Poll Nuance and False Assumptions

[SoCalDem]

Every day we hear how this poll says "X", and that poll says "Y".

Polls can (and do) deliver just about any result the poll-takers intend for it to deliver.

A few examples:

If a pollster asks:

"Do you think that lawbreakers should be harshly dealt with?"...most people would say "yes"

BUT

which law?
exactly what is meant by "harsh"?
should judges be forced to adhere to strict guidelines..with no exceptions?

Let's take a look at looting..

Looting is a crime, BUT in cases of dire emergency, like Katrina or any major disaster, looting WILL happen, and poor people are most likely to be the ones who will be caught & punished.

why?

They are the ones without funds to get AWAY from whatever the disaster may be, and they are also the ones with a "funds-deficit", that makes it all but impossible for them to get the things they need in order to survive, after the fact.

Stores will probably be closed-for-business in the aftermath, but the merchandise may still be there. In small businesses in poor areas, the proprietors may also be victim to the disaster, so they may be unable to get to their stores, even if they had planned to be "open".

Insurers surely don't like to pay off for damaged/lost merchandise, but the business owners probably have been paying for insurance, so the merchandise would probably have been replaced/reimbursed anyway, because it was damaged.. (looted or not)


......

Later in the poll, the pollster may ask if being an undocumented worker is "illegal" and deserving of punishment.

On its face, that too is a valid question, but the answer requires nuance.

A person brought here as a 2 yr old, may (as a 16 yr old) not even know their parents were not "legal". Should that 16 yr old be "deported" to a country they may have never even been to? Should that 16 yr old be deprived of college, and doomed to a shadowy life in low wage jobs because his/her parents never legalized their status?


Polls usually offer a variety of similar, but ambiguous choices:

agree
somewhat agree
disagree
somewhat disagree
no opinion

depending on the "reporter", a poll with those choices is skewed immediately, since the agree will almost always be offset by FOUR negatives, since all but agree will be lumped together.

Now, if we accept the fact that republicans are probably less than 30% of the population, even if 75% of them are "for" something or "against" something, it's still only 75% of less than 30%.

Republicans are highly susceptible to codewords, and pollsters know this, so their polls are "liberally" salted with those hot-words.

Pollsters also know which regions will yield more republicans (or dems), so when they report the polling sample, that may also be skewed. example:

"polled 1000 southern californians" ...

Orange County?
LA County?

each place would probably yield very different replies to questions.

Bottom line:

Polls would be better off kept for internal review, but the media loves them, so we will always be hearing about this poll or that poll, but in the end, they are all suspect.

[madokie]

That how I see it too

rec

[OHdem10]

1. Do you think the country is going in the right direction.? Democrats

could answer this "NO" and mean something entirely different
from a Republican who would answer "No" to the same question.

Yet when the Pundits report it--breathlessly, the country
is going in the wrong direction. They imply all persons
agree with the GOP version.

[madokie]

2. The problem in America right now is the main stream media

who for the most part are owned by the rightwingers

[Scuba]

3. Rupert Murdoch defines truth in America.

[madokie]

4. so true

The military industrial complex picks up any slack