Date: Thursday, February 6, 2025 at 12:00 pm - 1:00 pm ET
The COPSS-NISS Leadership webinar focused on the importance of effective communication of Statistics to diverse audiences, with panelists Dr. Francesca Dominici (Clarence James Gamble Professor of Biostatistics, Population, and Data Science at Harvard University) and Sir David Spiegelhalter (Emeritus Professor of Statistics at the University of Cambridge) sharing their experiences and insights. The discussion covered the challenges and opportunities of communicating Statistics to diverse audiences, especially in a phase characterized by an increasing attention and debate on AI, along with the need for improved training and incentives for statisticians interested in public communication. The conversation ended with a video of David's participation in in the first (and only) series of Winter Wipeout on BBC1, humorously highlighting the challenges and his eventual elimination from the competition.
Daniele introduced the Committee of Presidents of Statistical Societies (COPSS) and its Emerging Leader Award, which recognizes early career statistical scientists with potential for leadership. The purpose of the award is to share the vision of leadership within the field and organize a webinar series. The topic of the webinar was leadership in the communication of Statistics, with a focus on the value of Statistics in an era of increasing attention on AI, and how to communicate this value effectively to diverse audiences. The panelists were Dr. Francesca Dominici, Clarence James Gamble Professor of Biostatistics at Harvard University, and Sir David Spiegelhalter, Emeritus Professor of Statistics at the University of Cambridge. Francesca briefly discussed her role in leading Data Science at Harvard and her efforts in communicating the value of Data Science to solve societal problems.
Data Science and Societal Impact
Francesca discussed the importance of Data Science in addressing societal problems and the AI revolution. She highlighted her work in providing data-driven studies to inform regulatory policy, particularly in the context of air pollution and climate change. She mentioned a study linking air pollution exposure to increased Covid-19 incidence and mortality, which led to changes in air quality standards in the US. Francesca also shared her ongoing work on the environmental impact of cryptocurrency and AI, including a Data Science pipeline to quantify the carbon footprint of AI. She emphasized the need for effective communication of Data Science findings, especially in the face of vested interests.
David's Science Communication Journey
David shared his experiences in science communication, highlighting his transition from a traditional academic role to a more public-facing one. He discussed his work on Covid-19 statistics, including a column he wrote with Anthony Masters for a major Sunday newspaper, and his involvement in public inquiries such as the Bristol public inquiry and the infected blood inquiry. David emphasized the importance of clear communication of Statistics and the challenges of presenting complex data and findings to the public. He also mentioned some of his recent books, including “The Art of Statistics”.
Transitioning to Communication and Media
David discussed his transition from a standard academic career in Statistics to one more focused on communication, particularly in the context of legal inquiries and media engagement. He emphasized the importance of effective communication, especially when dealing with sensitive topics and high-profile cases. David also shared his experiences with media interviews, highlighting the need for control and preparation. He expressed his enjoyment of writing articles and his belief in the value of communication skills in Statistics training. Lastly, he hinted at his experience in the Winter Wipeout event, offering to share more if there was interest. Daniele ended the conversation by inviting questions and preparing for a moderated Q&A session.
Statistics Communication and Data Science
Daniele initiated a discussion on the core aspects of Statistics that should be communicated as a community. Francesca emphasized the importance of Data Science as a foundation for both Statistics and AI, highlighting the need for good data quality and interpretability. David agreed, adding that Statistics is a complex and beautiful discipline that requires simplification, clarity, creativity and a well-designed storytelling when communicating. The group also discussed the importance of understanding the audience for effective communication.
Communicating Statistics to General Audience
David and Francesca discussed the challenges of communicating Statistics to a general audience, particularly on sensitive topics. They emphasized the importance of preparation, understanding the audience, and avoiding manipulation. Francesca shared her experience in engaging with journalists and the need to remain objective. She also stressed the importance of understanding the rules of this engagement and the possible unexpected challenges that may arise. David suggested preparing a rough script for interviews and not being distracted into areas outside one’s expertise, while also being open to adjusting the script based on the conversation. The team also discussed the importance of learning from mistakes and being aware of their own limitations as statisticians when communicating with the public.
Miscommunication and Large Language Models
Francesca shared her experience on the cascading effects that a minor mistake in communicating statistical findings to the media may have. David agreed, further stressing the importance of avoiding pedantry and using jargon, which can lead to misunderstandings. He suggested that using longer explanations and storytelling techniques could improve communication. Daniele then raised the topic of the impact of large language models on effective communication, questioning whether they would enhance or hinder creativity and diversity in communication.
AI's Role in the Communication of Statistics
David and Francesca discussed the potential of AI in their field, emphasizing the importance of a human filter to ensure accuracy and trustworthiness. They agreed that AI is a powerful tool that will continue to grow and evolve, but it should not replace human judgment and expertise. Francesca highlighted the need for better training and opportunities for statisticians that are interested in communicating their work to the general public. David echoed this sentiment, stressing the importance of empowering people to make better decisions through informed understanding. They also discussed the need for a higher profile for statisticians in society, not as policy advocates, but as experts providing evidence-based insights to improve general understanding and counter polarization.
Webinar Gratitude and Conclusions
Daniele expressed gratitude to Francesca and David for their insightful discussion. He thanked the Organizing Committee of the COPSS-NISS leadership Webinar series for the opportunity to organize and moderate the webinar. David then shared a video of his participation in the first (and only) series of Winter Wipeout on BBC1. He humorously recounted his experience, highlighting the challenges and his eventual elimination from the competition. The conversation ended with Daniele, Francesca and David expressing their appreciation for the session.
Key Takeaways from the Webinar:
1. Effective communications of Statistics requires preparation, simplification, clarity, creativity and a well-designed storytelling.
2. Establish more opportunities to support training in communication of Statistics.
3. Prepare scripts and key messages before media interviews, and set clear ground rules when engaging with journalists. Be prepared to politely refuse to answer irrelevant questions.
4. AI tools can be useful for summarization and drafting scripts for communication, but human oversight is required.
5. Promote statisticians as impartial, evidence-based voices in society.
6. It is important to strategically position statisticians alongside the AI advancements in data analysis and communication.
Thanks and Recognition
We sincerely thank our esteemed speakers, Francesca Dominici and David Spiegelhalter, for sharing their invaluable insights on the critical role of effective communication in Statistics. Their expertise illuminated strategies for engaging with the public, the media, and policymakers while maintaining clarity and accuracy. We also extend our gratitude to our moderator, Daniele Durante, for skillfully guiding the discussion and facilitating a thought-provoking conversation. This webinar highlighted the importance of developing communication skills within the statistical community and provided actionable steps for statisticians to enhance their impact. Thank you to all who attended and contributed to this engaging session!