Types of Potential Bias in Testing

Intrinsic test bias refers to the inherent characteristics of a test that may favor or disadvantage certain individuals or groups. For example, if a test contains questions that rely heavily on cultural knowledge that is more prevalent among a specific group, it can lead to bias.

Item bias occurs when specific test items consistently produce different results among different groups, even though the groups being tested have the same underlying trait. This can indicate that the item is biased towards or against certain groups.

Method bias refers to systematic errors that occur in the administration or scoring of a test, leading to biased results. For example, if the instructions given to different participants vary in clarity, it can lead to differences in their performance.

Construct bias occurs when the latent trait being measured by a test is not equally relevant or appropriately captured for different groups. This can result in different interpretations or predictions for different groups, even when they have the same test scores.

Extrinsic bias refers to factors external to the test itself that may influence test results. For example, if a test is administered in an environment that is biased against certain individuals or groups, it can impact their performance.