-
Prospective Students
-
Faculty and Staff
-
Current Students
-
Alumni and Friends
Sexual and gender minority (SGM; e.g., lesbian, gay, bisexual, transgender) individuals are recognized as a health disparity population due to the undue burden of mental and physical health disorders among this population. The National Institutes of Health Sexual and Gender Minority Research Office (NIH-SGMRO) generated a social-ecological research framework for SGM health disparity research, articulating need for further research in the areas of (a) minority stress, (b) resilience, (c) violence and discrimination, and (d) intersecting identities. Informed by this research framework, the current dissertation contains three studies attending to these four research areas. Study one is a grounded theory of SGM suicide; 30 interviews with SGM adults in the United States lead to the co-construction of the SGM Suicide Risk and Protection (SuRAP) which outlines the impact of minority stress on suicide outcomes for SGM adults. Study two is a psychometric evaluation of the Brief Resilience Scale (BRS) among a sample of alternative sexuality community members (e.g., persons engaging in non-monogamy and kink), validating use of the BRS in future resilience-based research among this population. Study three is an examination of the mental health outcomes of sexual harassment, using a Psychological Mediation Framework to assess the ways in which social support, emotion regulation, and internalized minority stress explain the sexual harassment-mental health linkage among trauma-exposed sexual minority women. Taken together, findings indicate that therapeutic modalities such as Affirmative Dialectical Behavior Therapy may be of clinical utility.
Higher order Turing formulae, denoted as T_r for r ∈ Z+, are a powerful result allowing one to estimate the total probability associated with words from a random piece of writing, which have been observed exactly r times in a random sample. In particular T_r estimates the probability of seeing words not appearing in the sample. To perform statistical inference, e.g., constructing the asymptotic confidence intervals, the asymptotic properties of the higher Turing formulae need to be studied.
In this thesis we extend the validity of the asymptotic normality beyond the previously proven cases by establishing a sufficient and necessary condition for the asymptotic normality of higher order Turing formulae when the underlying distribution is both fixed and changing. We also conduct simulation studies with the complete works of William Shakespeare and data generated from different underlying distributions to check the finite sample performance of the derived asymptotic confidence interval.
Based on our theoretical results we also developed two methodologies for authorship detection with real twitter data analysis.
In this collection of manuscripts, I develop a deeper understanding and insight into how the Coronavirus Disease 2019 (COVID-19) pandemic and subsequent transition to telehealth impacted 1) clinical electronic health record (EHR) data quality and data entry patterns, 2) provider perceptions of the EHR’s influence on care delivery, and 3) patient perceptions on barriers related to pandemic-induced telemedicine.
The COVID-19 public health crisis has disproportionately affected individuals and populations historically marginalized in healthcare and public health, including racial and ethnic minorities and individuals with low-income status. The COVID-19 pandemic has drawn new attention to and compounded the existing health and digital disparities in healthcare, with Black Americans being almost 4 times more likely to die from the virus than White Americans. Racial and ethnic health disparities have been historically unwavering and persistent within the United States. Furthermore, this crisis has ignited rapid implementation of digital healthcare solutions such as virtual healthcare (telehealth and telemedicine capabilities) and health information technology (HIT) accessed via mobile applications or online platforms. When assessing HIT’s effectiveness, efficiency, quality, safety, and equity, it is important to consider the reciprocal relationship between HIT and the COVID-19 pandemic. This is of marked significance, considering that virtual care technologies have been shown to exacerbate the digital divide and worsen disparities in a patient’s ability to access high-quality care.
The research in this dissertation is informed by the socio-technical and complex systems perspectives of improved human health via high-quality, safe, HIT-driven care, which maintains two central concepts: 1) multiple levels of influence affect a patient's health outcomes, such as care quality, costs, and patient safety; and 2) complex adaptive systems occur when many agents work together within an organization and patterns materialize as the agents adopt, "simple rules" that optimize outcomes, such as the patient experience and the clinical team’s performance. Understanding how these HIT-related behaviors and perceptions multidimensionally affect care delivery is imperative to maximizing the potential benefits of technology and data in healthcare and promoting the need for a concerted effort to ensure safe, high-quality, and equitable care delivery.
Chapter 1 reviewed literature on the relationships between HIT and care quality, patient safety, health equity, biases, and discrimination. In Chapter 2, we assessed the influence of external, societal factors on disparities in data quality and data entry patterns. We found that an external change to healthcare operations – which modifies clinical practice – was correlated with clinical data entry patterns. Also, we found significant differences between departments within the healthcare organization, suggesting there were data entry differences based on distinct care goals housed within different units. These findings underscore some of the conclusions found in Chapter 3 where we determined the multidimensional relationship between HIT processes and patient safety and quality by exploring how healthcare provider demographic and health system-related characteristics were associated with their perception of the EHR’s impact on care delivery.
Perception disparities were present by providers based on sex, age, race, ethnicity, board certification, telemedicine utilization, and years of EHR experience. The results from this research are striking - we uncovered that providers using the EHR and telemedicine were roughly 20 times more likely to perceive the EHR as beneficial for patient safety (OR=20.25; p<0.001), compared to approximately only 4 times more likely for care quality (OR=4.48; p<0.05). Despite providers reporting that they found the EHR more beneficial for patient safety than care quality – we found conflicting practical evidence when assessing patient perceptions of telemedicine barriers and their reported outcomes.
Chapter 4 assessed the effect of demographic and healthcare-related factors on patient perceptions of telemedicine barriers. We found that 76% of patients reported facing at least one telemedicine barrier, and 66% reported experiencing a medical error via telemedicine during the pandemic. Similarly, we uncovered patients were more likely to report experiencing a telemedicine barrier if they utilized the patient-facing EHR (OR=27.72), had been diagnosed with one to two chronic conditions (OR=10.06) and experienced a medical error (OR=1.22). Interestingly, patients were less likely to report experiencing a telemedicine barrier if they identified as Black (OR=0.10; p<0.001), Female (OR=0.06; p<0.05) and reported three to four diagnosed chronic conditions (OR=0.10; p<0.01). These findings align with prior literature indicating the historically pervasive inequities and disparities amongst these subpopulations. This has been shown to lead to less patient engagement and activation, specifically in Black women, as well as those considered as “super-utilizers” of the healthcare system, often due to complex physical, behavioral, and social needs.
Collectively, these studies advance our understanding of how external factors such as COVID-19, modified workflows, demographic, health system, and healthcare-related characteristics impact health information technology and data perceptions and behaviors. Our findings suggest that these perceptions influence diagnostic EHR data entry, technological utilization, digital care barriers, and corresponding patient outcomes. This dissertation contributes to the public health and healthcare literature by providing practical implications for health systems, clinicians, care teams, and patients. Especially those who interact with technology and data in healthcare settings that affect the efficiency, safety, quality, and equity of care delivery, as well as generated clinical and population health data. Our findings underscore the need for further analysis to understand the interactions between the environment, processes, workflows, technological designs, patients, and the core operative nature of the system itself. Health administrators, policymakers, and researchers must acknowledge that technology and data can act as a roadblock to achieving health equity throughout this nation’s healthcare systems if human and information technology systems continue to co-exist but not co-evolve concurrently.
In policy and practice, we must pull back the curtain and recognize and address the many forms of coded inequity that is present throughout our healthcare systems by becoming more aware of the social dimensions of technology that generate dominant and discriminatory structures encoded in apps, algorithms, and payment data used in health and healthcare.
Improving elementary student reading achievement has been a well-defined goal of many federal and state educational initiatives over the last several decades. To that end, vast amounts of resources have been funneled into professional development for literacy teachers to solve the problem of students who are unable to read proficiently on grade level through a focus on improving teacher literacy instruction. The employment of literacy coaches into elementary schools has been an embedded professional development strategy implemented to combat these low literacy achievement rates across the country. However, the effectiveness of literacy coaches in positively influencing teacher instruction has not been rigorously investigated and many questions still remain about their true impact of improving student reading achievement.
This qualitative case study examined how teachers perceive a change in their instructional beliefs and practices through intervention with a literacy coach and which factors of coaching teachers report as most influential on their classroom practice. The implications of coach/teacher interplay were investigated through the entrenched professional development experiences of three elementary literacy teachers and their school-based literacy coach to determine the specific strategies, conversations, and interventions that brought about a possible change in teacher beliefs and practices relating to the teaching of literacy in their classrooms. A within-case analysis revealed changes in teacher beliefs and practices surrounding small group reading activities including phonics integration as well as changes in whole group phonemic awareness (PA) instruction. A cross-case analysis uncovered meaningful themes highlighting coach proximity to practice including shared responsibility, the coach as a sounding board, and scaffolding coaching cycles. Detailed results combined with recommendations for future research work to advance the field in an effort to develop an effective model of literacy coaching implementation.
Despite recurring arguments over the course of a century, intentional education geared toward the whole child in schools has not occurred (Khalsa & Butzer, 2016; Sabey, 2019). Consequently, children often emerge from high school exhibiting sufficient academic content knowledge applicable towards a successful career path, but lack social emotional skills essential for the development of optimal mental health and well-being (Butzer et al., 2016). Birth to age eight is precisely the time when the foundation of the whole child originates and when the building blocks for future academic success and social emotional well-being are established (NAEYC, 1986; Copple & Bredekamp, 2009). As a result of an existing gap underlining early elementary educator perceptions and experiences of social emotional learning (SEL), the purpose of this study was to discover the perceptions and experiences of full-time lead educators and paraprofessionals who teach SEL in Kindergarten through Second grade classrooms. Data was collected through a qualitative multiple-perspective case study design using a semi-structured interview process. Interview transcripts were analyzed and coded using a within-case analysis. Data analysis led to the development of seven themes: (1) Defining SEL, (2) Preparedness in Teaching SEL, (3) Barriers of SEL, (4) Educator Roles and Responsibilities, (5) High Priority of SEL, (6) SEL as a Positive Influence/Impact on Students, and (7) Evidence of SEL Skills. The findings of this study suggest that educators in K-2 classrooms (1) place SEL as a high priority in their classrooms, (2) perceive that SEL has a positive impact and influence on students based on observations, and (3) indicate how barriers such as under preparedness and lack of support inhibit SEL teaching in their classroom whereas positive school culture and pertinent resources greatly assist in effective facilitation of SEL.
Short-term load forecasting (STLF) is a conventional process at power companies to serve for better decision-making in their daily operations. Weather factors play a key role in STLF. In practice, an online STLF system typically requires the use of weather forecasts as input when projecting the future load, with associated weather forecast errors. This type of forecasting is known as ex-ante forecasting. Nevertheless, most existing academic literature developed load forecasting techniques under the ex-post forecasting settings, where the actual weather information is used in the forecast period. Meanwhile, the robustness of STLF models to the real weather forecast errors has rarely been studied in the literature. The gap between the practice and the research study is often due to the shortage of historical weather forecasts. In this research, we aim to close this gap by proposing two new frameworks to select better models in short-term ex-ante load forecasting. Compared to the conventional research which focuses on ex-post load forecast accuracy in the model development, both frameworks consider the impact of real weather forecast errors and are better fitted to field practices.
The effectiveness of the proposed frameworks is confirmed using an empirical case study at a medium-sized US utility with load data from multiple supply areas and real temperature forecasts. Compared to a state-of-the-art benchmark that uses the historical ex-post load forecast accuracy for model selection, the first framework leads to 2.4% improved accuracy on average. A further study among the weather sensitive hours (i.e., the hours when a smaller error in the temperature forecast may lead to a greater inaccuracy in the load forecast) suggests that the first framework outperforms the benchmark by 3.1% on average, although its performance is subpar when the predicted temperature forecast accuracy gets worse. The second framework addresses this issue effectively and improves the accuracy of the first framework by 7.4% for the hours with worse predicted temperature forecast accuracy. Overall, the second framework leads to an average of 0.8% improvement over the first framework and 3.9% improvement over the benchmark among the weather sensitive hours.
The importance of flexible customer-side resources in transitioning to a clean energy future is becoming increasingly apparent. Flexible customer-side resources can resolve most issues associated with intelligent and low carbon power grids and, in the process, unlock new value streams for both resource owners and load-serving entities (LSEs) with access to those resources. However, most LSEs with access to numerous flexible customer-side resources often use them for single applications when these resources can provide multiple value streams simultaneously. This dissertation focuses on developing models and frameworks to help LSEs simultaneously capture multiple value streams from customer-side resources within their jurisdiction.
Firstly, a stochastic equivalent battery model (EBM) that provides a simple yet accurate representation of the overall power consumption flexibility associated with a commercial building is proposed. The proposed stochastic EBM combines model-based functional simulations and optimization techniques to quantify the overall flexibility of a commercial building with flexible resources such as heating, ventilation, and air-conditioning (HVAC), electric water heater (EWH), battery, and electric vehicle charging stations. Illustrative case studies showcasing how the proposed model fits into complex resource scheduling problems whose objectives either maximize or minimize some value reflecting the LSE’s intended outcomes are also considered.
Secondly, a stochastic optimization framework is proposed to help an LSE capture value streams involving bulk power system support services from its residential customer-side resources. The specific value streams of interest are energy arbitrage, peak shaving, and market-based frequency regulation, while the customer-side resources are residential HVACs, EWHs, and behind-the-meter (BTM) storage. A resource type-centric clustering method is employed. The proposed framework contains
two parts. The first part involves a day-ahead resource scheduling problem that captures uncertainties in energy prices, regulation prices, and frequency regulation signals. A voltage sensitivity matrix-based approach is proposed to capture the impacts of resource control actions on system voltages. The second part includes
two real-time resource dispatch algorithms capable of eliciting fast responses from the resources to frequency regulation signals from the market operator with minimal voltage violations. The scheduling model and dispatch algorithms are evaluated using a HELICS-based co-simulation platform and real-world market data from New York Independent System Operator (NYISO).
Thirdly, a stochastic optimization framework is proposed to help an LSE capture multiple value streams focused on distribution system operations from its residential customer-side resources. The value streams of interest are peak shaving, energy arbitrage, ramp rate reduction, loss reduction, and voltage management. The framework captures the impact of third-party aggregators on the LSE’s network and includes two dispatch algorithms - decision rule-based dispatch and optimal real-time dispatch.
Finally, a framework to help LSEs compensate owners of customer-side resources for multiple value streams is proposed. The compensation sharing approach classifies the LSE’s realized value into three categories - additive, super-additive and subadditive. The appropriate compensation-sharing mechanism is then defined for each value category. A special component of the compensation sharing mechanism that provides additional social benefits, specifically credit rating improvement, for low and medium-income flexible resource owners is also proposed.
There is little consensus in the literature as to which approach for classification of Whole Genome Shotgun (WGS) sequences is most accurate. In this defense, two of the most popular classification algorithms, Kraken2 and Metaphlan2, were examined using four publicly available datasets. Surprisingly, Kraken2 reported not only more taxa but many more taxa that were significantly associated with metadata. By comparing the Spearman correlation coefficients of each taxa in the dataset against more abundant taxa, it was found that Kraken2, but not Metaphlan2, showed a consistent pattern of classifying low abundance taxa that were highly correlated with the more abundant taxa. Neither Metaphlan2, nor 16S sequences that were available for two of four datasets, showed this pattern. These results suggest that Kraken2 consistently misclassified high abundance taxa into the same erroneous low abundance taxa. These “phantom” taxa have a similar pattern of inference as the high abundance source. Because of the ever-increasing sequencing depths of modern WGS cohorts, these “phantom” taxa will appear statistically significant in statistical models even with a low classification error rate from Kraken2. These findings suggest a novel metric for evaluating classifier accuracy.
Customer churn leads to higher customer acquisition cost, lower volume of service consumption and reduced product purchase. Reducing the outflow of the customers by 5% can double the profit of a typical company. Therefore, it is of significant value for companies to reduce customer outflow. In this dissertation research, we mainly focus on identifying the customers with high chance of attrition and providing valid and trustworthy recommendations to reduce customer churn.
We designed and developed a customer attrition management system that can predict customer churn and yield actionable and measurable recommendations for the decision makers to reduce customer churn. Moreover, reviews from leaving customers reflect their unfulfilled needs, while reviews of active customers show their satisfactory experience. In order to extract the action knowledge from the unstructured customer review data, we thoroughly applied aspect-based sentiment analysis to transform the unstructured review text data into a structured table. Then, we utilized rough set theory, action rule mining and meta-action triggering mechanism on the structured table to generate effective recommendations for reducing customer churn. Lastly, in practical applications, an action rule is regarded as interesting only if its support and confidence exceed the predefined threshold values. If an action rule has a large support and high confidence, it indicates that this action can be applied to a large portion of customers with a high chance. However, there is little research focused on improving the confidence and coverage of action rules. Therefore, we proposed a guided semantic-aided agglomerative clustering algorithm to improve the discovered action rules.
Each year in the United States (U.S), one in five adults experience mental illness and one in six youth ages 6-17 experience a mental disorder (NAMI, 2020). While mental illness can affect individuals at similar rates, minority populations suffer from existent disparities in mental healthcare that have been exacerbated by the impact of COVID-19. Help-seeking behaviors of racial and ethnic minorities in the US have historically been influenced by the lack of trust in the medical system. When experiences of prejudice and discrimination are present in the counseling experience, they lead to damaging outcomes for minorities including misdiagnosis, receipt of less preferred forms of treatment, increased rate of premature termination, and overall dissatisfaction with service delivery in minority clients (Ridley et al., 2010; Rutgers University, 2019). Counselors who do not address biases, assumptions, and their own epistemological views risk operating within the oppressive framework of the dominant culture (Katz, 2014; Owen, 2017; Owen et al., 2018; Sue et al., 1992). Despite the growing support of cultural humility as complementary or even an alternative to cultural competence in counselor multicultural pedagogy, little has been examined about the ways in which this perspective can be enhanced in counselor education programs. Therefore, a standard multiple regression was utilized to examine the impact of intrinsic spirituality, common humanity, and affective empathy on cultural humility in counseling students. Results indicated that common humanity contributed significantly to the prediction of cultural humility accounting for 16% of the variance. Implications, limitations, and recommendations for future research are discussed.
