OMB Meeting Book_9-11-14
Table 18: Reproducibility data for starch-based infant formula test sample
Sample ID: Sample 2 S-1111/S-2222 Analyte CMP
UMP
GMP
IMP
AMP
(mg hg -1 )
(mg hg -1 )
(mg hg -1 )
(mg hg -1 )
(mg hg -1 )
Laboratory
Rep-1
Rep-2
Rep-1
Rep-2
Rep-1
Rep-2
Rep-1
Rep-2
Rep-1
Rep-2
1 2
10.75 10.49
10.73
4.03 3.91 3.14 a 3.90 3.44 b
4.14 3.78 3.11 a 3.90 4.29 b
1.59 1.78
1.60 1.78
1.51 1.82
1.49 1.84
3.45 3.53
3.48 3.52
9.43
3
9.90
9.98
1.63
1.63
1.58
1.59
3.38
3.42
4
11.90
12.30
1.70
1.70
1.70
1.70
3.60
3.70
5
11.92
11.92
1.67
1.69
2.05
2.07
3.56
3.79
6
11.27
11.20
4.23
4.24
1.63
1.60
1.55
1.56
3.52
3.46
7
11.28
11.22
4.02
3.89
1.65
1.60
1.47
1.44
3.59
3.58
8
11.10
10.93
4.04 1.19 c
3.77 0.00 c
1.67 0.00 c
1.62 0.00 c
1.61 0.00 c 1.63 1.80 b
1.61 0.00 c 1.63 1.71 b
3.55 0.00 c
3.42 0.00 c
2.77 c
3.08 c
9
10
11.44
11.44
3.91
3.93
1.72
1.79
3.58
3.54
11
11.03
11.84
3.73
3.91
1.78
1.75
3.56
3.75
12
9.91
9.79
3.93
4.17
1.64
1.58
1.62
1.59
3.46
3.42
a Identified as Grubbs outlier; results kept in data set for statistical analysis. b Identified as Cochran outlier; results kept in data set for statistical analysis. c Problems were identified by participants and as Grubbs outlier results excluded from data set for statistical analysis.
Table 19: Reproducibility analysis for starch-based infant formula test sample
Sample ID: Sample 2 S-1111/S-2222
CMP
UMP
GMP
IMP
AMP
Analyte Symbol
Seq.
Item
Value
Value
Value
Value
Value
11 a
11 a
11 a
11 a
11 a
1
Total number of laboratories
n
2
Total number of replicates
2n
22
22
22
22
22
3
Overall mean of all data (grand mean)
10.99
3.88
1.67
1.66
3.54
4
Repeatability standard deviation
S r
0.30
0.21
0.03
0.02
0.08
5
Reproducibility standard deviation
S R
0.81
0.31
0.07
0.17
0.11
6
Repeatability limit in SMPR
≤6%
≤6%
≤6%
≤6%
≤6%
7
Repeatability relative standard deviation
RSD r
2.7%
5.4%
1.6%
1.4%
2.1%
8
Reproducibility limit in SMPR
≤11%
≤11%
≤11%
≤11%
≤11%
9
Reproducibility relative standard deviation
RSD R
7.4%
8.4%
4.2%
10.3%
3.0%
10
HorRat value
0.9
0.9
0.4
1.0
0.3
a Less than 2/9 laboratories removed from statistical analysis.
Made with FlippingBook