<s>
In	O
statistics	O
,	O
the	O
Sobel	B-General_Concept
test	I-General_Concept
is	O
a	O
method	O
of	O
testing	O
the	O
significance	O
of	O
a	O
mediation	O
effect	O
.	O
</s>
<s>
The	O
Sobel	B-General_Concept
test	I-General_Concept
is	O
basically	O
a	O
specialized	O
t	B-General_Concept
test	I-General_Concept
that	O
provides	O
a	O
method	O
to	O
determine	O
whether	O
the	O
reduction	O
in	O
the	O
effect	O
of	O
the	O
independent	O
variable	O
,	O
after	O
including	O
the	O
mediator	O
in	O
the	O
model	O
,	O
is	O
a	O
significant	O
reduction	O
and	O
therefore	O
whether	O
the	O
mediation	O
effect	O
is	O
statistically	O
significant	O
.	O
</s>
<s>
The	O
Sobel	B-General_Concept
test	I-General_Concept
is	O
testing	O
how	O
large	O
area	O
c	O
is	O
.	O
</s>
<s>
Alternative	O
methods	O
of	O
calculating	O
the	O
Sobel	B-General_Concept
test	I-General_Concept
have	O
been	O
proposed	O
that	O
use	O
either	O
the	O
z	O
or	O
t	O
distributions	O
to	O
determine	O
significance	O
,	O
and	O
each	O
estimates	O
the	O
standard	O
error	O
differently	O
.	O
</s>
<s>
One	O
strategy	O
to	O
overcome	O
the	O
non-normality	O
of	O
the	O
product	O
of	O
coefficients	O
distribution	O
is	O
to	O
compare	O
the	O
Sobel	B-General_Concept
test	I-General_Concept
statistic	O
to	O
the	O
distribution	O
of	O
the	O
product	O
instead	O
of	O
to	O
the	O
normal	O
distribution	O
.	O
</s>
<s>
Another	O
approach	O
that	O
is	O
becoming	O
more	O
popular	O
in	O
the	O
literature	O
is	O
bootstrapping	B-Application
.	O
</s>
<s>
Bootstrapping	B-Application
is	O
a	O
non-parametric	O
resampling	O
procedure	O
that	O
can	O
build	O
an	O
empirical	O
approximation	O
of	O
the	O
sampling	O
distribution	O
of	O
αβ	O
by	O
repeatedly	O
sampling	O
the	O
dataset	O
.	O
</s>
<s>
Bootstrapping	B-Application
does	O
not	O
rely	O
on	O
the	O
assumption	O
of	O
normality	O
.	O
</s>
