Polls – Are They Useful?
Saturday, August 1st, 2009
The answer to the question depends a lot on who you ask. There is no doubt that, for some people, polls are invaluable. These people are ususaully part of one of these groups: Politicians, reporters, political activists, and those who can’t think for themselves.
However, as I see it, for the average, thinking, reasonably intelligent American voter, polls are actually a detriment. Yes, I believe that polls are a big part of the problems we’re seeing in this country at the present time.
Polls are not hard numbers, they are probabilities, yet some folks (see the first sentence of this post) throw them around as evidence to further whichever agenda they’re pushing.
EVERY SINGLE POLL IS WRONG.
That is a factual statement. Easily proven:
- Every poll is published with a “margin of ERROR”
-When the same question is asked in multiple polls, the results are virtually never the same (just take a look at the Presidential Approval Poll. There is, as of today, 8 points difference between the Rasmussen poll and the NY times poll). If polls were accurate, every poll would have the same results (duh)
-The results are published in terms of the entire country, when in fact, most polls are based on an extremely small sampling of people. Based on the list of polls mentioned above, FOX news asked 900 people. NPR asked 850 people. The largest sampling was done by Gallup (1547 people). Most others are at about 1000.
-After-the-fact results are often very different than what polls predicted. Even exit polls have proven to be inaccurate.
POLLS ARE MEANT TO MISLEAD
As mentioned above, polls are usually based on sampling of around 1000 people. Yet those who find polls to be useful tools for pushing their agenda/party/cause will often state poll results in a misleading fashion.
Let’s assume a sampling of 1000 people, and 50 percent of the respondents answer “Yes” to the polling question. (Lets say the question is “Do you think religion should play a part in government?”)
These are things you are likely to hear from talking heads on cable news:
“50 percent of Americans think religion is important”
“Half of America says religion should not play a part in government”
Neither of those statements is the truth, yet some people who hear those statements will repeat them as if they were absolute, factual statements.
Every one of these statements is more true, but less likely to be used, than those above:
“500 people think religion should be part of government”
“Half of the 1000 people surveyed don’t want religion in government”
THE NUMBERS ARE SKEWED
Of course, they’ll never mention that the results are based on only those that were willing to be polled. That thousands of people they called actually refused to answer the questions. So the numbers are skewed by that as well. And there are those who will lie, or don’t know anything about the question, but are more than willing to give an opinion.
THINK INTELLIGENTLY
Next time you hear or read someone reporting on polls, please remember that they are best guesses, based on a sampling of about 1000 people who were willing to answer the poll.
Comments