Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 30 of 63
Full-Text Articles in Digital Humanities
Language And Polarization: A Psycholinguistic Analysis Of Political Communication During Trump's Second Term, Andrew Vera
Language And Polarization: A Psycholinguistic Analysis Of Political Communication During Trump's Second Term, Andrew Vera
IPHS 300: AI for Humanity
This project uses psycholinguistics—the systematic measurement of psychological orientation through language —to examine political communication from when Trump took office in his second presidency to the start of the Iran war—January 20, 2025 through February 28, 2026. I asked whether elite political language during this period reflects what Robert Putnam calls bonding social capital (inter -group reinforcement) or bridging social capital (cross-group connection), a dichotomy relevant to theory of democratic depolarization. To answer that, I scraped 16,707 documents —7,409 of Trump's Truth Social posts and 9,298 White House, State Department, and Department of Defense press releases —an d scored each …
The Digital Supreme Court: A Meta-Analysis Of Antitrust Retrospectives Using Frontier Llms, Pablo Portillo
The Digital Supreme Court: A Meta-Analysis Of Antitrust Retrospectives Using Frontier Llms, Pablo Portillo
IPHS 300: AI for Humanity
This project examines whether frontier Large Language Models (LLMs) can serve as structured “expert critics” in economic and legal reasoning. I built a panel of AI judges composed of GPT-5.5, Opus 4.7, and DeepSeek V4, to evaluate three antitrust merger retrospectives: Carlton et al. (2022) on AT&T/Time Warner, Hazlett and Crandall (2024) on T-Mobile/Sprint, and Argentesi et al. (2021) on Facebook/Instagram plus Google/Waze. Using a 10-factor weighted rubric, the panel scored each paper on empirical alignment, parameter robustness, counterfactual validity, dynamic adaptability, and analytical rigor. I then synthesized the nine evaluations as the “Highest Judge,” comparing where the models converged, …
Confident And Wrong : How Llm Miscalibration Mirrors The Dunning-Kruger Effect Across Ambiguity Levels, Maggie Grespin
Confident And Wrong : How Llm Miscalibration Mirrors The Dunning-Kruger Effect Across Ambiguity Levels, Maggie Grespin
IPHS 300: AI for Humanity
Psychology gave us the Dunning-Kruger effect, the observed pattern that a person is often overconfident when they fall in the lowest quartile of that skill’s proficiency. This research is interested in seeing if large language models (LLMs) suffer the same miscalibration of metacognitive skills. Using thirty deliberately ambiguous college-level math question prompts, this research assesses Claude Sonnet 4.6’s ability to interpret each question, provide an answer, and rate its own confidence without asking for clarification. Results show that while confidence appropriately drops at the moderate tier, it rebounds on the hardest questions (0.840) even as accuracy falls to 0.450. This …
Whose Vietnam War? How Llms Answer Contested History In Three Languages, Melissa Vu
Whose Vietnam War? How Llms Answer Contested History In Three Languages, Melissa Vu
IPHS 300: AI for Humanity
As a Vietnamese international student studying in the US, I grew up learning one version of the Vietnam War -or, as we call it in Vietnamese, the American War -in Vietnamese history classes, only to encounter different perspectives in American college classrooms. Taking Chinese language courses and spending a semester abroad in China sharpened this further as the same historical events looked different again through Chinese sources. These experiences made me aware of how language shapes the way information and historical knowledge get presented and understood. Since LLMs are trained on data reflecting many cultures and languages, I wanted to …
Know Thyself: A Conversation That Reveals How You Think: Adaptive Ai Assessment Across Eight Psychometrically Grounded Metrics, Adam Leider
IPHS 300: AI for Humanity
AI's ability to assess human cognitive and psychological patterns through natural conversation is an underexplored area with significant potential, particularly for college students navigating a demanding and overwhelming transitional period. This project involved selecting eight psychometrically grounded metrics, designing an adaptive question bank and detailed system prompt to govern the assessment conversation, building a second AI model to review and correct results before they are shown, and deploying the full application as a publicly accessible web tool. The result is a conversational assessment that delivers a personalized psychological profile, including strengths, areas for growth, metric interactions, and actionable suggestions in …
Face-Work In Frontier Models: Benchmarking Sycophantic Behavior In Llm-Based Educational Tutoring, Nick Chen
Face-Work In Frontier Models: Benchmarking Sycophantic Behavior In Llm-Based Educational Tutoring, Nick Chen
IPHS 300: AI for Humanity
Since OpenAI’s famous rollback of GPT-4o for being “overflattering” in April 2025, sycophancy in large language models, which reflects the tendency to prioritize user satisfaction over truthfulness, has become a major concern in AI alignment. Sycophancy is widely recognized as a byproduct of Reinforcement Learning from Human Feedback (RLHF) and deserves attention in educational contexts where such responses to student errors will reinforce misconceptions. This project builds on the social sycophancy framework introduced by Cheng et al. (2025) in ELEPHANT, which applies sociologist Goffman's classical theory of face work to LLM behavior, extends it to educational settings, and raises the …
Whose Ethics? Auditing The Moral Defaults Of Large Language Models Against Ubuntu Philosophy, Davelle Akosua Ampofo-Twumasi
Whose Ethics? Auditing The Moral Defaults Of Large Language Models Against Ubuntu Philosophy, Davelle Akosua Ampofo-Twumasi
IPHS 300: AI for Humanity
As LLMs are deployed globally, the moral framework they u se matters. This study tested if four frontier LLMs default to Western ethics and how naming Ubuntu, a communal Southern African philosophy, changes their moral reasoning. Five dilemmas were presented to Claude Opus 4.7, GPT-5.4, Gemini 2.5 Flash, and DeepSeek V3, unprompted and with Ubuntu named. Responses were scored on seven Ubuntu dimensions (e.g., relational inclusion, dignity, reciprocity). All models defaulted to Western individualist reasoning when unprompted. With Ubuntu named, DeepSeek (57/70) and Claude (56/70) significantly outperformed GPT and Gemini (32/70 each). Gaps were largest in reciprocity and relational inclusion. …
Ai Mcat Tutor: Personalized Study Planning, Daily Check-Ins, & Slideshow Generator For Pre-Medical Students, Luke Latterman
Ai Mcat Tutor: Personalized Study Planning, Daily Check-Ins, & Slideshow Generator For Pre-Medical Students, Luke Latterman
IPHS 300: AI for Humanity
Preparing for the Medical College Admission Test (MCAT) is among the most demanding academic challenges facing pre-medical students, yet access to professional tutoring remains prohibitively expensive for many. This paper describes the design, implementation, and iterative evaluation of an AI-powered tutoring system built using Claude Code, Anthropic’s command-line interface for the Claude large language model, intended to serve pre-medical students directly, whether or not they have access to a professional tutor. The system integrates three components: a personalized multi-phase study schedule generator, a daily check-in assistant, and an AI-driven instructional slideshow generator for targeted content review. Each component was designed …
Translating Economics For Stem Minds: An Ai Socratic Tutor That Speaks Their Language, Andre Mccloud
Translating Economics For Stem Minds: An Ai Socratic Tutor That Speaks Their Language, Andre Mccloud
IPHS 300: AI for Humanity
STEM students frequently struggle with economic thinking due to the particular framework of solving problems they acquire from their respective academic disciplines in undergraduate programs. This AI Socratic tutoring system is designed to detect and address discipline-specific misconceptions and gaps frequently holding students back in economics. Using Google’s Gemini Flash model, this tutor can give students hints and feedback and even reroute students thinking when they are confused. It does so by providing manageable explanations tailored to the individual student’s background across STEM fields like mathematics, physics, engineering, and other STEM fields. This system utilizes a customized Q&A bank pre-loaded …
Changing Narratives Of Self-Improvement: Self-Help Literature Across Two Eras, Ayesha Aslam
Changing Narratives Of Self-Improvement: Self-Help Literature Across Two Eras, Ayesha Aslam
IPHS 300: AI for Humanity
This project explores how the narratives of personal growth in self-help literature have evolved between the 1970s and the 2010s. The goal was to identify which themes persisted across eras and how others shifted in response to broader cultural and technological changes. I hypothesized that 1970s texts would emphasize self-control and traditional success, while 2010s texts would highlight flexibility, mental well-being, and authenticity, with some shared focus on motivation and habit-building. To test this, I selected five bestselling books from each decade and applied Latent Dirichlet Allocation (LDA) topic modeling using a Colab-based notebook. The top 30 most salient terms …
Ai's Creative Boundaries: A Cross-Model Pattern Analysis Of Identity-Based Narratives, Maisie Jane Brigham
Ai's Creative Boundaries: A Cross-Model Pattern Analysis Of Identity-Based Narratives, Maisie Jane Brigham
IPHS 300: AI for Humanity
This research project examines creative writing outputs from five leading large language models (Claude 3.7 Sonnet, ChatGPT 4o, Grok 3, Gemini 2.5 Flash, and Deepseek R1) as they generate and respond to prompts focused on diverse identity experiences. By analyzing stories centered on immigrant, Black, LGBTQ+, transgender, and Indigenous experiences, I identify recurring patterns, tropes, and limitations across different AI systems. This study illuminates how algorithmic storytelling currently relies on a limited repertoire of narrative elements, raising important questions about the representation of diverse human experiences in AI-generated creative content and the potential impacts of these patterns as AI writing …
Biolinkbert-Pr: A Large Language Model For Diagnosing Pediatric Rheumatological Disorders, Juliette Lowe
Biolinkbert-Pr: A Large Language Model For Diagnosing Pediatric Rheumatological Disorders, Juliette Lowe
IPHS 300: AI for Humanity
Pediatric rheumatology is an understudied and underprovided field in medicine. Currently, there are only 850 pediatric rheumatologists in the United States (1). This can make it incredibly difficult to get a diagnosis, since pediatric rheumatologic disorders are rare and thus easily missed. For example, eight states have no pediatric rheumatologists at all, and only 25% of children with arthritis are able to see a rheumatologist (2). AI applications in pediatric rheumatology are lacking, likely due to the combination of a lack of interest and a lack of data. This project utilizes a large language model (LLM) to create an AI …
Can Advanced Chatbots Help Us Navigate Educational Advocacy? Understanding The Potential Of Large Language Models As Assistants And Guides Within An Iep (Individualized Education Program) Meeting Context, Nava Bahrampour
IPHS 300: AI for Humanity
Hundreds of thousands of parents and students in New York City’s school district alone (Fancsali and Farley 2018) go through the Individualized Education Program (IEP) process every year, attending meetings and evaluations in order to determine adequate services and students with disabilities’ educational trajectories. However, most parents, especially those new to public education policy, report difficulty navigating the system or acquiring the resources necessary to successfully advocate for their children at IEP meetings (Advocates for Children 2025, Kurth et al. 2020). The rise of generative artificial intelligence and increased linguistic capabilities of interactional chatbots provokes questions about such technologies’ ability …
Following The Crowd? A Topic Modeling Analysis Of Twitter Discourse And Supreme Court Decisions, Jessica Daughterty
Following The Crowd? A Topic Modeling Analysis Of Twitter Discourse And Supreme Court Decisions, Jessica Daughterty
IPHS 300: AI for Humanity
This project uses topic modeling to compare the top topics from Twitter posts with the text of the Supreme Court’s Dobbs v. Jackson ruling, which overturned Roe v. Wade in 2022. I analyzed whether the Court’s reasoning reflected public concerns or was disconnected. While some analysis of the Twitter data exists on the Kaggle dataset I used, I conducted this research independently of those results. By putting the Court’s language and public sentiment side by side, this project explores whether institutions are responding to what people care about.
Exploring The Potential Of Ai For Swim Technique Evaluation And Athlete-Centered Coaching, Gwen Eisenbeis
Exploring The Potential Of Ai For Swim Technique Evaluation And Athlete-Centered Coaching, Gwen Eisenbeis
IPHS 300: AI for Humanity
This project explores the use of AI-powered video analysis to simulate elite-level coaching in butterfly stroke technique. Using Google’s Gemini 2.5, the goal of the project is to determinewhether a large multimodel, trained on technique videos of elite swimmers, can generate detailed and personalized feedback that mirrors the insight of a well-informed human swim coach. Training videos featuring Olympic-level butterfly swimmers were used to establish an internal reference model of an ideal butterfly technique. The model was then presented with videos of collegiate swimmers and prompted to deliver coaching-style feedback. Overall, this project demonstrates the potential for AI systems to …
Ai Proof Benchmarking: Evaluating Mathematical Reasoning In Open-Source Llms Via Taylor Series Analysis, Godwin Idowu
Ai Proof Benchmarking: Evaluating Mathematical Reasoning In Open-Source Llms Via Taylor Series Analysis, Godwin Idowu
IPHS 300: AI for Humanity
This research project investigates large language models’ (LLMs) abilities to develop conceptually and mathematically correct proofs by using a benchmark based on the Taylor Series representation. The task examines LLM models for their capacity to adhere to definitions, theorems, and calculus. A range of models of varying sizes was tested, including Qwen, Gemma and LLaMA. Models under 2B parameters demonstrate poor understanding of Taylor and geometric series and apply wrong theorems while lacking logical reasoning about convergence. Models with at least 27B parameters typically generate proofs that are both coherent and almost complete. The research identifies existing constraints in symbolic …
Authenticity Under Review: How Well Does Genai Write College Admissions Essays?, Adrian Mangine
Authenticity Under Review: How Well Does Genai Write College Admissions Essays?, Adrian Mangine
IPHS 300: AI for Humanity
With nearly 30% of people ages 14-22 reporting frequent AI use, increasing AI assistance in college admissions essays poses a new consideration with ethical AI use (Rubin et al., 2024). This study investigates how well three generative AI models write college admissions essays: ChatGPT o3, Gemini 2.5 Pro, and Grok. Each model responded to a Common App 2025-2026 application prompt as three distinct applicants: Elijah, Grace, and Malik (Common App, 2025). Each student represented a common college applicant archetype: Elijah, an academically successful dual sports recruit; Grace, a near-perfect award-winning scholar; and Malik, a gifted student from an underrepresented background. …
Using Ai To Predict Nba Player Statistics Chatgpt Vs Google Bard, Rishil Kondapaneni
Using Ai To Predict Nba Player Statistics Chatgpt Vs Google Bard, Rishil Kondapaneni
IPHS 300: AI for Humanity
No abstract provided.
Ai Scouts In Baseball: How Ai Can Potentially Revolutionize Scouting In Professional Baseball, Parker Gibbons
Ai Scouts In Baseball: How Ai Can Potentially Revolutionize Scouting In Professional Baseball, Parker Gibbons
IPHS 300: AI for Humanity
No abstract provided.
Cross-Cultural Differences In Product Marketing, Junaid Yeasir Fahim
Cross-Cultural Differences In Product Marketing, Junaid Yeasir Fahim
IPHS 300: AI for Humanity
No abstract provided.
Comparing State-Of-The-Art Convolutional Neural Network Performance In Cancer Prediction Tasks, Cael Elmore
Comparing State-Of-The-Art Convolutional Neural Network Performance In Cancer Prediction Tasks, Cael Elmore
IPHS 300: AI for Humanity
No abstract provided.
Can Gpt-4 Fool Turnitin? Testing The Limits Of Ai Detection With Prompt Engineering, Abigail Foster
Can Gpt-4 Fool Turnitin? Testing The Limits Of Ai Detection With Prompt Engineering, Abigail Foster
IPHS 300: AI for Humanity
No abstract provided.
Evaluating The Potential Of Ai In Sports Consulting: Investigating Chatgpt- 4'S Ability To Consult An Mlb Team, Drew Robinson
Evaluating The Potential Of Ai In Sports Consulting: Investigating Chatgpt- 4'S Ability To Consult An Mlb Team, Drew Robinson
IPHS 300: AI for Humanity
No abstract provided.
Should Ai Have Intellectual Property Rights? An Analysis Of Copyright Law On Generative Ai, Ani Parnagian
Should Ai Have Intellectual Property Rights? An Analysis Of Copyright Law On Generative Ai, Ani Parnagian
IPHS 300: AI for Humanity
No abstract provided.
Can Large Language Models Replace The Role Of An Executive During An Oil & Gas Company’S Earnings Call?, Dillon Cleary
Can Large Language Models Replace The Role Of An Executive During An Oil & Gas Company’S Earnings Call?, Dillon Cleary
IPHS 300: AI for Humanity
No abstract provided.
The Woman’S Metamorphosis: A Time-Traveling Tale Of Image And Text, Grace Sparks
The Woman’S Metamorphosis: A Time-Traveling Tale Of Image And Text, Grace Sparks
IPHS 300: AI for Humanity
No abstract provided.
National Security: Does The Talk Match The Dollars? Topic Modeling To Map Presidential Speeches To Defense Spending, Casualties, And Freedom, Ben Gimbel
IPHS 300: AI for Humanity
No abstract provided.
Evaluating Gpt-4’S Consistency And Capacity To Assess And Self- Assess Literary Texts, Annalia Fiore
Evaluating Gpt-4’S Consistency And Capacity To Assess And Self- Assess Literary Texts, Annalia Fiore
IPHS 300: AI for Humanity
No abstract provided.
Decoding Emotions Beyond Words: A Holistic View Of Multimodal Sentiment Analysis, Samyak Shrestha
Decoding Emotions Beyond Words: A Holistic View Of Multimodal Sentiment Analysis, Samyak Shrestha
IPHS 300: AI for Humanity
No abstract provided.
Freedom, Democracy, And Well-Being: A Comparative Analysis Of Global Progress Indexes Using K-Means Clustering, Maxim Farkhat
Freedom, Democracy, And Well-Being: A Comparative Analysis Of Global Progress Indexes Using K-Means Clustering, Maxim Farkhat
IPHS 300: AI for Humanity
No abstract provided.