EC
E
T
t /
2

SEET
t /
2

SEEC
Overlap
of
95%
CIs
for
E
C
and
E
T
(
but
E
C
not
in
95%
CI
for
E
T
and
vice
versa):
[
t /
2
(
 )

SE
EC
+
t /
2
(
 )

SE
ET
]
 
(
E
T
 
E
C
)

Significant
difference
between
E
C
and
E
T
at
5%
level
of
Type
I
error:

(
E
T
 
E
C
)
 
t /
2
(
 
p
)

[
(
SE
ET
)
2
+
(
SE
EC
)
2]
½

(
N­
1)
 
v
p
 
2

(
N
 
1)

Does
the
first
condition
imply
the
second?:

[
t /
2
(
 )

SE
EC
+
t /
2
(
 )

SE
ET
]
 
t /
2
(
 
p
)

[
(
SE
EC
)
2
+
(
SE
ET
)
2]
½

F (
1,
 )

(
SE
EC
+
SE
ET
)
2
 
F (
1,
 
p
)

[
(
SE
EC
)
2
+
(
SE
ET
)
2]

F (
1,
 )/
F (
1,
 
p
)
 
[(
SE
EC
)
2
+
(
SE
ET
)
2]/(
SE
EC
+
SE
ET
)
2
But:
F (
1,
 )
 
F (
1,
 
p
),
so
F (
1,
 )/
F (
1,
 
p
)
 
1;

and
[(
SE
EC
)
2
+
(
SE
ET
)
2]/(
SE
EC
+
SE
ET
)
2
=

[(
SE
EC
)
2
+
(
SE
ET
)
2]/{[(
SE
EC
)
2
+
(
SE
ET
)
2]
+
2

SE
EC

SE
ET
}
 
1
Consequently,
overlap
of
the
95%
CIs
for
E
C
and
E
T
(
but
E
C
not
in
the
95%
CI
for
E
T
and
vice
versa)
demonstrates
that
the
difference
between
E
C
and
E
T
is
significant
at
the
5%
level
of
Type
I
error.
i
0
100
..

:=
RSX
i
i
100
:=

n
5
:=

p
0.05
:=

nu
i
n
1
 

(
)
RSX
i
(
)
2
1
+

  
  
2
RSX
i
(
)
4
1
+

  
  

 

:=

Crit
i
qF
p
1
,
n
1
 

,

(
)

qF
p
1
,
round
nu
i
(
)

,

(
)

   
 

  

1
RSX
i
+

(
)

:=

Crit
0
1
=
0
0.5
1
0.5
0.6
0.7
0.8
0.9
1
1.1
Crit
i
RSXi
"
RSX"
is
the
ratio
of
the
SEE2
of
two
estimates.

"
n"
is
the
sample
size
for
each
of
the
estimates.

"
nu"
is
the
degrees
of
freedom
of
the
test
for
a
significant
difference
between
the
two
estimates.

"
Crit"
is
a
criterion
whose
value
is
<
1
when,
given
overlap
of
the
95%
CIs
of
the
two
estimates
sufficient
that
one
(
or
both)
includes
the
estimate
for
the
other,

the
estimates
are
not
significantly
different
from
one
another.

Thus,
overlap
of
the
95%
CIs
for
two
estimates,
such
that
the
95%
CI
for
one
estimate
includes
the
other
estimate,
is
necessary
to
demonstrate
that
the
difference
between
the
estimates
is
nonsignificant.
i
0
100
..

:=

EE
i
0.5
i
100
+

:=

AEE
i
2
1
EE
i
 

:=
1
1.2
1.4
1.6
1.8
2
5
10
15
20
25
30
1
AEE
i
1
EE
i
EE
25
0.75
=

1
EE
25
1.333
=
1
EE
25
   
   

1
AEE
25
   
   
0.889
=

1
AEE
25
1.5
=

Extraction
efficiency,
EE,
is
an
overestimate;
the
actual
efficiency
is
closer
to
AEE.

(
1/
EE)/(
1/
AEE)
is
the
fraction
by
which
"
X
EEC"
estimates
based
on
EE
should
be
reduced.
2600X
DDD
46X
EEC
36X
EEC
11.2X
EEC
15.6X
EEC
12.3X
EEC
10.6X
EEC
37X
EEC
1400X
DDD
Margin
of
Exposure
(
ME)
2100X
DDDa
Mouse
38X
EECa
Earthworm
30X
EECa
Honeybee
9.2X
EECa
Carabid
beetle
13X
EECa
Rove
beetle
10X
EECa
Ladybird
beetle
8.8X
EECa
Flower
bug
30X
EECb
Trout
1150X
DDDa
Quail
Adjusted
ME
(
a0.89

ME,

b0.82

ME)

Test
Species
Variance
in
Trout
Weight
0
0.2
0.4
0.6
0.8
1
1.2
0
10
20
30
Days
Variance
Trt
Con
Quail
Toxicity
Test:
Change
in
Weight
(
g)

Treatment
Mean
+­

SEM
0
1
2
3
­
5
0
5
10
Female
Male
C
T
X
XX/
XX/

X/
XX
X/
X/

X
XXX
X
XX
Statistics
Problem(
s)

X
WI
Quail
Tox
NCSU
Chicken
Tox
X/

AZ/
AZ
Trout
Tox/
Diet
X
MT
Honey
bee
X/
X
MT/
SB
Lady
Beetle
Tox/
Diet
X/
X
MT/
SB
Rove
Beetle
Tox/
Diet
XX/

MT/
SB
Ground
Beetle
Tox/
Diet
MT/
SB
Flower
Bug
Tox/
Diet
X/

MT/
SB
Earthworm
Tox/
Diet
X
SB
Soil
Degradation
SS
MCRY3A
vs
LPMIR604
SS
Conc.
mCry3A
in
MIR604
SS
Characterize
MCRY3A
Design
Problem(
s)

Lab
Study
SS:
Syngenta
Seeds
SB:
Syngenta
Biotech.

MT:
Mombo­
Tox
BE:
Brixham
Environ.

WI:
Wildlife
International
