Worked example · Geography · Year 11
DCF: Data and computational thinking → Problem-solving and modelling
A secure Year 11 response, annotated against the success criteria, and your answer key. Key figures: chance of at least one flood in ten years 65%. EAD today £171,000. BCRs: A 1.20, B 1.96, C 4.77, B+C 2.09. Weighted favourites: finance officer C (38), residents A (36, but over budget), ecologist B+C (40). With doubled chances: do nothing £342,000, A’s BCR 2.40, B+C’s BCR 4.19. Pupils may recommend a different option. It is only wrong if their own figures do not support it.
(a) “1-in-10-year” is a probability, not a timetable. The chance is 10% every year, whatever happened last year, just as rolling a six does not make the next roll less likely to be a six. Zone 1 could flood again this winter.
(b) Chance of no flood for ten years = 0.910 = 0.349. Chance of at least one flood = 1 − 0.349 = 0.651, so about 65%. A Zone 1 family is more likely than not to be flooded in the next ten years.
Treats the return period as an annual probability and handles “at least once” correctly using the complement. Evidences criterion 1.
| Zone | Working | EAD today (£/year) |
|---|---|---|
| 1 | 0.1 × 40 × 30,000 | 120,000 |
| 2 | 0.02 × 60 × 30,000 | 36,000 |
| 3 | 0.005 × 100 × 30,000 | 15,000 |
| Total if the council does nothing | 171,000 | |
Zone 1 has the fewest homes, but it costs the most: £120,000 a year, 70% of the total. That is because it floods five times as often as Zone 2 and twenty times as often as Zone 3. How often a place floods matters more than how many homes it has.
The formula is applied to every zone with the working shown, and the pupil explains what the model is saying rather than only giving the numbers. Evidences criterion 2.
| Option | Zone 1 | Zone 2 | Zone 3 | Total EAD |
|---|---|---|---|---|
| A — wall | 0.01 × 40 × 30,000 = 12,000 | 0.01 × 60 × 30,000 = 18,000 | 15,000 | 45,000 |
| B — natural | 0.05 × 40 × 30,000 = 60,000 | 0.0125 × 60 × 30,000 = 22,500 | 0.004 × 100 × 30,000 = 12,000 | 94,500 |
| C — resilience | 0.1 × 40 × 9,000 = 36,000 | 36,000 | 15,000 | 87,000 |
| B+C | 0.05 × 40 × 9,000 = 18,000 | 22,500 | 12,000 | 52,500 |
| Option | Annual benefit | 50-year benefit | 50-year cost | BCR | Within £2.0m? |
|---|---|---|---|---|---|
| A — wall | 171,000 − 45,000 = 126,000 | 6,300,000 | 4,000,000 + 1,250,000 = 5,250,000 | 1.20 | No |
| B — natural | 76,500 | 3,825,000 | 1,200,000 + 750,000 = 1,950,000 | 1.96 | Yes |
| C — resilience | 84,000 | 4,200,000 | 480,000 + 400,000 = 880,000 | 4.77 | Yes |
| B+C | 118,500 | 5,925,000 | 1,680,000 + 1,150,000 = 2,830,000 | 2.09 | Yes |
Option C is by far the best value: every £1 spent prevents £4.77 of damage. But C does not stop a single flood. It only makes Zone 1 homes cheaper to repair, and it does nothing at all for the 60 homes in Zone 2. The wall prevents the most damage of all (£126,000 a year) but has the worst BCR, and it costs twice the budget.
Every EAD shows the new input that changed. Benefits are measured against the do-nothing baseline, and 50-year costs include upkeep. The best BCR is questioned, not just accepted. Evidences criteria 2 and 3.
| Option | Homes flooded in a 1-in-50 | Value | Homes | Env. | Soon | Comm. | Finance | Residents | Ecologist |
|---|---|---|---|---|---|---|---|---|---|
| A — wall | 0 | 2 | 5 | 1 | 3 | 2 | 26 | 36 | 23 |
| B — natural | 40 (Zone 1 now floods in a 1-in-20) | 3 | 3 | 5 | 1 | 4 | 31 | 29 | 37 |
| C — resilience | 100 (Zones 1 and 2) | 5 | 1 | 3 | 5 | 3 | 38 | 26 | 32 |
| B+C | 40, with resilience fitted | 3 | 3 | 5 | 4 | 4 | 34 | 35 | 40 |
Working for one total, the residents on the wall: 2×1 + 5×5 + 1×1 + 3×2 + 2×1 = 2 + 25 + 1 + 6 + 2 = 36.
Each stakeholder prefers a different option: the finance officer C, the residents A and the ecologist B+C. The residents' favourite costs £4.0m and the budget is £2.0m, so it cannot be chosen. Their next choice is B+C, only one point behind. B+C is the only option that comes first or second for all three people and fits the budget.
Both calculated scores follow the stated rules. For the 1-in-50 flood, the pupil works out which zones are still inside the flood for each option. The weighted totals are checked with one full line of working, and the budget is applied as a hard limit, not as another score. Evidences criterion 4.
| Doubled chances | Total EAD | Annual benefit | 50-year benefit | BCR |
|---|---|---|---|---|
| Do nothing | 342,000 | — | — | — |
| A — wall | 0.02×40×30,000 + 0.02×60×30,000 + 0.01×100×30,000 = 90,000 | 252,000 | 12,600,000 | 2.40 |
| B+C | 0.1×40×9,000 + 0.025×60×30,000 + 0.008×100×30,000 = 105,000 | 237,000 | 11,850,000 | 4.19 |
Doubling every chance doubles every EAD, because EAD is directly proportional to the chance. So every benefit doubles and the costs stay the same. B+C is still better value than the wall (4.19 against 2.40), so the order does not change. What does change is that the wall now easily pays for itself. In a wetter future the residents' case for it gets much stronger, but it still costs £4.0m up front, so it is still unaffordable. The recommendation survives the test, and that makes me more confident in it.
The model is re-run with changed inputs and the effect is explained from the structure of the formula. The pupil also separates “does the ranking change?” from “does the decision change?”. Evidences criterion 5.
I recommend B+C: natural flood management in Upper Cwm Rhaf together with flood doors and resilience measures for the 40 Zone 1 homes. It costs £1.68m, inside the £2.0m budget. It cuts the town's expected flood damage from £171,000 to £52,500 a year, and prevents £2.09 of damage for every £1 spent. It is the ecologist's first choice (40) and second for both the finance officer (34) and the residents (35). With doubled flood chances its BCR rises to 4.19. The finance officer will be disappointed: C alone is £1.2m cheaper and has a BCR of 4.77. Zone 2 residents may be too, because they wanted a wall. Zone 1 will still flood about once in 20 years on average, so the council should also make sure every Zone 1 household is signed up for flood warnings.
The model is simpler than reality in at least three ways. First, it gives every flooded home the same £30,000 of damage and counts nothing else: no shops, roads, stress or lost school days. That understates the benefits, so every option looks worse than it really is. Second, it does not discount future money. Most of the benefit arrives over 50 years while most of the cost is paid on day one, so leaving out discounting makes every option, especially the wall, look better than it is. Third, it assumes everything works perfectly from year one: the trees give their full effect immediately, and every family puts up its barriers. The trees need 10–15 years to grow, and if only 24 of the 40 families use their barriers, C's BCR falls from 4.77 to 2.86. So B and C both look better in the model than they would in practice.
The recommendation draws on the BCR, the matrix, the budget and the scenario, and names who loses out. The evaluation names three assumptions and gives the direction of the bias for each, one with a calculated example. Evidences criteria 3, 4 and 5.