March 2007 Vol. 19/No. 2
By Diana Franklin
Last year, a colleague in my department approached me with the following quandary: Why did his female student have difficulty working independently on her senior project, despite her demonstrated ability in his class?
When we delved further, we discovered it was merely fear of failure and the need for reassurance, not a lack of ability, that caused her to give this impression. This led me to two questions—had she not been in his class, how would he have recognized her potential? How can he develop in her the confidence and independence necessary to succeed at competitive levels in academia?
In this article, I discuss some of the societal causes of this apparent lack of independence, as well as other communication patterns that make females appear to have less ability than their male counterparts. I conclude with a set of suggestions for professors and mentors to help females develop the skills they need to succeed at the higher levels of this field.
Let us begin with two caveats. First, these are statistical observations, and the difference within the group is often larger than between groups. There are females who exhibit none of these characteristics, and those are the most likely to have succeeded to date. My concern is that we are unintentionally discouraging a large, talented segment of our population because they do not appear as capable as they are.
Second, behaviors that are a product of society change through time. What was true in this generation may not be true in the next. If we are aware of these societal effects, we can both develop the students who were influenced by them and change how we treat young females in order to prevent them from experiencing the same fate.
The difference in society’s treatment of males and females begins by pre-school. It has been observed that females are subtly punished by both their fathers and peers for playing with “boys’” toys 4. In elementary school classrooms, teachers ask boys more questions 2, and the type of question is more open-ended and challenging than those asked of girls 7. Thus, females have less practice with open-ended problems, leading to less confidence in these types of tasks.
Confidence is further diminished during adolescence. Adolescence is a difficult time for all children, especially those who are academically gifted. For females, the social structure is very important to self-confidence, so this period is especially hard on self-esteem. Females have been socialized to understand subtle social cues, so they are painfully aware of their social failures, such as when well-meaning relatives ask if they have a boyfriend 3. From adolescence to early adulthood, female self-esteem tends to decrease, whereas that of males increases 1. Furthermore, valuing independence and autonomy was strongly correlated with lowered self-esteem in females 1. Thus, by the time they enter college, females have much lower self-esteem than males.
Computer science is an excellent major for the male students to fit in, but for females this can be worse than high school, since it is the first time they are a minority in academic classes. In addition, males are more likely to have programming experience entering college 5, magnifying this sense of isolation for females. As a result, females have less confidence than their male counterparts and transfer out of the major, citing failure as the major reason, at higher GPAs than males 5,6.
Concrete Suggestions for Teaching and Mentoring
In only four years of teaching, I have seen the trends above expressed in many ways that would hurt recognition of the students’ accomplishments as well as damage their ability to compete at the highest level. These behaviors of high-achieving females include choosing less ambitious open-ended projects, asking many questions in class and/or office hours, and exhibiting a lack of persistence in independent work. We need to distinguish between perception and ability, and develop those students who have the potential to succeed. The idea is to begin the process as early as possible in students’ education in order to reverse the societal influences by the time students exit, or even enter, graduate school. These suggestions form a progression, from careful attention, to determining a student’s knowledge, to explicitly developing her self-confidence and independence.
Ask a student to guess the answer to their question to gauge their knowledge. Females are more likely to ask questions to reassure themselves they are on the right track, rather than because they have no idea of the answer. Determining the level of their knowledge is important in how you view them. I had a female student who, despite getting top grades on all the tests, did not convince her professor she was smart until the second quarter she had him; because she asked so many questions, he thought her grades were a fluke.
Provide mentoring. Because of the social isolation in computer science for females, it is more important to have a mentor relationship. Several times, I have shared stories with students and former students. The farther they get in their education, the more they identify with the stories. They have expressed relief and renewed confidence that what they are experiencing is normal.
Do not believe seemingly self-aware expressions of self-doubt. Females are likely to minimize their accomplishments. One student had so little confidence that she often expressed her doubts in her ability to succeed at open-ended, creative research problems. Despite the fact that the professor’s only direct evidence of her ability had been stellar, this led to a somewhat negative letter of recommendation.
Make expectations explicit. Open-ended projects should not be used to determine who is the most capable. Students with low self-confidence are more risk-averse, leading to less ambitious projects. If you want to see if a student can do something more challenging, assign it.
Assign leadership. In many large research groups, the professors let the leaders “naturally emerge” from the group. In group settings, men are more likely to interrupt females than males, and females are more likely to acquiesce than males when interrupted 8. This and other factors mean that a hands-off approach is very unlikely to result in a female leader. If you want to see how a female performs as leader, declare her leader and counsel her as to what level of authority that leadership gives her.
Assign increasingly risky projects. The student needs to be told explicitly that the intent of research is exploration, sometimes failure, and eventual success. Making failure an intermediate goal helps risk-averse students try ambitious projects. Beginning with a very high-risk project that fails, on the other hand, may reinforce the feelings of inadequacy.
Although it is easier to teach only technical content and make the students learn assertiveness and confidence on their own, this does a disservice to the field. Males and females alike can benefit from these techniques. Furthermore, as society matures, more and more females will be treated as equals from a young age, so one hopes that these techniques will be necessary for fewer students. In this era of global competition, we need to train those who will contribute the most, not just the ones who are the easiest to teach.
1 Block and Robins, “A Longitudinal Study of Consistency and Change in Self-Esteem from Early Adolescence to Early Adulthood,” Child Development, 1993, 64, 909-923.
2 French and French, “Gender imbalances in the primary classroom: an interactional account.” Educational Research, 1984, 26, pp. 127-136.
3 Holland and Eisenhart, Educated in Romance: Women, Achievement, and College Culture (1990, University of Chicago Press).
4 Langlois and Downs, “Mothers, Fathers, and Peers as Socialization Agents of Sex-Typed Play Behaviors in Young Children,” Child Development, 1980 – JSTOR.
5 Margolis and Fisher, Unlocking the Clubhouse: Women in Computing (2002, MIT Press).
6 Patterson and Trasti, "Women Students in Computer Science: Student Perspectives of Faculty Bias as a Possible Influence on Student Retention," http://www.multicultural.vt.edu/proceedings/Women_Students_in_Computer_.pdf
7 Swann & Graddol (1998), “Gender inequalities in classroom talk,” English in Education, 22, pp. 48-65.
8 West, Candace and Don I. Zimmerman, "Small Insults: A Study of Interruptions in Cross-Sex Conversations between Unacquainted Persons.” In Barrie Thorne, Cheris Kramarae, and Nancy Henley (Eds.), Language, Gender and Society. (Rowley, Mass.:Newbury House, 1983, pp. 102–118).
Diana Franklin is an Assistant Professor at Cal Poly, San Luis Obispo, who researches computer architecture.