Response to NEPC Report on K12 Inc.

K12 Inc. Public Affairs — 703/436-3168
July 18, 2012

  • The NEPC report claims that students attending schools managed by K12 are "falling behind," "falling further behind," and "more likely to fall behind" in reading and math scores compared to students in brick-and-mortar schools.  However, the report provides no evidence backing up this claim. To make such conclusions, one would need to know the academic starting point of students, in this case, test scores from a prior school year when they were enrolled in a brick-and-mortar school. That test data is largely unavailable and is certainly not included in the NEPC report. The report simply compares static, end-of-the-year test data among all schools in a state, without regard to comparisons of similar students or student populations. The report also neglects to provide any data or analysis on measures of student academic growth.
  • As K12 has previously explained, since the 2008-09 school year, we have administrated the Scantron Performance Series exam to students in grades 3-10 as a way to gauge growth over the academic year. Student growth is compared to the Scantron norm group, which is comprised of thousands of students who represent the national demographics in terms of socio-economic status and ethnicity. Our students have consistently outperformed the norm group in the majority of grade levels in both math and reading.
  • K12 also recently cited an independent analysis by researchers at the University of Arkansas, who compared students enrolled in the Arkansas Virtual Academy (ARVA), a K12-managed school, with their traditional school peers. The study matched ARVA students to two students from the same home district and with similar benchmark scores on the Arkansas state criterion reference exam, and with similar ethnicity and socioeconomic status. Using three years of state test data (2009, 2010, and 2011), the study concluded that in math and literacy, the ARVA students outperformed their traditional public school peers over the time period studied.
  • Recent work undertaken by K12, in cooperation with West Coast Analytics (WCA), equates Scantron Performance Series scores with state test scores in California, Ohio and Pennsylvania in an effort to predict student performance on the state standardized tests based on a school administration of the Scantron Performance Series tests.  By performing this equating, K12 was able to determine that a majority of incoming (i.e., new to the schools) students in the Ohio Virtual Academy, the Agora Cyber Charter School (PA), and the California Virtual Academies would have to make two or more times the expected annual gain during the school year to be proficient on the spring administration of the state standardized tests, the same tests used to determine the AYP rating for schools.
  • Understanding that K12-managed schools are serving large numbers of students who enter behind grade level in math and reading, K12 has undertaken several new initiatives specifically targeted to help these students reach grade level.
    • For example, K12 recently released a new math curriculum and a remedial reading course, both based on the latest educational research and pedagogical methods.  Also, for school year 2011-12, K12 launched a program called National Math Lab. In addition to their regular math coursework, students in grades 5-11 who are two or more grade levels behind in math are attending synchronous mathematical instruction provided by highly-trained math teachers four days per week.
    • K12 has also worked with the schools it manages to launch several physical learning centers using blended learning programs designed to provide additional support to students who need it most — severely "at risk" students living in urban areas. These programs assist students who have previously dropped out of high school using competency-based education methods to recover credits so they can graduate. In many cases, social services and legal services are also provided for these students.
  • NEPC correctly observes that the schools managed by K12 generally operate on less public revenue than brick-and-mortar charter or district schools, and that those schools spend more on direct student instructional costs, and less on teacher salary and benefits, than traditional schools.  However, the report fails to make clear that K12 is a vendor to public schools. Independent, non-profit charter or school district boards make decisions regarding the allocation of public funds. K12 does not "spend money" on anything - non-profit school boards make those decisions
  • The report claims that schools managed by K12 spend "little or nothing on...acilities and maintenance, transportations, and food services." While it is true that virtual schools do not use traditional brick-and-mortar school buildings, the report does not cite the unique costs in the virtual model, e.g., costs related to the shipping of computers and material to and from student's homes, costs associated with obtaining physical locations for proctored administration of standardized testing, internet and phone costs for teachers, software licenses for web-based classroom tools.
  • The authors of the NEPC report cite an "existing body of evidence," on academic performance, including a study of student in Pennsylvania virtual charter schools conducted by the Center for Research on Education Outcomes (CREDO) at Stanford University. However, it is likely that the CREDO study included students who had been enrolled less than a full academic year and did not disaggregate the data to show performance of different virtual school providers. Therefore, one cannot be sure of K12's specific rating in the CREDO study. In the 2009-10 school year, which was during the time period of the CREDO study, Agora, the only cyber school in Pennsylvania managed by K12, met 97% of its AYP targets.
  • The report correctly diagnoses many of the problems and shortcomings with the Adequate Yearly Progress (AYP) measure of school progress. Nevertheless, the authors then go on to use the admittedly flawed accountability measure as a way of criticizing the performance of schools managed by K12. As K12 has explained, AYP is structured to reward schools that have small, stable, and/or homogenous populations. AYP is not structured to recognize or reward academic growth; it is an annual "snapshot in time" look at student performance, thus favoring schools with steady populations. Hidden beneath the binary "AYP Yes-or-No" rating are the complexities that include a closer look at specific subgroup or target performance. For the 2010-11 school year, schools managed by K12 met 75% of the AYP targets.
  • In attempting to explain what it incorrectly concludes is K12's "weak performance," the NEPC report maligns teachers in K12 manage schools by suggesting that there is "inadequate or insufficient instruction." K12 manages and/or employs more than 3,500 public school teachers - general education teachers, special education teachers - in addition to counselors, advisors, family coaches, and paraprofessionals, who all work directly with students enrolled in the schools. All of the teachers meet each state's highly qualified guidelines under the federal No Child Left Behind (NCLB) law. 97% of the teachers have prior brick-and-mortar teaching experience, with 66% having taught for at least five years in a traditional setting. 54% have a Master's degree, over 80% report in surveys that they are satisfied with the K12 curriculum product, and more than 90% report they are satisfied with the training offered by K12.
  • NEPC correctly points out that some families approach the virtual schools as a temporary solution to academic, social, behavior, medical, or other challenges. In almost all of these cases, the traditional local brick-and-mortar was failing that family. Students who are academically successful in their current school are not usually shopping for other educational options, thus it is reasonable to assume that the majority of students enrolling in virtual schools are not experiencing success in the traditional model.
  • As K12 has explained previously, we conduct extensive customer satisfaction surveys to better understand why families enroll students, what expectations they are bringing, how satisfied they are with the programs and services, and why they choose to withdraw. 76% of parents surveyed cited concerns with the traditional school environment as the primary reason they chose a K12-managed online school. Parents report that they are highly satisfied with the K12 curriculum and with the services they receive at K12-managed schools. Overall satisfaction is at 90%+, and 88% of parents say they would recommend the K12 program to other parents.
  • The NEPC report suggests that K12 enrolls students with disabilities at rates moderately below public school averages, and claims that the company spends "half as much per pupil as charter schools overall spend on special education instruction and a third of what districts spend on special education instruction." However, the report fails to consider that the responsibility for special education services varies from state to state. In some cases, the local district retains responsibility, in which case it would appear in financial audits that the school managed by K12 spends little or nothing. Additionally, excess costs of special education services, which constitute encroachment to the general fund of the school, do not show as special education expenditure, but as general fund expenditure. Regarding special education, in the most recent round of gains testing, students with disabilities made 90% of the Scantron national norm group gain in math, and 200% of the norm group gain in reading.
  • The NEPC report uses incomplete or incorrect school data that should cast doubt on the conclusions regarding student characteristics.
    • According to NEPC, data was included from "K12's 48 full-time virtual schools." But the list of schools in the appendix includes a number of schools where K12 provides curriculum but is not responsible for managing the school. These schools include the Pennsylvania Virtual Charter School (PAVCS) with more than 3,300 students, the Lawrence Virtual School with 1,200 students, and Beacon Academy, with more than 600 students. Conversely, the report omitted data from the Georgia Cyber Academy, a fully managed statewide school with more than 10,000 students.
    • NEPC selected K12-managed schools in five states (AR, AZ, ID, OH, PA) to conduct an analysis of school finances. Taking those same five states, we find that the combined average percentage of students in the K12-managed schools who are eligible for Free and Reduced Lunch (FRL) is 58%, compared to the average of 45% among all schools in those states, and compared to the national average of 45.4%.
    • The report claims that the percentage of students who qualified for Free and Reduced Lunch for school year 2010-11 at the Nevada Virtual Academy was 0%. The actual percentage was 32%. The report also claims that the FRL population at Agora Cyber Charter School for 2010-11 was 35.1%. In fact, it was 60%.


Reference for academic data: "K12 Virtual Academies Academic Performance Trends," April 9, 2012, available as a PDF at