FGPS - Electronic Theses and Practica

Permanent URI for this collectionhttp://hdl.handle.net/1993/6

Overview

This collection contains University of Manitoba electronic theses and practica.

Before You Submit

The Faculty of Graduate Studies requires that all graduate students submit a copy of their thesis or practicum to this collection. Consult FGS Submitting your thesis or practicum to MSpace and Thesis/Practicum Deposit Step-by-Step for instructions and/or more information. Go to My MSpace to begin the submission process and, when prompted, choose the FGS - Electronic Theses and Practica collection. Unfinished or rejected submissions can be restarted by accessing My MSpace.

Browse

Now showing 1 - 20 of 27436

Access status: Embargo ,
Transsynaptic organization underlying a cerebellar gateway
(2026-08-11) Siddiqui, Mohd Sarim; Wang, Jun-Feng (Pharmacology and Therapeutics); Eftekharpour, Eftekhar (Physiology and Pathophysiology); Siddiqui, Tabrez
Transsynaptic adhesion molecules coordinate the precise alignment of presynaptic release machinery with postsynaptic receptor nanodomains, serving as fundamental architects of neural circuits. Among these, leucine-rich repeat transmembrane neuronal protein 3 (LRRTM3) is an understudied postsynaptic organizer and a genetic risk factor for autism spectrum disorder. Despite its known role in excitatory synapse development, how LRRTM3 governs nanoscale synaptic organization within the cerebellar glomerulus, a key gateway for sensory integration, remains poorly understood. To address this gap, I used two novel mouse models generated using CRISPR Cas9 genome editing, an endogenous 2xHA-tagged-LRRTM3 knock-in line for protein visualization and a global knockout line Lrrtm3⁻/⁻ for functional analysis. Immunohistochemical mapping showed that LRRTM3 is broadly expressed across the cerebral cortex, striatum, hippocampus, and cerebellum, colocalizing extensively with excitatory postsynaptic markers. LRRTM3 expression follows a tightly regulated spatiotemporal pattern. In the hippocampus, it emerges by postnatal day 12 and refines to the outer molecular layer of the dentate gyrus by day 21. In the cerebellum, LRRTM3 initiates at day 7 within the granule cell layer. Using super-resolution airyscan imaging, I found that LRRTM3 progressively tightens its association with the postsynaptic density in cerebellar glomeruli, becoming significantly enriched by postnatal day 30, marking it as a key player in circuit stabilization. In Lrrtm3⁻/⁻ mice, I confirmed no compensatory upregulation of related family members, establishing its non-redundant functional role. Expansion microscopy revealed a specific and bidirectional disruption of transsynaptic nanocolumn alignment, with increased distance between AMPA receptors and presynaptic release sites paired with compression of the Homer1 to active zone axis, while NMDA receptor associated organization and inhibitory axis remained intact. This indicates pathway-specific regulation of receptor-release site geometry. Electron microscopy further revealed coordinated reductions in excitatory and inhibitory synapse numbers, with the excitatory-inhibitory balance preserved. Surviving excitatory synapses showed shortened postsynaptic densities, reduced active zone lengths, and fewer docked vesicles, without changes to cleft width or PSD thickness. Together, these findings identify LRRTM3 as an essential, pathway-specific nanocolumn organizer, offering a mechanistic link between transsynaptic adhesion disruption and the sensory and cognitive phenotypes observed in neurodevelopmental disorders.
Access status: Embargo ,
The agri-food supply chain and short-line railways: an examination of benefits and challenges through a social capital perspective
(2026-07-15) DE ALWIS, DANDENIYAGE THARAKABHANU; Appadoo, Srimantoorao (Supply Chain Management); Gajpal, Yuvraj (Supply Chain Management); Ramachandran, Divya (Marketing); Malalgoda, Narendra
This study examines the social mechanisms underlying the formation and operation of community-led short-line railways within the agrifood supply chain in Canada following the institutional withdrawal of national Class I railways, while simultaneously providing a mathematically rigorous framework for macro-economic investment prioritization. While the technical and economic impacts of short-line railways are documented, the role of rural community collective action in the resilience of rural transport infrastructure remains understudied. Utilizing a sequential mixed-methods design, the first phase of this research conducts a deductive content analysis of Canadian short-line operators, providing evidence on how aspects of social capital facilitate collective action amidst institutional withdrawal. The findings reveal the presence of trust, associative density, and civic participation in narratives on the formation of Canadian short-line railways. Furthermore, this study identifies perceived environmental, social, and economic benefits and challenges of Canadian short-line railways, which are subsequently discussed through the forms of social capital. Although strong bonding capital can be attributed to the beneficial outcomes, the sector faces critical challenges rooted in a deficit of bridging and linking social capital, leading to technical resource gaps and regulatory navigation challenges. Crucially, the study also argues that this reliance on local communities represents an unfair burden where the responsibility of maintaining the supply chain is offloaded onto rural communities. To address these institutional constraints, the second phase of this research develops a quantitative Multi-Criteria Decision-Making (MCDM) model to evaluate infrastructure investment readiness across the four Western Canadian provinces. By integrating the Entropy and Variance Weight methods, the model establishes objective criteria weights derived directly from regional data behavior across five indicators: Agrifood Sector GDP, Road Network Density, Existing Short-Line Rail Density, Agrifood Export Values, and Agricultural Employment. These weights are funneled into the Weighted Aggregated Sum Product Assessment (WASPAS) framework, where a robust sensitivity analysis confirms mathematical invariance under parameter variations, yielding a stable prioritization hierarchy with Saskatchewan ranking first, followed by Alberta, Manitoba, and British Columbia. This research advances social capital theory by applying the framework to the novel context of rural transport infrastructure resilience, providing evidence of the underlying social and relational mechanisms. Ultimately, the study demonstrates that rural transportation infrastructure is a critical part of the rural community, couples’ grassroots resilience with a stable optimization model, and provides empirical recommendations on cultivating social capital and targeted capital allocation for rural transport infrastructure resilience.
Access status: Open Access ,
Essays on CEO vocal emotions, financial disclosure, and capital market responses
(2026-07-30) Jin, Linlin; Wang, Jun (Accounting and Finance); Wang, Xikui (Warren Centre for Actuarial Studies and Research); Wood, Geoffrey (Western University); Wu, Zhenyu
This dissertation examines whether CEOs’ vocal emotions during earnings conference calls convey value-relevant information beyond traditional financial disclosures and textual content. Using Layered Voice Analysis (LVA) on earnings call recordings of S&P 500 firms, I investigate how non-verbal vocal cues reveal managerial beliefs, shape capital market participants’ interpretations, and predict future firm outcomes. The dissertation consists of three essays. The first essay examines CEO excitement and distinguishes between authentic and inauthentic emotional expression. The findings show that performance-consistent excitement is associated with favorable stock market reactions and lower perceived risk, whereas excessive or misaligned excitement is interpreted as less credible and is associated with weaker subsequent performance and higher risk perception. The second essay focuses on CEO vocal stress as an involuntary signal of private concern. I find that high vocal stress predicts more negative analyst forecast revisions and greater analyst forecast dispersion, suggesting that analysts partially recognize stress as an adverse signal. However, vocal stress also predicts negative future earnings surprises and post-earnings announcement drift, indicating that analysts and investors do not fully incorporate the information embedded in CEO voice. The third essay develops a multidimensional framework of CEO emotional incongruence. Using principal component analysis, I identify two dimensions of vocal negativity: Cognitive Burden, reflected in stress, hesitation, and sadness, and Defensive Arousal, reflected in uneasiness and aggression. These vocal signals predict future earnings management and financial reporting quality. A two-stage decomposition further shows that residual vocal signals, after purging current earnings management, retain predictive power for future reporting outcomes. Together, the three essays show that CEO voice is an economically meaningful information channel. This dissertation contributes to the literature on corporate disclosure, behavioral finance, analyst underreaction, post-earnings announcement drift, and financial reporting quality by demonstrating that non-verbal executive communication reveals hidden information about firm risk, future performance, and reporting credibility.
Access status: Open Access ,
Smartwatch intervention for stress management: a pilot study amongst undergraduate students
(2026-08-10) Sitka, Madissen; Roos, Leslie (Psychology); Kim, Sunmee (Psychology); Giuliano, Ryan
Experiences of stress result in the activation of a variety of psychophysiological systems designed to address the increased demands of the environment that exceed routine demands. While such activation is initially productive in supporting one's ability to manage and address stressors, prolonged activation as a result of exposure to chronic and unpredictable stressors can lead to maladaptive physiological and psychological outcomes. Here, we investigated how a smartwatch psychoeducational intervention grounded in key principles of biofeedback might impact sleep duration and cardiovascular reactivity and recovery from a laboratory acute stressor. Thirty-three undergraduate students were randomized to either a self-guided smartwatch intervention or control group, and completed a longitudinal intervention, with pre- and post-intervention laboratory assessments of cardiovascular responsivity to acute stress. In total, 14 participants had useable data and were included in analysis. Permutation-based mixed-model analyses revealed no significant time-by-condition change in HRV responsivity, phasic HRV, or HR (p’s > 0.05) during the second administration of the acute stressor task. Furthermore, no time by condition specific improvement in sleep duration or negative affect was evident (p’s > 0.05). Feasibility analysis indicated that 54% of participants evidenced improvement in their sleep hygiene knowledge with 27% indicating a perceived improvement in sleep quality. Although results indicate promising effects of the intervention on participants perception and knowledge of sleep, findings suggest more research is needed to explore how to effectively translate key concepts of biofeedback interventions into a light-touch smartwatch intervention.
Access status: Open Access ,
The role of visual and proprioceptive feedback for motor imagery tasks
(2026-06-17) Wood, Stephen; Greening, Steven (Psychology); Glazebrook, Cheryl (Kinesiology and Recreation Management); Marotta, Jonathan
Motor imagery refers to the internal simulation or rehearsal of movements without physical execution. These internal action simulations are thought to evoke the same cognitive and neuronal processes involved in motor performance. One significant difference between motor imagery and motor execution is the availability of online sensory feedback for performance. The present study evaluated whether online sensory feedback systems are involved in motor imagery. Fifty participants completed the Hand Laterality Judgement Task (HLJT) in four conditions that altered the availability of visual and proprioceptive feedback from participants’ hands during task performance (control, no vision, disrupted proprioception, no vision + disrupted proprioception). The results indicate: 1) Online sensory feedback is used to simulate actions effectively; 2) implicit motor imagery continues to be used to identify the laterality of hands when sensory feedback is disrupted; 3) both visual and proprioceptive feedback contribute to motor imagery performance in the HLJT, which indicates that participants use both internal and kinesthetic imagery to make hand laterality judgements.
Access status: Open Access ,
Reimagining justice: exploring the possibilities and pitfalls of alternative approaches for addressing sexual and gender-based violence (SGBV) in Bangladesh
(2026-08-08) Meghla, Mansura Akter; Woolford, Andrew (Sociology & Criminology); Anderson, Kjell (Law); Laperrière, Marie
Sexual and Gender-Based Violence (SGBV) remains a pervasive issue in Bangladesh despite decades of legal reforms. While these reforms have expanded criminal sanctions, existing literature suggests that concerns regarding survivors’ access to meaningful justice persist nonetheless. Informed by anti-carceral feminism, intersectionality and postcoloniality, I examine the perceived inadequacies of Bangladesh’s current punitive approaches to SGBV. And consequently, explore how SGBV advocacy stakeholders in Bangladesh understand and imagine alternative approaches to justice within the country’s socio-legal context as a response to the existing system’s inadequacies. Using a qualitative research design, I interviewed lawyers, researchers, NGO practitioners, and caseworkers. The findings of my study are twofold. First, the analysis revealed that Bangladesh’s criminal justice system, like many criminal justice systems globally, is constrained by procedural delays, institutional shortcomings, and broader structural barriers. Second, despite such intersecting socio-legal and structural inequalities that continue to shape survivors’ ability to access justice, imagining justice beyond punishment remains profoundly difficult (even among SGBV advocacy stakeholders who are deeply aware of the limitations of punitive justice). This tension suggests that the challenge is not simply the absence of alternatives but the broader social, legal, and historical conditions that constrain justice imagination in the first place.
Access status: Open Access ,
Criminal justice system involvement among children and youth with autism spectrum disorder in Manitoba
(2026-08-12) Yousefiasl, Sepideh; Ricci, M. Florencia (Pediatrics and Child Health); Hamad, Amani (College of Community and Global Health); Urquia, Marcelo L.
Background: Autism Spectrum Disorder (ASD) is a neurodevelopmental condition that affects social communication and interaction and is often accompanied by restricted or repetitive behaviors. Earlier studies suggested that certain ASD-related behaviors may increase the likelihood of involvement in the criminal justice system (CJS), whereas more recent studies have reported contrasting findings. However, in Manitoba, Canada evidence is limited. This study had two objectives: (1) to compare ASD case-identification algorithms using administrative data available at the Manitoba Centre for Health Policy (MCHP) and identify the algorithm that best captures individuals with ASD, and (2) to examine age- and role-specific frequency and time to first criminal justice system involvement (CJSI) among individuals with ASD compared with matched controls in Manitoba, Canada. Methods: Linked administrative data housed at the Manitoba Centre for Health Policy were used. Two ASD case-identification algorithms were compared: a restrictive algorithm requiring more ASD-related physician claims and a relaxed algorithm requiring fewer claims, while all other criteria remained the same. A retrospective matched cohort study was then conducted, matching everyone with ASD to three individuals without ASD on age, sex, and Regional Health Authority. Cox proportional hazards models were used to examine time to first CJSI, and zero-inflated negative binomial models were used to assess CJSI frequency. Results: The restrictive algorithm yielded higher diagnostic certainty and positive predictive value, whereas the relaxed algorithm demonstrated higher sensitivity. Based on these findings, the restrictive algorithm was selected for the primary analyses. Individuals with ASD had a significantly lower risk of first CJSI and fewer CJS contacts than matched controls. This association was consistent across involvement roles and age groups. Sensitivity analyses supported the robustness of the findings. Conclusions: These findings highlight the importance of algorithm selection when identifying ASD using administrative data. In this Manitoba population, ASD was not associated with an increased risk of CJSI and was linked to lower overall involvement in the criminal justice system. Further research is needed to better understand discrepancies across studies and jurisdictions.
Access status: Open Access ,
Remote and proximal sensing for potato nitrogen status assessment and yield prediction in Manitoba, Canada
(2026-08-05) Chatraei Azizabadi, Ehsan; Lobb, David (Soil Science); Thiessen Martens, Joanne (Soil Science); Kishchuk, Barbara (Soil Science); Shirtliffe, Steven (Plant Sciences, University of Saskatchewan); Badreldin, Nasem
Potato production requires accurate and timely assessment of crop nitrogen (N) status and yield potential, yet conventional field sampling is labour-intensive, spatially limited, and often inadequate for representing within-field variability. This dissertation examined the use of remote and proximal sensing, integrated with machine learning (ML) and deep learning (DL), for potato N assessment and yield prediction under Manitoba field conditions. It combined critical synthesis of the Canadian evidence base with empirical evaluation of sensing approaches of increasing spectral and structural complexity to clarify how different forms of sensor-derived information contribute to prediction under field conditions. The review component showed that remote and proximal sensing have clear value for potato nutrient and yield management in Canada, but that the available literature remains geographically concentrated and methodologically inconsistent. Empirical analyses then demonstrated that unmanned aerial vehicle (UAV)-derived multispectral imagery can support in-season estimation of petiole nitrate-N (NO₃⁻-N), with the best Random Forest model reaching R² = 0.571 and RMSE = 0.459% after recursive feature elimination, indicating that prediction depended on both algorithm selection and feature composition. The analysis was extended through visible and near-infrared to shortwave infrared (VNIR-SWIR) hyperspectral reflectance spectroscopy acquired under field and laboratory conditions. In Chapter 4, the strongest petiole NO₃⁻-N prediction was obtained from laboratory fresh-leaf spectra using a one-dimensional convolutional neural network (1D-CNN; R² = 0.90, RMSE = 0.22%), compared with R² = 0.77 and RMSE = 0.27% for the best field canopy model. Calibration transfer improved field-based prediction, increasing 1D-CNN performance to R² = 0.82. The dissertation further evaluated UAV-derived light detection and ranging (LiDAR) and multispectral data within single-task and multi-task DL frameworks for simultaneous estimation of tuber yield and petiole NO₃⁻-N. Multi-task learning improved yield prediction more consistently than petiole NO₃⁻-N prediction, increasing test-set yield R² from 0.26 under the best single-task model to 0.46–0.60 under multi-task models. Yield prediction relied more broadly on spectral and LiDAR-derived structural features, whereas petiole NO₃⁻-N prediction was mainly driven by spectral indicators. Overall, the findings clarify how sensing depth, data type, and modelling framework shaped potato N status assessment and yield prediction under the conditions investigated.
Access status: Open Access ,
Fourier coefficients of Eisenstein series and explicit computations of Whittaker functionals
(2026-07-23) Senturk Avci, Sueda; Schippers, Eric (Mathematics); Chipalkatti, Jaydeep (Mathematics); Sankaran, Siddarth
This expository thesis bridges classical modular forms and the modern adelic framework, focusing on its application to the arithmetic Siegel-Weil formula. Transitioning to $\text{GL}(2,\mathbb{A})$, we explicitly construct non-holomorphic Eisenstein series and the local Weil representation. Our primary objective is the explicit evaluation of local Whittaker functionals for unramified binary quadratic spaces. Using Yang's local density formulas, we compute the central derivatives to verify an identity, providing a complete SageMath implementation for verification.
Access status: Open Access ,
Potential of digital image analysis in investigating concrete behavior at the mesoscale and macroscale
(2026-07-15) Khan, Filza Ahmed; Bassuoni, Mohamed (Civil Engineering); Wells, John (Civil Engineering); Svecova, Dagmar
This research investigates the potential of digital image analysis techniques to understand the multiscale behavior of concrete at both the mesoscale and macroscale. The study integrates fracture surface image analysis with macroscale fatigue response evaluation to establish relationships between concrete failure mechanisms and fatigue performance. In addition, Digital Image Correlation (DIC) techniques are employed to evaluate the strain behavior of concrete subjected to monotonic compression loading. A MATLAB-based image analysis workflow was developed to quantify fracture surface characteristics, including fractured aggregates, pullout aggregates, cement paste, and total aggregate area. The developed methodology was validated against manual measurements and demonstrated its capability to efficiently identify and quantify fracture surface parameters while capturing smaller and partially fractured aggregates that are difficult to detect through manual analysis. The quantified mesoscale parameters were subsequently correlated with fatigue life, cyclic strain response, stiffness degradation, and energy dissipation characteristics obtained from fatigue-tested concrete specimens. The results indicate that specimens exhibiting a higher percentage of pullout aggregates dissipated more energy during cyclic loading and experienced larger cyclic strains. In contrast, specimens with a greater proportion of coarse fractured aggregates generally exhibited lower energy dissipation and reduced strain levels, highlighting the significant influence of fractured mechanisms on concrete fatigue performance. A strong negative correlation was observed between average cyclic strain and fatigue life, indicating that specimens subjected to larger strains failed after fewer loading cycles. Based on this relationship, a fatigue strain–life model was developed using average cyclic strain values and demonstrated reasonable predictive capability for estimating fatigue life. To further evaluate image-based techniques, Digital Image Correlation (DIC) was employed to investigate failure development and stiffness response in aged and non-aged concrete specimens subjected to monotonic compression loading. The strain fields obtained through DIC successfully identified critical cracking regions and showed good agreement with the observed failure patterns. Although DIC generally produced higher stiffness values than conventional LVDT measurements, it effectively captured deformation trends and provided valuable insight into localized failure mechanisms. Furthermore, the results highlighted the importance of speckle pattern quality, image resolution, and overall experimental setup in achieving reliable and accurate DIC measurements. While the conclusions are limited to the concrete mixtures, specimen geometry, and loading conditions considered in this study, the results provide evidence of the underlying mechanisms governing fatigue failure and highlight the potential of image-based approaches for future concrete research and structural assessment applications.
Access status: Open Access ,
Using a novel virtual assessment to explore preschoolers’ emotion expression and mental health in the context of a dual-generation intervention for maternal depression
(2026-08-03) Stienwandt, Shaelyn; Newton Montgomery, Janine (Psychology); Giuliano, Ryan (Psychology); Stewart-Tufescu, Ashley (Social Work); Prime, Heather (York University); Roos, Leslie
Emotion regulation (ER) is understood to be a core component of normative mental health. Research suggests that preschool-aged children who experience difficulty with ER may be particularly at risk to develop emergent mental health (MH) symptoms. Therefore, it is critical to be able to accurately assess ER development in young children to identify at-risk children and families who may need additional support. This dissertation aimed to understand the feasibility of assessing one behavioural indicator of children’s ER, emotion expression (EE), using virtual methods and understand pathways of MH symptom change over time during their mother’s participation in a clinical intervention trial. Data was drawn from a three-arm clinical trial implemented toward the end of the COVID-19 pandemic. In study 1, pre-intervention data from mothers with depression symptoms and their preschool-aged child (N = 206) was examined. The majority (73%) of mother-child dyads chose to complete the virtual assessment rather than a brief online check-in without their child, indicating strong feasibility. T-tests showed that mothers who opted to complete the virtual assessment endorsed fewer harsh parenting and more positive parenting behaviours compared to mothers who declined to participate. Linear regressions showed that older child age and less parent-child dysfunctional interactions predicted completion of the virtual assessment. Child physiological data (i.e., heart rate or sleep) was not a significant predictor of assessment completion. Further, linear regression showed older child age predicted more positive EE during the virtual assessment. In study 2, pre-intervention data were first used to test moderation models examining whether child physiology or maternal variables altered associations between child EE and MH symptoms. Significant pre-intervention moderation models were carried forward into longitudinal analyses using pre-post change scores to evaluate whether these associations held over the course of the trial. Results of exploratory bivariate correlations demonstrated positive associations with change variables of child MH symptoms, maternal MH (i.e., anxiety and anger), harsh parenting, and child sleep problems from pre-intervention to post-intervention. At pre-intervention, children’s resting HR was found to moderate the positive association between children’s post-win EE and MH, such that children with higher HR had stronger associations between positive EE and fewer MH symptoms. This moderation did not replicate with change scores of post-win EE and MH. There was no significant impact of the maternal MH intervention on children’s EE and limited results for children’s MH. Results highlight the feasibility of virtual assessment methods for young children, the need to adapt virtual assessments to support families who endorse harsh parenting, and revised iterations of dual-generation interventions to support parent-child MH.
Access status: Open Access ,
Genomic identification of rare and low-prevalence enteric bacterial pathogens in Canada: implications for taxonomy and reference laboratories
(2026-08-07) Donahue, Julie; Christianson, Sara (National Microbiology Laboratory); Bay, Denice (Medical Microbiology and Infectious Diseases); Graham, Morag (Medical Microbiology and Infectious Diseases); Nadon, Celine
Accurate identification of enteric bacteria is essential for public health surveillance, outbreak detection, and clinical decision-making. While whole genome sequencing (WGS) has increasingly allowed us to replace traditional diagnostic and characterization methods for common pathogens, its applications for rare and low-prevalence enteric bacteria remains poorly characterized. Traditional approaches such as 16S rRNA gene sequencing and phenotypic profiling, still used by diagnostic and reference laboratories, often lack sufficient resolution for species-level identification of closely related or uncommon taxa. Furthermore, WGS-based workflows introduce challenges related to sequencing platform choice, bioinformatic tool selection, reference database stability, and taxonomy. Evidence guiding the implementation of WGS for the identification of uncommon enteric bacteria in reference laboratory settings is limited. This study evaluated the performance of multiple WGS-based workflows relative to traditional identification methods using a retrospective collection of 331 rare and low-prevalence bacterial isolates submitted to Canada’s National Microbiology Laboratory for identification from 1995 to 2023, together with 34 type strains. Isolates were sequenced using Illumina short-read and Oxford Nanopore long-read platforms and analyzed with bioinformatic workflows incorporating multiple taxonomic classification tools. Results were compared with 16S rRNA gene sequencing and biochemical identification methods. Concordance analyses assessed agreement across methods, characterized discordant classifications, and evaluated taxonomic resolution and accuracy. Results from WGS-based approaches consistently achieved higher taxonomic resolution than traditional methods, particularly at the species level. However, classification outcomes varied substantially across sequencing platforms and bioinformatic workflows. Discordance was frequently observed among closely related taxa and organisms with limited or unstable representation in reference databases. Illumina-based taxonomic classifications demonstrated higher concordance and more consistent species-level classifications, whereas Nanopore-based workflows produced a greater proportion of genus-level, ambiguous, or unresolved classifications. Across all methods, unresolved and discordant results were associated with read or assembly quality, taxonomic reclassification, nomenclatural inconsistencies, or incomplete reference database coverage. Overall, these findings show that while WGS-based methods offer advantages for identifying rare enteric bacteria, diagnostic utility strongly depends on the sequencing platform used, its analytical workflow design, and reference database curation quality. Together, these data identify the technical and taxonomic priorities that must be addressed for reliable implementation of WGS-based identification in public health and reference laboratories.
Access status: Open Access ,
Helping Older Adults Promote Emotional Well-being (HOPE): a co-designed intervention to reduce internalized stigma of seeking mental health services
(2026-06-15) Kaur, Jagpreet; Koven, Lesley (Clinical Health Psychology); Reynolds, Kristin (Psychology); Mackenzie, Corey
Mental health problems are common among older adults, yet this age group is the least likely to seek mental health services. Internalized stigma of seeking help, defined as the internalization of negative societal beliefs about seeking mental health services, is a key barrier. While various interventions exist to reduce internalized stigma among younger populations, only two have been specifically designed for older adults, both of which were developed before the new zeitgeist of involving individuals with lived experience in the intervention design process. Following this new zeitgeist, we used a community-based participatory research (CBPR) framework and co-design methods to guide the development of an anti-stigma intervention for older adults. Eleven team members, including four academic researchers and clinicians, five older adult community team members, and two video design experts, leveraged their expertise to collaboratively develop this brief video intervention over the course of eight monthly co-design meetings. These meetings were audio-recorded and transcribed verbatim. Resulting transcripts, field notes, and reflexive journaling helped document the shared decision-making process and how the principles of authentic partnership were enacted. Additionally, team members completed the patient engagement evaluation tool (PEET-6) anonymously at the end of each meeting, with results shared at the start of subsequent meetings to monitor and evaluate the ongoing quality of the collaborative partnership. Finally, we used structured co-design reporting guidelines to document our co-design process across three stages: planning, conducting, and evaluation.
Access status: Open Access ,
Advancing extended redundancy analysis via generalized estimating equations (ERA-GEE): integrating model-based recursive partitioning (MOB) for subgroup detection
(2026-08-05) Sichkaruk, Sam; Kruk, Richard (Psychology); Lix, Lisa (College of Community and Global Health); Kim, Sunmee
Extended redundancy analysis (ERA) combined with generalized estimating equations (GEE)—termed ERA-GEE—provides a robust framework for analyzing high-dimensional predictors and multivariate outcomes in large-scale observational studies. ERA reduces high-dimensional predictors into low-dimensional components that maximize variance explained in the outcome variable(s), while GEE accounts for within-subject correlations across repeated or multivariate outcomes. A key limitation of ERA-GEE is its assumption that all parameters are homogeneous across individuals, which may not hold when associations vary across sociodemographic subgroups. To address this, model-based recursive partitioning (MOB) was integrated into the ERA-GEE framework (termed MOB-ERA-GEE), enabling data-driven identification of subgroups with distinct parameter estimates. Simulation studies evaluated the performance of the method across varying conditions, including partitioning covariate type, subgroup balance, sample size, and magnitude of parameter heterogeneity. Performance was evaluated using Type I error, power, classification accuracy, and subgroup size recovery. Results from the simulation demonstrate that MOB-ERA-GEE reliably detects subgroup heterogeneity, and highlighted under which conditions optimal performance occurred. The practical utility of MOB-ERA-GEE was illustrated using data from the Canadian Longitudinal Study on Aging to examine heterogeneity in the social isolation–health relationship. Social isolation was measured using 29 interrelated predictors, summarized into three theoretically defined components, and examined in relation to executive functioning and grip strength. Results from the application demonstrate that MOB-ERA-GEE yields subgroup-specific insights into the social isolation–health relationship that conventional modelling approaches may overlook. These findings highlight MOB-ERA-GEE as a flexible, data-driven framework for uncovering population heterogeneity in high-dimensional observational data.
Access status: Open Access ,
Bleeding hearts: missions, philanthropy, gender, and Indigenous peoples in Saskatchewan, 1880-1920
(2026-08-04) Brglez, Karen; Thorpe, Jocelyn (History/Women's and Gender Studies); Carleton, Sean (History/Indigenous Studies); Fontaine, Lorena Sekwan (Indigenous Studies); Korinek, Valerie (History, University of Saskatchewan); Perry, Adele
Beginning in the 1870s, white women in Ontario and Quebec began organizing into women’s missionary societies to support Protestant missionary work at home and abroad. These groups organized the sending of female missionaries to work at Indian Residential Schools in Saskatchewan. The purpose of female missionary work was to convert ‘heathen’ children to Christianity as part of the Canadian state’s goal of ‘civilizing’ Indigenous peoples. Examining female missionaries and female supporters of missionary work as a single network explores how Anglo-Canadian women were connected to settler colonial processes. This dissertation investigates how sentimentality directed towards Indigenous peoples at residential schools and mission sites in Saskatchewan became a vehicle for expanding colonialism from 1880 to 1920. White women created an emotional culture around sentimentality which made colonialism appear as a humanitarian project on the Canadian prairies. Missionary women promised to love, care, and have compassion for Indigenous children and they expressed these sentiments to their female supporters in eastern Canada. The promises of sentimentality fused white women into a cohesive Christian emotional community which produced racial difference on the Canadian prairies. Race, gender, and religion were woven together creating the identity of the white female subject as generous and loving, and the Indigenous ‘other’ as a pitiable creature. Using diaries, letters, government records, and missionary memoirs, this study argues that as women began missionary work in Saskatchewan, they initially portrayed Indigenous peoples as ‘foreign heathens’ that they needed to convert. As settlers began pouring into Saskatchewan around the turn of the century, missionary discourse shifted and Indigenous people became ‘our Indians,’ a possessiveness rooted in paternalism and white supremacy. By centering emotion, this study draws out the tensions between the ways missionaries presented missions as benevolent, humanitarian efforts and the violence that underwrote colonialism.
Access status: Embargo ,
A threefold sociotechnical view of engineering to move beyond the culture of compliance
(2026-07-24) Bezerra Rodrigues, Renato Alan; Zacharias, Kari (Engineering Education); Hauseman, Cameron (Educational Administration, Foundations and Psychology); Friesen, Marcia (Engineering Education); Chichekian, Tanya (Biosystems Engineering); Jesiek, Brent (Purdue University); Seniuk Cicek, Jillian
The engineering profession in North America is marked by the narrative that engineers work for “societal benefit”. However, there is a disconnect between this narrative and engineers’ historical purposes and practices. Consequently, those who attempt to address social and environmental injustice through “neutral” practices may inadvertently further them. This disconnect stems primarily from an artificial separation of technical and social dimensions of the profession, commonly discussed as “socio-technical dualism” and “culture of disengagement”. This dissertation examines, through four studies, how cultural and systemic barriers prevent engineers from engaging with social and environmental justice issues. The first study uses inferential statistics to analyze 340 surveys from engineering students at the University of Manitoba (UM) to understand how they perceive the ranking engineering competencies. The second study – a scoping literature review of “sociotechnical thinking” in engineering education – maps the existing pedagogical approaches, successes, and challenges to researching and teaching this topic. The third study proposes a conceptual framework that conceptualizes engineering as a three-dimensional sociotechnical process. It uses a theoretical approach to synthesize three different, but complementary perspectives on the sociotechnical nature of engineering: its social construction, sociotechnical translation, and system closure. The fourth study employs critical discourse analysis on 18 student co-op reports from the UM to investigate if and how the ideology of technological neutrality informs their understanding of their work experiences. The findings from these studies indicate that: UM engineering students perceive the frequency and criticality of technical competencies as inverse from the impact of engineering; the ideology of technological neutrality is a barrier to engineering educators and students when engaging with social and environmental justice topics; and that this ideology manifests in a restrictive “culture of compliance”, in which students portray a depersonalized professional identity, decontextualized practice, and hierarchical relationships and specialized knowledge within their work. Based on these findings, I advocate for the use of the three-dimensional framework to scaffold educational content and pedagogical initiatives, intentionally redesign professional socialization, and take a political stance to move engineering education from a paradigm of compliance toward one of critical awareness and agency.
Access status: Open Access ,
Exploring athletes' organic self-talk in competitive sport settings
(2026-06-29) Garcia Reyes, Estefania; Strachan, Shaelyn (Kinesiology and Recreation Management); Leslie-Toogood, Adrienne (Canadian Sport Centre Manitoba); Strachan, Leisha
Background. The area of self-talk has many applications to sport settings, from psychological skills interventions programs as well as the actual use of self-talk by athletes in competition. Overall, self-talk in sports appears to serve several cognitive and behavioral functions that are related to performance, emotions, and self-regulation. However, many questions remain open regarding how, why, where, and under what situations self-talk changes, especially from a qualitative perspective. Aims. This research seeks to explore athletes' experiences in a competitive environment to better understand how self-talk changes before, during, and after competition. Methods. Semi-structured and open-ended interviews were conducted with competitive athletes from individual sports, using a video-assisted approach. Data were analyzed using reflexive thematic analysis. Implications. This study could offer insights into athletes lived experiences using self-talk in competitive settings and may serve as a valuable resource for athletes, coaches, and practitioners in developing strategies that could help redirect self-talk towards more constructive patterns, promoting positive experiences and healthier practices during competition.
Access status: Open Access ,
The lived experiences of Filipino Canadians in Manitoba who pursued social work education in Canada
(2026-08-09) Laureano, Rafael; Batac, Monica (Social Work); Luo, Hai (Social Work); Milliken, Eveline
This thesis explores the lived experiences of Filipino Canadians in social work education in Canada, to address the limited literature on their unique experiences within a Manitoban context. This research implements anti-oppressive practice (AOP) and anti-oppressive social work research (AOSWR) as its theoretical framework and constructivist grounded theory (CGT) as its methodological framework. This research used transcripts of 12 participant interviews, analyzed through open, axial, and selective coding, supported by analytical memos in an iterative process to develop a theoretical description of their experiences, encapsulated in five core themes: family and cultural influences, pathways into social work education, mental health and self-reflection, barriers, and educational experiences. From these themes, an analysis was constructed that highlights the additional burden Filipino Canadians face, such as cultural negotiation in white spaces with few Filipino role models, limited understanding of social work within their Filipino families and communities, and internalized oppression that undermines their sense of self in their academic experiences. This research also offers recommendations, based on excerpts from participants, to address these issues for researchers, government and policymakers, social work faculty members and educators, and current and future students.
Access status: Open Access ,
Evaluating and improving missing-data handling in ERA-GEE: a simulation and empirical study with high-dimensional longitudinal data
(2026-08-05) Phillips, Bryce; Kruk, Richard (Psychology); Jiang, Depeng (College of Community and Global Health); Kim, Sunmee
Extended Redundancy Analysis combined with Generalized Estimating Equations (ERA-GEE) provides a flexible framework for modeling high-dimensional longitudinal data by jointly summarizing complex predictor domains and accounting for correlated outcomes. A key limitation of ERA-GEE, however, is its reliance on a complete dataset for estimation. In practice, missing data are typically handled through preprocessing approaches such as listwise deletion (LWD) and multiple imputation (MI) prior to model fitting. Because these procedures are detached from the component construction, estimating-equation structure, and correlation assumptions that govern ERA-GEE, they can induce biased estimation and inefficient inference. This thesis proposes a least-squares imputation (LSI) strategy that is embedded directly within the ERA-GEE estimation procedure, yielding a unified framework for simultaneous parameter estimation and missing-data management. Rather than deleting or imputing observations prior to analysis, the proposed LSI-ERA-GEE approach iteratively updates model parameters and missing values within a single estimation criterion. A Monte Carlo simulation study evaluates the performance of LSI-ERA-GEE relative to LWD and MI across a range of missing-data conditions, including varying proportions of missingness and multiple missing mechanisms. Results show that across missing-data conditions, LSI-ERA-GEE reduces bias and improves estimation accuracy relative to competing approaches, particularly under low-to-moderate proportions of missingness. An empirical application using cognitive outcomes measured at three timepoints from the Canadian Longitudinal Study on Aging further demonstrates that parameter estimates, statistical significance, and predictive performance are highly sensitive to the choice of missing-management method. Overall, LSI-ERA-GEE produces the most precise estimates and lowest prediction error.
Access status: Open Access ,
“Take the picture or not take the picture?”: children’s awareness of when private information should be protected
(2026-05-01) Petriw, Sarah; Goulding, Brandon (Psychology); Greening, Steven (Psychology); Nancekivell, Shaylene
Children develop an understanding of privacy through real and virtual experiences. As digital sharing becomes increasingly common, so does the risk of privacy violations, highlighting the growing need for researchers to study children’s reasoning about when and how to protect their personal information. The present study examined how children reason about the sharing of personal images and how this reasoning changes with age. Two within-subjects experiments were conducted with children aged 5–8 years old. In Experiment 1 (N=82), children were told that a girl named Sally should “keep things private” while playing an online game and then decided whether the game should take pictures of either real Sally or a cartoon avatar in different household settings. Younger children (5–6 years old) were generally uncertain about whether images should be shared, whereas older children (7–8 years old) were more likely to endorse withholding images of real Sally but remained uncertain about cartoon images. An exploratory analysis further indicated that younger children who understood the privacy prompt were more likely to endorse withholding images than those who did not. Experiment 2 removed the cartoon avatar to test whether children’s responses were driven by the image content. Results (N = 87) replicated the pattern of withholding images of real Sally while showing greater acceptance of sharing cartoon rooms, suggesting that with age children increasingly consider the personal nature of information when reasoning about privacy protection.

Browse

Recent Submissions