Speaker: Ranjit K. Singh, GESIS - Leibniz Institute for the Social Sciences
Survey modes can impact response behavior in many ways. As surveys plan to switch modes or introduce mixed-mode approaches, this raises questions of data quality and comparability.
Ideally, we would like to retain seamless time series despite a mode change. And despite gathering responses in different survey modes, we would still like to provide researchers with a homogenous data product.
In this talk, Ranjit K. Singh will address this issue from the perspective of ex-post harmonization, which deals with assessing comparability issues and then attempting to mitigate them by transforming already collected data.
The focus will lie on single-item questions capturing concepts such as attitudes, values, or subjective assessments.
The presentation will focus on a pragmatic framework for assessing and improving comparability, which focuses on three issues:
- Is the same concept measured across modes?
- Is it measured with similar levels of precision (i.e., similar reliability)?
- Is it measured with similar numerical units?
For each issue, Ranjit K. Singh will discuss possible approaches to assessing comparability across survey modes as well as some ideas on how to mitigate comparability problems.
About the speaker
Ranjit K. Singh is a postdoctoral scholar at GESIS - Leibniz Institute for the Social Sciences, where he practices, researches, and consults on survey design and survey data harmonization. He has a background in both social sciences and psychology.
Research interests include measurement quality of survey instruments as well as assessing and improving survey data comparability with harmonization.