Sampling Bias

Remember that the ideal sample should be representative of the entire population.

sampling bias

In statistics, a sampling bias is created when a sample is collected from a population and some members of the population are not as likely to be chosen as others (remember, each member of the population should have an equally likely chance of being chosen). When a sampling bias happens, there may be incorrect conclusions drawn about the population that is being studied.

Main Sources of Bias

Here are the four main sources of bias to consider when sampling from a population:

• Undercoverage occurs when some groups of the population are left out of the sampling process, and the individuals in these groups do not have an equal chance of being selected for the sample. For example, a sample survey of households in a country may miss people who are homeless, prison inmates, or students living in dorms.
• Non-response bias occurs when an individual chosen for a sample cannot be contacted or decides not to participate in the study or research. This type of bias occurs after the sample has been selected and can create potential bias in the data collected.
• Response bias is defined as a systemic pattern of inaccurate responses to questions. This type of bias can occur when a person does not understand a question or feels influenced to respond to a question in a certain way. Response bias can also occur as a result of the wording of questions that are of a sensitive nature.
• A voluntary response bias is another form of bias because the sample is not random or representative of the population. The people who volunteer for a study or survey may be more inclined to respond to questions or report certain behaviors.

When we say that a random sample represents the population well, we mean that it doesn’t have inherent bias. Inherent bias means there isn’t anything about the method itself that unfairly favors some groups over others. Every member of the population has the same chance of being selected.

But remember: random doesn’t mean perfect. Different random samples can give different results, just by chance.

• One random sample might be very close to the population proportion.

• Another random sample might be “off” in either direction.

And in practice, we usually only get to look at one sample. That means we never know exactly how much it differs from the population.

Whenever we take a sample from a population, there is the potential of introducing sampling bias. It is important to be aware of potential sources of bias and take steps to minimize the chance that sampling bias is present in the way that we sample.