Inside Online Charter Schools

Transcript

1 A REPORT OF THE NATI ONAL STUDY OF ONLINE CHARTER SCHOOLS Online Charter Inside Schools October 2015 Brian Gill Lucas Walsh Claire Smither Wulsin Holly Matulewicz Veronica Severn Eric Grau Amanda Lee Tess Kerwin Submitted to: Walton Family Foundation P.O. Box 2030 Bentonville, AR 72712 Project Officer: M arc Holley Contract Number: 12- 02 2013- Submitted by: Mathematica Policy Research 955 Massachusetts Avenue Suite 801 Cambridge, MA 02139 Telephone: (617) 491- 7900 Facsimile: (617) 491- 8044 Project Director: Brian Gill

2 This p age has been left blank for double- sided copying .

3 ACKNOWLEDGEMENTS This report is one of three produced by the National Study of Online Charter Schools. We gratefully acknowledge the input of colleagues working on the other two reports n Lake of —Robi the Center on Reinventing Public Education and Macke Raymond and Lynn Woodworth of the —who provided helpful suggestions on Center for Research on Educational Outcomes (CREDO) y findings. Phil Gleason of the content of our principal surveys and on the interpretation of surve Mathematica also gave thoughtful feedback on an earlier draft of this report, helping us improve it substantially. John Kennedy expertly edited the report. Early in the study, to inform the survey design, we solicited input from a group of stakeholders with knowledge of online charter schools. We are grateful for the input of Greg Richmond from the National Association of Charter School Authorizers; Bruce Friend from the -12 Online Learning (iNACOL) ; Deanna Rowe from the Arizona International Association for K State Board for Charter Schools; Christy Wolfe from the National Alliance for Public Charter Schools; Mary Gifford from K12 Inc.; and Pat Laystrom from Connections Education. This study could not have been conducted without t he willing participation of more than a hundred principals of online charter schools across the country; we thank them for their participation. iii

4 This p age has been left blank for double- sided copying .

5 CONTENTS ... xi EXECUTIVE SUMMARY INTRODUCTION I. ... 1 NE CHARTER SCHOOLS A THE UNIVERSE OF ONLI ... 3 II. ND THEIR STUDENTS Online charter schools and enrollments, by state ... A. 3 ... 4 B. Grades served by online charter schools Students of online charter schools ... 5 C. CURRICULUM, INSTRUCT ION, ASSESSMEN T, AND TECHNOLOGY ... 9 III. A. ... 9 Delivery of instruction Class size and student ... 13 B. –teacher ratio Instructional and other support services ... 15 C. Curriculum and specialized programs ... 17 D. E. ... 17 Student assessment 18 ... F. Technology IV. STUDENT AND PARENT E NGAGEMENT ... 21 A. Student engagement ... 21 B. ... 22 Parent engagement TEACHERS AND STAFFIN SCHOOLS ... 25 V. G AT ONLINE CHARTER Staffing levels and teaching loads ... 25 A. Teacher hiring and experience ... 26 B. C. ... 27 Teachers’ responsibilities and expectations 29 ... D. Monitoring and evaluating teachers E. Teachers’ compensation ... 31 F. Professional development ... 31 VI. ... 33 SCHOOL LEADERS 33 ... A. Previous experience B. Training and professional development ... 33 C. Roles and responsibilities ... 33 D. Challenges ... 34 E. Principal evaluation and compensation ... 35 v

6 CONTENTS ESEARCH MATHEMATICA POLICY R 37 ... VII. GOVERNANCE, MANAGEMENT, AND FUNDING VIII. CONCLUSION ... 39 REFERENCES ... 41 APPENDIX A ... 43 : SURVEY METHODS APPENDIX B : SUPPLEMENTAL TABLE ... 49 vi

7 TABLES ... 18 III.1 Frequency of student assessments Technology provided to students of online charter schools ... III.2 18 V.1 Mean and median number of support staff per school ... 26 V.2 Te achers’ responsibilities in online charter schools ... 29 V.3 ... 29 Teacher evaluation methods and frequency V.4 Factors that affect teachers’ compensation in online charter schools ... 31 VI.1 Online charter principals’ greatest challenges ... 35 vii

8 This p age has been left blank for double- sided copying .

9 FIGURES Number of online charter schools operating by state, with statewide student II.1 ... enrollments 4 ... 5 II.2 Percentage of online charter schools serving grades 4, 7, and high school ... 6 Race and ethnicity of online charter students and other students in their states II.3 III.1 -paced instruction ... 9 Percentage of online charter schools exclusively using self ... 10 III.2 Instructional methods used frequently in online charter schools Median weekly time spent in synchronous instruction ... 11 III.3 Distribution of time spent in synchronous instruction in online charter high III.4 ... 12 schools Synchronous instruction support tools 12 ... III.5 III.6 ... 13 Asynchronous instruction support tools Median class sizes ... 13 III.7 ... 14 Percentage of schools with typical class size of 50 or more III.8 Students p er full -time teacher ... 15 III.9 Percentage of online charter schools using instructional staff in addition to the III.10 ... 15 teacher in a typical class III.11 on -one instructional support in online charter schools ... 16 Staff responsible for one- Monitoring students’ participation 21 IV.1 ... ... 22 IV.2 Roles of parents in online charter schools Frequency of communication wi th parents ... IV.3 23 V.1 Percentage of online charter schools with different numbers of full -time ... 25 equivalent teachers V.2 Most important factors in teacher hiring in online charter schools ... 27 V.3 Percentage of online charter schools expecting teachers to be available at different hours of a weekday ... 28 ... 30 V.4 Most important factors for teacher evaluation V.5 Frequency of teachers’ participation in professional development ... 32 VI.1 Percentage of online charter principals’ time spent on various tasks ... 34 VI.2 Performance measures used for compensation of princi pals ... 36 VII.1 Support from affiliated online charter school management organizations ... 38 ix

10 This p age has been left blank for double- sided copying .

11 EXECUTIVE SUMMARY charter schools — charters or cyber charters —are publicly Online also known as virtual funded schools of choice that eschew physical school buildings and use technology to deliver education to students in their own homes. These schools typically provide students with computers, software, and network- s to teachers via based resources, while also providing acces or tele email, telephone, web, and/ conference. From one perspective, online schools represent the kind of innovation that proponents of ve envisioned since their inception: online charter schools deliver instruction charter schools ha a radically different approach than using conventional public schools. Nonetheless, critics of online charter schools worry that they might not be effective in promoting student learning. mbitious and This report and its companion volumes describe the findings of the most a comprehensive study of online charter schools to date, conducted jointly by Mathematica Policy Research, the Center for Research on Education Outcomes (CREDO) at Stanford University, and the Center on Reinventing Public Education (CRPE) at the University of Washington. In this provide the first nationwide data on the operations and instructional approaches of volume, we online charter schools from a survey completed by a majority of all online charter schools across the country. In the second volume (Pazhouh et al. 2015) , CRPE describe s the policy environments of online charter schools and provide s recommendations to state policymakers. In the third volume (Woodworth et al. 2015) , CREDO measures the achievement effects of online charter schools. with a snapshot of online charter schools operating across the country, This volume begins e, and the students they serve. We then describing their numbers, the states in which they operat describe the instructional programs of online charter schools ; meth ods used to engage students and parents, along with expectations of parental involvement ; the teachers and principals of management and governance. online charter schools; and their The universe of o nline charter schools About 200 online charter schools ar e operating in the United States, serving about 200,000 students . Student enrollment in online charter schools is highest in Ohio, Pennsylvania, and California, each of which had more than 25,000 students enrolled in 2012– 2013; together those three states account for half of the online charter enrollments nationwide. Figure ES.1 shows the , alongside the total number of number of online charter schools operating in 2013–2014 by state students they enrolled in the preceding year. Nearly all online charter schools serve high school grades, but large numbers serve middle and elementary grades as well. More than half (56 percent) of online charter schools serve students in all three grade ranges — elementary, middle, and high. Individual online charter schools vary widely in size. Many are small, but a handful of large schools dominate. Almost a quarter (24 percent ) of online charter schools enrolled more than 1,000 students in 2012–2013, account ing for 79 percent of total enrollment in the sector. xi

12 EXECUTIVE SUMMARY ESEARCH MATHEMATICA POLICY R state ES.1. charter schools operating by nline , with Number of o Figure statewide student enrollments 50 40,000 45 35,000 40 30,000 35 25,000 30 25 20,000 20 15,000 15 10,000 online charter schools 10 5,000 5 Number of online charter schools Total number of students served by 0 0 IN NH DC FL NM KS AR HI MI LA OK MN SC WI OR TX NV CO AZ GA CA ID PA OH UT Number of online charter schools (2013-2014) Total number of students served (2012-2013) Students of online schools charter Online charter schools might be attractive to various kinds of students and families, including homeschoolers, rural students, student s with disabilities, highly mobile students, and percent ) of students who are not well suited to conventional schooling. A large majority (90 10 reas online charter schools reported that they serve a general population of students, whe ly on serving a specific population of students with particular needs. The percent focus primari data available to this study do not identify the reasons that students enroll in online charter schools, but they provide a broad snapshot of the characterist ics of online charter s tudents. re Nationally in 2012–2013 , 71 percent of online charter students we re white, 14 percent we black, and 12 percent we re Hispanic, with very small percentages of other groups (Figure ES.2) . Online charter schools have an overrepresentation of white s tudents and an underrepresentation of Hispanic students. xii

13 EXECUTIVE SUMMARY MATHEMATICA POLICY R ESEARCH ethnicity of online charter students and other students and Figure ES.2. Race in their states 100 90 80 71% 70 60 49% 50 40 31% Percentage 30 15% 14% 20 12% 5% 10 2% 1% 1% 0 Hispanic Asian or Asian Pacific White Black American Indian, Islander Alaska Native All public schools (includes only states with online charter schools) Online charter schools Students with disabilities are represented in online charter schools at approximately the earners are substantially same rate as ; E nglish l in public schools overall (14 percent) In the 10 states with available data, 0.4 percent of online charter students were underrepresented. identified as English learners, compared with 4.3 percent of all public school s tudents. The average length of enrollment for a typical student in an online charter school is about two years. instruction, , and t assessment Curriculum, echnology Online schooling creates both constraints and opportunities for delivering instruction. Thre e- quarters (76 of online charter schools include courses that are self -paced rather than tied percent) -third of online charter schools rely exclusively on self -paced courses . to the calendar. One the in -paced courses, Consistent with the prevalence of self structional method used most -driven independent study ( Figure frequently in online charter schools is individualized, student , students and that is -guided synchronous discussion ( ). Schools reported that teacher ES.3 teachers participating in discussion at the same time) is the next most frequently used lectures is used less frequently, and ollaborative learning instructional method for all grades. C . -fourth of online charter schools at any grade level are not used frequently in more than one xiii

14 EXECUTIVE SUMMARY ESEARCH MATHEMATICA POLICY R ES.3 . Ins Figure in online charter schools tructional methods used frequently 100 90 80 70 60% 59% 55% 60 50 42% 36% 40 32% Percentage 30% 27% 30 21% 20% 19% 20 11% 10 0 Individualized, student- Collaborative learning Teacher-guided Lecture synchronous discussion involving two or more driven independent study students working together 4th grade 7th grade High school As might be expected, in most online charter schools synchronous instruction occupies less ypical time than it does in conventional schools. The difference is dramatic: students in the t less synchronous instructional time in a week than students in a brick online charter school have in a day and mortar school have . Online charter schools report ed a median of four hours per week spent in synchronous instruction in 4th grade, three hours in 7th grade, and three hours in ). Larger online charter schools tend to have somewhat more high school (Figure ES.4 synchronous instructional time: the median online charter student is in a school that provides five hours of synchronous instruction per week in 4th and hours in high school. 7th grade s and six ES.4 weekly time spent in synchronous instruction . Median Figure 4.5 4 4 3.5 3 3 3 2.5 2 1.5 1 Median hours per week 0.5 0 4th grade 7th grade High school xiv

15 EXECUTIVE SUMMARY MATHEMATICA POLICY R ESEARCH Schools support synchronous instruction in several ways, the most common of which are phone calls ( ferencing ( 84 percent ), videoconferencing ), screen sharing or web con 88 percent tele ). Schools also reported using online chat 75 percent (76 percent), and audioconferencing ( -on- one chats, and text messages. forum s, instant messaging or other one ed typical class sizes of 25 students at the elementary and Online charter schools report middle school levels, rising to 30 students at the high school level. At the high school level, more than one ed typical class sizes of 50 or more. Data from the -third of online charter schools report U.S. Department of Education show substantially larger student –teacher ratios in online charter -mortar schools (Figure ES.5). On average, online charter schools and schools relative to brick- -time teacher, compared with 20 in brick -and- mortar charter schools have 30 students per full and 17 in conventional public schools in the same states . ES.5 -time teacher . Students per full Figure 35 29.9 30 25 20.0 17.4 20 15 10 Number of students 5 0 Brick-and-mortar charters Online charters Conventional public schools —83 percent at the elementary level, 91 percent at Most online charter s middle the chools one instructional support for students. school level, and 89 percent in high school —offer one -on- -on- But ed medians of only 45 to 60 schools that offer one one instructional support report one instructional time per week . In other words, the small amount of -on- minutes of one synchronous instructional time provided by most online charter schools is not coupled with a large amount of one -on- one interaction with teachers. This suggests that most online charter schools expect that the bulk of learning will occur during a student’s individual engagement with . the cou rse material, perhaps with the help of a parent ed offering a variety of other support services, including Most online charter schools report behavioral health services (65 percent), speech and language therapy (87 percent), and dropout -third (35 percent) of online charter t recovery programs (63 percent). But one prevention or credi schools do not offer special instruction for English learners, perhaps helping to explain why such students are underrepresented in the sector. -quarters of on In more than three purchased from a vendor a are line charter schools, curricul or provided by the management organization with which the school is affiliated. Most schools ed that their curriculum offerings include programs in fine arts (80 percent) and music (61 report (83 percent) offer percent) , as well as core academic subjects. M ost online charter high schools , but AP enrollments are usually small, with a median of only Advanced Placement (AP) courses xv

16 EXECUTIVE SUMMARY ESEARCH MATHEMATICA POLICY R h schools report ed AP 10 students participating. Nonetheless, 5 percent of online charter hig enrollments that exceed 600 students. Student assessment is an important aspect of the instructional program in online charter schools. Most schools conduct a diagnostic assessment of students’ skills when they initially he schools administer frequent assessments to track students’ progress in the enroll, and t curriculum. ed that they conduct student About two- thirds of online charter schools report assessments with weekly or greater frequency at all grade levels. computers in only 56 percent of online schools, and another 33 percent All students receive . Eleven percent of schools expect families to supply a computer to some but not all students Internet access. Most schools provide a computer, and 40 percent expect families to provide by remote control of their computers, but 13 provide technical support to students, usually percent of schools do not provide live technical support to students. arent Student and p engagement When asked an open -ended question about their greatest challenges in leading online charter schools, principals identified student engagement most often—nearly three times as often as any other issue . This challenge is inherent to online schooling, because the school has no way to ensure that students are “in their seat s” and focused on their coursework. The challenge is likely to be particularly acute for the subset of students enrolled in online charter schools because they were not fully engaged in conventional, brick- -mortar schools. Recognizing this, most online and charter schools monitor student engagement and participation through completion of course assignments (93 percent), activity in the online system (94 percent), and (less often) participation in synchronous interaction with the teacher (58 percent). Most online charter schools have substantial expectations of parents, surely necessitated in part by the limits of the schools’ tools for keeping students engaged. Expectations vary for particular monitoring and support tasks (Figure ES.6). Most notably, many onl ine charter schools —ranging from 43 percent in high school to 78 percent in elementary school —expect parents to actively participate in the student’s instruction (Figure ES.6). The schools are clearly aware that the success of their approach depends on substantial parental support a few —which is even more critical when the school provides only hours of synchronous instructional time each week. The expectation of active parental involvement in instruction m ight also explain why the great majority of online c harter schools ask parents to participate in training sessions. xvi

17 EXECUTIVE SUMMARY MATHEMATICA POLICY R ESEARCH Figure ES.6. Role of parents in online charter schools Percentage of online charter schools expecting parents to play particular roles 96% 96% 96% 100 90 80% 78% 80 70% 64% 70 59% 56% 60 50% 44% 50 43% 40 Percentage 30 20 10 0 Verify seat time Participate in parent Make sure the student Actively participate in the training sessions keeps up with assignments student's instruction High school 4th grade 7th grade Teachers and principals ring, online charter schools share one top priority with a In decisions about teacher hi -and management -mortar charter schools affiliated with a charter school national sample of brick organization (CMO ): willingness to work hard in support of the school’s mission is one of the two mos t important considerations. The online schools report ed that their second most important consideration in hiring is the teacher’s certification status —which distinguishes the online schools from the brick -and- mortar CMO schools, for which performance teaching a sample lesson is more important. Teachers in online charter schools tend to have more responsibility for individual attention to students than for developing curriculum, lesson planning, and lecturing—areas in which some schools minimize teacher resp onsibilities by purchasing curriculum and relying extensively on self -paced courses that involve no lectures. The most important factors that schools consider in s’ achievement growth, a evaluating teachers are observations of teacher s’ instruction, student nd teacher s’ accessibility to students. Nearly all schools (92 percent) said their teachers participated in professional development either online or in person (or both). Most but not all schools (89 percent) said they provided teachers with paid time for professional development. Almost half ( 48 percent) of the principals of online charter schools had no prior experience teaching in an online environment . Principals are responsible for a wide variety of tasks. According to survey responses, principals spend the largest part of their time —an average of 30 percent of their work week —performing internal administrative tasks related to issues such as human resources, regulations, reports, and school budget s. Not surprisingly, principals of online s spend a relatively small percentage of their time —11 percent —interacting with charter school students, far less than the 39 percent reported by principals of public schools nationally in 2011– 2012 according to federal data. xvii

18 EXECUTIVE SUMMARY ESEARCH MATHEMATICA POLICY R Governance and management More than half (57 percent) of online charter schools are affiliated with school management organizations . Almost half of the schools affiliated with a management organization are affiliated with one of the two largest online school management organizations, K12 and Connecti ons. Most of the schools affiliated with management organizations received several commonly received services were professional kinds of services from the organization. The most development for teachers (received by 91 percent of affiliated schools); curri culum and instructional materials (86 percent); and diagnostic assessments (75 percent). Conclusion school sector barely existed before 2000 and has grown rapidly since , The online charter Online charter sch now enrolling about 200,000 students nationally. ools serve a reasonably Hispanic students and English learners are underrepresented, perhaps in diverse population, but part because many of the schools do not offer instruction targeted to students whose first language is not English. In addition, the fac t that most online charter schools expect families to provide Internet access (and many do not uniformly provide computers to all students) m ight prevent some prospective students from enrolling. On average, students in online charter schools remain enroll ed for about two years. Online schooling creates both constraints and opportunities for modes of delivering instruction. In most online charter schools, a substantial amount of coursework is self -paced. Online charter schools have substantially less synchr onous interaction between teachers and students than conventional schools. This could be partly related to the higher ratio of students to teachers in online schools. With a limited number of live contact hours and a lean staffing model, online charter schools place substantial expectations on parents, who are expected not only to ensure that students keep up with assignments but also to participate in training sessions and, in a large number of schools, to actively participate in the student’s instruction. The substantial burden placed on parents is presumably a response to the issue that online charter school principals regard as their greatest challenge: keeping students engaged. The challenge is partly inherent to online schooling, because the schools ha ve no way to ensure that students are in their seats and ready to learn. It is probably partly due to online schools serving as -and - schools of last resort for a subset of their students who previously disengaged from brick mortar schools. And it is surely exacerbated by a high student –teacher ratio and a small number of live contact hours. All of these are reasons for concern about whether the sector is likely to be effective in promoting the achievement of its students —an issue addressed in de pth in the third volume of this study (Woodworth et al. 2015). xviii

19 I. INTRODUCTION charter schools — charters or cyber charters —are publicly Online also known as virtual funded schools of choice that eschew physical school buildings and use technology to deliver education to students in their own homes. These schools typically provide students with computers, software, and network- based resources, while also providing access to teachers via email, telephone, web, and/ or tele conference. From one perspective, online schools represent the kind of innovation that proponents of charter schools have envisioned since their inception: schools instruction using a n approach that differs radically from instruction online charter deliver rapid growt in conventional public schools. And the h of online schools over the past decade (including state -operated online schools as well as online charter schools) - and district considerable demand for online instruction. ton demonstrates that there is Indeed, Clay Christensen, the originator of the con cept of disruptive innovation, has predicted that half of all school courses will be conducted online by 2018 (Christensen et al. 2008). high Nonetheless, critics of online charter schools worry that they might not be effective in promoting student learning. The small number of studies that have attempted to measure the effects of online schools on student achievement have provided some evidence supporting this concern: students in online charter schools in Ohio (Zimmer et al. 2009), Pennsylvania (Raymond 2011), and California (Zimmer et al. 2003) have fallen short of their peers in terms of student achievement growth. O ne study (Ritter and Lueken 2013) found more favorable results in Arkansas. Before the current study, no one has conducted a rigorous, syste matic examination of the achievement effects of online charter schools across multiple states. This report and its companion volumes describe the findings of the most ambitious and comprehensive study of online charter schools to date Mathematica Policy , conducted jointly by Research, the Center for Research on Education Outcomes (CREDO) at Stanford University, and the Center on Reinventing Public Education (CRPE) at the University of Washington. In this volume, we provide the first nationwide data on the operations and instructional approaches of online charter schools from a survey completed by a majority of all online charter schools across In the second volume (Pazhouh et al. 2015) , our colleagues at CRPE describe the the country. policy environments of online charter schools and provide recommendations to state policymakers. In the third volume (Woodworth et al. 2015) , CREDO measures the achievement effects of online charter schools and connects the student achievement data with our survey data and state policy data to assess whether specific practices of online charter schools or specific state policies are associated with better achievement impacts. In this volume , we describe the universe of online charter schools, relying primarily on our arter school principals, supplemented by publicly available school -level data from survey of ch the U.S. Department of Education. We begin with a general snapshot of online charter schools operat e, and the operating across the country, describing their numbers, the states in which they students they serve (in terms of characteristics such as race and ethnicity, special needs, and English learner status ). We then describe characteristics of the instructional programs of online charter schools, including methods of delive ry of instruction, class sizes, methods of assessin g student learning, curriculum and support services, and technology. The following section addresses methods used to engage students and parents, along with expectations of parental 1

20 I. INTRODUCTION MATHEMATICA POLICY R ESEARCH involvement. The report then turns to the teaching staff in online charter schools, describing hiring expectations, teaching loads, professional development opportunities, methods of evaluating performance, and compensation. We continue with a section on the principals of online charter schools, describing their prior experience, training, and responsibilities, and the issues they perceive as the greatest challenges for online charter schools. T he volume then includes a section on management and governance , before a summary of co nclusions . 2

21 II. THE HARTER SCHOOLS AND T HEIR STUDENTS UNIVERSE OF ONLINE C , by state A. Online charter schools and enrollments ing a precise count of the number of online charter schools in operation across the Ob tain —and there are two types of hybrids. Some country is difficult, because some schools are hybrids schools offer blended instructional programs in which instruction is partly online and partly in person. Other schools offer different types of instructional programs, with some students enrolled onl ine and other students enrolled in person. We define an online charter school as a school that students are enrolled entirely online. If a offers an instructional program in which some or all school offers an online program alongside a program that offers conventional, in- person instruction, we attempted to include it in our study for the subset of students participating in the fully online program. About 200 online charter schools operat the United States in the 2013–2014 school ed in . We received surv year ey responses from 126 of these schools. Survey weights were used to Details on the survey methodology are available in A ppendix A . adjust for nonresponse. The U.S. Department of Education’s Common Core of Data (CCD) that the sector indicates served about 200,000 students in the 2012–2013 school year (the most recent year for which 1 (U.S. Department of Education 2013- 14) nationwide data are available) . This represented an increase of an order of magnitude since the turn of the millennium, when fewer than 20,000 2 Molnar et al. 2015). students were enrolled; most of the growth has occurred since 2005 ( Consistent with this rapid growth, most currently operating online charter schools are relatively he median online charter school operating in 2014 had opened only six years earlier . About new: t had initially opened before 2001, and more than one -third had opened 5 percent of the schools since 2010. Student enrollment in online charter schools is highest in Ohio, Pennsylvania, and California, each of which had more than 25,000 students enrolled in the 2012–2013 school year ; together , th ose three states account for half of the online charter enrollments nationwide. Figure II.1 shows the number o f online charter schools operating in the 2013–2014 school year , by state the total number of students they enrolled in the preceding year. alongside States are sequenced from those with the largest enrollments in online charter schools to those with the smallest enrollments. 1 The figure is an estimate because enrol lment data were missing for 12 schools and dates were missing from earlier years for 6 additional schools. 2 Molnar et al.’s (2015) count includes noncharter and charter schools, but 84 percent of those counted as 2013 enrollees were enrolled in charter schools. 3

22 II. UNIVERSE OF ONL INE CHARTER SCHOOLS MATHEMATICA POLICY R ESEARCH I.1. charter schools operating by s tate , with Figure I Number of online student enrollments statewide 50 40,000 45 35,000 40 30,000 35 25,000 30 25 20,000 20 15,000 15 Number of online charter schools 10,000 10 5,000 5 Total number of students served by online charter schools 0 0 KS AR HI MI LA OK UT MN NH IN ID SC WI OR DC TX FL NM NV CO AZ GA CA PA OH Number of online charter schools (2013-2014) Total number of students served (2012-2013) B. Grades served by online charter schools school grades, but large numbers serve middle Nearly all online charter schools serve high and elementary grades as well. The survey grade at the elementary asked respondents about one level (4th) and one grade at the middle school level (7th) to avoid potential ambiguities about the definitions of elementary and middle grades, and to ensure that respondents could answer questions specifically when there might be There was no need to differences across grade levels. specify the grade level for questions about high school classes. Figure II.2 shows the percentage of online charter schools serving students at each of the three grade ranges in the 2013–2014 56 percent half ( ar, according to survey responses ) of online charter schools school ye . More than —elementary, middle, and high. serve students in all three grade ranges 4

23 II. UNIVERSE OF ONL ESEARCH INE CHARTER SCHOOLS MATHEMATICA POLICY R serving grades 4, 7, and h II.2. Percentage of online charter schools Figure igh school 96% 100 90 82% 80 70 58% 60 50 40 Percentage 30 20 10 0 4th grade High school 7th grade Individual online charter schools vary widely in size. Many are quite small: 42 percent had year , according to national data . But a fewer than 100 students enrolled in the 2012–2013 handful of large schools dominate the sector. Almost a quarter (24 percent ) of online charter schools enrolled more than 1,000 students, and these accounted for 79 percent of total enrollment in the sector . Indeed, in the absence of physical constraints, a few 2013 school year in the 2012– online charter schools have grown to sizes that would dwarf almost every conventional, brick- in and -mortar school: w e identified seven charter schools that enrolled more than 5,000 students . Despite this, (located in Georgia, Ohio, Pennsylvania, and Texas) the 2012–13 school year schools reported that state or local laws or policies impose d constraints on almost 60 percent of their growth, consistent with CRPE’s finding that 14 states restrict charter enrollments (Pazhouh . et al. 2015) schools C. Students of online charter In principle, online charter schools might be attractive to various kinds of students and families . Families who wish to school their children at home might be attracted by the staff, curriculum, and technology resources that online charter schools can provide. Rural students who -and liv e far from the nearest brick -mortar school could find online schooling appealing. Some students with disabilities might find online education more accessible than conventional schools. Online education could provide stability in programming for highly mobile students. Students with social or behavioral difficulties, or with learning styles not well suited to conventional schooling, might find refuge in online educational environments. And online charter schools might be schools of last resort for students who have been unsuccessful in conventional schools, whether for academic, social, or behavioral reasons. The data available to this study do not enable us to identify the reasons that students enroll in online charter schools, unfortunately, but n provide a broad snapshot of the characteristics of online charter students. they ca 5

24 II. UNIVERSE OF ONL INE CHARTER SCHOOLS ESEARCH MATHEMATICA POLICY R Across the sector, the racial and ethnic distribution of students in online charter schools in . differed somewhat from that of the public schools of thei the 2012–2013 school year r states , 71 percent of online charter students we re white, 14 13 school year in the 2012–20 Nationally re black, and 12 percent we re Hispanic , with very small percentages of other groups percent we tates where online charter schools operat (Figure II.3). S e tend to have an overrepresentation of s. The data do not provide reasons white students and an underrepresentation of Hispanic student for this, but it is possible that the overrepresentation of white students could be due in part to the desirability of online schools to rural students (who are likely to have fewer conventional school options near home). The underrepresentation of Hispanic students might be due to lower ight be immigrants and not speak knowledge of options among students and parents who m En - glish as a first language. The underrepresentation of Hispanic students is not true for brick -mortar charter schools, which have a slightly higher percentage of Hispanic students relative and to the national average (Condition of Education 2015). I.3. Race and ethnicity of o nline charter students and o ther students Figure I tates in their s 100 90 80 71% 70 60 49% 50 40 31% Percentage 30 15% 14% 20 12% 5% 10 2% 1% 1% 0 Hispanic Asian or Asian Pacific White American Indian, Black Alaska Native Islander All public schools (includes only states with online charter schools) Online charter schools Nine ed that they serve a general population of online charter schools report ty percent of reas 10 percent focus primarily on serving a specific population of students with students, whe particular needs. Most of the 10 percent of schools serving special populations target dropouts or over -age/undercredited students. More than a third (35 percent ) of online charter school students are reported to be low - gible for free or reduced . This number is lower than the income students eli -price lunch (FRL) ight in their states (53 percent) , but it m average percentage for all public schools underestimate the true percentage of those who live in poverty: l acking facilities in which to d eliver lunch, online charter schools are presumably unlikely to participate in the federal FRL program, in which case they might not collect the relevant data. 6

25 II. UNIVERSE OF ONL INE CHARTER SCHOOLS MATHEMATICA POLICY R ESEARCH learners and students with disabilities are available in federal databases at Data on English the district level, but not the school level. Because charter schools constitute their own school districts in many states, it is possible to calculate the percentage of English learners and students 3 with disabilities in the online charter schools operating in those states. In the 10 states where it ), students with was possible to calculate these percentages (for the 2012–2013 school year disabilities were represented in online charter schools at approximately the same rate as in public : 14.4 percent of online charter students had disabilities ( that is , they had schools overall ducation plans) ; 13.9 percent of public school students overall in the same states individualized e and- had disabilities. This result contrasts with findings from various studies of brick- mortar ound an underrepresentation of students with disabilities charter schools, which usually f ( for . Online charter schools m ight provide unique example, Government Accountability Office 2012) opportunities for some students with disabilities who are not well served in conventional schools. In contrast to students with disabilities, English learners appear to be substantially underrepresented in online charter schools relative to their percentages in the larger public school population. In the 10 states with available data, 0.4 percent of online charter students were identified as English learners , compared with 4.3 percent of all public school students. Again, this differs from their representation in brick -and age of -mortar charter schools, where the percent learners nationally is approximately equivalent to their representation in conventional English U.S. Department of Education 2013- 14). public schools ( reported that the Online charter schools average length of stay for a typical student is a little more than two years , which is consistent with what CREDO f ound in student records dat a (Woodworth et al. 2015) . Nonetheless, the average school reported that nearly 40 percent of i ts students who took the state assessments at the school in the spring of 2014 had not been enrolled in the school for the entire school year. 3 The 10 states include Arizona, Georgia, Idaho, Indiana, Louisiana, Michigan, Minnesota, Ohio, Pennsylvania, and This variable was calculated at the district level, whereas other CCD data w ere calculated at the school level. Utah. most onli Because s, this comparison is valid. ne charters are their own school district 7

26 This p age has been left blank for double- sided copying .

27 III. ION, ASSESSMENT, AND TECHNOLOGY CURRICULUM, INSTRUCT A. Delivery of instruction Online schooling creates both constraints and opportunities for modes of delivering an instructional program. One potential advantage of online schooling is the opportunity to tailor the In most online charter an individual student. pace of instruction to the needs and desires of -paced. Three- quarters (76 percent) schools, a substantial amount of coursework is self of online -paced rather than tied to the calendar. Three -fifths charter schools include courses that are self percent ) reported that half or more of their courses a re self -paced —including one -third (3 (60 3 percent) of online charter schools that only self -paced instruction (though these enroll only offer . Schools relying exclusively on self 17 percent of the students in online schools) -paced instruction a re not limited to those serving middle or hig h school students: they include more -quarter of schools serving 4th grade as well as about one than one (34 percent) of schools -third serving 7th grade and high school grades (Figure III.1). In addition, the large majority of schools allow students to earn course credits by demonstrating mastery regardl ess of seat time , either in all courses ( of schools courses, subjects, or grades (45 percent of 40 percent ) or in selected ). schools Figure - III .1. Percentage of online charter schools exclusively using self paced instruction 100 90 80 70 60 50 40 34% 34% Percentage 28% 30 20 10 0 7th grade High school 4th grade -paced courses, the instructional method used most Consistent with the prevalence of self frequently in online charter schools is individualized, student ( Figure -driven independent study 55 percent of schools use it ich common in 4th grade ( III.2). Independent study is nearly as in wh frequently) as in 7th grade (59 percent) and high school (60 percent ). For younger students in particular, this suggests that online charter schools must rely on considerable parental oversight. schools. , we discuss parental responsibilities in online charter Later 9

28 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY ESEARCH MATHEMATICA POLICY R that is , students and teachers Schools reported that teacher -guided synchronous discussion ( participating in s the next most frequently used instructional a discussion at the same time) i method for all grades, used frequently in 32 to 42 percent of schools, depending on grade level ollaborative learning involving two or more students working together i s used (Figure III.2). C frequently in 30 percent of schools for 4th graders, 27 percent for 7th graders, and 21 percent for nts, ac cording to survey responses . high school stude ional, brick- One way in which online charter schools can differ substantially from convent is that they rely very little on traditional lectures. Survey respondents and -mortar schools -fourth of online charter schools re not used frequently in more than one indicated that lectures a (Figure III.2). A higher proportion of schools reported using lectures more at any grade level grade (19 percent) levels than in 4th grade (11 frequently at the high school (2 0 percent) and 7th- (59 percent) of online charter s chool s reported using lectures rarely or percent) . More than half never. The full list of responses on the use of different instructional approaches is noted in ndix B, Table B .1. Appe III .2. Instructional methods used frequently in online charter schools Figure 100 90 80 70 60% 59% 55% 60 50 42% 36% 40 Percentage 32% 30% 27% 30 21% 20% 19% 20 11% 10 0 Individualized, student- Teacher-guided Collaborative learning Lecture synchronous discussion driven independent study involving two or more students working together 4th grade 7th grade High school As might be expected, in most online charter schools, synchronous instruction (with teacher time than it does in conventional schools and students working at the same time) occupies less . In fact the difference is dramatic: students in the t ypical online charter school have less synchronous instructional time in a week than students in a brick and mortar school have in a day . Online charter s hours per week spent in synchronous four chools reported a median of grade, and three hours in high school instruction in . III.3) ( Figure 4th grade, three hours in 7th Larger online charter schools tend to have somewhat more synchronous instructional time : the hours of synchronous instruction median online charter student is in a school that provides five 7th grade per week in 4th and s, and six hours in high school. These numbers do not include 10

29 III. CURRICULUM, INSTRUCTIO ESEARCH N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R contact that occurs asynchronously (such as through email), which is not readily measurable in . hours and minutes III .3. Median weekly time spent in synchronous instruction Figure 5 4 4 3 3 3 2 Median hours per week 1 0 7th grade 4th grade High school 6 hours or fewer of synchronous instruction weekly. Most online charter schools have y online III.4 shows the distribution of weekly synchronous instructional time reported b Figure . In sum, the total amount of contact fewer ave 6 hours or quarters h charter high schools. Three- hours for students in most online charter schools is quite small. As the figure indicates, however, a small number of online charter schools provided substantially more synchronous instructional time: about 5 percent reported more than 20 hours weekly for high school students. Distributions of weekly time in synchronous instruction for 4th and 7th grade s (not shown) are similar to the high school distribution. t synchronous instruction in several ways, the most As shown in Figure III.5, schools suppor -sharing or web conferencing (84 common of which are telephone calls (88 percent), screen percent), videoconferencing (76 percent), and audioconferencing (75 percent). Schools also one chats (63 -on- at forums (68 percent), instant messaging or other one reported using online ch percent), and text messages (59 percent). A few online schools (6 percent) reported they used none of these supports. chools reported using many different to ols to support Similarly, online charter s asynchronous instruction (with students and teachers working at different times) . T he most common tools reported were email ( 96 percent of schools ), websites with instructional focus or interactive online exercises 4 percent) III.6). Many schools content (9 5 percent), and (Figure (9 recordings of lectures (79 percent), discussion (83 percent), also reported using online textbooks forums or groups (72 percent), physical textbooks ( ). 60 percent social media ( 64 percent), and reported using at least one support method. ( 99 percent) Almost all schools 11

30 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY ESEARCH MATHEMATICA POLICY R III .4. Distribution of Figure time spent in synchronous instruction in online charter high schools 40 35 30 25 23% 18% 20 17% 17% Percentage 15 11% 10 8% 5% 5 1% 0 15.1– 15.0 10.1– 30.1+ 30.0 0 0.1– 2.0 2.1– 4.0 4.1– 6.0 6.1– 10.0 Average hours per week III .5. Synchronous instruction support tools Figure 100 88% 90 84% 76% 75% 80 68% 70 63% 59% 60 50 40 Percengate 30 20 10 0 Screen Phone calls Video Online chat Instant Audio Text messaging sharing/web conferencing conferencing messaging (IM) forum conferencing (Skype, or other one- FaceTime, and on-one chats so on) 12

31 III. CURRICULUM, INSTRUCTIO MATHEMATICA POLICY R ESEARCH N, ASSESSMENT AND TECHNOLOGY Figure III .6. Asynchronous instruction support tool s 96% 95% 94% 100 90 83% 79% 80 72% 64% 70 60% 60 50 40 Percentage 30 20 10 0 Other Email Interactive Recordings of Online Discussion Physical Social media websites with online forums or lectures textbooks (paper) (blogs, wikis) instructional exercises threaded textbooks focus or discussion content groups B. Class size and student –teacher ratio Online charter schools reported median class sizes of 25 students in 4th grade in both math III.7) (these data exclude schools that rely entirely on ELA and English/Language Arts ( ) ( Figure individualized, self -paced courses for which class size has no clear meaning) . Similarly, in 7th 25 students in grade, median class sizes were also reported to be both math and ELA . Median ise to 30 students in both subjects in high school. reported class sizes r Figure III.7. Median class sizes 35 30 30 30 25 25 25 25 25 20 15 10 Number of students 5 0 4th grade 7th grade High school Math English/language arts . Classes of Median class sizes obscure wide variation among different online charter schools more than 50 students are relatively common in middle and high school grades (Figure III.8). 13

32 III. CURRICULUM, INSTRUCTIO ESEARCH N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R Figure III .8. Percentage of schools with typical class size of 50 or more 100 90 80 70 60 50 39% 35% 40 Percentage 27% 25% 30 19% 16% 20 10 0 7th grade High school 4th grade Math English/language arts Larger class sizes are more common in larger online schools, so the percentage of students enrolled in schools with typical class size s of 50 or more students is greater than the numbers shown in Figure III.8. One -quarter o f 4th -grade students grade and high school and half of 7th- students are enrolled in schools with typical class sizes of 50 or more. Shifting our attention from reported class size to the ratio of students to full -time teachers and facilitates -mortar charter schools and to a comparison of online charter schools to brick- –teacher rat , it is possible to compare student ios conventional public schools. Using the CCD across sectors. Figure III.9 shows the average student –teacher ratio in online charter schools -mortar charter schools and conventional public schools in the states pared with brick- com and where the online charter schools operat e. Online charter schools have nearly 30 students per full - and time teacher, substantially more than the 20 students per teacher fo und in brick- -mortar . charter schools and 17 students per teacher in conventional public schools in the same states About one -quarter of online charter schools report ed that their typical class includes instructional staff in addition to the teacher. As seen in Figure III.10, there is not much difference across grades and subjects in the proportion of schools using additional instructional staff. Moreover, schools with larger classes are no more likely to use additional instructional staff to support the teacher of record. Among the schools reporting typical class sizes of more than 50 students, no more than 36 percent (and fewer in most grades and subjects) reported using supplemental instructional staff in addition to the teacher of record (not reported in the figure). 14

33 III. CURRICULUM, INSTRUCTIO MATHEMATICA POLICY R N, ASSESSMENT AND TECHNOLOGY ESEARCH teacher III .9. Students per full Figure -time 35 29.9 30 25 20.0 20 17.4 15 Number of students 10 5 0 Online Charters Brick-and-Mortar Charters Conventional Public Schools . Percentage of online charter schools using instructional staff in III.10 Figure addition to the teacher in a typical class 100 90 80 70 60 50 40 Percentage 29% 27% 27% 25% 30 21% 19% 20 10 0 4th grade 7th grade High school English/language arts Math C. Instructional and other support services Nearly all online cha rter schools ( 97 percent ) indicated they provide support services to -skills classes; two students 5 percent) -thirds (6 : a lmost three ( 73 percent) offer study -quarters offer mental or behavioral health services ; 87 percent offer speech and language therapy ; and 63 percent have a dropout prevention or credit recovery program for students who previously o. Among the schools that have a drop- dropped out of school or are at risk of doing s out ng students is 20. Only prevention or credit recovery program, the median number of participati schools offer special instruction for English learners , two -thirds ( 65 percent) of online charter 15

34 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R ESEARCH . E xtra help for perhaps helping to explain why such students are underrepresented in the sector e form of supplemental group instruction in 84 percent of struggling students is available in th schools and one -on- . one tutoring in 90 percent of schools, according to their reports -on- one instructional he majority of schools report More generally, t ed that they provide one support to students. Eighty -three percent of online charter elementary schools, 91 percent of -on- middle schools, and 89 percent of high schools reported a typical student spends time in one -on- one one interaction with a teacher or tutor in an average week. Schools that offer one reported medians of 45 to 60 minutes of one -on- one instructional time per instruction al support el. In other words, the small amount of week, with only small differences by grade lev synchronous instructional time provided by most online charter schools is not coupled with a -on- one interaction with teachers. large amount of one This suggests that most online charter schools expect that the bulk of learning will occur during a student’s individual engagement with of a parent. the course material, perhaps with the help -on- one instructional support, most expect the teacher of Among the schools providing one record for the course (Figure III .11 ). Many schools also use tutors or coaches to provide support to provide one -on- one instructional support, with the f raction rising from one -third in 4th grade to half in 7th grade and high school . Figure III .11. Staff responsible for one- on-one instructional support in online charter schools on- one instructional support Percentage of online charter schools using each type of staff for one- 100 87% 90 78% 80 69% 70 60 53% 52% 50 35% 40 Percentage 30 20 14% 12% 10% 9% 8% 6% 5% 5% 10 5% 5% 5% 3% 2% 1% 1% 0 Teacher of Other Tutor/coach Other teacher Special One-on-one Parent record for the instructional education support not course staff faculty provided 7th grade High school 4th grade 16

35 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R ESEARCH pecializ ed programs D. Curriculum and s - Most online charter schools do not develop their own curriculum. For more than three quarters of online charter schools, curricul s ( 51 percent) or a are purchased from outside provider provided b y the school’s affiliated management organization (27 percent). Some schools (19 percent) create their curricul a in -house, which is either developed by individual course instructors (1 and shared across related courses 2 percent) or developed for the school as a whole schools reported their curricul a are a blend of purchased curricul a (7 percent). A small number of and in- house development. schools report ed that they offer students opportunities for instruction in fine arts (8 0 Most music (6 1 pe rcent). Sixty percent of schools have a talented or gifted program. percent) and -quarter s (73 ) of schools offer clubs or activities for students. Almost three percent Many schools offer additional specialized programs for high school students, mainly in the form of Adv , offered by 83 percent of online charter high schools. anced Placement (AP) courses But in most of these schools, only a few students are enrolled in AP courses : median AP school s reported enrollment among schools with AP courses is 10. At the high end, 5 percent of 600 students in AP courses. A few more than ools (3 percent) offer an online charter sch -school students . International Baccalaureate (IB) program to high E. Student assessment Most online charter schools conduct a diagnostic entry assessment for students who ha ve just enrolled (not an assessment to determine eligibility for admission) . An overwhelming majority of schools (95 percent) report ed that they pull the student’s record from the previous nt say that they visit the student’s home school, 81 percent call the household, and 28 perce . In the entry assessments, 79 percent of schools reported they measure academic skills, 6 6 percent measure English language skills, and 60 percent measure potential barriers for online learning. alf of scho ols (4 9 percent) reported they assess the level of parent al or other home Almost h -nine supports for online learning. Fifty percent of schools assess students for learning 6 percent assess for other disabilities. disabilities, and 4 nline charter schools asses s their students Most o thirds of schools frequently. About two- report ed that they conduct student assessments with weekly or greater frequency at all grade levels ( Table III.1) ; assessments are especially frequent in larger schools . The percentage of schools conducting assessments at least weekly varies only slightly by grade and subject, from 63 to 71 percent. These numbers include 1 percent of schools that conduct assessments in 1 to 15 daily ELA and math classes . In middle and high school math and ELA classes, about a quarter of schools reported that the frequency of student assessments varie d based on student pace, whe reas only 13 percent of 4th- grade math and ELA classes had assessments that varied in frequency depending on the student’s pace . Eighty -seven per cent of online charter schools reported that preparation for state assessments is embedded in regular courses. Three -fifths of online charter schools (enrolling 73 percent of online charter students) reported requiring students to take a separate test prep aration course. Intensive, targeted support for students who have difficulty achieving proficiency 17

36 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R ESEARCH standards on state assessments occurred in 77 percent of schools (enrolling 96 percent of online charter students), according to respondents. Table III.1. Frequency of student assessments of online charter Percent assessing students with each frequency age schools Varies based on Less often than student pace Weekly monthly Daily Monthly rade, ath g 13 5 8 M 17 0 13 4th g rade, ELA 13 5 6 1 8 0 4th 13 7th rade, M ath 1 2 53 9 0 25 g g 1 1 53 7th 9 0 2 7 rade, ELA s M ath 15 5 2 High 6 1 2 6 chool, High s chool, ELA 1 2 5 1 7 1 2 8 F. Technology chools provide a variety of technology to at least some of Most online charter s their rs are not uniformly provided to all students . All , in many schools compute ; however students computers in only 56 percent of online schools (serving 35 percent of students students receive in the sector) a computer to some but not all students (Table , and another 33 percent supply Internet III.2). service is provided (directly or by subsidizing purchase from an outside provider) to all students in 29 percent of schools and to some students in an additional 31 percent of ; the remaining 40 percent of schools expect families to secure schools access . Computer Internet accessories such as webcams, microphones, headsets, printers, or scanners are provided to all 5 percent of schools and to some students by 34 percent of schools. Table III.2 students by 4 on technology provision. shows the full list of school responses of online charter schools Table III.2. Technology provided to students age of online charter Percent schools School provides to School provides to some but not all Provided by all students students students’ families for example, Internet Internet connection ( 29 31 40 service or subsidy for I nternet service, modem, router, and/or hot spot) Computer ( laptop or desktop 11 5 6 3 3 for example, computer, or tablet computer such as iPad) Computer a ccessories ( for example, 2 2 4 5 3 4 webcam, microphone, hea d set, CD/DVD drive, printer, or scanner) Online charter s -related technical support to students in a variety chools provide technology of ways. The most common technical support that schools provide is troubleshooting via remote control of the s tudent’s computer ( 61 percent) . Forty -two percent of schools use an online ticketing system to handle students’ technical problems of schools (4 2 percent) . Fewer than half report ed that they provide in- person set up of student computers. Eighty -seven percent of schools 18

37 III. CURRICULUM, INSTRUCTIO N, ASSESSMENT AND TECHNOLOGY MATHEMATICA POLICY R ESEARCH that provide live technical support offer live technical support to students. Of these schools , the is available to students outside of standard business hours on weekday evenings in live support and on weekends in 31 pe 46 percent of schools . rcent of schools Most but not all schools also provide technical support to teachers. Seventy -five percent of schools report ed live support via tele phone or chat is available to teachers, and 49 percent provide manuals, technical guides, or frequently ed-question documents to teachers . In -ask addition to standard business hours, 28 percent of schools report that l ive tech nical support is available to teachers during weekday evenings and 18 percent of schools make it available to teachers on weekends . 19

38 This p age has been left blank for double- sided copying .

39 IV. STUDENT AND PARENT ENGAGEMENT A. Student engagement When asked an open -ended question about their greatest challenges in leading online charter schools, principals identified student engagement most often—nearly three times as often as any his challenge is inherent to online schooling, because the school has no way to other issue. T ensure that the students are in their seats and focused on their coursework. The challenge is likely to be particularly acute for the subset of students enrolled in online cha rter schools because they were not fully engaged in conventional, brick -and -mortar schools. online charter schools report ed schoolwide policies spelling out expectations for Most s (99 percent) report students. Almost all school ed schoolwide policies for the completion of assignments, and 91 percent of schools have policies for class participation. Nearly three- quarters of schools (74 percent ) have schoolwide policies on attendance in synchronous instruction. In addition to setting expectations for students, schools also use online tools to monitor student s’ attendance and participation. Almost all schools monitor students’ completion of course assignments (9 3 percent ) and activity in the online system ( 94 percent ), and the majority (58 percent ) monitor stude nts’ seat time in synchronous work with a teacher (Figure IV .1) . Figure I s’ participation V.1 . Monitoring student Percentag e of online charter schools using each approach to monitor student s’ participation 100 94% 93% 90 80 70 58% 60 50 Percentage 40 30 20 10 0 Activity in the online system Pace of student’s completion of Seat time involved in synchronous work with a teacher course assignments 21

40 IV. STUDENT AND PAR ENT ENGAGEMENT MATHEMATICA POLICY R ESEARCH most schools first respond by contacting a When a school identifies a student as disengaged, report ed that they place personal tele phone parent. The vast majority of schools (97 percent) calls to a parent when a student is disengaged, and 95 percent email the parent. Forty percent of ed tha report schools report ing social ed contact t they actually visit the home, and 47 percent . Half of schools report ed offering services organizations on behalf of disengaged students participation incentives to students identified as disengaged. B. Parent engagement antial expectations of parents —surely necessitated in Most online charter schools have subst part by the limits of the schools’ tools for keeping students engaged, but perhaps also a side effect of the small number of contact hours they provide for students. Not s urprisingly , parental responsibilities are greatest at the elementary level, but online charter schools expect parents to play a role even for high school students. Nearly all online charter schools, at all grade levels, expect that parents will make sure the student keeps up with (Figure school assignments ). Expectations vary for other monitoring and support tasks. IV.2 About half of online charter schools expect parents to verify students’ seat time . M ore than half of schools at all grade levels (including 8 0 percent of schools at the elementary level) expect parents to participate in training programs. Figure I in online charter schools V.2 . Roles of parents Percentage of online charter schools expecting parents to play particular roles 96% 96% 96% 100 90 80% 78% 80 70% 70 64% 59% 56% 60 50% 50 44% 43% 40 Percentage 30 20 10 0 Verify seat time Participate in parent Actively participate in the Make sure the student student's instruction training sessions keeps up with assignments 4th grade 7th grade High school Perhaps most nota bly, many online charter schools expect parents to actively participate in the student’s instruction. At the upper elementary (4th- grade) level, nearly four of five online charter schools (78 percent) expect parents to participate actively in the student’s instruction 22

41 IV. STUDENT AND PAR ENT ENGAGEMENT ESEARCH MATHEMATICA POLICY R (Figure IV.2). This percentage declines for older students, but even at the high school level, 43 percent of online charter schools expect parents to participate actively. The schools are clearly s on substantial parental support aware that the success of their approach depend —which is surely even more critical when the school provides only a few hours of synchronous instructional ight time each week. The expectation of active parental involvement in instruction m also explain why the great major ity of online charter schools ask parents to participate in training sessions. s’ progress to parents via email, regularly communicate student Schools reported that they -third of schools (35 percent tele ne charter phone, or postal mail. One , serving 25 percent of onli ) send information to parents once a week, with the next most reported frequencies being students about once a month (27 percent) and at the end of each quarter or semester (2 2 percent), as seen in Figure percent of schools (serving 5 percent of students) reported contacting IV.3 . Only 9 -quarters (77 percent ) of More than three parents more frequently, twice or more every week. schools provide parents with a measure of student engagement or participation. IV.3 . Frequency of communication wi Figure th parents 100 90 80 70 60 50 35% 40 Percentage 27% 30 22% 20 9% 5% 10 2% 0 At the end of When each This school does About once a About once a Twice or more each quarter or course is week every week not actively send month semester completed student progress information to parents —report ed monitor ing teacher s’ contact with Most online charter schools —94 percent students and parents. Almost two -thirds (64 percent ) of online charter schools indicated that teachers’ contact with students and parents monitoring is a respons ibility of the principal. Smaller numbers of schools gave other staff the responsibility for monitoring contact. Six percent of schools do not formally monitor teacher s’ contact with students and parents. 23

42 This p age has been left blank for double- sided copying .

43 V. TEACHERS AND STAFFIN OOLS G AT ONLINE CHARTER SCH teaching l oads A. Staffing l evels and , but the great majority of a large number of teachers online charter schools employ A few re small. Online charter schools report an average of 34 full -time equivalent these schools a s skewed, as can be seen in Figure V.1 . Four percent of (FTE) teachers, but the distri bution i more than 300 FTE teachers. The number of FTE teachers at the median online schools employ -eight percent of online charter schools employ 30 or fewer charter school i s only 10. Seventy chers . FTE tea , but these serve only 27 percent of online charter students Figure V.1 . Percentage of online charter schools w ith different numbers of -time e quivalent teachers full 35 29% 30 25 21% 20% 20 15 Percenage 10 8% 4% 5 2% 2% 2% 2% 2% 1% 1% 1% 1% 1% 0% 0% 0% 0% 0% 0% 0% 0% 0 65 percent) rely on both ful Most online charter schools ( l- and part , but across -time teachers the sector, schools report -time. Only 8 ed that four of every five teachers (81 percent) are full rel y excl usively on part -time percent of schools (serving 1 percent of online charter students) teachers. These schools a re small, having on avera ge 21 part -time teacher s. Most schools do not have many additional instructional or support staff, relative to the number of teachers on staff. The median online charter school does not use any tutors or teacher aides/instructional assistants, and ha s only one guidance counselor and one other instructional support person (Table V.1 ). Mean numbers of instructional support staff are only slightly higher. 25

44 V. TEACHERS AND STA FFING MATHEMATICA POLICY R ESEARCH V.1 . Mean and median number of support staff Table per school Mean Median Teacher aides/instructional as sistants 1.7 0 1.6 0 Tutors 2.1 1 Guidance counselors 3. 1 Other instructional support staff 3 In conventional schools, teachers in higher grades tend to be responsible for larger numbers y. T he of students due to subject specialization and the teaching of multiple class sections each da specialize in younger online environment could, in principle, make it possible to for teachers to . In practice, o nline charter schools show grades as well as higher grades a similar pattern of g more students at higher grades. Among the schools serving the specialized teachers servin elementary level, 89 percent report -grade core academic teachers are generalists ed that most 4th who are responsible for multiple subjects. , full -time In the median online charter high s chool s a re expected to teach 12 5 students, more than the medians for teacher -grade teacher s ( 50 7th students) and grade teacher s (4 0 students). Loads are slightly heavier in larger online charter 4th- schools. s In addition, the maximum student load reported for a typical high school teacher i greater (500 students), compared wi th 7th- grade teachers (270 students) and 4th- grade teachers (100 students). B. Te acher hiring and experience Online charter s chools reported the most important factor when hiring a teacher i s the ranked most teacher’s commitment to the school’s mission and willingness to work hard ( and as one of the top three factors by 88 percent of schools ) important by 51 percent of schools V.2 nd to be the most important (Figure ). Commitment to the school’s mission was similarly fou factor in hiring decisions in a national study of charter school management organizations -and -mortar charte r schools (Furgeson et al. 2012) . A teacher’s (CMOs) that operate brick t factor certification status was the second most importan in hiring in online charter schools and in the top three factors by 81 percent (ranked as most important by 36 percent ). In this -and -mortar CMO -affiliated schools respect, online charter schools differed from brick , for which performance in teachi ng a sample class was the second most important factor in hiring determinations. Online charter schools reported that a ll other factors ranked well behind the first two in importance in hiring decisions general experience teaching was deemed a top- three . A teacher’s three factor in 3 7 factor in 40 percent of schools, and experience teaching courses online a top- . Three other candidate factors —performance in teaching a sample lesson , percent of schools other factors, and college major in content area to be taught —were in the top three factors for 16, 16, and 12 percent of schools , respectively . Two factors listed —college grade point average and holding a master’s degree —were each chosen among the three most important factors for fewer than 3 percent of schools . Two factors — core on a test ( such as Praxis) and the quality of candidate’s preservice teacher training program —were never chosen among the three most important factors for teacher hiring and are not shown in the figure . 26

45 V. TEACHERS AND STA FFING MATHEMATICA POLICY R ESEARCH ed that the median t years. The median Schools report eacher has taught in the school for four among schools of different sizes. tenure is similar across schools serving different grades and Given that these schools are quite new, it is not surprising that this is a shorter duration than public school teachers nationally, for whom the average tenure in their current schools was eight years according to the Department of Education’s 2011–2012 Schools and Staffing Survey . (SASS) V.2 . Most important factors in t eacher hiring in online charter schools Figure 100 90 80 70 60 50 40 30 20 10 0 Certification Commitment General Experience Performance Other College College Master’s status (holds experience teaching in teaching factors, not major in to this grade point degree a valid as a teacher courses sample class listed content area average school’s teaching online mission / to be taught certificate) willingness to work hard First, second, and third most important factors 2nd 1st 3rd C. Teacher s’ responsibilities and expectations Even though students’ time in synchronous instruction is limited in most online charter expect teachers to be available to students during the hours when schools, the schools (Figure V.3 ). conventional schools are typically operating, from 8:00 a.m. until 3:00 p.m. :00 p. percent -one m. Forty -two percent of schools expect teachers to be available after 5 Twenty teaching. of schools reported that their teachers come to a central location to do most of their 27

46 V. TEACHERS AND STA FFING MATHEMATICA POLICY R ESEARCH V.3. Percentage of online charter schools Figure teachers to be expecting available at different hours of a weekday 100 93% 93% 93% 93% 92% 90% 89% 90 78% 80 70 60 50 38% 40 Percentage 30 21% 20 15% 14% 7% 10 3% 3% 1% 0% 0 10:00 11:00 12:00 1:00 2:00 6:00 4:00 10:00 7:00 8:00 9:00 5:00 6:00 7:00 8:00 9:00 3:00 a.m. p.m. a.m. p.m. a.m. p.m. a.m. p.m. a.m. p.m. a.m. p.m. p.m. p.m. p.m. p.m. p.m. For the classes they teach, t eachers in online charter schools tend to have more responsibility for individual attention to st , lesson planning , and than for developing curriculum udents . Teachers are responsible for identifying struggling learners, grading student work, lecturing , and tutoring (Table V.2 ). communicating with parents , according to 86 to 96 percent of schools Meanw , two -thirds (68 percent) of schools reported the teacher of record for a particular hile class is responsible for its lesson planning ; 63 percent made teachers responsible for lecturing (which, as previously noted, is not used at all in some schools) ; and only 3 8 percent of schools (serving 26 percent of students) reported the teacher is responsible for developing the curriculum. The relatively large number of teachers who are not even expected to plan their own lessons m ight be due partly to the fact that curricul a are typically purchased or provided, and perhaps is also related to the fact that so many courses are self -paced. In most schools (86 percent) teachers are also responsible for managing online learning environments. In 38 percent of schools, teac hers are also responsible for troubleshooting technical issues (Table V.2) . 28

47 V. TEACHERS AND STA FFING MATHEMATICA POLICY R ESEARCH V.2 . Teachers’ responsibilities Table in online charter schools Percentage of schools in which teachers are responsible for activity Identifying struggling learners 96 Grading work 9 2 student s’ 91 Communicating with parents - - one tutoring 8 6 on One for example, Managing online learning environments ( 86 online forums or discussion boards) 6 8 Lesson planning Lecturing 63 38 Developing curriculum 8 ues Troubleshooting technical iss 3 D. Monitoring and evaluating teachers In almost all online schools (93 percent), a principal or other administrator is expected to (Table V.3 ). Observation necessarily means something observe and provide feedback to teachers nstruction than it does in conventional classrooms, but the same technology different for online i that enables teachers to provide instruction online can enable administrators, instructional coaches, or teaching colleagues to observe the online interactions between teachers and students. A majority of online charter schools indicated that other educators, in addition to the principal, -eight conducted observations and feedback. Fifty percent of schools reported that their teachers are observed by peers, and 59 percent (serving 74 percent of students) said their teachers are observed by a master teacher or coach . Table V.3 . Teacher evaluation methods and frequency Percentage of schools reporting number of times per year each event occurs 8 or more 2 or 3 4–7 times Once Not at all times times 42 17 27 8 6 Observed by and received feedback from a peer 41 1 3 Observed by and received feedback from a 26 16 5 master teacher or someone else who coaches teachers 15 7 Observed by and received feedback from a 4 3 13 22 eone else who principal, administrator, or som monitors performance Asked to submit lesson plans to a master 3 1 5 5 7 19 6 teacher, department chair, principal, or other administrator for review In the median school, teachers are observed by principals or administrators tw o to three times annually (Table V.3) . This is less frequent than reported by principals of brick- and -mortar charter schools in CMOs, most of whom reported that their teachers were observed by 29

48 V. TEACHERS AND STA FFING ESEARCH MATHEMATICA POLICY R times annually (Furgeson et al. 2012). That study also included a administrators nearly eight comparison group of nearby (mostly urban) conventional public schools in wh principals ich conducted observations four to seven times annually —less frequently than their colleagues in CMOs, but still more frequently than online charter principals. Two to three times annually was also the median number of observations by master teachers or coaches and by peers, in the subset of online charter schools in which those staff conducted observations. Only 4 5 percent of schools (servin g 57 percent of students) expected teachers to submit lesson plans for review, perhaps in part because curricul a were often centralized and teachers did not have responsibility for planning lessons. More detail on observations and review included in Table V .3. of lesson plans is chools reported the most important factors they consider when evaluating Online charter s their teachers. By far, the most important factors considered are observations of teacher’s s’ achievement growth, and the teacher’s accessibility to students ( instruction, student Figure V.4 ). Feedback from colleagues or coaches was rarely used in teacher evaluation, despite the fact that most schools said that colleagues and coaches observed teachers’ instruction. . Most important fact ors for teacher evaluation V.4 Figure Percentage of schools ranking each factor first, second, or third in importance 100 90 First, Second, and Third Most Important Factors 80 2nd 1st 3rd 70 60 50 40 30 20 10 0 Students' Observations Teacher’s Meeting Students' Feedback Feedback Portfolio of achievement expectations course of teacher’s from students accessibility from other examples of growth for student completion or parents teachers or students' instruction to students engagement rate or instructional work progress coaches toward course completion 30

49 V. TEACHERS AND STA FFING MATHEMATICA POLICY R ESEARCH s’ compensation E. Teacher tion The most common factors that schools consider when determining teacher s’ compensa are teaching experience ( ) and whether the teacher coaches or mentors other teachers 55 percent of schools (46 percent) indicated that teachers are (53 percent) , as seen in Table V.4 . Almost half d- -quarter of schools indicated that either filling a har paid more for advanced degrees. About one to-staff position, additional certifications, number of students taught, student s’ achievement s’ proficiency levels and course growth, or teacher evaluation reports are also considered. Student completion rates were not common factors given by schools that responded to the question ( 16 and 9 percent of schools, respectively). . Factors that affect teacher s’ compensation in online charter Table V.4 schools Percentage of online charter schools using as a Factors that a ffect teacher factor in c ompensation s’ compensation Teaching experience 55 3 Serving as a mentor or coach to other teachers 5 Advanced degrees, such as master’s or doctoral degrees 4 6 - to Filling a hard staff position 2 8 - Additional certifications 2 7 Number of students taught 26 Student s’ ac hievement growth 2 5 Teacher evaluation results 24 s’ proficiency levels 1 6 Student Course completion rates of students 9 These findings suggest that the same characteristics that are most important in determining compensation of teachers in conventiona l public schools are most often used in online charter schools. Nonetheless, it is notable that nearly half of online charter schools do not consider teaching experience in determining compensation, and more than half of online charter schools do not consi der advanced degrees in determining compensation. In a few online charter schools, teacher s’ compensation is also determined in part by collective bargaining agreements and the ability of teachers to earn tenure. Teachers are covered by a coll ning agreement in 9 percent of online charter schools (serving only 5 ective bargai , and teachers can earn tenure in 1 percent of students) of the schools . 1 percent F. Professional development In the 2013–2014 school year, online charter schools reported that teachers p articipated in a range of online and in- person professional development and faculty meetings. Nearly all schools (92 percent ) said their teachers participated in professional development either online or in person (or both). Most but not all schools ( 89 pe rcent) said they provided teachers with paid time off for professional development. More than half of the schools (51 percent ) serving three- quarters (74 percent) of students reported that synchronous, online professional development involving groups of te achers took 31

50 V. TEACHERS AND STA FFING ESEARCH MATHEMATICA POLICY R V. 5) . About 13 percent of schools reported teachers participated in place at least monthly (Figure online sessions at least weekly, and 38 percent -monthly participation. reported monthly or twice The remaining schools — schools —reported online professional which tended to be smaller less often. One -quarter (25 percent ) of schools reported their development activities occurr ing to four times per year year. Thirteen teachers participated two and 11 percent reported once a percent reported their teachers did not attend online professional development sessions with other teachers from the same school. ed in -person professional As might be expected, most online charter schools report development activities occurring less frequently than online professional development. Still, 34 percent of schools reported in- person professional development sessions at a central location at most ( 66 percent , serving 89 percent of students ) schools reported least once a month, and person professional developm ent more than once a year (Figure V.5 ). Only 9 conducting in- -person professional development activities for teachers . percent of schools reported no in . Frequency of t V.5 Figure eachers’ participation in professional development 100 90 80 70 60 50 38% 40 32% Percentage 26% 25% 30 20 13% 13% 11% 9% 8% 8% 10 0 Weekly or more than Monthly or twice per 2-4 times per year Not at all Once per year month once a week Online In-Person 85 percent ) indicated their staff have opportunities to take on Most online charter schools ( ) of schools give additional responsibilities to advance their careers. More than half (55 percent staff the opportunity to supervise junior teachers (as a department chair or lead teacher) and 64 t allow teachers to become instructional coaches or master teachers. Thirty -nine percent of percen schools allow staff to take on additional teaching load, teaching more or larger classes. Almost three- quarters (72 percent ) offer opportunities for teachers to lea d professional development groups for other staff. - Most o nline charter schools report ed holding regular faculty meetings, either online or in ) of schools (serving 46 percent of students) reported their percent person. Almost a third (31 teachers attend faculty meetings at least once per week, 11 percent reported teachers attend aculty meetings occur red meetings twice a month, and 40 percent reported monthly meetings. F , and in 7 percent of schools these meetings two times per year at 11 percent of schools to four frequent faculty meetings tend to be happen only once per year or not at all ; schools with less smaller schools . 32

51 VI. SCHOOL LEADERS A. Previous experience On average, the principals in the survey sample had led their school s for three years before -fourth (28 percent) of principals had the 2013–2014 school year. Nonetheless, more than one newly taken over as leader — a substantially higher of their schools for the current school year number than typical in public schools nationally, where only 9 percent of principals were newly . About half hired in the current school year in 2011–2012 ( U.S. Department of Education 2012) had prior experience as a principal at another school. of online charter principals (49 percent) Nearly all of those with prior experien ce ( 92 percent of those with prior experience and 46 percent of all current online principals) had been principals at conventional, brick- and -mortar -fifth (18 percent) schools. Fewer than one current online charter principals had previously of served as pr incipals of other online schools. In addition to prior experience as a principal, the survey asked about teaching experience in general and specifically in online schools. The vast majority of principals reported prior teaching experience, with an average of 9.3 years of elementary or secondary teaching experience in any school setting. But almost half of principals ( 48 percent) reported no prior online teaching experience. The principals with experience averaged 5.5 years of online teaching experience. development B. Training and professional in some form of training and Nearly all responding principals reported participating professional development in the previous year. A large majority (89 percent ) of principals participated in workshops or conferences . More than three -quarters (78 percent ) of principals participat ed in a school leader network event ( for example, a group of school leaders organized by an outside agency or through the Internet) -thirds of principals (6 3 percent) visited . Nearly two hools improving their leadership skills. Similarly, 56 percent of other sc , with the aim of principals received mentoring, peer observation, or coaching by the leader of another school, or ls said they presented they performed these services for another principal. Almost half of principa of principals ( 39 percent) undertook a at workshops, conferences, or trainings. More than a third university course related to their role as a school leader. Principals of schools affiliated with management organizations participated in these activities at rates that were similar to those of principals of unaffiliated schools. The survey also asked about specific principal training programs. Three -quarters of principals had participated in a specialized principal training program. C. Roles and responsibilities . According to Principals in online charter schools are responsible for a wide variety of tasks d the largest part of their time —an average of 30 percent of survey responses, principals spen week their work nistrative tasks related to issues such as human —performing internal admi resources , regulations, reports, and school budget s (Figure VI.1 ). One -fifth (20 percent ) of principal s spend half or more of their time on administrative tasks. 33

52 VI. SCHOOL LEADERS MATHEMATICA POLICY R ESEARCH . Percentage of online charter pri various s’ time spent on VI.1 Figure ncipal tasks Other 6% Parent interactions 8% Administrative tasks 30% Work with students 11% Evaluate students' data 13% Observe teachers 11% Professional development Work with internal 8% school leaders 13% The rest of principals’ time i s divided among many different kinds of tasks. T ied for second re t wo tasks that each t : working with place a ake about 13 percent of principals’ time, on average , i ncluding instructional coaches, grade leaders, departmental leaders, and internal school leaders other instructional leaders s’ achievement data. Principals reported that , and reviewing student ed observing teachers took about 11 percent of their time , although 25 percent of principa ls report that they do not observe teachers at all. W orking directly with students on issues such as discipline and academic guidance fills about 11 percent of their time shows the . Figure VI.1 average distribution of principal In comparison, principals of public s’ time for a normal week. interacting with students in 2011– schools nationally spent an average of 39 percent of their time U.S. Department of Education 2012) 2012 ( . D. Challenges -ended questi on about the greatest We asked principals of online charter schools an open s of online schools challenge they face as the leader e received a wide variety of responses . W , but , identified as the greatest challenge by one issue far outpaced all others: student engagement - one ) of principals third (33 percent most frequent responses, cited by 11 to 13 he next ). T ble VI.1 (Ta , were administrative challenges . parent engagement , public perception, and percent of principals Principals mentioned various other challenges less frequently; several, like the top challenge, are 34

53 VI. SCHOOL LEADERS ESEARCH MATHEMATICA POLICY R related to the difficulty of engaging students or parents. Only 4 percent of principals identified recruiting effective teachers as the greatest challenge. (Note that the percentages in the table sum to more than 100 percent because some principals ide ntified multiple challenges or a challenge that fits in multiple categories). ’ greatest challenges VI.1 Table . Online charter principals age identifying issue as one of Percent of principals the greatest challenges 33 Student engagement School/CMO/distri ct/state administration 13 11 Public perception Parent engagement 11 10 Finding and analyzing meaningful data School structure 8 remotely from students 7 Being located 7 Accountability systems Funding 7 6 Challenging student population Want more t – student interaction 6 eacher Getting my job done in the available time 6 Finding quality teachers 4 Student recruitment and retention 3 School culture 2 E. Principal evaluation and c ompensation Most online charter schools (61 percent ) use a measure o f student s’ achievement in . Forty percent of schools use student achievement levels in principal s’ evaluating principals evaluation s and 60 percent use student growth ( 39 percent use both achievement s’ achievement levels and achievement growth). Student s’ achievement also plays a role in determining principal s’ compensation in some, but not all, online charter schools. One -third (32 percent) of schools (serving 48 percent of students) tie principa ls’ compensation to student s’ achievement levels on state ass essments ; 37 percent (serving 51 percent of students) tie compensation to student s’ achievement growth (or added) VI.2 ). Almost a third (32 percent ) of schools include one or both of these value (Figure ’ compensation. student achievement factors in determining principals Some schools use other performance measures to inform principal s’ compensation (Figure VI.2 ). The number of enrolled students is used t o determin e compensation in 37 percent of online charter schools, matching the percentage of schools using student s’ achievement growth. About 30 percent of schools use the reenrollment of current students across school years and about the same number use the school’s operating profit when determining the principal ’s compensation. From 10 to 20 percent of sch ools use student s’ course completion rates, retention of teaching staff, or other measures for principal s’ compensation. Forty -two percent of online 35

54 VI. SCHOOL LEADERS MATHEMATICA POLICY R ESEARCH charter schools do not use any of these performance measures for purposes of determining principals’ compensation. VI.2 of principals Figure . Performance measures used for compensation s’ compensation Percentage of online charter schools using each measure to inform principal 100 90 80 70 60 50 37% 37% 40 32% Percentage 30% 28% 30 18% 15% 20 10% 10 0 Number of Students’ Students’ Reenrollment School’s Student Other Retention of enrolled of current course teaching staff achievement test -score operating students students completion growth on levels on profit or loss across school standardized rates state years assessments assessments or the school’s value added 36

55 VII GOVERNANCE, MANAGEME NT, AND FUNDING . uctures. Sixty -two of responding schools Schools reported a range of governance str percent -three percent . Ninety are their own local education agencies of schools reported that charter reported that their authorizers monitor authorizers monitor their state test scores and 85 percent bout half of schools reported that authorizers monitor their reenrollment their attendan ce rates. A 52 and 50 percent , respectively ). Only 5 or course completion rates ( of schools reported percent their school authorizer does not monitor any of these student outcomes. re than half ) of online charter schools are affiliated with school management Mo (57 percent that provide curriculum or instructional support services . Almost half of the organizations schools affiliated with a management organization are affiliated with one of the two largest online school management organizations, K12 and Connections. Among schools affiliated with a management organization , nearly all reported receiving some form of support from the management organization’s central office; most reported receiving three or more services (Figure VII.1 ). A large majority (91 percent ) of schools affiliated with management organizations reported receiving professional development for teachers, such as workshops and in- , from the management service training programs . The second organizations most provided support reported was curriculum and instructional materials, which 86 percent of schools reported receiving from their affiliated management -five percent of affiliated schools reported receivi organization. Seventy system of diagnostic ng a or formative s tudent assessments and results . S lightly fewer (73 percent ) received technical assistance, support, or resources in areas in whic h student s’ test scores we re weak . Only half of the affiliated schools reported access to instructional coaches through their management . organizations The survey included a limited number of questions on the funding of online charter schools. More than half of online charter schools (55 percent, serving 71 percent of students) participate in the federal Title 1 program. Eighty -three percent of schools receive designated funding for special education services. Only 12 percent of schools reported that the total number of courses completed by students could affect their funding. 37

56 VII. GOVERNANCE, MANAGEMENT AND FUNDING ESEARCH MATHEMATICA POLICY R VII.1 Figure . Support from affiliated online charter school management organizations Among schools affiliated with management organizations, percentages receiving each of the following services from the management organization 100 91% 86% 90 75% 80 73% 70 60 49% 50 40 Percentage 30 20 10 0 Professional Curriculum and A system of Access to Technical assistance, development for instructional diagnostic or instructional coaches support, or resources teachers materials formative student assessments and results 38

57 VIII. CONCLUSION ter schools operat e in the United States, serving about 200,000 About 200 online char students. They serve students at elementary, middle, and high school grades, and they vary enormously in size, with 24 percent of schools serving more than 1,000 students each and collectively 79 percent of total enrollment in online charter schools. More than accounting for half are affiliated with school management organizations. The sector barely existed before the turn of the millennium, and it has grown rapidly since. s might appeal in principle to a wide variety of different kinds of Online charter school students, and they in fact serve a reasonably diverse population, with African American students and students with disabilities represented at levels similar to those in other public schools. Hispanic students and English learners are underrepresented in online charter schools, however, perhaps in part because many of the schools do not offer instruction targeted to students whose st online charter schools expect families first language is not English. In addition, the fact that mo to provide Internet access (and many do not uniformly provide computers to all students) m ight On average, students in online charter schools prevent some prospective students from enrolling. out two years. remain enrolled for ab Online schooling creates both constraints and opportunities for modes of delivering In most online charter schools, a substantial amount of coursework is self -paced. instruction. Consistent with this, the instructional method used most frequently in online charter schools is individualized, student -driven independent study . Given the ubiquity of independent study, it is not surprising that online charter schools have substantially less sync hronous interaction between teachers and students tha n conventional schools, and the difference is dramatic: s tudents in the typical online charter school have less synchronous instructional time in a week than students in a brick and mortar school have in a day. This higher ra tio of could be partly related to the students to teachers in online schools, which average 30 students per teacher, compared with 20 in brick- -mortar charters and 17 in conventional public schools . ( This could be the result of and the time available to teachers or because limited time provided in the model the ratio limiting makes the higher ratio possible) . With a limited number of live contact hours and a lean staffing model, online charter schools place substantial expectations on parents, who are expected not only to ensure that students keep up with assignments but also to participate in training sessions and, in a large number of schools, to actively participate in the student’s instruction. The substantial burden placed on parents is presumably a response to the issue that online charter school principals regard as their greatest challenge: keeping students engaged. The challenge is partly inherent to online schooling, because the schools have no way to ensure that y due to online schools serving as students are in their seats and ready to learn. It is probably partl schools of last resort for a subset of their students who previously disengaged from brick -and - mortar schools. And it is surely exacerbated by a high student –teacher ratio and a small number of live contact hours. All of these are reasons for concern about whether the sector is likely to be effective in promoting the achievement of its students —an issue addressed in depth in the third volume of this study (Woodworth et al. 2015) . 39

58 This p age has been left blank for double- sided copying .

59 REFERENCES hael B. Horn, and Curtis W. Johnson. Disrupting Class: How Christensen, Clayton M., Mic -Hill, Disruptive Innovation Will Change the Way the World Learns . New York: McGraw 2008. . Gill, J . Haimson, A . Killewald, M Furgeson, J . Nichols -Barrer, B . Teh, N . ., B . McCullough, I -Savi tz, M . Bowen, A . Demeritt, P . Hill, and R . Lake. “Charter -School Verbitsky Management Organizations: Diverse Strategies and Diverse Student Impacts.” Cambridge, MA: Mathematica Policy Research , 2012. harter Schools eral Attention Needed to Government Accountability Office. “C : Additional Fed Help Protect Access for Students with Disabilities.” Washington, DC: U.S. Government Accountability Office, June 2012. Molnar, A., Huerta, L., Rice, J. K., Barbour, M.K., Miron, G., Shafer, S. R., Gulosino, C., Horvitz, B. “Virtual Schools in the U.S. 2014: Politics, Performance, Policy, and Research Evidence.” Boulder, CO: National Education Policy Center, March 2014. Molnar, A., Huerta, L., Barbour, M. K., Miron, G., Shafer, S. R., Gulosino, C. “Virtual Schools in the U.S. 2015: P olitics, Performance, Policy, and Research Evidence.” Boulder, CO: National Education Policy Center, March 2015. Pazhouh, Rosa, Robin Lake, and Larry Miller. “The Policy Framework for Online Charter Schools.” Seattle, WA: Center on Reinventing Public Educa tion, University of Washington Bothell, 2015. Raymond, M. Charter School Performance in Pennsylvania . Stanford, CA: Center for Research on Education Outcomes, Stanford University , 2011. -Added in a Virtual Learning Environment: An Evaluation Ritter, G., and M.F. Lueken. “Value of the Arkansas Virtual Academy.” Presentation at the annual meeting of the Association for Education Finance and Policy, New Orleans, March 2013. U.S. Department of Education, National Center for Education Statistics. Schools and Staf fing NCES 2014- 356. Washington, DC: U.S. Government Printing Office , 2012. Survey. U.S. Department of Education, National Center for Education Statistics, Common Core of Data (CCD), "Public Elementary/Secondary School Universe Survey", 2013- 14 v.1a. Availa ble at [https://nces.ed.gov/ccd/elsi/]. Accessed February 1, 2015. Woodworth, J.L., M.E. Raymond, K. Chirbas, M. Gonzalez, Y. Negassi, W. Snow, and C. Van Donge. Online Charter School Study 2015 . Stanford, CA: Center for Research on Education Outcomes, Sta nford University, 2015. Zimmer, R., R. Buddin, D. Chau, G. Daley, B. Gill, C. Guarino, L. Hamilton, C. Krop, D. Charter School Operations and Performance: McCaffrey, M. Sandler, and D. Brewer. “ Evidence from California.” Santa Monica, CA: RAND Corp., 2003. Zimmer, R. , B . Gill, T.K . Booker, S . Lavertu, T .R. Sass, and J . Witte. “ Charter Schools in Eight States: Effects on Achievement, Attainment, Integration, and Competition .” Santa M onica, CA: RAND Corp., 2009. 41

60 This p age has been left blank for double- sided copying .

61 APPENDIX A SURVEY METHODS 43

62 This p age has been left blank for double- sided copying .

63 APPENDIX A ESEARCH MATHEMATICA POLICY R forts to identify universe of online charter schools. Working with our Screening ef ) to identify the universe colleagues at the Center for Research on Education Outcomes ( CREDO of online charter schools operating in the United States, we began with lists compiled by the International Association for K -12 Online Learning (iNACOL) and the National Education Policy Center ( Molnar et al. 2014), supplemented with data from state education departments, the National Center for Education Statistics, and web searches. We compile d an initial nationwide list of 897 schools that could potentially be identified as online charter schools. The aim was to include not merely a sample, but the entire population of online charter schools operating in the United States. Survey e ligibility c riteria . We identified several eligibility criteria for inclusion in the en a public charter school that delivered instruction during the study. First, a school must have be . Next, it must have served 2013–2014 school year time students. Th ird, the school at least 20 full- 4 ed a fully online program of study for its full time students. should have deliver Because the initial list did not indicate whether the schools met these criteria, we contacted schools by onfirm eligibility. Whe n eligibility was telephone with a brief screening questionnaire to c confirmed, we sought to collect or confirm contact information for the lead administrator of the school. methodology. We conducted a careful review to identify and confirm potential Screening duplicates, which we subsequently removed from the list. Online searches were then conducted, using the C ommon C ore of Data to (1) identify schools that were not public charters and (2) to obtain contact information for school administrators . From there, professional interviewer s conducted telephone outreach with the remaining 597 schools to complete a brief eligibility screening questionnaire. Screening efforts substantially reduced the number of schools from the ter schools or did not serve full initial list, as the majority were either confirmed not to be char - time students. When it was not possible to determine the number of students served, the school was included in the study universe. Further refinement of the sample came from combining cases in wh ich one head of school over saw a group of schools that were initially listed as separate entities. As a result of the screening, we confirmed that 63 percent were ineligible (567 of the , and 147 schools were confirmed as eligible. T 897 schools) remaining schools (183) could not he be confirmed as eligible or ineligible during the screening process, so we included them in the survey sample and assessed eligibility during the survey proc ess . Eligibility was confirmed for all respondents to the survey through the use of screening questi ons at the beginning of the questionnaire. These items confirmed that the school was (1) a public school , (2) a charter school , and (3) enrolled any students whose program of study was delivered entirely online. Survey p re- In preparation for data col lection, Mathematica conducted a pre- test with a test. convenience sample of four online school leaders in April 2014. This provided an opportunity to ensure the wording was clear, the items were easy to follow, the sequence of items flowed well, and to estimate the time needed to complete the survey . Based on feedback provided, we refined item wording and reduced the number of questions to reduce respondent burden. 4 Schools that required students to come on site for testing, lab work, noncore classes, or for classes as part of an individua lized education program were considered eligible for the study. 45

64 APPENDIX A ESEARCH MATHEMATICA POLICY R We secured endorsements for the survey from the iNACOL, the Conducting the survey. Arizona State B oard for Charter Schools, the California Charter Schools Association, and the In addition, senior leadership from the two largest Ohio Alliance of Public Charter Schools. online school management organizations, K12 and Connections Education, reviewed the s tudy 5 and provided their endorsement , encouraging all of their school leaders to take part. The survey field period ran for 19 weeks, from June 6 to October 14, 2014. The target respondent was the leader of the school or a designated proxy who was most knowledgeable about school operations, instructional practices, and policies. Respondents were offered two incentives to complete the 45 -minute questionnaire : a payment of $50 and an invitation to our presentation of the study findings. ly deployed on the web, with an invitation sent by mail, followed by The survey was initial six email reminders to all nonresponders. Telephone follow -up continued across the field period. questionnaire to all nonresponders via In week 14, Mathematica sent a paper version of the Priority M ail. After careful consideration of the trade -offs between unit and item nonresponse , minute version of the questionnaire by tele phone and fax in we deployed an abbreviated, 10- week 18. The abbreviated version was offered to all 92 of the presum ed-eligible nonresponders and was completed by 14 of them . Across the field period, staff continued efforts to confirm eligibility for schools that had not completed screeners. In addition to seeking this information directly from the schools (by tele phone ), Mathematica used administrative data from state boards of education to (1) confirm charter school status, (2) verify that the school operated an instructional program entirely online, and (3) identify schools t hat were no longer in operation. Through this process, contact information for school leaders was updated, as needed. At the end of the field period there Response rates and nonresponse weight adjustment. were 114 completed cases, 13 partial completes, 60 nonresponses, and 143 schools determined to be ineligible: 17 were not public schools, 95 were not charters, and 31 did not operat e a full -time program fully online. The final response rate for the survey was 67.9 percent of the population of 6 (187 schools) . schools ultimately deemed to be eligible The 127 schools with survey responses included at least one school from every state where online charter schools operat ed. On average, both responding and nonresponding schools had years at the time the survey was conducted. The a verage percentage of been operating for seven students certified for free and reduced -price lunch was nearly identical in responding and nonresponding schools ( 35 and 37 percent , respectively ). The average number of students enrolled in responding schools (800) was smaller than the average number in nonresponding schools (1,266), but the median was larger in responding schools (313) than in nonresponding 5 one message to their school leaders across the field period, we did not Although both of these organizations provided more than a list of nonresponding schools, so as to respect the schools’ right to voluntary participation in the study. give the organizations 6 The survey response rate is calculated as follows: (completed cases + partials) / total eligible sample (total sample – ineligibles). 46

65 APPENDIX A ESEARCH MATHEMATICA POLICY R schools (213). This occurred because a handful of very large schools (each with at least 9,000 students) were among the nonrespondents. All results reported are adjusted to account for observed differences between responding and nonresponding schools. Nonresponse weights were created by taking the inverse of response rates within weighting cells defined by variables that were associ ated both with outcomes of interest and with the response indicator. To ensure stable adjustments, each cell had to have at least 25 sample members. Because of the small number of cases, we were limited to a maximum of seven th five . After removing variables that we were certain weighting cells; we ended up wi would be unrelated to the outcomes of interest, we reviewed cross -tabulations of the remaining variables against the response indicator. We selected a small subset of variables in wh ich there was subst antial variation in the response rate across levels of the variables, indicating that response was associated with this variable. Variation in response rates for the student teacher percent (with low student –teacher ratios) to 75 percent (with high ratio variable ranged from 58 student –teacher ratios). There was also considerable variation in response rates across states and regionally. California was the only state with enough sample members to include in the adjustments. For schools with student –teacher r atios that were neither high nor low, response rates in California schools were 68 percent ; those not in California had a response rate of 61 percent . 47

66 This p age has been left blank for double- sided copying .

67 B APPENDIX SUPPLEMENTAL T ABLE 49

68 This p age has been left blank for double- sided copying .

69 APPENDIX B ESEARCH MATHEMATICA POLICY R B.1. Frequency of curriculum delivery method, by grade Table (percent ages) Used Used frequently Used rarely Not used at all occasionally 4th grade 11 28 34 Lecture 27 - guided synchronous 15 42 35 Teacher 8 discussion 23 39 30 9 Collaborative learning involving two or more students working together nt - driven 8 55 Individualized, stude 7 29 independent study 7th grade 19 38 Lecture 19 24 Teacher - guided synchronous 12 36 37 15 discussion 20 40 27 12 Collaborative learning involving two or more students working together - driven 7 5 9 Individualized, student 5 30 independent study High School 20 28 Lecture 24 28 Teacher - guided synchronous 21 32 31 16 discussion 21 38 Collaborative learning involving two or 21 19 more students working together - driven 60 22 3 15 Individualized, student independent study 51

70 This p age has been left blank for double- sided copying .

71 www.mathem atica -mpr.com Improving public well -being by conducting high quality, objective research and data collection PRINCETON, NJ ■ ANN ARBOR, MI ■ CAMBRIDGE, MA ■ CHICAGO, IL ■ OAKLAND, CA ■ WASHINGTON, DC ® Mathematica is a registered trademark of Mathematica Policy Research, Inc.

Related documents