<s>
Bivariate	B-General_Concept
analysis	I-General_Concept
is	O
one	O
of	O
the	O
simplest	O
forms	O
of	O
quantitative	O
(	O
statistical	O
)	O
analysis	O
.	O
</s>
<s>
Bivariate	B-General_Concept
analysis	I-General_Concept
can	O
be	O
helpful	O
in	O
testing	O
simple	O
hypotheses	O
of	O
association	O
.	O
</s>
<s>
Bivariate	B-General_Concept
analysis	I-General_Concept
can	O
help	O
determine	O
to	O
what	O
extent	O
it	O
becomes	O
easier	O
to	O
know	O
and	O
predict	O
a	O
value	O
for	O
one	O
variable	O
(	O
possibly	O
a	O
dependent	O
variable	O
)	O
if	O
we	O
know	O
the	O
value	O
of	O
the	O
other	O
variable	O
(	O
possibly	O
the	O
independent	O
variable	O
)	O
(	O
see	O
also	O
correlation	O
and	O
simple	B-General_Concept
linear	I-General_Concept
regression	I-General_Concept
)	O
.	O
</s>
<s>
Bivariate	B-General_Concept
analysis	I-General_Concept
can	O
be	O
contrasted	O
with	O
univariate	B-General_Concept
analysis	I-General_Concept
in	O
which	O
only	O
one	O
variable	O
is	O
analysed	O
.	O
</s>
<s>
Like	O
univariate	B-General_Concept
analysis	I-General_Concept
,	O
bivariate	B-General_Concept
analysis	I-General_Concept
can	O
be	O
descriptive	B-General_Concept
or	O
inferential	O
.	O
</s>
<s>
Bivariate	B-General_Concept
analysis	I-General_Concept
is	O
a	O
simple	O
(	O
two	O
variable	O
)	O
special	O
case	O
of	O
multivariate	O
analysis	O
(	O
where	O
multiple	O
relations	O
between	O
multiple	O
variables	O
are	O
examined	O
simultaneously	O
)	O
.	O
</s>
<s>
If	O
the	O
dependent	O
variablethe	O
one	O
whose	O
value	O
is	O
determined	O
to	O
some	O
extent	O
by	O
the	O
other	O
,	O
independent	O
variable	O
is	O
a	O
categorical	O
variable	O
,	O
such	O
as	O
the	O
preferred	O
brand	O
of	O
cereal	O
,	O
then	O
probit	O
or	O
logit	O
regression	O
(	O
or	O
multinomial	B-General_Concept
probit	I-General_Concept
or	O
multinomial	O
logit	O
)	O
can	O
be	O
used	O
.	O
</s>
<s>
If	O
both	O
variables	O
are	O
ordinal	B-General_Concept
,	O
meaning	O
they	O
are	O
ranked	O
in	O
a	O
sequence	O
as	O
first	O
,	O
second	O
,	O
etc.	O
,	O
then	O
a	O
rank	O
correlation	O
coefficient	O
can	O
be	O
computed	O
.	O
</s>
<s>
If	O
just	O
the	O
dependent	O
variable	O
is	O
ordinal	B-General_Concept
,	O
ordered	O
probit	O
or	O
ordered	O
logit	O
can	O
be	O
used	O
.	O
</s>
<s>
If	O
the	O
dependent	O
variable	O
is	O
continuouseither	O
interval	O
level	O
or	O
ratio	O
level	O
,	O
such	O
as	O
a	O
temperature	O
scale	O
or	O
an	O
income	O
scalethen	O
simple	B-General_Concept
regression	I-General_Concept
can	O
be	O
used	O
.	O
</s>
<s>
If	O
both	O
variables	O
are	O
time	O
series	O
,	O
a	O
particular	O
type	O
of	O
causality	O
known	O
as	O
Granger	B-General_Concept
causality	I-General_Concept
can	O
be	O
tested	O
for	O
,	O
and	O
vector	O
autoregression	O
can	O
be	O
performed	O
to	O
examine	O
the	O
intertemporal	O
linkages	O
between	O
the	O
variables	O
.	O
</s>
<s>
Graphs	B-Application
that	O
are	O
appropriate	O
for	O
bivariate	B-General_Concept
analysis	I-General_Concept
depend	O
on	O
the	O
type	O
of	O
variable	O
.	O
</s>
<s>
For	O
two	O
continuous	O
variables	O
,	O
a	O
scatterplot	B-Application
is	O
a	O
common	O
graph	O
.	O
</s>
<s>
When	O
one	O
variable	O
is	O
categorical	O
and	O
the	O
other	O
continuous	O
,	O
a	O
box	B-Application
plot	I-Application
is	O
common	O
and	O
when	O
both	O
are	O
categorical	O
a	O
mosaic	B-Application
plot	I-Application
is	O
common	O
.	O
</s>
<s>
These	O
graphs	B-Application
are	O
part	O
of	O
descriptive	B-General_Concept
statistics	I-General_Concept
.	O
</s>
