National Educational Longitudinal Survey Research Articles

Measurement bias and error correction in a two-stage estimation for multilevel IRT models.

Among current state-of-the-art estimation methods for multilevel IRT models, the two-stage divide-and-conquer strategy has practical advantages, such as clearer definition of factors, convenience for secondary data analysis, convenience for model calibration and fit evaluation, and avoidance of improper solutions. However, various studies have shown that, under the two-stage framework, ignoring measurement error in the dependent variable in stage II leads to incorrect statistical inferences. To this end, we proposed a novel method to correct both measurement bias and measurement error of latent trait estimates from stage I in the stage II estimation. In this paper, the HO-IRT model is considered as the measurement model, and a linear mixed effects model on overall (i.e., higher-order) abilities is considered as the structural model. The performance of the proposed correction method is illustrated and compared via a simulation study and a real data example using the National Educational Longitudinal Survey data (NELS 88). Results indicate that structural parameters can be recovered better after correcting measurement biases and errors.

Correction for Item Response Theory Latent Trait Measurement Error in Linear Mixed Effects Models

When latent variables are used as outcomes in regression analysis, a common approach that is used to solve the ignored measurement error issue is to take a multilevel perspective on item response modeling (IRT). Although recent computational advancement allows efficient and accurate estimation of multilevel IRT models, we argue that a two-stage divide-and-conquer strategy still has its unique advantages. Within the two-stage framework, three methods that take into account heteroscedastic measurement errors of the dependent variable in stage II analysis are introduced; they are the closed-form marginal MLE, the expectation maximization algorithm, and the moment estimation method. They are compared to the naïve two-stage estimation and the one-stage MCMC estimation. A simulation study is conducted to compare the five methods in terms of model parameter recovery and their standard error estimation. The pros and cons of each method are also discussed to provide guidelines for practitioners. Finally, a real data example is given to illustrate the applications of various methods using the National Educational Longitudinal Survey data (NELS 88).

Third Generation Disadvantage among Mexican Americans

Among Mexican Americans, generational differences in education do not fit with assimilation theory’s predictions of significant improvement from the second to third generation; instead, education for third generation remains similar to the second generation and falls behind that of non-Hispanic whites. Scholars have not examined this educational gap for recent cohorts, nor have they considered a wide range of economic outcomes by generation. Using a nationally representative sample of young adults from the National Educational Longitudinal Survey, we examine various educational and economic outcomes among second- and third-generation Mexican Americans and compare it to whites and blacks. We find that third-generation Mexican Americans have similar outcomes to the second generation and lower education and economic levels than whites and blacks, even when controlling for key factors. Our findings reveal limitations to assimilation theory and suggest that the persistent low status of third-generation Mexican Americans may be largely due to their racialization. These findings coupled with prior research on Mexican Americans point to a consistent pattern of third generation disadvantage, which stands in contrast to second generation advantage.

The Returns to Education at Community Colleges: New Evidence from the Education Longitudinal Survey

I use nationally representative data from the Education Longitudinal Survey (ELS) to update the literature on returns to community college education. I compare the experiences of the ELS cohort that graduated high school in 2004 with those of the National Education Longitudinal Survey (NELS) cohort that graduated high school more than a decade earlier, in 1992. I estimate that community college students from the ELS cohort were more likely to be employed, and that those who were earned about 21 percent more than comparable peers with only a high school education. This estimate is at least as large as that observed for the NELS cohort, though I find some evidence that the value of an associate's degree is smaller for the more recent cohort. I compare these results with those from the burgeoning body of research using state administrative data to answer similar questions.

Will Moving to a Better Neighborhood Help? Teenage Residential Mobility, Change of Context, and Young-Adult Educational Attainment

Research suggests that growing up in more affluent neighborhoods improves educational attainment. But would it help adolescents to move to relatively more affluent neighborhoods, as theories of neighborhood effects anticipate? Does it depend on the magnitude of the change of context? To answer these questions, we use data from the National Educational Longitudinal Survey and the 1990 Census to estimate models using propensity score methods. We found that both upward mobility and change of context during adolescence had small effects on long-term educational attainment that varied by race, socioeconomic status, transfer status, and the social class of starting neighborhoods. Importantly, upward moves and positive changes in context reduced African-Americans’ chances of completing high school.

High School Dropout of Mexican Origin and White Non-Hispanic Youths in United States

The aim of this paper is to analyze the factors and reasons causing high school dropout of a cohort of Mexican and white non-Hispanic students. The source of data used is the National Education Longitudinal Survey (NELS 88-92), which is representative at national level. The main findings are: student’s age, labor force participation, school absence, extracurricular activities, socioeconomic strata, parent´s education, family structure, number of siblings and type of school are influencing high school dropout for Mexican origin and White non-Hispanic youths. With respect to reasons for high school dropout, we found they are related with early adulthood. For Mexican origin young people, the first child or pregnancy is the main reason for school dropout, while white non-Hispanic is labor force entry.

MISBEHAVIOR, EDUCATION, AND LABOR MARKET OUTCOMES

Using data on young men from the National Education Longitudinal Survey, this paper investigates the relationship between childhood misbehavior and later education and labor market outcomes. The main finding is that eighth-grade misbehavior is important for earnings over and above eighth-grade test scores. Moreover, controlling for educational attainment, childhood misbehavior is associated with earnings at all educational levels, whereas achievement test scores are related to earnings only for young men with postsecondary degrees. Possible explanations for the association between eighth-grade misbehavior and economic success are explored.

Race/Ethnicity, Attitudes, and Living With Parents During Young Adulthood

Non‐White young adults are more likely to live with their parents throughout their 20s, more likely to return home after going away to college, and less likely to leave again after returning. Scholars have speculated that subcultural differences in attitudes toward marriage and family play a key role in generating racial/ethnic differences in rates of coresidence with parents among young adults. Data from the National Education Longitudinal Survey of 1988 (N = 11,228) were analyzed in order to test this hypothesis. Attitudes toward marriage and family were significantly associated with coresidence, especially among young men, but did not substantially account for racial/ethnic differences in living arrangements. Among young non‐White women and young Black men, higher rates of coresidence were related to differences from Whites in socioeconomic or marital status (and sometimes both) that were largely independent of differences in attitudes toward marriage and family.

Academic achievement trajectories of adolescents from Mexican and East Asian immigrant families in the United States

Drawing on the National Educational Longitudinal Survey 1988 (NELS:88), this study identified (1) the growth pattern of academic achievement of adolescent children from Mexican and East Asian immigrant families; (2) investigated to what extent ethnicity and family capital influenced the trajectories in the academic achievement of children from Mexican and East Asian immigrant families in the United States. The growth curve model analysis showed that ethnicity had a significant main effect on academic trajectories. Taking into account various forms of family capital, however, there was no significant ethnicity effect on the rate of change. In addition, compared to parents with a lower level of family capital, those whose families had higher levels of capital generally did better at eighth grade and these students’ achievement accelerated over time. Implications and limitations are also discussed.

Just the Facts, Ma'am: Postsecondary Education and Labor Market Outcomes in the U.S.

In this paper, we provide a comprehensive and up-to-date snapshot of the most important postsecondary education and labor market outcomes in the U.S. using two nationally representative sources of data: The Survey of Income and Program Participation (SIPP) and The National Educational Longitudinal Survey (NELS). We find that postsecondary educational attainment has risen modestly among young Americans over the past two decades, with greater gains in BA attainment in the 1990s and in certificate and AA attainment since 2000 (though attainment rose during the Great Recession at all levels). Both younger and older cohorts of blacks and Hispanics have made relative progress in the attainment of certificates and AAs but still lag behind whites in the entry into and completion of BA programs; completion rates in BA programs also lag substantially for those from low-income families or with weak academic achievement in high school. Young women (especially among whites and blacks) now achieve sub-BA and BA degrees at higher levels than their male counterparts and generally have higher achievement.There are labor market returns for all postsecondary credentials, including certificates and AA degrees, though these vary across field of study. Large gender gaps exist in field of study, with men more concentrated in high-paying fields. Lastly, we find that high school achievement measures can account for much of the racial gaps in BA completion and earnings and some of the gaps by family background, though they account for little of the continuing gender gap favoring young men in earnings.

Access Without Equity

The competitive dynamics that sustain stratification among postsecondary institutions have reinforced racial inequality in selective college enrollment between 1972 and 2004. Using a data set constructed from four nationally representative surveys (National Longitudinal Survey 1972, High School & Beyond 1980, National Educational Longitudinal Survey 1988, and Educational Longitudinal Survey 2002), the authors model how escalating admissions standards—including academic preparation and the growing importance of SAT scores and extracurricular leadership—effectively maintain racial inequality in selective college enrollment over time. Black and Latino students have made strides in their pre-collegiate academic preparation. Nevertheless, although access to postsecondary education has expanded since 1972 for all ethnic groups, Black and Latino students’ odds of selective college enrollment have declined relative to White and Asian American students.

Parental Involvement and the Academic Achievement of Mexican American Youths: What Kinds of Involvement in Youths' Education Matter Most?

Parental involvement in education is a key focus of current policies and programs aimed at improving the academic outcomes of students at risk for academic underachievement. This study examines six forms of parental involvement in education to determine which forms of involvement have the strongest relationships with youths' academic outcomes. Using nationally representative data (N = 1,609) from the National Education Longitudinal Survey, this study focuses specifically on Mexican American families and youths, a population at high risk for academic underperformance. Findings show that the positive effects of parental involvement among Mexican American parents occur through involvement in the home, whereas parental involvement in school organizations is not associated with youths' achievement. Parents' investment of financial resources in their children's education was found to have a somewhat higher impact on achievement than forms of involvement that require parents' investment of time. Findings also suggest that the impact of these forms of parental involvement occurs prior to high school. KEY WORDS: academic achievement; Mexican American youths; parent involvement ********** Policy and program interventions aimed at improving children's academic outcomes often focus on increasing parental involvement. The federal No Child Left Behind Act of 2001 (RL. 107-110) highlights parental involvement as a key factor in improving academic outcomes, particularly for children attending schools that serve high proportions of low-income children (Title 1 schools). Indeed, parental involvement appears to be an important factor in improving children's academic outcomes, and a sizeable body of work generally points to the positive effects of parents' naturally occurring involvement (for a recent review, see Pomerantz, Moorman, & Litwack, 2007). However, less is known about the effects of parental involvement on academic outcomes among children of color, particularly Latino children (Jeynes, 2003). Effects of parental involvement in academics appear to differ among racial-ethnic groups (Desimone, 1999) and among different Latino nationality groups (Figueroa-Moseley, Ramey, Keltner, & Lanzi, 2006). In addition, different types of parental involvement have distinct relationships with academic outcomes (Domina, 2005; Jeynes, 2003; Pomerantz et al., 2007). Consequently, to understand the role of parental involvement and its potential utility for intervention in promoting academic achievement among children of color, and particularly Latino children, more studies that carefully examine the role of multiple forms of involvement with diverse samples are needed. Of critical importance are studies focused on Mexican American families, both those that have recently migrated and those that have resided in the United States for generations. Mexican Americans are by far the largest and fastest growing population of Latinos in the United States, accounting for 65% of U.S. Latinos and 9.7% of the entire U.S. population (U.S. Census Bureau, 2009). Among Latino groups, Mexican Americans are at gravest risk for living in poverty, in part because of lower rates of high school completion and college attendance within this group (U.S. Census Bureau, 2002). The likelihood of high school dropout is two to four times higher for Mexican American students than for Cuban and South American youths, even after controlling for factors such as socioeconomic status (Driscoll, 1999; Landale, Oropesa, & Llanes, 1998). Indicators of academic achievement, such as grades and performance on standardized tests, are, on average, lower among Mexican American children than among children from other immigrant and native-born groups (Ferguson, 2001; Kao & Thompson, 2003; Portes & Rumbaut, 2001). Thus, interventions to address the academic disparities faced by Mexican American children and youths are much needed. …

Maternal Employment and Adolescent Self-Care

Mounting evidence shows that self-care produces deleterious consequences for adolescents in the U.S. Since desscriptive evidence suggests that maternal employment is the primary explanation for adolescent self-care, maternal employment, it is frequently argued, is harming children. Heretofore, very little empirical research has actually investigated the impact of maternal employment on adolescent self-care, however, calling into question this assertion. This paper aims to fill this gap. The author uses the National Education Longitudinal Survey of 1988 supplemented by the National Longitudinal Survey of Youth-1979 to estimate the relationship between maternal employment and adolescent self-care. Unlike prior research, the author employs a variety of fixed effects models to account for omitted variables that may be related to maternal employment and adolescent self-care. Findings suggest that the adolescents of mothers who work full-time spend an additional 43 minutes per week in self-care compared to the adolescents of mothers who work part-time. Further, a standard deviation increase in the number of weeks a mother works during the year increases the probability that her child will be unsupervised by 27 percent. These effects are not constant across socio-economic groups: affluent families have strong effects, while the relationship is more tenuous among low-income families. This finding has important implications for pro-work social welfare policies in the United States.

Routine activities as determinants of gender differences in delinquency

This study examined the extent to which gender differences in delinquency can be explained by gender differences in participation in, or response to, various routine activity patterns (RAPs) using data from the second and third waves of the National Education Longitudinal Survey of 1988. While differential participation in routine activities by gender failed to explain males’ high levels of deviance relative to females, two early RAPs moderated the effect of gender on subsequent deviant behavior. Participation in religious and community activities during the sophomore year in high school decreased, while unstructured and unsupervised peer interaction increased, levels of delinquency two years later substantially more for males than for females, suggesting there are gender differences in reactivity to contextual opportunities for deviance during early high school with effects that persist over time.

THE EARNINGS EFFECT OF EDUCATION AT COMMUNITY COLLEGES

In this paper, I make use of data from the 2000 follow‐up of the National Education Longitudinal Survey postsecondary education transcript files to extend what is known about the value of education at community colleges. I examine the effects of enrollment in community colleges on students' subsequent earnings. I estimate the effects of credits earned separately from credentials because community colleges are often used as a means for students to engage in study not necessarily leading to a degree or certificate. I find consistent evidence of wage and salary effects of both credits and degrees, especially for women. There is no substantial evidence that enrollment in vocational rather than academic coursework has a particularly beneficial effect, however. (JEL I2, J24)

Violence, Older Peers, and the Socialization of Adolescent Boys in Disadvantaged Neighborhoods

Most theoretical perspectives on neighborhood effects on youth assume that neighborhood context serves as a source of socialization, but the exact sources and processes underlying adolescent socialization in disadvantaged neighborhoods are largely unspecified and unelaborated. This paper proposes that cross-cohort socialization by older neighborhood peers is one source of socialization for adolescent boys in such neighborhoods. Data from the National Educational Longitudinal Survey suggest that adolescents in disadvantaged neighborhoods are more likely to spend time with older individuals. Qualitative interview data from 60 adolescent boys in three neighborhoods in Boston are analyzed to understand the causes and consequences of these interactions and relationships. I find that some of the strategies these adolescents employ to cope with violence in disadvantaged neighborhoods promote interaction with older peers, particularly those who are most disadvantaged, and that such interactions can expose adolescents to local, "unconventional," or "alternative" cultural models.

Multilevel Regression Mixture Analysis

SummaryA two-level regression mixture model is discussed and contrasted with the conventional two-level regression model. Simulated and real data shed light on the modelling alternatives. The real data analyses investigate gender differences in mathematics achievement from the US National Education Longitudinal Survey. The two-level regression mixture analyses show that unobserved heterogeneity should not be presupposed to exist only at level 2 at the expense of level 1. Both the simulated and the real data analyses show that level 1 heterogeneity in the form of latent classes can be mistaken for level 2 heterogeneity in the form of the random effects that are used in conventional two-level regression analysis. Because of this, mixture models have an important role to play in multilevel regression analyses. Mixture models allow heterogeneity to be investigated more fully, more correctly attributing different portions of the heterogeneity to the different levels.

The Sanctions-Perceptions Link in a Model of School-based Deterrence

This paper reports a study examining associations between objective indicators of the level of discipline within schools and students’ perceptions of the strictness of discipline. Data were analyzed from the National Education Longitudinal Survey (NELS), a nationally representative panel study of eighth grade students attending public and private schools in 1988. We find evidence for an association between objective and perceived risk of discipline in models that examine the covariation of these two constructs at several cross sections, and in models of change in perceptions as a function of change in school sanctioning climate. Moreover, these associations were strongest in small and less disordered schools.

Political Participation and Cumulative Disadvantage: The Impact of Economic and Social Hardship on Young Citizens

In this article, we seek to provide the most comprehensive exploration to date of the effects of economic and social disadvantage on young citizens' voter turnout. We look at four overlapping domains of hardship—those rooted in (a) the family context, (b) the community context, (c) the school context, and (d) major events and life transitions. Our conceptual model of cumulative advantage/disadvantage identifies the different ways in which disadvantages can have cumulative effects on social outcomes generally and on civic participation in particular. Our framework also suggests how patterns of advantage and disadvantage may be mediated by social institutions. Using data from the National Education Longitudinal Survey, we show that disadvantages rooted in the family have major impact on all groups of young citizens and that family disadvantage interacts with school disadvantage for whites. We also show that the effects of early parenthood, being arrested, and dropping out of school have differential effects depending on race. We also show that community colleges function as civic leveling institutions—especially for African American youth.

School and Parental Influences on Sociopolitical Development Among Poor Adolescents of Color

Sociopolitical development represents a motivation to reduce sociopolitical inequity, a healthy sense of the self, and perceived agency in the context of structural oppression. Sociopolitical development has been associated with progress in career development, school engagement, and healthier sexual behavior among oppressed and marginalized adolescents. However, knowledge of outcomes that sociopolitical development may facilitate has outpaced knowledge of what may facilitate sociopolitical development. This paper examines key aspects of adolescents' context—the school and parents—and their potential influence on the sociopolitical development of poor adolescents of color. This research question is examined with 2,078 National Educational Longitudinal Survey participants. The obtained structural equation model suggests that parental support and student racial relations are predictive of sociopolitical development. These results provide empirical support for extant sociopolitical development models and illuminate our understanding of a critical social justice issue. These results inform sociopolitical development interventions, counseling practice, and scholarly understanding of how oppressed adolescents cope with sociopolitical inequity.