Raphael Heiko Heiberger

Abstract

The rise of populism concerns many political scientists and practitioners, yet the detection of its underlying language remains fragmentary. This paper aims to provide a reliable, valid, and scalable approach to measure populist rhetoric. For that purpose, we created an annotated dataset based on parliamentary speeches of the German Bundestag (2013–2021). Following the ideational definition of populism, we label moralizing references to “the virtuous people” or “the corrupt elite” as core dimensions of populist language. To identify, in addition, how the thin ideology of populism is “thickened,” we annotate how populist statements are attached to left-wing or right-wing host ideologies. We then train a transformer-based model (PopBERT) as a multilabel classifier to detect and quantify each dimension. A battery of validation checks reveals that the model has a strong predictive accuracy, provides high qualitative face validity, matches party rankings of expert surveys, and detects out-of-sample text snippets correctly. PopBERT enables dynamic analyses of how German-speaking politicians and parties use populist language as a strategic device. Furthermore, the annotator-level data may also be applied in cross-domain applications or to develop related classifiers.

Abstract

In this paper, we investigate how protective effects of intergenerational closure correspond with conflict networks in school classes. Taking a multilevel ecological perspective, we also consider networks' socio-spatial conditions. In a first step, we use ERGMs to analyze the association between parental contact and students' friendship ties, i.e., intergenerational closure (IC). Then, we utilize spatial regressions to analyze direct and moderating effects of a school's neighborhood on conflicts in the 135 German class networks (N~=~3143 student measurements). In accordance with Coleman's theoretical notions, we find consistent negative effects of IC on the (standardized) density of the conflict networks. Moreover, we show that IC's impact is particularly strong in neighborhoods with a relatively high concentration of minorities. The results are in line with our theoretical considerations on multilevel network ecologies and the selective pressure of IC against conflict ties. Practically, our results provide evidence that fostering connections among parents (e.g., by implementing opportunity structures for parents to meet) might help to prevent deviating behavior in schools, especially in neighborhoods with relatively large proportions of ethnic minorities.

Abstract

Policy actors (PAs) like nongovernmental organizations, political parties or governmental institutions strategically communicate on social media to gain attention and thus influence the public agenda. We argue that networks of PAs engaged in the same issues (i.e., a PA’s peer network) are crucial to attracting the interest of a broad audience. Drawing on network theory, we posit that (i) ideological homophily, and (ii) the centrality and embeddedness in a PA’s peer network increase the attention received from all Twitter (now X) users. We investigate these premises by analyzing the European migration discourse on Twitter (2014–2020). The results of our study reveal that the centrality of PAs in their peer networks and ideologically similar relations considerably increase attention from the whole Twittersphere. These findings provide strong evidence that a PA’s role in its organizational peer network on social media governs the attention received in the overall discourse.

Abstract

In this chapter, Raphael Heiberger and Michael Windzio examine which topics are important for major education international organization (IOs). IOs in the field of education follow different ideological paradigms in the global education discourse. Yet, it is an open question as to whether different types of IOs focus on different topics and thereby support different paradigms of education. Based on more than 1000 documents with over 40 million words published by the World Bank, UNESCO, the ILO, the OECD, ISESCO, and SEAMEO, they explore education issues addressed in this sample. Using standardized methods of quantitative text analysis and topic modeling, Heiberger and Windzio reveal that major topics found in these documents do indeed differ between the different types of organizations.

Abstract

The European Union’s common public sphere project dates back to the 1960s and relies on Europeanisation through the gradual eradication of communication boundaries between its member countries. However, it is evident by now that Europeanisation of national public spheres is hard to achieve by increasing overlaps between national public spheres, synchronisation of news reporting across national boundaries, or diffusion of Europeanist norms into national politics. The European Union’s common public sphere project may hence be in danger. This calls for explorations of other imaginable models of the public sphere for Europe. Are there traces of other modes of transnational public sphere emerging in Europe? In this article, we explore a models of the transnational public sphere which is based on an alternative concept of Europeanisation derived from the cleavage theory. By drawing on social media data and employing tools of social network analysis, we demonstrate the empirical possibility of a cleavage model of the European public sphere.

Abstract

Applications of machine learning (ML) in industry and natural sciences yielded some of the most impactful innovations of the last decade (for instance, artificial intelligence, gene prediction or search engines) and changed the everyday-life of many people. From a methodological perspective, we can differentiate between unsupervised machine learning (UML) and supervised machine learning (SML). While SML uses labeled data as input to train algorithms in order to predict outcomes of unlabeled data, UML detects underlying patterns in unlabeled observations by exploiting the statistical properties of the data. The possibilities of ML for analyzing large datasets are slowly finding their way into the social sciences; yet, it lacks systematic introductions into the epistemologically alien subject. I present applications of some of the most common methods for SML (i.e., logistic regression) and UML (i.e., topic models). A practical example offers social scientists a ``how-to'' description for utilizing both. With regard to SML, the case is made by predicting gender of a large dataset of sociologists. The proposed approach is based on open-source data and outperforms a popular commercial application (genderize.io). Utilizing the predicted gender in topic models reveals the stark thematic differences between male and female scholars that have been widely overlooked in the literature. By applying ML, hence, the empirical results shed new light on the longstanding question of gender-specific biases in academia.

Abstract

We map the topic structure of psychology utilizing a sample of over 500,000 abstracts of research articles and conference proceedings spanning two decades (1995–2015). To do so, we apply structural topic models to examine three research questions: (i) What are the discipline’s most prevalent research topics? (ii) How did the scientific discourse in psychology change over the last decades, especially since the advent of neurosciences? (iii) And was this change carried by high impact (HI) or less prestigious journals? Our results reveal that topics related to natural sciences are trending, while their ’counterparts’ leaning to humanities are declining in popularity. Those trends are even more pronounced in the leading outlets of the field. Furthermore, our findings indicate a continued interest in methodological topics accompanied by the ascent of neurosciences and related methods and technologies (e.g. fMRI’s). At the same time, other established approaches (e.g. psychoanalysis) become less popular and indicate a relative decline of topics related to the social sciences and the humanities.

Abstract

Governments around the globe use non-pharmaceutical interventions (NPIs) to curb the spread of coronavirus disease 2019 (COVID-19) cases. Making decisions under uncertainty, they all face the same temporal paradox: estimating the impact of NPIs before they have been implemented. Due to the limited variance of empirical cases, researchers could so far not disentangle effects of individual NPIs or their impact on different demographic groups. In this paper, we utilize large-scale agent-based simulations in combination with Susceptible-Exposed-Infectious-Recovered (SEIR) models to investigate the spread of COVID-19 for some of the most affected federal states in Germany. In contrast to other studies, we sample agents from a representative survey. Including more realistic demographic attributes that influence agents' behavior yields accurate predictions of COVID-19 transmissions and allows us to investigate counterfactual what-if scenarios. Results show that quarantining infected people and exploiting industry-specific home office capacities are the most effective NPIs. Disentangling education-related NPIs reveals that each considered institution (kindergarten, school, university) has rather small effects on its own, yet, that combined openings would result in large increases in COVID-19 cases. Representative survey-characteristics of agents also allow us to estimate NPIs' effects on different age groups. For instance, re-opening schools would cause comparatively few infections among the risk-group of people older than 60 years.

Abstract

We investigate how sociology students garner recognition from niche field audiences through specialization. Our dataset comprises over 80,000 sociology-related dissertations completed at U.S. universities, as well as data on graduates’ pursuant publications. We analyze different facets of how students specialize—topic choice, focus, novelty, and consistency. To measure specialization types within a consistent methodological frame, we utilize structural topic modeling. These measures capture specialization strategies used at an early career stage. We connect them to a crucial long-term outcome in academia: becoming an advisor. Event-history models reveal that specific topic choices and novel combinations exhibit a positive influence, whereas focused theses make no substantial difference. In particular, theses related to the cultural turn, methods, or race are tied to academic careers that lead to mentorship. Thematic consistency of students’ publication track also has a strong positive effect on the chances of becoming an advisor. Yet, there are diminishing returns to consistency for highly productive scholars, adding important nuance to the well-known imperative of publish or perish in academic careers.

Abstract

Working with text poses important conceptual and methodological challenges. Topic models are a popular tool to reduce texts’ complexity and find meaningful themes in large corpora. After an overview of existing work, we explain how to employ structural topic models, one of the variations of topic modeling of most relevance to social researchers. In particular, however, this chapter emphasizes the selection of an appropriate number of topics K and its relation to preprocessing. We investigate the influence of preprocessing decisions on (i) the choice of K and (ii) the quality of a topic model (i.e., its predictive power and consistency). For that purpose, we examine a multitude of model setups by employing both established metrics and innovative measures. From our empirical results, we derive several practical recommendations for researchers and provide easy-to-use code to approximate an appropriate number of topics and test the robustness of one’s choice. We develop these arguments with comprehensive data on over 137,000 education-related dissertations completed at U.S. universities.

Abstract

Democratic politics builds on both clear differences and shared common ground. While the rise of digital media may have enabled more differences to be articulated, common ground is often seen as threatened by fragmentation of political debate, which some see as driven by news media. The relative importance of political actors (parties and politicians) in driving fragmentation has received less attention. In this paper, we compare how news media and political actors contribute to the fragmentation of online political debate on the basis of analysis of almost half a million election-related tweets collected during the 2017 French, German, and U.K. national elections. We employ a structural topic model to reduce online political debate to networks of topic overlap. Across the three countries with different political and media systems, we find news media are by far the most important actors in terms of creating and maintaining a common space of online political debate on Twitter. Our results also show that political actors, with some variation from country to country, contribute more to fragmentation as they focus on different topics while articulating clear differences. These findings underline the importance of complementing structural analysis of the rise of digital and social media with analysis of how important elite actors like news media and political parties/candidates use these media in different ways. Overall, we show how at least on Twitter, across three different countries with different media systems and political systems, news media create connection that contributes to commonality while political actors lay out clear differences that drive fragmentation.

Abstract

Media discourse is often seen as an important condition of people's attitudes and perceptions. Despite a rich literature, however, it is not well understood how media exposure influences attitudes towards immigrants. In contrast to previous studies, we argue that people rely on `availability heuristics' shaped by mass media. From that point of view, it is the specific content of media discourse on immigration that affects people's concerns. We use structural topic models to classify media content of more than 24,000 articles of leading German newspapers from 2001 to 2016. Utilizing linear fixed-effects models allows us to relate a person's concern towards immigration, as reported in the German Socioeconomic Panel, to prevalent topics discussed in print media while controlling for several confounding factors (e.g., party preferences, interest in politics, etc.). We find a robust relationship between topic salience and attitudes towards integration. Our results reveal that specific topics with negative contents (e.g., domestic violence) increase concerns, while others (e.g., scientific studies, soccer) decrease concerns substantially, underlining the importance of available information provided by media. In addition, people with higher education are generally less affected by media salience of topics.

Abstract

With regard to residential energy use in the European Union (EU), most studies consider potential adopters of the technology (e.g., private owners) as being the sole decision-makers in the technology selection. However, during an integrated decision-making process (e.g., a construction project) multiple stakeholders will interact, influencing each other’s judgement, thereby making it difficult to discern who is affecting the final decision, and to what extent. The goal of this study is to outline the full network of stakeholders involved in the decision-making process, along with their degree of power and interaction in the technology choice. For this purpose, empirical evidence from a multi-country survey is examined using social network analysis (SNA). The information is compared across building typologies, project types and countries (i.e., Italy, Spain, Germany, Poland, the United Kingdom, France, Belgium and the Netherlands). The results demonstrate that, in EU residential buildings, potential adopters of the technology are not the only stakeholders involved in the technology selection. They are in all instances in communication with multiple stakeholders, some of whom also hold a high level of power in the decision (i.e., key persuaders). Furthermore, their level of power and communication varies substantially across building typologies, project types and countries.

Abstract

Education entails conflicting perspectives about its subject matter. In the late 1980s, the conflict developed into a war between interpretive and causal paradigms. Did the confrontation result in a balance between these warring sides? We use text analysis to identify research trends in 137,024 dissertation abstracts from 1980 to 2010 and relate these to students’ academic employment outcomes. Topics associated with the interpretive approach rose in popularity, while the outcomes-oriented paradigm declined. Academic employment remained stably associated with topics in the interpretive approach, but their effect is moderated by the prestige of the students’ institutions. The relation between topic popularity and employability provides insight into field change and how the benefits of cultural shifts fall along the lines of institutional power.

Abstract

Social network analysis is playing an increasingly important role in sociological studies. At the same time, new technologies such as wearable sensors make it possible to collect new types of social network data. We employed RFID tags to capture face-to-face interactions of participants of two consecutive Ph.D. retreats of a graduate school on climate research. We use this data in order to explore how it may support ethnographic observations and to gain further insights on scholarly interactions. The unique feature of the data is the opportunity to distinguish short and long conversations, which often have a different nature from a sociological point of view. Furthermore, an advantage of this data is the availability of socio-demographic, research-related, and situational attributes of participants. We show that, even though an interaction partner is often found rather randomly during coffee breaks of retreats, a strong homophily between participants from the same institutions or research areas exists. We identify cores of the networks and participants who play ambassador roles between communities, e.g., persons who visit the retreat for the second time are more likely to be ambassadors. Overall, we show the usefulness and potential of RFID tags for scientometric studies.

Abstract

In diesem Beitrag betrachten wir die Einbettung des internationalen „Debt Security“-Marktes (DSM) in gesellschaftliche Teilbereiche aus einer feldtheoretischen Perspektive. Der DSM ist einer der größten Finanzmärkte und versorgt Staaten, Unternehmen und Banken mit Liquidität. Spätestens seit der letzten Finanzkrise ist klar geworden, dass eine rein wirtschaftswissenschaftliche Analyse von Finanzmärkten zu kurz greift und wichtige soziale Korrelate unberücksichtigt lässt. Aus diesem Grund untersuchen wir die Schuldverhältnisse aus einer netzwerkanalytischen Perspektive, berücksichtigen zugleich jedoch auch deren strukturelle Bedingungen und Voraussetzungen in Form von relationalen Einbettungsverhältnissen in einen globalen Sozialraum. Hierzu werden die aus den Schuldbeziehungen des DSM resultierenden Netzwerkpositionen mittels geometrischer Datenanalysen in ihrem Korrespondenzverhältnis zu sozialen Feldern und zu einem globalen Feld der Macht lokalisiert. Es zeigt sich, dass der DSM eng mit verschiedenen extraökonomischen Einflussbedingungen verwoben ist (etwa Militär, Rechtssicherheit, Wissenschaft, etc.) und sich dabei – wie der gesamte globale Sozialraum – durch feldübergreifend akkumulierende Vorteilsmechanismen auszeichnet.

Abstract

Reconstructing scientific networks from the past can be a difficult process. In this paper, we argue that eponyms are a promising way to explore historic relationships between natural scientists using taxonomy. Our empirical case is the emerging community of malacologists in the 19th century. Along the lines of pivotal concepts of social network analysis we interpret eponyms as immaterial goods that resemble the proporties of regular social contacts. Utilising Exponential Random Graph Models reveals that the social exchange underlying eponyms follows similar rules as other social relationships such as friendships or collaborations. It is generally characterized by network endogenous structures and homophily. Interestingly, the productivity of authors seems to be well recognised among contemporary researchers and increases the probability of a tie within the network significantly. In addition, we observe an epistemological divide in the malacological research community. Thus even in the 19th century, at a time when science was just emerging as a differentiated social system, epistemological distinctions have been a defining concept for scientific contacts.

Abstract

Das Ziel dieses Beitrags ist es eine computergestützte Methode zur Identifizierung von parteispezifischen politischen Themen und Trends im Zeitverlauf vorzustellen und deren Nützlichkeit zu demonstrieren. Als Datengrundlage dienen die Plenarsitzungen und die darin enthaltenen Debatten der Mitglieder des Bundestages von 1990 bis 2013. Zur Aggregation der Daten wurde die Latent Dirichlet Allocation, eine gängige Methode für Topic Modeling in der quantitativen Textanalyse, verwendet. Eine Modifikation der Methode erlaubt das Finden und einen Vergleich von Themen und Trends über die jeweiligen Legislaturperioden hinweg, d. h. wie populär bestimmte Themen im Bundestag besetzt werden. Mit der Auswertung der Ergebnisse wird gezeigt, dass diese Vorgehensweise relevante Rückschlüsse über lange Zeiträume Bundespolitik zulässt. Die Reduktion von Texten auf Themen und Trends bietet großes Potential für weitere Anwendungsgebiete mit großen Textmengen, es wird allerdings auch kritisch auf die Möglichkeiten und Grenzen der Methode eingegangen.

Abstract

Networks derived from stock prices are often used to model developments on financial markets and are tightly intertwined with crises. Yet, the influence of changing market topologies on the broader economy (i.e. GDP) is unclear. In this paper, we propose a Bayesian approach that utilizes individual-level network measures of companies as lagged probabilistic features to predict national economic growth. We use a comprehensive data set consisting of Standard and Poor’s 500 corporations from January 1988 until October 2016. The final model forecasts correctly all major recession and prosperity phases of the U.S. economy up to one year ahead. By employing different network measures on the level of corporations, we can also identify which companies’ stocks possess a key role in a changing economic environment and may be used as indication of critical (and prosperous) developments. More generally, the proposed approach allows to predict probabilities for different overall states of social entities by using local network positions and could be applied on various phenomena.

Abstract

Heimentgelte stationärer Pflegeeinrichtungen sind in Deutschland von einer hohen regionalen Varianz gekennzeichnet. Vor dem Hintergrund einer marktorientierten Ausgestaltung des Pflegesektors durch die Gesetzliche Pflegeversicherung und der damit einhergehenden Reproduktion sozialer Ungleichheiten untersucht die vorliegende Studie, welche Erklärungsfaktoren diese regionalen Preisunterschiede bedingen. Dazu wird der Einfluss von vier Variablengruppen – die sozioökonomischen Rahmenbedingungen einer Region, ihre Pflegeinfrastruktur, die Pflegenachfrage vor Ort sowie die Pflegeevaluationen – auf die Kosten stationärer Pflege in multiplen Regressionsanalysen getestet. Als Kostendaten liegen rund 12.000 Einzelpreise von Pflegeheimen vor. Die Analyse zeigt, dass regionale Unterschiede von Heimentgelten in hohem Maße durch marktförmige Wettbewerbsstrukturen zu erklären sind. Aufgrund der bestehenden rechtlichen Regulierung findet jedoch keine klassische Preisbildung auf freien Märkten statt. Vielmehr muss von einer zum Teil gesteuerten Bildung der Heimentgelte auf einem Quasi-Markt für Pflege ausgegangen werden, die die bestehenden sozialen Ungleichheiten reproduziert und vertieft.

Abstract

Physics is one of the most successful endeavors in science. Being a prototypic big science it also reflects the growing tendency for scientific collaborations. Utilizing 250,000 papers from ArXiv.org a prepublishing platform prevalent in Physics we construct large coauthorship networks to investigate how individual network positions influence scientific success. In this context, success is seen as getting a paper published in high impact journals of physical subdisciplines as compared to not getting it published at all or in rather peripheral journals only. To control the nested levels of authors and papers, and to consider the time elapsing between working paper and prominent journal publication we employ multilevel eventhistory models with various network measures as covariates. Our results show that the maintenance of even a moderate number of persistent ties is crucial for scientific success. Also, even with low volumes of social capital Physicists who occupy brokerage positions enhance their chances of articles in high impact journals significantly. Surprisingly, inter(sub)disciplinary collaborations decrease the probability of getting a paper published in specialized journals for almost all positions.

Abstract

Today’s world allows people to connect over larger distances and in shorter intervals than ever before, widely monitored by massive online data sources. Ongoing worldwide computerization has led to completely new opportunities for social scientists to conceive human interactions and relations in unknown precision and quantities. However, the large data sets require techniques that are more likely to be found in computer and natural sciences than in the established fields of social relations. In order to facilitate the participation of social scientists in an emerging interdisciplinary research branch of “computational social science,” we propose in this article the usage of the Python programming language. First, we carve out its capacity to handle “Big Data” in suitable formats. Second, we introduce programming libraries to analyze large networks and big text corpora, conduct simulations, and compare their performance to their counterparts in the R environment. Furthermore, we highlight practical tools implemented in Python for operational tasks like preparing presentations. Finally, we discuss how the process of writing code may help to exemplify theoretical concepts and could lead to empirical applications that gain a better understanding of the social processes initiated by the truly global connections of the Internet era.

Abstract

We propose a field-theoretical elaboration of the habitus-field-theory towards the level of the ‘world-society’. For this purpose we analyze an integrated data set from different social fields on country level (n = 181 countries) applying multiple factor analysis. We demonstrate that the global field of power can be described by two dimensions: meta-capital and internal functionality. Especially internal functionality can provide a valuable analytical frame of reference for comparative (survey) research.

Abstract

In this paper, we combine network analytical methods to understand the structure of financial markets with recent research about collective attention shifts by utilizing massive social media data. Our main goal, hence, is to investigate whether changes in stock networks are connected with collective attention shifts. To examine the relationship between structural market properties and mass online behavior empirically, we merge company-level Google Trends data with stock network dynamics for all S&P 100 corporations between 2004 and 2014. The interplay of massive online behavior and market activities reveals that collective attention shifts precede structural changes in stock market networks and that this connection is mostly carried by companies that already dominate the development of the S&P 100.

Abstract

The amendment of the German Stock Companies Act in 1998 regarding the repurchase of shares makes it possible to study the acceptance of a capital market strategy at managerial level as well as its effects on investors. Both are sociologically intertwined with the belief of the agents in the efficiency of share repurchases. In the U.S., this pattern has been observed ever since the prevailing of the “Agency Logic”. Based on the largest data set so far, the paper shows that repurchase programs have also been well established in the German business world in the meantime. Short-term market reactions are significantly positive and imply an informal, yet institutionalized rule of action for investors. However, both trends apply especially in times of negative markets and, generally, not in the long-run—which makes them a “bad signal” for private investors. The results present robust evidence in central aspects against the dominant economic explanation (the “Signalling Theory”) and suggest practical actions for politicians, entrepreneurs and investors.

Abstract

Today´s connected world allows people to gather information in shorter intervals than ever before, widely monitored by massive online data sources. As a dramatic economic event, recent financial crisis increased public interest for large companies considerably. In this paper, we exploit this change in information gathering behavior by utilizing Google query volumes as a "bad news" indicator for each corporation listed in the Standard and Poor´s 100 index. Our results provide not only an investment strategy that gains particularly in times of financial turmoil and extensive losses by other market participants, but reveal new sectoral patterns between mass online behavior and (bearish) stock market movements. Based on collective attention shifts in search queries for individual companies, hence, these findings can help to identify early warning signs of financial systemic risk. However, our disaggregated data also illustrate the need for further efforts to understand the influence of collective attention shifts on financial behavior in times of regular market activities with less tremendous changes in search volumes.

Abstract

Abstract Despite many efforts crises on financial markets are in large part still scientific black-boxes. In this paper, we use a winner-take-all approach to construct a longitudinal network of S&P 500 companies and their correlations between 2000 and 2012. A comparison to complex ecosystems is drawn, especially whether the May–Wigner theorem can describe real-world economic phenomena. The results confirm the utility of the May–Wigner theorem as a stability indicator for the US stock market, since its development matches with the two major crises of this period, the dot-com bubble and, particularly, the financial crisis. In those times of financial turmoil, the stock network changes its composition, but unlike ecological systems it tightens and the disassortative structure of prosperous markets transforms into a more centralized topology.

Abstract

This article describes patterns of scientific growth that emerge in response to major research accomplishments in instrumentation and the discovery of new matter. Using two Nobel Prize-winning contributions, the scanning tunneling microscope (STM) and the discovery of Buckminsterfullerenes (BUF), we examine the growth of follow-up research via citation networks at the author and subdiscipline level. A longitudinal network analysis suggests that structure, cohesiveness, and interdisciplinarity vary considerably with the type of breakthrough and over time. Scientific progress appears to be multifaceted, including not only theoretical advances but also the discovery of new instrumentation and new matter. In addition, we argue that scientific growth does not necessarily lead to the formation of new specialties or new subdisciplines. Rather, we observe the emergence of a research community formed at the intersection of subdisciplinary boundaries.