programme course and the daily practice of every role in the ladder. GATE DA data handling (present)
How our questions for ASRB NET spread across the units of Exploratory Data Analysis & Statistical Graphics. Difficulty: 3 easy, 17 medium, 5 hard.
| Graphics for designed experiments: interaction plots and residual plots | 7 | |
| Transformations to symmetry and constant variance | 4 | |
| Tables for reports and publications | 4 | |
| Types of data and measurement scales | 2 | |
| Tidy data: variables, observations and tables | 2 | |
| Scatterplots, smoothing (LOESS) and scatterplot matrices | 2 | |
| Five-number summaries, box plots and robust statistics (median, MAD, trimmed means) | 1 | |
| Histograms, density estimates and quantile-quantile plots | 1 | |
| Outliers and influential observations | 1 | |
| Graphical perception and honest chart design | 1 |
Choose an answer, then open the working. Each explanation says why the right answer is right and why the others are not.
Percentage germination is computed for seed lots tested with 25 seeds in some dishes and 400 in others. What should EDA note about this derived variable?
A. Percentages from 25 seeds are far less precise than from 400 seeds
The SE of a proportion is √(p(1 − p)/n), so 25-seed tests are four times less precise in SE than 400-seed tests. Keeping the counts and denominators allows a binomial analysis that weights correctly.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Farm size and district are typed on every plot row of a 3,000-plot on-farm trial file. Why is this considered untidy?
A. Two unit types share one table, so a farm's correction must be repeated on many rows
Farm-level variables belong in a farm table keyed by farm identifier. Repeating them invites inconsistent values when one copy is corrected and others are not.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
A scatterplot of farm size against yield is crowded near small farms with a few very large farms stretching the axis. Which change is most useful?
B. Plot farm size on a logarithmic scale
Positive, multiplicative variables such as farm size are spread evenly by a log scale, which also often linearises the relation. Deleting large farms discards real data.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
In an interaction plot of yield against nitrogen dose with one line per variety, the lines are nearly parallel. What does this suggest?
D. Little or no interaction between nitrogen and variety
Parallel lines mean the difference between varieties is the same at every dose. Their slope shows the nitrogen effect and their separation the variety effect.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Residuals from a field trial are mapped on the plot layout, and positive residuals cluster along one edge of the field. What does this show?
D. A spatial trend that blocking did not remove
Residuals should look like random noise on the field map. Patches indicate spatial variation, which spatial analysis (row and column terms or correlated errors) can model, and which better blocking can avoid next time.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Germination percentages in a trial range from 5% to 98%. Which transformation is traditional before ANOVA?
C. Angular (arcsine square root)
Binomial proportions have variance p(1 − p)/n, small near 0 and 100%; arcsin √p equalises it. It matters most when proportions span a wide range, as here.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Bacterial counts 2, 8 and 32 (thousands per gram) are analysed on the log scale. The mean of the logs is back-transformed. What value results?
A. 8
Back-transforming the mean log gives the geometric mean: (2 × 8 × 32)^(1/3) = 512^(1/3) = 8. The arithmetic mean, 14, is larger, as it always is for unequal positive values.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
What is the main drawback of mean-separation letters (means sharing a letter do not differ significantly)?
A. They hide effect sizes and depend on the procedure
Letters compress many pairwise tests, but readers cannot see how large the differences are, and changing from LSD to Tukey's test changes the letters. The procedure and level must be stated.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Data from a log-scale analysis are reported in a table. Which presentation is correct?
C. Back-transformed geometric means with asymmetric limits
Back-transforming log means gives geometric means; limits back-transform to asymmetric intervals. A log-scale SE cannot be exponentiated into an original-scale SE, and mixing raw means with log-scale LSDs is inconsistent.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Which practice makes the tables in an analysis report reproducible?
B. Generating every table by script from the cleaned data, with code and software versions archived
Scripted tables update automatically when data are corrected and can be regenerated by reviewers; hand-typed numbers introduce errors that are hard to trace.
Further reading: Tukey, Exploratory Data Analysis; Cleveland, Visualizing Data; Wickham, Tidy Data; Tufte, The Visual Display of Quantitative Information
Ten here, 25 in the app. The Laisha app has every practice question we have written to ASRB NET 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.