17179 research outputs found
Sort by
Development of conversational AI for mental health support
No full textNo abstract prepared.Engineerin
Feasibility, acceptability, and initial efficacy of providing personalized feedback on real-time monitored data to reduce mental health symptom severity
Get PDFMental health problems, including depression, anxiety, and suicidal ideation, are prevalent concerns for college students.
• Despite the presence of accessible university mental health services, fewer than half of students with diagnosable conditions typically engage in mental health treatment.
• The use of technology-based and/or single-sessions interventions may address low rates of help-seeking.
• For instance, ecological momentary assessment (EMA) is increasingly used to assess participants’ current experiences, contexts, and symptoms.
• However, no studies have tested whether providing participants with personalized feedback on their self-monitored data patterns is clinically useful.Psycholog
Mutual Aid and Crisis Response: How Austin Mutual Aid Addressed the Shifting Needs of Residents During Winter Storm Uri
No full textIn February 2021, the city of Austin, Texas, experienced widespread electrical outages, water shortages, and infrastructural failures as Winter Storm Uri overwhelmed the city’s capacity to respond (Martinez 2021). Although emergency response systems are designed to provide rapid aid during emergencies, government and official aid organizations were slow to address the scale of necessity (Haddow et al. 2018; Tiedmann et al. 2023). As delays in immediate response from public agencies became apparent, community-based mutual aid networks quickly stepped in to fill the gaps, providing emergency resources and support where traditional aid was lacking (Martinez 2021). Austin Mutual Aid (AMA), established the previous year in response to the COVID-19 pandemic, quickly mobilized volunteers. As part of their efforts, volunteers transported people to hotels and shelters, collected donations, and distributed essential supplies such as food, water, and blankets to those in need (“Financial Transparency Pre August 2021 | Austin Mutual Aid.” 2021). Grassroots organizations have historically played a pivotal role in mobilizing local resources to help communities recover from various crises. These grassroots efforts often bridge the gaps in critical aid, providing essential needs and services that become inaccessible to communities after severe events overwhelm municipal infrastructure and official emergency responses. Although existing literature examines the disruptions and impacts of Winter Storm Uri on Texas communities, none provide data that reflect the individual needs of residents from the Greater Austin Area (GAA). During its emergency activation, AMA published an aid request form that was active from February 17th - March 11th, providing a broader perspective on the impact of Winter Storm Uri on the GAA. Quantitative and qualitative data extracted from these responses were utilized to conduct statistical, content, and visual analyses. Two distinct waves of requests were observed, and the needs of survey respondents were found to evolve over the course of the Winter Storm Uri response. Residents living in apartments and east of IH-35 reported a higher overall level of need, accounting for the majority of requests made. Analysis of individual responses highlighted the broader systemic failure to equip public services, as even first responders were among those who had to request aid.
These findings are critical for understanding the lived experiences of individuals in the aftermath of disasters and for recognizing the recurring patterns of need that emerge in response. By analyzing the distribution patterns of aid, this study aims to illustrate the effectiveness of decentralized, community-based responses in mitigating disaster impact when formal systems are overwhelmed or fail to reach marginalized groups. The goal is to provide insights into how mutual aid can serve as a model for more equitable disaster response strategies.Geography and Environmental Studie
Beyond the Playing Field: Exploring the Interplay of Academic Support, Socio-Ecological Factors, and Institutional Perspectives on Collegiate Athletes' Academic Success
Get PDFThis dissertation investigates the academic success of collegiate athletes, with a particular focus on students from racially and ethnically marginalized backgrounds. socio-ecological factors, campus support systems, and institutional practices that shape academic outcomes for community college and four-year collegiate athletes. The dissertation consists of three studies: a secondary data analysis, a meta-analytic investigation, and a practitioner-focused manuscript.
The first study utilizes secondary data from the Community College Survey of Student Engagement (CCSSE) to identify predictors of academic success among community college athletes. Results show that high school GPA, self-efficacy, and first-generation status significantly influence college GPA, while developmental education enrollment presents a negative predictor to academic achievement. This study emphasizes the need for targeted academic support interventions for underrepresented athletes in two-year institutions.
The second study presents a meta-analytic investigation into the relationship between campus engagement, climate-based factors, and academic achievement among collegiate athletes. By synthesizing existing research, the meta-analysis identified positive correlations between faculty interactions, educational and student activities, and campus climate and belong with academic performance. However, it also identifies gaps in the literature of non-Division I athletes, underscoring the need for more inclusive research across diverse institutional contexts.
The third paper focuses on actionable recommendations for academic support professionals, advocating for asset-based, culturally responsive strategies to enhance collegiate athlete success. The findings suggest the importance of building supportive, holistic academic environments that bridge the gap between athletics and academics, emphasizing integrated advising, mentorship, and institutional commitment to student well-being.
Collectively, these studies provide a comprehensive framework for understanding the academic success of collegiate athletes, particularly those from marginalized backgrounds, and offer evidence-based recommendations for enhancing institutional support systems. The dissertation underscores the importance of addressing the evolving needs of collegiate athletes, particularly those from marginalized backgrounds, and calls for continued research into effective support systems in higher education.Curriculum and Instructio
Scalable and Robust Solutions for NP-Hard Problems in Large Signed Networks
No full textSigned graphs provide a robust framework for modeling complex relationships and interdependencies among entities. Negative edges in these graphs offer deeper insights into social dynamics and nuanced interactions. A highly efficient method for discovering the fundamental cycle basis in signed graphs has been introduced, which requires linear time and memory to compute the necessary labels and balance the signed graph with a spanning tree. However, this method has not been comprehensively evaluated against contemporary baselines designed to address specific signed graph-related problems. Moreover, existing machine learning models often focus on single-modal approaches, limiting their ability to generalize effectively to new and unseen data distributions. In this dissertation, we introduce three scalable heuristic algorithms designed to address common NP-hard problems encountered in signed networks. Our empirical evaluations demonstrate the untapped potential of these algorithms, showing their competitive edge over established baselines in terms of accuracy and computational efficiency. Finally, we propose the innovative integration of the Kolmogorov-Arnold Neural Network (KAN) and its variants into signed Graph Neural Networks (GNNs). We assess the effectiveness of this integration in tasks such as signed clustering and link sign prediction, demonstrating that KANs enhance the quality of embeddings produced by signed GNNs. This novel approach provides a new avenue for improving the scalability, robustness, and performance of machine learning models applied to complex signed network tasks.Computer Scienc
The Teaching and Learning of Small Gas Engines for SBAE Teachers
No full textNo abstract prepared.Agricultural Science
Enhancing Fire Resistance of Cement-Based Materials through Nanoparticles and CO2 Curing
No full textAs climate change increases the risk of residential fires, there is growing need for fire-resistant materials. This study examines different materials and methods to improve the fire resistance of cement-based materials while incorporating sustainability materials. Through a series of laboratory experiments and analytical techniques, this study investigates the temperature resistance performance of polypropylene (PP) fiber-modified cement-based materials combined with nanomaterials (colloidal nano-silica (CNS) and nano calcium-carbonate (NCC)).Also, the effect of CO2-curing on thermal resistance of cementitious materials was evaluated while incorporating supplementary cementitious materials (SCMs): Slag Cement and Fly-ash cenosphere (FAC), at 25% and 50% percentage by mass. The study employs methodology involving material characterization, mechanical property evaluation, and durability performance. Results revealed that mortar samples with PP fiber and nanomaterials exposed to peak temperatures of 300°C for 2 hours showed notable strength improvements compared to unheated samples. Specifically, specimens with 0.6% PP fiber and 2% NCC exhibited a 19% increase in flexural strength, while those with 0.6% PP fiber and 0.5% NCC showed a 32% increase in compressive strength. In contrast, at 600°C, there was a major decline in mechanical properties in all samples, with samples containing NCC exhibiting the lowest amount (average of 61% reduction in flexural strength and 32% decrease in compressive strength). In general, a higher dosage of nanoparticles and lower PP fiber content will be better to resist temperatures up to 600°C.Results also indicate that at low FAC and SC content (25%), CO2 curing process is more beneficial in forming a stable cementitious matrix, reducing the thermal stress at elevated temperature. Moreover, above 400°C, the effectiveness of CO2 curing on providing thermal stability to the pastes becomes minimal.Engineering Technolog
In the Cockpit with Copilot: The Impact of Copilot Conversations on Mental Health Outcomes Across Gender, Ethnicity, and Year in College
No full textThe study aimed to understand if conversations with these chatbots and AI companions (AICs) were associated with improvements in mental health and whether people feel comfortable talking with a chatbot comfortably. To address our aims, the methodology observed how participants differed in transdiagnostic factors related to mental health after having a week of daily conversations with Copilot. An additional aim of the study was to compare how differences in these factors before and after conversing with Copilot were impacted depending on their year in college, gender, and ethnicity. 72 Texas State University students made up the final sample comprising of 67.6% cisgender women and 50.7% White/Caucasian. Statistical analyses were ran to test whether variables of interest (rumination, suicidal ideation, thwarted belongingness (TB), perceived burdensomeness (PB), hopelessness, and mindfulness) changed significantly from baseline to follow-up after conversations. Further analyses were conducted for any main effects or interactions on the relationships between baseline and follow-up. Additionally, to help better understand their perceptions of conversations with Copilot, participants reported on metrics and open-response questions related to their experience. No gender main effects or interactions with time were observed, and neither were observed for ethnicity/race as well with the exception of a significant main effect and interaction in PB which implied that white individuals had experienced improvements in PB whereas for non-white individuals, the effect was negligible. On average, participants reported a positive experience and found Copilot to be supportive and knowledgeable. Some participants also found Copilot to be lacking of empathy. Several limitations included lack of strong internal consistency in some of the measures, difficulty maintaining initial methods due to Copilot updates, and no control group to make causal claims. Findings highlighted potential benefits of conversing with AICs as well as possible ethnic differences between white and non-white groups in interacting with chatbots.Psycholog
Resilience in the Face of Change: The Past, Present, and Future of Hispanic Female Doctors in the United States
No full textThis work examines the prevalence of female Hispanic doctors in the United States. While it appears that there is a lack of Hispanic women in the medical field currently, Hispanic women were providing, and still are providing, health care services to the Hispanic community through several traditional intervention strategies of medicine. With a multi-disciplinary approach, I explore the history of Latin American women in medicine, the social/psychological barriers for Latinas in medicine today, and the increasing need for Hispanic women in medicine. Data was collected from historical books and articles, as well as psychology journals and medical/physician data to assess the role of Hispanic women in medicine over time, and the need for them in our society. Evidence from these sources provides insight to current gaps in serving Latina healthcare needs. Hispanic female doctors are needed to help address the many growing health concerns of Hispanic women, which include neural tube defect risks, cervical cancer rates, and diabetes. This paper also addresses the barriers and challenges Hispanic women going into the medical field face, and how this will possibly evolve in the future.Psycholog
Exploring the Impact of Self-Efficacy, Social Support, and Gym Intimidation on Physical Activity Levels among Female College Students
Get PDFPhysical activity (PA) is defined as any bodily movement produced by skeletal muscles that requires energy expenditure (WHO, 2024). Engaging in PA reduces the risk of all-cause mortality and chronic diseases such as obesity, diabetes, and cardiovascular disease while also lowering the risk of cancer. Additionally, PA provides several mental health benefits, including improved sleep quality, reduced symptoms of depression and anxiety, decreased stress levels, and enhanced concentration (Sharma, 2006). In a university setting, Student Recreation Centers (SRCs) serve as a primary resource for student health and wellness. SRCs promote PA by offering diverse facilities and programs, ensuring that students of all backgrounds have opportunities to engage in PA. Despite the well-documented benefits of PA and the resources available through SRCs, limited research has explored the factors influencing PA participation among college students—particularly female college students. This study examines the relationship between self-efficacy, gym intimidation, and social support in relation to PA participation at SRCs among female college students, as well as the types of physical activities they engage in. This study hypothesizes that female college students with higher self-efficacy and greater social support will report higher PA levels, whereas gym intimidation will negatively predict PA participation. The research employs a correlational design using a one-time survey that took participants, on average, 8 minutes to complete. Participants included female college students currently attending Texas State University who were 18 years of age or older. A total of 78 female students attending Texas State University completed the survey with an average age of 22 years old. Additionally, many participants identified time and motivation as primary barriers to PA participation, while fewer cited money, knowledge, and support as barriers. Regression analyses revealed that family support and punishment positively predicted PA. Self-efficacy, gym intimidation, and friend support were not significant predictors. Ultimately, a multifaceted approach that addresses both the supportive role of family and the barriers created by gym intimidation, while fostering empowering environments, is essential to promoting lasting, positive behavior change and increasing physical activity participation among female student at student recreation centers.Health and Human Performanc