Thursday, September 30, 2010

"compared with" vs "compared to"

Because the outcome difference between treatment arms is what's being discussed, I've used "compared with," but the client keeps changing it back to "compare to." What authority should I use to demonstrate that "with" really is preferred?
See page 390 in in the 10th edition AMA manual (or 249 in the 9th):
One thing or person is usually compared with another when the aim is to examine similarities or differences in detail. An entity is compared to another when a single striking similarity (or dissimilarity) is observed, or when a thing of one class is likened to one of another class, without analysis (ie, one entity is comparable to another).
Thus AMA confirms what we already know; the more important question is whether AMA is considered authoritative. Does it carry weight?

Some years ago I was confronted by a relatively new but senior member of a company's client service team—a physician with considerable clinical experience and pile of publications:

"Look," he said a little too brusquely, "why do you keep doing this to P values? Nobody knows what this is—I don't think I've ever seen them formatted this way."

I answered that the account used AMA style, then tried to give AMA's rationale for probability value formatting but never got a chance to finish.

"AMA?" he interrupted. "That's JAMA, right?"

I nodded.

"I haven't published anything in JAMA in a decade, at least. JAMA? Who publishes in JAMA anymore?"

If AMA's take on compared with fails to inspire sufficient confidence, you can also cite the ACS Style Guide (page 48) or the 15th edition of Chicago (page 206 and, yes, I know I need to upgrade) or the AP Stylebook and Briefing on Media Law (page 54).

No comments: