The magnitude 4.8 Tewksbury earthquake surprised millions of people on the U.S. East Coast who felt the shaking from this largest instrumentally recorded earthquake in New Jersey since 1900.

But researchers noted something else unusual about the earthquake: why did so many people 40 miles away in New York City report strong shaking, while damage near the earthquake’s epicenter appeared minimal?

In a paper published in The Seismic Record, YoungHee Kim of Seoul National University and colleagues show how the earthquake’s rupture direction may have affected who felt the strongest shaking on 5 April.

Kim and her colleague and co-author Won-Young Kim of Lamont-Doherty Earth Observatory of Columbia University became curious about the strange pattern of shaking after visiting the epicenter area of the earthquake just eight hours after the mainshock.

“We expected some property damage—chimneys knocked down, walls cracked or plaster fallen to the ground—but there were no obvious signs of property damage,” the researchers said in an email. “Police officers within a couple of kilometers from the reported epicenter calmly talked about the shaking from the main shock. It was a surprising response by the people and houses for a magnitude 4.8 earthquake in the region.”

“This contrasted with the wide and huge response from the residents in and around the New York City area, some 65 kilometers from the epicenter,” they added.

The earthquake garnered more than 180,000 felt reports—the largest number ever for a single earthquake received by the U.S. Geological Survey’s “Did You Feel It?” app and website, according to a second paper published in The Seismic Record by USGS seismologist Oliver Boyd and colleagues.

Boyd and colleagues said the earthquake was felt by an estimated 42 million people between Virginia and Maine.

The reports from people southwest of the epicenter, toward Washington, D.C., indicated “weak” shaking on the scale that the USGS uses to measure an earthquake’s intensity, while people reporting from northeast of the epicenter felt “light to moderate” shaking.

Based on previous models of magnitude and earthquake intensity developed for the eastern U.S., however, a magnitude 4.8 earthquake should produce very strong shaking within about 10 kilometers or about six miles from its epicenter.

With this pattern in mind, Kim and colleagues wanted to look closer at the directivity of the earthquake’s rupture. To model the rupture, they turned to a kind of seismic wave called Lg waves, due to the lack of nearby seismic observation at the time of the mainshock. Lg waves are shear waves that bounce back and forth within the crust between the Earth’s surface and the boundary between the crust and mantle.

The resulting model indicated the earthquake rupture had propagated toward the east-northeast and down on an east-dipping fault plane. The direction of the rupture might have funneled the earthquake’s shaking away from its epicenter and toward the northeast, the researchers concluded.

In general, earthquakes in the northeastern U.S. take place as thrust faulting along north-south trending faults. The New Jersey earthquake is unusual, Kim and colleagues noted, because it appears to have been a combination of a thrust and strike-slip mechanism along a possible north-northeast trending fault plane.

“Earthquakes in eastern North America usually occur along the pre-existing zone of weakness—that is, existing faults,” the researchers explained. “In the Tewksbury area, a hidden fault plane trending north-northeast and dipping moderately can be mapped from the numerous small aftershocks detected and located” after the Tewksbury mainshock.

Boyd and colleagues noted that some damage was documented by a reconnaissance team deployed by the Geotechnical Extreme Events Reconnaissance Association and the National Institute of Standards and Technology. Along with cracks in drywall and objects falling from shelves, the team documented the partial collapse of the stone façade of Taylor’s Mill, a pre-Revolutionary War structure near the town of Lebanon, New Jersey.

The researchers have not yet attributed the earthquake to a particular fault but the locations of the mainshock and aftershocks suggest that the area’s well-known Ramapo fault system was not active during the earthquake.

The findings could “help us identify new earthquake sources and rethink how stress and strain are being accommodated in the eastern United States,” Boyd said.

He noted that some seismometers that were rapidly deployed to the region by the USGS will remain in place for at least five months.

“This can help us study, for example, mechanisms related to how the crust responds to the stress of a mainshock in the region, and how productive aftershock sequences can be in the eastern United States,” Boyd explained.

“Good station coverage can also allow us to observe how earthquake ground motions vary across the region as a function of magnitude, epicentral distance, and Earth structure. And each of these examples can help us better appreciate potential seismic hazards.”

In an examination of thousands of human-written college admissions essays and those generated by AI, researchers found that the AI-generated essays are most similar to essays authored by students who are males, with higher socioeconomic status and higher levels of social privilege. The AI-generated writing is also less varied than that written by humans.

“We wanted to find out what these patterns that we see in human-written essays look like in a ChatGPT world,” said AJ Alvero, assistant research professor in the Department of Information Science in the Cornell Ann S. Bowers College of Computing and Information Science. “If there is the strong connection with human writing and identity, how does that compare in AI-written essays?”

Rene Kizilcec, associate professor of information science in Cornell Bowers CIS is a co-author of “Large Language Models, Social Demography, and Hegemony: Comparing Authorship in Human and Synthetic Text,” published Sept. 27 in the Journal of Big Data.

This research stemmed from Alvero’s dissertation work at Stanford University. Part of his research involved an analysis of approximately 800,000 college admission essays written from 2015–17 by prospective students in the University of California system.

“We consistently found that there was a strong connection between the profiles of the applicants—their test scores, their demographic information, even the high schools they were applying from—and their admissions essays,” Alvero said. “The relationship was so strong that we were consistently able to predict an applicant’s SAT score, within about 120 points.”

“The ways that we speak can encode and contain information about our past and who we are,” he said, “and it’s very similar in writing, at least with personal statements.”

For this work, Alvero and the team compared the writing style of more than 150,000 college admissions essays, submitted to both the University of California system and an engineering program at an elite East Coast private university, with a set of more than 25,000 essays generated with GPT-3.5 and GPT-4 prompted to respond to the same essay questions as the human applicants.

For their analysis, the researchers used the Linguistic Inquiry and Word Count, a program developed in the mid-1990s by University of Texas social psychologist James W. Pennebaker that counts the frequencies of writing features, such as punctuation and pronoun usage, and cross-references those counts with an external dictionary.

“One of the first big-data analyses of college admissions essays was done about a decade ago by Pennebaker,” Alvero said, “and we wanted to try to build a robust understanding of these patterns across institutions, across time, and we did that through using the same method that they used.”

Alvero, Kizilcec and the team found that while LLMs’ writing styles don’t represent any particular group in social comparison analyses, they do “sound,” in terms of word selection and usage, most like male students who came from more privileged locations and backgrounds.

For example, AI was found on average to use longer words (six or more letters) than human writers. Also, AI-generated writing tended to have less variety than essays written by humans, although it more closely resembled essays from private-school applicants than those from public-school students.

Additionally, humans and AI tend to write about affiliations (with groups, people, organizations and friends) at similar rates—despite the AI not actually having any affiliations. As LLMs like ChatGPT become more popular and more refined, they will be used in all sorts of settings—including college admissions.

“It’s likely that students are going to be using AI to help them craft these essays—probably not asking it to just write the whole thing, but rather asking it for help and feedback,” Kizilcec said. “But even then, the suggestions that these models will make may not be well aligned with the values, the sort of linguistic style, that would be an authentic expression of those students.

“It’s important to remember that if you use an AI to help you write an essay, it’s probably going to sound less like you and more like something quite generic,” he said. “And students need to know that for the people reading these essays, it won’t be too difficult for them to figure out who has used AI extensively. The key will be to use it to help students tell their own stories and to enhance what they want to convey, not to replace their own voice.”

Alvero and Anthony Lising Antonio, associate professor of education at the Stanford University Graduate School of Education, are co-corresponding authors.