Data Analysis Shows Men More Likely to be Accused of Shaking Infants
McCormick graduate students collaborate with journalism students on investigative project
An analysis of thousands of police, court, and medical records conducted by Northwestern University students has shown that men are nearly three times more likely than women to be accused of violently shaking an infant.
The study was conducted by The Medill Justice Project, a Northwestern investigative journalism enterprise, in partnership with graduate students from the McCormick School of Engineering’s Master of Science in Analytics (MSiA) program.
Shaken-baby syndrome crimes involve caregivers who are accused of inflicting severe head trauma on children, typically under the age of 2, causing a triad of symptoms: brain bleeding, brain swelling, and bleeding within the eye. Out of nearly 3,000 cases nationwide, 72.5 percent of those accused of shaken-baby syndrome crimes are men, while 27.5 percent are women, the study found.
Research in recent years has raised some fundamental questions about the medicine and science behind shaken-baby diagnoses, leading to the Northwestern project, said Alec Klein, professor of journalism at Medill and director of The Medill Justice Project.
“It was once believed that the symptoms of shaken-baby syndrome would manifest immediately, but some research in recent years has shown that there can be what’s called a ‘lucid interval’ between the time of an infant’s head injury and when the infant becomes unresponsive — which could be a period of hours, days, or even weeks,” Klein said. “That research calls into question whether the last caregiver with the child would be the guilty party.”
The project, which has been under way for more than a year, will result in a national shaken-baby syndrome database that could help researchers, journalists, and the public gather insight into the crime.
As an extracurricular project during the 2012-13 academic year, four MSiA graduate students worked with undergraduates from the Medill School of Journalism, Media, Integrated Marketing Communications to identify and confirm 3,600 cases of shaken-baby syndrome. This required running defendant names and other identifiers through proprietary legal databases, cross-referencing them with police, appellate court and medical records, where available.
Amassing data about these deaths can be complicated, as police and court records often use general terms such as “head trauma” instead of explicitly stating that the baby was injured by shaking.
“No one else is publicly making this information available, so we consider it groundbreaking,” Klein said. “We’re hoping it serves the public to better understand this area of criminal justice.”
Supervised by professor Diego Klabjan, the graduate students — David Cooperberg, Justin Kim, Qifan Wu, and Yan Xue — created a special query tool to enable the journalism students to mine thousands of PDF documents for keywords. They also developed ways to visualize the data on a searchable Internet database.
“This is not the kind of analytics problem you often encounter,” said Klabjan, professor of industrial engineering and management sciences at McCormick and director of the MSiA program. “These students had a fascinating opportunity to use their skills on an important public service project.”
Now entering its second year, the MSiA program is a unique, 15-month analytics program housed in McCormick’s Department of Industrial Engineering and Management Sciences. Students are exposed to multiple aspects of analytics — such as big data, unstructured data, and data visualization — through a practicum, internship, capstone project, and coursework.