Plus, the polls are done on landlines, not cell phones.
So whatcha basically have is a bunch of pollsters calling people in the daytime on landline phones who are hanging out watching Oprah or “The Young and the Restless.”
Dandy sample there!
Ya know, that’s an aspect of polling not talked about and may be a prime reason for the inaccuracies of elections polls.
I think the pollsters intentionally call in the evening hours when people are more likely to be home.
Your point that the polls are garbage, because the samples are garbage is spot on, however.
Reasons for this could include but are not limited to:
people who screen their phone calls and don't bother answering the "number not available" calls
people who have gotten jaded by being endlessly push polled and don't believe that they are being asked questions by a real pollster
people who still believe in a secret ballot and don't wish to share their feelings
people who work in the evening
people who don't have a phone
people who don't have high speed internet and still use dial-up (basically, all the bitter hillbillies and rednecks)
Additionally, a random sample is only valid if every member of the poplulation has an equal chance of being included. People who are married usually share one landline. Therefore, single people (more likely to vote Obama) will have twice the chance of being called as married people (more likely to vote McCain). People who live with parents or adult children will have even less chance of being called.
For all of these reasons and a whole lot more, the pollsters have to heavily massage the numbers to have any chance of making an accurate prediction. There is nothing inherently wrong with their methodology. The polls only enter the realm of evil when they pretend to be "scientific" because they use "statistical methodology". The reporting of margins of error in these polls is an example of the pollsters joining the dark side. Including the margin of error makes the poll sound more like it is done using valid statistical methods when this is completely not the case.
Also, if cellphones were included you would open up a whole new Pandora's box. Single people with a cell phone for work, a cell phone for personal use and a landline would then have 6 times the chance of being included in the sample as a married person with no cell phone.
nor does it impress me that more rats are sampled....there are more rats out there than pubs....
no...I simply believe that the polls are wrong....at my house, we get calls for surveys, not necessarily political ones, but we never do them...
I don't think family people or conservatives care much about wasting time with surveys....
plus, I maintain, that if these samples are random, then the samples represent MORE people from NYS and CAli and ILL than smaller states....there's simply more names to pull from in NYS and Cali.....
polls be damned.....
MCCAIN AND SARAH PALIN ARE WINNING THIS WHOLE THING!