Zhanna Alexeyeva, Health Policy | Virginia Tech

"Quitting Motives, Strategies, and Support Among Lower-Income Adults who Smoke Cigarettes in Virginia"

Advisor: Dr. Morgan Snell

Co-Authors: R. Imran

Smoking remains a public health concern, but little is known about how people with lower income and education try to quit or what barriers they face. This study examines their motivations, strategies, and views on alternative tobacco products to inform more tailored cessation efforts. Adults 21+ who smoked ≥10 cigarettes per day were recruited through Craigslist and an IRB-approved registry. Six participants with lower income and education completed semi-structured interviews about their reasons and strategies for quitting, experiences with tobacco products, and needed support; interviews are ongoing. Interview transcripts informed a codebook; data were coded in NVivo software.  All participants were non-Hispanic men (mean age 36). They wanted to quit mainly for health reasons (breathlessness, declining fitness), dislike of smoke/smell, and concerns about work, family, and image. Quit approaches ranged from cold turkey to gradual reduction, sometimes using supposedly less harmful products short term. E-cigarettes, nicotine pouches, and cigars were tried for curiosity, convenience, cost, or peer advice, not always to quit. Many had quit temporarily—often with vaping—but relapsed and regretted restarting. They wanted clear evidence on safer options, peer support, help setting realistic goals, and stable, supportive life circumstances. Health, well-being, and social factors motivated quitting. Tailoring cessation to individual motivations, alternative product use, and available resources may better support lower-income smokers.

Zhanna Alexeyeva is a second-year Ph.D. student in Health Care Policy and Research at Virginia Commonwealth University. She has a broad interest in public health, particularly in addiction and the health economic factors that influence it. Her specific research interests focus on combustible and non-combustible nicotine products and their associated health harms.

Pacifique Dusabeyezu, Health Policy | Brandeis University

"Work-related Musculoskeletal Pain Among Factory Workers in Kigali: Prevalence and  Coping Strategies"

Advisor: Dr. Sagahutu Jean Baptiste

Co-Authors: K. Joanitah, S. Jean Baptiste

Work-related musculoskeletal pain (MSP)—pain affecting muscles, joints, and tendons—is a common problem among factory workers who perform repetitive or physically demanding tasks. Such pain can lower productivity, increase absenteeism, and affect overall well-being. This study examined how widespread musculoskeletal pain is among factory workers in Kigali and explored the strategies they use to cope with it. A cross-sectional, quantitative study was conducted with 148 factory workers (29 women and 119 men) from two factories. Participants completed a structured questionnaire that collected demographic information, identified pain in nine body regions using the Modified Nordic Musculoskeletal Questionnaire, and assessed coping strategies through the Coping Strategies Questionnaire. Data were analyzed using descriptive statistics with SPSS version 25. Nearly half of the workers (49.3%) reported experiencing musculoskeletal pain. The lower back (60.8%) and shoulders (48.0%) were the most commonly affected areas, followed by the neck (41.2%), upper back (36.5%), knees (39.2%), and ankles or feet (26.4%). To manage their pain, most workers relied on rest (62.8%) or medication (45.9%), while fewer reported using ergonomic adjustments or physical therapy. These findings highlight the significant burden of musculoskeletal pain among factory workers and the need for workplace health interventions. Reducing dependence on medication and promoting preventive measures, ergonomic practices, and rehabilitation support could improve workers’ health and productivity.

Pacifique Dusabeyezu is an M.S. student in Global Health Policy and Management at The Heller School for Social Policy and Management at Brandeis University. His research focuses on health systems strengthening, occupational health, rehabilitation and disability policy, and the use of data analytics to inform equitable health workforce and service delivery. He is interested in bridging clinical rehabilitation practice with evidence-based global health policy in low- and middle-income settings.

Amber Grady| Bellevue University

"From Anti-Partisanship to Anti-Personhood: The Relationships Between Political Animosity, Dehumanization, and Self-Censorship"

Advisor: Dr. Jerome Lewis

In the United States, rising tensions between political parties has led to animosity. This animosity has led many Americans to view political opponents as less than human. Because Americans are aware of these tensions, they may avoid social interactions to prevent being dehumanized. Past research shows that individuals dehumanize political opponents based on perceptions of conflicting moral values between political parties. Less is known about how people respond to perceived dehumanization by political opponents. This gap limits our understanding of how clashes in political identity may lead to perceptions of political animosity through perceived dehumanization. Thus, we conducted a study (n = 180) to explore the relationships between these variables. We found that perceived dehumanization, discrimination, self-censorship, and warmth and competence were correlated. Dehumanization scores correlated with discrimination (r = 0.237), self-censorship scores (r = 0.147), and warmth and competence scores (r = -0.571). When correlations were conducted within each political party, the strength of the relationship between variables changed. Previous research shows these variables may operate differently for Republicans, Democrats, and Independents. Our results will be shared within the framework of recent findings regarding political animosity and self-censorship.

Amber Grady is a recent MBA graduate from Bellevue University. She holds a B.S. in Psychology and has been a research assistant for the Victim Role Research Group in Bellevue University’s Psychology Department since 2021. Her work investigates how perceived political animosity may influence self-censorship behaviors.

Najib Ullah, History | James Madison University

"Political Parties, Ideological Contestation and the Role of History"

This study attempts to examine how members of different political parties, having ideological differences, contest history to make, argue for, and justify their current political stand and its existence in Pakistan. It seeks to theorize how history is contested thus instrumentalized by these political parties in their forums of indoctrination to argue in support of and justify their current politics and justify their ideology and existence. This qualitative study is based on the data collected from about 36 respondents/members of different political parties, through semi-structured in-depth interviews and purposive sampling who are serving as opinion makers, and are well-versed in the party’s ideological understanding and politics. The respondents were primarily based in Swat (researcher’s home district) allowing him to play an “interpretive role in meaning-making” to theorize the question under investigation. The paper adopted Constructivist Grounded Theory to explore how political parties construct, shape their members’ understanding (questioning epistemology) and theorize how history plays the role of an instrument for these parties to argue in support of and justify their position, ideology, and existence. The findings suggest that these political parties “construct” the understanding of their members and do contest the history of Pakistan to instrumentalize it for their current political stand, ideological position, and existence of the party. Each of the parties has its own “story,” heroes, and theory of the creation of Pakistan, which serves its current objectives.

Najibullah is a first year graduate student in the History Department at James Madison University. His areas of interest include contemporary Pakistan, history of the British Raj, colonial knowledge and political structures and Post-Colonial Studies. He is currently working on the British colonial experience in the former North West Frontier region of British India. 

Tabassum Islam Tamanna, Genocide and Mass Atrocity Prevention | Binghamton University 

"A Case Study on Doe v. Meta; Facebook's Impunity for Inciting Genocide against the Rohingya Minorities"

Doe v. Meta is an unprecedented class action lawsuit that was filed to hold Facebook (Meta) responsible for indemnifying the Rohingya Muslim minority for its fostering of ethnic atrocities in Myanmar. It was noted that Facebook's main business strategy, behavioral advertisements, contributed to the 2017 Rohingya atrocities by escalating the conflict at large. In 2018, the United Nations human rights inspectors warned that Facebook was encouraging hate speech and violent attacks against Myanmar's Muslim minorities. Amnesty International, in one of their studies, added that content that incited violence, racism, and prejudice against the Rohingyas was deliberately amplified and pushed on Facebook by Meta's tracking-based business model, which it claims still feeds off provocative, conflictual, and damaging content. Facebook is implicated in this dynamic for deliberately instigating violence against the Rohingyas due to its emphasis on engagement for profit. Therefore, the main focus of this paper will be on how Meta promoted anti-Rohingya content and how Facebook's seeming incapacity to control online hate speech and misinformation has led to a genocide. Social media's impunity in terms of its inability to regulate the narrative is alarming, and in order to prevent its business model from inciting further violence in the future, social media platforms must be held accountable.

Tabassum Islam Tamanna is a graduate student  at Binghamton University and a prospective master's student at the State University of New York at Binghamton (Binghamton University) studying genocide and mass atrocity prevention. She was previously employed at Manarat International University as a law lecturer.  

Sarah Allen, Marine Science

"Who's the Dad? Gracilaria vermiculophylla Paternity Analyses"

Advisor: Dr. Stacy Krueger-Hadfield

The reproductive mode partitions genetic variation within and among populations. Many eukaryotes undergo both sexual and clonal reproduction but quantifying sexual and clonal rates is challenging. Moreover, life cycle variation, such as the haploid-diploid life cycle found in macroalgae, creates further complexities. For example, the proportion of haploid individuals in a population has a disproportionate influence on genetic diversity. Reproductive system variation has been studied across the native and non-native range of the red alga Gracilaria vermiculophylla in which diploid sporophytes alternate with haploid gametophytes. In soft bottom habitats without hard substrate, the life cycle is disrupted whereby gametophytes are absent and sporophytes undergo thallus fragmentation. Conversely, in hard bottom habitats, reproduction is sexual, although variation occurs in outcrossing and selfing rates. Our current knowledge relies on indirect methods, such as genotype frequencies. Paternity analyses, on the other hand, directly quantify the number of fathers and relatedness among mating pairs. In red algae, the zygote is retained on the female and develops into a structure called the cystocarp. Gametophytes are haploid and the contents of the cystocarp are diploid because of the fusion of maternal and paternal genes. Using microsatellites, we will genotype cystocarps and directly assess selfing versus outcrossing rates in populations along the Eastern Shore of Virginia, providing a better understanding of the reproductive mode variation in G. vermiculophylla.

Sarah Allen is a first year M.S. student at William & Mary's Batten School and Virginia Institute of Marine Science. Her broad interests include ecology, evolution, and population genetics. She is currently studying reproductive mode variation in a non-native red alga species along the Eastern Shore of VA.

Elena Hoang, Marine Science

"Mapping Restored Reef Complexity and Effects on Juvenile Oyster Growth Using High-Resolution 3D Scanning "

Advisor: Dr. Rochelle Seitz

Habitat structure is a foundational driver of ecological processes on oyster reefs, influencing recruitment, growth, and the provision of ecosystem services. As artificial reef structures become increasingly central to restoration efforts in Chesapeake Bay, understanding how reef structural complexity contributes to restoration success and long-term stability is critical. Oyster planting is often used alongside the construction of artificial reefs to accelerate recovery, yet how habitat structure and seeding practices jointly influence the growth and reef-building potential of juvenile oysters (Crassostrea virginica) remains unclear. Although advanced technologies can quantify structural complexity, their use in restoration ecology has been limited. This study uses high-resolution 3D scanning to examine how artificial reef structure and seeded-spat density affect early oyster growth and morphology. Structural complexity was quantified using metrics such as rugosity, interstitial space, and surface area-to-volume ratio from 3D models of three reef types (C-Dome, X-Reef, and Granite) at two York River, Virginia sites. Experimental seeding of spat-on-shell was conducted across four density levels (High: 100+, Medium: 50–75, Low: <25, and Natural Settlement: 0 spat/shell). Results show that X-Reefs exhibited the greatest rugosity, and shells seeded at the lowest density and placed on X-Reefs had the highest growth. Findings will inform restoration strategies by identifying reef designs and seeding densities that enhance reef-building efficiency and ecological resilience.

Elena Hoang is a second year M.S. student in Coastal & Marine Sciences at William & Mary and the Virginia Institute of Marine Science (VIMS). She holds a B.A. in Marine Biology and Environmental Studies from Rollins College. Her research interests include community ecology and oyster reef restoration, and her thesis uses 3D scanning to study how reef structural complexity affects juvenile oyster growth and survival. 

Ada Li, Chemistry

"Investigation of the Pqse-Rhlr Interaction as a Target for Inhibiting Quorum Sensing in Pseudomonas Aeruginosa"

Advisor: Dr. Isabelle Taylor

Co-Authors: K. Smith

Multi-drug-resistant bacterial infections are a growing problem in hospitals, especially for patients with weakened immune systems, such as those undergoing chemotherapy. One major concern is Pseudomonas aeruginosa, a pathogen that can resist many antibiotics and cause serious illness. It uses a communication system called quorum sensing to coordinate group behaviors that enhance its pathogenicity. The system depends on signaling molecules that bacteria release into the environment to “talk” to each other. When enough of these molecules are present in the surrounding environment, this activates genes controlling group behaviors such as toxin production and biofilm formation. A key part of the quorum sensing system involves the protein PqsE, which works with another protein RhlR. When these two proteins interact with each other, this activates disease-related genes. Disrupting this interaction presents an attractive therapeutic avenue for treating infections by inhibiting virulence rather than killing the bacteria. By not placing a selective pressure on bacteria, spurring the evolution of resistant mutations, this strategy could avoid contributing to the rise of antibiotic resistance. In the Taylor Lab, we developed a model bacterial system in E. coli to identify molecules that interfere with the PqsE–RhlR interaction. We screened a library of ~800 FDA-approved compounds and had identified several candidates that may disrupt the PqsE-RhlR interaction. These candidate molecules are now being further studied for their potential as anti-virulence agents in P. aeruginosa.

Ada Li is a second-year master’s candidate in the Chemistry Department at William & Mary. Her research focuses on biochemistry, specifically the quorum sensing system of Pseudomonas aeruginosa. She earned a B.S. in Chemistry and Critical and Creative Media from Wake Forest University. 

Alexander Raffetto, Biology

"The Role of Subgenome Dominance In Monkeyflower Speciation Dynamics"

Advisor: Dr. Joshua Puzey

Flowering plants are essential for human society, incredibly diverse, and dominate significant portions of terrestrial and aquatic ecosystems. Genome duplications and hybridization between species are major factors that help explain their worldwide success and array of shapes, colors and forms. This project seeks to explore precise mechanisms behind how hybridization and gene duplications impact the evolution of flowering plants. Monkeyflowers (genus Mimulus) are a colorful group of flowering plants that make an excellent system for studying how plants diversify. We have sequenced the entire genomes of three similar monkeyflowers from Chile, which have undergone genome duplications and  frequently hybridize in their natural range. The goal of this project is to scan these genomes for characteristic patterns of evolution, and ask if factors that influence Mimulus evolution are restricted to one of the two duplicated genomes. The outcome will give insights into some of the most evolutionarily important processes driving flowering plant diversity. 

Alexander Raffetto is a first year M.S student in the Biology Department at William & Mary. He is broadly interested in plant science, with specific interests in evolutionary and population ecology and genetics. His current research employs various bioinformatics techniques to investigate a phenomenon known as subgenome dominance in a group of plants called Monkeyflowers. 

Sophia Tearman, Marine Science

"Life Cycle, Interrupted: the Ecological Genetics of the Invasion of Soft Bottom Habitats"

Advisor: Dr. Stacy A. Krueger-Hadfield

Life cycle variation exists across the tree of life, yet we lack empirical data to understand how this diversity is generated and maintained. Life cycles are the result of sexual reproduction in which there are two phases due to meiosis and fertilization. Seaweeds are promising organisms to address long-standing questions in life cycle evolution. The red seaweed Gracilaria vermiculophylla has a complex life cycle in which two isomorphic phases alternate with one another. This species has invaded near shore marine habitats throughout the Northern Hemisphere during which the sexual phase – the gametosporophyte – is lost from its life cycle. We will use a combination of genotyping and phenotyping to explore life cycle dynamics in this seaweed. We will explore reproductive mode variation, and the consequences of asexual reproduction, using single nucleotide polymorphisms (SNPs), thereby increasing the number of sampled regions across the genome. The different phases will be subjected to salinity and temperature stress experiments and herbivory assays to investigate differences between phases and sexes. Together, these genetic and ecological approaches will enable a holistic view of the mechanisms underlying the maintenance of this complex life cycle in G. vermiculophylla. Moreover, these data will contribute to our understanding of the evolutionary maintenance of sex across eukaryotes by expanding the taxa for which we have ecological genetic data. 

Sophia Tearman is a first year Ph.D. student in Marine Science at William & Mary's Batten School and Virginia Institute of Marine Science. Her research areas include evolutionary ecology and phycology. Her research will address the maintenance of sexual reproduction and life cycle variation in eukaryotes.

Morgan Wood, Biology

"Opportunistic Fungal Pathogens: The Role of Amoeba Predation in Incidental Virulence"

Advisor: Dr. Helen Murphy

The incidence of human fungal infections have been rapidly increasing in recent years. Interestingly, most of these infections arise from opportunistic pathogens, many of which are environmental fungi that have acquired the ability to propagate in humans. Where and how these pathogenic fungi obtain the ability to exploit humans is an open question. The Amoeba Predator-Animal Fungal Virulence (AP-FAV) hypothesis posits that amoeba predation selects for dual-use traits. Amoeba phagocytosis shares functions similarly to human macrophages, indicating potential evolutionary implications. Fungi may evolve dual-use traits that provide both predation resistance and macrophage evasion. This project will perform a long-term evolution experiment directly testing the AP-FAV hypothesis by co-culturing Dictyostelium discoideum (amoeba) and Saccharomyces cerevisiae (yeast) for 240 generations. Yeast populations will be assessed for predation and virulence rates, predicting decreased predation and increased virulence in final, evolved populations. Ideally, a better understanding of a mechanism underlying opportunistic pathogenicity in fungi will be concluded.

Morgan Wood is a first year master's candidate in Biology at William & Mary. Her research areas include environmental pathogens, fungal virulence, and immunology. Her project focuses on how environmental selection pressures (e.g. amoeba predation) drive fungal pathogenicity in human hosts. She holds a B.S. in Biology from William & Mary. 

Spencer Allison, Education

"How School Counselors Support Students of Color in Public K-12 Schools with Predominantly White Student Populations"

Advisor: Dr. Bianca Augustine 

Students of color are a diverse group of students with many backgrounds, experiences, and cultures. Students of color have unique and varying needs within a school setting, however, they may have the collective experience of racial marginalization. This experience may be exacerbated in schools with a predominantly white student population and little representation. School counselors support all students holistically and their academic, social/emotional, and career development. School counselors are equipped to mitigate the impact of marginalization and work towards dismantling policies and practices in school settings that result in inequities. A qualitative study will seek to explore the following question: what are the perspectives of school counselors who support students of color in predominantly white public schools? Individual and focus group interviews with school counselors working in these settings will be thematically analyzed. Themes will be identified and reported related to school counselors’ perspectives, supports, interventions, and advocacy for students of color and their academic, social/emotional, and career development. Implications for both school counselors and educators will be explored to promote equity and justice within school settings with predominantly white school settings. 

Spencer Allison is a second year Ph.D. student in the Counselor Education & Supervision program at William & Mary and school counselor at a local private school. His research areas relate to school counseling, specifically equitable counseling practices, wellness, and preparation. He holds a B.S.Ed. (Science Ed), an M.Ed. (Prof. Counseling), and an Ed.S. (Prof. School Counseling) from the University of Georgia.

Kim Hughes, Counselor Education

"LEAN Model: Interventions to Address Socioeconomic Disparities in the Treatment of Eating Disorders"

Advisor: Dr. Stephanie Dorais

Co-authors: A. Kauffman

The longstanding perception that eating disorders (EDs) primarily affect individuals of higher socioeconomic status (SES) has obscured clinical conceptualizations and care. An increasing body of research underscores that EDs are not only experienced across all SES strata, but ED-related symptomology presents at a higher rate among individuals who have experienced socioeconomic deprivation. ED-related risk further increases for individuals with multiple marginalized identities. The erroneous belief that EDs are primarily a “disease of [the] affluent” undercuts the experiences of individuals most affected by ED symptomatology, and mitigates access to often-costly care. Thus, the individuals most affected by EDs are often the first forgotten, leaving people in lower SES strata overlooked and undersupported. The LEAN Model is a framework for clinicians to use in supporting individuals of lower SES in navigating ED symptomology. Integrating aspects of liberation psychology (L), equity (E), advocacy (A), and narrative therapy (N), this framework is appropriate in both counseling and community settings. Collectively, the steps within this framework support an unlearning of ED stereotypes related to SES while enhancing accessibility to ED-related care. Specific opportunities for equity and advocacy will be outlined to increase access to care by reducing prejudice and increasing awareness. Lastly, the integration of narrative therapy supports the development of new, affirming storylines in which individuals of lower SES do not feel negated or forgotten. 

Kim Hughes is a second year doctoral student at William & Mary, pursuing her Ph.D. in Counselor Education. Kim’s research and clinical interests center around anti-fat bias and body dissatisfaction, with an emphasis on social justice. Kim has presented at the state and national levels on anti-fat bias, sexual desire and temperament, body presentation and disordered eating, and race-based trauma. 

Alexis Kauffman, Counselor Education

"Community of Comparison: Addressing the Influence of Social Media on Maladaptive Body Image "

Advisor: Dr. Stephanie Dorais 

Co-Authors: K. Hughes

The proliferation of social media has added a new dimension of consideration to the treatment of eating disorders (EDs) and body image. At the onset of the internet, the ED community expanded with the advent of “pro-ana” or “pro-ED” websites, promoting tips and strategies for maintaining ED-behaviors. Social media has perpetuated physical evaluation through AI-generated images that promote unrealistic body images and comparative commentary. A building body of research underscores how the explosion of digital engagement coincides with growing body discontentment. In fact, studies suggest that it is not only considered “normal” for individuals to dislike their physical form, but is a mechanism for connection. For clinicians, these trends present a critical and urgent need to address the influence of social media on maladaptive body image. This poster presentation outlines empirical insights related to social media and body image, while also extending accessible interventions to support positive self-image in an increasingly digital landscape. Instead of resisting social media, novel interventions acknowledge the inevitability of digital exposure and include adaptive digital skills, such as reality-checking and media literacy specific to body image. Grounded in positive and liberation psychology, this presentation provides tangible, pro-social tactics for body image management in an increasingly digital world. 

Alexis Kauffman is a second year M.Ed. clinical mental health counseling student in the Counselor Education Department at William & Mary. Her research and clinical focus areas include eating disorders, body image, and perinatal mental health. She holds a B.A. from Penn State University. 

*Jennifer Powers, Strategic Leadership | James Madison University

"Breaking Barriers: Predicting Leadership Role Attainment Among Women and Gender-Expansive Disadvantaged College Students"

Advisor: Dr. Nara Yoon

*GRS Visiting Student Award for Excellence in Scholarship in the Humanities and Humanistic Social Sciences

Persistent disparities in leadership role attainment continue to affect women and gender-expansive college students, particularly those from first-generation and low-income backgrounds. Despite growing institutional commitments to equity, systemic barriers rooted in historical and structural inequities continue to limit access to positional leadership roles in higher education. This study investigates five key predictors: first-generation status, socioeconomic background, observational awareness, leader self-efficacy, and socially responsible leadership, to assess their influence on leadership outcomes. Grounded in intersectionality, feminist theory, cultural and social capital theory, and the Social Change Model, the study applies a gender-restricted, equity-centered lens to examine how identity-based and leadership-related factors intersect to shape students' leadership trajectories. Logistic regression analyses reveal that observational awareness, leader self-efficacy, and socially responsible leadership significantly predict formal leadership attainment, while first-generation and low socioeconomic status are negatively associated with access. Interaction effects suggest that high observational awareness can partially offset structural disadvantage. Findings aim to inform more inclusive leadership development practices and support institutional efforts to dismantle exclusionary norms, cultivate equity, and reimagine leadership pathways in higher education.

Jennifer L. Powers is a Ph.D. candidate in Strategic Leadership at James Madison University. Her dissertation examines how structural inequities and intersecting identities shape leadership attainment in higher education for women and gender‑expansive students. She seeks to advance inclusive leadership pathways by integrating theory and practice with servant‑leadership principles of empathy, awareness, trust, and growth.

Andrew Singh, Business Analytics | Georgetown University

"Data-Driven Approaches to Educational Staffing Equity in NYC Public Schools"

Advisor: Dr. Emisa Nategh

New York City Public Schools (NYPS) face persistent staffing shortages in two critical areas: certified STEM teachers and mental health professionals. These shortages disproportionately affect schools serving economically disadvantaged and racially diverse student populations, limiting access to rigorous academic opportunities and adequate emotional support. This research investigates how operations management principles of optimization, resource allocation management, and process improvement through a data analytic and effective decision-making process could remediate these inequities. Using datasets from NYCPS, this study leverages demographic and staffing data from 2020-2022 to identify factors influencing counselor-to-student ratios and STEM teacher distribution across boroughs. The analysis suggests that inequities persist because staffing decisions rely on human decision making based on lack of effective human resource management systems. Building on these findings, this research proposes a practical decision framework that extends traditional regression analysis by incorporating reinforcement learning and bandit algorithms—providing public agencies facing similar staffing shortages with an adaptive, data- driven approach to continuously improve resource allocation and promote equity. While this analysis is exploratory, the robustness of the analysis is only limited to yearly-released publicly available NYCPS data. Additionally, variations in yearly school-specific data collection, quality control, and reporting protocols may skew the decision-making recommendations.

Andrew Singh is a second-year M.S. student in Business Analytics at the McDonough School of Business, Georgetown University. He is a nonprofit consultant and data analytics instructor specializing in data visualization and applied analytics for education and workforce organizations. His work focuses on using tools such as Tableau, Excel, and Salesforce to support program evaluation and equity-focused decision-making.

Chenghao Du, Computer Science

"Attack in Your Pocket: Understanding Security Threats in Mobile LLM Agents"

Advisor: Dr. Yue Xiao

Large Language Models (LLMs) have transformed software development, enabling AI-powered applications known as LLM-based agents that promise to automate tasks across diverse apps and workflows. Yet, the security implications of deploying such agents in adversarial mobile environments remain poorly understood. In this paper, we present the first systematic study of security risks in mobile LLM agents. We design and evaluate a suite of adversarial case studies, ranging from opportunistic manipulations such as pop-up advertisements to advanced, end-to-end workflows involving malware installation and cross-app data exfiltration. Our evaluation covers eight state-of-the-art mobile agents across three architectures, with over 2,000 adversarial and paired benign trials. The results reveal systemic vulnerabilities: low-barrier vectors such as fraudulent ads succeed with over 80% reliability, while even workflows requiring the circumvention of operating-system warnings, such as malware installation, are consistently completed by advanced multi-app agents. By mapping these attacks to the MITRE ATT&CK Mobile framework, we uncover novel privilege-escalation and persistence pathways unique to LLM-driven automation. Collectively, our findings provide the first end-to-end evidence that mobile LLM agents are exploitable in realistic adversarial settings, where untrusted third-party channels (e.g., ads, embedded webviews, cross-app notifications) are an inherent part of the mobile ecosystem.

Chenghao Du is a Second year Ph.D. student in Computer Science Department at William & Mary. His research area include Android Security, Large-Language Model Agent Security, and Cryptography Misuse Detection. He holds a B.S. from Ohio State University and a M.S. from New York University. He is currently exploring the validation of LLM-powered CryptoAPI misuse detector.

Xuehang Guo, Data Science

"Domain-Aware Agent for Multimodal Scientific Reasoning"

Advisor: Dr. Qingyun Wang 

Recent advances in enhancing the multimodal reasoning and problem-solving capabilities of multimodal large language models (MLLMs) have fundamentally benefited not only AI research but also a wide range of other domains. However, effective application in these fields requires not only complex multimodal reasoning but domain-specific knowledge that general-purpose MLLMs usually lack. Moreover, scientific research in specialized domains relies on both expert-level understanding and multimodal analysis of domain data presented in heterogeneous formats, such as tables, figures, and document layouts. To address these challenges, we introduce DomainBench, a comprehensive benchmark spanning diverse research domains that depend on domain-specific expertise to analyze and generate professional responses. Reflecting the heterogeneous complexity of real-world scientific research, DomainBench encompasses multiple data sources and formats. In addition, we present CliAgent, an agentic framework that augments MLLMs with domain-specific knowledge acquisition and interactive environments for task-driven exploration. Our results demonstrate that CliAgent achieves up to a 4.72% improvement in overall accuracy for clinical multimodal understanding, underscoring the significance of combining problem-solving with interactive exploration for effective domain-specific research.

Xuehang Guo is a first-year Ph.D. student in the School of Computing, Data Sciences & Physics at William & Mary. Her research focuses on NLP, Agentic AI, and AI for Science. She is particularly interested in designing adaptive autonomous AI systems capable of complex reasoning, multimodal understanding, long-horizon planning, and tool-interactive problem solving.

Robiul Islam, Computer Science

"Leveraging LLMs with Spatial Reasoning to Decide Edge-Cases in Autonomous Driving ODDs"

Advisor: Dr. Trey Woodlief

Driving Automation Systems (DAS) are designed to operate only within a defined Operational Design Domain (ODD), which specifies the environmental, roadway, and traffic conditions where the system can function safely. However, existing driving datasets often include scenes outside a DAS’s ODD which interfere with development. Removing these requires costly manual review since ODD’s use vague or context based terms such as “heavy rain” or “bright light”. LLMs can help resolve these nuances by connecting natural language descriptions with the visual sensor data of the scene. Building on this idea, we introduce ODD-Quadrant, an automated method that leverages LLMs to verify whether a dataset complies with ODD requirements. To improve reliability, we use spatially guided prompt engineering based on established domain knowledge in roadway safety. Each image is partitioned into spatial regions to localize visual context and highlight the vehicle’s forward path. A feedback loop analyzes false positives and negatives to iteratively refine prompts and capture challenging edge cases, improving model consistency across iterations. Compared to prior work, which re- lied on expensive commercial-grade LLMs, and achieved a 147% improvement over human inspection. ODD-Quadrant achieves a 205% improvement, yielding a significant boost in detection efficiency and nearly doubling the true-positive cases while using free, open-source models. Finally, our results demonstrate that using spatial prompting and structured validation enhances the accuracy and consistency of ODD compliance checks.

Robiul Islam is a second year PhD student in Computer Science at William & Mary. His research areas includes Robotics, Software Engineering, Large Language Models. He holds BSc from Uttara University, Dhaka, Bangladesh.

*Alvaro Martin Grande, Computer Science

"Optimizing Lustre's Metadata Management with CXL-based Shared Memory"

Advisor: Dr. Jie Ren

*GRS Award for Excellence in Scholarship -- School of Computing, Data Sciences, and Physics

Data centers face exponential growth in computing demands while struggling with memory costs and storage bottlenecks. Lustre, the dominant parallel distributed file system in High-Performance Computing, is widely deployed in national research facilities, and commercial cloud and AI infrastructures. Designed for massive scalability, Lustre manages petabytes of storage and delivers terabytes/s of throughput. However, Lustre’s metadata coordination, managed by centralized or distributed Metadata Servers (MDSs), suffers across distributed nodes, causing scalability and performance bottlenecks. Compute Express Link (CXL), an emerging interconnect standard, enables multiple hosts to pool heterogeneous memory devices as a shared memory, beyond local DRAM limits. This research explores integrating CXL-based shared memory into Lustre’s metadata management layer to overcome these bottlenecks. Our solution offloads metadata handling to CXL memory shared by all MDS nodes. By caching and synchronizing metadata updates in shared memory, Lustre reduces inter-MDS communication delays, reduce lock contention, and accelerate recovery after failures. Preliminary evaluations show metadata operations are a major scalability barrier in Lustre. Our work enables metadata caching over CXL, avoiding traditional network-based coordination. The main challenge lies in ensuring coherent, low-latency synchronization across MDS nodes while managing performance gaps between DRAM and CXL memory. By reducing synchronization overhead, our approach improves metadata throughput and scalability.

Alvaro Martin Grande is a third-year Ph.D. student in Computer Science at William & Mary. His research focuses on computer systems, including CXL-enabled disaggregated memory, distributed file systems, and operating-system support for tiered memory. He explores hardware–software co-design to improve file system performance through emerging memory architectures. He holds a B.A. in Computer Science from Augustana College.

Mehedi Sun, Computer Science

"Why Was This Code Written Like This? Informing Developers on Code Change Rationale"

Advisor: Dr. Oscar Chaparro

Understanding why the source code changes is essential for many software engineering tasks, e.g., refactoring and reviewing the code, debugging, and implementing new features. Unfortunately, locating this information (code change rationale) can be extremely difficult for developers because the information is often fragmented, inconsistently documented, and scattered across different artifacts, e.g, commit messages, issue reports, and pull requests, etc. We address this challenge in two major steps. First, we conduct an empirical study of 63 code changes from five projects to systematically analyze where key rationale components, for example, what the change aims to achieve (Goal), why it was needed (Need), and alternatives that were considered (Alternatives), are recorded. We find that rationale is highly fragmented (commit messages mostly describe Goals, while Needs and Alternatives appear more often in issues and pull requests) and, importantly, no single artifact type consistently captures all rationale components. Second, we introduce Argus, an LLM-based approach that identifies sentences expressing Goal, Need, and Alternatives across a commit's related artifacts and synthesizes short, readable summaries. In evaluation on 50 additional commits, Argus achieves high coverage of rationale sentences (93% recall), and the generated summaries were rated useful and effort-saving by developers in a user study. Our results show that multi-document reasoning can make the "why" of code changes more accessible to developers for better understanding and maintaining software systems.

Mehedi Sun is a third year Ph.D. student in Software Engineering at William & Mary. He research focus on how and why software changes over time and how to help developers better understand the reasons behind code updates, bug fixes, and design decisions. His work aims to make software development more transparent, reliable, and easier to maintain.

Muneeb Ahmad, Engineering | Virginia Tech

"Development of Neuromuscular Interface for Assistive Exoskeleton Control in Essential Tremor"

Advisor: Dr. Sujith Vijayan

Co-Authors: J. Decker

Essential tremor (ET) is one of the most prevalent neurodegenerative diseases in the elderly population: it is estimated to affect 5.79% of world population above the age of 65. Pharmacological treatments have limited efficacy in the long run and the implantation of brain stimulation devices is expensive and carries risks. Therefore, the development of an affordable exoskeleton is a promising solution. Traditional exoskeletons rely on muscular activity, referred to as electromyography (EMG), and motion data for device control. The utilization of brain signals has not been well explored because of practical limitations. Recent advancements in portable electroencephalography (EEG) devices have increased the feasibility of using brain signals with exoskeletons. Since neural signals lead muscular signals by 11–27ms in ET individuals, we expect that neuromuscular assistive devices that employ EEG in addition to EMG will be responding faster to wrist ET. Our objective is to investigate the relationship between ET neural dynamics, EMG and motion data. In this study, EEG, EMG and motion data were collected from control and ET participants during multiple arm movement tasks. By performing spectral analysis on motion and EMG data, we have identified localized tremor activity in ET participants. We are now examining coherence across all three signal modalities. We expect to identify the brain regions involved in tremor generation, leverage this knowledge to anticipate the tremor before its manifestation in the wrist, and use neuromuscular exoskeleton to counteract the tremor movement.

Muneeb Ahmad is a second-year Ph.D. student in Mechanical Engineering at Virginia Polytechnic Institute and State University. His research focuses on human–robot interaction and multi-robot systems. His current research involves the use of brain signals to control exoskeletons for tremor. He earned his B.Sc. and M.Sc. in Mechatronics Engineering from the University of Engineering and Technology (UET) Lahore, Pakistan.

Jasmine Albert, Physics

"Generating Anti-Coherent States Using Random One-Axis Twisting for Axis-Agnostic Precision Metrology"

Advisor: Dr. Gregory Bentsen

Co-Authors: J. Bringewatt, L. Zaporski, M. Radzihovsky, A. Gorshkov

In the field of metrology, the ultimate goal is to develop and implement techniques and technologies  to measure things as precisely as possible. However, any measurement one makes in the lab is fundamentally limited by the Heisenberg uncertainty principle. Despite this, we can use different techniques that involve sacrificing the resolution of one of one axis in favor of bolstering the precision of another, enhancing our measurements while still obeying the uncertainty principle. Traditionally, the states metrologists work with require a preferred axis of rotation for measurements. One way to work around this problem is the use of anti-coherent states, which inherently have no preferred axis of rotation. In this project, we look at different, experimentally-viable methods for creating these anti-coherent states, the main one being the Random One-Axis Twisting (ROAT) method. This process entails repeated redistribution of the quantum information throughout the state, resulting in highly-detailed states with extremely small angular precision. These methods will allow us to precisely measure magnetic fields without needing any specific measurement axis, which can be extremely useful in many real-world applications.

Jasmine Albert is a second year Ph.D. candidate in the Physics Department at William & Mary. Her research areas include quantum information and precision metrology. She is currently looking at methods of generating highly-sensitive states to reach quantum limits of measurement resolution.

Samuel Bevins, Physics

"Searching for Quantum Error-Correcting Codes via Measurement-Driven Dynamics"

Advisor: Dr. Gregory Bentsen

When we run a quantum system, two things can happen at once: random operations spread connections between its bits, while measurements tend to break those connections. These quantum states survive the measurement process because there is an error-correcting code hiding inside the underlying structure that describes the dynamics of the system. As we turn up the rate of measurements, the system passes a tipping point where those connections shrink dramatically. This tipping point is called a measurement-induced phase transition. We study this tipping point using fast, exact simulations of simple one-dimensional quantum circuits. First, we pinpoint where the transition happens by running many simulated experiments and using standard statistical checks to make sure our estimate is reliable. We are interested in examining this semi-natural phenomenon to see if we can generate codes with desirable properties. To do this, we look at the patterns left behind at the end of each run to automatically build candidate error-correcting codes. These error-correcting codes are recipes for protecting fragile quantum information. For each candidate, we record how much data it can carry, how tough it is against mistakes, and how complex its basic checks are. We also compare different circuit designs to see which ones most often produce useful codes. As a next step, we plan to test these codes with practical decoders under common noise models. Our goal is a simple, reproducible pipeline that spots a source of quantum codes that could work on real hardware.

Samuel Bevins is a second year Ph.D. candidate in the Physics Department at William & Mary. His interests lie in the intersection of Quantum Information Science and the geometric and algebraic structures that underlie this broad field of both theoretical and mathematical physics. He holds a B.S in Physics and a B.S. in Pure Mathematics from Virginia Commonwealth University.

Yamil Cahuana Medrano, Physics

"Gaussian Processes à la Feynman: Extracting PDFs from Lattice QCD"

Advisor: Dr. Kostantinos Orginos

In this work, we develop a method to reconstruct fundamental functions known as parton distribution functions (PDFs), which describe how the building blocks of matter—quarks and gluons—are distributed inside protons and neutrons. The data we use come from high-performance simulations of quantum chromodynamics (QCD), the theory of the strong force, performed on a discretized spacetime grid known as Lattice QCD. Reconstructing PDFs from these simulations is a challenging task because the data are limited, noisy, and expressed in a form different from the one in which PDFs are usually defined. To solve this problem, we apply a Bayesian statistical approach using Gaussian Processes (GPs), which offer a flexible way to model uncertainty and incorporate prior knowledge about the expected shape and behavior of these distributions. We explore different models and strategies to balance fitting the data and avoiding overfitting. To evaluate how much new information the data provide, we use a metric called Kullback–Leibler divergence. Finally, we combine multiple models through a selection and averaging procedure to improve the robustness and interpretability of the results.

Yamil Cahuana is a fourth-year Ph.D. candidate/Fulbright Scholar in the Physics department at William & Mary. His research areas include Bayesian and Machine learning reconstruction of parton distribution functions, and multi-grid algorithms in the context of lattice field theory. He is currently exploring the possibility of speeding up algorithms with neural networks.

Muhammad Nehal Khan, Physics

"Three-Photon Rydberg Excitation for local Electric-field Quantum Sensing"

Advisor: Dr. Irina Novikova 

This project aims to develop a quantum-sensing technique for localized electric-field measurement using laser-driven Rydberg excitations in rubidium vapor. The central question is whether a three-photon excitation process can produce a stable and spatially selective signal that can be used for electric-field detection. I have developed a collinear setup where all three lasers propagate in same direction and measure changes in the transmitted light under different field conditions. I also studied the blue fluorescence produced during the excitation; it shows multi-photon features that respond to applied laser field. The next step is to develop a star-configuration, where the beams intersect at angles, which is expected to improve the overall sensitivity and spatial resolution of the technique. The long-term goal is to establish a simple, high-resolution method for mapping electric fields using atomic quantum states.

M. Nehal Khan is a third year Ph.D. candidate in Physics Department at William & Mary. His research areas include experimental quantum optics, theoretical quantum information science, and quantum simulations. His current research focused on developing the quantum sensing technique to measure localized electric field. He holds two master degrees, one from William & Mary and another from PIEAS (Pakistan).

Malik Tasneem Raza, Physics

"Simulating Pion-Induced Backgrounds for the MOLLER Experiment"

Advisor: Dr. David Armstrong

The MOLLER experiment (Measurement of Lepton-Lepton Elastic Scattering) at Jefferson Lab is a precision experiment designed to test the Standard Model of particle physics, which describes how fundamental particles interact. In this experiment, high-energy electrons will be scattered off electrons in a hydrogen target to measure a tiny effect known as Parity-Violating asymmetry. This effect helps to determine the weak charge of the electron, a parameter that could help in probing for new physics beyond the Standard Model. One major challenge is to separate the desired electron signals from the unwanted background signals caused by the short-lived particles known as pions produced during high-energy collisions with protons in the hydrogen target. These pions can decay into other particles (muons) that also mimic electron signals, reducing the accuracy of the measurements. To study and reduce this background, the experiment uses a system of pion detectors placed behind the main electron detectors. It is made of lucite (a transparent material) with photomultiplier tubes (PMTs) attached to detect Cherenkov light, a glow produced when charged particles move faster than light travels through that material. I created a  Monte Carlo simulation of this detector to study how different particles, such as electrons & pions, produce Cherenkov light inside it. Electrons generate light even at low energies, while pions need higher energy. By comparing these light signals, we can identify and separate pion events from electron events, helping the experiment achieve the precision needed.

Malik Tasneem ul Raza is a Physics graduate student at William & Mary, working on the MOLLER experiment at Jefferson Lab. Her research focuses on the pion detector, where she does simulations of this detector to study how different particles, such as electrons and pions, produce Cherenkov light inside it and distinguish pion backgrounds from electron signals.

Shira Godin, Psychological Sciences

"The Roles of Attachment Security and Social Support in Buffering Lifespan Adversity and Intergenerational Trauma"

Advisor: Dr. Madelyn Labella and Dr. Shaina Kumar 

Adverse childhood experiences (ACEs) predict greater adulthood adversity and offspring trauma exposure (Narayan et al., 2017). Information is needed regarding potential moderators of this link to guide preventive interventions (CDC, 2025). Nurturing relationships have been shown to protect against negative health outcomes of ACEs (Lou et al., 2025), suggesting that related constructs may buffer continuity in adversity over the lifespan and across generations. The current project examines attachment security and social support as moderators of lifespan adversity and intergenerational trauma, using data from a longitudinal study of early social-emotional development in a socioeconomically diverse community sample (n = 205 parent-child dyads). At T1 (children aged 1–2 years), parents completed self-report measures of childhood and adulthood adversity, a semi-structured interview on perceived social support, and a story stem task assessing attachment security. At T2 (children aged 3–4 years), parents completed self-report measures of recent adversity and their child exposure to ACEs. Planned analyses include a series of moderated regressions. We expect that parental ACEs will positively predict adulthood adversity and child ACEs, and that attachment security and social support will weaken those links, consistent with protective effects. Findings may identify protective processes that buffer pathways from history of ACEs to adult adversity and intergenerational trauma, informing interventions that disrupt continuity in adversity by enhancing secure and supportive relationships.

Shira Godin is a first-year MS student in the Psychological Sciences Department at William & Mary. She is interested in developmental psychology, specifically in examining the role of parent-child relationships in a child's psychological well-being. She is currently exploring moderating factors of the link between parent and child adverse childhood experiences. She holds a B.S. from the University of Maryland.

Gabrielle Heard, Psychological Sciences

"Attentional Biases and Depression: Elucidating the Behavioral and Neural Correlates of Negative Bias"

Advisor: Dr. Paul Kieffaber

Depression reduces quality of life, accelerates biological aging, and shapes how individuals perceive and interpret experiences (WHO, 2025; Remes et al., 2021). These effects contribute to major public health costs. While research often focuses on symptoms and medication, cognitive biases also play a role. Chadwick and Taylor (2000) showed that inferential biases intensify when events are threatening and filtered through maladaptive schemata. Yet the illusory correlation (IC) effect, perceiving relationships between unrelated events, remains understudied in depression (Matute et al., 2015). EEG, a non-invasive method with millisecond precision, can reveal neural mechanisms of IC (Zhang et al., 2023). Event-related potentials (ERPs) index attention, expectancy violation, and semantic processing (Cohen, 2014). This study examines ERP components (N1, P1, P2, N400, MMN) to test whether higher depressive symptoms predict stronger IC for negative pairings. Sixty undergraduates will complete traditional and reversed IC tasks using validated trait datasets (Goldberg, 1990; Anderson, 1968). ERPs indexing attention (N1, P2) and expectancy violation (MMN, N400) will be combined with behavioral data. Depressive symptoms (CES-D; cutoff = 16) will be tested for links to IC magnitude and ERP responses. We predict stronger IC effects and larger ERP amplitudes for negative traits in individuals with higher depressive symptoms. This integration of IC and EEG will clarify how depression biases information processing and interventions for cognition. 

Gabrielle “Gabby” Heard is a Georgia native and Psychological Sciences Master's student at William & Mary (Class of 2027). She researches brain health and quality of life, focusing on alternative and preventative treatment in Dr. Paul Kieffaber’s Cognitive Psychophysiology Lab. Gabby plans to become a clinical neuropsychologist. 

Sarah Liu, Psychological Sciences

"Autistic Traits and Unhealthy Eating Habits: Biased Attention Pattern towards Food Stimuli "

Advisor: Dr. Catherine Forestell

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition characterized by social and communicative difficulties and restrictive and repetitive patterns of behavior – potentially including eating behavior (Adams et al., 2024). Since Gillberg (1983) first brought attention to the overlap between eating problems and ASD, substantial evidence has supported this link. Mayes & Zickgraf (2019) reported that autistic individuals exhibit more selective eating, food neophobia, and food fussiness. In addition, atypical eating behaviors are 15x more common in autistic children than in typically developing children. Despite the high prevalence of eating problems, the reason for the association between autism and eating problems is poorly understood, which restricts the efficacy of clinical care.  

The primary objective of our study is to understand the cognitive factors that contribute to the eating habits of individuals with high levels of autistic traits. We will first examine the association between autistic traits and eating habits (food neophobia, picky eating, and dietary variety seeking). Moreover, two implicit cognitive tasks will be administered: the dot probe task to assess the sense of threat towards disliked food items (i.e., vegetables), and a food flanker task to assess attentional distraction by unhealthy foods. Understanding the underlying cognitive mechanisms of autistic eating habits will not only fill a critical research gap, but also provide insights into more effective intervention strategies for eating problems in the autistic population.

Sarah Liu is a first-year master’s student in Psychological Sciences at William & Mary. Her research interests include eating behaviors, neurodivergence, and interoception. Her dissertation focuses on eating difficulties in autistic individuals, examining how attention toward food may differ from those of neurotypical peers.

*Kira Sturgess, Psychological Sciences

"Negative Affect Model of Alcohol: A Cross-National and Longitudinal Examination Among College Students"

Advisor: Dr. Adrian Bravo

*Carl J. Strikwerda Award for Excellence in the Humanities and Humanistic Social Sciences -- College of Arts & Sciences

Negative reinforcement models of addiction posit that negative affect leads to problematic substance use via cognitive mechanisms (e.g., ruminative thinking) and coping motivations (e.g., drinking to cope). The present study aimed to longitudinally replicate and extend prior findings by examining the relationships between depressive symptoms, rumination, and problematic drinking among college students from six countries (i.e., USA, Argentina, Spain, South Africa, England, and Canada). Participants at baseline were 2,538 college students (71.1% female; mean age = 20.87) who regularly consumed alcohol within a typical week and completed an online survey (longitudinal analytic sample at T3 follow-up n = 505, 78.5% female). Within our results, we found that greater depressive symptoms both cross-sectionally and longitudinally was positively associated with greater ruminative thinking (particularly problem-focused thoughts), which in turn was associated with more drinking to cope motivations, which in turn was associated with greater alcohol use quantity and related consequences. However, culture moderated many of the associations, highlighting the importance of contextualizing these relationships within cultural norms and practices. Overall, our findings suggest depression relates to higher scores of problem-focused thoughts, predicting more alcohol related consequences and problematic drinking. Aligning with prior literature, these findings enrich current research on negative affect models of addiction with the addition of cross-national and longitudinal data. 

Kira Sturgess is a first year Master's student in Psychological Sciences at William & Mary. Her research examines the interplay of substance use and chronic pain/illness, and how mindfulness, spirituality, and religiosity impact this relationship. Her thesis focuses on the development of a measure assessing expectancies for marijuana analgesia. She holds a B.S. from Colorado State University.

Natalie Tuinstra, Psychological Sciences

"Multi-Informant Assessment of Adolescent Outcomes Following a Modified Tuning in to Teens® Program"

Advisor: Dr. Janice Zeman

Early adolescence is a critical period when anxiety and depression often increase (Cohen et al., 2017). Emotional competency, including awareness and regulation, is essential for healthy development (Saarni, 1999; Lau & Wu, 2012), and difficulties managing anger, sadness, and worry can heighten risk for psychological problems (Zeman et al., 2002). Interventions that strengthen emotion skills may enhance regulation and reduce psychopathology (Moltrecht et al., 2021). This study evaluates a modified Tuning in to Teens® (TINT) program (Kehoe et al., 2013), adapted from a parent-focused program for direct delivery to middle-school adolescents in six weekly two-hour sessions. 54 youth (Mage = 11.75, 31 girls) and their parents completed questionnaires assessing anxiety, depression, externalizing symptoms, and emotion coping, regulation, and inhibition at T1 (first session), T2 (last session), and T3 (one month post-program). Parents reported improvements in internalizing and externalizing symptoms, coping, and regulation after six weeks, while youth reported more gradual changes after ten weeks. Findings suggest the modified TINT program may improve psychological and emotional functioning in early adolescence, though further research is needed to examine gender and age differences and generalizability to diverse samples.

Natalie Tuinstra is a second year master’s student in Psychological Sciences at William & Mary. Her research focuses on adolescent social and emotional development, adolescent psychopathology, and the effectiveness of interventions. She holds a B.S. in Psychology and a B.S. in Information Science from UNC-Chapel Hill.

Tingyi Zhang, Psychological Sciences

"Inspiration Across Cultures: Exploring Universal and Culture-Specific Aspects of an Optimal Motivational Experience"

Advisor: Dr. Todd Thrash

Inspiration is a key motivational construct linked to creativity and well-being. However, research on this topic has largely been confined to Western populations, leaving questions about its cross-cultural validity. The present study examines whether inspiration can be measured without bias across three cultural contexts: the United States, China, and India. Approximately 350 students per country will participate. Participants will complete online surveys assessing inspiration, awe, positive and negative emotions, life satisfaction, self-actualization, individualism–collectivism, and cultural tightness–looseness. We aim to evaluate whether the inspiration items function similarly across these groups and to explore how cultural contexts shape the ways individuals experience inspiration. We expect that inspiration will emerge as a universal but culturally shaped construct, as reflected in similar measurement properties across the three countries. In all three countries, inspiration is expected to be positively associated with well-being constructs such as positive emotions, life satisfaction, and self-actualization, positively related to individualism, and negatively related to cultural tightness. An exception is that we expect the relationship between inspiration and negative emotions to be negative only in U.S. sample. This research contributes to cross-cultural psychology by establishing whether inspiration can be measured without bias across cultures, and will also establish the universal significance of inspiration while being sensitive to cultural differences.

Tingyi Zhang is a second-year M.A. student in Psychological Sciences at William & Mary. Her research focuses on inspiration and positive emotions, with interests in cross-cultural psychology. She uses experience sampling and multilevel structural equation modeling to examine within- and between-person processes related to meaning and well-being.