Announcements

Anna Wilcoxen - Dissertation for Ph.D. in Communication Studies
Title: Nowhere People: Working-Class Academics and the Changing Financial Landscape of Higher Education
Major Professor: Sandy Pensoneau-Conway
Committee Members: Rebecca Walker Anderson, Nilanjana Bardhan, Allison Hammer, Jacob Juntunen
Date: October 14, 2024
Location: Virtual on Zoom 
Time: 11:30 am

Abstract: Historically, attending college and pursuing a graduate degree has been associated with greater economic opportunity. However, research reveals that the relationship between higher education and social mobility has shifted over time. The current context of rising student loan debt, the higher cost of education, more tenuous job markets, and stagnating wages diminishes the association between higher education and economic advancement, particularly for students who come from a working-class background. The cost to attend state universities has risen over 200% since the 1980s (Collinge), while graduate assistant salaries and the federal minimum wage remain comparatively unchanged. Additionally, many with a graduate degree are increasingly left to piece together a living wage through multiple adjunct instructor positions or employment in the service industry. While attending graduate school can be seen as a transitional -- and sometimes ambiguous -- space between student and professor, if PhDs are not able to secure full-time, steady employment post-graduation, then their ambiguous existence becomes prolonged, creating both a financial crisis and crisis of identity. In the context of the shift in the relationship between mobility and higher education, my research addresses class as an identity while also accounting for the complicated understandings of that identity during this historical shift. Thus, the dissertation responds to the following research questions: (1) How has the relationship between social mobility and higher education shifted over time? (2) How does this change in relationship systemically and philosophically impact those who earn graduate degrees? and (3) How can academics adapt our pedagogical practices and institutional policies to address this historical shift and its impact on graduates? Accounting for socioeconomic class more thoroughly and in a contemporary context, this research develops a theory of class identity that builds toward social justice praxis at the intersection of socioeconomic class and education, contributing both to the field of communication studies and to broader academic and social spheres. I develop and use collaborative autoethnographic interviewing (CAEI) as a dialogic method by which to gather the stories of working-class academics who have not achieved social mobility through attending graduate school. By seeking participants and gathering stories of graduate degree-holders who have experienced the shift in relationship between higher education and social mobility first-hand, this research provides a better understanding of coalitional opportunities that can be forged between upper, middle, and lower-classes; the educated and uneducated; and those invested in social justice who have not yet had the opportunity to expand their work to include economic equity. 

Spencer Schuchman - Thesis for M.S. in Plant, Soil, and Agricultural Systems
Title: The Role of Methyl Jasmonate in Modulating Trichome Density, Growth, Yield, and Cannabinoid Accumulation in Cannabis sativa L.
Major Professor: Jose Leme
Committee Members: Karla Gage, Alan Walters
Date: October 18, 2024
Location: Agriculture Building, Room 209
Time: 3:00 pm

Abstract: The relaxation of laws surrounding the cultivation of hemp (Cannabis sativa L.) in recent years has sparked revitalization of hemp’s use for fiber, food, medicine, as well as recreational purposes. Although the cultivation of hemp is now legal with proper documentation, growers face challenges associated with decades of prohibition, creating an evident gap in knowledge and need for novel cultivation techniques to improve production standards. A significant determinate of yield quality in high cannabinoid hemp is the density of glandular trichomes and concentration of cannabinoids these structures contain. Techniques that could manipulate these properties would be inherently valuable to cultivation outcomes. A collection of outdoor and indoor studies revealed that the exogenous application of methyl jasmonate (MeJa) yielded significant increases in trichome density. However, counter intuitively, no increases in cannabinoid concentration were observed to accompany this increase in glandular trichome density.

These results contradict similar works where increases in cannabinoid content were observed but not increases in trichome density. This raises the question of how genetic predisposition influences the activity of the plant mechanisms enacted upon by MeJa that ultimately modulate the upregulation of associated genes that determine trichome induction and secondary metabolite synthesis. Unexpected interactions between MeJa and insecticide, as well as ammonium bicarbonate further illuminate the need for further research to determine the genetic factors involved in modulating the plant’s response to MeJa. The current research sheds light of the importance of genotype in concern to hemp cultivation, where cultivar selection strongly determines cultivation outcomes. Although the current research outlines the promising potential of the stimulatory effects of methyl jasmonate, further research should be conducted to reveal the complexities of its effects on important hemp attributes and how its addition into a production system interacts with other important management strategies.

Erin R. Todd - Thesis for M.S. in Forestry
Title: Effects of Fire Management Treatments on Plethodon neomexicanus Aboveground Habitat Valles Caldera National Preserve, N.M.
Major Professor: Eric Holzmueller
Committee Members: Clay Nielsen, Laura Trader
Date: October 25, 2024
Location: Virtual
Time: 11:00 AM

Abstract: The federally endangered Jemez Mountains salamander (Plethodon neomexicanus) is a geographically isolated species with limited dispersal ability found in northcentral New Mexico. Alterations to the historic fire regime of the area have changed the forest structure and vegetative composition of P. neomexicanus habitat, threatening the future survival of this species. This study focuses on the use of prescribed broadcast fire on previously thinned P. neomexicanus habitat in the Valles Caldera National Preserve, New Mexico to restore historic habitat conditions. Results found that fire was highly effective in reducing fuel loads, while forest structure and composition remained essentially unchanged. The volume and count of cover objects were significantly reduced. Analysis of vegetation and substrate burn severity revealed nearly half of the of sampled points were left unburned, indicating that substantial fire refugia was left on the landscape post-fire. Despite minor short term negative impacts to cover object availability, the positive impacts on forest health seem to outweigh the drawbacks. These results provide support for the use of broadcast burning in previously thinned areas, as a highly effective tool in managing fuel load to reduce the risk of stand replacing fire, which is considered the most pressing threat to the persistence of P. neomexicanus. This study concludes that these fire management treatments are compatible with P. neomexicanus habitat management and should be considered as future management option in this area.

Zachary Sapienza - Dissertation for Ph.D. in Mass Communication & Media Arts
Title: Transmission Rituals & Ritual Transmissions: A Reclamation of James Carey's Dichotomy
Major Professor: Walter Metz
Committee Members: Peter Simonson, William Freivogel, Robert Spahr, Anneke Metz
Date: October 25, 2024
Location: via Zoom
Meeting ID: 94237045551   Passcode: 1p7gqY
Time: 12:45 - 2:45 pm

Abstract: In a contrarian rehabilitation project, this dissertation rewrites the history of the transmission model. In grounding the concept in the field of telecommunications, I demonstrate how the transmission model is a conception of communication that has historically been driven by humanity's dream of communicating in different places at the same time. In doing so, I trace its origins to numerous optical-based and messenger-based relay networks including the Great Wall of China, the Persian Empire's Angarium, and the Roman Empire's Cursus Publicus. I contend the transmission model's rise to prominence in Western culture is notably due to the gospel of Jesus and the spread of Christianity.

Sanjana Sakinala - Thesis for M.S. in Computer Science
Title: Integration of Blockchain Technology in Industrial Applications
Major Professor: Henry Hexmoor
Committee Members: Bidyut Gupta, Koushik Sinhai
Date: October 28, 2004
Location: Computer Science Conference Room
Time: 2:15 pm

Abstract: This research explores the integration of Hyperledger Fabric., a framework of
blockchain in industrial applications, particularly focusing on supply chain management.
By addressing the limitations of traditional blockchains - such as high transaction
latency, scalability challenges, and privacy concerns - this research demonstrates how
Hyperledger fabrics overcome them. Through simulation, we show significant
enhancements in transaction speed, data privacy, and scalability, making the Fabric a
viable solution for industrial systems.

Nishanth Sagar Panthangi - Thesis for M.S. in Computer Science
Title: Harnessing Federated Learning for Stock Analysis and Portfolio Optimization
Major Professor: Khaled Ahmed
Committee Members: Bidyut Gupta, Koushik Sinha
Date: October 30, 2024
Location: Computer Science Conference Room
Time: 2:00 pm

Abstract: The field of financial technology has experienced rapid transformation, with
machine learning playing an increasingly critical role in stock analysis and portfolio
optimization. Traditional stock analysis techniques often face challenges related to data
privacy, as pooling data from multiple institutions or traders raises significant
confidentiality concerns. Federated Learning (FL) offers a promising solution by
enabling institutions to collaboratively train a global model without sharing sensitive raw
data. Rather than moving data to the model, FL takes the model to the data, ensuring
data remains securely within each owning entity. This thesis explores the application of
Federated Learning in stock analysis, focusing specifically on optimizing portfolio
strategies through client-specific trading data.
The proposed approach adopts a federated learning framework that allows individual
entities to develop independent models based on their trading strategies while
safeguarding data privacy. By combining insights from these diverse strategies, the
federated system builds a global model capable of supporting informed portfolio
decisions across different market conditions. This collaborative architecture enables the
global model to benefit from a wide range of risk-return perspectives, enhancing its
adaptability and predictive accuracy. The evaluation of this approach demonstrates that
federated learning can achieve robust, privacy-preserving performance improvements,
even in the face of market volatility. Notably, the global model achieved a 25.08%
Compounded Annual Growth Rate (CAGR) and a Sharpe ratio of 4.46, significantly
outperforming individual client models, which had Sharpe ratios ranging from 1.75 to
2.67 and CAGR values ranging from 12.5% to 18.3%. This study reveals the potential
for federated learning to drive innovation in financial modeling, allowing institutions or
individual traders to achieve shared insights without compromising data security. Such
advancements highlight federated learning’s relevance in meeting the evolving
demands of privacy-conscious financial technology.
This study highlights federated learning’s potential to deliver scalable and
privacy-focused solutions for collaborative stock analysis and portfolio optimization. By
bridging advanced predictive capabilities with privacy standards, this framework
exemplifies how federated learning can transform financial decision-making, yielding
higher risk-adjusted returns. The federated model's superior Sharpe ratio and reduced
maximum drawdown illustrate its effectiveness in optimizing trader outcomes. The
implementation involved each participating trader training their local model using their
proprietary trading data, which was kept private throughout the process. These local
models were then aggregated into a global model using a federated averaging
technique. This collaborative process enabled the global model to harness diverse
trading strategies, resulting in more robust predictions and improved performance
metrics, such as risk-adjusted returns and Sharpe ratios, across different market
conditions. Such a framework not only enhanced the profitability for individual traders
but also demonstrated the scalability and adaptability needed to cater to varying
financial environments, paving the way for broader applications in privacy-conscious
industries.

Ali Shahidy - Thesis for M.A. in Psychology
Title: The Role of Grammatical Gender and Religiosity in Shaping Implicit Gender Attitudes: An Investigation into Pashto and Dari Languages
Major Professor: Usha Lakshmanan
Committee Members: Reza Habib, You-Jung Choi
Date: November 1, 2024
Location: Life Science II, Room 295
Time: 10:00 am - 12:00 pm

Abstract: This thesis contributes to the ongoing debate on the relationship between language and cognition by evaluating the impact of language-specific features, such as grammatical gender, on non-linguistic cognitive processes like implicit gender attitudes. Grammatical gender is a formal linguistic feature that categorizes nouns into classes such as masculine and feminine (with some languages, like Russian, having more than two gender categories) based on linguistic rules, including morphology and phonology. Previous studies have highlighted the influence of grammatical gender on various cognitive abilities, such as object categorization, semantic categorization, or gender perception. However, little attention has been given to the effect of grammatical gender on sexism and gender bias, with existing studies relying solely on explicit measures (e.g., Wasserman & Weseley, 2009). This thesis addresses this gap by examining the role of grammatical gender in implicit gender bias in two languages spoken in Afghanistan: Dari, a genderless language, and Pashto, a language with grammatical gender. The study employed the Gender-Career version of the Implicit Association Test (IAT) (Greenwald et al., 1988), which measures implicit biases by assessing the strength of associations between target concepts and attributes based on response times. It was predicted that, after controlling for age, sex, and religiosity, L1 gender-type (Dari vs. Pashto) would significantly explain the variation in gender bias, with Pashto speakers exhibiting higher implicit gender bias than Dari speakers.

To conduct this study, two online experiments were designed on Qualtrics: one in Dari and one in Pashto. Each experiment was comprised of a Consent Form, a Participant Background Questionnaire with items related to participants’ linguistic background, the Muslim Religiosity Scale, the IAT, and a debriefing statement. A total of 96 Afghan participants were recruited through flyers distributed on social media platforms—57 completed the Dari experiment, and 39 completed the Pashto experiment. A hierarchical regression analysis was conducted to examine the effect of Age, Sex, Religiosity, and L1-Gender type on participants implicit gender bias measured by the IAT. Model 1 of the analysis comprising Age and Sex revealed that Sex, but not Age, was the only significant predictor, with male participants showing lower IAT scores compared to female participants. Model 2, which added Religiosity, showed that Religiosity was not a significant predictor of the IAT scores while Sex remained a significant predictor. Finally, Model 3 which introduced the primary independent variable, L1-Gender type, revealed that L1-Gender type emerged as the only significant predictor of variance in IAT scores with Pashto speakers exhibiting lower IAT scores compared to Dari speakers while other predictor variables (Age, Sex, and Religiosity) remained non-significant. Contrary to the study predictions, the results revealed that the Pashto language, a language with grammatical gender, leads to less gender bias compared to Dari, a language without a grammatical gender system. These findings challenge the current assumptions about the role of grammatical gender on gender attitudes and biases.

 Maggie Ratcliff - Thesis for M.S. in Behavior Analysis and Therapy
Title: A Focus on the Social Validity of Treating Stereotypy Using Response Interruption and Redirection
Major Professor: Lesley Shawler
Committee Members: Ryan Redner, Paige Boydston
Date: November 1, 2024
Location: Virtual
Time: 2:00 - 4:00 pm

Kennedy Cloe - Thesis for M.S. in Behavior Analysis and Therapy
Title: Comparing Procedures within Matrix Training: A Systematic Replication
Major Professor: Lesley Shawler
Committee Members: Ryan Redner, Eric Jacobs
Date: November 1, 2024
Location: Life Science II Building
Time: 3:15 - 5:15 pm

Shu Chen - Dissertation for Ph.D. in Environmental Resources & Policy
Title: A Comprehensive Assessment and Analysis of City Dynamic Evolution and Driving Factors for Detroit
Major Professor: Guangxing Wang
Committee Members: John Groninger, David J Gibson, Jonathan WF Remo, Ruopu Li
Date: November 4, 2024
Location: Virtual https://zoom.us/j/5992881679
Time: 9:00 am

Abstract: In this study, a methodological framework was proposed based on multi-source remote sensed images and social-economic data by developing a novel mixed training sample-based spectral unmixing analysis method and two new comprehensive assessment indices. The former was used to detect city land use and land cover (LULC) changes, while the latter is utilized to monitor the dynamics of city social-economic aspects. A mixed training sample-based Convolutional Neural Networks (CNN) model was proposed for spectral unmixing analysis. Two comprehensive assessment indices consisted of an urban decline or decay index (UDI) and an image-derived comprehensive assessment index. The mixed training sample-based spectral unmixing analysis method was compared with traditional pure training sample-based spectral unmixing method through four models including multiple linear regression (MLR), random forest (RF), artificial neural network (ANN) and CNN. Moreover, the mixed training sample-based CNN spectral unmixing was integrated with the classification rule based on maximum fractional abundance, which led to a CNN-MMA classification method. The CNN-MMA classification was then compared with traditional maximum likelihood (ML) classification, RF classification, ANN classification, and an integration of pure training sample-based CNN spectral unmixing and the classification rule based on maximum fractional abundance (CNN-PMA) to detect the changes of LULC types. The methodologic framework was examined to detect and monitor Detroit decline from 1985 to 2020 using social-economic data including population, unemployment rate, poverty rate and per capital income (PCI), and multi-source remotely sensed images (Landsat, Sentinel, aerial photographs, etc.), including 400 subjectively selected pure training samples, 400 randomly drawn mixed training samples and 400 validation samples. Several conclusions were drawn.

Raymond Bieri - Thesis in Forestry
Title: Community Response to Floods in the Cache River Watershed of Southern Illinois: An Application of a Proposed Framework for Assessing General and Specified Community Resilience
Major Professor: Kofi Akamani
Committee Members: Logan Park, Leslie Duram
Date: November 4, 2024
Location: Agriculture Building, Room 209
Time: 10:00 am

Abstract: Recent years have seen increased awareness about the potential adverse impacts of climate change on hydrologic systems and human communities. The concept of community resilience – the capacity of communities to respond to drivers of change in a manner that maintains or improves upon community well-being – is widely recognized as a suitable framework for informing climate change adaptation policies. Emerging insights from the community resilience literature highlight the importance of the distinction between general resilience, which refers to the resilience of social-ecological systems to all kinds of drivers of change, and specified resilience, which is concerned with the resilience of social-ecological systems to particular drivers of change. While capital assets and institutions are generally considered to be the contextual factors shaping general resilience, the roles of these contextual factors in shaping specified resilience have not been adequately explored. To fill this knowledge gap, we developed a synthesized conceptual model that describes how community contextual factors influence the process and outcomes of community responses to specific drivers of change. This qualitative study was designed to test the proposed conceptual model by analyzing the responses of two rural communities (Belknap and Tamms) to floods in particular, and climate change impacts in general in the Cache River watershed of Southern Illinois. Both communities selected for the study have historical experience with flooding and consist of different population sizes. Hence, they are likely to have different levels of endowment of the assets and institutions that shape community resilience. Data for the study were collected through document review, as well as semi-structured interviews with 23 purposively sampled key informants representing various sectors of the local society (e.g. commerce, education, religion, local government). The data were analyzed using deductive coding in NVivo software based on analytical constructs derived from the proposed conceptual model. The results showed that communities in the Cache River watershed draw from a network of formal and informal institutions at multiple levels (local, state, national) in their responses to floods, although responses to other climate change impacts are less noticeable. Community capacity to respond to these impacts are also enabled by the availability of high levels of some capital assets (such as social capital), but also constrained by low levels of other community capital assets (such as economic capital and physical capital). Community assets and institutions also seemed to shape community perceptions and responses to future climate change scenarios where community relocation may be a viable response option. In all, our results provide rich insights on how community capital assets and institutions interact to either constrain or enhance the awareness, motivation, capacity, and opportunities that influence community responses to current and future drivers of change. Future research should seek to further refine and test this and related frameworks, as well as develop relevant indicators for further understanding the general and specified resilience of communities. With regard to policy, the results highlight the need for policies on climate change impacts and other disasters to move towards more targeted approaches to prepare communities to deal with specific threats in addition to those policies that focus on the general well-being and capacity of communities to deal with all kinds of drivers of change. Building the specified resilience of communities will require the utilization of existing scientific and non-scientific knowledge on the types of institutions and assets that are critical to community responses to the driver of change under consideration in a given context.

Erdem Anil - Thesis for M.S. in Agricultural Sciences - Plant Genetics
Title: Characterization of Raffinose and Stachyose Synthase Genes by Using TILLING by Sequencing+
Major Professor: Khalid Meksem
Committee Members: Karla Gage, Alan S. Walters, Amer AbuGhazaleh
Date: November 7, 2024
Time: 12:15 - 1:15 pm

Abstract: Soybean (Glycine max) is one of the most widely grown crops in the world. Soybeans are used as food and feed products due to their high protein, oil, and carbohydrate seed contents. Soybeans are one of the most important protein sources presenting several benefits for human and animal consumption. Moreover, the carbohydrate content of soybean is important since it contributes to the taste of soyfood such as tofu, soymilk, natto, and livestock. Sucrose is the predominant carbohydrate found in soybeans which is followed by other raffinose family oligosaccharides (RFOs) such as raffinose and stachyose. Higher sucrose contents contribute to improving the digestibility of products. However, higher raffinose and stachyose contents make soymeal indigestible and can cause undesirable flatulence in humans and animals. When these RFOs are consumed by animals they cannot be hydrolyzed in monogastric animals due to the absence of alpha-galactosidase activity in their body. Therefore, it would be desirable to reduce these polysaccharides in seeds for easier consumption, while also investigating the impact of oligosaccharide content changes on seed germination and vigor. TILLING has proven to be effective on different agronomic crops such as soybean, maize, rice, wheat, etc. TILLING mutants carrying mutations at the raffinose and stachyose synthase genes may reduce their activity which may impact sucrose content in the seeds without causing unwanted poor agronomic features. Hence, developing high sucrose lines with reduced RFOs for commercial usage will be useful for developing better cross lines for the future and will be healthier and desirable for human and animal consumption. In the current study, we developed an EMS soybean mutagenized population and then TILLING by sequencing was used to study the function of previously discovered raffinose synthase and stachyose synthase genes in the isolated soybean mutants.

Dounya Knizia - Dissertation for Ph.D. in Agricultural Sciences
Title: Improving Soybean Seed Sugar Content and Deciphering the Sugar Biosynthesis Pathway
Major Professor: Khalid Meksem
Committee Members: My Abdelmajid Kassem, Amer Abughazaleh, Alan S. Walters, Karla Gage
Date: November 7, 2024
Location: Room 209
Time: 3:00 pm

Jerome Specht - Dissertation for Ph.D. in Geosciences
Title: Acid Alternations of Clay Minerals and Implications for Martian Surface Alteration Processes
Major Professor: Liliana Lefticariu
Committee Members:  Tanya Peretyazhki (NASA), Daniel Hummer, Derek Gibson, Martin Pentrak (IL State Geological Survey), Vincent Chevrier (Arkansas Center for Space and Planetary Sciences)
Date: November 21, 2024
Location: Parkinson Conference Room and Virtual on Zoom
Zoom Link: https://zoom.us/j/91569778828
Time: 12:00 - 2:00 pm

Stellah Nakiranda - Research Paper for M.S. in Professional Media & Media Management Studies
Title: How Do Southern Illinois University Students Perceive TikTok's Influence on Their Academic Success?
Major Professor: Cinzia Padovani
Committee Members: Kavita Karan, Bridget Lescelius
Date: November 12, 2024
Location: Communication Building, room 1032
Time: 9:30 am