In research, the observer bias is a form of detection bias originating at a study’s stage of observing or recording information. Different observers may assess subjective criteria differently, and cognitive biases (including preconceptions and assumptions) can affect how a subject is assessed.

For example, being aware of a subject’s disease status may introduce a bias in how the outcome is assessed. Observer bias can also occur when the subject knows they are being examined (sometimes referred as the Hawthorne effect).

When a subject knows they are being observed, it can cause them to act differently from how they normally would, which could interfere with the experiment.

Another example examines police work, where police officers change their behavior based on who is watching.

Blinded experiments are used to limit observer bias. Observer bias can also be avoided or limited by having researchers work independently of one another.

 

Note:   The above text is excerpted from the Wikipedia article Observer bias, which has been released under the GNU Free Documentation License.