Wrap Your Naked Statistics in a Warm Blanket of Meaning
June 19, 2008 by John Hewitt
Article By Andrew Dlugan
Including facts and statistics in your article lends credibility to your assertions and grounds them in reality. Quoting a statistic from a credible source means that your arguments are no longer just your
arguments: you stand united with experts.
However, a naked statistic - one provided without any meaningful context - leads to confusion rather than clarity. Numbers are often too large to grasp by themselves. Unless your audience are experts in the field, they won’t be able to intelligently interpret the statistic. You might get a momentary “wow” factor for a big number, but it won’t be memorable.
For your statistic to penetrate more, you must provide a useful context for the numbers, particularly if that context makes the numbers personal for your audience. Let’s consider a couple of examples to illustrate this.
Suppose you are writing an article about cancer deaths to encourage more research funding.
- The assertion: Many people die from cancer every year.
- The naked statistic: It is estimated that 565650 people in the United States will die from cancer in 2008.
- Statistic with context: It is estimated that 565650 people in the United States will die from cancer in 2008. That’s about 1 of every 532 citizens.
- Statistic with context made relevant to audience: It is estimated that 565650 people in the United States will die from cancer in 2008. That’s about 1 of every 532 citizens. You are among 2000 visitors to Poewar each day. Assuming these visitors are representative of the general population, statistics tell us that 4 of these visitors will die from cancer this year.
Source: American Cancer Society
Let’s tackle another example. Suppose you are writing an article about the need for youth mentors.
- The assertion: There is a dire need for mentors to help youth make smart life decisions.
- The naked statistic: Across the United States, 14.6 million children need mentors.
- Statistic with context: Across the United States, 14.6 million children need mentors. This is 1 of every 2 children in the country.
- Statistic with context made relevant to audience: Across the United States, 14.6 million children need mentors. This is 1 of every 2 children in the country. That’s 15 of the 30 students in your son’s class. That’s 8 of the 16 children invited to your daughter’s last birthday party.
Source: Mentoring.org
A well-placed statistic will enhance your article, but only if you wrap it with a meaningful context that relates to your audience.
———————————
Andrew Dlugan is an award-winning public speaker and speech evaluator, and the author of the Six Minutes Public Speaking Blog. He’s also a mentor and a former cancer researcher.
Andrew,
Thanks for writing about a subject we don’t often read about or think about. Often we just throw out numbers without relevant context, which is, as you said so well, what makes the statistics meaningful.
Andrew, I’ve been looking forward to your post. In my limited time in the blogosphere, this is the first time I’ve read a post on using numbers and statistics in writing. You make excellent suggestions for providing context for those numbers and statistics.
While using numbers and statistics can bolster your argument, misusing them can compromise credibility as well as create confusion. For several years I worked in public health at a state government agency. We wrote lots of reports, heavily laden with statistics, and we were continually challenged to present those statistics in a way that would be comprehensible to the average citizen (or, dumbing down even more, to state legislators :-))
You provide a great example with cancer research data. The simple number of how many people may die from cancer in any given year is meaningless without knowing the size of the population you are talking about. Putting the number in a context that your audience can understand goes a long way toward educating people, rather than unduly alarming them.
Writing with numbers and statistics is a real challenge. Even one of your examples of context–the 2000 visitors to Poewar–has to be approached cautiously because those 2000 visitors may not be representative of the general population. There are a lot more tricky issues with using statistics: with cancer data, incidence and prevalence are often conflated; survey sample statistics are actually estimates that have to be analyzed for statistical significance … and then the reader’s eyes glaze over.
Thanks for the great post, Andrew!
Marie Ann Baileys last blog post..Elderbloggers
Thanks for the comments, Lillie and Marie.
Instead of viewing this as “dumbing down” the statistics, I prefer to look at it from the perspective of lifting your audience up to understand it by making it meaningful to them in their world.
Providing context for statistics is just one aspect of a three-part challenge when citing statistics.
1. Make sure the statistic is valid. This is really the responsibility of the people who are conducting the study or producing the statistic. Were proper protocols followed? Were systematic sources of bias eliminated (or accounted for)? Were appropriate steps taken to assure data integrity?
2. Make sure you aren’t deceiving the audience by misuse of the statistic. The old cliche “Lies, damn lies, and statistics” addresses this point. It is common for marketing departments to take this approach to serve their agenda. The hope is that the audience won’t notice, but this places your reputation on shaky ground.
3. Make sure the statistic is understandable. This is the point which the article above addresses.
Andrews last blog post..Wrap Your Naked Statistics in a Warm Blanket of Meaning
[...] caveat that this population may not be representative of the larger general population). Click here to read his full post. [...]
Andrew, your points are well-taken. The primary purpose of the reports I used to write was to educate the general public on specific public health concerns. Much of our data was based on surveys, so we had the additional challenge of explaining how two statistics that looked different could in fact not be different at all (overlapping confidence intervals, yikes!). An even greater challenge was delivering this information to state legislators who, I am sorry to say, often did want to misuse the statistics to support whatever platform they happened to be on at the time. It was very frustrating and once a report left our hands, we could not always control how it was represented to a particular constituency. I will say, however, we never published a report that we couldn’t stand behind.
OK, I’ll get off my soapbox now
Thanks again for your post, Andrew.
Marie Ann Baileys last blog post..Writing With Statistics and Numbers
I’ve just come across an article from Daphne Gray-Grant on the topic of putting numbers in context: Do you have your readers’ numbers?
Andrews last blog post..Wrap Your Naked Statistics in a Warm Blanket of Meaning
Andrew, good post. As an adjunct and graduate student, I am all too aware of statistics being carelessly used. I think a variety of people in higher education could benefit from reading your post.
Morgans last blog post..Nature, Interconnectivity, and Community
Morgan:
Thanks for your comment. When I was a graduate student, it became painfully clear that making statistics relevant and understandable is a huge challenge for a considerable fraction of those in academia.
In scientific papers, you can usually get away without paying too much attention to providing a context for statistics because the audience is generally a highly focused niche of experts. But when writing or giving a presentation for a broader audience, there is a greater need to “wrap up” those statistics.
Too often, the focus is on “deliver as many results/numbers/statistics as possible” when the focus should be on selecting the most important values and conveying those in a way that makes them memorable. A phrase I would like to read/hear more often is “… which means that… ” to link the statistic to the context/explanation.
Andrews last blog post..Wrap Your Naked Statistics in a Warm Blanket of Meaning
Andrew, thanks for the sharing the link to Daphne Gray-Grant’s article. She, like you, offer some excellent ideas for making numbers (especially very big numbers) meaningful for everyone.
Marie Ann Baileys last blog post..Let a Newbie Stimulate Your Creativity
Andrew, thank you for the insightful post on what some people would consider dry and dull: numbers! In the medical editing that I do, which is not usually for a lay audience, there are many instances of these sort of dead statistics to which you refer. They’re not really put in context; they’re simply stated and left alone, which is usually fine for this audience.
With a knowledge of statistics, though, a person would be well equipped to go into scientific or medical editing–not to mention writing for these genres. I think the warm blanket aspect is moreso necessary for the lay audience, as you allude, to really bring the numbers home to them (eg, in a fundraising piece or other call-to-action). Newspaper (or magazine) reporters would do well to have at least a basic grasp of it, too!
Leighs last blog post..The Music of Words and Other Matters
Leigh:
Thanks for your comments. You touch on two very important situations.
Fundraising: I think the primary goal of those who run the fundraising campaign is to give meaning to the statistics. Statistics alone may get headlines, but people are more apt to open their wallet when the connection between the stats and their personal situation is made clear.
Reporters: I agree with you that a wise reporter should acquire some basic statistical knowledge, particularly when reporting on products and services. In my view, the ideal reporter is one who can see through the marketing blanket (provided by the vendor), and translate it into a more objective context. A company may say that product A makes my life happier, but is there really something of substance there? To be fair, the definition of “reporters” here could be extended to include both mass media as well as independent media (including bloggers).
Andrews last blog post..Best Public Speaking Articles: Weekly Review [2008-06-21]
Andrew,
Great points! Statistics can often be downright boring — or at the very least lacking in real impact — when not presented in a manner that helps the reader relate them to his or her experience.
You’ve presented some very practical advice in this post that every writer would do well to remember.
Thanks!
Jeanne
Good point! For me though, there’s an extra point to make… it’s not just about presenting the statistic it’s about presenting it in a way that an audience can cope with in a presentation. (I’m a presentation skills trainer and I often have to help people present statistica/technical stuff.)
The examples you give above are great - but sadly I’ve seen them taken more than a bit too literally in presentations and the *whole sentence* up on the screen (in a bad font and too small, to add insult to injury!). As a presenter you should *say* your statistic with a relevant context but if you’re using PowerPoint etc you should *show* something interesting - not the words.
Use an image to illustrate your point nicely - provide the context - but not to simply quote the stats. Pleeaaaasseeee!
S