<s>
In	O
statistics	O
,	O
the	O
White	B-General_Concept
test	I-General_Concept
is	O
a	O
statistical	O
test	O
that	O
establishes	O
whether	O
the	O
variance	O
of	O
the	O
errors	O
in	O
a	O
regression	O
model	O
is	O
constant	O
:	O
that	O
is	O
for	O
homoskedasticity	B-General_Concept
.	O
</s>
<s>
This	O
test	O
,	O
and	O
an	O
estimator	O
for	O
heteroscedasticity-consistent	O
standard	O
errors	O
,	O
were	O
proposed	O
by	O
Halbert	O
White	O
in	O
1980	O
.	O
</s>
<s>
In	O
cases	O
where	O
the	O
White	B-General_Concept
test	I-General_Concept
statistic	O
is	O
statistically	O
significant	O
,	O
heteroskedasticity	B-General_Concept
may	O
not	O
necessarily	O
be	O
the	O
cause	O
;	O
instead	O
the	O
problem	O
could	O
be	O
a	O
specification	O
error	O
.	O
</s>
<s>
In	O
other	O
words	O
,	O
the	O
White	B-General_Concept
test	I-General_Concept
can	O
be	O
a	O
test	O
of	O
heteroskedasticity	B-General_Concept
or	O
specification	O
error	O
or	O
both	O
.	O
</s>
<s>
If	O
no	O
cross	O
product	O
terms	O
are	O
introduced	O
in	O
the	O
White	B-General_Concept
test	I-General_Concept
procedure	O
,	O
then	O
this	O
is	O
a	O
test	O
of	O
pure	O
heteroskedasticity	B-General_Concept
.	O
</s>
<s>
If	O
cross	O
products	O
are	O
introduced	O
in	O
the	O
model	O
,	O
then	O
it	O
is	O
a	O
test	O
of	O
both	O
heteroskedasticity	B-General_Concept
and	O
specification	O
bias	O
.	O
</s>
<s>
If	O
the	O
error	O
term	O
in	O
the	O
original	O
model	O
is	O
in	O
fact	O
homoskedastic	B-General_Concept
(	O
has	O
a	O
constant	O
variance	O
)	O
then	O
the	O
coefficients	O
in	O
the	O
auxiliary	O
regression	O
(	O
besides	O
the	O
constant	O
)	O
should	O
be	O
statistically	O
indistinguishable	O
from	O
zero	O
and	O
the	O
R2	O
should	O
be	O
“	O
small	O
"	O
.	O
</s>
<s>
Conversely	O
,	O
a	O
“	O
large	O
"	O
R2	O
(	O
scaled	O
by	O
the	O
sample	O
size	O
so	O
that	O
it	O
follows	O
the	O
chi-squared	O
distribution	O
)	O
counts	O
against	O
the	O
hypothesis	O
of	O
homoskedasticity	B-General_Concept
.	O
</s>
<s>
An	O
alternative	O
to	O
the	O
White	B-General_Concept
test	I-General_Concept
is	O
the	O
Breusch	B-General_Concept
–	I-General_Concept
Pagan	I-General_Concept
test	I-General_Concept
,	O
where	O
the	O
Breusch-Pagan	B-General_Concept
test	I-General_Concept
is	O
designed	O
to	O
detect	O
only	O
linear	O
forms	O
of	O
heteroskedasticity	B-General_Concept
.	O
</s>
<s>
If	O
homoskedasticity	B-General_Concept
is	O
rejected	O
one	O
can	O
use	O
heteroskedasticity-consistent	O
standard	O
errors	O
.	O
</s>
<s>
In	O
R	B-Language
,	O
White	B-General_Concept
's	I-General_Concept
Test	I-General_Concept
can	O
be	O
implemented	O
using	O
the	O
white	O
function	O
of	O
the	O
skedastic	O
package	O
.	O
</s>
<s>
In	O
Stata	B-Algorithm
,	O
the	O
test	O
can	O
be	O
implemented	O
using	O
the	O
estat	O
imtest	O
,	O
white	O
function	O
.	O
</s>
