Undergraduate Research and Creative Activities Journal

Language is tied directly to identity formation, especially in the way individuals express themselves and are perceived by others. This extends to language being proven to have the ability to change an individual’s personality depending on the language they are speaking (Ramirez-Esparza et al. 2006; Pavlenko, 2006; Wedérus, 2017). This study expands on these findings by comparing these theories to the experiences of culturally and linguistically diverse multilinguals. This study is particularly concerned with the experiences of bilingual immigrants in the US and how issues of acculturation and xenophobia may affect these phenomena. This project aims to help better understand how language shapes identity, as well as gain insight into the challenges of immigration and multilingualism.

Recent advances in Deep Neural Networks (DNN's) over the last decade have allowed modern neural networks to be reliably deployed "on the edge" in countless applications ranging from computer vision to natural language processing. Existing hardware is capable of running complex models with low latency, but a problem occurs when applications are scaled to require cheaper hardware with shallower memory resources or minimal latency. The goal of model compression is to take popular pre-trained deep neural networks and reduce their size to allow them to be readily deployed in areas requiring "on-device" inference such as self-driving vehicles and A.I. assistants. This paper covers recent advances in the field of model compression that has allowed us to create a 100x smaller model in terms of memory storage, while maintaining stable F1, Precision and Recall scores.

The COVID-19 pandemic shed light on many of the inequalities people face because of their socioeconomic status. Some of these disparities include accessible health care, medicine, consistent income, and the everyday risk of exposure. However, this division between the upper and lower socioeconomic classes is not new. This can be seen time and time again throughout history, especially in the cases of past pandemics in early modern Europe. Diseases such as the Black Plague, leprosy, and cholera infiltrated and affected communities throughout Europe and Italy, the latter being the focus of this article. In it, we compare the ways that past pandemics affected people in cities like Rome and Venice, and who was affected most and why. Did the same disparities between socioeconomic classes exist as they do today with the current COVID-19 pandemic? Have we progressed much as a society? It is important to explore how these socioeconomic differences emphasize the inequalities within society when it comes to basic human needs and rights, like safety and public health, and how they track from centuries ago to the present day.

A first-generation college student is the first of their family to enter higher education (RTI International, 2019). This study examined whether first-generation college students adopt avoidance goals (i.e., goals focused on avoiding negative outcomes) more often than continuing-generation college students, and how these goals may impact their campus resource utilization. We hypothesized that first-generation college students at UCSB would report less resource utilization compared to continuing-generation students, and that this association is mediated by the strength of their approach goal orientation. For first-generation college students, we expected higher endorsement of avoidance goals and lower endorsement of approach goals, compared to continuing-generation college students. Lastly, we hypothesized that for those with higher agreement towards avoidance goals and lower agreement towards approach goals, these students would utilize fewer campus resources. In our study, generational status was associated with goal orientation and campus resource utilization. However, generational status and goal orientation together did not appear to predict campus resource utilization. Additionally, first-generation college students used significantly more academic resources compared to continuing-generation college students. Future studies should continue to investigate the relationship between generational status, goal orientation, and resource utilization to develop more insight on how to better support minoritized communities.

Drosophila melanogaster, commonly known as the “vinegar fly”, is a model organism for studying olfaction-induced behavioral activity. The behavior of positive chemotaxis or attraction from the activation of odorant receptors such as Or42a are well characterized through extensive prior research. However, the behavior from the activation of aversive odorant receptors like Or49a are not well understood. To characterize aversion and to test whether aversion and attraction have equal and opposite behaviors, I utilized the PiVR tracking system to simulate several odor conditions by applying light gradients on optogenetically modified third-instar larva. I have concluded that the characteristics of aversive behavior are not directly opposing the characteristics of attractive behavior through the analysis and comparison of turn rate modulation, trajectories, and preference indexes between Or42a and Or49a light-stimulated larva.

Over 100 million people worldwide use oral hormonal contraception (OC), and yet there is still little knowledge surrounding the consequences of contraceptives on the human brain. In particular, the intersection of autobiographical memories, stress, and OC is important to study for real-world applicability. Previous research has shown that women on OC demonstrate a negativity bias when recalling an event by reporting more information for negative experiences compared to other emotional situations. This negativity bias could be an indication of the ease of recall which is defined by the speed, accuracy, and intensity of the memory search. The present study examined the reported ease of autobiographical recall of OC users compared to those who are naturally cycling (NC) for negative and neutral events. It was predicted that those on OC would report an easier time recalling a stressful event and report less difficulty for the neutral event compared to NC women. There were no significant differences between those who are NC and on OC in terms of ease of recall ratings for either event. However, within NC women rated remembering negative events more easily than neutral. Suggestions for future studies are discussed.

Teenage depression has long been a significant yet underreported and therefore undertreated disease. In recent years depression-centric narratives like The Perks of Being a Wallflower, Dear Evan Hansen, and All the Bright Places have garnered both attention and controversy in print, film, and even Broadway, bringing the conversation around teenage mental health into the spotlight. While some of these depictions have been praised for promoting empathy and understanding, others have been criticized for including graphic or even sensationalized representations of teen depression and suicide. These depictions, fictional as they may be, contribute to the larger societal discourse on teen mental health. This paper examines one of the most influential works concerning adolescent mental health in recent years: Netflix’s Thirteen Reasons Why, with the aim of exploring how its status of both an “activist” work of art and a product affect its depiction of depression and anxiety — and how these depictions might affect a population that, in the wake of the pandemic, reports higher rates of depression and anxiety than ever.

Given most people gain information about COVID-19 from news media, it is important to understand if news framing of COVID-19 can influence people’s intentions to take preventive behaviors and their actual behaviors, which may affect public health and many people’s life safety. Based on framing and functional emotion theory, this research examined how exposure to differently framed news (threat, positive future, or neutral news) influenced emotional reactions, intentions to take preventive action, and actual subsequent protective behaviors. 196 undergraduates participated in a two-part online experiment. First, they read two COVID-related news stories appropriate to their condition, and then reported their emotional reactions and behavioral intentions. Two days later, they reported their COVID-related protective behaviors. Results indicated that threat news in the frame of fear evoked fear as expected and positive future frames in the frame of hope evoked hope, as expected. Although frames did not directly influence intentions or behaviors, indirect effects were found such that threat frames generated fear, fear influenced behavioral intentions, and intentions predicted behavior 2 days later.

The Klamath river—the second longest river in California, stretching 257 miles from South Crescent City to Oregon—has been an object of environmentalist and humanitarian concern since the 1970s. And it was long before the acknowledgement of the Klamath’s worsening state that climate change, along with anthropogenic factors such as dam implementation, agricultural runoff, commodified farming, and racist governmental policy, had begun to irreversibly damage the once flowing water supply and diverse flora and fauna that used to characterize the Klamath area. These long-standing issues have culminated in mass environmental degradation in the Klamath basin—drought, poisonous algae blooms, mass fish kills, pollution—that threaten the Klamath ecosystem at large. Indigenous tribes like the Yurok people, who have lived in the Klamath area for decades, have been disproportionately negatively affected by the environmental degradation of the Klamath.

We have been presented the properties of asset return by simulation within the empirical data. However, is it possible to illustrate properties by statistical analysis? Most currently existing models fail to reproduce all these statistical features. In this paper, we will elaborate the properties by applying different statistical models: Geometric Brownian Motion and Ehrenfest URN. We will focus on the following properties: distributional properties, tail properties and extreme fluctuations, path-wise regularity, linear and nonlinear dependence of returns in time and across stocks. In this project, I will use S&P 500 index return as the data and apply it with the models to compare the results with empirical data.