One might include independent prior evidence (Kleywegt, Brown @ Ramaswami) showing that in general most other quality indicators are worse for high impact journals.
So, as a frequentist I agree that his correlation is significantly weak, as a Bayesian I say it is reasonably probable. Cheers, BR -----Original Message----- From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of Ethan Merritt Sent: Thursday, October 18, 2012 11:11 AM To: [email protected] Subject: Re: [ccp4bb] PNAS on fraud On Thursday, October 18, 2012 10:52:48 am DUMAS Philippe (UDS) wrote: > > Le Jeudi 18 Octobre 2012 19:16 CEST, "Bernhard Rupp (Hofkristallrat a.D.)" <[email protected]> a écrit: > > I had a look to this PNAS paper by Fang et al. > I am a bit surprised by their interpretation of their Fig. 3: > they claim that here exists a highly signficant correlation between > Impact factor and number of retractations. > Personnaly, I would have concluded to a complete lack of correlation... > Should I retract this judgment? Fang et al. claim that R^2 = 0.0866, which means that CC = 0.29. While a correlation coefficient of less than 0.3 is not "a complete lack of correlation", it's still rather weak. The "highly significant" must be taken in a purely statistical sense. That is, it doesn't mean "the measures are highly correlated", it means "the evidence for non-zero correlation is very strong". Ethan
