There's an oft-quoted aphorism that I deeply hate. It's attributed to 19th century Brittish Prime Minister Benjamin Disraeli and goes, "There are three kinds of lies: lies, damn lies, and statistics." Statistics aren't automatically mendacious just becuase they're statistics. They're first and foremost a means of cataloging observations about the real world. You can also think of them as a means of communicating those observations. It's in this realm that I take issue with Disraeli. As with any other form of communication, it can be earnest or it can be deceitful. It can also be misunderstood.
It's in this last category that the next story comes. The results of a study were misunderstood because the distinction between proportions and overall numbers was forgotten. I present this as a cautionary tale as I'll let the authors explain. One other thing to note here is that the authors of the original study themselves failed to notice or failed to communicate what was in their own data.
No comments:
Post a Comment
Please, no personal attacks, insults or rants.