82837 research outputs found
Sort by
Intentions to Use Evidence-Based Practices: Preservice Teachers’ View
No full textOver the past two decades, evidence-based practices (EBPs) have been increasingly emphasized in inclusive education to support diverse student needs and improve outcomes. However, implementation literature suggest that educators use EBPs infrequently and often rely on their personal experience and preference instead. Teachers have indicated that limited knowledge, low efficacy, time constraints, and burnout pose as barriers to EBP use. To counter these barriers, educating preservice teachers about EBP and supporting their beliefs to adopt EBPs before they enter the workforce is an important initiative. This exploratory study investigated preservice teachers’ knowledge of and familiarity with instructional practices, as well as if motivational factors around EBP differ based whether or not preservice teachers anticipate relying on evidence and personal preference when making instructional decisions. In part I, I used a pre-post quasi-experimental survey design to collect data from preservice teachers (N = 25) to assess their familiarity with and intentions to use various instructional practices. The results suggest that while the majority of participants (at least 60%) were familiar with all practices, they overestimated the supporting evidence in eight out of the 10 instructional practices. Additionally, results indicate that exposure to accurate evidence levels led to changes in intention to use some, but not all, practices – and to varying degrees. In part II, I used a correlational survey and found between-group differences in motivational constructs depending on preservice teachers’ anticipated reliance on evidence, but not personal preference. The results are discussed in terms of implications for researchers, education stakeholders, and psychological clinicians for promoting EBP adoption among future educators
Designing a Digital Health Platform for Women with PCOS: An Empowerment Approach for Women in Their Reproductive Years
No full textPolycystic Ovary Syndrome (PCOS) is a common yet under-diagnosed hormonal condition that affects approximately 1 in 15 women worldwide, leading to a wide range of physical and emotional challenges. Despite its prevalence, many women with PCOS face difficulties in diagnosis and management due to a lack of awareness, both among healthcare providers and the general public. This research explores the critical gap in PCOS knowledge and the need for accessible, user-friendly resources to support women in managing their condition. The aim of this study is to design and develop an informative platform that empowers women with PCOS through comprehensive education, self-management tools, and community support.
The research investigates medical journals, research papers and existing online platforms related to PCOS and women’s health, identifies knowledge gaps, and explores the effectiveness of digital tools that could empower women to better understand and manage their symptoms. Through a combination of research methods, including surveys, expert interviews, and user studies, this project aims to create a platform that offers tailored guidelines on diet, exercise, and lifestyle modifications, along with features for tracking menstrual cycles, symptoms, and health metrics. Additionally, the platform will facilitate access to healthcare resources and foster a supportive community where women can share experiences and advice.
By addressing these needs, the platform has the potential to improve health outcomes and the overall quality of life for women living with PCOS, ensuring they have the knowledge, tools, and support necessary to navigate their condition with confidence
In-Sample Deep Reinforcement Learning
No full textReinforcement learning (RL) is a paradigm of learning through interaction with the environment. Despite significant breakthroughs when combined with deep neural networks, deep RL still suffers from sample inefficiency. This thesis studies deep RL with in-sample update, a methodology that focuses on bootstrapping only from well-supported state-action pairs within the replay memory. We study two fundamental challenges: (1) How can we maximize the utility of existing data for \textbf{learning?} (2) How can we design effective \textbf{exploration} strategies to gather high-quality data for subsequent iterations?
To tackle the learning challenge, we treat the replay memory as an \emph{empirical MDP} and solve it using an in-sample Bellman update. Building on this, we develop regularizers and integrate them into value-based methods. This approach is further generalized into the Empirical MDP Iteration (EMIT) framework. This framework provides an iterative learning pathway that uniquely solves each empirical MDP and incrementally approaches the original MDP through new data collection. For the exploration challenge, we propose learning a behavior policy to estimate the probability of actions taken at each state. By constructing a policy set that balances state-action coverage and overestimation correction, we enable flexible and interpretable exploration. Finally, we explore the role of in-sample data in ensuring the safe deployment of RL policies. By framing this as an anomaly detection problem, we introduce a distance-based anomaly detection framework specifically designed for deep RL.
Our results demonstrate how in-sample deep RL boosts learning, exploration, and safety, advancing RL toward more reliable and effective methods
Developing Stable and Resilient Anammox-based Reactors Treating Low-Strength Wastewater
No full textIn efforts to advance sustainable wastewater management, the anaerobic ammonium oxidation (anammox) process is recognized as a more efficient alternative to traditional nitrification-denitrification systems, due to its lower requirements for energy and organic carbon. However, the slow growth rate of anammox bacteria and their sensitivity to environmental changes present significant challenges in practical applications. This study innovatively proposes a method to accelerate the start-up of autotrophic anammox reactor by adjusting the nitrogen source ratio in the influent. Additionally, this research focuses on enhancing adaptability to complex environmental conditions. For instance, in response to reduced ambient temperatures or substrate scarcity in the influent, we employed an optimal ratio of NH4+/NO2⁻/NO3⁻ in partial denitrification-anammox (PD/A) approach to improve the system's stability and resilience against external changes, thereby ensuring consistent nitrogen removal performance. Importantly, the research also investigates the functional differentiation of anammox granules with varying particle sizes. Granules of different sizes were found to exhibit distinct microbial activity levels, community compositions, and functional gene expression patterns. Larger granules (> 2 mm) demonstrated superior anammox performance and metabolic diversity, suggesting that granule size is a critical factor influencing overall reactor efficiency and should be considered in reactor design and sludge management strategies. Additionally, this study emphasizes exploring the role of granular activated carbon (GAC) in enhancing the efficiency of upflow anaerobic sludge blanket (UASB) anammox reactors. GAC is known to provide a conducive environment for biofilm formation and microbial interactions, which are critical for the stability and activity of anammox processes. We investigated whether the effects of GAC in the upstream anammox reactor could be effectively transmitted to downstream anammox units and how GAC might influence the functionality of these reactors by regulating the microbial extracellular secretions. This research could potentially aid in optimizing the performance of anammox in real-world wastewater treatment scenarios.
The key objectives outlined in this project include: 1) exploring rapid enrichment strategies for anammox bacteria and developing reactor startup strategies; 2) evaluating the reactor's adaptive response to external environmental fluctuations, such as temperature reductions and substrate deficiencies, focusing on nitrogen removal performance, variations in extracellular polymeric substances (EPS), and microbial community dynamics; 3) elucidating the role of granule size in shaping the functionality and stability of anammox systems and 4) investigating the specific effects of GAC on the performance of anammox reactors. The expected results of this research aim to significantly enhance the scalability and applicability of anammox technology in wastewater treatment, making it a viable and eco-friendly solution for global implementation
Conceptual analysis of induced seismicity risks associated with Hydraulic Fracturing (HF) and Wastewater Disposal (WWD) in Cadomin Formation A Formation-Adaptive Risk Assessment Framework
No full textThis research investigates the risks of induced seismicity associated with hydraulic fracturing (HF) and wastewater disposal (WWD) in the Cadomin Formation, a subsurface unit shared between Alberta and British Columbia, Canada. Recent seismic events linked to injection activities highlight the limitations of current regulatory frameworks, which rely on generalized thresholds and reactive protocols rather than formation-specific risk assessments.
To address this gap, the study develops a conceptual and exploratory framework that integrates Bow-Tie Analysis and Layers of Protection Analysis (LOPA) to systematically identify threats, barriers, and consequences, with barrier performance classified within the Hierarchy of Controls. The framework is demonstrated through Cadomin Formation, while comparative insights from the Duvernay Formation illustrate how lithological and stress-regime differences shape barrier reliability and initiating frequencies. Screening-level LOPA tables were applied as semi-quantitative tools to show how geological and operational conditions influence the reliability of preventive and mitigative safeguards. These outputs are not definitive probability calculations but structured indicators of relative risk reduction.
In the absence of direct operational data, this research adopts a conceptual but rigorous approach grounded in geological and engineering principles. The bow-tie analyses directly informed the positioning of risks within both the risk matrix and LOPA screening, ensuring consistency across tools. The results emphasize that while reactive regulatory protocols such as the Traffic Light System provide short-term guidance, they do not fully address long-term liabilities or cumulative risks. Proactive monitoring, advanced modeling, and improved regulatory clarity are needed to strengthen resilience.
Ultimately, this study contributes a formation-specific, auditable framework that advances induced seismicity risk management from reactive thresholds toward predictive, adaptive strategies. It offers methodological insights rather than prescriptive standards, providing a foundation for future research, regulator–operator collaboration, and the development of standardized, formation-specific protocols across HF, WWD, and emerging subsurface injection technologies
Ruhina Rana - Abstract 60 - Innovate Conference 2025
No full textTo address the cultural erasure of nurses’ experiences of Covid-19, we offer an arts-based study that creates space for discourse. There is currently a ripple effect of anti-intellectualism/anti-science that has permeated our culture following the pandemic that requires attention and reflection. To this end, a poem play was created that builds on the experiences of front-line nurses. This project shines a light on the silencing of nurses and the impact this has on patients, families, and the public
From Fermenter to Farm: Downstream Processing of Single Cell Protein
No full textThe global demand for meat is rapidly increasing, causing extensive overfishing and deforestation to provide livestock with high value feed and protein supplements. Alternative protein sources have been sought to reduce the environmental impact of these activities and redirect crop and food production toward humans. One such product is single cell protein, which is comprised of unicellular microorganisms that are high in protein. These products are produced through fermentation and reduce water, land, and nutrient requirements compared to conventional protein sources such as soymeal. In this work, single cell protein produced through methanol fermentation of Methylophilus methylotrophus AS1 was processed by different separation and drying methods to determine their efficacy and effect on the quality of the final product.
Proximate analysis of Methylophilus methylotrophus AS1 biomass determined its composition of 79% amino acids, 16% lipids, and 4.7% ash. Microbial cells were visualized by scanning electron microscopy and the average cell length of 1200 ± 200 nm and width of 300 ± 30 nm were determined for the first time. It was found that increasing substrate feed rate increased the cell width but not length, and stressful growth conditions caused cell lysis and formation of a white debris layer. Additionally, prolonged storage caused deformation, lysis, and clumping of the cells.
Flocculation and cross-flow membrane filtration separation methods were investigated and compared to centrifugation as a reference method. Flocculation was induced by heating at 70 °C and acidification to pH 3, and cross-flow filtration utilized a 0.2 μm PVDF flat sheet microfiltration membrane. The highest biomass recovery was achieved through flocculation, which also concentrated the material to a greater extent than cross-flow filtration. However, cross-flow filtration was more selective in separating the cells from the spent media. Overall, both methods produced a final product comparable to centrifugation, and product quality was maintained.
Spray drying was utilized as a test method against freeze drying as a control. Spray drying at 200 °C inactivated originally viable cells and produced a fine, light, free-flowing, homogenous powder with moisture contents between 4.3-7.4%. Most importantly, the product composition and amino acid content of the spray dried material was retained.
Overall, it was determined that single cell protein from Methylophilus methylotrophus AS1 was resistant to protein degradation and compositional changes from exposure to heat, acidification, and shear stress induced by the separation and drying methods tested. The results of this work support further development of downstream processing methods and enable greater freedom in selection of separation methods for industrial applications. Due to the stability and robustness of the product through the separation and drying methods tested, process development activities may focus on other parameters such as efficiency, energy requirements, environmental impact, and safety. Additionally, the high amino acid content indicates Methylophilus methylotrophus AS1 is an excellent organism to produce single cell protein for animal feed
Seismic Performance Evaluation of Steel Buckling-Restrained Braces Subjected to Subduction Interface Earthquake Demands
No full textSteel Buckling-Restrained Braced Frames (BRBFs) are commonly used as lateral-force resisting systems in multi-storey buildings located in high seismic regions of North America. The reliable seismic performance of BRBFs is attributed to their Buckling-Restrained Braces (BRBs), which provide high ductility and energy dissipation capabilities. BRBs were first developed in Japan in 1988, later adopted into the AISC Seismic Provisions (AISC 341) in 2005, and subsequently incorporated into the Canadian steel design standard (CSA S16) in 2009. Canadian standards have since referenced the BRB prequalification provisions from AISC 341, which require two tests: (1) a uniaxial test to evaluate the performance of the device, and (2) a subassemblage test to assess stability under additional rotational demands introduced by the connections. Both tests must be conducted under a standardized symmetrically increasing loading protocol, which was primarily developed from analytical studies of near-field and far-field seismicity in Los Angeles, California. The loading protocol development may not account for the demands due to more severe, long-duration subduction interface earthquakes of the Cascadia Subduction Zone. The adequacy of the prequalification loading protocol to capture the demands imposed by subduction interface earthquakes is therefore uncertain. In addition, studies specifically examining BRB demands under such events remain scarce. This gap in understanding emphasizes the importance of further investigation on the seismic response of BRBs under subduction interface earthquake demands. This M.Sc. research project aims to examine the seismic demands imposed on BRBs in low- to mid-rise steel buildings equipped with BRBFs subjected to seismicity expected in the west coast of Canada, develop a subduction interface-consistent loading protocol, and experimentally evaluate the seismic performance of steel BRBs subjected to subduction interface earthquake demands. Five different prototype BRBF buildings are first designed under Canadian standards. A fibre-based numerical model is developed for one of the BRBFs in each building and subjected to Nonlinear Response History Analyses (NLRHA) using three suites of 11 historical records each: shallow crustal, in-slab, and subduction interface events. The results show that the largest demands imposed on BRBs pertaining to peak strain demands, force overstrength, cumulative inelastic ductility, and fatigue life, are consistently largest under subduction interface events. The results of one of the prototype buildings are then statistically analyzed to develop a symmetrically increasing, subduction-consistent loading protocol. Following the analytical evaluation, an experimental program consisting of five full-scale BRB specimens is developed. Two of the BRB specimens have been previously tested and strain-aged for two years and eight months, while the remaining three are new. Both strain-aged specimens and one new specimen are subjected to direct brace deformation histories representing two historical subduction earthquake events. The other two new specimens are tested under symmetric loading protocols: the AISC 341 loading protocol and the subduction loading protocol developed in this study. The results of the experimental tests show that BRBs subjected to symmetric loading protocols exhibit post-yield envelope behaviour that is in good agreement with each other, irrespective of protocol. BRBs subjected to direct brace deformations confirm significant strain capacity and fatigue life available in BRBs. Furthermore, BRB force overstrength can generally be reproduced by post-yield slopes of symmetric loading protocols studied here. The strain-aged BRB shows a noticeable increase in peak forces, however, additional test data is required to verify this effect, particularly since the large tensile strain generated by the monotonic push in the strain-aged BRB was not imposed in the replica new BRB