Two points here. First, I'd 100% endorse what Doug has said here - clarity is essential and often the argument that a term is commonly understood in a particular sense isn't really adequate.

Also, re Ken's point - in many research domains this is regarded as unacceptable - now. Back in the 1970s a clipboard-wielding person approached people in a busy shopping street, and asked their opinion on some current political issues. But this wasn't the real purpose of the study. All interactions were observed discreetly through binoculars from an overlooking window, to determine what proportion of people approached actually complied. On some occasions the interviewer appeared as her normal self, on other occasions the occlusion of her front teeth was deliberately altered in some way (I've no idea how you do this in vivo). The dentists who did this study would obviously argue that this was the only way to determine to what degree orthodontic factors would affect human interaction. This 'fly-on-the-wall' research mode is clearly scientifically advantageous. But current research governance procedures - in the healthcare domain at least - would make it quite unacceptable. 


Robert G. Newcombe PhD CStat FFPH
Professor of Medical Statistics
Department of Primary Care and Public Health
Clinical Epidemiology Interdisciplinary Research Group
Cardiff University
4th floor, Neuadd Meirionnydd
Heath Park, Cardiff CF14 4YS

Tel: (+44) 29 2068 7260
Fax: (+44) 29 2068 7236

Home page http://tinyurl.com/7mr754
For location see http://www.cardiff.ac.uk/locations/resources/Faxable%20Map-07.pdf


>>> Doug Altman <[log in to unmask]> 17/04/10 17:19 >>>

Labels are convenient but are untrustworthy for scientific 
communication. Many terms are widely misused even when in principle 
they have a clear definition (although many don't even have that).

Several terms used in clinical trials fall into this category - as 
well as "single blind" and indeed "double blind" and "triple blind", 
other examples include "randomised" and "intention to treat". When it 
really matters, eg in a publication or grant application, all such 
terms are inadequate and the authors need to say exactly what they 
did or plan to do.


>>> Ken Masters <[log in to unmask]> 17/04/10 05:23 >>>

In a similar vein, what is it called when the subjects are put through a
trial of some sort, and the real reason and purpose for the experiment
is withheld from the subjects completely?  I'm thinking of something
like Stanley Milgram's "pain" experiments.  Subjects believed that they
were acting as a "teacher," inflicting pain on a third person (a
"learner") through electric shock when the "learner" gave incorrect
answers to questions posed by the person running the experiment  But
actually, the participants themselves were the subjects of an
experiments to see how far they would go in their obedience to
authority.  (The "learner" was not actually receiving any shocks, but
was using recoded sounds, and would add to the general confusion by
banging on the wall).

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.