The manufacturer runs
the control over and over and then computes the 95% Confidence interval (± 2 S.D.). This value is then supplied along with each lot of controls and has to be specific for the instrument on which it was run.
The lab pools
specimens together and runs the control over
and over and then computes the 95% Confidence interval (± 2 S.D.). This is done on the same make of instrument for which the controls are being developed.