Working with your Mitacs Business Development Director
Mitacs offers numerous funding opportunities to students and faculty in all academic disciplines, including mathematical disciplines like ours. These opportunities are a great resource for research supervisors looking to expand students’ --and their own-- exposure to industrial research questions, and allow supervisors to extend their student funding potential. However, from my experience, few members of the statistics and actuarial science communities take advantage of these opportunities. Many researchers cite that finding an industry partner is the biggest challenge in engaging in this type of research (CANSSI IIC (2016). “Results from CANSSI Industrial Innovation Committee Survey on Mitacs Participation.” Liaison, 30:4, pp37-39.).
Mitacs understands this, and has business development directors (BDs) across Canada that can assist with the matchmaking process. Part of the BD’s job is to identify industrial projects and sponsors that can be developed into fruitful partnerships with academia. They serve as a focal point for organizations that are looking to tap into the intellectual resources at a university, and therefore BDs often accumulate a variety of projects whose sponsors seek academic expertise. As a result, a BD can be a terrific resource for a faculty member or a department looking to branch out or offer on-the-job training to students.
When a new Mitacs BD, Allison Brennan, arrived at SFU, I first met her through a mutual acquaintance at a coffee shop. We chatted briefly about my department’s low uptake of Mitacs opportunities and she admitted that, with a PhD in Psychology, the difference wasn’t always clear to her between a statistical research question and research requiring statistical analysis. Many of the companies with whom she had spoken were convinced they needed a statistical scientist, but Allison thought that the problems they described might also fit with other disciplines including mathematics, computing science, psychology, sociology, or education to name a few. We set up a meeting to discuss it further.
At that meeting, I described how statisticians view the continuum between statistical consulting and statistical research. We discussed the range from the application of simple tools to a data set, through novel applications of existing tools to complex data and development of new models, and into full-on methodological development. I explained that our internal and external evaluation systems tend to focus heavily on publications in the last category, and hence that kind of project might pique a statistician’s interest more. However, we agreed that industrial partners may not be able to judge when this would happen --indeed, we cannot always know this until we are deep into a project ourselves! Finally, I explained the different research areas in our discipline, and that we are all specialists in some methods or application areas but may know little about others.
At the end of our conversation, Allison and I agreed that it would be helpful to share some potential industry project proposals, so we could discuss their merit as statistical sciences research and develop shared understandings. The first two proposals clearly fell under data science or predictive analytics and were not considered suitable statistical research. The projects also required more computing skill than statistics students typically have, but also more statistical skill than is typical of computer science students. (I note that some industry partners may not be fully aware of the skill sets that statistics students possess. They may ask for students from other disciplines like computer science when our students might serve them at least as well.)
I gave her my detailed perspectives on the proposals, differentiating between statistics work --analyzing the data to find answers-- and computing science work --extracting and preparing the data and building digital tools. In one case, I suggested that the partner might split their project in two so that students from different departments could be recruited to work together.
Although these specific projects were not suitable for statisticians, I considered the first discussions with Allison to be a success. A few months later, Allison sent another proposal for me to review. This project was more interesting and asked for an approach that lay somewhere between an advanced statistical application and applied statistics research. It definitely had the potential to be a research collaboration for a person with the right interests.
Following from these interactions, Allison tells me that she has a stronger sense of the continuum between statistical research and statistical consulting. In my department, she is now positioned as an asset to anyone interested in finding projects and developing collaborations outside academia. She is also better able to advise the companies and community organizations seeking academic research expertise when it could involve statistical research.
If you are interested in helping to improve your, or your department’s, chances of finding non-academic research collaborations through Mitacs connections, I would suggest having a similar discussion with your local business development director about your specific needs and interests. One could offer to look at available project proposals and to pass them along to colleagues. Allison assures me that doing so could be a tremendous help to increase the chance that the projects she brings forth may be of interest to SFU statistical and actuarial scientists.
For me, the collaboration with Allison has been pleasant and reviewing the projects did not take too much time. As a bonus, I got to learn more about the potential collaborations that industry partners seek through Mitacs so I can position myself and my students accordingly should I start taking advantage of these opportunities.
Tom Loughin, Simon Fraser University