Samples containing coarse gold can lead to poor replicate statistics in the assay process. Our labs review the analytical data and when a replicate falls outside our data acceptance criteria typically schedule duplicate retesting of the same pulps. If retesting shows that the sample is affected by coarse gold we need to report this data to our clients in a way that does not affect the statistical reliability of the client’s data set.
Responsibilities
This procedure ensures that our labs are not introducing a systematic bias into our client’s data set and gives the client the information they need to properly process the analytical data and enter it into their database. It is the responsibility of all SGS personnel to follow and adhere to this policy.
Procedure
Do not substitute an unknown result without auditable proof that it is invalid.
Do not make judgement calls on unknown vs. replicate/duplicate data for gold where a third check result may support one of the original values over the other: they are all probably valid data points (refer to #1).
In situations where a client has called for a screen fire check of high grade samples or to help resolve noisy replicates be cautious about changing an unknown gold result because a screened fire check does not agree with it – it may still be statistically valid.
Do not average the retest data, instead report all data in the order obtained.
Notify the client that we have found a sub sampling problem and that the data will be reported to them in the order of analysis date/time.
Contact
Why not contact us for more information? We would love to hear from you!
