Environmental and Occupational Health Sciences Institute
Rutgers University Community RepositoryNot a member yet
58345 research outputs found
Sort by
Babylon on the bayou
No full textThis thesis is a collection of poetry and prose inspired by Black life and culture in Houston, Texas. It explores the conditions and experiences of Black people in Houston. The collection follows a Black male youth speaker as he encounters community, strangers, and self in his quest to develop his sense of identity. The speaker travels across the realm of Houston, struggling against the limits of poverty, anti-black racism, gender and sexual identity, mental illness, etc. What is within this collection is an intimate view into one of the most culturally unique and diverse cities in the world that has gone mis- and underrepresented (ironic given its sprawling largeness) in the literary landscape.M.F.A.Includes bibliographical reference
Interrelationships among RN staffing, perceived workload, practice environment, and burnout among emergency department nurses
No full textBurnout among registered nurses (RN) is an ongoing nurse workforce issue. Yet, there are no quantifications of the impact of RN staffing, perceived workload, and practice environment on burnout among registered nurses in emergency departments.
Purpose: This study examined the interrelationships among RN staffing, perceived workload, practice environment, and burnout among hospital-based emergency department nurses in New Jersey (NJ).
Hypotheses: Six hypotheses were tested: 1) Low RN staffing is significantly associated with high burnout; 2) High perceived workload is significantly associated with high burnout; 3) Unsupportive nurse practice environments are significantly associated with high burnout; 4) Low RN staffing, high perceived workloads, and unsupportive practice environments are independent predictors of high burnout; 5) Unsupportive practice environments mediate the relationship between a) RN staffing and burnout, and b) perceived workload and burnout.
Methods: This study used a cross-sectional, correlational, survey design. Using a publicly available list of RNs licensed in NJ, potential participants were recruited via an email invitation that included a link to the electronic survey. A single-item measure of burnout was used. RN staffing was assessed as patient-to-RN ratios with a single item: How many patients were assigned to you on your last day of work? The Perceived Workload subscale of the Individual Workload Perception Scale was used to measure RN workload. The Practice Environment Scale was used to measure practice environment support.
Results: One hundred eighty-eight hospital-based emergency department RNs comprised the study sample. Seventy-two percent of participants reported moderate to complete burnout. One out of three nurses (34%) reported sustained or complete burnout. Patient-to-RN ratios were not significantly associated with burnout. Perceptions of high workloads and unsupportive practice environment ratings had significant independent direct effects on the odds of high burnout. An unsupportive practice environment significantly mediated an indirect relationship between high perceived workload and high burnout. The practice environment did not mediate an indirect effect of patient-to-RN ratios on burnout.
Conclusion: To reduce burnout among hospital-based emergency department RNs, findings from this study point to a pressing need for hospital and emergency department leadership to implement strategies designed to improve the practice environment and decrease RN workloads.Ph.D.Includes bibliographical reference
Representation in postsecondary education: effects on minority, low-income, and first-generation students
No full textDespite increased access to postsecondary education, historically underrepresented college students–minority students, low-income students, first-generation college students–still do not realize the same level of college success as their more privileged peers. While many factors drive outcome disparities, this dissertation investigates the ways in which a college campus, and the demographic composition of that campus affects the success of historically underrepresented students. While much theory, research, and supreme court litigation has focused on the compelling state interest of diversity for all students, this dissertation examines whether and how representation of underrepresented groups has effects on underrepresented students. Utilizing data from the NCES Educational Longitudinal Study as well as the Integrated Postsecondary Education Data System, I examine how representation impacts historically underrepresented students, whether those impacts are non-linear, and whether those impacts are mediated by or interact with high-school representation. The findings of this research are mixed. Ultimately, results suggest that representation matters for each examined group (minority students, low-income students, and first-generation college students). But, the degree to which representation matters is subtle; and the contours vary by outcome. This research contributes to the literature by adding an additional study of racial/ethnic representation, and less common quantitative studies on representation of low-income and first-generation students. This study also provides a combined analysis of the effects of high school and postsecondary representation.Ph.D.Includes bibliographical referencesIncludes vit
Chris and Jerel’s Initial Idea of Fairness in Ordinary Dice Games (Part 3 of 3)
No full textThis analytic is the third of three analytics that showcase the formulation of two students’ ideas of mathematical fairness. The analytics follow the argumentation of two students, Chris and Jerel, throughout two after-school sessions and one interview session as they investigate what makes a game involving rolling one or two dice fair. The first after-school session occurred on April 29, 2004. The second after-school session occurred on May 5, 2004, with the interview happening directly after the second session on the same day. The first analytic begins with Chris and Jerel discussing the fairness of a game involving a single die during the first session. By the end of the first session, Chris has designed his own game involving rolling two dice. The second analytic looks at Chris and Jerel’s work during the second after-school session and interview session where the pair explore a task investigating the fairness of a game involving rolling two dice. This game is different from the one Chris creates in the first analytic. The third analytic includes an investigation on the probability of rolling a sum of 7 versus a sum of 6 when rolling two dice that occurs during the interview.
In this analytic (the third of three analytics), Chris and Jerel are sixth-grade students discussing a problem about fairness with a game using two fair dice during the interview with Researcher Powell. Gameplay involves rolling two dice and assigning points to either Player A or Player B based on the sum of the rolled dice. Player A gets one point (and Player B gets 0) for sums of 2, 3, 4, 10, 11 and 12. Player B gets one point (and Player A gets 0) for sums of 5, 6, 7, 8 and 9. The first player to accumulate ten points is the winner. The questions that students are asking about this game are: 1) Is this a fair game? Why or why not? 2) Play the game with a partner. Do the results of playing the game support your answer? 3) If you think the game is unfair, how could you change it so that it would be fair?
During gameplay, Shay (2008) explains that researchers had encouraged students to record the outcome of each roll of the dice. This naturally led to students beginning to list out the sample space for the game. During the interview, Chris and Jerel show the sample space they developed. It is important to note that Chris and Jerel have a sample space of 21 outcomes rather than 36. Shay (2008) attributes this to students not considering symmetric pairs as separate events. For example, rolling a 3 and 4 versus rolling a 4 and 3. This is significant because in the sample space Chris and Jerel have created 6, 7, and 8 all have a 3/21 (approximately 0.143) chance of occurring, while in the sample space of 36 outcomes 6 and 8 have a 5/36 (approximately 0.129) chance of occurring and 7 has a 6/36 (approximately 0.167) chance of occurring. Due to the incomplete sample space, Chris and Jerel are investigating why the sum of 7 occurs more frequently, leading to a question in the fairness of rolling a single die.
At the end of the second analytic. Chris and Jerel have expressed the idea that rolling a 7 occurs more frequently than rolling a 6 even though, according to their notes, each number can be the result of exactly three outcomes. The boys claim that this is because the sums of 7 have more “large numbers” referring to rolling a 4, 5, or 6 on one of the two dice. During the third analytic Chris and Jerel carry out two trials to investigate the claim that when rolling a single die the “large numbers” (4, 5, and 6) occur more frequently than the “small numbers” (1, 2, and 3). The first trial results in more small numbers being rolled than large numbers, while the second trial results in more large numbers being rolled than small numbers. Combining the results of both trials reveals that 12 small numbers were rolled while 10 large numbers were rolled. This evidence causes Chris and Jerel to tentatively reject their claim that large numbers are more likely to occur than small numbers.
After viewing the analytic, it is worth reflecting on what might be a follow up task to challenge Chris and Jerel to find ALL possible outcomes.
Problem Task:
Dice game 2:
Roll two dice. If their sum is 2, 3, 4, 10, 11, or 12, player A gets one point (and player B gets 0). If their sum is 5, 6, 7, 8, or 9, player B gets one point (and player A gets 0). Continue rolling the dice. The first person to get ten points is the winner. (1) Is this a fair game? Why or why not? (2) Play the game with a partner. Do the results of playing the game support your answer? Explain. (3) If you think the game is unfair, how could you change it so that it could be fair?
[Note: The game favors Player B with a ⅔ probability of winning a point and a probability of approximately .935 of winning a game.]
Videos Referenced:
Title: B90, 46a, Probability problems: Dice games for two players (Student view), Grade 6, May 5, 2004, raw footage https://rucore.libraries.rutgers.edu/rutgers-lib/70978/
Title: B91, 46b, Probability problems: Dice games for two players (Work view), Grade 6, May 5, 2004, raw footage https://rucore.libraries.rutgers.edu/rutgers-lib/70979/
References:
Shay, K. (2008). Tracing Middle School Students’ Understanding of Probability: A Longitudinal Study. Rutgers University
“We at war:” The Bureau of Special Services and the surveillance of New York’s Black Left during the era of the urban rebellions
No full text“We at War,” examines the centrality of race to the history of American countersubversion and political repression through the first full length study of the New York Police Department’s political intelligence unit, or “red squad,” the Bureau of Special Services (BSS). Tracking the unit from its racialized origins as the “Italian Squad” at the turn of the century, it shows how BSS’s outsized influence on American countersubversive politics reached its apogee during the urban rebellions of the 1960s. When the first major urban rebellion of the 1960s erupted in Harlem in 1964, BSS already had Black undercover officers embedded in radical groups. Black communist Bill Epton’s conviction for “criminal anarchy” after the uprising, secured through testimony from a Black BSS infiltrator, fueled a conspiratorial narrative that Black leftists were fomenting urban disorder at the behest of foreign communist states. Over the next three years, BSS agents systematically infiltrated New York’s Black radical milieu, entrapping its young, impressionable members in outlandish plots. These covert operations shaped both state repression and conservative politics at the end of the 1960s. BSS acted as a clearinghouse for countersubversive intelligence, passing information to other municipal red squads through formal liaisons and influencing the targets and tactics of the FBI’s infamous “Black Nationalist-Hate Groups” Counterintelligence Program. Furthermore, BSS was an important political actor in the emerging coalition of the New Right. The grassroots right featured BSS operations against New York’s Black left in its alternative press, BSS agents promoted punitive policy responses to the urban crisis and discredited the Great Society in Congressional hearings, and it worked closely with private, right-wing intelligence operatives to demonize and surveil Black radical organizations. This project was made possible by the recent release of hundreds of thousands of surveillance files from the Bureau of Special Services. This dissertation is the first book project based on these previously secret intelligence files, which reorients our understanding of the surveillance and repression of the Black Liberation Movement. If, as Ellen Schrecker has argued, the FBI was the “bureaucratic heart” of McCarthyism, police intelligence units like BSS were its veins and arteries, channeling information through a self-reinforcing circulatory system that connected grassroots right-wing activists to the highest centers of power in the internal security state. While federal agencies like the FBI played an important role in this history, this project shows that racial disparities in the intelligence field were driven by a conspiratorial culture shaped as much on the streets of New York as it was in the halls of Washington D.C.Ph.D.Includes bibliographical reference
A multifaceted bioengineering toolbox to enable in vivo and in vitro characterization of polysaccharide synthases
No full textPolysaccharides are a major class of natural polymers found abundantly across all forms of life and play a critical role as structural, metabolic, or functional components in biomolecular processes. For example, the most abundant renewable polymer on earth, viz., cellulose, act as storage reserves or cell walls in plants and microbes that could potentially be converted into biofuels or useful bioproducts. Similarly, hyaluronan or hyaluronic acid is a multifunctional high molecular weight polysaccharide found in eyes and synovial fluid in joints that provides support and anchorage for cells, cell adhesion, and proliferation and facilitates movement and migration. Both cellulose and hyaluronan are synthesized by membrane-integrated glycosyltransferases (GT), namely cellulose synthase (CesA) and hyaluronan synthase (Has), respectively. Although these enzymes have been purified and characterized to some extent, there are huge gaps in our mechanistic, structural, and holistic understanding of these systems. Hence, in this work, we were interested in using protein engineering approaches and reconstitution strategies to understand these enzymes' biochemical and structural features in vivo and in vitro. Firstly, we studied the dynamics of cellulose synthesis in Arabidopsis plant protoplast cell walls in vivo. We designed different tandem carbohydrate-binding modules and biochemically characterized them to find the most suitable probe for real-time visualization of regenerating plant cell walls. In vivo cell wall regeneration observed using these protein probes provided fundamental insights into the cellulose synthesis mechanism, movement, and alignment on the surface of the plant membrane. Using these key events, a model was proposed for the regeneration of cellulose in plant cell walls. Then, cellulose synthases involved in primary (CesA5 from Physcomitrella patens) and secondary (CesA8 from Populus tremula x tremuloides) cell wall synthesis were heterologously expressed in Pichia pastoris, followed by purification and reconstitution into liposomes to study their substrate utilization, kinetics, and product characterization using biochemical assays. However, the expression of GTs in cell-based systems suffered from low yields because of numerous factors making them less suitable for structural characterization and nanodisc reconstitution. Fortunately, the wheat-germ-based cell-free expression system offered an alternative to the expression of these GTs. The cell-free expression system gave a higher yield, involved lesser steps, and was biochemically active. Mutations in the conserved residues of bacterial hyaluronan synthase (from Streptococcus equisimilis) showed an impact on substrate binding and polymer synthesis. Overall, this work demonstrates the development of various protein engineering toolkits that could be used to study and unravel the mechanism underlying polysaccharide synthesis with bioenergy and biomedical applications both in vivo and in vitro.Ph.D.Includes bibliographical reference
Parents’ lay beliefs about interracial contact and behaviors aimed at raising antiracist children
No full textInterracial contact is a key tool for reducing racial prejudice, including reducing children’s prejudice. Parents play a vital role in exposing their children to diversity and in the development of children’s racial attitudes. Thus, in the present study, I examined how White parents’ lay beliefs about interracial contact (i.e., whether parents are aware of the positive benefits of interracial contact for intergroup relations) relate to parents’ behaviors, including exposing their children to interracial contact. In a correlation study, 197 White parents completed a survey assessing their lay beliefs about interracial contact, indirect and direct diversity exposure behaviors, the diversity of the child’s and the parent’s social network, frequency of race discussions with one’s child, strategies to raise an antiracist child (general and interracial contact specific), importance of antiracism, bias awareness, and White privilege awareness. Analyses revealed that White parents’ lay beliefs about interracial contact are directly correlated with parents’ reports of greater engagement in indirect and direct diversity exposure behaviors, a greater frequency of race discussions and reporting a greater number of general and interracial contact specific strategies for raising an antiracist child. Lay beliefs about interracial contact were not related to the diversity of the child’s and the parent’s own social networks. Importance of antiracism and bias awareness did not moderate the relationships between lay beliefs about interracial contact and diversity exposure behaviors. This work suggests multiple avenues of future research exploring how parents’ lay beliefs may relate to reduced prejudice among their children.M.S.Includes bibliographical reference
Variation in state tax burden and family well-being
No full textAlthough taxes in the United States are generally thought of as progressive, state sales taxes and many state income taxes are in fact regressive, further depleting the resources of low-income households. Accordingly, evidence suggests that state taxes are associated with increases in mortality, non-marital births, crime, and high school dropout. Between states, there is significant variation in tax policies and the resulting tax burden on low-income families. However, very little research has been done on the effect of these policies on the well-being of families and children, and even fewer studies examine these policies in combination.
This dissertation aims to describe state tax policies over time and examine the effects of policies – singularly, in combination, and in totality – on two robust measures of family well-being, food insecurity and child birthweight. This dissertation leverages a new dataset, collected by the author, describing state tax policies over 20 years (1990 – 2019). Results from descriptive analyses of these policies demonstrate the significant variation in tax burden, such that a household at the Federal Poverty Line, in one state, could receive a refundable tax credit increasing their resources by 25%, while the same household, in another state, could lose 8% of their resources to state income and sales taxes.
This policy dataset is next linked to individual-level data from two nationally representative datasets to examine the effects of tax policies on household food insecurity and child birthweight. To examine the relationship between state taxes and food insecurity, the state tax policy data are merged with the nationally representative Current Population Survey Food Security Supplement (2002-2019). Results indicate that state sales taxes are particularly detrimental for household food security, while state EITCs are protective against household food insecurity. To examine the relationship between state taxes and child birthweight, the state tax policy data are merged with the Pregnancy Risk Assessment Monitoring System (PRAMS). These results indicate that state sales taxes on food and overall tax burden are associated with increased risk of mothers having a low birthweight infant. Results are contextualized with simulations of the effects of policy changes on the well-being outcomes.Ph.D.Includes bibliographical reference
Mood, media, and mental health: a study of therapeutic mood mediating technologies
No full textThis dissertation shows how digital therapeutic technologies, including popular mood-tracking apps and algorithmic social media, work to construct branded psychiatric identities and calibrate the mood of contemporary mental health discourse and practice.This mixed-methods qualitative study questions what happens when media platforms are afforded diagnostic gazes, and if these technologies collapse the already marketized mood that is tracked in work and retail spaces into the mood that is implicated in good or poor mental health. A critical psychology and media studies orientation examines today’s hyper-mediated mental health consumer as politically and (pop) culturally situated.
Therapeutic mood tracking technologies across media forms are re-shaping the ways we conceive of and experience mental health and mental health care in ways that largely align with political-economic and platform logics. While some tech startups exploit a mental health crisis, the technology cannot critiqued without critically addressing the sociocultural and historical conditions that underpin contemporary notions of mood and mental health. Common frameworks like “misinformation” in digital diagnostic cultures (such as those that result in self-diagnosis) are inadequate and misleading.
The project contributes the analytical framework of Therapeutic Mood (Mediating) Technology (TMT). TMT enrolls mood and affect as technical and discursive objects to be manipulated toward various ends, conflates modes of care with technological and economic models, and shapes mental health subjectivities. In the work chapters, this analysis highlights the collapse of conceptions of both mood and mental health into contemporary workplace and economic imperatives. Here, moods and affects are shown to be used socially in order to reconfigure workplace organization. Moody women’s apps are found to refute while reifying pathologized understandings of feminine mood swings, mental health, and their underlying assumptions. As digital mood tracking technologies, they serve an ongoing project of harnessing feminized emotionality while delimiting acceptable parameters and expressions of mood, mental health, and of gender. In the social media realm, TikTok users were found to imbue an ostensibly mood-reading algorithm with a clinical gaze, capable of accessing users’ subconscious selves and diagnosing them with a mood-related mental health disorder. TMT is often bolstered by a particular sensibility that runs through mood tracking and communication technologies historically–one that sees them as containing a compelling mixture of scientific and spiritual elements.This draws out important elements in the conception, implementation, and adoption of such technologies. All in all, the buy-in of these mood mediating technologies itself can be viewed as serving a therapeutic function—a threat response and a resolution of tensions inherent in power’s particular operations on our contemporary lives through digital spaces. Critical psychology frameworks are offered in order to situate the contemporary digital diagnostic cultures emergent in this study as historically bound and culturally contingent.Ph.D.Includes bibliographical reference
Trustworthy machine learning for securing cyber-physical systems
No full textThe safe operation of cyber-physical systems (CPS) is crucial in nearly all applications, from power-grid and chemical treatment systems to autonomous robots and traffic lights. Securing cyber-physical systems require addressing not only their physical components and controlling software, but also practical system constraints not present in conventional computer security. The increasing frequency and sophistication of attacks on these systems highlight the challenging nature of securing CPS and the need for better defense mechanisms.
In this thesis, we investigate the both the application of machine learning (ML) techniques for securing CPS, as well as the security of the ML models themselves. In the context of industrial CPS, we study the use of out-of-band side-channels for verifying the integrity of cyber-physical processes, and security risks these side-channels may pose. In the case of robotic vehicles like drones, we investigate the use of ML for finding vulnerabilities. And in the context of emerging edge and embedded cyber-physical applications, we propose a method for creating more robust and compact ML models for these systems to aid in their practical deployment.
First, we present a method for remote attestation of legacy industrial CPS which leverages physics-based fingerprinting as a makeshift root-of-trust resistant to replay spoofing attacks. Using an ML classifier to distinguish between correct and incorrect physical fingerprints, the verifier can ensure the integrity of software running on the controller during the attestation process. We demonstrate the feasibility of our framework by implementing it on a real CPS controlling a robotic arm.
Next, we show that out-of-band physical channels can be used for covert data exfiltration on CPS, despite the presence of state estimation monitors. This highlights a vulnerability which could allow attackers to establish a foothold for reconnaissance, which is an essential component of sophisticated persistent attacks. We demonstrate the feasibility of this method in several use cases.
Additionally, we conduct a security analysis of physical side-channel monitoring, an ML-based anomaly detection technique for industrial CPS. Unlike software-based methods for intrusion detection, monitoring hardware-based physical phenomena like power consumption or electromagnetic emanations is believed to increase security, as these signals are immutable and inaccessible to attackers. However, we find that despite the inherent challenges that physical side-channel monitoring poses, the technique remains vulnerable to attack in the form of maliciously-crafted adversarial examples which evade detection by the ML classifier. In our analysis, we demonstrate the process for creating these adversarial examples, evaluate the practicality of such an attack, and recommend additional mitigations.
Next, we propose a vulnerability assessment framework to explore robotic autonomous vehicle (RAV) security from a combined cyber-physical perspective. We present a data-driven method to illustrate that, despite state-of-the-art memory isolation efforts, RAV systems are still vulnerable to physics-aware data manipulation attacks. Utilizing a reinforcement learning-based method, we show how an attacker can exploit memory bugs and parameter defects in a legitimate memory view to elaborately craft adversarial variable values and disrupt an RAV's safe operations. We demonstrate the feasibility of this approach on the widely-used ArduPilot RAV framework. Our empirical evaluation shows that the attacker can leverage these vulnerable state variables to achieve various RAV failures during real-time operation, and even evade existing defense solutions.
Finally, we propose a framework for producing more robust and compact neural networks to improve practical deployment in embedded/edge CPS. Our approach uses model pruning to decrease model complexity while increasing robustness to attack. Specifically, we look to mitigate membership inference attacks (MIA), which violate privacy by determining a data point's membership in a model's training set. This work provides an entry point for addressing growing privacy concerns in new and emerging CPS applications.Ph.D.Includes bibliographical reference