10 rules for communicating statistical concepts to a non-technical audience

1. Know your audience

Every effective talk needs to reach its audience, hold their attention, and keep them thinking. Even if the subject is the same, the talk you give at a research conference must be couched differently from the talk you give to a group of middle-school aged kids.

2. Know your subject fully

Simplicity proves mastery. Or in the words of some well-known experts:

• “If we can’t explain it to an undergraduate student then we don’t understand it fully.” Richard Feynman

• “You don’t know something well if you can’t explain it to a child.” Albert Einstein

Before you present your research at a conference, put your talk to the test. Try to explain the core of your presentation to an undergraduate student, your children, someone who has limited formal education, or someone who doesn’t work in your field.

You’ll know your talk is truly great if they all understand your delivery and are eager to learn more.

3. Use large font sizes

This rule seems simple, but it’s crucial for both teaching the material effectively and getting practical results from your presentation. Make sure all attendees can easily read your slides by using a minimum font size of 12pt. When you prioritize readability, you make it easy for your entire audience to pay attention and increase the likelihood that your talk will be successfully understood by everyone.

4. A picture is worth a thousand words

Inspire your use of pictures and diagrams by reading Edward Tufte’s beautiful book, The Visual Display of Quantitative Information.

Tufte describes Charles Minard’s brilliant map of Napoleon’s march on Moscow in 1812. The map provides a two-dimensional representation of six types of data: the number of troops, distance, temperature, latitude and longitude, direction of travel, and location.

5. Be lively and energetic throughout your talk

If your audience seems sleepy or bored, then the reason is almost certainly that you look sleepy or bored!

To keep energy levels high, link your statistical analysis continuously to its impact on daily life. Every good analysis has real-world consequences, and audiences will stay awake if you repeatedly describe the practical outcomes of your analysis.

My personal favorite trick is to relate the analysis results to something instantly engaging, like casino games. The excitement and the temptations of “free money” is usually guaranteed to keep audiences alert and attentive.

6. Use simple, plain, everyday language

There’s no need to show-off. You won the prize the moment you were invited to present the talk.

Strunk and White’s The Elements of Style and Hayakawa’s Language in Thought and Action give great advice on the use of simple, everyday language. Follow their lead!

7. Avoid compound sentences, redundant slide references, and overuse of statistical jargon

Assume that your audience’s attention span isn’t as wide as you would like it to be.

Jargon is often used when speakers to try to hide what they don’t know or understand. Even if jargon is necessary, explain what it means in basic, everyday language.

8. Reread the masters of statistical exposition

When you’re building a presentation, keep the work of the great statistical communicators handy and reread them.

My personal favorites include David Freedman’s Statistical Models: Theory and Practice (2^nd edition. 2009) and the textbook Statistics (4^th edition. 2007) by Freedman, Pisani, Purves. In my opinion, Freedman-Pisani-Purves’ book is the best undergraduate statistics textbook that will ever be written!

While reading the masters, you’ll find many ideas for how to tighten and improve your own talk. So, revise, revise, and revise again.

You can (almost) never be finished polishing your presentation. Make it a habit to wake up each morning and think about additional ways to make your talk clearer or more compelling.

9. Summarize your talk with a one-line description of each main point

Present the main points in simple, everyday language. Emphasize the power of statistical thinking in words that show how the statistical analysis affects daily lives.

10. The acid test of a great talk

How do you know if your talk was a success? The acid test is simple: there should be an animated discussion and lively questions from the audience at the end.

Whenever I receive questions that are unclear, I usually offer to meet with the questioner to discuss the question after the talk is over.

As for questions that I simply don’t know how to answer, I’ve always found it a winning strategy to say, “I don’t know the answer right now, but I’ll investigate it and be happy to write to you when I find the answer.” It’s honest and shows commitment to the topic.

In conclusion, statistical communication is an art, not a science, and it takes continual practice. But when you deliver it well, the benefits to the audience, to the statistician, and to the field are truly immense.