Artificial intelligence and formal models of international behavior

During the past five years, models and techniques developed in artificial intelligence (AI) have been applied to a wide variety of topics in international relations (IR). These applications build on a theoretical base established at MIT during the 1970s, and the expansion of the applications reflects both changes in AI research and limitations of alternative models such as statistical modeling and rational choice approaches. Current applications in IR/AI tend to fall into three categories. First, there have been extensive applications of if-then rules in expert systems and simulations. Second, many models emphasize the role of historical precedent in IR decision making. Third, a variety of approaches employing natural language are under development. These AI techniques have extended considerably the range and complexity of formal models of international behavior. He has spent the past four years working on applications of artificial intelligence to the problem of making short-term predictions of international behavior. He is author ofMicrocomputer Methods for Social Scientists (Sage 1987) and numerous articles on mathematical modeling.     

MODELS FOR GENERALIZED LOCATION SYSTEMS

A formal framework is introduced for a general class of assignment systems that can be used to characterize a range of social phenomena. An exponential family of distributions is developed for modeling such systems, allowing for the incorporation of both attributional and relational covariates. Methods are shown for simulation and inference using the location system model. Two illustrative applications (occupational stratification and residential settlement patterns) are presented, and simulation is employed to show the behavior of the location system model in each case; a third application, involving occupancy of positions within an organization, is used to demonstrate inference for the location system. By leveraging established results in the fields of social network analysis, spatial statistics, and statistical mechanics, it is argued that sociologists can model complex social systems without sacrificing inferential tractability.     

“Did I just do that?”—Six-month-olds learn the contingency between their vocalizations and a visual reward in 5 minutes

It has been shown that infants can increase or modify a motorically available behavior such as sucking, kicking, arm waving, etc., in response to a positive visual reinforcement (e.g., DeCasper & Fifer, 1980; Millar, 1990; Rochat & Striano, 1999; Rovee-Collier, 1997; Watson & Ramey, 1972). We tested infants to determine if they would also change their vocal behavior in response to contingent feedback, which lacks the social, emotional, and auditory modeling typical of parent-child interaction. Here, we show that in a single five-minute session infants increase the rate of their vocalizations in order to control the appearance of colorful shapes on an iPad screen. This is the first experimental study to demonstrate that infants can rapidly learn to increase their vocalizations, when given positive reinforcement with no social element. This work sets the foundations for future studies into the causal relationship between the number of early vocalizations and the onset of words. In addition, there are potential clinical applications for reinforcing vocal practice in infant populations who are at risk for poor language skills.     

Gender and Children's Housework Time in China: Examining Behavior Modeling in Context

Differentiated gender roles in adulthood are rooted in one's gender role socialization. In order to understand the persistence of gender inequalities in the domestic sphere, we need to examine the gendered patterns of children's housework time. Although researchers have identified behavior modeling as a major mechanism of gender role reproduction and characterized gender socialization as a contextually embedded process, few have investigated contextual variation in behavior modeling, particularly in non‐Western developing countries. Analyzing data from the China Family Panel Studies 2010, the author examined the differences in behavior modeling between boys and girls age 10–15 from 2‐parent families (N = 1,903) in rural and urban China. The results revealed distinctive gendered interplays in the way parental housework and employment behavior helps shape children's housework time. This analysis is a crucial illustration of how the distinctive sociocultural contexts of rural and urban China moderate the effects of housework‐behavior modeling on intergenerational gender role socialization.     

Talking Prose All These Years: Agent‐Based Modeling as Process‐Oriented Analysis

This article considers the implications of an approach to computer simulation called agent‐based modeling for process‐oriented analysis. It argues that many theoretical and methodological debates found in the latter field can be effectively advanced by the former. The argument is presented and then extended using a ubiquitous agent‐based model proposed to improve understanding of ethnic residential segregation. The argument has three strands. The first is that theoretical and methodological debates are unlikely to progress unless they can be “cashed out” empirically. The second is that agent‐based modeling (and its distinctive methodology) has capabilities to do this that existing research methods lack and, in fact, that agent‐based models are a natural way to represent “social process” as apparently conceived by process‐oriented analysis. The third is that possibilities exist for productive synthesis between agent‐based modeling and process‐oriented analysis with the former clarifying, instantiating, and perhaps even testing notions of process developed by the latter.     

On Parallel Process in Social Work Supervision

The recognition and study of parallel process in social work supervision offers both the social work supervisor and supervisee a rich learning opportunity at the moment when an impasse seems unworkable. The parallel process is an unconscious replication in the supervisory session of therapeutic difficulties which a supervisee has with a client. This replication may originate with the supervisor unwittingly modeling behavior that is then taken by the social worker into the therapeutic interaction with the client. This paper reviews the need for social workers to grasp the dynamics of the parallel process, discusses the literature for the historical development of the phenomenon, addresses supervisory methods that will uncover the process, and illustrates the supervisor's stages of exploration and modeling in addressing the parallel process.     

A Piagetian approach to social work practice with children and adolescents

While the field of social work is heavily committed to the treatment of children and adolescents with behavior problems, it has given little attention to the application of Piaget's egocentrism theory to these problems. This paper outlines the theory proposed by Piaget and reviews subsequent empirical validation and clinical applications of his theory. The role of social work in developing more sophisticated applications of Piaget's theory is discussed.     

A Framework For The Study of Individual Behavior and Social Interactions

Recent work in economics has begun to integrate sociological ideas into the modeling of individual behavior. In particular, this new approach emphasizes how social context and social interdependences influence the ways in which individuals make choices. This paper provides an overview of an approach to integrating theoretical and empirical analysis of such environments. The analysis is based on a framework due to Brock and Durlauf (2001, forthcoming). Empirical evidence on behalf of this perspective is assessed and some policy implications are explored.     

Assessing program impact using latent growth modeling: a primer for the evaluator

Latent growth modeling (LGM) has emerged as a flexible analytic technique for modeling change over time because it can describe developmental processes at both the inter- and intra-individual levels. The LGM method can also provide a means for testing the contribution of other variables in order to explain variability in growth trajectories. This paper didactically illustrates the use of LGM as an analytical tool in program evaluation. Specifically, a hypothetical evaluation of a high school drug prevention program was used to demonstrate: (a) how LGM can be used to assess the longitudinal impact of a prevention program by comparing treatment and control populations with respect to individual differences in initial status and in rate of change; and (b) how predictors of initial status (post-intervention) and growth selected on the basis of a particular program theory can be incorporated in the model to explain program impact. Some advantages and limitations of using LGM in program evaluation are highlighted.     

Transactional Process of African American Adolescents’ Family Conflict and Violent Behavior

This is the first longitudinal study of urban African American adolescents that has examined bidirectional effects between their family conflict and violent behavior across all of high school. Structured interviews were administered to 681 students each year in high school at ages 15, 16, 17, and 18 years. We used structural equation modeling to test a transactional model and found bidirectional effects between family conflict and violent behavior across the middle years of high school, while accounting for sex and socioeconomic status. Findings suggest a reciprocal process involving interpersonal conflict in African American families and adolescent engagement in youth violence.     

Academic and health-related trajectories in adolescence: the intersection of gender and athletics

This study takes a life course approach to adolescent academic and health-related behavior. Researchers have often studied academic achievement and substance use as unrelated, static, and de-contextualized phenomena. Applying latent growth curve modeling to a sample of 2,651 high school freshmen and sophomores in California and Wisconsin (1987-1990), this study treats achievement and substance use as dynamic trajectories, inter-related to each other, and influenced by proximate and structural contexts. In general, achievement is high at the start of high school and then declines. This downward trend is driven by male non-athletes. Substance use is initially low and then increases, with boys driving this increase. Finally, academic and health-related trajectories are most related among athletes, who experience long-term academic consequences from risky health-related behavior. Overall, gender-athletic status can serve as either protective or risk factor, depending on the behavior and the temporal perspective.     

Random-Effects Modeling of Categorical Response Data

In many applications observations have some type of clustering, with observations within clusters tending to be correlated. A common instance of this occurs when each subject in the sample undergoes repeated measurement, in which case a cluster consists of the set of observations for the subject. One approach to modeling clustered data introduces cluster-level random effects into the model. The use of random effects in linear models for normal responses is well established. By contrast, random effects have only recently seen much use in models for categorical data. This chapter surveys a variety of potential social science applications of random effects modeling of categorical data. Applications discussed include repeated measurement for binary or ordinal responses, shrinkage to improve multiparameter estimation of a set of proportions or rates, multivariate latent variable modeling, hierarchically structured modeling, and cluster sampling. The models discussed belong to the class of generalized linear mixed models (GLMMs), an extension of ordinary linear models that permits nonnormal response variables and both fixed and random effects in the predictor term. The models are GLMMs for either binomial or Poisson response variables, although we also present extensions to multicategory (nominal or ordinal) responses. We also summarize some of the technical issues of model-fitting that complicate the fitting of GLMMs even with existing software.     

Simulation as a Social Process in Organizations

An emergent research literature is starting to cohere on simulation as a sociological process within organizations. This paper shines a spotlight on this scholarship, and offers new ways to think about the dizzying array of simulation we encounter in our organizational, institutional, and everyday lives. I define simulation as an empirical social process and show how they vary in consequence by their experiential modality, their referential frame, and their perceived realism. I then document three conceptual trends in the literature: (i) treating simulation as an organizational technique for risk management; (ii) a focus on virtual reality, video games, and moral ambiguity; and (iii) studies of the impact of computer simulation on scientific knowledge production and the reorganization of some technical fields, such as weapons research, artificial intelligence, and meteorology. Organizational uncertainty tends to coalesce around disputes about the appropriate qualities and functions of a given simulation technique or technology. I conclude the paper by identifying how the sociology of simulation can connect with more established areas of contemporary research within organizations, work and occupations, and institutional sociology.     

The Factors Influencing Saving in a Matched Savings Program: Goals,Knowledge of Payment Instruments,and Other Behavior

This paper explores the saving behavior of the group of low-income households that have participated in the Australian Saver Plus matched savings program. The paper finds that a dynamic panel model represents a good technique for modeling the savings account balances of such participants. We find that, even after controlling for the unobservable individual response to the program incentive, the saving goal and education/financial literacy variables play a positive role in encouraging saving behavior. More importantly however, we find that programs such as Saver Plus are able to modify the saving behavior of individuals to the point where their prior behavior and their inherent attitude towards saving no longer play a significant role.

Decade Review: Observing Marital Interaction

This article reviews the advances made in the decade of the 1990s in observing marital interaction. Many technological advances in data collection, including synchronization of physiology, behavior, and cognition, and advances in data analysis such as sequential analysis, have yielded new understanding and advances in prediction of marital outcomes. The advances have also included the study of developmental processes, including the transition to parenthood and the study of midlife and older marriages. Central advances have been made in the study of affect and the study of power and in their integration. This advance has included the mathematical modeling of interaction using nonlinear difference equations and the development of typologies. There has been an added focus on health outcomes and the bidirectional effects of marriages on children. There has been an expansion of the study of marital interaction to common comorbid psychopathologies. Most important has been emergent theorizing based on the interaction of behavior, perception, and physiology, as well as their predictive power.     

Environmental Racial Inequality in Detroit

This study uses industrial pollution data from the Environmental Protection Agency's Toxics Release Inventory (TRI) and tract-level demographic data from the 2000 U.S. census to determine whether environmental racial inequality existed in the Detroit metropolitan area in the year 2000. This study differs from prior environmental inequality research in two important ways. First, it offers a positive rationale for using hazard proximity indicators. Second, it uses a distance decay modeling technique to estimate hazard proximity. This technique weights each hazard's estimated negative effect by distance such that the estimated negative effect declines continuously as distance from the hazard increases, thus providing more accurate estimates of proximity-based environmental risk than can be obtained using other variable construction techniques currently found in the literature. Using this technique, I find that Detroit's black neighborhoods were disproportionately burdened by TRI facility activity in 2000 and that neighborhood racial composition had a strong independent effect on neighborhood proximity to TRI activity.     

The economics of housespousery: An essay on household work

The economic model of the household has been used extensively to examine questions relating to work within the household. This essay takes a critical look at this model and these household work applications. It concludes that the model provides several useful insights regarding factors that have influenced the decline in women's household work time during this century and the changes in the overall economic value of household work. In other areas (e.g., parental child care), the model has rarely been used although it has the potential to provide useful insights. Finally, in certain applications, such as investigations of the gender-related specialization of function within the household, the model relies on unrealistic assumptions that need to be modified if it is to prove useful. His research interests include the consumption, investiment, and time-use behavior of families and family members. Her research interests include family time allocation, household structure and economic well-being, and consumer policy.     

Exploring Gender Differences on Internalizing and Externalizing Behavior Among Maltreated Youth: Implications for Social Work Action

This purpose of this study was to explore the moderating influence of gender on the relationship between child maltreatment and internalizing symptoms (e.g., affective and somatic problems) and externalizing behavior (e.g., rule breaking behavior and aggression) among children aged 7–12 years old. Using a longitudinal comparison group design and a sample of 300 youth of which 56% (n = 168) had substantiated cases of child maltreatment, results of a structural equation modeling revealed that internalizing symptoms exerted a mediating influence that was conditioned by gender. Only girls’ internalizing symptoms were found to mediate the link between child maltreatment and externalizing behavior while a direct relationship between maltreatment and externalizing behavior was found among boys. These findings provide evidence for gender differences in the pathways between being child maltreatment and maladaptive thoughts, feelings, and behaviors. Thus, adapting evidence-based strategies that target gender specific internalizing behaviors and externalizing behaviors among maltreated youth may significantly reduce the risk of short and long-term maladaptive behavior.     

Adolescents' Prosocial Behavior Toward Family,Friends, and Strangers: A Person‐Centered Approach

This study examined longitudinal change in adolescents' prosocial behavior toward family, friends, and strangers. Participants included 491 mother–child dyads (average age of child at Time 1 = 11.5, 67% European American). Growth mixture modeling suggested that prosocial behavior toward family was generally stable or decreased over time, while prosocial behavior toward friends increased over time. However, findings highlighted unique developmental trajectories within subgroups of adolescents for prosocial behavior toward family and friends and found that maternal warmth and adolescent sympathy, self‐regulation, and gender consistently distinguished between groups. Discussion focuses on the need for a more multidimensional approach to prosocial development.