SYMPOSIA PAPER Published: 01 January 1960
STP48514S

### Proposed Procedure for Determination of Precision of Committee D-19 Methods

Source

Each method of analysis, determination, or measurement requires a statement of the precision attainable by the method. For Committee D-19 methods, this statement must conform with the definition of precision in the Definitions of Terms Relating to Industrial Water and Industrial Waste Water (ASTM Designation: D 1129). This means that the task group proposing a new or revised method must carry out suitably planned trials from which the precision (over-all, estimated standard deviation) of the method can be calculated. This method provides general guidance to task groups in planning and conducting such a determination of precision. Distribution of identical, stable samples is essential for this general plan; suggestions for methods for which such samples are impossible are given at the end of this procedure. Considerable pilot work on a method must precede the determination of its precision. This pilot work should explore such variables as reaction time and temperature, concentration of reagents, interferences, calibration, and sample size. All potentially significant factors must be prescribed and controlled closely in advance because cooperative identification and definition of them can be expensive. Disregard of such factors may introduce so much variation among co-operators that results are misleading or inconclusive. Only after a proposed method has been tried, proved, and reduced to unequivocal written form should a determination of its precision be attempted. The measure of precision that is sought is the over-all or total standard deviation of the results yielded by a method. Since no method can be tried in all conceivable applications, the standard deviation calculated from the results of any planned program can be only an estimate of the universe standard deviation. For this reason the symbol s (sample standard deviation) is used here. The total standard deviation, sT, of the planned sample is the result of many sources of variation—both recognized and unknown. The mathematical relationship is: sT2=sA2+sC2+sO2+sL2+sX2 where the subscripts have the following significance: T = total, A = apparatus, C = cooperating laboratory, O = operator, L = level, and X = unknown.

# Author Information

Committee D-19
Price: \$25.00
###### Reprints and Permissions
Reprints and copyright permissions can be requested through the