« Budget Cuts: Mine Safety Will Suffer | Main | UPDA »

December 18, 2008


Feed You can follow this conversation by subscribing to the comment feed for this post.

The so-called poll is useless, since it was not a controlled sampling and reflects only the opinions of participants, which might or might not have been organized to take part by proponents of MTR. It may only accidentally reflect the actual opinion of folks in eastern Kentucky -- the key word being "accidentally."

Get a statistics class at Morehead State U. or Eastern Kentucky U. or one of the eastern Kentucky community colleges to scientifically poll eastern Kentucky (perhaps only the counties where MTR is practiced and those downstream in the affected watersheds) and then report those results. They may be the same as your "poll's" results, but at least they will be valid results.

No, it sampled a bunch of people who say, and it is confirmed, they live in Eastern Kentucky, they answered some questions about mountain top removal. So these things can go either way. But it was controlled per region, I don't really care what Louisville thinks in regards to coal mining. I care about what we think and we sampled 500 people from Eastern Kentucky, with a 6 or 7% margin of error. I bet I am with 7%, either way, IN EASTERN KENTUCKY. If any does not believe this issue has truly polarized an entire region, is in for a big surprise.

To be statistically valid and therefore truly meaningful, the sampling must be controlled and not left to chance participation (or to not-so-chance participation if promoted or organized by proponents or opponents).

Ask any statistician. Ask any local high school mathematics teacher.

I made no reference to sampling opinion outside eastern Kentucky, certainly not to sampling Louisville. (Louisville?) Repeat -- "to scientifically poll eastern Kentucky (perhaps only the counties where MTR is practiced and those downstream in the affected watersheds)."

Joe is right - the word is "selection bias", and this poll certainly has it. A straw poll isn't scientific, no matter how much you'd like it to be, even if you didn't intentionally distort the results.

You are damn right, selection bias, I wanted Eastern Kentucky participants and I got them. I guess that is bias, exactly what I wanted. I wanted moderates, I got moderates. Eastern Kentucky is full of coal miners and moderates. I want to know what they think. That's why it is a survey and not a poll.

The fact that 85% of your participants were from EKY isn't selection bias if that is who you wanted to sample - but the fact that your site is more widely read by people who agree with you would mean selection bias. Your 420 participants from Eastern Kentucky can't be said to be indicative of the whole region - to do that, you'd have to call people randomly.

oh by the way, a poll is a kind of survey, and it is exactly the kind of survey that you did.

About half of those sampled have never read my blog. I have very few who agree with me here on MTR. All I did is prove 62% +/-7% approve of MTR in Eastern Kentucky. Thats all I care about and I know I am within 7%.

Shack - from a political scientist friend of mine, you can't prove any polling, especially a margin of error, without random sampling. Which is, coincidentally, exactly what I said previously. I'd like to see you take this straw poll to anyone with experience in polling and try to pass it off as fact...

even if, hypothetically, 62% of Eastern Kentuckians support MTR, your dime store poll doesn't prove it.

The comments to this entry are closed.