19304 research outputs found
Sort by
Bicultural Bilingual Music Teacher Identities: A Multiple Case Study
No full textHispanic music educators navigate bilingual and bicultural identities in an educational system largely rooted in English monolingualism and Western European music traditions. There is a need to better understand how Hispanic bicultural bilingual music teachers negotiate their own identities and how these identities shape their professional practices. This multiple case study examines the identities of four Hispanic bicultural bilingual in-service music educators in Texas. The study is guided by two research questions: (1) How do music teachers describe their bicultural bilingual identities? and (2) How do teachers feel their bicultural bilingual identities interact with their professional music teacher identities? Data was collected through semi-structured interviews, documents, and artifacts. Individual and cross-case analyses were organized using McClellan’s (2017) Social Cognitive Framework of Music Teacher Identity Construction. Findings indicated that participants’ bicultural bilingual identities were complex, multifaceted, and influenced by their life experiences. Their bicultural bilingual identities shaped instructional choices, classroom language practices, and relationships with students. Participants navigated systemic constraints such as language policies and curricular obligations while leveraging their bilingualism as an instructional and socioemotional support tool. They implemented culturally sustaining practices to create affirming classroom environments and viewed their bicultural bilingual identities as professional assets that enhanced their teaching efficacy and cultural responsiveness. The findings suggest the need for teacher education programs and school systems to recognize and support the social and linguistic capital that bicultural bilingual teachers bring to the field of music education more fully
Two Particle Angular Correlation Functions of Neutral and Charged Kaons in Pb–Pb collisions at √ sNN = 5.02 TeV with ALICE Experiment at the Large Hadron Collider
No full textIn ultra-relativistic heavy-ion collisions, hadronic matter undergoes transition to a deconfined phase of quarks and gluons, a state of matter widely known as quark-gluon plasma (QGP). It is believed that the universe existed in this new state of QCD matter microseconds after the Big Bang. The QGP state is transient, undergoing collective expansion and eventually hadronizing. Phase transitions in QCD are also realized in terms of chiral symmetry breaking/restoration. In the confined hadronic phase, chiral symmetry is broken and it is expected to be restored in the deconfined QGP phase. This was verified by Lattice QCD calculations at finite temperatures and zero densities. There have been several experimental evidences for the deconfinement phase transition while the chiral phase transition remains as a mystery for high energy physicists. Observing signals of chiral phase transition is as fundamental a feature of QCD as quark or color confinement and asymptotic freedom. Recent ALICE measurements have demonstrated large dynamical correlations between produced neutral and charged kaons in Pb--Pb collisions at TeV. These integrated correlations cannot be described by conventional heavy-ion models, such as HIJING, EPOS-LHC and AMPT. On the other hand, the ALICE measurements can only be described by invoking the presence of condensate. Two candidates for such a condensate are the Disoriented Chiral Condensate (DCC) and Disoriented Isospin Condensate (DIC). They both arise from chiral symmetry restoration in the QGP, which breaks during the phase transition to form a condensate that coherently emits hadrons. To further investigate these anomalous kaon correlations, a differential measurement of two-particle angular correlation functions of charged and neutral kaons as a function of and in Pb--Pb collisions at TeV is performed. The correlations involving oppositely charged kaons were computed as a baseline. These experimental correlations were then compared with HIJING and AMPT model predictions to determine if the angular correlations exhibited any anomalous behavior
Point Set Identification of Genetic Sequences
No full textThe human genome requires high fidelity of intron and exon identification for transcription, splicing, and translation. These steps are capable of identifying the exons (expressed portion) and introns (non-expressed portion) that constitute a gene. Splicing is of particular interest as it targets and removes introns, which comprise 97% of the length of a gene. The regulation of splicing is not fully understood: many genes have the possibility to be expressed in multiple ways through the process of alternative splicing, greatly increasing the complexity and number of proteoforms within the genome. Failures within this process however can result in tumorigenesis, Parkinson’s disease, and a variety of other genetic diseases. In an effort to understand and explore the possible regulations that influence splicing and expand our understanding of cancer, a simple numerical transform to represent the alphabet of nucleotides is introduced. This simple transform, inspired by studies of the H´enon attractor, looks posterior and anterior from a given location, creating point sets of the sequence. These point sets allow for calculating the General Moments, General Dimensions, and F(α) curves of the genetic sequences to study their multifractal nature. Additionally three machine learning algorithms are built to classify the points sets: one to distinguish between exons and introns, and two to distinguish between cancerous fusion genes and normal junctions, one looking only a few nucleotides away from the breakpoint and one looking many nucleotides away. Multifractal comparisons of exons, introns, and intragenic regions (the regions in between genes) show significant differences between them, and building a machine learning algorithm to classify them was able to achieve a 98% accuracy. Normal and cancerous breakpoints were then compared, looking at 20 and 250 nucleotides on either side of the exon-exon boundary (for normal sequences) and the cancer breakpoint (for cancerous ones). They too show significant differences. With that established, their machine learning algorithms were able to achieve 97% and 99% accuracy in identifying cancer vs normal images
The Two-Dimensional Nature of Political Cynicism: Formation and Effects of Cynical Views
No full textPolitical cynicism, once thought to remain static in the public, has been steadily increasing along with levels of affective polarization and negative partisanship. Such strong attitudes of politicians of one party over the other call into question the efficacy of survey items which ask about politicians without prompting partisanship. I argue that political cynicism is a two-dimensional construct, with respondents differing in their cynical evaluations of the two parties. From this I advance partisan measures of political cynicism which makes explicit the party to survey respondents. I show that broad-based measures of political cynicism primarily capture feelings of the out-party. This finding raises questions as to the causes and effects of political cynicism. Evidence from two experiments where I manipulate the coverage of a fictitious mayoral campaign provide further justification for these measures and adds to our understanding of the formation of cynicism in a polarized era. Specifically, I find that candidate-specific cynicism drastically increases in response to game-frame coverage while increases in broad-based measures are minimal. Lastly, I address a debate in the literature as to the effect of political cynicism on voter turnout. Drawing on theories of political alienation and spatial proximity, I argue that partisan cynicism is instrumental in understanding when and how cynicism reduces voter turnout. When cynicism is one-sided it serves as a provocation to vote by clarifying the choices between two candidates. However, as cynicism becomes balanced toward both parties and candidates, this provocation is lost
Frailty Mediating the Progression of Mild Cognitive Impairment to Dementia: A Comparative Study
No full textBackground: As the aging population continues to diversify, there is an increased public health concern regarding the health disparity gap among underrepresented populations. Studies have investigated frailty status as a risk factor for Mild Neurocognitive Disorder (MCI) and Major Neurocognitive Disorder (dementia). However, limited research exists regarding the conversion of MCI to dementia comparatively among racial/ethnic groups using frailty as a mediator. Purpose: To investigate frail status (prefrail and frail combined) as a mediator of MCI progression to dementia comparing White Americans and underrepresented groups with aims to: 1) determine the frequency of frail phenotype and the relationship of frail status and MCI and dementia and 2) to determine the conversion rate from MCI to dementia, investigating frail status as a mediating variable with age, sex, education, and race/ethnicity. Methods: Data was obtained from the Baylor College of Medicine Neuropsychology Clinic and the Alzheimer’s Disease and Memory Disorder’s Center registries. Frailty was based on Fried et al.’s (2001) frailty phenotype and derived from participants’ neurological evaluation and items from self-report measures. Descriptive statistics, prevalence rates, and percentage of those who converted from MCI to dementia were examined. Chi-square test was used to examine frailty status and cognitive status (MCI, dementia). Binary logistical regression was used to examine these variables across racial groups. Mediation analysis was used to investigate demographic variables, frailty status, and cognitive status. Results: Frailty status was not associated with cognitive status. Frailty was not a significant mediator between sex, education, and racial/ethnicity as predictors, and the conversion from MCI to dementia. Age was found to be a significant predictor of non-frail (robust) and frail status for White Americans. Conclusion: Multiple factors (e.g., underpower) contributed to the null findings. However, age predicting frail status for White Americans may suggest a positive effect of early interventions in this cohort
Molecular Structure-Property Relations in Chitosan Nanocomposite Bulk and Thin Films
No full textNatural biopolymers such as chitin have become promising replacements for synthetic petroleum-based polymers in food packaging and bioengineering fields. Amine based chitin and its deacetylated derivative chitosan are paving the research and development fields for their stable physical properties and biodegradable nature. Chitosan has randomly distributed β-linked D-glucosamine and N-acetyl-D-glucosamine and the protonation of NH2 groups increases affinity towards moisture. As the research focus of my Ph. D. work, we proposed chitosan-based nanocomposite thin films as a sustainable alternative to thermoplastic and inorganic semiconductors-based humidity sensors. Here, thin chitosan nanocomposite films of 50-400nm thickness were coated on Si substrate. Rapid swelling of the thin films in humid environment was spontaneously identified with visible changes in color explained by thin-film interference phenomenon. Over the full relative humidity range of 95%, film thicknesses increase 50% compared to dry state, confirmed by in-situ interferometric techniques. The response to humidity change was ultrafast (~5s) and the absorption-desorption of moisture exceeded 10 cycles with confidence. The moisture absorption kinetics followed non- Fickian type diffusion pathway with increasing trend of diffusion coefficient relative to thin film thickness. By blending nanofillers like graphene oxide, polyhedral oligo silsesquioxanes, and nanoclay in a small amount (0.5-5wt%), the rapid moisture absorption had similar diffusion kinetics while maintaining film stability owing to interaction of oxygen-rich groups in nanofillers and NH2 groups of chitosan. The nanocomposite thin films had reliable and stable Young’s modulus measured and mapped by Peak Force Tapping mode of Atomic Force Microscopy. Also, as part of my Ph. D. work, we developed a sustainable route to extract chitin and emphasized modifying the hydrophilic nature of bulk chitosan films of ~80µm to preserve its tensile properties by controlled chemical crosslinking using glutaraldehyde and reducing water uptake by 1000%. The dependable mechanical properties of the bulk films and the highly sensitive humidity dependent colorimetric property of chitosan nanocomposite thin films enables its potential as a biodegradable food packaging material and as a sensor for monitoring quality assurance systems benefiting several industries
Evaluating Pavement Deterioration Rates Due to Flooding Events Using Explainable AI
No full textFlooding can damage pavement infrastructure significantly, causing both immediate and long-term structural and functional issues. This research investigates how flooding events affect pavement deterioration, specifically focusing on measuring pavement roughness by the International Roughness Index (IRI). To quantify these effects, we utilized 20 years of pavement condition data from TxDOT’s PMIS database, which is integrated with flood event data, including duration and spatial extent. Statistical analyses were performed to compare IRI values before and after flooding and to calculate the deterioration rates influenced by flood exposure. Moreover, we applied explainable artificial intelligence (XAI) techniques, such as Shapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME), to assess the impact of flooding on pavement performance. The results demonstrate that flood-affected pavements experience a more rapid increase in roughness compared to non-flooded sections. These findings emphasize the need for proactive flood mitigation strategies, including improved drainage systems, flood-resistant materials, and preventative maintenance, to enhance pavement resilience in vulnerable regions
Houston's Current Resourcing Landscape from the Perspective of Local Community Health Workers
No full textCommunity Health Workers (CHWs) play a vital role in facilitating healthcare access and resource navigation within underserved communities. However, limited research examines their experiences within local U.S. contexts. This study explores the role of CHWs in Houston, Texas, focusing on their impact, training, and challenges in resource navigation. Through semi-structured interviews with CHWs from the University of Houston Community Health Worker Initiative (UH CHWI), this research identifies key themes, including the effectiveness of CHW training, trust dynamics with community members, and systemic barriers to resource accessibility. The study highlights the need for improved training programs, streamlined administrative processes, and enhanced resource distribution strategies to optimize CHW effectiveness. Future research aims to expand the sample size, incorporate focus groups with key stakeholders, and conduct cross-analyses of CHW experiences and resource availability data. By addressing these challenges, this research seeks to inform policy and programmatic interventions that enhance community health strategies, ultimately fostering more equitable and sustainable healthcare outcomes.Honors Colleg
Evaluating KeyBERT for Keyword Extraction in Lecture Videos
No full textExtracting key concepts from lecture videos is essential for efficient learning, but manual keyword identification is time-consuming. This study evaluates KeyBERT, a BERT-based keyword extraction model, against VideoPoints, an automated lecture video management system, to assess their effectiveness in extracting keywords from OCR-derived text of biology lecture videos. The extracted keywords were compared to instructor-provided ground truth to measure performance. Results show that KeyBERT performs well for broad anatomical terms but struggles with fine-grained biological classifications and multi-word key phrases. While lemmatization and stemming improved KeyBERT's recall and F1-score, VideoPoints consistently outperformed it across all evaluation metrics, suggesting that structured keyword extraction is more effective for lecture content. KeyBERT also exhibited limitations in recognizing domain-specific biological terms and generated redundant or fragmented phrases. Challenges include OCR limitations, KeyBERT's context sensitivity, and subjectivity in ground truth annotations. Future work will explore fine-tuning BERT models for educational content, improving OCR preprocessing, and integrating a hybrid approach combining KeyBERT's semantic extraction with VideoPoints' structured segmentation to enhance accuracy. This research highlights the need for domain adaptation in NLP-based keyword extraction to improve lecture video summarization and navigation.Computer Science, Department ofHonors Colleg
Evaluating the effect of weight loss on releasing chronic rhinosinusitis symptoms in obese population using machine learning
No full textChronic Rhinosinusitis (CRS) is an inflammatory disease of paranasal sinuses that substantially reduces patients' quality of life and poses a significant economic burden. Emerging evidence has identified a correlation between obesity and increased risk of CRS; however, the effect of weight loss on CRS symptom relief remains underexplored. This study investigates whether significant weight reduction is associated with decreased CRS symptom burden in obese individuals, leveraging data from the National Institutes of Health's All of Us Research Program. Using a cohort of 1,836 obese individuals with confirmed CRS diagnoses, this study analyzes changes in CRS diagnosis frequency before and after significant weight loss (defined as a decrease in BMI of at least five units). CRS symptom burden was quantified using the CRS diagnosis ratio (number of diagnoses per year), and statistical testing was performed using paired t-tests across multiple observation windows. Additionally, machine learning models-including Logistic Regression, Decision Tree, Random Forest, and XGBoost-were developed to predict symptom improvement using demographic and comorbidity features. Mutual information was employed for feature selection, and models were optimized through hyperparameter tuning. Results demonstrated that weight loss is statistically associated with a reduction in CRS diagnoses, especially within a two-year post-weight-loss window. However, the machine learning models struggled to accurately predict symptom improvement, with limited AUC-ROC scores despite improvements in accuracy and precision after tuning and feature selection. This study underscores the potential benefits of weight loss in managing CRS symptoms and highlights challenges in prediction modeling. Future research should incorporate a broader range of clinical and behavioral features to improve predictive performance and deepen understanding of the mechanisms linking obesity and CRS.Honors CollegeIndustrial and Systems Engineering, Department o