Steven Mesquiti

PhD Student · Department of Psychology · Princeton University

CV Scholar GitHub Bluesky Email

I build and evaluate language-based tools for mental health assessment, asking when they work, for whom, and why.

I am a third-year PhD student in the Department of Psychology at Princeton University, advised by Erik Nook in the Logic of Emotion Lab. My research sits at the intersection of computer science, natural language processing, and clinical psychology, with a focus on the psychometric validity and generalizability of language-based measures.

My current work examines whether large language models predict depression and anxiety equivalently across racially and ethnically diverse samples, how the semantic content of first-person pronoun use in psychotherapy transcripts tracks symptom severity and treatment response, and whether domain-adapted clinical language models can reliably predict treatment outcomes.

Before Princeton, I was a Lab Manager in Emily Falk’s Communication Neuroscience Lab at the University of Pennsylvania. I have also had the privilege of collaborating with the Pennebaker Language Lab and researchers from the World Well-Being Project.

My research has been supported by the Princeton Data Driven Social Science Graduate Fellowship, the New Jersey Health Foundation, and Princeton Precision Health.

Selected Publications

Under Revision ·The Meaning of “I”: Semantic Self-Representation Tracks Internalizing Symptoms Within and Across Persons in Psychotherapy Mesquiti, S., Stade, E. C., Hull, T. D., & Nook, E. C. Preprint

2026 · Annual Review of Biomedical Data Science Large Language Models in Mental Health Treatment and Research Mesquiti, S. & Nook, E. C. DOI

2026 · Communications Psychology Predicting Psychological and Subjective Well-being through Language-based Assessments of Well-being Mesquiti, S., Cosme, D., Nook, E. C., Falk, E. B., & Burns, S. DOI

Full list on Google Scholar and in my CV.

Research Interests

  • Development of language-based clinical assessments
  • Language as a transdiagnostic index of psychological distress
  • Domain adaptation and evaluation of large language models for mental health
  • Computational psycholinguistics and large-scale text analysis

Education

  • Ph.D., Psychology — Princeton University 2024–present
  • M.A., Psychology — Princeton University 2024–2026
  • M.A., Psychological Research — Texas State University 2019–2022
  • B.A., Psychology (Minor: Spanish) — Southwestern University 2015–2019