Grubbs T-Test has been consistenly used in startistical analysis for rejection of outliers in ASCOPE Laboratory Test Correlation Programme.The present paper compares the effects of using other statistical test, namely Ferguson b2-Test and Dixon r-Test on the rejection of Laboratory test resultsAlthough these methods gave conflicting verdicts on some of the laboratory results, it seems that the use of Grubbs T-Test is quite satisfactory for the purpose of ASCOPE Laboratory Test Correlation Programme.
Copyrights © 1991