Day 070 - Cheating

Submitted by Sam on 29 July, 2011 - 18:02

From an economic perspective, cheating is a simple cost-benefit analysis, where the probability of being caught and the severity of punishment must be weighed against how much stands to be gained from cheating. Behavioural economist Dan Ariely has conducted experimental studies to test whether there are predictable thresholds for this balance, and how they can be influenced.

In one study, Ariely gave participants twenty maths problems with only five minutes to solve them. At the end of the time period, Ariely paid each participant one dollar for each correctly answered question; on average people solved four questions and so received four dollars. Ariely tempted some members of the study to cheat, by asking them to shred their paper, keep the pieces and tell him how many questions they answered correctly. Now the average number of questions solved went up to seven; and it wasn't because a few people cheated a lot, but rather that everyone cheated a little.

Hypothesizing that we each have a “personal fudge factor”, a point at which we can still feel good about ourselves despite having cheated, Ariely ran another experiment to examine how malleable this standard was. Before tempting participants to cheat, Ariely asked them to recall either ten books they read at school or to recall The Ten Commandments. Those who had tried to recall the Commandments – and nobody in the sample managed to get them all – did not cheat at all when given the opportunity, even those that could hardly remember any of the Commandments. When self-declared atheists were asked to swear on the Bible before being tempted to cheat in the task, they did not cheat at all. Cheating was also completely eradicated by asking students to sign a statement to the effect that they understood that the survey falls under the “MIT Honor Code”, despite MIT having no such code.

In an additional variant of the same experiment, Ariely tried to increase the fudge-factor and to encourage cheating. A third of particpants were told to hand back their results paper to the experimenters, a third were told to shred it and ask for X number of dollars for X completed questions, and a third were told to shred their results and ask for X tokens. For this last group, tokens were handed out, and the participants would walk a few paces to the side and exchange their tokens for dollars. This short disconnect between cash and token encouraged cheating rates to double in this last group.

Putting these results in a social context, Ariely ran yet another variant of the experiment, to see how people would react when they saw examples of other people cheating in their group. Subjects were given envelopes filled with money, and at the end of the experiment they were told to pay back money for the questions that they did not complete. An actor was planted in the group, without the knowledge of the other participants. After thirty seconds the actor stood up and announced that he had finished all of the questions. He was told that the experiment was completed for him, and that he could go home (i.e. keeping the contents of the envelope). Depending on whether he was wearing a shirt identifying him as from the same university as the rest of the students in the test or not, cheating went either up or down respectively. Carnegie Mellon students would cheat more if he was identified as a Carnegie Mellon student, whilst cheating would decrease if he was identified by a University of Pittsburgh shirt.

Ariely's results show that the probability of getting caught doesn't influence the rate of cheating so much as the norms for cheating influence behaviour: if people in your own group cheat, you are more likely to cheat as well. If a person from outside of your group cheats, the personal fudge factor increases, and the likelihood of cheating drops, just as it did with the Ten Commandments experiment, reminding people of their own morality.

The stock market combines a worrying cocktail of features from these experiments. It deals with 'tokens', stocks and derivatives and not 'real' money. Stocks are many steps removed from real money, and for long portions of time. This encourages cheating. Any enclaves of cheating will be reinforced by people mirroring the behaviours of those around them, and this is precisely what happened in the Enron scandal.

Attribution Noncommercial Share Alike
This text, Day 070 - Cheating, by Sam Haskell is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike license.
Drupal theme by Kiwi Themes.