Marketing
Saved & recent

Sampling bias

Sampling bias occurs when the way participants are selected systematically favors some members of the intended population, distorting the conclusions. Who responds can introduce additional nonresponse bias.

A retailer studying all customers by inviting only loyalty-app users may miss occasional shoppers. Increasing that same sample’s size does not by itself repair the coverage problem; random sampling error is a different source of uncertainty.

Read first
Survey research
Related
Quantitative research
Further reading
Pew Research Center — What Low Response Rates Mean for Telephone Surveys
8 app users 12 nonusers app-only invitation sample of 6 outside the invitation 20 customers total
Hypothetical population: 8 app users and 12 nonusers. Six of the eight eligible app users are sampled. The 12 nonusers have no route into the sample, regardless of sample size.

View in the full Marketing collection