Review Team (February 2016).pdf

646 / 670

Fonterra Co-operative Group
Page 12
Table 9: Vitamin D 2 recovery of spiked samples at 50% of typical concentrations
Sample
Mean
Recovery
(RSD%)
Child Formula Powder 97 (5.0)
Infant Elemental Powder 98 (2.2)
Adult Nutritional RTF High Protein 98 (1.3)
Adult Nutritional RTF High Fat 98 (2.6)
Infant Formula RTF, Milk-Based 97 (4.4)
Table 10: Vitamin D3 recovery of spiked samples at 100% of typical concentrations
Sample
Mean
Recovery
(RSD%)
Child Formula Powder 101 (0.8)
Infant Elemental Powder 98 (1.5)
Adult Nutritional RTF High Protein 98 (0.9)
Adult Nutritional RTF High Fat 101 (1.2)
Infant Formula RTF, Milk-Based 98 (0.6)
Table 11: Vitamin D3 recovery of spiked samples at 50% of typical concentrations
Sample
Mean
Recovery
(RSD%)
Child Formula Powder 100 (3.1)
Infant Elemental Powder 96 (0.8)
Adult Nutritional RTF High Protein 98 (1.2)
Adult Nutritional RTF High Fat 101 (4.1)
Infant Formula RTF, Milk-Based 94 (0.6)
Bias Against CRM
Bias is a measurement of systematic error representing the difference between measured value and true
value. However, as the true value is unknown, an accepted reference value substituted for true value. A
certified reference material (CRM) contains assigned values accompanied by an uncertainty and stated level
of confidence.
Bias is evaluated by replicate analyses of a CRM with a declared concentration and uncertainty.
Differences between the measured value and certified value are determined with the mean and standard
deviation of the differences, and the test-statistic (Tstat) calculated (Equation 8). The null hypothesis is that
there is no difference between the measured results and the certified value (H 0: ݀ ൌ Ͳ). At the α = 0.05 level
VitD-18 (February 2016)
FOR ERP USE ONLY
DO NOT DISTRIBUTE