During our summer schools, participants and experts work together through the entire research process in projects on a specific research topic with specific CSS methods.

Apply to participate by February 29, 2020: Call for Participation.

The following projects will be offered:

  1. Using Twitter to measure social cohesion and attitudes towards immigration (Attitudes immigration)
  2. Königsteiner Schlüssel. Topic model analysis (Königsteiner Schlüssel)
  3. HUMAT – a mix of games and ABM (Gaming HUMAT)
  4. Cultural contingencies of social cohesion (Cultural contingencies)
  5. The emergence of formal governance institutions (Governance institutions)
  6. Predicting the Trust Radius with Machine Learning: To What Extent Are Those “Most People” Out-Groups? (Trust radius – Machine Learning)
  7. Structure and Dynamics of Gender Differences in Cooperative Citizenship Behaviors (Gendered Dynamics)
  8. Social cohesion in face-to-face interaction (Face-to-face)


1. Using Twitter to measure social cohesion and attitudes towards immigration

Project leader: Eduardo Graells-Garrido & , Francisco Rowe

Short title: Attitudes immigration

Research question:

  • What shapes public views of immigration?
  • How do they differ across socio-demographic characteristics (gender, age, education) and geographical location?
  • How can we measure attitudes towards immigration and social cohesion using social media data? How do local attitudes towards immigration relate to local levels of social cohesion?

We will leverage Twitter data to capture social cohesion and attitudes towards immigration using machine learning, spatial modeling, text mining, and information visualization techniques.

We will run a workshop on crawling and processing data from Twitter, and we will make available four years of political discussion in a Latin-American country (2015-2019, Chile). Also, we will make available intra-city mobility data for Santiago, the capital of Chile.

Traditional measures of social cohesion (segregation) are `static’. They implicitly assume people do not come in contact with people living in different neighborhoods. Yet people within cities often move for work, shopping and leisure activities, creating opportunities for interaction with individuals from other neighborhoods. We propose two indices to measure social cohesion based on human mobility patterns. We will seek to develop (1) an index -labeled equitable mobility- based on the degree to which the share of visits to other neighborhoods is equal; and, (2) an index, labeled concentrated mobility, based on the extent to which travels are concentrated in a handful of neighborhoods.

Then, we will use Twitter data to derive indices to characterize its users’ attitudes towards immigration. First, we seek to apply topic models, to identify the underlying semantic structure of tweets by quantifying the importance of representative themes. Second, we seek to characterize the overall profile of attitudes towards immigration using two commonly used metrics of sentiment analysis: tendency and polarity. Third, we characterize individual groups of tweets belonging to specific profiles according to their tendency and polarity. Fourth, we seek to understand the diffusion of information of immigration perception through the Twitter network. We will analyze the structure of two networks: (1) the mention network; and, (2) the retweet network estimating the assortativity coefficient of numerical attributes between pairs of linked nodes. Finally, we aim to model the relationship between our measure of social cohesion and immigration sentiment in a Bayesian multi-level modeling framework to determine whether local levels of immigration are positively related to local levels of social cohesion.

2. Königsteiner Schlüssel. Topic model analysis

Project leader: Reinhold Sackmann & Christian Papilloud

Short title: Königsteiner Schlüssel

Research question:

  • How political instances have extended the application of the indicator Königsteiner Schlüssel to migration politics in Germany and in Europe?

Topic Model Analysis

238 German newspaper articles (FAZ, ZEIT), 58 protocols as well as 116 reports of the Deutscher Bundestag going from 1951 until 2018.

Three kind of topic analysis will be applied based on three different algorithms resting on linear algebra (NMF), statistic probabilities (LDA), and neuronal networks, i.e. deep learning (word2vec) performed with the MTA software (, as well as low-level tools for specific analysis of semantic chains in textual data either in a static or in a longitudinal perspective.

3. HUMAT – a mix of games and ABM

Project leader: Patrycja Antosz & Timo Szczepanska

Short title: Gaming HUMAT

Research question:

  • How to use games to extract information about relevant actions and social influence to inform agent-based modeling assumptions?

Multi-method research design that combines the benefits of agent-based modeling (ABM) framework with serious games.

Data will be collected in game sessions during the workshop.

The goal of the project is to teach the conceptual foundations and methodology of agent-based modeling as it applies to understand social cohesion in a chosen area of application. The course leaders suggest three possible areas of application to choose from:
1) voting in a referendum to close the nearby park (Noordenplantsoen park in Groningen) for car traffic (focus on the emergence of social innovations in cities);
2) shifts in dietary preferences – more plant-based communities (focus on understanding the role of meso-level social structures);
3) value chains in fishing trade markets (focus on representing interest groups and their interactions in complex systems).
However, if depending on the interests of students participating in the project, a different case study/research problem related to broadly understood social cohesion is attainable.

Students will work with a multi-method approach that combines the benefits of an agent-based modeling (ABM) framework with serious games. The course organizers provide a toolbox that includes a HUMAT Netlogo architecture and a HUMAT serious game, which will be used to implement a study of a social cohesion phenomenon. HUMAT is an architecture designed in the H2020 project SMARTEES, which has the aim of understanding the diffusion of social innovations in local communities. Students will be presented with a Netlogo implementation of HUMAT architecture, combining cognitive and social aspects of individual decision-making processes. The case of HUMAT presents an approach of evidence-based model design, utilizing theoretical approaches from cognitive psychology and sociology, available secondary data, and primary qualitative and quantitative data to inform model assumptions, parametrization, calibration, and validation. While the HUMAT ABM is used to structure, formalize, and analyze social processes, the serious game provides the flexibility to implement our methodology to different topics of application. Overall, serious games provide interactive, engaging environments to study human interactions. Additionally, they are increasingly used as an innovative method in scientific and civic education, and as a tool fostering stakeholder involvement. Multi-method research projects combining serious games and agent-based modeling allow for identifying action-relevant knowledge and dynamics of social influence. Subsequently, these findings can be transferred to agent-based models linking behaviors to attitudes and social network activities. On top of the ability to comprehend the dependencies between research methods, the project will provide students with hands-on experience in understanding, building and analyzing agent-based models and simulations.

4. Cultural contingencies of social cohesion

Project leader: Gert Jan Hofstede & Geeske Scholz

Short title: Cultural contingencies

Research question:

  • How to represent group belonging and sociality using social identity notions?
  • How are resources, and norm- or value-based mechanisms to divide those resources, integrated with this social identity world?
  • Can we represent generic aspects of social cohesion that always play a role, e.g., how inequality in resource links to othering?

Agent-based models implemented in NetLogo. We are thinking of two ABM versions to develop during the summer school:

    1. Social Identity world The first version would have no content other than group belonging. It requires us to precisely formulate how we conceptualize sociality and belonging. The resulting model can be seen as a refinement of the GRASP world (Gert Jan Hofstede & Liu, 2019) using SIA notions.
    2. Social Resource world The second version would have resources, as well as a norm- or value-based mechanisms to divide those resources. The main question we wish to address is how inequality in resources links to othering. We hypothesize that this relation is mediated by values (as not-necessarily-conscious mental attributes in the minds of individuals) and norms (as rules for distributing resources shared in groups). For instance, unequal resource distribution is seen as intrinsically a proper thing in hierarchical societies, but not in egalitarian ones.

We attempt to use the following data sources:

      • World Values Survey, This is an ongoing research collaboration between countries all around the world, now in its 7th wave of data collection. It can be mined for longitudinal trends or cross-national comparisons.
      • Hofstede 6D model, (> Research and VSM > Dimension Data Matrix). This is the coherent database of Hofstede dimension scores, shown to have continued validity for understanding cross-national differences in all domains of life, from suicide rates, school performance, economic success, to political and governance systems.
      • World Bank open data,, with e.g. GINI coefficients and world development indexes.

Description :

We believe that there are generic aspects about social cohesion that always play a role, albeit in the background of proximate factors. These are for instance visible in the political dynamics of various societies around the world that show historic continuity across generations. Our ideas about generic issues of social cohesion find strong backup in the continuity and explanatory value of culture at the level of society (with nation as a proxy), as apparent from the work of scholars on culture such as Hofstede (G. Hofstede, Hofstede, & Minkov, 2010) and Minkov (Minkov & Hofstede, 2012). Although change is fast and ubiquitous, history repeats itself quite a bit, e.g. in politics, and this is because cultures are persistent as drivers of collective behavior. Culture is relevant for all walks of life, in particular for defining the social structure of a society as more like a market, a family, a machine or a pyramid. This structure in turn is important for the dynamics of social cohesion. How can one integrate or split a market, how a pyramid?
Based on these ‘facts of international history’, we would like to create a foundational model of social cohesion. In everyday life, culture is ‘carried’ by social identity. People are not usually conscious of their culture in the academic sense, but they are acutely aware of their social identities and the behaviors they are supposed to show based on those identities. Identities are usually theorized about by assuming that people have diverse identities. These identities can occur simultaneously: one can belong to several smaller groups as well as one larger society, and each of these identities can be salient depending on context. Social identities can motivate intergroup phenomena as response to status inequality due to social comparison (Brown, 2000). Here, the Social Identity Approach (SIA) will provide us with a conceptual toolkit for refining our model. SIA proposes that social identification, and the perception of people as fellow group members (or outsiders), is a fundamental basis for collective behavior (Tajfel, 1978; Tajfel & Turner, 1979; Tajfel & Turner 1986; Turner, Hogg, Oakes, Reicher & Wetherell, 1987). SIA can help to explain how social context affects a certain identity and how this influences which normative in-group behaviors are salient. As groups have their own social norms and expected behaviors, when a particular social identity is salient the group members are expected to act within those norms. For our purposes, the processes that make people’s attributions of group membership shift are particularly important. This includes both me believing to be part of a group, and others doing so (Dignum, Hofstede, & Prada, 2014). Building upon an ongoing review of the current state of ABMs that implement parts of SIA in model applications, we will discuss and include formalizations of social categorization, saliency, identity, and the accompanying behaviors.

Aditional material:  Cultural contingencies of social cohesion

5. The emergence of formal governance institutions

Project leader: Seth Frey & Michael Mäs

Short title: Governance institutions

Research question:

        • Among amateur-run, small-scale, self-governing online communities, what types of rule systems do they develop?
        • How diverse are these rule systems?
        • How do their rules change over time?
        • How does the structure of their rule systems relate to community outcomes?

Web scraping and web science Data science and statistical learning Quantitative institutional analysis Large-n comparative analysis at the unit of analysis of the social system

The rules from all subreddits on Reddit (~70K, about 10%, have posted rules), or some other system of online communities, coded by rule type. Measures from the same system of rule changes, community effectiveness, and relevant covariates.

Institutional structure and governance style are major drivers of social outcomes. But quantitative frameworks for generally representing governance structure remain a frontier. We will be introduced to the Institutional Analysis and Design framework, and its various taxonomies of rule types in small-scale governance systems.

Simultaneously, we will learn to extract data from a rich source of largely independent small-scale governance systems: the Internet. This will permit us to perform large-n studies not over individuals, but over governance systems, and more rigorously determine the extent of variation in communities’ governance inputs and “societal” outcomes, as well as correlations between these inputs and outputs.

The project leader, Seth Frey, is an interdisciplinary social scientist who uses large datasets and computational methods to isolate the decision processes behind complex, large-scale, and real-world social phenomena. He specializes in studying the self-organization of governance and cultural systems at a whole-system scale by using “designed societies” like sports matches, theme parks, multiplayer video games, and online communities. By working at the intersection of many methods and disciplines, he has introduced approaches that provide new insights into governance processes in small-scale online communities and the science of social system design generally.

6. Predicting the trust radius with machine learning: To what extent are those “Most People” out-groups?

Project leader: Wahideh Achbari

Short title: Trust radius

Research question:

        • To what extent can Generalized Trust be predicted by prejudice?
        • Can we cross-validate the findings with data that have less information on implicit prejudice and intergroup attitudes?

Supervised Machine Learning Ensembles

Project Implicit | World Values Survey | European Social Survey | LISS Panel | …

Generalized (social) trust (hereafter GT) or trust in “most people” is a conspicuous indicator in social survey research on intergroup relations and social cohesion. The survey question is as follows: “would you say that most people can be trusted, or would you be careful in dealing with them?” Recent meta-analyses show that researchers often assume GT taps into an evaluation of the trustworthiness of unknown people or even ethnic out-groups. While this prior research has debated the negative link between ethnic diversity and GT as a proxy for social cohesion, with some notable exceptions, relatively few have so far focused on systematic response bias in answers to GT. These mixed findings may well be due to the operationalization of social cohesion by GT. For example, a study using think-aloud protocols have demonstrated that the majority of respondents high in GT think “most people” refers to people they know, whereas a high proportion of those who are low in GT think about strangers. Following the homophily principle – the tendency of individuals to associate and bond with similar others – we can expect that people known to the respondents are ethnic in-groups, and strangers are more likely to be ethnic out-groups, although formally we do not know this. A faceless stranger one hypothetically meets for the first time could well be an ethnic in-group in homogenous settings as much as a person known to the respondent (friend, neighbor, family member) can be an ethnic out-group in diverse settings. In this project, we, therefore, return to some of the overlooked basics in this literature: the measurement and conceptualization of GT, which is allegedly in decline by ethnic diversity. The proposed project aims a) to examine the validity of the Generalized Trust (GT) question as a measure of out-group attitudes and implicit race bias in an unprecedentedly broad manner using Supervised Machine Learning Ensembles, and b) to cross-validate the results with 3 existing large-scale social surveys (WVS, ESS, LISS).

By employing Machine Learning, we propose to quantify complexity instead of relying on ex-ante model sparsity and favoring a set of variables of interests over others. Other advantages of ML over conventional statistical analyses are its flexibility to select variables when many potential predictors are available; its ability to model nonlinear relationships; and that there are fewer limits to the number of datasets, observed cases, interactions between variables, and hence modeling strategies. Finally, the results are less tainted by researcher degrees of freedom in preferring a scale or measure over another. Our goal with prediction, however, remains in line with what social science generally attempts to do: to get good out-of-sample predictions and to avoid overfitting. Therefore, we propose to train, validate, and test the model again (holdout) with differently sized random subsamples of the data. In addition, we then cross-validate the results using social surveys that contain a limited set of measures of intergroup attitudes.

Additional material:  Trust Radius – Machine Learning

7. Structure and dynamics of gender differences in cooperative citizenship behaviors

Project leader: Diane Bergeron & Corinne Coen

Short title: Gendered Dynamics

Research question:

        • How do differences in men and women’s helping behavior generate distinct patterns of cooperation?
        • How do distinct patterns of reciprocation to men and women’s helping behavior alter the benefits of cooperating?

The work will take place in two parts. First, data from past empirical studies on differences in men and women’s cooperative behavior and reciprocation behavior will be compiled and analyzed. The remaining time will be spent in hands-on workshops designing and building agent-based models of gendered agents enacting citizenship behaviors. Identified dynamics may focus on those resulting from structural features of the context, such as fixed relationships (e.g., networks or neighborhoods) or reward structures [e.g., pay or payoff frameworks (as in social dilemmas)], distinct mixing patterns based on locality or status, or features of the agents (e.g., prosocial orientation). These dynamics may drive opportunities for rewarding mutual cooperation rather than unrewarded individual cooperation. Data analysis will be performed as time permits.

The primary data for this study are based on empirical results in key published studies on gendered dynamics and sex differences in cooperative helping behaviors (i.e., organizational citizenship behavior).

Preliminary version, will be finished by 2020-01-10

8. Social cohesion in face-to-face interaction

Project leader: Michele Starnini & Jan Lorenz

Short title: Face-to-face

Research question:

  • How can we measure the social cohesion of crowds larger than small groups, based on their face-to-face interactions?
  • Do crowds on scientific conferences become more socially cohesive over the days of the conference?
  • Does the social and scientific attractiveness of participants play a role in enhancing the social cohesion of crowds?

Identification of key concepts to characterize social cohesion in interacting crowds. Definition of a measure of social cohesion to apply on face-to-face interaction data. Social network analysis, including the temporal dimension (temporal networks) and agent-based modeling.

Data of face-to-face interactions among the participants of a three-day scientific conference, collected by the Sociopatterns collaboration. Data measured the proximity patterns of individuals with a space-time resolution of 1 meter and 20 seconds by using wearable active radio-frequency identification devices. Data includes personality traits and self-perceived social and scientific attractiveness of participants, collected by surveys.

Uncovering the dynamics of face-to-face interactions is pivotal to understand the mechanisms that shape the formation of social ties, as well as the formation, segregation, and social cohesion of groups of individuals. The goal of this project is to characterize the social cohesion of interacting crowds by the definition of quantities that can be measured in empirical data. To this aim, we will use real face-to-face interactions recorded at a three days scientific conference, including personality traits as well as self-perceived social and scientific attractiveness of participants, collected by surveys. In order to characterize the social cohesion of crowds over time, we will model data by employing methods from network science and agent-based modeling.