Exponential Family Models for sampled and census Network data

Much progress has been made on the development of statistical methods for network analysis in the past ten years, building on the general class of exponential family random graph (ERG) network models first introduced by Holland and Leinhardt (1981) . Recent examples include models for Markov graphs, "p*" models, and actor‐oriented models. For empirical application, these ERG models take a logistic form, and require the equivalent of a network census: data on all dyads within the network. In a largely separate stream of research, conditional log‐linear (CLL) models have been adapted for analyzing locally sampled ("egocentric") network data. While the general relation between log‐linear and logistic models is well known and has been exploited in the case of a priori blockmodels for networks, the relation for the CLL models is different due to the treatment of absent ties. For fully saturated tie independence models, CLL and ERG are equivalent and related via Bayes' rule. For other tie independence models, the two do not yield equivalent predicted values, but we show that in practice the differences are unlikely to be large. The alternate conditioning in the two models sheds light on the relationship between local and complete network data, and the role that models can play in bridging the gap between them.     

Deciding Not to Un–Do the “I Do:” Therapy Experiences of Women Who Consider Divorce But Decide to Remain Married

This study explores women's experience of marital therapy while they navigated decision making around divorce. A qualitative method was used to gain a deeper understanding of the participants' therapy and relationship decision‐making experiences. How are women's decisions whether or not to exit their marriage affected by therapy? The researchers interviewed 15 women who had considered initiating divorce before they turned 40 and had attended at least five marital therapy sessions but ultimately decided not to divorce. In general, participants reported that the therapy was helpful to them, their decision‐making process and their marriages. Five main themes emerged from the interviews: Women Initiated Therapy, Therapist Was Experienced as Unbiased, Therapy was Helpful, Importance of Extra‐therapeutic Factors, and Gradual Process.     

The disagreeable behaviour of the kappa statistic

It is often of interest to measure the agreement between a number of raters when an outcome is nominal or ordinal. The kappa statistic is used as a measure of agreement. The statistic is highly sensitive to the distribution of the marginal totals and can produce unreliable results. Other statistics such as the proportion of concordance, maximum attainable kappa and prevalence and bias adjusted kappa should be considered to indicate how well the kappa statistic represents agreement in the data. Each kappa should be considered and interpreted based on the context of the data being analysed. Copyright © 2014 JohnWiley & Sons, Ltd.     

The problem of social generations: a critique of the new emerging orthodoxy in youth studies

Over the previous seven years the application of a social generation paradigm or ‘theory’ has gained increasing currency as a method in analysing young people's relationship with the life course. Whilst not a new concept or approach its resurgence and reconfiguration to ‘new’ times has seen some writers positioning it as a ‘new orthodoxy’ or ‘consensus’ within youth studies. In this it is seen as providing a conceptual framework that better helps us understand the complexity of circumstances and conditions that shape youth identities in late modern society. In this paper we examine and explore the underlying assumptions and claims that are made by those advocating the social generational paradigm, raising questions and seeking further clarification on a number of key themes. We accept youth studies needs to move beyond ‘old models’ that define and understand social context as a simply a tension between ‘structure or/and agency’ or as a ‘flavour’ to social action. To conclude therefore we propose the need to have an approach that is ecological and both accepts ‘social change’ and ‘continuity’ as critical parts of the life course, one that recognises the nature and influence of power and social reproduction, especially for different social classes, in shaping the experience of being young.     

The Greater Complexity of Lived Race: An Extension of Harris and Sim*

Objectives. Harris and Sim (2002) recently demonstrated the complexity of lived race by exploring patterns of racial self‐identification. They raised important sociological questions about the role of context in racial self‐identification, but offered an incomplete picture of ethnic fluidity by excluding Hispanics from their analyses. We address this limitation with data on Hispanics from the National Longitudinal Study of Adolescent Health (Add Health). Methods. Our social‐psychological approach, using compositional analysis, focuses on the lived experience of race and ethnicity as qualitatively similar conceptual categories. Results. Informed by the cognitive process of social categorization, we find that considerably more individuals show fluidity in racial and ethnic self‐identification across contexts than suggested by Harris and Sim. Conclusions. Harris and Sim's thesis is even more strongly supported by these findings than in their original analysis, and our findings reinforce their challenge to the assumed stability of racial and ethnic measurement in sociology. We conclude by proposing a change in the measurement of race/ethnicity in America.     

Family support and family therapy – same difference?

This article discusses the relationship between family support and family therapy and the historical ambivalence between them. Decisions on what form of help to offer families based on effectiveness, efficiency and acceptability are therefore crucial. A review of the characteristics and measures of effectiveness is used to articulate the similarities in both interventions. By illustrating the false dichotomy between them, practitioners in both modes can learn from each other to benefit families.     

The Myth of “Security Moms” and “NASCAR Dads”: Parenthood,Political Stereotypes,and the 2004 Election*

Objective. The objective of this article is to explore the impact of being a parent on political views and to test the accuracy of the “Security Mom” and “NASCAR Dad” labels that were pervasively applied in the context of the 2004 presidential election. Methods. The methods we employed consist of using data primarily from the 2004 National Election Study to determine whether parents differ from nonparents in terms of their political attitudes on a wide range of issues. We argue that parenthood affects women and men differently, so we analyze the sexes separately. Results. Mothers have important political distinctions from women without children, mostly in the case of social welfare issues. They do not, however, appear to be “Security Moms” in the post‐9/11 world. As for fathers, for the most part it makes little sense to talk about dads as a distinct voting bloc, as in very few cases do their political attitudes differ from men without children. Conclusions. The media's use of the labels “NASCAR Dad” and Security Mom” promoted an inaccurate understanding of the political preferences of parents. Beyond debunking these media myths, our results provide one of the first comprehensive looks at the impact of parenthood on political attitudes, an overlooked area in the adult socialization literature.     

Differences in Actual and Perceived Online Skills: The Role of Gender*

Objective. The literature on gender and technology use finds that women and men differ significantly in their attitudes toward their technological abilities. Concurrently, existing work on science and math abilities of students suggests that such perceived differences do not always translate into actual disparities. We examine the yet‐neglected area concerning gender differences with respect to Internet‐use ability. In particular, we test how self‐perceived abilities are related to actual abilities and how these may differ by gender. Methods. We use new data on web‐use skill to test empirically whether there are differences in men's and women's abilities to navigate online content. We draw on a diverse sample of adult Internet users to investigate the questions raised. Results. Findings suggest that men and women do not differ greatly in their online abilities. However, we find that women's self‐assessed skill is significantly lower than that of men. Conclusions. Women's lower self‐assessment regarding their web‐use skills may affect significantly the extent of their online behavior and the types of uses to which they put the medium. We discuss the implications of these findings for social inequality.