what factors can make a poll inaccurate?

Answer 1

who cares about polls just look at the last two presidential elections, you think they were accurate?

Answer 2

As well as being inaccurate, polls can also be misleading. It depends on how you ask the question.

If a pollster asks you: "Do you think President Bush has been honest with the American people?" the percentages saying yes may be different than if the pollster asks: "Some say the lack of evidence of WMDs in Iraq or a connection between Saddam Hussein and Al Queda are signs that President Bush lied about his reasons for going to war. Do you agree?"

One is open ended and the other is leading. The same thing happens with taste tests. I once took a public blind taste test between Pepsi and Coke. Though the two glasses of soda were unmarked, there was an obvious difference in them: one was cold and bubbly and the other was warm and flat. It was such obvious manipulation that I chose the warm, flat one just to mess with their results.

Pollsters are often engaging in this type of manipulative marketing. If you precede the questions with information that could prejudice the result, you have a demonstrable effect on the outcome.

It's all just marketing then.

Answer 3

A poorly worded question (e.g. "Have you stopped
beating your wife today": Having no wife, the correct
answer is "no")

Poorly designed media (consider Florida's hanging chads).

Performing the poll in some way that would inherently
skew the results. For instance, if you cold-call people
between 9am to 4pm, your answers will be skewed against
people who work normal hours.

Some how informing the pollee what you want the answer to
be. Showing up at a door to poll with a huge "Vote for XXX"
sign isn't going to get you an accurate count of who is going
to vote for XXX.

Polling immediately after some event that might briefly
change the poll base's opinion. For instance, asking if
all Arabs should be shot immediately after 9/11 is bound
to give you a pretty skewed result of America's actual
prolonged opinion.

Answer 4

If the sample size is too small. For example, only taking a poll of 100 people to represent a country.

If the sample is taken of specific people or places, instead of a general population. For example, taking an on-line poll. Not everyone has access to internet, so it is not truly representative.

If the questions are skewed. For example, asking "How is the president handling the disaster in Iraq" instead of "How is the president handling Iraq".

Answer 5

The major problem is Poll size. It takes at least 1,500 pollees to make up a representative-enough sample of the relevant population..

So when you read about a "poll "of 150 people.,,
Ignore it.

Of course the other factor is Honesty. People answer polls, based more on Wishful thinking, than facts based on rigorous self-analysis.They don't MEAN to lie:-- it's just that they don't even know HOW to tell the truth.

Answer 6

The demographics of the people polled. i.e. A poll taken in Orange County, California will most likely result in different answers/conclusions than one taken in San Francisco, California

In order for a poll to be accurate it should be taken across a wide cross section of the population

Answer 7

1. Who is asked (is it really a diverse, unbiased sample?)

2. How the questions is phrased (is it biased)

3. Who is doing the asking (is it a person/group that those answering the questions have an opinion about?)

4. How are the results compiled?

Answer 8

Using a sample that's not representative of the population. Phrasing questions in a biased way. Not accounting for statistical error properly.

Answer 9

Water!

Answer 10

People who aren't aware of all the information about an issue, or if not enough people are asked to vote.