62970 research outputs found
Sort by
In Principals We Trust: A Three-Part Study of Trust in Catholic School Leadership
No full textTrust is an essential component of school improvement (Bryk & Schneider, 2002). Trust can influence the relationships between students, teachers, administrators, and parents, serving as the foundation and cornerstone for effective collaboration, communication, which can lead to overall school improvement. High levels of trust within school communities contribute to better academic outcomes, improved school leadership, and stronger school communities (Bryk & Schneider, 2002; Tschannen-Moran, 2014a). Understanding how trust is influenced, built, and then maintained among stakeholders in school settings is crucial for identifying practices that improve schools.
This dissertation seeks to explore the construct of trust within Catholic schools, with a particular focus on the dynamics between teachers, parents, and their school leaders. Building on the foundational work of scholars such as Bryk and Schneider (2002) and Tschannen-Moran (2014a), this mixed-methods study aims to examine how trust functions in Catholic schools, where community values and religious beliefs may play an additional role in shaping relationships. By exploring a descriptive study of a trust survey and accompanying interviews with principals, teachers, and parents of high trust schools, the study will provide insights into the unique factors that contribute to high trust levels in Catholic schools and how these practices can be built and maintained by principals with parents and teachers
The Impact of SEC v. Jarkesy on Civil Tax Fraud Penalties
No full textThis article explores the implications of the Supreme Court’s decision in SEC v. Jarkesy for the enforcement of civil tax fraud penalties by the IRS. The article analyzes how the Court’s reasoning—particularly its limits on agency adjudication and emphasis on the Seventh Amendment right to a jury trial—could affect the constitutionality of IRS-imposed penalties. Camp considers whether such penalties must now be litigated in federal court rather than through administrative proceedings. He examines the potential consequences for tax administration, taxpayer rights, and the broader regulatory framework. Ultimately, the article highlights Jarkesy as a pivotal case that may reshape how civil tax fraud cases are prosecuted and resolved
Integrated Microbial Assessment and Quantification Methodologies for Salmonella Control in Pet Treats and Poultry Processing: From Thermal Efficacy to Biomapping Analysis Across Multiple Regions
No full textFoodborne illnesses remain one of the most pressing public health challenges worldwide, with Salmonella spp. representing a leading cause of bacterial outbreaks linked to poultry and animal-based food products. In response to this complex issue, this thesis presents a multidisciplinary approach to improving microbial food safety by integrating regional biomapping, predictive surface area modeling, and comparative evaluation of pathogen detection and quantification methodologies.
The research begins by establishing the global epidemiological burden of Salmonella and analyzing its routes of transmission through food, water, animal reservoirs, and environmental sources. Emphasis is placed on the poultry production continuum, which plays a significant role in Salmonella prevalence, particularly in low and middle-income countries. Current regulatory frameworks and industry practices were reviewed, highlighting the need for enhanced microbial risk assessment tools tailored to diverse processing environments.
Chapter II introduces a novel methodology for converting microbial loads from CFU/mL to CFU/cm² by developing predictive equations based on 3D scanning (Polycam™), mesh analysis (MeshLab 2.0), and physical measurements of poultry parts. A strong correlation between weight and surface area was observed (R² = 0.92 for tenders; R² = 0.86 for wings), enabling more accurate standardization and cross-regional comparisons of contamination data. Application of this model to biomapping studies conducted in North, Central, and South American poultry plants revealed regional variability in Salmonella load reduction across processing steps, particularly highlighting discrepancies in final product safety in South America.
Chapter III compares traditional and alternative sampling methodologies Composite Rinse (USDA-FSIS approved), MicroTally® Mitt, and MicroTally® Swab using both naturally contaminated and inoculated chicken samples. These were paired with microbial indicator assessments (Aerobic counts and Enterobacteriaceae) and four quantification techniques: spread plating, spiral plating, and the GeneUp® Salmonella system with and without incubation. The MicroTally® Mitt demonstrated consistent recovery of microbial indicators and Salmonella spp., often outperforming or equaling the Composite Rinse method, suggesting its potential for routine use in commercial processing plants due to its operational efficiency, ergonomic design, and surface coverage.
Chapter IV evaluates the efficacy of heat treatment cycles in industrial ovens against Salmonella spp. in commercial pet treats. Four thermal processing cycles were tested across three product types (tenders, fillets, and food pellets), and results showed that moisture content, product geometry, and cycle structure significantly influenced bacterial reduction. The most effective cycle achieved a > 6-log reduction in Salmonella, aligning with international safety standards for thermal inactivation. Notably, dry heat without intermediate humidity phases proved less effective, emphasizing the synergistic role of relative humidity in enhancing microbial lethality.
Finally, the thesis concludes by outlining practical implications for the food industry, emphasizing the importance of surface area-based microbial enumeration, harmonized sampling protocols under the USDA Salmonella Framework, and validated thermal processing regimes in the pet food sector. The findings provide a scalable and science-driven foundation for improving microbial safety across animal-based food supply chains. The study advocates for a One Health approach, integrating human, animal, and environmental health perspectives to design sustainable, regionally adaptable food safety interventions
Exploring Climate-Smart Agriculture Communication Efforts in Texas A&M AgriLife Extension: A Mixed-Methods Approach
No full textClimate change poses significant threats to agricultural productivity through extreme weather events such as droughts, floods, storms, and heat waves (Diehl et al., 2017). Climate-smart agriculture (CSA) offers a foundational approach to address these challenges by (1) improving food security, (2) enhancing climate resilience, and (3) reducing greenhouse gas emissions. In the U.S., the USDA and land-grant institutions like Texas A&M AgriLife have invested in CSA implementation. While Texas A&M AgriLife County Extension Agents are trusted sources of information, their communication may be influenced by personal beliefs and social norms. This study involved two phases. Phase one used a quantitative inventory audit of online communication materials produced by Texas A&M AgriLife. A codebook and key terms such as “climate-smart agriculture,” “sustainable,” and “regenerative agriculture” were used to identify CSA-related content. Research articles were the most common resource, with few materials being action-oriented or clearly tied to CSA practices. Phase two, guided by the Theory of Planned Behavior, examined how attitudes, subjective norms, and moral norms influence extension agents’ CSA information-seeking and sharing behaviors. Results revealed that existing materials did not align closely with the types of communication agents sought or shared. Strengthening the connection between CSA researchers and agents, especially those with lower climate change attitudes, may increase sharing intentions. Future efforts should focus on translating research into accessible materials distributed through internal announcements, publications, online learning databases, and social media. Expanding this research to other agricultural practices could uncover similar communication ga
Machine Learning-Based Predictive Modeling for OTC and exercise recommendation for knee joint pain with long-tail classification
No full textKnee osteoarthritis is a widespread chronic condition that significantly impacts the quality of life for many individuals, particularly older adults. There are many over-the-counter (OTC) medications and exercise plans can be offered to patient to manage the disease, but without personalized guidance, patients may struggle to find effective options. In this study, we analyzed clinical and demographic data from 1,687 patients and applied machine learning techniques to assist in making more meaningful treatment recommendations. Notably, some patients did not report a preferred OTC medication or exercise regimen (i.e., labeled as NaN). Through further analysis, we found that even these incomplete data records can still contribute to narrowing down potential treatment options by analyzing other features. We experimented with several models, including Random Forest, Logistic Regression, Focal Loss and LightGBM. To further address the long-tail distribution in the data, we used resampling and reweighting strategies, which helped improve model performance. Our results show that the best performance models achieved an accuracy of approximately 47% in predicting appropriate OTC medications and 63% in exercise plan recommendations. In addition, we observed that using tokenization slightly outperformed both BERT-based embeddings and the approach of omitting textual information entirely in terms of prediction accuracy. While the performance still can be improved, the approach demonstrates the potential of machine learning in providing practical, data-driven support for knee pain management. By narrowing the range of options, this system may help patients make more informed decisions and reduce unnecessary trial and error in selecting treatments
Optimizing Long-Term Agricultural Production to Estimate Nitrogen Fertilizer Demand
No full textNitrogen fertilizer is a significant and necessary component of modern-day farming practices and has consistently used the Harber-Borsch process to manufacture nitrogen since the early 20th century. Nitrogen fertilizer demand continues to rise, while recent world events such as the 2008 Recession, The Coronavirus Pandemic, and well as the Russo-Ukranianin War have led to vast increases in nitrogen fertilizer prices. Based on these rising prices and ever-growing demand it is imperative to achieve estimations of nitrogen demand. This study focused on forecasting nitrogen fertilizer demand for the Texas High Plains region over the next 50 years. In addition, determining future crop production based on groundwater availability will be analyzed. The benefits of sustainable nitrogen manufacturing developed by CASFER will also be discussed. An optimization model was critical to generate water availability and crop acreage forecasts over the 50-year period. Agronomic, economic, and hydrological data was inputted into the model to generate these projections. Using the model’s output nitrogen fertilizer demand could be estimated. Results from the model indicated that groundwater availability will decline in both District 1 and 2 of the Texas High Plains, vastly impacting the crop mix percentage over the forecasted period. Nitrogen fertilizer demand increased in District 1, though at a decreasing rate. District 2 saw similar diminishing growth, until the later part of the study until demand began to decline. These results allow insights to fertilizer manufacturers and CASFER on how nitrogen demand in the Texas High Plains will change. The results can aid in production schedule planning, inventory management, price setting and revenue management within these companies. Based on declining water availability and growing demand for nitrogen, CASFER’s method of harnessing nitrogen sustainably, could become a viable alternative to current manufacturing practices
Effects of feeding a Saccharomyces cerevisiae fermentation product compared to a direct-fed microbial in finishing diets of beef × dairy crossbred steers fed in the Pacific Northwest
No full textThe objective of this study was to evaluate the effects of a Saccharomyces cerevisiae fermentation product (SCFP) compared to a direct-fed microbial (DFM) on growth performance, health, carcass characteristics, and liver abscess prevalence in beef × dairy crossbred steers. Two thousand steers [50% beef, 25% Holstein, 25% Jersey genetics; initial shrunk body weight (SBW) = 288.2 ± 8.0 kg] were blocked by arrival date and randomly assigned to receive 1 of 2 treatments: 1) SCFP supplied in the starter diet at 12 g per steer daily and then 9 g per steer daily in the finishing diet (NS; NaturSafe™, Diamond V, Cedar Rapids, IA) or 2) DFM fed at 50 mg per steer daily throughout the feeding period (BD; Bovamine Defend, Chr. Hansen, Milwaukee, WI). Pen served as the experimental unit (200 steers/pen), with 5 pens per treatment. Data were analyzed as a randomized complete block design in R 4.2.2. with the main effect of treatment and random effect of block included in the model. Results were reported on a deads-in basis unless otherwise stated. Cattle were fed for a total of 275 ± 6.2 d. Initial and final SBW did not differ (P ≥ 0.84) by treatment. Initial treatment pulls were observed more frequently for NS compared to BD cattle (29.43% vs. 21.67%; P < 0.01). However, NS cattle had a lesser rate of repulls as a proportion of initial pulls (10.08% vs. 16.61%; P = 0.03). Fewer (P < 0.01) bullers were reported amongst NS cattle. Cattle supplemented with NS had a lower case fatality rate (6.08% vs. 11.96%; P < 0.01) and tended to have a lower total mortality rate (1.60% vs. 2.70%; P = 0.09) than BD. With deads included, average daily gain (ADG) tended (P = 0.06) to be greater for NS cattle. Dry matter intake did not differ (P = 0.99) by treatment; however, NS cattle had a numeric advantage in feed efficiency (G:F) nearing a tendency (0.132 vs. 0.130; P = 0.11). On a deads-out basis, ADG and G:F were similar (P ≥ 0.85). Dressing percentage tended (P ≤ 0.10) to be greater for NS carcasses. Cattle fed BD had a greater (P = 0.03) proportion of USDA Prime carcasses. While treatment had no impact on liver abscess severity or total abscess occurrence, NS cattle tended to have less A- abscesses (1.72% vs. 3.87%; P = 0.10). In this large-pen comparison, SCFP supplementation improved feedlot cattle health and positively influenced performance compared to a DFM.Diamond V provided retrospective funding for the product and for collection of performance data, carcass data, and analyses of results
Modeling Pulsed Magnetic Core Behavior in LTspice
No full textThis work demonstrates a modeling technique focused on reproducing the behavior of magnetic cores subject to high voltage pulses. The working principle of the model is based on a magnetic circuit with additional elements that influence the model’s behavior. The elements include a function that defines the response of the model depending on the applied pulse voltage and a component that dominates the transient response. These elements are necessary to replicate the experimentally observed behavior of magnetic cores. The model was developed based on the measured behavior of three nanocrystalline magnetic materials subject to a range of pulse voltages. This modeling technique was created to address the limitations of other models in accurately capturing fast pulse responses. The key limitation of traditional modeling techniques that the proposed model addresses is their inability to capture variations in core response under different applied pulse voltages (magnetization rates). The proposed model has been shown to produce accurate results for magnetization rates between 1 T/μs and 8 T/μs, with potential for further expansion. Implemented in LTspice, this model is both fast and accurate, effectively replicating the behavior of the magnetic core while maintaining simplicity. This work outlines the foundation of this modeling technique, the trends in the parameters that influence its behavior, and its application within a simple pulsed power system. The most notable feature of this model is its ability to operate across a wide range of pulse voltages without requiring adjustments to the model parameters.This work was done under the auspices of Mission Support and Test Services, LLC, under Contract No. DE-NA0003624 with the U.S. Department of Energy and the National Nuclear Security Administration’s Office of Defense Programs, and supported by the Site-Directed Research and Development Program. DOE/NV/03624--2162
A Reinforcement Learning Approach for Malware Detection
No full textMachine learning and artificial intelligence are increasingly seen as important tools for cybersecurity defense systems. The field promises to improve threat
detection through automated analysis of attacks and development of security systems that can adapt to changing threats. Researchers and security organizations are actively exploring smart approaches to defend networks and systems. However, this field faces substantial limitations that prevent machine learning from having real operational impact. Standard machine learning approaches require balanced datasets to work well, but cybersecurity environments have severe data imbalances where malicious activities represent a tiny fraction of total network traffic, making it difficult for learning algorithms to develop effective threat detection capabilities. In particular, network-based malware is one of the most common ways attackers infiltrate systems and exploit their vulnerabilities. With the growth of cloud computing, internet devices, and remote access, every system and network must be secure and should be able to detect intrusions adaptively. Modern attackers continuously change their methods to avoid established detection systems, requiring security solutions that can learn and respond to new attack patterns.This thesis presents the application of reinforcement learning to network-based malware detection, addressing the fundamental training challenges that have prevented effective deployment of learning-based approaches in realistic cybersecurity scenarios. It provides a solution through innovative environment design that combines statistical outlier detection, balanced episode construction, and hybrid reward functions that incorporate cybersecurity domain knowledge, presented as
a systematic approach to enable reinforcement learning in imbalanced security
environments. This work demonstrates that adaptive learning systems can be
successfully applied to network security challenges, providing a foundation for
intelligent defense mechanisms that can evolve with emerging cyber threats