Welcome to UCLA Undergraduate Research Week 2025!

Thank you for visiting the 2025 Undergraduate Research and Creativity Showcase. This Showcase features student research and creative projects across all disciplines. As a university campus, free expression is encouraged, and some content may not be appropriate for all ages. Visitors under the age of 18 are encouraged to explore these presentations with a parent or guardian. The views and opinions expressed here are those of the participants and do not necessarily reflect UCLA or any policy or position of UCLA. As a visitor, you agree not to record, copy, or reproduce any of the material featured here. By clicking on the "Agree" button below, you understand and agree to these terms.

Communications, Economics, and Geography: Prerecorded - Panel 2

Monday, May 19 12:01AM – 11:59PM

Location: Online - Prerecorded

Presenter 1
CHANAPORN TOHSUWANWANICH
This paper examines how affluent neighborhoods control their spaces, from minimum lot size requirements to gatekeeping public parks in the City of Beverly Hills, California. Utilizing spatial analysis and ethnographic observations, the study reveals that restrictive land-use policies, such as minimum lot size requirements, limit public space availability, reinforcing socio-economic disparities. Public parks in northern areas often function as semi-private extensions of luxury properties, with numerous regulations catering to aesthetic preferences rather than inclusivity. Parks in the southern parts of Beverly Hills, though more community-oriented, still face regulatory barriers, revealing the city’s prioritization of property values and homeowner interests over accessible public spaces. Through analysis of city commission meetings and planning decisions, this paper demonstrates how the influence of affluent residents on urban policy perpetuates exclusivity, as public parks serve dual roles in enhancing property values and acting as extensions of private green spaces. The findings underscore the need for urban planning reforms that address these inequities, advocating for public spaces that foster inclusivity and meet the needs of diverse communities.
Presenter 2
ARAVINTH RUPPA, Adrianna Lleras-Muney
Cancer remains one of the leading causes of death worldwide, with new cases projected to reach 35 million by 2050 [1]. As a result, significant resources and time have been devoted to developing new therapeutics. In 2024 alone, the National Cancer Institute allocated $7.22 billion to cancer research [2]. However, even after a discovery is made, many obstacles remain before a drug reaches patients. The FDA enforces a rigorous approval process through clinical trials to ensure safety and efficacy. This process is lengthy, costly, and only the first hurdle. Once approved, drugs enter a complex market influenced by numerous factors that affect adoption rates. Despite a drug’s availability, variables such as pricing, reimbursement rates, cancer incidence, and supporting clinical literature can drastically impact its uptake. There remains a gap in knowledge regarding which of these factors most significantly influence adoption. To investigate this, we collected data from the CMS on reimbursement rates and claims for novel cancer drugs approved around 2018, using claims volume as a proxy for clinical adoption. We also gathered cancer incidence data from the SEER database, PubMed literature mentions for each drug, and pricing data from the NADAC database. Using this dataset, we aimed to develop a model to identify which variables most strongly affect the rate of claim increases. Our goal is to highlight key areas that can be targeted to improve clinical adoption and ensure patients has access to novel cancer therapies.
Presenter 3
JORDEN PORK
This research paper aims to identify the role of emitted emotions in predicting engagement levels on TikTok videos that are product-related and aimed at women. TikTok has a large influence on the female buying process, especially through the TikTok Shop part of the platform. Understanding what emitted emotions can increase engagement and views has practical importance for creators, marketers, and brands since this is something that they have control over and can employ consciously with their content. The research asks: Which emitted emotions and emotional drivers most effectively predict increased engagement in product-related TikTok videos aimed at women? To address this, a mixed-method approach to content analysis will be conducted on 1,000 high-performing videos, split between TikTok Shop native videos and user-generated content. The videos were qualitatively coded with the emitted emotions they contained, quantitive analysis will then be employed using ANOVA, SPSS, and Natural Language Processing to identify recurring emitted emotions and their associated levels of engagement. Preliminary results show that it is not just one emotion that correlates with and increase in engagement but a combination of 3: desire-centric, informational, and relatable emotions. Ultimately, this paper provides actionable insights for enhancing content strategy in female-focused commerce on TikTok, offering a deeper understanding of how emotional cues drive viewer interaction and the inception of content virality.
Presenter 4
LETICIA RODRIGUES GUIMARAES BARBOSA
Phishing is a significant cyber threat which exploits human vulnerabilities to manipulate individuals into disclosing sensitive information – typically done in the form of fraudulent emails or messages. Research suggests that generational differences may play a role in phishing susceptibility, with older adults being more vulnerable due to factors such as the age-positivity bias and cognitive decline. This study investigates how age influences susceptibility to phishing attacks by examining the effects of visceral triggers (eg. urgent tone) and phishing deception indicators (e.g., spelling errors, suspicious URLs) on participants’ responses to a set of closed and open-ended questions. Using a sample of 200 participants from the United States, this study employs a survey design with six conditions based on email source (bank, government, company) and authenticity (phishing, authentic). Preliminary results indicate that participants are better able to differentiate between phishing and authentic emails in the bank and company conditions than in the government condition.
Presenter 5
KENNY GUO, TERRENCE YU, Peter Mannino
In this project with the California Policy Lab, we study the Unemployment Insurance (UI) Program in the United States. In particular, there has been much coverage on the recent inflation on workers and consumers, however less so on how inflation affects social programs like UI and the individuals under its reach. We primarily study the Weekly Benefit Amount (WBA) of the UI program, which is the dollar amount a UI claimant can receive a week for the duration of the program. We assess both long-run and recent short-run inflation erosion of the WBA, revealing drastic decreases in real WBA and the lack of adjustments by state policymakers across the nation. We also study whether changes in real WBA amounts are related to changes in the recipiency rate--the fraction of UI-eligible individuals receiving benefits--to discover how inflation affects UI and its dependents.
Presenter 6
JEANNINE XU

As AI systems increasingly influence everyday decisions, questions arise about how people perceive fairness and bias in human-AI interactions. This study examines how encounters with AI systems that reflect underlying social biases affect user trust, emotional responses, and perceptions of agency or fault. Using a mixed, survey-based experimental design, participants were randomly assigned to one of six conditions across two scenarios: a career advice scenario involving gender bias and a travel recommendation scenario involving cultural bias. Surveys assessed trust in the AI, emotional reactions, attribution of bias, and prior AI familiarity. Preliminary results show that biased responses reduced trust and increased frustration compared to neutral conditions. Bias against male users was more readily perceived, while participants with higher education or AI familiarity were less reactive and more likely to attribute bias to technical flaws. These findings suggest that users extend social expectations to AI systems and that trust is shaped by both content and context. This research supports the development of more equitable AI and highlights the influence of education and familiarity in shaping how users respond to perceived unfairness.