For researchers eager to improve the position of women at their own institutions, scientists from the Massachusetts Institute of Technology (MIT) offer some hard-won advice.

  • Start From Grass Roots: An edict from above is not enough. "Women have to organize themselves," says Robert Birgeneau, dean of sciences at MIT. Such organization puts pressure on the administration to act.

  • Safety in Numbers: The great lesson for Nancy Hopkins, the biologist who kicked off the MIT effort, is the power of solidarity. "Depend on the power of the group, because it works," she says. "If you all go to the administration, they can't say there's no problem." That means first building trust with one another, adds MIT biologist and engineer Penny Chisholm. "Suspend suspicions among yourselves, overlook cultural and departmental differences, and get to the common experiences."

  • Find an Administrative Ally: "You have to have a Dean Birgeneau," says Hopkins, lest the effort languish in the university bureaucracy. MIT's dean of science was the hero of the MIT saga, according to many women involved. He took their claims seriously, gained the support of MIT's president, and worked tirelessly to forge innovative compromises.

  • Include Men: "It's very important to have well-respected male faculty on board," says Birgeneau. They add credibility, defuse tensions, and can help win over male colleagues. "At first I was against [including men]," says Hopkins, "but the dean turned out to be absolutely correct."

  • Collect Data, Not Enemies: "A spirit of cooperation works better than confrontation," says Birgeneau. Gathering data and quietly discussing it was more effective than loud protests, say some MIT faculty members. And data gathered internally are likely to be more valuable than those produced by outside consultants, who may have a hard time penetrating departmental cultures, says Hopkins.

  • Get the Resources: Hopkins skipped teaching for 2 years while she spent 30 to 40 hours a week gathering data and conducting interviews. "The institution has to say this is important enough to give you relief from teaching," she says. And Birgeneau says if he had to do it over again, he would hire a full-time assistant to help.

  • Seek Out Personal Stories: Some science faculties, such as that of Harvard University, have too few women to make statistical analysis meaningful, but individual interviews can provide important insights, too. Allowing women to speak confidentially helps ease fears of retribution, Hopkins adds.

She says that an organized, visible effort is far more effective than fighting individual battles. "Changing hearts and minds one by one is much too slow," she says. "You have to change the institution, and the hearts and minds will follow."