That risk
is very much reduced if data is collected by self response, confidential questionnaires, for example. My concern about truthfulness would be less than my concerns about sampling or definitions.
( Dekazer, Mon 4 Jul 2005, 10:35, archived)

In most cases
I agree with you. But not where teenagers are concerned. doesn't matter that it is confidential self response, they'll lie, especially the boys. It's one of the founding principles of the universe...
( the mighty badger Aphrodite, on a bar stool, by your side, Mon 4 Jul 2005, 10:37, archived)

I really disagree
I work on a youth crime study, and we've triangulated results sufficiently to let us conclude that our lot are enormously truthful.

It all depends on your methods - not on the truthfulness or otherwise of young people.
( Dekazer, Mon 4 Jul 2005, 10:39, archived)

Tie them to a chair, and give them the rubber hose treatment, until they confess?
Or just interpret the data in such a way that "I carry an AK-47, and lead the East Cheam Massive on regular crime sprees" becomes "I stay at home and play far too much X-box"?

Either way, I salute your dedication to the truth!
( monster munch person, man, woman... camera... TV?, Mon 4 Jul 2005, 10:48, archived)

well, this is your area,
so I'll bow to you on this one. :P
( the mighty badger Aphrodite, on a bar stool, by your side, Mon 4 Jul 2005, 10:57, archived)