Presentations at International Conferences
OPUS Graduate Seminar Series
- Populism monitoring via computer-based methods for data extraction, cleaning and analysis.
Nowadays, the increase of internet use has led to the massive production of internet-based big data. The idea of the vast amount of unexploited information has motivated the academia, companies and institutions to search for the best practices to make the most out of this data. Most of the time these data are unstructured or semi-structured. The most common form of unstructured data in the Internet is text. This has led to the development of innovative methods and practices by the scientific community for the effective and comprehensive extraction, transformation, visualization and analysis of text data, a field described as “text analytics”.
For the purposes of this paper, my aim is to study populism on 6 Greek news companies’ websites utilizing various text analytics methods and tools. I have selected six Greek news media companies and I have collected 4778 articles in total from their websites in the period from 1 to 27 of February of 2022. Data collection was made through the development and application of a web scraping algorithm, which extracts and stores the content from all the links contained in each website.
After that I used text cleaning methods in order to clean the data from much “noise” such as punctuation, stop words etc. and prepare them for analysis. Then I used a series of computer-based approaches for the most effective detection of articles containing populist content, namely: sentiment analysis, word frequencies and an HITL (Human-In-The-Loop) machine learning algorithm.
My aim is to find the best available method that scholars could use for the automated detection of populism on the Internet, by highlighting the positives and negatives of each one.
Populism monitoring via computer-based methods for data extraction, cleaning and analysis.
ECPR General Conference, 22 – 26 August 2022, University of Innsbruck
- Unpacking the interplay between populism and euroscepticism: towards a new operationalisation.
Populism and euroscepticism are frequently studied together as the two phenomena often appear inextricably intertwined in contemporary European politics. Even so, the linkage between populism and euroscepticism has yet to be thoroughly and convincingly identified and analysed in empirical research. The main aim of our paper is to investigate this linkage by focusing on survey items that have been used over time in different studies that measure both populism and euroscepticism. We focus on relevant items that have been included in DATAPOPEU surveys, fielded as part of the DATAPOPEU Research Project, an endeavor which focuses on the interplay between the two phenomena. Based on the analysis of these items, we construct two indices, one for populism and one for euroscepticism. Then, we depict ways in which populist attitudes align with eurosceptic attitudes. Our findings reveal a strong correlation between populism and euroscepticism for Greek voters in 2019. The correlation appears stronger when political interest is high and among those that negatively evaluate government performance..
Unpacking the interplay between populism and euroscepticism: towards a new operationalisation
- Comparison of the positions of voters and candidates on populism: the Greek case
Populism has most often been examined from the supply-side of politics, as scholarship primarily focused on the rhetoric of populist movements, parties, and leaders. Most empirical research before the 2010s was based on qualitative approaches (e.g., Betz, 1994; Taggart, 2000; Mudde, 2007). Since then, an increasing number of studies take a more quantitative approach, as they delve into the study of both the supply- and demand-side of populism by using survey items to measure populist attitudes. Based on an ideational approach, that is, the belief that populism is a set of ideas, populism can be measured as an attitude that individuals can possess to a greater or a lesser extent (Rooduijn 2018: 364). This approach paves the way for examining populism, both at the demand and the supply side, with quantitative methods. Recent scholarship uses survey items to gauge levels of populism amongst individuals. These individual-level studies on populist attitudes form the foundation of the elite surveys and the expert surveys on populism.
This paper intends to study and compare the positions of Greek voters, candidates and political parties on a number of issues related to populism through quantitative research. Relying on the data of 2019 Greek national election studies (i.e., candidate and voter surveys) and an expert survey we aim to answer the following research question: do voters share the same views on issues related to populism as the candidates and political parties they vote for? In addition, studying their opinions on immigrants, minority rights, social liberalism etc. we will identify right-wing and left-wing populism in both voters and political parties.
However, as with any survey, some survey items may not work, so this paper’s goal is also to identify the survey items of populism that work and those that do not.
In this paper we use Greece as a case study; however, as many of the survey items used in these post-election studies have been used in surveys in other countries, this work contributes to the comparative research of populism and aims at encouraging scholars from other countries to include more survey items related to populism in their future election studies.
Comparison of the position of voters and candidates on populism: the Greek case
- Populism monitoring: web (text) data collection, cleaning and analysis.
Nowadays, the increase of internet use has led to the massive production of internet-based big data. The idea of the vast amount of unexploited information has motivated the academia, companies and institutions to search for the best practices to make the most out of this data. In most cases, data in big data environments (or data lakes) are not necessarily structured, meaning they do not follow a relational structure (i.e. tables with columns and rows). Most of the time they are unstructured or semi-structured. The most common form of unstructured data in the Internet is text. This has led to the development of innovative methods and practices by the scientific community for the effective and comprehensive extraction, transformation, visualization and analysis of text data, a field described as “text analytics”.
For the purposes of this paper we aim to study populism on social media (Twitter) and news websites, utilizing various text analytics methods and tools. We have chosen six Greek newspapers and we have collected articles from their websites and tweets posted by their official Twitter accounts in the same period of time. In this manner, we are able to compare the findings of media monitoring when we use two different data sources.
At first, we collected tweets and news articles containing the stem words “people” and “popul”. Twittter data were collected through API requests and news websites’ data through web scraping. Then, we transformed the collected data into a tidy format and finally we stored the transformed data into a local database.
After that we used text cleaning methods in order to clean the data from much “noise” such as punctuation, stop words etc. and prepare them for analysis. Then we used a series of computer-based approaches for the proper classification of the articles as populist (or not), namely: sentiment analysis, word frequencies and an HITL machine learning algorithm.
Our aim is to find the best available method that scholars could use for the automated detection of populism on the Internet, by highlighting the positives and negatives of each one. Also by comparing two different data sources (Twitter and websites) we show the differentiation in populist content someone could extract from those two.
Populism monitoring: web (text) data collection, cleaning and analysis
- Measuring Populism and Euroscepticism with surveys: The DATAPOPEU Political Compass
For many years studies on populism and euroscepticism operated predominantly on the basis of applying text analysis methods on party manifestos and speeches by party leaders. Only recently, there have been studies that try to cover both the supply and the demand side of populism and euroscepticism by including batteries of items in survey questionnaires, while there is also an increasing number of expert surveys focusing on both research directions At the focus of the DATAPOPEU Research Project funded by the Hellenic Foundation for Research & Innovation is the interplay between populism and euroscepticism. In this paper, we examine the survey items that have been used in several surveys for various target groups (voters, candidates and experts) within the DATAPOPEU project and we present the Populism and Euroscepticism Political Compass that is based on a selection of the best survey items. This is a web tool that people all over the world can use to find their own position on the populist scale while people who live in one of the EU member states or who are interested in the EU can use it to find their position both on the Populism and on the Euroscepticism scale.
Unpacking the interplay between populism and euroscepticism: towards a new operationalisation
ECPR Virtual General Conference, 24 – 28 August 2020
- Presentation at the ECPR virtual conference on conducting online experiments for the 2019 Parliamentary Candidate Survey in order to collect high quality data.
Web-surveys are being used more and more often in social sciences, as a fast and low-cost mode of data collection. However, there are some serious drawbacks which are mainly related to the absence of an interviewer such as low response rates. Furthermore, the length of the survey instrument affects considerably the response behavior. There is evidence that lengthy online questionnaires lead to lower response rates and lower quality responses. The aim of this paper is to study whether different web survey designs affect the response rate of a candidates’ survey. Specifically, during data collection for the Greek candidate study of 2019 we conducted three web experiments to test the impact on response behavior of: i) questionnaire length ii) web survey layout and iii) candidates recruitment. In order to test the impact of each factor we have manipulated the conditions of the survey in several ways. First, we divided the questionnaires into two or three parts in order to test whether shorter questionnaires increase response rate (splitting design). Second, we conducted an experiment on whether the layout of the survey affects the response rate. Specifically, there were two different designs of the questionnaire namely single page question and grid. Furthermore, some surveys were optimized for mobile devices to test if this affects the response rate. The final experiment is about the candidates’ recruitment. Most of the Greek candidate MPs have e-mail addresses which are available online especially during the period of electoral campaign. We collected their e-mail addresses using search engines and visiting websites related to the candidates or to the Greek elections in general. We also got in contact with the candidates whose e-mail could not be collected, through their Facebook pages and personal accounts.
Using Web Surveys to Collect Data from Candidates: Experience from the Greek Candidate Survey
- Presentation at the ECPR virtual conference on congruence between candidates MPs and voters in Greece over the last decade.
Congruence between voters’ and candidates’ policy position is a key element of political representation. This paper intends to study congruence between candidate MPs and their party voters in Greece during the last decade (2009-2019). After a rather stable period (after the democratic transition in 1974), the Greek party system in the last decade has experienced many alterations and transformations. Old traditional parties have lost their electoral power while new and, in some cases, more extreme parties have appeared in the Greek party system, especially after 2012. These changes could have affected the linkage between political elites and the electorate and as a result the level of congruence in Greece could be at stake. Examining this political phenomenon on different dimensions of political competitiveness (economic Left/Right, GALTAN, pro/anti-European) we answer the following research question: how has congruence been developed in the Greek political context during the last ten years? In the past congruence was typically studied by comparing the attitudes of voters with what opinion polls or panels of experts considered to be the attitudes of politicians or the positions of parties. In 2010, Golder and Stramski introduced a many-to-man approach, and then many recent studies have followed this approach. Relying on the data of Comparative Candidates Survey (CCS), in line with CSES voter study we study congruence between the voters’ and candidates’ policy position. Specifically, we include the Greek studies of 2009, 2012, 2015 and the recent study of 2019. In this paper we use Greece as a case study; however, this work contributes to the comparative research on congruence and to a further development of the CCS project. Constituting an archive with common variables, coding and data of all the CCS studies in line with voter studies is an important element for other potential studies and comparative analysis in the future.
Exploring Congruence in Greece (2009-2019).
- Presentation at the ECPR virtual conference on the impact of MP tweets on their electoral success
As candidates are increasingly using Twitter in their political communication, a question about its effectiveness as a political marketing tool is raised. Of course, along Twitter usage, there are more “traditional” political campaigning tools such as door-knocking, visiting businesses and social organizations etc. The aim of this paper is to study the factors that could influence the electoral performance of the candidate MPs and whether Twitter usage is one of them. To this aim we use information about candidates’ previous political activity, electoral campaign tools, money spent on campaign etc acquired through survey data. We combine this dataset with data produced by candidates’ Twitter activity, to study their effectiveness on the 2019 Greek Parliamentary election results. We are studying which are the factors with the greater influence on preferential votes each candidate gets in his constituency. The first data source is the Hellenic Candidate Study 2019. This dataset provides useful information about candidates’ political campaigns such as campaign spending, campaign means etc. The other source is Twitter activity data. We also measure candidates’ popularity using two different approaches: i) Wikipedia (number of hits on each candidate’s Wikipedia page), ii) experts’ evaluation. We are particularly interested in candidates’ Twitter audience. We apply an innovative method of analysis of candidate’s network starting from retweeters for each candidates’ post. We also take into account the followers of each retweeter and the candidate’s followers. In this way we explore whether the size of the Twitter audience of a candidate is related to the electoral success, or in other words, whether larger network sizes are associated with more successful candidates relatively to the party share in each electoral constituency. First indications suggest that the size of candidates’ Twitter audience affect the preferential votes that a candidate receives in his constituency.
Measuring the impact of candidates’ tweets on their electoral results