The sampling method described suffers from a form of bias. Identify the bias, tell how the sample statistic might differ significantly from the population parameter, and suggest how the problem might be corrected.A city supervisor would like to assess the level of tax payer support for improved services for the homeless. He uses the phone book to obtain a random sample of residents and finds that only 26% think that services for the homeless should be improved.

Question

What will be an ideal response?

Answer 1

There is a certain amount of undercoverage bias. The sample will exclude those who don't have a landline phone. In particular, the homeless will be excluded as well as many young people who tend to use cell phones rather than landlines. A more accurate estimate could be obtained by choosing a random sample from a list of all taxpayers in the city.