On Fri, 10 Jan 2014, Riccardo (Jack) Lucchetti wrote:
Of course, generating data which _does_ contain heterskedasticity is
totally
irrelevant here: if I understand the point correctly, the issue here is on
whether a constant should be present or not in the auxiliary regression for
the White test. And the answer is: yes, the constant should be there. And
that's what gretl does:
Whoops. Oh, no, you're right, sorry. The constant isn't there, but there
should be one. The right form of White's test should be as follows.
<hansl>
nulldata 100
set seed 987
x = uniform()
e = normal()
y = x + e
ols y x
modtest --white # wrong
u2 = $uhat^2
x2 = x*x
ols u2 const x x2
White = $T * $rsq # right
</hansl>
-------------------------------------------------------
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
-------------------------------------------------------