The finding contrasts with a number of other national polls, and it comes amid concerted Democratic efforts to prevent Kennedy from harming Biden’s campaign.
Why don’t you educate me? This is for sure not 95% confidence interval nor 3 sigma. This might be 80% but this is close enough to one sigma.
The reason why I said it is std, is because suppose that you have a single person instead of 1000. If we expect the actual numbers to be about 50% for Biden or Trump, then with one person you get 100% or 0%, which is +/-50% error over 50% median. Which gives std of 1. After that, std decreases as 1/sqrt(1000).
I understand that I might miss there small factors, but I could not be that far from correct answer. Where do I went wrong?
They are talking about couple percent lead one way or another. When there is 3.1% standard deviation. In short, it is in the noise.
3.1% is not the standard deviation
It is 1/sqrt(1000), which I think is std or close to it
You know just enough statistics to be confidently incorrect
Why don’t you educate me? This is for sure not 95% confidence interval nor 3 sigma. This might be 80% but this is close enough to one sigma.
The reason why I said it is std, is because suppose that you have a single person instead of 1000. If we expect the actual numbers to be about 50% for Biden or Trump, then with one person you get 100% or 0%, which is +/-50% error over 50% median. Which gives std of 1. After that, std decreases as 1/sqrt(1000).
I understand that I might miss there small factors, but I could not be that far from correct answer. Where do I went wrong?