Sampling Bias and Its Types

Overview

This lesson reviews common sources of bias in population sampling and explains how different biases can invalidate a study's results.

Sampling bias occurs when the sample is not representative of the population.
Voluntary response bias happens when only volunteers are included, often leading to bias.
Self-interest study bias arises when researchers have a vested interest in study outcomes.
Response bias is when survey responders provide inaccurate answers for any reason.
Perceived lack of anonymity can cause people to hide the truth from fear of negative consequences.
Loaded (leading) questions have wording that influences or guides responses.
Non-response bias occurs when those who refuse to participate skew the study results.

A pesticide company conducting its own study is an example of self-interest study bias.
Asking “Don’t you think 18-year-olds should be able to drink?” is a loaded question and creates response bias.
Questions like “When was the last time you ate pizza?” may result in response bias if respondents don’t remember accurately.
Sensitive questions (“Have you ever cheated on a test?”) may yield inaccurate replies, causing response bias.
If many respondents skip a question, like “When is the last time you read a book?” the results reflect non-response bias.
Work surveys about sensitive topics can be biased by perceived lack of anonymity.
Question order can also influence survey outcomes.

Sampling bias — when a sample does not accurately represent the population.
Voluntary response bias — bias caused by only including people who volunteer to participate.
Self-interest study — bias from researchers with a personal stake in outcomes.
Response bias — inaccuracies in answers given by respondents for any reason.
Perceived lack of anonymity — fear of being identified influences honest answers.
Loaded question — a question designed to influence the respondent’s answer.
Non-response bias — bias from people refusing to answer or participate.