It appears to be a coinage by Andrews, meaning the
total hypotheses space. I checked some of the
statistical references and did not find it used.
It also does not appear in the Combined Index to
Statistics. It would be worthwhile to examine some
of the Econometrica references -- I suspect that
he defines it in one of his earlier papers. Better
yet, send him an e-mail.
Alexander Tsyplakov wrote:
>
> I think you are right that this is not generally recognized term like
> null or alternative. But it is sometimes used. Just put "maintained
> hypothesis" into Google search engine.
>
> "Null hypothesis" is only one possible interpretation of "maintained
> hypothesis". There is at least one other interpretation that
> "maintained hypothesis" = "null hypothesis" + "alternative
> hypothesis". Of course, strictly speaking "null hypothesis" +
> "alternative hypothesis" is almost always true by assumption. But it
> could be false if the model itself is false. We just always assume
> (maintain) that the model is true. So this could be called (and
> sometimes is called) "maintained hypothesis".
>
> Good example is a paper by Donald Andrews
> http://cowles.econ.yale.edu/P/cd/d12a/d1229.pdf
>
> But Andrews is an econometrician. May be for statisticians this
> meaning is not familiar.
>
> -----------------------------
> Alexander Tsyplakov
> Novosibirsk State University
> http://matrixer.narod.ru/
>
> Bob Wheeler wrote...
> > I don't believe that this is a generally
> > recognized technical term. I suppose that its use
> > is an attempt to avoid the word "null" in "null
> > hypothesis," which is what the writers mean. The
> > adjective "maintained" in this context corresponds
> > to the Russian "itberzhdati." (not sure of the
> > approved transliteration)
> >
> > Alexander Tsyplakov wrote:
> > >
> > > The meaning of the term "maintained hypothesis" differs in the
> > > literature. I'm not sure which meaning is correct (or more widely
> > > used). Can anybody help? I need this for English-Russian
> dictionary.
--
Bob Wheeler --- (Reply to: [EMAIL PROTECTED])
ECHIP, Inc. --- (302) 239-6620, voice and
FAX
724 Yorklyn Rd., Hockessin, DE 19707
Randomness comes in bunches
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
