Reply #72: Well, Mitofsky's MOE is still 1%. No mention of cluster.

[TruthIsAll]

Does MP know that Mitofksky claims the national exit poll is a RANDOMLY_SELECTED sample?

And by so doing, confirms the simple MOE calculation that I have been posting for months:
MOE = 1/sqrt(n).

Here n= 13,047.
So the MOE is 0.88%.
I won't quibble.

And that this also would apply to state polls, except that the MOE would be higher (1.8% - 4.0%) based on the sample size.

Face it. The key arguments of MP (non-randomness and MOE "cluster" effect) are effectively demolished by Mitofsky in the notes accompanying the National Exit Poll.

That MOE sure is a tight noose, isn't it?

Koufax award?
Sandy was perfection.

MP is far from perfect.







Sorry, no 5% MOE.