Victoria Gamerman, Boehringer Ingelheim; John E. Kolassa, Rutgers, The State University of New Jersey; Jim Z. Li, Viatris; Fanni Natanegara, Eli Lilly and Company; Kimberly F. Sellers, Georgetown University; Aniketh Talwai, Medidata; Kelly H. Zou, Viatris
During the 2021 International Chinese Statistical Association (ICSA) Applied Statistics Symposium, several panelists discussed the key elements of forming and sustaining successful partnerships and collaborations, along with the challenges and barriers. This is Part 1 of a two-part panel that includes six experts from either academia, industry, or consulting, with Kelly Zou serving as the moderator. Look for Part 2 in the February issue.
Partnerships and collaborations come in all shapes. As Martin Luther King Jr. said, “We may have all come on different ships, but we’re in the same boat now.”
Frequently, sharing ideas between stakeholders from different organizations leads to exchange visits, support for graduate students, consulting jobs, grant support, and continuing education opportunities for statisticians or data scientists outside academe.
For statistics and data science to be impactful, sharing ideas and knowledge is key. The themes discussed here apply broadly, with deep implications for the impact and perception of our field.
How would you define “partnership,” since it may be of many shapes?
According to Fanni Natanegara of Eli Lilly and Company, partnership is an association between two or more individuals who pool their resources and skills to achieve a common goal. The partnerships she has seen in her role as a pharmaceutical statistician and ASA Statistical Partnerships Among Academe, Industry, and Government (SPAIG) Committee member have taken many forms. Formal partnerships can involve signing a contract such as a “master service agreement,” where academicians partner with industry statisticians to solve a research problem that could end up in method and tool development, manuscripts, presentations, and dissertation topics. This type of partnership could also involve monetary transactions. Informal partnerships may not necessarily include monetary exchange and can still produce manuscripts and presentations.
Is “collaboration” the same as “partnership”? Why?
John E. Kolassa of Rutgers sees “collaboration” and “partnership” as separate concepts on a spectrum of collaboration intensity. At one end of this spectrum is partnership, where the statistician’s name is on every publication, the statistician contributes to the grant application by planning the analysis and details of data collection, and the statistician is involved all the way through to the final publications. At the other end of the spectrum, a colleague describes a type of data and solicits an opinion on data analysis, experimental pitfalls, and other aspects to provide valid scientific reports. This may be considered as “collaboration,” which is less intense. It is important to keep in mind that many lab science collaborators treat their studies with a more proprietary spirit and recognizing this spirit helps avoid conflict.
What are benefits of collaborations and partnerships in statistics/data science?
Victoria Gamerman of Boehringer Ingelheim noted data science is a varied field that requires multifaceted roles. For successful collaborations and partnerships, there are different types of roles to consider and balance, depending on the desired outcome of the relationships: specialists and generalists. The specialization concept is key to finding which collaborator or partner has subject matter expertise on which topic or topics. This expertise strength from one partner should then be paired with another who has a strength in a different and complementary area.
These partnerships and collaborations also allow us to look outside our own area of expertise and way of thinking to focus on the outcome in the short term, with the perspectives of how to scale it either as part of the relationship or beyond. This ability to move from use case or single problem solution to generating sustainable solutions that are scalable within one of the partner organizations is where the benefit of these coming together moments will be seen and valued by all parties.
Through this approach of finding complementary expertise, bringing together diverse approaches, and valuing short- and long-term outcomes, the pressure of finding a single candidate who is a unicorn capable of ‘doing it all’ is not necessary. Instead, organizations can identify external experts related to its needs, work with these specialists through the partner organization, and achieve the desired goals with broader benefit.
What are some requirements for fruitful collaborations or partnerships?
Gamerman emphasized that a key element for the parties involved in the collaboration or partnership is understanding each other’s drivers. For example, one organization may emphasize getting regulatory-grade evidence to support a molecule through development while another organization may place an emphasis on external scientific publications related to the disease area in which the molecule works. Different motivators and drivers have their place, and both need to be taken into consideration when defining what a successful collaboration or partnership looks like. Having an aligned understanding of what is important to the other parties will allow for an early and proactive assessment of potential challenges (e.g., which priorities to pursue) and define an approach for handling them (e.g., a governance or steering committee).
Aniketh Talwai of Medidata suggests the following:
- Establishing formal stakeholder management mechanisms (e.g., governance committees; go vs. no-go; and gating sessions, project charters, and contractors) ahead of launch
- Adhering to these mechanisms over the course of the effort to provide transparency and lend clarity to scope, contributions, permissions, and resourcing. Such an action works to prevent misunderstandings and helps resolve issues before they escalate into conflicts.
- Adopting a modular, stepwise approach to project delivery, taking quick wins up front and having proof-of-success at each step, as opposed to trying to solve all the hardest challenges or complete everything perfectly in one go, not only provides for the learning curve needed for all new collaborations but also helps guard against disillusionment leading to premature abandonment of the effort.
- Having the collective periodically step back and consider the interests of all stakeholders, not just those of the immediate partners, and proactively involve them helps to bring in a diversity of viewpoints and forestall potential constraints.
Don’t miss Part 2 in the February issue of Amstat News. The panelists offer communication tips and best practices for navigating multidisciplinary collaborations.
Editor’s Note: The views expressed here are the authors’ and do not necessarily represent those of their employers.
Leave a Reply