The message I had entered was too short. I was pleased to lengthen my message to at least 2 characters.
How accurate are internet polls? I’d say they’re pretty near 100% accurate at recording the opinions of the participants. But they’re hideously biased in terms of who the participants are, which means you can’t count on them for much of anything.
I voted for all of the last five options except the lying one, because I think they all apply.
I selected “Really darn accurate” and “not at all accurate”.
I think you’re me.
In fact, judging by the result so far, I think we’re the majority…
But, I’m sure the results of this internet poll will be accurate!
I clicked all the checkboxes! Because I could!
You forgot to include the options “Baba Booey” and “Stephen Colbert.”
I found this sometime ago and kept it around because it expresses what I have felt in my gut for several decades.
“There is often a vast gulf between how people say they behave and how they actually behave. … The fibs might even be subconscious, with the subject simply saying what the surveyor wants to hear.” - Steven D. Levitt, professor of economics at the University of Chicago, along with former New York Times Magazine editor Stephen J. Dubner
They’re worthless. No control over sample = worthless poll.
Well, if (for example) you only want to poll the set of SDMB posters who sometimes hang out in IMHO, then you *can *get a tolerably accurate result, I’m sure.
The problem is that many polls we see here are of a type which can’t even accomplish that, for one of two reasons: they either invite unintentional misinformation (“What is your IQ?” “Are you a better than average driver?”) or because the question asked will disproportionately encourage a certain segment of the population to volunteer an answer (“Have you ever ridden a horse?” “Are you a medical professional?” “What is your IQ?”).
They are worthless and I do my best to keep them that way.