Love Data Week 2026.

Join us Friday February 13 for in-person events in the Student Union Theatre

Join us February 13, 2026, as we celebrate the University at Buffalo's in-person Love Data Week event. We'll showcase data backed research of undergraduate and graduate students from a diverse academic background. Information on student presentations can be found below. Reserve a spot today! We look forward to seeing you.

2026 Love Data Week Student Presentations

Presenter’s Name: T Maphosa

Major/Department: Biological Sciences

Title: Belonging in Biology: Inclusive Factors on Faculty Webpages

Topic: Biology, Biology Education, STEM Education, STEM Diversity & Equity, Science Belonging

Abstract: Faculty websites are often the first entry point for students seeking research opportunities, yet they vary widely in showing inclusive values. We examine how biology faculty websites at Minority Serving Institutions (MSIs) and non-MSIs include elements that welcome students from marginalized backgrounds. The main focus is the presence and content of inclusivity statements, referencing diversity, equity, and inclusion, provide resources, or support underrepresented students. Using qualitative coding, we analyze websites from a random sample of biology departments, examining inclusive factors such as lab member representation, personal information, and explicit anti-discrimination language. Results show that inclusivity statements remain rare overall, with minimal differences between MSI and non-MSI websites. By raising awareness of the role of faculty webpages in shaping belonging, this project advocates for intentional, equitable, and welcoming online spaces in biology education.

Presenter’s Name/s: Siobhan Lanich

Major/Department: Bachelor of Science in Biological Sciences

Co-Authors: Isabel Porto-Hannes, Corey Krabbenhoft

Title: Lampsilis siliquoidea (freshwater mussel, family Unionidae) placement for reduction of E. coli in Blasdell Creek at Woodlawn Beach State Park

Topic: Ecology & Environment

Abstract: Urban waterways are prone to poor water quality, including high concentrations of the bacteria Escherichia coli, largely due to wastewater effluent and runoff. Freshwater mussels of the family Unionidae have been shown to remove E. coli from water via filter feeding. We assessed the feasibility of using the unionid Lampsilis siliquoidea (Fatmucket) as a biocontrol agent to reduce E. coli in urban streams. Six hundred mussels were placed in containment at two sites in Blasdell Creek (Hamburg, NY), and E. coli was measured at multiple locations upstream and downstream of the array to determine the effect of the placed mussels on E. coli concentrations. While there was no apparent trend in E. coli concentrations attributable to mussels, high survival of mussels (99.3%) indicates promise for this approach. Future research is needed to determine the density of mussels necessary to cause a meaningful reduction of E. coli concentrations in this stream. 

Presenter’s Name: Jasmine Epps

Major/Department: Engineering Education and Industrial Engineering

Title: Engineering Institutional Predictors of Pell Grant Recipients' Student Loan Repayment Rates

Topic:  Learning more about the systemic disparities of student loan repayments (Student Loan Repayment, Pell Grant, Machine Learning)

Abstract: Federal student loan repayment is a key indicator of equity and institutional accountability in U.S. higher education, with persistent disparities between Pell Grant recipients who borrow federal student loans and non-Pell borrowers. Prior research suggests that engineering programs enroll fewer Pell-eligible students than many other fields, reflecting structural barriers in high-cost, credit-intensive STEM pathways.

Using the U.S. Department of Education’s College Scorecard Most Recent Cohorts dataset (2025), this study examines institutional, financial, and demographic predictors of disparities in the federally reported three-year student loan repayment rate for Pell Grant recipient borrowers and non-Pell Grant borrowers. Repayment gaps are defined as the difference between the three-year repayment rates for Pell and non-Pell recipients; negative values indicate lower repayment among Pell recipients. The analysis integrates exploratory data analysis with four modeling approaches, Ordinary Least Squares, LASSO regression, Random Forests, and Gradient Boosting Machines, each evaluated using 10-fold cross-validation.

Presenter’s Names: Srishti Arora, Gianna Minnuto, Sarah Ghadersohi, Isabella Kujawinski

Major/Department: Jacobs Medical School

Topic: This presentation is about how medical students’ preference in residency changes throughout their time in school.

Abstract: Medical school is a period of exploration and growth; however, many students enter training with an initial idea of the residency they wish to pursue. This study tracks student responses from their first year (M1) through their graduating year to assess how residency preferences evolve over time and to identify factors contributing to specialty changes. The data provide a longitudinal view of residency decision-making and highlight trends in specialty interest shifts. These findings can be used to visualize changes in residency preferences across medical school, inform curriculum planning, support institutional mission goals such as increasing primary care representation, and guide students who are considering more competitive specialties.

Student Bios:

  • Srishti Arora is a third year medical student at SUNY Jacobs.
  • Gianna Minnuto is a second year medical student at SUNY Jacobs.
  • Sarah Ghadersohi is a fourth year medical student at SUNY Jacobs.
  • Isabella Kujawinski is a first year medical student at SUNY Jacobs.

Presenter’s Name: Faheem Khawar

Major/Department: Department of Engineering Education

Title: Modeling first-year engineering students’ success: a focus on metacognition skills

Topic: Modeling strategies for student data, multilayered modeling, machine learning.

Abstract: The first year of college is challenging and often involves the development of many skills for learning how to learn. This project explores the role of metacognition skills, the ability to think about your own thinking, in the achievement of first-year engineering students at the University at Buffalo. Established instruments were used to measure metacognition, and their validity was evaluated through Exploratory Factor Analysis; then a variety of machine learning models such as regressions and ensemble trees were explored to evaluate the influence of metacognition skills in the success of students across the semester. The different models are evaluated for performance and usefulness. Analysis showed changes in perceived metacognitive ability in students throughout the semester as well as being important predictor for grades. Controlling variables were used to both isolate effects of key predictors as well as examine disparities between students. 

Presenter’s Names: Madison L. Williams, David Alonso, Joe Binkley, John Hayes, Michael Stebler, Emanuela Gionfriddo

Major/Department: Department of Chemistry

Title: Behind the Paint: Nontarget Perfluoroalkyl Substance Detection in Household Paints using Solid Phase Microextraction and Two-Dimensional Gas Chromatography 

Topic: Nontargeted detection of volatile “forever-chemicals” or perfluorinated alkyl substances (PFAS) in consumer paints using a sample preparation technique called solid phase microextraction (SPME) and comprehensive two-dimensional gas chromatography (GCxGC).

Abstract: Exposure to per- and polyfluoroalkyl substances (PFAS) can occur through multiple indirect and/or direct pathways. Recent work has demonstrated that paint can be a source of exposure to neutral volatile PFAS such as fluorotelomer alcohols (FTOHs), which biotransform into toxic perfluoroalkyl acids. Standardized methods for analyzing volatile PFAS in complex matrices are lacking, and trace-level concentrations coupled with analyte volatility complicate conventional sample preparation.

This work demonstrates a nontargeted workflow using a solid phase microextraction (SPME) arrow coated with hydrophilic–lipophilic balance (HLB)/polydimethylsiloxane (PDMS) to effectively preconcentrate volatile PFAS from paint. Comprehensive two-dimensional gas chromatography (GC×GC) provided the resolving power necessary to separate PFAS from co-eluting matrix components, enabling confident detection and identification. Across five paints, 230–830 analytes were identified, with perfluorinated compounds comprising 0.9–3.0%. This workflow can be readily extended to other consumer products and matrices, enabling comprehensive characterization of volatile PFAS sources.

Presenter’s Name: Rebecca Korsh

Major/Department: Physics

Title: Reprograming the disordered prion-like domain of FUS protein to regulate its pathological aggregation

Topic: Protein aggregation in neurodegeneration

Abstract: FUS protein aggregation, or fiber formation, is associated with neurodegenerative diseases, including ALS and FTD. The intrinsically disordered prion-like domain of FUS (FUSPLD) mediates self-interactions that can drive both solid aggregate formation and phase separation into dynamic condensates essential for cellular processes. The sequence features governing these distinct assemblies, and whether they are coupled, remain unclear. Fiber formation has been linked to short segments, called steric zippers, within the disordered region. Using the ZipperDB database, we identified two alternative sequences with higher predicted zipper propensity than the native FUSPLD zippers. Substituting these variants enhanced fiber formation without significantly affecting phase separation, indicating a decoupling of the underlying sequence determinants. The variants also cross-seeded each other and formed co-aggregates but did not seed native FUSPLD. Together, these results highlight how subtle changes in IDR sequence grammar can shift FUS assemblies from functional condensates toward pathological aggregates.

Presenter’s Names: Saana Mankotia, Maddie Taggart and Dr. Lindsay Hahn

Major/Department: Communication

Title: Revisiting the Suspense Paradox: Does Repeated Viewing Resolve Narrative Uncertainty?

Abstract: Suspense is a popular technique used to engage audiences in narrative entertainment. Given that suspense is characterized by concepts that hinge on audiences’ lack of knowledge about the narrative’s ending, we might expect that knowledge of the narrative’s resolution would minimize feelings of suspense during exposure. Yet despite this expectation, previous work has shown that audiences’ self-reported suspense remains stable on repeated viewings even when they know how the narrative ends (Zillmann, 2013). To understand the cognitive mechanisms behind it (e.g., Carroll, 2013, we investigate audiences' (N=29) responses to repeated viewings of a suspenseful Alfred Hitchcock film. We predicted that participants’ (1) suspense will remain stable after both viewings and, (2) uncertainty will decrease between the first and the second instance of viewing the stimulus. Results of paired-sample t tests revealed that uncertainty decreased from the first (Mt1=5.28, SDt1=0.94) to the second viewing (Mt2=3.86, SDt2=1.28; 5.10, p < .001), as did overall suspense (Mt1=5.05, SDt1=1.30; Mt2=3.61, SDt2=1.75; 4.39, p < .001). These preliminary results do not show evidence of the suspense paradox. Data collection is ongoing, and we are also planning to add measures of heart rate and skin conductance.

Presenter’s Name: Aisha Makama 

Major/Department: Communications

Title: Interaction Between Selfie Exposure and Contingencies of Self-Worth, And Psychological Well-Being: A Policy Proposal for More Transparent Algorithms

Abstract: The rise of social media has shifted individuals' self-presentation to digital form. Studies show that frequent exposure to idealized content, particularly selfies, has an impact on individuals' well-being through social comparison and external validation. This study employs a cross-sectional survey to examine social media behavior, selfie posting, selfie exposure, and their impact on psychological well-being. We aim to examine how different forms of selfie exposure—whether from peers or self-posted content—affect well-being as a function of self-worth. In this study, self-worth is operationalized based on appearance, competition, and social approval. Additionally, this study seeks to determine whether social media engagement moderates these effects, investigating whether higher levels of interaction (e.g., liking, commenting, or receiving engagement) intensify the relationship between selfie exposure and psychological outcomes. These research findings will provide insight into the psychological implications of social media engagement, contributing to broader policy-level discussions on the consequences of digital self-presentation.

Love Data Week

2025 - Inaugural UB event hosted by University Libraries

War of the Viz

All students are invited to participate in War of the Viz. Create a compelling visualization and for a chance to win prizes and have your work featured! Using this dataset from Kaggle*, show us your best data visualization. Submissions will be due by 11:59 pm ET on Monday, February 9, 2026 to oia.lovedataweek@buffalo.edu. An award ceremony will be held on Friday, February 13, 2026 in the Student Union. 

For questions or how to get involved, contact

Jordan Knutsen
jknutsen@buffalo.edu

Love Data Week is brought to you by OIA in collaboration with the University Libraries Research Data Services Department
Office of Institutional Analysis.
University Libraries.