The Philosophy of Statistical Bias Testing Revisited and Applied to Automatic Samplers

Following GyÆs earlier proposals for the use of StudentÆs t-test, a simplified, effective method and philosophy are proposed to test for maximum tolerable bias in data sets as well as in sampling devices. The philosophy of testing is particularly emphasised to derive important practical implications in terms of decision-making and risk control. Conclusions are offered which should equally matter to manufacturers and users of sampling equipment.