0.5% strongly approve, rounds up to 1%. 39.5% somewhat approve, rounds up to 40%. The proportion of people who, to varying degrees, approve is then listed as 41%, because otherwise people would see it as them saying 1+40=40.
People are going to say they don’t understand math either way, so they just need to pick a convention and stick to it.
Ok, fair enough, yet again you can’t divide the group into smaller groups and by rounding up ending up with a larger group. Percent means per 100. It can never end up being more than 100, except for Trump who is going to reduce medicine prices 1200%. I mean, Americans use weird measurements already, their education system is already fucked and getting worse, I can understand they fail at primary school level math. But this is a British news paper.
Since the original 5 numbers only add up to 95.5, you have a leftover 0.5 to distribute between them. After doing so, you would have what could have been the original results before rounding.
Approve and disapprove are generally a collection of different possible response (for example strongly approve, approve, slightly approve) which all can be rounded. When you round the result of an addition of rounded numbers, the result can be slightly off, without changing the significance of the result.
It’s called rounding, and it’s quite common.
40.5 + 56.5 + 3.5 = 100.5
Something’s not right.
Unless 40.5, 56.5 and 3.5 are themselves additions of rounded numbers. It’s generally how that works.
Using strongly/somewhat for approve and disapprove:
39.5 + 0.5 + 55.5 + 0.5 + 3.5 = 99.5
That could work.
=40, not 41
=56, not 57
Its only possible if you round up 40.1 to 41, which is also bad math. When there are 3 variables properly rounded up, you can only have 3x0.5 added.
What I say is that there is more than three variables. The three variables are already the result of addition of other variables rounded.
So there were other options you mean? Like 0.5% couldn’t read, 0.5% drew a picture of a gun instead of answering, etc?
0.5% strongly approve, rounds up to 1%. 39.5% somewhat approve, rounds up to 40%. The proportion of people who, to varying degrees, approve is then listed as 41%, because otherwise people would see it as them saying 1+40=40.
People are going to say they don’t understand math either way, so they just need to pick a convention and stick to it.
Ok, fair enough, yet again you can’t divide the group into smaller groups and by rounding up ending up with a larger group. Percent means per 100. It can never end up being more than 100, except for Trump who is going to reduce medicine prices 1200%. I mean, Americans use weird measurements already, their education system is already fucked and getting worse, I can understand they fail at primary school level math. But this is a British news paper.
39.5 = 40
0.5 = 1
40 + 1 = 41
———————-
55.5 = 56
0.5 = 1
56 + 1 = 57
———————-
3.5 = 4
Since the original 5 numbers only add up to 95.5, you have a leftover 0.5 to distribute between them. After doing so, you would have what could have been the original results before rounding.
Yet this is based on the assumption there were more options, plus the total should still not exceed 100%
What they should have done was adding up all the combined answers first before rounding up or down, this would make the total 100.
That is doing it wrong. You should only ever round as the last step before display.
Could they have 40.74 and rounded it down to the nearest .5?
The bigliest and bestest President the world has ever seen obviously surpasses the wokist 100% barrier! First to ever do that!
Edit: Accidentally replied to myself instead of you
It can’t happen with honest rounding
What rounding algorithm was used? I have a hard time coming up with values or methods to be more than a single unit off.
A wrong one clearly haha
Approve and disapprove are generally a collection of different possible response (for example strongly approve, approve, slightly approve) which all can be rounded. When you round the result of an addition of rounded numbers, the result can be slightly off, without changing the significance of the result.