About

Paul Perrin is a Professor of Data Science and Psychology at the University of Virginia and a core faculty member in the Clinical Psychology PhD Program. He is a clinical health and rehabilitation psychologist with a focus on social justice in disability and health, emphasizing disparities as a form of oppression and the role of academic and medical communities in addressing these issues. His research employs modern analytic techniques and community-based participatory research approaches to identify sources of disparities and potential solutions, particularly in medically underserved and minority populations with neurological conditions. Perrin's work encompasses cultural, familial, and international approaches to disability rehabilitation and adjustment, social determinants of health such as stigma and access to care, and social justice strategies to understand and dismantle oppression. He serves as Co-Director of the Polytrauma Rehabilitation Center Traumatic Brain Injury Model Systems Program at the Central Virginia Veterans Affairs Health Care System and is editor-in-chief of Rehabilitation Psychology. Perrin is dedicated to mentoring students across psychology, data science, and allied fields to become agents of social change, with an emphasis on disability and health. He teaches courses in multivariate statistics, behavioral research methodology, data science methodology, health disparities, health psychology, multicultural psychology, and community intervention.

Research topics

• Medicine
• Psychology
• Clinical psychology
• Psychiatry
• Physics

Selected publications

• Predictors of Anxiety Symptoms Over the 10 Years After TBI in Asian Americans/Pacific Islanders: A Model System Study

  Neurorehabilitation · 2026-04-09

  articleSenior authorCorresponding

  BackgroundAnxiety symptoms are common after traumatic brain injury (TBI), and its impact over time may be influenced by demographic and injury-related factors. Specifically, there is a need to characterize post-injury anxiety symptom trajectories after TBI in understudied racial/ethnic groups.ObjectiveTo examine demographic and injury-related predictors of anxiety trajectories over the first 10 years following TBI in an Asian American and Pacific Islander (AAPI) sample.MethodsParticipants were 272 individuals enrolled in the U.S. TBI Model System study. Participants self-identified as AAPI and completed the Generalized Anxiety Disorder-7 scale (GAD-7) at 1, 2, 5, and/or 10-year post-injury follow ups. Over the 10-year period, unconditional growth models assessed curvature of GAD-7 score trajectories, and hierarchical linear modeling (HLM) identified baseline predictors of those trajectories. Secondary HLMs assessed changes in anxiety trajectory slopes over time as a function of significant baseline predictors.ResultsClinically significant anxiety symptoms affected 10.9-13.4% of the participants across the four time points. Over 10 years, anxiety was characterized by a linear, flat/stable, low-level trajectory. A higher overall anxiety trajectory was predicted by lower educational attainment and pre-injury history (vs. no such history) of mental health treatment, with no significant interactions between these predictors over time.ConclusionsEducational background and pre-injury mental health history are risk factors in the development and maintenance of anxiety symptoms in AAPIs with TBI and should be considered in culturally responsive screening and intervention. The disaggregation of AAPI subgroups may be help reframe and target treatment efforts in this heterogeneous population.

  PublisherDOI

• Life satisfaction trajectories over the 10 years post-TBI among Asian Americans and Pacific Islanders: A model systems study.

  Rehabilitation Psychology · 2026-03-02

  articleOpen accessSenior author

  PURPOSE/OBJECTIVE: Asian Americans and Pacific Islanders (AAPIs) have been underrepresented in rehabilitation research, resulting in a limited understanding of the long-term effects of traumatic brain injury (TBI) in this population. The current study bridges this gap by examining demographic and injury-related factors predicting life satisfaction trajectories over the first 10 years after TBI in AAPIs. RESEARCH METHOD/DESIGN: The sample consisted of 381 AAPI-identifying individuals with moderate-to-severe TBI who were enrolled in the National Institute on Disability, Independent Living, and Rehabilitation Research-funded TBI Model Systems national study and had data for at least one Satisfaction with Life Scale total score at any time point (i.e., years 1, 2, 5, or 10). Hierarchical linear models examined baseline predictors of life satisfaction trajectories over the 10 years post-TBI and whether these predictors interacted with time. RESULTS: Overall, life satisfaction remained stable over time. Higher overall life satisfaction trajectories were seen among AAPIs who had been married at baseline, had higher educational attainment, had been employed at injury, had low annual earnings, and had no pre-TBI mental health treatment history. Nativity/country of birth was not a significant predictor of life satisfaction trajectories when controlling for other demographic factors. None of these predictors interacted with time, suggesting no differential change in life satisfaction as a function of these predictors. CONCLUSIONS/IMPLICATIONS: The findings provide valuable insights into culturally sensitive rehabilitation approaches for AAPIs with TBI by highlighting key risk and protective factors associated with long-term life satisfaction. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

  PublisherOA PDFDOI

• The Disability Identity Development Scale–Short Form: Development and validation in general and TGNB disability samples.

  Rehabilitation Psychology · 2026-02-12

  articleSenior author

  PURPOSE/OBJECTIVE: Disability identity is a multifaceted construct associated with important psychosocial outcomes, yet few psychometrically validated instruments exist to assess it comprehensively and parsimoniously. This study shortened the original Disability Identity Development Scale (DIDS) to create and validate the DIDS-Short Form to improve the feasibility and accessibility of measuring disability identity dimensions across diverse populations. RESEARCH METHOD/DESIGN: = 289). RESULTS: An exploratory factor analysis on DIDS responses from half of the general sample suggested the retention of 17 items and three subscales: Disability Identification, Contributions to the Disability Community, and Values and Advocacy. Two separate confirmatory factor analyses on the 17 items from the remaining half of the general sample and the TGNB sample both suggested good model fit. Total and subscale scores were positively correlated with life satisfaction in general and TGNB samples, particularly the Disability Identification and Contributions to the Disability Community subscales, and internal consistency was adequate across the total and subscales scores (αs = .75-.94). Conclusion/Implications The DIDS-Short Form demonstrates strong psychometric properties and offers a practical tool for assessing disability identity in both research and clinical contexts. Its brevity and conceptual clarity make it well suited for use with diverse disabled populations, including TGNB individuals. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

  PublisherDOI

• Identifying mechanisms of posttraumatic growth and recovery following moderate–severe traumatic brain injury: A mixed methods analysis.

  Psychological Trauma Theory Research Practice and Policy · 2026-03-12

  article

  OBJECTIVE: Traumatic brain injury (TBI) is a serious public health concern due to its potential lifelong consequences to physical, mental, and cognitive and social function (Haarbauer-Krupa et al., 2021). Comprehensive data highlighting the multidimensional and dynamic processes that underlie personal strengths and psychosocial growth following TBI are lacking. This study identified key mechanisms of posttraumatic growth after TBI by qualitatively mapping categories of posttraumatic growth development and quantitatively examining their relationships with growth and rehabilitation outcomes. METHOD: = 40) who received rehabilitation services for moderate-severe TBI completed subjective and objective assessments of global function and recovery, psychosocial health and coping strategies, and a free-response question about significant life changes following injury. Using a convergent parallel design, quantitative assessments were compared against categorical drivers of growth derived from content analysis. RESULTS: Qualitative analysis revealed five mechanism domains: presence of TBI sequelae; behavioral changes and goal-direct adaptations, changes in motivation and emotion, environmental and social contexts, contextual factors, and appraisal of themselves or from others. Joint assessment of data highlighted the role of event centrality, acceptance and active coping strategies, and contextual factors in the development of posttraumatic growth. CONCLUSIONS: Results demonstrate the need to incorporate sociocognitive processes absent in rehabilitation frameworks to account for the scientific development of psychosocial growth and healing in the context of TBI. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

  PublisherDOI

• Neuroscience beliefs in education among teachers in Argentina and other Latin American countries: predictors and comparisons to experts’ perspectives

  Trends in Neuroscience and Education · 2026-01-30 · 1 citations

  articleOpen access
  PublisherOA PDFDOI

• Disability level and visibility: Associations with unmet academic accommodation needs and attitudes toward requesting accommodations

  PLoS ONE · 2026-02-09

  articleOpen accessSenior author

  PURPOSE/OBJECTIVE: Despite legal mandates to provide appropriate accommodations for students with disabilities in the U.S., many report gaps between what they need and receive. This study examined the role of disability level, visibility, and demographic factors in predicting unmet academic accommodation needs and attitudes toward requesting accommodations. RESEARCH METHOD/DESIGN: A sample of 409 adults who had had disabilities during their school years and who still had them completed an online survey assessing their current disability level, current disability visibility (invisible, semi-visible, or visible), unmet academic accommodation needs across all levels of schooling in aggregate, and current attitudes toward requesting academic accommodations. RESULTS: Individuals with invisible disabilities in comparison to those with semi-visible or visible disabilities reported unmet needs for having a quiet or sensory room, extended time to take tests and exams, sensory objects (e.g., fidget toys), and an Individualized Education Plan. However, those with visible and semi-visible disabilities reported unmet need for having an educational assistant or tutor, recording equipment or a portable notetaking device, a modified or adapted course curriculum, and a computer, laptop or tablet with specialized software or apps. After controlling for disability severity and demographic variables, individuals with more visible disabilities had lower unmet academic accommodation needs compared to those with an invisible disability, as well as more positive attitudes toward requesting accommodations. CONCLUSION/IMPLICATIONS: Assisting students with disabilities-especially those with invisible disabilities-may enhance disabled students' experience of academic accommodations and empower them to advocate when those needs are unmet.

  PublisherDOI

• Handling Missing Data in Longitudinal Rehabilitation Research: A Methodological Demonstration With Functional Trajectories of Older Adults With TBI

  Journal of Head Trauma Rehabilitation · 2026-05-18

  articleSenior author

  OBJECTIVE: This study compared the effects of 3 different approaches to handling missing data (listwise deletion of participants with missing data, mean imputation, and full information maximum likelihood [FIML]) when predicting functional independence trajectories over 10 years in older adults after traumatic brain injury (TBI). SETTING: Twenty-three TBI Model Systems (TBIMS) inpatient rehabilitation facilities in the United States. PARTICIPANTS: Adults who sustained a complicated mild, moderate, or severe TBI at age 60 years or older and needed inpatient rehabilitation. They had to meet all eligibility criteria and have one or more functional independence measure (FIM) scores at 1, 2, 5, or 10 years post-TBI from the TBIMS national database. DESIGN: Retrospective analysis of observational data using hierarchical linear models. MAIN MEASURES: FIM total scores at 1, 2, 5, and 10 years post-TBI. RESULTS: Different missing data approaches led to drastically different findings. Model comparisons supported a quadratic effect of time only in the listwise deletion model and found no other significant predictors. Linear trajectories were found in the mean imputation and FIML models. For both these models, older age, underrepresented minority status, unemployment at injury, longer posttraumatic amnesia duration, and pre-injury limitations all predicted lower overall FIM trajectories. However, when compared with the mean imputation model, the FIML-estimated b-weights were larger with smaller P-values. Years of education significantly predicted higher overall FIM trajectories in the mean imputation model but not the FIML model, likely because of the artificial shrinking of the estimated b-weight standard errors in mean imputation. History of mental health treatment predicted lower FIM trajectories only in the FIML model. CONCLUSIONS: These findings show that it is critical to use appropriate modern methods to handle missing data because the method can affect outcome trajectory shape and identification of relevant predictor variables. Using older methods for handling missing data, such as listwise deletion, greatly reduces predictive ability, resulting in less generalizability and imprecision in longitudinal rehabilitation research.

  PublisherDOI

• Mapping the structure of internalized ableism: psychometric network analysis comparing cisgender and TGNB disabled samples

  International Journal of Transgender Health · 2026-01-02

  articleSenior authorCorresponding
  PublisherDOI

• Comparing Acute and 1-Year Outcomes Between Fall- and Motor Vehicle–Related Traumatic Brain Injury

  Neurology · 2026-03-10

  articleOpen access

  BACKGROUND AND OBJECTIVES: Traumatic brain injury (TBI) mechanisms are often grouped together in research. Differences in acute and long-term outcomes across mechanisms of injury (MOIs) remain unclear, partly because of confounding by age. Modeling MOI-specific effects can inform clinical triage and prognostication. We examined the relationship between motor vehicle accidents (MVAs) vs falls, the 2 most common MOIs, and acute and 1-year post-injury outcomes, after rigorous control of demographic and preinjury personal factors. METHODS: Data were analyzed from individuals with moderate-to-severe TBI requiring inpatient rehabilitation from the TBI Model Systems National Database, a multicenter prospective longitudinal cohort study. The analytic sample was restricted to individuals aged 16-79 years with an MOI due to MVA or fall occurring between April 2010 and January 2023. We used inverse probability of treatment weighting, based on propensity scores, to adjust for 14 demographic and preinjury personal characteristics and estimate the causal effect of MOI on acute and 1-year outcomes after TBI. Acute hospital and rehabilitation outcomes included the following: Glasgow Coma Scale (GCS), sedation, intubation, post-traumatic amnesia duration, time to follow commands (TFC), length of hospital stay (LOS), and Functional Independence Measure (FIM) cognitive and motor scores. One-year outcomes included the following: Disability Rating Scale and Participation Assessment with Recombined Tools Objective. RESULTS: = 0.014). At 1 year after injury, disability levels and community participation did not differ. DISCUSSION: MVA-related TBI was associated with worse acute outcomes. However, by 1 year after injury, disability level and community participation do not differ. This work highlights novel findings in short-term and long-term outcomes after falls and MVAs, the leading TBI causes, which are not explained by confounders such as age. Findings may not generalize beyond patients receiving inpatient rehabilitation for TBI.

  PublisherOA PDFDOI

• Are Artificial Intelligence, Machine Learning, and Data Science all the Same? The Public Sure Thinks So, and Why That’s Problematic

  Digital Society · 2026-04-01

  articleOpen access1st authorCorresponding

  Public discourse increasingly treats artificial intelligence (AI), machine learning (ML), and data science as interchangeable concepts, obscuring critical distinctions among their histories, goals, and methodological foundations. This article examines how this terminological conflation emerged, tracing the conceptual evolution of AI, ML, and data science and clarifying their overlapping yet distinct domains. It argues that linguistic imprecision shapes research agendas, funding priorities, regulatory frameworks, ethical debates, and public perceptions of technological capability and risk. Drawing on historical, technical, and sociotechnical perspectives, the article shows how AI has become an umbrella term encompassing diverse computational practices, while ML has driven much of AI’s recent practical success, and data science has reframed scientific inquiry around large-scale data analysis. The widespread collapse of these distinctions—reinforced by media narratives, industry marketing, and academic incentives—has contributed to inflated expectations, misplaced fears, and blurred accountability in the governance of algorithmic systems. The article contends that terminological precision is essential for responsible innovation and democratic oversight of emerging technologies. By articulating clearer conceptual boundaries among AI, ML, and data science, scholars, policymakers, and practitioners can foster more transparent communication, more coherent regulation, and more informed public understanding of the technologies increasingly shaping social, economic, and political life.

  PublisherOA PDFDOI

Recent grants

• A Culturally Sensitive Intervention for TBI Caregivers in Latin America

  NIH · $243k · 2014–2017

Frequent coauthors

• Daniel W. Klyce

  719 shared

• Shannon B. Juengst

  Southwestern Medical Center

  691 shared

• Janet P. Niemeier

  503 shared

• Ross Zafonte

  Brigham and Women's Hospital

  500 shared

• Amanda R. Rabinowitz

  Walker (United States)

  498 shared

• Kelli W. Gary

  Virginia Commonwealth University

  496 shared

• Flora M. Hammond

  University of South Florida

  490 shared

• Amy K. Wagner

  University of Pittsburgh

  488 shared

Education

• PhD, Counseling Psychology, Psychology

  University of Florida

  2011