Costello Research Data Analytics

Costello Research Artificial Intelligence

Artificial Intelligence

Costello Research Machine Learning

Costello Research Sustainable Operations

Information Systems and Operations Management Area

Costello College of Business health care research puts “patients at the center”

Jennifer Anzaldi — Tue, 11 Mar 2025 15:55:18 +0000

Costello College of Business health care research puts “patients at the center” Jennifer Anzaldi Tue, 03/11/2025 - 11:55

In This Story

Nirup Menon

Like virtually every other industry, health care is increasingly prioritizing digital transformation. The sector is unique, however, in that its results are measured not only in business terms but also tangible outcomes for people—often, literal life and death. So are newly acquired technologies actually paying off for patients?

Nirup Menon, a professor of information systems at the Donald G. Costello College of Business at ��, says that the answer is “not always.”

Nirup Menon

His recently published paper in Decision Support Systems tackles the so-called “HIT paradox,” or the widespread perception that health information technologies (HIT) have not yet moved the needle on important outcomes such as productivity, quality of care, and patient safety.

Menon co-authored the paper with Costello colleagues Amitava Dutta and Sidhartha Das.

Based on comprehensive survey data from approximately 6,000 U.S. hospitals, the research team looked into whether those that adopted Clinical Decision Support Systems (CDSS) saw lower mortality rates for cardiac patients.

“CDSS is not only for cardiologists,” Menon explains. “It is hospital-based—a system that helps with clinical decision-making. But we know that many cardiac patients may not necessarily have cardiac as their only problem. There are probably decisions being made about them using all kinds of ailments and medications, and so on.”

The basic idea behind CDSS is to use technology to mine actionable insights from a wealth of patient data, giving clinicians key tools to make informed decisions at the point of care. Theoretically, a hospital with CDSS solutions should be much better equipped to handle complex cases—such as a heart-attack sufferer with diabetes or another comorbidity—in real time than one without.

However, Menon and his co-authors discovered that when it came to preventing deaths from cardiac emergencies, the impact of CDSS was context-specific. Their paper finds a number of complementary effects suggesting that health care technologies need help from their environment in order to be most effective. For example, the presence of cardiac medical services (CMS), e.g. diagnostic catheterization and electrophysiology, was unsurprisingly associated with lower mortality rates—but CMS combined with CDSS was more impactful than either on its own.

“The labor force—by which I mean the physician and the entire team of nurses and technicians—should be trained to use this technology appropriately,” Menon summarizes. “You also need real-time integration between CDSS and other IT systems, because if it’s not well-integrated, the provider will not have all the data at their fingertips. If you don’t provide the right inputs into a CDSS, it’s not going to give you the right outputs.”

Menon points out that the “HIT paradox” isn’t limited to CDSS or any single technology. President Obama’s 2009 economic stimulus package, after all, included tens of billions in financial incentives for health care providers to digitize their patient records. By 2017, 95 percent of U.S. hospitals had adopted electronic patient records. Yet, as Menon tells it, “hospitals are just chugging along. The quality remains the same and the costs are just increasing. Or you might see improvements in one small department. So we are trying to find the variables that create complementarities within large samples.”

Menon knows, however, that the applications of health care tech can be closely targeted to relatively tiny patient populations, too. Another recent paper of his, published in JMIR Medical Informatics, uses causal survival forests, a machine-learning algorithmic technique, to determine which of two chemotherapy drugs promoted the most longevity for terminal prostate cancer patients. Taking into account age, race and comorbidity symptoms, their analysis produced an easy-to-use prescription policy tree that, by itself, could extend patients’ lives by almost two months—if the test sample, comprised of 2,886 veterans treated at VA health centers, was representative of the wider patient population.

“If you go down every branch of the policy tree, the numbers become very small,” Menon says. “It almost becomes like personalized medicine, because you can factor in age, race, gender—although gender didn’t matter in our study—PSA numbers, bilirubin numbers, etc.”

Menon has ongoing research projects aimed at improving health care through technology, at both the patient level (a la the prostate cancer study) and the ecosystem level (a la the CDSS study). One paper in progress focuses on Covid-19 and how the data-sets research scientists selected for their studies influenced their findings. Another looks at telemedicine’s effects on quality of care.

“My foray into health care began with my PhD dissertation, which was on IT in hospitals,” Menon says. “At that time, I was working primarily from a hospital administration point of view. As a business school researcher, it seemed logical to stay there. But as you come across more problems, and you read more, you realize that the patient is the center of everything, not the hospital.”

Topics

Costello Research Health & Well-being at Work

Costello Research ICT

George Mason professor furthers impact of telemedicine in Ukraine

Jennifer Anzaldi — Tue, 14 Jan 2025 22:39:16 +0000

George Mason professor furthers impact of telemedicine in Ukraine Jennifer Anzaldi Tue, 01/14/2025 - 17:39

Ukraine’s health care system has been hit hard amid the ongoing war. Power outages, staffing shortages, and the destruction of hospitals have added up to a drastic reduction in available care for the already-vulnerable population. In a desperate attempt to bridge the gap, Ukraine’s Ministry of Health opened the country to telehealth solutions from overseas. But will these prove to be a successful substitute for at least some necessary services, or turn out to be no better than a tech Band-Aid?

Answering that question is where Mariia Petryk, assistant professor of information systems and operations management at the Costello College of Business at ��, comes in. In her spare time, she works as volunteer director of analytics for TeleHelp Ukraine (THU).

Mariia Petryk

Founded by a cross-disciplinary group of Stanford students shortly after the war’s inception, THU was designed to succeed where other telemedicine initiatives in crisis-affected areas have failed. The founders worked tirelessly to assemble an international volunteer network comprising medical professionals, translators, interpreters and administrative “health navigators.” Aware that medical consultations were only part of the patient journey, THU’s founders sought to address the entire continuum of care.

Petryk stresses that while the project originated at Stanford, the technical team included “people from Chicago, Boston, other California schools…some very active volunteers were in Australia, South Korea, Canada and other countries.”

Petryk, herself of Ukrainian descent, was honored to lend her data science expertise to this worthy project. As analytics director, she manages a dozen or so number-crunching volunteers who measured and documented THU’s impact upon Ukraine’s displaced population during the initiative’s first full year.

As Petryk explains, “The Russian invasion created a humanitarian crisis where a lot of people were internally displaced. And when people relocate to a new place, they don’t know where to go for health care. They also are at higher risk for many issues, including mental health problems. And they don’t know where to turn to treat chronic diseases they may have.”

THU’s primary focus during its first year was delivering much-needed services to this population of war-ravaged internal exiles.

Petryk’s analytical work gave rise to a recent case study of THU published in Journal of Global Health. The paper’s other lead author was Aditya Narayan, a Stanford medical student and THU’s director of implementation and evaluation.

Their findings describe some impressive early successes. THU facilitated more than 1,200 virtual patient appointments from May 2022 to May 2023 alone. Despite often-chaotic conditions, patient attendance rates were above 70 percent for nine of the 13 months studied. As the first year wore on, the THU team found ways to prevent no-shows—for example, employing the popular texting platform Viber to communicate with patients and assigning an individual health navigator to each patient.

Even more impressively, 96 percent of patients reported that their health complaints were at least partially resolved during their visit.

The paper argues that aspects of THU’s model could be adapted for use in other humanitarian contexts. In its initial growth phase, THU had access to advanced technological infrastructure and a wide network of medical providers, by dint of its academic origins. This implies that partnerships with academia could be critical to replicating THU’s success outside Ukraine.

Petryk remains proud of THU’s impact and her role in helping define it. “Based on actual appointments and how much that amount of care would cost at a hospital, THU delivered an estimated $1 million worth of services in its first 13 months,” she says.

Looking ahead to THU’s future, she says, “I can only wish to see this ‘start-up,’ as it were, go for the IPO.”

For more information and to explore volunteering opportunities, visit THU’s website.

In This Story

Mariia Petryk

Connect with the Costello College of Business

Topics

Costello Research Health & Well-being at Work

Costello Research ICT

Costello Research Digital Platforms

Faculty and Staff News

Off the Clock

STEM outreach

MS in Finance students answer burning investment questions

Jennifer Anzaldi — Tue, 22 Oct 2024 17:08:48 +0000

MS in Finance students answer burning investment questions Jennifer Anzaldi Tue, 10/22/2024 - 13:08

In This Story

Derek Horstmeyer

�� Master of Science in Finance students have the rare opportunity to use their budding analytical skills to solve some of the mysteries of today’s financial markets. Moreover, their faculty-supervised research projects can land their names in the pages of The Wall Street Journal.

To date, these projects have spawned nearly 40 WSJ articles, not to mention dozens of write-ups in outlets such as MarketWatch, The Conversation, and the CFA Institute blog. They have also received television coverage on CNN and NewsNation.

SMIF officers Yashkaran Sidhu, Jonathan Pino, Matthew Rickard, Raheeg Joari, and Reema Hammad at the Chicago SMIFC conference.

As Costello College of Business finance professor Derek Horstmeyer tells it, the process started taking shape organically. “I had started writing for the CFA and The Wall Street Journal. In my talks with my students, some of them revealed they had research interests. It just kind of took off from there—I would bring it up in class and offer to supervise a research project that would last about a week.”

Student researchers are mainly recruited from Horstmeyer’s Montano Student Managed Investment Fund (SMIF) class, a hands-on exercise in security analysis and portfolio construction that serves as a capstone to the MS in finance degree.

While many student research projects originate with questions suggested by the students themselves, Horstmeyer helps the student teams devise their scope and direction.

“They might say, for example, that they’re interested in crypto. So I’ll show them previous articles on crypto and we kind of develop a new idea from there,” Horstmeyer says.

By the time they get to SMIF, students are usually comfortable getting their hands dirty with data. Finding the appropriate data-set for the question they’ve posed, however, can be a challenge. That’s another area where Horstmeyer’s support becomes critical. “The big bottleneck is they just don’t know where the data is,” he says. “We as faculty have access to these private databases. So I’ll guide them or set up the data.”

Horstmeyer also introduces the student teams to the best methods for parsing the data, while leaving students enough leeway to conduct independent follow-up analyses if needed.

Of the dozens of projects completed thus far, there are a few that, for Horstmeyer, typify the process at its best. For example, finance students Matthew Rickard and Camila Marín Builes investigated how stock-market trends fare once they have been familiarized enough to warrant a nickname. Among the eight named trends included in the study were FAANG (Facebook, Amazon, Apple, Netflix and Google), and “The Magnificent Seven” (Nvidia, Tesla, Meta, Apple, Alphabet, Amazon, and Microsoft).

“We did an extensive Google search to find the biggest named trends. Then we used the Wayback Machine to figure out when the trend was first coined, like the exact date. We did all sorts of tests to determine what kind of abnormal returns you can expect if you stick with that trend,” Horstmeyer explains.

The students found that newly christened trends experience an extended honeymoon period of about 12 months, amounting to an average of 13 percentage points in excess returns. But in their second year, trends start to go stale, underperforming the S&P 500 by an average of two points.

The Wall Street Journal published Horstmeyer’s article “Once a Hot Stock-Market Trend Has a Name, Its Best Days Are Likely Past”—published under his byline with the student researchers mentioned by name early on—in March 2024.

Horstmeyer says that participating in these projects helps students hone skills and techniques that enhance their careers and give them a competitive edge in the job market. “I get employers reaching out and saying, ‘I saw this line item on a candidate’s resume saying they did some pretty high-level data analysis for you. Is it true?’ They’re kind of surprised the student contributed to the CFA or a Wall Street Journal article.”

Melvin Jonathan Reyes Echeverria, who completed his MS in finance degree in May 2024, says, “I assisted Professor Horstmeyer with data analysis for the WSJ and the experience was amazing…I was able to leverage everything that I was taught in my MS in finance classes. It was a privilege to be published and something that was unexpected to receive at George Mason.”

Topics

Finance - Costello

Costello Research Corporate Finance

Master's in Finance Program

Finance Faculty Research

Students

Impact Fall 2025

The NYPD gave officers iPhones. Here’s what we learned about race and policing

Jennifer Anzaldi — Tue, 04 Jun 2024 16:50:43 +0000

The NYPD gave officers iPhones. Here’s what we learned about race and policing Jennifer Anzaldi Tue, 06/04/2024 - 12:50

In This Story

Brad Greenwood

The controversy about biased policing seems to draw endless fuel from race-based differences in public perception. Simply put, the vast majority of White citizens in the United States believe the police are doing a good job, including on issues of racial equality, while a similar percentage of Black citizens hold the opposite opinion. And while a growing number of studies have indicated persistent patterns of racial discrimination in policing, an emergent concern among scholars is that the data these papers rely on are also subject to baked-in biases, since they often derive from officers’ self-reports of their own behavior.

Brad Greenwood

Enter Brad Greenwood, professor of information systems and operations management at the Donald G. Costello College of Business at ��. One of his research interests lies in how digital technologies are bringing unprecedented transparency to police practices. For example, Greenwood’s 2022 paper documented how the introduction of body-worn cameras for the New York Police Department (NYPD) resulted in a significant reduction in abuse-of-authority complaints.

His latest work on policing is forthcoming in Proceedings of the National Academy of Sciences. Along with Gordon Burtch from Boston University and Jeremy Watson from the University of Minnesota, Greenwood examined the recent rollout of iPhones across the NYPD, which included a series of digital tools designed to replace the handwritten memo books officers previously relied on. Instead of scribbling in the physical books, which NYPD officers were required to hang onto even into retirement, officers could log their activities directly into a centralized database maintained by the NYPD. These detailed digital records shed fresh light on how cops spend their time—and attention—on the beat.

The researchers tracked data on NYPD stops and complaints in 2017 and 2018, the period when iPhones were being rolled out across precincts in New York City. A curious pattern emerged. There was an 18% increase in reported stops after a precinct received iPhones, which would be consistent with the digital tools making it easier for officers to report a citizen interaction. Further, the researchers discovered that this increase resulted in neither more arrests nor more complaints from the public. It wasn’t, therefore, that the phones were somehow causing the police to stop people more often, but rather that so-called “unproductive stops”—those leading to no further action—were being reported more often.

However, when breaking the results down across White and non-White citizens, the researchers found that unproductive stops involving non-White citizens were entirely responsible for the increase. In other words, the observed changes were based on police encounters with non-White members of the public, that would likely have gone unreported in the days of pen and paper. More specifically, after switching to the smartphone system, officers logged 22% more stops involving non-White citizens, while the number of reported stops of White citizens remained unchanged. These are statistical averages—the pattern was more marked in high-crime neighborhoods and those with a greater proportion of non-White residents.

Greenwood offered an interpretation of the finding: “The concern here is that we have an underreporting, which is concentrated in certain groups and means that we need to be cautious when interpreting prior work. On the one hand, it opens the door to bias in police interactions with civilians being worse than initially anticipated, at least based on the frequency of stops. On the other hand, it could mean that older data doesn’t accurately reflect the likelihood of an arrest once a stop occurs. And we need to be doubly cautious, because we don’t know if officers are reporting stops more frequently just because it is easier, or for some other reason.”

Greenwood cautions against making sweeping conclusions based on the study. “The only thing we know for sure is that more and deeper work is needed by scholars and policy makers to ensure transparency between law enforcement and the people they are charged to protect,” he said.

On the whole, however, the study raises the possibility that race-based disparities in policing are not only very real, but may have been underestimated thus far because of reporting gaps.

As police officers are not obligated to document all civilian interactions, their decisions regarding what—and what not—to report can be biased. The introduction of new technology, as in the case of the NYPD, can help counter such biases, but is not the only avenue worth pursuing. The researchers recommend that police departments “investigate the appropriate organizational complements (i.e., policies and procedures) necessary to uncover and eliminate such biases.”

Topics

Costello Research Cybersecurity

Costello Research ICT

Costello Research Digital Platforms

impact fall 2024

Cryptocurrency’s surprising transparency advantage

Marianne Klinker — Tue, 02 Jan 2024 18:18:45 +0000

Cryptocurrency’s surprising transparency advantage Marianne Klinker Tue, 01/02/2024 - 13:18

Despite the fears of regulators and skittish investors, clear and accurate signals of cryptocurrency quality may be hidden in plain sight.

As perhaps befits a product of the post-2008 economy, the cryptocurrency space has never known normalcy. In a mere 13 years, crypto went from an untried software innovation of mysterious origin to being touted as the future of investing by major movie stars in Super Bowl commercials. Soon thereafter, of course, came the “crypto winter” of 2022, which began well before, but was surely deepened by, the downfall of FTX and disgraced wunderkind Sam Bankman-Fried.

Forthcoming in Journal of Management Information Systems, a recent paper is the first to examine whether the intensity of developer engagement with a cryptocurrency could indeed be related to its quality.

The crypto backlash appears rooted in the belief that the sector is too rife with bad actors, and too technologically complicated, to be worthy of public trust. The consensus in the crypto community, however, is that accurate clues to crypto quality are hidden in plain sight. The majority of cryptocurrencies use open source platforms such as GitHub for the development of their software. Seasoned investors will routinely inspect the publicly available development history of a cryptocurrency as a basic aspect of due diligence.

Forthcoming in Journal of Management Information Systems, a recent paper co-authored by Mariia Petryk, an assistant professor of information systems at the Donald G. Costello College of Business at ��, is the first to examine whether the intensity of developer engagement with a cryptocurrency could indeed be related to its quality. (Her collaborators were Liangfei Qiu and Praveen Pathak of University of Florida.)

Mariia Petryk

“This paper is about how we evaluate reputation in a setting where traditional reputational instruments are not precise,” Petryk explains. “When we choose a doctor, we often rely on an endorsement from someone we trust. But in lieu of that, we look into the history of the doctor—education, training, etc.” By the same token (no pun intended), crypto investors could “judge quality based on the total effort developers made to write the code.”

Using GitHub data for 559 cryptocurrencies over the period August 2016 to December 2019, the researchers zeroed in on five open-source activities. They classed three under the heading of “quality enhancement.” Commits are code modifications submitted by developers and eventually added to the crypto source code. Before official adoption, proposed modifications exist as pull requests awaiting assessment. Issues are questions, bugs, or problems raised by developers or users for discussion within the open source platform.

The remaining two activities fall under “diffusion of software,” essentially a proxy for developer attention. Forks are mirrored copies of the original code attached to a developer’s account. Watches are a way for developers to “subscribe” in order to more closely follow the discussion within the community around a cryptocurrency.

Across the data-set, the researchers found that a one-standard-deviation increase in forks and watches would equate to a 0.56% price increase per month, or 6.7% per year.

As for the three “quality enhancement” metrics, increases in issues exerted upward pressure on token price, but pull requests and commits displayed the opposite effect. A one-standard-deviation increase in issues led to a 4.3 percent higher price over one year, while an equivalent increase in pull requests was linked to annual losses of five percent.

These contradictory results perhaps point to uncertainties about the length of time it would take for crypto admins to evaluate and implement proposed changes. It goes without saying that greater numbers of requests for revision would heighten such uncertainties and their possible implications for cryptocurrency quality.

Overall, these findings imply a “virtuous circle” whereby more promising cryptocurrencies attract more developer attention, which in turn produces quality improvements that are reflected in the token price. While the expansion of the developer community creates the potential for rising pull requests and commits to lower the token price, the price-positive impact of forks and watches was approximately six times stronger. As a general rule, then, intensity of developer attention and engagement could be viewed as a leading indicator of how the market values the tokens—with the very important caveat that broad statistical patterns may have little to no bearing on specific investment decisions.

Therefore, information from GitHub and other open-source platforms could be an important reference point for policymakers looking to evolve nuanced regulatory approaches. “One purpose of regulation is to create equal opportunity,” Petryk says. “Transparent mechanisms allow investors to learn about underlying assets and properties and make their own judgments. Open source platforms like GitHub are open to everyone, but not everyone thinks to use it. In terms of info disclosure, this could be an important factor.”

More Costello College of Business Faculty Research

In This Story

Mariia Petryk

The hidden cost of environmental policy uncertainty
October 24, 2025
George Mason marketing professor receives prestigious research funding
October 10, 2025
‘Nothing to see here’: How corporate spin confuses Wall Street
October 8, 2025
Costello College of Business welcomes eight new research-focused faculty
October 1, 2025
George Mason researcher is using AI to identify human trafficking hot spots
September 29, 2025

Topics

Costello Research Blockchain

Costello Research Fintech

Costello Research Risk Management

Future of Work & Leadership - Costello

Mason professor contributes to groundbreaking mega-study in behavioral economics

Marianne Klinker — Wed, 09 Aug 2023 13:44:11 +0000

Mason professor contributes to groundbreaking mega-study in behavioral economics Marianne Klinker Wed, 08/09/2023 - 09:44

A new "mega-study" consisting of dozens of simultaneous, independently designed experiments–shows that competitions have no automatic impact on our morality.

Academic research is tapping into the power of the crowd, thanks to an emerging paradigm known as the “mega-study.” Instead of one experimental design with a single set of parameters, mega-studies utilize a diverse array of simultaneous studies submitted by independent research teams, with a shared research question and participant pool.

In addition to efficiencies of scale, mega-studies can help scholars more quickly come to grips with the slippery issue of generalizability–i.e., the degree to which findings in one context apply to other contexts. For example, something that motivates people to do one prosocial behavior (e.g., recycling) might be different from what motivates people to do another prosocial behavior (e.g., donating to charity). Observing the full range of results across dozens of studies provides a clearer sense of how different environmental conditions may affect research outcomes.

The mega-study model is ideal for the field of behavioral economics, which explores how various factors in the world around us influence our daily decision-making. Human behavior, after all, is extremely complicated and changeable.

Einav Hart

Einav Hart, an assistant professor of management at �� School of Business, participated in the first-ever crowdsourced mega-study in behavioral economics, recently published in Proceedings of the National Academy of Sciences (PNAS). Hart’s experimental design was one of 45 selected for the mega-study. The broad research question was “Does competition erode, promote or not affect moral behavior?”

The mega-study was conducted to better understand and reconcile previous findings about the relationship between competition and moral behaviors. Previous (individual) studies have yielded mixed results. While some studies show that competition promotes moral behaviors such as trust and reciprocity, others point to a moral erosion.

In the mega-study, 18,123 online participants were randomly assigned to the 45 research designs. Notably, the 45 experiments were quite eclectic in their interpretations of “competition” and “moral behavior”. One research team, for instance, proposed a game in which participants solved puzzles either in or out of a competitive scenario, and were asked to self-report their performance. In this case, the honesty or dishonesty of their self-reported scores was used to measure “moral behavior”. Another proposed design was an online game in which an “investor” granted points to an “investee”, who could then choose whether or not to return the favor.

The pooled data from all 45 experiments showed that competitive conditions led to a minor decline in participants’ moral behavior. However, the size of the decline varied considerably from study to study. Further analysis revealed that the majority of the variance was due to the design differences among the 45 experiments.

The authors conclude that when it comes to broad research questions that admit a wide range of interpretations, there may be limitations to what a single study with a specific context–even one with a very large sample size–can reveal. Hart says, “This shows that in any individual study, you could observe very different results even with the same ‘ground truth,’ and this variance is largely dependent on how competition and moral behavior are operationalized.” In these cases, true generalizability may arise only from a greater diversity of experimental approaches, such as the mega-study and other crowdsourced research paradigms. As the PNAS paper states, “Our findings provide an argument for moving toward much larger data collections and more team science.”

Topics

Costello Research Competitive Strategy

Management Faculty Research

School of Business Faculty Research

More School of Business News

The hidden cost of environmental policy uncertainty
October 24, 2025
George Mason marketing professor receives prestigious research funding
October 10, 2025
‘Nothing to see here’: How corporate spin confuses Wall Street
October 8, 2025
Camp Costello gives first-year students a head start in their college careers
October 1, 2025
Costello College of Business welcomes eight new research-focused faculty
October 1, 2025

In This Story

Einav Hart

How machine learning improvements are helping to fight human trafficking

Marianne Klinker — Wed, 22 Feb 2023 13:09:33 +0000

How machine learning improvements are helping to fight human trafficking Marianne Klinker Wed, 02/22/2023 - 08:09

Human trafficking is a global crisis of overwhelming scope. Fortunately, anti-trafficking organizations can use AI to predict the criminals’ next moves–with the help of a �� professor.

The fight against human trafficking has a David and Goliath aspect to it. Trafficking rings are a global scourge with an estimated 25 million victims and $150 billion in annual profits. Agencies and NGOs tasked with stopping the traffickers, however, are typically small and under-resourced. Recently, a technological tool has come along to help close this gap–machine learning (ML) algorithms.

Primed with enough relevant data, these algorithms can find hidden clues to how traffickers operate and what they are likely to do in the future. In principle, anti-trafficking agencies can use ML predictions to direct their scant resources to the areas of greatest risk.

Abhishek Ray

Abhishek Ray, an assistant professor in the Information Systems and Operations Management area at the �� School of Business, has added his expertise to this effort. With collaborators Viplove Arora (post-doc researcher at SISSA, Italy), Kayse Maass (of Northeastern University) and Mario Ventresca (of Purdue University), Ray developed a model that reduces the guesswork of implementing ML predictions.

As Ray explains it, the model supplies a “layer on top” for improving outcomes of existing ML solutions. Armed with success and failure rates (i.e., true/false negatives and positive trafficking detections) of an anti-trafficking agency, it looks for scenarios in which the agency could achieve the same number of successes with fewer errors.

In a recently published paper, Ray and his co-authors apply their model to two real-life agencies: Global Fishing Watch (GFW), which targets trafficking in the global seafood industry, and Love Justice International (LJI), which monitors transit stations on the India-Nepal border.

The ML solution employed by GFW cross-references a large database of fishing vessel activity, which is obtained via satellite, with law enforcement data on human trafficking. The resulting predictions indicate the vessel types, behaviors, and features that the agency should watch most closely.

But criminals sometimes learn faster than machines. “If someone is caught on a long trawler, and they know they can get caught, they will never go on the long trawler again,” Ray explains. Traffickers’ evolving tactics may not be fully captured in vast, wide-ranging data-sets, such as the ones derived from satellite vessel surveillance. Changes in agency success rates, on the other hand, are closer to the action and may pick up the latest stages of this cat-and-mouse game.

Looking back over agency data from 2012-2018, the researchers found that had their model been used alongside Machine Learning, the agency Global Fishing Watch would likely have discovered more instances of trafficking while committing far fewer false negatives and false positives.

Looking back over agency data from 2012-2018, the researchers found that had their model been used alongside ML, GFW would likely have discovered more instances of trafficking while committing far fewer false negatives and false positives. In addition, monitoring recommendations shifted considerably over the six years. For example, under certain conditions, drifting longline vessels were the most likely suspects for the years 2012-2017. In 2018, however, squid jiggers replaced them and were singled out as among the most suspicious. These shuffling priorities may reflect the traffickers’ changing their vessel of choice to evade detection.

In the case of Love Justice International, Ray’s model could serve as a workaround for constraints such as a shortage of well-trained staff. LJI could input a target outcome into the algorithm, such as a baseline number of positive trafficker IDs, and receive recommendations for how to make the best use of strained resources.

For Ray, the fight against human trafficking is personal. As a child growing up in Kolkata, India, he was cared for by a household employee who, as his parents later found out, had been trafficked when she herself was a child. “I learned this when I was doing my PhD in the U.S.,” he says. “It’s like having a part of your childhood taken away. You came into contact with something criminal and had no understanding.”

“If I could make a contribution towards solving the problem, I’d consider it sort of giving back what I had taken unknowingly.”

To take full advantage of Ray’s “layer on top” ML solution, anti-trafficking agencies would need to create effective systems for collecting and centralizing data. Depending on the context, this might present challenges ranging from resource limitations to a political climate where trafficking data could be used against the already-victimized. “The maximum value from our framework is when it’s acting in real time with real-time data,” Ray says.

Topics

Costello Research Artificial Intelligence

Costello Research Machine Learning

Costello Research Cybersecurity

Tech Talent Investment Program

Artificial Intelligence

More School of Business Faculty Research

The hidden cost of environmental policy uncertainty
October 24, 2025
George Mason marketing professor receives prestigious research funding
October 10, 2025
‘Nothing to see here’: How corporate spin confuses Wall Street
October 8, 2025
Costello College of Business welcomes eight new research-focused faculty
October 1, 2025
George Mason researcher is using AI to identify human trafficking hot spots
September 29, 2025

In This Story

Abhishek Ray

Four stop signs at intersections are one too many, suggests Mason researcher

Jennifer Anzaldi — Wed, 19 Oct 2022 14:18:05 +0000

Four stop signs at intersections are one too many, suggests Mason researcher Jennifer Anzaldi Wed, 10/19/2022 - 10:18

In This Story

Jiasun Li

For most drivers in the United States, obeying a stop sign upon approaching an intersection is an unavoidable annoyance. But for �� Finance Professor Jiasun Li, it’s a problem waiting to be solved.

His recent working paper proposes a simple and economical improvement: removing one stop sign from every four-way intersection. According to his calculations, this would boost not only driver safety, but environmental sustainability as well.

Li specializes in game theory, which models strategic interactions where rational agents seek – as humans generally do–to optimize outcomes for themselves. As he drove around Fairfax, Virginia, Li could not help but view four-way intersections through this academic lens.

He was struck by the suspicion that having four stop signs at an intersection was a flawed way of preventing traffic accidents. In effect, they lowered the potential cost of not stopping at the intersection, because drivers could assume that motorists from other directions, should there be any, would come to a stop. Drivers turning right, a shallower maneuver with less exposure to oncoming traffic, have the least risk to begin with and would have the greatest incentive to ignore the sign.

Li surmised that the outcome of all drivers obeying the sign fell short of a Nash equilibrium–game theory’s term for a stable set of norms that all parties are incentivized to follow.

His working paper presents mathematical models that support his intuition. Comparing the risks of collision against the gains from ignoring the sign (i.e. a smooth driving experience or conserving gasoline), he finds that a symmetric equilibrium under the current four-sign mechanism to be one in which left-turners and straight-line travellers honor the stop sign but right-turners do not.

In real life, of course, fear of getting slapped with a moving violation increases the likelihood of compliance with stop signs. But the need for law enforcement is less than ideal, Li says. If the four-way stop system were self-enforcing, expensive police resources could be directed toward serious crimes and away from preventive punishment for traffic violators.

Li’s working paper first considers what would happen if traffic laws were changed to allow right-turners to run the stop sign legally. In an intersection with four stop signs, this would indeed prevent accidents. However, many U.S. intersections have only two signs, which would place right-turners at risk of colliding with traffic coming from the left. In addition, Li argues that old habits die hard, including behind the wheel.

“If you need people to change their behavior, that’s going to be difficult,” he says.

Instead, Li recommends removing one sign–any sign–from among the four. In that case, drivers would know that running a stop sign could send them straight into the path of an oncoming car that has not been directed to stop. The universal fear would incentivize compliance without relying on the threat of police punishment. “Importantly, although the safety and incentive analysis rely on game theoretical reasoning, once implemented, a three-sign mechanism does not require any changes in behaviors for law-abiding drivers,” Li says.

Further, Li’s back-of-the-envelope math suggests that his relatively mundane change could be surprisingly impactful. Based on official statistics from various U.S. authorities, he estimates there are about one million four-way intersections with stop signs in the U.S., each one being traversed by more than 760,000 vehicles per year. That amounts to more than 760 billion (760,000 x one million) stops and starts in a year, one-fourth of which (190 billion) would be avoided in a three-sign setup.

According to Li’s calculations, assuming it takes 124.9 kilojoules of work to bring a car of average weight to a full stop, and the same amount to resume the previous level of speed, adopting a three-sign system would save a total of 118.65 trillion kJ in the U.S. per year. That equates to one billion gallons of gasoline, or about 2.7 days’ worth of national gas consumption.

Looking farther out on the technological horizon, Li speculates in the paper that “when self-driving cars finally become real, our new simpler mechanism may also economize AI computational costs.”

Li has no illusions about the resistance his recommendation may provoke. “You’re challenging people’s conventional wisdom. If there’s criticism, that’s to be expected. For example, some cops might not like this idea, because I’m taking away one of their income sources, even though there could be an overall gain to the whole society.”

This paper is only one manifestation of what Li sees as a scholarly mission to apply novel perspectives – in this case, an economics viewpoint derived from game theory – to real-world problems, especially those that have been taken for granted for so long that they are not even recognized as problems.

“Economics thinking should be applicable to a lot of different fields and improve our daily lives,” Li explains.

What seems totally natural in one context, after all, may be foreign in another. Many countries–including China, where Li was born–don’t use four-way stop intersections. “If I grew up with these stop signs, I probably wouldn’t think about it,” he says. This is an example, Li suggests, of how diverse experiences and ideas imported via immigration can enrich host nations. Far from wanting the last word on the issue, Li is hoping to inspire wide-ranging conversation about how to change environments and communities for the better.

“I would rather have the discussion ongoing and people give me ideas,” he says. “Only through debate will people understand it better.”

Topics

Finance - Costello

Finance Faculty Research