www.wallettest.com (http://www.wallettest.com)

Well, one could question the scientific validity of the study, but interesting nonetheless.

Well, one could question the scientific validity of the study
Sample size was far, far too small.

They definitely needed to widen the range of time, and use a lot more people in different areas also... In some other parts of town the outcome may have been very different.

Someone tell me the outcome, the voice overs were too irritating.

There is a link that'll take you to the result page.

The result from the 100 people they tested yielded this:

Females are more honest.
Old people are more honest.
White people are more honest than black people.
Overall 1 out of 4 people was not honest.

Doesn't look like a very rigorous experiment to me. But there is nothing really shockingly unusual with this particular result set either.