Gender, Academic Performance, and STEM Career Choice: Evidence from Large-Scale National Assessments in Colombia
This study of over 126,000 Colombian students reveals that gender differences in STEM career selection persist regardless of academic performance in mathematics, natural sciences, and language, indicating that non-academic factors like self-efficacy and stereotypes, rather than preparation, drive the enrollment gap.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Every year, millions of young people around the world stand at a crossroads, deciding which path to take in their higher education. For decades, researchers and policymakers have worried about a persistent imbalance in these choices: far fewer women than men choose to study science, technology, engineering, and mathematics, often called STEM fields. This gap matters because it limits the pool of talent available to solve complex global problems and shapes the future economic opportunities for women. A common assumption has been that this disparity exists simply because girls are not as prepared academically as boys in the subjects required for these careers. If the data showed that women consistently scored lower in math and science during high school, the explanation would be straightforward. However, the reality is far more complex, and the reasons behind the gap have remained a subject of intense debate.
To understand why this question is so difficult to answer, one must look at how career choices are made. Psychologists and educators have long proposed that a student's decision is not just about how well they can solve an equation, but also about how they feel about their own abilities and what they believe society expects of them. This is known as the idea of "self-efficacy," or the belief in one's own capacity to succeed, and "stereotypes," which are the widely held but often unexamined beliefs about what men and women are naturally good at. When a student believes that a field requires a special, innate genius that they do not possess, or when they feel that a career does not align with their personal values, they may step away from it, regardless of how high their test scores are. The critical question for researchers has been whether these non-academic factors are strong enough to override academic preparation, or if the gap is simply a reflection of different levels of skill.
A team of researchers in Colombia set out to settle this question with a massive, detailed look at the lives of over 126,000 students. They used a unique national system that links two major events in a student's life: a standardized high school exit exam taken by everyone, and a professional competency test taken by almost everyone who graduates from college. By connecting the scores from the high school exam directly to the specific majors these students chose in university, the researchers could see exactly what happened to students who had the exact same academic grades. They did not just look at averages; they looked at the entire range of performance, from the lowest-scoring students to the very top achievers. They examined whether a woman who scored in the top 20 percent of her class in mathematics was just as likely to choose a STEM major as a man who scored in that same top 20 percent.
The results were clear and consistent. The study found that academic preparation alone does not explain the gender gap. Even when women and men had identical scores in mathematics, natural sciences, and language, the women were still significantly less likely to enroll in STEM programs. This pattern held true across the entire spectrum of academic ability. Whether a student was in the lowest performing group or the highest, the probability of a woman choosing a STEM career was lower than that of a man with the same grades. The researchers calculated that the gap was substantial, often amounting to a difference of more than ten percentage points in enrollment rates between men and women who were otherwise academically indistinguishable.
To dig even deeper, the researchers asked a more specific question: how much better would a woman need to perform than a man to have the same chance of choosing a STEM career? They compared women who were one step higher in the performance rankings than men to see if that extra academic boost would close the gap. In almost every case, it did not. A woman who scored in the second-highest group of students was still less likely to choose a STEM major than a man who scored in the third-highest group. The only time the gap nearly disappeared was at the very top of the mathematics performance scale. Even there, the effect was fragile; a woman needed to be in the absolute highest group to match the enrollment rates of a man who was just one step below her. For natural sciences and language, the gap remained wide even when women outperformed men by a full performance level.
These findings suggest that the barrier to entry for women in STEM is not a lack of skill or preparation. The data indicates that something else is at work, pushing women away from these fields even when they are fully qualified. The researchers point to factors that exist outside the classroom, such as the belief that success in these fields requires a rare, innate brilliance that society often associates with men. They also highlight the influence of stereotypes that suggest women are less interested in engineering or that these careers do not support communal goals, which many women value. The study also noted an interesting twist with language skills. While high scores in reading and language generally led students away from STEM, this effect was weaker for women than for men, suggesting that women with strong verbal skills do not feel the same pressure to leave science fields as men do, yet they still choose to do so at lower rates.
The implications of this study are significant for how we think about educational policy. If the problem were simply that girls were not learning math well enough, the solution would be to improve math instruction. But since the gap persists even among the best students, the solution must address the cultural and psychological environment. The researchers argue that efforts to increase the number of women in STEM must go beyond the classroom. They need to tackle the stereotypes that take root in early childhood, change the way teachers and parents talk about ability, and create supportive environments that make women feel they belong in these fields. The evidence from Colombia shows that academic excellence is not enough to guarantee equal participation; the path to a STEM career is shaped by a complex mix of confidence, belief, and social expectation that affects women differently than men, regardless of their grades.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.