What is your current research topic?
I am especially interested in survey methodology and measurement error, i.e., how to survey persons, which challenges and error sources this entails, how we can improve our surveys, and not least, how we can generate better quality data. Currently we are for example looking at participation and attrition across the entire runtime of the GIP – almost 12 years now – to see if we are losing certain groups of respondents at different rates and whether certain characteristics of the panel waves are associated with higher attrition rates. We of course hope that we may be able to identify certain red flags for imminent attrition and may then be able to intervene before respondents decide to no longer participate.
For those who have not yet delved deeply into the topic of Data Science: How would you explain to a child what you are working on?
Researchers from different disciplines would like to ask the people in Germany about their opinions on various topics. The GIP team works on making this possible. To do that, we regularly interview the same people online, who were randomly selected from the population. We also help researchers develop their idea into a finished questionnaire, monitor that everything works as intended, and give the people’s answers to the researchers for their analyses.
Everyone talks about Data Science – how would you describe the importance of the topic for yourself in three words?
Discovering hidden patterns
What points of contact with Data Science does your work have? Which methods do you already use, and which would be interesting for you in the future?
In the GIP we come into contact with a diverse set research projects from different disciplines and with different substantive interests and methodological approaches, from conventional survey studies to complex modulation using open answers or paradata. Besides the opportunity to include new survey questions in the GIP, there is also a lot left to discover in the available GIP data. For example, participation patterns or the open comments on the survey: Which points are brought up more often, which may be related to higher levels of attrition?
How high is the value of Data Science for your work? Would your research even be possible without Data Science?
Collecting survey data is only useful if somebody wants to work with them in the end. Points such as long-term archival, documentation, availability, and usability of data for secondary analyses and replications, in other words FAIR data principles, are also gaining more traction recently; and rightfully so. Because of this, the usefulness of survey data is in many cases no longer limited to a single research project.
What development opportunities do you see for the topic of Data Science in relation to your field?
Recently, connecting survey data to other types of data, such as digital behavioral data has become a huge topic of interest. Researchers hope that these passively collected data may give us a more complete and maybe more objective picture of our respondents than their self-reports while at the same time reducing the amount of surveying and thus the cognitive burden answering questionnaires puts on our respondents.
Automated text analysis has been another big topic for a while now, for example from social media posts or open comments on the survey to find out which points are brought up more frequently, which things bothered respondents, etc. The sheer amount of data often prohibits us from having a human analyze these “by hand”.