Google

A study on alcohol consumption by Lahaut and colleagues illustrates the danger of non-response bias. After receiving responses from only about half of their mail survey sample, the researchers conducted unannounced home visits to the non-responders. They discovered that these original non-responders included a high proportion of abstainers. If the researchers had relied solely on the original responders, their estimates of alcohol consumption would have been artificially inflated.

Alcohol Consumption Survey Example

In a study on alcohol consumption that used a mail survey, researchers followed up with non-responders through unannounced home visits and discovered that these non-responders included a disproportionately high number of people who abstained from alcohol.

Learn Before

Related