Here are some other reasons that a sample may not be representative of a population.

• Underrepresentation: Underrepresentation occurs when some parts of the population are left out of the process of choosing the sample.
• Lack of response: This occurs when people who are contacted for a survey can't be reached or refuse to respond.
• Misrepresentation: This occurs when survey participants give dishonest replies. For instance, many people will say they voted in the last election, even though they didn't.

Gallup public opinion polls have a reputation as reliable and objective because of the high number of people polled from a random sample. This organization’s polls are often referenced in the media. You can visit Gallup.com to see current polls on topics in politics, the economy, well-being, and the world.

Sample answer:

Here are some other reasons that a poll may go wrong.

• A significant number of people being polled may intentionally lie to the pollster.
• An Internet survey will not reach anyone who does not use a computer. For instance, an online survey that asks "Have you ever used the Internet?" should have a 100% response rate of "yes," but it is not true that 100% of the population has used the Internet.
• If a poll is being done by mail-in survey, a subpopulation that feels more strongly about an issue may be more likely to respond and may overrepresent itself. For instance, consider a local election for mayor in which a fervent grassroots minority responds strongly to a mail-in poll from the newspaper asking voters which candidate they will support. Even though the poll shows that the grassroots candidate will win by 58%, he only gets 40% of the vote and loses because the other candidate's supporters did not feel as strongly about mailing in their surveys.