Mentoring the Future "Gatekeepers of Science"

“As technology progresses, the ability to store and share vast amounts of information is becoming easier and easier,” says Lucas Mentch, assistant professor in the University of Pittsburgh Department of Statistics. “Our smartphones, TVs, laptops, and wearable devices (like Apple Watches and Fitbits) are constantly collecting data on us. Throughout the course of a normal day, most of us are providing loads of data to dozens of organizations, typically without knowing or realizing it.”

This data collection, according to Mentch, has led to an increased need for statisticians who can process and extract insights from that data. Enter the next generation of statisticians, whom Mentch is deeply committed to teaching by serving as an undergraduate research mentor.

“The undergraduate students I work with are often doing the vast majority of the applied work on research projects,” he says. “I guide the process and steer them in the right direction, but they actively engage in discussions with scientists and practitioners, and in most cases, with a little experience, they begin to develop ideas and directions on their own. The data analysis skills possessed by many of our upper-level undergraduates surpass those of the domain scientists we collaborate with, so they are able to make genuine contributions.”

In some cases, Mentch invites undergraduates to help with his own research projects. (He’s involved in 18 such projects at the University, so there’s no shortage of opportunities.) Other times, students bring ideas for their own independent research projects to him, as was the case with Nicholas Kissel, an undergraduate student pursuing a double major in statistics and applied mathematics.

Kissel first became involved in research because of something he actually did not find. As part of the group project in Mentch’s data science class, Kissel and his classmates were unable to find any good predictors of hospital readmissions for diabetes patients. When Kissel came across an article claiming to have found a valid predictor variable later that same year, his curiosity compelled him to reevaluate his initial research. Kissel reached out to Mentch and together they designed a new research project, this time incorporating more advanced techniques and providing a thorough analysis of which variables might be important for which patients.

“The biggest value for undergraduates, I think, is the experience itself,” says Mentch. “The student has a chance to make a real contribution to a project that can potentially have a very real impact. Of course, it’s also something that employers and graduate schools look very favorably upon. Being able to demonstrate that you’ve already successfully carried out advanced research projects as an undergraduate carries a lot of weight.”

While working with Kissel on the hospital readmissions project, Mentch noticed that many of the skills Kissel was developing were similar to those needed to support another research project. That project, a collaboration with Xi Mi, focuses on sports medicine and aims to predict the likelihood of injury among individuals in the U.S. Army Special Forces.

“I enjoy working with undergraduates like Nick,” Mentch says. “They’re very bright and capable individuals, and with them, you can accomplish a great deal.

“At the end of the day, it’s up to us to provide the theory and methods necessary to ensure that scientific conclusions are valid and appropriate,” says Mentch. “Doing our part to enforce a high standard is more important than ever in the current age of ‘big data,’ where more and more interest is being shown in data-based decision making. I see our role as being essentially the gatekeepers of science, and new gatekeepers must be ready for what lies ahead."