Cultivating Insights from Harvest Day 2024

Research & Innovation

Cultivating Insights from Harvest Day 2024

posted

Share

Copy Link Facebook Linkedin X Email

As technology continues to rapidly transform how companies do business, faculty across the Stevens School of Business are planting new ideas and harvesting novel breakthroughs in research across AI, large language models (LLM), leadership, teamwork and other essential topics.

Faculty shared the bounty of their work and progress with one another and Ph.D. students during SSB's fourth annual Harvest Day. In addition to showcasing research, the event also identified areas of research overlap, provided vital feedback and introduced new ideas to take research projects to the next level.

"Harvest Day is a celebration of research in the School of Business," explained Assistant Professor Jordan Suchow, the organizer of the event. "The goal of Harvest Day is for faculty from different areas to learn about what everyone works on and is excited about in their research. The event is well attended by faculty and Ph.D. students."

2024 HARVEST DAY PRESENTATIONS

Not Good Enough, But Try Again! The impact of improved rejection communications on contributor retention and performance in open knowledge collaboration

Presenter: Aleksi Aaltonen

Open knowledge collaboration often requires rejecting contributions made in good faith to maintain quality. However, rejections can demotivate new contributors from attempting further contributions. This research aims to determine if more informative rejection notices can encourage new contributors to try again by reducing uncertainty about future interactions with the system.

Sentimental Waves: Unraveling the Enchanting Forces Shaping M&A Success

Presenter: Suman Banerjee

Mergers and Acquisitions (M&As) are pivotal, transformative forces in the corporate sector, signaling significant strategic shifts and the promise of growth. However, they can be challenging due to excessive payments, regulatory hurdles and integration of diverse corporate cultures. This research tests if sentiment, as expressed in corporate communications and media narratives, influences the dynamics of M&A engagements and outcomes.

The Performance Consequences of (Mis)allocation of Human Capital

Pranav Garg

When an over-qualified employee is dispatched to tackle a problem, firms might forgo higher revenue-generating opportunities or pay a higher cost for that employee's given work. In addition, the employee may not develop their skills or feel challenged, causing them to leave the company. This research looks to calculate a gap between a role's required and the expertise deployed.

Financial Flexibility in the Gig Economy: The Impact of Ridesharing on Mortgage Loan Terms

Presenter: Balbinder Singh Gill

Ridesharing has made an undeniable impact in expanding a user's ability to get from point A to point B, while also providing an economic boost to the drivers. This research investigates the impact that the supplementary income provided by being a rideshare driver can have on the driver's mortgage loan terms.

Digital Platform: The Performance Implication of Alignment Between Public and Private Regulators

Presenter: Katia Meggiorin

Whether it's Instagram requiring age verification to create an account or Airbnb requiring hosts to get a city license for a short-term rental property, platforms have user regulations. This research examines if platform performance is higher when regulations from policymakers and platforms are aligned because the lower regulatory uncertainty increases the platform's user participation.

Check out Harvest Day 2023

The East Asian CEO Pay Discount and the Bias-Mitigating Role of Diverse Boards

Presenter: Ann Murphy

Recent research has found that East Asian MBAs received lower starting salaries, a phenomenon explained by having a lower negotiation propensity. This has prompted a look into whether these same inequalities persist at the CEO level.

Dialogic Organizations: Designing with Agents

Presenter: Jeffrey Nickerson

While information systems have always been designed through a dialectic process, machines and new technologies have altered the process and are now in the conversation. This research designs a meta system that is an organization of synthetic personae that engage in synthetic conversation to produce a system.

Interdicting Dense Clusters in Networks

Presenter: Foad Mahdavi Pajouh

In complex systems modeled as networks, specific structural properties are crucial to understanding how these systems function, and researchers are interested in strategically identifying and manipulating these critical network components. The goal of this research is to develop an optimization approach that determines the most cost-effective way to disrupt dense clusters within a network, thereby altering its fundamental characteristics while minimizing the resources required for intervention.

Trading TP2 Option Violations

Dan Pirjol

Total positivity (TP) has previously been applied to problems in mathematical finance. This research uses the TP properties in the strike and maturity dimensions.

Redundancy or Cross-Validation of Environmental Governance? ISO 14001 and Environmental Stringency for Emission Reduction

Presenter: Jose Tribo

ISO 14001 is an internationally recognized standard for environmental management systems (EMS) that many companies around the world use to demonstrate their environmental responsibility and improve their performance. This research examines if there is a cross-validation effect between environmental stringency and ISO 14001 in reducing corporate emissions.

When is the Right Time? AI Assistance, Information Sharing and Team Decision-Making

Presenter: Bei Yan

Virtually every working person today is part of a team. Teams can sometimes make poorer decisions when they only focus on shared information while ignoring unshared information. This research looks to see if integrating AI into a team setting at a specific time can improve their decision-making process.

Counterfactual Analysis via Large Language Models

Presenter: Zonghao Yang

Counterfactual analysis is a method of examining hypothetical scenarios and is studied across various subjects. This research uses the techniques via a large language model to analyze the results of a scenario.

Predicting and Regulating Inappropriate Behavior of LLMs Before They Happen

Presenter: Denghui Zhang

As LLMs have grown in power and scale, their outputs can be toxic, misleading, unsafe or biased. The research examines if using internal states can predict LLMs undesired behavior and prevent bad results from being shared early on.

Back to School of Business News