Statistical Selection
Identification of anomalous values in a set of replicate laboratory measurements requires a rigorous statistical test. By identifying Grubbs test outliers, laboratory technicians detect a single extreme value that deviates significantly from the sample mean. This analysis is necessary when evaluating physical strength tests where a single slip or yarn defect can skew the average.
Mathematical Procedure
The test compares the difference between the suspect value and the sample mean to the standard deviation of the dataset. If the calculated ratio exceeds the critical value for a chosen significance level, the data point is discarded as an outlier. This calculation must be done before calculating the final tensile strength or yarn count of the shipment.
Replicate Assessment
Anomalous results may arise from sample preparation errors, such as uneven clamp tension or fiber misalignment during mounting. Removing these non representative points ensures that the reported data reflects the true physical properties of the bulk textile production rather than experimental artifacts. The standard allows only one such removal per data series.
Report Integrity
Quality managers document each exclusion to maintain the transparency of the testing history. This documentation ensures that audited test certificates remain reliable and trace back to correct practices.
