HUGE problems. The most obvious one is that hardly anyone will even answer when a stranger calls anymore. They don’t recognize the number, so to voicemail it goes! I know I usually don’t answer unknown numbers unless I am expecting a call. Heck, the other day I very nearly refused to answer a call–it was a scheduler for my upcoming MRI. So I’m glad I did, but I almost didn’t!
I think most people don’t even take the call or hang up immediately once they realize it is nothing they actually care about. I’ve heard the reports of a 2% response rate.
That means that you didn’t hear the opinions of 98%. So how do you know that that 2% is even slightly representative of the other 98%? Are people who answer calls from strangers different from the broader population? Maybe just old people who are lonely and desperate to talk to anyone will take those calls. And are they representative at all of the broader population? Hmmmmm.
And many of us use cell phones ONLY. I don’t even HAVE a land line!
Add in the “shy” Conservative voters (and almost ALL Conservative voters are “shy”), and there seems no way to get a representative sample that responds accurately.
So, for example, Emerson polling measured Trump winning in Iowa by 1. He actually won by 8. Think about that for a moment. Is that span reliable across groups and states? I don’t know. But I do know that was a gargantuan miss!
In short, the more accurate a poll is, the more intrusive it is. And the “feedback loop” is that the more intrusive it is, the lower the response rate (and the greater the +/- or accuracy).
So what is left is mainly “push” polls that really don’t tell you anything and are just preaching to the choir.
It seems to me that the problems in polling are now pretty much insurmountable. An online survey is quite problematic, but MAY give a general sense of which way the wind is blowing. But that may be the best you can get!