MS-QA_06-2019 rev5

ANSI/AARST MS-QA 2019 7 Radon Measurement Systems Quality Assurance Definitions RPD (Relative Percent Difference): A statistic used to evaluate the difference between two measurements when there is no evidence to support one being more accurate than the other. As with RPE, the RPD normalizes the difference between two measurements by dividing by the best estimate of the true value, which in this case is the mean of the two results. The difference is normalized (compared as a fraction) to the mean of the two results as there is no reason to assume that one measurement is more accurate than the other, and over time a set of RPD values can be used as an estimate of imprecision. RPD = [(A – B) / mean] * 100 (6) where A = the larger result, B = the smaller result, and mean = the average of the two results. Informative advisory —It is important to record whether one measurement is consistently greater or less than the other (as applicable to the method being used), as that may indicate a systematic calibration or background problem that can be corrected. Note—A 10% CV for two measurements corresponds to a 14% RPD. RPE (Relative Percent Error): A statistic used to evaluate the difference between a measurement and the conventionally true value, which may be a more recently calibrated CRM or a chamber concentration. The RPE is the degree from which a single measured value (X) deviates from the conventionally true value (T) and is sometimes called IPE. The RPE is used as an estimate of total error, combining both imprecision and bias, and RPE statistics gathered over time can be used to estimate and correct system biases. RPE = [100 (X – T) / T] (7) where X = Measured value (Bq/m 3 , pCi/L, Bq-h/m 3 , or pCi-d/L) T = Conventionally true value (in the same unit as X) Significant Type Difference : Each measurement system uses calibration factors and algorithms to calculate the measured radon concentration from the measured and reported parameters. A significant type difference is defined as the difference between systems that requires type-specific calibration factors and/or algorithms to meet the requirements of MS-PC . Significant type differences include: • Any factor that results in a difference in the efficiency of the detection mechanism (e.g., a change in a charcoal batch that is significant enough to warrant a change in the calibration factors, or a difference between diffusion barrier materials or design that affects radon adsorption); • Changes in recommended measurement durations that require changes in calibration factors or algorithms; • Expansion of the manufacturers’ recommended range of temperature, pressure, relative humidity or radon concentration within which it should be used. Note—Manufacturers periodically change device user interfaces, software, internal electrical components and instrument housings; such changes do not generally change the method’s response characteristics and are not considered a significant change. In addition, determinations of significant change may require the evaluation of type-test data and be evaluated on a case-by-case basis rather than as categorical assessments. Spike: Spikes are devices or materials that are exposed in a STAR to known radon concentrations for duration or integrated exposures normally encountered in field measurements, and which are recommended by the manufacturer and agreed to between the supplier of the devices and the exposure facility; temperatures between 60–80 ° F; and relative humidities between 10–75%. Spikes are submitted blind to the analyst and are not for calibration purposes but are necessary to verify and document the accuracy of the continued