Like A Girl

Pushing the conversation on gender equality.

Code Like A Girl

From Obstacles to Opportunities: How to Transform Girls’ Participation in Computer Science

by Elizabeth K. Lawner, M.A., and Harriet S. Mosatche, Ph.D.

When I (Elizabeth) was offered the opportunity to participate in an innovative computer science program during my first year as an undergraduate at Duke University, I decided to give it a try. At that point, I was undecided about my major so I signed on. While I got A’s in the courses and felt that I understood the logic underlying coding, I struggled with coding itself and often had to get help from a teaching assistant to get my code to work. I assumed those failures meant that I’d never be good at computer science and turned away from the field. My perspective changed when, as a doctoral student in social psychology at UConn, I became deeply involved in studying the barriers women face in pursuing underrepresented areas of STEM, such as computer science. As a result of that research, I finally understood why I turned my back on computer science.

My (Harriet) experience with the field of computer science has been two-fold. As a graduate student in developmental psychology, I learned about the value of computer science for quickly analyzing the data I had collected for my research studies, including my dissertation. This was at a time when one computer took up a large room and had far less memory than today’s cell phones. I entered my data on punch cards, which were then fed into that huge computer. Years after earning my Ph.D., I began to develop and evaluate after-school programs related to STEM and observed first-hand the obstacles girls faced in engineering and computer science, the two STEM fields in which women are most underrepresented.


The focus of this article is to provide an overview of some of the obstacles that continue to hinder females’ participation in computer science and to outline the many opportunities and practices that may lead to full engagement in this field by girls and women. Many of the research findings and suggestions that are discussed below are adapted from the book we wrote with Susan Matloff-Nieves, Breaking Through! Helping Girls Succeed in Science, Technology, Engineering, and Math (2016).

What are the Obstacles?

The obstacles to achieving parity in computer science vary from an inadequate supply of appropriate role models to the stereotyping of those who are interested in the field. This section describes the barriers that continue to prevent computer science from fully engaging females.

Invisible Achievements. Women who’ve worked in computer science have often not been given the credit or recognition they deserve. Ada Lovelace, whose achievements only recently have been publicized, created the first computer logarithm more than 170 years ago. The aptly named book and movie Hidden Figures highlighted three of the women whose computer science and engineering work made possible important achievements in early space exploration.

During World War II, women made up the majority of workers in the computer field in positions that were then considered clerical. As computer science became more professionalized and the field became subsumed under engineering, academic and hiring practices began to favor men and masculine qualities. One rarely mentioned pioneer from the 1970’s was Radia Perlman, a computer scientist and engineer who programmed a robot that could teach preschoolers.

Lack of Diversity. When we look at the proportion of women who are employed by tech companies, it is evident that the numbers continue to favor men. This disparity doesn’t only adversely affect the women who are deprived of enriching job opportunities, but it impedes the quality of problem-solving. After reviewing years of research studies, Katherine Phillips of Columbia University concluded that diversity of all kinds, including gender, leads to improved problem-solving and decision-making. But the field of computer science continues to be male-dominated.

While the ratio between male and female video game players is about equal, only about 22 percent of those who develop games are female, according to a 2015 survey by the International Game Developers Association. While that percentage is significantly higher than it was years earlier, women’s contributions in this field are still generally invisible to the public.

Continuing Stereotypes. A series of studies conducted by Sapna Cheryan of The University of Washington and her colleagues indicated that when college women interacted with a role model who had stereotypical interests (such as video games and reading Electronic Gaming Monthly) and wore stereotypical clothing (such as a tee shirt that referenced coding) they were less likely to think they could be successful in computer science or desire a career in that field. Believing they were unlike those stereotypical individuals, women were less able to see a path to success in computer science for themselves.

A phenomenon that has been shown to hurt girls’ performance is stereotype threat, which occurs when someone’s behavior could inadvertently confirm a negative stereotype about her group. The original studies on this phenomenon and much of the research since then has focused on females’ performance on difficult math tests, particularly when they are made aware of their gender or gender stereotypes, such as when they are one of the few females in the room or have to indicate gender before completing the test. The stress associated with stereotype threat seems to hurt performance, which may well lead females to leave a field.

Unconscious Bias. While people are less likely to openly discriminate today than in previous decades, bias continues to adversely impact females, although the bias typically happens in subtler ways. For instance, numerous studies have found that teachers and parents are likely to attribute girls’ success in math to effort while boys’ success is perceived as an indication of ability. The opposite pattern has been noted for failure with girls’ poor performance attributed to lack of ability and boys’ failures due to lack of effort. The implication for females in computer science classes is that boys are more likely to be encouraged to work harder when they don’t do well, while girls might be seen as just not having the ability to do better and therefore not urged to persist on a task they’re struggling with. And even when they succeed, girls are viewed as having lower abilities than boys since their achievements are perceived as being due to effort, not talent. Other research has demonstrated gender bias in hiring in STEM, due to women being perceived as having lower abilities, even when their qualifications are the same as men’s.

Imposter syndrome. More than four decades ago, researchers Pauline R. Clance and Suzanne A. Imes described a concept that, unfortunately, still holds some women back today. The “imposter syndrome” represents the fear that others will discover that an achievement or status was attained “by accident” and not due to their strengths. Experiences with discrimination in school and in workplaces sometime foster this belief. For instance, if a college student taking her first computer science class thinks that the high grade she received on a test was a fluke rather than a result of her ability, strategy, and effort, she becomes burdened by the possibility that her teacher or classmates will find out that she is not really as capable as her performance just showed. That burden can create enough stress and anxiety to decrease the odds of success in the future. The imposter syndrome is one of the reasons some women leave computer science behind — whether they are students or employees.

Where are the Opportunities?

While the obstacles described above are very real and deeply ingrained in society and in females, there is good news as well. Research has pointed to some promising practices that can improve gender equity in computer science academic fields and workplaces. Whether you’re a parent, grandparent, manager, teacher, youth worker, or another interested stakeholder, below are ideas you can use to change the picture of the computer science field.

Role Models and Mentors. Successful women in computer science can act as role models, inspiring girls and increasing or sustaining their interest in the field. There’s even some evidence that female role models can combat stereotype threat, increasing girls’ performance in STEM fields, though this has not specifically been tested in computer science. Female role models work by demonstrating that women can succeed in computer science and that negative stereotypes about women’s abilities are not true. Role models also illustrate what careers in computer science are really like and can convey all the benefits of that career path.

However, female role models are not always effective, and positive effects seem to be larger in controlled lab experiments than they are when role models are used in the real world. Moreover, in the few studies that focused specifically on computer science, the role model’s gender didn’t make a difference, but how stereotypical they were did matter. These findings make more sense when we consider two key factors that make role models inspirational: attainability and identification. In other words, girls have to feel that they could one day achieve the same level of success, and they need to connect with a role model, often by seeing themselves as similar in some way. So as counterintuitive as it might seem, it might not inspire girls to become computer scientists themselves if all of their role models are portrayed as geniuses and superstars. Similarly, role models who are close in age can make girls feel like they won’t be able to achieve the same level of success in such a short amount of time. In terms of identification, gender is not the only way for role models to be similar to girls. And as the role model research with computer science demonstrates, similarity on one dimension, like gender, can be outweighed by dissimilarity in another dimension, like personality and interests, and similar interests or personality can outweigh demographic differences.

One reason role models may work better in lab studies is that researchers can carefully control how a role model is portrayed, ensuring that they are similar to participants and that they don’t seem too perfect. Programs in real life that involve role models generally don’t have that luxury, and often use one role model to connect with many girls with different backgrounds, personalities, and interests. Even if a role model is similar to her audience, she might not have experience interacting with kids and might not know how to connect with them.

If you’re a parent, you can take advantage of these findings to create the best role model experience by doing the following:

• Browse books (non-fiction and fiction), online resources, and even TV shows and movies, to find role models with whom your daughter has something in common — whether it’s appearance, interests, personality, where she’s from, or even her name. Even better are role models who show their success as coming from hard work.

• Use your social and professional networks to find people working in computer science who might be willing to talk to your daughter (in person, on the phone, or online) about their experience. Ask them questions to find something your daughter can connect with them on, and suggest that they mention that topic in their conversation with your daughter. Ask them to talk to her about what sparked their interest in computer science and what they were like when they were younger.

Mentors are similar to role models, but the relationship is generally more involved, long-term, and one-on-one. Mentors also go beyond inspiring and providing information about their field; they give advice and can connect girls to opportunities. For that reason, mentors can be more helpful as girls get to their teen years and beyond, as they look for internships, choose courses, and start thinking about college and careers. The same considerations for role models apply to mentors as well. However, since the mentoring relationship is deeper, good mentors will find ways to connect with their mentees over time, even if they initially don’t seem to be particularly similar. And, they might discuss their own struggles in the course of giving advice, making it clear to your daughter that she, too, can be successful in computer science, even if she faces some obstacles.

Power of Language. One of the key ways girls can override the obstacles described in the previous section is by developing and using strong communication skills. Fortunately, through practice (as often as possible) and constructive feedback, girls can learn to use assertive language in a variety of situations. Perhaps she needs to confront a teacher who doesn’t think she should pursue a computer science major or she is interrupted by the boys in her class before she has finished her thought (something that happens far more to females than males in group situations). Girls, particularly those who are reticent in groups, need to be encouraged to speak often and with evident power. They can eliminate the questioning tone at the end of a sentence that suggests uncertainty rather than confidence. They can practice removing tentative phrasing (such as “I don’t know but it could be….”) from their conversations.

Body language, too, needs to show strength so that girls and women are not easily dismissed at school or in the workplace. Whether it’s making eye contact, shaking hands firmly, or not slouching into the scenery, they are communicating a power that cannot be easily dismissed. Encouraging assertive behavior and language is much more effective than giving belittling criticism.

Growth Mindset. Growth mindset, a term coined by developmental psychologist Carol Dweck, is the idea that intelligence and abilities are not innate, but rather can grow with effort and experience. Unfortunately, many people believe that intelligence is fixed, and the more people in a field believe that the ability to succeed in that field is innate, the fewer women are in that field. Encourage girls to instead adopt a growth mindset by communicating the importance of hard work and improvement over time. This allows them to view any challenges, particularly when learning something new, as opportunities to grow and improve, rather than as an indication that they’re not smart. Research shows that kids with a growth mindset are more likely to choose challenging tasks that allow them to learn, and they’re more likely to view failure as an indication that they need to try harder, persisting longer when they’re struggling. As a result, they tend to learn more and perform better in the long run. A growth mindset also allows girls to see superstar role models and still believe that they can achieve the same level of success.

You can model a growth mindset for girls by trying something new with them, a way to show them that it takes time and practice for everyone to get good at something new. And when girls excel, be sure to recognize the improvement, rather than commenting on their ability. That will reinforce the idea that talent can be developed and strengthened over time.

Creative Problem-solving. While creativity may not seem like a major part of computer science, which is built around logic, creative problem solving is key to the innovation that truly lies at the heart of the tech field. The type of creativity that’s most closely related to the problem solving used in computer science is the ability to think outside the box and see things in new ways. You can incorporate this way of thinking into activities girls already do, such as by challenging them to make toys, crafts, or even musical instruments out of everyday objects. There are also a lot of existing board games that involve creative problem solving or where taking a different perspective can help you win. Or you can make a game out of a common method of testing creativity in psychological studies: provide “players” with a set amount of time to come up with as many different uses as possible for a particular object, such as a tennis ball. The more people practice being creative, the more they can apply that strategy to problems in real-life and at work.

Advocacy. Recognizing the obstacles to gender equity in the computer science field is only the first step in transforming the field. Action by parents, teachers, employers, and workers in technology fields must follow. Some stakeholders are in positions that allow them to make direct changes. For instance, teachers can assess their teaching style and alter their approach to be more encouraging to girls, and managers can make sure their expectations for success are not biased due to gender. However, often people don’t have the authority to make changes directly, but they can become advocates for change. Advocacy can and should occur in schools and universities, in the community, in workplaces, and in the media. When girls themselves become advocates, not only are they enabling change to happen in their immediate environment, but they are also changing themselves — becoming confident, assertive leaders who will not back down from challenges.

Parents can advocate for gender-neutral school environments. When they visit their children’s schools, they should observe the following:

• whether girls and boys have equal access to technology equipment (including computers)

• the degree to which helpful and detailed feedback is given to all children in an equitable way (not just reacting to the students who demand the most attention)

• the extent to which tests are assessed blindly so that implicit bias doesn’t creep in

• “micro aggressions,” such as demeaning comments by students

• how well female achievements in STEM, particularly in underrepresented fields such as computer science and engineering are depicted in posters, notices, and books in classrooms and libraries.

Noticing is not enough; suggesting remedial action (and doing so in respectful but not aggressive ways) comes next. A committee of parents, teachers, and students who care about increasing gender equity can work together to implement changes at the individual school or district level.

Another learning environment that would benefit from advocacy is after-school or community programs. They are very often open and ready to implement innovative projects that serve a neighborhood need. Corporate foundations and even local retail stores might be willing to donate computers and volunteers, particularly if advocates can arrange for news coverage.

Advocating for gender equity is also important in workplaces. If you’re working at a tech company that does not already have a committee or department devoted to increasing diversity (and many small ones do not), volunteer to start one. Some of the ideas listed for schools can be implemented in work environments. A group might check out recruitment and retention data, policies, and practices, as well as instances of sexual harassment and microaggressions. Once valid data have been collected, they can be presented to management along with suggestions for change.

Closing Thoughts

As researchers who have a keen interest in applying data to real-life problems, we know that the gender equity gap in computer science can be reduced if enough people understand the underpinnings of the inequity and are ready to apply the research. To decrease that gap, areas such as deeply ingrained employment norms, parenting practices, and teaching biases must all be addressed, not at all an easy task. But all of society — not just women — will benefit when the field of computer science, which is so vital to the world’s future, becomes more open to the full diversity of talent.