Glossary » selection bias

selection bias

in choosing test users for a usability study, selection bias is any imperfection in the selection process that gets either the wrong types of users (people who aren’t in your target audience) or a sample of users that is not random and therefore is biased toward one type of user rather than another (e.g. even though you want to target both novice and expert users, you only manage to recruit expert users, so your results will be skewed). The most common example is a volunteer bias, since people who volunteer to be in a study may not be representative of your overall target audience; they may, for instance, have more free time, have a greater interest in technology, and be more outgoing than those who wouldn’t volunteer.