Students’ choice of which university and programme to apply to isn’t a simple financial decision. There’s more at stake than finance and information gaps. Behavioural constraints play an important role. Camille Terrier shares her work on the impact of self-confidence on college access, highlighting an intervention that helps reduce disparities.
Camille Terrier will be sharing her research at this event for LSE students and staff, ahead of International Women’s Day.
Access to prestigious universities and high-paying careers remains uneven, varying substantially by gender and socioeconomic background. In a recent study, Rustamdjan Hakimov, Renke Schmacker and I explored how confidence affects students’ application decisions, using survey and administrative data from France. We also explore confidence gaps in relation to gender and socioeconomic status (SES).
With no centralised entrance exams and most students enrolling in free public institutions, France provides a useful case to examine confidence gaps and their impact on college access. Social inequalities in access to higher education are also comparable to those seen in other developed countries (like the US), making our findings widely relevant.
Why do we care?
Understanding how confidence affects college choice is vital from both efficiency and equity perspectives. From an efficiency standpoint, confidence can impact the student-college match. Underconfident students might avoid applying to prestigious colleges, mistakenly believing they have little chance of acceptance. They may later realise they could have succeeded had they applied, leading to regret and suboptimal educational outcomes. Conversely, overconfident students might aim too high and end up without a place.
Studying the relationship between confidence and college applications also matters from an equity standpoint as women and students from low socioeconomic backgrounds tend to exhibit lower levels of self-confidence compared to their male and high-status peers. We need to understand whether this confidence gaps explain why women and low-background students apply less to prestigious colleges.
In our study, we observe that among top students receiving the highest honours, women are 20 percentage points less likely to apply to elite programmes, and poorer students are 14.7 points less likely to do so compared to their male and wealthier peers. These gaps are particularly concerning, given that enrolment in prestigious colleges offers higher returns in terms of income and career opportunities.
Women and poorer students
To investigate the link between confidence and college applications, we conducted a large-scale survey of high-school seniors, typically 18-year-olds, participating in France’s national college admission process in 2021. In the weeks leading up to the application deadline, we collected data on students’ confidence, but also on their intended application lists and their perceived chances of admission to each programme.
We measured students’ confidence in their academic ability by asking them to report their grade point average (GPA). We then asked them to guess their national rank. As French students do not have the official ranking information, their guesses reveal over- or under-confidence. Some students guess a “better” rank than their real rank, while others guess a “lower” rank.
The survey revealed significant misperceptions among students regarding their GPA ranking. On average, those in the bottom half of the GPA distribution are overconfident, while those in the top half are underconfident. Most importantly, we find large gender and social gaps. Consider two high-achieving students who have the same GPA: female students position themselves 8.5 percentiles lower than their male counterparts. Similarly, students from low socioeconomic backgrounds also displayed greater under-confidence, estimating their rank 4.7 percentiles lower than wealthier peers.
Confidence and college application
After documenting confidence gaps, we investigate whether confidence affects college applications. To do so, we matched our survey data with administrative records on student real college applications, offers received, and final admissions. Our analysis now consists in comparing the applications of students who have the same GPA but whose confidence differs. We find that confidence is strongly correlated with the prestige of colleges to which students applied and were admitted. For example, being 10 percentiles less confident reduces the probability of applying to one of the elite French programmes (Classes Préparatoires aux Grandes Écoles, ou CPGE) by 3.3 percentage points.
Mitigating the confidence problem
These first findings raise an important question: Could we reduce the role played by confidence by correcting student misperceptions of their ranks? To answer this question, we designed an intervention that makes students aware of their over- and under-confidence and corrects it. We randomly split the participants to our survey into a treated group that receives feedback on their real rank in the GPA distribution and a control group that receives no feedback. This intervention aimed to determine whether accurate feedback could reduce the impact of mis-confidence on college choices and narrow gender and social gaps in admission. And the answer is yes.
Providing feedback decreased the role of confidence in selecting top-ranked programmes by 80 per cent, the likelihood of applying to an elite programme by 39 per cent, and the chance of being admitted to an elite programme by 72 per cent. Said differently, for students who received feedback on their real rank, their college applications were based on their GPA, but no longer on their (often incorrect) perception of their GPA rank.
For high-achieving students, correcting confidence also substantially reduced gender and social gaps in college applications and admissions. Rank feedback closed 61 per cent of the gender gap in application to an elite programme and 73 per cent of the gap in admission. It completely eliminated these gaps for poorer students.
Policy implications
Traditional policies aimed at boosting enrolment through preferential admissions, financial aid and information campaigns often overlook behavioural constraints. This study introduces a new, cost-effective method that is easy to implement and scale to reduce gender and social aspiration gaps.
Moreover, the results suggest that the timing of college applications relative to the release of academic results can influence these gaps. In countries where students apply after receiving exam scores, the gaps tend to be smaller, as students have a clearer understanding of their academic ranking. Recent initiatives give students personalised feedback on admission chances in various educational contexts. While calculating precise admission probabilities requires rich data, simply informing students of their relative academic rank can significantly influence their application behaviour.
Greater equity in higher education
Our study highlights the critical role of self-confidence in shaping educational trajectories and underscores the need for targeted interventions in addressing gender and social disparities in college access. By providing accurate feedback on academic standing, we can empower underconfident students – particularly high-achieving women and poorer students – to apply to prestigious programmes, unlocking opportunities that might otherwise remain out of reach.
As we get ready to celebrate International Women’s Day, our findings serve as a reminder that gender inequality exists and is perpetuated by our behaviour as much as by structural factors. Fostering confidence in all students can help us move to a more equitable system, where all young people, regardless of gender or background, are given the best foundation for success: belief in themselves.
Sign up for our weekly newsletter here.
- This blog post is based on Confidence and College Applications: Evidence from a Randomized Intervention.
- The post represents the views of its author(s), not the position of LSE Business Review or the London School of Economics and Political Science.
- Featured image provided by Shutterstock.
- When you leave a comment, you’re agreeing to our Comment Policy.
