Activity Number:
|
114
- Survey Design and Data Adjustment Decisions in Mixed-Mode Surveys
|
Type:
|
Topic Contributed
|
Date/Time:
|
Monday, July 30, 2018 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #328757
|
|
Title:
|
Evaluating Data Quality in a Randomized Sequential Mixed-Mode Survey Experiment
|
Author(s):
|
Joseph Sakshaug* and Alexandru Cernat
|
Companies:
|
German Institute for Employment Research and University of Manchester
|
Keywords:
|
sequential mixed-mode;
total survey error;
measurement error;
nonresponse error
|
Abstract:
|
Mixing multiple modes of survey interview (e.g., telephone, mail, face to face) is becoming a standard practice in survey research. Nevertheless there are still open questions regarding the impact on data quality and best practices. In this paper we answer three essential questions related to using mixed mode surveys: 1) which is the best first mode to use in a sequential approach? 2) does implementing a second mode sequentially improve data quality? 3) does the mode order of the sequential design impact data quality? In order to answer these questions we used an experiment that randomized two sequential designs: mail-telephone and telephone-mail. Additionally we make use of administrative data and previous information to calculate data quality from a Total Survey Error perspective. We find that single mode mail outperforms telephone while the mail-telephone sequence has better estimates of data quality compared to telephone-mail. The analysis also indicates that the second mode improves the overall quality of the survey for both sequential designs.
|
Authors who are presenting talks have a * after their name.