Projects & Working Papers

Believing and Sharing Information by Fake Sources: An Experiment (Download)

The increasing spread of false stories (“fake news”) represents one of the great challenges societies face in the 21st century. A little understood aspect of this phenomenon, and the processing of online news in general, is how sources influence whether people believe and share what they read. In contrast to the pre-digital era, the Internet makes it easy for anyone to imitate well-known and credible sources in name and appearance. In a pre-registered survey experiment, we first investigate the effect of this contrast (real vs. fake source) and find that subjects, as expected, have a higher tendency to believe and a somewhat higher propensity to share news by real sources. We then expose subjects to a number of reports manipulated in content (congruent vs. incongruent with individuals' attitudes), which reveals our most crucial finding. As predicted, people are more likely to believe a news report by a source that has previously given them congruent information. However, this only holds if the source is fake. We further use machine learning to uncover treatment heterogeneity. Effects vary most strongly for different levels of trust in the mainstream media, and having voted for the populist right.

Visualizing causal scenarios and planned|realized measurements [interactively] (Download)

We explore ways of visualizing scenarios of and data on the temporal relation between treatment and response variables across time, across units or groups. The graph we develop may be used for systematic assessments of the role of time in causal relationships. We illustrate various insights it may produce and test potential applications relying on a combination of hypothetical examples and classics from the causal inference literature (e.g. Lalonde 1986, Card and Krueger 1994). As a proof-of-concept we supplement our study with an open-source application (based on R, Shiny and Plotly). This app should allow users to visualize examples as well as ‘draw’ their own scenarios interactively. Most of the graphs in this paper were made using the app and you can find an online version that is under development.

The “quality” of citations: Measuring the qualitative impact of research (Download)

The quantity of citations (“times cited”) has evolved into an influential indicator of scientific impact both in itself and packaged into other metrics (e.g. h-index, impact factor). In this study we contrast the idea of “quantity” with the idea of the “quality” of citations, i.e. the “quality” of impact. We develop and present methods that can be used to move from a superficial assessment of citation quantity to a more nuanced view of the quality of citations. We illustrate these methods using six highly cited study in the fields of political science, economics and sociology. In the future this more nuanced view and the data we are generating should allow for testing various hypotheses linked to the reception of scientific works and the sociology of science more generally. Our study is complemented by opensource code (based on R) that shall be collected in a R package CitationsR that allows other researchers to pursue their own analyses of the quality of impact of one or several studies.

Conceptualizing Trust and Trustworthiness (Download)

This study explores the meaning of the concepts of trust and trustworthiness. Despite the concepts' popularity and indisputable relevance, interested scholars face a conceptual 'jungle' that is hard to pervade. Building on and summarizing previous definitions and research, we attempt to provide a general definition for both concepts. This conception may serve as a starting point for future research, as well as a basis on which to analyse research done thus far. It is flexible enough to describe a wide variety of situations in which both concepts play a role and sets a clear boundary between the concepts themselves, their causes and their consequences. In addition, it helps to isolate trust and trustworthiness from other closely linked concepts (e.g. trusting behavior) and to systematically classify different subconcepts of trust.

Writing a reproducible paper in R Markdown (Download)

The present paper provides a template for a reproducible scientific paper written in R Markdown. Below I outline some of the "tricks"/code (e.g., referencing tables, sections etc.) I had to figure out to produce this document. The underlying files which produce this document can be downloaded here (see paper). I think I got pretty far but there is always room for improvement and more automatization, in parallel to the incredible developments in R and Rstudio (bookdown etc.). I intend to update this file when I discover more convenient code.

Does Suffering Suffice? An Experimental Test of Moore’s Retributivism


Michael S. Moore is among the most prominent normative theorists to argue that retributive justice, understood as the deserved suffering of offenders, justifies punishment. Moore claims that the principle of retributive justice is pervasively supported by our judgments of justice and sufficient to ground punishment. We offer an experimental assessment of these two claims, (1) the pervasiveness claim, according to which people are widely prone to endorse retributive judgments, and (2) the sufficiency claim, according to which no nonretributive principle is necessary for justifying punishment. We test these two claims in a survey and a related survey experiment in which we present participants (N = ~900) with the stylized description of a criminal case. Our results seem to invalidate claim (1) and provide mixed results concerning claim (2). We conclude that retributive justice theories which advance either of these two Moorean claims have weak evidential support.

Heterogeneity, Polarization and Conflict: Evidence from School Classes

The increasingly popular concept of polarization is used to describe various social phenomena such as ethnic, political, and income polarization. Scholars study ethnic polarization because they assume that it is linked to conflict. So far this relationship has been investigated relying on cross-country data. Evidence is mixed but suggests that bipolarity of equally-sized ethnic groups (polarization) is the most conductive scenario for conflict. While current research concentrates on the macro level, we argue that the assumed causal link is best studied in the setting of small groups. Consequently, we study the link on the level of school classes and analyze data from the Children of Immigrants Longitudinal Survey in Four European Countries (CILS4EU). Our sample contains around 800 high school classes located in the Netherlands, Germany, England, and Sweden and we explore how the ethnic composition of school classes relates to intergroup conflict. We construct a set of treated - ethnically polarized - and a set of - non-polarized - control units and estimate the average treatment effect on the treated (ATT). Our preliminary findings suggest that there is a causal effect of ethnic polarization on the prevalence of conflict among classmates. The results also suggest that it is ethnic polarization, rather than heterogeneity per se, that is the driving force behind intergroup conflict. A real-time observatory

A pipeline that scrapes politicians' twitter data and produces a real-time report analyzing their activity and the content of their tweets. The project feeds the twitter bot tp_stats that tweets statistics on a daily basis.

Applied Causal Analysis (with R) (link)

An increasingly elaborate manuscript to teach students how to investigate causal questions using R. Includes some "novel" gimmicks such as 3d distributions to convey complex concepts.

Conceptualizing and measuring polarization: A review (Download)

The increasingly popular concept of polarization is used to describe various social phenomena such as political, opinion, health and income polarization. Despite this popularity it is still debated within disciplines how polarization should be conceptualized and how it should be measured. In this study I attempt to provide a systematic, interdisciplinary overview of the conceptual and measurement literature on polarization. I first describe the challenge of conceptualizing polarization, a task that requires taking decisions about how to aggregate individual positions on one or more scales. Different distributional aspects may matter and the concept’s meaning is related to the measurement level of the underlying scales. Subsequently, I review various polarization measures that have been developed during the last decades.