Abstract
A case study in Statistical Quality Control highlights the relevance of sensitivity, specificity, positive predictive value, and negative predictive value for In-Control/Out-of-Control diagnosis. Suspicious data probability patterns are computed and assessed through confusion matrices and (mis)classification performance metrics, showing how pattern contributions influence average false discoveries and false omissions. These findings support the effectiveness of the monitoring strategy in balancing the risks of unnecessary process interruptions against overlooked anomalies. Daily scoring and 15-period moving averages are also examined, establishing the prognostic value of 0-upcrossings in moving averages as reliable indicators of impending process halts.