Women who work in the high-tech industry gather for LeadOn, a conference held in February in Santa Clara, California.CREDITPHOTOGRAPH BY MARLA AUFMUTH/GETTY FOR LEADON

Amy Wibowo, a graduate of M.I.T. and a software engineer at Airbnb, wears pastel-colored floral dresses, makes lavish use of emoji on Twitter, and gets around on a mint-green bicycle. She presents herself like this—as “feminine,” in her words—not only because she feels like it but also, she recently explained, “to say a big FUCK YOU to the patriarchy.” Wibowo made that comment in January, during a speech at an Oakland gathering of AlterConf, an event series focussed on diversity in the tech and gaming industries. In early March, she posted a transcript to the Web site Medium, and it became a minor viral hit. Encouraged by this success, on Monday morning, Wibowo launched a Kickstarter campaign to publish a series of zines about computer science that would be geared toward high-school students. The proposed publication reflects her own sensibility and, she believes, that of many girls. One sample cover, for an issue about the computer-security practice known as cryptography, features a drawing of a cat whispering into another cat’s ear, and the words “Secret Messages” written in bubble letters and surrounded by a heart. Wibowo also wants to emphasize applications of computer science that she feels will appeal to girls, like using natural-language processing to research the needs of women in developing countries. Her message resonated; within four hours of launching her fund-raising campaign, she had reached her goal of ten thousand dollars. By Thursday, people had pledged more than twenty thousand dollars. Wibowo told me that she had decided to leave Airbnb to focus on the zine and other educational initiatives.

Airbnb, like many Silicon Valley companies, employs few women in technical positions; women make up thirteen per cent of its engineering team, a spokeswoman told me. Not long ago, such statistics involving women in Silicon Valley were little discussed; some people seemed to accept the dearth of women as the natural order of things. Then came a series of much-publicized stories about instances of overt sexism in tech circles, ranging from a couple of men getting onstage at the TechCrunch Disrupt conference and describing an app called Titstare to the C.E.O. of Uber joking about a hypothetical women-on-demand app called “Boob-er” to the C.E.O. of Microsoft remarking to an audience of female engineers that women who don’t seek raises have better “karma” than those who do. Then, last year, a female engineer at Pinterest, Tracy Chou, helped persuade several of Silicon Valley’s most prominent companies, including Google and Facebook, to reveal the percentages of women they were employing in tech positions. In most cases, the figure was below twenty per cent. Other research has shown that the percentage of women in computer science has declined in recent decades, while the proportions have risen in many other scientific fields.

It has started to dawn on Silicon Valley executives that this state of affairs is unacceptable. In January, Intel announced a three-hundred-million-dollar investment, over five years, to make its workplace more diverse; its C.E.O., Brian Krzanich, said at the time that not having a workforce that more closely mirrors the broader population meant that the company had been missing out on business opportunities by failing to understand what its customers sought. (Airbnb, the spokeswoman said, feels its proportion of women in technical roles is unacceptable and is trying several measures to improve the number.) Partly because the problem was ignored for so long, though, there has been relatively little research to explain why it exists and how to address it.

On Thursday afternoon, the American Association of University Women published a hundred-and-sixty-page report, “Solving the Equation,” that attempts to do just that. Some of the findings are unsurprising: for instance, that women engineers who reported being “frequently belittled, patronized, or systematically undermined” were likelier to intend to leave their jobs than those who don’t. But, while much of the public conversation about gender diversity in Silicon Valley has focussed on getting rid of detrimental behavior and policies—calling out sexist behavior, for instance, and encouraging hiring and promotion practices that minimize the influence of subtle gender biases—the “Solving the Equation” report also introduces research that suggests a different approach: Along with minimizing the aspects of engineering education and culture that make girls and women feel unwelcome, what about proactively changing engineering culture so that it’s more attractive to them in the first place?

The report highlights past studies by a Miami University psychology professor, Amanda Diekman, and her colleagues, who examined the relationship between gender, professional options, and two types of goals: “communal” goals, related to collaborating and helping others, and “agentic” ones, having to do with independence and self-advancement. In a survey of undergraduates, they found that women, on average, gave higher ratings to the importance of communal goals than men did, and that men gave higher ratings to agentic goals than women did.

The same undergraduates also rated the likelihood that various fields would help a person meet communal and agentic goals. On average, they rated engineering, computing, and science fields as more likely to fulfill agentic goals than communal ones. In other words, satisfaction in those fields is perceived as coming less from collaborating and helping people than from attaining individual self-advancement. “If you’re choosing your career, it matters if you care a little bit more about making a social contribution than you do about mastering some skill set,” Christianne Corbett, a co-author of “Solving the Equation,” told me. She said the study’s findings suggest that employers and universities could attract more women to engineering by emphasizing the communal aspects of the field—“both in how the fields are sold and marketed, but also in the substance.” Teachers and professors, for instance, might emphasize the social good that computer scientists can bring about, while companies might better incorporate societal goals into their missions and better communicate their commitment to those goals.

Corbett told me that she and others spent a great deal of time discussing whether even to raise this suggestion. The notion risks playing into stereotypes about feminine and masculine behavior and values (there are, of course, many women who, quite reasonably, care a great deal about self-advancement), and could also draw attention away from other important factors, such as gender-biased hiring practices, that have been shown to diminish the numbers of women in tech. Corbett and her co-author, Catherine Hill, who discuss those other factors elsewhere in the report, ultimately decided to include Diekman’s research but directly acknowledge the tensions. They quote Diekman’s remark to them: “One of my fears is that this research will be interpreted as suggesting that gender differences in communal goals are the only thing that matters.”

Corbett and Hill also raise another fraught possibility, which is that computer science simply might not fulfill communal goals as well as other science professions: “What if the stereotype that these are solitary jobs that don’t provide opportunities for making a social contribution is mostly accurate? If that’s true, then recruiting communally oriented women and men into engineering and computing through marketing campaigns touting the collaborative, helpful aspects of these fields will result in an exodus from the field later. The reality, of course, is more important than public perception.”

Some anecdotal evidence seems to align with the more optimistic suggestion that the field can be made more genuinely relevant to women’s professional goals. One chapter of the report discusses the success of Harvey Mudd College, in Claremont, California, at increasing its percentage of female computer-science graduates from fourteen per cent, in the Class of 2009, to thirty-eight per cent, in the Class of 2014. (The national average is eighteen per cent.) Harvey Mudd attributes its improved numbers to three major changes: reworking its introductory engineering class to focus more on the real-world applications and “social relevance” of engineering; providing research opportunities to women after their freshman year “to expose them to real computer-science problems”; and bringing women to the Grace Hopper Celebration of Women in Computing—a conference for women engineers and, incidentally, the event at which Microsoft’s C.E.O. made the controversial comment about raises for women—where they could “see a computing culture that is different from the stereotypical culture they might expect.”

With her zine, Wibowo is taking a similar approach. “It’s really awesome to have social issues that you care about and be able to do something about them with computer science,” she told me. But those social issues, she said, “are not the first thing people talk about” when they discuss computer science. She added, “For whatever concept is being covered in the zine, I want to talk about the range of applications.” She also sees value in talking about her own experiences as a woman engineer who presents in a “feminine” manner—not to suggest that all female engineers should wear floral dresses and speak softly, like she does, but to promote it as no less valid than turning up at work in a hoodie and jeans and using a loud voice.

During her AlterConf speech, Wibowo recounted a story told to her by Chou, of Pinterest, who, after going to a tech conference wearing a dress and having people assume that she wouldn’t understand technical matters, returned the following day in jeans and a T-shirt and was accepted as someone with a technical background. “If we have more women and marginalized people participating in computer science, but the price of doing that is that they have to fit into the stereotypes of what computer scientists are, then that’s not diversity,” Wibowo told me. “So I feel like however you want to present, whether that is super-feminine or anything else, if you show up at a conference and give an amazing talk, presenting the way that you like to present, then you’re making a statement about what computer scientists look like.”