Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

Applying to College: The Role of Family Resources in Academic Undermatch

by Josipa Roksa & Denise Deutschlander - 2018

Background/Context: While K-–12 research places parents at the center of understanding students’ educational outcomes, empirical analyses of academic undermatch, and transition into higher education more broadly, have focused primarily on students’ attitudes and behaviors. Family is implicitly present in the background but rarely brought to the fore. In this article, we integrate K-–12 and higher education literatures to illuminate how family social and cultural capital are related to the probability of academic undermatch and to social class inequality in this outcome.

Research Questions: We address three related questions: what is the relationship between family social and cultural capital and the probability of academic undermatch? To what extent is that relationship explained by students’ college-going attitudes and behaviors? Finally, how do family social and cultural resources contribute to social class inequality in academic undermatch?

Research Design: We use recent data from the Educational Longitudinal Study (ELS). ELS is a nationally representative sample of students who were 10th graders in 2002 and have been followed through the end of their high school education and into college. The analytical sample includes 5,370 students. The outcome examined is academic undermatch in college application, which occurs when a student applies to colleges at a selectivity level below the selectivity of colleges the student is academically prepared to attend.

Results: Family social and cultural capital play an important role in academic undermatch at the point of applying to college. More specifically, they influence students’ attitudes regarding what is important to consider when choosing colleges (such as college costs and living at home) and students’ college-going behaviors (primarily the variety of information sources consulted and the number of applications submitted). These variables collectively account for approximately 40% of the socioeconomic status (SES) gap in academic undermatch, net of controls. Moreover, we find no statistically significant interactions with SES, indicating that family resources, as well as specific attitudes and behaviors examined, benefit all students equally.

Conclusion: Students’ attitudes and behaviors related to college-going are deeply embedded in family contexts. Understanding academic undermatch, and college decisions more broadly, necessitates an explicit attention to family social and cultural resources, and mechanisms through which those resources are translated into specific educational outcomes. Policies and practices that aim to reduce social class inequality in college access would benefit from engaging parents, not only students.

Ample research has documented that students from socioeconomically disadvantaged backgrounds are less likely to enter higher education or attend more selective institutions (for reviews, see Deil-Amen & Turley, 2007; Grodsky & Jackson, 2009). This is the case even when researchers control for an array of academic preparation indicators, and even when students are highly academically prepared (Hoxby & Avery, 2012; Radford, 2013). Persisting inequality conditional on academic preparation has led researchers to turn their attention to “academic undermatch”—the likelihood that students attend institutions that are less selective than they are academically prepared to attend.  

Research on academic undermatch is nascent, and thus most studies have focused on describing the magnitude of the phenomenon, with very few studies examining mechanisms leading to undermatch. When mechanisms are considered, it is often in the form of considering whether students have completed particular steps on the path to college, for example, whether they have seen a high school counselor or applied to multiple institutions (e.g., Belasco & Trivette, 2015; Roderick, Coca, & Nagaoka, 2011). What is missing in this research is explicit attention to parental involvement, or more broadly, social and cultural resources embedded in students’ families.

We place family at the center of educational decision-making.1 Instead of considering only what students do, we trace the sources of their actions to family resources. More specifically, we propose that family social and cultural resources shape students’ attitudes and behaviors related to college-going, which in turn affect their outcomes, such as academic undermatch. We evaluate these propositions using data from the Education Longitudinal Study (ELS). The results confirm our propositions, indicating that students’ college-going attitudes and behaviors are deeply embedded in family contexts. Understanding academic undermatch, and college decisions more broadly, thus necessitates explicit attention to family social and cultural resources and the mechanisms linking those resources to student outcomes.  

This study makes several contributions to the existing literature. First, it is the first study of academic undermatch we are aware of that explicitly examines the role of family social and cultural resources. Second, very few studies in higher education in general, not just those examining academic undermatch, have considered the role of parental engagement (for recent exceptions, see Engberg & Allen, 2011; Perna & Titus, 2005; and Plank & Jordan, 2001). We extend this literature in two ways. We broaden the concept of family context by examining both cultural capital and a distinct form of social capital reflected in network closure. Moreover, we consider the mechanisms that connect family resources to student outcomes by examining the role of students’ attitudes and application behaviors. Presented findings have important implications for policy and practice. Helping students navigate the transition to college necessitates not only providing information and resources to students, but also engaging with their parents. Moreover, since social and cultural capital, which shape students’ attitudes and behaviors, are deeply embedded in the family, enhancing students’ educational outcomes will require a long-term collective effort.   


Research on academic undermatch, and entry into higher education more broadly, has typically focused on students—examining what students know and what steps they take on the path to college. This research has paid relatively little attention to the role of parents in the process of postsecondary entry. We thus begin by reviewing the recent research on academic undermatch, and then shift to the literature on K–12 education to make the case for the importance of two specific types of family resources: social and cultural capital. In the process, we also highlight how our study contributes not only to the research on undermatch, but also to the study of transition into postsecondary education more broadly.


Studying college transitions of Chicago Public School students, Roderick and colleagues (2008, 2011) first brought scholarly and policy attention to academic undermatch. In addition to outlining the leaky pipeline from college aspirations to enrollment, Roderick and colleagues noted that academically prepared students were not enrolling in colleges with selectivity levels that they were academically qualified to attend. Subsequent research using national datasets has confirmed that a substantial proportion of students undermatch by enrolling in institutions of a lower selectivity than they are academically prepared to attend (Belasco & Trivette, 2015; Smith, Pender, & Howell, 2013).2

The focus on undermatch rests in part on the prior literature indicating that students who attend more selective institutions are more likely to persist and graduate (e.g., Bowen, Chingos, & McPherson, 2009; Ishitani, 2006; Titus, 2004). Graduation is a crucial outcome given earning differentials between college graduates and high school diploma holders (Goldin & Katz, 2009). Moreover, some research indicates that graduates of more selective institutions have better labor market outcomes, particularly if they come from disadvantaged backgrounds (Dale & Krueger, 2011). The link between selectivity and labor market outcomes may in part operate through skill development. Although the literature is mixed (e.g., see the review in Pascarella & Terenzini, 2005), several recent studies show that selectivity is related to the development of critical thinking skills (Kugelmass & Ready, 2011; Roksa & Arum, 2015), and that those skills matter for early labor market outcomes (Arum & Roksa, 2014). Apart from the findings regarding selectivity, Bowen and colleagues (2009) report that students who undermatch are less likely to complete bachelor’s degrees and take longer to do so, and the effects are sizable: the graduation rate gap between those who undermatch and those who do not is 15 percentage points.

In addition, prior research indicates that students from socioeconomically disadvantaged backgrounds have a substantially higher probability of experiencing academic undermatch (Bowen et al., 2009; Deutschlander, 2017; Roderick et al., 2008; Smith et al., 2013). These findings highlight differential opportunities and decision-making processes of students from different socioeconomic backgrounds. This phenomenon thus warrants further investigation, as it potentially illuminates another form of stratification in higher education.   

Given the nascent character of the research on undermatch, most studies to date have focused on describing the phenomenon. When considering factors associated with undermatch, the literature has often focused on the steps students take on the path to college. Roderick and colleagues (2011) showed that students who applied to more institutions and completed FAFSA (Free Application for Federal Student Aid) were less likely to undermatch. Similarly, using a nationally representative dataset, Belasco and Trivette (2015) noted that seeing a high school counselor, completing FAFSA, and applying to multiple institutions reduced the probability of undermatch. In addition, student perceptions about the importance of expenses and living at home, were associated with whether students undermatched. Although these factors contribute to the socioeconomic inequality in academic undermatch, students from socioeconomically disadvantaged backgrounds continue to disproportionately undermatch even net of these factors (see Belasco & Trivette, 2015; Deutschlander, 2017). We consider whether family social and cultural capital can help to further illuminate these patterns.


Social and cultural capital have a long history in K–12 research. Building on Bourdieu’s work (Bourdieu, 1973; Bourdieu & Passeron, 1977), cultural capital is often understood as knowledge, dispositions and practices that facilitate successful interaction with dominant social institutions such as schools (see Lareau & Weininger, 2003). Recent work by Lareau (2011) has been particularly influential in shaping the study of cultural capital in schools. Studying elementary school children, Lareau  described social class variation in parenting practices. Middle class families engaged in what Lareau referred to as “concerted cultivation”: They worked to cultivate their children’s talents and abilities, and they were highly involved in schools (often questioning or challenging school decisions, as they considered themselves equal partners in the educational process). Several recent studies have aimed to quantitatively capture this complex set of processes and have reported a positive relationship between concerted cultivation and students’ academic success in K–12 education .

Similarly, prior research indicates that social capital is consequential for children’s educational success. K–12 research in particular has focused on Coleman’s notion of social closure. Coleman (1988, 1990) argued that communities characterized by intergenerational social closure—that is, communities in which parents and children are highly interconnected—are especially conducive to forming social capital among their members. Since connected parents are able to discuss their children’s activities, develop common evaluations of these activities, and exercise sanctions that guide and constrain these activities, they are able to supply sanctions and monitoring in support of pro-achievement norms (Coleman & Hoffer, 1987). Subsequent research has shown that social closure among parents is positively related to a range of students’ academic outcomes in K–12 education .

In the studies of transition into higher education, the boundary between social and cultural capital tends to be more blurred than in K–12 research. Scholars of higher education often focus on the question of college knowledge—i.e., whether students have the necessary information to navigate the college choice process (e.g., Kaufman & Gabler, 2004; Perna, 2000; Plank & Jordan, 2001). Indeed, in this context, the boundary of social and cultural capital is more difficult to determine as knowledge would typically be considered cultural capital, but knowledge is often obtained through social ties to others (i.e., social capital). Scholars in higher education thus often refer to social and cultural capital interchangeably or together. However, none of the higher education studies we are aware of formally examine the role of intergenerational social closure—the form of social capital most often examined in the K–12 literature. This is an important omission as social closure may be particularly influential in the college application process. Being a member of a tight network, which can provide information as well as structure norms and expectations, can have an important influence on whether and where students apply to college.

Moreover, very few studies of transition into higher education have considered the role of parental involvement or family social and cultural resources more broadly. The original college-choice model focused on students as autonomous actors (Hossler & Gallagher, 1987). Students were conceived of as progressing through different stages, from developing a disposition for college to engaging in the search process and subsequently making a choice as to where to apply and eventually enroll. Family is absent from this model. Although an updated college-choice model (e.g., Cabrera & La Nasa, 2000; Perna, 2006) explicitly includes family support and involvement, empirical research has lagged behind. Recent empirical research focuses on modeling students’ knowledge and activities. Families may be implicitly present in the background, but their role is not explicitly integrated in the understanding of students’ transition into college.

A few recent studies have begun to examine the role of family resources in the college-going process. Perna and Titus (2005), conceptualizing parental involvement as a form of family social capital, showed that different indicators of parent–student and parent–school involvement were related to the probability that a student would attend a four-year as well as a two-year institution, relative to not enrolling in college (see also Plank & Jordan, 2001). At the same time, focusing only on low-income students, Engberg and Allen (2011) reported that neither parent–parent nor parent–school involvement was related to whether students enrolled in two-year or four-year institutions. This study, however, found that other indicators of cultural capital, such as parental encouragement and cultural activities with children, were related to college-going in at least some of the models.

Although quantitative results are mixed, several recent qualitative studies have highlighted the importance of studying parental involvement in the transition to college. Following students from Unequal Childhoods, Lareau and Weininger (2008) reported that the concerned cultivation style of parenting, characteristic of the middle class, continued through the college search process. Middle class parents were involved in seeking information as well as visiting and selecting institutions. Working class and poor parents, on the other hand, were less engaged and often let children navigate the process more independently. Similarly, in case studies of 15 high schools, Bell, Rowan-Kenyon, and Perna (2009) found that family members were the primary sources of college information for most students. Although 11th graders relied on family less than ninth graders, parents were cited most frequently as sources of college information. Indeed, a recent study of high school valedictorians illuminated how students depend on their families to make crucial decisions about the transition to college, which disproportionately affects students from disadvantaged backgrounds whose families lack the necessary resources (Radford, 2013). We build on these insights to consider how family social and cultural resources are related to the probability of academic undermatch.  


In this study, we used recent data from the Educational Longitudinal Study (ELS). ELS is a nationally representative sample of students who were 10th graders in 2002 and have been followed through the end of their high school education and into college. The key outcome is academic undermatch, which occurs when a student applies to colleges at a selectivity level below the selectivity of colleges the student is academically prepared to attend.

The undermatch variable was created through a multistep process. First, colleges were assigned to a selectivity category. The restricted ELS data includes Barron’s categorization for each four-year institution.3 Barron’s categorizes schools based on the SAT/ACT scores, GPA, and class rank of accepted students, as well as the school’s admission rate. Students who apply to colleges ranked as “special” by Barron’s classification are excluded, since this is a general category for a variety of subject-specific schools that are difficult to classify. We collapsed Barron’s remaining categories into four selectivity categories: Very Selective (Barron’s Most Competitive and Highly Competitive), Selective (Barron’s Very Competitive), Less Selective (Barron’s Competitive), Nonselective (Barron’s Less Competitive and Noncompetitive), and included Two-Year Colleges as a separate category (indicator provided by ELS).

Second, following procedures outlined by Smith et al. (2013), we predicted the probability that a student would have access to an institution at each selectivity level. To do this, we used the application and admission data for each student provided by ELS. We constructed separate groups of students based on the selectivity level of the schools to which they applied. For example, all students who applied to very selective institutions were coded 1 if they were accepted and 0 if they were not. We used this information to run five probit analyses, one for each selectivity level, predicting whether a student who applied to a school at a specific selectivity level was admitted or not. These probit models included the following academic performance measures: honors-weighted GPA, ACT/SAT scores (similar to undermatch research by Bowen et al., 2009; Roderick et al., 2008), and student participation in AP/IB coursework (similar to Smith et al., 2013). Students who did not take the SAT/ACT were coded as 0, with an indicator noting that this substitution was made. Missing data on all academic measures were imputed using multiple imputation (with five imputed datasets). Based on the regression coefficients from these probit analyses and the individual academic performance measures for each student, we predicted individual students’ probability of being accepted to each selectivity level.

We used these predicted probabilities of being accepted to schools at different selectivity levels to define academic undermatch. To decide whether a student had access to a particular selectivity level, we used a 90% probability threshold (see also Bowen et al., 2009; Smith et al., 2013). In other words, students were grouped into selectivity categories based on a 90% likelihood of being admitted to that selectivity level. The 90% threshold provides a conservative estimate of undermatching since a student with an 85% likelihood of admittance to a school could feasibly gain admittance.

Finally, we compared the highest selectivity level that students were predicted to have access to with the highest selectivity level of the postsecondary institutions to which they applied. Students were then labeled as “matching” if the two selectivity levels were the same or if their predicted selectivity level was below that of the institutions they applied to (indicating the potential for overmatch). Students were coded as “undermatching” if their predicted selectivity level was higher than the institutions they applied to.

To define the sample for analysis of academic undermatch, we restrict it to students who were four-year eligible, meaning that they were classified as having access to at least one four-year institution. Smith and colleagues (2013) presented models with the same restriction. Others have restricted their analyses in similar ways—e.g., by restricting based on college aspirations (Roderick et al., 2008).4  A small proportion of students in the sample undermatched by not applying at all (1%). We conducted sensitivity checks by running models with and without these students, and all of the reported results are substantively identical. The analytical sample for the study includes 5,370 students.5

Although much prior research on academic undermatch considers enrollment, our analyses focus on application. One of the key issues in the process of academic undermatch is that some students do not even apply to selective institutions despite a high level of academic preparation (Hoxby & Avery, 2012; Roderick et al., 2011). If they do not apply to those institutions, they will inevitably undermatch at the point of enrollment, which makes application a crucial step in the process. Moreover, research on college choice indicates that socioeconomic inequality is particularly pronounced in the application phase (e.g., Cabrera & La Nasa, 2001; Roderick et al., 2011; Turley, Santos, & Ceja, 2007). In other words, students from socioeconomically disadvantaged backgrounds are more likely to attend less selective or two-year institutions than their more socioeconomically advantaged peers largely because of where they apply in the first place. Finally, focusing on the application phase of the college-going process is warranted given our attention to family social and cultural capital—if those resources matter, they should be most prominent in the application phase. After that, a number of other factors play a role, from whether students are accepted to whether and what kind of financial aid they receive. Moreover, “summer melt” research shows that some students who apply and are accepted still do not enroll in the fall (Castleman & Page, 2013). Studying enrollment combines all of those steps, including whether students applied, were accepted, and actually enrolled—some of which (e.g., acceptance and financial aid) are beyond the student’s or the family’s control. To understand how the family’s social and cultural resources matter in the transition to college, application is a crucial part of the process.

We run the models sequentially, starting with a model that includes family socioeconomic status (SES) and a range of controls. Next, we add family social and cultural capital. In the next set of models, we sequentially add students’ attitudes and the activities they pursue on their path to college. This sequential modeling allows us to consider the extent to which attitudes and behaviors explain the relationship between family social and cultural capital and undermatch, as well as to examine how much these different factors contribute to social class gaps in this outcome. Due to the impact of unobserved heterogeneity and rescaling effects (Mood, 2010), comparing coefficients across logistic regression models can lead to inaccurate conclusions. We have therefore also estimated average marginal effects, which confirm the described patterns and are available upon request from the authors.

Missing data was handled using a multiple imputation (MI) command in Stata with five imputed datasets. Following von Hippel (2007), we use the “multiple imputation, then deletion (MID)” method. In this approach, the dependent variable is used in the imputation equation but analyses are estimated on the non-missing values of the dependent variable. All models are also weighted and adjusted for clustering of students within schools using the SVY command.6


While there has been an increasing interest in academic undermatch in both academic and policy circles, efforts to estimate undermatch are not without challenges. Bastedo and Flaster (2014) raise three specific critiques of undermatch research. First, because undermatch measures rely on Barron’s categories (or a similar rating of institutions), undermatch estimates assume that selectivity differences matter across the entire range of postsecondary institutions. In contrast, Bastedo and Flaster argue that the differences in outcomes are observed primarily at the extremes (i.e., elite institutions or two-year colleges). Prior research indicates that selectivity is related to persistence and graduation, whether studies use categorical indicators (e.g., Ishitani, 2006) or continuous measures (e.g., Titus, 2004). Therefore, although the effects of selectivity may be stronger at the extremes, they are observed across the distribution. Moreover, national datasets cannot effectively capture outcomes of elite institutions due to the limited number of students attending those institutions, whether studies investigate college entry in general or undermatch in particular. Notably, the majority of students who undermatch in national datasets do so either by not enrolling or by enrolling in two-year institutions (Deutschlander, 2017). Estimates of enrollment undermatch using national datasets thus largely capture one end of the distribution—the end that Bastedo and Flaster (2014) argue is worthy of attention.

The undermatch procedure is based on a statistical model. In the social world, no statistical model provides a perfect (or even close to a perfect) prediction, which is a source of the second critique by Bastedo and Flaster (2014). Because no regression model of college outcomes, and by extension undermatch, provides perfect predictions, undermatch studies use a very high threshold (typically 90%).7 The strength of the undermatch approach is not in making perfect predictions but in presenting more compelling statistical comparisons. The 90% probability threshold does not mean that students will (or should) enroll in those institutions, but that they have a 90% likelihood of having access to a specific selectivity level based on their academic qualifications. An undermatch analysis thus compares outcomes of students who are academically qualified to enter particular types of institutions.

The final critique offered by Bastedo and Flaster (2014) is that undermatch research encourages greater reliance on test scores and assumes that fostering fit between students’ test scores and institutional selectivity will reduce inequality.8 Our reading of the undermatch research does not lead to these implications. Much research, often by the same authors, has shown that the “fit hypothesis” (the presumption that students perform best when attending institutions matching their abilities) is not accurate. In particular, there is convincing evidence that overmatching does not have any negative consequences for student outcomes (Bowen & Bok, 1998; Bowen et al., 2009; Kurlaender & Grodsky, 2013). Undermatch research does not argue for creating a tight fit between individual test scores and institutional selectivity. Instead, this research draws our attention to a specific aspect of the college enrollment process, and another aspect of inequality. If students are, statistically speaking, similarly qualified to attend a more selective institution, why are students from socioeconomically disadvantaged backgrounds less likely to do so? This is a social phenomenon that, like so many others (e.g., why students from disadvantaged backgrounds are more likely to attend two-year institutions or forego college altogether), deserves our attention because it offers insights into how students are sorted into institutions and how inequality is manifested across different dimensions of higher education.


Family socioeconomic status (SES) is a continuous measure developed by ELS. This measure includes parental education, parental occupational prestige (from the 1989 General Social Survey), family income, and an inventory of household items (such as newspaper, computer, Internet, DVD player, dishwasher, fax machine, etc.).

To understand students’ likelihood of undermatch, we begin by considering family social and cultural resources. In a recent review of the literature, Lareau and Weininger (2003) convincingly argue that focusing on highbrow cultural capital, such as going to a museum or concert, or taking arts classes, does not adequately capture the impact of cultural capital on educational outcomes in the U.S. context. Instead, the concept of cultural capital is better understood as a resource that can facilitate action within the education system, providing access to scarce rewards. In describing concerted cultivation, Lareau (2011) discussed several relevant factors, such as parent–child interactions and parent–school interactions. When considering parental involvement in higher education, scholars have tended to consider those factors independently—i.e., including responses to individual questions as variables in the models (e.g., Engberg & Allen, 2011; Perna & Titus, 2005; Plank & Jordan, 2001). In sociological studies of K–12 education, however, scholars often combine those indicators with the aim of operationalizing the concept of concerted cultivation (e.g., Bodovski & Farkas, 2008; Cheadle, 2008; Roksa & Potter, 2011).

In this study, we follow the latter approach as our interest is not to identify specific parental actions that may be related to undermatch, but rather to understand how family cultural capital, broadly conceived, may be related to this educational outcome. Even when entered in the models individually in prior research, specific indicators of parental involvement conceptually represent broader underlying preferences, tendencies, and approaches that facilitate children’s educational success. We created a measure of family cultural capital by combining information on educational resources in the student’s home, parent–child interactions, and parental interaction with the student’s high school. These questions were combined into a scale with a Cronbach’s alpha of 0.753.  

To define family social capital, we rely on Coleman’s (1990) definition of intergenerational social closure, capturing relationships between students, their friends, their parents, and their friends’ parents. This measure is based on surveys of both parents and students.  Parents were asked whether they know their child’s friends, and the parents of those friends. Students were also asked whether they know their friends’ parents and whether their parents know their friends’ parents. The questions were asked for students’ three closest friends. These questions were combined into a scale with a Cronbach’s alpha of 0.804.

The second broad category of variables of interest includes specific attitudes and actions characterizing the college transition process, namely students’ attitudes about the importance of different factors in the college choice process and specific activities they engage in as they pursue higher education. Prior studies have suggested that college cost (e.g., Paulsen & St. John, 2002; Tierney & Venegas, 2009) as well as distance from home (e.g., Belasco & Trivette, 2015; Turley, 2009) are important in the college-choice process, and that they are more prominent in deliberations of less advantaged students. We thus consider whether students reported that cost and distance (living at home) were very important in their college decisions.  

Second, we examine specific activities often regarded as important on the path to college: whether students consult with high school counselors, how many other sources of information they consult, and how many schools they apply to. Although the research on high school counselors is not entirely conclusive (Smith, 2011), a number of recent studies indicate that seeing a counselor is positively associated with transition to college (e.g., Belasco, 2013; Bryan, Holcomb, Moore, & Day, 2011; Robinson & Roksa, 2016). Belasco and Trivette (2015) also show that seeing a counselor is predictive of whether students undermatch.

In addition, counselors do not have a monopoly on information—students seek information through a number of different sources, which may inform their decisions about higher education. Students can seek information from other members of their school community, such as teachers and coaches. They can also seek information outside of school from their social networks (e.g., friends and relatives) or more general sources, such as college websites. Although the effects of specific sources of information on college-going are mixed, recent studies point to the importance of considering this broader array of information sources (e.g., Engberg & Allen, 2011; Galotti & Mark, 1994; Robinson & Roksa 2016). We thus include in our models the total number of sources of information students consulted, in addition to the counselor. The more information students consult, the better they may be positioned to make informed choices about postsecondary education.  

We also include in the models the number of applications students submitted. Some scholars have suggested that this measure reflects college knowledge (e.g., Hoxby & Avery, 2012), and prior studies of undermatch have reported that this measure is related to the probability of undermatch: Students who apply to more institutions are less likely to undermatch (e.g., Belasco & Trivette, 2015; Roderick et al., 2011).9 Correlations between key predictors, revealing expected relationships, are presented in Table 1.

Table 1. Correlations Between the Key Independent Variables










1. SES



2. Family cultural capital




3. Family social capital





4. College costs very important






5. Living at home very important







6. Number of information sources








7. Sought information from counselor









8. Number of college applications








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


In addition to these key indicators of interest, regression analyses control for students’ race (dummy variables for African American, Hispanic, Asian, and other racial/ethnic groups, with White students serving as a reference) and gender (dummy variable for females). We also include measures of several family characteristics: number of siblings, language spoken at home (dummy variable for non-English language spoken at home), and family structure (dummy variables for stepfamilies and other family types, including single-parent families, with two-parent biological families serving as a reference). We also control for academic preparation, including high school GPA, SAT scores, and total number of AP courses taken.

In addition to their families, students are embedded in broader social contexts (Perna, 2006; Perna & Thomas, 2008), among which schools have received the most attention. With respect to students’ transition into higher education, a number of studies have highlighted the importance of what has been termed “the college-going” culture, which reflects opportunity structures and resources that promote college-going (Engberg & Gilbert, 2014; McDonough, 1997; Roderick et al., 2011). We thus control for the proportion of students who attend four-year institutions at the students’ high school: high—75% or more students attending four-year institutions; moderate—50%–75% of students attending four-year institutions; and low (reference)—less than 50% of students attending four-year institutions. In addition, we control for several other school characteristics that may be related both to the college-going culture and students’ college transitions: school type (dummy variable for private high school); percent of students receiving free/reduced lunch; urbanicity (dummy variables for suburban and rural areas, with urban serving as a reference); percent of students who are racial/ethnic minorities; and region of the United States (dummy variables for South, Midwest, and West, with Northeast serving as a reference).


Overall, 18% of students undermatched at the point of applying to college, meaning that they applied only to schools with a lower level of selectivity than they were academically qualified for. In other words, they did not apply to a single institution that had a selectivity level at or above the level they were academically qualified for. Although students across SES groups undermatched, this outcome was much more pronounced for students from socioeconomically disadvantaged families: 24% of students in the lowest SES quartile undermatched and only 11% of students in the highest SES quartile did so.

These descriptive differences remain in multivariate analyses. The first model in Table 2 shows that students from socioeconomically advantaged backgrounds are less likely to experience academic undermatch. An increase of one standard deviation in SES lowers the odds of undermatching in the application process by nearly 25%. This is the case, even though the model controls for academic preparation, a host of other individual attributes, as well as high school characteristics.    

Table 2. Logit Models Predicting the Probability of Academic Undermatch in College Application



Model 1

Model 2

Model 3

Model 4

Model 5




Socioeconomic status (SES)












Family resources


   Family cultural capital











   Family social capital













   College costs very important









   Living at home very important











   Number of information sources







   Sought information from counselor







   Number of college applications









Individual-level controls


Race [Ref: White]


   African American




































   Other racial/ethnic group
























Family structure [Ref: two biological parents]














   Other family structure












Number of siblings












Non-English language at home




































Number of AP courses












School-level controls


Urbanicity [Ref: urban]


























Region [Ref: Northeast]






































Private high school












Free/reduced lunch (%)












Racial/ethnic minority (%)












College attendance [Ref: less than 50% attend four-year college]


   50%–74% attend four-year college












   75% or more attend four-year college

























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


Note: Analyses are weighted and adjusted for clustering of students within schools.

Note: SES and family social and cultural capital are standardized with a mean of 0 and standard deviation of 1.


n = 5,370. The number of cases is rounded to the nearest 10.


The second model indicates that family social and cultural capital play an important role in academic undermatch at the point of applying to college. Students embedded in families with more social and cultural capital are less likely to undermatch. The magnitude of the coefficients is particularly notable—cultural and social capital coefficients are approximately two thirds the size of the SES coefficient.10 Moreover, the coefficients for both family social and cultural resources are statistically significant, even though the two are moderately positively correlated (r = 0.312). Students thus benefit from intergenerational social closure, even net of their family’s cultural resources.   

The central question in this study is how these family resources translate into educational outcomes. More specifically, we consider whether family social and cultural capital are related to students’ attitudes and application behaviors, which in turn influence their educational decision-making and outcomes. Model 3 adds dummy variables for whether students regarded college cost and living at home as very important in their college choice process. Both of these measures are strong predictors of undermatch in the application phase. Students who regarded cost as very important had 32% higher odds of undermatching and those who believed it was very important to live at home had 1.7 times higher odds of undermatching. Including these measures in the model does not change the coefficient for family cultural capital, but it does reduce the coefficient for social closure by approximately 20%. Therefore, part of the influence of social closure, and the corresponding norms and expectations associated with closed social networks, is that it shapes what students believe is important on the path to college. Social closure thus forecloses opportunities for students in part by shaping their attitudes about what is important.     

Model 4 indicates that seeking college information from a high school counselor is not related to the probability of undermatching in the application phase. In a supplemental model, we found that seeing a counselor is related to undermatch when other sources of information are not included in the model. Students who seek college information from counselors also consult many more sources of information (see correlations in Table 1).11 Thus, in a model that controls for the number of sources of information consulted, seeing a counselor is not related to undermatch.  The number of sources of information students consult in their college search process is predictive of undermatch. For every additional source of information students consult, their odds of undermatching decrease by 7%.  

The number of applications students submitted as they applied to college is highly predictive of whether students undermatch—the more schools students apply to, the less likely they are to undermatch. Applying to one additional school decreases the odds of undermatching by over 60%. Submitting more applications represents a crucial step in increasing the likelihood that students will submit at least one application to a school with a selectivity level that matches their level of academic preparation. The three indicators of students’ behavior—consulting a counselor, the total number of sources of information consulted, and the number of applications submitted—explain just over 40% of the relationship between family cultural capital and academic undermatch, and render the family cultural capital coefficient statistically insignificant. This pattern of results indicates that while social networks are influential primarily through their relationship to students’ attitudes, cultural capital is mainly associated with students’ behaviors.   

The final model in Table 2 includes family cultural and social resources along with students’ attitudes and behaviors. In this final model, neither family social nor cultural resources are statistically significant. This indicates that family social and cultural resources influence academic undermatch largely through their relationship to students’ attitudes and behaviors.  

It is also worthwhile to note that the addition of behaviors in the final model substantially reduces the coefficient for students’ perceptions about the importance of staying close to home (by nearly 40%). This finding indicates that students’ actions, in terms of their application behaviors, are substantially related to their perceptions, especially the importance of staying close to home. Students who believe that it is important to stay close to home consult fewer sources of information and apply to fewer institutions (see correlations in Table 1). These underlying beliefs thus influence their application decisions, and subsequently, the probability of undermatch.  


In addition to examining whether and how family social and cultural capital are related to academic undermatch, we consider how different factors may contribute to social class gaps in this outcome. Although attention to social and cultural capital is theoretically rooted in concerns about social class inequality, most higher education studies report only the final model (i.e., the model controlling for all relevant factors), making it difficult to discern how much specific factors contribute to explaining the social class gaps in educational transitions. Presented models indicate that approximately 20% of the SES gap in application undermatch is related to family social and cultural capital (comparing Model 2 to Model 1). This may appear relatively low given extensive research on social and cultural capital, but this reduction is net of an extensive set of controls for academic, demographic, and high school factors. Moreover, the magnitude of the change is consistent with research in K–12 education (see estimates in Cheadle, 2008; Potter & Roksa, 2013).  

What students regard as important further contributes to social class gaps in academic undermatch, as do the actions students take on the path to postsecondary education. The SES coefficient drops slightly across each model. By the final model, the SES coefficient is reduced by nearly 40%. These patterns reveal the complexity of influences that lead students from socioeconomically disadvantaged backgrounds to undermatch in the application process. Some of the gap reflects differences in family social and cultural capital, and some of it reflects the steps students take on the path to college. The complexity of these social influences has implications for social policies aimed at reducing social class gaps in academic undermatch, a point that we address in the conclusion.

We have also considered the possibility that SES interacts with our key predictors. One could hypothesize that our key independent variables may differentially benefit students from different SES backgrounds. Higher SES students, for example, may get more specific information from the sources they consult. Our analyses, however, find no statistically significant interactions between SES and any of our key indictors. Students from more and less socioeconomically advantaged families thus seem to benefit equally from their family social and cultural capital in the application process. Similarly, consulting more sources of information and applying to more institutions benefit everyone equally. Finding that students from different SES backgrounds benefit to the same extent is consistent with a number of studies in K–12 education. These studies have argued that social and cultural resources can be valuable to all students, and indeed can provide an avenue of upward mobility for socioeconomically disadvantaged students [39_22150.htm_g/00001.wmf](De Graaf, De Graaf, & Kraaykamp, 2000; DiMaggio, 1982; Dumais, 2006).


While K–12 research places family at the center of understanding students’ educational outcomes, empirical analyses of the transition into higher education have focused primarily on students’ attitudes and behaviors. Family is implicitly present in the background, but rarely brought to the fore. In this study, we bring family to the center of educational decision-making by examining how family social and cultural resources are related to academic undermatch.   

The literature on academic undermatch is quite nascent, and thus most studies have focused on describing the magnitude of the phenomenon (Bowen et al., 2009; Smith et al., 2013).  A few studies that have begun to consider the mechanisms leading to undermatch have focused primarily on students’ attitudes and behaviors (Belasco & Trivette, 2015; Roderick et al., 2008). None of the studies we are aware of have considered how family resources, along with students’ attitudes and behaviors, affect undermatch. By considering family resources, and showing how they are related to the often-examined student attitudes and behaviors, we advance the understanding of the mechanisms leading to academic undermatch. These findings indicate that understanding the process of academic undermatch necessitates appreciating the embedded nature of students’ attitudes and behaviors, and the central role that family plays in shaping students’ educational decisions.  

While we focus on academic undermatch, the insights gained can be applied to studying students’ transitions into higher education more generally. Most research on higher education transitions has not explicitly examined the role of family resources. Although recent revisions of the college-choice model (e.g., Cabrera & La Nasa, 2000; Perna, 2006) include family support and involvement, empirical literature has yet to dedicate much attention to family resources. Only a few quantitative studies of the transition into higher education have examined the role of family social and cultural capital (Engberg & Allen, 2011; Perna & Titus, 2005; Plank & Jordan, 2001). These studies, however, do not consider the complex relationships examined herein—including different types of family resources (cultural capital and social closure), as well as students’ attitudes and behaviors.

Presented findings also have important implications for policy and practice. A recent large-scale intervention aiming to reduce academic undermatch among highly academically prepared students experienced very low take-up rates, meaning that most of the respondents targeted for the intervention did not participate (Hoxby & Turner, 2013).12 Our findings offer a plausible explanation as to why this may be the case: Students first and foremost rely on their family’s social and cultural resources for guidance and decision-making about college (see also Bell et al., 2009; Radford, 2013). External interventions may thus be quite difficult to implement unless they involve families in a more explicit and elaborate way.

Our analyses support the value of engaging with parents and emphasize the importance of doing so early. This has notable implications for high school counselors and college access organizations working with students from socioeconomically disadvantaged backgrounds. Since social and cultural capital, which shape students’ attitudes and behaviors, are deeply embedded in the family, influencing students’ college decisions—whether it means encouraging students to enroll at all or to consider a range of postsecondary institutions—will necessitate reaching out to students earlier than is currently typical, as well as engaging with their parents. Many counselors in public high schools, especially those attended by socioeconomically disadvantaged students, are already overburdened. In these cases, college access organizations can play an important role. College access organizations working with socioeconomically disadvantaged students have traditionally focused primarily on students. Their efforts are likely to be more effective if they include parents. The same holds for higher education institutions that aim to recruit students from socioeconomically disadvantaged backgrounds. College recruiters typically go to high schools to meet with students. But broadening their reach to also meet with parents, whether in high schools or community centers, could be more effective in recruiting, and perhaps subsequently retaining, students.

Similarly, college access interventions to date largely omit parents. Current programs typically focus on guiding students through various steps, such as choosing a list of schools, completing applications, writing essays, and completing FAFSA. Despite this focus on students, there is reason to believe that involving parents may increase the effectiveness of these programs. For example, Avery (2010) found that over one third of students who received college counseling did not follow through on all recommendations they received. One counselor remarked: “I pushed him towards Carnegie Mellon and he didn’t apply. He would have gotten in and I would have put him in touch with the minority recruiter, but it may have seemed too far away to him.” This suggests that students ignore advice because of preferences often developed with and among family members. Other college access programs have had limited to mixed success using financial incentives to motivate students (e.g., Carrel & Sacerdote, 2013). Including parents in the college exploration process may help ensure that students follow advice from high school counselors and consider a wider range of college options.

A small number of interventions that involve parents suggest that this may be a fruitful strategy. Bettinger Long, Oreopoulos, and Sanbonmatsu (2009) found that providing assistance to parents for completing FAFSA during the process of filing taxes was more effective than simply providing students with this information. Moreover, a qualitative evaluation of a college-transition program including parents indicated that the intervention encouraged parents to broaden their understanding of the steps involved in the college-choice process, develop a more proactive role in this process, and recognize the desirability and viability of college for their families (Auerbach, 2004). Finally, one effective high school intervention, unrelated to college, successfully engaged parents in decreasing high school student absenteeism by sending short messages from teachers to parents, in turn influencing parent–student conversations (Kraft & Rogers, 2015). Overall, these interventions provide suggestive evidence that engaging with parents may be effective in reducing inequality in college access and choice.

Moreover, college access activities—whether they include high schools, colleges, or college access organizations—would benefit from being comprehensive and considering the factors identified in this study. For example, individuals advising students on the transition to college could ask students the types of questions identified in this study regarding their attitudes and behaviors. By having explicit discussions with students and their parents about different preferences and dimensions of the college-choice process, they may be able to provide more appropriate and effective guidance. Attention to the various dimensions identified in this study is important, as social class inequality in college destinations cannot be addressed without a comprehensive approach that considers the complexity of the various factors influencing students’ postsecondary decisions. The foundation of this more comprehensive approach rests on engaging families in college decision-making and enhancing social and cultural resources at the family (as opposed to the individual) level.


The authors would like to thank Matea Pender and the ETS team for their assistance in developing the undermatch measure. In addition, we would like to thank the participants in the Center for Policy Analysis Seminar Series at Stanford University for their valuable feedback on an earlier draft of this paper. While working on this project, the second author was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant #R305B090002 to the University of Virginia. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.


1. Throughout the manuscript, the term “family” refers to “parents.”

2. The actual estimates of the proportion of students who undermatch depend on the definitions used (see a discussion in Rodriguez, 2015b).

3. ELS uses Barron’s Profiles of American Colleges to rank college competitiveness (Schmitt, 2009). Although not without criticism (e.g., Hess & Hochleitner, 2012), this resource is a common measure of school selectivity used by researchers in higher education (Bowen et al., 2009; Roderick et al., 2008).

4. Over 90% of students in our sample expected to earn a bachelor’s degree. Sensitivity analyses show that excluding those who did not expect a bachelor’s degree does not substantively alter the results.

5. Numbers of cases are rounded to the nearest 10 due to use of restricted data.

6. An alternative estimation strategy would involve using HLM. We are using the SVY command for several reasons. First, ELS is a stratified clustered sample, and the SVY command adjusts for both strata and clusters, producing robust standard errors. Second, as noted in the text, our analysis rests on comparing coefficients across models. The SVY command allows for estimation of average marginal effects, which we use as a robustness check for the reported patterns.

7. One could potentially improve the model by moving beyond academic factors (see Bailey, Belfield, Jenkins, & Kopko, 2015; Howell, Kumar, & Pender, 2015). However, no existing publically available dataset allows for that.

8. Changing demand without changing supply would necessitate realignment, but this is not unique to undermatch research. The same is the case for research comparing two-year to four-year institutions (and arguing that starting at a four-year institution is more conducive to completion), or for arguments regarding free college or calls for elite institutions to enroll more Pell-grant recipients.  

9. Another relevant factor sometimes considered in research on college-going is whether students completed FAFSA. If our models focused on less-advantaged students, FAFSA would be an important variable to consider (i.e., within SES categories, variation in undermatch would likely be substantially affected by FAFSA application). However, in a nationally representative sample, FAFSA application is negatively related to SES. Statistically speaking, this implies that higher SES students are at a disadvantage, when it actually reflects the fact that those students do not need to apply for FAFSA in the first place. When included in the models, FAFSA slightly increases the SES coefficients—inaccurately implying that higher SES students are disadvantaged in this step of the process. None of the other key predictors are affected by including FAFSA in the models.

10. Family social and cultural capitals as well as SES are standardized with a mean of zero and standard deviation of 1.

11. This measure, which captures whether students consulted with a counselor regarding college in the 12th grade, is also not correlated with SES. Students from more socioeconomically advantaged backgrounds have likely consulted with counselors earlier, thus eliminating the SES differences in the 12th grade.

12. While the work by Hoxby and Turner has garnered much attention, others have argued for the importance of examining undermatch among a wider range of students, not only those who are highly academically prepared (e.g., Rodriguez, 2015a).


Arum, R., & Roksa, J. (2014). Aspiring adults adrift: Tentative transitions of college graduates. Chicago, IL: University of Chicago Press.

Auerbach, S. (2004). Engaging Latino parents in supporting college pathways: Lessons from a college access program. Journal of Hispanic Higher Education, 3(2), 125–145.

Avery, C. (2010). The effects of college counseling on high-achieving, low-income students (No. w16359; NBER Working Paper No. 16359). Cambridge, MA: National Bureau of Economic Research.

Bailey, T., Belfield, C., Jenkins, D., & Kopko, E. (2015). Matching talents to careers: From self-directed to guided pathways. Washington, DC: American Enterprise Institute.

Bankston, C. L., & Zhou, M. (2002). Social capital and immigrant children’s achievement. Research in Sociology of Education, 13, 13–39.

Bastedo, M. N., & Flaster, A. (2014). Conceptual and methodological problems in research on college undermatch. Educational Researcher, 43, 93–99.

Belasco, A. S. (2013). Creating college opportunity: School counselors and their influence on postsecondary enrollment. Research in Higher Education, 54, 781–804.

Belasco, A. S., & Trivette, M. J. (2015). Aiming low: Estimating the scope and predictors of postsecondary undermatch. The Journal of Higher Education, 86, 233–265.

Bell, A. D., Rowan-Kenyon, H. T., & Perna, L. W. (2009). College knowledge of 9th and 11th grade students: Variation by school and state context. The Journal of Higher Education, 80, 663–685.

Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2009). The role of simplification and information in college decisions: Results and implications from the H&R Block FAFSA experiment (NCPR Working Paper). New York, NY: National Center for Postsecondary Research.

Bodovski, K., & Farkas, G. (2008). Concerted cultivation and unequal achievement in elementary school. Social Science Research, 37, 903–919.

Bourdieu, P. (1973). Cultural reproduction and social reproduction. In R. Brown (Ed.), Knowledge, Education and Cultural Change (pp. 71–112). London, England: Tavistock.

Bourdieu, P., & Passeron, J-C. (1977). Reproduction in education, society and culture. Beverly Hills, CA: Sage.

Bowen, W. G., & Bok, D. (1998). The shape of the river: Long-term consequences of considering race in college and university admissions. Princeton, NJ: Princeton University Press.

Bowen, W., Chingos, M., & McPherson, M. (2009). Crossing the finish line: Completing college at America’s public universities. Princeton, NJ: Princeton University Press.

Bryan, J., Holcomb, C., Moore, C., & Day, N. (2011). School counselors as social capital: The effects of high school college counseling on college application rates. Journal of Counseling and Development, 89, 190–199.

Cabrera, A. F., & La Nasa, S. M. (2000). Understanding the college-choice process. New Directions for Institutional Research, 107, 5–22.

Cabrera, A. F., & La Nasa, S. M. (2001). On the path to college: Three critical tasks facing America’s disadvantaged. Research in Higher Education, 107, 119–149.

Carbonaro, W. J. (1998). A little help from my friend’s parents: Intergenerational closure and educational outcomes. Sociology of Education, 71, 295–313.

Carrel, S., & Sacerdote, B. (2013). Late interventions matter too: The case of college coaching in New Hampshire (NBER Working Paper No. 19031). Cambridge, MA: National Bureau of Economic Research.

Castleman, B. L., & Page, L. C. (2013). A trickle or a torrent? Understanding the extent of summer melt among college-intending high school graduates. Social Science Quarterly, 95, 202–220.

Cheadle, J. (2008). Educational investment, family context, and children’s math and reading growth from kindergarten through the third grade. Sociology of Education, 8. 1–31.

Coleman, J. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95–120.

Coleman, J. (1990). Equality and achievement in education. Boulder, CO: Westview.

Coleman, J., & Hoffer, T. (1987). Public and private high schools. New York, NY: Basic Books.

Dale, S., & Krueger, A. (2011). Estimating the return to college selectivity over the career using administrative earnings data (NBER Working Paper No. 19031). Cambridge, MA: National Bureau of Economic Research.

De Graaf, N. D., DeGraaf, P. M., & Kraaykamp, G. (2000). Parental cultural capital and educational attainment in the Netherlands: A refinement of the cultural capital perspective. Sociology of Education, 4(1), 92–111.

Deil-Amen, R., & Turley, R. N. L. (2007). A review of the transition to college literature in sociology. Teachers College Record, 109(10), 2324-2366.

Deutschlander, D. (2017). Academic undermatch: How general and specific cultural capital structure inequality. Sociological Forum, 32(1), 162–185.

DiMaggio, P. (1982). Cultural capital and school success: The impact of status culture participation on the grades of U.S. high school students. American Sociological Review, 47, 189–201.

Dumais, S. A. (2006). Early childhood cultural capital, parental habitus, and teachers’ perceptions. Poetics, 34, 83–107.

Engberg, M. E., & Allen, D. J. (2011). Uncontrolled destinies: Improving opportunity for low-income students in American higher education. Research in Higher Education, 52. 786–807.

Engberg, M. E., & Gilbert, A. J. (2014). The counseling opportunity structure: Examining correlates of four-year college-going rates. Research in Higher Education, 55, 219–244.

Galotti, K. M., & Mark, M. C. (1994). How do high school students structure an important life decision? A short-term longitudinal study of the college decision-making process. Research in Higher Education, 35(5), 589–607.

Glanville, J. L., Sikkink, D., & Hernandez, E. I. (2008). Religious involvement and educational outcomes: The role of social capital and extracurricular participation. The Sociological Quarterly, 49(1), 105–137.

Goldin, C. D., & Katz, L. F. (2009). The race between education and technology. Cambridge, MA: Harvard University Press.

Grodsky, E., & Jackson, E. (2009). Social stratification in higher education. Teachers College Record, 111(10), 2347–2384.

Hess, F., & Hochleitner, T. (2012). College rankings inflation: Are you overpaying for prestige? Washington, D.C.: American Enterprise Institute.

Hossler, D., & Gallagher, K. S. (1987). Studying student college choice: A three-phase model and the implications for policymakers. College and University, 62, 207–221.

Howell, J., Kumar, A., & Pender, M. (2015). Academic match and fit: What can we learn from stated preferences, student actions, and college outcomes? Washington, DC: American Enterprise Institute.  

Hoxby, C., & Avery, C. (2012). The missing “one-offs”: The hidden supply of high-achieving, low income students (NBER Working Paper No. 18586). Cambridge, MA: National Bureau of Economic Research.

Hoxby, C., & Turner, S. (2013). Expanding college opportunities for high-achieving, low income students (Stanford Institute for Economic Policy Research Discussion Paper No. 12-014). Stanford, CA: Stanford University.  

Ishitani, T. (2006). Studying attrition and degree completion behavior among first-generation college students in the United States. The Journal of Higher Education, 77, 861–885.

Kaufman, J., & Gabler, J. (2004). Cultural capital and the extracurricular activities of girls and boys in the college attainment process. Poetics, 32, 145–168.

Kraft, M., & Rogers, T. (2015). The underutilized potential of teacher-to-parent communication: Evidence from a field experiment. Faculty Research Working Paper Series. Cambridge, MA: Harvard Kennedy School.

Kugelmass, H., & Ready, D. D. (2011). Racial disparities in collegiate cognitive gains: A multilevel analysis of institutional influences on learning and its equitable distribution. Research in Higher Education, 52, 323–348.

Kurlaender, M., & Grodsky, E. (2013). Mismatch and the paternalistic justification for selective college admissions. Sociology of Education, 86, 294–310.

Lareau, A. (2011). Unequal childhoods: Class, race, and family life. Berkeley, CA: University of California Press.

Lareau, A., & Weininger, E. B. (2003). Cultural capital in educational research: A critical assessment. Theory and Society, 32, 567–606.

Lareau, A., & Weininger, E. B. (2008). Class and transition to adulthood. In A. Lareau & D. Conley (Eds.), Social class: How does it work? (pp. 118–151). New York, NY: Russell Sage Foundation.

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

Mood, C. (2010). Logistic regression: Why we cannot do what we think we can do, and what we can do about it. European Sociological Review, 26, 67–83.

Pascarella, E. T., & Terenzini, P. (2005). How college affects students (vol. 2): A third decade of research. San Francisco, CA: Jossey-Bass.

Paulsen, M. B., & St. John, E. P. (2002). Social class and college costs: Examining the financial nexus between college choice and persistence. The Journal of Higher Education, 73, 189–236.

Perna, L. W. (2000). Differences in the decision to attend college among African American, Hispanics, and whites. The Journal of Higher Education, 71, 117–141.

Perna, L. W. (2006). Studying college choice: A proposed conceptual model. In J. C. Smart (Ed.), Higher education: Handbook of theory and research (Vol. 20, pp. 99–157). Dordrecht, Netherlands: Kluwer Academic Publishers.

Perna, L. W., & Thomas, S. L. (2008). Theoretical perspectives on student success: Understanding the contributions of the disciplines. ASHE Higher Education Report, 34(1). Hoboken, NJ: John Wiley & Sons.

Perna, L. W., & Titus, M. A. (2005). The relationship between parental involvement as social capital and college enrollment: An examination of racial/ethnic group differences. The Journal of Higher Education, 76, 485–518.

Plank, S. B., & Jordan, W. J. (2001). Effects of information, guidance, and actions on postsecondary destinations: A study of talent loss. American Educational Research Journal, 38, 947–979.

Potter, D., & Roksa, J. (2013). Accumulating advantages over time: Family experiences and class inequality in academic achievement. Social Science Research, 42, 1018–1032.

Radford, A. W. (2013). Top student, top school? How social class shapes where valedictorians go to college. Chicago, IL: University of Chicago Press.

Robinson, K., & Roksa, J. (2016). Counselors, information, and high school college-going culture: Inequalities in the college application process. Research in Higher Education, 57, 845–868.

Roderick, M., Coca, V., & Nagaoka, J. (2011). Potholes on the road to college: High school effects in shaping urban students’ participation in college application, four-year college enrollment, and college match. Sociology of Education, 84, 178–211.

Roderick, M., Nagaoka, J., Coca, V., & Moeller, E. with Roddie, K., Gilliam, J., & Patton, D. (2008). From high school to the future: Potholes on the road to college. Chicago, IL: Consortium on Chicago School Research at the University of Chicago.

Rodriguez, A. (2015a). A case for the study of average-performing students in college match.  Washington, DC: American Enterprise Institute.   

Rodriguez, A. (2015b). Tradeoffs and limitations: Understanding the estimation of college undermatch. Research in Higher Education, 15, 566–594.

Roksa, J., & Arum, R. (2015). Inequality in skill development on college campuses. Research in Social Stratification and Mobility, 39, 18–31.

Roksa, J., & Potter, D. (2011). Parenting and academic achievement: Intergenerational transmission of educational advantage. Sociology of Education, 84, 299–321.

Schmitt, C. (2009). NCES-Barron's admissions competitiveness index data files: 1972, 1982, 1992, 2004, 2008. National Center for Education Statistics.

Smith, V. (2011). Mediators of opportunity: High school counselors in the 21st century. Sociology Compass, 59, 792–806.

Smith, J., Pender, M., & Howell, J. (2013). The full extent of student-college academic undermatch. The Economics of Education Review, 32, 247–261.

Tierney, W., & Venegas, K. (2009). Finding money on the table: Information, financial aid, and access to college. The Journal of Higher Education, 80, 363–388.

Titus, M. A. (2004). An examination of the influence of institutional context on student persistence at 4-year colleges and universities: A multilevel approach. Research in Higher Education, 45, 673–699.

Turley, R. N. L. (2009). College proximity: Mapping access to opportunity. Sociology of Education, 82, 126–146.

Turley, R. N. L., Santos, M., & Ceja, C. (2007). Social origin and college opportunity expectations across cohorts. Social Science Research, 36, 1200–1218.

von Hippel, P. T. (2007). Regression with missing Ys: An improved strategy for analyzing multiply imputed data. Sociological Methodology, 37(1), 83–117.

Cite This Article as: Teachers College Record Volume 120 Number 6, 2018, p. 1-30
https://www.tcrecord.org ID Number: 22150, Date Accessed: 5/25/2022 12:50:58 PM

Purchase Reprint Rights for this article or review
Article Tools

Related Media

Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • Josipa Roksa
    University of Virginia
    E-mail Author
    JOSIPA ROKSA is Professor of Sociology and Education at the University of Virginia. Her research has centered on understanding the extent to which higher education amplifies, preserves, or reduces social inequality. In recent work, she has examined the role of cultural capital in facilitating student success, inequality in the development of critical thinking skills during college, the role of financial aid in fostering persistence of low-income students in STEM, and inequality in research skill development among graduate students. In addition to publishing in peer-reviewed journals in sociology and education, she is co-author of Academically Adrift: Limited Learning on College Campuses (University of Chicago Press, 2011) and Aspiring Adults Adrift: Tentative Transitions of College Graduates (University of Chicago Press, 2014).
  • Denise Deutschlander
    University of Virginia
    E-mail Author
    DENISE DEUTSCHLANDER is a Ph.D. candidate in the Department of Sociology and an Institute of Education Sciences Pre-doctoral Fellow in the Curry School of Education at the University of Virginia. Her research broadly investigates how family context and education intersect to enable mobility for some students and reproduce inequality for others. More specifically, her research currently explores the causal effect of advising interventions on student college choice, the effect of parent interventions on college student persistence, and the effect of universal pre-kindergarten on parental postsecondary educational attainment. Her most recent work, published in Sociological Forum, explored the heterogeneous nature of both family cultural capital and undermatch.
Member Center
In Print
This Month's Issue