Research

Growing up in Pavuna, Rio de Janeiro, I witnessed a relentless fight for the fundamental recognition of being “gente”, that is, to have the rights of personhood, to belong. My research agenda examines the complex dynamics of communities, exploring how social cohesion and fragmentation are shaped by social interactions and the physical and digital spaces where they take place. As a computational sociologist and demographer, I leverage quantitative, qualitative, and spatial methods to investigate how groups cultivate mutual support across neighborhoods in Rio de Janeiro and Chicago as well as online communities on Reddit and Twitter.

The Mechanics of Urban Social Capital

While classical urban sociology emphasizes how structural disadvantage erodes trust over time, my first line of research examines how dynamic, everyday interactions—daily mobility, real-time street conditions, familiarity with people and places, and shared experiences of hardship—cultivate conditions for trust and social cohesion.

  • “Turning a corner: Real-time experiences of informal control in public places” (with Erin York Cornwell). Published in Urban Studies.
    Investigating real-time neighborhood appraisals, we find that structural disadvantage is not directly associated with in-situ evaluations of informal social control. Rather, emergent micro-level conditions—such as physical disorder, friendly greetings, and the presence of familiar others (especially when individuals are farther from home)—drive how people evaluate informal control in public spaces.
    Journal Link
  • “Reading the Street: The Contextual Foundations of Neighborhood Trust” (Under Review).
    Pairing traditional surveys with repeated ecological momentary assessments (EMAs) from the Chicago Health and Activity Space in Real-Time (CHART) study, I use multilevel modeling and spatial statistics to examine localized trust. I demonstrate that neighborhood trust operates across two distinct dimensions: a reflective baseline and an experiential, in-situ reading activated through daily navigation. Real-time trust varies substantially within and across neighborhoods and responds far more strongly to immediate social cues—like prosocial behavior and familiar faces—than to static structural disadvantage.
    [Draft Available Upon Request]
  • “Perfect Storm: Uncovering the Latin American Cohesion Paradox” (In Preparation).
    Drawing on in-depth qualitative interviews in a disadvantaged Rio de Janeiro neighborhood facing poverty, violence, and recurring floods, I analyze how residents maintain social cohesion under acute stress. While everyday life features internal boundaries between “helpful” and “unhelpful” neighbors, environmental disasters dissolve these divisions into community-wide mutual protection—forging social bonds that persist well beyond the immediate recovery period.

Trust and Community Trajectories

My second line of research examines how localized perceptions of trust and mobility networks shape long-term neighborhood trajectories in crime, public health, and urban inequality.

  • “Residents Know Best: Real-Time Social Capital and Neighborhood Crime” (Working Paper).
    Linking real-time EMA trust assessments (2018–2019) from CHART with Chicago Data Portal crime records (2023), I compare how resident versus visitor ratings predict future neighborhood crime. While visitor assessments carry limited predictive weight, tracts deemed trustworthy by residents experience significant subsequent declines in violent, property, and vice crime, with the strongest crime reductions occurring where residents rate their neighborhood as more trustworthy than visitors do. In an ongoing extension, I examine how these divergent resident-visitor trust perceptions also predict long-term community health outcomes, including rates of coronary heart disease, diabetes, and stroke.
    Poster | Draft Available Upon Request
  • “CLUSTERTrust: Mapping Urban Trust Networks and Inequality” (Work in Progress).
    Scaling up to the metropolitan network level, this project maps the connectedness of Chicago block groups based on co-rating and co-visitation patterns. I find that high-trust neighborhoods are densely connected socially through mobility flows despite being geographically dispersed, exhibiting strong community partitioning driven by shared amenities and social ties.
    Poster 

Trust and Connectedness in Digital Space

Extending neighborhood concepts to “unplaced” digital spaces, I use Community-Based Participatory Research (CBPR) at the Citizens and Technology (CAT) Lab to examine how platform governance and technological disruption affect online cohesion.

  • Generative AI, Algorithmic Changes, and Community Trust on Reddit (Two Manuscripts in Progress with CAT Lab).
    Partnering directly with moderators across four major subreddits, we co-designed a survey to address community governance challenges and translated complex statistical models back to moderator teams. Across two papers, we demonstrate (1) how higher perceived prevalence of AI-generated content erodes user trust in information and community connectedness amidst changing LLM platform rules, and (2) how users’ perceptions of a subreddit’s core social function (e.g., public sphere vs. third space) shape norm enforcement and conflict.
    Information on the survey and community reports

Selected Predoctoral & Early Research

  • Dreaming within Limits: Intergenerational Socioeconomic Mobility of Brazilian Women (B.A. Thesis & Hanna Holborn Gray Fellowship, Bryn Mawr College).
    Combined qualitative interviews in Pavuna, Rio de Janeiro, with Two-Sample Instrumental Variables (TSIV) estimation using Brazil’s National Household Survey (PNAD) to measure mother–daughter intergenerational income mobility, highlighting the necessity of centering female heads of household in conditional cash transfer and educational policy design. 
    Read Full Project Archive →