Something Troubling About The Polls

Can anyone confirm or refute this? It was a LTTE in the Orlando Sentinel that claimed the pollsters weren't including seniors over 75 (or possibly younger). If this is the case, Obama may not really have a lead since seniors are more likely to actually vote and probably favor McSame.

During the past two weeks when I received three separate telephone calls from pollsters. I am a senior citizen, a proud American and a registered voter...I was asked first, "Are you a registered voter?" I passed that test, but then the second question was regarding age. I didn't seem to fit into any of the acceptable age ranges. Then I was asked my exact age. I replied, "I am 75." What I heard next was, "Thank you, goodbye."

Does anyone know anything about this?

How does the person know it was a Head to Head Poll?

When pollsters call they need a certain number of people in each demographic.
Perhaps they already had their allotment of calls to that age group.....
or they may not have even been polling seniors. The poll could have
been designed to only poll those 30 or under.

The letter means nothing.

I've been called to answer polling questions and been disqualified, and it's not because they hate my demographic, it's because they don't need any more of my demographic.

I know they can't fact-check every LTE but jeez. The really obvious ones, like this, deserve an "editor's note" underneath, something you almost never see any more.

But if they are looking to poll certain age groups that are harder to read or possibly for feedback on ad campaigns they may focus on certain age groups.

they may be randomly calling from phone books, but they are doing some type of sub-group sampling.

you dont get anything statistically sound if you just do a generic survey.

My company does Random sampling everyday (RDD.) It is very in-depth data generated purely by random sampling. We use weighting to correct the ups or downs of age cells, ethnicity, sex, gender, etc.

I'm sure they have to weight their polls with certain percentages of Democrats v. Republicans--according to the percentages of registration, as well as by age. Perhaps they already had the quota of those in your age range having answered their questions. Can't have 50% of the respondents be in the same age bracket, or it would skew the reliability of the poll.

:silly:

They include seniors. They also include them in the number that they vote. The problem maybe seniors are more likely to have landlines and more likely to answer them and therefore the sample of senior that were wanted for the polling formula was reached.

Or the person could be lying because they don't like the poll results and want to cast doubt about them

and the younger votes (and me) without landlines are not at all represented

polling to give seniors the very impression that the polls they see dont include them so to ignore them.

Dirty trick yes, but not above anything else phone jamming wise that has gone on before.

In the first scenario it could have been a poll on any political subject, local referendums, and so on.

and secondly, if that happened at all, it was likely a poll aimed at a particular targetted group. That happens to.

Thirdly, don't try and unnerve us. It won't work.

oops. nope, I'm sorry, my concern button is broken. Thanks for trying to push it, though, that tickled.

Jesus man, get a grip. Do you have any idea how many dumbass LTTEs get written every day? I don't either but it has to be a lot based on the small sample I read.

When you do a poll that seeks statistical validity, you basically have a bunch of sub-groups and respective weightings which seek to provide reliable data about each sub-group. Think of it this way.

If you have ten samples of 100 black women over 75...and they 99% vote the same way, you don't likely need to sample as many of them the next time because there is little to no real variation in their responses. You could arguably only sample 20 of them to confirm the pattern.

If you have ten samples of 100 white women over 75... and there is great variation, you need to keep doing more extensive sampling of them to be sure you're adequately capturing the variation within this sub-group.

There's a lot of math behind how this works, but you get the idea.

When they roll out their polling, they are in essence filling up statistically determined quotas for their sub-group categories. If you already have enough black women over 75, then you simply don't need to include more of them.

This is kinda abstract for a lot of people who think polls are just a "big vote from a big pool."

This does not mean that nobody over 75 gets polled, it means that their sampling methodology has a specific number of slots for 75+ registered voters.

it is probably offset by young first-time voters being correspondingly undersampled.

THREE calls in TWO weeks????? Geez, I've only gotten one call from a pollsters IN MY ENTIRE LIFE. There is no shortage of old people in Florida so why would they call the same person THREE times.

The ONLY call I ever got...they identified themselves as a polling company, they asked if I had about FIVE minutes.....sure, I was happy to do it. The first question was about abortion, I answered it and they said...thank you, good bye. So, oddly it only took 30 seconds!

problem.

Hey, let's pretend seniors are being dropped !! Let's circulate a story from an actual senior complaining about being hung up on by a pollster. Yeah. That oughta work.

Sounds like a Steve Schmidt Special.

So your concern is dutifully noted. ;)