Polls as manipulation

POLL: Ernst Breaks Away – http://huff.to/1sYQHLv

How often do the media announce that a poll greatly at variance with the prior polls is probably wrong?

The pollsters give us a range that’s supposed to be right 19 times out of 20.

The polls tend to reflect that, they’ll mostly be roughly consistent but maybe 1 in 20 will be anomolous.

But the 1 in 20 that are probably wrong are far more influential than the other 19 combined, because of articles like the linked one.

At the time they are released, they are probably false. 

But then they can become a self fulfilling prophesy if they are employed to create a bandwagon effect, so the polls create the outcome that they purport to be studying.

That is intentional.

There was a better HP article a few days ago about how polls are spun by right wing media, including organizations that used to be reputable.

The article was about one national paper that has been bought out by right wing interests on one recent occasion described a 2% drop in Obama’s approval as “plummeting”, then for the next month’s poll which showed Obama 3% up was described by the same paper as “statistically insignificant”.

Because polls only seem to count when they favor the right wing, a favorable poll for a democrat that is out of range may be used against him- if the next month’s poll is back within the established range, it would probably be depicted as the campaign falling apart.

The irony is the only time polls are really useful information is when one side is consistently burying the other, outside the margins of error.

When a 3.5 margin is a 7% range, and with the methodological problems with polling now that fewer people have land lines, a poll doesn’t mean much in a tight race- except that it can be used to manipulate the outcome, especially when used like this one.

Another thing, the poll is the property of the person who purchased it, who has no obligation to authorize release unless the contract states otherwise.

Anybody can commission a couple of dozen polls in order to produce an outlier, and only release the anomolous one.  That can be done month after month as well. 

So only polls for which there is a public commitment to release in advance, and publishing methodology in advance with an ability to audit for compliance, should be given any weight at all. 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: