Journal of Official Statistics, Vol.10, No.2, 1994. pp. 133–152
Cost-Efficiency and the Number of Allowable Call Attempts in the National Health Interview Survey
William D. Kalsbeek, Steven L. Botman, James T. Massey, and Pao-Wen Liu
Abstract:An important design decision for interview surveys is the number of call attempts allowed in obtaining a response from sample members. Considerations of survey cost favor a small number of allowable attempts, while the desirability of a high response rate points to a large number. An empirical assessment of the statistical efficiency for set cost among call attempt options sheds some light on this issue as it pertains to the U.S. National Health Interview Survey. Cost data were obtained from a special assessment on a subsample of the NHIS, and efficiency is measured by the estimated mean squared error of various estimates from the NHIS sample. While the current practice of allowing up to 15 call attempts is found to be the best overall, the cost-efficiency of cutoffs with as few as six attempts is nearly as good.
Keywords:Nonresponse; callbacks; survey cost; design optimization.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group