O722 - Safeguards in the Publicity Use of Vital Statistics

Not specifically about vaccines, but interesting nonetheless, because vital statistics are often used (or misused) in support of (or opposition to) vaccines. Also the author was the one who wrote the article for post 060 about measles attack rates.

Anyway, it's a good lesson and a good reminder about how it is difficult to convey accurate information with statistics, especially if you have some personal interest in what is being conveyed.

"To reason correctly from such complex data is not easy, and even the most acute and careful persons will at times fall into error in the process. But no confusion can be quite so profound as that of the individual who rushes into this maze with the intention of finding support of a preconceived idea."

Hedrich gives several examples of how statistics can go wrong that are worth checking out. It reminds me a bit of the book How to Lie with Statistics, though not quite as in-depth or entertaining.

But for we who want to examine the truth of a matter without bias, it is important to be twice as careful with our own statistics as we are when examining others', especially those on the other side of the issue.

"Public health certainly aspires to as high an ideal as is professed by business: Truth in Advertising. To that end, the first safeguard proposed in the publicity use of statistics is to try to be more critical of the materials and of the reasoning which favor our cause, than of those which oppose it—more critical, because otherwise we shall certainly not be critical enough; the instinct to defend one's beliefs and interests is so strong that desire too often pulls the wool over the eye of reason."

And keep in mind that finding the truth is not as simple as just finding a reasonable hypothesis and sticking with it:

"It is a fundamental theorem of scientific procedure that a given explanation cannot logically be accepted as the preferable one, unless all other reasonable hypotheses have been eliminated."

Anyway, it's a good article. Go check it out! It's even open access.

Reference: Hedrich, A. W. Safeguards in the Publicity Use of Vital Statistics. Am J Public Health Nations Health 24, 336–341 (1934).