University of Illinois at Chicago
University of Illinois at Chicago: UIC INDIGO (INtellectual property in DIGital form available online in an Open environment)Not a member yet
21439 research outputs found
Sort by
Evidence of Student Learning of Spectrophotometry and Intermolecular Forces in the Chemistry Laboratory
No full textThis dissertation investigates how undergraduate students engage with analytical methods and foundational concepts in the general chemistry laboratory with attention to meaningful learning and mechanistic reasoning. Using these frameworks three laboratory activities explored student perceptions and explanatory reasoning in the context of laboratory learning. The first strand of the research analyzed student perceptions of spectroscopy in contrast to gravimetric analysis through the methodological framework of phenomenography and then used the phenomenographic outcome space to redesign the curricular instruction and assessment. Analysis of student reports revealed four qualitatively distinct conceptions of spectrophotometry, with meaningful learning evident where students connected spectrophotometric measurements to light-matter interactions, expressed engagement with lab process, and articulated confidence in their findings. In a novel use of the phenomenographic outcome space, the laboratory activities were restructured through revised sequencing, added preparatory conceptual work, and procedural simplification guided by the categories of description from the phenomenographic study. Additionally, assessment was explicitly aligned with the outcome space. Comparative analysis of pre- and post-intervention showed marked improvement in students’ conceptual accuracy, procedural fluency, and attitude toward spectrophotometry. The second strand analyzed student reasoning and resource activation in a laboratory activity on intermolecular forces (IMFs) where they were asked to explain the appearance of water droplets on glass slides that were modified with alkoxysilanes. The lab engages students in constructing explanations that bridge macroscopic observations and submicroscopic representations, promoting causal reasoning grounded in molecular structure and surface chemistry. This laboratory development was followed by an analyzes of students’ written explanations of their observations using a causal mechanistic reasoning (CMR) framework. Findings indicate that while students often articulated appropriate CMR in homogeneous systems, their reasoning for heterogeneous systems generally lacked submicroscopic coherence. To support this, a CMR analysis map was developed to visualize conceptual resources, levels of representation, and explanatory structure. This tool offers new insight into how students construct mechanistic explanations of surface-liquid interactions. Collectively, this dissertation contributes to chemistry education research by advancing theoretical models of student cognition and demonstrating design principles for laboratory curricular instruction
Rgg2/Rgg3 Quorum Sensing in Streptococcus pyogenes and Innate Immune Cell Responses
No full textThe immune system and microbial pathogens work against each other in many ways resulting in a never-ending dance. This thesis investigates the relationship between macrophages, a type of innate immune cell, and Streptococcus pyogenes, an important human pathogen. In Chapter 2 of this thesis, it is shown that physiologically relevant concentrations of nitric oxide generates dinitrosyliron complexes which deplete the available intracellular iron in S. pyogenes and induces iron starvation. These results give more context to the functions of dinitrosyliron complexes in biological systems which previously were poorly understood. In Chapter 3 of this thesis, it is shown how S. pyogenes suppresses pro-inflammatory responses of macrophages. Reactive oxygen and reactive nitrogen species production, chemokine production, and neutrophil recruitment are all suppressed. Expression of the Rgg2/Rgg3 quorum sensing system allows S. pyogenes to suppress these innate immune responses. Altogether, the results from these chapters explores how the immune system and microbial pathogens interact with each other to try and come out on top
Large Language Models for Reasoning: From Indirect Supervision to Self-supervision
No full textIn recent years, large language models (LLMs) have demonstrated remarkable capabilities in natural language understanding and generation. However, challenges remain in guiding LLMs toward reasoning tasks that require complex decision-making and nuanced judgment across diverse domains. This dissertation explores novel techniques to enhance the reasoning capabilities of LLMs by shifting from traditional indirect supervision methods to more autonomous self-supervision paradigms.
The first part of this dissertation presents two published works that illustrate how task-specific training with indirect supervision improves LLMs’ reasoning ability on selection-based NLP tasks. These approaches demonstrate the power of leveraging indirect supervision to strengthen decision-making in multi-option tasks such as intent detection and question answering. The second part introduces a study on developing expertise-level benchmarks to assess and improve LLMs’ reasoning in scientific (meta-)review processes, paving the way for more accurate and robust evaluation of model performance in complex, expert-driven reasoning tasks. The third part presents a plug-and-play framework that adapts off-the-shelf pre-trained encoder–decoder models — leveraging their inherent pre-trained knowledge — to process long sequences and enhance long-context reasoning ability.
Together, these contributions provide architectures, benchmarks, and methodologies that strengthen the reasoning capabilities of NLP systems, highlighting the transformative potential of generative AI in both scientific research and real-world applications.
All experiments are conducted on public datasets (UFET, BANKING77, HWU64, CLINC150, MCTest, SummScreen, GovReport, BookSum) and do not need any Institutional Review Board (IRB)
Implementation of Intermittent-robust Computing on RISC-V Architecture
No full textThe growing deployment of batteryless Internet of Things (IoT) devices powered by ambient energy harvesting highlights the need for architectures capable of sustaining correct and reliable execution under frequent and unpredictable power interruptions. This thesis addresses this challenge by proposing RISE (RISC-V Intermittent System Extensions), an architectural and ISA-level framework that enables intermittent computing on a pipelined RISC-V processor.
The proposed approach introduces four lightweight hardware modules: the Intermittent Computing Register Wrapper (ICRW), which encapsulates the processor state and tracks modifications through dirty-bit management; the Power Control Unit (PCU), which performs selective background backups of modified registers; the Restore Control Unit (RCU), which reloads saved state upon power resumption; and the Dispatcher, which transparently arbitrates memory bus usage between normal execution and backup transfers. In addition, the instruction set architecture is extended with the .ICA primitive, which allows programmers to define atomic code regions that guarantee correctness and consistency despite intermittent power supply.
RISE is designed to preserve compatibility with standard RV32I pipelines, requiring only minimal modifications to the decode stage, while maintaining full portability across different RISC-V cores. The framework avoids reliance on non-volatile elements within the processor itself, ensuring CMOS compatibility and scalability. Backup and restore operations are executed concurrently with regular computation, thereby minimizing performance and energy overhead.
The framework was implemented in Verilog HDL and evaluated through simulation and synthesis using the Xilinx Vivado 2024.1 toolchain. Experimental results demonstrate that the proposed architecture achieves efficient and reliable execution in intermittent environments. In benchmark workloads, a complete processor state backup requires on average 203 cycles (0.203 µs at 1 GHz), confirming that RISE provides a practical and effective solution for sustainable intermittent computing in energy-harvesting IoT systems
Discovering Heterogeneous Causal Effects in Relational Data
No full textReal-world and online social and interaction networks are rich sources of human behavioral data. There is growing interest in deriving causal insights from such data which is inherently relational in nature. Causal inference in relational settings has to account for interference, where a unit's outcome may be influenced by the treatments or outcomes of other units. Despite recent progress in causal inference under interference research, there has been limited attention to heterogeneous effects when different units have different responses to treatment and/or a different susceptibility to peer influence, depending on the unit's contexts. This thesis addresses this gap by defining heterogeneous peer influence (HPI) as the general interference that occurs when a unit's outcome may be influenced differently by different peers based on some underlying mechanisms involving their attributes and relationships. Understanding heterogeneity in causal effects facilitates measuring impacts of treatment policies for different subpopulations and uncovering targeted intervention policies custom to relational domains such as viral marketing strategies, risk reduction interventions for infectious diseases, and awareness campaigns for vulnerable groups in social networks.
The main objective of this thesis is to develop a framework for expressive causal modeling, sound causal reasoning, and robust estimation of individual, i.e., unit-level, direct and peer effects in the presence of heterogeneous peer influence when the mechanism of influence is unknown. First, I present my initial research focused on developing an observational study design for testing a cause-effect hypothesis using data collected from Twitter, a popular online social network. Using Twitter opinions, I test whether recreational cannabis legalization impacts the development of pro-cannabis attitudes for the population in favor of tobacco vaping, and analyze heterogeneity of causal effects for different states implementing the policy and time since legalization. Second, I discuss my work on Network Structural Causal Model (NSCM) and Network Abstract Ground Graph (NAGG), a framework for expressive causal modeling and sound causal reasoning in networks, along with IDE-Net, an approach for robust individual direct effect estimation when underlying mechanisms of heterogeneous peer influence (HPI) are unknown. Third, I present EgoNetGNN, a novel graph neural network (GNN) architecture, to capture unknown HPI mechanisms that involve not only peer treatments but also attributes of the local neighborhood, including node, edge, and structural attributes for robust heterogeneous peer effect estimation. This thesis integrates causal inference, graph machine learning, and network science to uncover heterogeneous causal effects in complex network settings, paving the way for future multidisciplinary research directions and applications
Associations of Persistent Organic Pollutants Exposure with Coronary Heart Disease and Hearing Loss
No full textPersistent Organic Pollutants (POPs) are endocrine-disrupting chemicals associated with adverse health outcomes. Emerging evidence suggests that POP exposure may contribute to chronic diseases such as coronary heart disease (CHD) and hearing loss, potentially through inflammatory pathways. Inflammation is a key biological mechanism in the development of CHD, and several traditional CHD risk factors are also associated with hearing loss.
We examined the cross-sectional associations of (1) POPs with CHD, conducting sex-stratified analyses a priori to account for potential sex differences in POP metabolism, elimination, and toxicology; (2) POPs with hearing loss and hearing thresholds; and (3) inflammatory biomarkers with hearing loss and hearing thresholds among Hispanic/Latino adults. Data were obtained from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL) ancillary study, "Persistent Organic Pollutants, Endogenous Hormones, and Diabetes in Latinos," which included 2,333 participants aged 45–74 years. Multivariable logistics and linear regression models were used to estimate associations, taking into account the survey design and relevant covariates.
In the overall cohort, β-Hexachlorocyclohexane(β-HCCH) was significantly associated with increased odds of CHD. Among men, both β-HCCH and Hexachlorobenzene (HCB) were positively associated with CHD, while in women, mirex showed a significant positive association. Polybrominated Diphenyl Ethers (PBDEs) were positively associated with hearing loss and elevated hearing thresholds, while phenobarbital-type inducing polychlorinated biphenyls (PHBB-PCBs) demonstrated a non-linear relationship with hearing loss. Additionally, higher levels of the inflammatory markers high-sensitivity C-reactive protein (hsCRP) and gamma-glutamyltransferase (GGT) were associated with greater odds of hearing loss and increased hearing thresholds.
These findings add to the growing evidence linking POP exposure to CHD and hearing loss, particularly among Hispanic/Latino adults. Although additional longitudinal studies are warranted, the results underscore the need for public health strategies that aim to reduce POP exposure, enhance early monitoring of inflammation, and implement targeted interventions to mitigate the long-term burden of CHD and hearing loss
From Street Life to Community Life
No full textThis study strives to extend our knowledge of the ways in which young Black males develop and perform their understandings of the street life that they observe in their neighborhood. It involves a phenomenological investigation into how a group of eight young Black males make sense and meaning of street life and street gangs in their neighborhoods. It involved conducting one, two, or three interviews with each of the eight participants. The participants ranged in age from eighteen to twenty-five and all of them had grown up in working poor neighborhoods with a considerable amount of street life. The interviews were semi-structured with the questions progressing from basic background information such as “What neighborhood are you from?”; “How would you describe your neighborhood to a complete stranger who has never seen it?”; “What are the biggest strengths of your neighborhood?”; “What are the biggest challenges that it faces?”; “What were your earliest memories of being aware of street life?”; “How did/do you choose to be involved or to avoid it?”; and “Why did you make those choices?” to follow up questions to extend and clarify their original responses. Then, in another interview, a final question: “What would you want all youth workers who work with young Black people to understand more accurately or correctly?”
All of the interviews were then coded for major themes and sub-themes after member checking my understandings with the participants. The findings and their implications were then written up
Finite Element Analysis to Investigate Lipid Absorption in UHMWPE Under Mechanical Loading
No full textUltra-High-Molecular-Weight Polyethylene (UHMWPE) has long been considered the gold standard for tibial inserts in total knee arthroplasty (TKA) due to its mechanical strength, low friction, and biocompatibility. Despite advancements in crosslinking and post-irradiation stabilization, oxidative degradation remains a limitation, contributing to long-term implant failure. Recent studies have shown that in vivo oxidative degradation still occurs, suggesting alternative degradation pathways. One involves the diffusion of lipid species from synovial fluid into the UHMWPE matrix. Lipids diffuse particularly into amorphous regions and, by inter-
acting with them, alter the chemical and mechanical structure over time. The combination of cyclic mechanical stress and thermal effects from daily movements may increase polymer chain mobility, accelerating oxidative reactions and loss of integrity. This thesis presents two computational model to simulate the time-dependent diffusion of lipid species into crosslinked UHMWPE under mechanical loading. Two approaches were implemented to couple stress and diffusion finite element. The first approach involves two sequential finite element analyses in Abaqus: one to simulate the mechanical stress field, and the other to model mass diffusion using a stress-dependent diffusivity. A custom MATLAB script is used to iteratively couple the two simulations by updating diffusion parameters based on the local stress state, thus reducing computational cost. This method allows more flexibility in post-processing but requires iterative data exchange between Abaqus
and MATLAB. In contrast, the second approach offers a fully integrated simulation within Abaqus. This approach takes advantage of the mathematical similarity between mass diffusion and heat transfer equations, since the two formulations differ only for the variables involved. Abaqus fully coupled
temperature–displacement simulation is used to model mass diffusion by treating temperature as the equivalent for concentration. The model is applied to a simplified 2D geometry representing one-quarter of a symmetric experimental setup. Material properties, including diffusion coefficient (D), solubility (S) and
stress-diffusion factor (kp) were derived from literature and experiments and calibrated for virgin and remelted UHMWPE. Simulations predict palmitic acid accumulation and mass flux, identifying regions vulnerable to early softening or failure under high stress. These findings highlight the role of mechanical
loading in lipid transport and provide a predictive framework for long-term degradation under clinically relevant conditions
Causal Inference Methods for Microbiome Data
No full textThis dissertation introduces two complementary methodological frameworks for advancing causal mediation analysis in microbiome research. Microbiome data are characterized by high dimensionality, sparsity, overdispersion, and strong interdependencies, making conventional mediation approaches ill-suited for reliable inference. Existing methods often fail to deliver valid interval estimates or to capture longitudinal dynamics, especially when indirect effects are modest or when correlated mediator pathways are present. Our proposed framework addresses these challenges, providing tools that are broadly applicable across genomics, epidemiology, and other fields that require multiple mediator causal analysis.
The first framework develops a fiducial inference approach for baseline microbiome mediation. By integrating mixed-effects zero-inflated generalized linear models (MEZIGLM) with natural effects models, this method accounts for overdispersion, zero inflation, and subject-level heterogeneity while constructing generalized confidence intervals for natural direct and indirect effects. A second-order fiducial construction combines bootstrap-based covariance estimation with Monte Carlo sampling from asymptotic distributions, yielding robust interval estimates that outperform delta-method and bootstrap alternatives in finite samples. The second framework extends mediation analysis to longitudinal microbiome-outcome processes, introducing joint models that accommodate time-varying exposures, sequential
mediators, and dynamic outcomes. Factor-analytic random effects reduce the burden of high-dimensional correlations, ensuring unbiased estimation of direct and indirect effects. Simulations demonstrate that the proposed approach achieves stable coverage, improved sensitivity, and scalability in longitudinal contexts. Applications to data from the Cups and Community Health (CaCHe) study in Siaya County,
Kenya, highlight the real-world utility of these methods. At baseline, the fiducial framework identified taxa such as Lactobacillus crispatus, Atopobium vaginae, and Sneathia sanguinegens as key mediators linking sexual behavior to bacterial vaginosis (BV). Longitudinally, the joint mediation models captured hormonal and microbial pathways shaping community state type (CST) transitions, providing biologically interpretable insights into BV risk.
In summary, this dissertation makes two important contributions: (i) fiducial inference
for valid interval estimation in baseline mediation with microbiome data, and (ii) longitudinal joint mediation models for dynamic microbial processes. Together, these frameworks fill critical methodological gaps, enhancing the rigor and interpretability of causal mediation in complex, high-dimensional biomedical data
Communication Equity: Perspectives from Spanish-Speaking Families
No full textCommunication is a fundamental human right and critical for full participation in society. Yet communication is not equitably realized, particularly for people with disabilities who use augmentative and alternative communication (AAC). AAC encompasses a wide range of tools and strategies, from gestures to speech-generating devices, that support people with complex communication needs. Despite these supports, the responsibility of communication often rests disproportionately on AAC users, while their communication partners are ignored. Without adequate instruction, partners may unintentionally engage in behaviors that hinder, rather than support, communication—perpetuating inequities in access and outcomes. Communication Partner Instruction (CPI) is an evidence-based strategy to address these inequities. CPI explicitly teaches communication partners behaviors that support communication with and for AAC users. Although CPI has demonstrated success across various settings, its research and implementation have largely excluded culturally and linguistically diverse communities. Spanish-speaking Hispanics/Latinos, the largest ethnic and linguistic minority in the U.S., experience layers systemic of inequities that further compound access to and engagement in training, like CPI. In these communities, family members often serve as primary communication partners for AAC users and are critical to fostering dual-language development. Yet, CPI efforts have rarely been adapted to reflect their specific language and culture. This qualitative study, grounded in a human rights framework, explores the experiences and perspectives of Spanish-speaking family communication partners through in-depth interviews. Findings reveal multilevel barriers and facilitators to CPI access and efficacy, including the role of support networks, cultural-linguistic bias/discrimination and actionable community-driven recommendations for more inclusive CPI design and delivery. This research contributes novel insights into equity-driven service provision. It offers practical implications for policy, research, and clinical practice aimed at improving access to and the effectiveness of CPI for dual-language families, in turn supporting the communication rights of people with disabilities who use AAC
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇