Towards Determining an Optimal Contact Attempt Threshold for a Large‐Scale Personal Visit Survey

Adam Safir and Lucilla Tan

Abstract

While improving the design of the survey instrument and increasing the options for more convenient and/or efficient data capture are two important ways for maximizing returns on the survey operations budget, there is an additional dimension in personal-visit surveys for discretionary action that can make a difference - setting a maximum threshold for the number of contact attempts, or the level of effort, expended to resolve a case. In particular, the cost of contact attempts can be significant for personal-visit surveys with eligible sample units spread over an expansive geographic area. Before fielding a large-sample test of a proposed "optimal" contact attempt threshold, the authors initiated this study as a verification of findings from an earlier study (Safir and Tan 2009) that recommended 7 attempts as the threshold for resolving a sample unit's final disposition. Using more recent data and additional evaluation metrics, the authors perform a retrospective analysis of first wave panel data from the Consumer Expenditure Interview Survey (CEQ) collected between April 2012 and March 2014.