Parenting in Place: How Latina Mothers’ Mesosystems Shape Ethnic-Racial Socialization

Abstract

Objectives:

Ethnic-racial socialization is an important cultural-developmental process in U.S. Latinx families and can be influenced by the ethnic-racial compositions of family members’ ecologies, much of which extends beyond their neighborhoods. This study examined the ethnic-racial compositions of mothers’ mesosystems, operationalized using activity space methods, which capture the set of locations to which individuals are regularly exposed. For aim 1, we used a person-centered approach to identify profiles of mothers differentiated by the ethnic-racial composition of activity spaces they navigate. For aim 2, we explored how identified mothers’ activity space profiles predicted ethnic-racial socialization of their adolescents, including cultural socialization and preparation for bias.

Methods:

The sample included Latinx adolescents (N=547; M_W1age = 13.31 years; 55.4% girls; 89.6% U.S. born) and their mothers (n = 271 at W1) participating in the Caminos study in Atlanta, GA. The current study analyzed data from Wave 5 (2020) and Wave 6 (2020–2021).

Results:

We identified four profiles of mothers’ activity spaces, and these differentially predicted mothers’ ethnic-racial socialization.

Conclusions:

Moving beyond the examination of ethnic-racial socialization within singular microsystems (e.g., residential neighborhoods), this study indicates that day-to-day ethnic-racial exposures encountered by Latina mothers may influence how mothers socialize their adolescent children around issues of ethnicity and race.

In a racially stratified society such as the U.S., ethnic-racial socialization is a salient proximal process that influences youth cultural development and wellbeing (Hughes et al., 2006; Umaña-Taylor et al., 2004). Familial ethnic-racial socialization involves the transmission of knowledge from caregivers to children regarding their race and ethnicity (Hughes et al., 2006), and occurs within broader ecologies, including parents’ mesosystems. Mesosystems encompass two or more microsystems in which people have direct face-to-face contact (Bronfenbrenner & Morris, 2007; Witherspoon, White, Bámaca, et al., 2023). Prior work on ethnic-racial socialization has examined influences of a single microsystem (e.g., residential neighborhood; Y. Wang et al., 2022). Additionally, decade-old qualitative research suggesting that parents’ mesosystems influence their racial socialization messages has been informative about African American families (Winkler, 2012). Thus, we know little about mesosystem influences on ethnic-racial socialization practices among Latinx families, the largest ethnic minority group in the U.S. (Jordan, 2021).

Research informative about Latinx families’ microsystem exposures and ethnic-racial socialization has focused predominantly on established immigrant gateways (e.g., Pasco et al., 2021), characterized by historic settlement in ethnically segregated neighborhoods with predominantly Latinx residents (Jones & Pebley, 2014). This focus often precludes examination of more diverse environmental exposures that Latinx families may experience outside of established gateways. Indeed, recent demographic shifts in the U.S. have given rise to emerging immigrant gateways (Massey, 2008), where the settlement patterns of Latinx families are more varied (Frank & Akresh, 2016; Hall & Stringfield, 2014). A person-centered examination of Latinx parents’ mesosystems in understudied settlement areas can shed light on how different exposures to ethnicity and race in daily social environments may influence family practices of socializing youth about ethnicity and race. Given mothers’ salient role in familial ethnic-racial socialization (Umaña-Taylor et al., 2014), the current study explored how Latina mothers’ mesosystems, particularly the ethnic-racial compositions therein, shape ethnic-racial socialization.

Theoretical and Methodological Backgrounds

Place-based cultural-developmental frameworks posit that parental socialization is shaped by parental exposures across multiple microsystems, collectively forming their mesosystems (Witherspoon, White, Bámaca, et al., 2023). Indeed, the Integrative Model highlights that socialization in ethnically and racially minoritized families reflects the combined influences of heritage culture and parents’ experiences within their environments (García Coll et al., 1996). Depending on the sociodemographic characteristics (e.g., ethnic-racial compositions) of the locations where parents spend time, parents may perceive specific messages about ethnicity and race as crucial or choose particular socialization strategies believed to support their children (Winkler, 2012; Witherspoon et al., 2021, 2022). Recent scholarship advocates for integrating activity space frameworks with cultural-developmental research to better capture mesosystemic exposures (Witherspoon, White, Bámaca, et al., 2023). Based on this approach, mothers’ activity spaces serve as a methodological tool for capturing their mesosystems, encompassing a set of routine locations in mothers’ daily activities, such as home, workplace, children’s school, and church (Matthews & Yang, 2013; Witherspoon, White, Bámaca, et al., 2023).

Latinx families in emerging immigrant gateways can settle in neighborhoods where one ethnic or racial group makes up the majority (e.g., White, African American, Latinx; Brown & Brooks, 2006; Frank & Akresh, 2016) or, alternatively in neighborhoods where no single group holds the majority (i.e., ethnically and racially integrated; Hall & Stringfield, 2014; Jones, 2019). We know little about the ethnic-racial compositions of the extra-neighborhood locations that Latina mothers access. Although residential neighborhoods influence residents’ frequently visited locations (Browning et al., 2017; Browning & Soller, 2014), individuals within a given neighborhood often access different extra-neighborhood locations (Freisthler et al., 2016; Wolf et al., 2017). For example, one study found that residents of predominantly Latinx neighborhoods were less likely to travel to predominantly White neighborhoods compared to residents of predominantly White neighborhoods (Wang et al., 2018). Similarly, some African American families in Detroit sought leisure and food in affluent White suburbs due to underdevelopment in predominantly Black Detroit; other families in predominantly White neighborhoods sought out spaces with a stronger Black presence (e.g., Black churches; Winkler, 2012). Accordingly, Latina mothers may spend time in locations (outside of their residential neighborhoods) where the ethnic-racial composition is similar to or different from their residential neighborhood.

In studies focused on the residential neighborhood, researchers have used one of two approaches for measuring ethnic-racial composition. Ethnic-racial diversity assesses the extent to which multiple ethnic-racial groups reside in a neighborhood. Diversity scores range from 0 to 1, with higher scores indicating greater diversity. Ethnic-racial concentration assesses the proportion of neighborhood residents represented by a single ethnic or racial group. The scores have a theoretical range of 0% to 100% (see Witherspoon, White, Nair, et al., 2023, for details). Much of the work on neighborhood effects with Latinx families has focused on Latinx concentration, particularly in established immigrant gateways where Latinx and White concentrations are strongly negatively correlated (> −0.90; Seaton et al., 2022; White et al., 2018). Thus, in the bulk of extant work, predominantly White neighborhoods serve as the de facto comparison group. To capture the ethnic-racial compositions of mothers’ mesosystems, we adopted a comprehensive approach assessing both Latinx concentration and ethnic-racial diversity, while also considering levels of White concentration.

Given the limited research on mothers’ mesosystemic exposures and the potential for differential exposures across maternal groups (Witherspoon, White, Bámaca, et al., 2023), a person-centered approach (Howard & Hoffman, 2017) is valuable to identify subgroups of Latina mothers differentiated by their exposure to different mesosystems characterized by Latinx concentration and diversity levels. Notably, similar diversity values can reflect disparate ethnic-racial concentrations (e.g., a low diversity index may reflect predominantly White, or predominantly Black/African American neighborhoods; see Witherspoon, White, Nair, et al., 2023 for further discussion). Moreover, socioeconomic status (SES) often confounds with residential ethnic-racial patterns (Massey & Denton, 2018). Higher SES neighborhoods often have higher proportions of White residents, a combination that is often assumed to be ideal for raising children (Goyette et al., 2014; Pattillo, 2014). However, empirical research suggests that exposure to whiteness can present challenges for ethnically and racially minoritized groups (e.g., Spencer et al., 2019; White et al., 2022). Thus, to fully understand Latina mothers’ mesosystems, it is important to describe Latinx concentration and ethnic-racial diversity relative to additional neighborhood (e.g., Black concentration, White concentration, poverty) and family (e.g., immigration status, SES) demographics.

Mesosystems and Ethnic-Racial Socialization

Researchers have identified a variety of dimensions of ethnic-racial socialization (Hughes et al., 2006). Cultural socialization and preparation for bias have been central in both theoretical work (Hughes et al., 2006) and empirical studies of Latinx families (Araujo Dawson & Quiros, 2014; Rivas-Drake, 2011). Cultural socialization consists of practices that promote children’s knowledge about their history and heritage and that instill group pride. Examining cultural socialization based on parent-specific efforts (e.g., mothers’ engagement in cultural socialization; Pasco et al., 2021; Umaña-Taylor et al., 2014) and youth perception of family-wide efforts (e.g., Umaña-Taylor & Guimond, 2010) are both common approaches. Preparation for bias consists of parents’ efforts to promote their children’s understanding of and skills for coping with ethnic-racial prejudice and discrimination (Hughes et al., 2006). Considering that middle adolescence is a critical period for ethnic-racial identity development (Fuligni et al., 2009), and the substantial body of research examining ethnic-racial socialization during middle adolescence as identified by a recent decade review (Umaña-Taylor & Hill, 2020), understanding the ecological influences on ethnic-racial socialization during this developmental period is essential.

Parents’ exposures to day-to-day environments, particularly concerning the race and ethnicity of individuals therein, can influence the messages parents convey to their children about race and ethnicity. A recent decade review identified that studies on cultural-developmental family processes were primarily situated within residential neighborhoods, but none examined these processes within broader mesosystems or activity spaces (White et al., 2021). Studies centered on residential neighborhoods have shown significant but mixed associations between ethnic concentration and ethnic-racial socialization (Hughes et al., 2016; Witherspoon et al., 2021). In a qualitative study, a higher concentration of African Americans in a predominantly African American city was associated with less ethnic-racial socialization, including cultural socialization and preparation for bias, in African American families. This finding indicated that ethnic-racial socialization might be perceived by parents as less necessary in environments rich in ethnic-racial pride and community history (Winkler, 2012). Conversely, among African American college students, a higher proportion of African Americans in the neighborhood was concurrently associated with increased cultural socialization (Barr & Neville, 2013). In a longitudinal study in an established immigrant gateway, U.S. Mexican mothers’ perception of a stronger Latinx culture in the neighborhood, which can co-vary with higher Latinx concentration and lower white concentration, was associated with greater maternal cultural socialization (Pasco et al., 2021). Mixed findings also characterize studies of neighborhood ethnic-racial diversity. For example, greater neighborhood ethnic-racial diversity was associated with increased cultural socialization and/or preparation for bias in a longitudinal study with a nationally representative sample of ethnically and racially minoritized children (Y. Wang et al., 2022) and in cross-sectional studies of Latinx parents in an emerging immigrant gateway (Witherspoon et al., 2021) and of African American children and adolescents (Stevenson et al., 2005; Varner et al., 2021). Conversely, findings from a cross-sectional study of both Latinx and African American parents in established immigrant gateways found no association between neighborhood ethnic-racial diversity and parents’ cultural socialization or preparation for bias (Witherspoon et al., 2022). The predominantly cross-sectional nature of existing research underscores the need for more longitudinal studies.

Inconsistencies in prior findings could reflect that residential neighborhoods are not the only settings influencing ethnic-racial socialization. Consistent with the notion that people often spend time outside their residential neighborhoods (Browning & Soller, 2014; Wolf et al., 2017), several studies have demonstrated extrafamilial contexts shaping family processes: U.S. Mexican mothers’ workplace experiences impacted parent-child relationships (Wheeler et al., 2015); higher ethnic-racial diversity in children’s schools increased familial cultural socialization (Wang et al., 2022); and, the size of parents’ activity spaces influenced parenting (Freisthler et al., 2016). Accordingly, Latina mothers might carry out different ethnic-racial socialization strategies depending upon the ethnic-racial composition of the mesosystems they access.

The Current Study

This study examined how the ethnic-racial compositions of Latina mothers’ mesosystems shaped ethnic-racial socialization. By operationalizing mesosystems using activity space methods, this study captured ecological influences in ways that prior single-setting (primarily neighborhood) contextual research has not. For the first aim, we estimated the profiles of Latina mothers based on the Latinx concentration and ethnic-racial diversity of their activity spaces. To further describe the profiles with sociodemographic variables and facilitate links to prior neighborhood-focused research on ethnic-racial socialization (e.g., Pasco et al., 2021; Witherspoon et al., 2021), we described each profile in terms of Black and White concentration (the metro area’s largest groups) and neighborhood poverty. To identify key individual factors shaping profile membership, we also described the profiles relative to mothers’ education level, survey language, and nativity; the husband’s presence in the household, mother nativity, and adolescent ethnic origin.

For the second aim, we explored whether mothers’ activity space profiles predicted ethnic-racial socialization, including cultural socialization and preparation for bias, 6 months later. In line with prior studies of cultural socialization, we examined both adolescent-reported family-wide and mother-reported maternal cultural socialization. The study’s significance lies in exploring how ethnic-racial socialization behaviors in Latinx families are influenced by mothers’ mesosystemic exposures to ethnicity and race in an emerging immigrant destination, moving substantially beyond microsystem effects on ethnic-racial socialization (Witherspoon, White, Bámaca, et al., 2023) and cross-sectional examinations (e.g., Barr & Neville, 2013; Witherspoon et al., 2021).

Methods

Sample and Procedures

The current study used Waves 5 (W5) and 6 (W6) data from the ongoing Pathways to Health/Caminos al Bienestar (“Caminos”) study (Roche et al., 2020). At baseline in 2018 (W1), participants include a school-based sample of 547 Latinx adolescents and a random sample of 271 of their mothers recruited from middle schools in suburban Atlanta, Georgia, an emerging immigrant gateway (Singer, 2015). To ensure family, school, and community heterogeneity, students were selected from a stratified random sample of schools with a “low” (<13%), “moderate” (18–25%), or “high” (>40%) Latinx student concentration. Adolescents who self-identified as “Hispanic” on 2017–2018 school enrollment lists were screened for eligibility and selected at random from grade (6^th, 7^th, 8^th) and gender strata within each of the three Latinx concentration clusters. Adolescents were eligible if they were in the typical age range for their grade, did not have a severe emotional or learning disability (as indicated by an Individualized Education Plan), could read and write in either English or Spanish, self-reported being Latinx (or a related term), and did not have a sibling previously selected. The response rate among eligible adolescents whose parent was contacted and confirmed adolescent eligibility was 65.2% (547 of 839) and 95.3% (547 of 574) among contacted adolescents with parental permission.

Mothers were eligible to participate if they spoke English or Spanish and identified as Hispanic or Latina or as being from a Spanish-speaking national origin. A mother sub-sample was randomly selected using a planned missing data design (Little & Rhemtulla, 2013). Among the 386 eligible and reachable mothers of enrolled adolescents, 271 mothers enrolled and completed baseline surveys. Study materials were provided in Spanish and English (Knight et al., 2009). Adolescent surveys were self-administered on an electronic device using Qualtrics software (Qualtrics, 2005); mothers completed phone surveys. Data collection for W5 (adolescent age = 15.26) and W6 (adolescent age = 15.76) occurred approximately 2 and 2.5 years, respectively, after W1 (W5: 2020; W6: 2020 – 2021) and correspond to middle adolescence. Study procedures were approved by a University Institutional Review Board.

At baseline, adolescents were, on average, 13.31 years old (SD = 1.03), and the majority (n = 303, 55.4%) were girls. Most adolescents were born in the U.S. (89.6%) and completed surveys in English (95.4%). More than half of the adolescents’ mothers completed at least high school (51.9%). About two-thirds (67.1%) of adolescents reported living with two parents, with 19.2% living with a single parent or no parent, and 13.3% living with a stepparent. Most mothers completed surveys in Spanish (80.1%). The majority of the 271 mothers (90.8%) were born outside the U.S., including 59.8% mothers born in Mexico; 11.8% born in Central America, 11.4% born in South America, and 7.7% in other countries. A total of 133 of 271 mothers (49.1%) completed at least high school. The majority of mothers (n = 229, 84.5%) reported having husband living at home. Neighborhood characteristics of the metro area where the Caminos study is situated (Atlanta-Sandy Springs-Alpharetta) indicate that White concentration is 43.2%, Black concentration is 33.8%, Latinx concentration is 11.5%, and that 7.4% of families live below the federal poverty level (Census Reporter, 2022). The ethnic-racial diversity of the overall Atlanta metro area is 0.6 (U.S. News, 2020), indicating a 60% probability that two randomly selected individuals from the metro area would be from different ethnic-racial groups.

Measures

Mesosystem ethnic-racial compositions (W5).

Residential addresses were collected from participants. Mothers also reported three places they visited in a typical week (e.g., shopping, church, parks, see Table S2). These 4 locations were geocoded and matched to census block groups. The ethnic-racial compositions of block groups were obtained from the 2019 file of the ACS 5-year Data (U.S. Census Bureau, 2016). Latinx concentration was measured by the percentage of Latinx residents in each census block group. The ethnic-racial diversity index represents the probability that two randomly selected individuals in the same census block group are from different ethnic and racial groups. It was calculated with a formula (Budescu & Budescu, 2012) that considers the number of ethnic-racial groups as classified in the U.S. Census and the proportion of each group, where higher scores indicate greater diversity (U.S. Census Bureau, 2016).

Ethnic-racial socialization (W5 & W6).

Both adolescents and mothers completed the 12-item Familial Ethnic Socialization Measure (Umaña-Taylor et al., 2004), which assessed their cultural socialization (e.g., “[My family teaches me/ I teach my child] about the history of [my/our] ethnic/cultural background”). Items were scored on a 5-point Likert scale ranging from 1 (almost never or never) to 5 (almost always or always), with higher scores indicating higher levels of cultural socialization. Both adolescent (W5: α=.93, W6: α=.93) and mother reports (W5: α=.85, W6: α=.86) demonstrated good reliability (see also Table S3). Mothers completed 6-items from the Parent’s Messages to Children about Race/Ethnicity Scale (Hughes & Chen, 1997), which assessed their preparation for bias (e.g., “Talked to your child about discrimination or prejudice against your ethnicity”). Items were scored on a 5-point Likert scale ranging from 1 (never) to 5 (six or more times), with higher scores indicating higher levels of preparation for bias. The scale demonstrated acceptable reliability (W5: α =.84, W6: α=.85; Table S3).

Demographic covariates.

Adolescents reported mother education. Adolescents’ reports on parents’ countries of birth were used to code adolescent ethnic origin (Mexican, other Latinx) and mother nativity (U.S. born, foreign-born). Mothers reported on husband’s presence in the household and survey language preference (Spanish or English). We relied on the same 2019 ACS 5-year data files to assess additional block group demographics, including White concentration (% White non-Hispanic population), Black concentration (% Black or African American non-Hispanic population), and the percentage of families below poverty level.

Data Analyses

Preliminary analyses included consideration of missing data, sample design issues, and descriptive statistics (See Supplemental Methods). Addressing aim 1, we conducted latent profile analysis (LPA) of mothers based on the ethnic-racial compositions (Latinx concentration and ethnic-racial diversity) across the four locations in Mplus. Optimal model selection was guided by: lower values of the Akaike information criterion (AIC), Bayesian information criterion (BIC), and sample-size adjusted BIC (SABIC); the Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test (VLMR-LRT) and Bootstrapped likelihood ratio test (BLRT; whereby p-values above .05 supported solutions with one fewer profile; Lo et al., 2001); and entropy values > .80 (Asparouhov & Muthén, 2014). Interpretability of profiles and adequate sample size in each group also informed the final decision (Morin et al., 2020). After converging on an optimal LPA model, we utilized the BCH approach (Bolck et al., 2004) to establish profile membership probabilities, and describe profiles. We described mothers’ activity space profiles vis-à-vis additional location demographics (e.g., % White, % Black, % family below poverty) and family demographics (e.g., mother education, nativity, husband presence, survey language; adolescent ethnic origin). Family demographics differences across profiles were tested using pairwise comparisons with Mplus MODEL CONSTRAINT, adopting a significance level of p < .01 to address the potential for Type I errors due to multiple comparisons (Bender & Lange, 2001).

Addressing aim 2, we examined the effects of mothers’ activity space profiles (W5) on cultural socialization or preparation for bias (W6) using BCH (McLarnon & O’Neill, 2018). Specifically, we examined the differences in the intercept of the ethnic-racial socialization variables between the “target” profile and the reference profile. Mother education was included as a covariate, given education and related socioeconomic factors can influence families’ access to residential neighborhoods and other places (Browning et al., 2017). We controlled for cultural socialization or preparation for bias at W5, offering a prospective test.

Results

The final analytic sample included n = 516 participants (see Supplementary Methods). We compared 1- to 6-profile solutions (Table 1). Across the 1- to 5-profile solutions, AIC, BIC, and SABIC values consistently decreased. Despite both the 4- and 5-profile solutions showing significant BLRT and VLMR-LRT (p < .001), we selected the 4-profile solution. First, the improvement in fit indices from the 4- to 5-profile solution was less substantial compared to the improvements seen between the 3- and 4-profile solutions, and there was only a minimal improvement in entropy (Δentropy = 0.003). Second, the 4-profile solution yielded more informative and interpretable profiles with generally reasonable participant sizes, although it had one smaller profile (n = 57). In contrast, the 5-profile solution introduced two additional smaller profiles (n = 45 and 90), raising concerns about subsequent model convergence and interpretability. Thus, the 4-profile solution optimized model complexity, fit, and interpretability. The characteristics of the four latent profiles are illustrated in Figure 1. Additional demographics within each activity space profile are shown in Table 2.

Table 1.

Model Fit Indices

Profile solution	Log Likelihood	AIC	BIC	ABIC	Entropy	VLMR-LRT p	BLRT p
1 Profile	−3895.79	7823.59	7891.52	7840.74	–	–	–
2 Profiles	−3690.99	7431.97	7538.13	7458.77	0.917	<.001	<.001
3 Profiles	−3546.54	7161.08	7305.45	7197.53	0.885	0.003	<.001
4 Profiles	−3404.70	6895.41	7077.99	6941.50	0.942	<.001	<.001
5 Profiles	−3331.81	6767.63	6988.43	6823.37	0.945	<.001	<.001
6 Profiles	−3302.99	6727.98	6986.99	6793.36	0.854	0.048	<.001

Figure 1.

Latent Profiles of Latinx Concentration and Ethnic-Racial Diversity of Mothers’ Activity Spaces at W5 (N=516)

Notes. Latinx concentration and ethnic-racial diversity were measured on different scales. To facilitate activity space profile plotting, the scale of ethnic-racial diversity was transformed from 0 – 1 to 0 – 100%, aligning it with the scale of Latinx concentration for comparability. Location 1, location 2, and location 3 are referring to mothers’ three frequently visited locations. Residential = residential neighborhood. These profiles comprised 516 out of 547 families, as FIML handled missing data for up to 516 cases. Unstandardized results reported in these figures.

Table 2.

Additional Neighborhood Demographics within Each Activity Space Profile

Profile 1: Integrated, Latinx concentrated (n = 57)
	Profile Indicators		Additional Demographics
	% Latinx	ER Diversity	% White	% Black	% Family below Poverty
Location 1	32.2	0.61	34.51	26.88	11.39
Location 2	36.0	0.64	29.28	27.83	12.81
Location 3	24.7	0.62	36.14	31.71	11.30
Residential	57.8	0.58	18.50	19.62	18.88
Profile 2: Integrated (n = 117)
	Profile Indicators		Additional Demographics
	% Latinx	ER Diversity	% White	% Black	% Family below Poverty
Location 1	17.9	0.62	40.58	33.70	6.52
Location 2	24.9	0.62	32.08	35.59	10.82
Location 3	21.6	0.60	36.55	35.66	11.27
Residential	33.7	0.67	28.36	32.24	12.69
Profile 3: Predominantly White & Black (n = 219)
	Profile Indicators		Additional Demographics
	% Latinx	ER Diversity	% White	% Black	% Family below Poverty
Location 1	14.9	0.53	53.77	23.95	6.62
Location 2	15.5	0.56	50.56	25.64	7.86
Location 3	17.9	0.53	52.48	22.01	8.35
Residential	12.9	0.60	47.16	30.14	6.78
Profile 4: Predominantly White (n = 123)
	Profile Indicators		Additional Demographics
	% Latinx	ER Diversity	% White	% Black	% Family below Poverty
Location 1	10.1	0.48	64.16	16.28	8.49
Location 2	11.4	0.47	65.19	13.08	5.64
Location 3	11.2	0.46	65.36	13.99	6.40
Residential	5.2	0.35	78.60	8.87	2.51

Notes. Additional neighborhood demographics are shown alongside profile indicators to facilitate understanding and labeling of each activity space profile. ER Diversity = ethnic-racial diversity.

The activity space (AS) profiles were labeled based on profile indicators (Latinx concentration, ethnic-racial diversity) and an examination of additional demographics across the four activity space locations (Table 2). The profiles, from profile 1 to profile 4, were ordered from highest to lowest in terms of Latinx concentration and diversity levels. Profile 1, labeled as Integrated, Latinx Concentrated (n = 57; 11%), exhibited high ethnic-racial diversity and the highest overall Latinx concentration. The three frequently visited locations were represented by approximately one-third Latinx, one-third White, and one-third Black populations, indicating a high level of integration among the three groups. The residential neighborhoods predominantly consisted of Latinx residents (57.8%) and had the highest level of families below poverty (18.9%). Profile 2, labeled as Integrated (n = 117; 22.7%), exhibited high ethnic-racial diversity, featuring a high level of integration among Latinx, White, and Black populations across all locations, where no single group held the majority. Profile 3, labeled as Predominantly White & Black (n = 219; 42.4%), exhibited moderate ethnic-racial diversity with notable integration between White and Black populations. Here, White residents constituted about half of the population, Black residents about 25%, and Latinx residents about 15%, indicating a White plurality. Profile 4, labeled as Predominantly White (n = 123; 23.8%), included mothers whose routine locations, particularly their residential neighborhoods, exhibited low Latinx concentration and low ethnic-racial diversity, and were predominated by White residents. The residential neighborhoods had the most White residents (78.6%) and the least families below poverty (2.5%). The frequently visited locations had higher diversity and Latinx concentration, with lower but still majority White populations.

Table 3 displays family demographics for each activity space profile. Families in the Integrated, Latinx Concentrated and Integrated profiles exhibited similar characteristics, including lower education levels, a higher likelihood of completing the survey in Spanish and having a Mexican-origin adolescent, compared to mothers in Predominantly White & Black and Predominantly White profiles. Mothers in Predominantly White profile were more likely to be U.S. born than mothers in all other profiles.

Table 3.

Family Demographics in Each Activity Space Profile

	Mother Education	Survey Language	Husband in Household	Mother Nativity	Adolescent Ethnic Origin
	(Range: 1 to 6)	(1=Spanish, 0=English)	(1=yes, 0=no)	(1 = U.S. born, 0 = non-U.S. born)	(1=Mexican, 0=other Latinx)
Profile 1: Integrated, Latinx concentrated	2.10 a	0.95 a	0.75 a	0.09 a	0.85 a
Profile 2: Integrated	2.41 a	0.94 a	0.89 a	0.11 a	0.80 a
Profile 3: Predominantly White & Black	3.21 b	0.79 b	0.87 a	0.17 a	0.61b
Profile 4: Predominantly White	4.16 c	0.58 c	0.81 a	0.38 b	0.54 b

Notes. Mother education ranges from 1 to 6: 1 = 8^th grade or less, 2 = some high school, 3 = completed high school, 4 = some college, 5 = completed college, 6 = graduate or professional school after college. Within each column of family demographic variables, values sharing superscripts are not significantly different from one another, values that do not share superscripts are significantly different from one another (p < 0.01).

Given that predominantly White neighborhoods serve as the de facto comparison in extant work, we set the Predominantly White profile as the comparison profile to facilitate connection to prior work (Table 4). Compared to the Predominantly White profile, the Integrated, Latinx Concentrated (intercept Δ = 1.49, p = .08), Integrated (intercept Δ = 2.33, p = .001), and Predominantly White & Black (intercept Δ = 1.64, p = .02) profiles prospectively predicted increases in maternal cultural socialization. Compared to the Predominantly White profile, the Integrated profile prospectively predicted increases in maternal preparation for bias (intercept Δ = 1.02, p = .008; Table 4). Profiles did not predict differences in adolescent-reported family-wide cultural socialization.

Table 4.

Comparing Cultural Socialization and Preparation for Bias Levels across Profiles (N=516)

	Integrated, Latinx concentrated vs. Predominantly White			Integrated vs. Predominantly White			Predominantly White & Black vs. Predominantly White
	Estimate	SE	p-value	Estimate	SE	p-value	Estimate	SE	p-value
Family CS (W6) on Profile (W5)	0.46	0.71	0.52	0.03	0.49	0.95	0.67	0.50	0.18
Mom CS (W6) on Profile (W5)	1.49	0.85	0.08	2.33	0.68	0.001	1.64	0.71	0.02
Mom PB (W6) on Profile (W5)	0.47	0.69	0.50	1.02	0.39	0.008	0.39	0.39	0.32

Notes. CS = cultural socialization; PB = preparation for bias; CS/PB (W6) on Profile (W5) = difference in CS or PB levels between the treatment profile and the Predominantly White Profile. Mother education and cultural socialization or preparation for bias at W5 were included as covariates.

Discussion

We employed a combination of activity space and person-centered approaches to characterize the Latinx concentration and ethnic-racial diversity of Latina mothers’ mesosystems (Aim 1; Matthews & Yang, 2013; Witherspoon, White, Bámaca, et al., 2023). These methods offered a holistic view of the varied ethnic-racial exposures encountered by different groups of Latina mothers in their daily activities. Additionally, given the importance of ethnic-racial socialization in the racially stratified U.S. (García Coll et al., 1996), we examined how the ethnic-racial compositions of Latina mothers’ mesosystems influenced ethnic-racial socialization of adolescents (Aim 2). We found that mothers’ engagement in ethnic-racial socialization varied as a function of their earlier mesosystem exposures to ethnicity and race. This study moved beyond the examination of family processes within singular microsystems by examining ethnic-racial socialization within Latina mothers’ mesosystems.

Profiles of Latina Mothers’ Activity Spaces

To our knowledge, this study is the first to characterize Latina mothers’ mesosystems based on ethnic and racial compositions. Previous mesosystemic research that used activity space approaches was not specifically focused on Latinx families and often used a variable-centered approach, which examined (weighted) average characteristics across activity locations (e.g., Browning et al., 2017; Freisthler et al., 2016). Using a person-centered approach, the current study identified four profiles of Latina mothers with distinct ethnic-racial compositional patterns across their day-to-day lives.

The integrated, Latinx concentrated AS comprises mothers residing in predominately Latinx neighborhoods with higher poverty. In their daily activities, these mothers accessed ethnic-racially diverse locations outside their residential neighborhoods, characterized by integration among Latinx, White, and Black populations. The integrated AS comprises mothers residing in and accessing ethnic-racially diverse locations with integrated Latinx, White and Black populations. The predominantly White & Black AS comprises mothers residing in and accessing locations characterized by scant majority White and substantial Black populations. The predominantly White AS comprises mothers residing in and accessing locations with substantial majority White populations and lower ethnic-racial diversity. The residential neighborhoods in the predominantly White AS have the highest proportion of White residents and the lowest percentage of families below the poverty line, potentially indicative of segregated White neighborhoods (Beaulieu & Continelli, 2011).

Across the four AS profiles, profiles with higher Latinx concentration generally also had higher ethnic-racial diversity. For instance, in the integrated AS, the residential neighborhoods displayed average Latinx concentration at 33.7%, the peak across all four routine locations in this profile. This high average Latinx concentration co-occurred with the highest average ethnic-racial diversity level (0.67; Table 2) in this sample. This suggests that a growing Latinx presence in an emerging immigrant area like Atlanta, is likely enhancing ethnic-racial diversity in local communities (Holloway et al., 2012; J. A. Jones, 2019). In contrast, in established Latinx gateways, high scores on Latinx concentration are often indicative of a Latinx majority and correspond with low ethnic-racial diversity (Witherspoon, White, Nair, et al., 2023). Thus, consistent with prior recommendations (White et al., 2018, 2021), it is critical for researchers to consider the context-specific underlying ranges of each ethnic-racial group when theorizing neighborhood or activity space effects using ethnic-racial concentration and diversity variables.

This study sheds light on the residential patterns of Latinx families in an emerging immigrant gateway, offering a comparison to established gateways where historical trends point towards concentrated settlement in predominantly Latinx neighborhoods (Jones & Pebley, 2014). The Atlanta metropolitan area, as an emerging immigrant gateway (Singer, 2015), presents a context where Latinx families reside amidst the significant presence of both White and Black/African American populations (Census Reporter, 2022). This contrasts with some established Latinx areas, such as Phoenix, where the examination of neighborhood ethnic-racial composition is limited by the binary of low Latinx concentration equating to high White concentration or vice versa, without a substantially sized third group (White et al., 2018). In Atlanta, we identified the heterogeneity in the settlement landscape of Latinx families. Specifically, Latinx families resided in neighborhoods with predominantly Latinx, predominantly White, and integrated Latinx, White, and Black populations. It is noteworthy that Latina mothers who settled in predominantly Latinx residential neighborhoods comprise the numerically smallest profile in this study, which contrasts with the settlement patterns in established Latinx areas (M. Jones & Pebley, 2014). However, the absence of Latinx families settling in predominantly Black neighborhoods, despite the prevalence of such neighborhoods in Atlanta (Holloway et al., 2012), warrants further investigation. Future research can explore factors determining residential choices of Latinx families in neighborhoods with particular ethnic-racial compositions.

With respect to White concentration levels in mothers’ activity spaces, we found that mothers in predominantly White & Black AS and in predominantly White AS (who lived in scant or substantial majority White neighborhoods, respectively) tended to travel to other majority White locations in their day-to-day activities. Conversely, mothers in integrated, Latinx concentrated AS and in integrated AS (who lived in predominantly Latinx or integrated Latinx-White-Black neighborhoods, respectively) both tended to travel to integrated locations. This finding extends prior work showing that residents of predominantly Latinx neighborhoods were less likely to travel to predominantly White neighborhoods compared to residents of predominantly White neighborhoods (Q. Wang et al., 2018). The Wang et al. (2018) study, however, was not able to account for the race and ethnicity of the study participants. The current study employed an ethnically homogenous research design and found similar activity space patterns among Latina mothers in an emerging immigrant area. In terms of White concentration, this finding aligns with research showing that residential neighborhoods substantially shaped frequently visited location exposures (Browning et al., 2017; Browning & Soller, 2014), and highlighted a division in Latina mothers’ activity spaces between majority White locations and non-majority White locations.

With respect to ethnic-racial diversity levels, we found that some mothers traveled to locations with different diversity levels from their residential neighborhoods. Both Latina mothers in integrated, Latinx concentrated AS (those living in predominantly Latinx neighborhoods) and in predominantly White AS (those living in predominantly White neighborhoods) traveled to locations with more ethnic-racial diversity compared to their residential areas. For mothers living in predominantly Latinx neighborhoods, traveling to more diverse locations coincided with encountering less poverty. Similarly, prior work found that some African American families in Detroit sought out resources in affluent White suburbs (Winkler, 2012). Thus, residential segregation and uneven resource distribution for mothers living in predominantly Latinx neighborhoods in Atlanta may prompt Latina mothers to seek essential resources elsewhere. For mothers residing in predominantly White neighborhoods, traveling to more diverse locations coincided with encountering higher poverty, suggesting that alternative motivations beyond economic needs might drive these patterns. Common reasons mentioned by mothers for visiting these locations (e.g., shopping, church, and parks) might reflect desires for specific offerings: diverse groceries, co-ethnic churches, parks hosting community activities. This finding aligns with research on African American families in predominantly White neighborhoods who seek out spaces with a stronger Black presence (e.g., Black churches) for cultural affirmation and shared history (Winkler, 2012). Further research can explore mothers’ specific reasons for seeking locations with higher diversity and poverty, such as seeking social connections and cultural experiences. The nuanced nature of these findings stems from the study’s multifaceted approach to assessing ethnic-racial compositions, examining both concentration and diversity.

Activity Spaces and Ethnic-Racial Socialization

Building on research examining ethnic-racial socialization in a single microsystem (e.g., residential neighborhood; Y. Wang et al., 2022; Witherspoon et al., 2021), this study explored how various levels of Latinx concentration and ethnic-racial diversity within mothers’ mesosystems predicted variations in ethnic-racial socialization practices. Compared to mothers accessing Predominantly White AS and controlling for mother’s education and earlier levels of ethnic-racial socialization, mothers in the other three profiles reported greater increases in cultural socialization during middle adolescence. Extending prior findings that higher ethnic-racial diversity in extra-neighborhood locations increased familial cultural socialization (Y. Wang et al., 2022), Latina mothers accessing ethnically and racially integrated mesosystems where no single group holds a majority (Hall & Stringfield, 2014; J. A. Jones, 2019) may feel more comfortable and motivated to transmit cultural heritage to their children. Comparatively, mothers who accessed predominantly White AS engaged in less cultural socialization, possibly because these mothers did not perceive a supportive mesosystem for engaging in cultural socialization (Barr & Neville, 2013). Alternatively, these mothers may prioritize assimilation, seeking greater alignment with the predominantly White mesosystems (Feagin & Cobas, 2008).

Compared to mothers accessing Predominantly White AS, mothers accessing Integrated AS reported greater increases in preparation for bias. This suggests that consistent exposure to ethnic-racial diversity in mothers’ activity spaces, characterized by substantial and approximately equal presence of Latinx, White, and Black community members, could lead Latina mothers to prioritize preparing their children for navigating intergroup relations through parental socialization (Y. Wang et al., 2022; Witherspoon et al., 2021). Unlike predominantly White AS, the integrated AS was characterized by substantial presence of both Black and Latinx communities. This may provide Latina mothers with opportunities for social learning (Ogbu, 1981). Black communities, with their long history of navigating racial marginalization in the U.S., embodying Black joy (Stern et al., 2024), and generations of racial socialization strategies (Neblett et al., 2009; Stevenson & Arrington, 2009), could serve as role models for Latina mothers. Observing Black families’ coping strategies in their mesosystems could contribute to Latina mothers’ increased use of preparation for bias. Mothers who accessed Integrated, Latinx Concentrated AS and Predominantly White & Black AS did not differ from mothers who accessed Predominantly White AS in terms of preparation for bias. This may be due to the more limited presence of Black communities in those mesosystems who can serve as role models for Latina mothers.

In line with previous research, mothers reported on their own cultural socialization (e.g., Pasco et al., 2021; Umaña-Taylor et al., 2014) and adolescents reported on family-wide cultural socialization (e.g., Umaña-Taylor & Guimond, 2010). Unlike mother-reported cultural socialization, adolescents’ reports of family-wide cultural socialization did not vary by mothers’ activity spaces. This discrepancy suggests that ethnic-racial compositions in mothers’ mesosystems might not be the sole or the strongest influence on family-wide cultural socialization efforts. As shown in prior neighborhood research, ethnic-racial dynamics in residential neighborhoods influenced the cultural socialization practices of U.S. Mexican mothers, but not fathers (Pasco et al., 2021). The current and previous studies collectively suggested that some caregivers’ cultural socialization might be more sensitive to ecological exposures than others. Future research could benefit by examining the mesosystemic exposures of other family members, alongside mothers’, to gain a more comprehensive understanding of family members’ mesosystemic influences on cultural socialization practiced within the family.

Limitations, Future Directions, and Conclusions

This study demonstrates one method of capturing mesosystemic exposures by integrating activity space and person-centered frameworks. Building on this study, future research could consider additional sociodemographic characteristics within different family members’ mesosystems, the amount of time spent in different locations, and how shared and unshared mesosystems of different family members influence proximal processes and child development (see Witherspoon, White, Bámaca, et al., 2023 for recommendations). Additionally, the maternal cultural socialization difference observed in the Integrated, Latinx Concentrated vs. Predominantly White AS comparison was marginally significant, potentially due to the smaller sample size of the Integrated, Latinx Concentrated AS profile (n = 57). To better understand cultural socialization within these activity spaces, future research should replicate this study with a larger sample.

This study contributes to the growing body of research on integrating activity space frameworks into cultural-developmental research (Witherspoon, White, Bámaca, et al., 2023) by examining how Latina mothers’ ethnic-racial exposures at the mesosystem level shape ethnic-racial socialization. This study exemplifies a novel method of using activity spaces as a methodological tool for capturing mothers’ mesosystemic exposures (Matthews & Yang, 2013; Witherspoon, White, Bámaca, et al., 2023). Building upon prior research that examined ethnic-racial concentration and ethnic-racial diversity separately within a single microsystem (e.g., Y. Wang et al., 2022; Witherspoon et al., 2021), this study examined both constructs in mesosystems. By focusing on an emerging immigrant gateway with significant presence of White and Black/African American populations (Census Reporter, 2022), this study captured diverse residential and day-to-day activity patterns of Latina mothers in geographic locations with distinct ethnic-racial characteristics.

Public Significance Statement:

Groups of Latina mothers encountered varied ethnic-racial exposures in their day-to-day lives. These exposures shaped the ways mothers socialized their children about culture, ethnicity, and race. The findings can inform programs and policies aimed at supporting Latinx families in diverse places.