Question

Suppose a telephone poll is conducted by contacting U.S. citizens via landlines about their view of gay marriage. Suppose over 50% of those called do not support gay marriage. Does that mean that you can say over 50% of all people in the U.S. do not support gay marriage? Explain

Answer 1

Answer:

No, because it's not a representative sample of the people of the US.

Explanation:

In statistical analysis, a representative sample is a subset of a population that matches the characteristics of the whole population. A survey instrument (like a poll) used to collect information from a representative sample allows us to extrapolate the results of the survey to the whole population. However, if the sample is non-representative, this can't be done. In this case, the telephone poll is not representative of the whole population of the US, because this poll was conducted by contacting only people who have landline phones, thereby excluding people who only have mobile phones and people who have no phone at all. Therefore, no matter the results of the poll, you can't say that over 50% of all people in the U.S. do not support gay marriage.