The Current Industrial Reports (CIR) imputation research effort was designed to provide a general assessment of the effectiveness of the existing imputation methodology for surveys in the CIR series, and to identify, evaluate, and compare the potential statistical and practical effects of plausible alternative procedures. This report provides a summary of the principal results of the completed research effort and present specific recommendations for potential improvement in the current imputation methodology for the surveys in the CIR series.
A major conclusion of the CIR imputation research was that the presently-used screened sums imputation methodology is reasonably effective for many of the CIR surveys, particularly for the monthly surveys. Consequently continuation of that procedure for the monthly surveys has been recommended. Moreover, the use of two distinct ratio adjustments for both the quarterly and annual surveys is suggested.