Re:What we can learn from the latest indoor dining study
by szc on Saturday December 05, 2020 @01:15AM
(#25312261)
The way this is presented is going to be interpreted incorrectly. It is incorrect to compare percentages or probabilities like this.
Here is a different example to try and illustrate the problem:
Is 10% larger than 5%?
It is not possible to answer this question if the basis on which the percentages are calculated are different.
What if I said the 10% was 10% of whatever you last spent money on, but the 5% was 5% of Google's annual revenue. Which is larger?
The same thing is happening in the above - it isn't okay to calculate the probabilities in two situations and then try to cross compare them for size.
Sadly this error seems to be cropping up quite a lot! Matt Parker (Stand-up Maths) has an example of this and related to a recent election: https://www.youtube.com/watch?v=aokNwKx7gM8. His explanation is likely better than mine.
Here is a different example to try and illustrate the problem:
Is 10% larger than 5%?
It is not possible to answer this question if the basis on which the percentages are calculated are different.
What if I said the 10% was 10% of whatever you last spent money on, but the 5% was 5% of Google's annual revenue. Which is larger?
The same thing is happening in the above - it isn't okay to calculate the probabilities in two situations and then try to cross compare them for size.
Sadly this error seems to be cropping up quite a lot! Matt Parker (Stand-up Maths) has an example of this and related to a recent election: https://www.youtube.com/watch?v=aokNwKx7gM8. His explanation is likely better than mine.