Social desirability refers to the tendency for people to answer survey questions in the way that they feel is most socially appropriate and is not based on how they actually behave.
For example if we were to ask a random sample of teenagers how often they smoke marijuna-some teens will underexaggerate out of fear they will be punished, and some might overexaggerate to appear cooler to their peers. The fact that few would be answering honestly would be an example of social desirability.
How much of an impact do you feel social desirabilty might have on the reliability and validity of research into criminal justice? Do you feel this impact is greatest for online, telephonic, or face to face survey techniques?