As our University of Virginia (UVA) School of Data Science (SDS) grows by around 10-15 faculty per year, the question quickly becomes how to best organize ourselves? Putting the numbers aside for a moment, what are the organizational principles we should consider? On one hand, it is important for faculty to feel they belong to a group of colleagues that share their scholarly interests. On the other hand, while departments do that to some extent, they also create silos that work against collaboration as turf and funding battles creep in. Further, data science is particularly complex to organize in that it spans traditional disciplines and hence traditional departments. So how to move forward?

One way to start is by asking the question: What is the problem we are trying to solve? I see that problem as multifaceted as follows:

• Remove organizational barriers to foster collaboration
• Maximize communication across all faculty
• Reinforce faculty’s sense of scholarly worth 
• Provide labels/branding that have meaning across academia; For example, on the institutional affiliation found in papers, thereby providing the recognition needed when a faculty member seeks a position elsewhere
• Foster faculty equity and inclusion as part of an overall scholarly enterprise
• Minimize organizational hierarchies.

What organizational structure achieves all of the above? The answer is probably none, so what would be the most effective? 

One approach would be around how we at SDS define data science as 4 domains – Systems, Design, Analytics and Value, all applied to Practice across many disciplines (+1) – our so-called 4+1 model. Faculty would certainly align with one of those domains, but most would align with more than one. A few may claim to align with all four. This is not surprising since we actively seek faculty with a broad perspective. Then, there are faculty who have strong roots in one Practice, like myself and biomedicine.

Since Practice (the +1) is so broad, our school has chosen to focus, at least from a research perspective, on a subset of possible research areas – biomedicine, environment, democracy, business, digital humanities, cybersecurity. These research areas might align better with traditional disciplines, as broad as they are, but then we would lose the sense of expertise around the core underpinnings of data science exemplified by the “4.” 

The way forward would seem to be a rigorous debate as part of our shared governance model and determine if a consensus arises. We have begun that process through a new 5-year strategic plan which calls out the need for scholarly organization. Such organization will also influence our future hiring and presumably lead to a coalesced set of activities for which we can become national and international leaders in data science.

Concurrent with our broader strategic planning, we are also seeking grass roots get-togethers of folks around areas of interest. Let’s call them special interest groups. Perhaps an overall organization will be to some extent self-defining, although I can imagine some inside and outside our school would find that unworkable. 

Beyond the “it” – whether “it” be a Domain, a research area described by our definition of Practice, a special interest group, or something else – how to lead and o allocate resources appropriately? 

We are fortunate enough to have a number of unfilled endowed chairs that can be used as part of the process. We have already made a commitment to hire a Chair of Responsible Data Science, which is one example. These chairs are intended for leaders who will build out the enterprise by encouraging faculty to join in building out scholarship in common areas of interest. We also have visiting fellowships that enrich us and can be used to pollinate a variety of activities and try out new ideas without a tenure-level commitment. This does represent somewhat of a chicken and egg situation as by committing to such leaders and fellows we are to some extent defining our future organization. 

So, the first task would be to hold a rigorous debate by those already in the school as to how to organize ourselves and at the same time review what our peer institutions are doing. Stay tuned on where this leads us.