Patterns of Observed Dyadic Ethnic-Racial Socialization in a Sample of Black and Latinx Preadolescent-Caregiver Dyads

• In Event: Beyond the Survey: Innovative Methods for Measuring and Analyzing Ethnic Racial Socialization

Abstract

How parents and youth contribute simultaneously to conversations about race, ethnicity, bias, and discrimination remains an under-examined component of the literature on ethnic-racial socialization (ERS; Umaña-Taylor & Hill, 2020). Guided by the Theory of Racial Socialization in Action (Smith-Bynum et al., 2016), video-taped conversations among Black and Latinx parent-child dyads were examined for heterogeneous patterns of dyadic ERS. A minimum of three patterns were anticipated: one in which both family members displayed low engagement and warmth, one in which both showed high engagement and warmth, and one in which one member showed higher engagement than the other.
The sample was a multiply imputed, eight-wave longitudinal sample of 140 Black, 167 Latinx, and 46 Black or Latinx multiracial/multiethnic preadolescents (Mage= 11.19; 45.3% female) and their parents (94% mothers) with low socioeconomic standing. When youth were in 5th grade, 81 dyads (47% Black, 36% Latinx, 17% multiracial/multiethnic) were observed as they discussed how to respond to a hypothetical situation in which a White school counselor makes a derogatory statement to a student of color. Researchers trained on the Racial Socialization Observational Task and Coding Manual, Adapted for Ages 10-12 (RSOT; Smith-Bynum et al., 2019) rated conversations on 22 items from 0-very low/ not present to 4-very high addressing ERS content and frequency, as well as the emotional quality of the interaction. Estimated accuracies across raters ranged from 63%-99% (median = 90%; Bakeman, 2018). Missing data were multiply imputed with eight timepoints of auxiliary variables and 100 datasets using the PCAux package in R (Lang et al., 2015; Rhemtulla & Little, 2012). Density plots were used to assess the quality of the imputation of each RSOT variable. Intercorrelations and descriptive statistics of the imputed RSOT variables are presented in Table 1.
Results of a Latent Profile Analysis (LPA) suggested five patterns of dyadic ERS. Fitting with hypotheses, the Low Dyadic Engagement (n = 46) profile was marked by low engagement (averages < 2) by both dyad members, while the High Dyadic Engagement (n = 132) profile was distinguishable by the highest engagement of both members on nearly all items. Despite shared warmth and security between dyad members, parents in the Parent-Led (n = 65) profile took the lead in suggestion-giving and teaching. The defining features of the Justice Salient Advocates (n = 57) were their shared alertness to bias and injustice, as well as the highest rating on parent advocacy. Finally, although parents in the Child Dominant (n = 53) profile were still engaged, the child’s stated beliefs and suggestions were more frequent and emotionally emphasized than that of parents.

Very little research to date has examined ERS as a shared, dyadic process (exceptions: Aguayo et al., 2021; Dunbar et al., 2022; Johnson, 2005; Lewis, 1999). This presentation will discuss the benefits and pitfalls of imputing observed ERS data and analyzing parents and children’s contributions to ERS simultaneously via LPA. Similarities and distinguishing features among the identified profiles and prior research, as well as demographic differences among the profiles, will also be discussed.

Authors

• Kimberly R. Osborne, Arizona State University
  Presenting Author

• Margaret O'Brien Caughy, University of Georgia

• Mia A. Smith-Bynum, University of Maryland - College Park