62588 research outputs found
Sort by
Mechanical activation of von Willebrand Factor
No full textThe von Willebrand factor (vWF) is a critical multimeric glycoprotein that mediates hemostasis by facilitating platelet adhesion under shear stress. Its activation is tightly regulated by mechanical unfolding of distinct domains and interactions with various plasma proteins. This dissertation explores the biomechanical activation of vWF from both domain-specific and full-structure perspectives, using a combination of single-molecule force spectroscopy, structural modeling, and functional assays.Collectively, these studies offer a comprehensive framework for understanding how mechanical forces and domain-specific interactions regulate the primary hemostatic function of von Willebrand factor (vWF), and how dysfunction in this system can lead to a spectrum of bleeding and thrombotic disorders. The findings reveal that disease-associated mutations—whether within vWF itself or in its interacting plasma partners—can disrupt this finely tuned mechano-regulatory process. This work provides novel mechanistic insights with significant translational potential for the development of targeted therapies aimed at restoring or modulating vWF function in pathological conditions.</p
Spiritual Monarchs: The Nonjuring Movement and the Development of Separation of Church from State within the Church of England, 1691-1775
No full textAfter the Glorious Revolution, where the new king William sacked clergy who would not swear a new oath of loyalty, Nonjurors broke with the conforming Church of England. For these separatists, the Church of England spiritually autonomous society under bishops who were spiritual sovereigns. If the state infringed on the church, then there should be separation of church from state. These radical ideas called into question the establishment of religion and a national church, at the same moment that Enlightened theorists sought for a deeper basis for the establishment of society. The ideas of the Nonjurors not only influenced their sect, but churchmen who conformed to the post-revolutionary Church of England, who advocated for a complete divorce between spiritual and civil authority so they may remain untouched by secular powers, even as they claimed great, but unused, spiritual powers. Similarly, Nonjuring separatism also provoked a reaction from those who sought to establish a Protestant civil religion through godly state. These notions of separation of church from state spread across the Atlantic in North America. In planting the Church of England in English colonies, Nonjuror ideas influenced clergy who advocated for their own spiritual privileges against royal government and laity, pushing for a self-regulating, wholly spiritual, church. Similarly, in the bishop controversy of the 1760s, these rival visions of church and state relation flared, with supporters of a colonial bishop desiring an entirely detached, wholly spiritual, church within a British empire that was broadly tolerant, even to non-Protestants. These Anglicans pursued a form of religious liberty that mapped onto the denominationalism that prevailed throughout the future United States.</p
Long-Term Memory Transformation: Exploring the Conditions Fostering the Retention of Detail and Gist in Visual Recognition Memory
No full text{"value":"Memory consolidation involves the gradual transformation of episodic memories from detailed, context-rich recollections to more generalized, gist-like representations. One theory that addresses this process is the Competitive Trace Theory (CTT; Yassa & Reagh, 2013), which proposes that reactivation strengthens the overlapping elements of a memory while non-overlapping components, such as contextual or perceptual details, compete and degrade over time. As part of my dissertation, my two published studies (Zhang & Hupbach, 2020, 2023) tested key assumptions of CTT. The first study found that repeated encoding preserved detailed memory over time, challenging CTT’s claim that repeated reactivation leads to the degradation of perceptual details. The second study introduced context variability and found that perceptual memory was strengthened, while source memory degraded, partially supporting CTT’s view that non-overlapping contextual features are especially vulnerable to interference.Existing consolidation theories lack specificity about the conditions that promote gist extraction versus those that foster the preservation of detail. To address these gaps, I propose a new framework of memory transformation, specifying how different types of memory reactivation influence the long-term fidelity of episodic memory. I predicted that reactivation involving perceptual details would preserve specificity and improve lure discrimination, while reactivation emphasizing semantic content would accelerate transformation toward gist and increase false recognition of similar items.
To test these assumptions, I conducted two new experiments. Experiment 1 examined the impact of conceptual reactivation on memory, specifically testing whether multiple reactivations would accelerate gist extraction and reduce item specificity. The findings partially supported these predictions: only repeated reactivation, not a single instance, impaired lure discrimination, suggesting that repeated retrieval introduces competition among memory traces.
Experiment 2 explored how different types of practice (perceptual versus conceptual) affect long-term memory fidelity. I expected that retrieval requiring the recall of visual details would preserve memory specificity, while semantic-focused processing would integrate the object into existing knowledge at the expense of detail retention. However, the results did not support these predictions. Single conceptual reactivation improved target recognition but had no significant effect on target-lure discrimination. In contrast, perceptual reactivation failed to improve target memory and led to increased lure false alarm, likely due to the use of non-unique and repetitive perceptual cues that created interference, along with memory distortions from the lack of re-exposure.
These findings are discussed in relation to theories of memory consolidation, with a focus on how reactivation contributes to memory transformation. I also address the limitations of the proposed framework and the current studies and outline future directions for research aimed at clarifying the mechanisms by which different reactivation strategies shape long-term memory fidelity.","attr0":"abstract"
Apatite geochemistry of the 1875 caldera-forming eruption, Askja Iceland
No full textAskja is an on-rift caldera system in Iceland that erupts bimodal (rhyolite and basalt) compositions. The eruptive history of Askja is defined by two caldera-forming eruptions in 10 ka and 1875. Investigating volatile behavior preceding a cataclysmic eruption is invaluable for elucidating deep magmatic processes that drive the intensity of explosivity at the surface. This study’s approach in constraining volatile behavior at Askja preceding the 1875 caldera-forming eruption is apatite (Ca5(PO4)3(OH, F, Cl) geochemistry. Apatite analyzed in this study are free phenocrysts and inclusions in clinopyroxene and titano-magnetite. Halogen and trace element concentrations were determined with EPMA. Apatite chemistry is homogeneous in all eruptive units (Phreatic vs Plinian), analytical spots (core vs rim), and contexts (inclusion vs phenocryst) with no differentiation trends. EasyMELTS modelling indicates a liquidus temperature of 1031 ºC and apatite saturation at 875 ºC. Taken together, the 1875 eruption tapped one homogeneous silicic cap in which apatite grew in a period of equilibrium. The apatite in this study do not record changes to water saturation or redox conditions. Considering other Icelandic systems, Askja shows comparable LREE signatures to Hekla (transitional setting), both of which are depleted in comparison to Torfajökull (propagating rift setting). The LREE enrichment at Torfajökull suggests mantle plume enrichment, whereas Askja and Hekla chemistry is rift dominated. Future work is planned to analyze residual glass and apatite inclusions in orthopyroxene and feldspars to further clarify pre-eruptive trends at Askja.</p
Three Essays on Pre-login Search Behavior, Display Prominence, and AI-driven Beauty Filters on Online Platforms: Models and Evidence
No full text{"value":"This dissertation explores online consumer behavior and platform strategies within mobilecommerce and social media contexts. Employing empirical and analytical methodologies, it contributes
novel insights and implications to the Business Information Systems literature.The first chapter investigates how consumers’ pre-login and post-login search behaviors separatelyinfluence their purchase likelihood. While prior research addresses online search behavior, it rarely
differentiates between searches conducted before and after login. Using clickstream data, econometric
analyses reveal that both pre-login and post-login search efforts positively impact purchases. Surprisingly,
pre-login search effort has a greater impact than post-login search effort on consumer purchases,
demonstrating the importance of the pre-login search effort. I estimate that the potential effects of the pre-
login search effort range from 2.37 to 5.07 times greater than those of the post-login search effort. Further,
both search efforts exhibit diminishing returns, indicating that excessive searching can negatively influence
purchase decisions. Theoretical and practical implications are discussed.The second chapter employs a game-theoretic approach to examine online retailers’ strategies forprominently displaying product attribute buttons, given consumer heterogeneity. Particularly relevant to
mobile commerce contexts, consumers prefer retailers who streamline product navigation by prominently
displaying attribute-based search buttons (i.e., items in clickable pull-down menus, arrays of click-through
icons, and checkbox lists) of its faceted product-search system for the product attributes (i.e., product
features, brands, types, or categories) matching their preferences. Analyzing retailer behavior in a
duopolistic market, the study identifies factors, including consumer knowledge, retailer knowledge, and
consumer navigation costs, that influence whether retailers choose identical or differentiated the product
attribute buttons they prominently display. A key finding is that retailers strategically differentiate attribute
buttons, with one retailer targeting experienced consumers knowledgeable about specific product attributes,
and the other targeting a broader but less attribute-aware consumer segment.The third chapter shifts focus to social media, specifically examining User-Generated Videos(UGVs) through the lens of Self-Presentation Theory. Using multimedia analyses of over 15,000 UGVs, I
develop econometric models to quantify how facial appearance (i.e., smiling, direct gaze, mouth movement),
vocal appearance (i.e., singing quality, vocal brightness, vocal dynamics), and AI-driven beauty filters
impact user engagement (i.e., video likes). Key findings are: (1) Facial appearance such as smiling, direct
gaze, and mouth movement, vocal appearance such as singing in tune, and the use of AI-driven beauty
filters have positive effects on video likes, whereas wearing glasses has a negative effect on video likes.
The effects of facial appearance and AI-driven beauty filters on video likes are much greater than that of
the user’s vocal appearance, suggesting that, counterintuitively, the face is more important than the voice
even when it comes to music-based videos. (2) AI-driven beauty filters not only weaken the positive effect
of facial appearance on video likes but also weaken the negative effect of wearing glasses on video likes,
demonstrating a substitute effect. (3) There exists a time decay effect for video likes; that is, the number of
video likes is diminishing over time. Interestingly, the effects of facial appearance in terms of smiling and
direct gaze and vocal appearance in terms of singing in tune are greater immediately and decay faster over
time, whereas vocal appearance in terms of vocal brightness and vocal dynamics are weaker immediately
but stronger over time. Theoretical and managerial implications are thoroughly discussed.","attr0":"abstract"
The pH (Low) Insertion Peptide Selectively Delivers Immunotherapeutic Agents to Tumor Cells and Induces Immune Cell-Mediated Killing
No full textCancer immunotherapy agents are designed to engage potent immune system components to elicit an anti-tumor immune response. Existing immunomodulatory agents have limited efficacy and carry significant risks, including severe and life-threatening side effects. Targeted immunotherapies, including monoclonal antibodies, antibody-drug conjugates, and chimeric antigen receptor T (CAR-T) cell therapies, are effective in treating cancers that express targetable biomarkers. Engineered therapeutic antibodies and immune cells (e.g., CAR-T cells) rely on tumor-associated surface antigens to bind directly to tumor cells and initiate immune-mediated cytotoxicity. This approach is effective against a few cancers that express sufficient antigens. However, relying on tumor-specific antigen expression to achieve a therapeutic outcome severely limits the range of treatable cancers. Many cancers lack targetable biomarkers. Moreover, the heterogeneity and phenotypic plasticity inherent to cancerous cells provide opportunities for the selection of biomarker-deficient clones during therapy, which gives rise to drug resistance and treatment failure. To avoid these types of deleterious responses, new tumor-targeting strategies are warranted. Our approach herein relies on the pH-low insertion peptide (pHLIP), which can target solid tumors based on their inherent extracellular acidity, to deliver immunotherapeutic agents to tumor cells and facilitate immune cell-mediated killing. Specifically, we designed a series of antigen-pHLIP conjugates to decorate tumor cell surfaces and demonstrate that pHLIP can engage antibodies, CAR T/NK cells, and antibody-drug conjugates with tumor cell surfaces, leading to tumor cell death. These agents have the potential to reduce the toxicity and resistance associated with targeted immunotherapies and effectuate therapeutic outcomes in a wide range of solid cancers.</p
Topics in Mixture Regression Models: Parametric and Non-Parametric Approaches with High-Dimensional Covariates
No full textMixture regression models are useful statistical tools in a variety of applications and have been widely studied in the literature of statistics. In this dissertation, we in- vestigate some topics regarding extensions of finite mixture of linear regression. We first propose a fully parametric mixture of regression models allowing for covariates- dependent mixing proportions and variances. The EM algorithm is derived for pa- rameter estimation. We conduct simulations to examine the performance, and apply the proposed model to the analysis of a CO2-GDP dataset for illustration. We further study a nonparametric finite mixture of regression models where the mixing propor- tions, regression means and varaiances are all nonparametric functions of covariates. The model is originally proposed to deal with only one-dimensional covariates. We generalize the asymptotic results and the estimation procedure to multivariate cases. Monte Carlo simulations are conducted to examine the finite sample performance of estimates. The proposed model is illustrated by the analysis of the Pima Indian di- abetes data. We also investigate the variable selection problem in a semiparametric mixture of regression models to deal with high-dimensional covariates. The model under study assumes nonparametric mixing proportion functions and parametric lin- ear component regression. We propose a penalized local likelihood method within the framework of kernel regression for variable selection. A modified EM algorithm is developed and applied in the back-fitting estimation procedure for parameter estima- tion. The asymptotic properties of the resulting estimate are investigated. We derive1
the asymptotic bias and variance for the estimate and further establish the asymp- totic normality. The oracle property for variable selection is also established. The finite sample performance of the proposed method is inspected by a Monte Carlo sim- ulation study. The proposed method is demonstrated by an application to selecting important genes for breast cancer data.</p
Elementary Chronic Absenteeism: A Qualitative Approach to Policy into Practice
No full text{"value":"Chronic absenteeism, missing more than 10% of a school year for unexcused and excused reasons, has compounding negative impacts on the absent student, classmates, and society (Attendance Works, 2021; Pyne et al., 2021). One in six U.S. students were chronically absent during the 2017-2018 school year (Attendance Works, 2021), with only 17% of chronically absent kindergarten and first grade students reading on level in third grade (Aucejo & Romano, 2016), with approximately 11% of kindergarteners and 9% of first grades chronically absent (Romero & Lee, 2007). Thirty-six states under The Every Student Succeeds Act of 2015 selected chronic absenteeism as a measure of School Quality or Student Success, placing value on student attendance and holding schools accountable for their chronically absent students (Jordan & Miller, 2017; Kostyo et al., 2018). The factors influencing chronic absenteeism are interrelated across environmental, family/student and school categories, with no one specific factor impacting student chronic absenteeism (Lenhoff & Pogodzinski, 2018). School tracking systems are not equipped to properly measure chronic absenteeism (Lynch et al., 2015) and addressing policy into practice at the school level highlight the need for accurate, meaningful, and consistent measurement of school attendance data. This study qualitatively analyzed chronic absenteeism within one district’s four elementary schools during the 2018-2019 and 2019-2020 school years. Analysis of school level attendance data, school policy documents, school attendance improvement plans, and personnel interviews, show chronically absent students are overrepresented in minority racial groups and are more likely to qualify for services including the national school lunch program, English Language Learner supports or special education in comparison to their regularly attending peers. Interview data highlighted the reasons for absences consisting of maternal mental health and wellbeing, homelessness, transience, and medical treatment. Kindergarten and first grade students account for half of chronically absent students, and a third of chronically absent students met that threshold the following year. The conflict between PDE policy and school policy limit tracking and intervention for chronically absent student. Policy recommendations are included.
","attr0":"abstract"
A Mechanism for Adapting Theoretical Models of Trader Behavior to the Real-World
No full text{"value":"We adapt a class of theoretical models to a practical yet robust market environ-ment that more closely resembles the real-world. We introduce a model for a market
with many buyers and sellers each with a single valuation for an object drawn from
a continuous distribution and accommodate a model for bilateral trade between a
single buyer and seller each with many possible discrete valuations for an object.
The buyer and seller valuations are independent random variables unknown to the
counterparty at the time of trade. We formalize a comprehensive algorithm capable
of recreating the result that ex-post efficiency is generally impossible in a market
where participants have private information. We later show that efficiency can be
restored to the bilateral trade model through a market-maker who either maximizes
payoffs to the buyer and seller, or to themselves.","attr0":"abstract"
