Survey questionnaires can serve to collect information about respondents’ behaviours, characteristics or opinions, and to put psychological research into a social context. But if we want to be able to draw meaningful conclusions from the data, rather than just having an arbitrary set of numbers to create fancy charts, we should spend sometime thinking about who to ask and how to ask them.
In this seminar we will address these issues using the Total Survey Error (TSE) framework, which considers all potential sources of error that affect data quality, both on the ‘how’ side (measurement error) and the ‘who’ side (representation error). Depending on factors such as topic, survey population or available resources, different sources of error will pose the greatest risks to data quality, and approaches to mitigating them will need to be tailored to the specific setting. We will combine theoretical considerations on how to minimise TSE with practical experience, as conducting high quality surveys is as much acraft as it is a science.
In this seminar we will address these issues using the Total Survey Error (TSE) framework, which considers all potential sources of error that affect data quality, both on the ‘how’ side (measurement error) and the ‘who’ side (representation error). Depending on factors such as topic, survey population or available resources, different sources of error will pose the greatest risks to data quality, and approaches to mitigating them will need to be tailored to the specific setting. We will combine theoretical considerations on how to minimise TSE with practical experience, as conducting high quality surveys is as much acraft as it is a science.
- Kursverantwortliche/r: Arne Bethmann
- Kursverantwortliche/r: Entwicklungs- und Pädagogische Psychologie Moodle
- Kursverantwortliche/r: Alexander Schumacher
Semester: SoSe 2024