This is a very interesting video on intuition, experimentation, patient relations, and cheating.

He decided to conduct a series of experiments to understand cheating. He gave test subjects a math quiz with 20 problems, and promised to give a dollar for each correct answer. The problems weren't hard to solve, but Ariely imposed a five-minute time limit, making it impossible for anyone to complete the test. After five minutes, Ariely collected the test from the volunteers, scored them, and paid them for their correct answers. On average, volunters solved four questions correctly.

Next, he tempted people to cheat. He told a new group of test takers to score their own tests and tell Ariely how many questions they got correct. These volunteers reported, on average, that they solved seven questions. The interesting thing about this, says Ariely, was that the higher average wasn't because a  few people cheated a lot; rather, it was because a lot of people cheated a little. Equally interesting was the fact that the amount of cheating didn't change when the reward for a correct question was increased or decreased; nor did it change when the chances of being caught cheating were increased or decreased.