I am an Assistant Professor of Political Communication in the Department of Communication Science at the Vrije Universiteit Amsterdam. I received my PhD in Political Science from the Vrije Universiteit Amsterdam. Before joining the Department of Communication Science, I was a postdoctoral researcher at the Institute of Political Science at the University of Zurich, and visiting researcher of the Departments of Political Science at the University of North Carolina, Chapel Hill and the University of Oxford.
My research interest comprise the areas of political communication, political behavior, and computational social science. I am motivated by key societal challenges that face democracies today, such as the crisis of representative democracy and increasing political fragmentation. Specifically, I apply advanced computational approaches to study the communication and rhetoric of politicians, and how this affects political decision making and its electoral consequences in multi-party systems.
In my current work, I examine the legitimacy of political decision-making. In one project, I study the electoral ramifications of (un)compromising politicians. This research project is funded through a Innovation Grant (VENI) by the Netherlands Organisation for Scientific Research (NWO). In another project, I look at the perceived legitimacy of goverment communication, which is funded through a National Science Agenda Grant by NWO. Moreover, as part of the OPTED H2020 consortium funded by the European Research Council, I work on methods to compare textual data in multilangual settings.
Abstract: The process of news consumption has undergone great changes over the past decade: Information is now available in an ever-increasing amount from a plethora of sources. Recent work suggests that most people would favor algorithmic solutions over human editors. This stands in contrast to public and scholarly debate about the pitfalls of algorithmic news selection - i.e. so-called “filter bubbles”. This study therefore investigates the reasons and motivations which might lead people to prefer algorithmic gatekeepers over human ones. We expect that people have more algorithmic appreciation when consuming news to pass time, entertain oneself or out of escapism than when using news to keep up-to-date with politics (H1). Secondly, we hypothesize the extent to which people are confident in their own cognitive abilities moderates that relationship: When people are overconfident in their own capabilities to estimate the relevance of information, they are more likely to have higher levels of algorithmic appreciation, due to the third person effect (H2). For testing those two pre-registered hypothesis , we conducted an online survey with a sample of 268 U.S. participants, and replicated our study using a sample of 384 Dutch participants. The results show that the first hypothesis cannot be supported by our data. However, a positive interaction between overconfidence and algorithmic appreciation in case the gratification of news usage is surveillance (i.e. gaining information about the world, society, and politics) was found in both samples. Thereby, our study contributes to our understanding of the underlying reasons people have for choosing different forms of gatekeeping when selecting news.
Abstract: Ever since the seminal campaign of Lyndon B. Johnson in 1964 where Daisy was introduced, emotions are present in political campaigns and other political processes. Their effect used to be understudied by scholars of political communication, however, until the early 2000s. Scholars have demonstrated ample times that making emotional appeals matters for their electoral strategy. Moreover, it has been shown that specific emotions elicit behavioral effects on citizens. This review shows how emotions are applied by political elites, their effects on citizens as well as how to study these emotional appeals in political communication.
Abstract: The coverage of politics, and more specifically policies or political issues, in news media has been abundantly studied by scholars of agenda setting (see for example, McCombs and Shaw, 1972, 1993; Baumgartner and Jones, 2010, 1991; Soroka, 1999; Walgrave and Van Aelst, 2016; Vliegenthart and Walgrave, 2011; Walgrave and Van Aelst, 2016; Baumgartner et al., 2006). Building on Walter Lippmann’s (1957) argument of the media’s ability to construct social realities in the public mind, agenda setting refers to the transfer of often covered topics in news media to its salience in the public agenda. McCombs and Shaw (1972) pioneered this field, surveying voters in North Carolina (USA) on the most important political issues and comparing these results to a media content analysis of nine local news media outlets. This finding has been coined the first-level agenda setting theory. Ever since the seminal study of McCombs and Shaw (1972), this finding has been replicated hundreds of times all across the world – ranging from other locations in the USA, to Europe, Asia, Latin America and Australia – for both election and non-election settings over a broad range of public issues and other aspects of political communication. Moreover, the agenda-setting theory has been extended from objects of attention to attributes, known as the second-level (McCombs, 1992; McCombs and Shaw, 1993; McCombs et al., 2014). From the second-level, it became apparent that “the media not only can be successful in telling us what to think about, they also can be successful in telling us how to think about it” (McCombs, 2005, p.546, emphasis in original). In the early 2010’s, the theory extended with a third-level (Guo et al., 2012; Guo and McCombs, 2011). This level includes a network component to the theory. In this chapter, we will describe the state-of-the-art of agenda-setting theory for the coverage of politics, and especially policies and political issues in media in three trends. Thereafter, we discuss the most common used research designs (pp.5–8), and we conclude with the limitations and possible future directions of the field (pp.8–10).
Abstract: Sentiment is central to many studies of communication science, from negativity and polarization in political communication to analyzing product reviews and social media comments in other sub-fields. This study provides an exhaustive comparison of sentiment analysis methods, using a validation set of Dutch economic headlines to compare the performance of manual annotation, crowd coding, numerous dictionaries and machine learning using both traditional and deep learning algorithms. The three main conclusions of this article are that: (1) The best performance is still attained with trained human or crowd coding; (2) None of the used dictionaries come close to acceptable levels of validity; and (3) machine learning, especially deep learning, substantially outperforms dictionary-based methods but falls short of human performance. From these findings, we stress the importance of always validating automatic text analysis methods before usage. Moreover, we provide a recommended step-by-step approach for (automated) text analysis projects to ensure both efficiency and validity.
Abstract: Studies investigating gender gaps in the doctoral training of political science students have focused so far overwhelmingly on the US context. Although important research within this context has made strides in identifying the persistent challenges to women’s incorporation in political methodology, much remains unknown about whether women and men have different experiences in methods training during their PhD programs. We contribute to this debate by analyzing data from an original survey on the methods-training experiences of political science PhD students at different European universities. We assess whether gender gaps exist with respect to PhD students’ methods training and confidence in employing methods skills. Our findings show that women cover significantly fewer methods courses in their doctoral training. When women do participate in methods training, they show levels of method employment similar to their male colleagues. We discuss the implications of these findings in the context of European doctoral training.
Abstract: We present a new dataset of speeches given by Danish and Dutch politicians at party congresses between 1946 and 2017. The dataset is a unique collection of materials from different party archives and digital repositories. It offers a unique opportunity to analyse the issues discussed in these speeches, the positions taken and the rhetoric used by party elites over time and between countries. We describe the data and illustrate them with one application: a sentiment analysis that describes differences between parties and over time.
Abstract: P-values are the most frequently employed metric to assess the significance of statistical findings in the social sciences. Since the earliest years of their usage the meaning and usefulness of p-values were topics of heated discussion (Berkson 1942; Fisher 1935). Lately the reproduction/replication crisis resuscitated this debate (Benjamin et al. 2018; Gelman 2018; Lakens et al. 2018; McShane et al. 2019; Nuzzo 2014; Trafimow and Marks 2015). Meanwhile, the skepticism has not stopped at the gates of political science. Most prominently the journal “Political Analysis” banned p-values “in regression tables or elsewhere” after the new editor took over the board of editors in 2017 (Gill 2018: 1).1 Also political scientists contributed to a swelling debate suggesting to lower the threshold for p-values to 0.005 (Benjamin et al. 2018; Esarey 2017). This present debate seeks to contribute to the discussion on p-values by summarizing the main arguments of it, providing an encompassing discussion of p-values – also from an epistemological perspective – as well as advice for the discipline about the Do's and Don'ts for p-values. In doing so it contributes to two ongoing debates: First, the actual meaning of p-values – the mathematical definition. Second, the potential to misuse p-values – even if correctly understood and defined. In February 2018 the Department of Political Science at the University of Zurich held a workshop containing public lectures about the use and usefulness of p-values. The goal of the workshop was to cover a wide spectrum of opinions on p-values, covering frequentist, bayesian and more epistemological views. The present contribution summarizes these public lectures but also goes beyond them by giving each author the opportunity to engage with the position of the remaining authors on the one hand and by adding a critical discussion of all lectures in the conclusion on the other hand. Our introduction first gives a brief history of p-values and their definition before summarizing the discussion about the use and “misuse” of p-values in the discipline and situates this into the larger debate on the replication crisis; Vera Troeger (2019) discusses the logic of statistical inference and significance testing and the implications of significance testing for empirical research; Susumu Shikano (2019) provides a detailed insight into two Bayesian approaches to hypothesis testing; Marco R. Steenbergen (2019) takes a stronger epistemological view on to p or not to p; and finally, Simon Hug (2019) critically discusses the contributions of this debate by emphasizing that none of the discussed approaches appear to provide a “silver bullet” for the issues they seek to address.
Abstract: Analyzing political text can answer many pressing questions in political science, from understanding political ideology to mapping the effects of censorship in authoritarian states. This makes the study of political text and speech an important part of the political science methodological toolbox. The confluence of increasing availability of large digital text collections, plentiful computational power, and methodological innovations has led to many researchers adopting techniques of automatic text analysis for coding and analyzing textual data. In what is sometimes termed the “text as data” approach, texts are converted to a numerical representation, and various techniques such as dictionary analysis, automatic scaling, topic modeling, and machine learning are used to find patterns in and test hypotheses on these data. These methods all make certain assumptions and need to be validated to assess their fitness for any particular task and domain.
Abstract: Do parties change their platform in anticipation of electoral losses? Or do parties respond to experienced losses at the previous election? These questions relate to two mechanisms to align public opinion with party platforms: (1) rational anticipation, and (2) electoral performance. While extant work empirically tested, and found support for, the latter mechanism, the effect of rational anticipation has not been put to an empirical test yet. We contribute to the literature on party platform change by theorizing and assessing how party performance motivates parties to change their platform in-between elections. We built a new and unique dataset of >20,000 press releases issued by 15 Dutch national political parties that were in parliament between 1997 and 2014. Utilizing automated text analysis (topic modeling) to measure parties’ platform change, we show that electoral defeat motivates party platform change in-between elections. In line with existing findings, we demonstrate that parties are backward-looking.
We identify three gaps that hamper the utility and progress of computational text analysis methods (CTAM) for social science research. First, CTAM development has given insufficient attention to social scientists’ concerns about measurement validity. Second, we identify a mismatch between the focus of many computational tools and that of social science research. Third, we argue that the dominance of English language tools depresses comparative research and inclusivity towards scholarly communities examining languages other than English. To substantiate our claims, we draw upon a content analysis of all research published in the top ranked journals in communications, political science, sociology, and psychology. Identifying a total of 854 articles between 2016 and 2020 that use quantitative text analysis, we examined studies' reliance on CTAM, what variables were measured, what validation efforts were undertaken, and what languages were present in the studied materials. We show that each gap contributes to explaining the uneven uptake of CTAM, and discuss how each gap has implications for research practice in the social sciences. In order to address these gaps, we propose a research agenda for CTAM development.
Negative campaigning is widely studied in the context of American and Western European elections. Studies that seek to explain when political actors "go negative" often locate these determinants in contextual or party characteristics, such as parties’ position in the polls, their incumbency status or political ideologically. Against the backdrop of our current high-choice media environment, a crucial element that is often overlooked is that parties can also weigh their rhetorical strategies across communication channels, based on the amount of journalistic intervention that takes place and the audience political parties can reach. Focusing on the 2017 Dutch General Elections, we examine how Dutch political parties use negative campaigning strategies across different communication channels. Analyzing 1490 appeals that appeared in newspaper articles, talk shows and in parties' Facebook posts, we show that in increase in journalistic intervention is associated with more negative appeals. Our findings highlight the importance of studying communication channels in comparative perspective and add more nuance to our understanding of rhetorical strategies during campaigns.
Coalition governments are the norm in West European democracies. Yet, we do not know whether governing together affects parties' calculations about setting their policy position in their next election manifesto. Therefore we ask: do coalition parties drift apart or stick together after a spell in office? We hypothesize that government parties stick together in anticipation of government continuation. Coalition parties are more likely to expect continuation if they are popular, have no inter-party conflict and have experience in co-governing. If this is not the case, coalition parties drift apart as they seek new coalition formation opportunities. We empirically substantiate our hypotheses with an innovative measure of party platform change analyzing 1,193 platform changes in 8 European democracies between 1968 and 2013 using new opinion poll data and several existing data sets. We demonstrate that parties plant the seeds of future coalition participation in their platforms: the more experienced and the more popular coalition parties are, the more they converge. Encountering conflict, on the contrary, leads them to diverge.