<s>
In	O
statistical	O
hypothesis	B-General_Concept
testing	O
,	O
a	O
two-sample	B-General_Concept
test	I-General_Concept
is	O
a	O
test	O
performed	O
on	O
the	O
data	O
of	O
two	O
random	O
samples	O
,	O
each	O
independently	O
obtained	O
from	O
a	O
different	O
given	O
population	O
.	O
</s>
<s>
The	O
purpose	O
of	O
the	O
test	O
is	O
to	O
determine	O
whether	O
the	O
difference	O
between	O
these	O
two	O
populations	O
is	O
statistically	B-General_Concept
significant	I-General_Concept
.	O
</s>
<s>
There	O
are	O
a	O
large	O
number	O
of	O
statistical	O
tests	O
that	O
can	O
be	O
used	O
in	O
a	O
two-sample	B-General_Concept
test	I-General_Concept
.	O
</s>
<s>
Does	O
the	O
hypothesis	B-General_Concept
being	O
tested	O
apply	O
to	O
the	O
distributions	O
as	O
a	O
whole	O
,	O
or	O
just	O
some	O
population	O
parameter	O
,	O
for	O
example	O
the	O
mean	O
or	O
the	O
variance	O
?	O
</s>
<s>
Is	O
the	O
hypothesis	B-General_Concept
being	O
tested	O
merely	O
that	O
there	O
is	O
a	O
difference	O
in	O
the	O
relevant	O
population	O
characteristics	O
(	O
in	O
which	O
case	O
a	O
two-sided	B-General_Concept
test	I-General_Concept
may	O
be	O
indicated	O
)	O
,	O
or	O
does	O
it	O
involve	O
a	O
specific	O
bias	O
(	O
"	O
A	O
is	O
better	O
than	O
B	O
"	O
)	O
,	O
so	O
that	O
a	O
one-sided	B-General_Concept
test	I-General_Concept
can	O
be	O
used	O
?	O
</s>
<s>
Statistical	O
tests	O
that	O
may	O
apply	O
for	O
two-sample	B-General_Concept
testing	I-General_Concept
include	O
:	O
</s>
