Stories by the Numbers
“She was lost in her longing to understand.”
—Gabriel Garcia Marquez, Love in the Time of Cholera
There is nothing more beautiful than a well-turned phrase. Words conjure up deep feelings, unveil motives, and make the obscure explicable. But numbers can also reveal meaning in what we long to understand. Just ask Statistics in the Community (STATCOM), a student-founded and -run group supported and endorsed by the School of Statistics in the College of Liberal Arts.
Through STATCOM, Masters and Ph.D. students provide free statistical consulting services to local non-profit, governmental, and community-service organizations. The students get the benefits of tackling and solving actual problems on the ground and the clients gain information that helps them maximize their resources, run more efficient programs, and evaluate the effectiveness of a program or campaign.
A great example of STACOM’s work in the community is the Nice Ride MN bike sharing program. The model is simple. One essentially rents a bike when and where it is needed and returns it to any of the 190 stations in the system. But Nice Ride had a couple of issues they wanted to address. Enter STATCOM.
First, Nice Ride wanted to learn more about walk-up users. Registered Nice Ride members are pretty well known, but walk-up users are enigmas. And of course, the more businesses know about all of their customers, the better able they are to anticipate and meet their needs. Second, Nice Ride needed to address the conundrum of supply and demand. For example, one of the most used bike stations is at the IDS center in Minneapolis. But if all those bikes are rented in the morning and returned to other kiosks across the cities, there aren’t enough bikes at a site where there is heavy demand. Expand that availability challenge across the cities and there is a huge logistical problem. STATCOM is currently working on both issues. If STATCOM can develop a predictive model to figure out where Nice Ride needs to move bikes to prevent stations from becoming full or empty, they improve access to bikes and efficiency. As far as is known right now, Nice Ride will be the first in the nation to have a predictive model of this sort if STATCOM is successful.
“That’s our biggest expense, moving bikes around the city,” said Mitch Vars, IT director of Nice Ride. “These efforts have the potential to tell us much more about our customers and save us money.”
How does all this look as a practical matter? Well, according to the STATCOM team, messy. But that’s a good thing. Turns out crunching text book data is nothing like meeting up with data that’s not dressed up for the date with a statistician. Text book data is neat and easy and compliant. Real-world data is the cracked sibling who has to be talked down and then coaxed into cooperating.
“The most challenging part is figuring out how to change the question that the client is asking into an equation that can be answered using a statistical model,” said Yin-Ting Chou, a Master’s student who also sits on the STATCOM board. And the effort to drill down to just the right statistical question may take a few attempts
Compassionate Action for Animals (CAA), a Minneapolis nonprofit that educates the public about farm animal suffering and works to promote a vegan diet, wanted to determine whether its annual Veg Fest moved the needle on their stated goals. It took a bit of back and forth between the STATCOM team and CAA to hone the questions into ones that could be analyzed to gain more specific and refined knowledge. What CAA learned, according to communications and events coordinator Justin Leaf, is that their festival and public education efforts are impactful. But when STATCOM presented their findings to CAA, the team discovered another challenge: Statisticians obviously can read statistical equations and their conclusions. We regular people generally cannot.
Founder of the STATCOM chapter at the U and team lead on the CAA project, Brandon Whited, a Master’s student about to graduate puts it succinctly. “Statistics answer questions in a really scientific way, and not necessarily in the way people think.”
Added STATCOM board member Yun (Erin) You: “Non-statisticians like visuals, and not necessarily graphs,” she said with a laugh.
Professor Dennis Cook, director of the School of Statistics, is incredibly proud of STATCOM. “We’re trying to improve and expand real-world student experiences and STATCOM certainly does that. My role is to provide guidance and advice and I’ve called upon our entire faculty to do the same.” Faculty act as advisors and mentors, but the students lead the initiative.
STATCOM is working to become an interdisciplinary program in the future. Combining the obvious talents and insights of these statisticians with other student-led groups can only augur well for community members who need more clarity to understand their own stories.