Stern, J. A., Bailey, N. A., Costello, M. A., Hazelwood, O. A., & Allen, J. P. (2024). Fathers’ contributions to attachment in adolescence and adulthood: the moderating role of race, gender, income, and residential status. Attachment and Human Development . (Original work published 2024)

Fathers play a critical yet underappreciated role in adolescent development. To examine contributions of fathers’ parenting to attachment in adolescence and adulthood, this longitudinal study followed 184 adolescents from ages 13–24. At age 13, adolescents reported on their fathers’ parenting behavior and were observed in a father–teen conflict task; at ages 14 and 24, they completed the Adult Attachment Interview. Adolescents who lived with their father showed higher attachment security at age 14 (Cohen’s d = .72), compared to those with non-residential fathers. Fathers’ positive relatedness and support for teens’ psychological autonomy predicted attachment security at age 14. Fathers’ physical aggression predicted attachment insecurity in adolescence, whereas fathers’ verbal aggression predicted insecurity in adulthood, illuminating developmental shifts. Pathways to security were moderated by father residential status, adolescent gender, and race. Findings underscore the importance of fathers’ presence, autonomy support, and non-aggression in predicting adolescents’ state of mind in close relationships.

Allen, J. P. (2024). Rethinking peer influence and risk taking: A strengths-based approach to adolescence in a new era. Development and Psychopathology, 1-12. (Original work published 2024)

The ways that psychopathology manifests in adolescence have shifted dramatically over the past twenty-five years, with rates of many externalizing behaviors declining substantially while rates of anxiety and depressive disorders have skyrocketed. This paper argues that understanding these changes requires rethinking the field’s historically somewhat negative views of intense peer connections, peer influences, and adolescent risk-taking behavior. It is argued that intense peer connections are critical to development, and that peer influence and risk taking have important, often overlooked, adaptive components. The shift in observed manifestations of adolescent psychopathology over this period can be viewed at least partly in terms of a shift away from strong peer connections and toward greater risk aversion. Implications for research and intervention based on a focus on the adaptive aspects of peer influences and risk taking are discussed.

Stern, J. A., Bailey, N. A., Costello, M. A., Hellwig, A. F., Mitchell, J., & Allen, J. P. (2024). Empathy across three generations: From maternal and peer support in adolescence to adult parenting and child outcomes. Child Development. (Original work published 2024)

This study examined the development of empathic care across three generations in a sample of 184 adolescents in the United States (99 female, 85 male; 58% White, 29% African American, 8% mixed race/ethnicity, 5% other groups), followed from their family of origin at age 13 into their parenting years (through their mid-30s). Mothers' empathic support toward adolescents at age 13 predicted teens' empathy for close friends across adolescence (13–19 years). Participants' empathic support for friends in late adolescence predicted more supportive parenting behavior in adulthood, which in turn was associated with their children's empathy at age 3–8 years. Results suggest that individuals “pay forward” the empathic care they receive from parents, and that skills developed in adolescent friendships may inform later parenting.

Costello, M. A., Pettit, C., Hellwig, A. F., Hunt, G. L., Bailey, N. A., & Allen, J. P. (2024). Adolescent social learning within supportive friendships: Self-disclosure and relationship quality from adolescence to adulthood. Journal of Research on Adolescence. (Original work published 2024)

This study examines links between self-disclosure and relationship quality with close friends from adolescence to adulthood. A diverse community sample of adolescents (N = 184) participated in survey and observational measures annually from ages 13 through 29, along with their close friends and romantic partners. Random intercept cross-lagged panel modeling (RICLPM) was used to parse markers of within-individual change from age 13 to 18. Long-term longitudinal path models also investigated cascading associations among self-disclosure and relationship quality, on aggregate, from adolescence to adulthood. Adolescents who reported a higher-quality friendship in a given year demonstrated greater-than-expected increases in self-disclosure the following year, and an adolescent demonstrated high self-disclosure one year reported greater-than-expected increases in friendship quality the following year. Higher mean self-disclosure in adolescence predicted higher mean self-disclosure in adulthood. Results are interpreted as identifying high-quality adolescent friendships as key contexts for developing intimacy-building capacities (i.e. self-disclosure), which sets the stage for satisfying close relationships in adulthood.

Bailey, N. A., Costello, M. A., Stern, J. A., Davis, A. A., & Allen, J. P. (2024). Adolescent responses to paternal verbal aggression: Assessing spillover and compensatory processes using random intercept cross-lagged panel models. Journal of Adolescence. (Original work published 2024)


Prior research suggests several pathways through which verbal aggression manifests across adolescent relationship contexts, including spillover (continuity of aggression across different relationships) and compensation (offsetting an aggressive relationship with less aggression in other relationships). These pathways vary across timescales in ways that between-person analytic approaches are unlikely to adequately capture. The current study used random intercept cross-lagged panel modeling (RI-CLPM) to examine adolescents' spillover and compensatory responses to paternal verbal aggression.


Participants were 184 adolescents (53.2% female) from a United States community sample participating in a longitudinal study. Annually from ages 13–17, participants reported on their experiences of verbal aggression in their paternal and maternal relationships and participated in observed interactions with a close peer that were coded for aggressive behavior.


Spillover was observed from father-adolescent to mother-adolescent and adolescent-peer contexts in analyses at the between-person level, likely capturing long-term, cumulative effects of paternal aggression. Conversely, compensation was observed in analyses at the within-person level, likely capturing medium-term (i.e., year-to-year) adaptations to paternal aggression: Adolescents who experienced more aggression from their father than expected at a specific time point were less likely to both perpetrate and experience aggression in maternal and peer relationships the following year. Several findings differed across teen gender, with compensation more likely to occur in males than females.


These findings highlight the multiple pathways by which father-adolescent aggression may be linked to behavior in other relationships in the medium- and long-term. They also support the value of RI-CLPM in decomposing these effects.

Tang, D., Boker, S. M., & Tong, X. (2024). Are the Signs of Factor Loadings Arbitrary in Confirmatory Factor Analysis? Problems and Solutions. Structural Equation Modeling : A Multidisciplinary Journal. (Original work published 2024)

The replication crisis in social and behavioral sciences has raised concerns about the reliability and validity of empirical studies. While research in the literature has explored contributing factors to this crisis, the issues related to analytical tools have received less attention. This study focuses on a widely used analytical tool - confirmatory factor analysis (CFA) - and investigates one issue that is typically overlooked in practice: accurately estimating factor-loading signs. Incorrect loading signs can distort the relationship between observed variables and latent factors, leading to unreliable or invalid results in subsequent analyses. Our study aims to investigate and address the estimation problem of factor-loading signs in CFA models. Based on an empirical demonstration and Monte Carlo simulation studies, we found current methods have drawbacks in estimating loading signs. To address this problem, three solutions are proposed and proven to work effectively. The applications of these solutions are discussed and elaborated.

Szwedo, D. E., Davis, A. A., Fowler, C., Mikami, A. Y., & Allen, J. P. (2024). Social Media Posts from Friends during Late Adolescence as Predictors of Young Adult Physical Health. Journal of Youth and Adolescence. (Original work published 2024)

Although an increasing body of literature has linked social experiences to physical health, research has yet to consider how specific aspects of social experiences taking place on social media during late adolescence may predict future physical health outcomes. This study thus examined qualities of social media posts received from peers at age 21 as predictors of participants’ physical health (e.g., Interleukin-6 (inflammation), sleep problems, problems with physical functioning, and BMI) at age 28. Participants included 138 youth (59 men and 79 women); 57% of participants identified as White, 30% as Black/African American, and 13% as from other or mixed racial/ethnic groups. Posts from friends and participants at age 21 characterized by social ties predicted lower levels of future physical health problems, whereas socially inappropriate “faux pas” posts that deviated from peer norms by friends predicted higher levels of physical health problems at age 28. These associations were found after accounting for factors typically associated with physical health outcomes, including participants’ baseline social competence, internalizing and externalizing symptoms, alcohol use, observed physical attractiveness, and history of prior hospitalizations. The results of this study suggest the importance of both achieving social integration with peers online and adhering to peer norms in the online domain as key predictors of future physical health

Allen, J. P., Costello, M. A., Hellwig, A. F., & Stern, J. A. (2024). Pathways from adolescent close friendship struggles to adult negative affectivity. Development and Psychopathology.

This 19-year prospective study applied a social development lens to the challenge of identifying long-term predictors of adult negative affectivity. A diverse community sample of 169 individuals was repeatedly assessed from age 13 to age 32 using self-, parent-, and peer-reports. As hypothesized, lack of competence establishing and maintaining close friendships in adolescence had a substantial long-term predictive relation to negative affectivity at ages 27–32, even after accounting for prior depressive, anxious, and externalizing symptoms. Predictions also remained robust after accounting for concurrent levels of depressive symptoms, indicating that findings were not simply an artifact of previously established links between relationship quality and depressive symptoms. Predictions also emerged from poor peer relationships within young adulthood to future relative increases in negative affectivity by ages 27–32. Implications for early identification of risk as well as for potential preventive interventions are discussed.

Allen, J. P., Costello, M. A., Pettit, C., Bailey, N. A., & Stern, J. A. (2024). Unique Roles of Adolescents’ Friends and Fathers in Predicting Verbal Aggression in Future Adult Romantic Relationships. Development and Psychopathology. (Original work published 2024)

This 20-year prospective study examined verbal aggression and intense conflict within the family of origin and between adolescents and their close friends as predictors of future verbal aggression in adult romantic relationships. A diverse community sample of 154 individuals was assessed repeatedly from age 13 to 34 years using self-, parent, peer, and romantic partner reports. As hypothesized, verbal aggression in adult romantic relationships was best predicted by both paternal verbal aggression toward mothers and by intense conflict within adolescent close friendships, with each factor contributing unique variance to explaining adult romantic verbal aggression. These factors also interacted, such that paternal verbal aggression was predictive of future romantic verbal aggression only in the context of co-occurring intense conflict between an adolescent and their closest friend. Predictions remained robust even after accounting for levels of parental abusive behavior toward the adolescent, levels of physical violence between parents, and the overall quality of the adolescent’s close friendship. Results indicate the critical importance of exposure to aggression and conflict within key horizontal relationships in adolescence. Implications for early identification of risk as well as for potential preventive interventions are discussed.


Boker, S. M., Oertzen, T., Pritikin, J., Hunter, M. D., Brick, T. R., Brandmaier, A. M., & Neale, M. C. (2023). Products of Variables in Structural Equation Models. Structural Equation Modeling: A Multidisciplinary Journal , 30(5). (Original work published 2023)

A general method is introduced in which variables that are products of other variables in the context of a structural equation model (SEM) can be decomposed into the sources of variance due to the multiplicands. The result is a new category of SEM which we call a Products of Variables Model (PoV). Some useful and practical features of PoV models include the estimation of interactions between latent variables, latent variable moderators, manifest moderators with missing values, and manifest or latent squared terms. Expected means and covariances are analytically derived for a simple product of two variables and it is shown that the method reproduces previously published results for this special case. It is shown algebraically that using centered multiplicands results in an unidentified model, but if the multiplicands have non-zero means, the result is identified. The method has been implemented in OpenMx and Ωnyx and is applied in five extensive simulations.