Survey Method Shows That A Throw of the Dice Makes People More Honest

In general, the smart criminals among us know that the basic elements of getting away with an illegal act is to deny personal involvement. The old, “it wasn’t me” defense. But there are many circumstances where researchers and policy makers need accurate data about criminal acts in order to better understand how to prevent them from happening, in addition to understanding their magnitude.

One such instance is the illegal killing of leopards in South Africa. So, how can you get people to own up to the crimes they have committed? Researchers from Bangor University in the UK have a solution: Get the suspected criminal to throw dice. Led by Freya St John and Julia Jones, researchers have tested a randomized response technique based on throwing dice to get more accurate information from the public about the number of illegal leopard killings that take place in South Africa.

A randomized response technique is a survey method that makes it easier for people to answer questions that would typically cost them in some way — such as embarrassment, a fine, or jail time. For the University of Bangor study, a researcher asked respondents a question like, have you killed a leopard in the last year. The respondents would then throw dice before answering. The dice were hidden from the researcher so they could not see the number.

The rules of the survey have a built in safety for the respondent. It goes something like this: The respondent will be instructed to always answer yes if they throw a six or no if they throw a one, but all other times should tell the truth. The researcher never knows the number so they aren’t 100% sure if the respondent is lying or telling the truth.

So how exactly is this a better way of surveying people? If there is a built-in mechanism by which respondents are supposed to lie, that can’t be good for getting accurate information. But actually, it is. Other randomize response technique studies have shown that with the safety-net of the researcher never knowing if the respondent was lying, respondents are more inclined to tell the truth.

For the leopards, this is both good and bad news. The good news is that researchers now have more accurate information about the rate at which leopards are being killed in South Africa. The bad news is that this rate is extremely high. The answers produced by the dice-throwing survey suggest that 19% of respondents had killed a leopard in the last year, despite the practice being illegal.

The researchers also asked questions about the respondents’ attitudes toward the animals and how they felt their neighbors perceived the animals. They were able to correlate the negative view of the animals with the belief, correct or not, that respondents’ neighbors were killing leopards, and with the answer that the respondent had killed a leopard.

While it seems obvious that not liking leopards and killing leopards go hand in hand, establishing the relationship is actually an important step in designing effective conservation plans. According to the researchers, the information generated from this study could be used to help establish a compensation plan for ranchers who lose livestock to large carnivores like leopards. Targeting this plan to people most likely to kill leopards could make it more effective and improve the general attitude toward the animals.

The researchers say that as long as the information is not used to criminalize individuals the information generated is likely to continue to be very truthful. In the long run, this could be extremely beneficial for establishing an open dialogue about how to interact with wildlife going between researchers and the public most likely to be affected. The research was published in the Proceedings of the Royal Society B.

(via New Scientist, photo via Wikimedia Commons)