I work part time at UW Survey Center...

[WI_DEM]

which does the Badger Poll and many other surveys. I do data entry and am working on the Midus survey, but last night I heard the callers doing the "Badger Poll" calls and apparently this poll will be released on Friday. The last Badger Poll was an outlier with Kerry 12-pts down in late September here in Wisconsin--this one should be better.

But what struck me is that 90% of the people or more wouldn't participate. And when they do get someone the computer decides if they should speak with the man of the house or the lady of the house based on what they need. So if the man of the house is needed and isn't home, they don't interview the lady of house, but will try back.

It is kind of interesting to hear these college kids doing the polls. Most of them are for Kerry so when they hang up with a Bush supporter they shake their heads in disgust, but on the phones they are very professional.

But I find it interesting how many people will not be polled, so actually it is a small pool of people that pollsters actually do speak with.

[Was_Immer]

1. tell them to skew the polls Kerrys way so we

can get the REPUKES on the cable news to shut the hell up about the polls

[flowomo]

2. that's a very bad idea

[lakeguy]

3. it'd be interesting to see what percentage

of people that refurse the poll are dems...i know a few people that have refused polls during this election.

[MsUnderstood]

4. I think the process is at fault

Like when you call up and say "Hello can I speak to Mr. or Mrs. Smith" and you happen to be calling a lesbian couple. I automatically assume telemarketer and mess with them.

Or as you pointed out the computer only allows men or women at a certain time. I don't think it would be a problem for the computer program to allow the pollster to select the gender of the caller then be told if the gender is needed (instead of the computer telling the pollster this is the gender you must speak to before the call begins).

Computer programs are only as good as the programmers. The programmers need to make it easier to take the polls.

[chelsea0011]

5. I did polls for Carter( although they never said who I was working for)

in 1980. You are right about how few people actually will let you poll them. On a shift I would easily call 300 numbers. Most would not get an answer and out of all the numbers I got connected with someone, I would be lucky if I got 10 polls done a night.I was part of the poll done the night before the election when the undecideds all went for Reagan and Carter knew that night he couldn't win. It was quite stunning to see how many went over to Reagan that night after months of a high, steady undecided for months.

And how did I find out I was working for Carter? One night I did a poll with one question asked as "Do you think Reagan is more or less likely to start a nuclear war?" No kidding. That was the question.

[pelagius]

6. I screen my calls with Caller ID...

...and don't pick up "Private Caller" or "Out of Area" calls. Eight times out of ten it's an unsolicited call (telemarketers usually). Friends leave messages. Most of them actually call me on my cell, actually.

I think, due to changes in the technology and its use, traditional polling methods are becoming obsolete.