VTechWorks
VTechWorks provides global access to Virginia Tech scholarship, including journal articles, books, theses, dissertations, conference papers, slide presentations, technical reports, working papers, administrative documents, videos, images, and more by faculty, students, and staff. Faculty can deposit items to VTechWorks from Elements, including journal articles covered by the University open access policy. Email vtechworks@vt.edu for help.
Open Access Policy
Virginia Tech's open access policy enables researchers to deposit the accepted version of scholarly articles with no embargo.
Theses and Dissertations
Virginia Tech was first in the world to require ETDs in 1997, and continues to add scans of older theses and dissertations.
Open Textbooks
More than 50 freely available and openly licensed textbooks are among our most downloaded items.
Communities in VTechWorks
Select a community to browse its collections.
Recent Submissions
Assessing Level of Traffic Stress on Virginia’s Roadways
Hamm, Laura (National Surface Transportation Safety Center for Excellence, 2026-09-01)
This project developed and demonstrated a practical, scalable, and reproducible framework for calculating and applying Level of Traffic Stress (LTS) classifications across Virginia using publicly available and broadly accessible datasets. The work focused on developing a workflow that can be maintained and updated over time without reliance on proprietary software or specialized data collection efforts. The resulting LTS classifications were used to support analyses of bicycle network connectivity and accessibility and develop a prototype visualization tool to support planning and decision-making.
The browser-based prototype visualization tool using VTTI’s DataViz platform allows users to visualize LTS classifications throughout a network, compare different LTS calculation approaches, and conduct accessibility and routing analyses between selected origins and destinations. The application was intentionally designed to operate without specialized geographic information system software, improving accessibility for practitioners and stakeholders.
Overall, this project demonstrates that statewide and regional LTS analysis is technically feasible using scalable workflows and publicly available datasets. The resulting framework provides a practical foundation for identifying low-stress network gaps, evaluating bicycle accessibility, prioritizing infrastructure improvements, and supporting multimodal transportation planning. The work also highlights the importance of moving beyond isolated segment-level classifications toward network-based evaluations of accessibility that reflect how bicyclists experience the transportation system in practice.
Developing a Plan for Educating Roadway Users of Motorcycle-specific Safety Topics
McCall, Robert (National Surface Transportation Safety Center for Excellence, 2026-09-01)
This report documents the development of a Motorcycle Safety Awareness Program designed to educate middle and high school students not yet driving, as well as novice drivers, about the motorcycles’ unique vulnerabilities and operating constraints. The program builds directly on the proven structure and delivery model of the Virginia Tech Transportation Institute’s Sharing the Road initiative, which uses hands-on demonstrations to make roadway risks involving large trucks concrete and memorable. The motorcycle modules apply the same experiential learning logic: students first receive a short classroom lesson introducing core concepts (visibility limitations, braking complexity, rider vulnerability, and common conflict scenarios), then participate in a guided hands-on demonstration with a stationary, stabilized motorcycle. This physical perspective-taking component is intended to increase empathy, improve hazard anticipation, and reinforce simple behavioral commitments such as “look twice,” avoid left-turn conflicts, provide additional following distance, and respect a motorcycle’s full lane.
The proposed delivery format is intentionally brief and scalable. A typical session is designed as a 15-minute module that can be inserted into existing driver education classes, school assemblies, insurance safety events, and community outreach settings. The classroom component (≈ 7 minutes) uses visuals, statistics, and real-world examples to explain why motorcycles are harder to detect and why common driver assumptions about speed, distance, and braking do not reliably apply. The hands-on component (≈ 8 minutes) allows students to sit on the motorcycle (with the engine off) while an instructor guides them through: (1) a braking coordination exercise illustrating hand/foot control requirements; (2) a visibility and “head-check” exercise highlighting limited rearward awareness compared with passenger vehicles; and (3) a short reflection activity emphasizing the absence of a protective shell. Together, these elements aim to produce a more durable learning outcome than lecture-only messaging.
Finally, the report describes how the motorcycle module integrates with Sharing the Road to create operational efficiencies and a unified vulnerable road user safety message. Because both programs target new and soon-to-be new drivers and utilize similar delivery partners (educators, insurers, community safety staff), the motorcycle module can be deployed as a stand-alone session or paired with existing truck demonstrations, depending on scheduling and site needs. Interested parties have been identified that are willing to assist in a pilot deployment.
Self-Reported ADHD Symptoms and Intraindividual Variability in Momentary Cognition Among Young Adults
Mansoor, Marrium (Virginia Tech, 2026-08-31)
Previous research examining intraindividual variability (IIV) on cognitive tasks among individuals with ADHD has tended to focus on variability in reaction times on cognitive task trials. However, intraindividual variability that occurs in moment-to-moment daily cognition has not been examined thus far. This study utilized an Ecological Momentary Assessment (EMA) methodology to investigate within-individual variability across measurement occasions and its association with self-reported ADHD symptoms as well as daily lifestyle activities. Data was collected from 116 young adults (M age = 19.9 years, 68% female, and 44% White) who completed a baseline session in the lab, followed by three semi-randomized EMA sessions per day over two weeks. Each EMA session included three cognitive tasks and a counterbalanced selection of additional surveys. Data was analyzed using Mixed Effects Location Scale models. Results indicated that higher ADHD symptoms were associated with greater variability in reaction times for tasks requiring cognitive flexibility and working memory, but less variability in reaction times on an inhibitory control task. Several interaction models were then conducted to investigate if daily activities influenced the association between self-reported ADHD symptoms and cognitive variability. Several of these were found to be significant, however there was considerable heterogeneity in the results. This study is the first to examine intraindividual variability across occasions in the context of ADHD symptomology and represent an important step towards better understanding of daily cognition in young adults with ADHD symptoms.
Understanding First-Year College Students' Use of AI Chatbots for Mental Health
Ranjber, Samira (Virginia Tech, 2026-08-31)
Artificial intelligence (AI) mental health chatbots are becoming increasingly popular among college students seeking emotional support. Apps such as Woebot, Wysa, Repika, and Youper offer users immediate, low-cost, and private access to mental health tools through online conversation (Fitzpatrick, et al., 2017; Inkster, et al., 2018). First-year college students may turn to these chatbots as they adjust to academic pressure, social changes, and the emotional stress of transitioning to college life. As AI use grows, there is limited research on why first-year college students use these tools and how they emotionally experience communication with them (Beiter, et al., 2015).
Toward Trustworthy Health AI Systems: Advancing Clinician–AI Interaction Through Interpretability, Fairness, and Context-Aware Multi-Agent Safety Architectures
Nasarian, Elham (Virginia Tech, 2026-08-31)
AI has become increasingly integrated into healthcare, supporting clinical decision making, patient risk prediction, and patient-facing information systems. Despite substantial advances in predictive modeling and LLMs, widespread adoption of AI in healthcare remains constrained by challenges related to interpretability, fairness, safety, and human- AI interaction. Healthcare applications require not only accurate predictions and recommendations but also transparent, equitable, and clinically reliable systems that can be trusted by end-users. This dissertation advances the design of trustworthy health AI systems through three complementary studies focused on clinician–AI interaction, fairness-aware clinical risk prediction, and context-aware multi-agent safety architectures.
In essay 1, a systematic review of explainable AI (XAI) in healthcare synthesizes existing approaches and proposes a clinician-centered framework for improving collaboration between AI systems and healthcare professionals. The review identifies key challenges in implementing interpretable AI within clinical decision support systems and provides a roadmap for responsible deployment. In the second essay, a fair and clinically interpretable machine learning framework is developed to predict distinct opioid-related respiratory deterioration events among hospitalized patients receiving opioid therapy. Using electronic health record (HER) data, the proposed multiclass framework differentiates Naloxone intervention events, Blue Code respiratory arrest events, and Rapid Response events. The framework integrates explainability methods, fairness evaluation, and age-specific threshold optimization to improve detection of severe respiratory outcomes while maintaining clinical interpretability. Experimental results demonstrate substantial performance improvements compared with traditional clinical risk scoring approaches. The third essay, a context-aware multi-agent safety architecture, CareGuardAI, is introduced to address clinical safety risks and hallucination risks in patient-facing medical LLMs. The proposed system combines safety-constrained generation, risk assessment agents, and iterative refinement mechanisms to evaluate both medical safety and factual reliability before responses are delivered to patients. Across multiple healthcare safety and hallucination benchmarks, the architecture demonstrates improved performance relative to strong baseline models while maintaining bounded response latency. Collectively, these studies contribute a unified perspective on trustworthy health AI by integrating interpretability, fairness, and safety into the design of healthcare AI systems. The findings provide practical and methodological guidance for developing AI-enabled clinical decision support systems that improve effective human–AI interaction (clinicians and patients) and responsible deployment in high-stakes healthcare environments.