My View of Presenting Statistics

Sep. 6, 2007

Madam Toastmaster, fellow toastmasters, and honored guests:

The other day, I was reading ahead on the Toastmasters’ Competent Communication Manual. I was reading Project #7: Research Your Topic. I got charged up when I was on page 38. I thought this little nerd just got himself an opportunity to challenge the manual. Let me read it to you first, and then I’ll tell you why I take issues. This is what it says:

If your support material is complex or involves statistics, find the human interest side and include that information in your speech. For example, if your research shows that 20 percent of adults read at a 5th-grade level or lower, make this statistic more meaningful to your audience by saying, “One of five adults in this room reads at or below the 5th-grade level.”

I appreciate the advice: Find the human interest side and present it in a less nerdy way. After all, if the audience cannot understand you or fall asleep, then what’s the point of talking at all? However, as soon as the example was given, the speaker lost all credibility with me.
Why? Because it is simply wrong. Allow me to explain my view point:

Let’s assume this study is about the whole country. For a study to conclude that 20% of the adults read at or below 5th-grade level, it must have surveyed thousands of adults across the country; perhaps even more. The survey must include many different cities, suburban, and rural areas. It must include adults in all walks of life—farmers, accountants, professors, waitresses, etc.
<If you would pick up the hand-out.>

The triangles represent adults who read at or below 5th grade level; the small circles represent adults who read at a higher level. The big circle represents all the adults surveyed: out of the 100 adults surveyed there are 20 people who read at or below 5th grade level and 80 adults who read at a higher level. If the surveys were conducted correctly, it is appropriate to claim that 20% of the adults in the country read at or below 5th grade level. However, if you look at the smaller areas marked as A, B, C, or D this claim falls apart. Say, each of these areas represents a roomful of adults such as this room. As you can see, not all rooms are created equal—Room A has 10% triangles; room B has 50%; Room C has 0%. I am inclined to say, in this room is more like C than A or B.

What went wrong? Two things went wrong:

First: The speaker did a research on a large population and applied it to a much smaller population. This is not hard to demonstrate that it is inappropriate:
You flip a coin 1,000 times and get the conclusion that 50% of the time it will land on the tail; 50%, on the head. You then tell the audience to flip a coin 4 times and expect 2 heads and 2 tails. It won’t happen that way.

Second: The speaker did a research on a population and applied it to a population with different characteristics. The population surveyed includes many walks of life and different education levels. It cannot be true applying to a room of toastmasters and guests who, perhaps, are interested in becoming toastmasters themselves.

Why am I bent out of shape?

It is because I believe a responsible speaker should not put the sizzle ahead of the beef. A responsible speaker must always speak the truth, the whole truth and nothing but the truth. Only after the truth is conveyed correctly, clearly and succinctly can he/she worry about presenting it interestingly. I think compromising the correctness in favor of making a speech more understandable is, at best, putting the cart before the horse. At worst, it misleads the audience and does the audience a disservice. I believe this view is not nerdy. It is a must. My fellow toastmasters and honored guests, do you agree?

Madam Toastmaster.