Activity Number:
|
615
- Using Para-Data to Analyze the Determinants and Impact of Interview Length
|
Type:
|
Topic Contributed
|
Date/Time:
|
Thursday, August 2, 2018 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #330283
|
|
Title:
|
Two Short or One Long: An Experiment Comparing Survey Length vs. Quantity of Surveys
|
Author(s):
|
Rebecca Powell* and Paul Biemer and Sarah Cook and Kathleen Considine and Carolyn Halpern and Kathleen Harris and Sarah Dean
|
Companies:
|
RTI International and RTI Internatinoal and RTI International and RTI International and Carolina Population Center-UNC and Carolina Population Center-UNC and Carolina Population Center-UNC
|
Keywords:
|
Questionnaire Design;
Survey Length;
Response Rates;
Response Quality
|
Abstract:
|
As survey researchers, we aim to maximize response rates and data quality while minimizing respondent burden. The National Longitudinal Study of Adolescent to Adult Health (Add Health) has gathered survey data from a national sample of adolescents since 1994. The most recent wave of the Add Health study transitioned from a face-to-face to a mixed-mode web/paper survey. This switch from a 90-minute interviewer-administered to a 55-minute self-administered survey shifted all the burden of completing the survey to the respondent. In an effort to reduce perceived burden and increase response, we conducted an experiment comparing survey length and quality by randomly assigning sample members to one of two experimental conditions: 1. Singular condition where respondents were asked to complete the entire questionnaire in one sitting, or 2. Modular condition where they were asked to complete two shorter questionnaires with the same content in two sittings. This paper presents the findings of the experiment, including completion rates and data quality indicators.
|
Authors who are presenting talks have a * after their name.