When the 2007 Richard Tapia Celebration of Diversity in Computing convened on Monday morning, the difference was apparent. Most of the 400 people in the room were younger than 30, and pale-skinned guys were in the minority. There was a roughly equal mix of black and Latino students, and there were more women than men. They were stylish and fun, too. They were itching to go to Disneyworld, which was just across the street. “We’re not people with big glasses and pocket protectors who talk in a monotone,” said Christopher Harris, a grad student at UC Irvine.

This was a crowd with a mission: breaking in. The plenary speaker set the tone. “Fundamental ideas about the role of computing in society are going to change, and your contributions are going to be critical,” said Shirley Malcolm, head of the Directorate for Education and Human Resources Programs at the American Association for the Advancement of Science. The technical stuff isn’t the most important part of computing, she said. She encouraged the crowd to think of the machines as tools that can be used to create a better world.

Dr. Malcolm

Most college students in the U.S. are women, and 24 percent are black, Hispanic, or Native American. Yet the NSF reports that only 22 percent of doctorates in computer science were awarded to women in 1999, and only 10 percent were awarded to these minority groups. This is the lowest minority share for any scientific field measured by the NSF. Worse, the share of college freshmen who say they are interested in studying computer science has declined sharply in the last decade. “We were well on our way to gender parity in computing in the early 1980s,” said Malcolm. “Somehow the numbers just fell apart.”

In the book Unlocking the Clubhouse, Jane Margolis and Allen Fisher investigated the barriers that keep diverse students away from the computer science departments at Carnegie Mellon University. They found that recruiters were looking for students who fit the stereotype of a computer kid, instead of high school students who were inquisitive and interested in computational problems. They also found that recruiters were selling the promise of building bigger, faster machines, instead of the challenge of solving problems. When recruiters began looking for problem-solving ability instead of pocket protectors, the barriers started to fall.

Diversity in computing is important for several reasons, said Malcolm. First, research shows that a group with diverse backgrounds does a better job of solving problems than does a group where everyone shares the same background. One reason they do better is that the concerns of diverse groups are recognized and addressed. Malcolm showed a map of gerrymandered Congressional districts that dilute the voting power of minorities in Texas. “If you don’t know how to do this, it gets done to you,” she said. “If you do know how, you can get a seat at the table.”

Computing problems can have huge social impacts. Voting technology is a prime example, because issues of verifiability and reliability are central to democratic representation. Malcolm showed how remote sensing has been used to monitor human rights abuses in Zimbabwe, and how the enhancement of photographs has been used to catch criminals. Computer models predicted the site of Hurricane Katrina’s landfall with high accuracy, she said, but the computers in use in 2006 were not powerful enough to predict the strength of the storm at landfall. The next generation of machines will be able to do this, and therefore to give more powerful and accurate warnings to the public and leaders. “The point is to use computers to make a difference in the world,” she said. “They are a means to an end, not an end in itself.”