Many effects of interest to sociologists are nonlinear. Additionally, many effects of interest are interaction effects—that is, the effect of one independent variable is contingent on the level of ...another independent variable. The proper way to estimate, interpret, and present these two types of effects individually are well known. However, many analyses that combine these two—that is, tests of interaction when the effects of interest are nonlinear—are not properly interpreted or tested. The consequences of approaching nonlinear interaction effects the way one would approach a linear interaction effect are severe and can often result in incorrect conclusions. I cover both nonlinear effects in the context of linear regression, and—most thoroughly—nonlinear effects in models for categorical outcomes (focusing on binary logit/probit). My goal in this article is to synthesize an evolving methodological literature and to provide straightforward advice and techniques to estimate, interpret, and present nonlinear interaction effects.
Most analyses of sexual orientation and earnings find that gay men face a wage gap, whereas lesbian women earn higher wages than similar heterosexual women. However, analyses rarely consider bisexual ...men and women as a unique group separate from other sexual minorities. I argue that such binary views of sexual orientation—treating sexual minorities as a homogenous non-heterosexual group—have obscured understandings of the impact of sexual orientation on labor market outcomes. Specifically, I predict that unequal outcomes for gay men and lesbian women are partly due to the influence of family arrangements and their effects on earnings. In contrast, I argue that bisexual men and women should be the most disadvantaged in the labor market, due to particularly disadvantaging stereotypes, perceptions of choice to their sexual orientation, and prejudicial treatment. Using data from the General Social Survey (N = 13,554) and the National Longitudinal Study of Adolescent to Adult Health (N = 14,714), I show that family arrangements explain some of the observed earnings differentials for gay men and lesbian women. Bisexual men and women, in contrast, face wage penalties that are not explained by human capital differences or occupational characteristics. Perceptions of prejudicial treatment partially explain the observed wage gaps.
Many research questions involve comparing predictions or effects across multiple models. For example, it may be of interest whether an independent variable's effect changes after adding variables to ...a model. Or, it could be important to compare a variable's effect on different outcomes or across different types of models. When doing this, marginal effects are a useful method for quantifying effects because they are in the natural metric of the dependent variable and they avoid identification problems when comparing regression coefficients across logit and probit models. Despite advances that make it possible to compute marginal effects for almost any model, there is no general method for comparing these effects across models. In this article, the authors provide a general framework for comparing predictions and marginal effects across models using seemingly unrelated estimation to combine estimates from multiple models, which allows tests of the equality of predictions and effects across models. The authors illustrate their method to compare nested models, to compare effects on different dependent or independent variables, to compare results from different samples or groups within one sample, and to assess results from different types of models.
Are men and women categorized differently for similar sexual behavior? Building on theories of gender, sexuality, and status, we introduce the concept of precarious sexuality to suggest that ...men’s—but not women’s—heterosexuality is an especially privileged identity that is easily lost. We test our hypotheses in a series of survey experiments describing a person who has a sexual experience conflicting with their sexual history. We find that a single same-sex sexual encounter leads an observer to question a heterosexual man’s sexual orientation to a greater extent than that of a heterosexual woman in a similar situation. We also find that a different-sex sexual encounter is more likely to change others’ perceptions of a lesbian woman’s sexual orientation—compared to perceptions of a gay man’s sexual orientation. In two conceptual replications, we vary the level of intimacy of the sexual encounter and find consistent evidence for our idea of precarious sexuality for heterosexual men. We close with a general discussion of how status beliefs influence categorization processes and with suggestions for extending our theoretical propositions to other categories beyond those of sexual orientation.
Despite theoretical work suggesting that health behaviors should be considered in tandem rather than as individual and disconnected practices, little quantitative work has examined different ...lifestyles of health behavior practices. In addition, while a significant body of work has examined the association of holding multiple social roles and health outcomes, little work has examined how acquiring multiple roles in early adulthood influences health behavior.
This article (a) illustrates the utility of examining health lifestyles—defined as constellations of individual health behavior practices—and (b) contributes to the literature on how accumulating multiple social roles is associated with health.
Using two waves of the National Longitudinal Study of Adolescent to Adult Health (N = 12,098) a structural equation modeling approach is used to both (a) model latent health lifestyles from observed health behavior indicators, and (b) to predict health lifestyle membership based on changes in role-occupancy during the transition to early adulthood.
Results suggest that the type of social role matters, with intensive obligatory roles associated with lifestyles of less tobacco, alcohol, and marijuana use but also with physical inactivity. In contrast, voluntary roles are associated with more active lifestyles but increased alcohol use.
The results illustrate the importance of modeling overall health lifestyles rather than focusing only on individual health behaviors. The results also advance our understanding of how holding multiple roles is associated with health by extending the framework to an examination of health behavior.
•Health lifestyles can be modeled as constellations of health behaviors.•Accumulating social roles in early adulthood predicts health lifestyles.•Obligatory roles are associated with less substance use but physical inactivity.•Voluntary roles are associated with more substance use but physical activity.
In the midst of the current causal revolution, experimental methods are increasingly embraced across the social sciences. We first document the growth in the use of the experimental method and then ...overview the current state of the field along with suggestions for future research. Our review covers the core features of experiments that facilitate causal inference, and we offer numerous examples of different experimental designs and the types of research questions they are well-suited to test. We then harmonize popular but differing perspectives on validity in regard to experiments focused on theoretical considerations. Next, we detail a new framework of purposive sampling for social science experiments, suggesting researchers could benefit from moving beyond only two stock choices of representativeness or convenience. We then discuss issues of analysis unique to experimental data before ending with a critical discussion of null findings, open-science practices, social desirability, and other open debates in the experimental methods literature.
•Uses behavioral experiment to examine mental illness' stigma and status consequences.•Develops fictitious mental illness labels and descriptions of deviant behavior.•Mental illness labels do not ...independently affect status or stigma.•Descriptions of deviant behavior independently affect status, but not stigma.•Fear moderates the effect of labels and deviant behavior on stigma.
Despite the breadth of research on mental illness, there remains some ambivalence about the distinct and combined effects of mental illness labels and deviant behavior for stigma and status. To examine the effects of mental illness labels and deviant behavior on stigma and status, we use experimental methods. Because fear is tightly linked with understandings of mental illness, we also consider the role of fear in these processes. We find that absent behavioral descriptions, mental illness labels do not affect status; however, deviant behavior that is and is not labeled as a mental illness decreases status. Neither deviant behavior nor mental illness labels independently affect stigma; however, compared to unlabeled deviant behavior, deviant behavior that is labeled as a mental illness either increases or decreases stigma for fear-inducing and fear-neutral behavior, respectively. In summary, the effects of mental illness labels appear to depend on behavioral connotations, including fear.
Previous work shows that stereotypes influence women’s ability to attain and act in leadership positions, however much less work has examined the role that gendered expectations and stereotypes might ...play for subordinate behavior and how this might reinforce the gender leadership gap. Drawing on theories of gender and status, I predict gender differences in responses to—and behavior in—subordinate roles. In a series of experimental studies I find that men are more publicly critical of leaders and more willing to undermine leaders than are women. In two studies I find that men are more publicly critical in both high and low status subordinate roles, and under both men and women leaders. In a third study, I find that men and women do not differ in their private behavior or evaluations of a leader and that gender differences in subordinate behavior and evaluations only arise in publicly visible evaluations, behavior, and criticism. Gender differences in public behavior and evaluation may be due to men exaggerating their criticism to restore a threatened sense of status or masculinity, or due to women tempering their criticism due to gender stereotypes that discourage assertive and critical behavior for women. Either explanation suggests that men and women perform gendered expectations in public and behave in ways that can disadvantage women leaders.
Objective
This study examines whether parental leave‐taking affects perceptions of workers as good parents, and whether this relationship varies between mothers and fathers.
Background
Contemporary ...parenthood norms promote intensive parenting, but pressures to engage in intensive parenting are more pronounced for mothers than fathers. Paid parental leave policies have the potential to improve gender equality, but only if such policies help to change gendered parenting norms. We assess whether leave‐taking further entrenches gendered parenting norms within workplaces or promotes greater perceptions of good parenting among fathers.
Method
We use combined data on 3333 respondents from two survey experiments in which parental leave‐taking and parental gender were randomly assigned. This enables us to assess the causal effects of parental leave‐taking on perceptions of mothers and fathers as good parents.
Results
Mothers and fathers in married, different‐sex partnerships are more likely to be viewed as good parents when they take longer periods of leave. Also, the positive effects of leave‐taking on perceptions of workers as good parents are stronger for fathers than for mothers—but only for short amounts of leave taken.
Conclusion
Greater access to, and use of, paid parental leave may enable more parents to be perceived as fulfilling contemporary parenting norms, and may especially increase the likelihood that fathers are viewed as good parents. As such, these policies may help to change gendered perceptions of parenting and promote greater gender equality.
Americans do not always fully utilize available parental leave policies due (in part) to fear of a commitment penalty – where taking leave (or taking longer periods of leave) lowers perceptions of ...job commitment. Using a survey experiment (N = 1713) to identify whether organizational leave policies affect perceived job commitment, we find that leave-taking (and taking longer periods of leave) is negatively associated with perceived commitment. However, perceived commitment is higher when workers take leave under more favorable policies, and the effect of favorable policies on perceived commitment is greater for fathers than mothers. Overall, these results can help organizations design parental leave policies that increase the likelihood that workers can take needed leave without damaging their careers.
PDF
