Re: [Gretl-users] Possible redundancy of some function packages?
On Fri, 7 Nov 2014, yinung at Gmail wrote:
2014-11-06 0:14 GMT+08:00 Sven Schreiber <svetosch(a)gmx.net>:
> Hi everybody,
>
> 6) "JB" by Yi-Nung Yang. This does a Jarque-Bera test and even the help
> text mentions that it is built into gretl (as "normtest ... --jbera").
> So it's not obvious to me what the purpose of the package is.
>
"JB" package was written by me before it was built into gretl. Actually, I
am glad to see it as a built-in command now. However, there would be still
a possible small improvement on this built-in command.
JB is often used on checking the residuals of OLS in which the degree of
freedom should be taken into account in calculating the JB statistic. The
current "normtest" might consider adding a parameter specifying the number
of coefficients in OLS if it is used on checking the residuals. That would
make little numerical difference (for example, see codes below), but it
might be import in teaching undergraduate econometrics.
I beg to differ: the JB test is a LM test of the null of normality versus
the alternative hypothesis that the "real" distribution belongs to the
Pearson family. All its properties rely on asymptotics: it is _possible_
that some degrees-of-freedom adjustment may marginally improve its
size/power properties, but I'm not aware of any evidence on this.
-------------------------------------------------------
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
-------------------------------------------------------