Using Digital Trace Data in the Social Sciences, University of Konstanz (Summer 2018)

Instructor: Andreas Jungherr

Week 6: How to Find A Research Question? & Data Lab

After getting the hang of collecting data on Twitter and preparing them for analysis, it is time for you to design your own research project. As always, the best research questions are not purely data driven or solely motivated by opportunities provided by access to specific data sets. Instead, make sure to anchor your research design within larger questions connected with political or social science.

One approach to finding a promising question might be: “What aspect of social or political life is closely connected with political Twitter activity and might, therefore, be illustrated by data collected on Twitter?”

Also, in checking up on the promise of your project you might ask yourself:

“Given the findings presented in my awesome study, what have we learned about the world that we didn’t know before?”

(Also known in some quarters as the Rasmus-question).

In the beginning, this might seem a little awkward or challenging but stick with it. If you will not ask these questions someone else will. This way, you’ll make sure you have a good answer once you are asked in front of a room full of people. Also, choosing projects based on answers to these questions will make for rewarding projects and ultimately better chances for publication.

If you get stuck in thinking about promising research questions have a look at Howard S. Becker’s classic Tricks of the Trade: How to think about your research while you’re doing it. This should get you unstuck pretty quickly.

Required Readings:

Before you settle on a question, make sure to read up on what has already been done with Twitter data. A helpful overview on Twitter-based research on electoral campaigns can be found in the required readings for this session:

Background Readings:

How to Find a Research Question?

Conceptual Issues in Working with Digital Trace Data:

After getting a first overview on some of the work that has been done on and with Twitter in political contexts, it might pay off to read up on some of the conceptual issues in the use of digital trace data for social science research. Here is a short list of papers offering you a good window into current methodological debates.

Case Studies Illustrating Different Approaches to the Use of Twitter data:

Literature reviews and conceptual debates are all well and good but nothing stimulates your intuition as reading primary research directly. For this purpose, I provided you with a slightly longer list on innovative studies focusing on political uses of Twitter or using Twitter data in research. Of course, this can only be a small selection and is by no means exhaustive. Still, this should provide you with a running start.

Public Datasets:

In case you are interested in Twitter activities of German politicians there are potential some short-cuts available to you. The GESIS has published two dataset documenting all publicly available tweets published by candidates running in the 2013 and 2017 German Federal Election. The datasets also contains mentions by other Twitter users of these politicians and a set of tweets containing topically relevant hashtags.

Week 5Week 8