Rural-Urban Disparity Essay Example

gap has relied on census data such as the Current Population Survey (Gibbs 2003) and American Community Survey (Provisions et al. 2007). However, these studies have limitations. Many talented adolescents and young adults in rural areas often have to leave their communities to pursue educational and employment opportunities (Corbett 2000; Crockett, Shaman, and Jackson-Newswomen 000). This migration of youth from rural areas can impact how we estimate educational attainment levels both in rural and monaural areas (Howled and Gun 2004).

To accurately measure completion rates of postsecondary degrees in rural areas compared to other regions, it is crucial to use longitudinal data that track individuals from high school through their postsecondary education journey. Research using this type of data has revealed different patterns, particularly regarding differences between rural and monaural areas in terms of postsecondary attainment.A study by Blackwell and McLaughlin (1998) discovered that rural youth had initially lower educational ambitions than urban youth but were able to get closer to achieving their goals over time. Similar findings were reported by Bibcock and Delude (2005) and Dolman (2006), showing minimal disparities in bachelor's degree attainment between rural and urban areas. While these longitudinal studies provide valuable insights into educational differences between rural and urban communities, it remains unclear if residential or school location truly influences educational achievement or if observed differences reflect preexisting disparities among adolescents.

Previous research has employed regression-based methods to control for preexisting dissimilarities and eliminate selection bias, although statistical adjustment alone does not guarantee the removal of selection bias. Additional procedures can effectively model the selection process (Rosenberg and Rubin 1983, 1984). The GSM method (Rosenberg and Rubin 1983, 1984) offers an alternative

approach to address selection bias. It utilizes logistic regression to predict the probability (propensity score) of being in a specific group, ensuring equal chances of assignment to treatment conditions. Control students with similar propensity scores are then matched with treated students.This research aims to investigate disparities in educational attainment between rural and urban areas by utilizing the Global System for Mobile Communications (GSM) as an alternative approach. The study utilizes the NELL dataset, which is suitable for addressing selection bias issues and automating analysis of rural youth. Unlike other datasets like the Education Longitudinal Study (ELSE), NELL provides comprehensive information about students, parents, and longitudinal data tracking students from high school to postsecondary education. The objective is to expand upon previous research by thoroughly examining the influence of reality on educational attainment using advanced statistical models and longitudinal data. It does not aim to definitively establish causal effects but addresses limitations in prior studies exploring disparities in educational achievement between rural and urban areas. Various factors identified in previous research contribute to differences in educational attainment, including family background, demographic factors (such as gender and race/ethnicity), and school resources (like curriculum intensity).Socioeconomic status (SES), as measured by parental education and family income, is an influential factor in determining educational attainment. Numerous studies have shown that children from high-SES families are more likely to successfully complete high school, attend college, and earn a college degree. Family characteristics such as family structure and number of siblings also play a role in educational outcomes. Children from single-parent households have a lower likelihood of completing high school compared to those from two-parent households. Enrolling in postsecondary institutions is affected

by factors like family size and school performance. Students with more siblings tend to have lower school performance and higher dropout rates.

Demographics also impact educational attainment. Minority students, particularly black and Hispanic adolescents, are less likely to enroll in and complete college compared to white students. However, Asian American youth achieve higher levels of education than white children. Among NELL eighth graders, a study found that Asian students had the highest college enrollment rate at 95%, while white and black students had rates of 77% and Hispanics had a rate of 70%.The gender gap in educational achievement is highlighted by Buchanan and Dippier (2006) as well as the U.S. Department of Education (2004), who note a shift from favoring males to favoring females in college enrollment and completion. Furthermore, several studies (Dolman 1999, 2006; Goldbrick-Arab et al., 2007) demonstrate that tracking and academic coursework during high school have an impact on the transition to college and subsequent performance.

Dolman (1999) utilized data from the High School and Beyond (HAS) Survey to identify significant high school courses in math, science, and foreign language. The study revealed that students who took advanced levels of these courses had higher chances of attending a four-year institution and achieving success academically in college. This finding was also supported by Dolman's NELL study conducted in 2006.

Conversely, there has been extensive research on disparities between rural and urban areas regarding educational aspirations (Howled 2006; Huh 2003; Rosewoods 1996), as well as differences in school performance (Fan & Chin 1999; Howled & Gun 2004; Lee & McIntyre 2000). However, limited research specifically examines large-scale rural-urban discrepancies in educational achievement (Blackwell & McLaughlin 1998; Rosining

& Crowley 2001). Additionally, there is minimal research investigating variations between rural and urban areas in postsecondary attainment using longitudinal data.The role of reality in the postsecondary attainment of youth has not received much attention. Evidence suggests that family background and school resources are important factors in the rural-urban disparity in postsecondary attainment. Several studies, including those by Conger and Elder (1994), Hobbs (1994), Lighter and McLaughlin (1995), Provision et al.(2007),and Rosining and Crowley(2001) have highlighted the significance of these factors.

There are several important aspects to consider regarding the differences between rural and urban areas. Firstly, poverty rates have been increasing in rural areas compared to urban settings over the past decade (O'Hare 2009; Rogers 2005). Additionally, traditional family arrangements for rural children have become less common (O'Hare and Churchill 2008). Moreover, there is concern about limited opportunities for rural students to take advanced placement (AP) courses and a shortage of teachers with advanced degrees in rural schools (Graham 2009; Monk 2007; Provision et al. 2007).

Many rural communities have also experienced significant job losses across various occupations that were once integral to these communities for generations (Conger and Elder 1994; Gibbs, Siskin, and Corporate 2005; Hobbs 1994). Lighter and McLaughlin (1995) suggest that there may be a gap in educational aspirations and attainment among rural youth due to socioeconomic and educational factors specific to rural areas.Poverty, inadequate schooling conditions, low expectations from parents and teachers, and lower academic performance are factors often associated with rural students. However, there is empirical evidence that challenges this belief. A study analyzing NELL data found that rural students perform just as well as their urban counterparts in school. This

suggests that living in a rural area or attending a rural school does not necessarily put rural adolescents at a disadvantage. Additionally, research by Lee and McIntyre (2000) revealed that rural students actually outperformed their urban peers in math achievement based on NAEP assessment scores from 1992-1996. This supports the 'rural strength model' proposed by Edmonton and Koehler (1987). Nevertheless, limitations in previous research hinder making strong conclusions about the causal effects of reality on educational outcomes. Observational data lacks randomization and is therefore unsuitable for making causal inferences (Rosenberg and Rubin 1983, 1984).Comparing treatment and control groups directly in a non-governmental setting can lead to misleading results because of systematic differences between individuals exposed to different treatments (Rosenberg and Rubin 1983, 1984). The limitation of studying the rural-monaural achievement gap is evident due to systemic differences in background characteristics between rural and monaural adolescents. To examine the impact of reality on educational outcomes while controlling for student background characteristics, random assignment to families with only a difference in residence (rural vs. monaural) would be necessary. Unfortunately, such randomized assignment is not feasible. In a non-governmental setting, the lack of randomization makes it challenging to estimate causal effects of reality. However, it is still possible to draw causal inference from observational studies by addressing selection issues and establishing a causal relationship between reality and educational attainment. This study used statistical models, including the GSM technique, to address selection bias in observational data and make a causal inference regarding the role of reality in educational attainment. The research focused on comparing educational goals and achievement between rural and monaural areas using data from the National Education

Longitudinal Study conducted by the National Center for Education Statistics (ONCE) in 1988.The study followed approximately 25 eighth graders from around 1,000 middle schools until they reached ages 26 or 27 in both 1944 and 2000. The NELL:88-00 panel study involved around 12,100 students and utilized postsecondary transcript data from 1988 to 2000 for accurate information on postsecondary attainment. Family structure and number of siblings were obtained from the 1994 wave and combined with the transcript data. American Indian/Alaska Native and multiracial students were excluded due to small sample sizes. The final analytical sample included approximately 11,700 students, with about thirty percent being rural youth. Table one demonstrates the percentage distribution based on actual high school graduation status, postsecondary participation, and highest degree completed by the year 2000 among the high school seniors of '92. The data shows minimal differences in high school completion rates between rural and non-rural students. Out of approximately 3,630 rural students,80% earned an academic diploma while10% achieved a GED or equivalent qualification,and10% did not complete high school by the year2000.For non-rural students these rates were respectively recorded as80%,10%,and10%.Table one presents data on postsecondary participation and highest degree attainment for rural and non-rural students (referred to as monaural). In the year 2000, 70.1% of rural youth were enrolled in postsecondary institutions, while for monaural youth, this number was higher at 78.6%, representing an increase of around eight percentage points compared to rural youth.

In terms of highest degrees earned, among rural youth, 4% earned a certificate and 6.5% attained an associate degree. For monaural youth, these numbers were slightly lower at 3.5% and 5.1%. However, when considering bachelor's degrees or above, only

approximately 25.6% of rural youth achieved these degrees compared to about 30% for monaural youth.It is worth noting that approximately 31.6% of those who enrolled in postsecondary institutions but did not complete their studies were monaural youth, compared to 28% for rural youth. The multivariate analysis focused on postsecondary entrance status and highest degree status due to the lack of noticeable differences in high school completion between the two groups. Incomplete cases (7%) were excluded from consideration when analyzing highest degree status. To address missing data on educational attainment, reality, and gender, an alternative algorithm suggested by King and colleagues (2001) was used for imputation. The NELL school identifier was utilized in the multivariate models to account for initial sample clustering of students within schools (see Table 1). The study also took into account the longitudinal second follow-up to fourth follow-up panel weight. Two variables measured educational attainment: postsecondary participation and postsecondary degree attainment. Postsecondary participation determined whether respondents had enrolled in a postsecondary institution by 2000, while postsecondary degree attainment categorized the highest degree achieved by 2000 into four groups: no postsecondary enrollment, certificate/associate degree, bachelor's degree, and no postsecondary degree (referring to students who had enrolled in college but did not earn a degree by 2000). Both measures were based on postsecondary transcript data.According to Quality Education Data and NELL (Lippies, Burns, and Macarthur 1996), rural residence was determined based on the location of schools outside Metropolitan Statistical Areas. The initial responses representing rural, suburban, and urban backgrounds were combined into the binary category of rural vs. monaural. Therefore, it is essential to consider the socioeconomic and demographic backgrounds of young individuals in

order to accurately evaluate their educational achievement. Table 2 illustrates significant disparities in these backgrounds between rural and monaural youth. Hence, including measures of background variables that may impact educational attainment as controls becomes necessary for establishing a reliable connection between educational achievement and reality. In 1992, all background variables were measured during grade 12. Family background variables included parental education, family income, family structure, and family size. Parental education was determined based on respondents' highest level achieved ranging from less than high school graduation to a doctorate or other professional degree. These responses were condensed into two categories: some college or lower and bachelor's degree or higher with some college or lower being the reference category. Grade 12 family income was reported by parents and grouped into different categories based on income levels such as less than $25,000, $25,000 - $49,999,and $50,000 or more.
The income reference category was below $25,000. Family structure indicated whether students lived in two-parent families or other types of families. The number of siblings the student had at grade 12 was reported by the parent. Individual student characteristics included gender, race/ethnicity, and academic achievement. Gender was determined by the student's sex (female or male). Race/ethnicity options reported by the student included Asian, Hispanic black ,and white. White students served as the reference group for race/ethnicity. Academic achievement was measured using the math/reading composite score from a standardized test administered during 12th grade.

Table 2 presents weighted descriptive statistics for all indicators. The focus of our study is to compare background variables between rural and monaural youth by examining unadjusted differences in outcome variables shown in Table 1.

The results from Table 2

clearly indicate socioeconomic background disparities between rural and monaural youth. Specifically, there is a notable difference in parental education, with only 21 percent of rural youth having parents with a bachelor's degree or higher, compared to 32 percent of monaural youth – a difference greater than 10 percentage points. Furthermore, rural youth face disadvantages in terms of family income, as only 25 percent of those from families earning $50,000 or above annually are monaural, while this percentage increases to 36 percent for monaural youth.In contrast, the text discusses differences between rural and monaural youth populations in terms of family income, racial/ethnic background, family structure, and size. Among rural youth, 41 percent come from families earning less than $25,000 annually compared to only 30 percent for monaural youth. While there are minimal differences in family structure and size between the two groups, significant disparities exist in racial/ethnic background. Table 2 reveals that only 4 percent of rural youth are white, while a majority (68 percent) of monaural youth fall into this category. Additionally, the percentages of Hispanic and black individuals among rural youth are higher at rates of 7 percent each compared to monaural youth where these rates increase to approximately double at13 percent and14percent respectively. These rates approximately double those observed among rural youths.

The text also addresses the representation of Asian background in both populations as well as gender differences. Furthermore, an academic achievement gap favoring monaural youths is mentioned. The variations in background characteristics highlight selection effects that were addressed using multiple models including GSM technique.

To evaluate if rural-monaural differences remain significant when considering other factors such as parental education, family income, and race/ethnicity;

two analytical strategies were employed: conventional logistic regression and GSM technique were used to analyze college enrollment likelihoodsThe STATS survey commands were utilized to adjust stratification, clustering, and individual weighting in order to examine the validity of the observed relationship when students were matched based on their preexisting background characteristics. The propensity score indicated the likelihood of attending a rural high school considering observed backgrounds. To investigate GSM, logistic regression was initially conducted using Table 1 controls and associated covariates related to attending a rural high school (prior literature: Guy and Fraser 2009; Rosenberg and Rubin 1983, 1984). After obtaining propensity scores through logistic regression, a dataset with matched individuals was created using a one-to-one caliper matching procedure to ensure similar observed backgrounds among students. In this matching process, a caliper size equivalent to one-fourth of the standard deviation of estimated propensity scores was employed following the approach outlined by Rosenberg and Rubin in 1985. T-tests or chi-square tests were then performed on each covariate to assess balance within the matched sample. The results indicated no significant disparities in covariates between rural and monaural youth, suggesting that most of the selection bias in the matched sample had been effectively addressed by the estimated propensity scores. Finally, utilizing optimally matched samples, we replicated the logistic regression model described earlier.Using conventional multinomial logistic regression analysis, we assessed the likelihood of young individuals attaining different levels of postsecondary degrees in comparison to those who did not pursue postsecondary education. We presented two models: one comprising only reality arable and another with additional controls. To investigate the impact of reality on postsecondary attainment, we employed the GSM technique. It

is important to note that sample sizes varied in GSM analyses due to variations in outcome variables. In the following section, we present the findings regarding postsecondary participation by examining how reality relates to postsecondary enrollment while considering other influential factors affecting student participation. Table 3 presents easily understandable odds ratios, where a ratio above one indicates higher chances and a ratio below one suggests lower chances of being in the comparison category rather than the reference category.