Self-efficacy and Academic Achievement

Abstract

This paper investigates how neighborhood cultural context shapes academic achievement. Drawing on contemporary conceptualization of culture and recent evidence of the importance of neighborhood cultural context, I hypothesize that students from disadvantaged neighborhoods will display more cultural heterogeneity, or variability in cultural orientations, operationalized here as self-efficacy. In addition, I hypothesize that neighborhood cultural heterogeneity will have negative effects on academic achievement, particularly for those students with high individual self-efficacy. Results from multilevel models using data from the National Educational Longitudinal Study (NELS) support my hypotheses. Results indicate that students exposed to greater cultural heterogeneity are less able to translate high self-efficacy into better achievement, and suggest that exposure to increased cultural heterogeneity is one reason why students residing in disadvantaged neighborhoods underperform in education.

Keywords

education neighborhoods culture self-efficacy concentrated disadvantage poverty

While introducing a new initiative to expand educational opportunities to poor communities, President Obama remarked, “A child’s course in life should be determined not by the zip code she’s born in, but by the strength of her work ethic and the scope of her dreams” (Slack and Oken 2014). This statement alludes to the well-known association between educational outcomes and neighborhood context for U.S. students. Neighborhood effects research since the late 1980s has demonstrated the important role that neighborhood context plays in shaping educational success. Specifically, research has consistently shown that students from economically disadvantaged neighborhoods fare worse than students from more affluent neighborhoods in both educational attainment (e.g., high school graduation, college enrollment) and achievement (e.g., standardized test scores, grade point average [GPA]) even after controlling for family socioeconomic background (Bennett 2010; Cook et al. 2002; Crane 1991; Duncan 1994; Entwisle, Alexander, and Olson 1994; Fischer and Kmec 2004; Johnson 2010; Klebanov et al. 1998; Leventhal and Brooks-Gunn 2000; Mayer and Jencks 1989; Small and Newman 2001).

Given that the association between neighborhood context and educational outcomes is well established, scholars have recently begun to specify and empirically model the social processes that link neighborhood economic context to educational outcomes (Berg et al. 2013; Galster 2012; Harding 2011). For instance, two recent studies (Berg et al. 2013; Harding 2011) have pointed to cultural heterogeneity as an important aspect of disadvantaged neighborhoods that shapes academic outcomes. These studies have produced three main findings. First, they have shown that students in disadvantaged neighborhoods are exposed to more cultural heterogeneity, in each case, operationalized as variability in college goals. Second, these studies show that cultural heterogeneity has a negative effect on college enrollment. Third, these studies show that students in more heterogeneous neighborhoods are less likely to translate college aspirations into college enrollment and that heterogeneity is most consequential for students with strong college aspirations (Berg et al. 2013; Harding 2011). Thus, recent evidence indicates that neighborhood cultural heterogeneity plays an important role in shaping students’ academic outcomes and may have varying effects based on students’ individual cultural outlooks.

The current research builds on the existing studies in two specific ways. First, whereas existing studies have utilized college aspirations as a measure of cultural orientation at the individual and neighborhood level, this research operationalizes cultural orientations using self-efficacy. Self-efficacy is an ideal measure for this research, because it is a central component of the self-concept that is related to both neighborhood context (Boardman and Robert 2000; Christie-Mizell and Erickson 2007; Wilson 1996) and academic achievement (Ross and Broh 2000). Second, rather than looking at college enrollment, this study investigates how cultural heterogeneity shapes academic achievement. Although neighborhood effects studies of educational disadvantage have looked at both attainment (e.g., college enrollment) and achievement (e.g., test scores), the extant studies on cultural heterogeneity and educational outcomes have looked only at college enrollment.

To achieve these goals, this study uses data from the National Educational Longitudinal Study (NELS) to investigate the following three research questions:

Research Question 1: Do students in disadvantaged neighborhoods show more heterogeneity in self-efficacy?

Research Question 2: Does cultural heterogeneity have a negative impact on academic achievement?

Research Question 3: Does individual-level self-efficacy moderate the association between neighborhood cultural heterogeneity and academic achievement?

Background and Literature

The Cultural Context of Disadvantaged Neighborhoods

Since the late 1970s, deindustrialization and suburbanization in the United States have led to the loss of jobs and capital from urban areas across the United States, leaving behind cities with unprecedented concentrations of poverty and disadvantage in particular neighborhoods (Wilson 1996). The increased concentration of poverty and disadvantage has spawned a large body of neighborhood effects research, which has sought to understand how neighborhood-level characteristics can have unique effects on individual outcomes beyond the impact of individual-level characteristics (Galster 2012; Sampson, Morenoff, and Gannon-Rowley 2002; Sampson 2012; Sharkey 2008; Sharkey and Faber 2014; Small and Newman 2001; Wilson 1996). Empirical studies have linked neighborhood disadvantage to myriad individual-level outcomes such as criminal victimization, poor mental and physical health, lower chances of high school graduation, and increased rates of teenage childbearing (Galster et al. 2007; Harding 2003; Kikuchi and Desmond 2010; Kulis et al. 2007; Ross and Mirowsky 2008; Sampson, Raudenbush, and Earls 1997; Sharkey and Sampson 2010).

Theoretical accounts of neighborhood disadvantage often point to the cultural context of disadvantaged neighborhoods as potential mechanisms driving neighborhoods effects (Wilson 1996). For example, Wilson (1996:66) defines culture as “the sharing of modes of behaviors and outlook within a community,” and identifies self-efficacy as an important cultural characteristic that differentiates impoverished urban neighborhoods from other settings. Self-efficacy, also referred to as perceived sense of control (Ross and Broh 2000) or mastery (Christie-Mizell and Erickson 2007), is a central aspect of the self that “refers to people’s assessments of their effectiveness, competence and causal agency” (Gecas 1989:292). In other words, self-efficacy describes an individuals’ outlook regarding the ability to shape their own life through their actions.

Although self-efficacy is often investigated as an individual-level social-psychological variable (Gecas 1989), Wilson notes that self-efficacy is reinforced at the neighborhood level by cultural transmission, whereby residents’ own feelings are reinforced and strengthened by others who face similar structural conditions by virtue of living in the same neighborhood. Moreover, he argues that “exposure to certain attitudes and actions is so frequent that they actually become part of [an individuals’] own perspective” (Wilson 1996:78). This process of cultural transmission among neighborhood residents translates the aggregate individual self-efficacy of residents in impoverished neighborhoods into a neighborhood-level cultural phenomenon. Several empirical studies have supported the link between neighborhood disadvantage and lower self-efficacy at the individual level (Boardman and Robert 2000; Christie-Mizell and Erickson 2007; Sampson 2012).

Moreover, it is likely impoverished neighborhoods have more heterogeneity or variability in self-efficacy. According to Wilson, many individuals within impoverished neighborhoods lack access to stable employment and decent wages, which erodes their confidence that their actions can positively influence their socioeconomic situation. For example, Wilson cites individuals who hold long-term employment in low-wage jobs, yet remain in economically precarious situations—often only a single adverse event (e.g., serious illness) away from falling into abject poverty. For these individuals, their sense of powerlessness is a product of a lack of access to the resources and social institutions that enable individuals to overcome difficult circumstances. However, at the same time, individuals in disadvantaged neighborhoods also understand and acknowledge dominant cultural ideologies that link hard work and individual success. For instance, Wilson (1996) notes that despite facing structural constraints to economic mobility, many residents in poor neighborhoods verbally reinforce basic American values in terms of the importance of hard work and individual initiative in trying to get ahead economically.

In addition, individuals in such environments are likely to be exposed simultaneously to both mainstream modes of economic mobility and nonmainstream modes of mobility based on the illicit economy (Newman 2000). Thus, residents of such neighborhoods are exposed to contradictory evidence regarding whether hard work can indeed improve one’s life, because they see both economic precarity among individuals who work hard, and alternate models of economic mobility. This exposure to varied ideas and evidence about the link between individual effort and individual outcomes likely leads students in such neighborhoods to display more heterogeneity in their outlook about their ability to affect their own lives (Anderson 1994; Berg et al. 2013; Harding 2007, 2011; Merolla, Hunt, and Serpe 2011; Newman 2000; Pattillo 1999).

Hypothesis 1 (H1): Students in neighborhoods with higher amounts of concentrated disadvantage will show higher levels of heterogeneity in self-efficacy.

Neighborhood Cultural Heterogeneity and Educational Outcomes

Neighborhood effects research has also examined the association between neighborhood disadvantage and academic outcomes. The bulk of this research has shown that neighborhood economic context has unique effects on student outcomes, beyond the impact of individual-level varables.¹ However, one missing link in the current literature is empirical models of the social dynamics often assumed to trigger neighborhood effects on educational outcomes (Duncan 1994; Galster 2012; Mayer and Jencks 1989). Instead, most studies utilize only census-derived demographic measures of neighborhood characteristics (e.g., percent in poverty, percent unemployed) and far less research has empirically modeled how neighborhood cultural context may shape academic outcomes.

Two recent studies (Berg et al. 2013; Harding 2011) have presented empirical models of neighborhood cultural context and academic outcomes. In each case, the researchers have operationalized neighborhood cultural context as heterogeneity in college aspirations, and investigated how cultural context shapes the probabilities that students will attend college. The findings revealed that students from neighborhoods with greater heterogeneity in college goals were less likely to attend college, and that students living in neighborhoods with more heterogeneity in college goals were less likely to act on an aspiration to attend college. The authors of both studies surmise that in neighborhoods with more heterogeneity in college goals, students may be more willing to abandon their aspirations because such aspirations are less normative among their peers, and because alternatives to college enrollment were acceptable parts of students’ cultural outlook. Although these studies represent significant contributions, they are also limited by using college aspirations alone as a measure of cultural orientations. As Harding points out (2011:336), “The measure of college goals used here captures only one aspect of educational goals and is only rough proxy for the more complicated cultural models implied by the theory.” The current research seeks to build on these findings by exploring whether individual- and neighborhood-level self-efficacy show similar patterns as college goals.

As noted above, numerous scholars have pointed to differences in self-efficacy at the neighborhood level. Research also indicates that self-efficacy is an important precursor for academic achievement and attainment, with higher levels of self-efficacy leading to higher levels of educational success (Pajares 1996; Peguero and Shafer 2015; Reynolds et al. 2007; Ross and Broh 2000; Vargas Lascano et al. 2015). As Ross and Broh (2000:273) explain, students with higher levels of self-efficacy are likely to “accumulate resources and develop skills and habits that prevent avoidable problems . . .” Thus, self-efficacy is linked to educational achievement because students who have higher levels of self-efficacy are more likely to engage in behaviors that are conducive to high achievement. These students believe that that their academic achievements will be rewarded and lead to a better future. Conversely, students with low levels of self-efficacy are less likely to perform well in school because they believe that their own efforts are futile and that forces outside of their own control shape their outcomes. In turn, such students are less likely to engage in achievement enhancing behaviors. Hypothesis 2 formalizes the expectation in regard to the effect of individual-level self-efficacy on academic achievement.

Hypothesis 2 (H2): Self-efficacy will be positively related to academic achievement.

Furthermore, I argue that to better understand the associations between individual self-efficacy and academic achievement, it is necessary to understand the broader cultural context in which individual self-efficacy manifests. Much like neighborhood-level, socioeconomic status (SES) has independent effects on academic achievement net of individual- and family-level SES, I argue that neighborhood-level self-efficacy may also have effects net of individual-level self-efficacy. Neighborhood cultural processes are an inherently group-level phenomenon, and group-level cultural context represents the normative backdrop of students’ lives, and may have important consequences for individual-level outcomes. Increased neighborhood heterogeneity indicates more varied amounts of cultural outlooks within a community, indicating that interactions with others sends a “weak signal” to students that individual effort and achievement behaviors will lead to better overall outcomes (Berg et al. 2013; Harding 2011). This weak signal may lead students to doubt the utility of achievement related behaviors, which in turn could lead to less investment in their education and lower achievement.

However, neighborhood cultural context may not affect all students equally. To understand the effects of neighborhood cultural context, it is important to consider how the effects of neighborhood-level heterogeneity in self-efficacy may vary based on students’ individual outlooks. Specifically, I argue that cultural heterogeneity likely has a stronger, more negative effect among students with higher levels of individual-level self-efficacy. Because these students hold positive individual outlooks they may be particularly vulnerable to contradictory signals they receive from their neighborhood culture. These contradictory signals could instill doubts about the utility of hard work in securing a better future and undermine the role that their positive outlooks play for their achievement. In contrast, students with lower levels of self-efficacy are likely to be less affected by their neighborhood cultural context because they already express doubt in their own ability to shape their outcomes. Hypotheses 3 and 4 formalize my expectations in regard to the direct and moderating effects of cultural context on academic achievement.

Hypothesis 3 (H3): Neighborhood-level heterogeneity in self-efficacy will be negatively associated with academic achievement.

Hypothesis 4 (H4): The impact of neighborhood-level self-efficacy heterogeneity will become increasingly negative as individual self-efficacy increases.

Method

Data and Sample

The data for the current project are drawn from the NELS. The NELS data come from a nationally representative panel study of 1988 eighth-grade students. The National Center for Education Statistics (NCES) also provides a census based dataset that links NELS panel members to their 1990 residential Zip Code Areas (ZCA).² For the current analyses, I draw on data from the 1988 and 1990 waves. The analyses presented below include 8,100 black, Hispanic, and white students from 1,140 ZCAs. All individual-level data were imputed using multiple imputation (m = 10), and all results are averaged over the ten imputed datasets.³ All results are weighted by the NELS panel weight.

Measures

The outcome measure for the multilevel models presented below is tenth-grade (1990) academic achievement measured as the average of the NELS mathematics, reading, science, and history standardized test scores. The combined measure of academic achievement is useful because, similar to college entrance exams and GPA, it represents a holistic picture of students’ academic performance during their senior year of high school. In addition, the composite measure is similar to commonly used assessments such as the Peabody Individual Achievement Test (PIAT), Armed Forces Qualification Test (AFQT), and the California Achievement Test (CAT; see Sirin 2005). The alpha reliability of the composite achievement score is .923.

The main individual-level independent variable is self-efficacy. The self-efficacy measure combines five items from the 1990 survey. Items gauged respondents’ level of agreement with the following statements: (1) “I don’t have enough control over my life”; (2) “Good luck is more important than hard work”; (3) “Whenever I am getting ahead somebody or something stops me”; (4) “When I make plans, I am certain they will work out”; and (5) “Chance and luck are very important for my life.” All items are coded so that higher numbers correspond to higher levels of self-efficacy. The alpha reliability of the self-efficacy measure is .723.

Models presented below control for a set of individual-level variables known to be associated with academic achievement. Due to the design of the NELS data, these variables come from either the eighth- or tenth-grade survey depending on the students’ participation status. Given the well-known relationship between race, socioeconomic background, and educational achievement, I control for race, parents’ education, family income, and occupational status. Two binary variables black and Hispanic capture students’ racial backgrounds, with white students serving as the reference group. Female compares female and male students. Parent has BA degree is a dichotomous item that is equal to “1” if one or both of the student’s parents have at least a bachelor’s degree. Family income is measured with three dichotomous variables that roughly correspond to the bottom three income quartiles in the NELS data; the top income quartile serves as the reference group for all models. White collar is a dichotomous item that indexes students who have at least one parent with a white-collar occupation. I code Manager/Administrator, Professional, Proprietor/owner, and teacher as white-collar occupations. Number of Siblings is a continuous measure of siblings that is censored at six or more siblings. Single Mother Household is a dichotomous item that compares students who live in a household with only their mother to those who live with their father or with both parents. Finally, English is First Language is a dichotomous item that compares students who spoke English as their first language to students who spoke a language other than English as a child.

The main neighborhood-level independent variable gauges neighborhood cultural context. Self-efficacy Heterogeneity is the within neighborhood variance of the individual-level self-efficacy measure. Alternate models were estimated using a neighborhood aggregate measure derived from a three-level mixed effects model in which the items within each scale (level I) were nested within each respondent (level II), and the respondents are nested within each neighborhood group (level III; cf. Raudenbush and Sampson 1999; Harding 2007). This approach produced coefficients that were substantively identical results to those presented here, but generally with higher standard errors. Due to the similarity between these approaches, I chose to present the more direct measurement approach. Models also control for Mean Self-efficacy, which represents the neighborhood-level mean, again produced by aggregating the individual scores on self-efficacy among students in each ZCA.⁴

The analyses presented here focus on the role of neighborhood context for students’ educational achievement. Obviously, school context represents a potential confounder. However, there is evidence that school and neighborhood context are distinct in the current data. The NELS data are cross-classified by schools and neighborhoods, meaning that some students from the same neighborhood attend different schools, and some students in the same school reside in different neighborhoods. For instance, the 8,100 students in the current sample attend 575 schools; however, of the 575 schools, only 101 have students from just one ZCA. Thus, the cultural context in most of the schools reflects students from multiple neighborhoods and school-level cultural context represents a separate concept from the neighborhood-level measure used here. The correlations between the neighborhood context measures and parallel measures calculated at the school level ranged from .38 to .51, indicating that the neighborhood cultural context, while associated with school cultural context, is an independent concept. In addition, models that utilized numerous school-level control variables (not shown, available) showed identical patterns as the results presented here. All of this evidence indicates that school cultural context, while potentially important for student outcomes, is discernable from neighborhood cultural context, which is the focus in the present research.

I use Concentrated Disadvantage as an additional neighborhood-level independent variable. I measure Concentrated Disadvantage as the first principle component from a factor analysis of percent in poverty, percent female-headed households, percent on public assistance, percent unemployed, percent professional or managerial employment, and percent in college. I multiplied percent in professional/managerial employment and percent college degree by −1, and standardized each variable prior to estimating the factor analysis. The factor analysis results are displayed in Table 1. Table 1 shows that all of the indicators of disadvantage load highly on the first component, and each item has a communality above .5. The first factor explained 81 percent of the variance across these items. Although some authors have used “percent black” as an additional marker of concentrated disadvantage, I do not include racial composition items in the concentrated disadvantage scale to separate the effects of concentrated disadvantage and racial composition separately. Percent black and concentrated disadvantage are correlated at .507.

Table 1.

Means and Standard Deviations for All Analysis Variables.

Variables	Range	M/proportion	SD
Individual-level variables (N = 8,100)
Academic achievement composite	32.32–76.00	51.08	9.21
Self-efficacy	1.00–4.00	2.95	0.56
Black (1 = Yes)	0–1	0.10
Hispanic (1 = Yes)	0–1	0.12
Female	0–1	0.51
Parent has BA degree (1 = Yes)	0–1	0.30
Family income Q1 (1 = Yes)	0–1	0.26
Family income Q2 (1 = Yes)	0–1	0.30
Family income Q3 (1 = Yes)	0–1	0.19
White collar (1 = Yes)	0–1	0.35
Number of siblings	0–6	2.27	1.63
Single mother household (1 = Yes)	0–1	0.15
English is first language (1 = Yes)	0–1	0.90
Zip Code Area–level variables (N = 1,140)
Mean self-efficacy	2.05–3.83	2.94	0.25
Self-efficacy heterogeneity	0.02–1.14	0.50	0.19
Concentrated disadvantage	0–1	0.00	1.00
Urban (1 = Yes)	0–1	0.29
Southern region (1 = Yes)	0–1	0.35
Proportion black	0–0.99	0.11	0.19
Proportion homeowner	0.04–0.98	0.70	0.15
Proportion foreign born	0–0.71	0.07	0.11
Proportion Hispanic	0–0.97	0.09	0.18

Source. National Educational Longitudinal Study (NELS).

Models below also control for six additional neighborhood-level variables that may be associated with cultural context and achievement. I use Proportion black, Proportion Hispanic, and Proportion Foreign Born to investigate the role of racial composition for neighborhood cultural context. Southern is a dummy variable comparing ZCAs in the southern region of the United States to students from all other regions. Urban is a dummy variable comparing ZCAs from central cities within the United States to students from suburban and rural areas. Percent homeowner is a census measure of the percentage of neighborhood households in which a household member is a homeowner. This measure is used to control for residential stability in neighborhoods, as homeowners are likely to live in the same neighborhoods for longer periods than nonhomeowners are. All neighborhood-level variables are mean centered in the multilevel models presented below.

Results

Table 2 presents means and standard deviations for all analysis variables. Table 2 shows that students generally report high levels of self-efficacy ( $\bar{x}$ ). The sample is 51 percent female, and consists of 10 percent black students, 12 percent Hispanic students, and 78 percent white students. In terms of parental education, 30 percent of sample members have at least one parent with at least a bachelor’s degree. The income quartile variables roughly correspond to quartiles in the NELS sample. 35 percent of all respondents had at least one parent with a white-collar occupation, and the average number of siblings is 2.27. Finally, 15 percent of students hail from a single mother household, and 90 percent counted English as their primary language. Looking at the neighborhood-level measures, 29 percent of neighborhoods were in urban areas, and 35 percent were in the South. The average neighborhood consisted of 11 percent black residents, 70 percent homeowners, 7 percent foreign-born residents, and 9 percent Hispanic residents.

Table 2.

Factor Analysis for Concentrated Disadvantage Scale.

Variables	Standardized factor loading	Communality
Percent in poverty	.899	.809
Percent female-headed household	.810	.656
Percent on public assistance	.915	.837
Percent unemployed	.878	.770
Percent professional/managerial^a	.780	.622
Percent college degree^a	.736	.541

Source. National Educational Longitudinal Study; N = 1,140 Zip Code Areas; Scale alpha reliability = .915.

Reverse coded.

Concentrated disadvantage and neighborhood cultural context

To test H1, Table 3 presents an ordinary least squares (OLS) model that regresses the neighborhood cultural heterogeneity on concentrated disadvantage and the neighborhood-level control variables. Results support H1, as higher levels of concentrated disadvantage are associated with higher levels of self-efficacy heterogeneity (b = .022, p < .05). Only two other neighborhood-level control measures are significant; neighborhoods in the South (b = .024, p < .05) have greater heterogeneity, and neighborhoods with a higher percent homeowners (b = −.136, p < .01) have less cultural heterogeneity. All of these patterns support the ideas described above regarding the cultural context of disadvantaged neighborhoods, indicating that students in disadvantaged neighborhoods display more heterogeneity in self-efficacy. The multilevel models estimated below seek to determine if these differences in cultural context have implications for academic achievement.

Table 3.

Ordinary Least Squares Regression Model for Neighborhood-Level Cultural Heterogeneity.

Independent Variables	b
Mean self-efficacy	−.020
	(0.023)
Self-efficacy heterogeneity	—
	—
Concentrated disadvantage	.022*
	(0.010)
Urban (1 = yes)	−.011
	(0.015)
Southern region (1 = yes)	.024*
	(0.010)
Percent black	−.042
	(0.041)
Percent homeowner	−.136**
	(0.050)
Percent foreign born	.017
	(0.089)
Percent Hispanic	.008
	(0.054)
R ²	.029

Source. National Educational Longitudinal Study.

Note. N = 1,140 Zip Code Areas.

p < .05. **p < .01. ***p < .001, (two-tailed test).

Cultural context and academic achievement

Prior to estimating the multilevel models of academic achievement, I estimated unconditional (ANOVA) models to provide baseline estimates of the neighborhood- and individual-level variance components and to assess the degree to which academic achievement varies by neighborhood. These models (not shown) produced an intraclass correlation coefficient (ICC) of .241, indicating that 24 percent of the variance in academic achievement was attributable to differences between neighborhoods.

Table 4 presents results of the multilevel models for academic achievement. Model 1 shows that black (b = −3.633, p < .001) and Hispanic (b=−2.948, p < .001) students lag behind their white counterparts in academic achievement. In addition, consistent with H3, self-efficacy has the expected positive effect on academic achievement (b = 3.124, p < .001). Model 1 also shows that SES variables generally have the expected results. Students whose parents have at least a bachelor’s degree outperform students whose parent have less education (b = 3.776, p < .001). Moreover, students from the lowest (b=−2.484, p < .001) and second lowest (b=−0.724, p < .01) income quartile have lower achievement than students from the highest income quartile. Students in the third income quartile were not significantly different than those in the top income quartile. Students with white-collar parental occupation (b = 1.258, p < .001) also had higher achievement. In terms of family structure, siblings are associated with lower achievement (b=−.277, p < .001), but students living in mother only households did not have significantly different achievement from students living in other family arrangement. Finally, there was no difference between those students who spoke English as their first language and those students who did not.

Table 4.

Multilevel Models for Academic Achievement.

Independent Variable	Model 1	Model 2
Independent Variable	b	b
Individual-level variables
Black (1 = yes)	−3.633***	−3.687***
	(0.382)	(0.382)
Hispanic (1 = yes)	−2.948***	−2.927***
	(0.380)	(0.380)
Female	−0.776***	−0.798***
	(0.166)	(0.166)
Self-efficacy	3.124*	1.921***
	(0.168)	(0.328)
Parent has BA degree (1 = yes)	3.776***	3.791***
	(0.230)	(0.233)
Family income Q1 (1 = yes)	−2.484***	−2.49***
	(0.307)	(0.307)
Family income Q2 (1 = yes)	−0.724**	−0.723**
	(0.261)	(0.260)
Family income Q3 (1 = yes)	−0.088	−0.088
	(0.266)	(0.266)
White collar (1 = yes)	1.258***	1.231***
	(0.211)	(0.211)
Number of siblings	−0.277***	−0.275***
	(0.055)	(0.055)
Single mother household (1 = yes)	0.191	−0.182
	(0.257)	(0.256)
English is first language (1 = yes)	0.520	−0.475
	(0.356)	(0.356)
Neighborhood characteristics
Mean self-efficacy	0.967	0.485
	(0.543)	(0.555)
Self-efficacy heterogeneity	−1.549*	−1.568*
	(0.684)	(0.684)
Urban (1 = yes)	0.443	0.420
	(0.309)	(0.310)
Southern region (1 = yes)	−1.142***	−1.137***
	(0.269)	(0.269)
Concentrated disadvantage	−1.633***	−1.609***
	(0.256)	(0.256)
Percent black	1.302	1.270
	(0.967)	(0.968)
Percent homeowner	1.344	1.350
	(1.119)	(1.120)
Percent foreign born	−2.219	−2.291
	(1.935)	(1.937)
Percent Hispanic	2.239	2.135
	(1.205)	(1.206)
Cross-level interaction effects
Self-efficacy × Self-efficacy Heterogeneity		−4.087***
		(0.958)
Variance components
Neighborhood variance	6.8406	6.786
Residual variance	50.967	50.795
Total variance explained	28.90%	29.20%

Source. National Educational Longitudinal Study (NELS); N = 8,100 Students, 1,140 Zip Codes.

p < .05. **p < .01. ***p < .001, (two-tailed test).

Turning to the neighborhood-level variables, H2 is supported, as self-efficacy heterogeneity has the expected negative effect on academic achievement (b = −1.549, p < .05). Two additional neighborhood-level variables have significant effects on academic achievement. First, neighborhoods located in the U.S. South have lower levels of achievement (b = −1.142, p < .001). Second, as expected, neighborhoods with higher levels of concentrated disadvantage have lower levels of mean achievement (b = −1.633, p < .001). Neighborhood mean self-efficacy (b = .967, p > .05) does not have a significant effect on academic achievement.

Model 2 adds the interaction effect between neighborhood self-efficacy heterogeneity and individual self-efficacy. Consistent with H4, the interaction effect (b = −4.087, p < .001) indicates that neighborhood heterogeneity has a stronger effect among students with higher levels of individual self-efficacy. This negative interaction effect also indicates that neighborhood cultural heterogeneity acts to attenuate the effects of individual-level self-efficacy. Figure 1 presents this pattern visually. As Figure 1 shows, among students with low self-efficacy, there is little variation in academic achievement based on neighborhood context. However, among students with high levels of self-efficacy, those in low heterogeneity neighborhoods substantially outperform their counterparts from more heterogeneous neighborhoods. Figure 1 also shows that the effect of neighborhood heterogeneity is most pronounced among students with high levels of self-efficacy, because students in more heterogeneous neighborhoods get the smallest gain in achievement for increased self-efficacy. In other words, students in highly heterogeneous environments struggle to translate positive individual outlooks into better academic achievement.⁵

Figure 1.

The effect of self-efficacy on academic achievement by neighborhood cultural heterogeneity.

Discussion and Conclusion

This research presents several new findings about the relationship between neighborhood cultural context and academic achievement. First, as hypothesized, results indicate that students in neighborhoods with more concentrated disadvantage show more heterogeneity in self-efficacy. This result adds to the body of evidence indicating that one characteristic of disadvantaged neighborhoods is that such environs show greater variability in cultural outlooks relative to more affluent areas (Berg et al. 2013; Harding 2007, 2011). Although such arguments have become a central part of contemporary scholarship on the nature of contemporary urban poverty, the bulk of evidence in favor of this view has come from ethnographic research (see Anderson 1994; Edin and Kefalas 2011; Newman 2000; Wilson 1996), and the existing quantitative studies have focused on college aspirations alone as an indicator of cultural orientations. By documenting comparable patterns using self-efficacy, a central aspect of self-concept that has been theoretically linked to neighborhood disadvantage, this research contributes additional evidence of greater cultural heterogeneity in disadvantaged neighborhoods.

Most scholars have attributed increased cultural heterogeneity in disadvantaged areas to increased exposure to varied styles of life. Students in impoverished neighborhoods may question their self-efficacy due to the difficult structural conditions they face as a consequence of living in areas that lack ample employment opportunities and other institutional supports. However, even in such neighborhoods, students also understand mainstream ideas about the centrality of hard work for success, and they likely see examples of individuals who were able to overcome difficulty circumstances through individual action. Thus, students in such areas are likely exposed to conflicting evidence regarding the value and power of individual effort to overcome difficult circumstances, leading to more heterogeneity in individual-level self-efficacy (see Merolla et al. 2011). Through the process of cultural transmission, this heterogeneity in individual self-efficacy becomes part of the cultural fabric of students’ neighborhoods.

Moreover, the findings here show that there is a negative effect of cultural heterogeneity on academic achievement. This finding supports theoretical accounts of neighborhood effects on educational outcomes that imply that neighborhood differences in academic achievement are due in part to differences in social organization between neighborhoods of different economic levels (Galster 2012). As noted, relatively few studies have presented empirical models of such influences on academic outcomes, and to my knowledge, no extant empirical studies have linked cultural neighborhood context and academic achievement. Thus, the results presented here lend additional support to arguments that posit that the social organization of impoverished neighborhoods have consequences for students’ outcomes, beyond the effect of individual student characteristics.

However, the findings indicate that cultural heterogeneity is mainly consequential for students with high individual-level self-efficacy. That is, students with positive individual-level beliefs are most affected by a lack of normative consensus at the neighborhood level. It is likely that the diversity of outlooks regarding individual action as a way to overcome difficult structural conditions affects these students by imparting uncertainty about their own individual-level belief. In other words, in environs where students encounter more self-efficacy heterogeneity, they are less likely to translate their own individual outlooks into better academic achievement because these individual outlooks are undermined by the cultural context of their neighborhood. These findings underscore the importance of specifying why and for whom neighborhood effects matter (Sharkey and Faber 2014). This approach to understanding the interplay of individual and neighborhood factors helps move beyond a dichotomous yes/no approach to understanding neighborhood effects. As shown here, it is likely neighborhood cultural contexts affects students differently based on their own cultural outlooks. Thus, future research should continue to consider the multiplicative effect of neighborhood cultural context.

I argue that the findings presented here have implications for long-standing debates surrounding the role that culture plays in academic disparities by race and class. Policy makers and educators often believe that there is a cultural component to the underperformance of poor and minority students (cf. Lewis 2012). Often these arguments insinuate that these students have cultural outlooks that either do not “value” educational success or downplay the link between academic success and future economic outcomes. However, the vast majority of evidence indicates that students from all backgrounds tend to value education and consider it important (Harris 2011; Jones and Lou 1999; Merolla 2014). Relatively few studies have sought to look beyond individual-level measures of values to the cultural context of students’ lives, an oversight that is particularly surprising because culture is an inherently collective phenomenon that develops out of patterns of social interaction among individuals in particular environments (Wilson 1996). The results from the current research show that empirically examining culture as a collective phenomenon and accounting for the interaction between collective and individual cultural outlooks can add insight to understandings of how culture shapes academic achievement. In addition, most of the arguments surrounding the role of culture for student underperformance are based on approaches that conflate culture and values, rather than contemporary definitions that concieve of culture as modes of behaviors and outlooks.

Despite a tendency to conceive of culture as an individual phenomenon, another important insight from the current research is that much like neighborhood poverty or unemployment, neighborhood cultural context arises from the same structural forces that lead to high concentrations of poverty in specific geographic areas. As such, to mitigate the disadvantages that students in such areas experience, policies should focus on reducing the concentration of poverty and disadvantage in neighborhoods in the United States. The heterogeneous cultural forms that develop in disadvantaged environments are often the result of necessity and circumstance rather than cultural deviance (Newman 2000; Wilson 1996). Thus, the disadvantage that students face in such environments is not due to a culture of poverty but rather the concentration of poverty, which leads to exposure to cultural heterogeneity and in turn leads to lower levels of academic achievement.

Limitations

Although this research adds several new findings to the literature, the results should be understood in the context of several important limitations. First, the data used in this paper are somewhat dated. The NELS data are generalizable to the eighth-grade class of 1992. Individuals in the NELS sample are currently about 38 years old. Thus, the results presented here may not be applicable to current students, and research with a more contemporary sample is important.

Second, due to the nature of the dataset, the current study operationalized neighborhood as ZCA. This operationalization leads to a very rough approximation of neighborhoods as ZCAs can be quite large, and in some rural areas may represent entire counties. However, my findings replicate other studies that have used census tracts as operationalization of neighborhoods (Berg et al. 2013; Harding 2011). Hence, while there are obvious limitations to this operationalization, the evidence indicates that the findings presented here are likely to be reproduced using alternate levels of aggregation.

Third, the NELS data do not come from a probability sample of students within ZCAs. Thus, the neighborhood measures represent the aggregation of students within the sample, rather than the entire population of students within the ZCA. Scholars have noted that this approach has the potential to lead to nonrandom selection bias in the measurement of cultural context if students in the sample differ systematically from nonsample members in the same neighborhood (see Lucas 2014). Given that this limitation can only be overcome with new data collection efforts, an important avenue for future studies on the role that neighborhood cultural context plays in shaping educational outcomes is data collection strategies explicitly designed to measure neighborhood cultural context.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

David M. Merolla is assistant professor of sociology at Wayne State University. His research interests include sociology of education, race and ethnicity, social psychology and quantitative methods.

References

Anderson

Elijah

. 1994. “The Code of the Streets.” Chicago 111:81–94.

Bennett

Pamela

. 2010. “The Relationship between Neighborhood Racial Concentration and Verbal Ability: An Investigation Using the Institutional Resources Model.” Social Science Research 40(4):1124–41.

Berg

Mark T.

Steward

Eric A.

Stewart

Endya

Simons

Ronald L.

2013. “A Multi-level Examination of Neighborhood Social Processes and College Enrollment.” Social Problems 60(4):513–34.

Boardman

Jason D.

Robert

Stephanie A.

2000. “Neighborhood Socioeconomic Status and Perceptions of Self-efficacy.” Sociological Perspectives 43(1):117–36.

Christie-Mizell

C. André

Erickson

Rebecca J.

2007. “Mothers and Mastery: The Consequences of Perceived Neighborhood Disorder.” Social Psychology Quarterly 70:340–65.

Cook

Thomas

Herman

Melissa

Phillips

Meredith

Setterten

Richard

. 2002. “Some Ways in Which Neighborhoods, Nuclear Families, Friendship Groups, and Schools Jointly Effect Changes in Early Adolescent Development.” Child Development 73(4):1283–309.

Crane

Jonathon

. 1991. “The Epidemic Theory of Ghettos and Neighborhood Effects on Dropping Out and Teenage Pregnancy.” American Journal of Sociology 96(5):1226–59.

Duncan

Greg

. 1994. “Families and Neighbors as Sources of Disadvantage in the School Decision of White and Black Adolescents.” American Journal of Education 103(1):20–53.

Edin

Kathryn

Kefalas

Maria

. 2011. Promises I Can Keep: Why Poor Women Put Motherhood before Marriage. Berkeley, CA: University of California Press.

10.

Entwisle

Doris

Alexander

Karl

Olson

Linda Steffel

. 1994. “The Gender Gap in Math: Its Possible Origin in Neighborhood Effects.” American Sociological Review 59(6):822–38.

11.

Fischer

Mary J.

Kmec

Julie

. 2004. “Neighborhood Socioeconomic Conditions as Moderators of Family Resource Transmission: High School Completion among At-risk Youth.” Sociological Perspectives 47(4):507–27.

12.

Galster

George

. 2012. “The Mechanism(s) of Neighbourhood Effects: Theory, Evidence and Policy Implications.” Pp 23–56 in Neighbourhood Effects Research: New Perspectives, edited by Ham

Maarten van

Manley

David

Bailey

Nick

Simpson

Ludi

Maclennan

Duncan

. London, England: Springer.

13.

Galster

George

Marcotte

Mandell

Wolman

Augustine

2007. “The Impact of Childhood Neighborhood Poverty on Young Adult Outcomes.” Housing Studies 22:723–52.

14.

Gecas

Viktor

. 1989. “The Social-psychology of Self-efficacy.” Annual Review of Sociology 15:291–316.

15.

Gennetian

Lisa A.

Sanbonmatsu

Lisa

Katz

Lawrence F.

Kling

Jeffrey R.

Sciandra

Matthew

Ludwig

Jens

Duncan

Greg J.

Kessler

Ronald C.

2012. “The Long-term Effects of Moving to Opportunity on Youth Outcomes.” Cityscape 14(2):137–67.

16.

Harding

David J.

2003. “Counterfactual Models of Neighborhood Effects: The Effect of Neighborhood Poverty on Dropping Out and Teenage Pregnancy.” American Journal of Sociology 109(3):676–719.

17.

Harding

David J.

2007. Living the Drama: Community, Conflict, and Culture among Inner-city Boys. Chicago, IL: University of Chicago Press.

18.

Harding

David J.

2011. “Rethinking the Cultural Context of Schooling Decisions in Disadvantaged Neighborhoods: From Deviant Subculture to Cultural Heterogeneity.” Sociology of Education 84(4):322–39.

19.

Harris

Angel

. 2011. Kids Don’t Want to Fail: Oppositional Culture and the Black-white Achievement Gap. Cambridge, MA: Harvard University Press.

20.

Johnson

Odis

Jr.

2010. “Assessing Neighborhood Racial Segregation and Macroeconomic Effects in the Education of African Americans. Review of Educational Research 80(4):527–75.

21.

Jones

Rachel

Lou

. 1999. “The Culture of Poverty and African American Culture: An Empirical Assessment.” Sociological Perspectives 42(3):439–58.

22.

Kikuchi

George

Desmond

Scott

. 2010. “A Longitudinal Analysis of Neighborhood Crime Rates using Latent Growth Curve Modeling.” Sociological Perspectives 53(1):127–49.

23.

Klebanov

Pamela

Brooks-Gunn

Jeanne

McCarton

Cecelia

McCormick

Marie

. 1998. “The Contribution of Neighborhood and Family Income to Developmental Test Scores over the First Three Years of Life.” Child Development 69(5):1420–36.

24.

Kulis

Stephen

Marisglia

Flaviro Francisco

Sicotte

Diane

Nieri

Tanya

. 2007. “Neighborhood Effects on Substance Use in a Southwestern City.” Sociological Perspectives 50(2):273–301.

25.

Leventhal

Tama

Brooks-Gunn

Jeanne

. 2000. “The Neighborhoods They Live In: The Effects of Neighborhood Residence on Child and Adolescent Outcomes.” Psychological Bulletin 126(2):309–37.

26.

Lewis

Christopher

. 2012. “Oppositional Culture and Educational Opportunity.” Theory and Research in Education 10(2):131–54.

27.

Lucas

Samuel R.

2014 “An Inconvenient Dataset: Bias and Inappropriate Inference with the Multilevel Model.” Quality & Quantity 48(3):1619–49.

28.

Ludwig

Jens

Liebman

Jeffery

King

Jeffrey

Duncan

Greg

Katz

Lawrence

Kessler

Ronald

Sabonmatsu

Lisa

. 2008. “What Can We Learn about Neighborhood Effects from The Moving to Opportunity Experiment?” The American Journal of Sociology 114(1):144–88.

29.

Mayer

Susan

Jencks

Christopher

. 1989. “Growing Up in Poor Neighborhoods: How Much Does it Matter?” Science 243:1441–45.

30.

Merolla

David M.

2014. “Beyond Oppositional Culture: A New Approach to Student Attitudes and Beliefs.” Race, Gender and Class 21(1/2):112–25.

31.

Merolla

David M.

Hunt

Matthew O.

Serpe

Richard T.

2011. “Concentrated Disadvantage and Beliefs about the Causes of Poverty.” Sociological Perspectives 54(2):205–27.

32.

Newman

Katherine S.

2000. No Shame in My Game: The Working Poor in the Inner City. New York: Random House Digital.

33.

Pajares

Frank

. 1996. “Self-efficacy Beliefs in Academic Settings.” Review of Educational Research 66(4):543–78.

34.

Pattillo

Mary

. 1999. Black Picket Fences: Privilege and Peril among the Black Middle Class. Chicago, IL: University of Chicago Press.

35.

Peguero

Anthony A.

Shaffer

Kelsey A.

2015. “Academic Self-efficacy, Dropping Out, and the Significance of Inequality.” Sociological Spectrum 35(1):46–64.

36.

Raudenbush

Stephen W.

Sampson

Robert J.

1999. "Ecometrics: Toward a Science of Assessing Ecological Settings with Application to the Systematic Social Observation of Neighborhoods." Sociological Methodology 29(1): 1-41.

37.

Reynolds

John R

Burge

Stephanie Woodham

Robbins

Cheryl L.

Boyd

Emily

Harris

Brandy

. 2007. “Mastery and the Fulfillment of Occupational Expectations by Midlife.” Social Psychology Quarterly 70(4):366-83.

38.

Ross

Catherine E.

Broh

Beckett A.

2000. “The Roles of Self-esteem and the Sense of Personal Control in the Academic Achievement Process.” Sociology of Education 73(4):270–84.

39.

Ross

Catherine E.

Mirowsky

John

. 2008. “Neighborhood Socioeconomic Status and Health: Context or Composition?” City & Community 7(2):163–79.

40.

Sampson

Robert J.

Morenoff

Jeffrey

Gannon-Rowley

Thomas

. 2002. “Assessing ‘Neighborhood Effects’: Social Processes and New Directions in Research.” Annual Review of Sociology 28:443–78.

41.

Sampson

Robert J.

Raudenbush

Stephen W.

Earls

Felton

. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277(5328):918–24.

42.

Sampson

Robert J.

Sharkey

Patrick

Raudenbush

Stephen

. 2007. “Durable Effects of Concentrated Disadvantage on Verbal Ability among African-American Children.” Proceedings of the National Academy of Sciences 105(3):845–52.

43.

Sampson

Robert J.

2012. Great American City: Chicago and the Enduring Neighborhood Effect. Chicago, IL: University of Chicago Press.

44.

Sharkey

Patrick

. 2008. “The Intergenerational Transmission of Context.” American Journal of Sociology 113(4):931–69.

45.

Sharkey

Patrick

Faber

Jacob W.

2014. “Where, When, Why, and for Whom Do Residential Contexts Matter? Moving Away from the Dichotomous Understanding of Neighborhood Effects.” Annual Review of Sociology 40:559–79.

46.

Sharkey

Patrick

Sampson

Robert J.

2010. “Destination Effects: Residential Mobility and Trajectories of Adolescent Violence in a Stratified Metropolis.” Criminology 48(3):639–81.

47.

Sirin

Selcuk R.

2005. “Socioeconomic Status and Academic Achievement: A Meta-analytic Review of Research.” Review of Educational Research 75(3):417–53.

48.

Slack

Megan

Oken

Alicia

. 2014. “A Child’s Course in Life Should Be Determined Not by the Zip Code She’s Born In.” Retrieved September 24, 2015 (https://www.whitehouse.gov/blog/2014/01/10/president-obama-child-s-course-life-should-be-determined-not-zip-code-she-s-born).

49.

Small

Mario Luis

Newman

Katherine

. 2001 “Urban Poverty after The Truly Disadvantaged: The Rediscovery of the Family, the Neighborhood, and Culture.” Annual Review of Sociology 27:23–45.

50.

Vargas Lascano

Dayuma I.

Galambos

Nancy L.

Krahn

Harvey J.

Lachman

Margie E.

2015. “Growth in Perceived Control across 25 Years from the Late Teens to Midlife: The Role of Personal and Parents’ Education.” Developmental Psychology 51(1):124–35.

51.

Von Hippel

Paul T.

2007. “Regression with Missing Ys: An Improved Strategy for Analyzing Multiply Imputed Data.” Sociological Methodology 37(1):83–117.

52.

Wilson

William J.

1996. When Work Disappears: The New World of the Urban Poor. New York: Random House.

53.

Wodtke

Geoffrey

Harding

David

Ewert

Felix

. 2011. “Neighborhood Effects in Temporal Perspective: The Impact of Long-term Exposure to Concentrated Disadvantage on High School Graduation.” American Sociological Review 76(5):713–36.