University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
A qualitative study of college success for students with experience in foster care in Pennsylvania
Get PDFStudents who have experiences in the foster care system, group homes, or kinship care graduate college at a significantly lower rate than their peers. Most research indicates that no more than four percent of students with experience in the foster care system, who will be represented in this dissertation as SEFC, will complete any level of college degree (Jones & Dean, 2020).
The purpose of this study was to investigate the experiences of students with foster care backgrounds within a community college and four-year institution setting, to gauge the students’ educational confidence and their interest in using resources on campus. The study focused on two schools in Pennsylvania shortly after the adaptation of the Pennsylvania Fostering Independence Tuition Waiver that was passed in 2019 statewide, which provides tuition waivers for students with experience in foster care.
This study provides a unique focus in two ways: first, it examines SEFC and their college experience from a first semester perspective, to understand what students’ expectations and needs are as soon as they step onto their campuses. Second, while much of the research on SEFC is focused on data such as GPAs, retention, and graduation rates, this study’s interest was to hear the students’ voices, to learn how they speak about their schooling, and to hear any common language that was used by the participants.
The study finally aims to provide recommendations for institutions of higher education to help them to better support the academic achievement of SEFC as first-semester college students.
Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Amy Sosanko, accepted the attached license on 2025-11-26 at 14:06.The student, Amy Sosanko, submitted this Dissertation for approval on 2025-11-26 at 14:15.This Dissertation was approved for publication on 2025-11-26 at 15:37.DSpace SAF Submission Ingestion Package generated from Vireo submission #22968 on 2026-02-19 at 18:25:3
Developing and evaluating domain models for programming skills using learning curve analysis
Get PDFIdentifying key concepts in programming is important for accurately tracking skill development and designing better support mechanisms for students. Prior research has identified a plethora of skills at varying levels of granularity, from broad abilities such as code comprehension and tracing to fine-grained skills like using individual syntactic elements correctly. However, more evidence is required to understand the extent these skill models reflect the actual skill acquisition process of students.
Knowledge components, which are acquired units of skills and abilities inferred from the performance on a set of tasks, can be an appropriate starting point in a framework for evaluating these skill models and generating new models in a data-driven manner. These knowledge components can be evaluated by learning curve analysis, which is an educational data mining technique for modeling skill development using data on problem-solving performance of students. Yet, previous applications of learning curve analysis for programming could not identify a robust and interpretable skill model, which may imply that programming skills are more complex than initially assumed.
In this thesis, we summarize two studies where we evaluate two domain models that explain student skill development. The first study is a replication of prior work that proposed the use of syntactic structures in a programming language as individual skills in programming. The second study proposes a novel domain model that uses programming plans from computing education literature to model student skills. We evaluate the extent to which these domain models can explain students' development across homework assignments in an introductory programming course using data collected from seven semesters.
Our findings imply that learning curve analysis can be used to produce useful insights from solutions to open-ended code-writing exercises.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Mehmet Arif Demirtas, accepted the attached license on 2025-12-01 at 10:55.The student, Mehmet Arif Demirtas, submitted this Thesis for approval on 2025-12-01 at 11:02.This Thesis was approved for publication on 2025-12-01 at 14:33.DSpace SAF Submission Ingestion Package generated from Vireo submission #22992 on 2026-02-19 at 18:25:5
Designing and evaluating a fully interpretable neural network for learner behavior detection
Get PDFThe increasing complexity of machine learning models in education has created a "challenge of interpretability," where opaque decision-making processes risk undermining fairness, accountability, and trust, among other factors. This dissertation confronts this challenge by proposing and validating an alternative paradigm: developing neural networks that are interpretable by design. Through a series of three interconnected studies, this work demonstrates an end-to-end methodology for creating, evaluating, and validating a fully interpretable model for learner behavior detection.
The first study details the design of a novel, constraints-based convolutional neural network for identifying gaming-the-system behavior. By engineering the model's architecture and training process, its convolutional filters are made to function as explicit, human-readable behavioral patterns, ensuring that the evidence for its predictions has full explanatory potential and is directly tied to its inference process.
The second study presents a human-grounded evaluation to rigorously assess the model's explainability. The results demonstrate that education researchers, regardless of their machine learning expertise, could use the model's explanations to accurately predict its outputs and identify how to alter them. This provides strong evidence that the explanations are both faithful to the model's internal logic and intelligible to human users.
The third study validates the knowledge captured by the model through an interview with a subject-matter expert. The expert confirmed that the majority of the patterns learned by the model were valid indicators of gaming-the-system behavior, despite not being included in a cognitive model previously created by an expert. This highlights the potential of interpretable models to serve not only as predictive tools but also as instruments for knowledge discovery.
Taken together, these studies offer a proof-of-concept for a methodology that moves beyond the prevailing "black-box" paradigm. By shifting the focus from post-hoc explanations to interpretable-by-design architectures, this dissertation provides a framework for building more transparent, trustworthy, and insightful AI in education.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Juan Pinto, accepted the attached license on 2025-12-02 at 11:05.The student, Juan Pinto, submitted this Dissertation for approval on 2025-12-02 at 11:23.This Dissertation was approved for publication on 2025-12-04 at 17:16.DSpace SAF Submission Ingestion Package generated from Vireo submission #23017 on 2026-02-19 at 18:25:5
Health spectacle in American popular literature, 1790-1870
Get PDFThis project takes up health spectacle, which I identify as a prevailing concept for late eighteenth and nineteenth century American health history, first beginning in the 1790s and concluding in the 1860s and 70s. Through a study of popular literature of the period and other circulating medical thought in journals and newspapers, I investigate how instances of health spectacle shaped the development of professionalized health and medicine. Ultimately, I argue that reading for health spectacle in popular literature reveals a prevailing tension between medical expertise and medical democratization of knowledge in the United States.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Victoria Stewart, accepted the attached license on 2025-12-03 at 10:08.The student, Victoria Stewart, submitted this Dissertation for approval on 2025-12-03 at 10:14.This Dissertation was approved for publication on 2025-12-03 at 11:25.DSpace SAF Submission Ingestion Package generated from Vireo submission #23039 on 2026-02-19 at 18:26:3
Optimal control techniques for multimode propulsion mission design in cislunar space
No full textMultimode spacecraft propulsion is an emerging, enabling, and enhancing technology that combines two or more propulsive modes (e.g., chemical and electric) into one system using a single propellant. The primary benefit of this technology is that it can provide reduced propellant consumption for a given transfer in comparison to an all-chemical approach and reduced transit time in comparison to an all-electric approach. It can also significantly increase the maneuvering capability domain of a spacecraft in comparison to single mode (i.e., all-chemical or all-electric) propulsion systems. Additionally, multimode propulsion provides mission flexibility and adaptability because the propellant may, in principle, be used in any mode at any time. Despite these benefits, there is presently a lack of mature techniques for designing optimal multimode transfers—particularly for chemical-electric multimode systems.
In this dissertation, indirect optimal control techniques are developed for solving optimal multimode orbital transfers. Using an indirect approach results in the automatic selection of the burn sequence (i.e., which mode should be used at every instant). This yields more optimal transfers in comparison to the traditional approach in which the burn sequence is manually selected.
Two types of optimal control problems are solved: minimum-fuel and propellant-constrained minimum-time (PCMT). The former is the primary focus of this dissertation and examples are shown for interplanetary, geocentric, and cislunar transfers for systems with large performance differences (i.e., thrust and specific impulse) between the modes. Forward single shooting is used to solve the example interplanetary and geocentric transfers in combination with hyperbolic tangent smoothing (HTS) for the throttle. To solve multimode cislunar minimum-fuel problems, a new technique is developed using modified equinoctial elements (MEE), HTS, forward-backward multiple shooting, two reference frames (Earth-centered and Moon-centered), and highly consistent dynamics. The method is demonstrated with multiple two-mode transfers as well as a three-mode transfer.
The PCMT problem is particularly relevant for mission recovery efforts (e.g., after a launch injection failure) or spacecraft with hybrid propulsion systems (i.e., systems with two or more modes that are not integrated and do not share propellant) with fixed propellant budgets. To solve this problem, a penalty function approach was combined with HTS to enforce a maximum propellant usage constraint on the highest thrust, lowest specific impulse mode. The method is demonstrated with a three-dimensional transfer from a geostationary transfer orbit (GTO) to geostationary orbit (GEO) that was solved using MEE and forward single shooting.
Finally, note that the techniques developed here are agnostic to the propulsion system. As a result, the methods apply equally well for spacecraft with hybrid propulsion systems as to those with multimode propulsion. While multimode propulsion is still a maturing technology, there are many spacecraft currently flying with hybrid propulsion. Accordingly, the methods developed here are applicable not only to future multimode missions, but also to near-term missions with hybrid propulsion.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Bryan Cline, accepted the attached license on 2025-12-03 at 20:02.The student, Bryan Cline, submitted this Dissertation for approval on 2025-12-03 at 21:36.This Dissertation was approved for publication on 2025-12-04 at 11:23.DSpace SAF Submission Ingestion Package generated from Vireo submission #23050 on 2026-02-19 at 18:28:1
Oxidation driven degradation of silicon nitride based hot surface ignition assistant devices
No full textHot surface ignition assistant devices show great promise in enabling fuel flexibility in CI engines that otherwise require certain levels of reactivity in their fuels for safe and efficient operation. Previous research has shown that commercially available silicon nitride-based glow plugs can serve as effective ignition assistant devices in an engine environment as long as they remain continuously heated at elevated power levels that allow their heating tips to remain at ≥ 1350K. However, operating commercial glow plugs in this manner causes them to fail rapidly for a myriad of reasons, the most important among them being oxidation from species in the combustion environment. Literature shows that at temperatures necessary for ignition assistance, oxygen as well as water vapor are the two most active species in causing silicon nitride to oxidize and volatilize. In order to observe this behavior, a pressurized chamber and a flow facility were set up to expose Bosch and Beru glow plugs to an oxidizing environment containing both water vapor and oxygen. Testing shows that Bosch glow plugs are more resilient to oxidation-driven degradation than Beru glow plugs. While a Bosch glow plug’s conductive pathway is embedded within its ceramic tip, the Beru glow plug has its conductive pathway near the surface of its ceramic tip, where it is exposed to attacks from oxidative species. A mullite-based environmental barrier coating was applied on some of the Bosch and Beru samples. While the mullite coating was able to protect a Bosch glow plug sample from oxidation in a pressurized wet oxygen environment, repeated testing is required to fully determine its effectiveness.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2026-02-19 without embargo termsThe student, Seongyong Hong, accepted the attached license on 2025-12-08 at 09:06.The student, Seongyong Hong, submitted this Thesis for approval on 2025-12-08 at 09:09.This Thesis was approved for publication on 2025-12-08 at 09:41.DSpace SAF Submission Ingestion Package generated from Vireo submission #23096 on 2026-02-19 at 18:29:4
Static and dynamic models for rebalancing and charging battery-powered mobility
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Jesus Osorio Fuenmayor, accepted the attached license on 2025-04-28 at 15:41.The student, Jesus Osorio Fuenmayor, submitted this Dissertation for approval on 2025-04-28 at 15:57.This Dissertation was approved for publication on 2025-04-29 at 15:59.DSpace SAF Submission Ingestion Package generated from Vireo submission #22045 on 2025-10-19 at 19:16:27The number of trips by micromobility modes (such as dockless bikes and e-scooters) in North America, after their initial adoption in 2017, rapidly grew to 147 million by 2019. This trend of growth continued in recent years despite a significant drop in ridership due to the COVID-19 pandemic in 2020, and micromobility usage has since rebounded and continued to grow beyond pre-pandemic levels. The rapid growth of e-scooter sharing services, however, has come with significant operational challenges. Operators must manage tens of thousands of e-scooters in large cities such as Los Angeles, San Francisco, Washington DC, Austin, and Chicago, and the spatiotemporal demand heterogeneity requires that those e-scooters be frequently recharged and repositioned/rebalanced for future users. Proper strategies for e-scooter planning and operations, especially those related to charging and rebalancing, have become a critical challenge for cities and micromobility companies. Lack of such strategies has mandated many cities across the U.S. to impose licensing, parking, and pathway regulations to reduce the clutter caused by excess or uneven supply of e-scooters. Numerous charging technologies have emerged to enhance the rebalancing and charging operations of shared micromobility systems, including charging hubs, charging stations, and battery swapping. Recent innovations, such as portable batteries and fast-charging capabilities, also enable e-scooters to be charged quickly while being transported during rebalancing trips. These strategies provide new opportunities to improve operational efficiency and, when effectively implemented, may achieve stronger economies of scale. However, integrating these technologies into daily operations is nontrivial. Operators must create a daily charging-and-routing plan for all e-scooters that simultaneously accounts for their state of charge (SoC), required charging durations, energy supply resources, and a pickup/drop-off strategy, while balancing supply and demand across time and space. Furthermore, if rebalancing and charging occur during service operations, the strategy must accommodate temporal demand surges in real time, which increases the complexity of large-scale planning and operations. As such, this dissertation research focuses on developing new mathematical models and customized solution approaches that can effectively support decision-making for static and dynamic e-scooter rebalancing and charging operations in a variety of business scenarios. An on-board charging strategy for overnight operations, modeled as a discrete inventory routing problem, is first presented. E-scooters of different SoC are modeled as different commodities that can transition into each other while being transported by the rebalancing vehicles. To mitigate the complexity of such models, a discrete-continuous hybrid solution approach is developed for large-scale instances. The key idea is to embed a continuous approximation (CA) subroutine (which estimates costs for routing decisions in local areas) into a discrete mixed-integer inventory-routing problem formulation. A series of numerical experiments, including hypothetical problem instances and a full-scale case study for Washington DC, are conducted to demonstrate the applicability and effectiveness of the proposed hybrid approach. Then, we propose a capacitated charging hub strategy for overnight e-scooter rebalancing and charging. The problem is similarly modeled as a static inventory routing problem with SoC transitions at charging hubs (i.e., as satellite facilities). We propose a different discrete-continuous hybrid model reformulation and solution algorithm, in which the spatial region is partitioned into local zones with an embedded CA subroutine. This method is validated through hypothetical experiments and a case study to draw managerial insights. Finally, a family of dynamic decision support models are developed to evaluate and compare a range of charging strategies (e.g., on-board charging, battery swapping, and charging stations) during daytime operations. Under a unified Markov Decision Process (MDP) framework, we test different battery evolution rules and operation actions in a time-varying environment. A series of stochastic dynamic lookahead approximations are introduced to solve these MPD models. The models are initially tested on hypothetical networks and subsequently validated using real-world cases. The emphasis is on comparing their performance under various conditions to derive practical insights and estimate the benefits of each strategy
Protein digestibility and quality of selected traditional and alternative protein sources for canine and feline diets
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Julio Cesar Mioto, accepted the attached license on 2025-05-05 at 11:21.The student, Julio Cesar Mioto, submitted this Thesis for approval on 2025-05-05 at 11:31.This Thesis was approved for publication on 2025-05-07 at 15:57.DSpace SAF Submission Ingestion Package generated from Vireo submission #22205 on 2025-10-19 at 19:17:01The digestibility and quality of dietary proteins are crucial parameters in formulating nutritionally balanced canine and feline diets. Two experiments were conducted to evaluate protein digestibility and quality of alternative (black soldier fly larvae meal; BSFLM) and traditional (chicken-based) protein sources for canine and feline diets using the precision-fed cecectomized rooster assay. In experiment 1, twenty cecectomized roosters (5 roosters/protein source) were assigned to one of four protein ingredients: defatted BSFLM wheat-substrate (BSFLM-W), defatted BSFLM corn-substrate (BSFLM-C), chicken meal (CM), and whole egg powder (WEP). After a 26-hour fasting period, roosters were tube-fed test ingredients. Following crop intubation, excreta samples were collected for 48 hours. Endogenous loss corrections for amino acids (AA) were made using five additional cecectomized roosters to calculate standardized AA digestibility values. Generally, WEP had the highest (P<0.05) standardized indispensable AA digestibility, but BSFLM-C showed higher digestibility for certain AA, such as arginine, surpassing CM and BSFLM-W. Digestible indispensable AA score (DIAAS)-like values were calculated to determine protein quality according to the Association of American Feed Control Officials (AAFCO) nutrient profiles, European Pet Food Industry Nutritional Guidelines (FEDIAF) nutritional guidelines, and National Research Council (NRC) recommended allowances for adult dogs, adult cats, growing puppies, and growing kittens. The DIAAS-like values revealed methionine + cysteine and phenylalanine + tyrosine as limiting AA in BSFLM and CM, particularly in formulations for growing puppies, kittens, and adult pets, according to AAFCO, FEDIAF, and NRC nutrient profiles. Despite these limitations, both BSFLM sources showed high AA digestibility, suggesting their potential as viable alternative protein sources in pet diets. In experiment 2, standardized AA digestibility and DIAAS-like values were determined using the same procedures and reference nutrient profiles (AAFCO, FEDIAF, and NRC) as described in experiment 1. However, in this experiment, sixteen cecectomized roosters (n= 4 per treatment) were tube-fed test substrates along with corn: spray-dried chicken protein hydrolysate (CPH), spray-dried chicken meat (CC), spray-dried chicken broth (CB), and CM. Overall, CC showed the highest (P<0.05) standardized digestibility values for most AA, closely followed by CPH. CM showed intermediate digestibility, whereas CB presented the lowest digestibility among tested ingredients. Methionine, phenylalanine, valine, isoleucine, leucine, and lysine digestibilities were significantly higher in CC and CPH compared with CB and CM. DIAAS-like values, calculated according to AAFCO, FEDIAF, and NRC nutrient profiles, highlighted tryptophan or phenylalanine + tyrosine as generally the first limiting AA among chicken-based sources, with CC and CPH demonstrating superior protein quality relative to CB and CM. Overall, the results of both experiments emphasize the importance of AA digestibility and composition when evaluating protein ingredients for canine and feline diets. BSFLM appears promising as an alternative protein source, comparable to chicken meal. The traditional chicken-based sources showed variability, with CC and CPH demonstrating superior protein quality compared with CM and CB, highlighting the importance of ingredient processing on protein quality
Quantifying and visualizing deliberative dialogue quality in online social spaces
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Mukhil Umashankar, accepted the attached license on 2025-05-05 at 11:48.The student, Mukhil Umashankar, submitted this Thesis for approval on 2025-05-05 at 11:58.This Thesis was approved for publication on 2025-05-06 at 16:13.DSpace SAF Submission Ingestion Package generated from Vireo submission #22206 on 2025-10-19 at 19:17:02Social media spaces empower individuals to ideate, iterate, and refine their thinking through exposure to diverse perspectives and real-time feedback. However, in reality, these benefits are accompanied by toxicity, spam, and polarization. These negative features are especially prevalent in deliberative spaces. The effective functioning of these platforms is crucial for democracies as a substantial volume of public discourse on real-world issues takes place within them. This study offers a quantitative assessment of deliberative dialogue in news spaces, drawing its evaluative metrics from deliberative theory. A mix of linguistic and network-based methods is adopted to compute the indicators. The results of the assessment are presented through user-centric visualizations. The goal is to model deliberative quality and make it accessible to the general audience, recognizing their role in upholding high-standard deliberation
Characterization of infrared imaging systems
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Sandhya Ashok, accepted the attached license on 2025-05-06 at 11:13.The student, Sandhya Ashok, submitted this Thesis for approval on 2025-05-06 at 11:15.This Thesis was approved for publication on 2025-05-06 at 19:27.DSpace SAF Submission Ingestion Package generated from Vireo submission #22227 on 2025-10-19 at 19:17:05This thesis explores the characterization of infrared (IR) imaging systems by microfabricating and analyzing USAF targets with varying thicknesses (0.6 μm to 5.3 μm) using SU-8 photoresist. The targets were fabricated in a cleanroom environment to ensure controlled conditions and high precision. Designed with well-defined USAF patterns, they served as reliable test samples to assess spectral accuracy and resolution capability due to their tailored and precisely measured parameters. Comparative thickness measurements using surface profilometry (DekTak3ST) and laser confocal microscopy (Keyence VK-X1000) revealed that surface profilometry was more accurate for thin films (< 1 μm) due to challenges in resolving transparent film surfaces via confocal microscopy. High-resolution profiles of USAF targets demonstrated a superior edge quality in laser confocal profiling with 0.08 μm pitch size but were prone to negative artifacts due to the same resolution issue. The Varian 620-IR FTIR imaging system was used to acquire spectra, and 1608 cm⁻¹ peak (aromatic C=C) was analyzed, to evaluate spectral fidelity. While the results followed Beer-Lambert law by showing increasing absorbances for thicker films, thicker films presented baseline linearity challenges. This dataset of well characterized targets provides a strong foundation for validating future IR instrument accuracies as well as IR simulation models