Understanding how humans form and maintain interpersonal connections is paramount in an increasingly divided and digital world. What makes new acquaintances “click” with each other? When and how do relationships deepen to such an extent that people feel they have “merged minds” and report that they can “finish each other’s sentences”? And how do these connections influence people’s sense of certainty in their beliefs and their meaning in life? To answer these questions, my research introduces the novel construct of Generalized Shared Reality—the experience of sharing the same thoughts and feelings as an interaction partner about the world. This talk explores the role of shared reality in (1) forging and deepening interpersonal connections, (2) establishing a sense of certainty in one’s perceptions and beliefs, and (3) experiencing subjective well-being in daily life. This work suggests that shared reality bonds people to each other and shifts their perceptions of everyday experiences.

Much of the work on interpersonal symmetry has emphasized bivariate relationships, between two people. However, research in group dynamics suggests that symmetry might also be present at the group level. The Threeway Approximate Spatiotemporal Symmetry (TASS) algorithm was developed to capture symmetry between three signals in segments which correspond to either the presence or absence of symmetry. The TASS algorithm is an extension of the Pairwise Approximate Spatiotemporal Symmetry (PASS) algorithm (Sjobeck, Boker, Scheidt, & Tschacher, under review), which provides similar symmetry segmentation information between two signals. Like the PASS algorithm, the TASS algorithm captures association between signals using short windows of the time series which are lagged to capture non-simultaneous relationships. Additionally, the TASS and PASS algorithms both emphasize how the association pattern changes across time so as to determine which time points are and are not indicative of symmetry. Unlike the PASS algorithm, however, which uses correlation as a measurement of association, the TASS algorithm uses total correlation, an association metric from the field of information theory. To capture the underlying pattern of association between three signals, the TASS algorithm considers lags in all three time series. In order to understand how it performs under known conditions, the TASS algorithm will be used to capture potential symmetry relationships in a number of simulated examples. The results of these simulations will be discussed here. The relative strength of the TASS algorithm will be discussed within the context of these simulations.

One of the founding assumptions of social psychology is that groups influence human behavior—in particular, that an attempt to change a person’s behavior will fail in the long run if it does not involve her group. There has been enormous research interest in how groups motivate behavior change, but debates exist about the types of group structures that motivate change, and causal evidence with real world groups is rare. I conduct two field experiments in different contexts and with different populations to test the influence of increasing the participatory nature of groups over long-term behavior and attitudes. Study 1 experiments with 65 work group (1,792 workers) in a multinational factory in China. Study 2 experiments with 32 staff groups (172 workers) in an elite university in the US. In each experiment, half of the groups were randomly assigned to a 20-minute participatory meeting once per week for six weeks, in which workers were invited to speak and supervisors mandated to listen. The other half of the groups continued with status quo meetings. Participatory meetings led to a 10.6% increase in treatment factory workers’ productivity, which endured for 9 weeks after the experiment. I found that the frequency of voice within the group, rather than information or goals, drove the behavioral change. The treatment also led workers to be less authoritarian and more critical about societal authority and justice, and more willing to participate in political, social, and familial decision-making. Results in study 2 replicated such findings. This research highlights the power of participatory group dynamics in changing behavior and generalized attitudes across different contexts, both for theoretical understanding and pragmatic intervention in behavioral and attitudinal change toward social institutions and hierarchy.

Twins regularly score nearly a standard deviation below the population mean on standardized measures of cognitive development in infancy, but exhibit average abilities by early childhood. Building on early observations of increases in mean intelligence scores over time (Wilson, 1974), the present study applies contemporary nonlinear growth methods to quantify the rate and shape of recovery of cognitive abilities in a large prospective sample of twins from 3 months to 15 years. Polynomial, exponential, and sigmoid growth models were fit to the data. The s-shaped Gompertz model fit the data best, yielding information on the average lower asymptote, total growth, rate of approach to the asymptote, and age of steepest growth. Twins in the present sample exhibited initial cognitive abilities 0.89 standard deviations below the population mean, but scored at the population mean by age 6. Cognitive growth was most rapid at 3.27 years. Biometric analyses revealed that shared environmental factors accounted for the majority of the variance in initial cognitive abilities as well as growth in cognitive abilities. Twins born prematurely scored significantly lower than full-term twins on cognitive assessments in infancy, but caught up by three years. Family SES was positively related to total growth in cognitive abilities. Children in low-SES homes exhibited a two year delay in their most rapid growth relative to children in high-SES homes. This ultimately led to an 8-point difference (0.53 standard deviations) in cognitive abilities between children in high-SES and low-SES homes. Findings highlight the importance of prenatal factors and family economic resources in cognitive development.

Youth's emotion-related abilities are integral to their socioemotional health and academic success. Cultural context serves as a backdrop to children’s socioemotional development through the transmission of societal values of emotion (e.g., Halberstadt & Lozada, 2011). Yet, cultural expectations about youth's emotion varies across contexts (e.g., school vs home or intragroup vs intergroup interactions). This variation may be particularly salient for African American youth who engage with multiple cultural contexts with different expectations for their socioemotional behaviors. Such cultural navigation may foster advanced emotion skill integration and train youth to develop flexible emotional repertoires (e.g., “emotional codeswitching”). My work addresses conceptions of African American youth's emotional codeswitching through investigations of families' emotion-related beliefs and behaviors and youth's emotion-related abilities. I will summarize key findings from several of my quantitative and qualitative investigations, discuss the implications of considering intersecting contexts of emotion, and describe future directions for work on African American youths’ socioemotional development.

Many would argue that college is the great equalizer and once completed, individuals who hold a 4-year degree should be able to reap the rewards a college education has to offer. In this paper, we investigate what happens to first-generation (“first-gen”) college students when they disclose to potential employers that they are the first person in their family to go to college. Conventional wisdom reflects that many people in the United States are enamored with stories of people who pulled themselves up by the bootstraps, and the reality is that only 36% of adults over the age of 25 in the U.S. possess a bachelor’s degree, making completion of a college degree an elite marker that some might wish to guard. We contend that graduates who disclose that they are the first in their family to go to college are evaluated less favorably compared to equally-accomplished graduates who make no such disclosures, because many gatekeepers in mainstream, middle-class organizations tend to believe that the effects of people’s origins and initial circumstances in life are permanent and long-lasting. We test this theory with a large-scale randomized resume audit study across the United States (N = 1,785) and a large follow-up experiment (N = 5,013). Consistent with our social-deterministic account, our experiments revealed that applicants who disclosed that they were a first-generation graduate were less likely to receive callbacks than were applicants who made no such disclosure, and that these unfavorable evaluations emerge only when evaluators were personally inclined to believe in social determinism: that a person’s social character is shaped profoundly and permanently by their social background and upbringing.

Bayesian methods have been widely used to estimate models with complex structures. To assess model fit and compare different models, researchers typically use model selection criteria such as Deviance Information Criteria (DIC) and Watanabe-Akaike Information Criteria (WAIC), the calculation of which is based on the likelihoods of the models. When models contain latent variables, the likelihood is specified as conditional on the latent variables in popular Bayesian software (e.g., BUGS, JAGS, Stan). Although it reduces computation work and does not affect model estimation, our previous findings have shown that model comparisons based on the conditional likelihood could be misleading. In contrast, marginal likelihoods can be obtained by integrating out the latent variables and be used to calculate model selection criteria. In this study, we evaluate the effect of using conditional likelihoods and marginal likelihoods in model selection for a series of models (e.g., growth curve models, growth mixture models, etc.). Simulation results suggest that marginal likelihoods are much more reliable and should be generally used for models with latent variables.

Zoom Link, password: 828326

Having friends in high places is often considered necessary to achieve success. Indeed, connections with upper-class individuals have been identified as a key component of social capital. Despite the tangible benefits upper-class network contacts can offer, we find that these networks have a dark side: the increased potential for unethical behavior, over and above one’s own social class. We propose that because upper-class individuals are less constrained by social norms, individuals with many upper-class contacts will perceive their network as having looser social norms. As a result, individuals with upper-class network ties will view morality as more relative and will be more likely to engage in unethical behavior. To test our core hypothesis that having upper-class contacts increases unethical behavior, we conducted five multi-method (archival, field, quasi-experimental, and experimental) studies involving a range of samples (CEOs, nationally representative adults, student roommates) in multiple cultures. Overall, the current research takes a property of networks (its class composition), links it to perceptions of that network (the perceived norm looseness of one’s network contacts) and connects it to a psychological mindset (moral relativism) that ultimately affects unethical behavior. These findings demonstrate that the benefits of social capital also carry a moral cost.