It’s usually stated, and certainly implied, that any poll conducted here, or anywhere that voluntary posting is permitted, is not scientifically valid. That doesn’t slow news channels and newspapers and Hollywood mags and almost any publishing entity from securing the services of some polling expert to bolster their reading audience with “the latest findings.” This is not a revelation.
But when you see a poll being conducted here and some largish number of Dopers respond, where’s the crossover point for you between “Hmm, that’s interesting” and “Damn! That must be the case”?
Do numbers alone persuade you or must there be a segment of those Dopers who respond to the poll whose opinion you already value?
Does their membership status (including “Guest”) influence your acceptance or rejection of their poll response(s)?
Do you assume that there are a significant number of Dopers who are not responding that would support an unpopular opinion or stance if they just had the courage to reply?
What other reactions to Doper Polls do you typically have?
Unless the question is ‘strawberry vs. chocolate’ I don’t see how you ever can obtain a relavant sample here, there are just not enough RIBS (Republican Idiot Bush Supporters) posting here.
No poll conducted here rates more than a “That’s interesting” from me. Too small a group, and one that is badly skewed for sampling purposes. That said there a number of Dopers whose opinions I respect, even if I disagree with them whole-heartedly.
too small a sample, too self-selecting, with a strong tendency for over-representation of unusual viewpoints/experiences, and too many polls where one side becomes a landslide.
I appreciate posts which explain why someone reacts the way they do more than a post which just responds to the questions posed in the poll.
It’s not the sample size that’s the issue; it’s how the sample is obtained. (That’s an important statistical principle. If a sample is truly representative of the population it’s supposed to represent—if everyone in the population is equally likely to be included in the sample—you can get meaningful results from even a relatively small sample.)
No matter how many Dopers respond to a poll here on the SDMB, I would never assume the ones who respond are respresentative of the Dopers who didn’t respond (whether because they didn’t see the poll, or didn’t care enough to answer the question, or just aren’t the type to respond to polls)—let alone representative of the population at large.
I’ve used y’all once or twice when I need a contrarian vote – I’m thinking “who the hell does [whatever],” but before I put “no one does [whatever]” into print, I’ll ask here – and, yup, usually there are a few people who do [whatever].
This statement is a good example of how the wording of the poll’s issue can influence the responses. If it’s obvious that the poll is after negative reactions, it’s not hard to get some. If the opposite effect is wanted the wording can still direct traffic.
Some thread titles for polls likely to return skewed results:
– How often do you have sex with your blood relatives?
– Does God speak to you in your dreams?
– Do you ever tell your children where you buried their favorite kitty that you shot when it bit you?
– When you can’t get a sore to heal how many people do you show it to for advice?
– Are you embarrassed about being stupid?
Actually, Zeldar, I was thinking of the last time I did this, when the question was about whether kids play on jungle gyms in their backyards or not – but I guess your examples work equally well.
