Mariken van der Velden | Vrije Universiteit Amsterdam (original) (raw)
Papers by Mariken van der Velden
Computational Communication Research
Following the rapid expansion of digital textual data, which has permeated every last domain of s... more Following the rapid expansion of digital textual data, which has permeated every last domain of social reality, the comparative turn in textual analysis is finally picking up speed. Capitalizing especially upon recent advances in computational language technologies-notably, machine translation and large language models-social scientists focused on the study of textual data are beginning to discover the rich opportunities of cross-cultural, cross-national and cross-linguistic comparative research. Yet, as they do so, they are also rapidly discovering a whole host of challenges that arise from the study of multilingual text. Recent years have witnessed numerous important developments in the methodological literature on multilingual computational text analysis. These approaches range from dictionaries (e.g.
In this project we will test whether ethnic or civic conceptions of the 'people' matter f... more In this project we will test whether ethnic or civic conceptions of the 'people' matter for citizens' adherence to (the constitutive components) of populist attitudes. Whereas in some countries, the concept 'the people' has a positive or neutral connotation, in other contexts the concept of 'the people' is tainted by historical usage. In German, the term 'das Volk' is arguably tainted by its usage in Nazi-Germany. In this project, we conduct a wording experiment to test whether ethnic or civic wordings of items in the populist attitudes construct affect respondents' adherence, and ultimately, their degree of populist attitudes.
71st ICA Annual Conference, 2021
This project examines the effects of the willingness to accept political compromise on trust in p... more This project examines the effects of the willingness to accept political compromise on trust in political parties.
We present a new dataset of speeches given by Danish and Dutch politicians at party congresses be... more We present a new dataset of speeches given by Danish and Dutch politicians at party congresses between 1946 and 2017. The dataset is an unique collection of materials from different party archives and digital repositories. It offers an unique opportunity to analyze the issues discussed in these speeches, the positions taken, and the rhetoric used by party elites over time and between-country. We describe the data, and illustrate this with two applications: [1] a sentiment analysis describes differences between parties and over time; and [2] a topic model illustrates the increasing prevalence of the EU issue, and the rise and fall of the prevalence of economic issues
Oxford Research Encyclopedia of Politics, 2019
Analyzing political text can answer many pressing questions in political science, from understand... more 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.
Communication Methods and Measures, 2021
ABSTRACT We identify three gaps that limit the utility and obstruct the progress of computational... more ABSTRACT We identify three gaps that limit the utility and obstruct the progress of computational text analysis methods (CTAM) for social science research. First, we contend that CTAM development has prioritized technological over validity concerns, giving limited attention to the operationalization of social scientific measurements. Second, we identify a mismatch between CTAMs’ focus on extracting specific contents and document-level patterns, and social science researchers’ need for measuring multiple, often complex contents in the text. Third, we argue that the dominance of English language tools depresses comparative research and inclusivity toward scholarly communities examining languages other than English. We substantiate our claims by drawing upon a broad review of methodological work in the computational social sciences, as well as an inventory of leading research publications using quantitative textual analysis. Subsequently, we discuss implications of these three gaps for social scientists’ uneven uptake of CTAM, as well as the field of computational social science text research as a whole. Finally, we propose a research agenda intended to bridge the identified gaps and improve the validity, utility, and inclusiveness of CTAM.
While the Bayesian parameter estimation has gained a wider acknowledgement among political scient... more While the Bayesian parameter estimation has gained a wider acknowledgement among political scientists, discussion about the Bayesian hypothesis testing is still not enough visible. This paper introduces two Bayesian approaches to hypothesis testing: one based on the Bayesian credible intervals and the other based on the Bayes factor. By using an example based on a linear regression model, I demonstrate similarities and differences not only between the null-hypothesis significance tests and Bayesian hypothesis tests, but also those among two different Bayesian
Routledge International Handbook of Emotions and Media, 2021
Media and Communication, 2021
The process of news consumption has undergone great changes over the past decade: Information is ... more 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., the so-called “filter bubbles.” This study therefore investigates 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 to moderate 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...
Communication Methods and Measures, 2021
Sentiment is central to many studies of communication science, from negativity and polarization i... more 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-bystep approach for (automated) text analysis projects to ensure both efficiency and validity.
Swiss Political Science Review, 2019
PS: Political Science & Politics, 2019
Studies investigating gender gaps in the doctoral training of political science students have foc... more 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...
Research & Politics, 2019
We present a new dataset of speeches given by Danish and Dutch politicians at party congresses be... more 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.
Political Communication, 2017
Do parties change their platform in anticipation of electoral losses? Or do parties respond to ex... more 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 backwardlooking.
We do know people value algorithms as a selector of the news (Thurman et al. 2019). The recent st... more We do know people value algorithms as a selector of the news (Thurman et al. 2019). The recent study of Thurman suggest that people view algorithms as more neutral compared to editors and recommendation from friends. This is in contrast with what is typically argued by scholars. Namely, that algorithmic selection could feature into a so-called filter bubble or more generally, limit the diversity of a media diet. Yet, we don't know whether this algorithmic appreciation comes from from different needs to consume news. We aim to bridge people's algorithmic appreciation uses and gratification of news.
Computational Communication Research
Following the rapid expansion of digital textual data, which has permeated every last domain of s... more Following the rapid expansion of digital textual data, which has permeated every last domain of social reality, the comparative turn in textual analysis is finally picking up speed. Capitalizing especially upon recent advances in computational language technologies-notably, machine translation and large language models-social scientists focused on the study of textual data are beginning to discover the rich opportunities of cross-cultural, cross-national and cross-linguistic comparative research. Yet, as they do so, they are also rapidly discovering a whole host of challenges that arise from the study of multilingual text. Recent years have witnessed numerous important developments in the methodological literature on multilingual computational text analysis. These approaches range from dictionaries (e.g.
In this project we will test whether ethnic or civic conceptions of the 'people' matter f... more In this project we will test whether ethnic or civic conceptions of the 'people' matter for citizens' adherence to (the constitutive components) of populist attitudes. Whereas in some countries, the concept 'the people' has a positive or neutral connotation, in other contexts the concept of 'the people' is tainted by historical usage. In German, the term 'das Volk' is arguably tainted by its usage in Nazi-Germany. In this project, we conduct a wording experiment to test whether ethnic or civic wordings of items in the populist attitudes construct affect respondents' adherence, and ultimately, their degree of populist attitudes.
71st ICA Annual Conference, 2021
This project examines the effects of the willingness to accept political compromise on trust in p... more This project examines the effects of the willingness to accept political compromise on trust in political parties.
We present a new dataset of speeches given by Danish and Dutch politicians at party congresses be... more We present a new dataset of speeches given by Danish and Dutch politicians at party congresses between 1946 and 2017. The dataset is an unique collection of materials from different party archives and digital repositories. It offers an unique opportunity to analyze the issues discussed in these speeches, the positions taken, and the rhetoric used by party elites over time and between-country. We describe the data, and illustrate this with two applications: [1] a sentiment analysis describes differences between parties and over time; and [2] a topic model illustrates the increasing prevalence of the EU issue, and the rise and fall of the prevalence of economic issues
Oxford Research Encyclopedia of Politics, 2019
Analyzing political text can answer many pressing questions in political science, from understand... more 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.
Communication Methods and Measures, 2021
ABSTRACT We identify three gaps that limit the utility and obstruct the progress of computational... more ABSTRACT We identify three gaps that limit the utility and obstruct the progress of computational text analysis methods (CTAM) for social science research. First, we contend that CTAM development has prioritized technological over validity concerns, giving limited attention to the operationalization of social scientific measurements. Second, we identify a mismatch between CTAMs’ focus on extracting specific contents and document-level patterns, and social science researchers’ need for measuring multiple, often complex contents in the text. Third, we argue that the dominance of English language tools depresses comparative research and inclusivity toward scholarly communities examining languages other than English. We substantiate our claims by drawing upon a broad review of methodological work in the computational social sciences, as well as an inventory of leading research publications using quantitative textual analysis. Subsequently, we discuss implications of these three gaps for social scientists’ uneven uptake of CTAM, as well as the field of computational social science text research as a whole. Finally, we propose a research agenda intended to bridge the identified gaps and improve the validity, utility, and inclusiveness of CTAM.
While the Bayesian parameter estimation has gained a wider acknowledgement among political scient... more While the Bayesian parameter estimation has gained a wider acknowledgement among political scientists, discussion about the Bayesian hypothesis testing is still not enough visible. This paper introduces two Bayesian approaches to hypothesis testing: one based on the Bayesian credible intervals and the other based on the Bayes factor. By using an example based on a linear regression model, I demonstrate similarities and differences not only between the null-hypothesis significance tests and Bayesian hypothesis tests, but also those among two different Bayesian
Routledge International Handbook of Emotions and Media, 2021
Media and Communication, 2021
The process of news consumption has undergone great changes over the past decade: Information is ... more 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., the so-called “filter bubbles.” This study therefore investigates 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 to moderate 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...
Communication Methods and Measures, 2021
Sentiment is central to many studies of communication science, from negativity and polarization i... more 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-bystep approach for (automated) text analysis projects to ensure both efficiency and validity.
Swiss Political Science Review, 2019
PS: Political Science & Politics, 2019
Studies investigating gender gaps in the doctoral training of political science students have foc... more 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...
Research & Politics, 2019
We present a new dataset of speeches given by Danish and Dutch politicians at party congresses be... more 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.
Political Communication, 2017
Do parties change their platform in anticipation of electoral losses? Or do parties respond to ex... more 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 backwardlooking.
We do know people value algorithms as a selector of the news (Thurman et al. 2019). The recent st... more We do know people value algorithms as a selector of the news (Thurman et al. 2019). The recent study of Thurman suggest that people view algorithms as more neutral compared to editors and recommendation from friends. This is in contrast with what is typically argued by scholars. Namely, that algorithmic selection could feature into a so-called filter bubble or more generally, limit the diversity of a media diet. Yet, we don't know whether this algorithmic appreciation comes from from different needs to consume news. We aim to bridge people's algorithmic appreciation uses and gratification of news.