Statistics Collecting Data

researchers depend on sampling

because collecting data from entire population is often not practical

• Choosing a sample that represents the entire population fairly

• Collecting sample data in a way that ensures the data represent the population

To do their jobs well, researchers have two primary tasks:

sampling bias

when researchers do not randomly select the subjects in their sample.

Undercoverage bias

when certain members of the population are excluded, or likely to be excluded, from becoming part of the sample.

Nonresponse bias

when only a small portion of the sample responds and researchers believe that nonresponders would answer the questions differently.

A survey is sent to a random sample of 10,000 customers. Of the 500 surveys returned, 80% of them are under 20 years old. Because so many nonresponders are under age 20, the survey results may not be representative of the entire population.

Self-selection bias

Self-selection bias occurs when people with certain opinions or characteristics are especially likely to respond.