We spend our time bandying statistics about.
I'm a pollster by trade, so I'm into things that are actually quantifiable.
Polls are frequently taken to try to tease out or determine likely directions and trends, but once taken, they belong to the past, requiring that new polls be taken.
If you aren't taking a representative sample, you won't get a representative snapshot.
I know that a lot of people love to say that polls are wrong or don't matter, and from time to time they are - it depends on who they are polling.
Why are atoms so small? ... Many examples have been devised to bring this fact home to an audience, none of them more impressive than the one used by Lord Kelvin: Suppose that you could mark the molecules in a glass of water, then pour the contents of the glass into the ocean and stir the latter thoroughly so as to distribute the marked molecules uniformly throughout the seven seas; if you then took a glass of water anywhere out of the ocean, you would find in it about a hundred of your marked molecules.
The opinions of a convenience sample... may not represent all users.
Opinion polls are not worth the paper they are written on, in my view, through the conference season. They don't settle down again until November.
We look at all the polls, not just the Gallup Poll. So, it's kind of like if you have, you know, four out of five doctors agree that reducing cholesterol reduces your risk of a heart attack, Gallup is like the fifth doctor.
I'm a believer in the polls, by the way. Rarely do you see a poll that's very far off.
Definition of a Statistician: A man who believes figures don't lie, but admits than under analysis some of them won't stand up either.
I generally believe in polls. Scientific polls are accurate.
Any one game in baseball doesn't tell you that much, just as any one poll doesn't tell you that much.
When I started in the nineties, a sample size was a 4 and a 6.
Crazy old people are our entire source of polling information.
When you know absolutely nothing about the topic, make your forecast by asking a carefully selected probability sample of 300 others who don't know the answer either.
I view the work I've done related to statistics and economics as, roughly speaking, how to do something without having to do everything.
The reason is that in a group, individual errors on either side of the true figure cancel each other out.
