<s>
In	O
statistical	O
theory	O
,	O
Chauvenet	B-General_Concept
's	I-General_Concept
criterion	I-General_Concept
(	O
named	O
for	O
William	O
Chauvenet	O
)	O
is	O
a	O
means	O
of	O
assessing	O
whether	O
one	O
piece	O
of	O
experimental	O
data	O
an	O
outlier	O
from	O
a	O
set	O
of	O
observations	O
,	O
is	O
likely	O
to	O
be	O
spurious	O
.	O
</s>
<s>
The	O
idea	O
behind	O
Chauvenet	B-General_Concept
's	I-General_Concept
criterion	I-General_Concept
is	O
to	O
find	O
a	O
probability	O
band	O
,	O
centered	O
on	O
the	O
mean	O
of	O
a	O
normal	O
distribution	O
,	O
that	O
should	O
reasonably	O
contain	O
all	O
n	O
samples	O
of	O
a	O
data	O
set	O
.	O
</s>
<s>
By	O
doing	O
this	O
,	O
any	O
data	O
points	O
from	O
the	O
n	O
samples	O
that	O
lie	O
outside	O
this	O
probability	O
band	O
can	O
be	O
considered	O
to	O
be	O
outliers	O
,	O
removed	O
from	O
the	O
data	O
set	O
,	O
and	O
a	O
new	O
mean	O
and	O
standard	B-General_Concept
deviation	I-General_Concept
based	O
on	O
the	O
remaining	O
values	O
and	O
new	O
sample	O
size	O
can	O
be	O
calculated	O
.	O
</s>
<s>
This	O
identification	O
of	O
the	O
outliers	O
will	O
be	O
achieved	O
by	O
finding	O
the	O
number	O
of	O
standard	B-General_Concept
deviations	I-General_Concept
that	O
correspond	O
to	O
the	O
bounds	O
of	O
the	O
probability	O
band	O
around	O
the	O
mean	O
(	O
)	O
and	O
comparing	O
that	O
value	O
to	O
the	O
absolute	O
value	O
of	O
the	O
difference	O
between	O
the	O
suspected	O
outliers	O
and	O
the	O
mean	O
divided	O
by	O
the	O
sample	O
standard	B-General_Concept
deviation	I-General_Concept
(	O
Eq.1	O
)	O
.	O
</s>
<s>
is	O
sample	O
standard	B-General_Concept
deviation	I-General_Concept
.	O
</s>
<s>
In	O
order	O
to	O
find	O
the	O
standard	B-General_Concept
deviation	I-General_Concept
level	O
associated	O
with	O
,	O
only	O
the	O
probability	O
of	O
one	O
of	O
the	O
tails	O
of	O
the	O
normal	O
distribution	O
needs	O
to	O
be	O
analyzed	O
due	O
to	O
its	O
symmetry	O
(	O
Eq.3	O
)	O
.	O
</s>
<s>
is	O
the	O
standard	B-General_Concept
deviation	I-General_Concept
of	O
standard	O
normal	O
distribution	O
.	O
</s>
<s>
To	O
apply	O
Chauvenet	B-General_Concept
's	I-General_Concept
criterion	I-General_Concept
,	O
first	O
calculate	O
the	O
mean	O
and	O
standard	B-General_Concept
deviation	I-General_Concept
of	O
the	O
observed	O
data	O
.	O
</s>
<s>
From	O
there	O
we	O
see	O
that	O
and	O
can	O
conclude	O
that	O
50	O
is	O
an	O
outlier	O
according	O
to	O
Chauvenet	B-General_Concept
's	I-General_Concept
Criterion	I-General_Concept
.	O
</s>
<s>
Another	O
method	O
for	O
eliminating	O
spurious	O
data	O
is	O
called	O
Peirce	B-Language
's	I-Language
criterion	I-Language
.	O
</s>
<s>
It	O
was	O
developed	O
a	O
few	O
years	O
before	O
Chauvenet	B-General_Concept
's	I-General_Concept
criterion	I-General_Concept
was	O
published	O
,	O
and	O
it	O
is	O
a	O
more	O
rigorous	O
approach	O
to	O
the	O
rational	O
deletion	O
of	O
outlier	O
data	O
.	O
</s>
<s>
Other	O
methods	O
such	O
as	O
Grubbs	B-General_Concept
's	I-General_Concept
test	I-General_Concept
for	I-General_Concept
outliers	I-General_Concept
are	O
mentioned	O
under	O
the	O
listing	O
for	O
Outlier	O
.	O
</s>
<s>
Deletion	O
of	O
outlier	O
data	O
is	O
a	O
controversial	O
practice	O
frowned	O
on	O
by	O
many	O
scientists	O
and	O
science	O
instructors	O
;	O
while	O
Chauvenet	B-General_Concept
's	I-General_Concept
criterion	I-General_Concept
provides	O
an	O
objective	O
and	O
quantitative	O
method	O
for	O
data	O
rejection	O
,	O
it	O
does	O
not	O
make	O
the	O
practice	O
more	O
scientifically	O
or	O
methodologically	O
sound	O
,	O
especially	O
in	O
small	O
sets	O
or	O
where	O
a	O
normal	O
distribution	O
cannot	O
be	O
assumed	O
.	O
</s>
