RE: Son of unit roots -Reply -Reply -Reply -Reply -Reply

[Andy <soc40001@xxxxxxxxxxxxxx>]

However, analytical induction modeling utilizing verstehen (interpretive
understanding) does not use deductive theory.  One enters the field
with a few sensitizing concepts, but this is it.  One can make the
argument that the model being developed is actually the recording of lay
theory.  I, however, want to point out that all theory begins from the
point of induction.  One observes phenomena and tries to explain that
phenomena.  That explanation is the beginning of a theory.  From this
theory, hypotheses and models are drawn and then tested.  Of course, one
can fit observations into prexisting theory, but this theory began as an
attempt to explain a set of phenomena; theories just don't come out of
the blue.  Ultimately there are only two kinds of activities in this area:
1) model building and 2) model testing.  All statistical work can be
covered by these two activities.  It is obvious that one must have a
model to test, but that model had to be developed first.  This model was
developed from observations in the empirical world.  Now, we could take a
step back and argue that our assumptions about the nature of empirical
data (our sensitizing concepts) are prior to model building (well, of
course they are).  Granted.  But they are not theories.  Epistemologies
are not theories.  Moreover, at this point, we are entering into a
constructivist discussion (which is okay)...

Andy Austin
soc40001@xxxxxxxxxxxxxx




On Tue, 14 Feb 1995, Paul Davidson wrote:

> FROM:  Paul Davidson
> "      Holly Chair of Excellence in Political Economy
> "      University of Tennessee, Knoxville, Tn. 37996-0550
> Dear Louis-Phillipe: Unfortunately there are no facts without deductive theory.
> Merely to classify events when you collect data requires a taxonomy based
> on some deductive theory. All of this was spelled out in great detail by
> Charlie Leven in a chapter entitled "Social Accounts: Theory and Measurement",
> Chapter 15 in AGGREGATE SUPPLY AND DEMAND ANALYSIS, by P. Davidson and
> E. Smolensky (1964).
>   For example, how do you measure the quantity of money depends on the logical
> theory of what is money! Before a physical scientist can collect data and plook
>  for empirical regularities involving "quarks"she must have a theory about
> quarks.
>    Why is it that a whale looks like a fish, swims like a fish, and if out of
> water dies like a fish (all empirically observable regularities) but is not a f
> fish? It has to do with a thoery of biology!
>    Why is it that asteroids and comets fly around the sun, but are not planets?
>  Classification of empirical facts about planets have to do with a theory
> of the solar system!
>      There can be no classification without a theory of common properties
> and functions for each class! without classification there can be no
> collection of data -- other than a single entry listing of historical
> events. As we know, history never repeats itself -). But regularities
> appear. As the non-economist P. Billingsley noted if "the assumption of ergodic
> ity permits regularities to be perceived from what might at first sight be
> patternless fluctuations".  for more on this see "A Technical Definition of Unc
> ertainty and the Long-Run Non-neutrality of Money", CAMBRIDGE JOURNAL OF ECONOM
> ICS, September 1968.
>
> Have a good day! Paul
> email address: PB108928@xxxxxxxxxxxxxx
> fax# (615)974-1686
> phone# (615)974-4221
>