Abstract Further analyses were performed on the data obtained from the International Intercalibration and Intercomparison Programme for Radon, Thoron and Daughters Measuring Equipment, Part I - Radon Measurement. This project is sponsored jointly by the Nuclear Energy Agency of the Organisation for Economic Cooperation and Development, and the Radiation Protection Research Programme of the Commission of European Communities. In the report for Part I, conclusions were based primarily on the analysis of variance method, comparing within-laboratory (replication) errors to between-laboratory errors. Uncertainties were reported for each measurement; however, there was no additional use of this information. The reported uncertainties have been analysed more completely, leading to the conclusion that they are not a significant source of error. Further analyses were done using the two-way analysis of variance applied to the logarithms of the data. The findings reinforce and amplify those already reported: the results from three of the laboratories agreed within replication error, but the results from the fourth laboratory differed significantly (about 9%) from the other three. The overall unexplained variance, expressed as geometric standard deviation, was 1.05.
Read full abstract