20140415-Abu-Dhabi-Life-004
A group of men socialize in Abu Dhabi. Phillip Cheung / NYUAD

Designing meaningful experiments is a familiar challenge for scientists. In the fast-growing field of experimental social science, however, researchers may encounter difficulties quite different from those facing white-coated laboratory investigators.

As a member of NYU Abu Dhabi's Social Science Experimental Laboratory (SSEL), Jonathan Rogers has been seeking a way around one such constraint. His approach may prove useful for scholars throughout the Middle East and North Africa (MENA), and beyond.

Rogers, political science instructor, explained that in social science, "experiments let us study human behavior by stripping interactions down to their basic components."

For example, classic economic theory suggests that if you have $100 and are free to give some of it away, you'll give nothing. But when experimental economists ask test subjects to play this "dictator game", many do donate some money. This suggests that there can be a self-serving reason to donate; in the jargon, generosity gives you a "warm glow".

Often these games involve risk and reward. In the Bomb Risk Elicitation Task (BRET) you face a grid of 100 boxes; each one you open pays $1, but one contains a "bomb" – if you open that one, you lose everything and the game ends. How many boxes would you open before you quit?

Other experiments are less simple. "If we want to study cooperation," Rogers said, "we can put subjects in a situation where they only have the choice to either cooperate or not. Then we can change one aspect of the game and if we see different behavior, we can infer that the change caused that difference."

jonathan-rogers0810
Jonathan Rogers, NYUAD political science instructor

Many games test risk tolerance: are people more likely to enter a lottery if the chance of winning is high but the prize is small, or vice versa?

Such "risk elicitation mechanisms" are often "incentivized" – that is, test subjects can keep the small sums they may win. In other words, some games often involve gambling; others involve interest payments. This creates a problem for researchers in the MENA region: gambling and interest are both prohibited for Muslims.

In a paper now under review for publication Rogers tests a potential way around this difficulty: What if the reward goes to charity?

In 2014-15, 40 NYUAD student volunteers played the BRET game twice each, while 69 others played BRET and another game once each. Each student kept the winnings from one game; proceeds of the other were given by pre-arrangement to Operation Smile, a charity subsidizing surgery for children with facial deformities. Winnings averaged about AED 57, or $15.50 USD.

The results were clear: "Subject behavior under the two conditions is almost identical" Rogers's paper reports. That's good news, because it appears to mean that research done on this basis, among Islamic populations, can give useful results. "We still need to repeat this experiment with other subject pools, in different countries, and with different types of charities," Rogers said. But this is a first step.

20151114-Abu-Dhabi-Skyline-072
Jonathan Rogers is working in Abu Dhabi to come up with ways to study social behaviors that are respectful of religious beliefs. Phillip Cheung / NYUAD
quote

Academics want to study refugees, revolutionaries, people in developing countries, and attitudes toward extremist groups, but some of the tools that are considered standard in the West can come into conflict with cultures and norms in the places where these people live.

Jonathan Rogers, instructor of political science

SSEL members, Rogers said, hope to "attract researchers who want to conduct studies in MENA countries … Academics want to study refugees, revolutionaries, people in developing countries, and attitudes toward extremist groups, but some of the tools that are considered standard in the West can come into conflict with cultures and norms in the places where these people live. If researchers break local customs, people may refuse to participate."

"(But) If charitable incentives become accepted as an alternative for payments in risk experiments, maybe we can find a similar workaround for experiments about things like interest on investments."

By Brian Kappler for NYUAD Public Affairs