I’m not going to completely repost here everything that I write on The Lay Scientist, but I can’t help thinking it’d be a shame if I didn’t even mention them!
So, earlier this week I wrote a piece about David Spiegelhalter – if you don’t know who he is, then you’ll just have read it to find out!
Anyway, it’s more fun than your average blog post, since it involves a quiz – with a twist. The idea is that you answer questions and then give yourself a score based on how sure you are that you’ve got the right answer.
The idea is to get people to realise how difficult it is for statisticians to predict risks and that understanding uncertainty goes someway to make sure we have the correct level of confidence in a given statistic. A statement like ‘you’ve 2% more likely to get heart disease if you eat more than two bacon sandwiches a month (that’s a made up statistic by the way) is something we should understand as ‘true’ only with a specific degree of confidence.
And then of course there’s the question of ‘unknown unknowns’; what happens if there’s something about bacon sandwiches which we haven’t discovered yet which alters the result? We can’t take that into account in our risk analysis, can we?!
So if that whets your appetite, read more here. You can also see what results a cross-section or readers got when the took the how ignorant are you? test.