Applied Research
Research combining qualitative insight with performance and usage data to surface what is actually happening. The goal is not to collect everything, but to identify meaningful signals and separate them from noise.
Understanding Political Contagion and Strategic Intervention
A disaggregated quantitative network analysis of Donald Trump’s political coalition, distinguishing Core MAGA adherents from Non-MAGA transactional supporters and modeling each group’s distinct behavioral mechanics. Drawing on complex contagion theory (Centola), asymmetric hysteresis (Xie et al.), and preference falsification (Kuran), the briefing builds a longitudinal model spanning 2015 to the present and bifurcates into two forecast scenarios through 2029: Successor Continuity and Leaderless Fragmentation.
The central finding: the two coalition segments do not collapse through the same mechanism. Core MAGA experiences a non-linear preference falsification cascade when sanction capacity drops below the ~19% cascade floor. Non-MAGA supporters experience linear transactional drift governed by accumulated policy friction. Neither mechanism breaches the 19% threshold at the aggregate level in either scenario — the cascade trigger is the MAGA component specifically, not total approval. This structural distinction is the analytic contribution: collapsing both groups into a single approval line conceals exactly the dynamic worth understanding.
Methodological standards are maintained throughout. Where data is estimated, reconstructed, or illustrative rather than measured, that is stated plainly. The framework is offered as a rigorous theoretical structure consistent with available evidence, not as a fitted predictive model. Three formal arguments are structured for peer review: hysteresis and fold bifurcations, mechanistic divergence between contagion and sanction dynamics, and game-theoretic payoff asymmetry between the two coalition segments.
YourPeer: Campaign Performance and User Testing
Applied research across two concurrent tracks for YourPeer.nyc, a mobile-first web application developed by the non-profit Streetlives connecting unhoused youth in New York City to essential services including shelter, food, mental health, and employment. The campaign performance analysis compared Google Ads data across Q4 2024 and Q1 2025, revealing a 25% increase in conversions alongside a 25% reduction in Cost Per Acquisition, with the 18-24 age group showing the strongest engagement and conversion rates. A parallel user testing study on the MVP identified friction points and informed the next design iteration. Data sourced from Google Ads, GA4, and Looker Studio. Designer: Helena Cao. Researcher: Gavilán Longino. April 2025.
Campaign Performance Analysis — TableauQ4 ’24 to Q1 ’25 YourPeer Campaign Performance Analysis
Google Ads performance metrics, conversion trends, cost efficiency, and audience engagement data. Opens in Tableau Public.
Midtown Radio & Records: UX Research and Redesign
A UX research study applying a simplified Keystroke-Level Model (KLM) to evaluate and improve the click path efficiency of the Midtown Radio & Records website. Comparing the MVP against the final redesigned iteration, the study measured mental preparation time, scrolling, pointing, and click actions across participant sessions. The redesign resulted in a 172% increase in time spent on the website and sustained month-over-month sales growth from August 2021 through March 2022. A key finding: cooler color choices on the final site slowed cognitive processing in a way that increased engagement and improved conversion outcomes.
Before & After: MVP vs. Final Iteration
Policy Research: Congestion Pricing and CBDTPPH
Independent policy research applying data analysis to contested public policy questions. Work includes a critical examination of the NYC Central Business District Tolling Program and an independent data analysis challenging conventional framing of congestion pricing outcomes published on Medium.