In article <8q74pn$cg1$[EMAIL PROTECTED]>,
Alexander Tsyplakov <[EMAIL PROTECTED]> wrote:
>It's necessary to write an extensive journal article to
>answer the question. I just try to list some of my thoughts.
>In 1930-s and 1940-s the term "econometrics" also implied
>what is now called "mathematical economics" or "economic
>theory" (just look through any issue of "Econometrica").
>This usage do not contain any statistics, but now it is
>outdated.
There was statistics in econometrics back then; it was not
used as much, as even the theory behind the methods had
not yet been developed; it was in the mid 40's when those
who understood rigorous statistics first developed this,
and it may have been fortunate that too many techniques
were not in place, or the arguments about the dangers of
misusing them might not have been found.
BTW, statistics was not well developed then, either.
...............
>Some specific features of econometrics:
>Greater emphasis on various kinds of regression models,
There is far too much of this. Modeling is a good
idea, and linear models were used, and with worries
about their accuracy. One problem in econometrics
is that there is nowhere near the amount of data
needed to fit reasonable models. With the present
computer facilities, there is a rush to calculate,
and to uncritically accept the computer output.
This is also the case with other uses of statistics.
>comprehensive after-estimation specification testing
>("diagnostics"), preoccupation by nonstationarity of time
>series.
>Many statistical methods seem old-fashioned or useless to an
>econometrician. Something like MANOVA, PACF, factor
>analysis, KS test, comparing sample means, multidimensional
>scaling etc.
Statistics is universal, and statistical methods are
dangerous in the hands of those who do not understand
them. Most do not.
>The difference can become clear from comparison of
>statistical and econometric software (e.g., SPSS or
>Statistica vs. Eviews or Limdep).
If one has a slightly unusual problem in statistics,
it will be a major job to work with these. Instead,
wrong procedures are used.
>Note also econometix slang: proxy, dummy, unit root, LM
>test, encompassing, exogeneity/ endogeneity etc. I am sure
>many statisticians would not understand what some applied
>econometric papers are talking about (not to speak about
>some econometric theory papers).
I know some of these, but why should a statistician
know economic terminology? The econometrician should
understand enough probability to formulate the models
and interpret the results, and then can communicate
with the statistician who may not know economics.
I know; I was there.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
[EMAIL PROTECTED] Phone: (765)494-6054 FAX: (765)494-0558
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
