Re: [Gretl-devel] Signalling
On Mon, 7 Jan 2013, Lee Adkins wrote:
The publisher John Wiley & Sons surveyed the users of Hill et.
al.
Principles of Econometrics for software choices. As many of you know this
is an advanced undergraduate book--though many like me use it for masters
level classes (the latter chapters in it are particularly good for this,
IMO). The vast majority of the book's users opt for Eviews (43%) followed
by Stata and Excel (about 21% each). 11% reported using gretl. SAS, R,
and SPSS all report less than 1% each. I think their sample size was
around 6000 or so.
These are extremely interesting figures. Frankly, gretl's share is larger
than I had expected. For a free software project, which is always burdened
by the old myth "you get what you pay for", passing the 10% mark is
vitally important, as it means that it has become a recognised player in
the field. Moreover, I have the impression that gretl's presence is even
larger in non-English speaking countries, for obvious reasons. Also, I
wouldn't have expected Eviews' dominance, particularly compared to Stata's
share, but that's another matter.
My guess is that Excel is popular because the marginal cost of its
use is
zero for most adopters. Everybody has it already and sort of knows how to
use it. Gretl's cost, though not zero, is very low. The biggest
difference is that you have to install it and learn something new to use
it. Gretl is much easier to use and more purpose specific in all other
ways, so it theoretically should have an advantage if product awareness was
that of the others. That's my take, anyway..
I repeat: excel users are the "low-end" segment of the market. I don't
mean to diss these people, but my guess is that you'll never manage to
have them switch to anything else, nor that it's particularly important
for the project to do so.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------