Comparing Two Methods for Managing Telephone Interview Cases

Introduction

In an interviewer-administered survey, interviewers can have a notable effect on data quality (West and Blom 2017). Studies have shown that there is an association between unit noncontact and nonresponse rates and interviewer characteristics and behaviors, such as experience levels and approaches to making contact (Blom 2012; Groves and Couper 1998; Purdon, Campanelli, and Sturgis 1999).

A standard practice for reducing nonresponse rates in interviewer-administered surveys is to vary the times of day and days of week of contact attempts, thereby maximizing the chances of making contact (Groves and Couper 1998). In most telephone surveys, automated call scheduling systems are used to help mitigate potential impacts of interviewers' decision-making on contact and cooperation. These systems use algorithms based on information about prior contacts, the time elapsed since the last call attempt, the time zone in which cases are located, scheduled appointments, and the overall status of each case to provide interviewers with a sequence of cases to dial. They also dictate when to leave a voice message and how often to schedule callbacks. As such, they control the timing, frequency, and number of times each case is dialed to ensure that cases receive a consistent number of calls at optimal time points.

When an automated call scheduler is not available, a case ownership model is an alternative approach for case management. In a case ownership model, telephone interviewers are each assigned a set of individual sample units (cases) and encouraged to use their judgment and experience to determine how best to work the cases. Interviewers are responsible for all aspects of managing their assigned cases, including reviewing case histories, deciding which of their assigned cases and phone numbers to call at any given time, making judgments about how often to attempt contact with a particular case, scheduling callbacks, and conducting interviews.

This chapter will describe our experiences implementing a case ownership approach for part of a national study of young adults. We also will compare the case ownership approach to the automated call scheduler approach used for a later stage of the study. Specifically, we compare the number of call attempts per case, variability in the timing of call attempts between the two approaches, and number of calls per interviewer hour worked. We identify strengths and weaknesses of the case ownership approach vis-a-vis the automated call scheduler approach, provide lessons learned for other researchers planning to utilize this approach, and identify areas for future research. As described in further detail below, the decision to use the case ownership approach was made for pragmatic reasons; this study is observational, not experimental. However, the similarity of the two data collections within the same larger study involving the same study cohort provides a unique opportunity to compare the two case management approaches.

 
Source
< Prev   CONTENTS   Source   Next >