Good news/bad news/statistics

[Guest guest]

Saying, as someone just did, that there were '800 recoveries' can be wonderful,

depending on how many tried the stuff out.

If you got 800 out of 1000 who tried, that's wonderful! But 800 out of 10,000

means you have a 92% chance this will fail for you..

Similarly, 800 deaths from flu shot complications is awful. But out of

100,000,000 it's still not nice, but deaths from flu without the shot might have

been 80,000, which is still less than 0.1% of 100,000,000.

It's very important to know the size of the test, not just the number of

successes or failures, otherwise no estimate can be made whether it's any good,

neutral, dangerous, or what.

bG