Who They Become

The Stage

When I took on the role of the director of UC San Diego’s Early Research Scholars Program (ERSP), I knew I was inheriting something special. It turns out this first year has been incredibly rewarding.

ERSP is a year-long research experience for undergraduates. Students spend the fall quarter in a methods course where they learn how to read papers, manage projects, and collaborate in teams. In winter and spring, they leave the classroom and work on real research projects with faculty and graduate mentors. The year ends with a spring research symposium, where every team presents its results to the campus community.

This winter, I met with ERSPers, 54 students across 15 teams, for individual 20-minute check-ins. These conversations offered a close look at how students experience the shift from a structured class to open-ended research.

When the Training Wheels Come Off

The first shock they described was the ambiguity of the research process. No more structured assignments that tell them exactly what to do. Then there were the technical challenges. APIs disappeared. The datasets changed format. Weeks of code had to be thrown away when a model failed silently. Teams pivoted, scraped websites, rewrote parsers, and tried again. They learned that progress is rarely a straight line and that setbacks are part of the research process. Yet amid all the ambiguity, students found satisfaction in measurable progress. As one told me, “I love watching the model loss decrease”, the kind of tangible feedback that keeps you going.

What Kept You Going

Each group built its own system: Discord servers for late-night debugging, Git branches for safe merges, Zoom check-ins when midterms collided with deadlines. Some rotated leadership, others split into sub-teams. Teams that met weekly to discuss progress and maintained clear communication thrived, as one student noted, “We all know what’s going on!” When I asked students what kept them going when the going got tough, they consistently pointed to their teammates! The group kept each other’s spirits up, and peer accountability kept the work moving when classes, illness, or exams threatened to stall it.

Learning by Doing

Projects ranged from quantum algorithms to machine-learning pipelines to sound simulation. Students picked up tools and habits that no lecture can deliver: reading dense papers for usable ideas, tuning hyperparameters until a loss curve finally bent, and presenting messy results to a skeptical mentor.

Many also discovered that research writing, while challenging, felt different from academic essays. One student called proposal writing “hard but good practice,” recognizing it as a skill worth developing, even when it felt frustrating in the moment.

Building Confidence

Almost everyone talked about having built confidence. By spring, they could read new papers, ask insightful questions, and debug a stack they had never seen before. Time management stopped being a buzzword and became a survival skill. Many arrived unsure about research and left knowing they could handle it. By the end, students were already thinking beyond ERSP, applying for summer research programs, weighing research opportunities against industry internships, and some even preparing material for grad school.

Bottom Line: ERSP’s true measure isn’t what students learn, but who they become. After three quarters of steady work, students don’t just bring a poster to the spring symposium. They show up knowing they can handle ambiguous challenges, collaborate effectively, and persist when things get difficult.

Thanks: None of this would have happened without the dedication of our faculty and graduate student mentors, who guide every team week after week. Their time, expertise, and patience make ERSP possible.