THE REPRODUCIBILITY AND REPEATABILITY OF THE PSV DETERMINATION

This report describes, and gives the results of, an experiment designed to study the reproducibility and repeatability of the polished-stone value (PSV) determination when carried out according to the procedure given in BS 812 : 1967. Repeatability and reproducibility estimates were obtained from analyses of: (A) measured test results, and (B) test results 'corrected' by a factor based on the values obtained on Enderby control specimens. The analyses showed that the repeatability and reproducibility estimates on the measured test results for a single measurement were 4.9 and 6.0 units respectively, but for the 'corrected' test results both estimates were reduced to 4.2 units. These results confirmed the value (six units) for the reproducibility of the test which is quoted in the current edition of BS 812, but showed that the incorporation of a 'correction' technique into the procedures would appreciably improve its precision.