62588 research outputs found
Sort by
Calculus Gate: The Impacts of Self-Beliefs and High School Math Experience Upon Performance on College Math Placement Exams
No full textAbstractAlthough science and engineering (S&E) fields continue to grow, certain groups including first-generation students and women remain underrepresented among S&E degree recipients. Mathematics, specifically calculus, is often the gatekeeper to entering STEM majors that open the pathway to financially lucrative careers in S&E. Although incoming college students typically take math placement exams to determine their first college mathematics course, the use of noncognitive measures, such as math self-beliefs, in addition to cognitive skill assessments has been recommended due to high rates (25%-33%) of misplacement using solely skill assessments. A self-beliefs survey about math performance was collected for the 2022 cohort (N = 333) and 2023 cohort (N = 373) from incoming first-year college students at a small liberal arts institution. The first study compared a hypothesized four-factor model (self-efficacy, expectancy, value, and cost) and a three-factor model (ability beliefs, value, and cost) for the math self-beliefs data. The theoretical framework for the three-factor model was Eccles\u27s & Wigfield\u27s situated expectancy-value-cost theory (SEVT) while the four-factor model was a composite of SEVT theory and Bandura\u27s social cognitive theory. Confirmatory factor analysis indicated the four-factor model was a better fit of the data for the 2022 cohort, which was independently validated with the 2023 cohort. Factorial invariance was also tested across first-generation status and gender to ensure measurement equivalence for both cohorts. The analysis indicated that the four-factor model for self-beliefs achieved strict invariance indicating that the measure\u27s structure, factor loadings, item intercepts, and measurement errors were invariant across first-generation status and gender for the 2022 cohort and subsequently confirmed with the 2023 cohort. The second study utilized structural equation modeling (SEM) analysis with full information maximum likelihood estimation to compare two models representing the relationships between the self-belief latent factors (four-factor model from the first study), high school preparation, STEM interest, gender, first generation status, and performance on a math placement exam. Model 1 represented math performance by a total placement score (Part I algebra score combined with Part II emergent Calculus skills) while Model 2 represented math performance by an endogenous latent variable with three indicator variables (Part I Score, Part II Score, and Math Experience). SEM analysis indicated that Model 1 was a better fit for the data for the 2022 cohort and validated with the 2023 cohort. Model 1 was subsequently used to test for a moderation effect by first-generation status on the measurement weights and structural weights. No moderation effect for first generation status was discovered for Model 1 with either the 2022 cohort or the 2023 cohort. A follow up partial moderation analysis was run that supported the findings of the full moderation analysis for the 2022 cohort and the 2023 cohort. Keywords: Mathematics placement exam performance; Self-beliefs; Mathematics preparation; First-generation status; Gende
"Frogs don\u27t wear sweaters": The Influence of Anthropomorphized Characteristics on Children\u27s Biological Learning and Reasoning
No full textThis study aimed to better understand the influence of anthropomorphism on children\u27s learning outcomes. The findings in the literature are mixed, with some showing that anthropomorphism hinders, equally supports, or even enhances learning outcomes. The present study revisits this question by examining whether nuanced levels of anthropomorphism in language and illustrations may differentially influence children\u27s learning.We designed two children\u27s books to convey the biological principle that some animals migrate (1) due to food scarcity caused by season changes or (2) to find safe conditions to raise their young. A total of 80 4-year-olds and 76 5-year-olds were randomly assigned to one of six conditions. Children heard stories that either included (1) realistic or (2) anthropomorphized language paired with (1) realistic, (2) subtle anthropomorphized, or (3) extreme anthropomorphized illustrations. We presented the Food Scarcity and Raising Young stories one time for each child in counterbalanced order. Following the presentation of each story, children were given tasks that measured their Story Recall, their Generalization (applying the rule of migration to real animals), and Anthropomorphized Reasoning.The results suggest that children\u27s age and the difficulty of the educational materials were key factors impacting the relationship between levels of anthropomorphism and learning. The Food Scarcity story was somewhat easier for children than the Raising Young story, and there was some indication that the difficulty of the story influenced the relationship between anthropomorphism and learning. For the easier story, anthropomorphism generally led to equivalent learning outcomes as did realistic materials, and there was some indication that subtle anthropomorphism in the illustrations somewhat enhanced children\u27s learning. For the more difficult story, anthropomorphized language sometimes hindered 4-year-olds\u27 recall and generalization relative to realistic language. Finally, there was no evidence that exposure to anthropomorphism induced anthropomorphized reasoning about real animals.Overall, the results should be reassuring for parents and teachers who share anthropomorphized materials with their children. While anthropomorphism may hinder learning for some children (particularly if the content is difficult or the children are younger), this study provides evidence that anthropomorphism is generally supportive of children\u27s learning outcomes and does not induce anthropomorphized reasoning
Exploring the Food-Energy-Water Nexus: Insights from Co-evolution in Coupled Natural-Human Systems
No full textBalancing the rising demands of energy and food with sustainable water resource management under climate change presents a significant challenge. This complexity is heightened within a coupled natural-human systems (CNHS), where heterogeneous human activities affect the natural hydrologic cycle and vice versa. This dissertation explores the co-evolution in CNHS to advance our understanding of interactions between food, energy, water (FEW) sectors. This dissertation investigates the human dimension of the FEW nexus through a workshop with regional government agencies and a comprehensive survey of residents in a transboundary basin (the Columbia River Basin between the US and Canada), which illuminates both institutional and residential aspect of the FEW resources. The workshop offers insights into current policy and resource management, while the survey explores residents\u27 perceptions and understanding of the FEW nexus and associated regulations. By synthesizing information from these dual avenues, the result provides a complete picture of the human dimension of the FEW nexus and potentially reconciles varying stakeholder priorities, paving the way for enhanced FEW resource management. Moreover, this dissertation develops a distributed modeling framework that fully integrates an agent-based model (ABM, a human system model) which simulated spatially distributed human behaviors (i.e., heterogenous irrigation decisions), into a large-scale, process-based distributed hydrologic model (a natural system model) to consider endogenous human behaviors in the hydrologic cycle. In addition to simulating adaptive human behaviors under historical conditions, the integrated modeling framework is applied to evaluate the future FEW nexus. Utilizing the integrated modeling framework, this dissertation assesses the highly uncertain future climate change effects on FEW sectors in the Columbia River Basin at different spatial scales. The decision-scaling framework, an ex-post scenarios analysis method, is employed to quantify climate change uncertainties. This method identifies acceptable system performance in a broad range of future climate conditions, rather than focusing on predictions of future climate that are subject to various climate modeling and downscaling approaches\u27 uncertainties. The result of the FEW nexus under climate change offers insightful information for shaping long-term water management policies. In sum, this dissertation contributes to understanding the co-evolution in CNHS for the FEW nexus. It underscores the need for future research to establish a more comprehensive framework across diverse application topics to enrich the picture of this dissertation
A cloud-edge collaboration approach to drone cyber-physical systems
No full textCyber-physical systems (CPSs) play an important role in the industrial area. CPSis the organic integration of the computation and physical components. It consists of computing devices communicating with each other and interacting with the physical world through sensors and actuators. In my dissertation, we aim to develop a CPS for drones to facilitate the development of drone application scenarios using both edge servers and a cloud-based command center to schedule and coordinate the whole system. In our drone control CPS, drones are controlled by edge servers or ground stations, and ground stations are connected using TSN networks. Drones, edge servers, and ground stations are interconnected using local wireless networks in each zone. TSN networks are used to transfer the essential data for the handover process when drones fly across the boundaries of two edge servers. TSN is used to guarantee the real-time handover of control from one edge server to another edge server. A cloud command center is used to collect global information. The target of this design is to provide a drone control CPS that has real-time performance, high computational power, real-time monitoring and data collection, and long-range control to fulfill the increasingly stringent requirements of drone applications.Based on this design, we propose several research problems on different components of the system. These research problems include i)analyses of the real-time performance; ii) real-time handover process; iii) real-time scheduling for task execution.We use network calculus, which is a prevalent mathematical model to calculate the backlog for switches and the work-case delay bounds for flows, for the analysis of the real-time performance of the drone control system. To guarantee the real-time handover process, TSN-VM (violation mitigation) is proposed to monitor and mitigate the scheduling violations in a distributed manner in TSN networks. The last part of the research is the real-time scheduling for Same-Day Delivery (SDD) with dynamic numbers of drones. The novelty of this research is that we take the drone share and drone crashes into consideration. Both scenarios will force a changing number of drones during the shift. The consideration of drone crashes is more reality-conforming compared with the assumption of a fixed number of drones, while drone share among different depots increases drone utilization and, thus, the efficiency of the system.The research on the three aspects of the system. The target of the research is to prove the validity, robustness, stability, efficiency, and real-time performance of the system
Exploring Human Emotion & Vulnerability through Indian Classical Dance
No full textWhat does it mean to be human? Given the rise of Artificial Intelligence and conversations about machines replacing humans, it got me to reflect on what exactly it means to be human. For me, it means being flawed, emotional and vulnerable beings. In our fast paced lives, we often forget to pause and actually feel and experience whatever emotions we\u27re feeling – whether thats anger, sadness, beauty, love or gratitude and through my project, I want to explore these emotions through the medium of dance. I truly believe that our ability to feel these emotions and be vulnerable is what makes us human and special and I want to celebrate that through dance
Personalized Algorithms: AI\u27s Role in Information Overload
No full textIn today\u27s always-connected society, AI-driven recommendations in social media have become a staple of internet usage, significantly shaping users\u27 consumption experiences. This research investigates the extent to which such personalized algorithms may induce information overload and anxiety in regular users, contrary to their intended purpose. Drawing on prior literature exploring the potential negative effects of personalized algorithms, this paper presents findings from a controlled lab experiment conducted using the TikTok platform. Participants (N=44) engaged with their own personalized For You Page, as well as a neutral, human-curated feed, over the course of two separate days; measures of perceived state anxiety and information overload, as well as pre- and post-experiment heart rate, were recorded for each session. In support of the hypothesis, results indicate that participants experienced significantly higher levels of overload and anxiety when exposed to the personalized AI-driven recommendations, compared to the control condition. These findings emphasize the importance of understanding prominent algorithms and their effects on users\u27 well-being, essential both for platform development and the formation of technology policy
Surface Polaritons in Bilayer Graphenes: Cross–Correlation of Sub–Diffraction Optical Characterization Methods
No full textLayered 2D materials have been the focus of recent study due to the rich fundamental physics of these systems, as well as for their potential in surface andinterface engineering, and for novel quantum electronics (spin–, valley–, strain–,
and twist–tronics). Since their entire structure is that of an exposed surface, they
are also highly susceptible to local variance from environmental factors such as
strain and doping, allowing their electronic and phononic states to be readily modified. Graphene multilayers are excellent test–beds for studying these systems, as
graphene itself is relatively well understood. This thesis explores the electronic and
optical characteristics of graphene bilayers and trilayers in different stacking configurations, with particular attention on the near–field optical response and surface
bound modes. To this end, an sSNOM microscope was used to directly probe the
optical conductivity of graphene in different configurations, and a suite of homemade computational tools was developed to process the data, and map physical
properties to the imaged surface. Colocalized Raman microscopy was used in tandem to identify sample configurations, so that spectral characteristics from sSNOM
could be related to the correct graphene structures. It was also used to evaluate
several fabrication techniques used to produce new samples for study. A method to
measure the surface optical states of a novel system comprised of a gold grating on
top of a SiC/2D Ag/Graphene heterostructure is demonstrated. Through measuring
the polaritons in this sample with sSNOM, and processing the maps with a novel
computational procedure, it is shown to be possible to infer the excitations of 2D
Ag. Finally, a custom microscope is constructed to perform angle–resolved spectroscopy, with the ultimate goal of acting as a complementary system to sSNOM,
1with the ability to measure polariton dispersion in the far–field for comparison to
near–field measurements. This thesis will show that the host of tools, computational techniques, and methodologies presented herein form a cogent basis for the
measurement of 2D material physics.</p
Advancing Supercapacitive Swing Adsorption of Carbon Dioxide through Electrode Design, Charging Protocols, and Oxygen Stability Studies
No full textThe escalating levels of anthropogenic carbon dioxide in the atmosphere poses a significant challenge for our society. To address this challenge, it is imperative to swiftly develop efficient technologies that can capture and concentrate CO2 from dilute sources in a cost-effective, environmentally friendly, and energy-efficient manner. Current strategies for carbon capture and concentration have several limitations, including sorbent toxicity, energy intensive thermal and/or pressure swings, poor cyclability, selectivity, and capacity retention. These issues are mainly due to thermal degradation, volatility, and the reactivity of sorbent materials with oxygen. To address these issues, supercapacitive swing adsorption (SSA) of CO2 is of particular interest. SSA is an electrochemical carbon capture technology capable of capturing and concentrating CO2 from a gas mixture upon charging and discharging of the electrodes. SSA offers significant advantages over existing carbon capture methods, including high selectivity, longer sorbent lifetime, faster charge/discharge cycles, high round-trip energy efficiency, and the use of inexpensive and environmentally benign materials. However, the CO2 adsorption capacity of SSA reported prior to the research presented in this thesis was less than 100 mmol.kg-1, at least one order of magnitude lower than competing carbon capture technologies. Moreover, the energetic and adsorptive performance of SSA was only investigated with 15% CO2/85% N2 gas mixtures, and the influence of oxygen, a major component of flue gas and air, was not known. Advancing SSA technology required the development and investigation of new materials with greatly improved CO2 adsorption capacities, in-depth understanding of the factors necessary for performance improvements, understanding SSA performance under different voltage windows, and monitoring SSA performance under oxygen environments.This thesis discusses the development and characterization of new biomass and non-biomass-derived activated carbon electrodes for improved supercapacitive swing adsorption of carbon dioxide under different voltage windows with oxygen and without the presence of oxygen in the CO2/N2 gas mixture. Chapter 1 provides an overview of the increasing need to develop energy efficient and cost-effective CO2 capture technologies, existing carbon capture methods (non-electrochemical and electrochemical), challenges with existing carbon capture methods, the history of supercapacitors and supercapacitive swing adsorption, and the outline of this thesis. Chapter 2 covers the fundamental principles of different physicochemical and electrochemical characterization techniques used in this research to investigate and compare the surface area, porosity, surface functionalities, capacitances, and resistances of different types of carbons. Chapter 3 reports six different types of activated carbons derived from biomass, coal, coke, and carbide sources, and provides the relationship between higher capacitance and improved CO2 sorption capacities. Chapter 4 covers a simple, one-step synthesis procedure to prepare garlic roots derived activated carbons and investigates SSA at higher voltage windows. Chapter 5 reports the critical role of oxygen on the energetics and adsorptive performance of SSA. Chapter 6 provides the outlook and future directions to further advance SSA towards commercialization.</p
IMAGE GUIDED INTELLIGENT IN-SITU MICROFABRICATION
No full textMicrofluidics has provided the experimentalists with many advantages in fields of biology, chemistry, medicine, materials, etc. But it seems to be experiencing a noticeable disparity in its applicational advancement compared to the rapid refinements and improvement in conventional and competing technologies. A recent review by some of the pioneers in modern microfluidics concluded by saying that "In simple terms, a microfluidic tool must make a persuasive case for adoption on the basis of factors such as analytical performance, usability, and information yield." Indeed, a rather obvious but important fact is that, if users are not given a convincing edge over an existing methodology, it often becomes impractical to adapt. Not surprisingly, while applications like droplet microfluidics and DNA amplifications have seen a widespread adaptation and fast commercialization, researchers still prefer to stick to conventional methodologies while performing electrophoresis experiments. Apart from the convenience of experiment, microfluidics in this case simply does not facilitate any novelty for researchers to switch. It is equally important if not more to focus on the application rather than the technology itself while developing a new method. Against this backdrop, my research carves out a niche in the development of novel microfabrication and microfluidic systems.</p
Estimating the Internal Layers of Greenland Ice Sheets with Machine Learning Methods
No full textThe Greenland Ice Sheet, with its vast expanse and pivotal role in global climate dynamics, presents a significant challenge for accurate estimation of its internal layers. This thesis explores novel methodologies for improving the estimation of internal ice sheet layers, leveraging machine learning techniques and integrating physical principles into computational models. The research focuses on two primary approaches: Long Short-Term Memory (LSTM) networks and Neural Operators.The first part of the thesis investigates traditional LSTM models and Physics Informed LSTM variants, aiming to enhance the resolution and accuracy of internal layer estimations. The second part explores the emerging field of Neural Operators, including Fourier Neural Operators (FNO), Koopman Neural Operators (KNO), Tensorized Fourier Neural Operators (TFNO), and Spherical Fourier Neural Operators (SFNO), which offer a versatile framework for capturing complex spatiotemporal patterns in ice sheet data.
Through a comprehensive review of existing literature and the development of innovative methodologies, this thesis contributes to advancing our understanding of the Greenland Ice Sheet\u27s dynamics and its implications for global climate change. The findings provide valuable insights for climate scientists, policymakers, and stakeholders, guiding efforts towards informed decision-making and sustainable environmental management.
</p