The following errors might cause the bias observed in the determination of calcium in the reference serum:

• The standard concentrations might be incorrect. There could be the result of a calculation error, the material that was used to make the calcium standard (usually calcium carbonate) might have been impure and/or nonstoichiometric, or the dilutions could have been incorrect.
• The preparation of SRM 909b might have been incorrect. This material is freeze-dried serum that must be reconstituted with exactly 10.00 mL of distilled water. If it was reconstituted with too large a volume of water, the value would be lower than expected.
• There could be an error in the 50-fold dilution of the SRM to prepare the working solution. A dirty pipet might deliver significantly less than expected.
• There could be a serum matrix-induced suppression of the signal that would result in lower than expected results. This could be corrected using the method of standard addition. However, since the method used in this procedure is essentially that which was tested and recommended in the NIST Special Publication, it is highly unlikely that any matrix interference might occur.