原文也批评了该研究声称的Delta和Omicron住院率没有明显差别的说法。指出相差4倍。且统计上Significant。我计算

的住院率OR(Odds ratio)也差4倍。 The Imperial team, led by Professor Neil Ferguson, said there was little evidence to suggest that omicron was milder, in contrast to mounting real-world and laboratory data showing the opposite. Imperial’s own data, which looked at more than 200,000 cases between November 29 and December 11 also showed that the risk of needing hospital attention from omicron was 0.15 per cent compared to 0.66 per cent compared to delta – four times less.

所有跟帖： 

• 你用的是汇总的数据，他用的是回归分析，两者可以不一致。你仔细看数据，得O的年轻人的比重较高，而年轻人的住院率低，所以拉低了O的总 -skyline荷9- ♀ (553 bytes) () 12/20/2021 postreply 17:11:20

• 你算说到点子上了。我也有此猜疑。只是原文只提供了汇总数据。谢谢。 -value_hunter- ♂ (0 bytes) () 12/20/2021 postreply 17:20:59

• 这有点像田横赛马，有选择性地比较，违背了统计最基本的原则 : 随机原则。 -LingYuan- ♂ (0 bytes) () 12/20/2021 postreply 20:04:40

• 田忌赛马。嗯，现在不少都是选择性比较，统计也是，其结果会差很大甚至得出相悖结果。 -pickshell- ♀ (0 bytes) () 12/21/2021 postreply 06:17:53

• 他在混淆概念。同质之间的比较，怎么算做上士对下士。谁做统计不分组？难道将孩子与老人比？ -fuz- ♂ (0 bytes) () 12/21/2021 postreply 10:31:40