Rethinking School-Based Ties: Social Class and the Role of Institutional Agents in Adolescents’ College Plans

by Jessica Halliday Hardie - 2018

Background: Planning for college is an increasingly common rite of passage for high school students. Institutional agents—nonkin adults who possess institutional resources—are important sources of support and guidance in this process.

Purpose: This mixed-methods study examines social class differences in the involvement of school-based institutional agents such as teachers and school counselors in helping young people plan for college and the future.

Population: Interviews were conducted with 61 middle-class, working-class, and poor young women to collect information regarding their future plans, social ties, and role of social ties in guiding their plans. In addition, the author uses survey data from the Education Longitudinal Study (ELS) to examine the association between social class and the role of school-based ties in adolescents’ college planning.

Research Design: Analyses of in-person interview transcripts involved inductive coding and the development of effects matrices to compare coding output by class. Quantitative models were constructed based on qualitative findings. Based on these findings, analyses of ELS data used hierarchical models to estimate the association between social class background and receiving encouragement and information regarding college from social ties.

Findings: Analyses of interview transcripts reveal that disadvantaged young women see school-based ties as their primary means for college planning, whereas middle-class young women often discount advice from these ties when other sources of advice are available. Quantitative models also show that disadvantaged youth rely on school-based ties for information in the college planning process to a greater extent than do middle-class youth. However, disadvantaged youth receive less encouragement to attend college from school and nonschool ties, even after accounting for academic performance.

Conclusions: Inequality in access to college stems in part from differences in the resources available to high school students as they plan for the future. Disadvantaged youth look to schools to help them plan; if schools marshal their resources to assist these young people, they can help address existing inequality in access to college.

A primary aim of high schools in the United States is to prepare young people to pursue their educational and occupational goals, and much of this preparation is thought to take place through social ties encouragement, information, and guidance in the college planning process (Bettie, 2003; Bloom, 2007; Diamond, Randolph, & Spillane, 2004; Hardie, 2015; Holland, 2010; McDonough, 1997; Stanton-Salazar, 1997; Stanton-Salazar & Spina, 2000). Yet research suggests two alternative hypotheses regarding the relationship between social class and adolescent school-based social capital: first, that middle-class adolescents may form closer and more productive ties with teachers and school counselors than disadvantaged youth (e.g., Stanton-Salazar, 2011; Stanton-Salazar & Spina, 2000), and second, that disadvantaged youth may rely more heavily on school-based ties than their middle-class peers do (McDonough, 1997). Both may be true, but prior research has rarely explored these possibilities in conjunction with one another, nor how these differences may shape young peoples chances of following through on their educational and occupational goals.

The first argument is supported by a wealth of research on class privilege in schools. Researchers argue that privileged youth are more likely to develop and rely on relationships with teachers and administrators than their disadvantaged peers, thus improving their chances of developing ambitious goals and achieving them (Bloom, 2007; Stanton-Salazar, 1997). Indeed, evidence suggests that institutional agents, or high status nonkin adults (Stanton-Salazar, 2011), are more receptive to middle-class values and habitus (Lareau, 1987). Additionally, some scholars have argued that working-class and poor youth distrust their teachers and administrators, making it less likely that they will turn to them for support (Bettie, 2003; Stanton-Salazar & Spina, 2000; Willis, 1981). The second argument does not necessarily contradict the first but contends that schools may be particularly important sources of support for disadvantaged youth. Studies suggest that middle-class adolescents are connected to larger, more educated networks of adults outside of schools compared with disadvantaged youth, possibly making middle-class youth rely less on schools for future planning than their disadvantaged peers (Hardie, 2015; Lareau & Weininger, 2008). As McDonough (1997) has argued, adolescents who do not have the support of friends or family members in planning for the future may rely on school personnel more than their privileged peers precisely because their other sources of support are weak.

To investigate these alternative possibilities, I first analyze qualitative data from 61 in-depth interviews I conducted with junior and senior high school girls to examine social class differences in young womens access to school-based social ties in comparison with other venues (e.g. family, community) and the resources they receive through these ties in planning for college. Based on conclusions drawn from these interviews, I next build quantitative models to test these qualitative findings and expand on them using a nationally representative sample of young men and women from the Education Longitudinal Study of 2002 (ELS). Together, the elements of this mixed-methods study examine social class differences in young peoples use of social capital in planning for college and find that school-based ties are a particularly important, yet at times underproductive, form of support for poor and working-class youth.

THEORETICAL BACKGROUND

Social capital comprises the aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition (Bourdieu, 1986, p. 249). This definition of social capital incorporates both the size of ones social network and the resources available through it. Social class is expected to be associated with social capital because wealthier members of society are more likely to associate with others, like themselves, who have numerous resources and diverse social networks (Lin, 2000). Likewise, middle-class youth form ties through their neighborhoods, schools, religious organizations, and structured activities where others of similar social class standing live, study, worship, and participate (e.g., Crosnoe, Johnson, & Elder, 2004; Smith, 2003; Sun, 1999). Their parents can also draw on their own ties to provide information to their children regarding colleges and careers (Hardie, 2015; Kim & Schneider, 2005; Lareau, 2000). The parents of working-class and poor youth, on the other hand, have fewer ties and may be unable to afford enrollment in structured activities and access to high-resource schools (Covay & Carbonaro, 2010; Lareau, 2011; Stearns & Glennie, 2010). The neighborhoods in which they reside are often characterized by concentrated disadvantage and have few institutions, reducing the scope and effectiveness of networks in such communities (Briggs, 1998; Wacquant & Wilson, 1989; Wilson, 1987).

Institutions play a key role in facilitating access to social ties (Bourdieu, 1986; Coleman, 1988; Lin, 2001). They provide the organizational structure for members to meet regularly, share a common identity, trade information, and establish trust. For adolescents, institutional affiliations such as schools may be particularly important, and provide the opportunity to forge connections to high-status adult ties (those who possess a greater number of resources). These high-status ties can act as institutional agents who, by virtue of their own social, cultural, and human capital, transmit institutional resources and privileges to others (Stanton-Salazar, 2011). One key aspect of the social capital literature, however, suggests that it is not only access to high-status social ties that matters, but the mobilization of these ties (Lin, 1999); that is, an individual may have access to a tie but never ask for or receive resources through it. This mobilization relies on both parties: The individual must ask a tie for some kind of resource, and the tie must be willing to provide it. In examining social class differences in planning for college, therefore, it is important to examine both who young people turn to for help and who provides resources to them.

CLASS DIFFERENCES IN SCHOOL SOCIAL CAPITAL

Much research has found that middle-class youth are better positioned to access and mobilize resources in schools than working-class and poor youth (Bettie, 2003; Bloom, 2007; McDonough, 1997; Stanton-Salazar, 1997). Obtaining resources through ties requires establishing relationships and feeling comfortable asking for help, and prior research shows that middle-class youth are more comfortable asking for and receiving assistance than working-class and poor youth (Calarco, 2011; Lareau, 2011; Stanton-Salazar, 1997). Institutional agents are also more receptive to middle-class values and habitus, increasing the likelihood that they will provide such help (Lareau, 1987). Conversely, disadvantaged youth lack the cultural signals necessary to make claims on adults time and attention and may be reluctant to rely on others for help (Stanton-Salazar & Dornbusch, 1995; Stanton-Salazar & Spina, 2000). Some scholars have argued that working-class and poor youth distrust their teachers and administrators, making it less likely that they will turn to them for information and advice (Bettie, 2003; M. M. Holland, 2015; Stanton-Salazar & Spina, 2000; Willis, 1981). Finally, schools serving poor and working-class youth have larger school counselor caseloads than the national average (McDonough, 2004), which prior research has found is associated with students likelihood of transitioning to college (Woods & Domina, 2014).

Alternatively, schools may be an important source of high status social ties for disadvantaged youth because they are often the only setting in which these young people have access to college-educated adults. As McDonough (1997) has argued, adolescents who do not have the support of friends or family members in planning for the future may rely on school personnel more than their peers precisely because their other sources of support are weak. Furthermore, qualitative research shows that working-class parents and children rely on schools for information regarding college applications and admission because they see college advising as the purview of school staff (Lareau & Weininger, 2008; Mehan, Villanueva, Hubbard, & Lintz 1996; Weis, 1990).

In sum, prior research implicates social class as an important factor in two processes linked to the mobilization of social capital in the college-going process: young peoples likelihood of asking for help and young peoples likeliness to receive help from adults, whether asked or unasked. Help in this case is a general set of resources that can include encouragement to attend college, information about the college application process, or connections in the form of recommendation letters. However, it is unclear how social class structures whom high school students turn to for support and the resources they receive through these ties. On one hand, researchers often assume that privileged youth are more likely to develop and rely on relationships with teachers and administrators, thus improving their chances of receiving help. On the other, working-class and poor youth and their parents report relying on schools for help, particularly in the college planning process. Both may be true, if disadvantaged youth are more likely to turn to schools for help yet receive less assistance. Identifying disparities in the use of school-based institutional agents and in the proportion of school-based to family-based and friend-based ties will illuminate how class structures access to institutional agents and mobilization of the resources they provide, ultimately paving the way to the intergenerational transmission of social class.

CURRENT STUDY

This article aims to elucidate the ways that young people use (or do not use) the social ties they have available to them in school relative to other contexts (e.g., home or neighborhood), and the extent to which they report being provided resources through these ties. I focus on the college planning process because this is a time when young people need information and advice from knowledgeable adults. To examine this process, I first analyze interviews that I conducted with 61 high school girls about how they planned for college and other future goals. Next, I marshal rich data from the ELS 2002 data set to test and expand on my findings from the qualitative interviews. In particular, the quantitative data allow me to draw inferences from a nationally representative sample of both young men and young women.

Because I include both young men and women in the quantitative data, I examine whether the association between social class and social capital varies by gender. Prior literature on social class and the use of institutional agents for planning does not suggest gender differences (Lareau & Weininger, 2008; McDonough, 1997), and although there is a slightly higher likelihood of young women going to college than young men (DiPrete & Buchmann, 2013), this disparity is not large enough to expect different college planning processes. However, one study of ties to parents, between parents, and between peers suggests that some of young womens advantage in college-going is explained by these social networks (Klevan, Weinberg, & Middleton, 2016). Furthermore, schools are largely staffed by women, which may make it easier for young women to forge ties with institutional agents in schools than young men, potentially weakening the association between social class and mobilization of school-based ties among young women (for literature on gender homophily in social networks, see Marsden, 1987).

METHODS

To investigate the relationship between social class and the use of school-based ties for college planning, I analyze both qualitative in-depth interviews that I conducted with 61 junior and senior high school girls (ages 1618) in the Midwest as well as a nationally representative, longitudinal study of high school students. Theory and findings from the qualitative data informed the development of quantitative models. Both data sets are described next.

QUALITATIVE DATA

In the fall of 2008, I interviewed 61 junior and senior high school girls from two large midwestern high schools, one located in a working-class town and the other located in a middle-class suburb of a midsized city, although middle-class and working-class girls participated at both schools. Participants were selected based on the larger goals of my project: to understand class and race differences in educational and occupational future plans within a definable category of occupational interest, but across an occupational status range. Therefore, I selected young women with an interest in entering the health field, broadly construed.¹ The current article focuses on study participants educational plans and their use of school-based social ties to inform these plans.

Interviews lasted from 45 minutes to 2.5 hours, with an average of about 1.5 hours, and students received $15 for participating. Most interviews were audio recorded, except when the parent or student indicated they preferred the students interview not be recorded (nine cases) and in one case of recording malfunction. I took careful notes to reconstruct nonrecorded interviews.

Participants were interviewed about their aspirations; adults they were close to or could turn to for advice; and their knowledge of work, education, and family pathways. I asked participants to describe all of their adult ties, rather than just those from whom they received information or other resources, to examine differences in the availability and use of social capital by social class. After discussing the initial list of reported close ties, participants were asked about any additional close ties in several domains: family, family friends, school, community, and other organizations of which the respondent was a member. Participants were not asked directly about their ties roles in their future plans, so as to not guide their responses. Instead, they were asked to describe their relationship with the adult and what they usually talked about. If they reported information about college planning or colleges during the interview, they were asked how they had learned this information. Participants frequently described discussing their plans with adult social ties and receiving information through these ties. They were also asked who initiated these conversations.

Participants were also asked to respond to open-ended questions regarding four vignettes, ordered randomly, that recounted hypothetical conflicts young people may have over future plans: not having money for college, a young woman being pressured by her fiancée to leave a four-year college to move close to him and attend a community college, a young woman considering switching majors from pre-med to pre-nursing to accommodate plans to have children, and parents pressuring their daughter to follow a pre-med track instead of majoring in English. Participants answered questions about what they thought the young women in the stories should do and what kinds of people they could turn to for advice. Vignettes are useful in interviews as a way to elicit generalized attitudes and beliefs, compare responses across groups, and explore potentially sensitive subjects in a less personal context (Barter & Renold, 1999). I use these vignettes to compare both within subjects (participants responses to their own circumstances versus their objective response to anothers circumstances) and across subjects (how different groups respond to the same vignette) on the question of whom young people should turn to for advice. See Appendix A1 for an abbreviated copy of the interview guide, including all four vignettes.

Based on student reports of their parents educational and occupational attainment, respondents were first categorized into four social classes, following Thomas and Hickey (2007): upper-middle-class (parent/s employed in a professional occupation), lower-middle-class (parent/s employed in white-collar/semiprofessional occupation), working-class (parent/s worked in blue- or pink-collar jobs), and lower-class (parent/s worked in very poorly paid jobs or in insecure sectors where unemployment was common). However, I found that the approach these young women took to planning for the future, and the social resources they had available to them for this planning, was similar among the upper-middle-class and lower-middle-class. Similarly, working-class and poor young women described similar levels of support and strategies for future planning. Therefore, in this article, I focus on the major differences between upper- and lower-middle-class, referred to as middle-class for brevity (N=25), and working-class/poor families (N=36). Table 1 contains summary information for this sample on race/ethnicity, educational plans, and social ties by social class.

Table 1. Descriptive Statistics of Qualitative Sample (N = 61)

	Working-Class/Poor (N = 36)	Middle-class (N = 25)
Race/Ethnicity
White	26 (72.2%)	13 (52.0%)
Black or biracial (Black and White parents)	8 (22.2%)	11 (44.0%)
Latina	2 (5.6%)	1 (4.0%)
Educational Plans
Graduate/professional schooling	9 (25.0%)	16 (64.0%)
Four-year college degree	10 (27.8%)	6 (24.0%)
Two-year college degree	9 (25.0%)	2 (8.0%)
College (vague)	8 (22.2%)	1 (4.0%)
Total Number of Sources of Social Ties	6.5	8.5
School ties as a proportion	0.14	0.09
Proportion of Girls Who Claim Ties To. . .
Teachers	0.61	0.52
Administrators/counselors	0.17	0.16
Coaches/activity leaders	0.14	0.40

EDUCATION LONGITUDINAL STUDY

The Education Longitudinal Study (ELS) of 2002 is a longitudinal study of high school students conducted by the National Center for Education Statistics (NCES). This data set is ideal for the current studys purposes because it collects data on social ties, college planning, family background, and school characteristics from a nationally representative sample of high school students in the United States. In addition, respondents to this survey are from a similar age cohort as the qualitative sample members (ELS students were in their senior year of high school in 2004, compared with interview participants who were in their junior and senior years of high school in 2008).

The initial national probability sample of 752 public, Catholic, and other private high schools yielded a sample of 15,362 high school sophomores. Data for this study include the base year and first follow-up, in 2004, when respondents were seniors in high school. Respondents are retained if they participated in both waves (N = 12,842). Students who transferred to a new school between the base year and first follow-up were also dropped (N = 1,184) to take advantage of school measures from the baseline survey.² Missing data were imputed using chained equations, and results were obtained by averaging results across five imputed data sets (Little & Rubin 2002). Prediction models used all independent and dependent individual- and school-level variables, but the dependent variables were not imputed (Allison, 2002). Sensitivity analyses using listwise deletion and imputation via the multivariate normal model suggest that results are robust across imputation methods. The analytic sample size for descriptive statistics was 11,658. Analytic sample sizes for multivariate analyses varied by dependent variable, as described next.

Dependent Variables

The dependent variables were derived from two sets of questions asked when respondents were seniors. First, respondents were asked for their assessment of what seven different ties (school counselor, favorite teacher, coach, mother, father, relative, and friend) thought was the most important thing for the respondent to do right after high school (N = 11,555; 99.1% of eligible). Responses included go to college, get a full-time job, enter a vocational or technical program or apprenticeship, enter the military, get married, do whatever the respondent wants, and dont know. Using this information, I created dichotomous indicators of whether the social tie most wanted the teen to go to college (1) or not (0) for each tie. Dont know responses were coded as no because students who did not know whether a tie wanted them to go to college had most likely not been encouraged by that tie directly. Mothers and fathers were combined into parent (with either offering encouragement coded as 1, and 0 otherwise). For brevity, this outcome is referred to as receiving encouragement, on the assumption that in order to know that these ties wanted them to go to college, the respondents probably received either subtle or outright encouragement. Three dependent variables were constructed from these questions: (1) a summed measure of the number of sources of ties³ who provided encouragement (06); (2) the number of sources of school-based ties (school counselor, teacher, or coach) who encouraged college (03); and (3) the percentage of the total number of encouraging ties who were school-based ties (0100). To obtain this last variable, I divided the number of sources of encouraging school-based ties by the total number of sources of encouraging ties, then multiplied it by 100 to obtain the percentage of ties that were school based.

Second, respondents who reported that they planned to attend college at some point (N = 9,675; 83.0% of eligible) were asked, Where have you gone for information about the entrance requirements of various colleges? Respondents could indicate any (or all) of the following: school counselor, teacher, coach, parent, sibling, other relative, and friend.⁴ I refer to these as informative social ties. For the first dependent variable, I created a summed variable of the number of sources of social ties teens reported turning to for advice regarding college (07). The second dependent variable is an indicator of the number of sources of school-based ties the respondents reported consulting (03), and the third measure is the percentage of informative ties who are school based (0100).

Most correlations between the six dependent variables are positive, as expected. The correlation between the total number of informative ties and the total number of school-based informative ties, and the correlation between the total number of encouraging ties and total number of school-based encouraging ties were both strong and positive. Given that school-based ties comprise a large proportion of both measures of total ties, this is unsurprising. However, there is also a strong positive correlation between the total number of school-based encouraging ties and the percentage of encouraging ties that were school based (r = .75). This may suggest that both outcomes tap into the degree of encouragement respondents receive from school-based ties, rather than two distinctly different concepts.

Independent Variable

Social class was measured in the ELS data by socioeconomic status (SES) quartiles.⁵ This measure was compiled by study staff and based on parents reports of their education, income, and occupation, with information imputed from student surveys where missing from parents. Supplementary descriptive statistics (not shown) revealed that young people in the lowest socioeconomic quartile typically had parents who had not completed or just completed high school, were making a median income of $20,000 to $25,000, and frequently worked in service- and labor-related occupations. Those in the second quartile had parents who most frequently completed high school or attended some college, earned a median income of $35,000 to $50,000, and worked in clerical and crafts jobs most frequently. Parents of youth in the third quartile varied between completing high school and completing college, earned a median income of $50,000 to $75,000, and often worked in clerical work or as managers and administrators. Finally, those in the highest quartiles had parents who completed college or graduate school, earned a median income between $75,000 and $100,000, and frequently worked in professional or managerial/administrative occupations.

Control Variables

Demographic and family background factors were obtained in the first wave of data. Race/ethnicity was measured as: White (reference category), Black, Hispanic/Latino, and other race/ethnicity. An indicator of whether the respondent was male (1) or not (0) was included in all models. Family structure was measured in three categories: two biological or adoptive parents (reference category), stepfamily or two guardians, and single-parent or other family structure. Immigrant status was coded as first generation, second generation, or third and above generation (reference category). Three indicators of academic performance were also included: senior year GPA (recoded from a categorical indicator to numerical estimates of GPA based on the midpoints of each category)⁶ and reports of whether the student was enrolled in a college preparatory academic track and whether the student had received an award for high grades since the ninth grade, both measured in the 10th-grade survey.

Finally, I examined several features of respondents school settings at Wave 1. I converted a categorical measure of the percentage of students receiving free or reduced priced lunch to a linear estimate based on category midpoints. Next, I averaged a study staff created scale of studentteacher relations comprising five questions about how teachers got along with and treated students (α = .73) across all student respondents within each school. Study staff also created a measure of academic climate for each school using five administrator responses about students’ and teachers’ attitudes about teaching and learning (α = .86). Dummy variables indicated whether schools were public (reference category), Catholic, or some other kind of private school and by location (urban [reference category], suburban, or rural). Three variables assessed school readiness to support students college planning. Administrators at the schools indicated what proportion of students at the school participated in a college planning program: none (reference category), some, or all. Two measures of the number of students per full-time teacher and number of students per full-time school counselor were also constructed. These are approximations based on a categorical variable of number of students in the school.

ANALYSIS: QUALITATIVE DATA

Each transcribed interview was analyzed using Atlas.ti and Excel software. Throughout the coding process, I focused on participants access to and use of school-based ties compared with their access to and use of family- and community-based social networks, and how these social ties informed their future plans.

Coding the qualitative data proceeded in several steps. First, I constructed a preliminary coding list differentiating social ties by type, the resources available through these ties, and how (and if) these resources were transmitted to the adolescents. A research assistant and the author conducted this broader level of coding to ensure that the identification of social ties and their provision of social ties were coded reliably across coders.⁷ Both coders identified social tie type (e.g., parent, family, school) the same, and they agreed on nearly all examples of a social tie providing support. Additional codes were developed during the reading and rereading of transcripts, which were added to the initial coding list (see Appendix A2 for full coding list). One particularly interesting code that developed during this process was the distinction between types of resources garnered through ties: passive support (approval of participants goals without offering help) and active support (social tie offered participant information, resources, or access to a new social tie). The author conducted this more detailed level of coding alone. Table 2 provides examples of transcribed text that were coded as passive and active support (with active support delineated by subtype), and examples of types of ties. I was particularly interested in discovering whom students identified as their primary source of support and information, and under what circumstances they mobilized resources through these individuals.

Table 2. Sample Transcription Coding

Categories	Examples
Passive support	We have brief conversations, but theyre not really, like, detailed. Like, Ive told her, like, um, I heard that [LocalState] has a, like, a really good music program. I told her [MidCity] was, like, a good, I think, thinking about going there cause its in state, and, you know, it might not cost as much. So, weve had like little, brief discussions. And, she, I think cause shes, like, really not [pause], maybe cause she didnt have the same opportunities I have now when she was younger. So, she really is just like, doesnt know what to say.
Active support
Provide   information	I talk to her about, like, college, and, she taught, I mean she coached softball for a while, and I play softball, so, she gave me some little tips and, about scholarships, sports-wise, and things like that.
Provide   resource	[Interviewer: So, you went to some of the college expos?] Um-hmm. My mom went to em. [Interviewer: Oh, okay. So, when did your mom start going?] She started, it was, there was one, I believe it was the fourteenth; Im not for sure. But she went for me, cause I had to work that day.
Provide social   tie	I talked to, uh, [Calvin] which is my moms friend, I guess hes my friend. And hes the one I shadowed. And Im gonna shadow [Jackie], which is her coworker.
Sources of support
Parents	Well, college is already in my life because my mom. And, like, I see what she goes through, and how many papers she writes, and how long she studies, and gets up at 3:00 in the morning and just goes and does all her work, like. Its crazy. Im like, wow. And shes already, like, setting me up, telling me what Im gonna need to do, and how much work it is. So, shes giving me an understanding of whats gonna happen.
School	People that could help you find money for loans, like, counselors can help you find information for scholarships, or, you know, places for her to get loans. Cause like, I remember a counselor came in the other day, he was like, Watch out for the people who say they can get you loans, cause theyll just take your money and scam you.
Activity leaders	I can go to [my coach] for anything. Like, hes been helping me get a job. Giving me places to go, and then he would be my reference for me. Put in a good word for me.
Family	It has a wonderful nursing program. [Interviewer: Have you ever visited?] My aunt made me go, when I was about 7 years old. And then I went again last year. Shes making me go on a college visit this year-like, overnight.
Community	I looked it up, and a woman at my church went there. She has an RN degree from [St. Theresas], and she said I wouldnt have a problem getting a job if I get my degree from there. Its known. People are like, Oh, youre from [St. Theresas]? Okay.
Friends	Like, from my friends. Theyre all like, You should check this school out. You should check this school out. Its like a big thing with me and my friends. We all look at colleges, and we all kind of, like, trade.

Once I coded all the transcripts, I used the effects matrices approach (Miles & Huberman, 1994) to create tables in Excel that compared one level of coding with another. Specifically, I selected data by code and exported these data (which consisted of full excerpts associated with the code) into Excel spreadsheets. Next, I used these raw data to create spreadsheets containing the effects matrices. For example, I created one spreadsheet that summarized all participants discussion of their school-based ties and resources from their school relevant to college and career planning. Each participant was listed in a row, organized by social class, and several columns were labeled according to the primary ways schools were discussed: resources the school offered for planning, actions the students took to ask for or obtain resources from their school, students opinions of the school, and a miscellaneous column. My summaries of the participants responses were included in these columns. I took the same approach with other themes, such as resources from parents, educational aspirations and motivations, and the full range of adult social ties. This iterative process revealed patterns of social tie type, social tie helpfulness, and the degree to which participants turned to some ties over others in planning for college. In addition to this, I used Excel to create a small database of numeric information, such as the number of social ties each participant listed, the number of social ties in each category, and the number of social ties with a college education. As a last step, I searched for disconfirming evidence for each pattern I observed. As I note in my findings, there were participants whose experiences contradicted the general pattern. Their experiences are instructive.

QUALITATIVE DATA FINDINGS

All the study participants expressed an interest in attending college. However, there were clear social class differences in the specifics of their college plans and preparation to follow through on these plans. As shown in Table 1, slightly over one half of the working-class and poor adolescent girls planned to complete at least four years of college, while the remainder planned to attend a two-year college or university, or held inconsistent or unclear plans (e.g., were not sure or discussed college plans without making a distinction). Nearly all the upper- and lower-middle-class girls expected to complete a bachelors or graduate degree. There are also disparities in these girls college planning processes and social networks, which I discuss below.

WORKING-CLASS AND POOR GIRLS

Interviews with working-class and poor girls revealed two recurrent themes. First, they relied on institutional agents (school counselors and teachers) for advice and information about college, often (but not always) in the absence of significant parental support. Second, the resources they received through school-based ties were insufficient to close the gap between their circumstances at the beginning of their senior year and where they needed to be to successfully apply to and enroll in college.

Tonya, for example, was a White working-class senior who wanted to be a nurse or x-ray technician. She received inconsistent grades throughout high school, ranging from failing courses to a GPA of 3.7 in one semester. When asked about how she planned to get to college, she explained that she felt lost because no one in her family could help her. She described relying on school counselors for all college advice. Later in the interview, when asked to respond to a vignette about Alyce, a young woman being pressured by her parents to go into a pre-med program, I asked her what she would do in similar circumstances. She explained,

Um, I mean, like, my family doesnt have experience in that, so theres really nobody in my family, and theyre, I mean, going to look at which one pays more. . . . And thats whats important to them. Uh, I mean, I, I have to depend on school counselors a lot.

Thus, Tonya described herself as relying on her school counselor for advice for both college and a career path.

Despite this, the resources Tonya received through the school counselor were limited. The school counselor met with Tonya three times in the fall of her senior year, but the bulk of these meetings were spent dealing with graduation credits and other bureaucratic hurdles. By the time I interviewed her in October, Tonyas counselor had arranged to get Tonyas ACT test fees waived and showed her how to register for the exam online, but they had not discussed college programs or applications. Tonya explained, Im expecting to be called down again; I dont know when though.

Other working-class and poor teens reported similar doubts about the help their families could provide and hoped that school counselors would help them plan for college. Terry, a working-class White junior who aspired to be a massage therapist, had not yet talked to any adults about college. Terry reported five adult ties, only one of whom had attended college (a teacher). When I asked her whom she thought she would go to for advice about college, she replied, I dont know about my parents cause they wouldnt know . . . they didnt go to college. Maybe my counselor or somebody like that. She had not yet met with a school counselor about college, however, and did not know what type of college she wanted to attend. Chana, a working-class Black senior, also relied on her school counselor for advice. She was enrolled in general courses and received As and Bs. Her family encouraged her to go to college, but she did not talk with them about the specifics of her plans. She met once with her counselor for her senior talk, which Chana described: She was telling me where Im at right now, and what I need to graduate. And, she was asking also if I have looked through colleges. And I really havent. After that meeting, Chana turned to her older cousin for help: I brought it up to her because, like, I was so confused about the whole process, and how its stressful. And I wanted to know what did she do. Finally, Shauna, a working-class White young woman whose mother had not completed high school, said she had often spoken with her school counselor about college visits and scholarships during her junior year, but when I asked her what he had recommended for her, she said he gave me the number for [college].

Working-class and poor girls reported that they would turn to teachers for support and guidance. Although many of these young women described talking to teachers about personal or family issues, they also saw teachers as a potential source of information about college. When asked whom she would turn to about information for scholarships, Brit, a poor White junior, replied, Id probably ask a teacher before I asked anyone else, and just see if they knew, and if they didnt, then Id probably ask [my counselor] or, like, one of the principals. Similarly, Tiffany, a White working-class senior, explained that teachers were a potential source of information about paying for college: Um, probably, like, teachers, I mean, they could, they could help out, cause they, I know theres a lot of things that teachers know. Like, to help out, student loans, the FAFSA thing. Isabel, a Hispanic senior who was on the honor roll and actively involved in several school activities, also described relying on teachers for help with applications and guidance.

These girls relied on school-based ties in part because their families did not have the resources to help them plan for college. In many cases, they received only passive encouragement from their families and other adults. Dana, a working-class White junior who wanted to be a surgical nurse, had recently moved in with her father after many years of upheaval living with her mother. She received As and Bs in her classes, which were a mix of general and college prep. When I asked Dana about whether she talked to her father about college, she explained,

He just like, says that I should really go, because he dont want it to be, like, him and like my mom where like you have to work really hard to get money . . . I guess cause he didnt go to college. Like, he wants me to go.

When asked whether her parents helped her plan for college yet, Dana explained,

When I was living with my mom, like, tons of colleges were sending me stuff. . . . But, like, my mom would never really sit down and look at the papers with me or anything . . . I was gonna get em, and see if like my dad would look over with them. Like, I could see him doing it more than my mom, cause, like, my moms just like, she dont really care. Like, she thinks its good that I wanna go to college and stuff, but, like, she dont really like take time to actually sit down and talk about it and stuff.

Because she had recently moved schools, Dana had not spoken about her plans with a counselor. She explained that she met with a counselor when she first moved to the school in early fall but only to arrange her course schedule. When asked about the timeline for applying to college, she said, I really dont know . . . Im not really that knowledgeable about all that.

Nekes family also took a hands-off approach. Neke was a poor Black senior who aspired to be a doctor and was enrolled in general and remedial coursework. Neke reported that she told her mother and sister about her plans, explaining,

Neke: [My mom] didnt say anything. And I talked to . . . [my sister]. She asked, and I told her. Thats it.

Interviewer: Do you think your family wants you to go to college?

Neke: Yeah, they all want me to go. But I decided when I was little. My family didnt tell me to. But theyve always wanted me to go.

In contrast, Neke described talking to her counselor about her college applications, which she had partially filled out already and was waiting for a second meeting with the school counselor to complete.

Not all working-class and poor young women reported low levels of support from their parents and nonschool ties. Shannon, for example, was a working-class biracial senior who aspired to be a nurse, sports trainer, or physical education teacher. Shannon lived with her mother, a high school dropout who later completed a GED, and her father, a commercial driver who had completed two years of college. Her parents were deeply involved in Shannons planning. She explained,

They help a lot. They do a lot of research, look up a bunch of scholarships. . . . Theyll plan visits for me, and if, like, if it looks like a school that Im not interested in, just dont worry about it. Were not going to go; cancel it. Uh, theyve been real open-minded.

Thus, Shannons parents were exceptional in their involvement. This may have been facilitated by their financial and living situation; Shannons father made a steady income and her mother did not work, giving her more time to invest in Shannons future. Shannon was also adopted, and it is possible that their difficulty in conceiving a child influenced their parenting style. Or they may have other personal, economic, or educational resources not clear from Shannons interview. In any case, working-class/poor girls experiences were not identical. Although a sizeable number reported that their parents had little involvement and that they hoped their teachers and counselors would help them prepare for college, a handful reported more involvement from parents.

Overall, however, working-class and poor girls looked to school-based ties for help with college. Their reasoning for doing so was clear: They acknowledged that many of the adults close to them had not attended college and therefore did not possess the resources necessary to guide them in the college application and enrollment process. They saw schools as the primary location in which to plan for college. At school, however, bureaucratic complications often intervened. In some cases, counselors did not provide usable information. Additionally, these girls were frequently unprepared to meet with their counselors and reported meetings in which they were given generic information packets about college exams and financial aid. Finally, these young women appeared to wait for their teachers and school counselors to approach them. Although this could be interpreted as a lack of commitment or initiative on the part of these young women, prior research has documented that disadvantaged youth do not feel as comfortable directly requesting help from institutional agents as middle-class students do (e.g., Calarco, 2011; Lareau, 2011). Instead, they trusted schools to fulfill their role of guidance without intervention.

MIDDLE-CLASS GIRLS

Upper- and lower-middle-class girls had access to a large number of ties who freely offered information and advice useful to their future plans. In addition, middle-class ties information was often tailored to the girls occupational interests. Therefore, they were advised not only about general college admissions processes but also about the best colleges for the careers they wished to have. School-based ties often played a more limited role in their planning. While some young women reported talking to school-based ties, they often did so in the context of close relationships with a teacher who was also a coach or activity leader or with school counselors for very specific tasks (like writing recommendation letters).

For example, Jocelyn was a White middle-class senior who aspired to be a pharmacist. She took AP/Honors classes and received mostly As. Jocelyn reported 11 adult ties, most of whom took an active interest in Jocelyns future plans. Jocelyns mother facilitated one of these ties to encourage her daughters interest in pharmacy, a job she characterized as being flexible enough to fit Jocelyns future family plans. As Jocelyn reported, her mother told a family friend who had a degree in pharmacy that Jocelyn enjoyed chemistry and was considering pursuing a pharmacy degree. Diane invited Jocelyn to shadow her at her work, advising doctors and patients for an insurance company. Jocelyn also received advice from a family friend who had been promoted from pharmacy to management at a large grocery chain and shadowed a clinical pharmacist whom she was introduced to by a neighbor. All three contacts shaped Jocelyns plans, providing her with alternative career paths within pharmacy and informing her regarding college programs. Jocelyn visited the college that the clinical pharmacist she shadowed had attended; she applied to this school and several others, as well as for scholarships for these schools, in early November. She was particularly enthusiastic about a program that streamlined students bachelors and graduate degree coursework, allowing students to complete their studies in five years.

Despite showing enthusiasm for the guidance she received from professionals working in the field of pharmacy, Jocelyn appeared indifferent to receiving advice from her school counselor. When I asked Jocelyn whether she had talked to her school counselor, she explained,

I dont really ask her for that much advice about [college], just because she doesnt really know me. She doesnt, like, my last counselor . . . he had my brother, and he had me; he knew my family. He knew that I worked hard, you know, and like that kind of thing. And he would tell me, you know, Well, I think you can go here. And last year when I said I was looking at pharmacy, he goes, Are you looking at [Harper University]? And, you know. Like, he really, like, thought about it. And she doesnt really know me . . . shes not the first person I go to about it.

Jocelyn was in a position to evaluateand in some cases rejectresources available to her through school-based ties.

Most middle-class girls did not directly discount the role of school counselors, but either omitted them from their discussion of college or mentioned their role in broad strokes. Jessie, for example, was a middle-class junior who was enrolled in mostly advanced and Honors classes and received mostly As and some Bs. Jessie recounted eight ties, one of whom was an elementary school teacher she had seen weekly for several years when she returned to her school for music lessons. Otherwise, however, Jessie relied on her parents for guidance. Although she acknowledged that she would probably [ask] some of the counselors here if they know of any scholarships thatd be good at some point, she did not appear to see them as a primary avenue of information.

Some middle-class students also appeared dismissive of the role of teachers. Chelsea, for example, was a Black AP/Honors senior who received As and Bs, and one C in high school. She aspired to complete a Ph.D. in psychology and work with the mentally ill as a clinical counselor. Her parents and relatives provided Chelsea with a great deal of information about colleges and career plans. When asked whether she was close to any teachers, Chelsea explained,

Teachers at [Glenbrook] dont get involved . . . I guess, when youre in a high school, um, atmosphere. Teachers now, they try to stay [as uninvolved] as possible. But I know if they, um, they saw a problem, they would care, but&Its the toughness you have to be working in a high school.

Chelsea reported feeling close to a school administrator who was writing a recommendation letter for her, counseled her when she was feeling lost at school, and was close to her parents. However, she did not appear to look for information or advice from teachers or her school counselor.

Jean, a nearly straight-A honors student, attributed her initial interest in medicine to a teacher but downplayed the subsequent role of teachers in her plans, explaining, I had a good science teacher, who made it really interesting. And Ive had good science teachers since then. Mostly good. But after that, I was already interested, so it didnt matter if there was a bad teacher. Jean never mentioned speaking to a school counselor and reported that she was not close to any teachers. Despite this, she knew the process for getting into both college and graduate school and had a clear plan for doing so, informed by her father and brothers, who were all working in the science and engineering fields.

Middle-class young women reported feeling close to some teachers at their high schools. However, this was primarily within the context of extracurricular activities or other intensive experiences at the school. Janice, for example, a middle-class biracial senior, reported feeling very close to her choir teacher, explaining,

I see him every day feels like. We all talk to him, we dont even use Mr. Tally. We call him, like, Tally, Tally-poo, What are we doing today? And we have all these inside jokes and everything with all the choirs, so its pretty hard to, like, be with a group every day and not become kind of a family.

Kerry, a White senior, reported knowing Mr. Greene, the advisor for the schools Junior State of America team, quite well. She saw him twice a week at JSA meetings, and he knew her older brother, who was in his AP Government class. Kerry said she hoped to be in his AP Government class the next year, Theres only two teachers, uh, Ms. Binder and Mr. Greene. Theyre both supposed to be good, but I know Mr. Greene better. In many cases like these, middle-class girls described close relationships with teachers. However, these relationships were largely described as different from the kind of advising experiences that working-class and poor girls looked for from their teachers.

Middle-class girls experiences were not all identical. Gail, for example, was a middle-class White senior who turned to her counselor for help in planning for college and a career in pharmacy. She reported turning to her counselor for help both to fill out recommendation forms after she had conducted research on colleges and scholarships she wanted to apply to, and to ask for help finding a pharmacy internship. Her counselor responded enthusiastically, filling out forms and sending an email to all teachers in the school to get a list of potential internships for Gail. Gail was unique, therefore, in seeing her counselor as a potential source of support. Unlike working-class and poor young women, however, she came prepared to her meetings and asked for specific resources. She also reported receiving extensive information and help from nonschool ties. Her parents were less involved; they provided general advice but did not help Gail plan for college. This may have been idiosyncratic, or it may have been because they were occupationally on the lower end of the middle-class spectrum; her mother had attended college and worked as an administrative assistant, and her father had completed college and was the general manager at an automotive company. Thus, their own social ties may have been more working-class than other middle-class families.

In summary, most of the upper- and lower-middle-class young women I interviewed reported a much larger circle of adults to whom they could turn to for help in college planning than working-class and poor girls. The advice and information they received through these ties were specific to their interests and often comprehensive. This made school-based ties less central to their planning process. This did not mean school-based ties played no role in their lives, but teachers were important primarily when participants had long-term and intense contact with them, such as being in extracurricular activities led by teachers. In these cases, the teachers were viewed as friends or trusted leaders rather than sources of guidance for the future.

COMPARING ACROSS CLASS

These findings reveal social class differences in the kinds of social ties young women use to plan for college. While working-class and poor youth primarily rely on school-based ties to inform their future plans, particularly for college, upper- and lower-middle-class girls preferred to talk to their parents or other close adults about their plans. This did not reflect differences in their knowledge of schools as potential sources of advice. In response to vignettes, middle-class girls were more likely than working-class and poor girls to suggest talking to someone at school for advice. Thus, for imagined others, they saw schools as a potential source of support. However, they were also more likely to report that they spoke with family members and professionals; school ties were situated within a broad network that middle-class girls possessed in their own lives and saw as potential resources for others.

Data from the interviews discussed earlier provide portraits of the ways young women use and do not use school-based ties to plan for college and how this differs by class. By using qualitative data, I am able to explore differences not only in the number of ties but also in the ways that young women talk about these ties and the resources they receive through them. These distinctions shed light on the disconnect between a literature that depicts working-class and poor teens as distrustful of and disconnected from institutional agents in schools, and a second literature that suggests these institutional agents may be the sole source of support for these young women. I find evidence of both; working-class and poor young women recognize the absence of resources in their own families and trust that schools are the source of such resources in the planning process. Middle-class girls, on the other hand, see the richness of the resources available to them through their families and communities and largely ignore school-based ties as a source of support. At the same time, however, disadvantaged youth are not embedded in long-term associations with institutional agents, are unaware of when and sometimes how to begin working with counselors to plan for college, and are often bogged down by the bureaucratic steps that intercede prior to college applications and enrollment. Thus, school-based ties fall short of these young womens hopes and expectations.

These qualitative findings suggest additional questions, such as whether they are generalizable to the population, including high school boys. The next section takes advantage of quantitative data to examine whether class differences in access to and use of institutional agents, particularly those that are school-based, are evident among a nationally representative sample of both young women and young men.

ANALYSIS: QUANTITATIVE DATA

Quantitative models were constructed based on the qualitative findings regarding social class and the use of social ties within schools. The dependent variables (total ties, total school ties, and percentage of ties that were school based) were constructed based on findings from the qualitative data about the role of school-based ties in young womens college planning and how this role was related to their broader set of networks. In addition, the use of informative ties (whom the students turned to) and encouraging ties (an indicator of a resource received) reflects the distinction in the qualitative findings between young womens reliance on ties and the resources they received through them.

Descriptive statistics compare sources of encouragement and information by socioeconomic quartile. This detailed information offers an overview of disparities in encouragement and sources of information. Next, I use hierarchical modeling to examine six broad outcomes: total encouragement received from various sources (0 to 6); number of types of encouraging school-based ties (0 to 3); percentage of encouraging ties who were school based (0 to 100); total number of sources of informative social ties (0 to 7); number of types of school-based informative ties (0 to 3); and percentage of informative ties who were school based (0 to 100). I use a linear approach for two outcomesthe percentage of encouraging ties who were school based and the percentage of informative ties who were school basedbecause these outcomes range from 0 to 100. I use a Poisson link for the remaining outcomes because they are count variables in which the mean was roughly equivalent to the variance. Coefficients in Poisson models are equivalent to the change in the log of the expected counts of the dependent variable for each 1-unit change in the independent variable, holding all other variables constant. Because this is difficult to interpret, I also provide the incidence rate ratio in the text, which is equivalent to the expected change in the incidence rate of the dependent variable for each 1-unit change in the independent variable. In these models, student-level outcomes are a function of student- and school-level predictors. Hierarchical modeling accounts for the clustering of students within schools. This was warranted because intraclass correlation coefficients indicated that the variance to be explained at the school level was substantial for encouragement-related outcomes (11%16%), although it was low for informative tie outcomes (4%5%). All linear independent variables were grand mean centered, although the unstandardized versions were retained for descriptive statistics. Analyses conducted using ordinary least squares and Poisson regression with correlated errors produced consistent results. Because the qualitative sample comprised only young women, I tested for gender differences in the degree to which social class background was associated with each outcome.⁸ Where differences were present, I separate models by gender.

QUANTITATIVE DATA FINDINGS

Table 3 presents descriptive statistics for the ELS sample. Respondents reported an average of 3.04 informative ties, of whom slightly under half were school-based ties, and 4.27 encouraging ties, of whom about the same proportion were school based. Sample demographic and school characteristics are representative of students and schools in the early 2000s.

Table 3. Weighted Descriptive Statistics for All Independent Variables (N = 11,658)

Student Characteristics	Mean/%	Standard Deviation	Range
Social Ties
Number of sources of informative ties	3.04	1.77	0 to 7
Number of school-based informative ties	1.36	0.83	0 to 3
% of informative ties from school	45.65	29.01	0 to 100
Number of sources of encouraging ties	4.27	2.06	0 to 6
Number of school-based encouraging ties	2.13	1.27	0 to 3
% of encouraging ties from school	44.29	26.38	0 to 100
Socioeconomic Quartile
Lowest quartile	21.47%		0 to 1
Second quartile	24.08%		0 to 1
Third quartile	26.34%		0 to 1
Highest quartile	28.11%		0 to 1
Race/Ethnicity
White	65.06%		0 to 1
Black	12.25%		0 to 1
Hispanic	13.79%		0 to 1
Other race/ethnicity	8.90%		0 to 1
Male	49.39%		0 to 1
Family Structure
Two bio/adopt parents	61.62%		0 to 1
Stepfamily or two guardians	16.23%		0 to 1
Single parent/other	22.15%		0 to 1
Immigrant Generation
First generation	7.74%		0 to 1
Second generation	9.90%		0 to 1
Third+ generation	82.37%		0 to 1
Grade Point Average	2.82	0.69	0.5 to 3.75
Enrolled in College Prep	54.50%		0 to 1
Received Award for Grades	53.25%		0 to 1
School Characteristics
% of School Receiving Free/Reduced Price Lunch	24.24	23.28	2 to 88
StudentTeacher Relations	0.03	0.39	-1.05 to 2.14
Academic Climate	-0.01	0.16	-0.63 to 0.27
Type of School
Public	91.93%		0 to 1
Catholic	4.76%		0 to 1
Other private	3.31%		0 to 1
School Location
Urban	27.11%		0 to 1
Suburban	52.10%		0 to 1
Rural	20.79%		0 to 1
Students Engage in College Planning Program
No, students do not do this	26.96%		0 to 1
Some students participate	58.51%		0 to 1
All students participate	14.53%		0 to 1
Students per Full-Time Teacher	19.65	8.14	6 to 54
Students per Full-Time School Counselor	346.14	157.09	100 to 900

Note. Dependent variables and grade point average obtained in 12th-grade survey; all other variables obtained from the 10th-grade survey except for the small number (N = 202) of refreshed sample students who were added to the sample in the second wave.

Table 4 presents bivariate social class differences in informative social ties and encouragement from these ties. Those in the highest two socioeconomic quartiles are significantly more likely to report encouragement from each source than the two lower quartiles, and there is also a significant advantage of being in the top quartile compared with the third quartile. This difference in encouragement is particularly large for coaches (42.9% increase from the lowest to highest quartile) and parents (22.5%), and slightly smaller for friends, counselors, and teachers. There is also a very large difference in the percentage of young people who do not report any encouraging ties, with 15% of the first and second quartiles reporting this, compared with only 6% in the highest quartile. Overall, those in the lowest quartile report 3.42 sources of encouraging ties, on average, compared with 3.53 in the second quartile, 3.93 in the third quartile, and 4.29 in the highest quartile. Socioeconomic quartile is also positively associated with the percentage of reported ties that are school based.

Looking at informative ties, students in the highest quartile are slightly more likely to report speaking to a school counselor about college plans than those in the lowest socioeconomic quartile (an increase of 3.9%) and no more likely to speak to a teacher. They are much more likely to speak to a coach (50.0%) or parent (69.8%), however. They are also more likely to talk to a sibling (15.4%). There are no socioeconomic differences in young peoples reliance on friends as a source of information. Those in the lowest quartile of SES are more likely to speak to a relative outside their immediate family for information regarding college than better-off youth. Finally, the likelihood of having spoken to no social ties about college decreases as class status increases. There is a positive relationship between social class and number of sources of social ties, but young people in the lowest socioeconomic quartiles report that a higher percentage of the ties they speak to about college are from their schools (teacher, school counselor, or coach) than those in the higher socioeconomic quartiles.

Table 4. Weighted Group Differences in High School Seniors Reports of Sources of Encouragement (N = 11,555) and informative ties (N = 9,675)

	Lowest SES Quartile	Second SES Quartile	Third SES Quartile	Fourth SES Quartile	Difference (1st to 4th)
Received Encouragement
Counselor	0.60	0.59	0.66a, b	0.71a, b, c	+.11 (18.3%)
Teacher	0.58	0.59	0.65a, b	0.70a, b, c	+.12 (20.7%)
Coach	0.42	0.47a	0.53a, b	0.60a, b, c	+.18 (42.9%)
Parent	0.71	0.72	0.81a, b	0.87a, b, c	+.16 (22.5%)
Relative	0.61	0.65a	0.72a, b	0.80a, b, c	+.19 (31.1%)
Friend	0.51	0.51	0.55a, b	0.63a, b, c	+.12 (14.3%)
None	0.15	0.15	0.10a, b	0.06a, b, c	-.09 (23.5%)
All sources	3.42	3.53	3.93a, b	4.29a, b, c	+0.87 (25.4%)
School-based	1.60	1.66	1.85a, b	2.01a, b, c	+0.41 (25.6%)
% school-based	36.18	36.39	38.41s, b	40.49a, b, c	+4.31 (11.9%)
Informative Social Ties
Counselor	0.77	0.78	0.80a	0.80a	+.03 (3.9%)
Teacher	0.46	0.45	0.43	0.43	-.03 (6.5%)
Coach	0.10	0.12	0.13a	0.15a, b	+.05 (50.0%)
Parent	0.43	0.54a	0.63a, b	0.73a, b, c	+.30 (69.8%)
Sibling	0.26	0.23	0.27b	0.30a, b	+.04 (15.4%)
Relative	0.32	0.28a	0.29a	0.28a	-.04 (12.5%)
Friend	0.52	0.51	0.52	0.54	+.02 (3.8%)
None	0.09	0.08	0.08	0.07a, b, c	-.02 (22.2%)
All Social Networks	2.85	2.91	3.07a, b	3.23a, b, c	+0.38 (13.3%)
School-based	1.32	1.36	1.36	1.38a	-0.06 (4.5%)
% school-based	48.06	47.27	45.03a, b	43.23a, b	-4.83 (10.0%)

^aSignificant compared with the lowest quartile at the 0.05 level. ^bSignificant compared with the second quartile at the 0.05 level. ^cSignificant compared with the third quartile at the 0.05 level.

Table 5 presents the results of hierarchical models predicting the six outcome variables. The encouragement-related outcomes are separated by gender because interaction coefficients suggested that SES matters differently for boys and girls when predicting encouragement from social ties. For brevity, I display only the coefficients for socioeconomic quartiles here. Appendix B presents full models with all control variables.

Results from Table 5 show that SES is positively associated with receiving encouragement for both boys and girls, although the association is stronger for boys (b = -0.16, IRR = 0.85, p < .001 for the lowest SES boys compared with the highest SES boys; b = -0.10, IRR = 0.91, p < .001 for the lowest SES girls compared with the highest SES girls). SES was also positively associated with reporting a greater number of encouraging school-based ties (b = -0.13, IRR = 0.87, p < .001 for boys in the lowest quartile compared with the highest and b = -0.07, IRR = 0.93, p < .01 for girls). Finally, boys from the two middle socioeconomic quartiles reported that a significantly smaller percentage of their encouraging ties came from school as compared with those in the highest quartile (b = -2.16, p < .05 and b = -2.40, p < .05). No significant socioeconomic differences were evident for girls.

Results from regression analyses of the number of social tie sources whom respondents turned to for information show that SES was positively associated with the number of ties students turned to for help, even after accounting for academic performance and institutional factors (b = -0.08, IRR = 0.93, p < .001 for the lowest socioeconomic quartile compared with the highest). However, there are no socioeconomic differences detected in the number of school-based ties respondents turned to for information. Furthermore, disadvantaged youth reported that a higher percentage of their sources of informative ties were teachers, administrators, or coaches. Both the lowest socioeconomic quartile and the second-lowest socioeconomic quartile reported that a significantly higher percentage of their informative ties were school based (b = 4.31, p < .001 and b = 4.03, p < .001, respectively).

Table 5. Hierarchical Regression Results of Adolescent Social Networks

	Encouragement From Social Ties						Informative Social Ties
	# of sources (Boys)	# of sources (Girls)	# of school sources (Boys)	# of school sources (Girls)	% school-based (Boys)	% school-based (Girls)	# of sources	# of school sources	% school-based
SES Quartile
Lowest quartile	-0.16***	-0.10***	-0.13***	-0.07*	-2.33	0.52	-0.08***	-0.01	4.31***
	(0.02)	(0.02)	(0.04)	(0.03)	(1.19)	(1.02)	(0.02)	(0.03)	(0.99)
Second quartile	-0.11***	-0.07***	-0.09**	-0.06*	-2.16*	-0.51	-0.05**	0.01	4.03***
	(0.02)	(0.02)	(0.03)	(0.03)	(1.05)	(0.92)	(0.02)	(0.03)	(0.87)
Third quartile	-0.04*	-0.01	-0.05	-0.00	-2.40*	0.64	-0.04*	-0.02	1.25
	(0.02)	(0.02)	(0.03)	(0.03)	(0.96)	(0.86)	(0.02)	(0.02)	(0.80)
Highest quartile (reference)	----	----	----	----	---	----	----	----	----
N	5,671	5,884	5,671	5,884	5,671	5,884	9,675	9,675	9,675
Regression model	Poisson	Poisson	Poisson	Poisson	Linear	Linear	Poisson	Poisson	Linear

Note. Models control for race/ethnicity, sex (informative tie models only), family structure, immigrant generation, grade point average, enrollment in college prep courses, award received for grades, percentage of school receiving free or reduced price lunch, scale of teacherstudent relations, scale of academic climate, type of school, school setting, student involvement in college planning programs, students per full-time teacher, and students per full-time school counselor. Standard errors in parentheses.

^*p < 0.05. ^**p < 0.01. ^***p < 0.001

DISCUSSION

Adolescents rarely make it to college unless they first have a plan to do so. In addition to families, communities, and friends, schools are key factors in these plans. Moreover, schools are publicly funded institutions that are intended to be an equalizing force in the United States, giving all young people the chance to pursue their dreams. Previous research has documented some of the troubling inequalities present within schools (Bettie, 2003; Bloom, 2007; Diamond et al., 2004; McDonough, 1997; Stanton-Salazar, 1997; Stanton-Salazar & Spina, 2000). The current study contributes to this scholarship by answering two key questions: How do adolescents class backgrounds shape how they rely on schools for help in planning for college, and how is social class related to what they receive from school personnel in return? Answering these questions elucidates the process by which social capital facilitates some adolescents plans while inhibiting others plans.

Both the qualitative and quantitative data support the contention that working-class and poor adolescents expect school-based ties, or institutional agents, to play a larger role in their college planning than middle-class adolescents do, yet they receive fewer resources. This is an important point because it suggests both an explanation and potential solution for the commonly observed replication of inequality despite publicly funded educational institutions that are intended to alleviate class-based disparities among students. It is not that disadvantaged students distrust school officials, but that they trust those officials to provide them with information and resources for college planning, either because they lack other sources of information or because they view these school-based ties as the best available sources.

This echoes work by others who have found that working-class and poor parents view schools as being responsible for college planning, whereas middle-class parents are more actively engaged in their childrens future plans (Lareau & Weininger, 2008; McDonough, 1997). It counters other work suggesting that distrust of school officials on the part of disadvantaged youth may bar them from asking for help (Bettie, 2003; Stanton-Salazar & Spina, 2000; Willis, 1981). This does not suggest that disadvantaged youth have the resources they need to plan for college. As the qualitative evidence demonstrates, disadvantaged adolescents need more from schools. In particular, they need to receive information about college planning much earlier, both to give them time to prepare for the college search and application process and so school counselors can accomplish bureaucratic tasks that may delay college discussions. Furthermore, the quantitative evidence suggests that although poorer youth rely on schools for advice, they are less likely to receive encouragement from them compared with their better-off peers, even when accounting for academic performance. This is particularly true of high school boys. These findings underscore the role schools play in perpetuating inequality, even while working-class and poor young people rely on them for help. Although schools have the opportunity to alleviate class disparities, they are thus far unsuccessful.

This study advances prior work by directly comparing the use of school ties vis-à-vis other family- and community-based social ties in order to demonstrate class differences in the positioning of schools relative to youths broader social networks. Furthermore, it focuses on adolescent intentions and behavior in using social networks to help plan for college. Prior work has usually examined the availability and actions of school staff in assisting students (Stanton-Salazar & Dornbusch, 1995) or adolescents families actions regarding college planning (e.g., Kim & Schneider, 2005; Lareau & Weininger, 2008; McNeal, 1999) rather than students relative likelihood of receiving encouragement and seeking out advice from school personnel compared with other social ties. Findings illuminate young womens assumptions regarding the role of schools in the college planning process. I compare my qualitative findings for young women with a larger sample of both boys and girls in high school in order to estimate broad trends in the resources young people receive and their likelihood of turning to school-based ties by social class.

One interesting implication of the qualitative data is that schools may have little to do with the college planning process for many high school students. Although working-class and poor youth rely on school counselors and teachers for help, several report receiving relatively little. Middle-class students largely eschew school counselors assistance when adults with specialized resources are available. This is a rational response to the college application process, which relies on a series of personalized essays, letters, and narrative extracurricular accomplishments alongside grade point average and standardized test criteria. Yet this highly personalized system relies on young peoples access to adults who are both invested in their future and knowledgeable about the application process, access that many young people do not have. This raises important questions about what the role of schools should be in the college planning process and what policies and resources are needed to fulfill this role.

There are several limitations to the current study. First, it would be ideal to obtain qualitative study participants and survey respondents who were identical in age and were interviewed and surveyed at approximately the same time; nested data (where interview participants are selected from a larger sample of survey respondents; see Small, 2011) would have facilitated this timing. However, rules regarding confidentiality preclude access to most large, nationally representative samples such as those used by the ELS. Although the qualitative sample is not identical in age or distribution across the country, the data from both components of the study add to our understanding of how young people use and receive resources from school-based ties. Similarly, the qualitative sample was limited by its sample of young women rather than both young women and young men. Interactions were used in analyses of the quantitative data to test whether this limitation may have affected the results. Results of these interactions suggest that the relationship between social class and informative social capital does not differ by gender. However, social class has a stronger association with receiving encouragement from social ties and the percentage of encouragement received through school-based ties for young men than young women. Thus, some social class differences may have been muted in the qualitative study of young women.

Young women were categorized into two large classes for the purposes of this study: middle class and working class/poor. This was due to striking similarities in access to and use of social ties between upper- and lower-middle-class young women, and similarities in the same processes among the working class and poor. This decision is similar to other qualitative work that compares the experiences of the middle class with those of the working class and poor (e.g., Bettie, 2003; Calarco, 2011; Lareau, 2011). However, combining these groups may mask some differences. Upper-middle-class parents have greater economic resources than lower-middle-class parents. Poor young womens lives were more chaotic than working-class young womens lives. Although these differences did not appear to matter in young womens access to and use of social ties, they might matter in other ways not observed in this analysis, or later in life.

Both qualitative and quantitative data are also obtained from the students perspectives. The acquisition of social capital depends on both the sources and recipients of resources, which in this case are the students and institutional agents (teachers, counselors, and administrators). This study largely reflects students orientations to and experiences with institutional agents and cannot speak directly to institutional agents motivations or actions (for research on guidance counselors perspectives, see M. M. Holland, 2015; Shamsuddin, 2016). Finally, the quantitative data are limited to counting one type of social tie only. Therefore, the number of types of social ties and percentage of social ties from school outcomes are rough approximations of students full social networks. Better measures could aid this studys precision. However, given similar results from both studies, better measures might not drastically change the findings.

Despite these limitations, the current study contributes to the literature on schools, social capital, and inequality by using a mix of methods to provide a deeper understanding of how social class differences emerge within educational institutions in the networks forged and used for college planning. Findings build on prior work on school-based social capital and adolescent social networks (e.g., Bloom, 2007; McDonough, 1997; Mehan et al., 1996; Stanton-Salazar, 1997; Weis, 1990), demonstrating that working-class and poor youth must (and often do) rely on schools for information regarding college planning, yet they receive fewer resources from these ties than do middle-class youth.

These findings have important policy implications for schools. Schools have the potential to minimize social class differences in access to resources (Mehan et al., 1996). In college planning, this could be accomplished through making information about the postsecondary education system available to all, thus lessening adolescents dependency on informal networks. Schools could demystify the college planning process, ideally by preparing students early for many of the administrative tasks they will face in applying for and entering college, on top of the academic preparation needed to complete college-level work.

Notes

1. This design eliminates potential issues that may arise in comparing goals across different fields (e.g., teacher vs. dancer). The health field offers a wide range of jobs that vary by prestige and pay, but all require some specialized training. This allowed me to compare job requirements with young womens educational and occupational plans. The health industry is the largest occupational industry in the nation, encompasses 10 of the 30 fastest growing occupations, and is projected to have the highest annual growth in employment through 2020 (Sommers & Franklin, 2012). Supplemental analyses with the National Study of Youth and Religion suggests that health was the field most frequently aspired to among young women.

2. As a sensitivity check, these respondents were retained and models rerun, with no change in the results.

3. Respondents, of course, may have spoken to more than one teacher or family member, and so on, and therefore these sums do not necessarily represent the number of ties used but the number of types of ties used. ELS did not include information about the total number of people adolescents turned to for information about college. Throughout the article, I use types or sources to indicate that these numbers represent categories of tie rather than the absolute number of ties.

4. Respondents were also asked about college representatives, college publications and guides, and library sources. These ties are not included in the summary measures because they are generally not indicators of social capital.

5. I conducted sensitivity analyses by substituting a linear measure of SES for the categorical indicators and by substituting parental education and household income for SES. Results (available on request) were consistent with the findings presented here. I use SES because it includes occupational information (which could not be used as a separate variable because the public data only include crude categories, whereas the SES variable was constructed using a detailed measure), and it offers a more streamlined approach for easily comparing groups. I use the categorical version of the measure both to reflect the categorical nature of social class as revealed in the qualitative analyses and to identify nonlinear associations between SES and the outcome variables.

6. Unfortunately, there is no measure of 10th-grade point average available in the data. This is potentially problematic because students could become discouraged with school and withdraw in response to not receiving encouragement. Thus, this variable may be endogenous to the outcomes and may mask some of the true associations between variables of interest and the outcomes. I retain grade point average because prior research has found that the likelihood of forming school-based ties is predicted by grade point average, rather than the reverse (Crosnoe et al., 2004). I also conducted sensitivity analyses where I dropped grade point average from the models. Findings did not change, and in fact, the associations between SES and the outcome variables were stronger in these models.

7. The research assistant coded two thirds of the transcripts, and the author coded all of them.

8. As an additional supplemental analysis, I tested for differences in the association between SES and each outcome by school characteristics (results available on request). No interactions were significant when predicting the use of informative ties. For boys, the positive association between SES and encouraging ties was stronger in suburban schools as compared with urban schools. For girls, schools with stronger studentteacher relations lessened the association between SES and encouragement. In both cases, however, these differences were quite small.

References

Allison, P. D. (2002). Missing data. Thousand Oaks, CA: Sage.

Barter, C., & Renold, E. (1999, Summer). The use of vignettes in social science research. Social Research Update, 25. Retrieved from http://sru.soc.surrey.ac.uk/SRU25.html

Bettie, J. (2003). Women without class: Girls, race, and identity. Berkeley: University of California Press.

Bloom, J. (2007). (Mis)reading social class in the journey towards college: Youth development in urban America. Teachers College Record, 109, 343368.

Bourdieu, P. (1986). The forms of capital. In J. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp. 241258). Westport, CT: Greenwood Press.

Briggs, X. (1998). Brown kids in White suburbs: Housing mobility and the many faces of social capital. Housing Policy Debate, 9, 177221.

Calarco, J. M. (2011). I need help! Social class and childrens help-seeking in elementary school. American Sociological Review, 76, 862882.

Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95120.

Covay, E., & Carbonaro, W. (2010). After the bell: Participation in extracurricular activities, classroom behavior, and academic achievement. Sociology of Education, 83, 2045.

Crosnoe, R., Johnson, M. K., & Elder, G. H., Jr., (2004). Intergenerational bonding in school: The behavioral and contextual correlates of student-teacher relationships. Sociology of Education, 77, 6081.

Diamond, J. B., Randolph, A., & Spillane, J. P. (2004). Teachers expectations and sense of responsibility for student learning: The importance of race, class, and organizational habitus. Anthropology & Education Quarterly, 35, 7598.

DiPrete, T. A., & Buchmann, C. (2013). The rise of women: The growing gender gap in education and what it means for American schools. New York, NY: Russell Sage Foundation.

Hardie, J. H. (2015). The best laid plans: Social capital in the development of girls educational and occupational plans. Social Problems, 62, 241265.

Holland, M. M. (2015). Trusting each other student-counselor relationships in diverse high schools. Sociology of Education, 88, 24462.

Holland, N. E. (2010). Postsecondary education preparation of traditionally underrepresented college students: A social capital perspective. Journal of Diversity in Higher Education, 3, 111125.

Kim, D. H., & Schneider, B. (2005). Social capital in action: Alignment of parental support in adolescents transition to postsecondary education. Social Forces, 84, 11811206.

Klevan, S., Weinberg, S. L., & Middleton, J. A. (2016). Why the boys are missing: Using social capital to explain gender differences in college enrollment for public high school students. Research in Higher Education, 57, 223257.

Lareau, A. (1987). Social class differences in family-school relationships: The importance of cultural capital. Sociology of Education, 60, 7385.

Lareau, A. (2000). Home advantage (2nd ed.). Lanham, MD: Rowman & Littlefield.

Lareau, A. (2011). Unequal childhoods: Class, race, and family life (2nd ed. with an update a decade later). Berkeley, CA: University of California Press.

Lareau, A., & Weininger, E. B. (2008). Class and the transition to adulthood. In A. Lareau & D. Conley (Eds.), Social class: How does it work? (pp. 118151). New York, NY: Russell Sage.

Lin, N. (1999). Social networks and status attainment. Annual Review of Sociology, 25, 467487.

Lin, N. (2000). Inequality in social capital. Contemporary Sociology, 29, 785795.

Lin, N. (2001). Social structure: A theory of social structure and action. New York, NY: Cambridge University Press.

Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data (2nd ed.). Hoboken, NJ: Wiley InterScience.

Marsden, P. V. (1987). Core discussion networks of Americans. American Sociological Review, 52, 122131.

McDonough, P. M. (1997). Choosing colleges: How social class and schools structure opportunity. Albany: State University of New York Press.

McDonough, P. M. (2004). Counseling matters: Knowledge, assistance, and organizational commitment in college preparation. In W. G. Tierney, Z. B. Corwin, & J. E. Colyar (Eds.), Preparing for college: Nine elements of effective outreach (pp. 6987). Albany: State University of New York Press.

McNeal, R. B., Jr. (1999). Parental involvement as social capital: Differential effectiveness on science achievement, truancy, and dropping out. Social Forces, 78, 117144.

Mehan, H., Villanueva, I., Hubbard, L., & Lintz, A. (1996). Constructing school success: The consequences of untracking low-achieving students. New York, NY: Cambridge University Press.

Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook. Thousand Oaks, CA: Sage.

Shamsuddin, S. (2016). Taken out of context: Piecing together college guidance information in urban high schools. The Urban Review, 48, 10122.

Small, M. L. (2011). How to conduct a mixed methods study: Recent trends in a rapidly growing literature. Annual Review of Sociology, 37, 5786.

Smith, C. 2003. Theorizing religious effects among American adolescents. Journal for the Scientific Study of Religion, 42, 1730.

Sommers, D., & Franklin, J. C. (2012). Overview of projections to 2020. Monthly Labor Review, 135, 320.

Stanton-Salazar, R. D. (1997). A social capital framework for understanding the socialization of racial minority children and youths. Harvard Educational Review, 67, 140.

Stanton-Salazar, R. D. (2011). A social capital framework for the study of institutional agents and their role in the empowerment of low-status students and youth. Youth & Society, 43, 10661109.

Stanton-Salazar, R. D., & Dornbusch, S. M. (1995). Social capital and the reproduction of inequality: Information networks among Mexican-origin high school students. Sociology of Education, 68, 116135.

Stanton-Salazar, R. D., & Spina, S. U. (2000). The network orientations of highly resilient urban minority youth: A network-analytic account of minority socialization and its educational implications. The Urban Review, 32, 227-260.

Stearns, E., & Glennie, E. J. (2010). Opportunities to participate: Extracurricular activities distribution across and academic correlates in high schools. Social Science Research, 39, 296309.

Sun, Y. (1999). The contextual effects of community social capital on academic performance. Social Science Research, 28, 403426.

Thomas, W. E., & Hickey, J. V. (2007). Society in focus: An introduction to sociology. Boston, MA: Allyn & Bacon.

Wacquant, L. J. D., & Wilson, W. J. (1989). The cost of racial and class exclusion in the inner city. Annals of the American Academy of Political and Social Science, 501, 825.

Weis, L. (1990). Working class without work: High school students in a de-industrializing economy. New York, NY: Routledge.

Willis, P. (1981). Learning to labor: How working class kids get working class jobs. New York, NY: Columbia University Press.

Wilson, W. J. (1987). The truly disadvantaged: The inner city, the underclass, and public policy. Chicago, IL: University of Chicago Press.

Woods, C. S., & Domina, T. (2014). The school counselor caseload and the high school-to-college pipeline. Teachers College Record, 116, 130.

APPENDIX A1: Interview Guide

This is an abbreviated version of the interview guide used for this project. Additional questions about family plans and work/family balance were included (but are not germane to the present analysis).

Introductory Questions

1.

As things stand right now, how far in school do you think you will get?

2.

What kind of job do you expect or plan to have when you are 30 years old?

Social Network Ties

For the next section, I would like to get a list of all the adults you know. As you list them, I will copy their names into this sheet, according to their relationship to you. Feel free to use initials or nicknames if you do not wish to give me their names.

3.

Please tell me the names of any adults that you know well and could turn to for advice if you needed it. [LIST]

a.

Prompts: What is this persons relationship to you? How often have you seen him/her in the past year? How long have you known him/her? How close do you feel to him/her? What does s/he do for a living (what kind of job)? How much education did s/he get?

4.

Probe for additional ties: Are there any [adult family members, friends of your parents, teachers or school administrators, neighbors, parents of your friends, and leaders of organizations] that you have not mentioned, that you know well and could turn to for advice?

a.

Same follow-up prompts about relationships.

Educational Aspirations

5.

You said that you would like to [leave school/finish high school/go to college/finish college/go to graduate school/get a(n) X degree].

a.

Why do you want to get this amount of education? What kind of jobs might be available to you after you do this?

b.

Have you always wanted to [go to college/etc.]? When did you start thinking about doing this? Who has encouraged you to do this? How have they encouraged you?

c.

Have you talked to anyone about your plans? Who? Tell me about that.

6.

What kind of college would you like to go to? Why?

a.

What do you know about that college? How did you decide?

b.

Have you gone to any meetings about applying to colleges? Have you visited any schools? Asked for brochures? Talked to someone who has been to a college like this? Tell me about your experience.

c.

How do you plan to get to college? What obstacles might you face?

d.

What have your parents done to help you prepare for college?

e.

Have you talked to anyone about how to choose a college? Who? Tell me about your conversation.

7.

How many of the adults you listed before have attained this amount of education? How much education do you think they expect you to get?

Occupational Aspirations

8.

You said that you would like to work as a _________ when you are 30 years old.

a.

Have you always wanted to do this? When did you start wanting to do this? What else have you wanted to do? Why did you change your mind?

b.

Can you tell me a little about this job?

c.

What do you think about, when you think of yourself as being age 30?

9.

Do you know anyone who does this kind of work?

a.

What is your impression of their work?

b.

Have you ever talked to him/her about his/her work? Tell me about that.

c.

Have you ever talked to any other adults about the kind of job you would like to have? Tell me about that.

10.

Do any of your friends have similar aspirations? Have you ever talked to him/her about it?

11.

Have you ever had a job or internship that might prepare you to have this kind of job? How did you get this job? What was your experience like?

Life History Calendar

Now I would like to get some information about your life from eighth grade to now, to get a sense of things you have done during that time. I will fill out this sheet with the information that you give me. Please feel free to make corrections as we go along. [Spreadsheet included topics in rows and years in columns. The topics included in rows were: age; grade; place of residence; living arrangements (people lived with); school (courses, academic programs, extracurricular activities); grades; goals (career, educational, school, other); work; friends; influential people; and significant events (awards and celebrations, challenges and losses).

Card Sort

12.

Each of these cards lists an event that many people go through in their lives, like marriage, leaving school, and things like that. Please take a look at them, and then make a pile of all the cards that describe a life event that you expect to have in your lifetime. There are two blank cards that you may use if there are other major events you expect to experience that are not listed on any of these cards.

a.

Cards: move out of parents home, leave high school, enter college, leave college, marry, have first child, begin working, stop working, move in with a boyfriend/girlfriend, buy a house, join the military, leave the military.

b.

Now please put these cards in the order you would like these things to happen.

c.

Please describe the sequence you chose, and why. (For each card) At what age do you think this should happen? Why?

d.

If you had to remove one of these cards from the list of things you would like to happen to you in the future, what would it be? Why?

e.

Which of these events is the most important to you?

Vignettes

In this last section, I will read you four short stories about four teenagers who are facing decisions about what they want to do with their lives. After that, I will ask you a few questions about what you think they should do, and what you would do in a similar situation. There is no right or wrong answer, just your opinion. Okay? [Order of vignettes randomized]

13.

Shannon is about to enter college. She has always wanted to be a doctor, and she has been accepted to a large state university with an excellent pre-med program. Recently, however, she has started to think about her desire to raise a family and have three kids. She is worried that being a doctor will not allow her to spend enough time with her children or her husband. She is considering enrolling in a nursing program instead.

14.

Jennifer is a junior in high school and hopes to go to college. She is interested in being an x-ray technician, but her parents have told her that they cannot pay for her tuition to college. Because of this, she is thinking about not applying to colleges in the fall.

15.

Alyce is entering her senior year of high school. She has received high grades throughout high school in all her courses and has taken many Honors-level courses. Her favorite subject is English, and she hopes to apply to several liberal arts colleges with strong literature programs. Her parents, however, are pressuring her to apply to larger universities and enroll in a pre-med program of study.

16.

Katrina is enrolled in her junior year of college, and she is taking courses to become a physicians assistant. She is enjoying college but misses her boyfriend, who decided to start working after high school in their hometown, which is more than a six-hour drive from her university. When she returned home from Christmas break, her boyfriend asked her to marry him, and she said yes. They made plans to get married the next summer. Katrina realized that she would have trouble planning the wedding from far away, and her fiancé wants them to live together next year. He asked her to leave college temporarily and return to school after the wedding at the local community college in her town.

APPENDIX A2: Coding list

Aspirations

"

Educational aspirations

"

Occupational aspirations

"

Family plans

"

Preparation for aspirations

Social Ties

"

Parents

"

School

"

Activity leaders

"

Family

"

Community

"

Friends

Support

"

Passive support

"

Active support

o

Provide information

o

Provide resources

o

Provide social ties

"

Mobilization

o

Participant solicited support

o

Tie offered support

o

Other tie/organization facilitated support

APPENDIX A3: Sample coding

Note. Codes for this section include: Occupational aspiration=OA, Preparation = Prep, Parents = P, Provide information = Info, Mobilized by tie = MT. Codes frequently overlap.

[Start OA] Interviewer: You said you were also considering x-ray technician?

Christina: Yea, we were, me and my dad was thinking about that.

Interviewer: How come?

Christina: Well, it pays well, and its, its not boring; its interesting.

Interviewer: Whats interesting about it?

Christina: Just that you can go in and take x-rays of people. And, I dont know; I just find it to be easy cause its stuff Im interested in, like, science and. [End OA]

Interviewer: Cool. [Start Prep] So, do you take a lot of science classes here?

Christina: I did. I havent took one this year. Im about to get my schedule changed to take Anatomy.

Interviewer: Oh, okay. For next semester, or?

Christina: Yea. [End Prep]

[Start P] Interviewer: Okay, cool. Um, okay. So, you said you and your dad were thinking about it?

Christina: Um-hmm.

[Start MT] Interviewer: Um, did he bring up the idea, or did you?

Christina: [Start Info] Yea, he brought up the idea cause we were trying, I was thinking about majoring in business. And, he found, I told him I wanted to do something that was just really, was out there, really exciting, so [unintelligible] an x-ray technician. And hes like, They pay well. Theres not a lot of them, but if you do get a job, it will pay really well. And its interesting. [End P, MT, and Info]

APPENDIX B: Supplemental Tables

Table B1. Full Hierarchical Linear Regression Results for Models Predicting Encouragement

	# of sources (Boys)	# of sources (Girls)	# of school sources (Boys)	# of school sources (Girls)	% school-based (Boys)	% school-based (Girls)
SES Quartile (ref=highest quartile)
Lowest quartile	-0.16***	-0.10***	-0.13***	-0.07*	-2.33	0.52
	(0.02)	(0.02)	(0.04)	(0.03)	(1.19)	(1.02)
Second quartile	-0.11***	-0.07***	-0.09**	-0.06*	-2.16*	-0.51
	(0.02)	(0.02)	(0.03)	(0.03)	(1.05)	(0.92)
Third quartile	-0.04*	-0.01	-0.05	-0.00	-2.40*	0.64
	(0.02)	(0.02)	(0.03)	(0.03)	(0.96)	(0.86)
Race (ref=White)
Black	0.24***	0.10***	0.24***	0.09*	4.42***	0.51
	(0.03)	(0.02)	(0.04)	(0.04)	(1.32)	(1.16)
Hispanic	0.09**	0.01	0.08*	-0.01	1.01	-1.16
	(0.03)	(0.03)	(0.04)	(0.04)	(1.35)	(1.24)
Other	0.03	-0.01	0.00	-0.02	-1.48	-0.27
	(0.03)	(0.02)	(0.04)	(0.04)	(1.30)	(1.21)
Family structure (ref=married bio/adopt)
Stepfamily/two-parent	-0.01	-0.04*	0.01	-0.04	0.80	-0.98
	(0.02)	(0.02)	(0.03)	(0.03)	(1.02)	(0.89)
Single parent	-0.00	-0.02	-0.01	-0.02	0.38	0.30
	(0.02)	(0.02)	(0.03)	(0.03)	(0.92)	(0.80)
Immigrant status (ref=nonimmigrant)
First-generation immigrant	0.07*	0.04	0.02	0.00	0.49	-0.77
	(0.03)	(0.03)	(0.05)	(0.04)	(1.60)	(1.46)
Second-generation immigrant	0.04	0.05*	-0.00	0.00	-0.16	-1.44
	(0.03)	(0.02)	(0.04)	(0.04)	(1.34)	(1.29)
Student GPA (12th grade)	0.16***	0.08***	0.17***	0.12***	3.84***	3.28***
	(0.01)	(0.01)	(0.02)	(0.02)	(0.69)	(0.66)
Student enrolled in college prep classes (10th grade)	0.18***	0.09***	0.19***	0.10***	4.76***	2.07**
	(0.02)	(0.01)	(0.02)	(0.02)	(0.77)	(0.67)
Received award(s) for grades (10th grade)	0.04*	0.05**	0.05	0.06*	1.44	1.59
	(0.02)	(0.02)	(0.03)	(0.03)	(0.85)	(0.84)
Percentage of school receiving free/reduced price lunch	-0.00	0.00	-0.00	0.00	-0.00	0.03
	(0.00)	(0.00)	(0.00)	(0.00)	(0.03)	(0.03)
Studentteacher relations	0.02	0.03	0.04	0.08	1.18	2.88
	(0.03)	(0.02)	(0.05)	(0.04)	(1.55)	(1.54)
Academic climate	0.19*	0.06	0.29*	0.14	5.65	4.26
	(0.08)	(0.06)	(0.12)	(0.12)	(3.83)	(4.07)
School type (ref=public)
Catholic	0.09**	0.07*	0.12*	0.09	2.96	3.06
	(0.03)	(0.03)	(0.05)	(0.05)	(1.89)	(1.98)
Other private	0.04	-0.02	0.10	-0.01	3.66	0.89
	(0.04)	(0.03)	(0.06)	(0.06)	(2.15)	(2.24)
Community (ref=urban)
Suburban	-0.04	-0.05**	-0.03	-0.05	-0.52	-0.77
	(0.02)	(0.02)	(0.04)	(0.04)	(1.25)	(1.27)
Rural	-0.12***	-0.06*	-0.14**	-0.06	-3.35*	-0.96
	(0.03)	(0.03)	(0.05)	(0.05)	(1.66)	(1.71)
School-based college planning program (ref=none)
Some students	0.01	-0.01	0.01	-0.02	0.88	-0.95
	(0.03)	(0.02)	(0.04)	(0.04)	(1.30)	(1.31)
All students	-0.02	-0.02	-0.04	-0.05	-0.16	-2.31
	(0.03)	(0.03)	(0.05)	(0.05)	(1.70)	(1.79)
Number of students per teacher	-0.00	-0.00	-0.01	-0.00	-0.12	-0.15
	(0.00)	(0.00)	(0.00)	(0.00)	(0.10)	(0.11)
Number of students per school counselor	0.00	0.00	0.00	0.00	0.00	0.00
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.01)
Constant	1.21***	1.39***	0.42***	0.581***	35.18***	38.70***
	(0.03)	(0.03)	(0.05)	(0.05)	(1.78)	(1.76)
School variance	0.03		0.08		2.30
	(0.00)		(0.01)		(0.05)
Individual variance					3.22
					(0.01)
Observations	5,671	5,884	5,671	5,884	5,671	5,884
Number of groups	719	719	719	719	719	719
Regression model	Poisson	Poisson	Poisson	Poisson	Linear	Linear

Standard errors in parentheses.

^*p < 0.05. ^**p < 0.01. ^***p < 0.001.

Table B2. Full Hierarchical Linear Regression Results for Models Predicting Informative Ties

	Informative Social Ties
	# of sources	# of school sources	% school-based
SES Quartile (ref=highest quartile)
Lowest quartile	-0.08***	-0.01	4.31***
	(0.02)	(0.03)	(0.99)
Second quartile	-0.05**	0.01	4.03***
	(0.02)	(0.03)	(0.87)
Third quartile	-0.04*	-0.02	1.25
	(0.02)	(0.02)	(0.80)
Race (ref=White)
Black	0.14***	0.16***	2.50*
	(0.02)	(0.03)	(1.08)
Hispanic	-0.01	0.03	1.53
	(0.02)	(0.03)	(1.12)
Other	0.09***	0.06	-1.10
	(0.02)	(0.03)	(1.07)
Male	-0.03*	0.05**	2.41***
	(0.01)	(0.02)	(0.61)
Family structure (ref=married bio/adopt)
Stepfamily/two-parent	-0.03	0.01	1.26
	(0.02)	(0.03)	(0.87)
Single parent	-0.03	-0.01	0.45
	(0.02)	(0.02)	(0.77)
Immigrant status (ref=nonimmigrant)
First-generation immigrant	-0.02	0.01	1.96
	(0.03)	(0.04)	(1.25)
Second-generation immigrant	0.02	-0.00	-1.09
	(0.02)	(0.03)	(1.08)
Student GPA (12th grade)	0.03*	0.05**	1.47*
	(0.01)	(0.02)	(0.60)
Student enrolled in college prep classes (10th grade)	0.07***	0.07***	1.46*
	(0.01)	(0.02)	(0.64)
Received award(s) for grades (10th grade)	0.07***	0.08***	1.23
	(0.01)	(0.02)	(0.74)
Percentage of school receiving free/reduced price lunch	-0.00	-0.00	0.03
	(0.00)	(0.00)	(0.02)
Studentteacher relations	0.04*	0.06*	0.98
	(0.02)	(0.03)	(1.09)
Academic climate	0.03	0.03	2.31
	(0.05)	(0.07)	(2.82)
School type (ref=public)
Catholic	0.06**	0.02	-0.32
	(0.02)	(0.03)	(1.30)
Other private	-0.07**	-0.07	2.17
	(0.03)	(0.04)	(1.51)
Community (ref=urban)
Suburban	-0.02	-0.04	-0.77
	(0.01)	(0.02)	(0.87)
Rural	-0.08***	-0.06*	1.82
	(0.02)	(0.03)	(1.18)
School-based college planning program (ref=none)
Some students	-0.00	-0.00	-0.24
	(0.02)	(0.02)	(0.92)
All students	0.03	0.03	-0.13
	(0.02)	(0.03)	(1.17)
Number of students per teacher	-0.00	-0.00	-0.08
	(0.00)	(0.00)	(0.06)
Number of students per school counselor	0.00	-0.00	-0.00
	(0.00)	(0.00)	(0.00)
Constant	1.09***	0.20***	40.06***
	(0.03)	(0.04)	(1.37)
School variance	0.00	0.00	30.25
	(0.00)	(0.00)	(0.45)
Individual variance	---	---	801.46
			(0.21)
Observations	9,675	9,675	9,675
Number of groups	695	695	695
Regression model	Poisson	Poisson	Linear

Standard errors in parentheses.

^*p < 0.05. ^**p < 0.01. ^***p < 0.001.

Cite This Article as: Teachers College Record Volume 120 Number 7, 2018, p. 1-49 https://www.tcrecord.org ID Number: 22296, Date Accessed: 10/21/2021 11:23:22 PM Purchase Reprint Rights for this article or review