Open Leadership in Data Science
Open Leadership in Data Science#
Every team, every project and every collaboration has people who participate in a way that inspires others in the team to join efforts and accomplish shared goals together. These people take initiative, guide others, ensure that diverse perspectives are involved, and that everyone has the right opportunity to build on each others skills and expertise.
This is called leadership and these people are leaders. Specifically, in the context of open source/science and community projects, we call this phenomenon ‘Open Leadership’. This is also applicable to data science, research communities, and collaborations more broadly.
In formal terms, leadership is defined as specific skills “encompassing the ability of an individual or organisation to lead or guide other individuals, teams, or entire organisations.” (source: Wikipedia). Open leaders might not necessarily be the same people who have formal authority or the title of institutional leaders. Nonetheless, open leaders work towards facilitating an environment that empowers others to collaborate equitably, identify shared challenges, develop shared goals, and address them with a shared sense of agency. In the long term, their work contributes to the success of projects, and institutional goals and hence should be rewarded with formal titles, recognitions, and positive culture change.
Open Leadership Framework#
The Turing Way embodies Open Leadership and has its origin in Mozilla’s Open Leadership Framework.
Mozilla’s Open Leadership Framework describes open leadership as “a set of principles, practices, and skills people can use to mobilise their communities to solve shared problems and achieve shared goals” (See Mozilla Open Leadership framework).
The Open Organisation Resources defines open leadership as a mindset and behaviors that anyone can learn and practice.
Their roles may not be permanent or predestined but emerge as needed for the task at hand. Open leaders uphold transparency, inclusivity, adaptability, collaboration, and community.
The Open Leadership Framework states that leadership is about involving members and mobilising communities to solve problems and achieve goals that are beneficial for everyone.
Individuals who lead their work (project, organisation, communication, resources) openly design different aspects of their work to promote openness and social good throughout their work life cycle.
The key principles of openness are: design, build, and empower for understanding, sharing, and participation and inclusion.
Design: You plan projects for contributors and users’ specific needs and capacities.
Build: You create systems and solutions that maximize a project’s clarity, usability, and inclusiveness.
Empower: You help contributors own the work through transparency, accountability, and shared decision-making.
The emphasis is on ensuring that the members of an open project understand the project so well that the project can run without leaders because all the members are empowered to take on leadership tasks.
The Turing Way is built upon Open Leadership principles, which involve and support its members to apply these principles when developing content for its book or supporting its community.
Understanding Leadership in Data Science#
Data science is broadly and inclusively considered as study of data to draw meaningful insights, contribute to the technological advances and benefit society by informing decisions across all sectors.
Theoretical frameworks from business schools, executive training and resources often explain, define, or assess leadership skills as traits such as “popularity, power, showmanship, or wisdom” which are not really the essence of leadership (Harvard Business Review). There is no such thing as a fixed set of leadership skills and not all great leaders have the same traits, strengths or leadership styles.
In the context of data science, where we strive to advance our knowledge through validated scientific methods based on data, talking about leadership, essentially a nuanced and fuzzy concept, is challenging to define. In data science and research, discussions and upskilling in leadership are extremely important as:
Researchers and data scientists constantly navigate fast-changing demands of ‘innovative’ approaches in technological development.
They manage and coordinate collaborations with their peers as well as external stakeholders across multiple disciplines.
Leadership is one of the toughest non-technical skills to learn without guidance.
Leadership is rarely discussed openly in research environments, despite being everywhere.
(Open) Leadership is one of the keys to healthy and inclusive communities that we are striving to build in data science.
In this chapter, we explore different aspects of (open) leadership skills that are applicable in the data science and research contexts. After all, “technical skills are just one aspect of making data science research open, reproducible, inclusive and ethical for all” (as stated in the Welcome page of The Turing Way).
We hope that by discussing leadership here, we will:
Spotlight on important points that everyone in research should know
Inspire those who think leadership is not for them to review their assumptions (and maybe see themselves with abilities to lead a project).
Ignite reflection and spark conversations about the leaderships we want to promote in data science.
Evaluate if leadership in our workplace or communities are healthy, compassionate, and inclusive, and how we can improve it.
In this chapter, we will introduce and discuss:
important features of leadership when experiencing and exercising them in data science,
what we mean by “healthy” leadership, how we can build open and healthy leadership approaches and what they should avoid/address,
how we can create leadership opportunities in our projects,
finally, our community members provide examples of open leadership from their own professions and experiences. We invite you, the diverse group of leaders in The Turing Way (readers, users, contributors, and community members) to provide more examples from your contexts in this chapter.