The Human Side of Statistical Leadership

Thomas M. Menino Convention & Exhibition Center 

Room: CC-258B 

More than 40 years ago, James Boen and Douglas Zahn published the book The Human Side of Statistical Consulting. It served as a reminder to (consulting) statisticians that as much as we often prefer to focus on data, models, and code, there are a significant number of interpersonal skills necessary to consulting that require consideration and development for success. Over time, there has been greater emphasis on developing leadership and communication skills among statisticians within educational departments, as well as professional and scientific organizations. For example, many statistical and data science departments include courses in consulting and leadership, the American Statistical Association (ASA) offers the StatsForward Fellowship to develop leadership skills in young professionals, and the Leadership in Practice Committee (LiPCom) of the Biopharmaceutical Section develops content to promote and train our next generation of statisticians. Statisticians and data scientists know firsthand that every decision, even seemingly innocuous ones, can have major implications and consequences for all downstream activities, including the statistical analysis and subsequent communication of its findings. It is incumbent on us to ensure that statistical and data concerns are considered and addressed.
In recent years, there have been numerous methodological advances leveraging machine learning (ML) and artificial intelligence (AI), fundamentally changing the academic and professional environments in how we learn, work, and communicate. There has also been a necessary call to action among statisticians and data scientists, not only to understand the technical aspects of AI/ML methodologies, but to engage in strategic leadership to understand and communicate the benefits and limitations of AI/ML to maintain the rigor, reliability, and reproducibility of research. As a result of these efforts, there is a growing understanding that embracing AI suggestions and solutions at face value without sense checking, verification, or formal adjudication may perpetuate erroneous data and bad science. While there are benefits to AI/ML for efficiency, reducing the burden of trivial tasks, and implementing data-driven decisions, even in the context of strategic leadership, there are concerns that are perhaps less appreciated by the masses: that our reliance (and potential dependence) on AI tools can lead to diminished cognitive skills and leadership capabilities. Furthermore, aptitudes unique to human beings such as emotional intelligence, creativity, ethical reasoning, adaptability, and interpersonal communication are at risk of atrophy. There are additional concerns for AI/ML over the transparency in how decisions are made, accountability, responsibility, and ethics in decision making, all of which are important for effective leadership.
This panel discussion focuses on the growing concern and potential impacts of AI dependency and its potential effects on leadership and interpersonal skills from the perspectives of statisticians and data scientists. More importantly, we explore possible solutions to maintain the human side of statistical leadership.

Yes

Biopharmaceutical Section

Caucus for Women in Statistics and Data Science

Committee on Career Development

Professional Issues and Visibility Council