Data for a Better World

Five years ago, Kathleen Campbell Garwood, Ph.D., assistant professor of decision and system sciences, was teaching a summer class in advanced data analysis. Hoping to give the students more hands-on experience, she handed them a large set of data and some pointers on how to make the most out of it.

Nearly 4,000 miles to the south, in the mountainous South American nation of Bolivia, Fe y Alegría, a chain of schools that educates thousands of children from economically depressed neighborhoods, was trying to get a better handle on identifying which students needed the most help with specialized career training to escape poverty. The administrators had collected reams of data on these children but didn’t know what to do with it.

It turned out to be the start of a beautiful long-distance relationship between Saint Joseph’s and the South American educators that’s thrived over Skype and more than an occasional assist from Google Translate.

That summer in 2014, John Neiva de Figueiredo, Ph.D., professor of management, shared with Garwood that he had received all this raw data from Fe y Alegría that no one had really analyzed. Garwood gave her students the numbers, and they eventually narrowed down the factors that were common among the children who needed the most help, including whether their house had running water, how many people lived in the home and their father’s work status.

That information was so useful that the Bolivian educators keep coming back to Garwood and her classes — to help them determine whether they should close struggling schools or open new ones, or to assist in analyzing data from schools in other underprivileged countries like Haiti. Last year, three of Garwood’s students traveled to Bolivia and visited classrooms and homes to connect the data with real-life situations. Meanwhile, the output from the SJU data analysts — like heat maps tracking poverty in the neighborhoods around Fe y Alegría schools — continues to grow more sophisticated. 

"It empowers the students,” Garwood says. “They start to see that their skill set is about more than just getting a good job.”

Indeed, the students can use their increasing mastery of algorithms to undertake social projects like the work with Fe y Alegría — work that isn’t about making greater profits but about making the world a better place.

That’s not by accident. In an era when the availability of all sorts of data — from people’s social media habits to detailed Google Earth maps — and the remarkable predictive power of artificial intelligence are showing both the promise and the peril of the Information Age, faculty members in SJU’s Department of Decision and System Sciences are determined to crunch the numbers for social advancement.

Using Data for Good

Virginia Miori, Ph.D., department chair, says that using data for the betterment of society is “just part and parcel with the nature of the University” and in line with its official mission statement. These efforts are coming at a time of elevated societal awareness of how big businesses such as Facebook and Amazon are collecting and storing data about their users and how that information is being used to sell products or influence opinions. Miori sees the ongoing proof at Saint Joseph's as a reminder that data doesn’t have to be synonymous with bad intentions.

Over the last few years, professors in the department, and in some cases their students, have been working on projects to improve the quality of life and patient care in nursing homes, help drug-treatment centers improve outcomes, assist ride-sharing companies, operate more efficiently to curb traffic and pollution, create a more ethical pricing scheme for prescription drugs and help connect Pennsylvania farmers with towns where shoppers need better access to healthy produce.

Miori, who studied how to optimize business supply chains while pursuing a doctorate at Drexel University, has spent years on the front lines of analyzing data. In a world where piles of data are now not only accumulating at once unimaginable rates but so much easier for researchers to access, Miori and her colleagues have ramped up their efforts to show the power of smart analysis.

“I think the University is just becoming more aware of the importance of data and how to use it to improve all areas of life,” says Nicolle Clements, Ph.D., assistant professor of decision and system sciences, who also coordinates the master’s degree program in business intelligence and analytics. “You can learn so much from data, and here at Saint Joseph’s we’re well equipped with faculty experts.”

Clements is one of those experts. Over the last few years, she and Miori have developed a specialty in using satellite maps to study land use and agriculture practices, with the goal of helping farmers on the ground increase their crop yields. In a story similar to some of her colleagues, Clements said her work began about 10 years ago when she “stumbled” across a large cache of data — including aerial imagery — from agricultural regions of East Africa such as Kenya and Tanzania. The goal was to detect vegetation patterns and advise local farmers on sites that might yield better crops than those currently under harvest.

Now what she learned in Africa is being applied closer to home. Clements and Miori recently started working with the Pennsylvania Department of Agriculture’s PA Preferred program to use satellite images to better connect farms with nearby “food deserts” — communities where residents lack stores selling fresh produce.

“These tend to be areas where income levels are low, but where there’s a high death rate from obesity,” says Clements, who added that state officials only recently began to learn these areas are not just in the poorest neighborhoods of Philadelphia or Pittsburgh but also in smaller towns scattered across the state. The hope is that better mapping and analysis will help PA Preferred create new farm-to-market pipelines.

Meanwhile, the streets of larger cities are increasingly clogged by traffic caused by the surging popularity of ride-sharing apps such as Lyft and Uber. Sina Shokoohyar, Ph.D., assistant professor of decision and system sciences, has been working off of the voluminous data these companies produce in an effort to make urban transport more efficient.

“Uber is competing with public transportation,” explains Shokoohyar, so information about where residents are going and at what times of day can be incredibly valuable to city planners. Public transportation can use that data to provide better services, or even forge collaborations with the ride-sharing platforms. Shokoohyar is also studying how to make ride-sharing more environmentally sustainable and how to make their services function better during extreme weather events.

His colleague, Ronald Klimberg, Ph.D., professor of decision and system sciences, worked for a time with the U.S. Food and Drug Administration before moving into academia. Klimberg uses a technique called data envelopment analysis to sift through information from nursing homes in Massachusetts in order to develop more efficient methods of patient care.

That led to a side project in which he looks at techniques for nursing home staff to offer better personal attention to patients showing signs of dementia and, as a result, perhaps reduce pharmaceutical treatments and their side effects. “The whole purpose is to get their minds more active,” Klimberg explains.

Miori has been conducting somewhat similar data work for a drug-treatment program in South Florida, a region that has become a magnet for opioid abuse rehab centers as well as complaints that some facilities are not doing enough to prevent relapses. Her project, now in its third year, looks at the data from patient satisfaction surveys and follow-up calls by staff to better understand what factors best prevent patients from slipping back into addiction. The data, Miori explained, can help the center by showing that longer patient stays are more effective — information that could then convince insurance companies to pay for extended treatment.

The data also demonstrated that an alumni network that helped former patients stay away from drugs frequently needed to be more active in the period three-to-six months after the initial treatment — a time when relapses are common.

Navigating Data Challenges

For Miori and the department, the sum of this work is greater than its parts. That’s because, in a moment of heightened anxiety over privacy and abuses of data, the work that Saint Joseph’s faculty members and students are carrying out sends a message that data not only can but should be used for ethical purposes and to foster human growth.

“We are not seeking to use data to our advantage — nobody’s making any money off any of this,” Miori said. “We are using data for the common good and we’re using it ethically.” The challenge, she added, is for the next generation of business executives coming out of universities like Saint Joseph’s to hold onto those values. “It’s time for the people we put out into industry to put a greater ethical value on privacy and ethical use of data.”

Will Bunch is the national opinion columnist for The Philadelphia Inquirer and author of several books, including "Tear Down This Myth: The Right-Wing Distortion of the Reagan Legacy."