Laserfiche WebLink
SIGMA- ALDRICH <br />Definitions and Interpretation of Statistical Analysis: <br />Assigned Value: Value attributed to a particular quantity and accepted, sometimes by convention, as having an uncertainty <br />appropriate for a given purpose. See ISO /IEC 17043 for additional information. In general the assigned value is the value used to <br />assess proficiency and may or may not be the made to value (gravimetric value). <br />Accept. Window: The range of values that constitute acceptable performance for a laboratory participating in this PT study. <br />Z: A Z -Score tells how a single data point compares to normal data. A Z -Score says not only whether a point was above or below <br />average, but how unusual the measurement is. Generally, a method result with a Z -Score less than 121 is considered to be in control, <br />a Z -Score between 121 and 131 is considered 'Questionable', but still within control and a Z greater than 131 is considered not <br />acceptable and the method is out of control. Calculated as Z = (Reported Value - Assigned Value) / Proficiency Std. Dev. <br />Proficiency Std. Dev.: Standard deviation calculated based on Evaluation Criteria. <br />Study Mean: Statistical study mean calculated using a robust statisitical model (RTC employs the 'Biweight Program'). Robust <br />statistical techniques to minimize the influence that extreme results can have on estimates of the mean and standard deviation. <br />NOTE - These techniques assign less weight to extreme results, rather than eliminate them from a data set. <br />Study Std. Dev.: Standard deviation calculated from study data using robust statisicals (Biweight). <br />Gravimetric Value: The 'prepared to' value, determined by gravimetric means. The uncertainty associated to this value is standard <br />uncertainty and based on RTC's gravimetric tolerances. <br />Evaluation Criteria: <br />1 - Regression Equation - Acceptance windows based on TNI adopted equation of proficiency value +/- 3 proficiency standard <br />deviations and check limits of proficiency value +/- 2 proficiency standard deviations. Proficiency value and proficiency standard <br />deviation are calculated from gravimetric variables a, b, c, & d as proficiency value = a * gravimetric + b and proficiency standard <br />deviation = c* gravimetric + d. <br />2 - Study Robust Mean and c,d regression - Acceptance windows based on TNI adopted equation of proficiency value +/- 3 <br />proficiency standard deviations and check limits of proficiency value +/- 2 proficiency standard deviations. Proficiency value and <br />proficiency standard deviation calculated from robust study mean and variables c & d as proficiency value = robust mean and <br />proficiency standard deviation = c * proficiency value + d. <br />3 - Fixed Limits - Acceptance windows based on span of gravimetric percentage from gravimetric as gravimetric +/- gravimetric * <br />percentage. <br />4 - Adjustable Fixed Limits - Acceptance windows base on a span of gravimetric percentage from gravimetric as gravimetric +/- <br />gravimetric * IowPercentage where gravimetric < break and gravimetric +/- gravimetric * highPercentage where gravimetric >= break. <br />5 - Study Statistics - Acceptance windows based on a number of standard deviations span from the study mean as study mean <br />+/- (deviations * standard deviation). <br />6 - Log Transform Statistics - Acceptance windows based on lognormal distributed data. Acceptance windows = <br />mean(lognormal) +/- span * standard deviation(lognormal). <br />7 - Reserved <br />8 - Regression Equation 2SD - Acceptance windows based on EPA equation of proficiency value +/- 2 proficiency standard <br />deviations. Proficiency value and proficiency standard deviation are calculated from gravimetric variables a, b, c, & d as proficiency <br />value = a * gravimetric + b and proficiency standard deviation = c * gravimetric + d. Generally reserved for drinking water studies. <br />Proficiency Test Item Preparation, Homogeneity and Stability Assessment - RTC uses proprietary and published methods for <br />the manufacture, homogeneity and stability testing of proficiency test items. RTC's proficiency test materials meet requirements of <br />5/27/11 RT1014 WP11 -2 -75 <br />WP11 -2 -75 <br />Concluded 05/25/2011 <br />Final Report <br />Page 14 of 15 <br />