Christian Lemieux ’14 and the hunt for cost-effective weight-loss, smoking cessation programs
Obesity and smoking pose a healthcare double whammy, adding billions of dollars annually to the cost of treatment. So it is no surprise that governments, insurers, and employers take keen interest in weight-loss and smoking cessation programs that work.
Christian Lemieux ’14 (far left) took an interest, too. For his senior capstone project, he conducted a statistical analysis of data in order to identify which preventative programs would be most effective at reducing healthcare costs over the next nine years.
In textbooks, “every problem or situation is cut and dry. ... It isn't until you work with real data that you realize the many imperfections.”
Lemieux pored over data made available as part of an analytics competition sponsored by SAS, the leading provider of business analytics software and services. The competition offers university students and professors an opportunity to demonstrate their predictive modeling skills. The multi-year data sets included demographic and socio-economic details, healthcare costs, and claims information.
“Of all the work I had done at Bryant, this was by far the most daunting and research intensive,” says Lemieux, who graduated in May with a Bachelor of Science in Applied Mathematics and Statistics. “ There was no set specification; the outcome was entirely my own. The data was very detailed and involved, but it was up to me to figure out what I needed and how to combine elements to get them to make sense.”
No cut-and-dry solutions
The most striking aspect of the project was the challenge posed by real data, Lemieux says. In textbooks, “every problem or situation is cut and dry. The problems are meant to be perfect for the sake of explaining the concept. It isn't until you work with real data that you realize the many imperfections.”
It’s a challenge he now embraces as a data and reporting analyst at OnProcess Technology in Ashland, MA. “Digging through the data and understanding what is useful and what is important is the exciting part of my job,” says Lemieux, who also graduated with prestigious SAS data mining certification, offered at the undergraduate level exclusively at Bryant.
Lemieux praises his capstone mentor, Billie Anderson, Ph.D., assistant professor of statistics. “Although I was an undergraduate, she treated me as a peer and as a colleague,” he says. “She knew when to offer guidance, but also when to step back and let me make my own decisions.”
“What Christian did was graduate-level work,” says Anderson. “I pushed him beyond his comfort zone, but he is very motivated and loved the discovery process.”
Presenting at national conference
Anderson also praises Lemieux’s communication skills – highly prized by employers in technical fields. She should know: She worked in the SAS Research and Development Department for four years before coming to Bryant and still works closely with SAS to train Fortune 500 companies on how to successfully use SAS Enterprise Miner to build data mining models.
On Aug. 4 in Boston, Lemieux and Anderson will present “Finding Cost Effective Solutions to Healthcare Problems” during a poster session at the Joint Statistical Meetings, the largest annual gathering of mathematicians and statisticians in North America.