Allocating demonstrations when the old index and the pest bulletin disagree
The district extension office has one field vehicle and four staff-days left in the last week of October. Three mandals are competing for ten front-line cotton demonstrations. Mandal A has 1,850 ha of cotton, a 38% rainfall deficit, 50% attendance at the last kisan mela, 35% baseline adoption of seed treatment, and 90 minutes of travel. Mandal B has 2,400 ha, a 22% deficit, 35% attendance, 58% adoption, and 45 minutes of travel. Mandal C has 1,100 ha, a 45% deficit, 70% attendance, 29% adoption, and 2 hours 10 minutes of travel. The office uses an old composite index with weights of 40% rainfall deficit, 10% area, 20% attendance and 30% prior adoption, where low adoption means high need. On 24 October the state pest bulletin reports pink bollworm catches of nine moths per trap per week in Mandal B, above the district action threshold; A and C report two and one. The junior assistant has prepared a note recommending Mandal C. The sarpanch of Mandal B has telephoned twice.
1. Calculate the composite score for the three mandals using the stated weights. State your normalisation method, show the working, and say what the result does and does not justify.
Read the worked answer
Use min–max normalisation to 0–100, with higher score meaning greater need. Rainfall deficit (range 22–45): A = (38−22)/23×100 = 69.6; B = 0; C = 100. Area (range 1,100–2,400): A = 57.7; B = 100; C = 0. Attendance (range 35–70): A = (50−35)/35×100 = 42.9; B = 0; C = 100. Prior adoption is reversed because low adoption means high need (range 29–58): A = (58−35)/29×100 = 79.3; B = 0; C = 100.
Weighted totals: A = 0.40×69.6 + 0.10×57.7 + 0.20×42.9 + 0.30×79.3 = 27.8 + 5.8 + 8.6 + 23.8 = 66.0. B = 0 + 10 + 0 + 0 = 10.0. C = 40 + 0 + 20 + 30 = 90.0.
The old index therefore ranks C first, A second and B third, so the junior assistant's arithmetic is consistent. But the result only justifies prioritising C for chronic adoption work. It does not capture acute pest risk, and the weights are three years old and were never validated. B's low score is an artefact of being best on adoption and worst only on area; it hides the fact that B has the largest cotton area and the lowest meeting attendance, which makes it the most exposed if an outbreak starts. A score is a starting point for judgement, not a substitute for current field intelligence.
2. After the pest bulletin arrives, would you change the allocation? Give a specific distribution of the ten demonstrations and justify it against the index.
Read the worked answer
Yes. The composite index measures chronic structural need; the bulletin measures acute risk. B has 2,400 ha, only 35% meeting attendance and trap catches above the district action threshold - a combination that can produce rapid spread and heavy loss. Reallocate: six demonstrations to B with a pink bollworm scouting and management focus (pheromone traps, timely crop termination, destruction of affected squares, need-based insecticide); three to C to address the chronic adoption gap; one to A as a comparison site.
The vehicle constraint supports this rather than blocking it. B is 45 minutes away, so four staff-days go further there; C needs an overnight stay, so its three demonstrations can be clubbed into one trip. Before shifting, spend half a day in B with the agriculture extension officer to confirm trap counts and larval incidence in green bolls. If the counts are a trap-placement error or are localised to one pocket, keep four demonstrations in C. Record the reason for departing from the index so the decision is transparent and defensible in the next review.
3. Identify two important limitations of the data behind the index and state what you would do in the first week to reduce the risk of a wrong allocation.
Read the worked answer
Limitations: first, the adoption figure is a three-year-old baseline and may no longer reflect current practice, seed choice or labour availability. Second, attendance at a kisan mela is a weak proxy for need: it measures who came, not who is at risk, and may overstate need in a mandal with better mobilisation. Other gaps include the absence of current pest, soil and water data, and the fact that area alone says nothing about intensity or yield potential.
First-week actions: (1) ask the AEOs in all three mandals for a rapid transect of ten fields per mandal, recording plant stand, boll damage, trap counts and irrigation source; (2) verify the adoption baseline with a short random check of 15 farmers per mandal rather than relying on the old survey; (3) check whether the trap catches in B are from correctly placed traps and whether larvae are present in green bolls; (4) place the revised evidence before the office review and update the weights with written reasons. Do not spend the full four staff-days before this verification.