in the polls! No “buyer’s remorse” here, eh?
Archive for the ‘Polls’ Category
Things really start
15
Oct
to change once one starts to look at “likely voters.” Hope for change!
Keep it up. Already the dems are talking about how they can win through fraud..
Yeah, DUH!
03
Oct
Don’t kid yourself, most polls are biased. They are often (and usually) used to further a certain political agenda. This is NOT the place to look for truth. Keep that in mind as you watch the polls.
When response
02
Oct
rates are so very low, are the polls even accurate? I personally don’t answer calls I don’t recognize. Who does? And a lot of people I know use cell phones as their home phones, so they don’t even get asked….
There ARE some
28
Jul
points of good news. Now don’t get cocky, kid. There is a very long way until November. Things can and will change.
But still, this is good.
Even with
26
Jul
likely voters we are starting to see a Romney lead. Now it not time to get too exited. Things can and will change a bunch before November. But remember, this is registered voters and not likely voters (likely voters are far more disposed to conservatives). But even so, we are starting to see change. Hope and change…
Now don’t get cocky. These folks are more than able and willing to cheat. Plus, we don’t want complacency. Win by 10 percentage points and they can’t cheat. If it’s 1 percentage point, you get a second Obama term.
You know,
23
Jul
it’s certainly NOT over and you will see some fluctuations where Obama is on top and where Romney is on top, but this does look promising. Now I would like to see Romney over the 50% mark, because then it is a done deal (if one side has 51% there is no way the other side can win). But it is only July. There is a long way to go before the elections.
A word about the sample: Rasmussen uses a Likely Voter sample rather than what many other polls use, which is a Registered Voter sample. Why? Well, there are several reasons and in truth it is a bit of a judgment call, and there are pros and cons to either way you go. But likely voter samples have typically been more accurate, and registered voter samples are much more tilted in favor of Democrats. Plus, Rasmussen has been pretty dang accurate over the last couple elections.
Now I would have zero problem using a registered voter sample (with its tilt toward democrats) if that were more accurate. We need to see the data for what they are, and not what we wish them to be. But in fact, I’m not at all convinced that such samples are more accurate, and almost for sure they are less so. So I think Rasmussen is a good place to put your money.
I know that discussing samples is kind of “in the weeds” and I apologize. But it is important that we understand the very real differences in polls. That is just one issue (grossly over-sampling the left is another).
So why are certain
17
Jul
polls so obviously biased?
You would think that there would be a drive to call the thing correctly regardless of the facts. I think that factor exists, but on the whole if you think that they are motivated to tell the truth you would be wrong.
The truth is, I can get any result I want from a poll (yes, I have both done and taught polling at the university level). Depends on my sample and the way I phrase (or even order) a question. We need to understand the “halo effect” and many other issues. The bottom line is that any fool who has even dabbled in this area knows that these polls are invalid. So why aren’t these companies correcting the problems?
Part of the reason is pure wishful thinking. Part of it is that the bias is transparent to those creating the survey–it is part of who they are. Part of it is that they know that they will soon be gone (at another polling company) so they might as well “fudge” in the direction of their biases. It really is a big part insight (or lack thereof) into their own psychological processes. It’s just hard for them not to be a cheerleader. It’s not lying per se, it’s a mental disorder. And it is transparent to the perpetrator.