Abstract：APMP attenuation key comparison, APMP.EM.RF-K19.CL linked the comparison results to the corresponding CCEM comparison. The calculation method of weighted mean was used in the calculation of the linking correction. The method of median of absolute deviation was used for identifying statistical outlier from the linking data. It was considered whether measurement results in CCEM comparison of the linking laboratories were validity as well. The linking results show that the calculation result of linking correction is reliable and the evaluation of the uncertainties is reasonable.
［1］Delahaye F, Witt T J. Linking the Results of Key Comparison CCEM-K4 with the 10 pF Results of EUROMET Project 345［J］. Metrologia, 2002, 39:01005.
［2］Reedtz G M, Cerri R. Linking the Results of Key Comparisons CCEM-K8 and EUROMET.EM-K8［J］.Metrologia,2004,41(1A):01002.
［3］Randa J. Proposal for KCRV and degree of equivalence for GTRF key comparisons［R］. NIST,GT-RF/2000-12,2000.
［4］Randa J. Update to proposal for KCRV & degree of equivalence for GTRF key comparisons［EB］. http://www.bipm.org/wg/CCEM/GT-RF/Allowed/18/GTRF-05-04.pdf.
［5］Eio C,Howes J. Final report on CCEM key comparison CCEM.RF-K19.CL (GT-RF/00-2): Attenuation at 60 MHz and 5 GHz using a Type N step attenuator［J］.Metrologia,2009,46(1A):01006.
［6］Gao Qiu-lai, Liang Wei-jun, Lee Joo-gwang,et al. Final report on APMP attenuation key comparison APMP.EM.RF-K19.CL: Attenuation at 60 MHz and 5 GHz using a Type N step attenuator［J］. Metrologia,2004,47(1A):01015.
［7］Bachmair H. CCEM Guidelines for Planning, Organizing, Conducting and Reporting Key, upplementary and Pilot Comparisons［EB］.http://www.bipm.org/utils/common/pdf/ccem_guidelines.pdf.
2012, Vol. 33 
