Citation
Can rurarilty predict student achievement

Material Information

Title:
Can rurarilty predict student achievement
Creator:
Eveatt, Hailey
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of psychology)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
School psychology
Committee Chair:
Harris, Bryn
Committee Members:
Crepeau-Hobson, Franci
Hohnbaum, Colette

Notes

Abstract:
This study seeks to enhance the current literature on student achievement by examining a snapshot of Colorado’s academic success - the third grade reading scores on the 2014 Transitional Colorado Assessment Program - and discovering whether or not the degree of a school district’s rurality has an impact on student performance. This study utilized a non-experimental, quantitative research methodology to investigate both the relationship and the strength of the relationship between rurality and student achievement using correlation analysis and multiple linear regression. Pearson’s r correlation coefficient was used to measure the strength of the relationship between rurality and student achievement. Independent variables include rurality of a school district, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body. Rurality of a school district was defined using the guidelines set out by the Colorado Department of Education. Regression analysis indicated that the model significantly predicted roughly 26% of third grade 2014 TCAP reading scores. However, the correlation between a school district’s rurality and third grade TCAP reading scores could not be made for the selected Colorado school districts in the 2013 - 2014 school year.
General Note:
n3p

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Hailey Eveatt. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
CAN RURALITY PREDICT STUDENT ACHIEVEMENT?
by
HAILEY EVEATT B.A., University of Denver, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Psychology School Psychology Program
2018


11
© 2018
HAILEY EVEATT
ALL RIGHTS RESERVED


Ill
This thesis for the Doctor of Psychology degree by Hailey Eveatt has been approved for the School Psychology Program by
Bryn Harris, Chair and Advisor Franci Crepeau-Hobson Colette Hohnbaum
Date: May 12, 2018


Eveatt, Hailey (Psy.D., School Psychology)
Can Rurality Predict Student Achievement?
Thesis directed by Associate Professor Bryn Harris
ABSTRACT
This study seeks to enhance the current literature on student achievement by examining a snapshot of Colorado’s academic success - the third grade reading scores on the 2014 Transitional Colorado Assessment Program - and discovering whether or not the degree of a school district’s rurality has an impact on student performance. This study utilized a non-experimental, quantitative research methodology to investigate both the relationship and the strength of the relationship between rurality and student achievement using correlation analysis and multiple linear regression. Pearson’s r correlation coefficient was used to measure the strength of the relationship between rurality and student achievement. Independent variables include rurality of a school district, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body. Rurality of a school district was defined using the guidelines set out by the Colorado Department of Education. Regression analysis indicated that the model significantly predicted roughly 26% of third grade 2014 TCAP reading scores. However, the correlation between a school district’s rurality and third grade TCAP reading scores could not be made for the selected Colorado school districts in the 2013-2014 school year.
The form and content of this abstract are approved. I recommend its publication.
Approved: Bryn Harris


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION...........................................................8
Vignette: The Great Divide..........................................8
II. LITERATURE REVIEW....................................................12
No Child Left Behind Act...........................................12
Every Student Succeeds Act.........................................13
History of Standardized Assessment in Colorado.....................14
Predictors of Rural Student Achievement............................16
III. METHODS.............................................................20
Research Questions and Hypotheses..................................21
Target Population..................................................22
Sampling Method....................................................22
Data Collection and Analyses.......................................23
IV. RESULTS.............................................................25
Correlation and Multiple Regression Analysis : Research Question 1.27
Analysis of Research Question 2....................................31
V. DISCUSSION..........................................................33
Limitations........................................................34
Conclusion.........................................................37
REFERENCES
39


VI
LIST OF TABLES
TABLE
1. Rurality, Achievement, and Predictor Variable Descriptive Statistics.26
2. Regression Analysis Model Table......................................30
3. Correlation Coefficients.............................................32


LIST OF FIGURES
FIGURE
l. Normal Probability Plot


8
CHAPTER I INTRODUCTION Vignette: The Great Divide
The towns of Aspen and Center, Colorado are just 251 miles apart. They are both small, geographically isolated, rural communities nestled within the Rocky Mountains. Using just this information as context, a reasonable observer might assume that the two towns are fairly similar - comparable, even. But Aspen and Center are not just separated by the Continental Divide - they are also split by a decidedly less picturesque gap. To examine this great divide, we turn to the illuminating lens of public education. In 2013 -2014, there were 1,670 students enrolled in the Aspen 1 school district, in grades preschool through 12. Center 26 JT had 617 pre-kindergarten through twelfth grade students attending during that same school year. In the Spring of 2014, the state of Colorado’s testing window began for the annual Transitional Colorado Assessment Program (TCAP) exams. In Aspen, 88.89% of students received passing scores of proficient or advanced. In Center, 66.67% of their students earned a passing score. The differences get starker from there. Roughly 13% of Aspen’s student population was identified as a racial and/or ethnic minority, and 13.4% of their students were identified as English language learners. A mere 5.4% of their total student population qualified for free and reduced lunch. In Center, roughly 95% of their students were identified as racial and/or ethnic minorities. Over half of their students were identified as English language learners, and the vast majority of students (91%) qualified for free and reduced lunch. The divide between urban/suburban and rural communities is one that is often examined, and with good reason. Nevertheless, the differences found within rural


9
communities can also be vast - as illustrated by the case of Aspen 1 and Center 26 JT. If the denizens of urban/suburban and rural municipalities can be described as “of two worlds,” what phrase should be used to depict the cavernous split that separates towns like Aspen and Center? Of two worlds? Perhaps there are three.
This study seeks to enhance the current literature on student achievement by examining a snapshot of Colorado’s academic success - the third grade reading scores on the 2014 Transitional Colorado Assessment Program - and discovering whether or not the degree of a school district’s rurality has an impact on student performance. Degree of rurality was coded based on a school district’s size and geographic location, as either Urban/Suburban, Rural, or Small Rural. The Colorado Department of Education (CDE) determines rural school districts based on the geographical size of the district, the distance from the nearest urban area, and enrolling approximately 6,500 students or fewer. By this definition, 148 school districts in Colorado qualify as rural school districts. To complicate things further, 109 of the 148 meet the definition of small rural, meaning they enroll less than 1,000 students and are significantly geographically isolated. Eighty-eight of these small rural districts have a total student enrollment of less than 500 students. All together, these 148 rural districts account for 16% percent of Colorado’s total student population - or roughly 136,000 students.
The literature often describes the differences in rural and urban academic achievement as dichotomous, when this is not necessarily the case. The wide variety of academic achievement found in public schools across the nation might be better described as the rural to urban continuum. This classification takes into account the variation between the school experiences of students in rural, suburban, and urban districts. For this particular study, it’s important to note that both the resources


10
available to families and the stressors they encounter vary systematically across locations, and thus, influence and shape children’s development (Voltruba-Drzal, Miller, and Coley, 2016). Community resources, like child-care centers, museums, libraries, hospitals, and public transportation are generally greater in urban areas than in suburban or rural communities (Licter, 2012). Rural communities can be conceptualized as having limited resources and moderate level stressors, when compared to urban and suburban areas (Voltruba-Drzal, Miller, and Coley, 2016).
In public education, poorly performing schools tend to serve the most impacted populations in the United States. Throughout the country, “impacted populations” are often found in rural communities. Students who attend school in rural areas exhibit lower educational achievement and are more likely to not finish high school when compared to their suburban peers (Roscigno, Tomaskovic-Devey, Crowley, 2006). Families and schools in both inner city and remote communities lack many of the aforementioned resources that promote academic achievement and the attainment of higher education (Roscigno, Tomaskovic-Devey, Crowley, 2006). Rural schools in particular often have high concentrations of students living in poverty, fewer educational resources, and lower per-pupil expenditures (Roscigno, Tomaskovic-Devey, Crowley, 2006).
Early math and literacy skills are the strongest predictors of future academic success in young children, regardless of their geographic location (Duncan et al., 2007). Worryingly, children in rural areas appear to enter kindergarten with academic skills that are significantly behind their urban and suburban counterparts (Miller & Voltruba-Drzal, 2013). Following this trend, the reading levels of both rural and urban third graders are lower on average than those of children attending schools in suburban


11
districts (Graham & Teague, 2011). While reading achievement is important for every student, it appears to be particularly important for rural children. One study on early literacy found that rural students who were struggling readers when they started kindergarten demonstrated lower average reading achievement in third grade than both urban and suburban students. (Graham & Teague, 2011). This achievement gap between rural third graders reading abilities and their urban/suburban peers persisted even after controlling for socioeconomic status.
For the purposes of this study, the author limited the scope to include the aforementioned predictors of student achievement, as well as added the variable of total student enrollment. Total Student Enrollment was included as a variable to further parse out the difference between rural districts and small districts. While it is true that the majority of rural districts have smaller student enrollments than urban districts, not every small district in the state of Colorado can be accurately identified as rural. For example, while Englewood 1 had a total student population of 2,627 in the 2013 - 2014 school year, it is classified as an Urban/Suburban school district due to its close geographic proximity to Denver.
Rural student achievement is well documented by the literature, but in a dichotomous way that pits rurality against urbanity. There are relatively few studies that examine to which the degree of rurality has an effect on learning. Additionally, the majority of studies focused on rurality turn their attention to the rural southeast of the United States, and Rural Colorado is often overlooked. This research aims to add to the current knowledge on student achievement by determining if the degree of a school district’s rurality has a relationship with student achievement.


12
CHAPTER II LITERATURE REVIEW No Child Left Behind Act
Student achievement and school accountability have been the prevailing themes surrounding education reform for the majority of the 21st century. The discussion encompassing student achievement generally begins by acknowledging the far reaching effects of the No Child Left Behind Act (NCLB) (2001). The focus on student achievement measured via high-stakes testing was born out of NCLB’s mandate that public school students be regularly assessed in the areas of reading and mathematics. The NCLB Act significantly increased the federal government’s role in holding states and schools accountable for the academic performance of their students (2001). Critically, NCLB focused on ensuring that the performance of certain subgroups of students - children whose first language was not English, racial and ethnic minorities, children with disabilities, and children from low income families - was monitored (No Child Left Behind of 2001). The ultimate goal of NCLB was to improve test outcomes for the lowest achieving students and ensure that all students reached the “proficient level” on state tests by the 2013-2014 school year. By the 2013 deadline, no state was able to reach 100% proficiency for all of their students (McMurrer, J., & Yoshioka, N., 2013).
No Child Left Behind also mandated that schools hire “highly qualified” teachers and demonstrate “adequate yearly progress” in both reading and math achievement (2001). By definition, highly qualified teachers possess both a bachelor’s degree in the subject they teach and a current state teaching certificate. Adequate yearly progress (AYP) is measured by state defined targets of appropriate student growth. According to


13
NCLB (2001), in order for a district/school to make AYP, all of the following requirements must be met:
1. Achieve a 95% participation rate in state reading and math assessments.
2. Reach targets for either proficiency or decrease non-proficiency in reading and math.
3. Reach targets for one other indicator - advanced level of performance for elementary and middle schools in reading or math, or graduation rate for high schools.
NCLB has been historically criticized for overemphasizing standardized testing, lacking funding, and placing an unreasonable burden on rural schools (Reeves, 2003, Powell, D., Higgins, H. J., Aram, R., & Freed, A. (2009). In smaller districts, one student’s test score may have a disproportionately large impact on the district’s reported achievement (Powell, D., Higgins, H. J., Aran, R. & Freed, A., 2009). Throughout the United States, rural school districts have struggled with declining enrollment numbers and a difficulty with recruiting and retaining highly qualified teachers (U.S. Department of Agriculture, 2004). Therefore, rural schools found it more difficult than their suburban and urban counterparts to meet the mandates outlined by NCLB (Powell, D., Higgins, H. J., Aran, R. & Freed, A., 2009).
Every Student Succeeds Act
In response to the criticisms and shortcomings of the No Child Left Behind Act, it was replaced in 2015 by the Every Student Succeeds Act (ESSA). This legislation rolled back several of NCLB’s mandates and gave more power to the individual states to determine and define what success looks like for their student population (Every Student Succeeds Act, 2015). Key changes from NCLB include state designed


14
accountability goals, a new definition of “challenging” academic standards, and removing NCLB’s mandate for highly qualified teachers (Every Student Succeeds Act, 2015). States are still required to test students’ reading and math achievement in grades three through eight, as well as once in high school. Similar to NCLB, ESSA continues to obligate states to address students’ reading and math proficiency, high school graduation rates, English-language proficiency, and at least “one additional indicator” of student success (Every Student Succeeds Act, 2015). These other indicators can include student engagement, educator engagement, access to and completion of advanced coursework, college readiness, school climate/safety, or any other indicator proposed by the state. Low performing schools, as defined by schools in the bottom five percent or with graduation rates lower than 67%, are required to be identified every three years and placed on a comprehensive improvement plan (Every Student Succeeds Act, 2015). In order to receive federal funding, schools and districts must continue to demonstrate student growth and improvement. If after four years a school does not achieve measurable student gains, the state is required to step in and implement its own improvement plan (Every Student Succeeds Act, 2015). The changes outlined by the Every Student Succeeds Act take full effect in the 2017-2018 school year.
History of Standardized Assessment in Colorado The history of standardized assessment in the state of Colorado began in 1993, with the passage of House Bill 93-1313. This law mandated that the state develop content standards in multiple subject areas as well as a designated procedure for assessing student achievement (Colorado Revised Statutes). The Colorado Student Assessment Program (CSAP) was created as a direct response to HB 93-1313. Statewide, CSAP was first administered in 1997 to measure the reading and writing ability of fourth


15
graders. Student performance was defined as unsatisfactory, partially proficient, proficient, or advanced. CSAP results were then aggregated into district and statewide reports, which provided data on the percentage of students who were able to meet the proficient or advanced standard. This data was further broken down into other demographic variables, including gender, race/ethnicity, percentage of students receiving free and reduced lunch, percentage of English language learners, and percentage of students with disabilities. Until 2011, the CSAP was administered to students in grades three through ten in reading and writing, grades five through ten in math, and grade eight in science.
In 2011, the Colorado Student Assessment Program was replaced by the Transitional Colorado Assessment Program (TCAP) in response to new academic content standards developed by the Colorado Department of Education, the Colorado teaching community and CTB/McGraw-Hill (Colorado Department of Education, 2014). TCAP was designed to bridge the gap between assessment of the old Colorado Model Content Standards (1997) and the new Colorado Academic Standards, which were adopted in August 2011 (Colorado Department of Education, 2014). TCAP was administered on the same schedule as CSAP until 2014, when it was phased out in favor of an assessment developed by the Partnership for Assessment of Readiness for College and Careers (PARCC).
Started with seed money from the federal government, PARCC was one of two multi-state efforts that built new tests to measure how well students were learning the Common Core standards that were developed in 2009. In 2015, Colorado joined 12 other states in administering the inaugural PARCC designed exams. In Colorado, this exam was dubbed the Colorado Measures of Academic Success, or CMAS exam. CMAS


16
exams are currently administered to grades three through eight to measure math and English/Language Arts achievement, grades five, eight, and n to measure science achievement, and grades four and seven to measure achievement in social studies (Colorado Department of Education, 2017).
Starting in 2018, the state of Colorado has announced that it will be shifting away from standardized tests developed via PARCC and toward tests developed mostly by Colorado educators (Garcia, 2017). In addition, Colorado will once again re-examine its academic standards in 2018, which will likely influence additional changes to its standardized assessments. Due to multiple significant changes in how the state of Colorado measures student achievement throughout the past 20 years, it is difficult - if not impossible - to compare current and past student achievement.
Predictors of Rural Student Achievement Approximately 2.6 million - or 20 percent of the children in rural American communities are poor (U.S. Department of Agriculture, 2004). The poverty faced by these children in rural areas is especially fraught - it is often intergenerational, and disproportionately affects ethnic and racial minorities (Lichter and Johnson, 2007). Rural poverty has been shown to increase the odds that students struggle with educational problems, including underachievement and high school dropout (Crosnoe et al. 2002; Farmer et al. 2006; McLoyd 1998). Limited community resources, differing views on the value of higher education, geographic isolation, and family economic hardship often combine to negatively affect students in rural school districts (Crockett et al. 2000; Duncan 2001; Elder and Conger 2000; Farmer et al. 2006; Hardre 2009).
Conversely, rural schools also possess protective characteristics that can lead to more positive outcomes for students, despite low socioeconomic status (Barley and


17
Beesly, 2007). Rural schools often help promote a sense of community and foster supportive relationships between students, teachers, and local community members (Burney and Cross, 2006; Lyson, 2002). Additionally, smaller class sizes have shown to positively influence education outcomes for rural children (Nye et al., 2000). When compared to students who are not at-risk, some studies suggest that the teacher to student ratios found in smaller schools may be more beneficial for youth at-risk for underachievement - due to poverty, previous low achievement, and racial/ethnic minority background (Konstantopoulos and Chung, 2009, Nye et al., 2004).
The demographic composition of rural schools (e.g., percentage of students receiving free/reduced lunch, percentage of students who are racial/ethnic minorities) can also affect student achievement. In 2010,19% of rural children lived in poverty, compared to 15% percent of suburban children (National Center for Education Statistics, 2013). The proportion of rural students attending schools that have more than 50% of children eligible for free/reduced lunch is higher in remote rural areas (45%) than the national average (35%) (National Center for Education Statistics, 2013). However, it should be noted that there is a wide range of variability throughout America’s rural school districts (Provasnik et al., 2007). Additionally, a higher percentage of rural students from ethnic/racial minority backgrounds (African American, 60%; Hispanic/Latino, 54%; Native American 69%) attend schools that have more than half of their students eligible for free and reduced lunch when compared to their White rural peers (21%) (Provasnik et al., 2007).
Geographic isolation also plays a significant role in the academic achievement of rural students. Schools in remote locations often struggle to recruit and retain highly qualified teachers, especially in the areas of math, science, and special education (Barley


18
and Brigham, 2008; Monk, 2007). Rural youth also tend to have lower education aspirations because a college education is not required for many local job opportunities in rural economies. This effect is compounded by the fact that for most rural students, pursuing higher education comes at the cost of leaving their home communities. Literature on the subject suggests that students who attend rural schools may lower their aspirations in order to maintain their connections to their families and communities (Hektner 1995; Howley, 2006; Irvin et al., 2011).
When examining student achievement, the percentage of students identified with disabilities is also generally included, however it has been excluded from the data set used for this research. For the purposes of this study, the percentage of students identified with disabilities was removed as a variable because it is generally constant across state (10% of students identified, CDE 2016) and national lines (13% of students identified, NCES, 2015) and does not appear to change noticeably based on where students are located.
Third grade Transitional Colorado Assessment Program scores were chosen as the measurement of student achievement due the strong association between early reading success and long term positive academic achievement (Mol & Bus, 2011). Findings from Sparks et al. indicated that first grade reading comprehension was a significant predictor of eleventh grade reading achievement (2014). According to the literature, early reading skills appear to “snowball” over time and lead to more growth in the areas of language ability, vocabulary, spelling, and comprehension (Sparks et al., 2014). Reading comprehension was also found to be positively correlated with math achievement, which suggests that mastery of early reading skills is important for future success in math (Grimm, 2008). Mol and Bus discovered a moderate positive


19
relationship between print exposure and academic achievement, which suggests that more frequent readers become more successful students (2011). Additionally, the U.S. Department of Education has placed particular emphasis on reading achievement as “reading opens the door to learning about math, history, science, literature, geography and much more. Thus, young capable readers can succeed in these subjects, take advantage of other opportunities, and develop confidence in their own abilities” (The U.S. Department of Education, 2003, p. 28).


20
CHAPTER III METHODS
This study utilized a non-experimental, quantitative research methodology to investigate both the relationship and the strength of the relationship between rurality and student achievement using correlation analysis and multiple linear regression. Pearson’s r correlation coefficient was used to measure the strength of the relationship between rurality and student achievement. Additionally, Pearson’s r determines a positive or negative correlation between the variables. According to Chen and Popovich (2002), the Pearson’s r correlation coefficient is a critically important parametric measure of association. Additionally, Multiple linear regression analysis was used to illuminate the predictive strength of the independent variables on the dependent variable. Regression analysis provided a quantitative explanation as to the effect of the independent variables on the dependent variable. This methodology was chosen as it is commonly used in educational research and there are a wealth of studies that provide models for the design of this research (Dunagin, 2006; Pan et al., 2003; Siegrist, Weeks, Pate, & Monetti, 2009).
Independent variables include rurality of a school district, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body. Rurality of a school district was defined using the guidelines set out by the Colorado Department of Education. The dependent variable is student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). The statistical analyses used to examine the interplay of these variables provided the evidence required to answer the research questions posed by this study.


21
Research Questions and Hypotheses
Research Question 1: Can the set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body be used to predict student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP)?
Alternative Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Null Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are not significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Research Question 2: Does a positive linear relationship exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP)?


22
Alternative Hypothesis: A positive linear relationship does exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Null Hypothesis: A positive linear relationship does not exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Target Population
This study targeted third graders from Colorado’s rural public school districts. According to October Count 2014, Colorado is home to 178 school districts with a total student population of 889,006 (Colorado Department of Education, 2014). Districts ranged in size from five students (Agate 300 School District) to 81,131 students (Denver Public Schools) (Colorado Department of Education, 2014). This study’s target population focused on rural school districts in particular. The Colorado Department of Education (CDE) determines rural school districts based on the geographical size of the district, the distance from the nearest urban area, and enrollment data (approximately 6,500 students or fewer). By this definition, 148 school districts in Colorado qualify as rural school districts. Colorado’s rural school districts were home to 30,139 students in the 2013 - 2014 school year (Colorado Department of Education, 2014). According to 2014 October Count data, approximately 37% of Colorado’s total student population attended a rural school district in the 2013 - 2014 school year (Colorado Department of Education, 2014)
Sampling Method
All 178 school districts were available for inclusion in this study. However, due to privacy concerns, the Colorado Department of Education does not report scores on


23
standardized assessments if a cohort of 16 or fewer students took the state exam. Because of this, third grade achievement on the 2014 Transitional Colorado Assessment Program was not reported by CDE for 42 school districts. Each of these 42 school districts qualified as small rural districts, meaning they enroll less than 1,000 students and are separated by considerable distance from the nearest urban/urbanized area (Colorado Department of Education, 2017). After excluding the districts for which CDE does not report student achievement data, 133 school districts were analyzed. With that being said, this study examined the majority of the available school district data and can provide a generalization for all Colorado public school districts. However, it should be noted that a large amount of geographic, racial/ethnic, and socioeconomic diversity exists between school districts that meet the qualification of rural. The Colorado Department of Education supplies student achievement data aggregated via school districts on a yearly basis. Every school district in the state is required to provide CDE with the information contained in this study. No permission was required to acquire or analyze this data, as it is available to the public. No individual student records were used for this research.
Data Collection and Analyses
Transitional Colorado Assessment Program third grade reading scores for the Spring 2014 administration were accessed via CDE at
https://www.cde.state.co.us/assessment/cmas-dataandresults. The data was then downloaded in the form of an Excel spreadsheet. Historical supplementary data regarding percentage of students receiving free and reduced lunch, percentage of racial and ethnic minorities, and percentage of English Language Learners was accessed using CDE’s District Dashboard, which can be found at


24
http://www2.cde.state.co.us/schoolview/dish/dashboard.asp. In order to answer the two research questions, correlation and multiple linear regression were conducted. All data used was publicly accessible. These statistical analyses illuminated the relationship between rurality and student achievement, as well as the predictive intensity of the independent variables. SPSS 24 Regression was used to conduct the data analysis.


25
CHAPTER IV RESULTS
This study examined the student achievement of 133 Colorado school districts using the Spring 2014 Transitional Colorado Assessment Program (TCAP) administration. The predictor variables of rurality, total student enrollment, percentage of English language learners, percentage of racial/ethnic minorities, and percentage of students that qualify for free and reduced lunch were examined. Analysis was conducted on these variables to determine if they exert influence on overall student achievement. The range of district populations varied from a total student population of 173 (Moffat County RE 1) to 82,942 (Jefferson County School District). The mean student enrollment for the 2013-2014 school year was 6,139 students, with a median of 1,002 students. The dependent variable, student achievement, was measured using the percentage of third grade students in the district who earned a score of proficient or advanced on the 2014 TCAP examination. The percentage of students who scored proficient or advanced ranged from 25.9% (Lake County R1) to 100.0% (Swink 33). The mean for the selected districts was 72.93% with a median of 74.68% and a mode of 66.67%.
Districts were coded based on CDE’s definition of rural school districts. Depending on student population size and geographic location, districts were assigned a “1” for Urban/Metro, a “2” for Rural, and a “3” for Small Rural. Twenty-nine school districts, or 21.8% of the school districts examined were coded as Urban/Metro. Forty-three districts from the selected data set were coded as Rural, totaling 32.3% of the overall sample. Lastly, 61 school districts were coded as Small Rural, which accounts for 45.9% of the districts analyzed.


The overall percentage of English language learners ranged from o.o% (Big Sandy 100J, North Conejos RE lJ, Cheraw 31, Dolores County RE No. 2) to 57.9%
26
(Center 26 JT) with a mean of 12.74% ELL, a median of 7.7% ELL, and a mode of 0.0% ELL. The percentage of students eligible to receive Free and/or Reduced Lunch ranged from 5.6% (Aspen 1) to 91.7% (Center 26 JT) with a mean of 48.4%, a median of 48.2%, and a mode of 53.5%. The percentage of students identified as racial or ethnic minorities in the 133 districts selected ranged from 1.6% (Me Clave RE 2) to 94.7 (Center 26 JT). Table 1 shows the descriptive data for the aforementioned variables.
TABLE 1:
Rurality, Achievement, and Predictor Variable Descriptive Statistics
N Range Min Max Mean Std. Dv Var Skew Std. Err
Rurality 133 2 1 3 2.24 0.79 0.62 -0.45 0.21
TCAP 2014 Percent Proficient & Advanced 133 74-03 25-97 100.00 72.93 11.97 143.18 -0.85 0.21
Percent Free & Reduced Lunch 133 86.1 5.60 91.70 48.47 18.65 347-73 0.12 0.21
Percent Minority 133 93-t 1.6 94-7 36.55 22.01 484.71 0.74 0.21
Percent English Language Learners 133 57-9 0.0 57-9 12.74 13-65 186.22 1.48 0.21
Total Student Enrollment 133 82,769 173 82,942 6,189 13,852 191896 3-72 0.21


27
Correlation and Multiple Regression Analyses:
Analysis of Research Question 1
RQ 1: Can the set of variables including rurality of a school district, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body be used to predict student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP)?
Alternative Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Null Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are not significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP).
The regression equation used to analyze this research question was: 2014 TCAP third grade Reading Score (yi) = constant + rurality (coefficient) + total student enrollment (coefficient) + % of English language learner (coefficient) + % of students identified as racial/ethnic minorities (coefficient) + % of students qualified for Free and Reduced Lunch (coefficient). SPSS 24 was used to perform the data analysis.


28
For research question one, the null hypothesis was rejected and the alternative hypothesis was supported. The aforementioned independent variables are significant predictors of third grade reading scores as measured by the 2014 Transitional Colorado Assessment Program (TCAP). Assumptions were tested by visual examination of the normal probability plots of residuals and scatter plots of residuals versus predicted residuals. There were no identified violations of normality, linearity, and/or homoscedasticity (see Figure 1). The Durbin-Watson statistic of 1.878 revealed that the residuals were truly independent of one another.


Expected Cum Prob
29
Normal Probability P Plot of Regression Standardized
Residual
Dependent Variable: TCAP 2014 Percent Proficient and Advanced
Figure 1: Normal Probability Plot


30
Regression analysis indicated that the model significantly predicted 2014 third grade TCAP reading scores, F (10.448) = .000, p < .001 (see Table 2). As shown in Table 3, R2 for the model was .291, with an adjusted R2 of .264, which indicates that approximately 26% of the variance found in 2014 third grade TCAP reading scores can be predicted using the independent variables of rurality, total student enrollment, the percentage of students identified as racial/ethnic minorities, the percentage of students identified as English language learners, and the percentage of students qualified for free and reduced lunch. While the model is significant, 73% of the variance found in student achievement is not accounted for by this linear multiple regression model.
TABLE 2: Regression Analysis Model Table
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-
Square Estimate Watson
1 0.540 0.291 0.264 10.26866 1.878
a. Predictors: (Constant), Total Student Enrollment, Percent Free and Reduced Lunch,
Percent English Language Learners, Rurality, Percent Minority
b. Dependent Variable: TCAP 2014 Percent Proficient and Advanced
c.
ANOVA
Model Sum of Squares df Mean Square F Sig
Regression 5508.61 5 1101.72 10.448 0.000
Residual 13391-56 127 105-45
Total 18900.18 132


31
With regards to the individual relationships between the independent variables and third grade TCAP reading scores, three variables significantly predicted reading scores. Rurality (t = 2.263, P > -05), the percentage of English Language Learners (t = -2.764, p < .01), and the percentage of students eligible for free and reduced lunch (t = -3-358, p < .05) were all significant predictors of student achievement.
Analysis of Research Question 2
Research Question 2: Does a positive linear relationship exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP)?
Alternative Hypothesis: A positive linear relationship does exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Null Hypothesis: A positive linear relationship does not exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP).
Research question 2 examined the impact of one independent variable, rurality, on its dependent variable, 2014 third grade TCAP reading scores. In this hypothesis, the correlation was weakly positive (r = .169, p = .051) but failed to meet the p = <.05 threshold, as shown in Table 6. Therefore, the correlation between a school district’s rurality and third grade TCAP reading scores could not be made for the selected Colorado school districts in the 2013 - 2014 school year. For research question 2, the author fails to reject the null hypothesis.


32
Table 3: Correlation Coefficients
Correlation Coefficients
Rurality TCAP 2014 Percent Proficient & Advanced Percent Free and Reduced Lunch Percent Minority Percent English Language Learners Total Student Enrollm ent
Rurality Pearson Correlation 1 0.169 0.151 -.289** -.252** -.603**
TCAP 2014 Percent Proficient & Advanced Pearson Correlation 0.169 1 -.431** -.422** -.438** -0.043
Percent Free and Reduced Lunch Pearson Correlation 0.151 -431** 1 .678** .442** -0.126
Percent Minority Pearson Correlation -.289** -.422** .678** 1 .679** 0.153
Percent English Language Learners Pearson Correlation -.252** -.438** .442** .679** 1 .189s
Total Student Enrollment Pearson Correlation -.603** -0.043 -0.126 0.153 .189s 1
** indicates correlation is significant at the o.oi Level (2-tailed) * indicates correlation is significant at the 0.05 level (2 tailed)


33
CHAPTER V DISCUSSION
Research Question 1 focused on socioeconomic and linguistic influences that can predict outcomes on standardized assessments. Higher 2014 TCAP reading scores were associated with urban/suburban school districts, a smaller overall percentage of students receiving free and reduced lunch, and a smaller overall percentage of English language learners. While the overall model is significant, the variables of total student enrollment and percentage of racial and ethnic minorities were not significant predictors of student achievement. Overall, 26.4% of the variance in 2014 TCAP reading scores could be predicted using the variables of rurality, percentage of English language learners, and the percentage of students that qualify for free and reduced lunch. Seventy-three percent of the variance found in student achievement as measured by third grade TCAP reading scores is not accounted for by this linear multiple regression model. This study did not uncover other independent variables that exert the majority of influence over third grade reading scores. Additionally, the variables of rurality, the percentage of English language learners, and the percentage of students qualified for free and reduced lunch can only be considered as weak predictors (r squared = .264) of the variance in TCAP reading scores.
Research Question 2 asked if the degree of a school district’s rurality could be positively correlated with TCAP reading scores for third grade students in the 2013 -2014 school year. For this hypothesis, the correlation could not be made as the analysis indicated it narrowly missed the p< .05 threshold, at p = .051. This could be further examined using a bigger data set, for example, using TCAP reading scores from third through tenth graders. This research addresses a gap in the literature by indicating that


34
the degree of rurality for Colorado’s public school districts does not appear to be correlated with academic achievement.
While not directly related to Research Question 2, the correlation performed still provided interesting data to examine. The degree of rurality was associated with lower total student enrollment (r = - .603, p < .01), fewer English language learners (r = - .252, p < .01), and fewer students identified as racial or ethnic minorities (r = .289, p < .01). The variable of percent free and reduced lunch (r = -.431, p < .01), percent racial and ethnic minorities (r = - .422, p < .01)., and percent English language learners (r = -•438, p < .01) were moderately correlated with lower student achievement, as measured by third grade reading scores. According to the correlation analysis, increased percentages of students identified as racial or ethnic minorities was strongly correlated with increased percentages of students receiving free and reduced lunch (r = .678, p < .01) and higher numbers of English language learners (r = .679, p < .01). This tracks with demographic data from the state of Colorado, as Hispanic/Latinx students are the biggest minority group in the state. Additionally, the majority of English learners in Colorado identify as Hispanic/Latinx and share Spanish as a home language (Office of Culturally and Linguistically Diverse Education).
Limitations
The scope of this research was limited by time, duration, and definition. The questions posed were fairly narrow and only concerned the specific achievement of students (third grade reading TCAP scores), through the use of specific methodology (quantitative; correlation and multiple linear regression), for only one school year (
2013 - 2014), and for publicly funded school districts in the state of Colorado.


35
The multiple linear regression performed indicated a small influence of the chosen socioeconomic predictor variables on the dependent variable of student achievement. Examining other independent variables via correlation and multiple linear regression could lead to a better understanding of what ultimately impacts student achievement. This study in particular did not define what other factors may influence the predictability of academic achievement. While the quantitative methodology utilized in this study was able to identify relationships between variables, it does not explain why these relationships exist. For example, simple correlational analyses identified that rural school districts have relatively fewer amounts of students who are members of cultural and linguistic minority groups, but it does not illuminate the causation behind these relationships. In the future, researchers seeking more definitive answers to these questions should explore other analyses.
Utilizing publicly aggregated data, while convenient and expedient, was not without its faults. Due to student privacy concerns, CDE masks achievement data when they have less than 16 results per test, per district. When considering small districts, this amounts to a significant portion of the available achievement data and is the reason why 42 small rural school districts had to be excluded from this study’s data set. In 2015,
CDE also began using new, more stringent suppression rules for achievement data. Starting in 2015, if fewer than four students either did or did not meet the expectations of the test, the achievement data is not provided. According to CDE, this accounts for a missing 6% of district level student achievement data and a missing 9% of school level student achievement data. In 2016, 36% of school and district student achievement data was masked. This is especially concerning when you take into account that the achievement data of specific groups of students is not being released due to small cohort


36
sizes. While it is critically important to examine the student achievement of low-income students, English language learners, racial/ethnic minorities, and students identified with a disability, gathering this information is often at odds with students right to privacy.
In addition, using the percentage of students identified as racial/ethnic minorities and English language learners could be construed as an oversimplification of these two categories. Percentage of English language learners can be further broken down into percentage of Non-English Proficient (NEP), Limited English Proficient (LEP), and Former English language learners (FELL) students, which would more thoroughly examine the relationship between the levels of language acquisition and development with its impact on student achievement. Relatedly, using the percentage of racial/ethnic minorities in a district student population can be problematic for similar reasons. First, it assumes that the majority of students in a school district are identified as White/Non-Hispanic, which is not the case for 35 of the schools districts examined in this data set. Identifying all non-white students as simply “minority students” instead of using the more specific identifiers of Black, Hispanic/Latinx, American Indian or Alaskan Native, Asian, or Two or More Races also deemphasizes the differences in student achievement found within these racial/ethnic groups. Further research on the topic would benefit from a more specific breakdown of both racial and ethnic groups as well as differentiation between the levels of English language acquisition/development.
This current study was limited in scope as it only examined one measure of student achievement by using the 2014 TCAP reading assessment scores. Student achievement could also be defined using high school graduation rates, high school dropout rates, college readiness, ACT/SAT exam scores, or state standardized math,


37
science, or social studies scores. In addition, this study provided only a snapshot for student achievement by looking at one year’s results. Student achievement could also be measured using a longitudinal growth model that focuses on whether or not students, schools, and districts are making Adequate Yearly Progress (AYP).
Lastly, one of the central variables of this study - rurality - is an incredibly fraught concept to operationalize. CDE’s identification of rural/small rural was used, but there are also more detailed models of rurality that could have been utilized. For example, CDE also organizes their schools districts by settings that may prove to be more descriptive for further research. These settings include the classifications of Denver-Metro, Urban-Suburban, Outlying City, Outlying Town, Remote, Colorado Board of Collaborative Educational Services (BOCES), and Detention Centers. In the future, it could be illuminating to examine student achievement based on these more specific geographic and academic parameters.
Conclusion
This study explored socioeconomic and linguistic influences in relationship to student achievement outcomes. The purpose of this research was to determine if a correlation between the degree of rurality in Colorado’s public school districts and third grade reading achievement as measured by the Transitional Colorado Assessment Program (TCAP) could be made. A statistically significant correlation was not identified for the 2013 - 2014 school year. This suggests that there is not a significant relationship between the degree of rurality in Colorado’s public school districts and overall student achievement. The predictor relationship between five independent variables (degree of rurality, total student enrollment, percentage of racial and ethnic minorities, percentage of English language learners, percentage of students qualified for free and reduced


38
lunch) and student achievement was verified. The regression analysis indicated a weak level of predictability for the independent variables, and accounted for roughly 26% of the overall variance in student achievement. Lastly, considering the context of high-stakes testing and educational reform, this study calls for additional research to better understand the predictor variables and their relationship with Colorado’s student achievement.


39
REFERENCES
Barley, Z. A., & Beesley, A. D. (2007). Rural school success: What can we learn? Journal of Research in Rural Education, 22(1). Retrieved from http://jrre.psu.edu/ articles/ 22-1.pdf
Barley, Z. A., & Brigham, N. (2008). Preparing teachers to teach in rural schools (Issues & Answers Report, REL 2008-N0. 045). Washington, DC: U.S. Department of Education. Retrieved from
http://ies.ed.g0v/ncee/edlabs/regi0ns/central/pdf/REL_2008045_ sum.pdf
Bill, Colorado House Bill 93-1313. Colorado Revised Statutes, 22-53.
Burney, V. H., & Cross, T. L. (2006). Impoverished students with academic promise in rural settings: 10 lessons from project aspire. Gifted Child Today, 29(2), 14-21. doi:io.42i9/gct-20o6- 200
Colorado Department of Education. TCAP - Mathematics, English Language Arts, Science and Social Studies Data and Results. (2014). Retrieved December 4, 2017, from https://www.cde.state.co.us/assessment/cmas-dataandresults
Colorado Department of Education. Pupil Membership. (2016). Retrieved December 4, 2017, from https://www.cde.state.co.us/cdereval/pupilcurrent


40
Rural and Small Rural Designation List[XLSX]. (2017, January). Colorado Department of Education.
Colorado Senate. (2010). Senate Bill 10-191. Retrieved from
http://www.leg.state.co.us/clics/clics2010a/csl.nsf/fsbillcont3/EFs3BB67D4734 2CF 8972576A8ooo27Bo78?open&file=i9i_enr.pdf.
Crockett, L. J., Shanahan, M. J., & Jackson-Newsom, J. (2000). Rural youth: Ecological and life course perspectives. In R. Monte- mayor, G. R. Adams, & T. P. Gullotta (Eds.), Adolescent diversity in ethnic, economic, and cultural contexts: Advances in adolescent development (Vol. 10, pp. 43-74). Thousand Oaks, CA: Sage Publications.
Crosnoe, R., Mistry, R., & Elder, G. H., Jr. (2002). Economic disadvantage, family
dynamics, and adolescent enrollment in higher education. Journal of Marriage and Family, 64, 690-702. doi:io.nn/j.1741-3737.2002.00690.x
Dunagin, P. (2006). An analysis of specific factors that relate to the dropout rate of
Hispanic students in Texas public school (Doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Database. (3226795).
Duncan, C. M. (2001). Social capital in America’s poor rural communities. In S. Saegert, J. P. Thompson, & M. R. Warren (Eds.), Social capital and poor communities (pp. 60-86). New York, NY: Russell Sage Foundation.


41
Every Student Succeeds Act (ESSA). (2015). Retrieved December 4, 2017, from https://www.ed.gov/essa
Farmer, T. W., Leung, M.-C., Banks, J., Schaefer, V., Andrews, B., & Murray, R. A.
(2006). Adequate yearly progress in Small Rural Schools and Rural Low-Income Schools. Rural Educator, 27(3), 1-7.
Garcia, N. (2017, June 15). Colorado forges its own path with standardized tests, will stop giving PARCC English and math tests. Retrieved December 10, 2017, from https://www.denverpost.com/2017/06/15/colorado-standardized-english-math-tests/
Graham, S. E., & Teague, C. (2011). Reading levels of rural and urban third graders lag behind their suburban peers.
Grimm, K. J. (2008). Longitudinal associations between reading and mathematics achievement. Developmental Neuropsychology, 33(3), 410-426.
Hardre, P. L., & Hennessey, M. N. (2010). Two rural worlds: Differences of rural high school students’ motivational profiles in Indiana and Colorado. Journal of Research in Rural Education, 25(8). Retrieved from http://www.jrre.psu.edu/articles/ 25-8.pdf


42
Hardre, P. L., Sullivan, D. W., & Crowson, H. M. (2009). Student characteristics and
motivation in rural high schools. Journal of Research in Rural Education, 24(16). Retrieved from http:// www.jrre.psu.edu/articles/24-16.pdf
Hektner, J. M. (1995). When moving up implies moving out: Rural adolescent conflict in the transition to adulthood. Journal of Research in Rural Education, 11(1), 13-14.
Howley, C. W. (2006). Remote possibilities: Rural children’s educational aspirations. Peabody Journal of Education, 82, 62-80.
Irvin, M. J., Meece, J. L., Byun, S. Y., Farmer, T. W., & Hutchins, B. C. (2011). Relationship of school context to rural youth’s educational achievement and aspirations. Journal of youth and adolescence, 40(9), 1225.
Konstantopoulos, S., & Chung, V. (2009). What are the long-term effects of small classes on the achievement gap? Evidence from the lasting benefits study. American Journal of Education, 116,125-154. doi:io.1086/605103
Lichter, D. T., McLaughlin, D. K., & Cornwell, G. T. (1995). Migration and the loss of human resources in rural America. In L. J. Beaulieu & D. Mulkey (Eds.),
Investing in people: The human capital needs of rural America (pp. 235-256). Boulder, CO: Westview Press.


43
Lichter, D. T., & Johnson, K. M. (2007). The changing spatial concentration of America’s rural poor population. Rural Sociology, 72, 331-358. doi:io.1526/003601107781799290
Lichter, D. T. (2012). Immigration and the new racial diversity in rural America. Rural Sociology, 77(1), 3-35.
Logan, J. R., & Burdick-Will, J. (2017). School segregation and disparities in
urban, suburban, and rural areas. The ANNALS of the American Academy of Political and Social Science, 674(1), 199-216.
Lyson, T. A. (2002). What does a school mean to a community? Assessing the social and economic benefits of schools to rural villages in New York. Journal of Research in Rural Education, 17(3), 131-137. Retrieved from http://www.jrre.psu.edu/articles/ vi 7,n3,pi3i-i37,Lyson.pdf
McLoyd, V. C. (1998). Socioeconomic disadvantage and child development. American Psychologist, 53,185-204. dokio.1037/ 0003-066X.53.2.185
McMurrer, J., & Yoshioka, N. (2013). States' Perspectives on Waivers: Relief from NCLB, Concern about Long-Term Solutions. Center on Education Policy.


44
Miller, P., Votruba-Drzal, E., & Setodji, C. M. (2013). Family income and early
achievement across the urban-rural continuum. Developmental psychology, 49(8), 1452.
Mol, S. E., & Bus, A. G. (2011). To read or not to read: a meta-analysis of print exposure from infancy to early adulthood. Psychological bulletin, 137(2), 267.
Monk, D. H. (2007). Recruiting and retaining high-quality teachers in rural areas. The Future of Children, 17,155-174.
National Center for Education Statistics (2013). The Status of Rural Education. Retrieved January 7, 2018, from http: / / nces.ed.gov/ programs/coe/indicator_tla.asp
National Center for Education Statistics (2017). The Condition of Education : Children and Youth with Disabilities Retrieved December 7, 2017, from https: / / nces. ed.gov/ programs / coe/indicator_cgg. asp
Nye, B., Hedges, L. V., & Konstantopoulos, S. (2000a). Do the disadvantaged benefit from the small classes? Evidence from the Tennessee class size experiment. American Journal of Education, 109,1-26.
Nye, B., Hedges, L. V., & Konstantopoulos, S. (2004). Do minorities experience larger
lasting benefits from small classes? Journal of Educational Research, 98, 94-100.


45
Office of Culturally and Linguistically Diverse Education, (n.d.). Retrieved December 20, 2017, from https://www.cde.state.co.us/cde_english
Pan, D., Rudo, Z., Schneider, C., & Smith-Hanson, L. (2003). Examination of resource allocation in education: Connecting spending to student performance. Southwest Educational Development Laboratory, Austin, TX.
Powell, D., Higgins, H. J., Aram, R., & Freed, A. (2009). Impact of No Child Left Behind on Curriculum and Instruction in Rural Schools. Rural Educator, 31(1), 19-28.
Provasnik, S., KewalRamani, A., Coleman, M. M., Gilbertson, L., Herring, W., & Xie, Q. (2007). Status of education in rural America (NCES 2007-040). Washington, DC: National Center for Education Statistics.
Reeves, C. (2003). Implementing the No Child Left Behind act: Implications for rural schools and districts. Retrieved June, 14, 2011.
Roscigno, V. J., Tomaskovic-Devey, D., & Crowley, M. (2006). Education and the inequalities of place. Social Forces, 84(4), 2121-2145.
Siegrist, G., Weeks, W., Pate, J., & Monetti, D. (2009). Principals’ experience,
educational level, and leadership practices as predictors of Georgia high school


46
graduation test results. Journal of Philosophy and History of Education (59), 174-179.
Sparks, R. L., Patton, J., & Murdoch, A. (2014). Early reading success and its
relationship to reading achievement and reading volume: replication of‘10 years later’. Reading and Writing, 27(1), 189-211.
U.S. Department of Agriculture. (2004). Rural poverty at a glance (Report No. 100). Retrieved from http://www.ers.usda.gov/ publications/rdrrioo/rdrrioo.pdf
U.S. Department of Education, Office of the Under Secretary (2003). No Child Left Behind: A toolkit for teachers, Washington, DC.
U.S. Congress. (2001). No Child Left Behind Act. Retrieved from: http: / / www2.ed.g0v/ nclb/landing.jhtml.


Full Text

PAGE 1

CAN RURALITY PREDICT STUDENT ACHIEVEMENT? by HAILEY EVEATT B.A., University of Denver, 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Psychology School Psychology Program 2018

PAGE 2

ii ! 2018 HAILEY EVEATT ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Doctor of Psychology degree by Hailey Eveatt has been approved for the School Psychology Program by Bryn Harris, Chair and Advisor Franci Crepeau Hobson Colette Hohnbaum Date: May 12, 2018

PAGE 4

iv Eveatt, Hailey (Psy.D., School Psychology) Can Rurality Predict Student Achievement? Thesis directed by Associate Professor Bryn Harris ABSTRACT This study seeks to enhance the current literature on student achievement by examining a snapshot of Colorado's academic success the third grade reading scores on the 2014 Transitional Colorado Assessment Program and discovering whether or not the degr ee of a school district's rurality has an impact on student performance. This study utilized a non experimental, quantitative research methodology to investigate both the relationship and the strength of the relationship between rurality and student achie vement using correlation analysis and multiple linear regression. Pearson's r correlation coefficient was used to measure the strength of the relationship between rurality and student achievement. Independent variables include rurality of a school district , percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body. Rurality of a school district was defined using the guidelines set out by the Colorado Dep artment of Education. Regression analysis indicated that the model significantly predicted roughly 26% of third grade 2014 TCAP reading scores. However, the correlation between a school district's rurality and third grade TCAP reading scores could not be m ade for the selected Colorado school districts in the 2013 2014 school year. The form and content of this abstract are approved. I recommend its publication. Approved: Bryn Harris

PAGE 5

v TABLE OF CONTENTS CHAPTER I. INTRODUCTIONÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ.ÉÉÉ 8 Vignette: The Great Divi de ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. ÉÉ.ÉÉ 8 II. LITERATURE REVIEW ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ .. É 12 No Child Left Behind Act ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ . ÉÉ .. 12 Every Student Succeeds Act ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ ..É 13 History of Standardized Asse ssment in Colorado ÉÉÉÉÉÉÉÉÉÉÉ . É 14 Predictors of Rural Student Achievement ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. 16 III. METHODS ÉÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 20 Research Questions and H ypotheses ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 21 Target Populatio n ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 22 Sampling Method ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 22 Data Collection and Ana lyses ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ.. 23 IV. RESULTS ÉÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 25 Correlation and Multiple Regression Anal ysis : Research Quest ion 1 É.. 27 Analysis of Research Ques tion 2 ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ.É 31 V. DISCUSSION ÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉ.ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 33 Limitations ÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ. 34 Conclusion ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ ÉÉÉÉÉÉÉÉÉÉÉ.É. 37 REFERENCES ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ ÉÉ ÉÉÉÉ. 39

PAGE 6

vi LIST OF TABLES TABLE 1. Rurality, Achievement, and Predictor Variabl e Descriptive Statistics ÉÉÉÉ . 26 2. Regression Analysis Model Table ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ.É.. 30 3. Correlation Coefficients ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ 32

PAGE 7

vii LIST OF FIGURES FIGURE 1. Normal Probability Plot ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ ÉÉÉÉ. ÉÉÉÉÉÉÉ 2 9

PAGE 8

8 CHAPTER I INTRODUCTION Vignette: The Great Divide The towns of Aspen and Center, Colorado are just 251 miles apart. They are both small, geographically isolated, rural communities nestled within the Rocky Mountains. Using just this information as context, a reason able observer might assume that the two towns are fairly similar comparable, even. But Aspen and Center are not just separated by the Continental Divide they are also split by a decidedly less picturesque gap. To examine this great divide, we turn to t he illuminating lens of public education. In 2013 2014, there were 1,670 students enrolled in the Aspen 1 school district, in grades preschool through 12. Center 26 JT had 617 pre kindergarten through twelfth grade students attending during that same sch ool year. In the Spring of 2014, the state of Colorado's testing window began for the annual Transitional Colorado Assessment Program (TCAP) exams. In Aspen, 88.89% of students received passing scores of proficient or advanced. In Center, 66.67% of their s tudents earned a passing score. The differences get starker from there. Roughly 13% of Aspen's student population was identified as a racial and/or ethnic minority, and 13.4% of their students were identified as English language learners. A mere 5.4% of th eir total student population qualified for free and reduced lunch. In Center, roughly 95% of their students were identified as racial and/or ethnic minorities. Over half of their students were identified as English language learners, and the vast majority of students (91%) qualified for free and reduced lunch. The divide between urban/suburban and rural communities is one that is often examined, and with good reason. Nevertheless, the differences found within rural

PAGE 9

9 communities can also be vast as illustra ted by the case of Aspen 1 and Center 26 JT. If the denizens of urban/suburban and rural municipalities can be described as "of two worlds," what phrase should be used to depict the cavernous split that separates towns like Aspen and Center? Of two worlds? Perhaps there are three. This study seeks to enhance the current literature on student achievement by examining a snapshot of Colorado's academic success the third grade reading scores on the 2014 Transitional Colorado Assessment Program and discove ring whether or not the degree of a school district's rurality has an impact on student performance. Degree of rurality was coded based on a school district's size and geographic location, as either Urban/Suburban, Rural, or Small Rural. The Colorado Depar tment of Education (CDE) determines rural school districts based on the geographical size of the district, the distance from the nearest urban area, and enrolling approximately 6,500 students or fewer. By this definition, 148 school districts in Colorado q ualify as rural school districts. To complicate things further, 109 of the 148 meet the definition of small rural, meaning they enroll less than 1,000 students and are significantly geographically isolated. Eighty eight of these small rural districts have a total student enrollment of less than 500 students. All together, these 148 rural districts account for 16% percent of Colorado's total student population or roughly 136,000 students. The literature often describes the differences in rural and urban academic achievement as dichotomous, when this is not necessarily the case. The wide variety of academic achievement found in public schools across the nation might be better described as the rural to urban continuum. This classification takes into accoun t the variation between the school experiences of students in rural, suburban, and urban districts. For this particular study, it's important to note that both the resources

PAGE 10

10 available to families and the stressors they encounter vary systematically across locations, and thus, influence and shape children's development (Voltruba Drzal, Miller, and Coley, 2016). Community resources, like child care centers, museums, libraries, hospitals, and public transportation are generally greater in urban areas than in suburban or rural communities (Licter, 2012). Rural communities can be conceptualized as having limited resources and moderate level stressors, when compared to urban and suburban areas (Voltruba Drzal, Miller, and Coley, 2016). In public education, poo rly performing schools tend to serve the most impacted populations in the United States. Throughout the country, "impacted populations" are often found in rural communities. Students who attend school in rural areas exhibit lower educational achievement a nd are more likely to not finish high school when compared to their suburban peers (Roscigno, Tomaskovic Devey, Crowley, 2006). Families and schools in both inner city and remote communities lack many of the aforementioned resources that promote academic a chievement and the attainment of higher education (Roscigno, Tomaskovic Devey, Crowley, 2006). Rural schools in particular often have high concentrations of students living in poverty, fewer educational resources, and lower per pupil expenditures (Roscign o, Tomaskovic Devey, Crowley, 2006). Early math and literacy skills are the strongest predictors of future academic success in young children, regardless of their geographic location (Duncan et al., 2007). Worryingly, children in rural areas appear to ent er kindergarten with academic skills that are significantly behind their urban and suburban counterparts (Miller & Voltruba Drzal, 2013). Following this trend, the reading levels of both rural and urban third graders are lower on average than those of chil dren attending schools in suburban

PAGE 11

11 districts (Graham & Teague, 2011). While reading achievement is important for every student, it appears to be particularly important for rural children. One study on early literacy found that rural students who were str uggling readers when they started kindergarten demonstrated lower average reading achievement in third grade than both urban and suburban students. (Graham & Teague, 2011). This achievement gap between rural third graders reading abilities and their urban /suburban peers persisted even after controlling for socioeconomic status. For the purposes of this study, the author limited the scope to include the aforementioned predictors of student achievement, as well as added the variable of total student enrollment. Total Student Enrollment was included as a variable to further parse ou t the difference between rural districts and small districts. While it is true that the majority of rural districts have smaller student enrollments than urban districts, not every small district in the state of Colorado can be accurately identified as rur al. For example, while Englewood 1 had a total student population of 2,627 in the 2013 2014 school year, it is classified as an Urban/Suburban school district due to its close geographic proximity to Denver. Rural student achievement is well documented by the literature, but in a dichotomous way that pits rurality against urbanity. There are relatively few studies that examine to which the degree of rurality has an effect on learning. Additionally, the majority of studies focused on rurality turn their a ttention to the rural southeast of the United States, and Rural Colorado is often overlooked. This research aims to add to the current knowledge on student achievement by determining if the degree of a school district's rurality has a relationship with stu dent achievement.

PAGE 12

12 CHAPTER II LITERATURE REVIEW No Child Left Behind Act Student achievement and school accountability have been the prevailing themes surrounding education reform for the majority of the 21st century. The discussion encompassing student ac hievement generally begins by acknowledging the far reaching effects of the No Child Left Behind Act (NCLB) (2001). The focus on student achievement measured via high stakes testing was born out of NCLB's mandate that public school students be regularly as sessed in the areas of reading and mathematics. The NCLB Act significantly increased the federal government's role in holding states and schools accountable for the academic performance of their students (2001). Critically, NCLB focused on ensuring that th e performance of certain subgroups of students children whose first language was not English, racial and ethnic minorities, children with disabilities, and children from low income families was monitored (No Child Left Behind of 2001). The ultimate go al of NCLB was to improve test outcomes for the lowest achieving students and ensure that all students reached the "proficient level" on state tests by the 2013 2014 school year. By the 2013 deadline, no state was able to reach 100% proficiency for all of their students (McMurrer, J., & Yoshioka, N., 2013). No Child Left Behind also mandated that schools hire "highly qualified" teachers and demonstrate "adequate yearly progress" in both reading and math achievement (2001). By definition, highly qualified t eachers possess both a bachelor's degree in the subject they teach and a current state teaching certificate. Adequate yearly progress (AYP) is measured by state defined targets of appropriate student growth. According to

PAGE 13

13 NCLB (2001), in order for a distric t/school to make AYP, all of the following requirements must be met: 1. ! Achieve a 95% participation rate in state reading and math assessments. 2. ! Reach targets for either proficiency or decrease non proficiency in reading and math. 3. ! Reach targets for one other indicator advanced level of performance for elementary and middle schools in reading or math, or graduation rate for high schools. NCLB has been historically criticized for over emphasizing standardized testing, lacking funding, and placing an unreasonable burden on rural schools (Reeves, 2003, Powell, D., Higgins, H. J., Aram, R., & Freed, A. (2009 ). In smaller districts, one student's test score may have a disproportionately lar ge impact on the district's reported achievement (Powell, D., Higgins, H. J., Aran, R. & Freed, A., 2009). Throughout the United States, rural school districts have struggled with declining enrollment numbers and a difficulty with recruiting and retaining highly qualified teachers (U.S. Department of Agriculture, 2004). Therefore, rural schools found it more difficult than their suburban and urban counterparts to meet the mandates outlined by NCLB (Powell, D., Higgins, H. J., Aran, R. & Freed, A., 2009). E very Student Succeeds Act In response to the criticisms and shortcomings of the No Child Left Behind Act, it was replaced in 2015 by the Every Student Succeeds Act (ESSA). This legislation rolled back several of NCLB's mandates and gave more power to the individual states to determine and define what success looks like for their student population (Every Student Succeeds Act, 2015). Key changes from NCLB include state designed

PAGE 14

14 accountability goals, a new definition of "challenging" academic standards, and removing NCLB's mandate for highly qualified teachers (Every Student Succeeds Act, 2015). States are still required to test students' reading a nd math achievement in grades three through eight, as well as once in high school. Similar to NCLB, ESSA continues to obligate states to address students' reading and math proficiency, high school graduation rates, English language proficiency, and at leas t "one additional indicator" of student success (Every Student Succeeds Act, 2015). These other indicators can include student engagement, educator engagement, access to and completion of advanced coursework, college readiness, school climate/safety, or an y other indicator proposed by the state. Low performing schools, as defined by schools in the bottom five percent or with graduation rates lower than 67%, are required to be identified every three years and placed on a comprehensive improvement plan (Every Student Succeeds Act, 2015). In order to receive federal funding, schools and districts must continue to demonstrate student growth and improvement. If after four years a school does not achieve measurable student gains, the state is required to step in a nd implement its own improvement plan (Every Student Succeeds Act, 2015). The changes outlined by the Every Student Succeeds Act take full effect in the 2017 2018 school year. History of Standardized Assessment in Colorado The history of standardized ass essment in the state of Colorado began in 1993, with the passage of House Bill 93 1313. This law mandated that the state develop content standards in multiple subject areas as well as a designated procedure for assessing student achievement (Colorado Revis ed Statutes). The Colorado Student Assessment Program (CSAP) was created as a direct response to HB 93 1313. Statewide, CSAP was first administered in 1997 to measure the reading and writing ability of fourth

PAGE 15

15 graders. Student performance was defined as uns atisfactory, partially proficient, proficient, or advanced. CSAP results were then aggregated into district and statewide reports, which provided data on the percentage of students who were able to meet the proficient or advanced standard. This data was fu rther broken down into other demographic variables, including gender, race/ethnicity, percentage of students receiving free and reduced lunch, percentage of English language learners, and percentage of students with disabilities. Until 2011, the CSAP was a dministered to students in grades three through ten in reading and writing, grades five through ten in math, and grade eight in science. In 2011, the Colorado Student Assessment Program was replaced by the Transitional Colorado Assessment Program (TCAP) in response to new academic content standards developed by the Colorado Department of Education, the Colorado teaching community and CTB/McGraw Hill (Colorado Department of Education, 2014). TCAP was designed to bridge the gap between assessment of the old Colorado Model Content Standards (1997) and the new Colorado Academic Standards, which were adopted in August 2011 (Colorado Department of Education, 2014). TCAP was administered on the same schedule as CSAP until 2014, when it was phased out in favor of an assessment developed by the Partnership for Assessment of Readiness for College and Careers (PARCC). Started with seed money from the federal government, PARCC was one of two multi state efforts that built new tests to measure how well students were l earning the Common Core standards that were developed in 2009. In 2015, Colorado joined 12 other states in administering the inaugural PARCC designed exams. In Colorado, this exam was dubbed the Colorado Measures of Academic Success, or CMAS exam. CMAS

PAGE 16

16 exa ms are currently administered to grades three through eight to measure math and English/Language Arts achievement, grades five, eight, and 11 to measure science achievement, and grades four and seven to measure achievement in social studies (Colorado Depar tment of Education, 2017). Starting in 2018, the state of Colorado has announced that it will be shifting away from standardized tests developed via PARCC and toward tests developed mostly by Colorado educators (Garcia, 2017). In addition, Colorado will once again re examine its aca demic standards in 2018, which will likely influence additional changes to its standardized assessments. Due to multiple significant changes in how the state of Colorado measures student achievement throughout the past 20 years, it is difficult if not im possible to compare current and past student achievement. Predictors of Rural Student Achievement Approximately 2.6 million or 20 percent of the children in rural American communities are poor (U.S. Department of Agriculture, 2004). The poverty faced by these children in rural areas is especially fraught it is often intergenerational, and disproportionately affects ethnic and racial minorities (Lichter and Johnson, 2007). Rural poverty has been shown to increase the odds that students struggle with e ducational problems, including underachievement and high school dropout (Crosnoe et al. 2002; Farmer et al. 2006; McLoyd 1998). Limited community resources, differing views on the value of higher education, geographic isolation, and family economic hardshi p often combine to negatively affect students in rural school districts (Crockett et al. 2000; Duncan 2001; Elder and Conger 2000; Farmer et al. 2006; HardrÂŽ 2009). Conversely, rural schools also possess protective characteristics that can lead to more po sitive outcomes for students, despite low socioeconomic status (Barley and

PAGE 17

17 Beesly, 2007). Rural schools often help promote a sense of community and foster supportive relationships between students, teachers, and local community members (Burney and Cross, 2 006; Lyson, 2002). Additionally, smaller class sizes have shown to positively influence education outcomes for rural children (Nye et al., 2000). When compared to students who are not at risk, some studies suggest that the teacher to student ratios found i n smaller schools may be more beneficial for youth at risk for underachievement due to poverty, previous low achievement, and racial/ethnic minority background (Konstantopoulos and Chung, 2009, Nye et al., 2004). The demographic composition of rural sch ools (e.g., percentage of students receiving free/reduced lunch, percentage of students who are racial/ethnic minorities) can also affect student achievement. In 2010, 19% of rural children lived in poverty, compared to 15% percent of suburban children (N ational Center for Education Statistics, 2013). The proportion of rural students attending schools that have more than 50% of children eligible for free/reduced lunch is higher in remote rural areas (45%) than the national average (35%) (National Center fo r Education Statistics, 2013). However, it should be noted that there is a wide range of variability throughout America's rural school districts (Provasnik et al., 2007). Additionally, a higher percentage of rural students from ethnic/racial minority back grounds (African American, 60%; Hispanic/Latino, 54%; Native American 69%) attend schools that have more than half of their students eligible for free and reduced lunch when compared to their White rural peers (21%) (Provasnik et al., 2007). Geographic is olation also plays a significant role in the academic achievement of rural students. Schools in remote locations often struggle to recruit and retain highly qualified teachers, especially in the areas of math, science, and special education (Barley

PAGE 18

18 and Bri gham, 2008; Monk, 2007). Rural youth also tend to have lower education aspirations because a college education is not required for many local job opportunities in rural economies. This effect is compounded by the fact that for most rural students, pursuing higher education comes at the cost of leaving their home communities. Literature on the subject suggests that students who attend rural schools may lower their aspirations in order to maintain their connections to their families and communities (Hektner 1 995; Howley, 2006; Irvin et al., 2011). When examining student achievement, the percentage of students identified with disabilities is also generally included, however it has been excluded from the data set used for this research. For the purposes of this study, the percentage of students identified with disabilities was removed as a variable because it is generally constant across state (10% of students identified, CDE 2016) and national lines (13% of students identified, NCES, 2015) and does not appear to change noticeably based on where students are located. Third grade Transitional Colorado Assessment Program scores were chosen as the measurement of student achievement due the strong association between early reading success and long term positive acad emic achievement (Mol & Bus, 2011). Findings from Sparks et al. indicated that first grade reading comprehension was a significant predictor of eleventh grade reading achievement (2014). According to the literature, early reading skills appear to "snowbal l" over time and lead to more growth in the areas of language ability, vocabulary, spelling, and comprehension (Sparks et al., 2014). Reading comprehension was also found to be positively correlated with math achievement, which suggests that mastery of ea rly reading skills is important for future success in math (Grimm, 2008). Mol and Bus discovered a moderate positive

PAGE 19

19 relationship between print exposure and academic achievement, which suggests that more frequent readers become more successful students (2 011). Additionally, the U.S. Department of Education has placed particular emphasis on reading achievement as "reading opens the door to learning about math, history, science, literature, geography and much more. Thus, young capable readers can succeed in these subjects, take advantage of other opportunities, and develop confidence in their own abilities" (The U.S. Department of Education, 2003, p. 28).

PAGE 20

20 CHAPTER III METHODS This study utilized a non experimental, quantitative research methodology to investigate both the relationship and the strength of the relationship between rurality and student achievement using correlation analysis and multiple linear regression. Pearson' s r correlation coefficient was used to measure the strength of the relationship between rurality and student achievement. Additionally, Pearson's r determines a positive or negative correlation between the variables. According to Chen and Popovich (2002), the Pearson's r correlation coefficient is a critically important parametric measure of association. Additionally, Multiple linear regression analysis was used to illuminate the predictive strength of the independent variables on the dependent variable. R egression analysis provided a quantitative explanation as to the effect of the independent variables on the dependent variable. This methodology was chosen as it is commonly used in educational research and there are a wealth of studies that provide models for the design of this research (Dunagin, 2006; Pan et al., 2003; Siegrist, Weeks, Pate, & Monetti, 2009). Independent variables include rurality of a school district, percentage of English language learners, percentage of students qualified for free/red uced lunch, and the percentage of racial/ethnic minorities in a student body. Rurality of a school district was defined using the guidelines set out by the Colorado Department of Education. The dependent variable is student achievement as measured by the t hird grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). The statistical analyses used to examine the interplay of these variables provided the evidence required to answer the research questions posed by t his study.

PAGE 21

21 Research Questions and Hypotheses Research Question 1: Can the set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body be used to predict student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP)? Alternative Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). Null Hypothesis: The set of variables including rurality of a school district, t otal student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are not significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). Research Question 2: Does a positive linear relationship exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP)?

PAGE 22

22 Alternative Hypothesis: A positive linear relationship does exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP). Null Hypothesis: A positive linear relationship does not exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP). Targe t Population This study targeted third graders from Colorado's rural public school districts. According to October Count 2014, Colorado is home to 178 school districts with a total student population of 889,006 (Colorado Department of Education, 2014). Di stricts ranged in size from five students (Agate 300 School District) to 81,131 students (Denver Public Schools) (Colorado Department of Education, 2014). This study's target population focused on rural school districts in particular. The Colorado Departme nt of Education (CDE) determines rural school districts based on the geographical size of the district, the distance from the nearest urban area, and enrollment data (approximately 6,500 students or fewer). By this definition, 148 school districts in Color ado qualify as rural school districts. Colorado's rural school districts were home to 30,139 students in the 2013 2014 school year (Colorado Department of Education, 2014). According to 2014 October Count data, approximately 37% of Colorado's total stude nt population attended a rural school district in the 2013 2014 school year (Colorado Department of Education, 2014) Sampling Method All 178 school districts were available for inclusion in this study. However, due to privacy concerns, the Colorado Department of Education does not report scores on

PAGE 23

23 standardized assessments if a cohort of 16 or fewer students took the state exam. Because of this, third grade achievement on the 2014 Transitional Colorado Assessment Program was not reported by CDE for 42 school districts. Each of these 42 school districts qualified as small rural districts, meaning they enroll less than 1,000 students and ar e separated by considerable distance from the nearest urban/urbanized area (Colorado Department of Education, 2017). After excluding the districts for which CDE does not report student achievement data, 133 school districts were analyzed. With that being s aid, this study examined the majority of the available school district data and can provide a generalization for all Colorado public school districts. However, it should be noted that a large amount of geographic, racial/ethnic, and socioeconomic diversity exists between school districts that meet the qualification of rural. The Colorado Department of Education supplies student achievement data aggregated via school districts on a yearly basis. Every school district in the state is required to provide CDE w ith the information contained in this study. No permission was required to acquire or analyze this data, as it is available to the public. No individual student records were used for this research. Data Collection and Analyses Transitional Colorado Asses sment Program third grade reading scores for the Spring 2014 administration were accessed via CDE at https://www.cde.state.co.us/assessment/cmas dataandresults. The data was then downloaded in the form of an Excel spreadsheet. Historical supplementary data regarding percentage of students receiving free and reduced lunch, percentage of racial and ethnic minorities, and percentage of English Language Learners was accessed using CDE's District Dashboard, which can be found at

PAGE 24

24 http://www2.cde.state.co.us/schoo lview/dish/dashboard.asp. In order to answer the two research questions, correlation and multiple linear regression were conducted. All data used was publicly accessible. These statistical analyses illuminated the relationship between rurality and student achievement, as well as the predictive intensity of the independent variables. SPSS 24 Regression was used to conduct the data analysis.

PAGE 25

25 CHAPTER IV RESULTS This study examined the student achievement of 133 Colorado school districts using the Spring 2014 Transitional Colorado Assessment Program (TCAP) administration. The predictor variables of rurality, total student enrollment, percentage of English language learners, percentage of racial/ethnic minorities, and percentage of students that qualify for free and reduced lunch were examined. Analysis was conducted on these variables to determine if they exert influence on overall student achievement. The range of district populations varied from a total student population of 173 (Moffat County RE 1) to 82 ,942 (Jefferson County School District). The mean student enrollment for the 2013 2014 school year was 6,139 students, with a median of 1,002 students. The dependent variable, student achievement, was measured using the percentage of third grade students i n the district who earned a score of proficient or advanced on the 2014 TCAP examination. The percentage of students who scored proficient or advanced ranged from 25.9% (Lake County R 1) to 100.0% (Swink 33). The mean for the selected districts was 72.93% with a median of 74.68% and a mode of 66.67%. Districts were coded based on CDE's definition of rural school districts. Depending on student population size and geographic location, districts were assigned a "1" for Urban/Metro, a "2" for Rural, and a "3" for Small Rural. Twenty nine school districts, or 21.8% of the school districts examined were coded as Urban/Metro. Forty three districts from the selected data set were coded as Rural, totaling 32.3% of the overall sample. Lastly, 61 school districts wer e coded as Small Rural, which accounts for 45.9% of the districts analyzed.

PAGE 26

26 The overall percentage of English language learners ranged from 0.0% (Big Sandy 100J, North Conejos RE 1J, Cheraw 31, Dolores County RE No. 2) to 57.9% (Center 26 JT) with a mean of 12.74% ELL, a median of 7.7% ELL, and a mode of 0.0% ELL. The percentage of students eligible to receive Free and/or Reduced Lunch ranged from 5.6% (Aspen 1) to 91.7% (Center 26 JT) with a mean of 48.4%, a median of 48.2%, and a mode of 53.5%. The perc entage of students identified as racial or ethnic minorities in the 133 districts selected ranged from 1.6% (Mc Clave RE 2) to 94.7 (Center 26 JT). Table 1 shows the descriptive data for the aforementioned variables. TABLE 1 : Rurality, Achievement, and Predictor Variable Descriptive Statistics N Range Min Max Mean Std. Dv Var Skew Std. Err Rurality 133 2 1 3 2.24 0.79 0.62 0.45 0.21 TCAP 2014 Percent Proficient & Advanced 133 74.03 25.97 100.00 72.93 11.97 143.18 0.85 0.21 Percent Free & Reduced Lunch 133 86.1 5.60 91.70 48.47 18.65 347.73 0.12 0.21 Percent Minority 133 93.1 1.6 94.7 36.55 22.01 484.71 0.74 0.21 Percent English Language Learners 133 57.9 0.0 57.9 12.74 13.65 186.22 1.48 0.21 Total Student Enrollment 133 82,769 173 82,942 6,189 13,852 191896 3.72 0.21

PAGE 27

27 Correlation and Multiple Regression Analyses: Analysis of Research Question 1 RQ 1: Can the set of variables including rurality of a school district, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial/ethnic minorities in a student body be used to predict student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP)? Alternative Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are significant predictors of student achievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). Null Hypothesis: The set of variables including rurality of a school district, total student enrollment, percentage of English language learners, percentage of students qualified for free/reduced lunch, and the percentage of racial ethnic minorities in a student body are not significant predictors of student ach ievement as measured by the third grade standardized reading scores for the Spring 2014 Transitional Colorado Assessment Program (TCAP). The regression equation used to analyze this research question was: 2014 TCAP third grade Reading Score (y1) = constan t + rurality (coefficient) + total student enrollment (coefficient) + % of English language learner (coefficient) + % of students identified as racial/ethnic minorities (coefficient) + % of students qualified for Free and Reduced Lunch (coefficient). SPSS 24 was used to perform the data analysis.

PAGE 28

28 For research question one, the null hypothesis was rejected and the alternative hypothesis was supported. The aforementioned independent variables are significant predictors of third grade reading scores as measur ed by the 2014 Transitional Colorado Assessment Program (TCAP). Assumptions were tested by visual examination of the normal probability plots of residuals and scatter plots of residuals versus predicted residuals. There were no identified violations of nor mality, linearity, and/or homoscedasticity (see Figure 1). The Durbin Watson statistic of 1.878 revealed that the residuals were truly independent of one another.

PAGE 29

29 Normal Probability P Plot of Regression Standardized Residual Dependent Variable: TCAP 2014 Percent Proficient and Advanced Expected Cum Prob Observed Cum Prob Figure 1: Normal Probability Plot

PAGE 30

30 Regression analysis indicated that the model significantly predicted 2014 third grade TCAP reading scores, F (10.448) = .000, p < .001 (see Table 2). As shown in Table 3, R! for the model was .291, with an adjusted R! of .264, which indicates that approxim ately 26% of the variance found in 2014 third grade TCAP reading scores can be predicted using the independent variables of rurality, total student enrollment, the percentage of students identified as racial/ethnic minorities, the percentage of students id entified as English language learners, and the percentage of students qualified for free and reduced lunch. While the model is significant, 73% of the variance found in student achievement is not accounted for by this linear multiple regression model. TABLE 2: Regression Analysis Model Table Model Summary Model R R Square Adjusted R Square Std. Error of Estimate Durbin Watson 1 0.540 0.291 0.264 10.26866 1.878 a. ! Predictors: (Constant), Total Student Enrollment, Percent Free and Reduced Lunch, Percent English Language Learners, Rurality, Percent Minority b. ! Dependent Variable: TCAP 2014 Percent Proficient and Advanced c. ! ANOVA Model Sum of Squares df Mean Square F Sig Regression 5508.61 5 1101.72 10.448 0.000 Residual 13391.56 127 105.45 Total 18900.18 132

PAGE 31

31 With regards to the individual relationships between the independent variables and third grade TCAP reading scores, three variables significantly predicted reading scores. Rurality (t = 2.263, p > .05), the percentage of English Language Learners (t = 2.7 64, p < .01), and the percentage of students eligible for free and reduced lunch (t = 3.358, p < .05) were all significant predictors of student achievement. Analysis of Research Question 2 Research Question 2: Does a positive linear relationship exist between rurality and student achievement as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP)? Alternative Hypothesis: A positive linear relationship does exist between rurality and student achievemen t as measured by third grade reading scores on the Spring 2014 Transitional Colorado Assessment Program (TCAP). Null Hypothesis: A positive linear relationship does not exist between rurality and student achievement as measured by third grade reading score s on the Spring 2014 Transitional Colorado Assessment Program (TCAP). Research question 2 examined the impact of one independent variable, rurality, on its dependent variable, 2014 third grade TCAP reading scores. In this hypothesis, the correlation was w eakly positive (r = .169, p = .051) but failed to meet the p = <.05 threshold, as shown in Table 6. Therefore, the correlation between a school district's rurality and third grade TCAP reading scores could not be made for the selected Colorado school distr icts in the 2013 2014 school year. For research question 2, the author fails to reject the null hypothesis.

PAGE 32

32 Table 3: Correlation Coefficients Correlation Coefficients Rurality TCAP 2014 Percent Proficient & Advanced Percent Free and Reduced Lunch Percent Minority Percent English Language Learners Total Student Enrollm ent Rurality Pearson Correlation 1 0.169 0.151 .289** .252** .603** TCAP 2014 Percent Proficient & Advanced Pearson Correlation 0.169 1 .431** .422** .438** 0.043 Percent Free and Reduced Lunch Pearson Correlation 0.151 .431** 1 .678** .442** 0.126 Percent Minority Pearson Correlation .289** .422** .678** 1 .679** 0.153 Percent English Language Learners Pearson Correlation .252** .438** .442** .679** 1 .189* Total Student Enrollment Pearson Correlation .603** 0.043 0.126 0.153 .189* 1 ** indicates correlation is significant at the 0.01 Level (2 tailed) * indicates correlation is significant at the 0.05 level (2 tailed)

PAGE 33

33 CHAPTER V DISCUSSION Research Q uestion 1 focused on socioeconomic and linguistic influences that can predict outcomes on standardized assessments. Higher 2014 TCAP reading scores were associated with urban/suburban school districts, a smaller overall percentage of s tudents receiving free and reduced lunch, and a smaller overall percentage of English language learners. While the overall model is significant, the variables of total student enrollment and percentage of racial and ethnic minorities were not significant p redictors of student achievement. Overall, 26.4% of the variance in 2014 TCAP reading scores could be predicted using the variables of rurality, percentage of English language learners, and the percentage of students that qualify for free and reduced lunch . Seventy three percent of the variance found in student achievement as measured by third grade TCAP reading scores is not accounted for by this linear multiple regression model. This study did not uncover other independent variables that exert the majorit y of influence over third grade reading scores. Additionally, the variables of rurality, the percentage of English language learners, and the percentage of students qualified for free and reduced lunch can only be considered as weak predictors (r squared = .264) of the variance in TCAP reading scores. Research Question 2 asked if the degree of a school district's rurality could be positively correlated with TCAP reading scores for third grade students in the 2013 2014 school year. For this hypothesis, th e correlation could not be made as the analysis indicated it narrowly missed the p< .05 threshold, at p = .051. This could be further examined using a bigger data set, for example, using TCAP reading scores from third through tenth graders. This research a ddresses a gap in the literature by indicating that

PAGE 34

34 the degree of rurality for Colorado's public school districts does not appear to be correlated with academic achievement. While not directly related to Research Question 2, the correlation performed stil l provided interesting data to examine. The degree of rurality was associated with lower total student enrollment (r = .603, p < .01), fewer English language learners (r = .252, p < .01), and fewer students identified as racial or ethnic minorities (r = .289, p < .01). The variable of percent free and reduced lunch (r = .431, p < .01), percent racial and ethnic minorities (r = .422, p < .01). , and percent English language learners ( r = .438, p < .01) were moderately correlated with lower student achievement, as measured by third grade reading scores. According to the correlation analysis, increased percentages of students identified as racial or ethnic minorities was strongly correlated with increased percentages of students receiving free and red uced lunch ( r = .678, p < .01) and higher numbers of English language learners (r = .679, p < .01). This tracks with demographic data from the state of Colorado, as Hispanic/Latinx students are the biggest minority group in the state. Additionally, the ma jority of English learners in Colorado identify as Hispanic/Latinx and share Spanish as a home language (Office of Culturally and Linguistically Diverse Education). Limitations The scope of this research was limited by time, duration, and definition. The questions posed were fairly narrow and only concerned the specific achievement of students (third grade reading TCAP scores), through the use of specific methodology (quantitativ e; correlation and multiple linear regression), for only one school year ( 2013 2014), and for publicly funded school districts in the state of Colorado.

PAGE 35

35 The multiple linear regression performed indicated a small influence of the chosen socioeconomic p redictor variables on the dependent variable of student achievement. Examining other independent variables via correlation and multiple linear regression could lead to a better understanding of what ultimately impacts student achievement. This study in par ticular did not define what other factors may influence the predictability of academic achievement. While the quantitative methodology utilized in this study was able to identify relationships between variables, it does not explain why these relationships exist. For example, simple correlational analyses identified that rural school districts have relatively fewer amounts of students who are members of cultural and linguistic minority groups, but it does not illuminate the causation behind these relationshi ps. In the future, researchers seeking more definitive answers to these questions should explore other analyses. Utilizing publicly aggregated data, while convenient and expedient, was not without its faults. Due to student privacy concerns, CDE masks ach ievement data when they have less than 16 results per test, per district. When considering small districts, this amounts to a significant portion of the available achievement data and is the reason why 42 small rural school districts had to be excluded fro m this study's data set. In 2015, CDE also began using new, more stringent suppression rules for achievement data. Starting in 2015, if fewer than four students either did or did not meet the expectations of the test, the achievement data is not provided. According to CDE, this accounts for a missing 6% of district level student achievement data and a missing 9% of school level student achievement data. In 2016, 36% of school and district student achievement data was masked. This is especially concerning wh en you take into account that the achievement data of specific groups of students is not being released due to small cohort

PAGE 36

36 sizes. While it is critically important to examine the student achievement of low income students, English language learners, racial /ethnic minorities, and students identified with a disability, gathering this information is often at odds with students right to privacy. In addition, using the percentage of students identified as racial/ethnic minorities and English language learners c ould be construed as an oversimplification of these two categories. Percentage of English language learners can be further broken down into percentage of Non English Proficient (NEP), Limited English Proficient (LEP), and Former English language learners ( FELL) students, which would more thoroughly examine the relationship between the levels of language acquisition and development with its impact on student achievement. Relatedly, using the percentage of racial/ethnic minorities in a district student popula tion can be problematic for similar reasons. First, it assumes that the majority of students in a school district are identified as White/Non Hispanic, which is not the case for 35 of the schools districts examined in this data set. Identifying all non whi te students as simply "minority students" instead of using the more specific identifiers of Black, Hispanic/Latinx, American Indian or Alaskan Native, Asian, or Two or More Races also deemphasizes the differences in student achievement found within these r acial/ethnic groups. Further research on the topic would benefit from a more specific breakdown of both racial and ethnic groups as well as differentiation between the levels of English language acquisition/development. This current study was limited in s cope as it only examined one measure of student achievement by using the 2014 TCAP reading assessment scores. Student achievement could also be defined using high school graduation rates, high school dropout rates, college readiness, ACT/SAT exam scores, o r state standardized math,

PAGE 37

37 science, or social studies scores. In addition, this study provided only a snapshot for student achievement by looking at one year's results. Student achievement could also be measured using a longitudinal growth model that focus es on whether or not students, schools, and districts are making Adequate Yearly Progress (AYP). Lastly, one of the central variables of this study rurality is an incredibly fraught concept to operationalize. CDE's identification of rural/small rura l was used, but there are also more detailed models of rurality that could have been utilized. For example, CDE also organizes their schools districts by settings that may prove to be more descriptive for further research. These settings include the classi fications of Denver Metro, Urban Suburban, Outlying City, Outlying Town, Remote, Colorado Board of Collaborative Educational Services (BOCES), and Detention Centers. In the future, it could be illuminating to examine student achievement based on these more specific geographic and academic parameters. Conclusion This study explored socioeconomic and linguistic influences in relationship to student achievement outcomes. The purpose of this research was to determine if a correlation between the degree of rur ality in Colorado's public school districts and third grade reading achievement as measured by the Transitional Colorado Assessment Program (TCAP) could be made. A statistically significant correlation was not identified for the 2013 2014 school year. Th is suggests that there is not a significant relationship between the degree of rurality in Colorado's public school districts and overall student achievement. The predictor relationship between five independent variables (degree of rurality, total student enrollment, percentage of racial and ethnic minorities, percentage of English language learners, percentage of students qualified for free and reduced

PAGE 38

38 lunch) and student achievement was verified. The regression analysis indicated a weak level of predictabi lity for the independent variables, and accounted for roughly 26% of the overall variance in student achievement. Lastly, considering the context of high stakes testing and educational reform, this study calls for additional research to better understand t he predictor variables and their relationship with Colorado's student achievement.

PAGE 39

39 REFERENCES Barley, Z. A., & Beesley , A. D. (2007). Rural school success: What can we learn? Journal of Research in Rural Education, 22(1). Retrieved from http://jrre.psu.edu/articles/22 1.pdf Barley, Z. A., & Brigham, N. (2008). Preparing teachers to teach in rural schools (Issues & Answers Report, REL 2008 Ð No. 045). Washington, DC: U.S. Department of Education. Retrieved from http://ies.ed.gov/ncee/edlabs/regions/central/pdf/REL_2008045_ sum.pdf Bill, C olorado House Bill 93 1313. Colorado Revised Statutes, 22 53. Burney, V. H., & Cross, T. L. (2006). Impoverished students with academic promise in rural settings: 10 lessons from project aspire. Gifted Child Today, 29(2), 14 Ð 21. doi:10.4219/gct 2006 200 Colorado Department of Education. TCAP Mathematics, English Language Ar ts, Science and Social Studies Data and Results. (2014). Retrieved December 4, 2017, from https://www.cde.state.co.us/assessment/cmas dataandresults Colorado Department of Educa tion. Pupil Membership. (2016). Retrieved December 4, 2017, from https://www.cde.state.co.us/cdereval/pupilcurrent

PAGE 40

40 Rural and Small Rural Designation List[XLSX]. (2017, January). Colorado Department of Education. Colorado Senate. (2010). Senate Bill 10 191. Retrieved from http://www.leg.state.co.us/clics/clics2010a/csl.nsf/fsbillcont3/EFs3BB67D4734 2CF 8972576A800027B078?open&file=191_enr.pdf. Crockett, L. J., Shanahan , M. J., & Jackson Newsom, J. (2000). Rural youth: Ecological and life course perspectives. In R. Monte mayor, G. R. Adams, & T. P. Gullotta (Eds.), Adolescent diversity in ethnic, economic, and cultural contexts: Advances in adolescent development (Vol. 10, pp. 43 Ð 74). Thousand Oaks, CA: Sage Publications. Crosnoe, R., Mistry, R., & Elder, G. H., Jr. (2002). Economic disadvantage, family dynamics, and adolescent enrollment in higher education. Journal of Marriage and Family, 64, 690 Ð 702. doi:10. 1111/j.1741 3737.2002.00690.x Dunagin, P. (2006). An analysis of specific factors that relate to the dropout rate of Hispanic students in Texas public school (Doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Database. (3226795). Duncan, C. M. (2001). Social capital in America's poor rural communities. In S. Saegert, J. P. Thompson, & M. R. Warren (Eds.), Social capital and poor communities (pp. 60 Ð 86). New York, NY: Russell Sage Foundation.

PAGE 41

41 Every Student Succee ds Act (ESSA). (2015). Retrieved December 4, 2017, from https://www.ed.gov/essa Farmer, T. W., Leung, M. C., Banks, J., Schaefer, V., Andrews, B., & Murray, R. A. (2006). Adequate yearly progress in Small Rural Schools and Rural Low Income Schools. Rural Educator, 27(3), 1 Ð 7. Garcia, N. (2017, June 15). Colorado forges its own path with standardized tests, will stop giving PARCC English and math tests. Retrieved December 10, 2017, from https://www.denverpost.com/2017/06/15/colorado standardized english math tests/ Graham, S. E., & Teague, C. (2011). Reading levels of rural and urban third graders lag behind the ir suburban peers. Grimm, K. J. (2008). Longitudinal associations between reading and mathematics achievement. Developmental Neuropsychology, 33(3), 410 426. HardrŽ, P. L., & Hennessey, M. N. (2010). Two rural worlds: Differences of rural high school students' motivational profiles in Indiana and Colorado. Journal of Research in Rural Education, 25(8). Retrieved from http://www.jrre.psu.edu/articles/ 25 8.pdf

PAGE 42

42 HardrŽ, P. L., Sullivan, D. W., & Crowson , H. M. (2009). Student characteristics and motivation in rural high schools. Journal of Research in Rural Education, 24(16). Retrieved from http:// www.jrre.psu.edu/articles/24 16.pdf Hektner, J. M. (1995). When moving up implies moving out: Rural adolescent conflict in the transition to adulthood. Journal of Research in Rural Education, 11(1), 13 Ð 14. Howley, C. W. (2006). Remote possibilities: Rural children's educational aspirations. Peabody Journal of Education, 82, 62 Ð 80. Irvin , M. J., Meece, J. L., Byun, S. Y., Farmer, T. W., & Hutchins, B. C. (2011). Relationship of school context to rural youth's educational achievement and aspirations. Journal of youth and adolescence, 40(9), 1225. Konstantopoulos, S., & Chung, V. (2 009). What are the long term effects of small classes on the achievement gap? Evidence from the lasting benefits study. American Journal of Education, 116, 125 Ð 154. doi:10.1086/605103 Lichter, D. T., McLaughlin, D. K., & Cornwell, G. T. (1995). Mi gration and the loss of human resources in rural America. In L. J. Beaulieu & D. Mulkey (Eds.), Investing in people: The human capital needs of rural America (pp. 235 Ð 256). Boulder, CO: Westview Press.

PAGE 43

43 Lichter, D. T., & Johnson, K. M. (2007). The changin g spatial concentration of America's rural poor population. Rural Sociology, 72, 331 Ð 358. doi:10.1526/003601107781799290 Lichter, D. T. (2012). Immigration and the new racial diversity in rural America. Rural Sociology , 77 (1), 3 35. Logan, J. R., & Bu rdick Will, J. (2017). School segregation and disparities in urban, suburban, and rural areas. The ANNALS of the American Academy of Political and Social Science, 674(1), 199 216. Lyson, T. A. (2002). What does a school mean to a community? Assessing the social and economic benefits of schools to rural villages in New York. Journal of Research in Rural Education, 17(3), 131 Ð 137. Retrieved from http://www.jrre.psu.edu/articles/ v17,n3,p131 137,Lyson.pdf McLoyd, V. C. (1998). Socioeconom ic disadvantage and child development. American Psychologist, 53, 185 Ð 204. doi:10.1037/ 0003 066X.53.2.185 McMurrer, J., & Yoshioka, N. (2013). States' Perspectives on Waivers: Relief from NCLB, Concern about Long Term Solutions. Center on Educati on Policy.

PAGE 44

44 Miller, P., Votruba Drzal, E., & Setodji, C. M. (2013). Family income and early achievement across the urban Ð rural continuum. Developmental psychology, 49(8), 1452. ! Mol, S. E., & Bus, A. G. (2011). To read or not to read: a meta analysis of print exposure from infancy to early adulthood. Psychological bulletin, 137(2), 267. ! Monk, D. H. (2007). Recruiting and retaining high quality teachers in rural areas. The Future of Children, 17, 155 Ð 174. National Center for Education Statistics (2013). The Status of Rural Education. Retrieved January 7, 2018, from http://nces.ed.gov/pr ograms/coe/indicator_tla.asp National Center for Education Statistics (2017). The Condition of Education : Children and Youth with Disabilities Retrieved December 7, 2017, from https://nces.ed.gov/programs/coe/indicator_cgg.asp Nye, B., Hedges, L. V., & Konstantopoulos, S. (2000a). Do the disadvantaged benefit from the small classes? Evidence from the Tennessee class size experiment. American Journal of Education, 109, 1 Ð 26. Nye, B., Hedges, L. V., & Konstantopoulos, S. (2004). Do m inorities experience larger lasting benefits from small classes? Journal of Educational Research, 98, 94 Ð 100.

PAGE 45

45 Office of Culturally and Linguistically Diverse Education. (n.d.). Retrieved December 20, 2017, from https://www.cde.state.co.us/cde_english Pan, D., Rudo, Z., Schneider, C., & Smith Hanson, L. (2003). Examination of resource allocation in education: Connecting spending to student performance. Southwest Educational Development Laboratory, Austin, TX. Powell, D., Higgins, H. J., Aram, R., & Freed, A. (2009). Impact of No Child Left Behind on Curriculum and Instruction in Rural Schools. Rural Educator, 31(1), 19 28. Provasnik, S., KewalRamani, A., Coleman, M. M., Gilbertson, L., Herring, W., & Xie, Q. (2007). Status of ed ucation in rural America (NCES 2007 Ð 040). Washington, DC: National Center for Education Statistics. Reeves, C. (2003). Implementing the No Child Left Behind act: Implications for rural schools and districts. Retrieved June, 14, 2011. Roscigno , V. J., Tomaskovic Devey, D., & Crowley, M. (2006). Education and the inequalities of place. Social Forces, 84(4), 2121 2145. Siegrist, G., Weeks, W., Pate, J., & Monetti, D. (2009). Principals' experience, educational level, and leadership practices as predictors of Georgia high school

PAGE 46

46 graduation test results. Journal of Philosophy and History of Education (59), 174 179. Sparks, R. L., Patton, J., & Murdoch, A. (2014). Early reading success and its relationship to reading achievement an d reading volume: replication of Ô10 years later'. Reading and Writing, 27(1), 189 211. U.S. Department of Agriculture. (2004). Rural poverty at a glance (Report No. 100). Retrieved from http://www.ers.usda.gov/ publications/rdrr100/rdrr100.pdf U. S. Department of Education, Office of the Under Secretary (2003). No Child Left Behind: A toolkit for teachers, Washington, DC. U.S. Congress. (2001). No Child Left Behind Act. Retrieved from: http://www2.ed.gov/nclb/landing.jhtml.