koreanhistory1-8, Exercise 2: Why there is no single figure (a sample; yours will differ)

Model answer:

1. [Source] The lead gives a range (deaths 798 to 7,509; participants about 0.8 to 2 million), and the article says the statistics are "uncertain, and are a subject of controversy."
2. [Source] The 2019 NIKH database is intentionally conservative (deaths 798 to 1,170; participants 823,702 to 1,065,172), Park Eun-sik's 1920 count is higher (7,509 deaths; 2,023,098 participants), and the Japanese police report gives about 1.1 million participants.
3. [Source] The article says Japanese records were politically altered, and Kwon Tae-eok thought Japan had incentives to understate and found Park's estimate plausible.
4. [My own reading] Each count has a motive and a method (a conservative database, a contemporary Korean tally, a police report), so a range with named owners is more honest than a single number. I would trust the Japanese police report least on deaths, because the article describes the records as politically altered, but I would check how the tool-computed total differences in Park's table come out against the original.