Eunsaem Cho, a student in the Master of Data Science program at KDI School, has received the Outstanding Thesis Award for her thesis titled “Digital Divide vs. AI Divide: A Comparative Analysis of Structural Inequality.” Her research examines how traditional forms of digital inequality differ from emerging patterns of AI inequality, offering timely insight into one of the most important policy challenges in the age of artificial intelligence.

Before joining KDI School, Cho studied digital media design and worked as a UX designer for four years. That background shaped her interest in how people interact with technology, but her time at KDI School helped her look at those interactions through a broader policy and research lens. “When I was working as a UX designer, the goal was to solve user problems as quickly as possible,” she said. “Research is a different direction. My advisor once told me that research is not about solving a problem, but about understanding a phenomenon in a broader context. That really stayed with me.”

Cho’s thesis compares the digital divide and the AI divide within the same individuals. The study uses Korea’s annual National Digital Divide Survey, which includes 15,000 respondents. The 2024 survey was especially important because it included both digital and AI-related modules. This made it possible to measure digital inequality and AI inequality in the same sample, something that has rarely been done in previous studies.

According to Cho, much of the existing research on AI inequality has examined the AI divide separately from the digital divide. As a result, it has been difficult to understand exactly how AI inequality differs from earlier forms of digital exclusion. Her thesis was motivated by this gap. By directly comparing the two domains, the study provides a clearer picture of whether current digital inclusion frameworks are sufficient for addressing AI-related inequality.

The idea for the thesis began earlier than Cho expected. During her first semester, she had already been working with Professor Jaehyuk Park on AI divide research. Later, for her final project in R Fundamentals for Public Policy, she tried comparing the digital and AI divides within the same dataset. At first, she did not expect the project to become a thesis. However, Professor Park saw its potential and encouraged her to develop it into a full research paper. Professor Byungkoo Kim, who taught the R course, later joined as an advisor, and the thesis was completed under the guidance of both faculty members.

Cho said several first semester courses in the MDS program played an important role in helping her develop the thesis. R Fundamentals for Public Policy, Statistical Foundations for Data Scientists, and Introduction to Computational Social Science were particularly influential. These courses helped her connect data analysis with actual policy questions. She also emphasized the role of faculty support throughout the process. “My advisors guided the research from start to finish, both in shaping the direction and giving detailed feedback,” she said. “And it was not just my advisors. The MDS faculty in general were genuinely generous with their time, and that kind of environment really helped.”

The thesis writing process was not without challenges. Cho began the research in earnest around May and had a draft by early August. At that point, she had selected what she believed were the most representative variables for each domain, and the results were largely in line with her expectations. However, her advisors encouraged her to take a more systematic approach to variable selection. This meant restarting the analysis with nearly all available variables included.

The new results were much harder to interpret. Some patterns were unexpected, and Cho initially worried that the analysis had failed. “I genuinely thought the analysis had failed and did not even want to show my advisors,” she recalled. “But they told me not to judge the results myself before bringing them in, and that unexpected findings are often where things get interesting.” Looking back, she believes that difficult decision made the final paper much stronger.

To current KDI School students preparing their own thesis, Cho offers two pieces of advice. First, do not struggle alone. “When you get stuck, do not sit with it alone. Go talk to your advisor,” she said. Second, avoid comparing yourself too much with others. “Even I sometimes feel behind when I see people working with the latest methods or complex datasets. But if your message is clear and it covers something that has not been well addressed in prior research, I think that can actually make for a better paper.”