Standard: 2009 TNI
Section: V1M4, Section 18.104.22.168
Under Matrix Duplicates it states that the precision may be expressed as RPD or another statistical treatment. We do not do matrix duplicates, we perform sample duplicates only and use percent recovery. Are we required to run matrix duplicates, or are we okay running sample duplicates only? If we are allowed to run sample duplicates only, do we have to express precision as RPD, or can we stick to percent recovery? If we need to switch to matrix duplicates do we need to use RPD or can we use percent recovery?
TNI FINAL RESPONSE:
(Quality Systems Expert Committee and NELAP AC, 1-20-14)
Any laboratory-duplicated sample (however named, such as matrix duplicates, sample duplicates, or matrix spike duplicates) must be evaluated for precision. A percent recovery evaluates accuracy, but not precision. Precision must be evaluated by a statistical technique such as RPD, absolute difference or Percent relative standard deviation (% RSD).