Published January 15, 2013
A new game, loosely modeled after the spread of gonorrhea, offers a new way for researchers to study people's motivations for choosing whether or not to use condoms, get flu shots or otherwise protect themselves from epidemics.
The game may be the first specifically designed for studying the spread of diseases, its developers wrote in a paper published Jan. 9 in the journal PLOS ONE.
With the game, researchers can ethically study people's behaviors during epidemics. In real life, of course, performing certain experiments, such as releasing a disease or making vaccines more expensive for some people than others, would be deplorable and highly unethical.
"With this virtual world approach, we can do all kinds of policy experiments," said Frederick Chen, an economist at Wake Forest University in North Carolina and the lead researcher in the development of the gonorrhea game.
This isn't the first time researchers have tried using game versions of diseases as a proxy for real life. In 2007, some epidemiologists published studies of how "Corrupted Blood," a disease in the massive online game "World of Warcraft," mimicked real-life outbreaks. What's new about Chen's game is that it's a simple game built for doing controlled experiments.
"We wanted to strip the problem down into the most essential components and to analyze incentives," Chen said. "If we think about a multiplayer, online game like 'World of Warcraft,' you have a lot of other things going on."
In the future, games like Chen's could complement surveys that researchers already conduct, which ask what people thought and did during real disease outbreaks, Chen thinks. Together, insights from surveys and games could go toward the design of ads, laws and other campaigns for preventing flu epidemics like the one gripping the United States now.
Chen and his colleagues, including other economists and one computer scientist, recruited 102 people to play their game over 45 days. Every day, the study participants got emails saying whether or not they were sick and what their likelihood was of getting sick the next day. Healthy players could choose either to buy protection for the day or take their chances unprotected. "Sick" players could do nothing for a day. [SEE ALSO: Man Infects Himself with Computer Virus]
That simple disease model is reminiscent of chlamydia or gonorrhea, Chen and his colleagues wrote, because getting one of these illnesses doesn't give people immunity later, the way getting the flu does. The game-disease also doesn't kill people off. The team chose this model so that players could keep getting infected, and keep playing, giving the researchers plenty of data.
Participants earned more points by staying healthy, with the most points going to those who stayed healthy without protection because they didn't have to spend their virtual money. At the end of 45 days, each player's points translated into money on an Amazon gift card. The top players earned about $25, Chen said.
The game offered some insights not available from current mathematical models, another way to virtually and ethically study disease spread, Chen and his colleagues found.
The study of how people act during epidemics is a young field, however, so researchers tend to use simple models, Chen told TechNewsDaily. Such models assume, for example, that people always act rationally if they know all the rules of the game. [SEE ALSO:9 Super-Cool Uses for Supercomputers]
In Chen's game, however, that assumption didn't hold true. People who happened to get sick early in the game were more likely to buy protection later in the game, even though they knew they were just as likely as the non-diseased players to get sick on any given day.
Chen and his colleagues also found that people acted differently over time, depending on the prevalence of the game-disease. High prevalence led to more and more people buying protection as days passed, while low prevalence led to fewer and fewer people buying protection as time went on. The latter scenario reminded the researchers of the reductions in condom use that came with the fall of HIV/AIDS rates after the 1990s.
That means that letting people know their day-to-day risk of getting diseases such as the flu could encourage people to get vaccinated — up to a point. "It's actually difficult to eradicate disease" because of the way people act once they know disease risk is low, Chen said.
Experiments for the future
Chen's results often echoed what other researchers already knew from surveying people after real-life epidemics. That shows a game can be a realistic way to study what would happen in reality, Chen said. At the same time, games can go beyond surveys since they don't have to rely on people's poor memories and can avoid cases in which respondents don't follow through with their stated intentions, he said.
Next, Chen hopes to get funding to study more disease scenarios with games. For example, his team could test different illnesses, such as a flu-like scenario, in which people become immune after one infection, or an HIV/AIDS-like scenario, in which people don't know right away that they're infected.
"There are a lot of different kinds of diseases. There are a lot of different variations that we want to try," he said. "I think this virtual, online epidemic setting is really a great framework."
Copyright 2013 TechNewsDaily, a TechMediaNetwork company. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.