Search

This paper discusses the crucial but sometimes neglected differences between unconditional quantile regression (UQR) models and quantile treatment effects (QTE) models. We argue that there is a frequent mismatch between the aim of the quantile regression analysis and the quantitative toolkit used in much of the applied literature, including the motherhood wage penalty literature. This mismatch may result in wrong conclusions being drawn from the data, and in the end, misguided theories. In this paper, we clarify the crucial conceptual distinction between influences on quantiles of the overall distributions, which we term population-level influences, and individual-level QTEs. Further, we use data simulations to illustrate that various classes of quantile regression models may, in some instances, give entirely different conclusions (to different questions). Finally, we compare quantile regression estimates using real data examples, showing that UQR and QTE models differ sometimes but not always. Still, the conceptual and empirical distinctions between quantile regression models underline the need to match the correct model to the specific research questions. We conclude the paper with a few practical guidelines for researchers.

Exponential random graph models (ERGM) have been widely applied in the social sciences in the past ten years. However, diagnostics for ERGM have lagged behind their use. Collinearity-type problems can emerge without detection when fitting ERGM, skewing coefficients, biasing standard errors, and yielding inconsistent model estimates. This article provides a method to detect multicollinearity in ERGM. It outlines the problem and provides a method to calculate the variance inflation factor from ERGM parameters. It then evaluates the method with a Monte Carlo simulation, fitting 216,000 ERGMs and calculating the variance inflation factors for each model. The distribution of variance inflation factors is analyzed using multilevel regression to determine what network characteristics lend themselves to collinearity-type problems. The relationship between variance inflation factors and unstable standard errors (a standard sign of collinearity) is also examined. The method is shown to effectively detect multicollinearity and guidelines for interpretation are discussed.

Unpaid or ‘free’ labour is an important element of how precarity has been theorized. It is also an issue that is often seen as endemic to cultural and creative work, rightly attracting a range of criticism. Questions as to the role of unpaid work, for example internships, have become central to understanding the social exclusiveness of many cultural and creative jobs. This paper develops this existing analysis by comparing and contrasting the meaning of 'free' work over the life course of a range of creative occupations, historicising the impact of unpaid labour on the creative sector and showing how it has been stratified by social class, age and career stage. The paper uses two datasets drawn from the Panic! What happened to social mobility in the arts? project, to outline the differing experiences of unpaid labour in cultural and creative occupations. By demonstrating the stratification of unpaid work as a form of precariousness in cultural jobs, along with the social distribution of benign narratives of unpaid work, the paper aims to offer new empirical evidence for those seeking to resist precarious forms of labour.

Traveling to different destinations is a big part of our lives. We visit a variety of locations both during our daily lives and when we're on vacation. How can we find the best way to navigate from one place to another? Perhaps we can test all of the different ways of traveling between two places, but another method is to use mathematics and computation to find a shortest path. We discuss how to construct a shortest path and introduce Dijkstra's algorithm to minimize the total cost of a path, where the cost may be the travel distance, travel time, or some other measurement. We also discuss how to use shortest paths in the real world to save time and increase traveling efficiency., Comment: This is an article for teens and pre-teens. Let us know if you have any comments

We re-consider Schelling’s (1971) bounded neighbourhood model as put into the form of a dynamical system by Haw and Hogan (2018). In the case of a single neighbourhood we explain the occurring bifurcation set, thereby correcting a minor scaling error. In the case of two neighbourhoods we correct a major error and derive a dynamical system that does satisfy the modeling assumptions made by Haw and Hogan (2020), staying as close as possible to their construction. We find that stable integration then is only possible if the populations in the two neighbourhoods have the option to be in neither neighbourhood. In the absence of direct movement between the neighbourhoods the problem is furthermore equivalent to independent single neighbourhood problems.

Desarrollamos un modelo de adopción de lenguaje basado en un modelo evolutivo genera-do por un juego con dos poblaciones y tres estrategias y encontramos dos equilibrios estables, a saber: bilingüismo en ambas poblaciones y monolingüismo basado en la lengua vernácula de cada población.

The spread and transmission dynamics of directly transmitted airborne pathogens, such as SARS-CoV-2, are fundamentally determined by in-person contact patterns. Reliable quantitative estimates of contact patterns are critical to modeling and reducing the spread of directly transmitted infectious diseases. While national-level contact data are available in many countries, including the United States, local-level estimates of age-specific contact patterns are key since disease dynamics and public health policy vary by geography. However, collecting contact data for each state would require a very large sample and be prohibitively expensive. To overcome this challenge, we develop a flexible model to estimate age-specific contact patterns at the subnational level using national-level interpersonal contact data. Our model is based on dynamic multilevel regression with poststratification. We apply this approach to a national sample of interpersonal contact data collected by the Berkeley Interpersonal Contact Study (BICS). Results illustrate important state-level variation in levels and trends of contacts across the US.

Theoretical and Empirical Modeling of Identity and Sentiments in Collaborative Groups (THEMIS.COG) was an interdisciplinary research collaboration of computer scientists and social scientists from the University of Waterloo (Canada), Potsdam University of Applied Sciences (Germany), and Dartmouth College (USA). This white paper summarizes the results of our research at the end of the grant term. Funded by the Trans-Atlantic Platform’s Digging Into Data initiative, the project aimed at theoretical and empirical modeling of identity and sentiments in collaborative groups. Understanding the social forces behind self-organized collaboration is important because technological and social innovations are increasingly generated through informal, distributed processes of collaboration, rather than in formal organizational hierarchies or through market forces. Our work used a data-driven approach to explore the social psychological mechanisms that motivate such collaborations and determine their success or failure. We focused on the example of GitHub, the world’s current largest digital platform for open, collaborative software development. In contrast to most, purely inductive contemporary approaches leveraging computational techniques for social science, THEMIS.COG followed a deductive, theory-driven approach. We capitalized on affect control theory, a mathematically formalized theory of symbolic interaction originated by sociologist David R. Heise and further advanced in previous work by some of the THEMIS.COG collaborators, among others. Affect control theory states that people control their social behaviours by intuitively attempting to verify culturally shared feelings about identities, social roles, and behaviour settings. From this principle, implemented in computational simulation models, precise predictions about group dynamics can be derived. It was the goal of THEMIS.COG to adapt and apply this approach to study the GitHub collaboration ecosystem through a symbolic interactionist lens. The project contributed substantially to the novel endeavor of theory development in social science based on large amounts of naturally occurring digital data.

The social sciences face growing demand for reproducible tools for processing massive troves of often-complex text data (political speeches, medical notes, etc.). In response, we aim toward computational literature review by developing an inductive method of applying expert-built dictionaries for automated analysis of complex texts. Our workflow begins with developing dictionaries from foundational texts and domain expertise. Next, we apply text-analytic methods of differential domain-specificity and complexity to create vector-space representations of texts. Finally, we compare the validity of these methods by using regression models to evaluate relationships between their representations and ground truth. Taking as our use case a large corpus of academic articles in organizational science, we find that domain-specific, relatively simple embeddings were most valid--while the more sophisticated models were very weak. Thus, we suggest that social science workflows for learning from complex texts incorporate embeddings that are domain-specific and straightforward--rather thanconvoluted.

Homophily is one of the robust findings in social network studies. It persists even in a diverse population where the opportunity to develop homogeneous friendship is not greater than a mere chance and the process of developing heterogeneous friendship is facilitated. In this study, we introduce the Framework for Intergroup Relations and Multiple Affiliations Networks (FIRMAN), derived from social identity theory and social network framework, that can explain why that is the case. We begin by explaining its components: social identity space, social identity distance, length of ties, tie outreachability, and tie capacity. Then, through 7000 agent-based simulations, we demonstrate how the interaction of tie outreachability and tie capacity constraints heterogeneous friendship formations, which in turn make homophily inevitable even in a very diverse population. Surprisingly, the presence of even a small percentage (< 15%) of agents who can only develop homogeneous friendships can affect the whole population, preventing other agents from developing heterogenous friendships. We conclude by providing some directions for future research.

This preprint is an earlier version of the forthcoming article: Arseniev-Koehler, Alina. Theoretical foundations and limits of word embeddings: What types of meaning can they capture? Sociological Methods & Research. Measuring meaning is a central problem in cultural sociology and word embeddings may offer powerful new tools to do so. But like any tool, they build on and exert theoretical assumptions. In this paper I theorize the ways in which word embeddings model three core premises of a structural linguistic theory of meaning: that meaning is relational, coherent, and may be analyzed as a static system. In certain ways, word embedding methods are vulnerable to the same, enduring critiques of these premises. In other ways, they offer novel solutions to these critiques. More broadly, formalizing the study of meaning with word embeddings offers theoretical opportunities to clarify core concepts and debates in cultural sociology, such as the coherence of meaning. Just as network analysis specified the once vague notion of social relations (Borgatti et al. 2009), formalizing meaning with embedding methods can push us to specify and reimagine meaning itself.

People's evaluation of the relative position of their income is not as accurate as the relative income hypothesis assumes. It is observed from empirical survey data that income evaluation is concentrated in the middle. We develop a model that assumes income comparison on a subjective income reference distribution to explain the centralization phenomenon of income evaluation. We conduct theoretical analysis and empirical parameter estimation using Bayesian statistical modeling. The theoretical analysis shows that the centralization of income evaluation distribution occurs when the subjective reference distribution is more dispersed than the objective distribution. Empirical analysis using Japanese data from 2015 shows that the relationship between subjective and objective distributions differed depending on social categories with different social experiences. Women had a more ambiguous distribution than men. Among men, those aged 45--54 had a subjective distribution closest to the objective distribution. Thus, the subjective reference income distributions that potentially define people's evaluation of their income and their differences based on social category were only clarified by constructing the model.

How does society work? How do groups emerge within society? What are the effects of emotions and memory on our everyday actions? George Homans, like us, had a perspective on what society is, except that he was a sociologist. Homans theory, which is an exchange theory, is based on a few propositions about the fundamental actions of individuals, and how values, memory, and expectations affect their behavior. In this paper, our main interest and purpose are to find out whether these propositions can satisfy our conception of society and generate essential properties of it computationally. To do so, Based on Homans' prepositions, we provide the opportunity for each agent to exchange with other agents. That is, each agent transacts with familiar agents based on his previous history with them and transacts with newly found agents through exploration. One novelty of our work is the investigation of implications of the base theory while covering its flaws with minimal intervention; flaws which are inevitable in a non-mathematical theory. The importance of our work is that we have scrutinized the consequences of an actual sociological theory. At the end of our investigation, we propose another proposition to Homans theory, which makes the theory more appealing, and we discuss other possible directions for further research.

We reflect on the criticism that the methods in the natural sciences have been an object of and argue that these very methods, seen from the right perspective, are crucial to the progress of social science. We propose a knowledge generation framework that sets the minimum requirementsfor a claim to be considered scientific knowledge and sketchan ontology of the objects of study. We present science as an accumulating yet self-revising process and present examples in the social sciences.

The article discusses fascism's place on the political spectrum. At present, there is no consensus among political scientists and economists on that issue, as it has been extraordinarily politicized and distorted during ideological struggles among various currents of socialism. From the very beginning, fascism was depicted by Marxists as belonging to the Right, while Fascists themselves wanted to build a society that transcends the Left-Right paradigm. However, few voices in academia have noted that practical implementation of the fascists’ ideas, inherited from the works of revolutionary and national syndicalists, exhibited predominantly leftist characteristics.The ambiguity of placing fascism in its proper place on the political spectrum can be confidently resolved by applying three primary factors that govern political spectrum polarization: attitude to private property, scope of individual freedom, and degree of wealth redistribution. The article argues that fascism is a particular current of non-Marxian socialism that utilized collectivization of consciousness and wealth redistribution as the main paths toward socialism rather than outright expropriation of private property or means of production. Simultaneously, it is acknowledged that private property rights were inhibited by the fascist state, even though de jure they were permitted.The fascist ideal of the “alternate way” had a logical inconsistency that produced an unstable equilibrium between labor and capital as well as between the man and the state. The politico-economic structure predictably collapsed to the left in the course of building a new society. Therefore, fascism could be correctly called the Right of the Left.

The research framework of this study is related to the problem of why and how the population´s beliefs born, develop and, finally extinct. In previous work, the kinetics of the groups that remain constant in their empire's nationalistic ideologies was developed. Now, this research tries to calculate how beliefs diminish until disappearing. Founded in basic historical data of two ideologies, Christianism and Sovietism, the aim of this work is to establish that the populations increase is inversely proportional to the difference between that population and the maximum Nm that could reach, dN / (Nm - N), and the elapsed time 1dt, where1 is a coefficient greater than zero. As an eternal duration is impossible, a decrement factor Nmexp (-2t) is introduced. This produces a differential equation that can be interpreted mechanically as the initial impulse that pushes and made it works. The conclusion can be desegregated in three parts: first, in terms of formal equations; second, in sociological terms, to say, religious beliefs are more important than political beliefs; and finally, in terms of political change, that is, slow transformations last very much longer than rapid ones.

La détection de communautés est une question centrale en analyse de réseaux. Cet article combine une approche socio-historique à la reconstrucion expérimentale de programmes informatiques afin d’éclairer l’histoire des premiers algorithmes de détection de clique, détection qui compte encore aujourd’hui parmi les problèmes NP-complets non résolus. Restituer les recherches menées par l’archéologue Jean-Claude Gardin depuis les années 1950 sur le traitement de l’information non-numérique et l’analyse de graphes permet de mettre en évidence des contributions précoces à l’informatique menées depuis les sciences humaines et sociales. Le fait que ces applications originales de l’informatique aux humanités aient reçu une réception et une reconnaissance limitées est éclairé par deux facteurs, outre les effets des historiographies et des bibliographies sur la conservation, la découvrabilité, et la réutilisation des résultats scientifiques : 1) les politiques de financement, ayant motivé le transfert des efforts de recherche sur les graphes depuis éphémère espace interdisciplinaire vers des organisations de recherche en informatique, domaine alors émergeant ; 2) les carrières erratiques des algorithmes, où l’efficacité, les erreurs, les corrections, et le statut des auteurs ont été des facteurs déterminants.

This is a preprint for the published article: Arseniev-Koehler, A., Cochran, S. D., Mays, V. M., Chang, K. W., & Foster, J. G. (2022). Integrating topic modeling and word embedding to characterize violent deaths. Proceedings of the National Academy of Sciences, 119(10), e2108801119. There is an escalating need for methods to identify latent patterns in text data from many domains. We introduce a new method to identify topics in a corpus and represent documents as topic sequences. Discourse Atom Topic Modeling draws on advances in theoretical machine learning to integrate topic modeling and word embedding, capitalizing on the distinct capabilities of each. We first identify a set of vectors ("discourse atoms'') that provide a sparse representation of an embedding space. Atom vectors can be interpreted as latent topics: Through a generative model, atoms map onto distributions over words; one can also infer the topic that generated a sequence of words. We illustrate our method with a prominent example of underutilized text: the U.S. National Violent Death Reporting System (NVDRS). The NVDRS summarizes violent death incidents with structured variables and unstructured narratives. We identify 225 latent topics in the narratives (e.g., preparation for death and physical aggression); many of these topics are not captured by existing structured variables. Motivated by known patterns in suicide and homicide by gender, and recent research on gender biases in semantic space, we identify the gender bias of our topics (e.g., a topic about pain medication is feminine). We then compare the gender bias of topics to their prevalence in narratives of female versus male victims. Results provide a detailed quantitative picture of reporting about lethal violence and its gendered nature. Our method offers a flexible and broadly applicable approach to model topics in text data.

Why do people adorn themselves with elaborate body piercings or tattoos, wear obstructing garbs, engage in life-threatening competitions and other wasteful and harmful but socially stipulated practices? Norms of cooperation and coordination, which promote the efficient attainment of collective benefits, can be explained by theories of collective action. However, social norms prescribing wasteful and harmful behaviours have eluded such explanations. We argue that signalling theory constitutes the basis for the understanding of the emergence of such norms, which we call signalling norms. Signalling norms emerge as a result of the uncertainty about who is a friend and who is a foe. The need to overcome this uncertainty arises when different groups compete for scarce resources and individuals must be able to identify, trust and cooperate with their fellow group members. After reviewing the mechanisms that explain the emergence of cooperation and coordination norms, we introduce the notion of signalling norms as markers of group distinction. We argue that adherence to signalling norms constitutes a commitment promoting parochial cooperation rather than a quality-revealing signal facilitating partner choice. We formalize our argument in a game-theoretic model that allows us to specify the boundary conditions for the emergence of signalling norms. Our paper concludes with a discussion of potential applications of our model and a comparison of signalling norms with related concepts.This article is part of the theme issue ‘The language of cooperation: reputation and honest signalling’.

This research examines the geographical distribution of behavior in line with social norms that are spread and maintained primarily by the effect of social conformity. These include widely held norms that good citizens vote, don’t commit crimes, get flu vaccinations, abstain from binge drinking, and comply with census reporting. A partial differential equation model is used to determine whether such behavior may have attained a geospatial equilibrium in the United States. An equilibrium, as the end state of a diffusion process, has definitive mathematical properties that can be used to test for equilibrium. This is done using recent data for the 48 contiguous states. Results confirm that behavior for several important social norms fits the equilibrium model geographically. Policy implications are briefly discussed.

This paper introduces a demographic agent-based model of mobility in opportunity structures featuring vacancy chains, such as jobs in bureaucracies or homes in residential housing systems. The novel computational treatment of vacancy dynamics is superior to analytic variants by enabling us to 1) model off-equilibrium situations; 2) simulate multiple, overlapping, exogenous shocks of varying duration and intensity; 3) instantiate dynamic systems with an inherent tendency towards stability; 4) refine theory through pattern-oriented modelling; 5) increase realism by relaxing standard Markovian assumptions of embeddedness, homogeneity, and stationarity; and 6) generate confidence bands for point estimates. Parametrising these computational vacancy models with empirical data on professional mobility among judges in a post-revolutionary context reveals an unexpected career mechanism, the purge boomerang. Meta-theoretically, this paper furthers computational ABMs by introducing a model that explicitly codes vacancies as agents, a formalisation of absence akin to the zero symbol in arithmetic.

We develop a simple mathematical model describing the dynamics of the gender gap in a labor market niche and study the effects of bias, market size, and market dynamism in the evolution of the system. A high-level characterization of the system is suggested by studying a large sample of the parameter space and specific cases of interest to policymaking are explored.

I propose a theory of social structure that challenges the widely accepted role of preferential attachment and triadic closure as primary mechanisms of network formation. For this, I build upon Feld's concept of social circles, Breiger’s concept of the duality of actors and groups, and Hinde’s concept of interactions and relationships. The theory emphasizes that ties between actors arise and evolve according to social circles and social situations in which they participate, a notion straightforwardly modeled through two-mode and projected networks. Using recent results aided by analyses of empirical and artificial networks, I argue that structural properties such as tie strength, heterogeneity of popularity and strength among actors, clustering, community formation, and segregation emerge from homophily, jointly with overlap and social activity—mechanisms introduced in this study. The mechanisms form the two-mode network, and these structural properties naturally arise in the one-mode projection. The results show that social circle and social situation size distributions modulate network structures by interweaving with social activity distributions, and that overlap increases segregation from a network viewpoint. This theory’s implications are broad, affecting several social processes ranging from social cohesion, tolerance, and child development to the spread of infectious diseases.

Multi-layer networks arise when more than one type of relation is observed on a common set of actors. Modeling such networks within the exponential-family random graph (ERG) framework has been previously limited to special cases and, in particular, to dependence arising from just two layers. Extensions to ERGMs are introduced to address these limitations: Conway--Maxwell-Binomial distribution to model the marginal dependence among multiple layers; a "layer logic" language to translate familiar ERGM effects to substantively meaningful interactions of observed layers; and non-degenerate triadic and degree effects. The developments are demonstrated on two previously published data sets.

While networks have been the standard mathematical representation for detailedarrangements of relations in society, the simplicity of the representation limits whatinformation can be preserved in the abstraction, and subsequently, what can bemathematically analyzed. In this paper, we introduce a more rigorous and flexibleabstraction for representing arrangements of relations in society, the Messy Structure.After introducing the representational form and theoretically justifying it, we introducesome basic generative procedures for creating Messy Structures with “hierarchical”properties.

Population behaviours, such as voting and vaccination, depend on social networks. Social networks can differ depending on behaviour type and are typically hidden. However, we do often have large-scale behavioural data, albeit only snapshots taken at one timepoint. We present a method that jointly infers large-scale network structure and a networked model of human behaviour using only snapshot population behavioural data. This exploits the simplicity of a few-parameter, geometric socio-demographic network model and a spin-based model of behaviour. We illustrate, for the EU Referendum and two London Mayoral elections, how the model offers both prediction and the interpretation of our homophilic inclinations. Beyond offering the extraction of behaviour-specific network-structure from large-scale behavioural datasets, our approach yields a crude calculus linking inequalities and social preferences to behavioural outcomes. We give examples of potential network-sensitive policies: how changes to income inequality, a social temperature and homophilic preferences might have reduced polarisation in a recent election.

The metaphor of the organization as a garbage can is often invoked as a playful insult. However, as was recognized early on by management theorists studying garbage can ideas, the unpredictability arising from garbage can decision making has the potential to be adaptively rational for organizations facing complex task environments. The chaos produced by intraorganizational goal conflict and fluid participation in collective decision making can aid in search by enabling firms to escape local performance peaks or competency traps. The decades-old hypothesis that conflict and chaos could promote adaptively rational search, however, has largely been overlooked in research on organizational design. This paper uses an agent-based model to evaluate these competing views and, in the process, identify conditions under which garbage can decision making is indeed adaptively rational. In particular, I explore the conditions under which preference conflict and fluid participation in decision making promote effective search in an uncertain task environment. I show that the biased and chaotic outcomes that emerge as a result of garbage can decision making---the very features of garbage cans that lead them to be perceived to be dysfunctional---can facilitate short-term exploitation and long-term exploration of uncertain technical landscapes when organizations engage in serial judgment of local alternatives and internal conflict over desired outcomes is not too extreme. I conclude that decision making routines that encourage chaotic conflict are robust to bounded rationality and complex task uncertainty and thus should be included in the organizational designer's portfolio.

In an earlier article published in this journal ("Concept Mover’s Distance,” 2019), we proposed a method for measuring concept engagement in texts that uses word embeddings to find the minimum cost necessary for words in an observed document to "travel" to words in a "pseudo-document" consisting only of words denoting a concept of interest. One potential limitation we noted is that, because words associated with opposing concepts will be located close to one another in the embedding space, documents will likely have similar closeness to starkly opposing concepts (e.g., "life" and "death"). Using aggregate vector differences between antonym pairs to extract a direction in the semantic space pointing toward a pole of the binary opposition (following "The Geometry of Culture," American Sociological Review, 2019), we illustrate how CMD can be used to measure a document"s engagement with binary concepts.

When and how to come out are difficult choices. In this research, we examine one form of disclosure: the inclusion of an LGBTQ keyword in one's online social media profile. We construct daily time series of the prevalence of American Twitter users whose self-descriptions contain LGBTQ keywords. Further, we construct daily time series of inferred Add and Delete events - i.e. we make best estimates for how many users per day make an edit to include a previously absent word or remove a word previously present. These we compare to relevant annual and one-time events such as LGBTQ Pride Month and the date of the Pulse Nightclub shooting. We confirm two pre-registered hypotheses and explore several others.

In this study we propose a Dirichlet-multinomial regression model and a Bayesian estimation method to analyse dynamic migration networks whereby the nodes are the origins and destinations, and the edges are the number of people migrating between the two over time. We apply this model to analyse 25,632,876 migration instances that take place between Turkeys 81 provinces from 2009 to 2018. We quantify the associations of economic, physical, social, political, and network factors with migration. We find that economic prosperity and population in and the spatial distance between the origin and destination, as well as network characteristics such as reciprocity, province popularity and centrality are important predictors of migration. We also find that electoral distance between provinces and the strength of the AKP (Justice and Development Party) in a province are associated with migration. The proportion of Kurds in a province has no sizeable association with in- and out-migration.

Anecdotal evidence suggests the elite may be more likely to contract an infectious disease during a pandemic, especially earlier on. In this paper, I propose general patterns for how the elite may experience contact differently and then utilizing mathematical models, simulate how those general patterns may affect infection rates among the elite and general population during a pandemic.

This preprint is an earlier draft of the forthcoming article: Arseniev-Koehler, Alina and Jacob Foster. "Machine learning as a model for cultural learning: Teaching an algorithm what it means to be fat". Sociological Methods & Research. As we navigate our cultural environment, we learn cultural biases, like those around gender, social class, health, and body weight. It is unclear, however, exactly how public culture becomes private culture. In this paper, we provide a theoretical account of such cultural learning. We propose that neural word embeddings provide a parsimonious and cognitively plausible model of the representations learned from natural language. Using neural word embeddings, we extract cultural schemata about body weight from New York Times articles. We identify several cultural schemata that link obesity to gender, immorality, poor health, and low socioeconomic class. Such schemata may be subtly but pervasively activated in public culture; thus, language can chronically reproduce biases. Our findings reinforce ongoing concerns that machine learning can also encode, and reproduce, harmful human biases.

Gun violence is a central public concern in the United States, annually leading to the deaths of 36,000 individuals and the non-fatal injuries of 85,000 others. It has been called an epidemic and a public health crisis. In May of 2019, a diverse group of researchers participated in a workshop at the University of Tennessee, Knoxville. This workshop was sponsored by the Center for the Dynamics of Social Complexity (DySoC) and the National Institute for Mathematical and Biological Synthesis (NIMBioS). The objectives of this workshop were to review the existing approaches on the mathematics and modeling of gun violence, identify and prioritize areas in the field that require further research, develop cross-disciplinary collaborations to gain new perspectives, and suggest research and data collection that could assist evidence-based policy recommendations. The purpose of this report is to present some of the responses to the mentioned objectives and to suggest areas of future research .

This essay uses an example of Margaret Archer’s morphogenetic approach to sociodynamics to explore the explanatory gap indicated in the following propositions: 1. Individuals act and interact in variable ways from constantly shifting states-of-affairs in pursuit of variable and inconsistent interests; 2. Groups are made-up of individuals and their actions/interactions; 3. The vast majority of groups, current and historic, are stable. Together, the above statements comprise one of the central, perennial problems of sociology; given that the three statements above are empirically verifiable and thus objectively true, we live in stable societies made up of mutually inconsistent, constantly fluctuating individual actions. But how can this be? How can the patterned regularity of social life be aggregated from the disorder and seeming randomness of individual actions? When modeled, sampled, and plotted, the data set of statement 1 should have wide fluctuation in its mean; indeed, rational choice models have shown this to be the case. However the data set for statement 3 should not, yet because 3 is derived from 1 (via the axiom held at statement 2), we are confronted with the above problem. The answer lies in what I call the Predictability Hypothesis, consisting of two clauses: A) of the possible paths to achieving her desired goals, an individual will choose the most predictable path towards the most predictable desired goal; if the individual cannot sufficiently predict the behavior or attributes of their interest or the path that constitutes pursuit of its achievement, she is very unlikely to choose to pursue the interest or to follow that path to the interest, and will instead choose a path and/or interest with a higher predictability; B) for comparable societies, those that provide more choice to their members simultaneously provide more predictability and will, ceteris paribus, be more stable than those that provide less choice and, thus, less predictability. I contend that this hypothesis allows the resolution of the above difficulty: the analytical centrality of (neo-Bayesian) predictability to every choice no matter the context bridges the explanatory gap between individual actions and the large-scale sociological phenomena of social stability; in the same way, the analytical centrality of stability to groups and group structures allows us to identify the predictable paths of agency. This analytical dualism is used to identify the mutual morphogenesis of both sociodynamic ‘poles’ in the example...individual and collective, ‘micro’ and ‘macro’, agent and structure-culture, in a way applicable to a wide range of rigorous sociological inquiry.

Cascades over social networks can spread information, beliefs, diseases, technologies, and behaviors. Simple cascades spread from mere contact and produce submodular influence curves. Complex cascades assume agents with thresholding behavior and may produce non-submodular influence curves. In this study, we run three experiments that request charitable donations from human participants and experimentally manipulate whether and where their peers donate. We find evidence that we can (1) direct donations to an otherwise unpopular charity and (2) elicit complex contagion as evidenced by a non-submodular influence curve. The findings represent the most straightforward evidence to date of treatment-induced complex contagion - explicitly and formally defined - in human decision-making.

Coefficient plots are a popular tool for visualizing regression estimates. The appeal of these plots is that they visualize confidence intervals around the estimates and generally center the plot around zero, meaning that any estimate that crosses zero is statistically non-significant at at least the alpha-level around which the confidence intervals are constructed. For models with statistical significance levels determined via randomization models of inference and for which there is no standard error or confidence intervals for the estimate itself, these plots appear less useful. In this paper, I illustrate a variant of the coefficient plot for regression models with p-values constructed using permutation tests. These visualizations plot each estimate's p-value and its associated confidence interval in relation to a specified alpha-level. These plots can help the analyst interpret and report both the statistical and substantive significance of their models. Illustrations are provided using a nonprobability sample of activists and participants at a 1962 anti-Communism school.

The article introduces a new approach to the problem of political spectrum polarization. Political science has introduced a multitude of spectrograms based on different factors, dimensions, axes, and cardinal points. Most often the graphics do not complement each other, and it seems that each of them describes a completely different reality. There was an urgent need to conduct an objective analysis of political philosophies and find the factors that influence the political spectrum polarization. For these purposes, the Qualitative Comparative Analysis (QCA) was used, as it allows to introduce a fraction of objectivity due to the use of a formalized mathematical apparatus of the theory of sets. Thirteen main political doctrines were analyzed and obtained that spectrum polarization depends on three conditions: attitude to private property, individual freedom, and wealth redistribution. As the factors that influence the political spectrum polarization were firmly determined, it became possible to build a spectrogram unambiguously.

Ego network data have a long history in the social sciences, acting as a bridge between traditional statistical techniques and network analysis. Ego network data provide personal network information, as the data are based on a sample of individuals. This chapter details the basic features of such data, describing the advantages, disadvantages and potential applications of using ego network data. Ego network data remain a popular choice, despite the growing availability of full networks from automated sources (such as cell phone records). This is in large part because ego network data are easy to collect but still provide a surprisingly large amount of network information. Ego network data are also quite flexible, with past work using the same basic data structure for widely different purposes: from measuring social support and social boundaries; to inferring global network features from a sample; to acting as a corrective for other sampling techniques (such as respondent-driven sampling). Ego network data are thus easy to collect and useful for a wide variety of substantive and methodological problems. Given the ease of collection and the flexibility of use, there is every reason to believe that ego network data will continue to be a useful option for network scholars. I end the chapter by discussing possible avenues for future work.

NOTE: A heavily revised version of this paper was later published under a different title. The revised paper is available at this link: https://journals.sagepub.com/doi/full/10.1177/0049124120986178 ABSTRACT: Statistical network methods have grown increasingly popular in the social sciences. However, like other nonlinear probability models, statistical network model parameters can only be identified to a scale and cannot be compared between groups or models fit to the same network. This study addresses these issues by developing methods for mediation and moderation analyses in exponential random graph models (ERGM). It first discusses ERGM as an autologistic regression to illustrate that ERGM estimates can be affected by unobserved heterogeneity. Second, it develops methods for mediation analysis for both discrete and continuous mediators. Third, it provides recommendations and methods for interpreting interactions in ERGM. Finally, it considers scenarios where interactions are implicated in mediation analysis. The methodological discussion is accompanied with empirical applications and extensions to other classes of statistical network models are discussed.

I investigate experimentally the effects of heterogeneity in social identities and inequality in resources on cooperative behavior in a social dilemma. The experiment also varies the overlap between social identity and resources. The results show that both heterogeneity in social identities and inequality in resources reduce cooperation. When social identity and unequal resources are crossed orthogonally, the existence of composite group memberships mitigates the negative effects of inequality, but not of heterogeneity in social identities. When social identity and unequal resources overlap, cooperation decreases substantially among the participants with low resources, in both ingroup and outgroup interactions. My results also show that participants who are assigned to advantaged roles cooperate significantly less than those who are assigned to disadvantaged roles.

Although agent-based models (ABMs) have been increasingly accepted in social sciences as a valid tool to formalize theory, propose mechanisms able to recreate regularities, and guide empirical research, we are not aware of any research using ABMs to assess the robustness of our statistical methods. We argue that ABMs can be extremely helpful to assess models when the phenomena under study are complex. As an example, we create an ABM to evaluate the estimation of selection and influence effects by SIENA, a stochastic actor-oriented model proposed by Tom A. B. Snijders and colleagues. It is a prominent network analysis method that has gained popularity during the last 10 years and been applied to estimate selection and influence for a broad range of behaviors and traits such as substance use, delinquency, violence, health, and educational attainment. However, we know little about the conditions for which this method is reliable or the particular biases it might have. The results from our analysis show that selection and influence are estimated by SIENA asymmetrically and that, with very simple assumptions, we can generate data where selection estimates are highly sensitive to misspecification, suggesting caution when interpreting SIENA analyses.

This study provides an extension of the social value orientation model and a tool, other-other Decomposed Games, to quantify the influence of social identity on social value orientations. Social identity is induced experimentally using the minimal group paradigm. Subsequently, the weights subjects add to the outcomes of outgroup others relative to ingroup others and to the absolute difference between the outcomes of ingroup and outgroup others are estimated. Results are compared to a control condition in which social identity is not induced. Results show that when the outgroup is better off than the ingroup, the average subject is spiteful: they derive negative utility from the outcomes of the outgroup other. When the ougroup is worse off than the ingroup, the average subject attaches similar weights to the outcomes of outgroup and ingroup others. There is also significant variation across subjects with respect to the level of ingroup bias.

Mixed-effects multilevel models are often used to investigate cross-level interactions, a specific type of context effect that may be understood as an upper-level variable moderating the association between a lower-level predictor and the outcome. We argue that multilevel models involving cross-level interactions should always include random slopes on the lower-level components of those interactions. Failure to do so will usually result in severely anti-conservative statistical inference. We illustrate the problem with extensive Monte Carlo simulations and examine its practical relevance by studying 30 prototypical cross-level interactions with European Social Survey data for 28 countries. In these empirical applications, introducing a random slope term reduces the absolute t-ratio of the cross-level interaction term by 31 per cent or more in three quarters of cases, with an average reduction of 42 per cent. Many practitioners seem to be unaware of these issues. Roughly half of the cross-level interaction estimates published in the European Sociological Review between 2011 and 2016 are based on models that omit the crucial random slope term. Detailed analysis of the associated test statistics suggests that many of the estimates would not reach conventional thresholds for statistical significance in correctly specified models that include the random slope. This raises the question how much robust evidence of cross-level interactions sociology has actually produced over the past decades.

The structure of social networks is usually inferred from limited sets of observations via suitable network sampling designs. In offline social network sampling, for practical considerations, researchers sometimes build in a cap on the number of social ties any respondent may claim. It is commonly known in the literature that using a cap on the degrees begets methodologically undesirable features because it discards information about the network connections. In this paper, we consider a mathematical model of this sampling procedure and seek analytical solutions to recover some of the lost information about the underlying network. We obtain closed-form expressions for several network statistics, including the first and second moments of the degree distribution, network density, number of triangles, and clustering. We corroborate the accuracy of these estimators via simulated and empirical network data. Our contribution highlights notable room for improvement in the analysis of some existing social network data sets.

Behavioral theories of organizational decision making emphasize that organizations are political coalitions. Despite considerable recent qualitative research in management and organizational theory on the role of politics in decision making and managing organizational change, quantitative research in this area has stalled. The reason for the lack of progress is not theoretical, but rather methodological; researchers lack empirical tools for understanding basic processes of coalition formation, evolution, and conflict in organizations. We introduce a novel method for modeling politics in organizations that builds on the model of intra-organizational conflict in March (1962), which we call “subcoalition cluster analysis” (sCCA). The main contribution of sCCA is that it identifies subcoalitions with consistent preferences that are in conflict without placing additional restrictions on the structure of individual preferences. We apply sCCA to two cases, Wikipedia and the Baseball Writers’ Association of America and show how leadership would benefit from conceiving of their membership as competing subcoalitions instead of individuals with idiosyncratic preference disagreement. Finally, we compare the performance of sCCA to Principal Component Analysis (PCA) and k-means clustering and demonstrate that sCCA does a better job identifying latent structure in the data when the organization consists of more subcoalitions, when individual preferences are not perfectly aligned with those of their subcoalition, and when observations are missing.

How people connect with one another is a fundamental question in the social sciences, and the resulting social networks can have a profound impact on our daily lives. Blau offered a powerful explanation: people connect with one another based on their positions in a social space. Yet a principled measure of social distance, allowing comparison within and between societies, remains elusive. We use the connectivity kernel of conditionally-independent edge models to develop a family of segregation statistics with desirable properties: they offer an intuitive and universal characteristic scale on social space (facilitating comparison across datasets and societies), are applicable to multivariate and mixed node attributes, and capture segregation at the level of individuals, pairs of individuals, and society as a whole. We show that the segregation statistics can induce a metric on Blau space (a space spanned by the attributes of the members of society) and provide maps of two societies. Under a Bayesian paradigm, we infer the parameters of the connectivity kernel from eleven ego-network datasets collected in four surveys in the United Kingdom and United States. The importance of different dimensions of Blau space is similar across time and location, suggesting a macroscopically stable social fabric. Physical separation and age differences have the most significant impact on segregation within friendship networks with implications for intergenerational mixing and isolation in later stages of life., Comment: Article: 23 pages, 3 figures. Supplementary material: 8 pages, 1 figure

This article asks why and how governments keep secrets from publics, journalists and politicians using the strategy of ‘cover storying’. To develop a theory of cover storying, insights are drawn from political sociologies of state secrecy and from recent sociological examinations of secrecy and deception in organisations. This theory is illustrated by analysing Cobra Mist, a secretive and deceptive Anglo-American Cold War intelligence operation. Examining recently declassified documents, this article develops a framework for the analysis of five interrelated narrative conditions that shape social processes of cover storying: correspondence; plausibility; accountability; constraint; and durability. In conclusion this article reflects on the broader implications of this analysis for contemporary state and organisational theories and understandings of secrecy.

The present work introduces and mathematically formalizes the identity labeling problem - given an individual in a social situatio, can we predict what identity(ies) he or she will be labeled with by someone else (the “labeler”)? We argue that existing predictive models of identity labeling are theoretically incomplete and provide results from a survey-based stimuli that confirm this intuition. We then introduce a novel approach to the identity labeling problem, which we call Latent Cognitive Social Spaces (LCSS). We demonstrate that LCSS is a better predictor of identity labeling in survey data than previous models and discuss its theoretical implications for linking the definition of the situation to identity theories.

Network models of diffusion predominantly think about cultural variation as a product of social contagion. But culture does not spread like a virus. In this paper, we propose an alternative explanation which we refer to as associative diffusion. Drawing on two insights from research in cognition—that meaning inheres in cognitive associations between concepts, and that such perceived associations constrain people’s actions—we propose a model wherein, rather than beliefs or behaviors per-se, the things being transmitted between individuals are perceptions about what beliefs or behaviors are compatible with one another. Conventional contagion models require an assumption of network segregation to explain cultural variation. In contrast, we demonstrate that the endogenous emergence of cultural differentiation can be entirely attributable to social cognition and does not necessitate a clustered social network or a preexistingdivision into groups. Moreover, we show that prevailing assumptions about the effects of network topology do not hold when diffusion is associative.

Recent methodological advances in indirect migration and mortality estimation reveal important unforeseen patterns underlying these population processes, yet accurate indirect estimation of fertility remains difficult. The primary fertility index for a population, the total fertility rate (TFR), requires data on births disaggregated by mother’s age and thus cannot be calculated for the many areas and time periods that lack such information. Here we discuss a universal methodological framework for estimating TFR using inputs as minimal as the age-sex structure of a population. We show that the implied total fertility rate (iTFR) accurately estimates fertility from a population's age-sex structure in a wide range of scales, time periods, and even species. We also discuss two extensions of the iTFR that offer improved accuracy with minimal additional data requirements. To demonstrate the utility of this approach, we produce the first complete county-level map of U.S. fertility, reconstruct historical TFRs for 1000 additional country-years up to 150 years prior to the collection of birth records, and estimate TFR for the U.S. conditioned on household income, a variable unrecorded on U.S. birth records. Given its parameter-free nature, the method has wide applicability across space and time. We anticipate that our methodological framework will allow extension of fertility analysis to new sub-populations, time periods, and geographies, expanding our ability to understand fertility processes.