Research on bias in Surveys
Together with Jocelyn Evans, I have developed a new measure for bias in the survey-based measurement of multinomial variables (ungated pre-print of our article in Political Analysis). We have also created a free add-on for Stata, surveybias. The latest version (1.4) has many new features and runs considerably faster than the original one and is documented in our article in the Stata Journal. Click here to see what surveybias can do for you. One of these days, I will start porting surveybias to R.
Spatial Analysis in Political Science
It is debatable whether all politics is local, but local living conditions (in the broadest sense) should have an impact on political attitudes and political behaviour. This is why I am very much interested in research methods that bring spatial information into models of electoral choice. Jocelyn Evans and I have published an analysis of the effect of candidate localness in the 2010 General election in England, and another one on the same phenomenon (physical distance between candidates and voters) in the 2013 County Council election. Together with Rosie Campbell and Phil Cowley, we have also replicated and extended our analyses for the 2015 General election in England.
In more recent work, we study (together with teams in Amsterdam, Leeds and Nice) the effect of local living conditions on anti-immigrant attitudes, populism, and radical right-wing voting.
Sociology of Knowledge
I have been interested in citation analysis for a very long time. Citation research is an excellent means for uncovering social and intellectual structures in science. Together with Harald Schoen, I have analysed the structure of publication networks in German Political Science. More recently, I have looked into the usage of the terms “extreme” and “radical” by various authors in a bid to ascertain whether there are separate schools in European Radical Right Studies (TM).
Measurement and Postal Panel Response Rates
Some of my very early work is on the adequate measurement of value orientations and on improving the response rate in mail panel surveys (that was before the internet).