Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

One of these things is not like the other: Predictors of core and capital mentoring in adolescence

One of these things is not like the other: Predictors of core and capital mentoring in adolescence INTRODUCTIONMentors can be a critical asset in supporting young people as they transition from adolescence to adulthood. Through informal and formal relationships, mentors provide a wide range of social supports, network on their behalf, and ultimately help protégés reach their goals (Allen & Eby, 2011). Informal mentoring relationships occur naturally within one's social network, while formal mentoring occurs through intentional effort, often within organized programs and initiatives (Inzer & Crawford, 2005). Informal mentoring has wide‐ranging benefits, including those psychosocial, socioemotional, educational, and vocational in nature (Van Dam et al., 2018). Psychosocial outcomes associated with mentoring include lower stress levels, higher life satisfaction, and lower rates of depression (Chang et al., 2010; DuBois & Silverthorn, 2005a, 2005b; Munson & McMillen, 2009). Socioemotional outcomes include an increase in interpersonal skills, perceived social support, and higher self‐esteem (Miranda‐Chan et al., 2016; Van Dam et al., 2018). Youth who receive mentorship are more likely to feel connected to their school (Black et al., 2010), have better grades (Chang et al., 2010), attend college (DuBois & Silverthorn, 2005a, 2005b; Reynolds & Parrish, 2018), receive a bachelor's degree (Erickson et al., 2009; Miranda‐Chan et al., 2016) and have higher access to employment after graduation (DuBois & Silverthorn, 2005a, 2005b).Researchers have long since acknowledged that not all mentors are equal and some can be more impactful on certain desired outcomes than others. Hurd and Sellers (2013) established that there are two types of informal mentoring relationships, naming them “more connected” versus “less connected,” based on how long the pair had known each other, how often they saw or communicated with each other, and how close they felt to one another. Liao and Sánchez (2019) then built on that two‐category distinction, adding that the more connected relationships were also more growth‐oriented. Recent work has continued to build out these two types of informal mentors, naming them core mentors and capital mentors. These two types of mentors come from different domains of a youth's life and may provide different types of support. Core mentoring relationships provide emotional support and are typically with a mentor from within the extended family and most‐immediate social circle (Gowdy & Spencer, 2021). On the other hand, capital mentors provide advice and guidance and are from outside of the young person's family unit (Gowdy & Spencer, 2021). Although previous studies have established the validity of these typologies (Gowdy & Spencer, 2021) for both general populations of young people and those specifically linked to the foster care system (Gowdy & Hogan, 2021), studies have not yet begun to explore which qualities of young people (e.g., demographics) or their environment (e.g., neighborhood resources) dictate what type of informal mentorship they receive.Potential predictors of core and capital mentoringStanding on the well‐established and wide‐ranging benefits of having an informal mentor (see Van Dam et al., 2018), there are growing efforts to build support for mentoring relationships (Schwartz et al., 2018; Spencer et al., 2018). Often, interventions seek to scaffold informal mentoring relationships for the ultimate goal of either educational/vocational support (see Schwartz et al., 2018) or psychosocial/emotional support (see Spencer et al., 2018). Gowdy and Spencer (2021) used cluster analysis to categorize informal mentoring relationships into either core or capital in nature, using a longitudinal nationally representative dataset. They found that those who reported a capital mentor in adolescence (often a teacher or workplace‐based mentor who had not known the young person for as long, but provided informational support and bridging capital) were likelier to be economically mobile in adulthood. They hypothesize that while the supports provisioned by capital mentors are linked to educational, vocational, and ultimately economic outcomes, core mentorship likely supports psychosocial outcomes. Therefore, it is important for us to have a better understanding of which young people are likelier to have core versus capital mentoring so practitioners can better build interventions supporting these different targeted outcomes.Previous literature suggests that one potential predictor of the type of mentoring a young person receives may be linked to the nature of the young person's social network. Broad social networks made up of people from multiple contexts can provide a young person with a wider array of capital (Coleman, 1988). Access to a broader network provides opportunities for capital mentoring via connections with adults who have unique knowledge, skills, and relationships. Those with more homogenous social networks are likelier to have a core mentor, someone from within their network already, as these networks yield fewer opportunities to connect to new resources one would not have had access to before. Those with more heterogenous networks, then, are likelier to have a capital mentor.Previous studies have demonstrated both that young people of Color are less likely to have an informal mentor (Fruiht et al., 2022) and that communities of Color tend to have more homogeneous social networks. It is likely, then, that when young people of Color do report informal mentorship, the mentors are likely “core” in nature, existing in the young person's social network already. Given the propensity for intergeneration familial and fictive‐kin relationships among communities of Color, and that prior research clearly demonstrates that Black and Latino/a youth are more likely to be mentored by a close relative (Hurd & Sellers, 2013; Liao & Sánchez, 2019; Wittrup et al., 2019), analyses utilizing a predominantly White sample may not detect the unique demographic and environmental characteristics of the youth of Color who receive core versus capital mentoring unless analyses are conducted separately for different racial/ethnic groups.Looking beyond race and ethnicity, it is also important to consider the makeup of a young person's social network in terms of its breadth and diversity as it relates to core versus capital mentorship. The networks of support, not just individual mentors, that surround a young person provide them access to greater social capital (Higgins & Kram, 2001), as well as provide connections and access to a wider network. Because young people tend to access social networks primarily through their parents (White & Glick, 2000), and because low‐income communities also have more homogenous social networks (Putnam, 2015), indicators of socioeconomic status such as parental education and parental employment also predict more heterogenous networks (Erola et al., 2016; McDonald & Lambert, 2014). Similarly, it is valuable to consider other indicators of parental social capital, including a parent meeting their child's best friend or talking to their neighbor about a child in trouble (Ashtiani & Feliciano, 2018) in predicting the nature of the mentoring they receive.Young people also access resources and relationships through schools and neighborhoods (White & Glick, 2000), so qualities like prosocial behavior and positive social experiences at school will also be important to consider, as previous literature has established that they are associated with informal mentorship (Gowdy et al., 2019; Hagler, 2017). Neighborhood socioeconomic status (e.g., local employment rate, poverty rate) and neighborhood social capital (e.g., neighbors looking out for each other) could be important in which type of mentor a young person has, as they have been linked to both indications of network heterogeneity and access to mentorship in the past (Erickson et al., 2009; Putnam, 2015). Similarly, a young people's involvement in the community, such as attendance in religious services, could lead to capital mentoring, as these activities often diversify their social network (Abelev, 2009). Finally, qualities of the young person's personal resources such as GPA should be included, as they have been associated with having an informal mentor (Erickson et al., 2009).Present studyThis study is one of the first to examine the emerging typology of core versus capital mentorship and is the first published study using this typology to focus specifically on demographic characteristics and the qualities of a young person's environment that are predictive of the type of mentorship they receive. While there are a growing number of studies focus on predictors of who receives mentorship broadly (see Fruiht et al., 2022; Gowdy et al., 2019; Hagler, 2017) the present study builds upon past work by utilizing the breadth of variables and indicators captured in prior analyses to understand differences in access to mentoring specifically to parse the differences in core versus capital mentoring. Furthermore, results will build upon the work of Fruiht et al. (2022) by widening the scope of dependent variables utilized in these models utilizing the more complete restricted access Add Health dataset. Because racial and ethnic differences often intersect with other demographic and social characteristics including SES, neighborhood characteristics, and the nature of social support systems, we parse these findings separately by racial/ethnic group to understand differences in the qualities that predict core versus capital mentoring not only in the larger population but within different racial/ethnic groups. In sum, this study will thus make an important contribution to the literature by asking the following research question: Which qualities of a young person (i.e., demographics and personal resources) and their environment (i.e., parent‐level, peer‐level, school‐level, and neighborhood‐level) are associated with having a core versus capital mentor?METHODSDataFor the present study, we used the National Longitudinal Study of Adolescent Health (Add Health). Add Health is a multiwave longitudinal, nationally representative study of youth who have been followed from adolescence through to adulthood. The Add Health respondents were originally recruited in the 1994–1995 academic year and followed over time. Our study used two waves of data from this study, Waves 1 and 3. The first wave of data collection focused on the youth's socioeconomic status, social capital, and other related variables. This wave collected when most respondents were between 11 and 19 years old (n = 20,745 youth). This study also used information from the third wave of in‐home interview data, namely questions on informal mentoring, which were collected in 2001 and 2002 when the youth (N = 15,197) were 18–26 years old. Samples used for analyses, further described below, only included those who reported a mentor in Wave 3, had nonmissing information on all mentor‐related Wave 3 questions, and had nonmissing information on the dependent variable.There are two samples used throughout the study, one for regression‐based analyses and another for Conditional Inference Tree (C‐Tree) plots (both analytic approaches are described in further detail below). The final sample size for regression‐based analyses is 2294 participants while the sample size for our C‐Tree plots is 4226. The regressions' sample size represents every observation that has nonmissing information on all predicting variables, described below, in addition to nonmissing information on the independent and dependent variables of interest. C‐Trees analyses, conversely, do not require nonmissing information on predicting variables. Thus, we could include a higher number of observations for this set of analyses.VariablesPredictor variablesThere are several variables to consider when hypothesizing which qualities of a young person are associated with core versus capital mentorship. The present study included variables based on conceptual relevance and previous research, all from Wave 1 of data collection (see Gowdy et al., 2021). In addition to demographic variables under consideration (racial‐ethnic status, age, and sex of young person), all included matching variables are organized into five types of resources: parental, peer, school, neighborhood, and personal. Regarding parental resources, both indications of socioeconomic status (e.g., parental education, household income) were considered in addition to indications of parents' social capital (e.g., parents' involvement in the community, meeting their children's friends.) Some parental resource questions were asked of the young person (e.g., do you feel your parent cares about you?) while some were asked of the parent directly (e.g., how many of your child's friends have you met?).Indications of peer resources include the number of friends young people have and how often they see them, all directly to the young person. School resources include youth‐rated connectedness to school and impressions of their teachers. Neighborhood‐level resources include indications of socioeconomic health of the community (e.g., proportion of community living under poverty, proportion of community with a college education) and indications of community‐level social capital (e.g., if neighbors “look out” for one another, low rates of residential mobility). While some neighborhood resource questions were asked directly to the young person (e.g., do you talk with your neighbors?), others were calculated by linking responses to the Census tract they were reported in (e.g., local poverty rate). Finally, there are two indications of personal resources: GPA and Picture Vocabulary score. The GPA was calculated from grades in Math, History, English, and Science the year before Wave 1 data collection, while the Add Health Picture Vocabulary Test Raw Score stems from an abridged version of the Peabody Picture Vocabulary Test. Originally developed in 1959, the Peabody Picture Vocabulary test is a widely‐used proxy for verbal IQ and involves showing a participant four pictures and a word, then asking the participant to identify the picture most closely associated with the word (Hoffman et al., 2012; Krasileva et al., 2017). This vocabulary test is the only operationalization of verbal IQ in the dataset.Dependent variableInformal mentorship is captured in this study by using the following retrospective question from Wave 3 of the Add Health data: “Other than your parents or step‐parents, has an adult made an important positive difference in your life at any time since you were 14 years old?” Respondents were then asked “How is this person related to you?” and given response options like “family,” “teacher/counselor,” “friend's parent,” “neighbor,” and “religious leader.” Following the guidelines of previous research (see Miranda‐Chan et al., 2016), we did not count spouses, partners, siblings, peers, or coworkers as informal mentors, and recoded the respondents that indicated such as not having an informal mentor.In accordance to previous studies, we ran a two‐step cluster analysis (Norusis, 2008) based on eight variables in Wave 3 of the Add Health data characterizing the mentoring relationship: mentor role, how youth met mentor, two indicators of relationship duration, two indicators of frequency of contact, two indicators of youth‐rated closeness to the mentor, and support provided by the mentor. This latter variable was derived from open‐ended responses in Add Health and coded for a previous study (Gowdy et al., 2020). Because this previous study then examined the association between mentor type and a Wave 4 dependent variable, only observations who had nonmissing information on this Wave 4 variable were coded up for this typology. The two clusters produced from this process were validated with three criteria: each cluster held at least 5% of all observations, clusters made conceptual sense, and there was distinct differentiations among the clusters (Liao & Sánchez, 2019). This process produced our dependent variable for this study: a binary variable indicating either core or capital mentorship.Both of our analytic samples have 45% of the observations categorized as a core mentor and 55% of the observations categorized as capital mentoring. Akin to previous research, the majority of core mentors are from within the extended family, while the majority of capital mentors were school personnel (Spencer et al., 2019). Core mentors had known their mentees for longer than capital mentors had, and see or speak to them more often. Young people in core mentoring relationships also report feelings closer to their mentor and are more likely to report that their mentor was still important at time of data collection. While core mentors were more likely to provide emotional support and instrumental support, capital mentors were more likely to provide valued advice.AnalysisWe used two forms of analysis for this study. The first was a series of logistic regressions predicting core or capital mentorship, first on our whole sample, then segmented by racial‐ethnic status. This analytic approach was chosen as recent work has underscored the fact that the relationship between the qualities of a young person and their likelihood to have an informal mentor likely varies greatly by their racial‐ethnic status (see Fruiht et al., 2022; Gowdy et al., 2022). The second set of analyses utilized C‐Trees, a nonparametric modeling technique that relies on machine learning to build models that allow complex and infinite interactions between a large number of independent variables. The C‐Tree algorithm tests the global null hypothesis of independence between all predictors and the outcome variable, ultimately created by recursively testing models until the null hypothesis is rejected. Using recursive partitioning, each possible interaction between youth demographic and social factors that were entered into the model was modeled to build a categorization model to describe the characteristics that best predict core versus capital mentoring. This analytic strategy results in a tree‐shaped model with splits, called branches, representing variables that significantly contributed to the categorization of the dependent variable. Continuous variables are split into two ranges by the algorithm during the modeling process, based on their contribution to the model, so branches represent both continuous and categorical variables. Each branch also has an associated p value, representing sample‐specific permutation distributions of the test statistics (Hothorn et al., 2006).Both analytic approaches were used and included here as they provide unique perspectives on our central research question. While logistic regression models are commonly used and a standard approach for this type of question, C‐Tree models add new and contextualized findings with their hierarchal approach. After stepping back to reflect on findings from each analytic approach together, this process can better inform the literature than either of them alone. Models were created to categorize mentorship into core versus capital mentoring both in the whole sample, and subsequently segmented by racial‐ethnic status. C‐Tree analyses are robust to missing data, but cannot handle missing data on the outcome variable, therefore models were produced from 4226 observations that had valid race and mentoring data.FindingsDemographicsOf our 4226 participants, a majority (58%) of the young people with a capital mentor were White non‐Hispanic, while only 46% of those with core mentors were White non‐Hispanic. A majority of young people were female (50% for those with capital mentoring, 58% for those with core mentoring) were female. Young people with capital mentors came from families making an average of $6000 more per year than those with core mentors. While 27% of those with capital mentors had parents with a Bachelor's degree or higher, the same was true for only 24% of young people with core mentorship. A complete list of these demographics for both of our analytic samples can be found in Table 1.1TableDescriptive statistics.Regression sample (n = 2294)C‐Tree sample (n = 4226)Core mentoringCapital mentoringCore mentoringCapital mentoring(n = 1037)(n = 1257)(n = 1911)(n = 2315)Demographics45.20%54.80%45.22%54.78%Age (at Wave 4)41.941.7441.831.7341.861.7141.811.73SexMale42.29%48.56%41.56%49.62%Female57.71%51.44%58.44%50.38%Race and ethnicityWhite non‐Hispanic45.92%57.08%46.43%58.40%Black non‐Hispanic27.12%14.46%28.31%14.66%Asian non‐Hispanic6.28%4.75%3.06%4.61%American Indian non‐Hispanic3.14%6.48%5.67%6.01%Other non‐Hispanic3.14%5.35%4.42%3.65%Hispanic14.40%11.87%12.11%12.67%Parental resourcesTotal household income (Wave 1)32,373.0017,643.0038,129.0015,132.0033,593.0016,985.0039,501.0014,531.00Parent educationLess than high school degree16.39%12.73%9.72%9.24%High school degree or equivalent33.68%31.75%32.41%29.62%Some college30.57%32.41%33.75%34.58%Bachelor's degree or equivalent12.86%14.86%15.44%16.03%More than a college degree6.48%8.25%8.67%10.53%Number of ways parent is involved in community0.720.890.980.920.840.950.880.95Parent/neighbor communication about neighbor's childNo14.89%15.79%15.35%14.50%Yes85.11%84.21%84.65%85.50%Parent/neighbor communication about own childNo26.37%28.22%25.26%25.73%Yes73.63%71.78%74.74%74.27%Parent meeting number of child's friends2.021.852.501.962.181.902.461.97Child feeling parent cares about themStrongly disagree0.31%0.26%0.10%0.15%Disagree0.73%0.56%0.86%0.53%Neutral2.26%2.42%1.81%1.91%Agree9.50%11.32%9.34%10.31%Strongly agree87.20%85.44%87.89%87.10%Parent born in neighborhood where child is livingPeer resourcesNumber of friends3.222.623.222.603.212.623.232.58How often they see friendsNot at all9.26%8.10%8.39%7.56%1–2 times22.96%24.56%23.74%25.50%3–4 times28.56%27.14%30.60%27.33%5 or more times39.23%40.20%37.27%39.62%Child feels friends care about themStrongly disagree0.37%0.39%0.19%0.38%Disagree2.48%1.77%2.10%1.91%Neutral12.50%11.49%11.82%10.99%Agree38.29%42.92%38.13%40.99%Strongly agree46.36%43.44%47.76%45.73%School resourcesChild feels part of schoolStrongly disagree2.84%3.05%2.48%2.52%Disagree7.40%7.36%7.15%6.49%Neutral14.43%14.63%14.59%14.66%Agree47.80%46.54%46.90%46.64%Strongly agree27.52%28.43%28.88%29.69%Child gets along with teachersEvery day1.98%2.26%2.00%2.14%Almost every day5.42%4.09%5.62%4.27%Once a week8.69%9.10%8.67%8.70%Just a few times42.06%44.71%42.33%44.20%Never41.85%39.83%41.37%40.69%Child thinks teachers treat students fairlyStrongly disagree4.19%3.44%3.72%2.98%Disagree14.76%15.02%14.97%14.50%Neutral22.49%22.33%23.16%22.06%Agree41.49%43.62%41.75%43.97%Strongly agree17.07%15.59%16.40%16.49%Neighborhood resourcesKnowing others in neighborhoodYes26.31%29.66%72.93%73.21%No73.69%70.34%27.07%26.79%Talking to others in neighborhoodYes18.66%19.05%81.70%82.29%No81.34%80.95%18.30%17.71%Looking out for others in neighborhoodYes25.70%25.93%75.69%76.56%No74.30%74.07%24.31%23.44%Proportion of community who has lived in same county in 19850.830.120.810.120.830.120.810.12Proportion of community 25+ with college degree0.210.150.250.130.220.140.230.13Proportion of community who is unemployed0.080.060.070.060.080.060.070.05Proportion of community whose household income is below 15k0.220.170.180.150.200.160.180.15Child employed at Wave 1Yes43.07%41.79%59.96%61.15%No56.93%58.53%40.04%38.85%Child attendance of weekly religious servicesNever10.56%10.88%9.53%10.00%Less than once/month20.36%20.18%18.40%19.92%At least once/month but less than once/week23.28%21.57%24.69%22.60%Once/week or more45.80%47.37%47.38%47.48%Child feels adult care about themStrongly disagree0.63%0.39%0.29%0.38%Disagree2.31%1.73%2.29%1.37%Neutral7.09%9.57%6.86%8.24%Agree30.36%34.42%29.27%34.43%Strongly agree59.61%53.90%61.30%55.57%Personal resourcesGPA2.760.752.910.762.820.752.950.75Add Health Picture Vocabulary Test63.8710.267.539.4864.499.9867.879.16Regression analysesOur regression analyses, seen in Table 2, found some significant predictors for capital mentorship in relation to core mentorship, and some changes in how the qualities of a young person impact their mentorship type by racial‐ethnic status. Our most consistent finding across sub‐analyses was that those who score higher on the Picture Vocabulary test were likely to report a capital mentor, in comparison to those who had lower scores. For our full sample, age, biological sex, and racial‐ethnic status mattered, with older respondents, female respondents, and Black non‐Hispanic respondents being more likely to report a core mentor (age: OR = 0.95, p = .08; sex: OR = 0.76, p < .001; racial‐ethnic status: OR = 0.52, p < .001 in comparison to White non‐Hispanic respondents). Some indications of parental social capital mattered as well, with respondents whose parents met a greater number of friends having an increase in the odds of reporting a capital mentor, compared to those whose parents had met fewer friends (OR = 1.05; p = .05). Respondents whose parents are from the same neighborhood arebeing more likely to report a core mentor in comparison to those who were not living where their parents grew up (OR = 0.79; p = .04). In consideration of neighborhood resources, living in a community with a higher proportion of college‐educated adults and lower proportion of residential community were both associated with core mentorship (education: OR = 0.35, p = .01; residential mobility: OR = 0.48, p = .07). Finally, higher GPAs and higher Picture Vocabulary scores were both associated with an increase in the odds of reporting a capital mentor, compared to those with lower GPAs and vocabulary scores (GPA: OR = 1.15, p = .04; Picture Vocabulary: OR = 1.03, p < .001).2TableRegression results.Full sampleWhite respondentsBlack respondentsMultiracial and Other respondentsHispanic respondentsN = 2294n = 1223n = 468n = 317n = 286ORSEORSEORSEORSEORSEAge (at Wave 4)0.95*0.270.91**0.040.970.070.950.081.070.11SexMaleRefRefRefRefRefRefRefRefRefRefFemale0.76***0.070.820.110.790.180.52**0.150.45***0.14Race and EthnicityWhite non‐HispanicRefRefNANANANANANANANABlack non‐Hispanic0.52****0.07NANANANANANANANAOther non‐Hispanic0.940.13NANANANANANANANAHispanic0.940.14NANANANANANANANAParental resourcesParent employedYes1.140.131.030.153.54**1.480.840.311.170.43Total household income10101.010.010.990.011.02**0.01Parent educationLess than high school degreeRefRefRefRefRefRefRefRefRefRefHigh school degree or equivalent0.770.130.60.190.610.280.750.40.950.38Some college0.790.140.630.210.670.310.890.460.960.4Bachelor's degree or equivalent0.760.150.650.230.750.370.750.420.680.39More than a college degree0.790.180.710.280.990.540.770.520.28*0.2Number of ways parent is involved in community0.980.0510.070.860.10.970.130.810.16Parent/neighbor communication about neighbor's childYes1.180.161.230.211.380.631.840.710.620.26Parent/neighbor communication about own childYes0.990.110.930.140.980.311.130.371.590.59Parent meeting number of child's friends1.05*0.0261.030.030.970.061.080.081.24**0.12Child feeling parent cares about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.980.571.251.62.362.720.80.76(empty)(empty)Neutral0.950.990.840.412.352.710.40.65(empty)(empty)Agree0.910.90.830.191.81.891.551.531.240.94Strongly agree0.920.890.860.31.491.460.850.351.870.82Parent born in neighborhood where child is living0.79**0.090.770.120.990.270.940.360.470.29Peer resourcesNumber of friends0.990.020.990.021.030.050.950.050.980.06How often they see friendsNot at allRefRefRefRefRefRefRefRefRefRef1–2 times1.020.191.140.310.780.310.750.420.940.523–4 times0.830.151.090.290.41**0.170.850.480.560.325 or more times1.110.191.350.350.570.221.170.650.960.54Child feels friends care about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.530.50.330.542.291.33(empty)(empty)(empty)(empty)Neutral0.480.430.520.81.330.43(empty)(empty)1.751.68Agree0.490.440.520.811.280.320.940.441.100.83Strongly agree0.430.390.430.672.241.360.680.21.080.35School resourcesChild feels part of schoolStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.980.331.280.580.430.350.650.718.880.65Neutral0.950.31.090.450.550.440.9811.391.01Agree0.910.281.190.470.290.221.041.041.190.86Strongly agree0.920.281.30.520.320.250.760.781.020.07Child gets along with teachersEvery dayRefRefRefRefRefRefRefRefRefRefAlmost every day0.670.240.40.240.780.581.11.141.120.82Once a week0.750.260.450.260.990.681.11.091.340.99Just a few times0.80.260.480.261.160.721.161.051.210.82Never0.750.250.450.251.190.761.211.111.210.82Child thinks teachers treat students fairlyStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree1.220.330.870.354.623.440.940.710.490.49Neutral1.080.290.790.313.092.31.10.80.350.35Agree0.220.321.020.393.772.781.050.750.340.33Strongly agree1.190.330.890.363.282.560.720.550.610.64Neighborhood resourcesKnowing others in neighborhoodYes1.020.121.10.180.770.231.210.390.870.33Talking to others in neighborhoodYes1.080.141.010.171.780.640.970.351.030.46Looking out for others in neighborhoodYes10.110.880.151.990.571.280.420.680.24Proportion of community who has lived in same county in 19850.35***0.140.360.21.231.130.010.020.450.6Proportion of community 25+ with college degree0.48*0.20.34*0.191.871.953.745.480.080.12Proportion of community who is unemployed0.820.872.915.270.020.051.183.815.2417.13Proportion of community whose household income is below 15k1.060.470.660.4811.259.880.771.070.630.94Child employed at Wave 1Yes0.90.850.930.130.870.191.190.330.56*0.17Child attendance of weekly religious servicesNeverRefRefRefRefRefRefRefRefRefRefLess than once/month0.980.171.040.231.030.620.720.41.750.9At least once/month but less than once/week0.940.160.960.220.770.430.80.441.850.98Once/week or more0.980.161.020.210.770.411.240.651.580.76Child feels adult care about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.80.670.550.743.542.10.580.440.450.92Neutral1.351.070.80.992.972.830.260.289.1115.52Agree1.260.980.8112.332.10.280.273.215.09Strongly agree1.020.790.630.771.781.60.260.252.33.66Personal resourcesGPA1.15**0.081.130.111.190.21.060.221.43*0.29Add Health Picture Vocabulary Test1.03****0.011.03****0.011.04****0.011.03*0.021.03**0.02Psuedo R20.06000.04000.10000.11000.1800*p < .10; **p < .05; ***p < .01; ****p < .001.Personal and neighborhood resources mattered for if White non‐Hispanic respondents reported a core or capital mentor, with age and living in an educated neighborhood being associated with core mentorship (age: OR = 0.91, p = .02; educated neighborhood: OR = 0.34, p = .05) and higher scores on the Picture Vocabulary test being associated with capital mentorship (OR = 1.03, p < .001).For our Black respondents, having a parent who was employed in Wave 1 was associated with an increase in the odds of reporting a capital mentor (OR = 3.54, p < .001), in comparison to those whose parents were unemployed in Wave 1. Peer‐level and school‐level resources also mattered for this particular racial‐ethnic group: seeing friends more often was associated with core mentorship (OR = 0.41, p = .03) while thinking teachers treated students fairly was associated with capital mentorship (OR = 4.62, p = .04 for “disagree” response, OR = 3.77, p = .07 for “agree” option, both with “strongly disagree” as the reference group). Several neighborhood‐level resources mattered for Black respondents: those who live in a neighborhood where people look out for one another and those who live in a neighborhood with a higher poverty rate are both more likely to report capital mentors (look out: OR = 1.99, p = .02; poverty: OR = 11.25, p = .01) while those who live in a neighborhood with high unemployment rates are likelier to report core mentors as opposed to capital mentors (OR = 0.02, p = .05). Finally, higher Picture Vocabulary scores were associated with capital mentorship (OR = 1.04, p < .001).Similar to our White non‐Hispanic respondents, our Multiracial and Other respondents had very few associations between their resources and which type of mentor they reported. While women were likelier to report a core mentor than men (OR = 0.52, p = .02), as were those living in areas of low residential mobility (in comparison to those in areas of higher residential mobility) (OR = 0.01, p < .001), those who have a higher Picture Vocabulary score were likelier to report capital mentoring (OR = 1.03, p = .06).In consideration of our Hispanic respondents, women were once again likelier than men to report a core mentor (OR = 0.45, p = .01). Parental socioeconomic status mattered more for this subgroup, as having a parent who was employed was associated with an increase in the odds for capital mentorship (OR = 1.02, p = .05) while having an educated parent was associated with core mentorship (OR = 0.28, p = .08). The more friends of the respondent that the parent had met, the likelier the respondent was to have a capital mentor (OR = 1.24, p = .03). Having a job at Wave 1 was also associated with core mentorship (OR = 0.56, p = .06). Finally, increased scores on both personal resources were associated with capital mentorship (GPA: OR = 1.43, p = .08; Picture Vocabulary: OR = 1.03, p = .04).C‐Tree findingsThe initial C‐Tree modeling using the entire sample resulted in a five‐branch model, driven initially by Picture Vocabulary score (p < .001), moderated by race (p < .001). Among participants with a Picture Vocabulary score of 60 or less, Black non‐Hispanic participants (68.9%) were more likely than participants of other ethnicities (51.8%) to have a core mentor. Similar trends were held among those with higher Picture Vocabulary scores. On this branch, 52.6% of Black non‐Hispanic participants had a core mentor, whereas, among those of other races, the tree split again on Picture Vocabulary score (p = .004), this time at a score of 70. That is, non‐Black participants with a Picture Vocabulary score between 60 and 70, 41.3% had core mentors, but among those with a score above 70, the model continued to split. The third branching point was at the number of friends that a participant's parent had met (p = .043) such that among these non‐Black participants with higher Picture Vocabulary scores, those whose parents had met more than 2 of their friends were the least likely to have a core mentor (30.2%) of any group in the model. Among those whose parents had met 2 or fewer of their friends, age further moderated the relationship (p = .011), such that 32.2% of younger participants had a core mentor, but 44.3% of older participants had a core mentor. See Figure 1. Taken together, these results speak to the contribution of Picture Vocabulary score and race, in particular as predictors of mentor type. Those participants with higher vocabulary scores were significantly less likely to have core mentors, and regardless of score, Black non‐Hispanic participants were more to have a core mentor than their counterparts of other races.1FigureFull sample (N = 4226).To further explore the predictors of mentor type among participants of different racial and ethnic backgrounds, C‐Tree models were created. Models were generally similar in their structure across racial and ethnic groups with Picture Vocabulary score (ps < .003) on all four models. The model for Black non‐Hispanic participants split only once, on Picture Vocabulary score, such that those with scores of 60 or lower were much more likely to have core mentors (67.8%) than those with scores over 60 (52.1%). Among White non‐Hispanic participants, 44.7% of those with a Picture Vocabulary of 70 or below had core mentors, whereas among higher scorers core mentoring was less common both in younger (30.3% core mentors) and older participants (42.3%; p = .035). Among Hispanic participants, the model split first on Picture Vocabulary score, again such that those with a score over 59 were less likely to have a core mentor. However, among lower scorers feeling like people in their neighborhoods looked out for each other predicted their mentor type (p = .012). Participants who agreed that their neighbors looked out for each other were much more likely to have a core mentor (65.6%) than those with less close neighborhoods (40.7%). Similarly, among participants of Multiracial and Other racial and ethnic groups, while Picture Vocabulary drove the initial split such that higher scorers were less likely to have core mentors (36.4%), residential mobility moderated that effect for lower scorers. Participants who lived in neighborhoods where about 76% or less of their neighbors were long‐term residents of the county were much less likely to have core mentors (30.2%) than those with a higher proportion of longer‐term residents (61.1%). Please see Figure 2 for all subsample C‐Tree plots.2FigureC‐Tree plots by subsample.DISCUSSIONAccess to mentoring relationships in adolescence can promote success and well‐being during the transition to adulthood. Both core mentoring from a caring adult committed to promoting socioemotional well‐being, and capital mentoring that generally comes from more distal connections who support academic and vocational development, can be rich assets for young people (Gowdy & Spencer, 2021). The present study aimed to identify the characteristics of youth who access more core versus capital mentoring to best identify areas for support and intervention. Findings demonstrate that participants' scores on a vocabulary test taken in adolescence were across the board the best predictor of capital mentoring. Furthermore, Black participants and females were more likely to receive core mentoring. Analyses investigating the unique predictors of core and capital mentoring among participants of different races shed some light on parental and neighborhood resources that may promote different types of mentoring, generally supporting prior findings that youth with more resources are more likely to report a more distal mentor (Fruiht et al., 2022; Raposa et al., 2018) who provides more academic/vocational support.Across both analytic strategies, Peabody Picture Vocabulary scores were the most consistent predictor of capital mentoring. That is, participants who had a stronger vocabulary in adolescence were more likely to report a capital mentor. Instead of translating this at face value, however, literature tells us to focus on the socioeconomic differences in vocabulary scores. Indeed, differences in vocabulary by socioeconomic status can be seen in children as young as 18 months (Fernald et al., 2012). The vocabulary measure for this study was collected during Wave 1, when participants were between 11 and 19 years old, meaning that the socioeconomic differences among our sample youth have likely shown themselves in differences in vocabulary scores. This then retranslates our most consistent finding as capital mentoring being most available to students of higher socioeconomic status, a finding in alignment with previous research (Gowdy & Spencer, 2021).Race was also a notable predictor across models, with Black non‐Hispanic participants being much more likely to receive to core mentoring. This is very much in line with prior research suggesting that Black youth have strong networks of family and fictive‐kin that provide them support and mentorship (Hurd & Sellers, 2013; Wittrup et al., 2019), as well as neighborhood‐level sociological research that demonstrates benefits and prevalence of intergenerational relationships in Black neighborhoods that depend not on assets coming in from outside of the neighborhood, but on the stability of a neighborhood itself (Sampson, 1999). Furthermore, in line with past findings suggesting that more resourced youth have more access to mentors in their larger communities (Erickson et al., 2009; Fruiht et al., 2022), parental resources predicted capital mentoring. Most consistently, participants whose parents had met more of their friends were more likely to have capital mentors, but more objective indicators such as income and education were also significant predictors of capital mentoring in regression models. Furthermore, gender was a consistent predictor; females were more likely to have a core mentor in these models. However, this finding did not hold in C‐Tree analyses.In addition to looking at predictors of core and capital mentoring in the entire sample, additional analyses subsetted the sample by race to more clearly understand the unique predictors of mentoring among youth in different racial and ethnic groups. Recent research has demonstrated the utility of this strategy to better parse the specific demographic and environmental factors that influence mentorship among youth from different cultural backgrounds and lived experiences. This methodology ensures that variables linked to systemic inequalities (e.g., parental income and education) do not overshadow potentially important trends that highlight the unique resources of people of Color (Fruiht et al., 2022). While findings from analyses of separate racial and ethnic groups were largely consistent with those from the larger sample in that Picture Vocabulary was a consistent predictor of capital mentoring, more nuanced findings about parental and neighborhood characteristics emerged for different racial and ethnic groups. Regression models generally demonstrated that youth with more resourced parents and those from more resourced neighborhoods (more educated, less poverty, low unemployment) were more likely to have capital mentors. One notable exception, however, was that among Hispanic participants, having a more‐educated parent was actually predictive of core mentoring. This finding may speak to the potential for more educated family members, who share a culture and life experiences with a Hispanic adolescent, to serve as mentors and role models, warranting further investigation.Furthermore, the more targeted approach of the C‐Tree analyses highlighted the ways that neighborhood factors support core mentoring relationships among some youth of Color. For example, among Hispanic participants with lower Picture Vocabulary scores, those who felt their neighbors looked out for each other were more likely to report core mentoring. Similarly, among Multiracial and Other race participants with lower Picture Vocabulary scores, lower residential mobility predicted core mentoring. Lower residential mobility and neighbors looking out for each other are likely both indicators of neighborhood‐level social capital. Therefore, these findings may speak to the potential for building close‐knit and emotionally supportive connections within one's community, and the benefit of residential stability in supporting that. Taken together, findings speak to the rich resources available in neighborhoods and the propensity of stable, closer‐knit neighborhoods to translate into mentorship.Although there is important literature establishing what qualities of a young person and their context lead to informal mentoring broadly (see Erickson et al., 2009; Fruiht et al., 2022), there have been no published studies before the present that focuses on predictors of core versus capital mentoring. While both types of mentoring provide supports that benefit a protégé, given the differences in the mentoring functions provided by core and capital mentors it is critical to understand the assets available to different youth. For instance, core mentoring may provide a protégé with socioemotional support and a deeper connection to the community that serves a protective function, particularly for youth of Color as they establish a sense of racial identity (Hurd et al., 2012). Conversely, because that capital mentoring is associated with economic mobility for low‐income youth and youth of Color (Gowdy & Spencer, 2021), it is important that we understand who is likely and unlikely to be in this type of mentoring relationship. Not only do we need to know who is unlikely to have a capital mentor for mobility‐focused intervention development, but researchers also need this knowledge to isolate the impact of mentoring alone, without the potentially confounding variables that predict both capital mentoring and the outcome of interest. Given that our most consistent finding throughout this study is that higher Picture Vocabulary scores predict mentoring, contextualized with our knowledge that these are likely an indicator of socioeconomic differences in early childhood, our findings potentially underscore the true impact capital mentoring can have on upward mobility for low‐income youth.LimitationsThe present study utilized a large nationally representative longitudinal dataset to investigate these factors. While data sets of this nature have many methodological strengths, they also come with limitations that impact the generalizability of our findings. Data collection for this longitudinal study began in 1994–1995 when most participants were in high school. As a result, these analyses capture the mentoring relationships of participants who are now well into adulthood and not the experiences of today's youth. Demographic and social shifts over the past three decades have impacted the experiences of youth of Color in their neighborhoods as well as access to higher education and mentoring in the context of higher education. Therefore, results must be considered through an appropriate historical lens. Furthermore, the Add Health dataset oversampled for highly educated Black families. While the general pattern of findings for Black youth was in line with overall trends, some caution should be used in generalizing findings about Black families, in particular, from these findings.Beyond the historical context of the data, there are issues in measurement. The primary question on informal mentoring is prone to recall bias, as are all of the follow‐up questions used to create the core and capital clusters. In addition, all of these informal mentoring questions are asked about only one person, eliminating our ability to understand how a young person reports more than one informal mentor. The Add Health study asked participants to report being either male or female, limiting our understanding of how gender nonconforming young people experience informal mentorship.Implications and future directionsFindings have significant implications for the development and implementation of mentoring programs for youth. Namely, they illuminate opportunities to supplement the naturally occurring and informal relationships that young people already have with formal mentoring that provides unique additional support. Black youth, for instance, are more likely to access core than capital mentoring across the board. While a stronger vocabulary and more resourced families and neighborhoods may promote access to capital mentors, our findings generally suggest that Black youth have access to stronger networks of family and friends who support them (Hurd & Sellers, 2013; Wittrup et al., 2019). While core mentoring relationships may help the youth of Color develop a sense of pride and identity (Albright et al., 2017; Hurd et al., 2012), Black communities may benefit from mentoring programs that supplement that core support with more capital mentoring to promote economic mobility. Similarly, participants of Multiracial and Other races and ethnicities who live in neighborhoods with low residential mobility may also have access to close, supportive core mentoring relationships. This suggests that neighborhoods with less mobility may be potentially opportune locations to promote capital mentoring, but that such efforts must be complimentary to and supportive of the core mentoring that these youth may already access.As mental health issues become increasingly commonplace among teenagers and young adults (Twenge et al., 2019) and today adolescents and young adults report unprecedented levels of loneliness (Demarinis, 2020), it is also important to consider the potential gaps in core mentoring for some young people. The Add Health interview collected data on just one significant nonparental relationship, therefore we cannot know the extent of the support that young people experience outside of this single mentoring relationship, however, our findings do suggest that some youth may be missing out on the benefits of core mentoring. More resourced communities may put undo pressure on teenagers to excel academically, providing a good deal of academic and career‐oriented support, without ensuring that they have nonparental adults to turn to that support the development of psychological well‐being. Just as access to capital mentoring may be useful to supplement the naturally occurring supports in less‐resourced communities, there may be a need to better support the socioemotional needs of some young people. The gender difference in mentor‐type demonstrated by our regression models may speak to the particular lack of these relationships among adolescent boys.CONCLUSIONAs they move through adolescence and into adulthood, young people learn to balance the competing needs for socioemotional support and instrumental support that promote their future career ambitions. Mentoring from committed adults can help young people master skills and pave the road to a happy and successful adult life, providing different types of support depending on the role they play in the life of a protégé. This study demonstrates some demographic and environmental differences between youth who report support from core versus capital mentors, highlighting that youth of higher socioeconomic status, who experience the cognitive benefits of being more resourced, are more likely to have capital mentors. Given prior findings that capital mentors can help promote economic mobility, these findings highlight the possibilities for the promotion of capital mentoring for less‐resourced youth, while still recognizing the asset of core mentoring from the communities and families in which these youth are embedded.REFERENCESAbelev, M. S. (2009). Advancing out of poverty: Social class worldview and its relation to resilience. Journal of Adolescent Research, 24(1), 114–141. https://doi.org/10.1177/0743558408328441Albright, J. N., Hurd, N. M., & Hussain, S. B. (2017). Applying a social justice lens to youth mentoring: A review of the literature and recommendations for practice. American Journal of Community Psychology, 59, 363–381. https://doi.org/10.1002/ajcp.12143Allen, T. D., & Eby, L. T. (2011). The Blackwell handbook of mentoring: A multiple perspectives approach. John Wiley & Sons.Ashtiani, M., & Feliciano, C. (2018). Access and mobilization: How social capital relates to low‐income youth's postsecondary educational (PSE) attainment. Youth & Society, 50(4), 439–461. https://doi.org/10.1177/0044118x15607163Black, D., Grenard, J., Sussman, S., & Rohrbach, L. (2010). The influence of school‐based natural mentoring relationships on school attachment and subsequent adolescent risk behaviors. Health Education Research, 25(5), 892–902.Chang, E. S., Greenberger, E., Chen, C., Heckhausen, J., & Farruggia, S. P. (2010). Nonparental adults as social resources in the transition to adulthood. Journal of Research on Adolescence, 20(4), 1065–1082. https://doi.org/10.1111/j.1532-7795.2010.00662.xColeman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120.Demarinis, S. (2020). Loneliness at epidemic levels in America. Explore, 16(5), 278–279. https://doi.org/10.1016/j.explore.2020.06.008DuBois, D. L., & Silverthorn, N. (2005a). Natural mentoring relationships and adolescent health: Evidence from a national study. American Journal of Public Health, 95(3), 518–524. https://doi.org/10.2105/ajph.2003.031476DuBois, D. L., & Silverthorn, N. (2005b). Characteristics of natural mentoring relationships and adolescent adjustment: Evidence from a national study. The Journal of Primary Prevention, 26(2), 69–92. https://doi.org/10.1007/s10935-005-1832-4Erickson, L. D., McDonald, S., & Elder, G. H. (2009). Informal mentors and education: Complementary or compensatory resources. Sociology of Education, 82(4), 344–367. https://doi.org/10.1177/003804070908200403Erola, J., Jalonen, S., & Lehti, H. (2016). Parental education, class and income over early life course and children's achievement. Research in Social Stratification and Mobility, 44, 33–43. https://doi.org/10.1016/j.rssm.2016.01.003Fernald, A., Marchman, V. A., & Weisleder, A. (2012). SES differences in language processing skill and vocabulary are evident at 18 months. Developmental Science, 16(2), 234–248. https://doi.org/10.1111/desc.12019Fruiht, V., Boeder, J., & Chan, T. (2022). Rising stars and underdogs: The role race and parental education play in predicting mentorship. Youth & Society, 54(4), 635–661. https://doi.org/10.1177/0044118x211004637Gowdy, G., Miller, D., & Spencer, R. (2019). Expanding and deepening our understanding of which young people have an informal mentor. Children and Youth Services Review, 108. https://doi.org/10.1016/j.childyouth.2019.104490Gowdy, G., Miller, D., & Spencer, R. (2020). Helping those who need it the least: A counterfactual and comparative analysis of whether informal mentoring promotes economic upward mobility for low‐ and middle‐income youth. Youth & Society, 53(7), 1152–1180. https://doi.org/10.1177/0044118X20959241Gowdy, G., Miller, D. P., & Spencer, R. (2021). Helping those who need it the least: A counterfactual and comparative analysis of whether informal mentoring promotes economic upward mobility for low‐ and middle‐income youth. Youth & Society, 53(7), 1152–1180. https://doi.org/10.1177/0044118x20959241Gowdy, G., & Hogan, S. (2021). Informal mentoring among foster youth entering higher education. Children and Youth Services Review, 120, 105716. https://doi.org/10.1016/j.childyouth.2020.105716Gowdy, G., Palmer, M. T., Saastamoinen, M., & Rivera, M. (2022). Using a social work perspective to understand contextual factors impacting access to informal mentorship for under‐resourced and minoritized youth. Child and Adolescent Social Work Journal. https://doi.org/10.1007/s10560-022-00838-4Gowdy, G., & Spencer, R. (2021). It's who you know that matters: Identifying which type of informal mentor is most likely to promote economic mobility for vulnerable youth. The Journal of Primary Prevention, 42(3), 217–238. https://doi.org/10.1007/s10935-021-00630-7Hagler, M. (2017). The influence of parents and natural mentors on young adults' substance use behaviors: Evidence from a national study [Master's thesis]. Western University.Higgins, M. C., & Kram, K. E. (2001). Reconceptualizing mentoring at work: A developmental network perspective. Academy of Management Review, 26(2), 264–288.Hoffman, L., Templin, J., & Rice, M. L. (2012). Linking outcomes from peabody picture vocabulary test forms using item response models. Journal of Speech, Language, and Hearing Research, 55(3), 754–763. https://doi.org/10.1044/1092-4388(2011/10-0216)Hothorn, T., Hornik, K., & Zeileis, A. (2006). Unbiased recursive partitioning: A conditional inference framework. Journal of Computational and Graphical Statistics, 15, 651–674.Hurd, N. M., Sánchez, B., Zimmerman, M. A., & Caldwell, C. H. (2012). Natural mentors, racial identity, and educational attainment among African American adolescents: Exploring pathways to success. Child Development, 83, 1196–1212. https://doi.org/10.1111/j.1467-8624.2012.01769.xHurd, N. M., & Sellers, R. M. (2013). Black adolescents' relationships with natural mentors: Associations with academic engagement via social and emotional development. Cultural Diversity and Ethnic Minority Psychology, 19(1), 76–85. https://doi.org/10.1037/a0031095Inzer, L. D., & Crawford, C. B. (2005). A review of formal and informal mentoring. Journal of Leadership Education, 4(1), 31–50. https://doi.org/10.12806/v4/i1/tf2Krasileva, K. E., Sanders, S. J., & Bal, V. H. (2017). Peabody picture vocabulary test: Proxy for verbal IQ in genetic studies of autism spectrum disorder. Journal of Autism and Developmental Disorders, 47(4), 1073–1085. https://doi.org/10.1007/s10803-017-3030-7Liao, C., & Sánchez, B. (2019). Mentoring relationship quality profiles and their association with urban, low‐income youth's academic outcomes. Youth & Society, 51(4), 443–462. https://doi.org/10.1177/0044118x16668058McDonald, S., & Lambert, J. (2014). The long arm of mentoring: A counterfactual analysis of natural youth mentoring and employment outcomes in early careers. American Journal of Community Psychology, 54(3–4), 262–273. https://doi.org/10.1007/s10464-014-9670-2Miranda‐Chan, T., Fruiht, V., Dubon, V., & Wray‐Lake, L. (2016). The functions and longitudinal outcomes of adolescents' naturally occurring mentorships. American Journal of Community Psychology, 57(1–2), 47–59. https://doi.org/10.1002/ajcp.12031Munson, M. R., & McMillen, J. C. (2009). Natural mentoring and psychosocial outcomes among older youth transitioning from foster care. Children and Youth Services Review, 31(1), 104–111. https://doi.org/10.1016/j.childyouth.2008.06.003Norusis, M. (2008). SPSS 16.0 Statistical Procedures Companion. Prentice Hall.Putnam, R. (2015). Our kids: The American dream in crisis. Simon & Schuster.Raposa, E., Erickson, L. D., Hagler, M., & Rhodes, J. (2018). How economic disadvantage affects the availability and nature of mentoring relationships during the transition to adulthood. American Journal of Community Psychology, 61(1–2).Reynolds, J. R., & Parrish, M. (2018). Natural mentors, social class, and college success. American Journal of Community Psychology, 61(1–2), 179–190. https://doi.org/10.1002/ajcp.12209Sampson, R. J. (1999). What Community Supplies. In R. F. Ferguson, & W. T. Dickens (Eds.), Urban Problems and Community Development (pp. 241–292). Brookings Institution Press.Schwartz, S. E. O., Kanchewa, S. S., Rhodes, J. E., Gowdy, G., Stark, A. M., Horn, J. P., Parnes, M., & Spencer, R. (2018). “I'm having a little struggle with this; can you help me out?”: Examining impacts and processes of a social capital intervention for first‐generation college students. American Journal of Community Psychology, 61(1–2), 166–178. https://doi.org/10.1002/ajcp.12206Spencer, R., Drew, A. L., Gowdy, G., & Horn, J. P. (2018). “A positive guiding hand”: A qualitative examination of youth‐initiated mentoring and the promotion of interdependence among foster care youth. Children and Youth Services Review, 93, 41–50. https://doi.org/10.1016/j.childyouth.2018.06.038Spencer, R., Gowdy, G., Drew, A. L., Abrams, E., & Rhodes, J. E. (2019). “Who knows me the best and can encourage me the most?:” Matching and early relationship development in youth‐initiated mentoring relationships with system‐involved youth. Journal of Adolescent Research. https://doi.org/10.1177/0743558418755686Twenge, J. M., Cooper, A. B., Joiner, T. E., Duffy, M. E., & Binau, S. G. (2019). Age, period, and cohort trends in mood disorder indicators and suicide‐related outcomes in a nationally representative dataset, 2005–2017. Journal of Abnormal Psychology, 128(3), 185–199. https://doi.org/10.1037/abn0000410Van Dam, L., Smit, D., Wildschut, B., Branje, S. J. T., Rhodes, J. E., Assink, M., & Stams, G. J. J. M. (2018). Does natural mentoring matter? A multilevel meta‐analysis on the association between natural mentoring and youth outcomes. American Journal of Community Psychology, 62(1–2), 203–220. https://doi.org/10.1002/ajcp.12248Wittrup, A. R., Hussain, S. B., Albright, J. N., Hurd, N. M., Varner, F. A., & Mattis, J. S. (2019). Natural mentors, racial pride, and academic engagement among black adolescents: Resilience in the context of perceived discrimination. Youth & Society, 51(4), 463–483. https://doi.org/10.1177/0044118X16680546White, M., & Glick, J. (2000). Generation status, social capital, and the routes out of high school. Sociological Forum, 15(4), 671–691. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Community Psychology Wiley

One of these things is not like the other: Predictors of core and capital mentoring in adolescence

Download PDF
17 pages

Loading next page...
 
/lp/wiley/one-of-these-things-is-not-like-the-other-predictors-of-core-and-gkfEcgzq9a

References (45)

Publisher
Wiley
Copyright
© 2023 Society for Community Research and Action
ISSN
0091-0562
eISSN
1573-2770
DOI
10.1002/ajcp.12627
Publisher site
See Article on Publisher Site

Abstract

INTRODUCTIONMentors can be a critical asset in supporting young people as they transition from adolescence to adulthood. Through informal and formal relationships, mentors provide a wide range of social supports, network on their behalf, and ultimately help protégés reach their goals (Allen & Eby, 2011). Informal mentoring relationships occur naturally within one's social network, while formal mentoring occurs through intentional effort, often within organized programs and initiatives (Inzer & Crawford, 2005). Informal mentoring has wide‐ranging benefits, including those psychosocial, socioemotional, educational, and vocational in nature (Van Dam et al., 2018). Psychosocial outcomes associated with mentoring include lower stress levels, higher life satisfaction, and lower rates of depression (Chang et al., 2010; DuBois & Silverthorn, 2005a, 2005b; Munson & McMillen, 2009). Socioemotional outcomes include an increase in interpersonal skills, perceived social support, and higher self‐esteem (Miranda‐Chan et al., 2016; Van Dam et al., 2018). Youth who receive mentorship are more likely to feel connected to their school (Black et al., 2010), have better grades (Chang et al., 2010), attend college (DuBois & Silverthorn, 2005a, 2005b; Reynolds & Parrish, 2018), receive a bachelor's degree (Erickson et al., 2009; Miranda‐Chan et al., 2016) and have higher access to employment after graduation (DuBois & Silverthorn, 2005a, 2005b).Researchers have long since acknowledged that not all mentors are equal and some can be more impactful on certain desired outcomes than others. Hurd and Sellers (2013) established that there are two types of informal mentoring relationships, naming them “more connected” versus “less connected,” based on how long the pair had known each other, how often they saw or communicated with each other, and how close they felt to one another. Liao and Sánchez (2019) then built on that two‐category distinction, adding that the more connected relationships were also more growth‐oriented. Recent work has continued to build out these two types of informal mentors, naming them core mentors and capital mentors. These two types of mentors come from different domains of a youth's life and may provide different types of support. Core mentoring relationships provide emotional support and are typically with a mentor from within the extended family and most‐immediate social circle (Gowdy & Spencer, 2021). On the other hand, capital mentors provide advice and guidance and are from outside of the young person's family unit (Gowdy & Spencer, 2021). Although previous studies have established the validity of these typologies (Gowdy & Spencer, 2021) for both general populations of young people and those specifically linked to the foster care system (Gowdy & Hogan, 2021), studies have not yet begun to explore which qualities of young people (e.g., demographics) or their environment (e.g., neighborhood resources) dictate what type of informal mentorship they receive.Potential predictors of core and capital mentoringStanding on the well‐established and wide‐ranging benefits of having an informal mentor (see Van Dam et al., 2018), there are growing efforts to build support for mentoring relationships (Schwartz et al., 2018; Spencer et al., 2018). Often, interventions seek to scaffold informal mentoring relationships for the ultimate goal of either educational/vocational support (see Schwartz et al., 2018) or psychosocial/emotional support (see Spencer et al., 2018). Gowdy and Spencer (2021) used cluster analysis to categorize informal mentoring relationships into either core or capital in nature, using a longitudinal nationally representative dataset. They found that those who reported a capital mentor in adolescence (often a teacher or workplace‐based mentor who had not known the young person for as long, but provided informational support and bridging capital) were likelier to be economically mobile in adulthood. They hypothesize that while the supports provisioned by capital mentors are linked to educational, vocational, and ultimately economic outcomes, core mentorship likely supports psychosocial outcomes. Therefore, it is important for us to have a better understanding of which young people are likelier to have core versus capital mentoring so practitioners can better build interventions supporting these different targeted outcomes.Previous literature suggests that one potential predictor of the type of mentoring a young person receives may be linked to the nature of the young person's social network. Broad social networks made up of people from multiple contexts can provide a young person with a wider array of capital (Coleman, 1988). Access to a broader network provides opportunities for capital mentoring via connections with adults who have unique knowledge, skills, and relationships. Those with more homogenous social networks are likelier to have a core mentor, someone from within their network already, as these networks yield fewer opportunities to connect to new resources one would not have had access to before. Those with more heterogenous networks, then, are likelier to have a capital mentor.Previous studies have demonstrated both that young people of Color are less likely to have an informal mentor (Fruiht et al., 2022) and that communities of Color tend to have more homogeneous social networks. It is likely, then, that when young people of Color do report informal mentorship, the mentors are likely “core” in nature, existing in the young person's social network already. Given the propensity for intergeneration familial and fictive‐kin relationships among communities of Color, and that prior research clearly demonstrates that Black and Latino/a youth are more likely to be mentored by a close relative (Hurd & Sellers, 2013; Liao & Sánchez, 2019; Wittrup et al., 2019), analyses utilizing a predominantly White sample may not detect the unique demographic and environmental characteristics of the youth of Color who receive core versus capital mentoring unless analyses are conducted separately for different racial/ethnic groups.Looking beyond race and ethnicity, it is also important to consider the makeup of a young person's social network in terms of its breadth and diversity as it relates to core versus capital mentorship. The networks of support, not just individual mentors, that surround a young person provide them access to greater social capital (Higgins & Kram, 2001), as well as provide connections and access to a wider network. Because young people tend to access social networks primarily through their parents (White & Glick, 2000), and because low‐income communities also have more homogenous social networks (Putnam, 2015), indicators of socioeconomic status such as parental education and parental employment also predict more heterogenous networks (Erola et al., 2016; McDonald & Lambert, 2014). Similarly, it is valuable to consider other indicators of parental social capital, including a parent meeting their child's best friend or talking to their neighbor about a child in trouble (Ashtiani & Feliciano, 2018) in predicting the nature of the mentoring they receive.Young people also access resources and relationships through schools and neighborhoods (White & Glick, 2000), so qualities like prosocial behavior and positive social experiences at school will also be important to consider, as previous literature has established that they are associated with informal mentorship (Gowdy et al., 2019; Hagler, 2017). Neighborhood socioeconomic status (e.g., local employment rate, poverty rate) and neighborhood social capital (e.g., neighbors looking out for each other) could be important in which type of mentor a young person has, as they have been linked to both indications of network heterogeneity and access to mentorship in the past (Erickson et al., 2009; Putnam, 2015). Similarly, a young people's involvement in the community, such as attendance in religious services, could lead to capital mentoring, as these activities often diversify their social network (Abelev, 2009). Finally, qualities of the young person's personal resources such as GPA should be included, as they have been associated with having an informal mentor (Erickson et al., 2009).Present studyThis study is one of the first to examine the emerging typology of core versus capital mentorship and is the first published study using this typology to focus specifically on demographic characteristics and the qualities of a young person's environment that are predictive of the type of mentorship they receive. While there are a growing number of studies focus on predictors of who receives mentorship broadly (see Fruiht et al., 2022; Gowdy et al., 2019; Hagler, 2017) the present study builds upon past work by utilizing the breadth of variables and indicators captured in prior analyses to understand differences in access to mentoring specifically to parse the differences in core versus capital mentoring. Furthermore, results will build upon the work of Fruiht et al. (2022) by widening the scope of dependent variables utilized in these models utilizing the more complete restricted access Add Health dataset. Because racial and ethnic differences often intersect with other demographic and social characteristics including SES, neighborhood characteristics, and the nature of social support systems, we parse these findings separately by racial/ethnic group to understand differences in the qualities that predict core versus capital mentoring not only in the larger population but within different racial/ethnic groups. In sum, this study will thus make an important contribution to the literature by asking the following research question: Which qualities of a young person (i.e., demographics and personal resources) and their environment (i.e., parent‐level, peer‐level, school‐level, and neighborhood‐level) are associated with having a core versus capital mentor?METHODSDataFor the present study, we used the National Longitudinal Study of Adolescent Health (Add Health). Add Health is a multiwave longitudinal, nationally representative study of youth who have been followed from adolescence through to adulthood. The Add Health respondents were originally recruited in the 1994–1995 academic year and followed over time. Our study used two waves of data from this study, Waves 1 and 3. The first wave of data collection focused on the youth's socioeconomic status, social capital, and other related variables. This wave collected when most respondents were between 11 and 19 years old (n = 20,745 youth). This study also used information from the third wave of in‐home interview data, namely questions on informal mentoring, which were collected in 2001 and 2002 when the youth (N = 15,197) were 18–26 years old. Samples used for analyses, further described below, only included those who reported a mentor in Wave 3, had nonmissing information on all mentor‐related Wave 3 questions, and had nonmissing information on the dependent variable.There are two samples used throughout the study, one for regression‐based analyses and another for Conditional Inference Tree (C‐Tree) plots (both analytic approaches are described in further detail below). The final sample size for regression‐based analyses is 2294 participants while the sample size for our C‐Tree plots is 4226. The regressions' sample size represents every observation that has nonmissing information on all predicting variables, described below, in addition to nonmissing information on the independent and dependent variables of interest. C‐Trees analyses, conversely, do not require nonmissing information on predicting variables. Thus, we could include a higher number of observations for this set of analyses.VariablesPredictor variablesThere are several variables to consider when hypothesizing which qualities of a young person are associated with core versus capital mentorship. The present study included variables based on conceptual relevance and previous research, all from Wave 1 of data collection (see Gowdy et al., 2021). In addition to demographic variables under consideration (racial‐ethnic status, age, and sex of young person), all included matching variables are organized into five types of resources: parental, peer, school, neighborhood, and personal. Regarding parental resources, both indications of socioeconomic status (e.g., parental education, household income) were considered in addition to indications of parents' social capital (e.g., parents' involvement in the community, meeting their children's friends.) Some parental resource questions were asked of the young person (e.g., do you feel your parent cares about you?) while some were asked of the parent directly (e.g., how many of your child's friends have you met?).Indications of peer resources include the number of friends young people have and how often they see them, all directly to the young person. School resources include youth‐rated connectedness to school and impressions of their teachers. Neighborhood‐level resources include indications of socioeconomic health of the community (e.g., proportion of community living under poverty, proportion of community with a college education) and indications of community‐level social capital (e.g., if neighbors “look out” for one another, low rates of residential mobility). While some neighborhood resource questions were asked directly to the young person (e.g., do you talk with your neighbors?), others were calculated by linking responses to the Census tract they were reported in (e.g., local poverty rate). Finally, there are two indications of personal resources: GPA and Picture Vocabulary score. The GPA was calculated from grades in Math, History, English, and Science the year before Wave 1 data collection, while the Add Health Picture Vocabulary Test Raw Score stems from an abridged version of the Peabody Picture Vocabulary Test. Originally developed in 1959, the Peabody Picture Vocabulary test is a widely‐used proxy for verbal IQ and involves showing a participant four pictures and a word, then asking the participant to identify the picture most closely associated with the word (Hoffman et al., 2012; Krasileva et al., 2017). This vocabulary test is the only operationalization of verbal IQ in the dataset.Dependent variableInformal mentorship is captured in this study by using the following retrospective question from Wave 3 of the Add Health data: “Other than your parents or step‐parents, has an adult made an important positive difference in your life at any time since you were 14 years old?” Respondents were then asked “How is this person related to you?” and given response options like “family,” “teacher/counselor,” “friend's parent,” “neighbor,” and “religious leader.” Following the guidelines of previous research (see Miranda‐Chan et al., 2016), we did not count spouses, partners, siblings, peers, or coworkers as informal mentors, and recoded the respondents that indicated such as not having an informal mentor.In accordance to previous studies, we ran a two‐step cluster analysis (Norusis, 2008) based on eight variables in Wave 3 of the Add Health data characterizing the mentoring relationship: mentor role, how youth met mentor, two indicators of relationship duration, two indicators of frequency of contact, two indicators of youth‐rated closeness to the mentor, and support provided by the mentor. This latter variable was derived from open‐ended responses in Add Health and coded for a previous study (Gowdy et al., 2020). Because this previous study then examined the association between mentor type and a Wave 4 dependent variable, only observations who had nonmissing information on this Wave 4 variable were coded up for this typology. The two clusters produced from this process were validated with three criteria: each cluster held at least 5% of all observations, clusters made conceptual sense, and there was distinct differentiations among the clusters (Liao & Sánchez, 2019). This process produced our dependent variable for this study: a binary variable indicating either core or capital mentorship.Both of our analytic samples have 45% of the observations categorized as a core mentor and 55% of the observations categorized as capital mentoring. Akin to previous research, the majority of core mentors are from within the extended family, while the majority of capital mentors were school personnel (Spencer et al., 2019). Core mentors had known their mentees for longer than capital mentors had, and see or speak to them more often. Young people in core mentoring relationships also report feelings closer to their mentor and are more likely to report that their mentor was still important at time of data collection. While core mentors were more likely to provide emotional support and instrumental support, capital mentors were more likely to provide valued advice.AnalysisWe used two forms of analysis for this study. The first was a series of logistic regressions predicting core or capital mentorship, first on our whole sample, then segmented by racial‐ethnic status. This analytic approach was chosen as recent work has underscored the fact that the relationship between the qualities of a young person and their likelihood to have an informal mentor likely varies greatly by their racial‐ethnic status (see Fruiht et al., 2022; Gowdy et al., 2022). The second set of analyses utilized C‐Trees, a nonparametric modeling technique that relies on machine learning to build models that allow complex and infinite interactions between a large number of independent variables. The C‐Tree algorithm tests the global null hypothesis of independence between all predictors and the outcome variable, ultimately created by recursively testing models until the null hypothesis is rejected. Using recursive partitioning, each possible interaction between youth demographic and social factors that were entered into the model was modeled to build a categorization model to describe the characteristics that best predict core versus capital mentoring. This analytic strategy results in a tree‐shaped model with splits, called branches, representing variables that significantly contributed to the categorization of the dependent variable. Continuous variables are split into two ranges by the algorithm during the modeling process, based on their contribution to the model, so branches represent both continuous and categorical variables. Each branch also has an associated p value, representing sample‐specific permutation distributions of the test statistics (Hothorn et al., 2006).Both analytic approaches were used and included here as they provide unique perspectives on our central research question. While logistic regression models are commonly used and a standard approach for this type of question, C‐Tree models add new and contextualized findings with their hierarchal approach. After stepping back to reflect on findings from each analytic approach together, this process can better inform the literature than either of them alone. Models were created to categorize mentorship into core versus capital mentoring both in the whole sample, and subsequently segmented by racial‐ethnic status. C‐Tree analyses are robust to missing data, but cannot handle missing data on the outcome variable, therefore models were produced from 4226 observations that had valid race and mentoring data.FindingsDemographicsOf our 4226 participants, a majority (58%) of the young people with a capital mentor were White non‐Hispanic, while only 46% of those with core mentors were White non‐Hispanic. A majority of young people were female (50% for those with capital mentoring, 58% for those with core mentoring) were female. Young people with capital mentors came from families making an average of $6000 more per year than those with core mentors. While 27% of those with capital mentors had parents with a Bachelor's degree or higher, the same was true for only 24% of young people with core mentorship. A complete list of these demographics for both of our analytic samples can be found in Table 1.1TableDescriptive statistics.Regression sample (n = 2294)C‐Tree sample (n = 4226)Core mentoringCapital mentoringCore mentoringCapital mentoring(n = 1037)(n = 1257)(n = 1911)(n = 2315)Demographics45.20%54.80%45.22%54.78%Age (at Wave 4)41.941.7441.831.7341.861.7141.811.73SexMale42.29%48.56%41.56%49.62%Female57.71%51.44%58.44%50.38%Race and ethnicityWhite non‐Hispanic45.92%57.08%46.43%58.40%Black non‐Hispanic27.12%14.46%28.31%14.66%Asian non‐Hispanic6.28%4.75%3.06%4.61%American Indian non‐Hispanic3.14%6.48%5.67%6.01%Other non‐Hispanic3.14%5.35%4.42%3.65%Hispanic14.40%11.87%12.11%12.67%Parental resourcesTotal household income (Wave 1)32,373.0017,643.0038,129.0015,132.0033,593.0016,985.0039,501.0014,531.00Parent educationLess than high school degree16.39%12.73%9.72%9.24%High school degree or equivalent33.68%31.75%32.41%29.62%Some college30.57%32.41%33.75%34.58%Bachelor's degree or equivalent12.86%14.86%15.44%16.03%More than a college degree6.48%8.25%8.67%10.53%Number of ways parent is involved in community0.720.890.980.920.840.950.880.95Parent/neighbor communication about neighbor's childNo14.89%15.79%15.35%14.50%Yes85.11%84.21%84.65%85.50%Parent/neighbor communication about own childNo26.37%28.22%25.26%25.73%Yes73.63%71.78%74.74%74.27%Parent meeting number of child's friends2.021.852.501.962.181.902.461.97Child feeling parent cares about themStrongly disagree0.31%0.26%0.10%0.15%Disagree0.73%0.56%0.86%0.53%Neutral2.26%2.42%1.81%1.91%Agree9.50%11.32%9.34%10.31%Strongly agree87.20%85.44%87.89%87.10%Parent born in neighborhood where child is livingPeer resourcesNumber of friends3.222.623.222.603.212.623.232.58How often they see friendsNot at all9.26%8.10%8.39%7.56%1–2 times22.96%24.56%23.74%25.50%3–4 times28.56%27.14%30.60%27.33%5 or more times39.23%40.20%37.27%39.62%Child feels friends care about themStrongly disagree0.37%0.39%0.19%0.38%Disagree2.48%1.77%2.10%1.91%Neutral12.50%11.49%11.82%10.99%Agree38.29%42.92%38.13%40.99%Strongly agree46.36%43.44%47.76%45.73%School resourcesChild feels part of schoolStrongly disagree2.84%3.05%2.48%2.52%Disagree7.40%7.36%7.15%6.49%Neutral14.43%14.63%14.59%14.66%Agree47.80%46.54%46.90%46.64%Strongly agree27.52%28.43%28.88%29.69%Child gets along with teachersEvery day1.98%2.26%2.00%2.14%Almost every day5.42%4.09%5.62%4.27%Once a week8.69%9.10%8.67%8.70%Just a few times42.06%44.71%42.33%44.20%Never41.85%39.83%41.37%40.69%Child thinks teachers treat students fairlyStrongly disagree4.19%3.44%3.72%2.98%Disagree14.76%15.02%14.97%14.50%Neutral22.49%22.33%23.16%22.06%Agree41.49%43.62%41.75%43.97%Strongly agree17.07%15.59%16.40%16.49%Neighborhood resourcesKnowing others in neighborhoodYes26.31%29.66%72.93%73.21%No73.69%70.34%27.07%26.79%Talking to others in neighborhoodYes18.66%19.05%81.70%82.29%No81.34%80.95%18.30%17.71%Looking out for others in neighborhoodYes25.70%25.93%75.69%76.56%No74.30%74.07%24.31%23.44%Proportion of community who has lived in same county in 19850.830.120.810.120.830.120.810.12Proportion of community 25+ with college degree0.210.150.250.130.220.140.230.13Proportion of community who is unemployed0.080.060.070.060.080.060.070.05Proportion of community whose household income is below 15k0.220.170.180.150.200.160.180.15Child employed at Wave 1Yes43.07%41.79%59.96%61.15%No56.93%58.53%40.04%38.85%Child attendance of weekly religious servicesNever10.56%10.88%9.53%10.00%Less than once/month20.36%20.18%18.40%19.92%At least once/month but less than once/week23.28%21.57%24.69%22.60%Once/week or more45.80%47.37%47.38%47.48%Child feels adult care about themStrongly disagree0.63%0.39%0.29%0.38%Disagree2.31%1.73%2.29%1.37%Neutral7.09%9.57%6.86%8.24%Agree30.36%34.42%29.27%34.43%Strongly agree59.61%53.90%61.30%55.57%Personal resourcesGPA2.760.752.910.762.820.752.950.75Add Health Picture Vocabulary Test63.8710.267.539.4864.499.9867.879.16Regression analysesOur regression analyses, seen in Table 2, found some significant predictors for capital mentorship in relation to core mentorship, and some changes in how the qualities of a young person impact their mentorship type by racial‐ethnic status. Our most consistent finding across sub‐analyses was that those who score higher on the Picture Vocabulary test were likely to report a capital mentor, in comparison to those who had lower scores. For our full sample, age, biological sex, and racial‐ethnic status mattered, with older respondents, female respondents, and Black non‐Hispanic respondents being more likely to report a core mentor (age: OR = 0.95, p = .08; sex: OR = 0.76, p < .001; racial‐ethnic status: OR = 0.52, p < .001 in comparison to White non‐Hispanic respondents). Some indications of parental social capital mattered as well, with respondents whose parents met a greater number of friends having an increase in the odds of reporting a capital mentor, compared to those whose parents had met fewer friends (OR = 1.05; p = .05). Respondents whose parents are from the same neighborhood arebeing more likely to report a core mentor in comparison to those who were not living where their parents grew up (OR = 0.79; p = .04). In consideration of neighborhood resources, living in a community with a higher proportion of college‐educated adults and lower proportion of residential community were both associated with core mentorship (education: OR = 0.35, p = .01; residential mobility: OR = 0.48, p = .07). Finally, higher GPAs and higher Picture Vocabulary scores were both associated with an increase in the odds of reporting a capital mentor, compared to those with lower GPAs and vocabulary scores (GPA: OR = 1.15, p = .04; Picture Vocabulary: OR = 1.03, p < .001).2TableRegression results.Full sampleWhite respondentsBlack respondentsMultiracial and Other respondentsHispanic respondentsN = 2294n = 1223n = 468n = 317n = 286ORSEORSEORSEORSEORSEAge (at Wave 4)0.95*0.270.91**0.040.970.070.950.081.070.11SexMaleRefRefRefRefRefRefRefRefRefRefFemale0.76***0.070.820.110.790.180.52**0.150.45***0.14Race and EthnicityWhite non‐HispanicRefRefNANANANANANANANABlack non‐Hispanic0.52****0.07NANANANANANANANAOther non‐Hispanic0.940.13NANANANANANANANAHispanic0.940.14NANANANANANANANAParental resourcesParent employedYes1.140.131.030.153.54**1.480.840.311.170.43Total household income10101.010.010.990.011.02**0.01Parent educationLess than high school degreeRefRefRefRefRefRefRefRefRefRefHigh school degree or equivalent0.770.130.60.190.610.280.750.40.950.38Some college0.790.140.630.210.670.310.890.460.960.4Bachelor's degree or equivalent0.760.150.650.230.750.370.750.420.680.39More than a college degree0.790.180.710.280.990.540.770.520.28*0.2Number of ways parent is involved in community0.980.0510.070.860.10.970.130.810.16Parent/neighbor communication about neighbor's childYes1.180.161.230.211.380.631.840.710.620.26Parent/neighbor communication about own childYes0.990.110.930.140.980.311.130.371.590.59Parent meeting number of child's friends1.05*0.0261.030.030.970.061.080.081.24**0.12Child feeling parent cares about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.980.571.251.62.362.720.80.76(empty)(empty)Neutral0.950.990.840.412.352.710.40.65(empty)(empty)Agree0.910.90.830.191.81.891.551.531.240.94Strongly agree0.920.890.860.31.491.460.850.351.870.82Parent born in neighborhood where child is living0.79**0.090.770.120.990.270.940.360.470.29Peer resourcesNumber of friends0.990.020.990.021.030.050.950.050.980.06How often they see friendsNot at allRefRefRefRefRefRefRefRefRefRef1–2 times1.020.191.140.310.780.310.750.420.940.523–4 times0.830.151.090.290.41**0.170.850.480.560.325 or more times1.110.191.350.350.570.221.170.650.960.54Child feels friends care about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.530.50.330.542.291.33(empty)(empty)(empty)(empty)Neutral0.480.430.520.81.330.43(empty)(empty)1.751.68Agree0.490.440.520.811.280.320.940.441.100.83Strongly agree0.430.390.430.672.241.360.680.21.080.35School resourcesChild feels part of schoolStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.980.331.280.580.430.350.650.718.880.65Neutral0.950.31.090.450.550.440.9811.391.01Agree0.910.281.190.470.290.221.041.041.190.86Strongly agree0.920.281.30.520.320.250.760.781.020.07Child gets along with teachersEvery dayRefRefRefRefRefRefRefRefRefRefAlmost every day0.670.240.40.240.780.581.11.141.120.82Once a week0.750.260.450.260.990.681.11.091.340.99Just a few times0.80.260.480.261.160.721.161.051.210.82Never0.750.250.450.251.190.761.211.111.210.82Child thinks teachers treat students fairlyStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree1.220.330.870.354.623.440.940.710.490.49Neutral1.080.290.790.313.092.31.10.80.350.35Agree0.220.321.020.393.772.781.050.750.340.33Strongly agree1.190.330.890.363.282.560.720.550.610.64Neighborhood resourcesKnowing others in neighborhoodYes1.020.121.10.180.770.231.210.390.870.33Talking to others in neighborhoodYes1.080.141.010.171.780.640.970.351.030.46Looking out for others in neighborhoodYes10.110.880.151.990.571.280.420.680.24Proportion of community who has lived in same county in 19850.35***0.140.360.21.231.130.010.020.450.6Proportion of community 25+ with college degree0.48*0.20.34*0.191.871.953.745.480.080.12Proportion of community who is unemployed0.820.872.915.270.020.051.183.815.2417.13Proportion of community whose household income is below 15k1.060.470.660.4811.259.880.771.070.630.94Child employed at Wave 1Yes0.90.850.930.130.870.191.190.330.56*0.17Child attendance of weekly religious servicesNeverRefRefRefRefRefRefRefRefRefRefLess than once/month0.980.171.040.231.030.620.720.41.750.9At least once/month but less than once/week0.940.160.960.220.770.430.80.441.850.98Once/week or more0.980.161.020.210.770.411.240.651.580.76Child feels adult care about themStrongly disagreeRefRefRefRefRefRefRefRefRefRefDisagree0.80.670.550.743.542.10.580.440.450.92Neutral1.351.070.80.992.972.830.260.289.1115.52Agree1.260.980.8112.332.10.280.273.215.09Strongly agree1.020.790.630.771.781.60.260.252.33.66Personal resourcesGPA1.15**0.081.130.111.190.21.060.221.43*0.29Add Health Picture Vocabulary Test1.03****0.011.03****0.011.04****0.011.03*0.021.03**0.02Psuedo R20.06000.04000.10000.11000.1800*p < .10; **p < .05; ***p < .01; ****p < .001.Personal and neighborhood resources mattered for if White non‐Hispanic respondents reported a core or capital mentor, with age and living in an educated neighborhood being associated with core mentorship (age: OR = 0.91, p = .02; educated neighborhood: OR = 0.34, p = .05) and higher scores on the Picture Vocabulary test being associated with capital mentorship (OR = 1.03, p < .001).For our Black respondents, having a parent who was employed in Wave 1 was associated with an increase in the odds of reporting a capital mentor (OR = 3.54, p < .001), in comparison to those whose parents were unemployed in Wave 1. Peer‐level and school‐level resources also mattered for this particular racial‐ethnic group: seeing friends more often was associated with core mentorship (OR = 0.41, p = .03) while thinking teachers treated students fairly was associated with capital mentorship (OR = 4.62, p = .04 for “disagree” response, OR = 3.77, p = .07 for “agree” option, both with “strongly disagree” as the reference group). Several neighborhood‐level resources mattered for Black respondents: those who live in a neighborhood where people look out for one another and those who live in a neighborhood with a higher poverty rate are both more likely to report capital mentors (look out: OR = 1.99, p = .02; poverty: OR = 11.25, p = .01) while those who live in a neighborhood with high unemployment rates are likelier to report core mentors as opposed to capital mentors (OR = 0.02, p = .05). Finally, higher Picture Vocabulary scores were associated with capital mentorship (OR = 1.04, p < .001).Similar to our White non‐Hispanic respondents, our Multiracial and Other respondents had very few associations between their resources and which type of mentor they reported. While women were likelier to report a core mentor than men (OR = 0.52, p = .02), as were those living in areas of low residential mobility (in comparison to those in areas of higher residential mobility) (OR = 0.01, p < .001), those who have a higher Picture Vocabulary score were likelier to report capital mentoring (OR = 1.03, p = .06).In consideration of our Hispanic respondents, women were once again likelier than men to report a core mentor (OR = 0.45, p = .01). Parental socioeconomic status mattered more for this subgroup, as having a parent who was employed was associated with an increase in the odds for capital mentorship (OR = 1.02, p = .05) while having an educated parent was associated with core mentorship (OR = 0.28, p = .08). The more friends of the respondent that the parent had met, the likelier the respondent was to have a capital mentor (OR = 1.24, p = .03). Having a job at Wave 1 was also associated with core mentorship (OR = 0.56, p = .06). Finally, increased scores on both personal resources were associated with capital mentorship (GPA: OR = 1.43, p = .08; Picture Vocabulary: OR = 1.03, p = .04).C‐Tree findingsThe initial C‐Tree modeling using the entire sample resulted in a five‐branch model, driven initially by Picture Vocabulary score (p < .001), moderated by race (p < .001). Among participants with a Picture Vocabulary score of 60 or less, Black non‐Hispanic participants (68.9%) were more likely than participants of other ethnicities (51.8%) to have a core mentor. Similar trends were held among those with higher Picture Vocabulary scores. On this branch, 52.6% of Black non‐Hispanic participants had a core mentor, whereas, among those of other races, the tree split again on Picture Vocabulary score (p = .004), this time at a score of 70. That is, non‐Black participants with a Picture Vocabulary score between 60 and 70, 41.3% had core mentors, but among those with a score above 70, the model continued to split. The third branching point was at the number of friends that a participant's parent had met (p = .043) such that among these non‐Black participants with higher Picture Vocabulary scores, those whose parents had met more than 2 of their friends were the least likely to have a core mentor (30.2%) of any group in the model. Among those whose parents had met 2 or fewer of their friends, age further moderated the relationship (p = .011), such that 32.2% of younger participants had a core mentor, but 44.3% of older participants had a core mentor. See Figure 1. Taken together, these results speak to the contribution of Picture Vocabulary score and race, in particular as predictors of mentor type. Those participants with higher vocabulary scores were significantly less likely to have core mentors, and regardless of score, Black non‐Hispanic participants were more to have a core mentor than their counterparts of other races.1FigureFull sample (N = 4226).To further explore the predictors of mentor type among participants of different racial and ethnic backgrounds, C‐Tree models were created. Models were generally similar in their structure across racial and ethnic groups with Picture Vocabulary score (ps < .003) on all four models. The model for Black non‐Hispanic participants split only once, on Picture Vocabulary score, such that those with scores of 60 or lower were much more likely to have core mentors (67.8%) than those with scores over 60 (52.1%). Among White non‐Hispanic participants, 44.7% of those with a Picture Vocabulary of 70 or below had core mentors, whereas among higher scorers core mentoring was less common both in younger (30.3% core mentors) and older participants (42.3%; p = .035). Among Hispanic participants, the model split first on Picture Vocabulary score, again such that those with a score over 59 were less likely to have a core mentor. However, among lower scorers feeling like people in their neighborhoods looked out for each other predicted their mentor type (p = .012). Participants who agreed that their neighbors looked out for each other were much more likely to have a core mentor (65.6%) than those with less close neighborhoods (40.7%). Similarly, among participants of Multiracial and Other racial and ethnic groups, while Picture Vocabulary drove the initial split such that higher scorers were less likely to have core mentors (36.4%), residential mobility moderated that effect for lower scorers. Participants who lived in neighborhoods where about 76% or less of their neighbors were long‐term residents of the county were much less likely to have core mentors (30.2%) than those with a higher proportion of longer‐term residents (61.1%). Please see Figure 2 for all subsample C‐Tree plots.2FigureC‐Tree plots by subsample.DISCUSSIONAccess to mentoring relationships in adolescence can promote success and well‐being during the transition to adulthood. Both core mentoring from a caring adult committed to promoting socioemotional well‐being, and capital mentoring that generally comes from more distal connections who support academic and vocational development, can be rich assets for young people (Gowdy & Spencer, 2021). The present study aimed to identify the characteristics of youth who access more core versus capital mentoring to best identify areas for support and intervention. Findings demonstrate that participants' scores on a vocabulary test taken in adolescence were across the board the best predictor of capital mentoring. Furthermore, Black participants and females were more likely to receive core mentoring. Analyses investigating the unique predictors of core and capital mentoring among participants of different races shed some light on parental and neighborhood resources that may promote different types of mentoring, generally supporting prior findings that youth with more resources are more likely to report a more distal mentor (Fruiht et al., 2022; Raposa et al., 2018) who provides more academic/vocational support.Across both analytic strategies, Peabody Picture Vocabulary scores were the most consistent predictor of capital mentoring. That is, participants who had a stronger vocabulary in adolescence were more likely to report a capital mentor. Instead of translating this at face value, however, literature tells us to focus on the socioeconomic differences in vocabulary scores. Indeed, differences in vocabulary by socioeconomic status can be seen in children as young as 18 months (Fernald et al., 2012). The vocabulary measure for this study was collected during Wave 1, when participants were between 11 and 19 years old, meaning that the socioeconomic differences among our sample youth have likely shown themselves in differences in vocabulary scores. This then retranslates our most consistent finding as capital mentoring being most available to students of higher socioeconomic status, a finding in alignment with previous research (Gowdy & Spencer, 2021).Race was also a notable predictor across models, with Black non‐Hispanic participants being much more likely to receive to core mentoring. This is very much in line with prior research suggesting that Black youth have strong networks of family and fictive‐kin that provide them support and mentorship (Hurd & Sellers, 2013; Wittrup et al., 2019), as well as neighborhood‐level sociological research that demonstrates benefits and prevalence of intergenerational relationships in Black neighborhoods that depend not on assets coming in from outside of the neighborhood, but on the stability of a neighborhood itself (Sampson, 1999). Furthermore, in line with past findings suggesting that more resourced youth have more access to mentors in their larger communities (Erickson et al., 2009; Fruiht et al., 2022), parental resources predicted capital mentoring. Most consistently, participants whose parents had met more of their friends were more likely to have capital mentors, but more objective indicators such as income and education were also significant predictors of capital mentoring in regression models. Furthermore, gender was a consistent predictor; females were more likely to have a core mentor in these models. However, this finding did not hold in C‐Tree analyses.In addition to looking at predictors of core and capital mentoring in the entire sample, additional analyses subsetted the sample by race to more clearly understand the unique predictors of mentoring among youth in different racial and ethnic groups. Recent research has demonstrated the utility of this strategy to better parse the specific demographic and environmental factors that influence mentorship among youth from different cultural backgrounds and lived experiences. This methodology ensures that variables linked to systemic inequalities (e.g., parental income and education) do not overshadow potentially important trends that highlight the unique resources of people of Color (Fruiht et al., 2022). While findings from analyses of separate racial and ethnic groups were largely consistent with those from the larger sample in that Picture Vocabulary was a consistent predictor of capital mentoring, more nuanced findings about parental and neighborhood characteristics emerged for different racial and ethnic groups. Regression models generally demonstrated that youth with more resourced parents and those from more resourced neighborhoods (more educated, less poverty, low unemployment) were more likely to have capital mentors. One notable exception, however, was that among Hispanic participants, having a more‐educated parent was actually predictive of core mentoring. This finding may speak to the potential for more educated family members, who share a culture and life experiences with a Hispanic adolescent, to serve as mentors and role models, warranting further investigation.Furthermore, the more targeted approach of the C‐Tree analyses highlighted the ways that neighborhood factors support core mentoring relationships among some youth of Color. For example, among Hispanic participants with lower Picture Vocabulary scores, those who felt their neighbors looked out for each other were more likely to report core mentoring. Similarly, among Multiracial and Other race participants with lower Picture Vocabulary scores, lower residential mobility predicted core mentoring. Lower residential mobility and neighbors looking out for each other are likely both indicators of neighborhood‐level social capital. Therefore, these findings may speak to the potential for building close‐knit and emotionally supportive connections within one's community, and the benefit of residential stability in supporting that. Taken together, findings speak to the rich resources available in neighborhoods and the propensity of stable, closer‐knit neighborhoods to translate into mentorship.Although there is important literature establishing what qualities of a young person and their context lead to informal mentoring broadly (see Erickson et al., 2009; Fruiht et al., 2022), there have been no published studies before the present that focuses on predictors of core versus capital mentoring. While both types of mentoring provide supports that benefit a protégé, given the differences in the mentoring functions provided by core and capital mentors it is critical to understand the assets available to different youth. For instance, core mentoring may provide a protégé with socioemotional support and a deeper connection to the community that serves a protective function, particularly for youth of Color as they establish a sense of racial identity (Hurd et al., 2012). Conversely, because that capital mentoring is associated with economic mobility for low‐income youth and youth of Color (Gowdy & Spencer, 2021), it is important that we understand who is likely and unlikely to be in this type of mentoring relationship. Not only do we need to know who is unlikely to have a capital mentor for mobility‐focused intervention development, but researchers also need this knowledge to isolate the impact of mentoring alone, without the potentially confounding variables that predict both capital mentoring and the outcome of interest. Given that our most consistent finding throughout this study is that higher Picture Vocabulary scores predict mentoring, contextualized with our knowledge that these are likely an indicator of socioeconomic differences in early childhood, our findings potentially underscore the true impact capital mentoring can have on upward mobility for low‐income youth.LimitationsThe present study utilized a large nationally representative longitudinal dataset to investigate these factors. While data sets of this nature have many methodological strengths, they also come with limitations that impact the generalizability of our findings. Data collection for this longitudinal study began in 1994–1995 when most participants were in high school. As a result, these analyses capture the mentoring relationships of participants who are now well into adulthood and not the experiences of today's youth. Demographic and social shifts over the past three decades have impacted the experiences of youth of Color in their neighborhoods as well as access to higher education and mentoring in the context of higher education. Therefore, results must be considered through an appropriate historical lens. Furthermore, the Add Health dataset oversampled for highly educated Black families. While the general pattern of findings for Black youth was in line with overall trends, some caution should be used in generalizing findings about Black families, in particular, from these findings.Beyond the historical context of the data, there are issues in measurement. The primary question on informal mentoring is prone to recall bias, as are all of the follow‐up questions used to create the core and capital clusters. In addition, all of these informal mentoring questions are asked about only one person, eliminating our ability to understand how a young person reports more than one informal mentor. The Add Health study asked participants to report being either male or female, limiting our understanding of how gender nonconforming young people experience informal mentorship.Implications and future directionsFindings have significant implications for the development and implementation of mentoring programs for youth. Namely, they illuminate opportunities to supplement the naturally occurring and informal relationships that young people already have with formal mentoring that provides unique additional support. Black youth, for instance, are more likely to access core than capital mentoring across the board. While a stronger vocabulary and more resourced families and neighborhoods may promote access to capital mentors, our findings generally suggest that Black youth have access to stronger networks of family and friends who support them (Hurd & Sellers, 2013; Wittrup et al., 2019). While core mentoring relationships may help the youth of Color develop a sense of pride and identity (Albright et al., 2017; Hurd et al., 2012), Black communities may benefit from mentoring programs that supplement that core support with more capital mentoring to promote economic mobility. Similarly, participants of Multiracial and Other races and ethnicities who live in neighborhoods with low residential mobility may also have access to close, supportive core mentoring relationships. This suggests that neighborhoods with less mobility may be potentially opportune locations to promote capital mentoring, but that such efforts must be complimentary to and supportive of the core mentoring that these youth may already access.As mental health issues become increasingly commonplace among teenagers and young adults (Twenge et al., 2019) and today adolescents and young adults report unprecedented levels of loneliness (Demarinis, 2020), it is also important to consider the potential gaps in core mentoring for some young people. The Add Health interview collected data on just one significant nonparental relationship, therefore we cannot know the extent of the support that young people experience outside of this single mentoring relationship, however, our findings do suggest that some youth may be missing out on the benefits of core mentoring. More resourced communities may put undo pressure on teenagers to excel academically, providing a good deal of academic and career‐oriented support, without ensuring that they have nonparental adults to turn to that support the development of psychological well‐being. Just as access to capital mentoring may be useful to supplement the naturally occurring supports in less‐resourced communities, there may be a need to better support the socioemotional needs of some young people. The gender difference in mentor‐type demonstrated by our regression models may speak to the particular lack of these relationships among adolescent boys.CONCLUSIONAs they move through adolescence and into adulthood, young people learn to balance the competing needs for socioemotional support and instrumental support that promote their future career ambitions. Mentoring from committed adults can help young people master skills and pave the road to a happy and successful adult life, providing different types of support depending on the role they play in the life of a protégé. This study demonstrates some demographic and environmental differences between youth who report support from core versus capital mentors, highlighting that youth of higher socioeconomic status, who experience the cognitive benefits of being more resourced, are more likely to have capital mentors. Given prior findings that capital mentors can help promote economic mobility, these findings highlight the possibilities for the promotion of capital mentoring for less‐resourced youth, while still recognizing the asset of core mentoring from the communities and families in which these youth are embedded.REFERENCESAbelev, M. S. (2009). Advancing out of poverty: Social class worldview and its relation to resilience. Journal of Adolescent Research, 24(1), 114–141. https://doi.org/10.1177/0743558408328441Albright, J. N., Hurd, N. M., & Hussain, S. B. (2017). Applying a social justice lens to youth mentoring: A review of the literature and recommendations for practice. American Journal of Community Psychology, 59, 363–381. https://doi.org/10.1002/ajcp.12143Allen, T. D., & Eby, L. T. (2011). The Blackwell handbook of mentoring: A multiple perspectives approach. John Wiley & Sons.Ashtiani, M., & Feliciano, C. (2018). Access and mobilization: How social capital relates to low‐income youth's postsecondary educational (PSE) attainment. Youth & Society, 50(4), 439–461. https://doi.org/10.1177/0044118x15607163Black, D., Grenard, J., Sussman, S., & Rohrbach, L. (2010). The influence of school‐based natural mentoring relationships on school attachment and subsequent adolescent risk behaviors. Health Education Research, 25(5), 892–902.Chang, E. S., Greenberger, E., Chen, C., Heckhausen, J., & Farruggia, S. P. (2010). Nonparental adults as social resources in the transition to adulthood. Journal of Research on Adolescence, 20(4), 1065–1082. https://doi.org/10.1111/j.1532-7795.2010.00662.xColeman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120.Demarinis, S. (2020). Loneliness at epidemic levels in America. Explore, 16(5), 278–279. https://doi.org/10.1016/j.explore.2020.06.008DuBois, D. L., & Silverthorn, N. (2005a). Natural mentoring relationships and adolescent health: Evidence from a national study. American Journal of Public Health, 95(3), 518–524. https://doi.org/10.2105/ajph.2003.031476DuBois, D. L., & Silverthorn, N. (2005b). Characteristics of natural mentoring relationships and adolescent adjustment: Evidence from a national study. The Journal of Primary Prevention, 26(2), 69–92. https://doi.org/10.1007/s10935-005-1832-4Erickson, L. D., McDonald, S., & Elder, G. H. (2009). Informal mentors and education: Complementary or compensatory resources. Sociology of Education, 82(4), 344–367. https://doi.org/10.1177/003804070908200403Erola, J., Jalonen, S., & Lehti, H. (2016). Parental education, class and income over early life course and children's achievement. Research in Social Stratification and Mobility, 44, 33–43. https://doi.org/10.1016/j.rssm.2016.01.003Fernald, A., Marchman, V. A., & Weisleder, A. (2012). SES differences in language processing skill and vocabulary are evident at 18 months. Developmental Science, 16(2), 234–248. https://doi.org/10.1111/desc.12019Fruiht, V., Boeder, J., & Chan, T. (2022). Rising stars and underdogs: The role race and parental education play in predicting mentorship. Youth & Society, 54(4), 635–661. https://doi.org/10.1177/0044118x211004637Gowdy, G., Miller, D., & Spencer, R. (2019). Expanding and deepening our understanding of which young people have an informal mentor. Children and Youth Services Review, 108. https://doi.org/10.1016/j.childyouth.2019.104490Gowdy, G., Miller, D., & Spencer, R. (2020). Helping those who need it the least: A counterfactual and comparative analysis of whether informal mentoring promotes economic upward mobility for low‐ and middle‐income youth. Youth & Society, 53(7), 1152–1180. https://doi.org/10.1177/0044118X20959241Gowdy, G., Miller, D. P., & Spencer, R. (2021). Helping those who need it the least: A counterfactual and comparative analysis of whether informal mentoring promotes economic upward mobility for low‐ and middle‐income youth. Youth & Society, 53(7), 1152–1180. https://doi.org/10.1177/0044118x20959241Gowdy, G., & Hogan, S. (2021). Informal mentoring among foster youth entering higher education. Children and Youth Services Review, 120, 105716. https://doi.org/10.1016/j.childyouth.2020.105716Gowdy, G., Palmer, M. T., Saastamoinen, M., & Rivera, M. (2022). Using a social work perspective to understand contextual factors impacting access to informal mentorship for under‐resourced and minoritized youth. Child and Adolescent Social Work Journal. https://doi.org/10.1007/s10560-022-00838-4Gowdy, G., & Spencer, R. (2021). It's who you know that matters: Identifying which type of informal mentor is most likely to promote economic mobility for vulnerable youth. The Journal of Primary Prevention, 42(3), 217–238. https://doi.org/10.1007/s10935-021-00630-7Hagler, M. (2017). The influence of parents and natural mentors on young adults' substance use behaviors: Evidence from a national study [Master's thesis]. Western University.Higgins, M. C., & Kram, K. E. (2001). Reconceptualizing mentoring at work: A developmental network perspective. Academy of Management Review, 26(2), 264–288.Hoffman, L., Templin, J., & Rice, M. L. (2012). Linking outcomes from peabody picture vocabulary test forms using item response models. Journal of Speech, Language, and Hearing Research, 55(3), 754–763. https://doi.org/10.1044/1092-4388(2011/10-0216)Hothorn, T., Hornik, K., & Zeileis, A. (2006). Unbiased recursive partitioning: A conditional inference framework. Journal of Computational and Graphical Statistics, 15, 651–674.Hurd, N. M., Sánchez, B., Zimmerman, M. A., & Caldwell, C. H. (2012). Natural mentors, racial identity, and educational attainment among African American adolescents: Exploring pathways to success. Child Development, 83, 1196–1212. https://doi.org/10.1111/j.1467-8624.2012.01769.xHurd, N. M., & Sellers, R. M. (2013). Black adolescents' relationships with natural mentors: Associations with academic engagement via social and emotional development. Cultural Diversity and Ethnic Minority Psychology, 19(1), 76–85. https://doi.org/10.1037/a0031095Inzer, L. D., & Crawford, C. B. (2005). A review of formal and informal mentoring. Journal of Leadership Education, 4(1), 31–50. https://doi.org/10.12806/v4/i1/tf2Krasileva, K. E., Sanders, S. J., & Bal, V. H. (2017). Peabody picture vocabulary test: Proxy for verbal IQ in genetic studies of autism spectrum disorder. Journal of Autism and Developmental Disorders, 47(4), 1073–1085. https://doi.org/10.1007/s10803-017-3030-7Liao, C., & Sánchez, B. (2019). Mentoring relationship quality profiles and their association with urban, low‐income youth's academic outcomes. Youth & Society, 51(4), 443–462. https://doi.org/10.1177/0044118x16668058McDonald, S., & Lambert, J. (2014). The long arm of mentoring: A counterfactual analysis of natural youth mentoring and employment outcomes in early careers. American Journal of Community Psychology, 54(3–4), 262–273. https://doi.org/10.1007/s10464-014-9670-2Miranda‐Chan, T., Fruiht, V., Dubon, V., & Wray‐Lake, L. (2016). The functions and longitudinal outcomes of adolescents' naturally occurring mentorships. American Journal of Community Psychology, 57(1–2), 47–59. https://doi.org/10.1002/ajcp.12031Munson, M. R., & McMillen, J. C. (2009). Natural mentoring and psychosocial outcomes among older youth transitioning from foster care. Children and Youth Services Review, 31(1), 104–111. https://doi.org/10.1016/j.childyouth.2008.06.003Norusis, M. (2008). SPSS 16.0 Statistical Procedures Companion. Prentice Hall.Putnam, R. (2015). Our kids: The American dream in crisis. Simon & Schuster.Raposa, E., Erickson, L. D., Hagler, M., & Rhodes, J. (2018). How economic disadvantage affects the availability and nature of mentoring relationships during the transition to adulthood. American Journal of Community Psychology, 61(1–2).Reynolds, J. R., & Parrish, M. (2018). Natural mentors, social class, and college success. American Journal of Community Psychology, 61(1–2), 179–190. https://doi.org/10.1002/ajcp.12209Sampson, R. J. (1999). What Community Supplies. In R. F. Ferguson, & W. T. Dickens (Eds.), Urban Problems and Community Development (pp. 241–292). Brookings Institution Press.Schwartz, S. E. O., Kanchewa, S. S., Rhodes, J. E., Gowdy, G., Stark, A. M., Horn, J. P., Parnes, M., & Spencer, R. (2018). “I'm having a little struggle with this; can you help me out?”: Examining impacts and processes of a social capital intervention for first‐generation college students. American Journal of Community Psychology, 61(1–2), 166–178. https://doi.org/10.1002/ajcp.12206Spencer, R., Drew, A. L., Gowdy, G., & Horn, J. P. (2018). “A positive guiding hand”: A qualitative examination of youth‐initiated mentoring and the promotion of interdependence among foster care youth. Children and Youth Services Review, 93, 41–50. https://doi.org/10.1016/j.childyouth.2018.06.038Spencer, R., Gowdy, G., Drew, A. L., Abrams, E., & Rhodes, J. E. (2019). “Who knows me the best and can encourage me the most?:” Matching and early relationship development in youth‐initiated mentoring relationships with system‐involved youth. Journal of Adolescent Research. https://doi.org/10.1177/0743558418755686Twenge, J. M., Cooper, A. B., Joiner, T. E., Duffy, M. E., & Binau, S. G. (2019). Age, period, and cohort trends in mood disorder indicators and suicide‐related outcomes in a nationally representative dataset, 2005–2017. Journal of Abnormal Psychology, 128(3), 185–199. https://doi.org/10.1037/abn0000410Van Dam, L., Smit, D., Wildschut, B., Branje, S. J. T., Rhodes, J. E., Assink, M., & Stams, G. J. J. M. (2018). Does natural mentoring matter? A multilevel meta‐analysis on the association between natural mentoring and youth outcomes. American Journal of Community Psychology, 62(1–2), 203–220. https://doi.org/10.1002/ajcp.12248Wittrup, A. R., Hussain, S. B., Albright, J. N., Hurd, N. M., Varner, F. A., & Mattis, J. S. (2019). Natural mentors, racial pride, and academic engagement among black adolescents: Resilience in the context of perceived discrimination. Youth & Society, 51(4), 463–483. https://doi.org/10.1177/0044118X16680546White, M., & Glick, J. (2000). Generation status, social capital, and the routes out of high school. Sociological Forum, 15(4), 671–691.

Journal

American Journal of Community Psychology – Wiley

Published: Jun 1, 2023

Keywords: Add Health; Peabody Picture Vocabulary; youth mentoring

There are no references for this article.