Serious video games may seem like an oxymoron, but for OU’s Norah Dunbar they’re anything but contradictory.
Dunbar, a professor in the department of communications, has spent the last few years working with a team to develop the “serious” games “MACBETH” and “MACBETH 2,” which have nothing to do with a murderous king in Scotland.
MACBETH, which stands for Mitigating Analyst Cognitive Bias by Eliminating Task Heuristics, was designed as a training tool that reduces cognitive biases in the player, Dunbar said.
The game tests for confirmation bias, fundamental attribution error, bias blind spot, anchoring bias, projection bias and representativeness bias, she said.
The game has been successful not only in effectively teaching the player a lesson, but also in making sure the lesson is remembered, she said.
Due to its educational nature, MACBETH is categorized as a serious game vs. an entertainment game, like Grand Theft Auto. Think of it as a really fun sort of homework, Dunbar said.
The game has been successful competitively as well. It won the 2013 Best Serious Game in the business category as well as the 2013 Best Game Special Emphasis Adaptive Force Award, judged by the Office of the Secretary of Defense, at the Serious Games Showcase and Challenge out of over 50 submitted serious games by commercial businesses, Dunbar said.
The game has also helped change biases about biases. There were some who told Dunbar her game wouldn’t work because biases are a part of human nature and therefore impossible to undo.
MACBETH proved them wrong, reducing confirmation bias alone by 21 percent, Dunbar said.
Her team has published two papers already and was just accepted for a presentation at the Persuasive Technology Conference in Italy, Dunbar said.
The first iteration of MACBETH was funded by a $10.7 million grant from the Intelligence Advanced Research Projects Activity with the goal of mitigating biases in intelligence analysts, according to Daily archives.
The team began work on the second version of MACBETH on another contract with the agency. But funding for that project was cut as a part of federal budget sequestration, Dunbar said.
By the time those cuts were made, the team was about 90 percent done with the project, Dunbar said.
Developers completed the project with funding from the OU office of the Vice President for Research as part of their Defense, Security and Intelligence, or DSI, Research Program.
Though words like “defense” and “security” might seem to suggest a narrow field of research in the fields of science and engineering to a covert group of researchers with high security clearances, DSI includes a broad range of projects from energy generation to human behavior, according to the office’s website.
Robert Palmer, associate vice president for research, commented on the variety within fields of research at OU, including Dunbar’s MACBETH.
“She’s doing some really interesting things along the lines of DSI, especially on the intelligence side,” said Palmer.
Losing government funding has one benefit: the rights to market the game in fields other than intelligence, Dunbar said.
Kelvin Droegemeier, vice president for research, said the office of the vice president for research is now looking for further ways to support Dunbar’s work.
Since MACBETH did so well in reducing confirmation bias, Dunbar thinks the training would help doctors avoid biases that overlook symptoms that don’t fit into a patient’s existing diagnosis. It would just take a small tweak in the game script, perhaps putting the player in an epidemic instead of espionage scenario.
The bias training in MACBETH is also related to Dunbar’s work in developing training for deception detection. Its focus is teaching people to use correct cues instead of relying on biases or myths, Dunbar said.
People in fields like law enforcement, intelligence and the military are hungry for proper training, Dunbar said. Teaching it in a game gives you a lot of cool things you can’t do in a workshop or an instructional video.
