BTW, sometimes the poll is halted for a time period. It will sometimes say "Continues soon" until it resumes.
crib notes for those of us who don’t want to click onto the NYSlimes??
ping
In very simple terms, it can't. A sample of 1,000 can accurately reflect the entire country, if a substantial majority of those contacted respond. A sample can even represent the entire country well when only a small minority responds - if those who respond are consistent from one election to the next. That's what we used to have.
These past three years, we have had a rabid left growing increasingly volatile, violent, and vindictive. Those willing to respond in that environment are no longer a predictable or consistent group. The bottom line for polling this year is simple. We don't know.
Q: Does the NY Times SUCK?
A: YES!
(Poll results based on one response)
They all routinely oversample democRATS to skew it to the democRATS
CNN’s poll two days ago was D+10 in the results yet it was D+10 in the sample.
For House seats, district polling is the only way they can have any accuracy. Nationwide polling does not work.
Polling seems to be halted according to time zone, and limiting calls from mid afternoon to ~11pm.
I think it’s worth looking at.
These are competitive races (tossups) the polling is being done from the voters list like the campaigns do.
And, they show that the REPUBs are in surprisingly good shape compared to these national “generic” polls showing a double digit lead.
Democrats would have to win half these races to take the House, and right now, by these polls, they are not doing that.
re: response rates of 1-2 percent
even the campaigns supposedly have response rates of around 10 percent, which they get from doing callback after callback.
Also, look at the variance in results for different turnout assumptions, as well as the issue polling
Things could still go south but I think this looks quite optimistic for Repubs
Conservatives aren’t responding to pollsters’ questions.
Given the low response rates, mid-term polls are way off.
No wonder CNN crowed the Democrats regained the House.
Oh, joy!
We may well see small Democrat gains in the House (single digit range) and a gain of a few Senate seats by Republicans.
2% response rate isn’t a surprise when you consider they may be calling while people are at work and a large percentage hang up rather than participate.
What is more important is that with a 2% response it is unlikely that the results of the pole will really have any direct correlation with the population in question.
I’m sure they will continue to call until they get whatever minimum number of responses they decided they needed beforehand, but the sample will be skewed dramatically to the subset willing to participate.
Thanks! Very interesting