Exponential random graph models of preschool affiliative networks

Exponential random graph models were used to assess the relevance of reciprocity, popularity, transitivity controlling for effects based upon ego and alter sex, and sex homophily, on the formation of affiliative ties in 19 Portuguese preschool peer groups. The number of times two children were recorded as nearest neighbors in focal samples was used as an indicator of the relationship's strength. Independent parameter estimates of the different models (one for each group) were summarized, separately for the three age groups (“3-year-olds”, “4-year-olds” and “5-year-olds”) using a multi-level approach to meta-analysis. Results showed that affiliative ties between children were sex segregated, highly reciprocal, more likely to be directed to a restricted number of children and with a tendency to create transitive triads. The structural processes underlying the formation of affiliative ties were quite stable between classrooms.     

Bayesian model selection for exponential random graph models

Exponential random graph models are a class of widely used exponential family models for social networks. The topological structure of an observed network is modelled by the relative prevalence of a set of local sub-graph configurations termed network statistics. One of the key tasks in the application of these models is which network statistics to include in the model. This can be thought of as statistical model selection problem. This is a very challenging problem—the posterior distribution for each model is often termed “doubly intractable” since computation of the likelihood is rarely available, but also, the evidence of the posterior is, as usual, intractable. The contribution of this paper is the development of a fully Bayesian model selection method based on a reversible jump Markov chain Monte Carlo algorithm extension of Caimo and Friel (2011) which estimates the posterior probability for each competing model.     

Methodological considerations in using complex survey data: an applied example with the Head Start Family and Child Experiences Survey

Complex survey data are collected by means other than simple random samples. This creates two analytical issues: nonindependence and unequal selection probability. Failing to address these issues results in underestimated standard errors and biased parameter estimates. Using data from the nationally representative Head Start Family and Child Experiences Survey (FACES; 1997 and 2000 cohorts), three diverse multilevel models are presented that illustrate differences in results depending on addressing or ignoring the complex sampling issues. Limitations of using complex survey data are reported, along with recommendations for reporting complex sample results.     

Bayesian analysis for partially observed network data,missing ties,attributes and actors

We consider partially observed network data as defined in Handcock and Gile (2010). More specifically we introduce an elaboration of the Bayesian data augmentation scheme of Koskinen et al. (2010) that uses the exchange algorithm (Caimo and Friel, 2011) for inference for the exponential random graph model (ERGM) where tie variables are partly observed. We illustrate the generating of posteriors and unobserved tie-variables with empirical network data where 74% of the tie variables are unobserved under the assumption that some standard assumptions hold true. One of these assumptions is that covariates are fixed and completely observed. A likely scenario is that also covariates might only be partially observed and we propose a further extension of the data augmentation algorithm for missing attributes. We provide an illustrative example of parameter inference with nearly 30% of dyads affected by missing attributes (e.g. homophily effects). The assumption that all actors are known is another assumption that is liable to be violated so that there are “covert actors”. We briefly discuss various aspects of this problem with reference to the Sageman (2004) data set on suspected terrorists. We conclude by identifying some areas in need of further research.     

Regression artifacts in time-series and longitudinal data

The problem of “regression artifacts” in causal inference, otherwise known as the problem of error and reliable irrelevant variance in “independent” variables used for matching or regression adjustment, is illustrated first in the time-series data where a treatment is triggered by an extreme measure. The “offset effect” in psychotherapy, Underwood's scalloped learning curve, and potential pseudo-effects in AIDS therapies are used as illustrations. The magnitude of such artifacts is computable if the autocorrelation pattern for various lags is known, and thus could be distinguished from genuine effects. For longitudinal studies in which a population of respondents is repeatedly measured, the problem of anchoring the matching or regression adjustments on a single wave of measurement (usually the first) is illustrated as affected by the proximally autocorrelated nature of such measures. Data from a famous study of the effects of job training are reinterpreted in light of this consideration.     

Using latent class analysis to detect behavioral patterns in systems of observational variables

Behavioral observation is widely used for data gathering in evaluation research. Yet it leaves the investigator with unique problems. Usually, multiple observations result in a hierarchical data set, where numerous data records exist for each subject. Researchers face data reduction problems at least at two levels. First, there is the well-known and often-addressed problem of reducing the number of variables in a data set with only one information record per individual. Second, there is the problem of summarizing data at the individual subject level. The easiest way to perform this latter type of aggregation involves using univariate summary measures as probabilities of “using” an item, means, or standard deviations for each item per subject. Other standard procedures include first order interactions between pairs of items. However, use of pair-wise interactions is restricted because of variable dependence within each subject (this Affects e.g. factor analysis), or because of the relatively high number of single observations (this Affects e.g., cluster analysis). In this paper we propose employing Latent Class Analysis (LCA) to reduce the amount of information in observational data sets. In a first step, LCA allows one to specify intraindividual behavior patterns. In a second step, LCA allows one to derive meaningful summary scores for each individual. The two steps are illustrated using data that describe peer play competence in Swedish toddlers.     

Policies over time and pareto optimality

This paper is concerned with sequences of policies that occur over time in voting models and planning procedures. The framework for our analysis includes assumptions that are satisfied by models in the corresponding literatures, together with other standard assumptions for microeconomic analysis that involve time. The starting point for our analyses is the prespective that results from combining the following (widely held) views: 1) certain voting models and planning procedures can be interpreted as being non-tatonnement or sequential processes (where each policy in the sequence that is generated is actually experienced by voters or consumers) and 2) an alternative being Pareto optimal in any given period (temporal Pareto optimality) is the appropriate efficiency criterion only if the alternative is the final outcome from a tatonnement process-and that, otherwise, one should examine the efficiency of the entire path (using intertemporal Pareto optimality). Our first observation about the planning literature is that is has (by and large) neglected the efficiency criterion that is appropriate for the discrete-time procedures that can be interpreted as non-tatonnement or sequential processes-and that, what's more, such trajectories will (in general) fail to meet this criterion. Our second observation identifies some results that can be used to establish that some of these trajectories will at least be ultimately intertemporally Pareto optimal. In our discussion of voting theory, we review Buchanan's opposition to requiring (social) choice consistency for voting procedures-and his argument for this position on (Pareto) efficiency grounds. We then consider voting procedures that can be interpreted as non-tatonnement or sequential processes and arrive at the conclusion that, in these cases, (i) majority rule cycles are intertemporally Pareto inefficient and (ii) achieving intertemporal Pareto optimality requires choice consistency. We then go on to show that related observations apply to Kramer's normative conclusions about his dynamical model of political equilibrium — and identify some further references where similar observations apply. In the final part of our discussion of voting models, we arrive at the further conclusion that, in the most relevant cases, a trajectory that stays in the temporal Pareto set is not necessarily more desirable (on efficiency grounds) than one that doesn't.This paper has been improved by helpful comments and suggestions that have been provided by Kenneth Arrow, an anonymous referee, and the editor who handled the paper.Peter Coughlin gratefully acknowledges financial support provided by (i) National Science Foundation Grant No. SES-8409352, and (ii) the Center for Advanced Study in the Behavioral Sciences at Stanford (with support from National Science Foundation Grant No. BNS-8700864 and the Alfred P. Sloan Foundation).     

Multilevel models for social networks: Hierarchical Bayesian approaches to exponential random graph modeling

In many applications, researchers may be interested in studying patterns of dyadic relationships that involve multiple groups, with a focus on modeling the systematic patterns within groups and how these structural patterns differ across groups. A number of different models – many of them potentially quite powerful – have been developed that allow for researchers to study these differences. However, as with any set of models, these are limited in ways that constrain the types of questions researchers may ask, such as those involving the variance in group-wise structural features. In this paper, we demonstrate some of the ways in which multilevel models based on a hierarchical Bayesian approach might be used to further develop and extend existing exponential random graph models to address such constraints. These include random coefficient extensions to the standard ERGM for sets of multiple unconnected or connected networks and examples of multilevel models that allow for the estimation of structural entrainment among connected groups. We demonstrate the application of these models to real-world and simulated data sets.     

Ordres socio-économiques et polarisation de la recherche dans l'agriculture : pour une critique des rapports science/société

What if farming were the exemplar of a knowledge-based society? Due to their many, ambivalent effects, research and innovation in agronomy lie at the center of trends now shaping the world of agriculture. To present this history, the types of pairing are examined between, on the one hand, processes for producing scientific knowledge and techniques and, on the other, contingent forms for organizing collective action. For this purpose, the notion of a “socioeconomic order” is introduced to account for the multiplication of collective spaces of action based on specific systems of norms and cooperation. Studying these pairings brings to light three “models of polarization of research” in France (the Colbertist model, academic polarization, and fragmented polarization), which have changed over the past fifty years. Accordingly, agriculture's recent history is interpreted as the joint transformation of socioeconomic orders and these modes of polarization. - Special issue on Agriculture and food.     

Linked Ego Networks: Improving estimate reliability and validity with respondent-driven sampling

Respondent-driven sampling (RDS) is currently widely used for the study of HIV/AIDS-related high risk populations. However, recent studies have shown that traditional RDS methods are likely to generate large variances and may be severely biased since the assumptions behind RDS are seldom fully met in real life. To improve estimation in RDS studies, we propose a new method to generate estimates with ego network data, which is collected by asking respondents about the composition of their personal networks, such as “what proportion of your friends are married?”. By simulations on an extracted real-world social network of gay men as well as on artificial networks with varying structural properties, we show that the precision of estimates for population characteristics is greatly improved. The proposed estimator shows superior advantages over traditional RDS estimators, and most importantly, the method exhibits strong robustness to the recruitment preference of respondents and degree reporting error, which commonly happen in RDS practice and may generate large estimate biases and errors for traditional RDS estimators. The positive results henceforth encourage researchers to collect ego network data for variables of interests by RDS, for both hard-to-access populations and general populations when random sampling is not applicable.     

Computing continuous core/periphery structures for social relations data with MINRES/SVD

When diagonal values are missing or excluded, MINRES is a natural continuous model for the core/periphery structure of a symmetric social network matrix. Symmetric models, however, are not so useful when dealing with asymmetric data. Singular value decomposition (SVD) is a natural choice to model asymmetry, but this method also requires the presence of diagonal values. In this paper we offer an alternative, more general, approach to continuous core/periphery structures, the minimum residual singular value decomposition (MINRES/SVD), where each node in the network receives two indices, an “in-coreness” and an “out-coreness.” The algorithm for computing these coreness vectors is a least squares computation similar to, but distinct from the SVD, again because of the missing diagonal values. And in contrast to the standard, symmetric MINRES algorithm, we can more accurately model asymmetric matrices. This allows us to distinguish, for example, countries in the world economy that are more in the exporting core than they are in the importing core. We propose two nested PRE (proportional reduction of error) measures of fit: (1) the PRE from the MINRES vector with respect to the data and (2) the PRE of the product of the two MINRES/SVD vectors. Applying the resulting method to citations between journals and to international trade in clothing, we illustrate insights gained from being able to model asymmetrical flow patterns. Finally, two permutation tests are introduced to test independently for the MINRES and MINRES/SVD results.     

Student Self‐Reported Positive Purpose Over Two Years in Urban Middle Schools

We used Developmental Systems Theory as a framework for understanding the role of contextual factors in the development of purpose in urban adolescents. The sample included primarily low‐socioeconomic students of color attending urban middle schools (n = 2,629; 10–16 years of age). Longitudinal data were collected at four time points across two years. We used hierarchical linear modeling to model change in purpose. On average, purpose declined over time. We also identified several predictors of intercept and slope. For example, Black youth had a higher average purpose intercept, as well as a steeper average decline than other racial/ethnic groups. Females demonstrated a higher average purpose intercept than males, but this effect disappeared when academic achievement was added to the model.     

Negative impact induced by foreign workers: Evidence in Malaysian construction sector

Though foreign workers served to overcome the labor shortage in the Malaysian construction market, over-dependence on foreign workers and the negative impacts induced have become a serious social problem. The aims of this research are to identify those negative impacts induced by foreign workers in the Malaysian construction market and to determine strategies in minimizing these negative impacts. Data were collected through a structured interview and survey. The questionnaires were delegated to professionals in construction projects who have direct contacts with foreign workers. Only respondents from those companies registered under the CIDB grade G7 were chosen for this survey. There were 117 sets of questionnaires completed and analyzed through the structured interviews. The three principal factors attracting foreign workers to the Malaysian construction market are “Preference of the employers”, “Lifestyle and working conditions”, and “Unattractiveness of the career pathway”. The main negative impacts induced by foreign workers are “Over-dependence on foreign workers”, “Increment in criminal activities or social problems”, and “Existence of illegal workers”. This research proposes that strategies such as “Attract local workers into the construction industry”, “Industrialized Building Systems”, “Eliminate illegal migration”, and “Improve governance structure” would be successful in minimizing negative impacts induced by foreign workers.     

Random errors in egocentric networks

The systematic errors that are induced by a combination of human memory limitations and common survey design and implementation have long been studied in the context of egocentric networks. Despite this, little if any work exists in the area of random error analysis on these same networks; this paper offers a perspective on the effects of random errors on egonet analysis, as well as the effects of using egonet measures as independent predictors in linear models. We explore the effects of false-positive and false-negative error in egocentric networks on both standard network measures and on linear models through simulation analysis on a ground truth egocentric network sample based on facebook-friendships. Results show that 5–20% error rates, which are consistent with error rates known to occur in ego network data, can cause serious misestimation of network properties and regression parameters.     

Norms,status and the dynamics of advice networks: A case study

The issue of the influence of norms on behavior is as old as sociology itself. This paper explores the effect of normative homophily (i.e. “sharing the same normative choices”) on the evolution of the advice network among lay judges in a courthouse. 0020 and 0025 social exchange theory suggests that members select advisors based on the status of the advisor. Additional research shows that members of an organization use similarities with others in ascribed, achieved or inherited characteristics, as well as other kinds of ties, to mitigate the potentially negative effects of this strong status rule. We elaborate and test these theories using data on advisor choice in the Commercial Court of Paris. We use a jurisprudential case about unfair competition (material and “moral” damages), a case that we submitted to all the judges of this court, to test the effect of normative homophily on the selection of advisors, controlling for status effects. Normative homophily is measured by the extent to which two judges are equally “punitive” in awarding damages to plaintiffs. Statistical analyses combine longitudinal advice network data collected among the judges with their normative dispositions. Contrary to what could be expected from conventional sociological theories, we find no pure effect of normative homophily on the choice of advisors. In this case, therefore, sharing the same norms and values does not have, by itself, a mitigating effect and does not contribute to the evolution of the network. We argue that status effects, conformity and alignments on positions of opinion leaders in controversies still provide the best insights into the relationship between norms, structure and behavior.     

The hidden curriculum of youth care interventions—A case study

Themes such as “effectiveness” and “evidence-based practice” currently dominate debates about the professional care of children and youth. Notwithstanding the importance of these themes, this rather narrow focus may leave the values and ideals underlying youth care interventions undiscussed. In this article we will present our case study of EQUIP as an example of the role of this “hidden curriculum” in youth care interventions. Qualitative research has been conducted, consisting of content analysis of documents, interviews and observations. For the analysis an existing theory on the content and structure of values is used (Schwartz, 1992). The results show that there are distinct differences in the way moral behavior is defined in a Dutch and American “version” of EQUIP. The results also suggest that a difference in value perspective might be related to the effectiveness of the program.     

Missing data in networks: exponential random graph (p1) models for networks with non-respondents

Survey studies of complete social networks often involve non-respondents, whereby certain people within the “boundary” of a network do not complete a sociometric questionnaire—either by their own choice or by the design of the study—yet are still nominated by other respondents as network partners. We develop exponential random graph (p¹) models for network data with non-respondents. We model respondents and non-respondents as two different types of nodes, distinguishing ties between respondents from ties that link respondents to non-respondents. Moreover, if we assume that the non-respondents are missing at random, we invoke homogeneity across certain network configurations to infer effects as applicable to the entire set of network actors. Using an example from a well-known network dataset, we show that treating a sizeable proportion of nodes as non-respondents may still result in estimates, and inferences about structural effects, consistent with those for the entire network.If, on the other hand, the principal research focus is on the respondent-only structure, with non-respondents clearly not missing at random, we incorporate the information about ties to non-respondents as exogenous. We illustrate this model with an example of a network within and between organizational departments. Because in this second class of models the number of non-respondents may be large, values of parameter estimates may not be directly comparable to those for models that exclude non-respondents. In the context of discussing recent technical developments in exponential random graph models, we present a heuristic method based on pseudo-likelihood estimation to infer whether certain structural effects may contribute substantially to the predictive capacity of a model, thereby enabling comparisons of important effects between models with differently sized node sets.     

AVERAGE PREDICTIVE COMPARISONS FOR MODELS WITH NONLINEARITY, INTERACTIONS, AND VARIANCE COMPONENTS

In a predictive model, what is the expected difference in the outcome associated with a unit difference in one of the inputs? In a linear regression model without interactions, this average predictive comparison is simply a regression coefficient (with associated uncertainty). In a model with nonlinearity or interactions, however, the average predictive comparison in general depends on the values of the predictors. We consider various definitions based on averages over a population distribution of the predictors, and we compute standard errors based on uncertainty in model parameters. We illustrate with a study of criminal justice data for urban counties in the United States. The outcome of interest measures whether a convicted felon received a prison sentence rather than a jail or non-custodial sentence, with predictors available at both individual and county levels. We fit three models: (1) a hierarchical logistic regression with varying coefficients for the within-county intercepts as well as for each individual predictor; (2) a hierarchical model with varying intercepts only; and (3) a nonhierarchical model that ignores the multilevel nature of the data. The regression coefficients have different interpretations for the different models; in contrast, the models can be compared directly using predictive comparisons. Furthermore, predictive comparisons clarify the interplay between the individual and county predictors for the hierarchical models and also illustrate the relative size of varying county effects.     

Le travail enseignant en lycée professionnel et ses paradoxes

The paradoxes in the work of teaching in French vocational education stem from the “symbolic closure” of these secondary schools. The latter, given the mass enrollment in general secondary schools, receive students marked by their failure in middle school. This has created a distance with the students' social and family environment and with firms, even though the latter are supposed to be the destination where these students are bound. The work of these teachers is strongly marked by “subjective ordeals” related to their subordinate position in the school system, to a concern for keeping peace in the schools and to doubts about integrating students in the world of work. Might the work of these teachers not lead paradoxically to reasserting the “weight of the educational form” by including therein actions that are supposed to help students “mature” and develop an “appropriate school identity”? In this case, the purpose of this work is the academic socialization and rehabilitation of young people as much as their vocational training as such.     

Violent children and structural violence: Re-signaling ‘RAD Kids’ to inform the social work professions

This article critically examines representations of children diagnosed with Reactive Attachment Disorder, or “RAD Kids”, and their construction as dangerous subjects. Based on ethnographic research within attachment therapy clinics, and among adoptive families, social workers, and medical professionals in the U.S. and Russia, the author suggests that notions of danger associated with “RAD Kids” actually reflect a social anxiety about the contexts of structural violence in which we are attempting to build families and raise children at the turn of the 21st century. The author culturally and historically contextualizes the signaling of “RAD Kids” as violent within literature on moral panics over children and youth. She explores how these representations function as an attempt to “resignal” public anxieties about the difficulties associated with building families through adoption, and especially, the adoption of formerly institutionalized children. The article provides a model for thinking about complex relationships between children, pathology, and power to inform the social work professions, and particularly practice with children diagnosed with RAD.