Environmental and Occupational Health Sciences Institute
Rutgers University Community RepositoryNot a member yet
58345 research outputs found
Sort by
Toxic floods: analyzing the preparedness of New Jersey’s contaminated sites for sea level rise & evaluating the use of plant materials to enhance remedial strategies
No full textLike other coastal states, New Jersey is particularly vulnerable to the consequences of climate change. Namely, sea level rise threatens coastal communities and habitat stability. The Garden State however is no stranger to anthropogenic environmental degradation, a history of rampant industrialism and unchecked urban sprawl in the eighteen and nineteenth centuries has created a legacy of contamination that is still being dealt with today. At the intersection of these issues are spaces that desperately require both speedy remediation as well as a strong development towards climate resilience. Considering these compounding factors, this thesis ultimately seeks to determine how many and what kind of contaminated sites in the state have a strong risk of future inundation from sea level rise, the physical ecological qualities of those sites and finally evaluating the plausibility of using plant material as a primary or supplemental method for the remediation and enhanced resiliency of these sites.M.L.A.Includes bibliographical reference
Flipping the switch on opioid use disorder: stress induced vulnerability to fentanyl requires activation of the lateral habenula.
No full textEndogenous opioids are involved in regulating stress responses, mood, pain, and reward. However, chronic activation of the endogenous opioid system following repeated stress or other homeostatic insults is maladaptive and can increase the risk of developing an opioid use disorder (OUD). Nonetheless, this risk affects only a subset of individuals, while others remain resilient. Therefore, our ability to understand the behaviors and brain circuits involved in opioid use vulnerability may be instrumental in preventing OUD, developing novel treatment strategies, and preventing relapse. In order to identify these behavioral features, we developed a comprehensive behavioral paradigm in mice to test for footshock stress induced maladaptive responses that could predict fentanyl preference (Aim 1). After identifying these predictors, we hypothesized that the lateral habenula (LHb), a major aversion center in the brain, was necessary in their development (Aim 2). We tested this through chemogenetic inhibition of the LHb during stress and tested mice using the same behavioral paradigm. In doing so, we observed that individual susceptibility to OUD is highly dependent on stress history. Following stress, we discovered that reward-seeking and heightened mechanical sensitivity are predictive of future fentanyl seeking, while deficits in reward value were more predictive in control mice. Moreover, we found that chemogenetic inhibition of the LHb vglut2 neurons could prevent the stress-induced mechanical sensitivity and could partially reverse stress-induced effects on reward. Furthermore, we observed that LHb inhibition also reduced future fentanyl intake and fentanyl preference. Together we take this to mean that stress acts to trigger individual differences that are predictive of developing a substance use disorder and that LHb activity is necessary for these long-lasting behavioral effects predictive of opioid seeking.Ph.D.Includes bibliographical reference
What makes a teacher? attitudes toward teachers’ language varieties in the Spanish language classroom
No full textThe present dissertation employs a mixed-methods approach to investigate the attitudes of teachers and learners of Spanish toward three ‘native’ varieties of Spanish (Mexican-accented, Peninsular-accented, and Puerto Rican-accented) and a ‘non-native’ variety (English-accented Spanish) in the context of the language classroom. To obtain a robust understanding of participants’ language attitudes, a Verbal Guise Test was used as an indirect method of eliciting language attitudes, while a Belief Questionnaire and semi-structured interviews were employed as direct methods. Quantitative data were analyzed in R, with the Verbal Guise Test analyzed through linear mixed models and the Belief Questionnaire through descriptive statistics. The qualitative data from the interviews were examined via thematic analysis. The Verbal Guise Test revealed that teachers’ attitudes tended to be more positive than learners’ and that ‘native’ varieties of Spanish were evaluated more favorably than the ‘non-native’ variety. Nonetheless, the ratings among the ‘native’ varieties varied, illustrating that each ‘native’ variety occupies a different ideological space in the mental linguistic repertoire of the respondents. Additionally, the Belief Questionnaire revealed that, while participants displayed more positive attitudes toward ‘native’ speakers, they also showed a preference for stigmatized varieties of Spanish. Lastly, the thematic analysis of the qualitative data revealed four major themes: ease of understanding Spanish, ‘native’ speaker-‘non-native’ speaker dichotomy, accentedness in Spanish, and instructors’ bilingualism as a resource.
This dissertation makes several significant contributions to the fields of language attitudes toward varieties of Spanish, as well as Spanish language education. First, it investigates the attitudes of both teachers and learners, thus offering a more holistic understanding of the ways teachers’ and learners’ attitudes interact. Second, along with attitudes toward ‘native’ varieties of Spanish associated with varying degrees of linguistic prestige, this dissertation also investigates attitudes toward a ‘non-native’ variety, namely English-accented Spanish, thus addressing a relatively underexplored area in the literature. Ph.D.Includes bibliographical reference
Dynamics of lateralized electrophysiological and cellular processes in the zebra finch auditory system
No full textSongbirds provide a powerful model for studying adult neuroplasticity in the auditory cortex as a function of recent auditory experience due to many parallels with the human auditory system, which is similarly tasked with processing complex conspecific vocalizations. As in human speech processing, lateralized auditory responses are evident in the songbird’s higher auditory cortex, NCM (caudomedial nidopallium), which encodes stimulus-specific auditory memories through a process of adaptation that leads to reduced responses to familiar sounds. The right NCM typically shows larger auditory responses and adaptation rates than the left for conspecific song, suggesting lateral differences in auditory representations and memory; this pattern of lateralization is known to depend on normal rearing conditions, however the ontogeny or the stability of auditory lateralization in adulthood have not been explored. Furthermore, the songbird brain incorporates new neurons in adulthood, including in NCM. In a series of coordinated experiments, Zebra Finches (ZFs; Taeniopygia guttata) were used to explore the effects of auditory exposure on learning and lateralized NCM responses via electrophysiology, immunohistochemistry, and behavioral assays. 1) we show that adult-like, right-lateralized auditory responses emerge out of left-biased patterns about halfway through development and that auditory experience in development shapes NCM responses in adulthood. 2) we document the time course of changes in lateralized auditory responses in adult ZFs exposed to a foreign, heterospecific (canary) acoustic environment; these changes are characterized as shifts in lateralization, whereby lateralization transiently reverses to a left-biased state, followed by a return to the original right-biased state after prolonged exposure; finches that experienced both the reversal and return to typical right-lateralized patterns of activity exhibited enhanced ability to behaviorally discriminate between test stimuli (canary songs). 3) these dynamic changes in the pattern of lateralized activity are shown to occur successively when ZFs are sequentially exposed to two different heterospecific environments (canary and budgerigar); in addition, we characterize how exposure to these environments leads to learning at the neural level (multiunit and single-unit in NCM). 4) we show that the shifts in lateralized activity manifest at the cellular level and potentially at different loci of the auditory afferent pathway. Furthermore, the exposure paradigm elicits shifts in the lateral distribution of new neurons in NCM, suggesting a possible neural substrate of lateralized neuroplasticity. 5) Finally, we provide evidence for learning in ZFs exposed to a heterospecific environment for different durations, using a novel behavioral method that explores learning in a consequence-free head-turning assay, Together, the results suggest that lateralization represents the current state of an organism, whereby adult-like lateralization is maintained by the current stimulus statistics, and dramatic changes in the stimulus statistics reverts the brain to a learning (developmental-like) state out of which (re)emerges the adult-like state; further, these changes in lateralization states are observed in loci along the afferent auditory pathway and we propose that they are a read-out of neurobiological substrates of learning.Ph.D.Includes bibliographical reference
Promising brighter futures: a mixed-methods analysis of the impact of promise programs on BIPOC student outcomes in California and New Jersey
No full textWith increasing concerns about rising college costs, policymakers have sought to identify ways to make college affordable. College affordability is particularly important to address equity concerns in both higher education and the workforce, as studies consistently demonstrate disparities in outcomes by race/ethnicity and socioeconomic status. One policy intervention to alleviate the financial burden of college for students has gained traction: tuition-free college—commonly known as "promise programs.” Promise programs are now present in all fifty states, each with its own criteria and benefits. Researchers have paid particular attention to two types of promise programs, first-dollar and last-dollar, due to their different structures. Focusing on concerns around racial/ethnic equity in higher education, this study examined how program structure influences enrollment and credential attainment for students who identify as Black, Indigenous, or People of Color (BIPOC). A mixed-methods design was used to determine the impact of two statewide promise programs—the California College Promise (CCP) and the New Jersey Community College Opportunity Grant (CCOG). Semi-structured interviews were conducted with BIPOC community college students in California and New Jersey to set a baseline understanding of how each promise program influences the student experience. Quantitative enrollment and graduation rate data from the federal Integrated Postsecondary Education Data System (IPEDS) was then analyzed to inform the qualitative themes identified from the interviews. Interview data demonstrated that students in California had a better understanding of financial aid, lower levels of financial stress, and higher campus engagement than students in New Jersey. IPEDS graduation rate data showed that while California has a better overall BIPOC graduation rate, the impact of New Jersey’s program was greater. The results are examined through a conceptual framework that pairs critical race theory and student development theory.Ph.D.Includes bibliographical reference
“With heart and head”: solving the “Resettlers Problem” in the Soviet Occupation Zone and German Democratic Republic
No full textThis dissertation examines the expulsion of “ethnic Germans” from parts of Eastern Europe after World War II and their integration in the Soviet Occupation Zone (SBZ) and German Democratic Republic (GDR). The victorious Allied leaders sanctioned the expulsion of approximately 12 million “ethnic Germans” from parts of what are now Poland, Romania, Hungary, the Czech Republic, Russia, and Yugoslavia. The expulsions occurred in various stages from 1945 until the early 1950s. Around 4.3 million of these people landed in the SBZ, constituting a quarter of the population. Despite their German heritage, members of this population, called resettlers in the “East,” were a heterogeneous group, with members speaking different dialectics and observing various cultural celebrations. Their arrival in the SBZ put a strain on a struggling postwar economy but also offered a new labor force and potential political allies. However, the occupying Soviet forces and German socialists realized there was a tension between resettlers and the “natives,” the receiving population, who viewed the newcomers as competitors for resources. This study examines the triangulated relationship between the resettlers, “natives,” and members of the state to illuminate how all three negotiated Germanness in the postwar world. Ultimately, this work challenges the assumption that resettlers were simply pawns of the state rather than active and vocal members of the population. Central to this study are questions of belonging, identity, and homeland. Focusing on resettlers illuminates the connection between material and emotional connections to a place. Moreover, resettlers highlight the ever-shifting nature of national and regional identities and how they interact with legal citizenship. The first three chapters establish the complexity of Germanness before WWII and how the division of Germany into occupied zones brought questions of who and where was German to the fore. Focusing on the Soviet Occupation Zone is an important step to looking at the large degree of improvisation around policies on resettlers. Chapter Three provides a pivot point, as the Soviets and socialists declared the “resettler problem” solved. The final two chapters prove that despite public silencing and the declaration that resettlers were fully integrated, members of the state, resettlers, and natives continued to contend with what it meant to become East German in light of expulsion.Ph.D.Includes bibliographical referencesIncludes vit
Authoritarianism and punitiveness
No full textAuthoritarianism has become an increasingly relevant phenomenon in recent years as many countries have experienced an expansion of authoritarian rule and an increase in democratic backsliding. For these reasons, it is important to further understand authoritarianism and its related phenomenon. In this dissertation, I investigated whether authoritarians (RWA for Republicans, LWA for Democrats) are more punitive towards people from the opposing side of the political aisle, whether this occurs because they engage in vilification of these political opponents, and whether they choose to punish people in ways that infringe on their democratic rights. Study 1 assessed correlational relationships between authoritarianism, political vilification of the opposing political party, support for infringing on democratic rights of the opposing political party, support for partisan violence, and affective polarization. In the Republican sample, most of the effects were related to RWA but were predicted more so by strength of political identity. It is possible that outcomes are more related to partisanship bias than authoritarianism in that sample. However, in the Democrat sample, LWA predicted most of the effects above and beyond strength of political identity. Study 2 used a minor political and criminal transgression to investigate the relationship between authoritarianism, vilification of the transgressor, punitiveness towards the transgressor, and support for infringing on democratic rights because of the transgression. While authoritarianism was an important predictor for most of the outcomes of interest in both Republican and Democrat samples, the transgressor’s political party was also an important predictor for Democrats high on LWA but not so much for Republicans high on RWA. In addition, political vilification was found to mediate the relationship between authoritarianism and punishment in both samples. Study 3 used a minor political, but non-criminal, transgression to investigate the relationship between authoritarianism, vilification of the transgressor, punitiveness towards the transgressor, and support for infringing on democratic rights of the transgressor. Authoritarianism was once again an important predictor for many of the outcomes of interest in both Republican and Democrat samples. In addition, the transgressor's political party was once again an important predictor for Democrats, but this time, it was also an important predictor for Republicans. Political vilification was also found to mediate the relationship between authoritarianism and punishment in the Republican sample. These studies together provide more information about some of the contexts in which authoritarians are more punitive, and about who they are more punitive towards.Ph.D.Includes bibliographical reference
Neural activity in the olfactory bulb evoked by aversive, non-olfactory stimuli and patterned by respiration
No full textSensory circuits in the brain change over time to reflect the organisms learned knowledge about the sensory world, including the incorporation of information about stimulus meaning or importance. For example, when a person or animal model learns that an odor predicts an aversive electric shock, the olfactory system itself undergoes extensive neuroplasticity including in the olfactory sensory neurons in the nose and in the brain’s olfactory bulb. This dissertation reports findings using optical neurophysiology in a mouse model to explore how various populations of neurons in the early olfactory know about aversive, non-olfactory stimuli and how they interact with respiratory activity. These experiments revealed that multiple populations of neurons in the mouse olfactory bulb respond strongly to aversive electrical stimulation on the tail. The earliest and largest responses were observed in populations of periglomerular (PG) interneurons, with later responses in short axon (SA) cells and mitral/tufted cells (M/T), showing that the tailshock-evoked activity propagates through the circuit and evolves over time. Similar results were observed using direct electrical stimulation of the trigeminal nerve, which communicates odor-evoked feelings of pain (e.g. ammonia) and temperature (e.g. warming cinnamon or cooling mint) from the nose to the brain via the trigeminal ganglion rather than the olfactory nerve. This illustrates that the somatosensory-evoked responses in the olfactory bulb can be evoked by facial, cranial nerve-mediated stimuli and non-facial, spinal nerve-mediated stimuli.Aversive stimulation typically evokes a sharp inhalation, even in these sedated mice, and the somatosensory-evoked responses observed in the olfactory bulb were frequently (though not exclusively) time-locked to this initial inhalation. However, the spatiotemporal patterns of these responses differed markedly from odor-evoked responses (including not being confined to specific olfactory bulb glomeruli) and no shock-evoked responses were observed in the population of axon terminals from the olfactory sensory neuron (OSNs) that compose the olfactory nerve. Somatosensory-evoked inhalation of background odors is thus an insufficient mechanism for the observed olfactory bulb response. However, respiratory-linked oscillations are a major feature of neural activity in the olfactory bulb, and all four neuronal populations (OSN, PG, SA, and M/T) exhibited robust oscillatory behavior in the absence of explicit odor presentation. This included prominent respiration-coupled oscillations that were reduced when intranasal airflow was eliminated by shunting the mouse’s breathing through a tracheal tube instead of through the nose, though surprisingly they were rarely eliminated entirely even in the OSNs themselves. Ongoing neural activity in the olfactory bulb is thus normally coupled to the respiratory rhythm through a combination of peripheral airflow and centrifugal signaling. Similarly, we observed that somatosensory-evoked activity in the olfactory bulb was usually hugely reduced in PG, SA, and M/T cell populations but that clear somatosensory-evoked responses sometimes persisted and remained respiration-locked despite the total absence of intranasal airflow. Unilateral naris occlusion also greatly reduced shock-evoked activity in the olfactory bulb ipsilateral to the occlusion, but the effect was similarly incomplete. Taken together these experiments demonstrate that like ongoing oscillatory activity, shock-evoked activity in the bulb may reflect the interplay of a centrifugally-originating signal with peripheral input from the nose.
The convergence of information about odors and aversive somatosensory stimuli like tailshock suggests that odor-specific, learning-induced neuroplasticity could occur via local interactions in the olfactory bulb. These interactions may also play a role in linking olfactory percepts with their chemesthetic features, such as the smell and burn of ammonia.Ph.D.Includes bibliographical referencesIncludes vit
Outlier detection and inference using modern assumption-light methods
No full textOutlier detection is a critical research area in modern statistics, machine learning, and computer science literature. It is not only essential for enabling valid and reliable statistical analyses by identifying and removing outliers from observations in advance, but it also helps uncover unusual patterns and defend against potential threats or attacks. In various applied fields, there is a strong demand for systematic and quantitative methods for outlier detection. Moreover, uncertainty quantification plays a crucial role in both outlier detection and subsequent decision-making processes. Misidentifying an outlier or anomaly can lead to significant costs and resource expenditure for institutions addressing the perceived threat. To mitigate such risks, there is a strong motivation to develop statistical methods that minimize the probability of false discoveries, specifically by controlling type-I errors. A major challenge addressed in this research project is the problem of outlier detection without a predefined reference set, i.e., clean data set without outliers. Conventionally, an outlier is qualitatively defined as a data point (or a set of data points) that lies far from the primary clusters in a dataset. However, the concept of an outlier is inherently vague and depends on the assumption that the main data points (the reference set) are known beforehand. In most practical scenarios, it is challenging to determine whether a data point originates from the "normal" reference distribution or from an outlier distribution. This study aims to address the motivation for defining outliers and developing methods to detect and infer them in the absence of reference sets. To tackle the challenges of outlier detection and inference, this thesis proposes two approaches using modern, assumption-light methodologies. The first part introduces novel theory and methods for high-dimensional linear models with contamination in the noise structure. The second part explores conformal inference for outliers, leveraging the -divergence condition, which corresponds to the non-existence of a finite second moment.Ph.D.Includes bibliographical reference
Explainable CNN-based ADHD detection using EEG data
No full textAttention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental condition marked by persistent symptoms of inattention, hyperactivity, and impulsivity, significantly affecting individuals across all age groups globally. Accurate and timely diagnosis is critical for effective intervention, yet current diagnostic methods often rely on subjective clinical evaluations and behavioral assessments, which can be inconsistent and prone to bias. To address these challenges, this study introduces an innovative data-driven approach for the automated detection of ADHD using Electroencephalography (EEG) data, leveraging Convolutional Neural Network (CNN) models integrated with explainability techniques.The proposed methodology employs advanced preprocessing techniques to extract meaningful features from raw EEG signals, capturing subtle neural activity patterns associated with ADHD. Utilizing a hybrid dataset comprising EEG recordings from both children and adults, the model demonstrates robust performance, achieving an accuracy of 98.91% on unseen test data. These results underscore the model's potential for precise and reliable ADHD detection, offering a significant improvement over traditional diagnostic methods.
To ensure transparency and interpretability in clinical applications, two state-of-the-art explainability techniques—Local Interpretable Model-agnostic Explanations (LIME) and SHAPley Additive Explanations (SHAP)—were employed. LIME approximates the model's behavior for specific data instances, identifying influential features in individual predictions, while SHAP provides a global perspective by quantifying feature importance across the dataset. These techniques validated the relevance of specific EEG channels and features in distinguishing ADHD, revealing critical biomarkers and enhancing model interpretability.
This study establishes a comprehensive framework for automated ADHD detection, integrating deep learning with robust explainability methods to ensure accuracy and transparency. By bridging the gap between advanced machine learning techniques and clinical applicability, this work promotes objective, early, and reliable ADHD diagnosis. Beyond ADHD detection, the framework's adaptability suggests potential extensions to other neurodevelopmental disorders, highlighting its broader implications in AI-driven healthcare solutions.M.S.Includes bibliographical reference