Numbers Don't Trick People, People Trick People

I find numbers to be wonderful things.  They are straightforward, inarguable, and don't have a bias.  Words can be used and abused, as honest as the person saying them.  But numbers tell what is or isn't there.

Or, do they?

If I have learned one thing from all of my science training, it is how to use numbers.  They are fantastic things.  Say you have discovered a new thing called apples that may be able to limit people's exposure to doctors.  So, you start conducting an experiment to see if an apple a day keeps the doctor away.  You get 20 people and divide them into two groups.  Everyone will be exposed to doctor attractant, and then 10 of those people will be given an apple everyday for 10 days.  The experiment is completed, and the results are in.  Of the 10 control group people, all attracted doctors.

Now, if of those who got an apple a day, nine didn't attract a doctor, fantastic! You can now market apples as great doctor repellants.

But, if only one person out of ten didn't attract a doctor, yikes!  Those aren't very good results.  You've used so much time and resources in developing these apples, and now the numbers don't support you.  No one can argue with the numbers.

I think you know where this is going.

Again, statistics are a very reliable way of supporting your views.  For every paper that I write, I try to go to the government sites or research facilities and get some good numbers for support.  I DO find numbers to be wonderful things.

I just want to make people aware that despite numbers' good intentions, they can still be used in a misleading way.

Here is one example...5 Hour Energy

The set up:

5HE has a commercial out talking about how 73% of 3,000 doctors would recommend 5HE.  They end with something along the lines of "So ask your doctor. We've already asked 3,000."

The BREAKDOWN:

First, THE WORDS.  

The doctors aren't recommending 5HE to just anyone.  In fact, they aren't recommending 5HE, only a "low calorie energy supplement," and only to their healthy patients already taking an energy supplement.  That means that these doctors see that they have no choice in if an energy supplement is taken or not.  The choice is between a low or high calorie energy supplement.  Well, duh!  Of course the lower calorie one would be recommended more than a high calorie one.  Words, people, words.

Next, THE NUMBERS.

73% of the reviewed surveys would recommend a low calorie energy supplement.  That's not 73% of the 3,000 doctors; it is 73% of those who actually turned in their surveys.  But, best case scenario, 2,190 doctors would recommend a low calorie energy supplement to their healthy patients already on one.

Now, thanks to DVR, I was able to pause the commercial to read the essay of fine print.  Here we go:

"Of the 73% of primary care physicians who would recommend a low calorie energy supplement to their healthy patients who use energy supplements, 56% would specifically recommend 5HE for their healthy patients who use energy supplements." 

Of the best case scenario 2,190 doctors who would recommend a low calorie energy supplement, only 1,226 would specifically recommend 5HE.  That turns into 40.8% actually recommending 5HE.  We just went from having less than 30% of doctors NOT recommending a "low calorie energy supplement" to just above 30% actually recommending 5HE.

And remember, these numbers are from the received surveys, not from the 3,000 doctors.  Goodness, they could have only received 100 surveys back, meaning that only 40.8 doctors recommended 5HE in a nation of 79,831 practicing primary care family practice physicians.

I'm not trying to get you to distrust everything that is presented to you.  Just remember to pay attention to the words around the numbers.  70% is a high percentage, but of what in what context?  70% will survive from drinking 3 low calorie energy drinks who aren't healthy, or 70% will develop a stomach ache from eating too many apples?