The 0.06% defect rate is calculated from 100% sort data.
There are True and False Call rates within manual sort, therefore we have to first question what level of the 0.06% do we estimate to be True defects.
It is also important to interpret the number of lots which contain the defect ....for example were all of the defects detected within one lot???
If we claim that all of the defects within the 0.06% calculation are considered to be TRUE defects and they are across multiple lots, then it would be fair to expect LAR = 82.0% at the n=315 audit
Since the actual LAR = 98%, and we claim to trust that the 0.06% defects are True ... then we must question why the defects are not detected at the n=315 final visual audit.