EMBARGOED UNTIL 2/1/2017, 10 am THE OUTLIERS THE STATE OF COLORADO SCHOOL DISTRICTS I 2016 endorsed by THE OUTLIERS, 2016 Acknowledgments A+ Colorado would like to thank the Colorado Department of Education and the Colorado Department of Higher Education for making accessible information about outcomes for students during and after their time in Colorado’s K-12 public education system. Additionally, we would not be able to produce this report without the support of our funders including The Anschutz Foundation, Carson Foundation, Daniels Fund, The Denver Foundation, The Donnell-Kay Foundation, The Gates Family Foundation, The Piton Foundation, Rose Community Foundation, and Walton Family Foundation. Most importantly we would like to thank the educators across our state who work daily to instill in our students a sense of curiosity and a love of learning. Endorsements Colorado Succeeds is a nonprofit, nonpartisan coalition of business leaders who have joined forces to make sure the education system works better and smarter for all the people of Colorado. 2 THE OUTLIERS, 2016 Table of Contents I. Introduction I 5 II. Demographics I 6 III. K-12 Achievement I 8 a. Who is Reading, Writing, and Doing Math at Grade Level? I 9 b. Who is Bucking the Trend? I 11 c. Who is Making Headway Year over Year? I 14 IV. College/Career Readiness I 18 a. Who is Opening Doors for Graduates? I 22 b. What Happens Beyond K-12? I 27 V. Conclusion I 28 VI. Appendices I 29 3 THE OUTLIERS, 2016 List of Figures 1. Districts with the Largest Increases in Student Enrollment (2011-2015) I 6 2. Districts with the Largest Changes in Proportion of Students Qualifying for Free or Reduced Price Lunch (percentage points) 2011-2015 I 6 3. Racial and Ethnic Makeup of Colorado Students I 7 4. Districts with the Largest Changes in Proportions of Racial and Ethnic Student Groups (percentage points) 2011-2015 I 7 5. Districts with the Largest Changes in Proportion of Emerging Multilingual Students Learning English 2011-2015 I 7 6. Districts with Big Changes in Relative Performance in Elementary English Language Arts 2013-2016 I 10 7. Districts with Big Changes in Relative Performance in Elementary Math 2013-2016 I 10 8. Districts with Big Changes in Relative Performance in Middle School English Language Arts 2013-2016 I 10 9. The Outliers: Elementary School English Language Arts Percent of Students who met Grade Level Standards Compared to Districts with Similar Demographics (2016) I 12 10. The Outliers: Elementary School Math Percent of Students who Met Grade Level Standards Compared to Districts with Similar Demographics (2016) I 12 11. The Outliers: Middle School English Language Arts Percent of Students who met Grade Level Standards Compared to Districts with Similar Demographics (2016) I 13 12. Growth by Race or Ethnicity: CMAS PARCC English Language Arts (2016) I 15 13. Growth by Race or Ethnicity: CMAS PARCC Math (2016) I 16 14. Growth by Free or Reduced Price Lunch Eligibility: CMAS PARCC English Language Arts (2016) I 17 15. Growth by Free or Reduced Price Lunch Eligibility: CMAS PARCC Math (2016) I 17 16. Districts with Improved Graduation Rates (2011-2015) I 18 17. Highest and Lowest Graduation Rates by Race or Ethnicity (2015) I 20 18. Highest and Lowest Graduation Rates for Emerging Multilingual Students (2015) I 21 19. Highest and Lowest Graduation Rates for Students Eligible for Free or Reduced Price Lunch (2015) I 21 20. Highest and Lowest Graduation Rates for Students with Disabilities (2015) I 21 21. ACT Scores by Race or Ethnicity I 24 22. ACT Scores by Free or Reduced Price Lunch Eligibility I 25 23. Schools with the Top Ten ACT Scores by Student Group I 26 24. Highest Matriculation Rates (Class of 2014) I 27 25. Lowest Matriculation Rates (Class of 2014) I 27 26. Highest Remediation Rates (Class of 2014) I 27 27. Lowest Remediation Rates (Class of 2014) I 27 4 THE OUTLIERS, 2016 Introduction Welcome to A+ Colorado’s first annual report exploring how districts across the Centennial State are serving diverse student populations. We envision a state where every student has access to a high quality education: an education that provides students with the skills, knowledge, and opportunities needed for success in the 21st century. The intent of The Outliers is to do just that— identify school districts that buck trends across the state. This provides a snapshot of where outcomes for students are different from the norm, where the promise of an excellent education shines bright, and where that light is still too dim. Above all this report identifies districts that are demonstrating success and merit further investigation. Why the focus on school districts? In Colorado, a state that embraces local control, school districts hold a particularly important position in providing a high quality education. School districts are where the rubber hits the road. For example, districts are responsible for choosing school leaders, they determine staffing models, and, they can dictate school operations, with decisions from how to distribute funds to curriculum selection. Importantly, districts are responsible for ensuring their schools meet student performance expectations. This report looks specifically at student performance outcomes at the district level. In compiling this report, we gathered publicly available information on all school districts across the state, and looked at trends over the past five years. Unless otherwise noted, data was gathered from the Colorado Department of Education. Some data, like disaggregated ACT scores, required a Colorado Open Records request. Other data, like disaggregated CMAS PARCC achievement data was not publicly available; therefore we are unable to provide insight such as which districts are closing the opportunity gap between students of different races and ethnicities, or different income levels for example. We focus our analysis on districts serving more than 1,000 students across pre-kindergarten to twelfth grade. This represents 76 of 186 school districts and BOCES,1 and 96 percent of all Colorado students. We made this decision not because any sized district matters more or less than others, but we recognize that smaller school systems face unique challenges that our research does not address. You can find the longitudinal data that has been used to develop this report for all school districts at apluscolorado.org. 1 Boards of Cooperative Educational Services, or BOCES, are administrative agencies that provide educational services to two or more school districts that find it either advantageous or cost-effective to share services. For more information, see coloradoboces.org 5 DEMOGRAPHICS THE OUTLIERS, 2016 Welcome to the West: How are School District Demographics Changing? Demographic shifts help us understand a piece of the diverse experiences of Colorado’s students. The state serves nearly 900,000 students, a number that has grown steadily over the past five years. The increase in the student population has primarily been driven by growth in the Denver Metro Area (which saw a 4.6 percent increase in students), urban-suburban communities outside of the Denver Metro area (which saw a 7.2 percent increase in students), and remote communities (which saw a 7.3 percent increase in students). Growth was not as strong in outlying towns and outlying cities. For a map of Colorado school districts, see Appendix A. % Increase in Enrollment Figure 1: Districts with the Largest Increases in Student Enrollment (2011-2015) 50% 43.5% 36.5% 25% 20.6% 15.8% 0% Charter School Institute Falcon 49 13.0% 12.6% Windsor Johnstown- St Vrain Milliken RE-4 Valley RE-5J RE 1J 12.3% 11.6% 11.3% Mapleton 1 Steamboat Denver Springs County 1 RE-2 Summit RE-1 Not Pictured Above: Byers 32J authorizes several multi-district online schools which accounts for a drastic increase in enrollment (514.4%) over the past five years: Colorado Digital Academy (beginning in 2014), Colorado Virtual Academy (previously authorized by Adams 12, authorized by Byers starting in 2014), Elevate Academy (starting in 2104), and Great Plains Academy (beginning in 2012). Despite the fact that Colorado has shown strong recovery from the recession of 2008, the recovery has not been evenly distributed. 7.8 percent more students qualified for free or reduced price lunch (FRL, a proxy measure for students from low-income families) in 2015 than in 2011, outpacing growth of the student population statewide. This is particularly true outside of the Denver Metro Area. During the same time period, the proportion of students qualifying for free or reduced price lunch jumped from 41 percent to 44 percent in urban-suburban communities outside Denver. In remote communities the proportion of students qualifying for free or reduced price lunch increased from 44 percent to 46 percent. Figure 2: Districts with the Largest Changes in Proportion of Students Qualifying for Free or Reduced Price Lunch (percentage points) 2011-2015 14.7 Trinidad 1 73.9% FRL 13.6 Falcon 49 34.6% FRL Mapleton 1 -7.4 Alamosa RE-11J -7.4 Charter School Institute -8.6 Johnstown-Milliken RE-5J -8.7 61.2% FRL 64.6% FRL 42.6% FRL 28.2% FRL -12 -8 -4 0 4 Change in Proportion 8 12 16 6 THE OUTLIERS, 2016 Figure 3: Racial and Ethnic Makeup of Colorado Students 900,000 Number of Students Racial and ethnic demographics are also changing: the number of Latinx students (a gender-neutral term for students who identify as either Latino or Latina) grew twice as much as the student population overall. Students who identify as two or more races represent a small, but also growing proportion of the student population. In exploring academic outcomes by racial and ethnic groups, this report focuses primarily on the four largest categories in Colorado: white students, Latinx students, black students, and multiracial students. 899,112 854,265 3.8% 0.2% 3.1% 0.2% 600,000 56.1% 54.1% 300,000 33.4% 31.9% Changes in racial and ethnic demographics have been far more rapid for specific districts, particularly in parts of the Colorado Springs metro area and the Denver metro area. 4.6% 3.1% 0.7% 4.8% 3.1% 0.8% 0 2011 2015 American Indian or Alaskan Native Black White Multiracial Asian Latino Native Hawaiian or Pacific Islander Figure 4: Districts with the Largest Changes in Proportions of Racial and Ethnic Student Groups (percentage points) 2011-2015 District Byers 32J Black Students District Latinx Students District White Students Denver County +2.3 +4 Byers 32J +40.5 Ellicott 22 +1.9 Trinidad 1 +8.9 Valley Re-1 Adams-Arapahoe 28J (Aurora) +1.1 Falcon 49 +7.2 Manitou Springs 14 +1 -0.2 Harrison 2 +6.9 State of Colorado -2 State of Colorado Englewood 1 -1 Cherry Creek 5 -1 Widefield 3 -1.7 Fountain 8 -1.7 Falcon 49 -1.8 Charter School Institute -2.1 Harrison 2 -2.7 Adams 12 Five Star Schools State of Colorado Multiracial Students Harrison 2 +4.8 Englewood 1 +2.4 +6.1 Harrison 2 -6.4 Pueblo County 70 Widefield 3 Cheyenne Mountain 12 +1.6 Platte Valley RE-7 -6.6 Fountain 8 -6.7 State of Colorado +2.4 +2.3 +2.2 +2 Garfield 16 -1.1 Adams 12 Five Star Schools Manitou Springs 14 -1.6 Trinidad 1 -7.3 Byers 32J -1.5 Archuleta County 50 JT -9.4 Ellicott 22 -1.7 Denver County 1 English Language Learners Colorado students bring a rich linguistic diversity to the classroom. Over 128,000 students speak 251 home languages other than English between them; 88% speak Spanish. While the number of emerging multilingual students who are learning English (English Language Learners or ELLs) increased over the past five years by 4.1 percent, the proportion statewide of ELL students stayed relatively consistent at 14.2 percent. The vast majority of ELLs (over 70 percent) live in the Denver Metro Area.2 ELL populations are significantly larger in a number of Colorado districts including Adams 14 (43%), Westminster 50 (41%), Adams Arapahoe 28 (Aurora) (39%), Sheridan 2 (38%), and Lake County (36%). Colorado Department of Education. “Culturally and Linguistically Diverse Learners in Colorado: State of the State” (2015) 2 +1.1 District -2 Byers 32J +0.8 -44.6 Figure 5: Districts with the Largest Changes in Proportion of Emerging Multilingual Students Learning English (percentage points) 2011-2015 Roaring Fork RE-1 33.2% ELL 5.2 Eagle County RE 50 33.8% ELL -4.1 Denver County 1 31.6% ELL -4.5 Fort Morgan RE-3 28.6% ELL -4.5 Mapleton 1 29.8% ELL -5.5 Monte Vista C-8 10.2% ELL -6 -6 -4.5 -3 -1.5 0 1.5 Change in Proportion 3 4.5 6 Not Pictured Above: Byers 32J authorizes several multi-district online schools which accounts for a drastic increase in ELL enrollment (23.9 percentage points) over the past five years: Colorado Digital Academy (beginning in 2014), Colorado Virtual Academy (previously authorized by Adams 12, authorized by Byers starting in 2014), Elevate Academy (starting in 2104), and Great Plains Academy (beginning in 2012). 7 K-12 ACHIEVEMENT THE OUTLIERS, 2016 Climbing to New Heights? Student Achievement in Colorado Districts Colorado school districts are diverse in their student populations and performance. Despite this diversity, Colorado’s constitution requires the “establishment and maintenance of a thorough and uniform system of free public schools.” The following section looks at student achievement outcomes and begs the question to what extent are our public schools providing a thorough and uniform education to every student? Are school districts improving outcomes for kids? Which systems are the most equitable, providing students, regardless of their background, an equal likelihood as their peers to be college and career ready? a state have determined to be important standards. With the adoption of the Colorado Academic Standards In 2009, Colorado set new expectations for what students should learn at each grade level across ten subjects. While Colorado has a twenty year history of measuring student mastery of content standards with standardized assessments, 2016 was only the second time Colorado students took CMAS PARCC, an assessment specifically aligned to these higher standards in English Language Arts and in Math. This assessment informs our understanding of where students are or are not in fact learning what we as The data presented on the next few pages raise the following questions: This summative assessment data is but one indicator of student performance. By no means does it exhaustively answer what is going on in certain school systems. And it certainly does not answer how education happens or is experienced in individual schools. The data should catalyze all of us to look at these systems to ask what is working, not working, and why. Educators, communities, and policymakers alike should be reflecting, learning, and sharing practices. » Which districts have made big improvements over the past four years in getting more students to meet grade-level standards that will prepare them for college and career? Which districts have fallen off the mark in fulfilling this promise? » In which districts are students outperforming their peers in districts serving similar students? Where are students underperforming relative to similar districts? Flying Blind? The Case of the Missing Data The data collected through the statewide assessment system has one primary purpose: learning. It should expose what is working, direct inquiry, and share lessons across the state. The data— and the successes it can uncover— should be informing improvement strategies for schools and districts across the state, so that they can deliver the high-quality education our students deserve. Unfortunately, this mission has become increasingly more difficult to achieve. The Colorado Department of Education masks much of the data where student populations are small. In 2015, CDE also introduced new suppression rules for CMAS PARCC achievement data, whereby the department masks all information if fewer than four students either did or did not meet the expectations of the test. This has meant that an additional 6 percent of district level proficiency data and 9 percent of school level proficiency data is masked due to suppression rules other than the small cohort rule. All told, 36 percent of 2016 school and district level proficiency data is masked. Privacy has come at the expense of a clear understanding of how students, schools, and districts are doing. In addition, due to the complications of the new suppression rules, the Department has not released any information about whether or not specific groups of students met expectations on the exam. This means it is impossible to compare performance of students of different races or ethnicities, or of students from different socioeconomic backgrounds. And that is a problem. It is no secret that the education system has historically been worse for students of color and low-income students than their white and more affluent peers. Yet Colorado has reverted back to a system where we are flying blind. This report explores what publicly available information exists on student performance. It creates proxies to measure how districts are serving low-income students, emerging multilingual students, and students in special education. Yet these cannot replace the invaluable information that Colorado used to provide that helped schools, districts, and communities identify and address opportunity gaps between students, and work to provide an equitable education to every student who stepped inside a Colorado school. 8 THE OUTLIERS, 2016 Who is Reading, Writing, and Doing Math at Grade Level Are there systems for illuminating and sharing lessons learned across districts? Since 1997, Colorado has administered a summative assessment to measure what students know. Starting in 2015, Colorado shifted tests from TCAP/CSAP to CMAS PARCC. Because these are two separate tests, the results cannot be directly compared. In order to understand how schools and districts were doing from one year to the next, A+ conducted a percentile analysis that measured relative rankings of each district and school based on percent of students meeting grade level standards on each test. For more information see Appendix B. The charts on the following page show those districts with the biggest changes in relative performance between 2013 and 2016. Spotlights Some Districts With Many Low Income Students Made Big Gains A variety of districts have made large gains in the percentage of students reaching proficiency benchmarks in multiple subject areas and school levels. Of particular note is the state’s largest district: Denver, which in 2013 was consistently in the 15th to 20th percentile across subjects, has moved closer to the state average, landing in the 43rd percentile in Elementary English Language Arts (ELA), the 41st percentile in Elementary Math, and the 56th percentile in Middle School ELA. Fort Morgan, a district of over 3,000 students, saw similar improvements to Denver in academic performance, with the percent of students meeting grade level standards at or above the state average in Elementary Math and Middle School ELA. Also impressively, Platte Valley, a district in Weld County with just over 1,000 students, moved from percentile ranks in the bottom third to half of the state in 2013 to now ranking in the top quartile. 3 Colorado Revised Statute. 22-30.7-102 4 James L. Woodworth, Margaret E. Raymond, Kurt Chirbas, Maribel Gonzalez, Yohannes Negassi, Will Snow and Christine Van Donge, “Online Charter School Study 2015,” Center for Research on Education Outcomes (CREDO), Stanford University (2015): i-104. Questions in Low Performing Districts There are several districts that have struggled to help students master grade level standards. For districts with relative declines in students meeting grade level standards compared to the rest of the state, has the shift to new standards been more difficult or not fully implemented? Is there sufficient support for students who are struggling? Are teachers supported? Are school improvement strategies in place? For districts where fewer students are mastering content than in similar districts, are there systems for sharing practices across districts? What is the state’s role in ensuring that these districts are supporting students to reach proficiency? A Case Against Some Online Schools? Byers 32J authorizes four multi-district online schools: schools that provide “full-time education… primarily through online digital learning.”3 Multi-district online schools, while authorized by a single entity (a school district, or state authorizer), may serve students across the state of Colorado. Indeed, Byers authorized schools serve over 2500 students from 107 different school districts. Student performance in Byers 32J and Colorado Digital BOCES mirrors significant research on the impact of online schools.4 Not only do online schools have lower academic results than comparable brick-andmortar schools, but online schools tend to have a negative impact on their students when compared to students from similar backgrounds and with similar past academic performance. Dramatic drops in the proportion of students mastering grade level standards when compared to the rest of the state, and substantially lower performance than districts serving similar students, continues to challenge the value of these particular school options. Highlighting Participation Participation is key to identifying and sharing best practices, as well as ensuring equitable access to a high quality education. The federal Every Students Succeeds Act (ESSA) requires 95% of students participate in statewide assessments. In 2015, participation rates in Colorado dropped dramatically, particularly in higher grades. Participation rates improved in 2016 due to legislation passed reducing testing time and other adjustments. Yet pockets of low 2016 participation rates in specific grades and schools threatens our ability to gauge student achievement. This is particularly true in higher grades. Ensuring assessments are relevant (for example, college-entrance aligned), results are timely, and the information is accessible, can improve the impact and meaningfulness of Colorado’s assessment program. 9 THE OUTLIERS, 2016 <3,000 students Figure 6: Districts with Big Changes in Relative Performance in Elementary English Language Arts 2013-2016 100 Percentile Rankings District Size Key: +28 +41 50 -39 -28 +31 +36 +26 0 -58 +28 25 2013 Ranking 2016 Ranking Valley Re-1 Platte Valley Re-7 0 Weld County Re-3J Denver County 1 Buena Vista R-31 Lamar Re-2 Byers 32J* Manitou Johnstown9 Springs Milliken Re-5J 14* Figure 7: Districts with Big Changes in Relative Performance in Elementary Math 2013-2016 Percentile Rankings 100 25,000-39,999 students Declining 75 3,000-9,999 students 10,000-24,999 students Improving Improving Declining 75 -34 +35 50 +44 +27 +38 +38 -86 -49 +32 -41 -36 25 0 40,000-79,999 students 2013 Ranking 2016 Ranking 0 Weld County Re-8 Lamar Widefield Platte Re-2 3 Valley Re-7 Weld Fort Trinidad Garfield Morgan County Byers 1 Re-2 Re-3J 32J* Re-3 Johnstown- Aspen 12 Milliken 1 Re-5J Figure 8: Districts with Big Changes in Relative Performance in Middle School English Language Arts 2013-2016 >80,000 students *Districts with less than 80% participation in given grade level and content area. These results may or may not be representative of the broader student population. Percentile Rankings 100 Improving Declining +31 +34 50 +41 +33 -27 -32 +25 +37 -41 -33 +31 25 0 -25 +23 75 -29 2013 Ranking 2016 Ranking Denver County 1 Roaring Trinidad 1 Ellicott Platte Weld Valley County Fork 22 Re-7 Re-1 Re-1 Fort Estes Valley Buena Bayfield Archuleta Manitou Windsor Morgan Park Re-1 Vista 10JT-R County Springs Re-4 Re-3 R-31 50 JT R-3 14* 10 THE OUTLIERS, 2016 Who is Bucking the Trend? The state has yet to release information about different student groups’ achievement in 2016. This information is incredibly important, as many groups of students—low income students, students with disabilities, students learning English as a second language—get left out of the best educational opportunities the state has to offer. This contributes to the well-documented correlation between economic advantage and student achievement. To uncover districts that buck this trend, A+ conducted an analysis to compare district performance to other districts with similar student demographic populations, creating a District Demographic Index based on the population of students qualifying for Free or Reduced Price lunch, the proportion of students learning English as a second language, the proportion of students with disabilities, and the rate of students moving in and out of the district. The following charts show the districts that perform outside of the trend. For the explanation of the methodology and selection see Appendix C. Spotlights Some Districts with Many Low-Income Students Buck Trends A district demographic analysis draws attention to school systems that buck trends based on the students they serve. While there is a strong correlation between students’ backgrounds (including family income, race and ethnicity, home language, and students with disabilities) and academic performance, there is significant variability in outcomes for students across schools and districts. Steamboat Springs and East Grand 2, both smaller districts, serving approximately 2,500 and 1,300 students respectively, have significantly higher proportions of students meeting grade level standards than other districts with similar student demographics: the proportion of students reaching grade level standards is 15 to 20 percentage points higher than similar districts. It is not just low poverty districts that achieve outlier results for students. Harrison 2 in Colorado Springs also significantly bucks the trend. Harrison serves a large proportion of students qualifying for free or reduced price lunch, emerging multilingual students (ELLs), special education students, and has a more mobile student population; the proportion of elementary students in Harrison meeting grade level expectations in English Language Arts is 15 points higher than similar districts. Other districts with higher proportions of at-risk students that also show signs of bucking demographic trends are Fort Morgan, Ellicott 2, Weld County S/D Re-8, Denver, and Delta. 11 THE OUTLIERS, 2016 Figure 9: The Outliers: Elementary School English Language Arts Percent of Students who Met Grade Level Standards Compared to Districts with Similar Demographics (2016) Figure 10: The Outliers: Elementary School Math Percent of Students who Met Grade Level Standards Compared to Districts with Similar Demographics (2016) 12 THE OUTLIERS, 2016 Figure 11: The Outliers: Middle School English Language Arts Percent of Students who Met Grade Level Standards Compared to Districts with Similar Demographics (2016) Note: Districts included on chart fall well outside the trend line and serve at least 1,000 students. For more information about the methodology, see Appendix C. 13 THE OUTLIERS, 2016 Who is Making Headway Year Over Year? Information about proficiency rates helps guide inquiry about whether schools and districts are successfully helping students reach grade level expectations. However, students start at very different places in terms of their mastery of grade-level standards at the beginning of a school year. The Colorado Growth Model is critical to understanding whether students are making progress or falling behind their peers. Growth provides an important measure of whether schools are delivering value regardless of whether or not students enter the year on, above, or behind grade-level. Growth is calculated by comparing a student’s performance on the assessment to her academic peers—other students who had the same test score the previous year, resulting in the individual student growth percentile. A school or district is measured by the median growth percentile (MGP) of all its students in a given subject. The following charts explore where students of different racial and ethnic backgrounds, and students from different family income levels, are progressing on the state assessment relative to their peers. This information is particularly important given that the most recent school and district-level proficiency data—the number and percent of students who have mastered content previously explored in this report—is not publicly available for different groups of students. In districts with small cohorts of specific student groups, each student’s performance carries more weight than in larger groups of students. This is important to keep in mind when reviewing and asking questions of the data. The following charts explore the districts with the highest and lowest median growth percentiles for different groups of students. It is clear that some districts are greatly accelerating the learning of some students relative to their academic peers. Growth is critical to getting students on the path to mastering grade-level content. However, where students are behind, as we see in many districts and schools, growth must be higher to ensure students can catch up to grade-level expectations. As a guidepost, based on this 2016 data, roughly fifteen percent of all Colorado schools had a median growth percentile of 65 or above, or “exceeded expectations” on growth on the state performance framework; roughly fifteen percent of all Colorado schools had a median growth percentile of 35 or below, or “does not meet expectations” on growth on the state performance framework. These are important scores to keep in mind while exploring growth results for different groups of students across the state. Spotlights Gaps in Growth Across subjects white students and students ineligible for free or reduced price lunch are more likely to show growth than students of color and students eligible for free or reduced price lunch. What these growth gaps show then is that schools and districts continue to underserve students of color and low-income students. These students are simply not mastering as much content as their white and more affluent peers. Of course, there are exceptions to this trend. East Grand 2 and Platte Canyon have median growth percentiles of 60 for their free and reduced price lunch students in English Language Arts and Math respectively. And while the student populations are small, black students show the highest growth in Fort Morgan and Lewis Palmer. Multiracial students post some of the highest growth percentiles and showed higher growth than their white peers in Roaring Fork, Cheyenne Mountain, Delta, Durango, and Summit. 14 THE OUTLIERS, 2016 Figure 12: Growth by Race or Ethnicity: CMAS PARCC English Language Arts (2016) Districts with the Highest Growth (MGP) State MGP for Student Group Districts with the Lowest Growth (MGP) Black Students 71.5 Fort Morgan Re-3 Lewis-Palmer 38 Harrison 2 Academy 20 Pueblo County 70 State of Colorado Douglas County Re 1 Boulder Valley Re 2 Adams County 14 Thompson R2-J Byers 32J 64 58 55 53 48 37 35.5 32 31 16 Latinx Students 72 East Grand 2 Platte Canyon 1 Trinidad 1 Durango 9-R Steamboat Springs Re-2 State of Colorado Montezuma-Cortez Re-1 Aspen 1 Colorado Digital BOCES Byers 32J Manitou Springs 14 63 60 60 67 48 36 35.5 35 34 32.5 White Students 65 65 64 63 63 Denver County 1 East Grand 2 Steamboat Springs Re-2 Cheyenne Mountain 12 Trinidad 1 State of Colorado Garfield Re-2 Johnstown-Milliken Re-5J Bennett 29J Colorado Digital BOCES Brush Re-2(J) Manitou Springs 14 51 38 38 36 35 32 30.5 Multiracial Students 73 71.5 71 67.5 65.5 Roaring Fork Re-1 Durango 9-R Summit Re-1 Cheyenne Mountain 12 Cañon City Re-1 State of Colorado Englewood 1 Eagle County Re 50 Mesa County Valley 51 Valley Re-1 Moffat County Re-1 51 44.5 40 40 39 28 0 25 50 75 100 Median Growth Percentile 15 THE OUTLIERS, 2016 Figure 13: Growth by Race or Ethnicity: CMAS PARCC Math (2016) Districts with the Highest Growth (MGP) State MGP for Student Group Districts with Lowest Growth (MGP) Black Students 64 60.5 60.5 Lewis-Palmer 38 Littleton 6 Fort Morgan Re-3 Mesa County Valley 51 Widefield 3 State of Colorado Adams County 14 Greeley 6 Westminster Public Schools Thompson R2-J Byers 32J 57 56 46 32 29.5 38.5 37 35.5 Latinx Students 59 58 58 57 56 56 56 Fort Morgan Re-3 Platte Canyon 1 Valley Re-1 Weld County S/D Re-8 Woodland Park Re-2 Salida R-32 Widefield 3 State of Colorado Adams County 14 Manitou Springs 14 Lake County R-1 Byers 32J Colorado Digital BOCES 46 37 37 36 29 33 White Students Steamboat Springs Re-2 Littleton 6 Denver County 1 Buena Vista R-31 Cheyenne Mountain 12 Roaring Fork Re-1 State of Colorado Englewood 1 East Otero R-1 Lake County R-1 Adams County 14 Byers 32J Colorado Digital BOCES 65 63 62 62 62 70.5 53 41 41 41 38 36 31 Multiracial Students 66 64 63 62.5 60 Roaring Fork Re-1 Littleton 6 Cheyenne Mountain 12 Delta County 50(J) Johnstown-Milliken Re-5J State of Colorado Colorado Spring 11 Garfield Re-2 Montrose County Re-1J Greeley 6 Adams County 14 51 46 42 41 37.5 27 0 25 50 75 100 Median Growth Percentile 16 THE OUTLIERS, 2016 Figure 14: Growth by Free or Reduced Price Lunch Eligibility: CMAS PARCC English Language Arts (2016) Districts with the Highest Growth (MGP) State MGP for Student Group Districts with the Lowest Growth (MGP) Students Eligible for Free or Reduced Price Lunch 62 East Grand 2 Trinidad 1 Woodland Park Re-2 Cheyenne Mountain 12 Summit Re-1 State of Colorado Montezuma-Cortez Re-1 Johnstown-Milliken Re-5J Colorado Digital BOCES Byers 32J Manitou Springs 14 58 57 57 57 47 37 36 35 33 29.5 Students Ineligible for Free or Reduced Price Lunch 74 Trinidad 1 East Grand 2 Durango 9-R Steamboat Springs Re-2 Denver County 1 State of Colorado Johnstown-Milliken Re-5J Garfield Re-2 Bennett 29J Colorado Digital BOCES Manitou Springs 14 67.5 66 65 64 52 40.5 40 37 36 32 0 25 50 75 100 Median Growth Percentile Figure 15: Growth by Free or Reduced Price Lunch Eligibility: CMAS PARCC Math (2016) Students Eligible for Free or Reduced Price Lunch Districts with the Highest Growth (MGP) State MGP for Student Group Districts with the Lowest Growth (MGP) 64 Platte Canyon 1 Fort Morgan Re-3 Steamboat Springs Re-2 Salida R-32 Woodland Park Re-2 Buena Vista R-31 State of Colorado Adams County 14 Lake County R-1 Montezuma-Cortez Re-1 Byers 32J Colorado Digital BOCES 59 59 57 57 57 46 37 36 36 31.5 28 Students Ineligible for Free or Reduced Steamboat Springs Re-2 Price Lunch Littleton 6 71 66 66 Weld County S/D Re-8 Buena Vista R-31 Cheyenne Mountain 12 Roaring Fork Re-1 State of Colorado Garfield Re-2 Sheridan 2 Byers 32J Adams County 14 Colorado Digital BOCES 62 62 62 53 42 41.5 40 36 35 0 25 50 75 100 Median Growth Percentile 17 COLLEGE/CAREER READINESS THE OUTLIERS, 2016 Manifest Destiny: Who is Prepared for College, Career and Life? The goal of our elementary and secondary education system should be to ensure that all students are prepared to succeed in college, career, and life. High school diplomas matter, but research shows that postsecondary education and credentials are increasingly critical for employment and earning a living wage today. this rate or better. This report focuses on four year graduation rates as it is the goal of the vast majority of schools and districts across the state to support students to complete high-school within this time-frame. This also reflects the expectations set for students across the country through the federal legislation. Colorado school districts have made significant progress in supporting more students to obtain a high school diploma. The four-year graduation rate in Colorado improved from 73.9% in 2011 to 77.3% in 2015. Some districts have made even greater strides in getting students that cap and gown. Figure 16: Districts with Improved Four-Year Graduation Rates (2011-2015) District Colorado Digital Boces Percent point improvement 54 Sheridan 2 39 Garfield Re-2 25 Monte Vista C-8 18 Mapleton 1 13 Adams 12 Five Star Schools 12 Adams-Arapahoe 28J 11 Brush Re-2(J) Archuleta County 50 JT 10 Durango 9-R 10 Eaton Re-2 The following charts explore the districts that are excelling at ensuring students receive a high school diploma, and those districts where students need more support to reach that milestone, by groups of students. There continue to be large disparities in the background of students who make it to a graduation ceremony. The guideposts set by both the federal government and Colorado are helpful in understanding the magnitude of success, and some of the gaps we see. 11 9 Denver County 1 9 Harrison 2 8 Pueblo City 60 Charter School Institute 8 8 What are the right expectations for graduation rates, particularly given the significant progress Colorado has made? A few attempts have been made to help navigate and set expectations. The most recent federal legislation, the Every Student Succeeds Act, requires states to focus efforts on improving schools with four-year graduation rates less than 67%. Colorado uses a “best-of” 4, 5, 6, or 7- year graduation rate for its accountability system. The state of Colorado has set the bar for exceeding expectations at the 95% best-of graduation rate because the top 15 percent of schools in Colorado achieve Spotlights Improved Statewide Graduation Rates Leave Some Groups Of Students Out Graduation rates have improved across the state, enough such that the Colorado State Board of Education decided to adjust expectations in the accountability system to reflect the improvement. In Colorado, half of all schools have at least a 93.9% graduation rate. Fifteen percent of schools have at least a 99.3% graduation rate. And yet, not all groups across the state and within districts make it to graduation at these high rates. While the data looks at four year graduation rates, as opposed to best-of graduation rates like the accountability system, it is clear that many students are not receiving the support they need to graduate on time. And it is critical that these diplomas are meaningful and that they communicate what students know and can do. There is not a single district that serves more than 1,000 students that meets the “exceeds expectations” graduation benchmark for Black students, or students qualifying for free or reduced price lunch. 18 THE OUTLIERS, 2016 Notable Graduation Trends for Black Students A couple districts were close to reaching the “exceeds expectations” graduation rate benchmark for black students: in 2015, St. Vrain Valley Re 1J and Fountain 8 had graduation rates for black students at 93% and 90% respectively. Cherry Creek 5, the district with the largest graduation base of black students other than Denver County 1, tied for the 5th highest graduation rate (84%) for black students amongst school districts in Colorado. St. Vrain Valley, School District 27J Boulder Valley, and Jefferson County R-1 were amongst the top five graduation rates for black students in 2015, having improved upon their 2011 graduation rate for those students by 16, 21, 24, and 11 percentage points respectively. High school diplomas should be meaningful and connected to a student’s achievement. A Few Outliers Serving Colorado’s Growing Latinx Population Three districts cleared the exceeds expectations benchmark for Latinx students. Statewide only 68% of Latinx students graduated on time in 2015. Yet in Eaton (24% Latinx students), Archuleta (27% Latinx students), and Platte Valley (34% Latinx students) that narrative was turned on its head with graduation rates of 97%, 96% and 96% respectively. Colorado Springs Districts Bucks the Trend for English Language Learners Three Colorado Springs area school districts— Lewis-Palmer 38, Widefield 3, and Harrison 2—had some of the highest graduation rates for students learning English. They had 2015 graduation rates of 91%, 86%, and 85% respectively for their emerging multilingual students. Eaton Re-2, though it has a small population of emerging multilingual students learning English, graduated all ten of their students in that cohort in 2015. Empty Graduation Ceremonies Several districts have low graduation rates for all students. Specifically, graduation rates in Charter School Institute, Mapleton, Englewood, Westminster 50, Falcon 49, and Adams 14 fall significantly below expectations for multiple groups of students in 2015. Within this group variable progress has been made in boosting these graduation rates in the past five years. For example, Mapleton has seen big gains in graduation rates for Asian students, black students, Latinx students, and emerging multilingual students; Westminster has improved for black students, and students with disabilities, has made little progress for Latinx students, and has seen declines in the graduation rate of white students; Falcon has seen a declining graduation rate for all groups of students. A Pause on Graduation Rates for Students with Disabilities Students with disabilities represents a set of students with incredibly diverse learning needs: data is not reported separately for students with mild-moderate learning disabilities and students with severe learning disabilities. Some students might be best supported in a center-based program that serves them until they age out at age 21. Other students, especially students with mild needs, should be able to graduate on-time with the right supports. Given the variability in learning needs, we cannot derive much from the graduation rate of students with disabilities, but it is an important group of students and we would be remiss to exclude the information. Make Sure Diplomas are Meaningful A high school diploma is incredibly valuable for students. Employment rates are higher, wages are higher, upward economic mobility is more likely, and health outcomes are better for high school graduates than their peers who don’t receive a diploma. However, high school diplomas should be meaningful and should communicate that students have received a high quality education that has taught them to think critically, problem solve, and be ready for their next steps after college and career readiness. Graduation rates are arguably the easiest metric of school success to improve. As such, it is critical that gains in graduation rates are linked to comparable gains in student achievement. 19 THE OUTLIERS, 2016 Figure 17: Highest and Lowest Graduation Rates by Race or Ethnicity (2015) Black Students Highest Graduation Rates Lowest Graduation Rates St Vrain Valley Re 1J Fountain 8 School District 27J Boulder Valley Re 2 Cherry Creek 5 Jefferson County 1 State of Colorado Adams-Arapahoe 28J Colorado Spring 11 Mapleton 1 Charter School Institute Englewood 1 State Graduation Rate by Student Group 93% 90% 88% 85% 84% 84% 70% 62% 55% 45% 41% 25% Latinx Students 97% 96% 96% 94% 91% 91% Archuleta County 50 JT Platte Valley Re-7 Eaton Re-2 Cheyenne Mountain 12 Manitou Springs 14 Ellicott 22 State of Colorado Adams-Arapahoe 28J Valley Re-1 Salida R-32 Colorado Digital Boces Falcon 49 Englewood 1 Charter School Institute 68% 55% 55% 55% 42% 42% 40% 32% White Students 98% 96% 96% 96% 95% 95% Summit Re-1 Cheyenne Mountain 12 Lewis Palmer 38 Aspen 1 Eaton Re-2 Roaring Fork Re-1 State of Colorado Westminster 50 Colorado Digital Boces Mapleton 1 Charter School Institute Englewood 1 83% 57% 57% 55% 53% 51% Multiracial Students 100% 100% 96% 96% 92% 92% Windsor Re-4 Montrose County Re-1J Lewis Palmer 38 Academy 20 Douglas County Re 1 Boulder Valley Re 2 State of Colorado Englewood 1 Woodland Park Re-2 Adams-Arapahoe 28J Mapleton 1 Charter School Institute 80% 64% 55% 53% 50% 0% 25% 43% 50% 75% 100% Graduation Rate 20 THE OUTLIERS, 2016 Figure 18: Highest and Lowest Graduation Rates for Emerging Multilingual Students (2015) Highest Graduation Rates Lowest Graduation Rates State Graduation Rate by Student Group 100% Eaton Re-2 94% Brush Re-2(J) 91% Lewis Palmer 38 86% Widefield 3 85% Harrison 2 85% Sheridan 2 61% State of Colorado Adams-Arapahoe 28J 47% Valley Re-1 45% Englewood 1 41% Falcon 49 34% Charter School Institute 0% 22% 25% 50% 75% 100% Graduation Rate Figure 19: Highest and Lowest Graduation Rates for Students Eligible for Free or Reduced Price Lunch (2015) Highest Graduation Rates Lowest Graduation Rates State Graduation Rate by Student Group 91% East Grand 2 91% Eaton Re-2 90% Archuleta County 50 JT 90% Cheyenne Mountain 12 88% Platte Valley Re-7 State of Colorado 66% Valley Re-1 54% Colorado Digital Boces 52% Englewood 1 43% Falcon 49 41% Charter School Institute 32% 0% 25% 50% 75% 100% Graduation Rate Figure 20: Highest and Lowest Graduation Rates for Students with Disabilities (2015) Highest Graduation Rates State Graduation Rate by Student Group Lowest Graduation Rates 100% Archuleta County 50 JT 92% Aspen 1 88% Summit Re-1 81% Steamboat Springs Re-2 80% Manitou Springs 14 State of Colorado 54% Denver County 1 37% Adams-Arapahoe 28J 35% Falcon 49 33% Chater School Institute 30% Platte Canyon 1 0% 27% 25% 50% 75% 100% Graduation Rate 21 THE OUTLIERS, 2016 Who is Opening Doors for Graduates? High school diplomas should be a ticket to a career or college. Not all students will choose a four-year college or university after high school, but all should have the opportunity. All students should be prepared to enter the workforce or a training program after high school. After all, by 2025 seventy percent of jobs in Colorado will require some additional training or education after 12th grade. The ACT assessment provides insight into how well students are prepared for college and career. Colorado requires all students to take the ACT during their junior year. While Colorado will shift to the SAT beginning in Spring 2017, the importance of a college admissions score is clear as it often acts as a gatekeeper to higher education. For example, 75% of admitted students at Metro State University in Denver score at least an 18 out of 36 on the ACT. At the University of Colorado-Boulder 75% of admitted students score at least a 24. Too many Colorado students are excluded from these institutions because they have not developed the skills and knowledge, as measured by the ACT, through their time in the K-12 education system. The charts on the following pages show districts with the highest and lowest average ACT composite scores (a combination of scores in English, Math, Reading, and Science) for different groups of students. For reference, a school with an average ACT composite score of less than 17 would earn a “does not meet” on the state accountability framework. An average score of 20 “meets expectations.” An average composite score above 22 would earn an “exceeds expectations” on the state accountability framework. Spotlights An Opportunity Chasm The opportunity gap for students of color and low income students is well documented across a variety of educational data. Looking at information about success on the ACT, which is rarely presented for different groups of students, underscores this reality. There is no district where the average ACT composite score for black students is over 22. There is no district where the average score for students qualifying for free or reduced price lunch is over 22. There is only one district that reaches this bar for Latinx students. And their white peers? There is no district where the average score for that group of students is below 17. Nor for multiracial students. And only two districts have average scores below 17 for students who are ineligible for free or reduced price lunch. College Readiness in Colorado Springs Area Districts Districts in Colorado Springs have some of the highest average composite ACT scores across student groups. Academy 20 has some of the highest scores for black students and Latinx students. Cheyenne Mountain 12, Lewis-Palmer 38, and Academy 20 have the three highest average ACT scores for Latinx students. Those three districts, joined by neighboring Manitou Springs 14, had some of the highest scores for students eligible for free or reduced price lunch. Cheyenne Mountain 12 and Lewis-Palmer 38 also had some of the highest average composite ACT scores for white students and students ineligible for free or reduced price lunch. Lewis-Palmer 38 also had one of the highest average scores for multiracial students. College Readiness in Suburban Communities In addition to the Colorado Springs area suburban districts, including Academy 20, Lewis Palmer 38, and Cheyenne Mountain 12, other districts in suburban communities had some of the highest average composite ACT scores across student groups. The five districts with the highest average composite ACT scores for black students are all urban-suburban districts. The same is true for Latinx students, white students, and multiracial students. The exception to this trend is Aspen 1 which appears on the top districts for Latinx and white students. In fact, with the exception of two outlying towns, Aspen and Estes Park, all districts with the highest average composite ACT scores across student groups are urban-suburban districts. 22 THE OUTLIERS, 2016 The School-Level Story Which schools do the best job of opening doors for their students after graduation? A look at average school level ACT scores shows schools that offer students a greater chance of accessing higher education opportunities. A few schools appear in the top ten schools by highest average composite ACT score for multiple groups of students. Two DSST campuses in Denver County appear in the top ten: DSST: Green Valley Ranch for black students and students eligible for free or reduced price lunch, and DSST: Stapleton High School for black students, Latinx students, White students, and for both students eligible and ineligible for free or reduced price lunch. Cherry Creek High School ranks in the top ten for black students, white students, multiracial students, and students ineligible for free or reduced price lunch. D’velyn Junior/Senior High School in Jefferson County ranks in the top ten for Latinx students (with a 27.4 average composite ACT score it is significantly higher than any other school for Latinx students), white students, and students ineligible for free or reduced price lunch. Four Cherry Creek 5 high schools (Cherry Creek High School, Grandview High School, Cherokee Trail High School, and Eaglecrest High School) are amongst the top schools for highest average composite ACT scores for black students. Four high schools in Jefferson County 1 (Evergreen High School, Dakota Ridge Senior High School, Ralston Valley Senior High School, and Green Mountain High School) have amongst the highest average composite ACT scores for students eligible for free and reduced price lunch. Gaps at the School Level DSST: Green Valley Ranch is the only school in the state where black students are scoring at least an average of 22 on the ACT. That is one school out of nearly 500 high schools in Colorado. Compare that to white students. There are over 90 schools in the state where white students score an average of at least 22, and white students score an average of 26 or above at the ten schools with the highest ACT scores for that group of students. 23 THE OUTLIERS, 2016 Figure 21: ACT Scores by Race or Ethnicity (2016) Highest Average Composite ACT Scores Lowest Average Composite ACT Scores State Composite ACT Score by Student Group Black Students 19.2 19 18.8 18.7 18.5 Poudre R-1 Academy 20 Adams 12 Five Star Schools Cherry Creek 5 Jefferson County R-1 State of Colorado Falcon 49 Adams-Arapahoe 28J Colorado Springs 11 Greeley 6 Pueblo City 60 17.4 16.7 16.7 16.4 15.6 14.8 Latinx Students 22 21.5 21 20.3 20.1 Cheyenne Mountain 12 Lewis-Palmer 38 Academy 20 Aspen 1 Douglas County Re 1 State of Colorado Boulder Valley Re 2 Sheridan 2 Westminster 50 Moffat County Re 1 Ignacio 11 JT 17.7 16.1 15.7 15.7 15.2 15.1 White Students 25.1 25.1 24.1 23.7 23.6 Cheyenne Mountain 12 Aspen 1 Lewis-Palmer 38 Boulder Valley Re 2 Cherry Creek 5 State of Colorado Weld County Re-1 Westminster 50 Weld County S/D Re-8 Ellicott 22 Adams County 14 21.9 17.9 17.9 17.8 17.6 17.4 Multiracial Students 24.4 24.3 23.7 23.3 22.7 Littleton 6 Lewis-Palmer 38 Thompson R2-J Adams 12 Five Star Schools Boulder Valley Re 2 State of Colorado Fountain 8 Colorado Springs 11 Harrison 2 Adams-Arapahoe 28J Widefield 3 21.4 20 19.6 19.5 19.3 18.2 0 12 24 36 ACT Score 24 THE OUTLIERS, 2016 Figure 22: ACT Scores by Free or Reduced Price Lunch Eligibility (2016) Highest Average Composite ACT Scores Students Eligible Lowest Average Composite ACT Scores for Free or Reduced Price Lunch Cheyenne Mountain 12 Lewis-Palmer 38 Manitou Springs 14 Estes Park R-3 Academy 20 State of Colorado Students Ineligible for Free or Reduced Price Lunch State Composite ACT Score by Student Group 21.6 20.4 20.1 19.9 19.6 17.6 Roaring Fork Re-1 Adams County 14 Lamar Re-2 Sheridan 2 Westminster 50 16.2 16.1 16 15.9 15.9 25.1 24.6 24.1 Cheyenne Mountain 12 Aspen 1 Boulder Valley Re 2 24 23.5 Lewis-Palmer 38 Poudre R-1 State of Colorado Monte Vista C-8 Weld County Re-1 21.8 17.8 17.7 17.6 Weld County S/D Re-8 Westminster 50 Adams County 14 16.9 16.7 0 12 24 36 ACT Score 25 THE OUTLIERS, 2016 Figure 23: Schools with Top Ten Average ACT Scores by Student Group Top Ten Schools for Black Students by Average ACT Composite Scores (2016) Top Ten Schools for Latinx Students by Average ACT Composite Scores (2016) School DSST: Green Valley Ranch High School DSST: Stapleton High School Cherry Creek High School Grandview High School School D'velyn Junior/Senior High School Denver School Of The Arts Ralston Valley Senior High School Jefferson Academy High School Air Academy High School Discovery Canyon Campus High School Thunderridge High ThunderRidge High School School Palmer Ridge High School Rock Canyon High School DSST: Stapleton High School Liberty High School Cherokee Trail High School Eaglecrest High School Harrison High School Thomas Jefferson High School Sand Creek High School District Denver County 1 Denver County 1 Cherry Creek 5 Cherry Creek 5 Academy 20 Cherry Creek 5 Cherry Creek 5 Harrison 2 Denver County 1 Average ACT Falcon 49 18.1 23.2 21.7 21.4 20.1 19.8 19.5 18.9 18.7 18.5 District Jefferson County R-1 Denver County 1 Jefferson County R-1 Jefferson County R-1 Average ACT 27.4 23.1 23.1 23 Academy 20 22.5 Academy 20 22.4 Douglas County Re 1 Lewis-Palmer 38 Douglas County Re 1 Denver County 1 22.4 22.3 22.2 21.9 Top Ten Schools for White Students by Average ACT Composite Scores (2016) Top Ten Schools for Multiracial Students by Average ACT Composite Scores (2016) School The Vanguard School (High) DSST: Stapleton High School Liberty Common Charter School Ridgeview Classical Charter Schools D'evelyn Junior/ Senior High School Peak To Peak Charter School Evergreen High School Telluride High School Cherry Creek High School George Washington High School School Fairview High School Palisade High School Legend High School Cherry Creek High School Littleton High School Arapahoe High School Fossil Ridge High School Grandview High School Fruita Monument High School Air Academy High School District Cheyenne Mountain 12 Denver County 1 Average ACT 29.1 28.9 Poudre R-1 28.1 Poudre R-1 27.5 Jefferson County R-1 Boulder Valley Re 2 Jefferson County R-1 26.4 Telluride R-1 26.2 Cherry Creek 5 Denver County 1 27.4 26.3 26 26 District Boulder Valley Re 2 Mesa County Valley 51 Douglas County Re 1 Average ACT Cherry Creek 5 25.1 Littleton 6 24.8 Littleton 6 24.6 Poudre R-1 24.2 26.2 25.8 25.3 Cherry Creek 5 24 Mesa County Valley 51 24 Academy 20 23.6 Top Ten Schools for Students Eligible for Free or Reduced Price Lunch by Average ACT Composite Scores (2016) Top Ten Schools for Students Ineligible for Free or Reduced Price Lunch by Average ACT Composite Scores (2016) School School District Cheyenne Mountain 12 21.2 The Vanguard School (High) Liberty Common Charter School D'evelyn Junior/Senior High School Ridgeview Classical Charter Schools DSST: Stapleton High School Peak To Peak Charter School 21.1 Fairview High School Evergreen High School DSST: Green Valley Ranch High School DSST: Stapleton High School Rock Canyon High School Dakota Ridge Senior High School Ralston Valley Senior High School Monarch High School Norwood Public Schools Fossil Ridge High School Green Mountain High School District Jefferson County R-1 Denver County 1 Denver County 1 Douglas County Re 1 Jefferson County R-1 Jefferson County R-1 Boulder Valley Re 2 Norwood R-2J Average ACT 25.9 22.3 22.2 22.1 21.4 21.1 Poudre R-1 21 Jefferson County R-1 21 28.2 Poudre R-1 28 Jefferson County R-1 27.7 Poudre R-1 27 Cherry Creek High School Denver County 1 Boulder Valley Re 2 Boulder Valley Re 2 Jefferson County R-1 Cherry Creek 5 Telluride High School Telluride R-1 Evergreen High School Average ACT 26.7 26.6 26.1 26 26 25.3 Would reach the “exceeds expectations” cut point on the State Accountability Framework (the 85th percentile of all schools) 26 THE OUTLIERS, 2016 What Happens Beyond K-12? Setting students up for success is a key goal of the education system, and a path to a two- or four-year postsecondary institution is better traveled in some districts. The following tables show the districts where the highest percentage of graduates matriculate to two- or four-year higher education institutions, and the proportion of those students who need remedial classes in at least one subject when they get there.4 Remedial coursework is high school-level work and is non-credit bearing, meaning students have to pay for these classes that do not count toward a degree. Spotlights A Clearer Path to College in Districts that Set Students Up for Success There is a clear relationship: those districts where the largest percentage of their graduates Figure 24: Highest Matriculation Rates (Class of 2014) Highest Matriculation Rates Lewis-Palmer 38 Littleton 6 Cheyenne Mountain 12 Douglas County Re 1 Valley Re-1 Steamboat Springs Re-2 Boulder Valley Re 2 Academy 20 Aspen 1 Summit Re-1 Trinidad 1 75.2% 72.9% 72.5% 70.9% 70.5% 70.1% 69.3% 68.7% 68.6% 68.3% 67.7% Figure 25: Lowest Matriculation Rates (Class of 2014) 4 Matriculation data was provided by the Colorado Department of Higher Education (CDHE). Unlike some CDHE reports which just include matriculation to public Colorado institutions, the matriculation data included in this report includes matriculation to public and private institutions both within and outside Colorado, as tracked by the National Student Clearinghouse. Remediation data was gathered from CDHE. The remediation rates included only capture students enrolled at public institutions. The district-level data was also aggregated from publicly available school-level data, some of which is masked due to small class sizes. The remediation data is thus an estimate, but important bellwether in understanding college readiness. Lowest Matricluation Rates Weld County S/D Re-8 Falcon 49 Adams-Arapahoe 28J Lake County R-1 Mapleton 1 Westminster 50 Englewood 1 Sheridan 2 Ellicott 22 Adams County 14 40.7% 38.8% 38.5% 37.1% 37.0% 33.9% 31.0% 30.9% 30.0% 27.1% matriculate to a higher education institution, also tend to have low remediation rates. In fact, half of the districts in the ten districts with the highest matriculation rates also have the lowest remediation rates in the state. Conversely, half of districts with the lowest matriculation rates in the state also have the highest remediation rates. What this suggests is that students from these districts often cannot access higher education, and when they do, they are often unprepared for the work. There are opportunities for better pathways in these districts and scaffolded support to ensure students are ready to succeed in higher education settings. An important location for this work could be Trinidad 1 which has both the eleventh highest matriculation rate in the state, and the highest remediation rate. Figure 26: Highest Remediation Rates (Class of 2014) Top Remediation Rates Trinidad 1 Weld County Re-1 Westminster 50 Fremont Re-2 Adams County 14 Monte Vista C-8 Lake County R-1 Englewood 1 Weld County S/D Re-8 ~ Canon City Re-1 69.4% 68.9% 65.4% 64.7% 64.1% 60.0% 60.0% 59.0% 58.5% 54.4% Figure 27: Lowest Remediation Rates (ClassLowest of 2014)Remediation Rates Bennett 29J Elizabeth C-1 Lewis-Palmer 38 Strasburg 31J Summit Re-1 Valley Re-1 Estes Park R-3 East Grand 2 Cheyenne Mountain 12 Aspen 1 18.5% 18.1% 16.2% 16.0% 15.5% 15.3% 14.7% 13.6% 13.3% 10.0% Both highest matriculation rate and lowest remediation rate Both lowest matriculation rate and highest remediation rate 27 THE OUTLIERS, 2016 Conclusion The Outliers is a report intended to raise questions by providing a foundation of information about what is happening in Colorado school districts. School district leaders, local boards of education, the state board of education, and policymakers across the state should be taking a hard look not only at local student achievement, but at trends across the state. There are clear cases of success and improvement across the state where more students are receiving the promise of a high quality education. But these successes are not consistent enough, nor do they reach enough students. To ensure high quality education is the norm across the state, A+ recommends the following: Build an improvement culture Across the state, education stakeholders, including educators, district leaders, policymakers, and communities, should be focused on continuous improvement within the district, asking questions about what is working well for students, and what is not working for students. A culture of improvement means there is a commitment to deep investigation about challenges within schools, and about potential solutions aligned to those challenges. Share information and be transparent Transparency about student and school performance with families and communities is important for a number of reasons. First, every family deserves to find the best educational fit for their child. That means that families should understand how schools are serving students, and whether students in the school are likely to meet grade-level standards. Second, better transparency with families and communities can lead to more community-driven strategies for school improvement. Schools are more likely to meet community and student needs when families are empowered to define, advocate for, and hold policymakers and school leaders accountable for meeting student needs. Ensure lessons are learned Questions can only be asked, solutions can only be understood, if there is information about both the problem, and the strategies for improvement. But the state has taken major steps back in the amount of information about student performance that is publicly available by masking data about performance in small schools, suppressing additional results, and not releasing disaggregated data about student groups. If made available, this information can be transformational for educators and communities alike. Keep equity front and center It is critical to focus attention on discrepancies in educational opportunities both within schools districts— between schools, regions, groups of students— and between school districts. As a state, policymakers, educators, district leaders, and communities should be concerned about which students access advanced coursework, show greater growth, and reach academic expectations. There should be concern about where academic opportunity is segregated. There should be equal awareness of places that are closing achievement gaps, and serving students who have historically been left out of educational opportunities, like students from low-income families and students of color, as well as their more affluent and white peers. We hope this report sparks questions, and ultimately contributes to a foundational conversation about how to ensure every student in Colorado receives an excellent education. 28 THE OUTLIERS, 2016 Appendix A: Colorado District Map RANGELY RE-4 DE BEQUE 49JT MESA COUNTY VALLEY 51 WEST END RE-2 NORWOOD R-2J MANCOS RE-6 DURANGO 9-R TELLURIDE R-1 GARFIELD RE-2 BAYFIELD 10 JT-R SILVERTON 1 OURAY R-1 RIDGWAY R-2 MONTROSE COUNTY RE-1J DELTA COUNTY 50(J) PLATEAU VALLEY 50 GARFIELD 16 MEEKER RE1 MOFFAT COUNTY RE NO 1 DOLORES RE-4A DOLORES COUNTY RE NO.2 MONTEZUMA-CORTEZ RE-1 IGNACIO 11 JT HAYDEN RE-1 ROARING FORK RE-1 STEAMBOAT SPRINGS RE-2 SOUTH ROUTT RE 3 POUDRE R-1 ESTES PARK R-3 WESTMINSTER 50 MAPLETON 1 SHERIDAN 2 JEFFERSON COUNTY R-1 WOODLAND PARK RE-2 GREELEY 6 PAWNEE RE-12 WIGGINS RE-50(J) PRAIRIE RE-11 FORT MORGAN RE-3 BYERS 32J WELDON VALLEY RE-20(J) BRIGGSDALE RE-10 WELD COUNTY SCHOOL DISTRICT RE-3J PLATTE VALLEY RE-7 EATON RE-2 AULT-HIGHLAND RE-9 WINDSOR RE- WELD COUNTY RE-1 WELD COUNTY S/D RE 8 STRASBURG 31J BRUSH RE-2(J) REVERE SCHOOL DISTRICT HAXTUN RE-2J YUMA 1 JULESBURG RE-1 HOLYOKE RE-1J WRAY RD-2 BURLINGTON RE-6J IDALIA RJ-3 BETHUNE R-5 LAMAR RE-2 VILAS RE-5 HOLLY RE-3 WALSH RE-1 GRANADA RE-1 PLAINVIEW RE-2 CHEYENNE COUNTY RE-5 STRATTON R- LIBERTY J-4 HI-PLAINS R-23 OTIS R-3 LONE STAR 101 FRENCHMAN RE-3 District Border County Border PLATEAU RE-5 VALLEY RE-1 BUFFALO RE-4J AKRON R-1 ARICKAREE R-2 ARRIBA-FLAGLER C-20 EADS RE-1 MC CLAVE RE-2 PRITCHETT RE-3 SPRINGFIELD RE-4 WILEY RE-13 JT KIT CARSON R-1 LAS ANIMAS RE-1 GENOA-HUGO C113 WOODLIN R-104 LIMON RE-4J DEER TRAIL 26J AGATE 300 BIG SANDY 100J KARVAL RE-23 CHERAW 31 CROWLEY COUNTY RE-1-J MANZANOLA 3J ROCKY FORD R-2 SWINK 33 KIM REORGANIZED 88 CAMPO RE-6 Produced by the Colorado Department of Education Web Management Team - April, 2014 EAST OTERO R-1 BRANSON REORGANIZED 82 HOEHNE REORGANIZED 3 FOWLER R-4J EDISON 54 JT MIAMI/ YODER 60 JT CALHAN RJ-1 KIOWA C-2 BENNETT 29J SCHOOL DISTRICT 27J SCHOOL D STRICT 27J ADAMS COUNTY 1 ADAMSARAPAHOE 28J ENGLEWOOD 1 CHERRY CREEK 5 LITTLETON 6 ELIZABETH C-1 ELBERT 200 PEYTON 23 JT FALCON 49 ELLICOTT 22 HARRISON 2 HANOVER 28 FOUNTAIN 8 CHEYENNE MOUNTAIN 12 WIDEFIELD 3 MANITOU SPRINGS COLORADO SPRINGS 11 1 ACADEMY 20 LEWIS-PALMER 38 DOUGLAS COUNTY RE 1 DENVER COUNTY 1 ST VRAIN VALLEY RE 1J JOHNSTOWNMILL KEN RE-5J THOMPSON R2-J Colorado School District Map EAST GRAND 2 ADAMS 12 FIVE STAR SCHOOLS BOULDER VALLEY RE 2 G LPIN COUNTY RE-1 CLEAR CREEK RE-1 PLATTE CANYON 1 PARK COUNTY RE-2 SUMMIT RE-1 NORTH PARK R-1 WEST GRAND 1-JT. EAGLE COUNTY RE 50 ASPEN 1 LAKE COUNTY R-1 BUENA VISTA R-31 CRIPPLE CREEK-VICTOR RE-1 CANON CITY RE-1 FREMONT RE-2 PUEBLO CITY 60 TRINIDAD 1 AGUILAR REORGANIZED 6 PUEBLO COUNTY 70 HUERFANO RE-1 PRIMERO REORGANIZED 2 LA VETA RE-2 CUSTER COUNTY SCHOOL DISTRICT C-1 COTOPAXI RE-3 MOFFAT 2 ALAMOSA RE-11J SANFORD 6J CENTENNIAL R-1 SIERRA GRANDE R-30 SANGRE DE CRISTO RE-22J SALIDA R-32 MOUNTAIN VALLEY RE 1 CENTER 26 JT SARGENT RE-33J NORTH CONEJOS RE-1J MONTE VISTA C-8 DEL NORTE C-7 GUNNISON WATERSHED RE1J HINSDALE COUNTY RE 1 CREEDE SCHOOL DISTRICT ARCHULETA COUNTY 50 JT SOUTH CONEJOS RE-10 29 THE OUTLIERS, 2016 Appendix B: Percentile Analysis The A+ percentile analysis in the Outliers report compares the relative performance of districts on previous and current tests: TCAP 2013 and 2014, and CMAS PARCC 2015 and 2016. The analysis includes results from Elementary English Language Arts, Elementary Math, and Middle School English Language Arts. Secondary math is not included given that students can choose between subject specific tests and are not necessarily comparable. High School English Language Arts is not included given low participation rates in many districts. Methodology Percentile ranks compare districts on the basis of the percent of students who met the grade-level benchmark (level 4 and above on PARCC; meets or exceeds expectations on TCAP) in a particular test and grade range on the 2013 TCAP, 2014 TCAP, 2015 PARCC, and 2016 PARCC assessments. Grades were grouped as follows: - 3-5 (elementary students) - 6-8 (middle school students) This analysis relied on publicly available data. The Colorado Department of Education implemented additional data suppression rules in 2015 and 2016. These rules include: - Minimum n-size = 16 (no reporting on cohorts of students with fewer than 16 students) - Minimum cell-size = 4 (no reporting when a single cell, or the difference between valid scores and results cell, is less than 4) For the 2016 analysis of PARCC scores, results from specific grades were included only if a) there were more than 15 valid scores, and b) results of the valid scores were reported. In 2015, results were included when a) there were more than 15 valid scores, b) results of the valid scores were reported or results could be estimated (this change in methodology in 2015 to 2016 is due to changed reporting rules from the Colorado Department of Education). Calculation of percent of students at benchmark: PARCC (Math and English Language Arts) and TCAP Math: N students at benchmark % of students at benchmark = N valid scores TCAP Reading and Writing (combined to provide a better comparison to 2015 PARCC English Language Arts exams): TCAP Reading and Writing % at benchmark = (N students at benchmark in Reading + N students at benchmark in Writing) (N valid scores Reading + N valid scores Writing) Selection Criteria for Inclusion as an Outlier For each subject area, A+ calculated the range of percentile changes from 2013 to 2016. Districts with sufficient data from TCAP 2013 and PARCC 2016 were included. “Outliers” were selected to be highlighted in the report if the percentile rank change was greater than one standard deviation from the average change. Those cut points are listed below: Elementary English Language Arts Average Percentile Change 2013-2016: 1.8 Standard Deviation (based on range of district percentile change 2013-2016): 22.8 Elementary Math Average Percentile Change 2013-2016: -0.5 Standard Deviation (based on range of district percentile change 2013-2016): 27.5 Middle School English Language Arts Average Percentile Change 2013-2016: -1.1 Standard Deviation (based on range of district percentile change 2013-2016): 23.8 While the percentile analysis, and change in percentile ranks, was calculated using the full set of districts with available data, A+ included only districts with an enrollment greater than 1,000 students in the reporting set. 30 THE OUTLIERS, 2016 Appendix C: District Demographic Analysis Methodology To better compare like-districts based on their demographics, every district was assigned a District Demographic Score. This methodology mirrors closely what Denver Public Schools uses to compare similar schools, and is based on research of student factors that are often correlated to academic performance on standardized tests. The Index was calculated according to the following formula: District Demographic = Index (40% X proportion of students qualifying for free or reduced price lunch) + (20% X proportion of emerging multilingual students (ELL)) + (20% X proportion of students receiving special education services) + (20% X district mobility rate) A+ then produced a correlation between student performance in the district (percent of students meeting grade-level standards on PARCC 2016) and the District Demographic Index. Those correlations are below: Correlation: Elementary English Language Arts and District Demographic Index r= -0.6 R2= .36 Correlation: Middle School English Language Arts Performance and District Demographic Index r= -0.5 R2= .25 Selection Criteria for Inclusion as an Outlier To identify “Outliers,” A+ compared actual performance in a district to the correlated value based on the District Demographic Index and performance in districts across the state. A+ calculated the range of the discrepancy between actual and correlated performance, and identified those districts that performed at least 0.8 standard deviations from the correlated value; 30-40% (depending on the subject area and grade level) of districts were identified as “Outliers,” falling outside the trend line. While the district demographic analysis, and correlation to performance, was calculated using the full set of districts with available data, A+ included only districts with an enrollment greater than 1,000 students in the reporting set. Correlation: Elementary Math Performance and District Demographic Index r= -0.5 R2= .27 31 THE OUTLIERS, 2016 ABOUT A+ COLORADO The mission of A+ Colorado is to sharpen public education by building public will and advocating for the changes necessary to dramatically increase student achievement in schools and districts in Colorado. We are an independent, nonpartisan 501(c)(3) organization working to bring the power of data and research to challenge ourselves, educators and policymakers to rethink public education. © 2017 A+ Colorado All rights reserved. A+ Colorado 1390 Lawrence St, Suite 200 Denver, CO 80204 Email: admin@apluscolorado.org Tel: 303.736.2549 www.apluscolorado.org 32