Why not subscribe?

Tuesday, July 05, 2011

No replication? No value.

Byron Sharp posted this:

If it hasn't been replicated it's probably not worth much

which is probably a good rule for research results.

This sounds like an echo of John P. A. Ioannidis' famous paper: "Why Most Published Research Findings Are False"http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.0020124

Ioannidis is discussing published medical research, but if most published medical research is false, it's doubtful marketing would be better.

On the one hand, there are much stronger financial incentives operating in the medical arena (particularly in terms of new drugs). On the other hand, there is the FDA, a set of strong journals, the ability to do double blind studies, and the idea that medicine is a life-and-death business.

Marketing is not a heavily regulated field, and doesn’t deal with issues of life and death.  There are incentives to shade results (in order to get published as a marketing academic; in order to impress the client / your boss if you are in an applied area – or to justify some course of action).