Revisiting fraud control in light of new research
written by Jackie Lorch
Dan Ariely’s new research on why we cheat has a lot of very relevant findings for the research industry, given the importance of honesty in our research participants. If we believe Ariely’s research, our industry may be investing too much time and money trying to catch the Bernie Madoff or Enron-size cheats, who are very rare and difficult to outwit, while there is much more we could be doing to minimize the more-common type of minor cheating.
In a recent Wall Street Journal article, Ariely says that everyone has the capacity to be dishonest, and almost everyone cheats—at least a little – if the conditions are right. He comes to this conclusion after a number of experiments which asked people to solve simple math problems, with a small reward offered each time. When allowed to add up their own correct answers and feed their papers into a shredder before being paid, they claimed to have solved just a couple more problems than the people who had to hand in their papers so couldn’t cheat. Many people cheated a little; very few people cheated a lot. Interestingly, the cheating behavior didn’t change whether people were being offered 50c or $10 for their efforts. In fact, cheating was slightly less frequent at the higher payment levels.
Factors that increased cheating included:
- Making the prospect of a monetary payoff more “distant,” in psychological terms. When students were tempted to cheat for tokens (which could immediately be traded in for cash) subjects cheated twice as much as those lying for direct cash
- Thinking that someone else was doing it. When an actor pretend to have solved all the problems in an unbelievably short time, the rate of cheating among others rose dramatically
- Being tired from the demands of a mentally difficult task
Is there a better way to keep people honest?
Ariely concludes that putting fines and other punishments in place to prevent cheating are an ineffective defence against “I’m only cheating a bit” and “everyone does it” thinking. In a later test, his team asked one group to think about the 10 Commandments before solving the math problems and the other to think about 10 books they read in High School. There was no cheating in the first group. This even worked when the test was re-run among people who said they were atheists! Reminders of morality at the point of decision seem to have a big impact on behavior.
What can researchers learn from these findings?
- Don’t underestimate the value of reminders at the point of decision. Wording about the need for good research data in making good decisions for society at the start of a survey and during a survey may work better than threats and legalese
- Keep reminding people of the value of honesty – not just the first time they join a panel
- Since cheating increases if people think others are doing it, find better ways to convince people that they are part of a community of trust
- Mental fatigue increases cheating. Another good reason to keep surveys short
About SSI KWB
- RT @CASRO: Hot off the Press: #CASRO Annual Journal is now available. Research on research, management tips, & much more http://t.co/yQiEd9… 27 minutes ago
- Thxs for joining us in preventing fraud, the presentation & video will be posted shortly. Visit us soon! bit.ly/XW4HXy #SSIwebinar 5 days ago
- For more info on respondent fraud or others, email us at: firstname.lastname@example.org #SSIwebinar 5 days ago