|
Activity Number:
|
128
- Impact of Data Collection Modes and Data Sources
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 31, 2017 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #324908
|
|
|
Title:
|
Adjusting for Mode Effects in Longitudinal Studies Utilizing a Bayesian Prior on Within Subject Correlation
|
|
Author(s):
|
Heather Kitada* and Sarah C Emerson and Claudio Fuentes
|
|
Companies:
|
Oregon State University and Oregon State University and Oregon State University
|
|
Keywords:
|
Mode Effects ;
Mode Adjustment ;
Bayesian ;
Longitudinal Studies
|
|
Abstract:
|
Mixed-mode surveys have grown in popularity due to concerns of falling response rates and coverage bias; however, treatment of the inherent mode effects threaten the integrity of these studies. There is an ongoing debate in the survey community about the validity of performing mode adjustments. While several diverse attempts have been made to adjust for survey mode effects with one time period, the literature reveals a demand for the development of theory which extends to longitudinal studies. Adjustment methods are based on the framework of creating counterfactuals of reassigning respondents to one mode in order to establish consistency for ease of comparison over time. These procedures assume either independence or maximal dependence between an individual's hypothetical responses to different modes. Although individual effects are inestimable within a single survey study, the definition and treatment of correlation structure may affect the conclusions we draw and how accurate they are. In this paper we explore a Bayesian prior on the within subject effect to create a flexible unified adjustment model that explicitly parametrizes the within-respondent correlation.
|
Authors who are presenting talks have a * after their name.
