When new evidence arrives, add it with a date and source, keep the old record visible, and let the evidence set the claim. Do not edit history so that an argument looks cleaner.
This is the last lesson in case-study reasoning, and it uses every earlier habit: verified sources, selected facts, linked chains and fair comparison.
What does a good update look like?
A good update is a new dated line, not an erasure. The ledger keeps one row per fact, with date, source and note. If a fact is replaced, the old row stays and is marked as superseded.
The evidence ledger is designed for this. It separates what was verified, what you interpret and what is still unanswered.
Worked example
All data invented. Your 2015 note says: “3,200 of 8,000 homes in Town K flooded, and the town had no flood wall.” A 2023 council report says a wall was built in 2020 and, in a 2023 storm, 900 homes flooded. Rainfall in 2023 was 140 mm against 280 mm in 2015.
Step 1, add dated lines. Keep the 2015 row. Add a 2020 row (wall built) and a 2023 row (900 homes flooded, 140 mm of rain), each with its source.
Step 2, calculate the change. 3,200 − 900 = 2,300 fewer homes. 2,300 ÷ 3,200 = 0.71875, so about 72% fewer.
Step 3, check what else changed. Rainfall halved (140 against 280 mm). So the fall in flooded homes cannot all be credited to the wall.
Step 4, write a measured claim. “Fewer homes flooded in 2023 (900 against 3,200), but rainfall was also lower, so the wall may have contributed; the evidence does not isolate its effect.”
The mistake to watch for
The common slip is to claim a cause the data cannot show.
Mistaken claim: “The flood wall reduced flooding by 72%.”
It ignores that the 2023 storm was half the size.
The correction is to report the change as observed, name the other factor and use cautious wording. If you need more, state what you would look for: a storm of similar size before and after the wall.
Check yourself
1. A count falls from 500 homes to 350. What is the percentage decrease?
Show answer
500 − 350 = 150. 150 ÷ 500 = 0.3, so 30%.
2. A 2018 note says “no warning system”. A 2022 source says one was installed in 2021. How should you record it?
Show answer
Keep the 2018 line, add a 2021 line with the 2022 source, and mark the old statement as superseded.
3. Why is “the policy worked” a risky conclusion from two data points?
Show answer
Two points cannot rule out other changes, such as a smaller event or a lower population. Use “may have contributed” and name what else to check.
Where this leads next
Test the whole module with the case-study reasoning practice set. The statistics and distribution explorer can help check percentage change. For notes that stay tidy as new evidence arrives, see online one-to-one Geography tuition.