62970 research outputs found
Sort by
American and Chinese Math Teachers’ Epistemological Beliefs: One of the Reasons for Students’ Discrepancies in Math Performances
No full textThis study conducted a narrative synthesis of 15 relevant studies to explore how mathematics epistemological beliefs contribute to differences in student mathematics performance between the United States and China—where China consistently outperforms the U.S. on standardized assessments, yet the U.S. produces more professional mathematicians. Using the Narrative Synthesis of Systematic Reviews (NSSR) approach, the findings were analyzed through the lens of Ernest’s framework of mathematics epistemological beliefs. Data were retrieved using the PRISMA protocol (Moher et al., 2015) and analyzed through thematic coding.
The findings, interpreted through Ernest’s perspectives on the problem-solving view, Platonist view, and instrumentalist view of mathematics, revealed notable differences in the mathematics epistemological beliefs among K–8 mathematics teachers in the two countries. Most American K–8 teachers tended to hold a problem-solving view of mathematics, while most Chinese K–8 teachers aligned more closely with a Platonist view. These divergent beliefs influenced their preferred instructional approaches, which
in turn contributed to differing student outcomes in mathematics performance.
This study aimed to encourage educators to broaden their perspectives on
mathematics and its teaching by considering alternative epistemological orientations. Cross-national learning between teachers may help improve instructional practices. Further research is recommended to explore how a balanced integration of problem solving and content mastery can be achieved in mathematics education
Unsupervised Deep Learning Image Segmentation Approaches for Echocardiogram Analysis and Clinical Risk Prediction
No full textThis dissertation presents two complementary approaches to artificial intelligence to advance cardiovascular medicine: an unsupervised deep learning method for segmenting echocardiogram images and a multimodal framework to predict survival outcomes in Left Ventricular Assist Device (LVAD) recipients.
The first contribution addresses the challenge of segmenting cardiac structures in echocardiographic images without relying on manually annotated datasets. We developed an unsupervised approach that combined a U-Net-like architecture with a composite loss function that includes reconstruction, contour regularization, and similarity components. The method was enhanced with 3D Watershed post-processing to isolate the left ventricle for the calculation of the ejection fraction. Evaluation in the EchoNet Dynamic dataset demonstrated competitive performance with supervised methods (Dice coefficient: 0.8225, IoU: 0.7071) while using significantly fewer labeled data points.
The second contribution explores the integration of Electronic Health Records (EHR) and echocardiogram video features to predict LVAD survival. A multimodal pipeline that extracts features from segmented echocardiogram videos using convolutional neural networks and combines them with preprocessed EHR data. Survival analysis was performed using Cox Proportional Hazards and Random Survival Forest models. Exploratory analysis in a cohort of 31 LVAD patients showed that models incorporating segmented echocardiogram characteristics achieved the highest predictive performance (C-index ~0.71), outperforming models using raw video features or EHR data alone.
The clinical implications of this work include the potential for automated echocardiogram analysis in resource-limited settings and improved risk stratification for LVAD candidates. The unsupervised segmentation approach addresses the scarcity of labeled medical imaging data, while the multimodal prediction framework demonstrates the value of integrating diverse data sources for personalized medicine.
Future research directions include expanding training datasets, investigating semi-supervised learning approaches, and validating the methods in larger, multicenter cohorts to enhance clinical applicability and generalizability
“Improved by Design?” Essays on the Causal Effects of Agricultural Policy, Women’s Representation, and Economic Reform
No full textGlobal development initiatives have long prioritized the identification of policies that promote economic growth and expand opportunities for historically marginalized populations. In pursuit of these goals, policymakers and international institutions have implemented various reforms to boost agricultural productivity, enhance women’s political participation, and stimulate broader economic development. However, the success of these reforms often hinges on how they interact with local institutions and contexts.
This dissertation critically investigates the causal effects of targeted policy interventions across three distinct yet thematically linked policy domains: agricultural policy, women’s political representation, and economic reforms associated with the Washington Consensus (WC). Each chapter employs a distinct causal inference method, including Synthetic Control, Callaway and Sant’Anna Difference-in-Differences estimator, and matching techniques to evaluate how policy interventions in three areas perform in practice.
By integrating causal inference methods with a focus on institutional and contextual adaptation, this dissertation contributes to understanding how international policy models vary in effectiveness depending on these contexts
Does the tail show when the nose knows? Artificial intelligence outperforms human experts at predicting detection dogs finding their target through tail kinematics
No full textDetection dogs are utilized for searching and alerting to various substances due to their olfactory abilities. Dog trainers report being able to ‘predict’ such identification based on subtle behavioural changes, such as tail movement. This study investigated tail kinematic patterns of dogs during a detection task, using computer vision to detect tail movement. Eight dogs searched for a target odour on a search wall, alerting to its presence by standing still. Dogs’ detection accuracy against a distractor odour was 100% with trained concentration, while during threshold assessment, it progressively reached 50%. In the target odour area, dogs exhibited a higher left-sided tail-wagging amplitude. An artificial intelligence (AI) model showed a 77% accuracy score in the classification, and, in line with the dogs’ performance, progressively decreased at lower odour concentrations. Additionally, we compared the performance of an AI classification model to that of 190 detection dog handlers in determining when a dog was in the vicinity of a target odour. The AI model outperformed dog professionals, correctly classifying 66% against 46% of videos. These findings indicate the potential of AI-enhanced techniques to reveal new insights into dogs’ behavioural repertoire during odour discrimination.S.M.P. was supported by the Austrian Science Fund (FWF) Grant DOI 10.55776/I5052. G.P. was supported by a doctoral grant from the University of Parma and the FIL Bando Ateneo 2022 fund from the University of Parma to Paola Valsecchi. This work has benefited from the equipment and framework of the COMP-HUB and COMP-R Initiatives, funded by the ‘Departments of Excellence’ program of the Italian Ministry for University and Research (MUR, 2023). G.M. was supported by the Data Science Research Center of the University of Haifa. T.M. was supported by the Austrian Science Fund (FWF) Grant DOI 10.55776/P37052 and by funds from FCT—Fundação para a Ciência e Tecnologia, I.P., in the context of the R&D Unit: UID/04810—William James Center for Research
When Confucianism Meets Neoliberalism — How Chinese Graduate Instructors Establish Teacher Identity in U.S. Writing Classrooms
No full textThis dissertation explores Chinese graduate instructors (GIs) teacher identity formation
from theoretical foundations of linguistic justice in higher education, systemic oppression in
neoliberal education setting, and approaches to linguistic and cultural differences in writing
studies. I developed this study in response to my own experiences as a Chinese GI in writing
studies, challenges of nonnative English speaking (NNES) instructors in developing their teacher
identity in US writing classrooms, and problems of writing program administration and
education for international graduate students.
Adopting a feminist methodology, this study tries to address these three research
questions: (1) How do Chinese GIs perceive their teacher identity as NNES writing instructors?
(2) What challenges have they experienced in establishing their identity as writing teachers?
How do they cope with them? And (3) How have they grown professionally and personally
through the process of teacher identity construction? Utilizing a qualitative research approach,
data collection comprises three phases: phase 1: short in-take survey with each of the 16
participants at the beginning of the research (n=16); phase 2: semi-structured interview lasting
60-90 minutes with each of the 16 participants (n=16); and phase 3: teaching documents from
the participants collected after the interview (n=7). Research data also includes informal
communication with the participants and my research memos based on ongoing analysis.
The research findings reveal how the intersection of Chinese GIs’ linguistic identity,
ethnic identity, cultural background, race, gender, and nationality play out in writing classrooms
on neoliberal campus, examine the challenges that NNES graduate writing instructors are facing
in their professional growth, and explore how NNES writing instructors’ teaching practices and
unique strengths contribute to the composition study. The dissertation concludes by offering
implications and suggestions for individual NNES GIs in their teacher identity formation and
writing programs in how they can better support their NNES GIs in their professional growth
Evaluating the Effectiveness of the PLC at Work® Process and Student Achievement in Texas
No full textThis report evaluates the impact of the Professional Learning Communities (PLC) at Work® process on student academic achievement over a three-year period in Texas public schools. Drawing on linked administrative data from the University of Houston Education Research Center (UH-ERC), the study uses a quasi-experimental approach to compare achievement trends in Model PLC at Work® Schools with those in non-designated schools. We estimate effects on math and reading performance using statewide standardized assessment data, disaggregating results by school level and student population. Findings indicate that gains in student achievement emerge gradually once schools begin their implementation of PLC at Work® practices, with the largest effects occurring in the designation year. Positive impacts are observed across our study sample (elementary and middle schools), with notable gains among economically disadvantaged students and English learners. These results suggest that the ongoing, sustained implementation of the PLC at Work® process by educators and administrators in the schools themselves is associated with measurable improvements in academic performance across a variety of grade levels, contents, and campus demographics. Given the observed trajectory of impact, school and district leaders may wish to consider long-term strategies for supporting and maintaining PLC at Work® implementation to promote continuous instructional improvement that leads to increased student achievement.
This research was supported by funding from Solution Tree. The findings and conclusions presented are those of the author(s) and do not necessarily reflect the views of the funding organization
Browse Nutritive Dynamics and Temporal Overlap of Ungulates in the Southern Cross Timbers and Prairies of Texas
No full textWhite-tailed deer (Odocoileus virginianus) are a species of ecological and economic importance in the Southern Cross Timbers Region of Texas. White-tailed deer are concentrate selectors and select for the most nutritious plant parts. However, these vary in availability and nutritive value across seasons. This study evaluated the seasonal nutritive dynamics of the preferential browse species post oak (Quercus stellata Wangenh.), skunkbush (Rhus trilobata Nutt.), winged elm (Ulmus alata Michx.), chittamwood (Bumelia lanuginosa (Michx.)) Pers. Var. albicans Sarg), elbowbush (Forestiera pubescens Nutt.), and live oak (Quercus virginiana Mill.). This study took place on a high fenced ranch near Cross Plains, Texas. Samples were collected across two ecological sites (loamy bottomland and clay loam) bimonthly for one year, starting in May 2023. Browse nutritional composition was evaluated including crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), fat, protein-precipitable phenolics (PPP), and in-vitro true digestibility (IVTD). Dormancy and growing period proved to be more influential than individual months on the nutritive dynamics across species. NDF and ADF increased in the dormancy period while IVTD decreased. During the growing season, IVTD increased and fibrous components decreased.
To compliment these nutritional findings and gain insight into habitat preferences and diel activity of native and exotic ungulates in Texas, motion activated game cameras were deployed across the ranch to monitor white-tailed deer, blackbuck (Antilope cervicapra), and gemsbok (Oryx gazella). Game cameras were monitored for a one-year period beginning in December of 2023 and ending in November 2024. Analysis of diel activity overlap between species indicated that there was no clear temporal avoidance among the species. The species diel activity was reflective of behavior in their native and free-roaming habitats. Results from our canonical correspondence analyses also indicated that environmental preferences differed among species. While white-tailed deer appeared to have no clear preferences among the environmental variables, blackbuck and gemsbok were likely to be in separate habitats that varied seasonally.
These findings increase our understanding of nutritional dynamics, habitat preferences, and diel activity of native and exotic species within niche high fence systems. Managing the environment to meet the preferential needs of multiple species can help to minimize interspecies competition, disease transmission, and sustain a healthy ecosystem
Toward a Relational View of Entrepreneurial Opportunities: A Three-Essay Dissertation
No full textAs a foundational concept in entrepreneurship research, the notion of entrepreneurial opportunity has long been criticized for relying heavily on abstract philosophical inquiry at the expense of empirical concreteness. Accordingly, the purpose of this dissertation is to advance a research program that grounds entrepreneurial opportunities in the concrete empirical world of ordinary language expressions. To do so, I utilize a three-essay format to 1) develop an empirically oriented theoretical framework for the linguistic phenomenon of entrepreneurial opportunities; 2) test the theoretical framework in an experiment; and 3) explain the results of the experiment as they apply to the concrete empirical world as represented in the language of practicing entrepreneurs. In developing the empirically oriented theoretical framework in Essay 1, I suggest that the social relations underlying the context of entrepreneurial opportunity-referenced language expressions serve as the basis for carefully constrained and disciplined theorizing about entrepreneurial opportunities. In Essay 2, I report and discuss results from a metric conjoint experimental study to establish the effect of social relations on the production of entrepreneurial opportunities as meaningful social realities for relevant entrepreneurial actions. In Essay 3, I employ a mixed-methods explanatory sequential design by analyzing secondary qualitative interview data to demonstrate how the language of practicing entrepreneurs concretely accords with and explains the theory (Essay 1) and its experimental substantiation (Essay 2)
