How to lie with statistics

Being very busy with work, I’ve fallen behind on my reading of health-related news.  Yesterday late afternoon, however, I spent a few minutes seeing what’s the latest.  I had lots to choose from but settled on what I’m about to share.

The headline was a clever one, “A Warming Planet Might Mean More Diabetes”, and it seduced me to read more.  This clever title would be inviting to both the global warming crazy-types and diabetic crowd.  While I’m in neither camp, and will do my best to never join, I was nonetheless intrigued.

Early in the article, I read the following: “The effects of climate change are far reaching, but new research suggests a surprising linkage to a warming Earth - - more cases of type 2 diabetes.”

Apparently, according to this research, for every 1.8 degrees’ Fahrenheit rise in environmental temperature, researchers calculated that there would be an increase of more than 100,000 new cases of type 2 diabetes in the United States.

Still curious, I read more.

Until I read this: “Before you start packing for colder climates, it's important to note that this study cannot prove a direct cause-and-effect relationship between warmer temperatures and the development of type 2 diabetes.”

Come on now, it either does or it doesn’t. Don't waste my time!

I suppose it depends on who’s paying for the research.  Anything is possible.  Many of these researchers are equals with actuaries, economists, expert witnesses and other paid liars.

Yes, anything is possible as my college statistics professor once told us on the first day of class, by referring to a book written on the subject, “How to Lie with Statistics”.

Pax Domini sit semper vobiscum