The experimental setting can produce misleading results due to its 'falseness'.
It only applies when testing for aggressive responses.
Isolating one variable for study is limiting.
Subject expectancy can produce misleading results.
It uses a lot of antiquated ideas.
In the experimenter's home
They concentrate on written reports rather than actual behaviour
They are often wildly inaccurate
People are often resistant to participating in survey-based studies
The preventative technique of non-bias
The single blind technique of control
Obfuscation scenario method
Here's an interesting quiz for you.