Letters to the Editor
-
margin of error
The margin of error is plus or minus 2.4 percentage points, meaning the Clinton-McCain numbers are really a tossup.
No it doesn't. The "margin of error" is perhaps the most abused epistatistic. Just because the difference between two candidates' numbers is less than the so-called "margin of error", it doesn't mean that it's "really" a tie. What it actually means (e.g., in this case) is that there's 97.6% confidence in the numbers. (That's still not entirely correct, actually.)
Say that in week 1 Candidate A is at 52% and B is at 48% and the "margin of error" is 5%. Then in week 2 Candidate A is at 51% and B is at 47% but the "margin of error" is still 5%, and in week 3 Candidate A is at 53% and B is at 49% and the "margin of error" is still 5%. You'd have to be really naive to call that a "statistical tie"; A is clearly ahead.
But everything is within that "margin of error"...