Through the dark forest of data: Inside the Statistical Consulting Center
What do Russian language education, the epidemiology of Turkey, microbiology, engineering, and nutrition sciences have to do with each other? At first these may seem like discordant subjects, but for Noah Feldman, a statistics doctoral candidate, these topics represent individuals he has worked with as a graduate consultant at the Statistical Consulting Center (SCC).
The Statistical Consulting Center at Penn State is comprised of a team of faculty, staff, and graduate students in the Department of Statistics, providing statistical support to those within and beyond Penn State.
Maggie Niu, the director of the SCC shared, “my team can assist researchers at any stage of their project, offering guidance on research planning, experimental design, survey sampling, statistical modeling, and results interpretation.”
Graduate students involved in the SCC are introduced to the center via an educational pedagogical model. Each semester, students in a master of science or applied statistics program and doctoral students serve as consultants. Through lectures, observation of live sessions, and mock round consultations, students learn how to consult on statistical problems from a variety of disciplines. Those students then report on the project and their proposed solution during an in-class discussion, led by two research assistants and a professor. The group then collaborates to come up with the best solution that meets the client’s needs and available time scale, which will be provided to the consultee in a written report.
Working as a graduate consultant at the SCC, Feldman helps advise scientists from across the university on how they can navigate their data, identify a research question, and provide feedback on their analysis plan.
“Helping clients to traverse the dark forest of their data and seeing them light up when we reach a solution is my favorite part,” said Feldman.
Consulting times are available every weekday on a first come, first served basis, though individuals are limited to three slots each semester.
There are two main avenues for collaboration with the SCC — “Stat-on-Call” which provides short-term consulting, or “Long-Term Collaboration,” where those with more in-depth work exceeding two weeks can collaborate for a fee with faculty consultants.
“Whether you need quick statistical guidance or in-depth project support, researchers complete a request form and schedule your appointment by visiting the SCC website,” said Niu.
For those who are interested in accessing the SCC’s services, Feldman has three pieces of advice:
- Come before you collect your data, if you can. The saddest thing that happens in the SCC is a scientist who comes in and explains a fascinating problem, but their data isn't set up to support an investigation of it. We can often help make sure your experimental design is appropriate before you've cultured your first petri dish.
- Have your scientific question in mind. We're happy to build a statistical approach from the ground up, or to modify an existing framework, but we can't do anything without understanding your problem and why it's important.
- Handouts! Give us something to look at and markup. This can be a copy of a survey you collected, a paper you're replicating, or even a print of part of your data.
“It sounds cheesy, but I didn’t really understand how someone could love their work until I started consulting,” said Feldman. “Every new client presents a fascinating new puzzle, and their passion for their work only makes the process more exciting! I’m really glad that the SCC gives me an opportunity to get involved with so many different projects.”